text
stringlengths 66
377k
|
|---|
The Rise of Cloud Computing: Data Protection, Privacy, and Open Research Challenges—A Systematic Literature Review (SLR)
Cloud computing is a long-standing dream of computing as a utility, where users can store their data remotely in the cloud to enjoy on-demand services and high-quality applications from a shared pool of configurable computing resources. us, the privacy and security of data are of utmost importance to all of its users regardless of the nature of the data being stored. In cloud computing environments, it is especially critical because data is stored in various locations, even around the world, and users do not have any physical access to their sensitive data. erefore, we need certain data protection techniques to protect the sensitive data that is outsourced over the cloud. In this paper, we conduct a systematic literature review (SLR) to illustrate all the data protection techniques that protect sensitive data outsourced over cloud storage. erefore, the main objective of this research is to synthesize, classify, and identify important studies in the field of study. Accordingly, an evidence-based approach is used in this study. Preliminary results are based on answers to four research questions. Out of 493 research articles, 52 studies were selected. 52 papers use different data protection techniques, which can be divided into two main categories, namely noncryptographic techniques and cryptographic techniques. Noncryptographic techniques consist of data splitting, data anonymization, and steganographic techniques, whereas cryptographic techniques consist of encryption, searchable encryption, homomorphic encryption, and signcryption. In this work, we compare all of these techniques in terms of data protection accuracy, overhead, and operations on masked data. Finally, we discuss the future research challenges facing the implementation of these techniques.
# Introduction
Recent advances have given rise to the popularity and success of cloud computing. It is a new computing and business model that provides on-demand storage and computing resources. e main objective of cloud computing is to gain financial benefits as cloud computing offers an effective way to reduce operational and capital costs. Cloud storage is a basic service of cloud computing architecture that allows users to store and share data over the internet. Some of the advantages of cloud storage are offsite backup, efficient and secure file access, unlimited data storage space, and low cost of use. Generally, cloud storage is divided into five categories: (1) private cloud storage,personal cloud storage, (3) public cloud storage, (4) community cloud storage, and (5) hybrid cloud storage.
However, when we outsource data and business applications to a third party, security and privacy issues become a major concern [bib_ref] Security and privacy in cloud computing, Xiao [/bib_ref]. Before outsourcing private data to the cloud, there is a need to protect private data by applying different data protection techniques, which we will discuss later in this SLR. After outsourcing the private data to the cloud, sometimes the user wants to perform certain operations on their data, such as secure search. erefore, while performing such operations on private data, the data needs to be protected from intruders so that intruders cannot hack or steal their sensitive information.
Cloud computing has many advantages because of many other technical resources. For example, it has made it possible to store large amounts of data, perform computation on data, and many other various services. In addition, the cloud computing platform reduces the cost of services and also solves the problem of limited resources by sharing important resources among different users. Performance and resource reliability requires that the platform should be able to tackle the security threats. In recent years, cloud computing has become one of the most important topics in security research. ese pieces of research include software security, network security, and data storage security.
e National Institute of Standards and Technology (NIST) defines cloud computing as"a model for easy access, ubiquitous, resource integration, and on-demand access that can be easily delivered through various types of service providers. e Pay as You Go (PAYG) mechanism is followed by cloud computing, in which users pay only for the services they use. e PAYG model gives users the ability to develop platforms, storage, and customize the software according to the needs of the end-user or client. ese advantages are the reason that the research community has put so much effort into this modern concept [bib_ref] An ADS-PAYG approach using trust factor Against economic denial of sustainability attacks..., Karthika [/bib_ref].
Security is gained by achieving confidentiality, integrity, and data availability. Cloud users want assurance that their data must be saved while using cloud services. ere are various types of attacks that launch on a user's private data, such as intrusion attacks, hacking, stealing the user's private data, and denial of service attacks. 57% of companies report security breaches using cloud services [bib_ref] Cloud computing and its role in the information technology, Alam [/bib_ref]. Data privacy is more important than data security because cloud service providers (CSPs) have full access to all cloud user's data and can monitor their activities, because of which the cloud user privacy is compromised. For example, a user is a diabetic, and the CSP is analyzing their activities, such as what he is searching for more and what kind of medicine he is using the most. Because of this access, CSP can get all the sensitive information about an individual user and can also share this information with a medicine company or an insurance company [bib_ref] Heterogeneity in mobile cloud computing: taxonomy and open challenges, Sanaei [/bib_ref]. Another problem is that the user cannot fully trust CSP. Because of this reason, there are many legal issues. Users cannot store their sensitive data on unreliable cloud services because of this mistrust. As a result, many users cannot use cloud services to store their personal or sensitive data in the cloud. ere are two ways to solve this problem. One is that the user installs a proxy on his side, and this proxy takes the user's data, encrypts and saves their data using some data protection techniques, and then sends it to the untrusted CSP [bib_ref] Cloud computing and big data: a review of current service models and..., Branch [/bib_ref]. e recent Google privacy policy is that any user can use any Google service free of cost; however, Google monitors their activity by monitoring their data to improve their services . In this paper, we compare different types of data protection techniques that provide privacy and security over the data stored on the cloud. Many papers discuss outsourcing data storage on the cloud [bib_ref] Cloud security issues and challenges: a survey, Singh [/bib_ref] [bib_ref] Attribute based encryption in cloud computing: a survey, gap analysis, and future..., Praveen [/bib_ref] , however, we also discuss how we can secure the outsourced data on the cloud. Most of the paper describes the data security on the cloud vs the external intruder attacks [bib_ref] A survey of security issues for cloud computing, Khan [/bib_ref] [bib_ref] A survey on cloud computing security: issues, threats, and solutions, Singh [/bib_ref]. is paper not only discusses the security attacks from outside intruders and securing mechanisms but also inner attacks from the CSP itself. Many surveys cover data privacy by applying cryptographic techniques [bib_ref] Ensuring security and privacy preservation for cloud data services, Tang [/bib_ref] [bib_ref] Practical secure computation outsourcing: a survey, Shan [/bib_ref]. ese cryptographic techniques are very powerful for the protection of data and also provide a very significant result. However, there is a problem as these cryptographic techniques require key management, and some of the cloud functionalities are not working on these cryptographic techniques. In this paper, we also discuss some steganographic techniques. To the best of our knowledge, no study discusses all the conventional and nonconventional security techniques. erefore, all the data protection techniques need to be combined in one paper. e rest of this paper is organized as follows: Section 3 of the paper describes the research methodology that consists of inclusion, exclusion criteria, quality assessment criteria, study selection process, research questions, and data extraction process. Also, we discuss assumptions and requirements for data protection in the cloud. Section 4 presents all the cryptographic and also noncryptographic techniques that are used for data protection over the cloud. Also, we discuss the demographic characteristics of the relevant studies by considering the following four aspects: (i) publication trend, (ii) publication venues (proceeding and journals), (iii) number of citations, and (iv) author information. Section 4 also compares all these data protection techniques. Lastly, in Section 5, we discuss results and present conclusion and future work.
# Related work
e first access control mechanism and data integrity in the provable data possession (PDP) model is proposed in the paper [bib_ref] Provable data possession at untrusted stores, Ateniese [/bib_ref] , and it provides two mobile applications based on the RSA algorithm. Like the PDP, the author in the paper [bib_ref] PORs: proofs of retrievability for large files, Juels [/bib_ref] proposed a proof of retrievability (PoR) scheme that is used to ensure the integrity of remote data. PoR scheme efficiency is improved using a shorter authentication tag that is integrated with the PoR system [bib_ref] Compact proofs of retrievability, Shacham [/bib_ref]. A more flexible PDP scheme is proposed by the author of the paper [bib_ref] Scalable and efficient provable data possession, Ateniese [/bib_ref] that uses symmetric key encryption techniques to support dynamic operations. A PDP protocol with some flexible functionality is developed, in which, we can add some blocks at run time [bib_ref] Dynamic provable data possession, Erway [/bib_ref]. A new PDP system with a different data structure is introduced, and it improves flexibility performance [bib_ref] A novel efficient remote data possession checking protocol in cloud storage, Yan [/bib_ref]. Similarly, another PDP model with a different data structure is designed to handle its data functionality [bib_ref] An efficient public auditing protocol with novel dynamic structure for cloud data, Shen [/bib_ref]. To improve the accuracy of the data, the author of the paper [bib_ref] MuR-DPA: top-down levelled multi-replica merkle hash tree based secure public auditing for..., Liu [/bib_ref] designed a multireplicas data verification scheme that fully supports dynamic data updates.
A unique data integration protocol [bib_ref] Identity-based distributed provable data possession in multicloud storage, Wang [/bib_ref] for multicloud servers is developed. e author of the paper [bib_ref] Efficient Identity-Based Provable Multi-Copy Data Possession in Multi-Cloud Storage, Li [/bib_ref] also considers the complex area where multiple copies are stored in multiple CSPs and builds a solid system to ensure the integrity of all copies at once. A proxy PDP scheme [bib_ref] Proxy provable data possession in public clouds, Wang [/bib_ref] is proposed, which supports the delegation of data checking that uses concessions to verify auditor consent. In addition, the restrictions of the verifier are removed that strengthened the scheme, and it proposes a separate PDP certification system [bib_ref] Identity-based privacy preserving remote data integrity checking for cloud storage, Li [/bib_ref]. To maintain the security of information, a concept for information security is proposed and a PDP protocol for public research is developed [bib_ref] Privacy-preserving public auditing for secure cloud storage, Wang [/bib_ref]. To resolve the certification management issue, the PDP system with data protection is introduced [bib_ref] Identity-based remote data integrity checking with perfect data privacy preserving for cloud..., Yu [/bib_ref].
Identity-based cryptography is developed, in which a user's unique identity is used as input to generate a secret key [bib_ref] Certificateless public integrity checking of group shared data on cloud storage, Li [/bib_ref]. Another PDP protocol is recommended to ensure confidentiality [bib_ref] Remote data possession checking with privacy-preserving authenticators for cloud storage, Shen [/bib_ref]. e author of the paper [bib_ref] Knox: privacy-preserving auditing for shared data with large groups in the cloud, Wang [/bib_ref] proposed a scheme, in which tags are generated through the ring signature technique for group-based data sharing that supports public auditing and maintains user privacy. A new PDP system is introduced for data sharing over the cloud while maintaining user privacy [bib_ref] Privacy-preserving public auditing for shared cloud data supporting group dynamics, Wang [/bib_ref]. Additionally, it supports the dynamic group system and allows users to exit or join the group at any time. Another PDP system [bib_ref] Mona: secure multiowner data sharing for dynamic groups in the cloud, Liu [/bib_ref] that is based on broadcast encryption and supports dynamic groups [bib_ref] Anonymous certificate-based broadcast encryption with personalized messages, Chen [/bib_ref] is introduced. e issue of user revocation has been raised [bib_ref] Panda: public auditing for shared data with efficient user revocation in the..., Wang [/bib_ref] , and to address this issue, a PDP scheme has been proposed, which removes the user from the CSP using the proxy signature method. A PDP-based group data protocol was developed to track user privacy and identity [bib_ref] Enabling public auditing for shared data in cloud storage supporting identity privacy..., Yang [/bib_ref]. A PDP system [bib_ref] Privacy-preserving cloud auditing with multiple uploaders, Wu [/bib_ref] is proposed for data sharing between multiple senders. e author of the paper [bib_ref] SEPDP: secure and efficient privacy preserving provable data possession in cloud storage, Nayak [/bib_ref] provides SEPDP systems while maintaining data protection. However, the author of the paper [bib_ref] Comments on" SEPDP: secure and efficient privacy preserving provable data possession in..., Yu [/bib_ref] proved that the scheme proposed in [bib_ref] SEPDP: secure and efficient privacy preserving provable data possession in cloud storage, Nayak [/bib_ref] is vulnerable to malicious counterfeiting by the CSP. A collision-resistant user revocable public auditing (CRUPA) system [bib_ref] CRUPA: collusion resistant user revocable public auditing of shared data in cloud, Mara [/bib_ref] is introduced for managing the data that is shared in groups. Another scheme [bib_ref] An efficient and secure data sharing scheme for mobile devices in cloud..., Lu [/bib_ref] is introduced as a way to ensure the integrity of mobile data terminals in cloud computing.
To address the PKI issue, identity-based encryption [bib_ref] Identity privacypreserving public auditing with dynamic group for secure mobile cloud storage, Yu [/bib_ref] is designed to enhance the PDP protocol and maintain user privacy in a dynamic community. Before sharing usersensitive data with third parties or researchers, data owners ensure that the privacy of user-sensitive data is protected. We can do this using data anonymization techniques [bib_ref] Utility-privacy tradeoffs in databases: an information-theoretic approach, Sankar [/bib_ref]. In recent years, the research community has focused on the PPDP search area and developed several approaches for tabular data and SN [bib_ref] Combined fuzzy clustering and firefly algorithm for privacy preserving in social networks, Langari [/bib_ref] [bib_ref] Novel trajectory privacypreserving method based on clustering using differential privacy, Zhao [/bib_ref] [bib_ref] An evaluation of edge modification techniques for privacy-preserving on graphs, Casas-Roma [/bib_ref] [bib_ref] Extended k-anonymity model for privacy preserving on micro data, Rahimi [/bib_ref] [bib_ref] An efficient privacy mechanism for electronic health records, Anjum [/bib_ref]. ere are two popular settings in PPDP: one is interactive, and the other is noninteractive [bib_ref] Differentially private data release for data mining, Mohammed [/bib_ref]. e K-anonymity model [bib_ref] k-ANONYMITY: a model for protecting privacy, Sweeney [/bib_ref] and its effects are most commonly used in the noninteractive setting of PPDP [bib_ref] Providing k-anonymity in data mining, Friedman [/bib_ref] [bib_ref] Anonymizing classification data for privacy preservation, Fung [/bib_ref] [bib_ref] Efficient multidimensional suppression for k-anonymity, Kisilevich [/bib_ref] [bib_ref] Information based data anonymization for classification utility, Li [/bib_ref] [bib_ref] Design of big data privacy framework-a balancing act, Geetha [/bib_ref]. Differential privacy (DP) [bib_ref] Differential privacy: a survey of results, Dwork [/bib_ref] and an interactive configuration of PPDP make extensive use of DP-based methods [bib_ref] Data mining with differential privacy, Friedman [/bib_ref] [bib_ref] Individual differential privacy: a utility-preserving formulation of differential privacy guarantees, Soria-Comas [/bib_ref] [bib_ref] Evaluating Laplace noise addition to satisfy differential privacy for numeric data, Sarathy [/bib_ref]. Meanwhile, several studies for a noninteractive setting reported a PD-dependent approach [bib_ref] Privacypreserving utility verification of the data published by noninteractive differentially private mechanisms, Hua [/bib_ref]. Researchers have expanded the concepts used to anonymize tabular data to protect the privacy of SN users [bib_ref] Differential privacy for edge weights in social networks, Li [/bib_ref] [bib_ref] Evolutionary algorithms for k-anonymity in social networks based on clustering approach, Yazdanjue [/bib_ref] [bib_ref] K-anonymity for social networks containing rich structural and textual information, Hao [/bib_ref].
Most images on the internet are in a compressed form. Hence, various studies design some techniques for AMBTCcompressed images. Data concealment has become an active research area. We can hide the data by adding confidential information to the cover image, and as a result, we get the stego image. ere are two types of data hiding schemes: one is irreversible [bib_ref] An improved scheme for data hiding in encrypted H. 264/AVC videos, Xu [/bib_ref] [bib_ref] Improving stego image quality in image interpolation based data hiding, Yang [/bib_ref] [bib_ref] High-capacity steganography: a global-adaptive-region discrete cosine transform approach, Rabie [/bib_ref] [bib_ref] Data hiding scheme improving embedding capacity using mixed PVD and LSB on..., Jung [/bib_ref] , and the other is a reversible data hiding scheme [bib_ref] Adaptive reversible data hiding by extending the generalized integer transformation, Qiu [/bib_ref] [bib_ref] High-fidelity reversible data hiding using directionally enclosed prediction, Chen [/bib_ref] [bib_ref] Separable reversible data hiding in encrypted JPEG bitstreams, Qian [/bib_ref]. A cipher text designated for data collection can be re-encrypted as designated for another by a semitrusted proxy without decryption [bib_ref] An improved proxy Re-encryption scheme for IoT-based data outsourcing services in clouds, Lin [/bib_ref]. e first concrete construction of collusion-resistant unidirectional identity-based proxy re-encryption scheme, for both selective and adaptive identity, is proposed in the paper [bib_ref] Collusionresistant identity-based proxy Re-encryption: lattice-based constructions in standard model, Dutta [/bib_ref]. One of the data hiding schemes is the histogram shifting scheme [bib_ref] A novel background-weighted histogram scheme based on foreground saliency for mean-shift tracking, Wang [/bib_ref] [bib_ref] Histogram shifting in encrypted images with public key cryptosystem for reversible data..., Li [/bib_ref] [bib_ref] Rate and distortion optimization for reversible data hiding using multiple histogram shifting, Wang [/bib_ref] , and it is the most widely used. A histogram-shifting data hiding scheme [bib_ref] Reversible data hiding, Ni [/bib_ref] that detects pixel histograms in the cover image is introduced. When big and diverse data are distributed everywhere, we cannot control the vicious attacks. erefore, we need a cryptosystem to protect our data [bib_ref] A cell-array-based multibiometric cryptosystem, Kumar [/bib_ref] [bib_ref] Ciphertext-only attack on linear feedback shift register-based Esmaeili-Gulliver cryptosystem, Lee [/bib_ref] [bib_ref] Efficient cryptosystem approaches: S-boxes and permutation-substitution-based encryption, Belazi [/bib_ref].
Some identity-based signature (IBS) schemes [bib_ref] Short signature from the bilinear pairing, Zhang [/bib_ref] [bib_ref] An efficient content source verification scheme for multi-receiver in NDN-based Internet of..., Hussain [/bib_ref] [bib_ref] Identity-based signature scheme using random oracle model, Sharma [/bib_ref] [bib_ref] A key management scheme based on pairing-free identity based digital signature algorithm..., Yuan [/bib_ref] are introduced that are based on bilinear pairing. However, the authentication schemes based on bilinear pairing over elliptic curve are more efficient and safer than traditional public key infrastructure [bib_ref] Efficient online/offline identity-based signature for wireless sensor network, Liu [/bib_ref] [bib_ref] PF-IBS: pairing-free identity based digital signature algorithm for wireless sensor networks, Sharma [/bib_ref]. e paper [bib_ref] PAC: privacy preserving proxy re-encryption for access control in public cloud, Chaudhari [/bib_ref] proposed a preserving proxy re-encryption scheme for public cloud access control. A differential attack is performed on one-to-many order preserving encryption OPE by exploiting the differences of the ordered ciphertexts in [bib_ref] Security analysis on one-to-many order preserving encryption-based cloud data search, Li [/bib_ref]. Another scheme is proposed, which consists of a cancelable biometric template protection scheme that is based on the format-preserving encryption and Bloom filters. Some of the researchers also use the concept of paring free identity-based signature schemes [bib_ref] TinyPK: securing sensor networks with public key technology, Watro [/bib_ref] [bib_ref] A method for obtaining digital signatures and public-key cryptosystems, Rivest [/bib_ref] [bib_ref] New directions in cryptography, Diffie [/bib_ref] [bib_ref] A dynamic user authentication scheme for wireless sensor networks, Wong [/bib_ref]. A lightweight proxy re-encryption scheme with certificate-based and incremental cryptography for fog-enabled e-healthcare is proposed in [bib_ref] A lightweight proxy Re-encryption approach with certificate-based and incremental cryptography for fog-enabled..., Hassan [/bib_ref].
# Research methodology
e objective of this SLR is to evaluate, investigate, and identify the existing research in the context of data storage security in cloud computing to find and evaluate all the existing techniques. SLR is a fair and unbiased way of evaluating all the existing techniques. is way provides a complete and evidence-based search related to a specific topic. At this time, there is no SLR conducted on data storage security techniques that explains all the cryptographic and noncryptographic techniques. Hence, this SLR fulfills the gap by conducting itself. is SLR aims to provide a systematic method using the guidelines of an SLR provided by Kitchenham. Furthermore, to increase the intensity of our evidence, we follow another study that is provided by [bib_ref] Impact analysis and change propagation in service-oriented enterprises: a systematic review, Alam [/bib_ref]. Our SLR consists of three phases, namely planning, conducting, and reporting. By following these three phases, we conduct our SLR, as shown in [fig_ref] Figure 1: Review procedure. [/fig_ref].
## Research questions.
e primary research question of this systematic literature review is "What types of data protection techniques have been proposed in cloud computing?" is primary research question is further divided into four RQs. All these four questions are enlisted below. ). is string/query is constructed using a population, intervention, control, and outcomes (PICO) structure that consists of population, intervention, and outcome. Database search queries are given in [fig_ref] Table 2: Databases search query [/fig_ref].
Population: "cloud computing" Intervention: "data security," "data privacy," "data integrity" Using the PICO structure, we construct a general query for the electronic database. Generic: (("Document Title": cloud * ) AND ("Document Title": data AND (privacy OR protect * OR secure * OR integrity * ))).
## Procedure of study
Selection. e procedure of study selection is described in [fig_ref] Figure 2: Study selection procedure [/fig_ref]. is procedure has three phases: the first one is exclusion based on the title, in which articles are excluded based on the title, and the relevant titles are included. e second is exclusion based on the abstract in which articles are excluded. By reading the abstract of the articles, the most relevant abstract is included, and the last one is exclusion based on a full text that also includes quality assessment criteria.
## Eligibility control.
In this phase, all the selected papers are fully readied, and relevant papers are selected to process our SLR further. [fig_ref] Table 3: Results from electronic databases [/fig_ref] shows the final selected papers from each database based on inclusion and exclusion criteria. e related papers are selected based on inclusion and exclusion criteria, which are given in [fig_ref] Table 4: Inclusion and exclusion criteria [/fig_ref].
## Inclusion and exclusion criteria.
We can use the inclusion and exclusion criteria to define eligibility for basic study selection. We apply the inclusion and exclusion criteria to those studies that are selected after reading the abstract of the papers. e criteria for inclusion and exclusion are set out in [fig_ref] Table 4: Inclusion and exclusion criteria [/fig_ref]. [fig_ref] Table 4: Inclusion and exclusion criteria [/fig_ref] outlines some of the conditions that we have applied to the articles. After applying the inclusion and exclusion criteria, we get relevant articles, which we finally added to our SLR. e search period is from 2010 to 2021, and most of the papers included in our SLR are from 2015 to onward.
We apply inclusion and exclusion criteria in the third phase of the study selection process, and we get 139 results. After that, we also apply quality criteria, and finally, we get 52 articles, which are included in this SLR. Most of the articles are taken from Elsevier and IEEE electronic databases. IEEE is the largest Venus for data storage security in Database name Search query IEEE xplore (("Document Title": cloud * ) AND ("Document Title": data AND (privacy OR protect * OR secure * OR integrity * ))) Wiley "Cloud computing" in Title and "data AND (privacy OR protect * OR secure * OR integrity * )" in Title Springer link (("Document Title": cloud * ) AND ("Document Title": data AND (privacy OR protect * OR secure * OR integrity * ))) ACM acmdlTitle:(+"cloud computing" +data privacy protect * secure * integrity * ) Elsevier ((Document Title: cloud computing * ) AND (Document Title: data AND (privacy OR protect * OR secure * ))) Hindawi (("Document Title" cloud) AND ("Document Title" data AND (privacy OR protect OR secure OR integrity))) Computational Intelligence and Neuroscience 5 cloud computing. e ratio of the selected articles from different electronic databases is shown in [fig_ref] Figure 3: Percentage of selected studies [/fig_ref].
## Quality assessment criteria.
Quality checking/assessment is done in the 3 rd phase of the study selection process. A scale of 0-1 is used for the quality assessment (QA) of the articles.
Poor-quality articles get 0 points on the scale, and goodquality articles get 1 point on the scale. e articles with 1 point on the scale are included in this SLR. Hence, by applying the quality checking/assessment criteria on all the articles, we finally get 52 articles. All the selected papers have validity and novelty for different data protection techniques, and also, we find the relevance of the articles in the quality assessment criteria, which ensures that all the articles are related to the SLR (data storage protection and privacy in cloud computing). e quality checking (QC) criteria are given in [fig_ref] Table 5: Quality checking criteria [/fig_ref].
## Taxonomy of the data protection techniques.
In this section, all the data protection techniques are depicted in . All the data protection techniques are arranged and classified in their related categories. e purpose of the taxonomy is to give a presentational view of all the data protection techniques. e data protection techniques are mainly divided into two categories, namely (1) noncryptographic techniques and (2) cryptographic techniques.
# Results and discussions
Data protection on the cloud is done by developing a thirdparty proxy that is trusted by the user. e trusted proxy is not a physical entity. It is a logical entity that can be developed on the user end (like on the user's personal computer) or at that location on which the user can trust. Mostly, all the local proxies are used as an additional service or as an additional module (like browser plugins). To fulfill the objective of data protection by proxies, some requirements are needed to fulfill necessarily. e requirements are given below:
(1) User privilege. ere are several objectives of user privilege or user empowerment, however, the main objective is to increase the trust of the users in data protection proxies used by the cloud.
(2) Transparency. Another important objective is that when users outsource their sensitive data to trusted proxies, their data should remain the same and should not be altered.
(3) Cloud computing provides large computing power and cost saving resources. However, one concern is that if we increase data security, computation overhead should not increase. We want to minimize the computation overhead over the proxies.
(4) Cloud functionalities preservation. Cloud functionalities preservation is the most important objective. e users encrypt their sensitive data on their personal computers by applying different encryption techniques to increase the protection of their data, however, by applying these different encryption techniques, they are not able to avail some of the
## Qc1
Are the goals and objectives of the paper described?
## Qc2
Are there any concise and clear limitations and statements? QC3
Does the research design support state objectives? QC4
Is the proposed technique providing any validation? [bib_ref] Heterogeneity in mobile cloud computing: taxonomy and open challenges, Sanaei [/bib_ref] Computational Intelligence and Neuroscience cloud functionalities because of compatibility issues [bib_ref] Outsourcing scalar products and matrix products on privacy-protected unencrypted data stored in..., Domingo-Ferrer [/bib_ref]. Hence, it is the main issue. [fig_ref] Figure 5: Data [/fig_ref] provides a data workflow for protecting sensitive data on the cloud using a local proxy. ere are different types of the assumption that are made for data protection, and some of them are discussed below.
(i) Curious CSPs, the most commonly used model in cloud computing, is given in the literature [bib_ref] Data security and privacy preservation in cloud storage environments based on cryptographic..., Kaaniche [/bib_ref]. e cloud service provider honestly fulfills the responsibilities, i.e., they do not interfere in the user activities, and they only follow the stander protocols. e CSP is honest, however, sometimes, it is curious to analyze the users' queries and analyze their sensitive data, which is not good because it is against the protocol. Also, by this, the privacy of the user is compromised. Hence, we can avoid these things by applying some data protection techniques on the user end to protect the users' sensitive data from the CSPs. (ii) In some cases, CSPs may collaborate with data protection proxies that are present on the users' sides to increase the level of trust between the users and CSPs because better trust can motivate more users to move to the cloud. is collaboration can be done if CSPs provide some services to the users with a stable interface for storing, searching, and computing their data. (iii) A multicloud approach to cloud computing infrastructure has also been proposed to improve their performance. In this regard, multiple cloud computing services are provided in the same heterogeneous architecture [bib_ref] Dynamic provable data possession, Erway [/bib_ref]. A multicloud gives the user multiple different places to store their data at their desired location. ere are several benefits to use a multicloud, e.g., it reduces reliance on a single CSP, which increases flexibility.
## Rq1: what type of data protection techniques has been
Proposed in Cloud Computing? In this session, we will discuss all the techniques for data storage security over the cloud. All these techniques are divided into two main categories, namely (i) cryptographic techniques and (ii) noncryptographic techniques. e local proxy uses different techniques to protect data that are stored on the cloud. Because of this reason, we cannot gain all the advantages of cloud services. erefore, we analyze and compare all these techniques based on different criteria. ese different criteria are as follows: (i) the data accuracy of all the techniques, (ii) the data protection level of all the techniques, (iii) all the functionalities these schemes allow on masked and unmasked data, and (iv) the overhead to encrypt and decrypt data over the cloud.
## Noncryptographic techniques.
ere are some noncryptographic techniques, and we discuss them in this paper as follows:
(1) Data Anonymization. Data anonymization is a data privacy technique used to protect a user's personal information. is technique hides the person's personal information by hiding the person's identifier or attributes that could reveal a person's identity. Data anonymization can be done by applying various mechanisms, for example, by removing or hiding identifiers or attributes. It can also be done by encrypting the user's personal information. e : Taxonomy of the data protection techniques.
Computational Intelligence and Neuroscience main purpose of performing data anonymization is that we can hide the identity of the person in any way. Data anonymity can be defined as the user's personal data being altered in such a way that we cannot directly or indirectly identify that person, and the CSP cannot retrieve any person's personal information. Data anonymization techniques have been developed in the field of statistical control disclosure. ese techniques are most often used when we want to outsource sensitive data for testing purposes. Data anonymization is graphically represented in [fig_ref] Figure 6: Data anonymization flow diagram [/fig_ref]. Data anonymization techniques are most often used when we want to outsource sensitive data for testing purposes. For example, if some doctors want to diagnose certain diseases, some details of these diseases are required for this purpose. is information is obtained from the patients that suffer from these diseases, but it is illegal to share or disclose anyone's personal information. However, for this purpose, we use data anonymization technique to hide or conceal the person's personal information before outsourcing the data. In some cases, however, the CSP wants to analyze the user's masked data. In the data anonymization technique, attributes are the most important part. Attributes can include name, age, gender, address, salary, etc. [fig_ref] Table 6: Identifiers classification [/fig_ref] shows the identifiers classification.
Data anonymization can be performed horizontally or vertically on this table and also on the record or group of records. e attributes are further classified into the following categories.
(i) Sensitive Attributes: sensitive attributes possess sensitive information of the person, such as salary, disease information, phone number, etc. ese attributes are strongly protected by applying some protection techniques. (ii) Nonsensitive Attributes: these types of attributes do not belong to any type of category. Hence, they do not disclose the identity of a person. (iii) Identifiers: identifier belongs to the identity of a person, such as Id card, name, social security number, etc. Because of the presence of these identifiers, the relationship between different attributes can be detected. Hence, these identifiers must be replaced or anonymized.
(iv) Quasi-Identifiers: quasi-identifiers are the group of identifiers that are available publicly, such as zipcode, designation, gender, etc. Separately, these identifiers cannot reveal the personal identity, however, by combining them, they may reveal the identity of the person. Hence, we want to separate these quasi-identifiers to avoid the discloser.
ere are two main categories of data masking: (1) perturbative masking and (2) nonperturbative masking.
(1) Perturbative Masking
In perturbation, masking data is altered or masked with dummy datasets. Original data is replaced with dummy data, however, this data looks like the original data with some noise addition. e statistical
[formula] Identifier Categorical Numerical Name ✓ × Age × ✓ Gender ✓ × Address ✓ × Zip-code × ✓ Designation ✓ × Salary information × ✓ Diseases ✓ × [/formula]
## 5.
Search on Data
# Masked results
Local proxy properties of the original data are present in the masked data, however, nonperturbative masking does not contain the statistical properties of original data, because in perturbation masking, data is altered or masked with physically same but dummy data. Some of the perturbative masking methods are given below.
(i) Data swapping In data swapping, the data is randomly changed with the same but dummy data between different records [bib_ref] Utilitypreserving privacy protection of nominal data sets via semantic rank swapping, Rodriguez-Garcia [/bib_ref]. However, if the numerical values are present in the dataset, then in certain limits, the values can be changed. Otherwise, the meaning of the data is changed. e masked data cannot look like the original data. For those attributes that can be ranked, the attribute is replaced with the nearby ranked attributes, and a very large difference between ranks is not suitable [bib_ref] Dependency-Aware tensor scheduler for industrial AI applications: dymem-an aggressive data-swapping policy for..., Rang [/bib_ref]. In data swapping, higher-level attributes are swapped [bib_ref] Reverse mapping to preserve the marginal distributions of attributes in masked microdata, Muralidhar [/bib_ref] and individual values are not changed.
## (ii) noise addition
In this mechanism, some noise is added to the original dataset to alter the original data. Noise is only added to the data that is continuous and divided into categories [bib_ref] A semantic framework for noise addition with nominal data, Rodriguez-Garcia [/bib_ref]. e noise is added into all the attributes that are present in the original dataset, such as sensitive attributes and also quasi-attributes.
## (iii) microaggregation
In this technique, all the relevant data is stored into different groups, and these different groups release average values from each record [bib_ref] Does $ k $-Anonymous microaggregation affect machine-learned macrotrends?, Rodríguez-Hoyos [/bib_ref]. If a large number of similar records is present in different groups, then more data utility is done. We can cluster the data in many ways, e.g., in categorical versions [bib_ref] Efficient k-anonymous microaggregation of multivariate numerical data via principal component analysis, Monedero [/bib_ref]. Microaggregation is done on a quasi-attribute to protect these attributes from reidentification, and the quasiattributes protect all the other attributes from reidentification. We can also minimize reidentification by data clustering [bib_ref] T-closeness through microaggregation: strict privacy with enhanced utility preservation, Soria-Comas [/bib_ref].
[formula] (iv) Pseudonymization [/formula]
In this method, the original data is replaced with artificial datasets [bib_ref] Secure and efficient pseudonymization for privacy-preserving vehicular communications in smart cities, Bouchelaghem [/bib_ref]. In this technique, each attribute present in the original data is a pseudonym, and by doing this, data is less identifiable.
(2) Nonperturbative Masking Nonperturbative masking does not change or alter the original data, however, it changes the statistical properties of the original data. Mask data is created by the reduction of the original data or suppressions of the original data. Some methods that are used for nonperturbative masking are as follows:
[formula] (i) Bucketization [/formula]
In this method, original data is stored in different buckets, and these buckets are protected through encryption [bib_ref] Range based queries over order preserving encrypted data, Vasanth [/bib_ref]. We can protect the sensitive attributes through bucketization.
## (ii) slicing
Data slicing is a method in which a larger group of data is divided into smaller slices or segments [bib_ref] Automated data slicing for model validation: a big data-AI integration approach, Chung [/bib_ref]. Hence, we can slice the data, and in this way, the sensitive attribute and the quasi-attributes are divided into different slices. By identifying the individual slice, the identity of the person cannot be disclosed.
[formula] (iii) Sampling [/formula]
Sampling is a technique in which the population and sample concept is present. e entire data is called population, and the masked data is called a sample. In this technique, we make different samples of the original data. A smaller data sample provides more protection [bib_ref] Bisample: bidirectional sampling for handling missing data with local differential privacy, Sun [/bib_ref].
[formula] (iv) Generalization [/formula]
It is a technique in which some additional attributes are added to the record. If the number of quasi-attributes is less rare, then some dummy attributes are added into the record, which look like the quasi-attributes. Hence, by doing this, reidentification becomes more difficult [bib_ref] Privacy protection of textual attributes through a semanticbased masking method, Martínez [/bib_ref]. By applying generalization on data, we can protect the identity of a person because it hides the relationship between the quasiattributes.
e summary of data anonymization techniques is given in [fig_ref] Table 7: e summary of data anonymization techniques [/fig_ref].
(2) Data Splitting. Data splitting is a technique in which sensitive data is divided into different fragments [bib_ref] Validation in prediction research: the waste by data splitting, Steyerberg [/bib_ref] to protect it from unauthorized access. In this technique, we first split the data into different fragments, then these fragments are randomly stored on different clouds. Even if the intruder gains access to a single fragment in any way, still the intruder will not be able to identify the person. For example, if an intruder gets a fragment from the cloud that contains the salary information of an organization, it is useless until he knows which salary belongs to which person. Hence, data splitting is a very useful technique for protecting data stored on the cloud.
Local proxies outsource data to the cloud without splitting the data, and they can also split the data first and then outsource to the same cloud using different accounts in the same CSP. It can also store data on different cloud platforms that run through different CSPs but provide some of the same services. Data is split before storing in different locations because even if some part or piece of data is known to an intruder, they will not be able to identify anyone.
Firstly, the local proxy retrieves sensitive data from the user and then calculates the risk factor for disclosure. In this method, the user can define the privacy level, and this privacy level provides information about all the sensitive attributes that can reveal someone's identity. ese sensitive attributes are called quasi-attributes or quasi-identifiers. Next, the local proxy decides the number of pieces into which the sensitive data will be split and the number of locations that will be needed to store those pieces. erefore, no one can reveal a person's identity, and all this information about the data splitting mechanism is stored at the local proxy. However, the system must be able to function properly and respond to the queries on time. After that, the local proxy stores these different data fragments in different cloud databases, and now, they are free from disclosure. e data-splitting mechanism supports almost all the functions of the cloud. Hence, we can use almost all the services provided by CSP using the data-splitting mechanism for storing data in the cloud.
When the users want to retrieve the original data, they process a query on a local proxy. e query is processed, and the data storage locations are retrieved from the local database. After that, the query is replicated as many times as the data is split into fragments, and these queries are forwarded to the relevant CSPs. As a result, each CSP provides a set of results that represent a partial view of the complete result. Finally, the proxy collects partial results according to the criteria used to split the data and provides the complete result to the user. Mostly, all these fragments are stored on different cloud databases in their original structure. erefore, computation on these fragments can be performed easily. However, there is a problem if we want to perform computation separately on the individual fragment.
en, there is no algorithm that exists for this computation. erefore, some algorithms are required to perform these types of computation as this computation requires communication between different CSPs. e redundancy of proxy metadata and backup policies must be essential to ensure the robustness of the mechanism. e data-splitting is graphically represented in [fig_ref] Figure 7: Data-splitting flow diagram [/fig_ref].
e summary of the data-splitting is given in [fig_ref] Table 8: e summary of the data-splitting techniques [/fig_ref]. Different data-splitting techniques are used for the protection of data stored on the cloud. Some of these are given below.
## (i) byte level splitting
In this type, all the sensitive data is converted into bytes [bib_ref] Data privacy protection using multiple cloud storages, Zhang [/bib_ref]. en, these bytes are randomly shuffled with each other. After that, all the bytes are recombined. Fixed length fragments are made, and then, these fragments are stored on a different cloud.
## (ii) privacy level splitting
In this mechanism, the user chose the privacy level of each file [bib_ref] An approach to protect the privacy of cloud data from data mining..., Dev [/bib_ref] that is to be stored on a cloud database. Hence, a privacy level is attached with the file that is to be stored on the cloud. Using this privacy level, the user can decide that the higher privacy level files should be stored on the trusted cloud.
## (iii) byte level splitting with replication
Byte-level data splitting is combined with data replication to improve both performance and security. e author of the paper [bib_ref] DROPS: division and replication of data in cloud for optimal performance and..., Ali [/bib_ref] proposed an algorithm to store the data fragments on different clouds, so that they are at a certain distance and by doing this; we can avoid confabulation attacks where the intruder can aggregate the split fragments. (iv) Byte level splitting with encryption Firstly, byte-level data splitting [bib_ref] Security-Aware efficient mass distributed storage approach for cloud systems in big data, Gai [/bib_ref] [bib_ref] A multi-cloud approach for secure data storage on smart device, Alqahtani [/bib_ref] is proposed. In this scheme, every fragment of data is encrypted to enhance the security of sensitive data. In this mechanism, the data is split into bytes, and these bytes are randomly shuffled and finally recombined. is type of data splitting is suitable for binary or multimedia files that are not processed through the cloud. Another problem is the length of a fragment in which we can say that the data cannot be reidentified or the identity of a person cannot be revealed. If the length is too short, then the probability of disclosure increases, and if the length is too long, then it is difficult to handle these fragments. Computational Intelligence and Neuroscience
Hence, it should have a certain length so that we can also protect the identity of a person. ere is another type of data splitting in which we split data into attributes. e attribute level splitting is performed in two ways: one is horizontal splitting and the second is vertical splitting. ese types of splitting are mostly done on structural databases, and they provide strong privacy.
(v) Vertical splitting
In vertical data splitting [bib_ref] A hybrid solution for privacy preserving medical data sharing in the cloud..., Yang [/bib_ref] [bib_ref] Distributing data for secure database services, Ganapathy [/bib_ref] , we divide quasi-identifiers or quasi-attributes in such a way that all the risky attributes are divided into different fragments to secure the reidentification. Some of the sensitive fragments required encryption on it. Hence, we can encrypt these fragments by applying some encryption algorithms or by applying some other privacy methods to increase the security level.
A solution for sensitive data splitting without performing encryption on fragments is proposed [bib_ref] Selective data outsourcing for enforcing privacy, Ciriani [/bib_ref]. is mechanism is suitable for data on which we want to perform some computation, because on encrypted data, we cannot perform computation directly. Another technique has been proposed [bib_ref] C-sanitized: a privacy model for document redaction and sanitization, Sánchez [/bib_ref] , which demonstrates the redaction and sanitization of a document that identifies all sensitive attributes and protects the data in most documents. e schemes that use vertical splitting to protect data are faster than other splitting techniques because data fragments consist of a single attribute or multiple attributes. It does not involve data masking or encryption. Hence, the computation is easy. ere is another type of encryption in which we do not encrypt and decrypt every time to perform computation. It is called homomorphic encryption. In this case, all data modification is done on encrypted data, and actual data is not changed, however, the final result is preserved [bib_ref] CryptDICE: distributed data protection system for secure cloud data storage and computation, Rafique [/bib_ref].
(3) Steganography. Steganography is the practice of concealing a message within another message or a physical object. In computing contexts, video, audio, image, message, or computer file is concealed within another image, message, or file. e steganography flow diagram is depicted in [fig_ref] Figure 8: Steganography [/fig_ref]. ere are two main types of steganography, namely (1) linguistic steganography and (2) technical steganography. ese techniques are given as follows:
(1) Linguistic Steganography is technique is further divided into two subtechniques, which are given below.
## (i) semagrams
It uses images and symbols alone to cover the data. ere are two types of Semagrams [bib_ref] Hiding short secret messages based on linguistic steganography and manual annotation, Muñoz [/bib_ref]. e first is a visual Semagram. In this type, we In this case, we hide the real message from the intruder by installing the original massage in an authorized carrier [bib_ref] RNN-stega: linguistic steganography based on recurrent neural networks, Yang [/bib_ref]. Open code technique is further divided into two types: one is jargon code, and the second is covered ciphers.
(2) Technical Steganography Technical steganography uses different cover media to hide secret data [bib_ref] An overview of steganography, Kessler [/bib_ref]. ere are different types of cover media, such as text, image, audio, and video. Different types of methods are applied on the covers to hide or embed the secret data into the covers.
## (i) covers
Cover data is like a vehicle for sending data from one place to another place. Secret messages in cover data exist in two forms: one is streaming, and the second is block [bib_ref] Image steganography using genetic algorithm for cover image selection and embedding, Shyla [/bib_ref]. Different types of covers exist, such as text, image, audio, and video. Four different types of steganographic methods are applied on the cover given as blow.
(a) Text steganography In this type, we change some textual characteristics of text, such as the font, color, or symbols of the text message [bib_ref] A compression-based text steganography method, Satir [/bib_ref]. ree coding techniques are used to change these textual features, which are as follows: (1) line-shift coding, (2) word-shift coding, and (3) feature coding.
## (b) image steganography
It is the most popular type of steganography. Image steganography refers to the process of hiding sensitive data inside an image file [bib_ref] Large-Capacity image steganography based on invertible neural networks, Lu [/bib_ref]. e transformed image is expected to look very similar to the original image because the visible features of the stego image remain the same. e image steganography is divided into three parts, namely (1) least significant bits coding, (2) masking and filtering, and (3)
## Transformations. (c) audio steganography
Audio steganography is a technique that is used to transmit secret data by modifying a digitalized audio signal in an imperceptible manner [bib_ref] Learning to generate steganographic cover for audio steganography using gan, Chen [/bib_ref]. Following types of audio steganography are given: (1) least significant bits coding, (2) phase coding, (3) spread spectrum, and (4) echo hiding.
## (d) video steganography
In video steganography, both image and audio steganography are used [bib_ref] IoT using machine learning security enhancement in video steganography allocation for Raspberry..., Karthika [/bib_ref]. A video consists of many frames. Hence, video steganography hides a large amount of data in carrier images. In this type of steganography, we select the specific frame in which we want to hide the sensitive data.
(ii) Methods Two types of methods are used in steganography:
(1) spatial domain and (2) frequency domain.
## (a) frequency domain
A frequency-domain steganography technique is used for hiding a large amount of data with no loss of secret message, good invisibility, and high security [bib_ref] Novel DCT and DST based video steganography algorithms over non-dynamic region in..., Patel [/bib_ref]. In the frequency domain, we change the magnitude of all of the DCT coefficients of the cover image. ere are two types of frequency domain: (1) discrete cosine transformation and (2) discrete wavelet transformation. (b) Spatial Domain e spatial domain is based on the physical location of pixels in an image [bib_ref] Review on spatial domain image steganography techniques, Sirisha [/bib_ref]. A spatial domain technique gives the idea of pixel regulation, which minimizes the progressions of a stego image created from the spread image. Some methods of the spatial domain are given as follows: (1) e summary of the steganographic techniques is given in [fig_ref] Table 9: e summary of the steganographic techniques [/fig_ref].
## Cryptographic techniques.
Cryptography is the most important and most widely used technique for security purposes. In cryptography, the plain text is converted into ciphertext using a key and some encryption algorithms. Cryptographic techniques are the most secure techniques among all the other security techniques. Hence, these cryptography techniques are widely used in data storage security over the cloud. e present day's cryptography techniques are more realistic. We can achieve different objectives by applying these cryptographic techniques, for example, data confidentiality and data integrity. Because of an increase in the number of data breaches in the last few years, some cloud service provider companies are shifting toward cryptographic techniques to achieve more security. e most commonly used cryptographic technique is AES. Key management is an important issue in cryptographic techniques because if the key is hacked by an intruder, then all the data will be hacked or stolen by this intruder. Hence, key protection or key management is a very important issue. erefore, it is mostly the responsibility of CSP to manage the key and also provide the protection of key. Cryptographic techniques also protect the user from an untrusted CSP because sometimes the CSP outsources sensitive data without taking the permission of users, and it is an illegal activity. Hence, to avoid these things and protect our sensitive data from untrusted CSPs, we use cryptographic techniques, and it is the best option for users. However, there are some difficulties the user has to face while using cryptographic techniques, i.e., if a user wants to update a small amount of data, the user needs to decrypt the data and then perform this minor update. Hence, this work is very costly. Over time, implementing cryptographic techniques gives us a higher level of security, however, we compromise on performance or speed. It all depends on the user, the standard, the performance, or the high level of security the user wants to achieve. In this paper, we are focusing on the four main functionalities that are required or needed on cloud computing when using cryptographic techniques. [fig_ref] Figure 9: Encryption flow diagram [/fig_ref] shows the flow diagram of encryption.
Some of the main functionalities of cryptographic functions are given below.
## (i) search on encrypted data
If a user wants to retrieve their data stored in a cloud database, they generate a query and run the query on a local proxy server and search for the data they want. Searching for encrypted data is a very important part of cryptography because every user who stores their sensitive data in a cloud database wants to retrieve it, and it is done by searching their sensitive data through queries. erefore, the procedure of retrieving their data is very difficult.
## (ii) storage control
Sometimes the user wants to store data in a desired location or trusted database. Hence, the user must have full control over the storage of data.
## (iii) access control
It is a very important control and is referred to as data access restriction. Sometimes, the user does not want to share a private file publicly. Hence, access control is an important functionality. (iv) Computation on data Data computation is the main functionality of cloud computing. Sometimes, the user wants to perform some computation on data that are stored on a cloud database. For example, if a user wants to perform computation on encrypted data that is stored on cloud databases, then there are two ways. One is that the user, firstly, decrypts the entire data, performs computation on the data, and finally, the user encrypts the entire data and stores on the cloud database. is process is very expensive in terms of computation.
Some of the cryptographic techniques are as follows:
(1) Homomorphic Encryption. Homomorphic encryption is a form of encryption that permits users to perform computations on encrypted data without decrypting it. ese resulting computations are left in an encrypted form, which, when decrypted, result in an identical output to that produced had the operations been performed on the unencrypted data.
ere are some types of homomorphic encryption that are described below.
## (i) partial homomorphic encryption
In partial homomorphic encryption, only one arithmetic function addition or multiplication is performed at one time. If the resultant ciphertext is the addition of the plain text, then it is called an additive homomorphic scheme, and if the resultant ciphertext is the multiplication of the plaintext, then it is called the multiplicative homomorphic scheme. Two multiplicative homomorphic schemes are given as in [bib_ref] Cloud-based quadratic optimization with partially homomorphic encryption, Alexandru [/bib_ref] [bib_ref] Privacy-aware quadratic optimization using partially homomorphic encryption, Shoukry [/bib_ref]. ere is one additive homomorphic scheme that is called Paillier [bib_ref] Practical implementation of privacy preserving clustering methods using a partially homomorphic encryption..., Catak [/bib_ref].
(ii) Somewhat Homomorphic Encryption is technique allows the user to perform the multiplication and subtraction mathematical operations. However, this scheme allows a limited number of arithmetic operations, because if it allows a large number of arithmetic operations, then it produces noise.
is noise changes the structure of the original data. Hence, limited numerical math operations are allowed. ere is a somewhat homomorphic encryption scheme that Computational Intelligence and Neuroscience 13 is presented by the authors of the papers [bib_ref] High-capacity reversible data hiding for encrypted multimedia data with somewhat homomorphic encryption, Xiong [/bib_ref] [bib_ref] High-performance FV somewhat homomorphic encryption on GPUs: an implementation using CUDA, Badawi [/bib_ref]. In this scheme, the time of encryption and decryption is increased when multiplication operations are increased. To avoid this increase in time, we allow only a limited number of mathematical operations. (iii) Fully Homomorphic Encryption is technique allows a large number of arithmetic operations, namely multiplication and subtraction. Multiplication and addition in this technique are performed in the form of XOR and AND gates [bib_ref] TFHE: fast fully homomorphic encryption over the torus, Chillotti [/bib_ref]. Completely homomorphic encryption techniques require a higher computation time to encrypt and decrypt data. erefore, this technique is not applicable in real-life applications for implementation.
is technique uses a bootstrapping algorithm when a large number of multiplication operations is performed on data and also for the decryption of the data it is used. Homomorphic encryption, on the other hand, represents the tradeoff between operations and speed performance. Only a limited number of arithmetic operations are allowed if someone wants low computation, and a large number of arithmetic operations are allowed if someone wants high security. It depends on the needs of the user.
(2) Searchable Encryption. A searchable encryption technique is proposed by the author of the paper [bib_ref] Blockchain-assisted Secure fine-grained Searchable Encryption for a Cloud-Based Healthcare Cyber-Physical System, Gupta [/bib_ref]. In this technique, before storing data on a cloud database, encryption is performed, and after that, it is stored on the cloud. e advantage of this technique is that when we search for some data over the cloud database, this technique provides a secure search over the cloud database.
## I) searchable asymmetric encryption
Over the past two decades, we have focused on searchable encryption. Much of the work is related to the multiwriter and single-reader cases. Searchable encryption is also called public keyword search encryption along with keyword search (PEKS) [bib_ref] Public-key authenticated encryption with keyword search achieving both multi-ciphertext and multitrapdoor indistinguishability, Pan [/bib_ref].
## (ii) searchable symmetric encryption
Symmetric-key algorithms use the same key for massage encryption and ciphertext decryption. e keys can be the same, or there can be a simple transformation to go between the two keys. Verifiable searchable symmetric encryption, as a key cloud security technique, allows users to retrieve encrypted data from the cloud with keywords and verify the accuracy of the returned results. Another scheme is proposed for keyword search over dynamic encrypted cloud data with a symmetric-key-based verification scheme [bib_ref] Towards achieving keyword search over dynamic encrypted cloud data with symmetrickey based..., Ge [/bib_ref].
(3) Encryption. In cryptography, encryption is the process of encoding information. is process converts the original representation of the information, known as plaintext, into an alternative form known as ciphertext. Ideally, only authorized parties can decipher a ciphertext back to plaintext and access the original information.
## (i) symmetric key encryption
Only one key is used in symmetric encryption to encrypt and decrypt the message. Two parties that communicate through symmetric encryption should exchange the key so that it can be used in the decryption process. is method of encryption differs from asymmetric encryption, where a pair of keys is used to encrypt and decrypt messages. A secure transmission method of network communication data based on symmetric key encryption algorithm is proposed in [bib_ref] A secure transmission method of network communication data based on symmetric key..., Cai [/bib_ref].
(ii) Public Key Encryption e public-key encryption scheme is proposed by the author of the paper [bib_ref] A practical public key encryption scheme based on learning parity with noise, Yu [/bib_ref]. In this scheme, a public key pair is created by the receiver. is public key pair consists of two keys. One is called a public key, which is known publicly to everyone, and the second is the private key, which is kept a secret. Hence, in this scheme, the sender performs encryption on the data using the public key of the receiver and then sends this encrypted data to the receiver. After receiving this encrypted data, the receiver can decrypt this data using the private key. Hence, in this way, we can perform secure communication between two parties.
## (iii) identity-based encryption
Identity-based encryption is proposed by the author of the paper [bib_ref] Identity-based encryption transformation for flexible sharing of encrypted data in public cloud, Deng [/bib_ref]. In this technique, a set of users is registered on the database and a unique identity is assigned to all the registered users by an admin that controls this scheme. e identity of the users can be represented by their name or their e-mail address. Just like in a public-key encryption, there is a public key pair that consists of one public key, which is the identity of the user, and one private key, which is a secret key. Just like in public-key encryption, the receiver cannot generate their public key in identitybased encryption. e identity cannot be generated by the user. ere is a central authority that generates and manage the user's identity. e identitybased encryption is improved by the author [bib_ref] Lattice-based proxy-oriented identity-based encryption with keyword search for cloud storage, Zhang [/bib_ref]. e main advantage of identity-based encryption is that anyone can generate the public key of a given identity with the help of the central main authority. (iv) Attribute-Based Encryption e authors of the papers [bib_ref] Attribute Based Encryption with Privacy protection and Accountability for CloudIoT, Li [/bib_ref] [bib_ref] An improved attribute-based encryption technique towards the data security in cloud computing, Namasudra [/bib_ref] propose a technique called attribute-based encryption. Similar to identity-based encryption, attribute-based encryption also depends on the central main authority. e central main authority generates the private key and distributes it to all the registered users. It can be encrypting the messages, however, if it does not have this designation, then it cannot be generating the messages. Attribute-based encryption is used when the number of registered users is very large. en, the attribute-based encryption is useful. e attribute-based encryption consists of two schemes, which are key policy and ciphertext policy.
(
## V) functional encryption
A functional encryption technique [bib_ref] From minicrypt to obfustopia via private-key functional encryption, Komargodski [/bib_ref] [bib_ref] Function-private functional encryption in the private-key setting, Brakerski [/bib_ref] consists of identity-based encryption, attribute-based encryption, and public-key encryption. All the functionalities of these three techniques combinedly make function encryption. In this technique, all the private keys are generated by the central main authority, which is associated with a specific function. Functional encryption is a very powerful encryption technique that holds all the functionalities of three encryption techniques. A functional encryption technique is used in many applications.
(4) Signcryption.
Cryptography is publicly open-source, and it functions simultaneously as a digital signature and cipher. Cryptography and digital signatures are two basic encryption tools that can ensure confidentiality, integrity, and immutability. In [bib_ref] An efficient and provable secure certificate-based combined signature, encryption and signcryption scheme..., Ullah [/bib_ref] , a new scheme called signature, encryption and encryption is proposed, based on effectively verifiable credentials. e system not only performs encryption and encryption but also provides an encryption or signature form only when needed [bib_ref] A lightweight and formally secure certificate based signcryption with proxy reencryption (CBSRE)..., Hussain [/bib_ref]. e paper proposes lightweight certificate-based encryption using a proxy cipher scheme (CSS) for smart devices connected to an IoT network to reduce computing and communications costs. To ensure the security and efficiency of the proposed CBSS project, we used a cipher system encoded with 80 bit subparameters. Reference [bib_ref] An Access Control Scheme Using Heterogeneous Signcryption for IoT Environments, Ullah [/bib_ref] proposes an input control scheme for the IoT environment using a cryptographic scheme corresponding to the efficiency and robustness of the UK security system. e proposed scheme shows that besides security services, such as protection against attacks, confidentiality, integrity, nonblocking, nondisclosure, and confidentiality, accounting and communication costs are low compared to the current scheme. Document [bib_ref] A lightweight and secured certificate-based proxy signcryption (CB-PS) scheme for E-prescription systems, Ullah [/bib_ref] gives the informal and formal security proof of the proposed scheme. Automated Validation of Internet Security Protocols and Applications (AVISPA) tool is used for formal security analysis, which confirms that the proposed CB-PS scheme can potentially be implemented for resource-constrained low-computing electronic devices in E-prescription systems. e proposed scheme [bib_ref] An Efficient and Secure Multi-Message and Multi-Receiver Signcryption Scheme for Edge Enabled..., Ullah [/bib_ref] introduced a new concept that does not require a reliable channel. e main production center sends a part of the private key to the public consumers. e summary of the cryptographic schemes is given in [fig_ref] Table 1: Databases sources. [/fig_ref].
All data storage protection on cloud computing is discussed in session 3.
ere are a lot of data protection techniques, however, all these techniques are only divided into three main categories, namely (i) data splitting, (ii) data anonymization, and (iii) cryptography. From different points views, we discuss all these techniques, e.g., overhead on the local proxy, computation cost, search on encrypted data, data accuracy all these techniques retained, and data protection level all these techniques have, and all the masked data techniques have the functionalities. ese are some different views, and by considering them, we can analyze all the data protection techniques. Cryptography provides highlevel security but limited cloud functionalities and a high cost of performing computation on cloud data. Data splitting provide low computation cost but a low level of security. Data anonymization is of two types: one is perturbative masking, and the second is nonperturbative masking.
Hence, in perturbative masking, data is altered with dummy data. Hence, security is high, however, we cannot perform some functionalities.
## Rq2: what are the demographic characteristics of the relevant studies?
We answer this question by considering the four following aspects: (i) publication trend, (ii) publication venues (proceeding and journals), (iii) number of citations, and (iv) author information. [fig_ref] Figure 1: Review procedure. [/fig_ref] shows all trends of all the publications Identity-based encryption [bib_ref] Identity-based encryption transformation for flexible sharing of encrypted data in public cloud, Deng [/bib_ref] [bib_ref] Lattice-based proxy-oriented identity-based encryption with keyword search for cloud storage, Zhang [/bib_ref] Use for data access control Required a valid password to access data Provides a high level of privacy Symmetric-key encryption [bib_ref] A secure transmission method of network communication data based on symmetric key..., Cai [/bib_ref] Encryption and Decryption of data using the same key
No functionality can be performed on encrypted data Provides a high level of privacy
Public-key encryption [bib_ref] A practical public key encryption scheme based on learning parity with noise, Yu [/bib_ref] Use for data access control Required a valid public key for encryption and private key for decryption
No key exchange is required. Provides a very high level of privacy Attribute-based encryption [bib_ref] Attribute Based Encryption with Privacy protection and Accountability for CloudIoT, Li [/bib_ref] [bib_ref] An improved attribute-based encryption technique towards the data security in cloud computing, Namasudra [/bib_ref] Data access control based on attributes
Less secure than public-key encryption Provides lesser privacy than public-key encryption Functional encryption [bib_ref] From minicrypt to obfustopia via private-key functional encryption, Komargodski [/bib_ref] [bib_ref] Function-private functional encryption in the private-key setting, Brakerski [/bib_ref]
## Publication venues.
ere are different types of publication venues, and some of them are book articles, conference proceedings, journals, workshop proceedings, and symposium proceedings. Hence, in our SLR, the number of publications in a different venue is given in [fig_ref] Figure 1: Review procedure. [/fig_ref]. We have a total of 52 papers after applying the inclusion and exclusion criteria in Section 2.
Out of 52 papers, 0 papers are published in book chapters. 1 paper is published in workshop proceedings. 0 papers are published in symposium proceedings. 43 papers are published in journals. 8 papers are published in conference proceedings. ere are some most active journals in cloud data security, which are enlisted in [fig_ref] Table 1: Databases sources. [/fig_ref]. e most active journal is the IEEE Access. In this journal, 6 papers are published. Journal of Cryptology is the second most active journal in the field of data storage, security, and privacy in cloud computing. In this journal, 3 papers are published. In the third journal, i.e., in the Journal of Information Fusion, 3 papers are published. e fourth journal is the Information Science. In this journal, 2 papers are published. e fifth journal is IEEE Transactions on Knowledge and Data Engineering, and in this journal, 2 papers are published. Most active conferences are given in [fig_ref] Table 1: Databases sources. [/fig_ref].
## Number of citations.
e number of citations of a paper also tells the quality of the paper. e more the number of citations, the higher the quality, and the fewer the number of citations of the paper, the lower the paper quality. [fig_ref] Table 1: Databases sources. [/fig_ref] shows the most influential authors, and [fig_ref] Figure 1: Review procedure. [/fig_ref] shows the number of citations of all the papers that we have used in this SLR. Few papers have citations of more than 100. Hence, it shows that papers have a very high quality, and hence, the citation of those papers is very high. ese papers are [bib_ref] T-closeness through microaggregation: strict privacy with enhanced utility preservation, Soria-Comas [/bib_ref] [bib_ref] A hybrid solution for privacy preserving medical data sharing in the cloud..., Yang [/bib_ref] [bib_ref] RNN-stega: linguistic steganography based on recurrent neural networks, Yang [/bib_ref] [bib_ref] TFHE: fast fully homomorphic encryption over the torus, Chillotti [/bib_ref].
## Author information.
Some authors are most active in their publication. To identify these authors, we enlist the names of the top 10 authors that are more active in the field of data protection and privacy in cloud computing. Hence, we enlist the names of the top 10 authors and also their numbers of publications in [fig_ref] Table 1: Databases sources. [/fig_ref].
## Rq3: which data protection technique provides more
Data Protection among all the Techniques? We answer this question by considering the following four aspects: (i) publication trend, (ii) publication venues (proceeding and journals), (iii) number of citations, and (iv) author information.
## Comparison of data protection techniques.
In this section, we compare all the data protection techniques that are discussed in this SLR, and finally, we review which technique is better and provides more protection among all these data protection techniques. We compare these techniques based on different functionalities, which are given as (i) local proxy overhead, (ii) data accuracy retain, (iii) level of data protection, (iv) transparency, and (v) operation supported, and finally, we discuss RQ2. [fig_ref] Table 1: Databases sources. [/fig_ref] depicts a comparison of all the data protection techniques and provides a brief comparison of all the data protection techniques discussed in this SLR. Now, we discuss all these five functionalities one by one in more detail.
## (a) local proxy overhead
(1) Encryption e overhead on the local proxy for encryption is very high because the data is encrypted. If the user wants to update the data, firstly, the user Computational Intelligence and Neuroscience decrypts the data and then updates the data. After that, the user encrypts the data again. Hence, this operation requires a lot of time, and all this work is performed by the local proxy. It is the reason the overhead on the local proxy for encryption is very high for encryption. (2) Data Splitting e overhead on a local proxy for data splitting is very low. e local proxy overhead remains constant while splitting data into fragments.
(3) Anonymization e overhead on a local proxy for anonymization is average because most of the anonymization methods require quasilinear computation in the number of records to generate the anonymized data set. Whenever the anonymized data is generated and stored in the cloud database, then there is no overhead on the local proxy. (4) Homomorphic Encryption e overhead on local proxies for homomorphic encryption is very high because homomorphic encryption involves a large number of mathematical operations. erefore, there is a lot of overhead on local proxies for homomorphic encryption. (5) Steganography e overhead on the local proxy for steganography is not too much as the data is concealed inside the cover for secure communication. However, based on the complexity of the operation in the transformed domain technique, the local proxy overhead is more than the spatial domain technique. (6) Signcryption e overhead on the local proxy for signcryption is high compared to the simple encryption because in signcryption, hashing and encryption are performed in a single logical step. Because of an extra operation in signcryption, the overhead on the local proxy is higher than the simple encryption.
## (b) data accuracy retain
(1) Encryption e data accuracy level for encryption is very high because data is encrypted by applying some algorithms. e sensitive data is encrypted by the sender, and this data is decrypted by the receiver using a key. is data cannot be read by anyone who does not have the secret key. erefore, data accuracy is very high for encryption.
(2) Data Splitting e data accuracy level for data splitting is average because data-splitting data is present in the form of fragments. erefore, CSP can easily access the fragments of data. Both encryption and data splitting are irreversible methods. Hence, we can retrieve the original data easily.
(3) Anonymization e data accuracy level for data anonymization is very low because anonymization is not irreversible. In anonymization, data is replaced with dummy data, and it cannot be retrieved back.
erefore, anonymization has a very low level of data accuracy. (4) Homomorphic Encryption e data accuracy level for homomorphic encryption is very high because data is encrypted by applying some algorithms. (5) Steganography e data accuracy level for steganography is very low as compared to the other cryptographic techniques because data is embedded inside the cover of another medium. Any change in the cover during transmission results in the change of the concealed data. erefore, it is hard to ensure a high accuracy level in steganography. e stego image contains the secrete data that is transmitted over the communication channel. Data concealed by the sender is extracted from the cover by the receiver. erefore, the concealment of data results in accurate data transmission. (6) Signcryption e data accuracy level for signcryption is also very high, because in signcryption, confidentiality and authentication are achieved. erefore, we can also verify the identity of the sender.
(c) Level of Data Protection (1) Encryption e level of data protection is very high for encryption techniques, because in encryption, data is changed into ciphertext, which cannot be understood. erefore, we can say that the identification of data is impossible without decryption using a secret key because encryption is a one-way function that is easy to execute in one direction, however, it is impossible to execute in the opposite direction.
(2) Data Splitting e level of data protection for data splitting is less high as compared to cryptographic techniques because data is split into different fragments, and these fragments contain original forms of data. Hence, if an intruder hacks or steal these fragments, then the untired data can be easily read. Hence, the data protection level is not high as compared to encrypted methods. (3) Anonymization e level of data protection for data anonymization is less high as compared to cryptographic techniques, because in anonymization techniques, quasi-identifiers are protected if the quasi-identifiers are not protected strongly.
en, there is a change in the reidentification of person-sensitive data. [bib_ref] An ADS-PAYG approach using trust factor Against economic denial of sustainability attacks..., Karthika [/bib_ref] Homomorphic Encryption e level of data protection is very high for homomorphic encryption techniques because Applicable when the user wants a medium level of data protection with low computation encryption data is changed into ciphertext, which cannot be understood. (5) Steganography e data protection level for steganography is medium because data is embedded inside the cover of another medium. e stego image contains the secrete data that is transmitted over the communication channel. Data concealed by the sender is extracted from the cover by the receiver.
erefore, the concealment of data results in secure data transmission. [bib_ref] Heterogeneity in mobile cloud computing: taxonomy and open challenges, Sanaei [/bib_ref] Signcryption e data protection level for signcryption is also very high, because in signcryption, both confidentiality and authentication are achieved. erefore, we can also verify the identity of the sender.
## (d) transparency
(1) Encryption ere is no transparency for the encrypted data, because in encryption, there is a need for key management. Hence, the local proxy needs to keep the records of all the keys and manage all these keys. erefore, there is no transparency for the encrypted data.
(2) Data Splitting ere is no transparency for the data-splitting mechanism, because in the data-splitting mechanism, data is split into different fragments, and the local proxy stores these fragments in different locations. Hence, there is a need to keep the record of the location of all the fragments that are stored on different locations.
## (3) anonymization
Anonymization is fully transparent, because in anonymization, there is no need to keep the record of data storage by the local proxy. In anonymization, data is statistically similar to the original data. Hence, CSP also performs computation and some analysis on the anonymized data. (4) Homomorphic Encryption ere is no transparency for the homomorphically encrypted data, because in encryption, there is a need for key management. Hence, the local proxy needs to keep the records of all the keys.
## (5) steganography
In steganography, as compared to other data protection techniques, the main aim is to transmit data without letting the attacker know about the data transmission as it is concealed inside the cover of another medium. e data transmission in steganography is fully transparent. No key management is required, and there is no need to keep track of data storage. (6) Signcryption ere is no transparency for the signcrypted data, because in signcryption, there is a need for key management. Hence, the local proxy needs to keep the records of all the keys and also manage all these keys.
## (e) operation supported
## (1) encryption
Only the data storage operation is supported on the encrypted data, because if the user wants to update some encrypted data that are stored on a cloud database, firstly, the user needs to decrypt this data, and then the user performs an update on this data. We cannot perform any modification operation on encrypted data. (2) Data Splitting
All the operations cloud be performed on data splitting, because in data splitting, the data is present in their original structure. Hence, we can perform data storage, search, data update, and also data computation.
## (3) anonymization
In anonymization, there are two types of data anonymization: one is data masking, and the second is data nonmasking. If data is nonmasked, then we can perform data storage and search on this data. Otherwise, we can only perform data storage.
## (4) homomorphic encryption
Only the data storage operation is supported on the encrypted data, because if the user wants to update some encrypted data that are stored on the cloud database, firstly, the user needs to decrypt this data, and then the user performs some updates on this data.
## (5) steganography
A stego image only supports data storage operations because if the user wants to update the data hidden in a stego image, the user, firstly, retrieves that data from the stego image, and the user can perform any modification on this data.
## (6) signcryption
Only the data storage operation is supported on the signcrypted data, because if the user wants to update signcrypted data that are stored on the cloud database, firstly, the user needs to unsign this data, and then the user can perform any update on this data. In this SLR, we have presented all the data privacy techniques related to data storage on cloud computing systematically, and we also present a comparison among all the protection techniques concerning the five finalities, which are the (i) local proxy overhead, (ii) data accuracy retains, (iii) level of data protection, (iv) transparency, and (v) operation supported.
## Conclusion and future work
ere are some research gaps we found in all these techniques of data splitting, anonymization, steganography, encryption, homomorphic encryption, and signcryption.
(i) ere is a very strong need to develop some ad hoc protocols for the communication of data splitting fragments that are stored on different CSPs, and also, there is a strong need to develop some protocol for the communication between different CSPs. Noncryptographic techniques are faster on different CSPs but do not provide enough security. Hence, we can improve security by developing some methods for data-splitting techniques. (ii) Anonymity techniques work very effectively on a small amount of data but not for big data. Hence, there is a search gap in which we can develop some anonymity techniques to achieve more efficient performance. erefore, some anonymous schemes need to be developed, which provide stronger protection to the quasi-identifier. Current anonymity techniques are very immature. (iii) One of the limitations of steganography is that one can only use it to defend against a third party who does not know steganography. If the third party knows steganography, it can extract the data in the same way that the recipient extracts it. erefore, we always use encryption with steganography. erefore, there is a need to develop such steganography techniques that can protect sensitive data from third parties. (iv) ere is a need to develop some cryptographic techniques that can take less time than the existing cryptographic techniques to perform search and computation operation on encrypted data. Cryptographic techniques provide high security but low computational utility. erefore, it is a search gap to develop some techniques that provide both high security with more efficiency. (v) e complexity of homomorphic encryption and decryption is far greater than that of normal encryption and decryption, and it is not applicable to many applications, such as healthcare and time-sensitive applications. erefore, there is an urgent need to develop such homomorphic encryption schemes that have low complexity and computation cost. (vi) Signcryption is used to verify and authenticate users. We can obtain confidentiality and authentication using signcryption, however, the main limitation of signcryption is that the calculation costs of the encryption algorithm used in signcryption are very high. erefore, there is a need to develop such signcryption schemes that use such encryption algorithms, which have low computation cost.
## Data availability
e data used to support the findings of this study are provided in this article.
## Conflicts of interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.
[fig] Figure 1: Review procedure. [/fig]
[fig] Figure 2: Study selection procedure. [/fig]
[fig] Figure 3: Percentage of selected studies. [/fig]
[fig] Figure 6: Data anonymization flow diagram. [/fig]
[fig] Figure 5: Data [/fig]
[fig] Figure 7: Data-splitting flow diagram. [/fig]
[fig] Figure 8: Steganography [/fig]
[fig] Figure 9: Encryption flow diagram. [/fig]
[fig] 4. 2 1: Publication Trend. From 2010 to 2021, we found 52 papers that were of top ranked journals and conferences. From 2010 to 2017, there is linear work in cloud computing, however, after 2017, a lot of work is done in cloud computing data security. From 2018 to 2021, 37 papers are published. After 2018, the trend about data security in cloud computing increased very vastly. Most of the work is done in 2021. High-ranked studies are published in 2021. [/fig]
[table] Table 1: Databases sources. [/table]
[table] Table 3: Results from electronic databases. [/table]
[table] Table 2: Databases search query. [/table]
[table] Table 4: Inclusion and exclusion criteria. [/table]
[table] Table 5: Quality checking criteria. [/table]
[table] Table 6: Identifiers classification. [/table]
[table] Table 7: e summary of data anonymization techniques. [/table]
[table] Table 8: e summary of the data-splitting techniques. [/table]
[table] Table 9: e summary of the steganographic techniques. [/table]
|
Pharmacokinetics and Bioequivalence of Two Formulations of Febuxostat 40-Mg and 80-Mg Tablets: A Randomized, Open-Label, 4-Way Crossover Study in Healthy Chinese Male Volunteers
The present study aimed to investigate the pharmacokinetic properties of febuxostat in healthy Chinese male volunteers and evaluate whether the two formulations of febuxostat 40-mg and 80-mg tablets are bioequivalent. A randomized, open-label, 4-way crossover study was conducted in healthy Chinese male volunteers under fasting conditions. 24 eligible subjects were randomized in a 1:1:1:1 ratio to receive a single dose of test or reference formulation of febuxostat 40-mg or 80-mg tablet. The washout period between each administration was 1 week. Plasma febuxostat was quantified by a validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) method. Tolerability was evaluated by monitoring adverse events, physical examinations, 12-lead ECG and laboratory tests. After single-dosing of 1 tablet of 40-mg febuxostat, the pharmacokinetic parameters of test and reference formulations were: T max 1.22±0.87 and 1.85±1.03 h, C max 1689.16 ±461.31 and 1613.80±608.43 ngÁmL -1 , AUC 0-t 5139.87±1349.28 and 5517.91±2024.26 ngÁmL -1 Áh, AUC 0−1 5263.06±1339.16 and 5640.48±2040.22 ngÁmL -1 Áh, t 1/2 4.82±2.61 and 4.85±1.78 h, respectively. After single-dosing of 1 tablet of 80-mg febuxostat, the pharmacokinetic parameters of test and reference formulations were: T max 1.71±1.21 and 2.23±1.55 h, C max 2744.47±1157.44 and 2998.17±1200.13 ngÁmL -1 , AUC 0-t 9634.03 ±2768.25 and 10467.95±3501.65 ngÁmL -1 Áh, AUC 0−1 9834.32±2730.51 and 10626.63 ±3504.08 ngÁmL -1 Áh, t 1/2 6.25±2.44 and 5.46±1.65 h, respectively. For single-dosing of 1 tablet of 40-mg febuxostat, 90% CIs for the test/reference ratio of AUC 0-t , AUC 0−1 and C max were 89.79 to 102.55, 90.14 to 102.56 and 93.99 to 129.63, respectively. For singledosing of 1 tablet of 80-mg febuxostat, 90% CIs for the test/reference ratio of AUC 0-t , AUC 0−1 and C max were 86.67 to 100.00, 87.50 to 100.51 and 79.48 to 105.99, respectively. This single dose study revealed similar pharmacokinetic properties in healthy Chinese male volunteers as those found in Caucasic population. The test and reference PLOS ONE | febuxostat tablets formulations met the regulatory criteria for bioequivalence at 40-mg and 80-mg strengths in fasting healthy Chinese male volunteers.Trial Registration: Chictr.org ChiCTR-TTRCC-14004288 NEB = nebulizer gas, CUR = curtain gas, IS = ion spray voltage, TEM = source temperature, DP = declustering potential, FP = focus potential, EP = entrance potential, CE = collision energy, CXP = collision cell exit potential.
# Introduction
Gout is the most common inflammatory arthritis which affects nearly 2% of the total population.The prevalence of gout has been increasing in many countries including the United States and China over the past decades. [bib_ref] Increasing prevalence of gout and hyperuricemia over 10 years among older adults..., Wallace [/bib_ref] [bib_ref] The prevalence of hyperuricemia in China: a meta-analysis, Hn [/bib_ref] Febuxostat is a novel xanthine oxidase (XO) inhibitor indicated for the chronic management of hyperuricemia in patients with gout. Many previous studies have indicated the efficacy and safety of febuxostat in lowering serum uric acid levels in patients with hyperuricemia and gout. [bib_ref] Febuxostat: a selective xanthine-oxidase/xanthine-dehydrogenase inhibitor for the management of hyperuricemia in adults..., Ernst [/bib_ref] [bib_ref] Febuxostat: a review of its use in the treatment of hyperuricaemia in..., Frampton [/bib_ref] [bib_ref] Efficacy and tolerability of febuxostat in hyperuricemic patients with or without gout:..., Ye [/bib_ref] The pharmacokinetic characteristics of febuxostat in human have been studied previously, including data available in Chinese population. [bib_ref] Metabolism and excretion of [14C] febuxostat, a novel nonpurine selective inhibitor of..., Grabowski [/bib_ref] [bib_ref] The effect of mild and moderate hepatic impairment on pharmacokinetics, pharmacodynamics, and..., Khosravan [/bib_ref] [bib_ref] Pharmacokinetics, pharmacodynamics and safety of febuxostat, a non-purine selective inhibitor of xanthine..., Khosravan [/bib_ref] [bib_ref] Effect of food or antacid on pharmacokinetics and pharmacodynamics of febuxostat in..., Khosravan [/bib_ref] [bib_ref] The effect of age and gender on pharmacokinetics, pharmacodynamics, and safety of..., Khosravan [/bib_ref] A previously published study compared the pharmacokinetics of febuxostat between Chinese and other races and indicated that some parameters in Chinese population appeared not to accord with the published literature in Caucasic population. [bib_ref] Pharmacokinetics of febuxostat in healthy Chinese volunteers, Liu [/bib_ref] Febuxostat was first developed in Japan and has been approved in many countries including the USA and the EU.Recently, febuxostat was approved in China and recommended at 40 mg or 80 mg once daily. The dosage forms and strengths of the branded formulation in China are tablet 40-mg and 80-mg. Before allowing the marketing of a generic formulation, bioequivalence evaluation of the generic and branded formulations is required by China Food and Drug Administration (CFDA). The present study aimed to investigate the pharmacokinetic properties of febuxostat in healthy Chinese male volunteers and evaluate whether the two formulations of febuxostat 40-mg and 80-mg tablets are bioequivalent.
This was a registered study approved by China Food and Drug Administration. It was also registered in the World Health Organization International Clinical Trials Registry Platform (Chictr.org identifier: ChiCTR-TTRCC-14004288).
# Subjects and methods
## Study design
This was a randomized, open-label, 4-way crossover study planning to enroll 24 healthy Chinese male subjects. Based on a computer-generated table of random numbers, 24 subjects were allocated in a 1:1:1:1 ratio to receive a single dose of test or reference formulation of febuxostat 40-mg or 80-mg tablet under fasting conditions. The washout period between each administration was 1 week. The study flowchart is shown in [fig_ref] Fig 1: Flowchart of the study [/fig_ref] The study protocol was approved by the Independent Ethics Committee of West China Hospital, Sichuan University (Chengdu, China). The protocol for this trial and supporting CONSORT checklist are available as supporting information (see S1, S2 and S3 Files). The approval letter of ethic committee is available as supporting information S4 and S5 Files.
## Subjects
Healthy, nonsmoking Chinese male volunteers were recruited according to the following inclusion criteria: aged from 18 to 40, body mass index between 19-24 kg/m 2 , healthy status confirmed by physical examination, medical history, 12-lead ECG and laboratory tests (hematology, blood biochemistry, urinalysis, tests for alcohol and other drugs abuse). Exclusion criteria included any allergic history or history of cardiac, pulmonary, hepatic, renal, hematologic or gastrointestinal disease or any other acute or chronic disease. Written informed consent was obtained from each subject before screening procedures.
## Test and reference formulations
## Drug administration and sampling
The study drug administration and blood sampling were conducted in the Phase I Unit of West China Hospital, Sichuan University. In each period, the subjects were given a single dose of test or reference formulation of febuxostat 40-mg or 80-mg tablet after an overnight fast (12 hours). The febuxostat tablets were administered with 200 mL water. Additional water intake was permitted 2 hours after dosing. Standard meals (Food energy~900 kcal; 30% protein, 60% carbohydrate, 10% fat) were offered 4 and 10 hours after dosing. Blood samples (~3.5 mL) were
## Assays of febuxostat
Plasma febuxostat was quantified by a liquid chromatography-tandem mass spectrometry (LC-MS/MS) method developed and validated before the clinical study. Chromatography was performed using a Shimadzu SIL-HTC system (Shimadzu, Kyoto, Japan) equipped with a Ultimate C18 analysis column (50×4.6 mm, 5 μm). Mass spectrometric detection employed an API 3000 mass spectrometer with the working station Analyst 1.4.2 (AB Sciex, Ont., USA). The mobile phase consisted of acetonitrile-10mM ammonium acetate in water and formic acid (70:30:0.05, v/v/v) was delivered at a flow rate of 0.35 mL/min. Each 100 μL plasma sample was spiked with 100 μL internal standard (IS) solution. Then 100 μL hydrochloric acid (1molÁL -1 ) was added to acidize the spiked sample before the following extracting process. After extracted by 3.5 mL dichloromethane and then centrifuged, the supernatant was evaporated in a 45°C water bath. The residue was dissolved in 100 μL mobile phase and injected 10 μL onto the column. Bezafibrate was used as internal standard (IS). Negative multiple reaction monitoring (MRM) model was used and transitions were at m/z 315.1!271.1 and 361.9!275.6 for febuxostat and IS, respectively. The detailed MS parameters are shown in [fig_ref] Table 1: The mass spectrometry [/fig_ref]. Typical chromatograms are shown in [fig_ref] Fig 2: LC-MS/MS of febuxostat of blank plasma solution [/fig_ref] The retention time for febuxostat and IS were 3.2 min and 2.3 min, respectively. The calibration curve was linear over the range of 10~4000 ngÁmL -1 . The lowest concentration of detection of febuxostat in plasma was 10 ngÁmL -1 . The method recovery was 98.6%~100.4%. The intra-day RSD were less than 3% and inter-day RSD were less than 5%. The results of all stability studies were qualified for requirements.
## Pharmacokinetics and bioequivalence analysis
WinNonlin Version 6.1 (Pharsight Corporation, Mountain View, California) was used to calculate the pharmacokinetic parameters of febuxostat with noncompartmental analysis method. C max and T max were obtained directly from the concentration-time data. Other pharmacokinetic parameters were calculated as previously reported [bib_ref] Pharmacokinetic Properties of Single-and Multiple-Dose Pitavastatin Calcium Tablets in Healthy Chinese Volunteers, Luo [/bib_ref] : AUC 0−t was calculated with the linear trapezoidal rule. AUC 0−1 was obtained as the sum of AUC 0−t and Ct/λ. Ct was the last measured concentration and λ was the slope of linear regression of the log-transformed concentration-time curve. t 1/2 was calculated as 0.693/λ. The relative bioavailability of the test formulation was calculated as AUC 0−t(test) / AUC 0-t(reference) × 100%. 90% CIs for the test/reference ratio of log-transformed C max and AUC were assessed by analysis of variance (ANOVA) using WinNonlin Version 6.1. Other pharmacokinetic parameters were analyzed using SPSS Version 18.0 (SPSS Inc. Chicago, IL, USA). T max was tested by Wilcoxon signed rank test for significant differences. The two formulations were considered to be bioequivalent if the 90% CI for AUC was located within
## Tolerability assessment
Tolerability was evaluated by monitoring adverse events, physical examinations, laboratory tests (hematology, blood biochemistry and urinalysis) and 12-lead ECG. All the laboratory tests were performed in the laboratory of West China Hospital, Sichuan University and the laboratory was authenticated by College of American Pathologists (CAP).
# Results
## Subjects
Totally 24 male subjects were enrolled in 2 weeks and completed the study. The demographics of the subjects (mean ± SD, standard deviation) were: age 22.8 ± 1.4 (range, 21.0~26.0) years, At both 40-mg and 80-mg strengths, the 90% CIs of AUC 0-t and AUC 0−1 were located within 80% to 125%, 90% CI for C max was within 70% to 143%. The two formulations met the regulatory criteria for bioequivalence and were bioequivalent at 40-mg and 80-mg strengths.
## Pharmacokinetics of febuxostat
## Bioequivalence evaluation
## Tolerability
Among the 24 subjects, only one reported abnormal laboratory tests (elevated transaminase) after administration, which was rated as mild and considered possibly associated with the study drug. No other Adverse Events were observed or reported. Single-dosing of febuxostat tablet at 40-mg and 80-mg strengths was considered to be well tolerated by healthy Chinese male volunteers.
# Discussion
Febuxostat is a novel xanthine oxidase (XO) inhibitor approved for the chronic management of hyperuricemia in patients with gout. The pharmacokinetic characteristics of febuxostat in human have been studied previously. After oral administration, febuxostat is well absorbed with the bioavailability about 84%. The time to reach the maximum plasma concentration is 1 to 1.5 hours and half-life is 5 to 8 hours. After a single of 40 mg and 80 mg of febuxostat, C max is approximately 1.6 ± 0.6 μg/mL and 2.6 ± 1.7 μg/mL, respectively; AUC is approximately 4.0 ± 1.9 μgÁh/mL and 9.1 ± 3.8 μgÁh/mL, respectively. [bib_ref] Pharmacokinetics, pharmacodynamics and safety of febuxostat, a non-purine selective inhibitor of xanthine..., Khosravan [/bib_ref] [bib_ref] Effect of food or antacid on pharmacokinetics and pharmacodynamics of febuxostat in..., Khosravan [/bib_ref] [bib_ref] The effect of age and gender on pharmacokinetics, pharmacodynamics, and safety of..., Khosravan [/bib_ref] Pharmacokinetic data of febuxostat is also available in Chinese population. A previous study reported the pharmacokinetics of febuxostat in healthy Chinese volunteers and concluded the pharmacokinetics of febuxostat in Chinese appeared not in accord with the published literature in other races. [bib_ref] Pharmacokinetics of febuxostat in healthy Chinese volunteers, Liu [/bib_ref] It was reported that C max increased 27-58%, AUC increased 70-117%, half-life increased 20-57%, and CL/F decreased 44-57%, respectively, at 40-80 mg dose level for Chinese subjects. The author deduced that body weight and genetic variability should be taken into account to explain the observed ethnic difference.
It was reported that age or gender had no clinically significant effect on the pharmacokinetics of febuxostat. [bib_ref] The effect of age and gender on pharmacokinetics, pharmacodynamics, and safety of..., Khosravan [/bib_ref] Co-administration with food reduced the rate and extent of febuxostat absorption, however, this food effect on absorption was not associated with a clinically significant pharmacodynamic change. [bib_ref] Effect of food or antacid on pharmacokinetics and pharmacodynamics of febuxostat in..., Khosravan [/bib_ref] Furthermore, febuxostat pharmacokinetic parameters for patients with hyperuricemia and gout estimated by population pharmacokinetic analyses were similar to those estimated in healthy subjects.Considering the pharmacokinetic properties of febuxostat and the dosage forms and strengths of the branded formulation, we designed this randomized, open-label, 4-way crossover bioequivalence study in healthy male volunteers under fasting conditions.Our study further investigated the pharmacokinetic properties of febuxostat in healthy Chinese male volunteers and evaluated the bioequivalence of two formulations of febuxostat 40-mg and 80-mg tablets. Our study indicated there were no significant differences between the pharmacokinetic parameters of the branded and generic formulations. The two formulations met the regulatory criteria for bioequivalence proposed by China Food and Drug Administration.
In the present study, we used a liquid chromatography-tandem mass spectrometry (LC-MS/ MS) method to quantify plasma febuxostat. The method validation results indicated that the specificity, precision, accuracy, stability and recovery of the present method were comparable to those previously reported methods for the determination of febuxostat, suggesting our method was suitable for the determination of febuxostat in human plasma. [bib_ref] A sensitive LC-MS/MS method for the quantification of febuxostat in human plasma..., Vaka [/bib_ref] [bib_ref] Development and validation of a liquid chromatographytandem mass spectrometry method for the..., Wang [/bib_ref] In the present study, after a single dose of 40 mg and 80 mg of febuxostat tablets, the pharmacokinetic parameters accorded with those reported in Caucasic population. [bib_ref] Pharmacokinetics, pharmacodynamics and safety of febuxostat, a non-purine selective inhibitor of xanthine..., Khosravan [/bib_ref] [bib_ref] Effect of food or antacid on pharmacokinetics and pharmacodynamics of febuxostat in..., Khosravan [/bib_ref] [bib_ref] The effect of age and gender on pharmacokinetics, pharmacodynamics, and safety of..., Khosravan [/bib_ref] Compared with the subjects in the above mentioned study, the subjects in the present study were of more mean body weight and body mass index, which were more similar to the subjects of Caucasic population. [bib_ref] Pharmacokinetics, pharmacodynamics and safety of febuxostat, a non-purine selective inhibitor of xanthine..., Khosravan [/bib_ref] [bib_ref] Effect of food or antacid on pharmacokinetics and pharmacodynamics of febuxostat in..., Khosravan [/bib_ref] [bib_ref] The effect of age and gender on pharmacokinetics, pharmacodynamics, and safety of..., Khosravan [/bib_ref] This might further confirm that the observed ethnic difference between Chinese and other races were caused by the subjects' body weight. Another aspect to explain the metabolic difference is the genetic variability between subjects, however, we did not include polymorphism analysis in the present study, which is a limitation of this study. Another limitation of this study is that it only enrolled Chinese male volunteers and the data could not be extrapolated in female subjects or other races, which caused the lack of generalisability of the present study.
# Conclusions
The pharmacokinetic parameters in the present study accorded with those reported in Caucasic population and the test and reference febuxostat tablets formulations met the regulatory criteria for bioequivalence at 40-mg and 80-mg strengths in fasting healthy Chinese male volunteers.
## Supporting information
[fig] Fig 1: Flowchart of the study. doi:10.1371/journal.pone.0150661.g001collected before and at 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 4, 6, 8, 10, 12, 16, 24, 36, 48 hours after dosing. All the drug administration and blood sampling processes were under continuous medical supervision. [/fig]
[fig] Fig 2: LC-MS/MS of febuxostat of blank plasma solution(A), reference standards solution (B), blank plasma with reference standards solution(C) and subject plasma solution (D). doi:10.1371/journal.pone.0150661.g002 Pharmacokinetics and Bioequivalence of Febuxostat in Healthy Chinese PLOS ONE | DOI:10.1371/journal.pone.0150661 March 14, 2016 weight 65.3 ± 7.5 (range, 50.0~76.0) kg, height 173.0 ± 5.0 (range, 160.0~179.0) cm, body mass index 21.8 ± 1.5 (range, 19.4~23.9) kg/m 2 . [/fig]
[fig] 2: shows the pharmacokinetic parameters of febuxostat in 24 healthy Chinese male volunteers after single-dosing of 1 tablet of 40-mg or 80-mg febuxostat. The mean plasma concentration-time profiles of the two formulations after single-dosing of 1 tablet of 40-mg and 80-mg are shown in Fig 3. [/fig]
[fig] Fig 3: Mean plasma concentration-time curves of test and reference formulations of febuxostat after single-dosing of 1 tablet of 40-mg (A) and 80-mg (B). doi:10.1371/journal.pone.0150661.g003 [/fig]
[fig] S1: File. CONSORT Checklist. (DOC) S2 File. English translation of study protocol abstract. (DOC) S3 File. Study protocol of febuxostat-Chinese version. (DOC) S4 File. English translation of EC approval letter. (DOC) S5 File. Approval letter of ethic committee. (PDF) [/fig]
[table] Table 1: The mass spectrometry (MS) parameters for assays of febuxostat.80% to 125% and C max within 70% to 143%, according to China Food and Drug Administration proposal.[18] The study's sample size was estimated with Software PASS 11.0 (NCSS Statistical Software, Kaysville, Utah) based on the China Food and Drug Administration proposed bioequivalence range of 80% to 125% for AUC and 70% to 143% for Cmax, a power of 80% at an α of 0.05 to show bioequivalence. The test/reference mean ratio was estimated to be 0.95 to 1.05 and the within-subject CV was 25%. [/table]
[table] Table 3: Bioequivalence evaluation of two formulations of febuxostat tablets after single-dosing in in healthy Chinese male volunteers. doi:10.1371/journal.pone.0150661.t003 [/table]
|
The modulation effect of non-invasive brain stimulation on cognitive function in patients with mild cognitive impairment: a systematic review and meta-analysis of randomized controlled trials
Background: To prevent and control dementia, many scholars have focused on the transition stage between normal ageing and dementia, mild cognitive impairment (MCI) which is a key interventional target for dementia. Studies have shown that non-invasive brain stimulation (NIBS) is beneficial to improve cognitive function of MCI patients. However, whether NIBS is conducive to the protection of cognitive ability in MCI patients remains unknown due to limited evidence. The aim of the study was to systematically evaluate the modulation effect of NIBS on cognitive function (global cognitive ability and specific domains of cognition) in patients with MCI.Results: A total of 11 RCTs comprising a total of 367 MCI participants. Meta-analysis showed that NIBS can significantly improve global cognition (n = 271, SMD = 0.94, 95% CI 0.47-1.41, p < 0.0001) and verbal fluency (n = 72, MD = 2.03, 95% CI 0.17-3.88, p = 0.03). However, there was no significant improvement in other domains of cognition.Conclusions: NIBS has a positive effect on improving global cognitive function and verbal fluency. At the same time, it has a small positive effect on improving executive function. However, these findings should be interpreted carefully due to the limitations of the study.
# Background
With age increasing, the risk of Alzheimer's disease (AD) is on the rise [bib_ref] Global prevalence of dementia: a Delphi consensus study, Ferri [/bib_ref] , however, the treatment of dementia is far from satisfactory [bib_ref] Alzheimer's disease facts and figures, Association [/bib_ref] [bib_ref] Pharmacological treatment of mild cognitive impairment as a prodromal syndrome of Alzheimer..., Karakaya [/bib_ref]. To prevent and control dementia, many scholars have focused on the transition stage between normal ageing and dementia, mild cognitive impairment (MCI) which is a key interventional target for dementia [bib_ref] The prevalence of mild cognitive impairment and its etiological subtypes in elderly..., Jia [/bib_ref]. The main characteristics of MCI are objective memory impairment and other cognitive deficits; however, aspects of daily living are not significantly affected [bib_ref] Mild cognitive impairment: an overview, Petersen [/bib_ref]. The incidence of MCI in people over 65 years of age is 10-20%, and more than half of them will progress into dementia within 5 years [bib_ref] Understanding memory dysfunction, Budson [/bib_ref] [bib_ref] Mild cognitive impairment, Bartley [/bib_ref]. Severe cognitive decline will have a huge impact on the daily lives of patients such as independent living ability losing, lower quality of life, and huge economic burden [bib_ref] Quality of life in patients with Dementia with Lewy Bodies, Lee [/bib_ref]. Thus, effective and timely interventions that aim to improve cognitive function or delay the process of cognitive decline will significantly benefit patients and their families.
Recently, there has been an increased interesting on the use of non-drug therapy to improve the cognitive function of MCI patients [bib_ref] Efficacy of cognitive rehabilitation therapies for mild cognitive impairment (MCI) in older..., Huckans [/bib_ref]. Studies have shown that as a new type of treatment, non-invasive brain stimulation (NIBS) is beneficial to improve cognitive function of MCI patients. NIBS can alter neuronal activity temporarily and affect behavioural performance [bib_ref] Modelling non-invasive brain stimulation in cognitive neuroscience, Miniussi [/bib_ref].
The two most commonly used techniques of NIBS are transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS). TMS modulates cortical activities by delivering strong magnetic pulses to the cortex through the scalp [bib_ref] Technology insight: noninvasive brain stimulation in neurology-perspectives on the therapeutic potential of..., Fregni [/bib_ref] [bib_ref] Parameterization of transcranial magnetic stimulation, Rubens [/bib_ref]. Different stimulation frequencies can enhance or inhibit cortical excitability in the target cortical region. Unlike TMS, tDCS delivers a continuous week currents (0.5-2.0 mA) to the scalp to modulate neuronal transmembrane potential toward hyperpolarization or depolarization, thereby altering plasticity in the stimulated brain regions [bib_ref] Potentials and limits to enhance cognitive functions in healthy and pathological aging..., Prehn [/bib_ref]. Although they are different in some respects, both tools can induce long-term after effects on cortical excitability and neuroplasticity.
Many studies [bib_ref] tDCS selectively improves working memory in older adults with more education, Berryhill [/bib_ref] [bib_ref] Non-invasive brain stimulation improves object-location learning in the elderly, Floel [/bib_ref] [bib_ref] Non-invasive brain stimulation of the right inferior frontal gyrus may improve attention..., Eliasova [/bib_ref] [bib_ref] A double-blind randomized clinical trial on the efficacy of cortical direct current..., Khedr [/bib_ref] have reported the treatment effect of tDCS and TMS on cognitive outcomes in various populations, such as AD and healthy adults. For example, a study reported that for older adults, compare to the anodal or placebo (sham) tDCS, anodal tDCS not only strengthened episodic memories, but delayed recall is enhanced after 48 h compared with placebo stimulation [bib_ref] Older adults get episodic memory boosting from noninvasive stimulation of prefrontal cortex..., Sandrini [/bib_ref]. Similarly, another study found up-regulation of the dorsolateral prefrontal cortex (DLPFC) led to improvements of everyday memory after 10-Hz TMS in MCI patients.
However, beneficial effects of NIBS are not always observed. A study [bib_ref] Anodal tDCS during face-name associations memory training in Alzheimer's patients, Cotelli [/bib_ref] showed that 2 weeks of tDCS did not show significant group differences in the face-name association task. In addition, Boggio et al. [bib_ref] Modulation of decision-making in a gambling task in older adults with transcranial..., Boggio [/bib_ref] found that tDCS over the prefrontal cortex increases high-risk behaviour in older adults. In particular, a study with a crossover design failed to induce positive or negative behavioral effects following either low-or high-frequency TMS in seven patients with vascular MCI [bib_ref] Neurocognitive effects of repetitive transcranial magnetic stimulation in patients with cerebrovascular disease..., Sedlackova [/bib_ref]. Therefore, the efficacy of tDCS or TMS as a treatment option remains controversial. Specifically, whether tDCS or TMS is conducive to the protection of cognitive ability in MCI patients remains unknown due to limited evidence.
This study was designed to systematically evaluate the effect of tDCS and TMS as an intervention on cognitive function of MCI patients, including global cognitive ability and specific domains of cognition, such as memory, executive function, attention, verbal fluency.
# Methods
## Protocol and registration
The protocol of this study was registered with the International Prospective Register of Systematic Review, PROSPERO, under the identification number CRD42018092620, and can be integrally assessed online (http://www.crd.york.ac.uk/PROSP ERO/displ ay_recor d.php?ID=CRD42 01809 2620).
## Literature search
We searched a total of seven electronic databases, including PubMed, EMBASE (OVID), SinoMed, China National Knowledge Infrastructure (CNKI), Wanfang degree and conference papers database, Chinese Science and Technology Periodical Database (VIP) from its inception to 31 January 2018, as well as the Cochrane Central Register of Controlled Trials (Cochrane Library, 2018, Issue 1). These databases were searched without language restrictions. Relevant keywords related to noninvasive brain stimulation as Medical Subject Heading terms and text words (e.g., 'noninvasive brain stimulation' , or 'transcranial direct current stimulation' , or 'transcranial magnetic stimulation') were used in combination with words related to mild cognitive impairment.
## Inclusion criteria
The trials selected in the study should met the following inclusion criteria: (1) published or unpublished randomized controlled studies; (2) participants diagnosed with MCI based on any diagnostic criteria, such as Petersen criteria, 2004 MCI Key Symposium criteria or other standards and consensus; participants with vascular cognitive impairment or other neurological disorders resulting from AD, dementia or Parkinson's disease were excluded; (3) the intervention of the experimental group was NIBS technique, regardless of the type; (4) the control group received basic intervention, sham stimulation, medication or other interventions; (5) outcomes included global cognitive ability and specific domain of cognition, which was measured by neuropsychological tests or other objective measurements. Studies without available data were excluded.
## Study identification and data extraction
Eliminate duplicate records with the reference management software (Note Express V.2.0). In the literature screening, the title and abstracts were read first to eliminate irrelevant studies. The studies that potentially met the inclusion criteria were independently screened, extracted, and cross-checked by two reviewers. Disagreements were resolved by discussion with a third reviewer. Extract the following information from eligible studies: study design, sample size, participants' characteristics, methodological information of research quality, experimental and control intervention, the duration, frequency, intensity and type of NIBS, outcomes, the time of followup, and adverse events.
## Assessment of risk of bias in individual studies
Two reviewers independently evaluated the scientific quality of the study according to the JADAD scoring manual [bib_ref] Assessing the quality of reports of randomized clinical trials: is blinding necessary?, Jadad [/bib_ref]. Evaluation content included description of random sequence generation, description of the doubleblind procedure, and description of withdrawals and dropouts. The total score was 5 points, where a score of 1-2 points was associated with low quality and a score of 3-5 points was deemed as high quality.
# Data analysis
A meta-analysis of outcomes for each study was performed using Review Manager 5.3 software and a twosided p < 0.05 was considered statistically significant. Data were summarized using relative risk with 95% CI for binary outcomes. The measurement data used the MD or standardized MD and the 95% CI as the effect amount. The meta-analysis used the I 2 test to observe the degree of statistical heterogeneity between studies. When I 2 ≤ 50%, a fixed-effect model was used, while a randomeffects model was used when I 2 > 50%. A parallel sensitivity analysis was used to find the source of heterogeneity. Studies with different interventions were divided into subgroups for subgroup analysis according to different factors, such as design options, treatment duration, and specific interventions. For studies with significant clinical and methodological heterogeneity, the outcome metaanalysis was not performed, and only general statistical descriptions were performed. Statistical heterogeneity among the included studies was assessed using a χ 2 test and Higgins I 2 value, with I 2 > 75% suggesting high statistical heterogeneity [bib_ref] Measuring inconsistency in meta-analyses, Higgins [/bib_ref].
# Results
## Study identification
According to the original search strategy, 510 studies were retrieved, and 72 repetitive studies were excluded. After reading titles and abstracts, two independent reviewers excluded 416 articles that did not meet the inclusion criteria. A total of 22 studies were further evaluated, excluding 1 repeated publication, while 2 did not meet the MCI diagnostic criteria, 4 had unacceptable cognitive outcomes, and 4 were non-RCT studies. A total of 11 studies [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] Changes in cerebral glucose metabolism after 3 weeks of noninvasive electrical stimulation..., Yun [/bib_ref] [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] [bib_ref] Effects of repetitive transcranial magnetic stimulation on cognitive function in patients with..., Zhang [/bib_ref] [bib_ref] Preliminary study of high-frequency repetitive transcranial magnetic stimulation in treatment of mild..., Linlin [/bib_ref] [bib_ref] Effect of repetitive transcranial magnetic stimulation combined with cognitive training on mild..., Rui [/bib_ref] [bib_ref] Impacts of repetitive transcranial magnetic stimulation on the brain network regulation and..., Shiyu [/bib_ref] [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] [bib_ref] Repetitive transcranial magnetic stimulation for apathy in mild cognitive impairment: a doubleblind,..., Padala [/bib_ref] were included in the metaanalysis finally.
The detailed screening flow used to find eligible studies is presented in [fig_ref] Figure 1: Flow diagram for searching and selection of the included studies [/fig_ref].
## Characteristics of included studies
This review included 11 RCT studies involving 367 MCI participants (175 males and 192 females, average age 66.52 years) and [fig_ref] Table 1: Characteristics of included studies in this systematic reviewBNT, Boston Naming Test [/fig_ref] shows the characteristics of each study. Six studies [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] [bib_ref] Effects of repetitive transcranial magnetic stimulation on cognitive function in patients with..., Zhang [/bib_ref] [bib_ref] Preliminary study of high-frequency repetitive transcranial magnetic stimulation in treatment of mild..., Linlin [/bib_ref] [bib_ref] Effect of repetitive transcranial magnetic stimulation combined with cognitive training on mild..., Rui [/bib_ref] [bib_ref] Impacts of repetitive transcranial magnetic stimulation on the brain network regulation and..., Shiyu [/bib_ref] [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref] were conducted in China, two [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] in Italy, one [bib_ref] Changes in cerebral glucose metabolism after 3 weeks of noninvasive electrical stimulation..., Yun [/bib_ref] in Brazil, one [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] in South Korea and one [bib_ref] Repetitive transcranial magnetic stimulation for apathy in mild cognitive impairment: a doubleblind,..., Padala [/bib_ref] in the US. All studies reported clear diagnostic, inclusion, and exclusion criteria.
The types of NIBS included tDCS [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] Changes in cerebral glucose metabolism after 3 weeks of noninvasive electrical stimulation..., Yun [/bib_ref] and TMS [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] [bib_ref] Effects of repetitive transcranial magnetic stimulation on cognitive function in patients with..., Zhang [/bib_ref] [bib_ref] Preliminary study of high-frequency repetitive transcranial magnetic stimulation in treatment of mild..., Linlin [/bib_ref] [bib_ref] Effect of repetitive transcranial magnetic stimulation combined with cognitive training on mild..., Rui [/bib_ref] [bib_ref] Impacts of repetitive transcranial magnetic stimulation on the brain network regulation and..., Shiyu [/bib_ref] [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] [bib_ref] Repetitive transcranial magnetic stimulation for apathy in mild cognitive impairment: a doubleblind,..., Padala [/bib_ref]. For tDCS, the frequency varied from three to five sessions weekly and 25-30 min per session. The duration of the intervention lasted 2-3 weeks. The stimulation site was in the dorsolateral prefrontal cortex (DLPFC), and the intensity was 2 mA. For TMS, the frequency varied from five to six sessions weekly and 15-45 min per session. The duration of the intervention lasted 2-16 weeks. The stimulation site was in the DLPFC, bilateral frontal area, and precuneus, with an intensity from 80 to 120% resting motor threshold (RMT). Of these 11 studies, 3 were followed up from 1 to 3 months [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref]. Two studies compared TMS with drug therapy [bib_ref] Effects of repetitive transcranial magnetic stimulation on cognitive function in patients with..., Zhang [/bib_ref] [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref]. Two studies combined TMS/tDCS with physical therapy (PT) or cognitive training to compare the effects of combination therapy with PT or cognitive training alone [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] Effect of repetitive transcranial magnetic stimulation combined with cognitive training on mild..., Rui [/bib_ref]. Other studies performed comparisons between true and shame-stimuli. There was a wide variety of cognitive measurement tools used in these studies, including Mini
## Risk of bias of included studies
The risks of bias for all studies are shown in [fig_ref] Table 2: Risk of bias summary [/fig_ref]. In these studies, 3 studies described the use of a random number table and Matlab software to generate random sequences [bib_ref] Preliminary study of high-frequency repetitive transcranial magnetic stimulation in treatment of mild..., Linlin [/bib_ref] [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref]. One study used the covariate adaptive randomization method for random assignment [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref]. The rest of the studies mentioned random grouping but did not describe the generation method of random sequences, and there was a potential risk of high selection bias. Five studies used a double-blind design for blindness to subjects and assessors [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] Preliminary study of high-frequency repetitive transcranial magnetic stimulation in treatment of mild..., Linlin [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] [bib_ref] Repetitive transcranial magnetic stimulation for apathy in mild cognitive impairment: a doubleblind,..., Padala [/bib_ref] ; therefore, their risk of detection bias was judged as low. A single-blind design was used in one other study [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref]. However, none of the other studies mentioned blindness, and thus have a potentially high risk of implementation bias and high detection bias. Four studies reported shedding and described the data processing method [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] Preliminary study of high-frequency repetitive transcranial magnetic stimulation in treatment of mild..., Linlin [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] [bib_ref] Repetitive transcranial magnetic stimulation for apathy in mild cognitive impairment: a doubleblind,..., Padala [/bib_ref] ; therefore, the risk of loss bias was deemed as low. Overall, most of the included studies were judged to have a high risk of bias with low methodological quality.
## Effect of interventions global cognitive function
Nine studies [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] [bib_ref] Effects of repetitive transcranial magnetic stimulation on cognitive function in patients with..., Zhang [/bib_ref] [bib_ref] Preliminary study of high-frequency repetitive transcranial magnetic stimulation in treatment of mild..., Linlin [/bib_ref] [bib_ref] Effect of repetitive transcranial magnetic stimulation combined with cognitive training on mild..., Rui [/bib_ref] [bib_ref] Impacts of repetitive transcranial magnetic stimulation on the brain network regulation and..., Shiyu [/bib_ref] [bib_ref] Effect of high-frequency repetitive transcranial magnetic stimulation on elderly patients with depression..., Wu [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] reported the effects of NIBS on global cognitive ability in participants with MCI by using MMSE, MoCA and Parkinson's Disease Cognitive Rating Scale (PD-CRS). The results showed that NIBS had a significant effect on improving global cognitive ability among participants with MCI, as demonstrated by significantly increased MMSE scores and MoCA scores (n = 301, SMD = 1.82, 95% CI 0.86-2.78, p < 0.0002, I 2 = 91%, the random-effect model; . One study [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] reported that tDCS participants had higher PD-CRS scores than those in the control group, the difference was significant (p = 0.041).
## Memory
Five studies [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] Changes in cerebral glucose metabolism after 3 weeks of noninvasive electrical stimulation..., Yun [/bib_ref] [bib_ref] Impacts of repetitive transcranial magnetic stimulation on the brain network regulation and..., Shiyu [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] involving 109 participants reported the effects of NIBS on memory ability by using the Rivermead Behavioural Memory Test (RBMT), Rey Auditory-Verbal Learning Test (RAVLT), Digit Symbol Substitution Test (DSST), Multifactorial Memory Questionnaire (MMQ), and Clinical memory scale (CMS). Three of the studies compared the effects of TMS stimulation and sham TMS stimulation. The other two articles compare the difference between anode tDCS stimulation and PT/sham stimulation. The results of meta-analysis showed no significant difference between the NIBS group and control groups (n = 109, SMD = 0.20, 95% CI − 0.18 to 0.58, p = 0.31, I 2 = 0%, the random-effect model; .
## Executive function
The effects of NIBS on executive function were evaluated in five studies [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] [bib_ref] Repetitive transcranial magnetic stimulation for apathy in mild cognitive impairment: a doubleblind,..., Padala [/bib_ref] using the Wisconsin card sorting test (WCST), Semantic fluency and Frontal Assessment Battery (FAB), and TMT part B (ms). Due to the use of these different tools, we performed a subgroup analysis. One study showed that TMS significantly improved executive ability by increasing WCST scores (SMD = 1.15, 95% CI 0.45-1.84, p = 0.001), while another reported no significant changes [fig_ref] Figure 4: NIBS versus other intervention [/fig_ref].
## Attention
Three studies [bib_ref] Mild cognitive impairment in Parkinson's disease is improved by transcranial direct current..., Manenti [/bib_ref] [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] [bib_ref] Transcranial magnetic stimulation of the precuneus enhances memory and neural activity in..., Koch [/bib_ref] reported the effects of NIBS on attention ability by using TMT part A (ms) and Digit Symbol Substitution Test (DSST). The results of metaanalysis showed no significant difference between the NIBS group and control groups on reaction time of TMT part A (n = 58, SMD = − 0.31, 95% CI − 1.01 to 0.38, p = 0.38, I 2 = 40%, the random-effect model). Regarding [fig_ref] Figure 5: NIBS versus other intervention [/fig_ref].
## Verbal fluency
Two studies [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] reported the effects of NIBS on verbal fluency ability. The results of meta-analysis showed a significant difference between NIBS and control group (n = 72, MD = 2.03, 95% CI 0.17-3.88, p = 0.03, I 2 = 7%, the fixed-effect model; .
## Adverse effects
Adverse effects were reported in 4 studies [bib_ref] The case-control study of the effect of repetitive transcranial magnetic stimulation on..., Han [/bib_ref] [bib_ref] Effect of repetitive transcranial magnetic stimulation combined with cognitive training on mild..., Rui [/bib_ref] [bib_ref] Impacts of repetitive transcranial magnetic stimulation on the brain network regulation and..., Shiyu [/bib_ref] [bib_ref] Repetitive transcranial magnetic stimulation for apathy in mild cognitive impairment: a doubleblind,..., Padala [/bib_ref] , where the main symptoms were dizziness, pain, and facial twitching. The results of meta-analysis showed no significant difference between the NIBS group and control groups on adverse event rate (RD = 0.25, 95% CI − 0.03 to 0.53, p = 0.08, I 2 = 89%, the random-effect model; [fig_ref] Figure 7: NIBS versus other intervention [/fig_ref]. Notably, two participants experienced severe pain during TMS stimulation, while one participant quit treatment.
# Discussion
Epidemiological studies show that 10-15% patients with MCI will transform to dementia [bib_ref] Memory impairment on free and cued selective reminding predicts dementia, Grober [/bib_ref]. It is very important to provide timely interventions for patients with cognitive impairment in this reversible phase. NIBS has become a potentially useful tool. In particular, NIBS can directly affect the memory mechanisms of young, elderly, and neurotic dysfunction patients, including working memory, episodic memory, and contact memory [bib_ref] Transcranial brain stimulation studies of episodic memory in young adults, elderly adults..., Manenti [/bib_ref] [bib_ref] Transcranial magnetic stimulation and transcranial direct current stimulation: treatments for cognitive and..., Elder [/bib_ref]. However, there is no consistent conclusion regarding whether NIBS can improve the cognitive function of MCI patients. Eleven studies involving 367 subjects were included in this review. All studies were designed to compare tDCS or TMS with a lack of specific stimuli. The results of the meta-analysis showed that NIBS was beneficial With functional magnetic resonance imaging (fMRI), multiple resting state networks such as default mode network, attention network and sensorimotor network have been affected in patients with MCI [bib_ref] Loss of intranetwork and internetwork resting state functional connections with Alzheimer's disease..., Brier [/bib_ref]. Meinzer et al. conducted a double-blind, cross-control study in which brain changes were recorded using task-related and resting fMRI during tDCS stimulation. fMRI data suggest that the low accuracy of semantic flow tests in MCI patients may be related to hyperactivity of bilateral prefrontal area. Anodic-tDCS significantly improved the accuracy of semantic fluency tests in MCI patients, reduced task-related prefrontal hyperactivity and facilitated the normalization of abnormal network structure in resting-state fMRI [bib_ref] Transcranial direct current stimulation in mild cognitive impairment: behavioral effects and neural..., Meinzer [/bib_ref]. Another critical one concerns with modulation of neurotransmitter levels [bib_ref] Determinants of the induction of cortical plasticity by non-invasive brain stimulation in..., Ridding [/bib_ref] [bib_ref] Modulation of human motor cortex excitability by single doses of amantadine, Reis [/bib_ref]. TMS may affect the regulation of cortical neuronal activity by altering the dynamics of excitation/inhibition of neurotransmitter systems, such as GABA and glutamate [bib_ref] Modulation of human motor cortex excitability by single doses of amantadine, Reis [/bib_ref].
Previous studies and reviews have summarized the main advantages of tDCS and TMS [bib_ref] Differential behavioral and physiological effects of anodal transcranial direct current stimulation in..., Heise [/bib_ref] : first, these methods are ideal for exploring brain plasticity throughout the life cycle. Second, they can regulate neurons bidirectionally, not only inhibiting the excitability of neurons but also enhancing the excitability of neurons. Third, these techniques can apply the stimulus only to the position we want without affecting other parts, so that we can treat the disease while avoiding side effects. This is beyond the ability of pharmacology or complementary therapy. In particular, the application of NIBS to dysfunctional neural networks could significantly enhance learning-and memory-related effects. Moreover, repeated use of NIBS for stimulation could make the effect longer lasting.
In terms of safety, there were several mild adverse events in the NIBS group and control group, such as temporary dizziness, pain, and facial twitching. However, these events could be recovered without special treatment. In addition, there was no significant difference in the incidence of adverse events between the two groups. However, it should be noted that there were 2 serious adverse events in the intervention group, of which one participant had severe pain during the treatment, where the pain was relieved after the stimulus was stopped, while the other patient quit treatment due to pain. Although the safety of NIBS has been recognized by most people currently, the individual safety should be monitored, and personal thresholds should be evaluated to determine appropriate parameters for each patient undergoing NIBS treatment [bib_ref] Consensus recommendations for the clinical application of repetitive transcranial magnetic stimulation (rTMS)..., Mcclintock [/bib_ref] [bib_ref] Use of repetitive transcranial magnetic stimulation for treatment in psychiatry, Aleman [/bib_ref] [bib_ref] A systematic review on reporting and assessment of adverse effects associated with..., Brunoni [/bib_ref].
This review only included randomized controlled trials, which implied that the included studies had rigorous research design. Regarding the participants, we more force on the patients with MCI. In order to reduce potential confusion and make the generalization of the findings more pertinent, we excluded participants with secondary cognitive impairment (e.g., vascular dementia) or severe cognitive impairment (e.g., AD). All these methods helped to support the NIBS-related causal hypothesis.
There were several limitations in the systematic review. (1) The stimulation parameters used in each study (including stimulation intensity, stimulation rate, stimulation site, and duration) were quite different, and the optimal parameters for NIBS treatment could not be determined. [bib_ref] Alzheimer's disease facts and figures, Association [/bib_ref] Only three studies were followed up for 1-3 months, suggesting that the effect of NIBS could be maintained for a long time; however, the conclusion should be verified by larger and longer follow-up studies.
(3) The control methods adopted in the study were different. Due to the small number of included studies, the systematic review included the results of all the studies together for analysis but did not analyse them separately according to the methodological differences. For example, there are studies that only used drugs as controls, and thus, the placebo effect could not be avoided, which may exaggerate the efficacy of NIBS. Moreover, some studies did not describe whether the subjects were combined with the basic treatment. Therefore, it is unclear whether there was a synergistic effect between the treatments. (4) The overall quality of the research included in this systematic review was not high. Most of the studies did not specifically describe the method of random allocation and did not perform allocation concealment, which may generate selection bias. (5) Eight studies used shame stimuli, but did not test whether the blindness was successful, which may increase measurement bias.
## Clinical implications and recommendations for future studies
Meta-analysis showed that in patients with MCI, NIBS seems to improve the overall cognitive function, verbal fluency, and executive function, which suggests that NIBS may be a potential intervention for patients with MCI. Because the quality of the studies is not high, the conclusions should be treated with caution. We also believe that NIBS may offer an exciting novel treatment option in patients with MCI.
Future studies in this field should explore appropriate NIBS treatment parameters (e.g., stimulus type, stimulus intensity, stimulation frequency, stimulation site) for patients with MCI. For example, Ahmed et al. [bib_ref] Effects of low versus high frequencies of repetitive transcranial magnetic stimulation on..., Ahmed [/bib_ref] conducted a TMS intervention trial on AD, suggesting that high-frequency stimulation is superior to low-frequency stimulation. Moreover, a more sensitive and objective measurement tool should be used to assess the overall cognitive ability and specific domains of cognition. As memory loss is a major manifestation of MCI, recent studies have concluded that delayed memory and semantic memory tests can better predict whether MCI will progress to AD compared to other memory tests [bib_ref] Neuropsychological predictors of conversion from mild cognitive impairment to Alzheimer's disease, Gainotti [/bib_ref].
There are different types of MCI, such as amnestic MCI and non-amnestic MCI as well as single domain impairment or multiple domain impairment [bib_ref] Mild cognitive impairment, Petersen [/bib_ref]. Whether the response of different types of MCI to NIBS treatment is different remains to be discussed. In addition, since the adverse events reported in these studies are mainly transient dizziness and headaches, we believe that the setting of NIBS parameters in future studies should be based on safety guidelines [bib_ref] ethical considerations, and application guidelines for the use of transcranial magnetic stimulation..., Rossi [/bib_ref] [bib_ref] A technical guide to tDCS, and related non-invasive brain stimulation tools, Woods [/bib_ref]. At the same time, followup studies are needed to assess the long-term risks and benefits of NIBS treatment. In addition, in order to better assess the quality of the research, the author should follow the CONSORT guidelines [bib_ref] statement: updated guidelines for reporting parallel group randomised trials, Schulz [/bib_ref] when reporting research.
# Conclusion
NIBS may benefit the improvement of global cognitive function and verbal fluency in patients with MCI, while it has a slight positive impact on executive function. However, considering the different types of NIBS, and the discrepancies in intensity, frequency, locus, duration of stimulation, as well as the limited number of studies and the small sample size, these findings must be carefully explained. More sample size trials are needed, with more rigorously randomized controlled trial designed and standardized training programmes to draw specific and accurate conclusions.
## Additional file
Additional file 1 Search strategy.
[fig] Figure 1: Flow diagram for searching and selection of the included studies [/fig]
[fig] Figure 4: NIBS versus other intervention: executive ability [/fig]
[fig] Figure 5: NIBS versus other intervention: attentionFig. 6 NIBS versus other intervention: verbal fluency to improve the cognitive function of MCI patients. In terms of global cognitive function, NIBS significantly improved the global cognitive function in MCI patients. Within a specific cognitive domain, NIBS has a slightly significant benefit on executive function and a potentially positive impact on verbal fluency ability. Nevertheless, as verbal fluency was measured only in two studies, the results of this measure should be taken with caution. [/fig]
[fig] Figure 7: NIBS versus other intervention: adverse effects [/fig]
[table] Table 1: Characteristics of included studies in this systematic reviewBNT, Boston Naming Test; CMS, Clinical Memory Scale; CGI, Clinical Global Impression; DLPFC, dorsolateral prefrontal cortex; DSST, Digit Symbol Substitution Test; ERP, Event-related Potentials; FAB, Frontal Assessment Battery; HVLT, Hopkins Verbal Learning Test; IADL, Instrumental Activities of Daily Living; MMSE, Mini-mental State Examination; MMQ, Multifactorial Memory Questionnaire; MoCA, Montreal Cognitive Assessment; PAL, Paired Associated Learning; PC, Precuneus; PD-CRS, Parkinson's Disease Cognitive Rating Scale; PET, Positron Emission Tomography; PT, Physical Therapy; RAVLT, Rey Auditory Verbal Learning Test; RBMT, Rivermead Behavioral Memory Test; RCFT, Rey Complex Figure Test; RMT, Resting Motor Threshold; rTMS, repetitive Transcranial Magnetic Stimulation; TMT, Trial Making Tests; tDCS, transcranial Direct Current Stimulation; VFT, Verbal Fluency Test; WAIS-III, Wechsler Adult Intelligence Scale III; WCST, Wisconsin Card Sorting Test; WMS, Wechsler Memory Scale [/table]
[table] Table 2: Risk of bias summary: review authors' judgements about each risk of bias item for each included study [/table]
|
Comparison of the most likely low-emission electricity production systems in Estonia
To meet targets for reducing greenhouse gas emissions, many countries, including Estonia, must transition to low-emission electricity sources. Based on current circumstances, the most likely options in Estonia are renewables with energy storage, oil shale power plants with carbon capture and storage (CCS), or the combination of renewables and either oil shale or nuclear power plants. Here we compare these different scenarios to help determine which would be the most promising based on current information. For the comparison we performed simulations to assess how various systems meet the electricity demand in Estonia and at what cost. Based on our simulation results and literature data, combining wind turbines with thermal power plants would provide grid stability at a more affordable cost. Using nuclear power to compliment wind turbines would lead to an overall levelized cost of electricity (LCOE) in the range of 68 to 150 EUR/MWh (median of 103 EUR/MWh). Using oil shale power plants with CCS would give a cost between 91 and 163 EUR/MWh (median of 118 EUR/MWh). By comparison, using only renewables and energy storage would have an LCOE of 106 to 241 EUR/MWh (median of 153 EUR/MWh). OPEN ACCESS Citation: Baird ZS, Neshumayev D, Järvik O, Powell KM (2021) Comparison of the most likely low-emission electricity production systems in Estonia. PLoS ONE 16(12): e0261780. https://doi.
# Introduction
Estonia has the goal of being a climate neutral economy by 2050. There is currently active discussion about what strategy Estonia should take to reduce greenhouse gas emissions to reach that goal at an acceptable price. The largest source of carbon emissions in Estonia is the oil shale industry, and in particular the oil-shale-fired power plants . Oil shale is a type of fossil fuel that has been the main source of electricity in Estonia for decades and has enabled Estonia to have more energy independence. However, there is increasing pressure to limit its use. Even Eesti Energia, the company that operates the oil shale power plants, has pledged to stop producing electricity from oil shale by 2030 and to be carbon neutral by 2045. The four most likely strategies that could be used to achieve these goals and reduce the carbon emissions of Estonia's electricity sector are Renewable energy sources now have costs low enough to compete with fossil fuels, but wind and solar power are intermittent and require additional infrastructure to achieve a stable grid. Onshore wind turbines are currently the cheapest low-emission energy source for Estonia. However, there is only a limited number of suitable locations on land in Estonia and often locals do not want them built near them. Offshore wind turbines, although more expensive, avoid some of these issues and can take advantage of the higher wind power density in the Baltic Sea. Due to the relatively poor solar resources and higher installation costs in Estonia, solar panels are less effective and cost more than wind turbines. In recent years the cost of renewable energy sources, such as wind turbines and solar panels, has significantly decreased. Now the cost of electricity from these sources is generally on par with that of fossil fuel power plants, which makes them an attractive alternative to oil shale. Indeed, a state-owned energy company, Enefit Green, is pursuing development of a 1100 MW offshore wind farm 12 km off the coast of Hiiumaa. This large installation would likely produce about 2.6 TWh of electricity per year, which is significant when considering that Estonia's annual demand is only 8.4 TWh. However, wind and solar are intermittent and cannot consistently meet demand on their own or maintain important grid parameters, such as the frequency.
The electricity supply could be stabilized if wind turbines and solar panels were coupled with a nuclear power plant, energy storage, or a smaller number of oil shale plants fitted with carbon capture. Carbon capture and storage is being investigated as a potential solution and there is also active discussion in Estonia about whether or not to construct a nuclear power plant. Although Estonia can import electricity from neighboring countries to somewhat compensate for variation in the production from wind turbines, this is not a realistic substitute for having domestic dispatchable power because relying on imports would reduce Estonia's energy independence and might lead to an increase in prices. Additionally, many neighboring countries may also decommission their fossil fuel plants, meaning that there may not be a stable supply in neighboring countries that Estonia could rely on for imports to stabilize its own grid. Therefore, it is likely that Estonia would need to pair wind and solar power with a dispatchable form of electricity generation or storage.
Here we compare these various potential energy systems by performing simulations to see to what extent they are able to meet demand. Based on the simulation results we also estimate the levelized cost of electricity in each of these systems.
# Methodology
To study the performance of a variety of electricity systems in Estonia, we developed code to simulate the electricity demand and production from various sources. Functions for modeling electricity storage were also included. Each simulation covered a full calendar year with data calculated at 10 second intervals. Our code is written in Python and C++ and is available online (https://github.com/zmeri/electricity-sim-estonia). The code relies on the Scipy, Numpy, Pandas, and Matplotlibpackages.
To account for the uncertainty in the parameters underlying these simulations, we chose to use a Monte Carlo method. In this method the expected range for key parameters was described using distributions. In each simulation, parameter values were randomly selected from these distributions. The exact parameters used for these distributions can be seen from the code used for our simulation. Then the simulation was run 2000 or 3000 times (depending on the analysis) to get a distribution for each of the summary metrics we used to assess the performance of the specified electricity system. This enables the uncertainty of the results to be taken into account and quantified, which can allow us to determine if a difference between any two scenarios is significant.
Representative data for a few days of a simulation is shown in More details about the different elements of the simulation are given in the following subsections.
## Demand
We used a discrete cosine transform (DCT) to model historical electricity demand data in Estonia, and then we selected similar coefficients and transformed them using an inverse DCT to generate realistic demand data for the simulation. Historical data from the five years of 2016-2020 was obtained from Elering's website. By performing the DCT on each year separately, we obtained a set of 5 values for each variable in frequency space. From this set of values we were able to calculate the mean and standard deviation of each variable. To generate new demand data for the simulations, we simply selected a value for each variable in frequency space from the normal distribution described by its mean and standard deviation. We only used the first 5000 cosine terms because including additional terms made no significant difference in the result. Then, we transformed these generated values back using the inverse DCT to obtain demand data for the simulation. The DCT functions in Scipywere used to perform these transformations. For more details, see our code that was used to implement the simulation.
Generated data was used instead of actual historical data because using actual data for a single year or even a few years does not capture the full range of variation that might occur in the future and could bias the results. The demand for a given day can vary quite a bit depending on the year. This is shown in which shows the historical data along with an example of generated demand data. For instance, on Day 234 the demand for 2020 was significantly lower than that of the previous 4 years. Such variations might be because the weather on a given day can be different from year to year or because some social factors could influence it (as an example, see how water usage in Canada changed significantly during a gold medal hockey game. If only actual historical data is used, then that might introduce a bias into the simulation results because the full range of potential variation is not being represented.
We visually observed the generated demand data for several simulations and found it to replicate actual demand data quite well. It followed the patterns expected in the Estonian electricity grid, such as a higher electricity demand during the cold dark winter months and a lower demand during the summer. The model was even able to replicate patterns that are not immediately obvious, for instance the consistently low usage of electricity on the 23rd and 24th of June in Estonia.
## Production
We considered electricity production from combinations of the following sources: Other fossil fuels were not considered because oil shale is locally available and importing other fossil fuels would reduce energy independence.
[formula] - [/formula]
The capacity for hydropower, biomass, and waste was essentially kept the same for all the simulations (i.e. they were always selected from the same distributions). The annual mean capacity for these sources was set to be the same as available in 2019, as reported by Statistics Estonia. For biomass and waste power plants, this was equal to 160 MW and for hydropower it was 6 MW. Although biomass and waste are generally combusted in combined heat and power plants in Estonia, to our knowledge they generally still produce year-round. Data on wood fuel usage also indicates this by showing that wood fuel is still used during the summer months and the dip is not so large . For this reason we kept production from these sources constant throughout the year.
Because wind and solar resources are intermittent and depend on the weather and solar irradiation, the capacity factors for these sources were modeled using an inverse DCT, implemented in Scipy. This allowed us to model the variability of these resources.
For the wind capacity factor we used the same method as used for modeling demand. First, actual wind production data for Estonia was obtained from Elering's website. A DCT was performed on each year separately, and from these results the mean and standard deviation of each transformed variable could be calculated. These statistics were then used to calculate a normal distribution from which new values could be selected to generate wind production data for the simulation. For wind we only used the first 50,000 cosine terms because this was enough to accurately represent wind production for a year. An inverse DCT was applied to then obtain the wind capacity factor over time.
For the solar capacity factor, the first term in the series was set so that the average annual capacity factor matched the actual value for Estonia. The average capacity factor for solar panels was estimated to be 11% using the solar potential for Estonia given by the World Bank Group's Global Solar Atlas. The coefficients for additional cosine terms were selected to give output during the day and no output at night. To take into account the effect of weather on solar output, we randomly selected a factor for each day to multiply by the capacity factor. This allowed us to simulate the large differences between neighboring days due to weather events, such as rain, clouds, or snow, which more closely resembles the trends for real solar panels.
Because oil shale and nuclear power plants are dispatchable, it was easier to model those. The output of such power plants is still limited by various issues, such as operational problems and the need for maintenance, and we took this into account by multiplying the nameplate capacity by a capacity factor. The distributions for the capacity factors were chosen based on data in a National Energy Technology Laboratory report. In recent years, the capacity factors for coal and oil shale plants have shown a decreasing trend, and we expect this is largely due to market and regulatory constraints that make their operation less economical . We assumed that these constraints would be alleviated if CCS was added to the oil shale plants because CCS would reduce emissions and because some sort of governmental financial support would probably be needed for CCS to even be constructed. Therefore, we selected an 82% capacity factor for the oil shale plants based on the data from NETL. Nuclear power plants generally have a capacity factor of about 93% .
Both oil shale and nuclear power plants can be operated flexibly, i.e., their output can be reduced during times when less electricity is needed. It may come as a surprise to some that nuclear power plants can also be operated flexibly, but EDF in France has operated its 58 reactors flexibly for more than 30 years. These plants have demonstrated the ability to ramp between 20% and 100% capacity within 30 minutes when in load following mode. Most modern nuclear reactors can operate flexibly, including many, if not all, of the small modular reactors currently being developed. Small modular reactors offer additional flexibility because individual reactors can be shut off to further reduce the output of the power plant as a whole. We expect that the oil shale power plants can adjust their output at a rate of approximately 2.5 MW e /min (or about 1-2%P nom /min). A recent IRENA report shows a similar rate for coal power plants. For nuclear power plants a rate in the range of 1-3% P nom /min is also common. Therefore, we included this flexibility in the production models for oil shale and nuclear power plants by decreasing their output during periods when production exceeded demand (including some potential exports from foreign demand). The rate of change was kept reasonable to make sure it corresponded to the response rates actually achievable with these types of plants. For oil shale power plants, it was assumed that the lowest possible output they could operate at was 40% of their full capacity. For nuclear power plants the minimum was 20% of their full capacity.
It is also worth noting here that although the majority of nuclear power plants worldwide have quite large capacities, there are also multiple types of small reactors in use commercially. For instance, in India many small reactors have been built starting in the 1960s. Many of them have been domestically manufactured pressurized heavy-water reactors that have a capacity of about 200 MW e. And in recent years, there has been increasing interest in small modular reactors. Therefore, there are vendors available that could supply a nuclear power plant with the 530 MW e capacity assumed here.
Pyrolysis gas is also used to produce between 700 and 1000 GWh of electricity in Estonia. This pyrolysis gas is a byproduct of the shale oil production process, and is sometimes called semicoke gas or generator gas. In Estonia, oil shale is also used to produce a synthetic crude oil, called shale oil, via pyrolysis, and in addition to the oil a gaseous byproduct is produced. Eesti Energia feeds its pyrolysis gas to its Auvere and Estonia power plants for co-combustion. VKG has a separate combined heat and power plant specifically for burning pyrolysis gas, and this complex has a capacity to produce 87 MW of electricity. Based on data from recent years about the amount of electricity produced from pyrolysis gas, we estimate that there is approximately 150 MW e of capacity from pyrolysis gas in Estonia. Eesti Energia is planning to build an additional shale oil plant in the next few years, and based on this information we assumed a slightly larger pyrolysis gas capacity of 180 MW e in the simulations. However, because pyrolysis gas is a byproduct, it cannot reasonably be assumed that it could provide enough base power to stabilize the Estonian grid on its own. Therefore, the capacity was kept constant at 180 MW e in all the simulations in which pyrolysis gas was included.
Transmission and distribution losses were also deducted from the amount of electricity produced. These losses are about 8% in Estonia.
Note also that in the simulations production was not required to exactly meet demand. This is because Estonia is connected to the Nordic electricity grid (Nordpool). It has connections with Latvia and undersea cables connecting it with Finland. Through Finland and Latvia it is also connected with Skandanavia and Lithuania. Therefore, surpluses and shortfalls can be compensated for by importing or exporting electricity to the Nordpool grid, which helps keep the Estonian grid in balance. When the simulation gave extremely large surpluses or shortfalls for a particular system, this indicates that the specific system would most likely not be realistic and would not be a good option.
## Storage
For some of the configurations, energy storage was included. This was modeled by sending excess electricity to the storage system when production exceeded demand and drawing on energy in the storage system when production fell short of demand. The amount of energy transferred was limited by the power rating of the storage system. Both the power rating and the energy capacity could be adjusted to simulate different types of storage systems. Efficiencies were also taken into account both when charging and discharging. The exact algorithm used can be seen from our code: https://github.com/zmeri/electricity-sim-estonia/blob/ master/storage_func/storage_base.cpp For the purposes of this simulation, we assumed that surplus electricity was first stored before exporting.
There are a variety of potential energy storage technologies. We chose to use underground pumped hydro storage in our simulations because it is currently the most promising technology for Estonia. Pumped hydro storage is a well-established technology that currently accounts for the majority of global energy storage. Because Estonia does not naturally have large changes in elevation, the height differential required could be achieved by placing the lower reservoir underground. Indeed, the company Energiasalv is attempting to construct such an underground storage facility in Northern Estonia near Paldiski. Eesti Energia is also planning a smaller 50 MW underground pumped storage project in Ida-Virumaa county in Northern Estonia. Also, pumped hydro storage is generally estimated to be one of the cheapest forms of long-term energy storage.
Many of the other potential energy storage technologies are currently much more expensive, such as battery technologies. For this reason, these technologies were not considered. Additionally, many storage technologies have yet to reach a mature stage of development, which is another reason we did not consider them for this study. We did perform a few simulations using hydrogen storage for comparison; however, the cost of hydrogen storage was significantly higher than that of pumped hydro storage. In large part this seems to be due to the fact that fuel cells and electrolyzers currently cost much more than pumps and turbines with the same power rating. Fuel cells and electrolyzers also generally have lower efficiencies than pumped hydro storage, which would also cause hydrogen storage to have a comparatively higher cost. Also, storing the hydrogen might be challenging because Estonia does not have suitable underground formations, such as salt caverns, so more energy intensive pressurized or cryogenic options would need to be used. Therefore, we chose underground pumped hydro storage as the technology for comparisons in this study.
The round-trip efficiency for underground pumped hydro storage was taken to be about 75%, so the efficiency for a single charge or discharge was set to be 86.5%.
## Import and export
We assumed that it is possible to export an amount equal to 30% of the electricity demand in Estonia. This value was chosen based on data from 2018 in which net exports from Estonia were about 26% of domestic demand. It is worth noting that in 2019 and 2020 production from oil shale power plants dropped, probably due to higher carbon emission prices, and currently Estonia is no longer a net exporter. However, if economically competitive electricity production would be available in Estonia, we assume that exports similar to those in 2018 could be achieved. The electricity connections with Latvia and Finland can handle up to 2463 MW of exports and 2275 MW of imports, as of January 2021, so it seems exports would be limited mainly by the demand in neighboring countries and not the power rating of the transmission cables.
Any surplus electricity that could not be exported was considered to be unused potential. In practice, this surplus is often handled by curtailing electricity production, such as by rotating the vanes of wind turbines to reduce production, or by paying other grids to accept the excess power.
## Cost estimation
Here we describe some of the assumptions and methods used to estimate the levelized cost of electricity (LCOE) for oil shale plants with CCS and alternative technologies. The Python code and other materials used in these analyses can be found in the Github repository for this project (https://github.com/zmeri/electricity-sim-estonia/).
In our calculations we used a Monte Carlo method. In essence, this involved specifying a distribution for the parameters affecting the cost of a given technology and then sampling many times from each distribution and combining to get an overall expected range for the levelized cost of electricity. Using this Monte Carlo method allowed us to take into account the uncertainty of the parameters we used. For the simulations investigating the energy storage required to achieve balanced electricity imports and exports (subsection "Energy storage required"), the simulation was repeated 1900 times for each set of parameters for the storage system. For the simulations comparing 8 different likely scenarios (subsection "Comparison of potential systems"), 3003 repeats were performed.
For all the calculations we performed we assumed a discount rate of 9% and 2020 euros was used as the base currency. The formula we used to calculate the LCOE is the one most frequently used. It is presented in an IRENA reportand elsewhere in the literature, and we show it here as Eq 1.
[formula] LCOE ¼ P n t¼1 I t þM t þF t ð1þrÞ t P n t¼1 E t ð1þrÞ tð1Þ [/formula]
In Eq 1 LCOE is the average lifetime levelized cost of electricity generation, I t are the investment expenses in year t, M t are the operation and maintenance expenses, F t are the fuel expenses, E t is the electricity generation for year t, r is the discount rate, and n is the lifetime of the system.
Oil shale. For the oil shale plants, we first needed to estimate the current cost of electricity. The overnight construction cost could be estimated based on the cost of the Auvere oil shale power plant, which was constructed in Estonia between 2012 and 2015. The total cost of the plant was 610 million euros and it has the capacity to produce 300 MW of electricity. To calculate the overnight construction cost, the interest during construction must also be excluded. The interest during construction includes both the cost of debt as well as the amount needed to provide an acceptable rate of return to equity investors. Using the formula given in Section 2.3.4 of reference, a construction time of 4 years, and the discount rate used throughout this study (9%), we estimated the interest during construction and subtracted that from the reported cost of the Auvere plant. We also accounted for inflation to update to 2019 euros using the Chemical Engineering Plant Cost Index. Based on this calculation, we estimated the overnight construction cost of a new oil shale power plant to be about 1870 EUR/kW e .
To estimate the operating expenses, we used information from the media and the literature. For instance, on July 2, 2019 Eesti Energia stopped electricity production because its production cost was higher than the market price. The wholesale market price was about 30-50 EUR/MWh at that time. Also, Eesti Eneriga has stated that the cost of CO 2 credits makes up more than half of the production cost of electricity for the oil shale power plants, and these credits cost about 25-27 EUR/MWh at that time because oil shale emits about 1 tonne of CO 2 per MWh produced. Additionally, it takes approximately 1 tonne of oil shale to produce 1 MWh of electricity, and oil shale costs about 12 EUR/tonne, which sets a lower limit for the operating costs. We also found a government document from 2005 that indicates the operating costs of the oil shale power plants was approximately 30 EUR/MWh. However, these older plants had a lower efficiency of about 26-30%, and a new plant would probably have an efficiency higher than 40%. With the higher efficiency the operating costs would also be expected to decrease to about 20 EUR/MWh. Based on this information, the operating costs for oil-shale-fired power plants is probably somewhere between 15 and 25 EUR/MWh.
We also included the cost of CO 2 credits when calculating the LCOE. The price of CO 2 on the European Emissions Trading System can change relatively rapidly. For instance, between November of 2020 and May of 2021 the price of CO 2 increased from about 27 EUR/tonne to about 50 EUR/tonne. This makes it difficult to estimate what the price might be in the future and adds uncertainty to our estimate of the LCOE for oil shale. We used an average value of 40 EUR/tonne and a wide standard deviation of 15 EUR/tonne on the distribution to account for this uncertainty in the price. When excluding the cost of CO 2 credits, we estimated the LCOE of an oil shale plant without CCS to be approximately 60 EUR/MWh.
The newer Auvere plant has also been designed so that it can use up to 50% biomass in its fuel mixture, and indeed, large amounts of biomass are currently used. In our analysis, though, we considered biomass and oil shale separately. There is not enough biomass in Estonia to provide the base load needed to stabilize the grid. Most of the wood available for energy use is already being used, and this capacity is already included in the simulations. Although Estonia has other potential biomass resources, such as manure or energy crops like grasses, the size of these resources is still smaller than Estonia's electricity demands and the cost of these resources would likely be prohibitively expensive. Therefore, it is unlikely that biomass could replace oil shale as the base load generator in Estonia.
To estimate the cost of electricity if CCS would be incorporated into the plant, we used the full range of cost estimates given in the literature. This data is presented in another article of ours and is also available on Open Science Framework (https://osf.io/8b2dw/ ). We also calculated a second estimate based on next generation capture technologies that are still being developed but are expected to have lower costs. Additionally, about 25 EUR/MWh was added to the cost to account for transportation and storage because, based on literature data, we expect that this would be the cost for transportation from Estonia to the North Sea for storage. We assumed that CO 2 emissions would be reduced by 90%, and accordingly we reduced the cost of the CO 2 credits in the LCOE estimate.
It should be noted that CCS technologies are still new and under development. The only technology that has so far been applied commercially is absorption with amines, and that has only been used at two power plants. Note that the cost estimates we used here were for an nth-of-a-kind plant, which means that several such commercial plants have already been built. If a first-of-a-kind CCS plant is selected instead, then the cost can be expected to be higher.
Solar. Although we could obtain data on the global cost of utility scale photovoltaics, we wanted to perform an estimate specifically for Estonia because it has poorer than average solar resources. For example, according to the World Bank's Solar Atlas, solar panels in Estonia produce about 1000 kWh annually per kW of installed capacity. By comparison, in Spain that value is about 1600 kWh/kWp and in the most sunny areas in the world the solar potential can exceed 2000 kWh/kWp.
To calculate the cost of solar electricity in Estonia specifically, we took estimates of the capital and operational cost of solar panels from the IRENA reportand the solar potential in Estonia from the Solar Atlas. Then we calculated the discounted costs and energy production over the lifetime of the solar panels (assumed to be 25 years).
Nuclear. Based on literature data for the cost of nuclear power plants, we estimated the LCOE for a nuclear power plant in Estonia. The construction cost is by far the largest contributor to the cost of electricity for nuclear power plants. In estimating the overnight construction cost, we focused on small modular reactors (SMR), which is the type that has most commonly been proposed in discussions in Estonia. More specifically we estimates for Generation III + SMRs. Unlike the large nuclear reactors that have traditionally been built, SMRs would have a capacity of only dozens to maybe a few hundred MW e . These smaller reactors have been touted as safer and less prone to the construction delays encountered with large reactors. Additionally, their smaller size makes it more feasible to implement them into smaller electricity grids, such as in Estonia.
To estimate the overnight construction cost of small modular nuclear reactors, we relied mainly on the expert assessments published by Abdulla et al.for Generation III+ SMRS, and also took into account data from other literature sources. Based on these sources, we used 4400 EUR/kW e as the average overnight construction cost for an nth-of-akind plant. This is in the same range as data for existing conventional nuclear reactors. On average, literature estimates seem to show that small modular reactors have a slightly higher overnight construction cost per kW than conventional large reactors. However, there is also the potential for cost savings with small modular reactors due to aspects such as the potential for factory fabrication and shorter construction times.
The actual cost of a nuclear reactor can vary widely, as can also be seen from historical data. The experts interviewed by Abdulla et al.also gave estimates ranging from 3200 to 7100 USD/kW e , which again highlights the variability in the cost. So, we used a wide distribution to account for this uncertainty. We chose to use a lognormal distribution to describe the potential range of the capital cost because the historical data seems to follow such a distribution. Using a lognormal distribution also allowed us to take into account the risk that the construction can go well over budget.
The estimated value used for fuel costs (including waste disposal) was 6 EUR/MWh and for operations and management expenses 10 EUR/MWh. These estimates were taken from D'haeseleerand were similar to those given elsewhere in the literature.
Note that small modular reactors are still a new technology that is actively being developed. One 70 MW plant is already in commercial operation in Russia. It is a floating nuclear plant in the Arctic Ocean that uses two KLT-40S reactors normally used for nuclear ice-breakers. This highlights how some of the small modular reactors simply use smaller scale versions of existing technology. To some extent, small reactors have already been around for decades because they have been used to power naval vessels and smaller 200 MW reactors have been built in India for decades. In 2021 a 210 MW plant is supposed to come online in China, and this will use a high-temperature gas-cooled reactor . Construction has also resumed on a 25 MW reactor in Argentina, which is expected to be completed in the next year or so. Several more small modular reactors are also planned or under construction. So, the technology is just now reaching commercial implementation and has not yet been widely adopted. If a small modular reactor is selected for Estonia that has not been built before at an industrial scale, then a higher cost should be expected than the estimates given in this article.
Pyrolysis gas. Pyrolysis gas is a byproduct of the pyrolysis process used to produce shale oil from oil shale. There are several shale oil plants in Estonia that produce the gas. Pyrolysis gas has a number of cost advantages over oil shale. First, because it is a byproduct it is essentially a free fuel. Second, we expect that the overnight construction cost of a pyrolysis gas power plant would be somewhat less than an oil shale power plant: we used 1600 EUR/kW e as the estimate for our calculations. Third, less CO 2 is emitted per unit of electricity produced, which means CCS would cost less.
For the CO 2 emission factor for pyrolysis gas, we used an average value of 0.2 tonne CO 2 / MWh. There are two different types of shale oil production technologies used in Estonia: gas generator and solid heat carrier retorts. These different technologies produce gas with different compositions and heating values. The pyrolysis gas from solid heat carrier retorts has a higher heating value and a lower CO 2 emission factor. Because gas from each type of retort is combusted in roughly equal proportions, we simply took an average of the emission factors of the two different types of pyrolysis gas.
To estimate the cost of CCS, we used the same estimates as used for oil shale power plants, but because pyrolysis gas emits less CO 2 the overall contribution of CCS to the LCOE is smaller for pyrolysis gas.
It should be noted that for pyrolysis gas to be available shale oil production must continue in Estonia. Although shale oil is largely exported, and so the CO2 emissions from burning it do not generally add to Estonia's reported total, it still affects the environment. In the future there may also be a desire to end shale oil production in Estonia, and this would remove pyrolysis gas as a potential resource. Currently though, shale oil production is still expected to continue.
Biomass and hydro. The estimated cost of electricity production from biomass was taken from an IRENA report, and the average estimate was 100 EUR/MWh. For hydropower the estimated average cost used was 80 EUR/MWh.
Note that we also calculated the LCOE of producing electricity in a power plant similar to the one used for oil shale. This was done because biomass is currently co-fired with oil shale in some power plants in Estonia and we wanted to compare the cost of such co-firing. For this reason, in the calculation we used essentially the same parameters as for an oil shale power plant except that we used a different price for fuel and a slightly lower efficiency (as would be expected for biomass). The LCOE calculated was added toHowever, in all the other simulations and analysis performed in this article we used the LCOE for biomass taken from the IRENA report. This was because much of the biomass in Estonia is used in combined heat and power plants and such plants have different costs.
Storage. We took data and estimates from the literature to estimate the cost of underground pumped hydro storage. Energiasalv has estimated that their proposed underground storage facility in Paldiski, Estonia would take 8 years to construct and have a lifespan of 60 years, so we used those parameters for our calculations. We took estimates for the capital costs of the reservoirs, pumps, turbines, and other equipment from Kapila et al.and Guo et al.. Estimates from Guo et al.indicate that underground pumped hydro storage is Comparison of the cost of various electricity production technologies. Global costs for solar and wind were taken from ref, and the average line for those is the global weighted average. The global residential price data is from.
https://doi.org/10.1371/journal.pone.0261780.g003
## Plos one
Comparison of low-emission electricity systems in Estonia roughly twice as expensive as conventional aboveground configurations, and this difference was taken into account when estimating the cost. The operational costs and scaling factors were also obtained from Kapila et al.. We also took into account the cost of electricity for charging and the potential revenue gained when discharging.
System LCOE. Because production from solar and wind is intermittent, these energy sources incur additional costs on the entire electricity system. Such costs include additional power generating capacity for when wind turbines and solar panels are not producing enough (balancing costs), additional transmission lines and more complex congestion management (grid costs), and loss of revenue due to overproduction or load reduction at dispatchable power plants (profile costs). These additional costs are not accounted for in the LCOE of solar and wind separately.
Therefore, to compare different potential electricity systems, we used the idea of a System LCOE, which was proposed by Ueckerdt et al., to take into account these additional costs. We calculated the balancing and profile costs as part of the simulations. The balancing costs were incorporated by using the simulation data to determine how much additional production or storage capacity would be needed to achieve a balanced level of imports and exports. The profile costs were also calculated automatically in the simulation because reductions in electricity production were directly accounted for in the calculation of the cost. For the grid costs, we used literature estimates of the specific cost of grid reinforcements for wind power. We multiplied this specific cost by the capacity of wind and solar installations and added the respective value to the capital cost of each.
## Job creation
We also briefly compared the various low-emission technologies based on other environmental and social factors. One of the metrics we used for comparison was the estimated number of jobs created. We calculated the number of permanent jobs created using information in the literature. For this analysis we only considered permanent jobs directly created at the energy companies producing the electricity. Additional jobs are created during construction, but because these jobs are only temporary, we left them out of the calculation. Also, many indirect jobs are also created in the surrounding economy, for example in companies that serve the energy company and in local businesses where employees of the company spend their money. However, these indirect jobs created were not included in this calculation because they are more difficult to quantify and doing so would require a separate detailed study.
For oil shale power plants with CCS, we estimated the number of jobs from mining and operations based on 2020 employment data for Enefit Power (formerly Enefit Energiatootmine AS and Enefit Kaevandused AS), which is the oil shale power plant operator in Estonia (see in. Enefit Power also produces shale oil, so we split the total number of employees between the electricity and shale oil businesses based on the percentage of the oil shale used for each product. Data for Enefit's 2019 oil shale mining and shale oil production was obtained from the yearbook for the Estonian Oil Shale Industry. The amount of oil shale used specifically to produce shale oil was then estimated by assuming that 0.2 tons of shale oil is produced from 1 ton of oil shale. Based on this calculation we estimate that about 70% of the oil shale is used for electricity production. Assuming that company employees are split between the two products by a similar proportion, we calculated that electricity production employed about 1050 people at Enefit Power in 2020. Given that the total installed capacity of Enefit Power's oil shale power plants was 1490 MW, this gives a specific ratio of 0.71 jobs per MW of capacity. Including a CCS plant would also add more jobs, and based on data from the Petra Nova plant in Texas, USA, a CCS plant would create an additional 0.071 jobs per MW. Adding these ratios together and multiplying by the 600 MW of oil shale plant capacity assumed in our scenario used for comparison gives approximately 470 jobs created.
For the nuclear plant, we directly used estimates provided by Hõrakin a study commissioned by Fermi Energia. They estimated that a 300 MW small modular reactor would require 170 employees. For our comparison scenario, 530 MW of nuclear power was assumed, which would create about 300 jobs.
For the energy storage technology, the jobs created by additional wind turbines and solar panels must also be included. This is because energy storage needs to be coupled with something that produces electricity, so to appropriately compare with oil shale and nuclear plants (which produce their own electricity), the additional wind turbines needed to produce the same amount of electricity were also included. For the pumped hydro storage, a ratio of 0.25 jobs/MW was calculated based on the number of people employed in operations and management at hydropower plants in the United States. For the 1300 MW of hydro storage in our comparison scenario, this equates to about 330 jobs. To this we added the operations and management jobs created by wind turbines. For wind turbines, 0.36 jobs are created per MW of capacity, based on an average of literature estimates. For our scenario being compared, 1655 MW of additional wind turbines would be needed to provide the same amount of electricity as the oil shale or nuclear plants. This means that about 600 jobs would be created by those wind turbines.
# Sensitivity analysis calculations
To estimate the sensitivity of the results to variations in individual parameters, we ran some simulations where only one parameter at a time was allowed to vary. The remaining parameters were assigned their average values. For each parameter, the simulation was run 200 times and by calculating the variation of the average system cost and the net surplus the effect of each parameter could be estimated. Parameters related to hydro and solar power were not included in the analysis because the capacities for these two energy sources were much smaller and adjusting their parameters did not have a significant impact on the simulation results. The code for this sensitivity analysis is also included in the Github repository for this project.
# Results and discussion
## Cost of electricity production alone
As a starting point, we compared the cost of various electricity production methods. The cost of different production technologies is generally compared using the LCOE (see Eq 1).
The comparison is shown inThe wholesale market price of electricity in Estonia (the Nordpool next day price from January 1, 2014 to September 17, 2021), as well as data on the residential price of electricity in different countries, have been added for reference.
One of the first observations is that the current cost of electricity for oil shale plants is already above the wholesale market price in Estonia and is in the range of residential electricity prices. Note that we estimated the current cost based on literature data and publicly available information since Eesti Energia does not disclose its exact production cost (see the Methodology section for details on all the calculations). However, our estimate seems to be in the correct range given that within the past few years the Eesti Energia oil shale power plants have even temporarily shut down because the price of electricity was lower than their costs. A large part of their costs is due to the CO 2 credits that must be purchased by the plants to offset their emissions under the EU emissions trading system, and the LCOE for oil shale power plants also includes the cost of those credits. When excluding those credits, we estimate the LCOE of an oil shale plant without CCS to be approximately 50 EUR/MWh. This shows that the current climate strategy of the EU is putting pressure on oil shale power plants to reduce their emissions, as is intended.
Although CCS could help oil shale power plants avoid the cost of those CO 2 credits, implementing CCS will still lead to a significant increase in the cost, which may be one of the reasons it has not yet been implemented. Here we collected a wide range of literature estimates of the cost of carbon capture and storage to estimate the additional cost of CCS (data available on Open Science Framework at https://osf.io/8b2dw/). As seen inadding CCS significantly increases the cost of electricity for oil shale. Indeed, this is often cited as a reason not to invest in CCS. For instance, in 2015 the UK government canceled a competition that would have provided 1 billion GBP for large scale CCS projects. One of the major justifications was that CCS would cost too much money. The next generation of CCS technologies is expected to have lower costs, as shown inbut more development will be required and there is the chance that those lower estimates will not be realized at the industrial scale.
But, the lack of strong support for CCS is about more than simply its cost. In many cases the decision has been to support renewable technologies instead of cleaning up existing fossil fuel operations. Large-scale commercial wind and solar installations have become a reality, and along with the increased adoption have come significant decreases in the cost of these renewables. By contrast, based on the literature data that we have collected, the cost estimates for CCS have not changed significantly in the past two decades (see the literature data we compiled: https://osf.io/na5gy/). This is likely because any new commercial projects have still used the same absorption technology. So far no one has been willing to put up the funding to construct an industrial scale plant using a new CCS technology.
The one situation in Estonia where CCS could be economically favorable is with combustion of pyrolysis gas. Pyrolysis gas is a byproduct of the production of shale oil. Shale oil is a synthetic crude oil that is also produced from oil shale via pyrolysis, and a mixture of gaseous compounds is also formed during the process. Because pyrolysis gas forms less CO 2 per unit of energy, the cost of CCS per MWh of electricity is significantly lower (see. Additionally, because pyrolysis gas is a byproduct it is essentially a free fuel for the oil shale companies. Between 700 and 1000 GWh of electricity is already produced annually from pyrolysis gas in Estonia, and based on its lower cost and carbon emission factor, it could be advantageous to continue using it and implement CCS for it.
Wind and solar power have also become quite affordable as a result of consistent cost reductions over the past decade. Now the LCOE for wind and solar technologies has fallen to the level where it can directly compete with current fossil fuel plants. Recent estimates of the cost of solar and wind from an IRENA reportare shown inhighlight how the global average cost for solar and onshore wind is even competitive with the current oil shale power plants without CCS. Granted, solar cells perform worse than average in Estonia due to the poor solar potential, so when we recalculated the LCOE for solar in Estonia specifically it is much higher. And if the choice is between implementing CCS or renewables, it does not appear that CCS has a cost advantage. Offshore wind can provide electricity at costs lower than an oil shale plant with CCS and solar has roughly the same LCOE. Given the negative public image that fossil fuels have, it could be difficult to gain support for CCS over renewables, especially if no cost benefit exists.
One important advantage that an oil shale plant has over wind and solar is stability. Renewable energy sources that are intermittent, such as solar and wind, need to be balanced with technologies that can stabilize the grid and reliably meet demand even when the sun is not shining and the wind is not blowing. Additionally, the grid needs to have some way of stabilizing important parameters of the power grid, such as the frequency. Currently, this is accomplished by large plants that produce a steady, stable output, which generally means coal, natural gas, nuclear, or hydroelectric power plants. In Estonia specifically, the oil shale power plants can provide this reliable base load needed by the power grid. If Estonia were to shut down its oil shale power plants, it would need either a nuclear power plant or utility-scale energy storage system to meet the load.
A nuclear power plant would likely cost less than implementing CCS, as illustrated inWe estimated the LCOE of a nuclear power plant based on literature data. We used data on the construction costs of small modular reactors, since these are the reactors currently being considered in Estonia, but using data for conventional large reactors gives a similar LCOE. Fermi Energia is also actively promoting the construction of a nuclear power plant in Estonia. For comparison, we used the budget they have published to calculate the LCOE of their proposal. Based on our calculations, the LCOE for the Fermi Energia nuclear plant would be 78 EUR/MWh, which is quite similar to the median LCOE we calculated based on literature data.
The LCOE is significantly less than that of an oil shale plant with CCS. Based on our cost estimates, there is a 94% chance that nuclear would have a lower LCOE than an oil shale plant with current CCS technologies. Even if a next generation CCS technology would be used, which are expected to be cheaper, but require more development, there is an 85% chance nuclear would be cheaper.
Other sources have also indicated that a nuclear plant could be preferable to a fossil fuel plant with CCS. For instance, the Intergovernmental Panel on Climate Change also estimated that a nuclear plant would cost less than a coal plant with CCS (see 19 in. The IEA, in their report "Projected Costs of Generating Electricity 2020", also estimated that using CCS with coal or lignite power plants would cost about 50 USD/MWh more than nuclear. And in a recent report, the Stockholm Environment Institute recommended transitioning Estonia from oil shale to electricity based on wind turbines and a nuclear power plant.
One final renewable resource is biomass, but it cannot be considered a viable replacement for the oil shale power plants. The reason is that there is not enough biomass to replace them. Although Estonia does have significant forest resources, more than half is used for non-energy purposes. Of the approximately 5.5 million m 3 that is currently used for energy production, about 2.3 million m 3 is burned directly in homes for heating. The remaining 3.2 million m 3 from the forest, along with about 2 million m 3 of waste from the wood industry, is already used for energy production, mainly in combined heat and power plants or for producing pellets . Therefore, to add the approximately 600 MW e of capacity needed to provide a stable base of dispatchable power, new biomass sources would be needed. The only other biomass resources with significant potential in Estonia are manure (for biogas production) and energy crops (grasses). However, the estimated potential for these sources is, respectively, 4.5 and 21 PJ. For comparison, Estonia uses about 8.4 TWh of electricity per year. Given the efficiency of generating electricity from thermal energy, which is generally 30-40% for biomass combustion, only about 2.5 TWh in new generation could potentially be obtained from biomass, which would be about 285 MW e of capacity. Thus, even if all the manure could be utilized and grasses could be grown as energy crops, they could not provide enough base capacity on their own. Additionally, collecting and utilizing large amounts of manure and grasses would be logistically challenging and likely expensive.
Furthermore, the LCOE for production from biomass is higher than that of oil shale (excluding CO 2 credits). This means that cofiring biomass with oil shale in a plant with CCS would cost even more than just using oil shale with CCS. Although biomass combined with CCS is often touted as a way to achieve negative life cycle emissions of CO2, it would be more costly. If in the future some sort of credit was provided to promote CCS with biomass, then it might make economic sense, but currently no such incentive is available in Estonia. And, it should be noted that the LCOE shown infor biomass used only for electricity production. The LCOE for a combined heat and power plant would be different and potentially higher. Due to the limited resources and higher cost, only the existing biomass capacity was included in further simulations and analyses.
## Energy storage required
The cost of electricity production alone is only one piece of the picture. Because wind power is non-dispatchable, energy storage also needs to be included in the grid if wind power is to make up a majority of the electricity production in Estonia.
We first investigated how much electricity storage would be needed to achieve a certain net surplus/deficit in Estonia. For these simulations, only renewable resources were included: so wind turbines (3800 MW) and the solar, biomass, and hydropower capacity already available. 3800 MW was selected as the wind turbine capacity because, at this scale, the raw amount of electricity produced would be approximately equal to the demand in Estonia, which means that any shortfalls are mostly caused by the intermittency of wind. We chose to model the energy storage as an underground pumped hydro system. We selected this technology because pumped hydro storage is mature and is currently one of the cheapest forms of energy storage. Also, two Estonian companies, Energiasalv and Eesti Energia, are currently attempting to construct underground pumped hydro storage facilities in Estonia.how an increasing amount of electricity storage affects the net electricity surplus/deficit, the amount of overproduction, and the total amount imported for a year with such a system. For reference, Estonia's electricity consumption for all of 2016 was about 8380 GWh. The amount overproduced is the electricity that probably could not be used domestically or exported. In reality curtailments would likely be used to require electricity producers to limit their production. Overproduction is a common problem with electricity grids that have high shares of non-dispatchable power, such as solar and wind.
Fromcan conclude that at a storage capacity of about 100 GWh (with 3800 MW wind turbines) Estonia would be close to balancing electricity imports and exports. To put this in perspective, 100 GWh would be enough to completely cover Estonia's average electricity demand for about 4 days. However, curtailments on the order of 5-10% would probably still be necessary to handle times when wind turbine production exceeds the storage capacity. That is, some of the wind turbines may need to be shut off at times to avoid producing more electricity than can be used, stored, or exported. The storage capacity would need to be at least an order of magnitude larger to avoid curtailments all together. Such a large storage capacity, however, would cost significantly more and probably would not make economic sense. On the other hand, a smaller storage capacity could be used if policymakers determine that a higher amount of imports and curtailments is acceptable. Also,that storage systems smaller than 1 GWh would have essentially no effect on the net electricity deficit.
The amount of storage needed would decrease if there was a larger capacity of wind turbines. For instance, with 4000 MW of wind turbines, only about 30 GWh of storage would be needed to balance imports and exports. This is because overall more electricity would be produced and provided to the system, reducing the need for storage. However, overproduction of electricity would also increase, leading to a higher level of curtailments. Curtailments increase the levelized cost of electricity from wind turbines because the same capital costs must be spread over a smaller volume of production. Therefore, increasing the number of wind turbines to reduce the need for storage would probably not reduce the cost of the overall system.
## Plos one
## Comparison of low-emission electricity systems in estonia
If there is less than 3800 MW of wind turbines, it is unlikely that enough electricity could be produced to meet Estonia's demand from wind turbines (along with the existing biomass and solar capacity).
We also investigated how the power rating of the storage system affects the amount of energy lost. As seen fromif the storage power is too low, then increasing the storage capacity has essentially no impact on the net electricity deficit. In other words, the storage power must be large enough to make use of a larger storage capacity.
It is also worth noting that at a certain point, increasing the storage capacity and storage power has basically no effect on the net deficit. For the given simulation, this occurred at about 100 GWh of storage and 800 MW of storage power. Above those levels, the system was instead limited by the installed wind turbine capacity, which was set at 3800 MW for this analysis. One potential way to use such excess capacity would be to import and store electricity from other countries. We did not investigate such scenarios in this study, though, because the more pressing question is how Estonia can still meet domestic demand when switching to a low-emission electricity system. Therefore, we assumed that the smallest viable storage system would be selected to reduce the overall cost of the storage in the electricity system.
## Comparison of potential systems
What we really want to know is the cost and performance of the system as a whole when taking into account the natural fluctuations in the grid and the need for energy storage. Although wind and solar power have competitive production costs, they are also non-dispatchable and must be coupled with another stable power supply to compensate for the fluctuations.
We selected 10 scenarios for comparison (see. To determine how much storage to include in each scenario, we used the results of the earlier analysis to estimate how much energy storage might be needed. And again, underground pumped hydro was chosen as the storage technology. For each scenario, a total capacity was selected that would produce about enough electricity to cover demand in Estonia, so the differences in the net surplus are mainly due to mismatches between supply and demand and not a lack of capacity. Only wind turbines were included as the additional renewable source because solar power in Estonia has a higher LCOE (see. Additionally, even in the "Oil shale with CCS" and "Nuclear" scenarios 400 MW of wind turbines was included because there is already about that amount of wind capacity installed in Estonia (315 MW, as of 2019. The results of the simulations for these scenarios are shown inand the raw data from the simulations is available on Open Science Framework (https://osf.io/8b2dw/). As can be seen, about 100 GWh of electricity storage would be needed to provide comparable stability to using an oil shale or nuclear power plant. With less storage there would be an electricity deficit and more electricity would need to be imported to make up the difference. The surplus duration curves shown inshow how having less storage leads to many more days in which a deficit would occur.
## Scenario wind turbines (mw) oil shale or nuclear (mw) pyrolysis gas (mw) storage capacity (gwh) storage power (mw)
## Renewables
For both the scenarios with 100 GWh of storage and those with a thermal power plant, the amount of imported electricity would be approximately in balance with the amount exported, on average. However, systems with electricity storage would cost significantly more, even when taking into account potential revenue from electricity arbitrage. Using 100 GWh of electricity storage to stabilize the grid would, by our estimates, cost about 40% more than using oil shale and pyrolysis gas power plants with CCS and about 50% more than a nuclear power plant. This finding is consistent with the results from Zappa et al.in which they estimated that a 100% renewable system in Europe would cost about 30% more than systems with nuclear or CCS technologies. Pleßmann et al.estimated that the global average cost of electricity for a 100% renewable grid would be 142 EUR/MWh, which is reasonably close to the mean of 152 EUR/MWh calculated here. However, Pleßmann et al. seem to have underestimated the cost of the power to gas technology used in their model. Pleßmann et al. assumed the capital costs would be 940 EUR/kW, but other sources show that even the electrolyzer alone would cost more than that. In summary, using only electricity storage and no thermal power plants would most likely lead to significantly higher electricity prices.
We also performed some calculations using hydrogen storage instead of underground pumped hydro, but hydrogen storage was significantly more expensive. This seems to be due to the higher cost of fuel cells and electrolyzers compared to the turbines and pumps used in a pumped hydro system. The lower efficiency of hydrogen energy conversion might also be a contributing factor, and large scale hydrogen storage may be more challenging in Estonia due to a lack of suitable geological formations. Battery technologies, including flow batteries, currently cost significantly more than either of these options.
Although it would be quite expensive if electricity were provided almost solely by wind power, having some amount of renewables actually should help lower the cost. This is shown bywhich shows simulation results for how the cost would change as power plant capacity is replaced by wind turbines. There is a drop in the cost for wind capacities up to about 400 MW because this is roughly the potential for onshore wind turbines in Estonia (Meeliste et al.estimated 500 MW as the maximum). As of 2019, there was already 315 MW of installed onshore wind turbines in Estonia. Onshore wind is significantly cheaper than the other sources of low-emission electricity in Estonia, so adding this helps reduce the levelized cost. Offshore wind turbines, although more expensive than onshore ones, are still expected to be less expensive than oil shale and probably will be cheaper than nuclear if the cost of wind power continues to decrease (see. The implication is that some amount of renewable electricity generation should be in the mix in Estonia because wind turbines (especially onshore wind turbines) are most likely the cheapest form of low-emission energy.
As the percentage of renewables in the grid increases, the net surplus of electricity decreases and the overall cost of electricity increases. As shown infor wind capacities up to about 50-60% of Estonia's demand, thermal power plants could still stabilize the grid enough to make Estonia a net exporter. However, at higher levels of wind penetration the power plants could not smooth the variations on their own, and additional energy storage or peaker plants (such as natural gas plants) would be needed. Also, at higher capacities, wind turbines would need to be built offshore or solar panels would need to be used, both of which are more
## Plos one
Comparison of low-emission electricity systems in Estonia expensive. These factors would increase the overall cost of electricity as more wind and solar power is added to the Estonian grid. This is why the cost of a system with wind turbines and nuclear power would increase as the proportion of wind power increases (see.
For oil shale plants with CCS, offshore wind turbines actually have a lower cost, so initially adding more wind turbines decreases the overall system LCOE. Additionally, as the electricity generating capacity of the oil shale industry would decrease, pyrolysis gas would account for a larger share of the capacity. Since pyrolysis gas is cheaper, this would decrease the overall cost of a system based on oil shale. However, at about 2000 MW of wind capacity, the cost levels off and even slightly increases for the scenario with oil shale and pyrolysis gas power plants. This is because a high penetration of renewables starts influencing the stability of the electricity supply, which increases the system LCOE due to potential overproduction and the resulting loss in electricity production. If the wind turbines or power plants have to curtail their production, then they must spread their fixed costs over a smaller amount of electricity sold, which increases the levelized cost of a unit of electricity. The same effect also occurs with the power plants if they need to be operated flexibly to balance the intermittent wind supply. This can again be seen fromThe "Renewables with oil shale and pyrolysis gas" and "Renewables with nuclear" scenarios have a larger share of wind power (roughly 50% of the generated electricity would be from wind turbines), and consequentially the "Renewables with nuclear" scenario has a higher cost compared to the "Nuclear" scenario. The cost in the "Renewables with oil shale and pyrolysis gas" scenario is also lower than the "Oil shale with CCS" scenario. When wind power accounts for more than about 50% of production, then the amount of potential overproduction starts to increase exponentially and Estonia would be a net importer of electricity (at the capacities used for these simulations).
Stable, dispatchable power could be provided by either oil shale power plants or a nuclear plant, but a nuclear plant is expected to cost less. To compare the two strategies, we included a scenario where all remaining oil shale plants are replaced by 530 MW of nuclear power ("Renewables with nuclear"). Nuclear has a higher capacity factor (about 93%) , so this would provide the same amount of electricity as the 600 MW of oil shale power plants in the "Renewables with only oil shale" scenario. By comparing these two scenarios we can see that nuclear would likely offer two advantages. First, nuclear power would most likely be cheaper than that from an oil shale power plant with CCS (see. Indeed, the "Renewables with nuclear" scenario does have a lower estimated electricity cost than the comparable oil shale scenario.
Producing electricity from pyrolysis gas helps the oil shale industry to lower its costs because pyrolysis gas is a byproduct and emits less CO 2 per unit of electricity. The oil shale industry already produces electricity from pyrolysis gas, and continuing to do so would reduce the amount of oil shale or nuclear needed and help reduce the overall cost of electricity, as shown by the scenarios that include pyrolysis gas inHowever, there is not enough pyrolysis gas available to completely provide the capacity necessary for stabilizing the grid, so an oil shale or nuclear power plant would still be needed, even when utilizing pyrolysis gas.
A nuclear power plant would also likely have the advantage of requiring somewhat smaller curtailments. This is because nuclear power plants can in general be operated more flexibly than the current oil shale power plants. For instance, nuclear power plants can generally be operated at as low as 20% of full capacity, but combustion boilers, like those in oil shale plants, generally have a lower limit in the range of 30-60%, although specialized modifications to the boiler can reduce this. The effect of the better flexibility of a nuclear plant was seen in the simulations because the estimated curtailments (overproduction) were lower for the "Renewables with nuclear" scenario. Based on our simulations, and the data and assumptions behind them, a nuclear power plant would likely provide better performance than a comparable oil shale power plant.
# Sensitivity analysis
To estimate how much various parameters affected the results of the simulations, we performed a simple sensitivity analysis. We ran simulations for the three main scenarios: "Renewables-100 GWh storage", "Renewables with oil shale and pyrolysis gas", and "Renewables, nuclear, and pyrolysis gas", but this time only one parameter at a time was allowed to vary. The remaining parameters were assigned a single mean value. The resulting variation in the average system cost and net surplus are shown ingive an estimate of how variations in each parameter affect the simulation results.
The results inshow that the cost of offshore wind power is the biggest source of variability in the average system cost for these scenarios. One reason is that in these scenarios wind power makes up a majority of the production capacity, and therefore, changes in the cost of wind power have a proportionally larger effect on the overall cost of electricity in the grid. A second reason is simply that there is large variability in the cost of different wind projects. So, uncertainty around the cost of offshore wind also affects the uncertainty of the simulation results. Because the cost of wind and solar have fallen in recent years, this means that a lower cost might be achieved than the simulation results shown here if that trend continues.
Changes in the construction/capital costs of storage and nuclear also have a meaningful effect on the system LCOE of electricity. This is expected for nuclear since construction costs are the largest component of the LCOE of nuclear. For energy storage, it seems the cost of storage per unit of energy has more of an impact than the cost of the energy conversion equipment (i.e. the cost per unit of power). For the underground pumped hydro technology specifically, this means variations in the cost of constructing the underground reservoirs have a bigger impact than variations in the cost of pumps and turbines used to move water between the lower and upper reservoirs. For oil shale and pyrolysis gas, there does not appear to be a single dominate factor that leads to variation in the system cost of electricity. In terms of the balance between production and demand, the capacity factor of the wind turbines is the most significant factor that can lead to variations in the net surplus of electricity, at least for these scenarios. Again, reasons behind this include both the large capacity of wind turbines in the scenarios and the higher variability in production from intermittent sources like the wind. The size of the biomass capacity in Estonia also would affect the net surplus of electricity, as would variations in electricity demand and the cost of grid investments needed for transitioning to using more wind power.
## Additional aspects
Other aspects of the energy system also need to be considered when deciding which to implement. The most significant other aspects are the environmental impact, safety, and the socioeconomic impact of the different systems. Although these aspects are not the focus of this article, we give a brief overview of some literature and data on these aspects to give a more well-rounded comparison of the technologies. A summary of some key metrics related to these aspects is given infor the three potential technologies that could provide stability to the Estonian grid.
One significant factor is the environmental impact of each scenario. All of these systems would have much lower lifetime carbon dioxide emissions than the current oil shale power plants, which is positive. However, an oil shale power plant with CCS will still have much higher lifecycle emissions than wind turbines or a nuclear power plant. CCS generally only captures about 90% of the carbon dioxide in the flue gas because capturing close to 100% usually increases the cost of CCS significantly. So, even with CCS, an oil shale power plant would emit about 100-200 gCO 2 eq/kWh, compared with a median emission rate of about 12 gCO 2 eq/kWh for a nuclear power plant or wind turbines. Emissions of other pollutants, such as SOx, NOx, and particulate matter, are also a concern with oil shale power plants. For a nuclear plant the spent fuel also needs to be handled, but this is already routinely done at nuclear plants and actually is only a small portion of the overall LCOE.
Safety is another important factor to consider. Fossil fuel sources have the highest death rates, much higher than nuclear, wind, or solar. Even when including estimated deaths from nuclear accidents at Chernobyl and Fukushima, as done in, the death rate is orders of magnitude higher for fossil fuels. The main cause of death from fossil fuels is air pollution. However, with nuclear power plants, the discussion around safety has become quite politicized and the actual risk level may be less important than the public's perception of the risk.
In addition to the safety risks of fossil fuels, job creation is also an important socioeconomic factor to consider. The complete impact of a project on jobs in a region can be difficult to assess because in addition to direct jobs at a specific facility, other jobs in the surrounding community can be benefited, or harmed, by a project. For simplification, here we estimated only the direct jobs that would be created by the facility providing dispatchable electricity. We did not include jobs needed during construction because these would only be created temporarily during the construction phase and would require a more complicated analysis to assess their levelized impact. The estimated number of direct jobs created is shown in. The estimates are based on information in the literature. For the electricity storage system, we also included the jobs created by the additional 1655 MW of wind turbines that would be needed to provide electricity for the storage system. For details on calculations we made, see the "Job creation" subsection. The data show that each of the systems would provide a significant number of jobs. Using energy storage would probably create the most jobs due to the increased amount of renewables that could be installed. Wind turbines create a larger number of jobs per unit of electricity than either fossil fuel or nuclear plants, even when excluding construction and considering only the jobs required for operations and management. However, if switching, the current workers in the oil shale industry might not necessarily be able to get those jobs created by other energy technologies. Retraining or relocation might be needed. Additionally, the oil shale industry is the largest employer in Ida-Virumaa county in Estonia. If energy sector jobs move from Ida-Virumaa to another place in Estonia, then active measures would need to be taken to compensate for the loss and to help develop the local economy there. Again, here we just give a brief overview of these additional aspects that should be considered when deciding which electricity system to use. However, other resources that provide more detailed analysis of these aspects should be considered as well. Our analysis here is by no means comprehensive and other factors related to socioeconomics, the environment, or safety that are not touched on here may also be important to consider.
# Limitations
Although we have put forth significant effort to ensure that our simulations correspond accurately to the situation in Estonia, it is important to recognize some of the potential limitations of this study.
One source of uncertainty is the behavior of the broader European electricity market, of which Estonia is a part. Estonia has electricity connections to Finland and Latvia and electricity is exchanged across the region. Modeling this complex network of interconnected local markets is beyond the scope of this study, and we made the simple assumption that up to 30% of the demand in Estonia could be exported at any given moment. This assumption is consistent with data for the year of 2018, when Estonia was still producing enough electricity to be a net exporter. However, the export potential likely fluctuates over time. If more can be exported, especially when wind turbines are producing a lot of electricity, then Estonia could export more electricity and have fewer curtailments than in the simulation results shown here. The opposite would also occur if there is less demand for electricity abroad.
The offshore wind turbines may also have a higher capacity factor than we used for the simulations here. In these simulations we only had data for onshore turbines because no offshore turbines have so far been built in Estonia, and therefore, we assumed the same production profile for both onshore and offshore. Some sources indicate that the offshore wind turbines could actually have a higher capacity factor, although it appears this is not always the case, which is why we did not assume a higher capacity factor for offshore in our simulations. If a higher capacity factor could be achieved offshore, then a smaller wind turbine capacity would be needed and it would be more likely that a lower cost could be achieved.
Additionally, the estimated costs for CCS and small modular nuclear reactors are based on the assumption that other plants would have been built before with the same technology. However, many of these technologies are still new and if Estonia built one of the first plants using a given technology, then the cost would likely be higher than the estimates given here.
Many of these energy infrastructure projects also involve large upfront capital costs, and the costs and complexities of such large projects mean there is a risk of going significantly over time and over budget. This has been the case with several recent large nuclear plants in Europe and America, such as the EPR reactors built at Olkiluoto in Finland and Flamanville in France. Similar overruns have also occurred with large CCS projects, such as the Kemper County IGCC plant in the United States. Therefore, it is worth remembering that additional factors, such as poor project management, opposition from the public, or regulatory uncertainty, could drastically increase the cost for any of the proposed energy projects discussed here.
It is also worth noting that various strategies and technologies have been proposed for improving the flexibility of electricity systems with high levels of wind and solar, and these could in turn reduce the System LCOE. Some ideas focus on using other infrastructure in the economy for additional energy storage, such as using the batteries of electric vehicles to help balance the grid. Other strategies would improve interconnectivity. This could be done by building new transmission lines to allow better balancing of supply and demand between regions, or increasing the performance of existing lines. One such example for increasing the performance of existing lines is dynamic thermal rating. The capacity of transmission lines is limited to ensure that the temperature of the lines does not exceed their rated value. Currently, a fixed, static thermal rating is generally used, but by dynamically adjusting the thermal rating based on actual environmental conditions, the capacity of the lines could be increased during some periods. This could reduce the need for and cost of grid improvements. Another often mentioned strategy is demand response, which involves enabling consumers to change their electricity usage based on the current state of their electricity market. For example, a consumer's air conditioner or water heater could be run when wind and solar are producing and the price is cheaper. Many of these strategies and technologies are still being developed and the economic feasibility of them is generally still unsure, and therefore, we did not include any of them in our simulations.
More generally, all estimates have uncertainty associated with them. We have quantified this uncertainty to the best of our ability by using a Monte Carlo simulation method. Using this method, parameter values are selected from the likely range of values, and repeating the simulation with a range of parameters gives us an estimate of how much the results might vary due to the uncertainty in the underlying model parameters. That said, the situation could change. For instance, the demand in Estonia might not follow historical trends. Also, the cost of various technologies will likely change over time. Furthermore, in this study, we relied often on literature estimates for cost and other parameters. Better accuracy could be achieved by performing more detailed engineering designs specific to Estonia and making estimates using those.
On the most basic level, there are simply a variety of perspectives and approaches that can be used to evaluate an electricity system. Here we evaluated the grid at a system-wide level and focused on how demand, production, storage, and the balance between them vary in time.
There are, however, other perspectives that could and should be considered as well. For instance, one could also look at these different scenarios in terms of grid operation and control because the behavior of some configurations may be more difficult to predict and manage. In addition, safety and environmental impacts should be considered. For instance, it has been shown that fossil fuels have much higher external costs (i.e. health and environmental impacts) than other energy production methods . In this study, we also did not look at social aspects, such as whether or not workers losing their jobs if oil shale plants closed would be able to take advantage of the jobs created by newer energy technologies. Although the results in this study are important to consider before choosing which strategy to pursue to reduce emissions, other perspectives and approaches should also be taken into account.
# Conclusions
Our simulations showed that to achieve the lowest possible LCOE Estonia should complement wind turbines with some sort of dispatchable power plant to provide stability to its electricity system. Using energy storage to provide stability is not a viable option in the foreseeable future because it is significantly more expensive. Our calculations for underground pumped hydro storage, which is one of the cheapest storage technologies, showed that the LCOE would be about 40-50% higher than a comparable system with a power plant. Based on literature estimates, other storage technologies, such as hydrogen storage or batteries, would cost even more. Energy storage might be useful on a smaller scale to provide important ancillary services to the grid, but at the scale required to stabilize an entire grid of wind turbines and solar panels, it is currently too expensive.
Stable dispatchable power could be provided by using oil shale, pyrolysis gas, or nuclear power plants, but using oil shale with CCS is more expensive than the other options. A small modular nuclear power plant (Generation III+) is expected to have a lower LCOE than an oil shale plant with CCS. Utilizing pyrolysis gas would likely have the lowest cost, but pyrolysis gas is a byproduct and is not available in sufficient quantities to provide the approximately 600 MW e of dispatchable power needed to stabilize the grid. Therefore, an oil shale or nuclear power plant would still be needed. Simulations also indicated that using nuclear power could reduce the amount of electricity overproduction. This is mainly due to the fact that nuclear power plants can be operated more flexibly than oil shale power plants, unless specific modifications are made to the oil shale boilers to improve their flexibility.
Additionally, although CCS significantly reduces CO 2 emissions, an oil shale power plant with CCS would still have probably an order of magnitude higher greenhouse gas emissions over its lifetime than a nuclear plant. Statistically, oil shale power plants also cause significantly more deaths than nuclear power plants, mainly due to air pollution.
Note also that all of the potential low-emission scenarios are expected to have a higher LCOE than the current wholesale price of electricity in the Nordpool market. This indicates that the market may need to be adjusted to compensate power plants or energy storage for the additional services they provide to the grid, such as enabling flexible production and ensuring a reserve capacity. Payment for such services could give these energy projects an additional stable revenue stream, which would be important for securing the large investment capital needed for power plants or large-scale energy storage projects.
In summary, transitioning to low-emission electricity production will most likely increase the cost of electricity in Estonia. However by balancing wind turbines with stable production from a smaller number of thermal power plants the country can achieve the transition at a lower cost.
# Author contributions
Conceptualization: Zachariah Steven Baird, Dmitri Neshumayev, Oliver Järvik. |
Lipomas of the central nervous system in the newborns – a report of eight cases
Background:Central nervous system lipomas are rare tumours. In most of the cases they are located in corpus callosum of the brain. The ultrasonographic image of lipomas tends to be quite characteristic. Final diagnosis is however done on a basis of brain resonance. The purpose of this work is to present proceeding in case of central nervous system lipomas with particular attention to diagnostic imaging. This work is based on own research.Case Report:There are eight patients with central nervous system lipomas described in this work. The ultrasonographic imaging performed upon patients' birth revealed features of agenesis of corpus callosum with presence of hyperechoic structure in the area of median line within corpus callosum. This image correlated with Nuclear Magnetic Resonance examination results. Our research confirms that patients with central nervous system lipomas represent rare diagnostic and therapeutic cases. Due to characteristic results of ultrasonographic imaging of the brain, recognition of agenesis of corpus callosum would not cause difficulties. However the presence of hyperechoic structure without vascular flow which may suggest lipomas of corpus callosum would require final verification of the diagnosis and wider assessment of brain with NMR examination. We did not recognize any relation between corpus callosum pathology and neuroinfection of cytomegalovirus etiology. In all of the eight research cases there were malformations diagnostics conducted. There were genetic irregularities in case of two of the neonates only. Until today, all of the patients remain under neurological care. Their psychomotor development is regularly controlled.Conclusions:Taking into consideration that numerous malformations occur altogether with brain lipomas, it is recommended to conduct appropriate diagnostics, to inform parents on an essence of diagnosis and on necessity of observing child's psychomotor development. Obviously, it is crucial to secure a patient with paediatric and neurological care.
## C a s e r e p o r t
spinal, cervical cord, respectively, considering the frequency of incidence. CNS lipomas are often found in more than one location [fig_ref] Table 1: Defects coexisting with cerebral lipoma, depending on the location [/fig_ref].
The exact etiology of lipomas remains uncertain. One of the theories point out the role of genetically determined disturbances of mechanisms controlling tissue differentiation. The early stage of neural tube differentiation is assumed to be the period of CNS lipoma formation [bib_ref] Dorsal mesencephalic lipoma with inferior collicular hypoplasia: a case report and review..., Catala [/bib_ref].
Hence, a lipoma is often accompanied by other CNS pathologies [fig_ref] Table 1: Defects coexisting with cerebral lipoma, depending on the location [/fig_ref]. Also a congenital cytomegalovirus infection is considered to be a potential cause of agenesis (ACC) and lipomas of the corpus callosum [bib_ref] Congenital CMV with callosal lipoma and agenesis, Mehta [/bib_ref]. Diagnosing brain lipomas is nowadays possible in the prenatal period. However, many a time the diagnosis is made after a postnatal ultrasonography performed for other indications. Verification is based on nuclear magnetic resonance imaging (MRI), and in selected cases computed tomography (CT) may be needed. A confirmed presence of CNS lipoma is an indication for further diagnostics considering possible coexistence of other congenital defects.
Neurological symptoms that affect more than 50% of patients with CNS lipomas may appear at different age, but rarely in the neonatal period. The presence of symptoms and their nature depend on the tumor size and location. The most common symptoms of CNS lipomas include seizures, impaired psycho-motor development, hormonal disturbances secondary to hypopituitarism and vision disturbances.
Surgery is undertaken very rarely and only in symptomatic cases. The decision on lipoma resection should take into account tumor location and its position relative to the surrounding tissue. Because of the nature of tumor growth, the aim of surgery is to decompress rather than to resect the tumor completely. After surgery, histological verification of every removed lipoma is required. Although the risk of malignancy is very low, it should always be taken into account [bib_ref] Ahmet Sari A: Sylvian fissure lipoma associated with cortical dysplasia and abnormal..., Ahmeto [/bib_ref]. Widespread access to ultrasonography and an increasingly frequent use of USG in pregnancy and neonatal period (from various indications), enhanced the likelihood of an early CNS lipoma detection, even before the clinical manifestations, and enabled to arrange an appropriate specialist care for the baby. The aim of the study was to present recommendations for management (based on own experience) in case of a suspected CNS lipoma in a newborn, with particular emphasis on diagnostic imaging.
## Cases report
In the years 2007-2010 in the Neonatology, Neonatal Pathology and Intensive Care Clinic of the Children's Memorial Health Institute in Warsaw, 8 patients diagnosed with lipoma of the central nervous system were hospitalized (six patients with brain lipoma and two with extramedullary lipoma). Detailed characteristics of patients were presented in [fig_ref] Table 2: The characteristics of patients with nervous system lipomas [/fig_ref]. Patients with brain lipoma were referred to the Clinic due to abnormalities in neonatal transfontanelle ultrasound imaging requiring further diagnostics. Only two cases of brain anomalies were detected with ultrasonography in the prenatal period, however relatively late (week 31 and 36 of pregnancy). In the postnatal ultrasound, the features of corpus callosum agenesis with coexistent midline hyperechogenic structure near corpus callosum prevailed [fig_ref] Figure 1: Lipoma of corpus callosum and choroid plexuses of the lateral ventricles -USG [/fig_ref]. Only in one case, the lesion was isolated (patient 5), whereas in other cases, similar hyperechogenic structures were reported also in lateral ventricles, in choroid plexuses (patient 1,2,3,4), in cingulate gyrus (patient 2) or in the frontal area (patient 4) [fig_ref] Figure 1: Lipoma of corpus callosum and choroid plexuses of the lateral ventricles -USG [/fig_ref]. Ultrasound results correlated with images obtained in MRI. In case of patient 4, due to an ambiguous picture on MRI scans, suggesting a potential vascular anomaly, an additional angio-CT was performed, which excluded vascular malformation and confirmed corpus callosum agenesis with the presence of lipoma of the corpus callosum and additional lipomas in the right choroid plexus, roof of the third ventricle, and the medial right frontal lobe [fig_ref] Figure 4: Lipomas of corpus callosum, right choroid plexus and 3 rd ventricle ceiling... [/fig_ref]. Before angio-CT results, because of diagnostic doubt, the patient was consulted by an oncologist and tumor markers were determined (negative results).
## Lipoma of corpus callosum
Agenesis of corpus callosum, hypoplasia of the mamillary bodies. Thalamic adhesion, agenesis of septum lucidum [bib_ref] Hypothalamic lipoma adjacent to mamillary bodies, Kurt [/bib_ref] [bib_ref] Corpus callosum lipoma and complex partial seizures, Toshiki [/bib_ref] Lipoma of cerebral and cerebellar hemispheres Hypoplasia, dysplasia, schizencephaly with malformed blood vessels within dysplastic areas [bib_ref] Ahmet Sari A: Sylvian fissure lipoma associated with cortical dysplasia and abnormal..., Ahmeto [/bib_ref] Dorsal brain stem lipoma Not coexistent with corpus callosum lipomas [bib_ref] Dorsal mesencephalic lipoma with inferior collicular hypoplasia: a case report and review..., Catala [/bib_ref] Lipoma of choroid plexuses (very rarely isolated) Corpus callosum lipoma [bib_ref] Isolated choroid plexus lipomas, Sener [/bib_ref] Lipoma of tectal plate and cerebellum coexisting with hypoplasia of these structures Tectocerebellar dysraphia [bib_ref] Occipital extra-and intracranial lipoencephalocoele associated with tectocelebellar dysraphia, Dehdashti [/bib_ref] Interhemispheric lipoma Upper lift cleft and polyps of facial skin -Pai syndrome [bib_ref] Pai syndrome: a report of a case and review of the literature, Al-Mazrou [/bib_ref] Posterior fossa lipomas Dandy-Walker syndrome and Arnold Chiari malformaltion [bib_ref] Occipital extra-and intracranial lipoencephalocoele associated with tectocelebellar dysraphia, Dehdashti [/bib_ref] [bib_ref] An infant with an intradural lipoma of the cervical spine extending into..., Kai [/bib_ref] cardiac arrhythmia were observed. Brain ultrasonography also revealed features of hydrocephalus. In addition, cerebellar vermis hypoplasia was suspected and hyperechogenic circumventricular areas were detected. Basing on ultrasonography, CNS lipomas were not suspected. MR of the brain and spinal cord was performed that revealed asymmetric hydrocephalus, the presence of developed, but lipomatosis (ECCL), so called, Haberland syndrome. Because of the location of lipomas, the patient will likely manifest a neurological syndrome and will require neurological treatment, possibly even a neurosurgical intervention.
Two other patients were admitted to the clinic because of sacral region abnormality (soft, palpable tumor) detected in the physical examination. In these patients, it was decided to perform the MRI of the spinal cord without initial ultrasound diagnostics. Patient 7 had intrathecal lipoma, adjacent to the spinal cord taper and developmental defect of the sacrum [fig_ref] Figure 6: Spinal canal lipoma at L1-L5 level, adjacent directly to the spinal cord... [/fig_ref]. In patient 8, MRI revealed extrathecal lipoma protruding into the spinal canal, with cord tethering inside the lipoma and a linear layer of fat tissue alongside the posterior cord.
# Discussion
According to our knowledge and observation, central nervous system lipomas constitute a very rare diagnostic and therapeutic problem. The ratio of patients with CNS lipoma hospitalized in our clinic was 0.21, which undoubtedly is a high result as compared to the incidence in general population, which is of course associated with specialization of our clinic. Among the presented patients, male gender predominated (75%), which correlates with data reported by others [bib_ref] Hypothalamic lipoma adjacent to mamillary bodies, Kurt [/bib_ref].
Among CNS lipomas the most frequent are the lipomas of the corpus callosum, coexisting with corpus callosum hypogenesis/agenesis. It was also confirmed in our observation (62.5% -5 patients with corpus callosum agenesis and lipomas). In our study, relatively frequently (in 4 out of 6 patients) lipomas of corpus callosum coexisted with lipomas in choroid plexuses. Similarly, Sener et al. reported that isolated choroid plexus lipomas occur extremely rarely [bib_ref] Isolated choroid plexus lipomas, Sener [/bib_ref]. In case of interhemispheric location with coexisting cleft of the upper lip and polyps of the facial skin, the patient is diagnosed with Pai syndrome [bib_ref] Pai syndrome: a report of a case and review of the literature, Al-Mazrou [/bib_ref]. However, in our patient, no additional abnormalities were observed.
Corpus callosum agenesis was not difficult to diagnose due to its characteristic features in transfontanelle USG.
However, the presence of an additional hyperechogenic lesion without vascular flows suggests corpus callosum lipoma. Small lipomas, located outside the corpus callosum, particularly in the posterior cranial fossa, are much more difficult to detect in transfontanelle ultrasonography. MRI is recommended for a final verification of the diagnosis and more comprehensive assessment of the brain regarding additional congenital malformations. Except for one case (patient 5), in which only CT was possible to perform, each of patients had brain MRI during spontaneous sleep. However, in one case (patient 4), because of the detected vascular structures of the lesion found in the area of hypoplastic corpus callosum, CT angiography was performed additionally, to exclude vascular malformations. Based on the analysis of 8 presented cases of neonates with CNS lipomas, it should be noted that MRI allows for obtaining a much more accurate image of midline structures, with a precise assessment of the corpus callosum. MRI in newborns can be performed in spontaneous sleep with no need for general anesthesia and without exposing the child to the X-rays. Thus, the use of CT in suspected CNS lipomas is limited only to situations in which the MRI is equivocal. Similar conclusions also appear in reports by other authors.
Since TORCH infection may lead to CNS malformations, serological testing, and in some cases, additional PCR were performed in our patients (except for one patient with spinal cord lipoma) to exclude congenital cytomegalovirus and Toxoplasma gondii infection. There is a theory indicating a relationship between corpus callosum pathology and CMV neuroinfection [bib_ref] Congenital CMV with callosal lipoma and agenesis, Mehta [/bib_ref]. However, no such link was found among our patients, which contradicts the above hypothesis. An additional argument may be that over many years, none of the patients hospitalized in our clinic due to cytomegalovirus neuroinfection had lesions suggestive of CNS lipoma.
In two of the presented patients, CNS abnormalities had been previously suspected during the prenatal period, but not before week 30 of pregnancy. This seems to confirm the observations of other authors that CNS tumors in prenatal ultrasound are diagnosed relatively late because of the period of their formation/grow. However, this contradicts the theory of brain lipoma formation in early embryogenesis. Due to the detected brain developmental pathology, it is recommended to perform diagnostics towards other congenital defects, as well as to provide continuous neurological care and observation of psychomotor development. All presented patients were diagnosed for other congenital defects. Only in two newborns (patients 5 and 6), genetic abnormalities were found: trisomy of 8 chromosome in one case and Haberland syndrome in the other.
During hospitalization, four out of six patients with cerebral lipomas (66%) manifested neurological abnormalities in the neonatal period. However, these were predominantly mild disturbances in the distribution of muscle tension. None of the patients had seizures. Although, in the patient with encephalocraniocutaneous lipomatosis (ECCL) due to the lipoma location, symptoms of compression of n. V and spinal cord can be expected in the future. All patients remain under neurological care and their psychomotor development is monitored in their local clinics. Apart from the delay in speech development in the patient with encephalocraniocutaneous lipomatosis, a normal psychomotor development is observed in all other children. As for now, the follow-up period is relatively short and amounts to 1-4 years, depending on the case.
It should be noted that due to the location of lipomas, patient 6 will probably present with neurological symptoms and require treatment or even neurosurgical intervention in the future.
# Conclusions
The diagnosis of cerebral lipoma can be made with high probability basing on transfontanelle ultrasonography (especially in the case of lipomas of the corpus callosum). However, final diagnosis should be confirmed in MRI. In case of spinal canal lipomas, MRI is the method of choice. Brain CT should be reserved only for cases of doubt, due to high radiation dose and less precision in the assessment of the corpus callosum.
Considering the fact of coexistence of different congenital defects along with brain lipomas, it is recommended to perform appropriate diagnostic procedures, provide parents with comprehensive information on the diagnosis and on the need for continuous monitoring of child's development and adequate pediatric and neurological care.
Neurosurgical intervention is needed in exceptional cases, in patients manifesting compression symptoms.
[fig] Figure 1: Lipoma of corpus callosum and choroid plexuses of the lateral ventricles -USG (patient 2). [/fig]
[fig] Figure 2: Corpus callosum lipoma -USG (patient 1). [/fig]
[fig] Figure 3: Corpus callosum lipoma and intraventricular lipomas -MRI scan (patient 2). [/fig]
[fig] Figure 4: Lipomas of corpus callosum, right choroid plexus and 3 rd ventricle ceiling -USG and CT (patient 4). [/fig]
[fig] Figure 5: Multiple intradural, extramedullary lipomas in the dorsal part of the spinal canal (patient 6). [/fig]
[fig] Figure 6: Spinal canal lipoma at L1-L5 level, adjacent directly to the spinal cord taper (patient 7). [/fig]
[table] Table 1: Defects coexisting with cerebral lipoma, depending on the location. , because of coexisting defects and/ or dysmorphic features were referred to genetic testing. In the physical examination of patient 5, characteristic dysmorphic features (microphthalmia, narrow palpebral fissures, recessed mandible) and the deep hand-and footprints were noted. Further diagnostics showed no other developmental abnormalities. In neurological examination, the patient presented: hypertonia of the lower limbs, increased deep tendon reflexes, incomplete Moro reflex, excessive Galant reflex, a tendency for opisthotonus in Landau reflex. Trisomy of chromosome 8 was detected prenatally. Patient 6 was admitted to hospital due to multiple developmental malformations. Facial dysmorphia with hypertelorism, a number of lumps in the skin of the eyelids, hydrocephalic cranial bones, widely spaced nipples, hypospadias and [/table]
[table] Table 2: The characteristics of patients with nervous system lipomas. [/table]
|
Prenatal adverse environment is associated with epigenetic age deceleration at birth and hypomethylation at the hypoxia-responsive EP300 gene
Background: Obstetric complications have long been retrospectively associated with a wide range of short-and long-term health consequences, including neurodevelopmental alterations such as those observed in schizophrenia and other psychiatric disorders. However, prospective studies assessing fetal well-being during pregnancy tend to focus on perinatal complications as the final outcome of interest, while there is a scarcity of postnatal follow-up studies. In this study, the cerebroplacental ratio (CPR), a hemodynamic parameter reflecting fetal adaptation to hypoxic conditions, was analyzed in a sample of monozygotic monochorionic twins (60 subjects), part of them with prenatal complications, with regard to (i) epigenetic age acceleration, and (ii) DNA methylation at genes included in the polygenic risk score (PRS) for schizophrenia, and highly expressed in placental tissue. Results: Decreased CPR measured during the third trimester was associated with epigenetic age deceleration (β = 0.21, t = 3.362, p = 0.002). Exploration of DNA methylation at placentally expressed genes of the PRS for schizophrenia revealed methylation at cg06793497 (EP300 gene) to be associated with CPR (β = 0.021, t = 4.385; p = 0.00008, FDR-adjusted p = 0.11). This association was reinforced by means of an intrapair analysis in monozygotic twins discordant for prenatal suffering (β = 0.027, t = 3.924, p = 0.001). Conclusions: Prenatal adverse environment during the third trimester of pregnancy is associated with both (i) developmental immaturity in terms of epigenetic age, and (ii) decreased CpG-specific methylation in a gene involved in hypoxia response and schizophrenia genetic liability.
robustly show high correlation across studies, the difference between both variables allows the estimation of the so-called age acceleration (i.e., when epigenetic age is higher than chronological age).
On the one hand, epigenetic age acceleration in adult subjects has been associated with cumulative lifetime stress, lifestyle, and all-cause mortality, among others, suggesting its utility as a better predictor for life expectancy than chronological age itself [bib_ref] Lifetime stress accelerates epigenetic aging in an urban, African American cohort: Relevance..., Zannas [/bib_ref] [bib_ref] DNA methylation-based measures of biological age : meta-analysis predicting time to death, Chen [/bib_ref] [bib_ref] Back to the future: epigenetic clock plasticity towards healthy aging, Declerck [/bib_ref]. On the other hand, epigenetic GA deceleration (i.e., when chronological age is higher than epigenetic age), as measured in cord blood, has been described in newborns born to women with low socioeconomic status, Sjögren syndrome, insulin-treated gestational diabetes mellitus, and experiencing antenatal depressive symptoms [bib_ref] An epigenetic clock for gestational age at birth based on blood methylation..., Knight [/bib_ref] [bib_ref] Associations between maternal risk factors of adverse pregnancy and birth outcomes and..., Girchenko [/bib_ref] [bib_ref] The epigenetic clock at birth: associations with maternal antenatal depression and child..., Suarez [/bib_ref]. Such findings suggest that newborns exposed to prenatal stressors are born in an immature state independently of their chronological GA. In this regard, boys-but not girls-who exhibited lower epigenetic GA at birth exhibited more internalizing problems, such as anxious-depressive symptoms or somatic complaints, at follow-up (mean age 3.7 years), suggesting they are born with a developmental disadvantage [bib_ref] The epigenetic clock at birth: associations with maternal antenatal depression and child..., Suarez [/bib_ref].
Nevertheless, there is a dearth of studies examining the putative relationship between ultrasound parameters acquired during pregnancy and epigenetic GA acceleration. In this regard, the cerebroplacental ratio (CPR) has been reported to be associated with adverse perinatal outcomes not only in growth-restricted fetuses, but also in low-risk population [bib_ref] The importance of the cerebroplacental ratio in the evaluation of fetal well-being..., Devore [/bib_ref] [bib_ref] Changes in fetal Doppler indices as a marker of failure to reach..., Morales-Roselló [/bib_ref]. Briefly, CPR is calculated by dividing the middle cerebral artery (MCA) pulsatility index (PI) by the umbilical artery (UA) PI [bib_ref] The Doppler cerebroplacental ratio and perinatal outcome in intrauterine growth restriction, Bahado-Singh [/bib_ref]. The PI is a parameter reflecting vascular impedance or resistance, i.e., decreased blood flow. Specifically, fetal brain blood supply is known to increase in front of hypoxic stimuli thus decreasing PI in the MCA [bib_ref] Blood flow to fetal organs as a function of arterial oxygen content, Peeters [/bib_ref] ; while placental insufficiency decreases umbilical blood flow hence increasing UA PI, and has been associated with both short-and long-term detrimental outcomes, including increased cardiovascular risk and deficits in cognition [bib_ref] Fetal umbilical artery Doppler pulsatility index as a predictor of cardiovascular risk..., Mone [/bib_ref] [bib_ref] Fetal umbilical artery Doppler pulsatility index and childhood neurocognitive outcome at 12..., Mone [/bib_ref]. Consequently, a decreased CPR reflects the combination of both alterations and is an indicator of fetal adaptation to adverse conditions [bib_ref] The importance of the cerebroplacental ratio in the evaluation of fetal well-being..., Devore [/bib_ref].
Obstetric complications (OCs) constitute one of the risk factors more reliably associated with psychopathology, particularly with neurodevelopmental disorders; specifically, the putative association between OCs and schizophrenia has been debated since the 1970s [bib_ref] A critical review of recent adoption, twin, and family studies of schizophrenia:..., Gottesman [/bib_ref] [bib_ref] Obstetric complications and schizophrenia : historical and meta-analytic review, Cannon [/bib_ref] [bib_ref] Early life programming and neurodevelopmental disorders, Bale [/bib_ref]. In this regard, a recent umbrella review evaluating all published meta-analysis regarding risk factors and biomarkers for schizophrenia spectrum disorders revealed a history of OCs to significantly increase the risk for developing the disorder with an odds ratio of 2 [bib_ref] Risk factors and peripheral biomarkers for schizophrenia spectrum disorders: an umbrella review..., Belbasis [/bib_ref]. Furthermore, exposure to severe OCs together with increased genetic vulnerability, as measured with the polygenic risk score (PRS) for schizophrenia, interact to increase the risk to suffer the disorder up to an odds ratio of 8.36 [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref]. In the same study, authors further explored the putative relevance of placental expression of genes included in the PRS; following this approach, they reported (i) an enrichment of PRS genes expressed in placental tissue and (ii) differential expression of PRS genes in placentae from complicated pregnancies (specifically in pre-eclampsia and intrauterine growth restriction). Specifically, the described gene-environment interaction between exposure to OCs and the PRS for schizophrenia was driven by those genes highly expressed in placenta and/or dynamically regulated in complicated pregnancies [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref]. Since CPR is a robust indicator of prenatal stress and a predictor of perinatal and long-term morbidity, DNA methylation analysis of genes included in the placental PRS for schizophrenia could shed light on the epigenetic mechanisms mediating the interaction between OCs and neurodevelopmental disorders.
Monozygotic twins have been instrumental for the elucidation of environmental and genetic risks in the etiology of complex traits and disorders. Actually, the differential role of the prenatal environment in shaping psychopathological proneness was first described thanks to monozygotic twin designs [bib_ref] Second-trimester markers of fetal size in schizophrenia: a study of monozygotic twins, Bracha [/bib_ref] [bib_ref] Prenatal origin of schizophrenia in a subgroup of discordant monozygotic twins, Torrey [/bib_ref] [bib_ref] Congenital dermatoglyphic malformations and psychosis: a twin study, Rosa [/bib_ref] ; these pioneering studies focused on dermatoglyphic measures assessed at birth, which can be used as surrogate measures of altered neurodevelopment during the second trimester of pregnancy [bib_ref] Dermatoglyphic a-b ridge count as a possible marker for developmental disturbance in..., Fañanas [/bib_ref]. Furthermore, monozygotic twin pregnancies and, more specifically, monochorionic twin pregnancies-i.e., those in which both fetuses share the placenta-are at a higher risk of obstetric complications, the more prevalent being twin-to-twin transfusion syndrome (TTTS) and selective intrauterine growth restriction (sIUGR) [bib_ref] Selective intrauterine growth restriction in monochorionic twins: pathophysiology, diagnostic approach and management..., Valsky [/bib_ref] [bib_ref] The hidden mortality of monochorionic twin pregnancies, Sebire [/bib_ref] [bib_ref] Specific complications of monochorionic twin pregnancies: twin-twin transfusion syndrome and twin reversed..., Chalouhi [/bib_ref]. Thus, the thorough and prospective ultrasound assessment of prenatal development through monochorionic twin pregnancies offers a quasi-experimental study design in which the genetic and environmental components of epigenetic variability can be dissected.
The objective of the current study was to investigate whether prenatal adverse environment (i) alters human development in terms of epigenetic age, and if (ii) it can get embedded through epigenetic mechanisms in genes previously identified as risk factors for schizophrenia acting during prenatal stages. We hypothesized that a higher exposure to prenatal adverse environment would be associated with (i) delayed development and (ii) DNA methylation at genes involved in the pathogenesis of schizophrenia. While CPR can have diverse effects on genome-wide DNA methylation, with potential relevance for a multitude of phenotypes, the present study a priori examined how CPR epigenetically regulates risk genes for schizophrenia, previously described to interact with the presence of OCs [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref].
# Results
## Ga estimation using knight's epigenetic clock
After exclusion of two twin pairs (see Methods section), the final sample size was 30 twin pairs. The mean GA at birth of our twin cohort (n = 30 twin pairs) was 35.3 weeks (range = 31.7-37.1) and the mean DNA methylation GA at birth was 35 weeks (range = 31.4-37.7). To validate the epigenetic clock predictor in our sample, DNA methylation-based GA was tested for correlation with chronological GA (r = 0.76, p = 1.68 × 10 −12 ; . The average absolute difference between epigenetic GA and chronological GA-hereinafter referred as ΔGA-was 0.9 weeks (range = 0.03-4.02), i.e., 6.3 days.
In agreement with previous studies, there was a significant negative correlation between ΔGA and chronological GA (r = − 0.47; p < 0.001).
Association between ΔGA and CPR ΔGA was tested for associations with CPR measured during the third trimester (mean = 33.8 weeks, range = 28.3-36.4), a few days before childbirth (median = 6.5 days). CPR was significantly associated with ΔGA (β = 0.21, t = 3.362, p = 0.002) when adjusting for sex, birthweight, diagnostic of either TTTS or sIUGR, surgery time interval (when laser fetoscopy had been applied), and gestational age at ultrasound as covariates. The positive association between CPR and ΔGA remained significant after correction for cell type proportion (β = 0.21, t = 2.616, p = 0.01). [fig_ref] Figure 2: Association between epigenetic age acceleration and cerebroplacental ratio measured during the third... [/fig_ref] shows the positive association between third trimester CPR and ΔGA.
Epigenetic exploration of placental PRS for schizophrenia with regard to CPR Following the approach developed by Ursini and collaborators (2018), association between CPR and DNA methylation was tested in all CpG sites included in the DNA methylation array located within genes of the PRS for schizophrenia expressed in placental tissue (placental PRS) [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref]. There were 1400 CpG sites annotated to placental PRS genes out of 866,091 CpG sites included in the array. After FDR correction for multiple testing, methylation at one single CpG site, cg06793497, was significantly associated with CPR (β = 0.021, p = 0.00008, t = 4.385; q FDR adjusted = 0.11; , such that increased Correlation between chronological GA and epigenetic GA. Chronological GA was calculated using first-trimester crown-rump length measurement of the larger twin, and epigenetic age was calculated based on DNA methylation-based Knight's clock. Both GA estimations were significantly correlated (r = 0.76; p = 1.68 × 10 −12 ) cg06793497 methylation was associated with increased CPR. The top 10 CpG sites yielded by this approach are summarized in [fig_ref] Table 1: Top 10 CpG sites of the PRS methylomic exploration in association with... [/fig_ref] (all q values > 0.75).
To further explore the association between cg06793497 methylation and CPR, it was analyzed in a monozygotic twin intrapair design. The intrapair twin design further allows controlling for chronological GA, sex, and timing of the Doppler ultrasound, since these variables are shared by co-twins of a pair. Four observations were removed from the analysis due to missingness for any of the variables in one of the co-twins of a pair. Thus, intrapair differences for these measures were calculated for all twin pairs of the sample were both measures were available for both twins of a pair (n = 27 twin pairs). Intrapair differences in cg06793497 methylation and CPR, measured during the third trimester, were significantly correlated (r = 0.64, p < 0.001; . The association between both variables remained significant after adjusting for cell type count intrapair differences (β = 0.027, t = 3.924, p = 0.001). Intrapair exploration of the top 10 CpG sites [fig_ref] Table 1: Top 10 CpG sites of the PRS methylomic exploration in association with... [/fig_ref] revealed significant associations between CPR and DNA methylation at CpG probes cg00262246 (β = 0.012, p = 0.029), cg01024069 (β = − 0.01, p = 0.033), and cg12955069 (β = − 0.021, p = 0.026).
## Dna methylation exploration of ep300 gene
To further explore the putative relevance of DNA methylation at other CpG sites located within the EP300 gene and its surrounding regions, DNA methylation at 27 CpG sites included in the array and annotated to this region was also explored with regard to CPR (see [fig_ref] Table 2: List of CpG sites included in the array located in the EP300... [/fig_ref]. All analyses were adjusted for cell sex, birthweight, gestational age at ultrasound, and cell type count. In addition to cg06793497, two additional CpG sites-cg12968540 and cg19011939-were significantly associated with CPR (p < 0.05); moreover, methylation at four additional CpG sites-cg04452260, cg24349919, cg11931284 and cg25888227-showed trend associations with CPR (p < 0.01). The intrapair approach was then applied for these newly identified six CpG sites revealing cg11931284 (β = 0.028, t = 2.985, p = 0.008) and cg19011939 (β = − 0.021, t = − 2.343, p = 0.03) to be significantly associated with CPR, when adjusting for cell types intrapair differences. Association between epigenetic age acceleration and cerebroplacental ratio measured during the third trimester. Epigenetic age delta (ΔGA) corresponds to estimated epigenetic age minus chronological age. Thus, ΔGA-positive values reflect epigenetic age acceleration while negative values point out the presence of epigenetic age deceleration. The cerebroplacental ratio (CPR) is calculated as the ratio between the MCAPI and UAPI. Both variables were significantly correlated when adjusting for sex, chronological gestational age, birth weight, and gestational age at ultrasound
# Discussion
To the best of our knowledge, this is the first study analyzing the epigenetic age in association with adverse prenatal environment as measured by a hemodynamic ultrasound parameter. Firstly, we describe the significant association between CPR measured during the third trimester of pregnancy with epigenetic age acceleration. Specifically, subjects exhibiting decreased CPR-exposed to prenatal adverse conditions-were born with decelerated epigenetic age, i.e., prenatally stressed subjects were born immature adjusting for their gestational age at birth. Additionally, methylomic exploration of schizophrenia PRS genes known to be expressed in placenta revealed the association between CPR and EP300 gene CpG-specific methylation, at the cg06793497 probe, in our monochorionic twin sample. Schizophrenia PRS methylation exploration with regard to cerebroplacental ratio measured during the third trimester. a Methylation at cg06793497 (EP300 gene) was significantly associated with CPR in the whole sample (n = 54 twin subjects). b Intrapair methylation difference at cg06793497 was significantly associated with intrapair CPR difference (n = 26 twin pairs) Developmental deficits and developmental delays have been previously described in children who would later develop schizophrenia [bib_ref] Static and dynamic cognitive deficits in childhood preceding adult schizophrenia: a 30-year..., Reichenberg [/bib_ref] ; although such prodromal symptoms were in accordance with the neurodevelopmental hypothesis for schizophrenia, biological mechanisms mediating these effects remain largely unknown. Epigenetic immaturity in response to prenatal stress could be contributing to this developmental delay. Interestingly, epigenetic age deceleration has been previously described in association with maternal pathologies during pregnancy, such as maternal depression or Sjögren's syndrome, suggesting it can be a robust biomarker of prenatal suffering [bib_ref] Associations between maternal risk factors of adverse pregnancy and birth outcomes and..., Girchenko [/bib_ref] [bib_ref] The epigenetic clock at birth: associations with maternal antenatal depression and child..., Suarez [/bib_ref]. It is worth noting that CPR was measured a few days prior to childbirth; thus, it can be used as a surrogate marker of prenatal adaptation to adverse conditions experienced at the end of the pregnancy, i.e., as a marker of perinatal risk.
Integration of the schizophrenia PRSwith obstetric and placental information [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref] , allowed the identification of E1A binding protein p300 (EP300) gene CpG-specific methylation as a putative marker of exposure to prenatal stress. Interestingly, the EP300 gene encodes a histone acetyltransferase (HAT) involved in several cell pathways such as cell proliferation and differentiation. Mutations at EP300 gene have been described to cause Rubinstein-Taybi syndrome, a rare autosomal dominant neurodevelopmental disorder characterized by intellectual disability, psychomotor and language delay, and facial dysmorphisms [bib_ref] Rubinstein-Taybi 2 associated to novel EP300 mutations: deepening the clinical and genetic..., López [/bib_ref]. Likewise, these symptoms, including developmental delay, learning problems, and cleft palate, characterize the 22q11.2 deletion syndrome, a well-defined congenital condition caused by the deletion of the 22q11.2 segment. Notably, this syndrome is associated with a higher risk to develop schizophrenia, among other psychiatric conditions [bib_ref] Update on the 22q11.2 deletion syndrome and its relevance to schizophrenia, Van [/bib_ref] ; interestingly, EP300 gene is located on chromosome 22 at position 22q13.2.
Further exploration of differential DNA methylation in and around the EP300 gene revealed cg19011939 to be differentially methylated in association with prenatal adversity. While higher exposure to a prenatal adverse environment, as reflected by lower CPR during the third trimester, is associated with decreased methylation at cg06793497 in the hypoxia-responsive EP300 gene, there appears to be increased methylation at cg19011039 at EP300-AS1 gene. Thus, we speculate that higher exposure to prenatal stress might be associated with reciprocal patterns of EP300 and EP300-AS1 epigenetic regulation that could act synergistically, a hypothesis that may be explored in future studies [bib_ref] Antisense RNAs and epigenetic regulation, Cui [/bib_ref].
Remarkably, EP300 has been identified as a co-activator of the hypoxia-inducible factor 1 alpha (HIF1A). In this regard, hypoxic conditions stimulate EP300 expression, which has a neuroprotective role [bib_ref] p300 expression is induced by oxygen deficiency and protects neuron cells from..., Tan [/bib_ref]. Accordingly, genetic variability at EP300 gene has been associated with human adaptations to high altitude regions, e.g., the Tibet [bib_ref] Genetic Variations in Tibetan Populations and High-Altitude Adaptation at the Himalayas, Peng [/bib_ref]. Likewise, pre-and peri-natal hypoxia have been associated with schizophrenia spectrum disorders, particularly by decreasing hippocampal volume [bib_ref] Contributions of genetic risk and fetal hypoxia to hippocampal volume in patients..., Van Erp [/bib_ref] [bib_ref] Fetal hypoxia and structural brain abnormalities in schizophrenic patients, their siblings, and..., Cannon [/bib_ref] ; complementarily, a decreased or impaired response to hypoxia via neurotrophic factors has also been implicated in the etiology of schizophrenia [bib_ref] Decreased neurotrophic response to birth hypoxia in the etiology of schizophrenia, Cannon [/bib_ref]. Furthermore, DNA methylation at the IGF2BP1 gene, also involved in prenatal development [bib_ref] Fetal DNA methylation associates with early spontaneous preterm birth and gestational age, Parets [/bib_ref] , has been associated with both adult working memory and birthweight [bib_ref] Birth weight, working memory and epigenetic signatures in IGF2 and related genes:..., Córdova-Palomera [/bib_ref] ; further highlighting the advantage of twin study designs to identify environmentally-driven epigenetic consequences of prenatal stress. Overall, these findings point to the existence of a GxE interaction between genetic vulnerability and exposure to prenatal hypoxia, as already highlighted by Ursini and DNA methylation at CpG sites highlighted in italics was significantly associated with CPR measured during the third trimester in an intrapair approach collaborators [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref]. In this framework, EP300 methylation could be one of the mediators of such interaction. A number of limitations of the present study should be noted. First, the moderate sample size (n = 60 subjects, 30 twin pairs) limits the statistical power of the analysis; however, smaller sample sizes (n = 22 MZ twin pairs) have been described to be sufficient to identify methylation differences of 6% with > 80% power [bib_ref] Disease-associated epigenetic changes in monozygotic twins discordant for schizophrenia and bipolar disorder, Dempster [/bib_ref]. Moreover, a lenient significance threshold after correction multiple testing was used; however, previous epigenetic studies have described FDR values between 5 and 20% as markers of medium-confidence sites [bib_ref] Epigenetic vestiges of early developmental adversity: childhood stress exposure and DNA methylation..., Essex [/bib_ref]. Another limitation regards the moderately small reported effect sizes (around 2%) questioning the biological relevance of our findings [bib_ref] DNA methylation: conducting the orchestra from exposure to phenotype?, Leenen [/bib_ref] ; however, these findings are in agreement with a larger body of evidence regarding cord blood methylation after exposure to a number of prenatal stressors. Such small DNA methylation changes may act in conjunction with a myriad of other epigenetic signatures and biological processes in order to maintain homeostasis in the face of threats. Additionally, although epigenomic information was available from a methylomic array including more than 800,000 CpG sites distributed throughout the whole human genome, only 1400 CpG sites were analyzed; alternative approaches including the total of CpG sites included in the array would have yield different findings, probably pointing to genes involved in other neurodevelopmental disorders besides schizophrenia. Furthermore, while the set of genes analyzed in the current approach were described to be highly expressed in placental tissue [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref] , placentae were not available for this sample and cord blood was thus analyzed as the proxy tissue of choice with regard to exposure to prenatal adversity. Finally, MZ twin pregnancies are characterized by lower gestational ages at birth than singleton pregnancies; besides, obstetric scales commonly used in psychiatric studies include twin pregnancies as an obstetric complication. Thus, findings derived from the present design might not be generalizable to the general population.
# Conclusions
Further studies are needed to test the time stability of the hereby identified methylation signature. It will be equally relevant to explore neurobehavioral correlates of EP300 methylation during early childhood along with its putative association with neurodevelopmental outcomes, including psychosis liability. Additionally, a longitudinal follow-up is required to test the role of postnatal environment in these phenotypes since both epigenetic age deceleration and CpG-specific differential methylation in association with CPR could return to basal levels after birth. Finally, genetic exploration of these subjects regarding schizophrenia PRS will be instrumental for the study of GxE interactions and genetic liability for an impaired hypoxia response during human development.
# Methods
## Study population
This was a prospective study including fetal pairs from monochorionic diamniotic twin pregnancies attended at Hospital Clínic de Barcelona (Spain) during a 2-year recruitment period. Monochorionic monoamniotic twin pregnancies were excluded from the present study to avoid putative confounding with regard to differential exposure to stress in both types of twin pregnancies. The study protocol was approved by the hospital ethics committee (HCB/2016/0046), and all patients provided written informed consent.
We included 32 monochorionic pregnancies (n = 64 samples). The sample was enriched for two monochorionic-specific severe obstetric complications: twin-to-twin transfusion syndrome (TTTS, n = 8) and selective intrauterine growth restriction (sIUGR, n = 9). All TTTS cases were treated upon detection by means of laser fetoscopy [bib_ref] Fetoscopic laser photocoagulation for twin-twin transfusion syndrome, Sago [/bib_ref].
Maternal age and pre-pregnancy BMI were retrieved from hospital records. Gestational age was dated using first-trimester crown-rump length measurement of the larger twin [bib_ref] Sonar measurement of fetal crown-rump length as means of assessing maturity in..., Robinson [/bib_ref].
## Fetal ultrasound assessment
Ultrasound assessment was performed on a Voluson Expert 8 (General Electrical Medical Systems, Milwaukee, WI, USA) or a Siemens Sonoline Antares (Siemens Medical Systems, Erlangen, Germany) with 8-to 4-MHz or 6-to 4-MHz curved array probes, respectively. All fetuses underwent detailed ultrasound evaluation including fetal anatomy and Doppler measurements such as UAPI, MCAPI and ductus venosus PI. All Doppler evaluations were acquired at a normal fetal heart rate (FHR) in the absence of fetal body or respiratory movements and at an angle of insonation as close to 0°as possible (but always < 15°), and the mechanical and thermal indices were maintained below 1. CPR was calculated as the ratio between MCAPI and UAPI, according to previous studies [bib_ref] The importance of the cerebroplacental ratio in the evaluation of fetal well-being..., Devore [/bib_ref].
## Dna methylation
Umbilical vein cord blood samples were obtained from the clamped umbilical cord immediately after delivery of the fetus. All blood samples were collected in EDTA-treated tubes and processed within 1 h. Plasma was separated by centrifugation at 3000 rpm for 10 min at 4°C, and stored at − 80°C until further use. Genomic DNA was extracted from fetal cord blood using QIAamp DNA Mini Kit (Qiagen). DNA quality and quantity were assessed by NanoDrop One (Thermo Scientific).
Genomic DNA was bisulfite converted using the Zymo EZ-96 DNA Methylation Kit (Zymo Research). Genome-wide DNA methylation levels were assessed over 850,000 CpG sites by means of the Infinium Methy-lationEPIC BeadChip Kit (Illumina Inc., CA, USA) according to the manufacturer's protocol. Pre-processing and normalization were performed using the Bioconductor minfi package [bib_ref] Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium..., Aryee [/bib_ref]. CpG probes containing common SNPs were discarded. All probes mapping to the X and Y chromosomes were also removed. Finally, cross-hybridizing probes as previously identified were excluded from further analysis [bib_ref] Identification of polymorphic and off-target probe binding sites on the Illumina Infinium..., Mccartney [/bib_ref]. All samples (n = 64) were run on the same plate.
Absence of maternal contamination was confirmed after retrieving DNA methylation values at 10 CpG sites previously described to identify sample contamination by maternal blood during sample collection [bib_ref] Maternal blood contamination of collected cord blood can be identified using DNA..., Morin [/bib_ref]. None of the samples assayed exhibited DNA methylation values above the threshold at 5 or more of those CpG sites (see Additional file 1 for specific methylation values). Two samples (from the same twin-pair) were excluded from further analyses due to lack of monozygosity as assessed by 59 SNPs included in the array. One of the samples was removed from analysis due to insufficient DNA concentration, the co-twin sample was also excluded from further analysis.
## Statistical analyses
All statistical analyses were conducted in R version 3.5.0. DNA methylation-based GA prediction was performed using the R code and statistical pipeline developed by Knight, based on the methylation profile of 148 CpG sites [bib_ref] An epigenetic clock for gestational age at birth based on blood methylation..., Knight [/bib_ref] ; this predictor was developed using 15 Illumina DNA methylation datasets (n = 1434 neonates). Following Simpkin et al. recommendations, the Knight clock was preferred for our analysis as it was developed and tested in preterm infants datasets such as our monozygotic twin population, characterized by a mean gestational age at birth of 35.3 weeks [bib_ref] Epigenetic clocks for gestational age: Statistical and study design considerations, Simpkin [/bib_ref]. The EPIC array lacks 6 of the CpG sites originally included in the Knight clock, these values were imputed manually as non-available. Interestingly, DNA methylation-based age estimation relying on EPIC array data has already been described to accurately predict age despite the lack of several CpG sites originally included in Horvath's and Hannum's clocks [bib_ref] Systematic evaluation of DNA methylation age estimation with common preprocessing methods and..., Mcewen [/bib_ref].
Gestational age acceleration (ΔGA) was calculated as the absolute difference between epigenetic GA and chronological GA. Since ΔGA was associated with chronological GA (r = − 0.47; p < 0.001), the latter was included as a covariate in all statistical models; this association has been already reported in prior studies exploring epigenetic-based GA estimations at birth [bib_ref] An epigenetic clock for gestational age at birth based on blood methylation..., Knight [/bib_ref] [bib_ref] Associations between maternal risk factors of adverse pregnancy and birth outcomes and..., Girchenko [/bib_ref] [bib_ref] The epigenetic clock at birth: associations with maternal antenatal depression and child..., Suarez [/bib_ref].
Cell counts of CD4 + T cells, CD8 + T cells, B cells, NK cells, granulocytes, monocytes, and nucleated red blood cells (nRBCs) were estimated using the R code and statistical pipeline developed by Houseman [bib_ref] DNA methylation arrays as surrogate measures of cell mixture distribution, Houseman [/bib_ref].
A multiple linear regression model was built to analyze the correlation between ΔGA and CPR. Fetal sex, birthweight, diagnostic of either TTTS or sIUGR (binary variable), post-surgery interval (in TTTS cases where laser fetoscopy had been applied), and gestational age at ultrasound were included as independent variables in the model as they are known to influence either DNA methylation (from which ΔGA is calculated) or CPR. This analysis was conducted in the total MZ twin sample (n = 60).
DNA methylation at CpG sites annotated to the 43 genes of the Placental PRS1 as described by Ursini et al. [bib_ref] Convergence of placenta biology and genetic risk for schizophrenia, Ursini [/bib_ref] was retrieved to test their association with CPR. A second multiple linear regression model was then designed to explore putative effects of CPR upon methylation of PRS genes, testing 1,400 associations. The aforementioned confounding variables along with cell types proportions (CD4 + T cells, CD8 + T cells, B cells, NK cells, granulocytes, monocytes, and nRBCs) were included as covariates, as they are known to affect methylation values. False discovery rate (FDR) correction for multiple testing was applied, considering q values under 20% to be indicative of medium-confidence probes following prior studies [bib_ref] Epigenetic vestiges of early developmental adversity: childhood stress exposure and DNA methylation..., Essex [/bib_ref].
A twin-based approach previously developed in our group [bib_ref] Further evidence of DEPDC7 DNA hypomethylation in depression: a study in adult..., Córdova-Palomera [/bib_ref] was also applied to refine the association between cg06793497 methylation and CPR. Briefly, intrapair differences for both variables of interest were computed for each twin pair; afterward, a regression model was fitted with an estimated intrapair cg06793497 methylation (Δmethylation) and intrapair CPR (ΔCPR). This last model was not adjusted for either sex or chronological gestational age since both variables are identical for both twins of a pair.
## Additional file
Additional file 1: DNA methylation values for CpG probes used to discard the presence of maternal contamination. (DOCX 29.6 kb)
[fig] Figure 2: Association between epigenetic age acceleration and cerebroplacental ratio measured during the third trimester. Epigenetic age delta (ΔGA) corresponds to estimated epigenetic age minus chronological age. Thus, ΔGA-positive values reflect epigenetic age acceleration while negative values point out the presence of epigenetic age deceleration. The cerebroplacental ratio (CPR) is calculated as the ratio between the MCAPI and UAPI. Both variables were significantly correlated when adjusting for sex, chronological gestational age, birth weight, and gestational age at ultrasound [/fig]
[table] Table 1: Top 10 CpG sites of the PRS methylomic exploration in association with CPR (1400 CpG sites tested)Refers to the statistics of each analysis in the first model encompassing 1400 CpG sites located in genes of the PRS for schizophrenia highly expressed in the placenta as described by Ursini et al.Refers to intrapair comparison of DNA methylation values and CPR. Thus, the n for these analyses was of 26 twin pairs [/table]
[table] Table 2: List of CpG sites included in the array located in the EP300 gene and its surrounding CpG island and antisense ncRNA (EP300-AS1) [/table]
|
Familial Interstitial Lung Disease Caused by Mutation of the STING1 Gene
# Introduction
In childhood, ILD refers to a heterogeneous group of rare pulmonary conditions with complex etiologies. Some gene mutations have been found to be related to systemic autoinflammatory diseases that cause ILD. In 2014, a novel autoinflammatory syndrome, a novel type I interferonopathy attributed to mutations in STING1, termed STING-associated vasculopathy with onset in infancy (SAVI), was initially reported. SAVI is characterized by early onset systemic inflammation related to cutaneous vasculitis and tissue damage in addition to interstitial lung disease. Because fewer relevant cases have been reported and the associated phenotype is highly variable, more studies should be conducted to delve into the disease. Furthermore, the phenotype of ILD and rheumatoid factor positive (RF+) polyarticular arthritis is reported even less frequently. Accordingly, this study reports three familial cases of this new phenotype, which is expected to expand the spectrum of this disease.
## Case presentation
# Methods
# Ethical approval
The study was approved by the institutional review board of the Children's Hospital of Hebei Province, and informed consent was obtained. Consent for minors was signed by their parents.
## Patient recruitment and clinical characteristics
The index case is a boy, age 9 when enrolled, who presented with a persistent cough after physical activity over the course of more than 8 years and joint pain for 4 months. He was the first full-term child born via vaginal delivery of unrelated parents of Chinese ethnic background with a body weight of 4.6 kg at birth. Shortly after birth, the patient showed fatigability after light activity and activity-induced tachypnea and cough. A chest X-ray and lung computed tomography (CT) showed bronchial pneumonia with interstitial changes, and percutaneous oxygen saturation was 95% in room air. Severe febrile attacks were rare, and the patient did not seek treatment regularly. At the time of enrollment, the patient weighed 32 kg and was experiencing no failure to thrive (FTT), no obvious abnormality in growth and development, and no tissue lesions involving the skin.
Four months before admission, the patient developed migratory polyarthralgia, affecting the bilateral fingers and toes, wrists, knees, and other joints with swelling at the proximal joints of the middle finger. After the administration of the musk rheumatism capsule (main ingredients include Sichuan aconite, scorpion, earth dragon, black bean, beehive, and artificial musk), the patient's symptoms were alleviated. However, the pain worsened 3 days before admission. Laboratory tests showed a rheumatoid factor (RF) of 47.9 IU/ml (normal: 14 IU/ml) and erythrocyte sedimentation rate (ESR) of 75 mm/h (prior to admission).
The proband's only sibling is a 4 year-old boy, who is asymptomatic; however, CT of the chest showed interstitial changes.
The proband's father presented with shortness of breath postexercise at the age of 18 after an episode of illness. Chest radiographs and CT showed persistent pathological changes of interstitial pneumonia. At the time of enrollment, the patient's percutaneous oxygen saturation was 93% in room air.
None of the patients needed supplemental oxygen or experienced tachypnea in daily life, but the father experienced mild dyspnea during heavy physical activity. The proband's mother is healthy. The proband's father has a sister from the same parents who has no similar symptoms. The proband's grandmother and grandfather were both healthy.
## Laboratory and other investigation
The proband and his father and brother underwent chest Xray or CT. In addition, the proband's pulmonary function and indicators of systemic inflammation were examined.
## Whole genome sequencing and sanger sequencing
Targeted exome sequencing (TES) was performed on DNA from peripheral blood cells. Genomic DNA was fragmented, ligated with paired-end adaptors, amplified, and purified. A total of 6,110 gene exons and their 50 bp adjacent introns were captured by a TES Kit (SureSelect Focused Exome, Agilent, USA). A DNA library was established by postcapture amplification and purification and then sequenced on Illumina HiSeq X Ten (Illumina, USA). NextGene V2.3.4 software (Softgenetics, USA) was used for sequencing data alignment to the human genome reference (hg19) and variant calling. The mean read depth was 134.01×, but reads reached up to 200× for 96.194% of the target sequences. Meanwhile, annotation information, including conserved nucleotides and amino acids; prediction of biological functions; frequency in healthy population (1000 Genomes, gnomAD, dbSNP database, and local specific databases); and data from HGMD, Clinvar, and OMIM, was added in NextGene V2.3.4, using self-constructed scripts in our lab.
Variants of pathogenicity were chosen according to standards and guidelines for the interpretation of sequence variants published by ACMG in 2015 with HGVS nomenclature. We used NM_198282.3 as the reference sequence for STING1 gene analysis. Potentially pathogenic mutations were verified using Sanger sequencing with forward primer 5 ′ -GGACTCTATCGTTACAGGCTGAGG-3 ′ , reverse primer 5 ′ -GCTCCATAGCCCCTTCTGACTCT-3 ′ , and a product length of 414 bp.
# Results
## Laboratory, imaging, and pulmonary function tests
After birth, chest X-ray or CT of the proband showed progressive pulmonary interstitial fibrosis. In addition, there were multiple cystic changes and emphysema in the lobular center of the lung.
Pulmonary function tests indicated mixed ventilation dysfunction. Bronchoscopic images of the index case have a coarse and pale appearance. Bronchoscopy found hyperemic tracheal mucosa as shown in the high-magnification smear. Cell classification of bronchoalveolar lavage fluid (BALF) showed 79% neutrophils and 21% mononuclear phagocytes. The high-magnification smear showed red blood cell +, white blood cell ++, epithelial cell +, gram-positive diplococcus+, and gram-negative diplococcus +. Bacterial cultures of alveolar lavage fluid showed normal respiratory flora.
Laboratory investigations revealed a normal white blood cell count in a routine blood examination (WBC 8.7 NE 4.27 * 10 ∧ 9/L LY4.01 * 10 ∧ 9/L). During hospitalization, the child developed fever, and the blood routine was reviewed; it showed WBC 12.3 * 10 ∧ 9/L NE 5.79 * 10 ∧ 9/L LY 5.83 * 10 ∧ 9/L systemic inflammation with elevations of C-reactive protein (CRP), high-sensitivity CRP (hs-CRP), and ESR levels (CRP of 39.5 mg/L, hs-CRP of 22.73 mg/L, and ESR of 41 mm/h upon reexamination after admission). Blood CD4+ T was normal. CD8+ T was 38.4% (normal: 20-35%). The ratio of CD4+ to CD8+ cells in the lymphocytes was normal. Total B cells (CD19+) were 18.6% (normal: 5-18%), which was slightly elevated, and natural killer (NK 1.7%) lymphocytes decreased. Hypergammaglobulinemia of IgA and normal IgM and IgG were detected. All autoantibodies were negative (anti-PO, anti-SSA, anti-CENPB, anti-AnuA, anti-AHA, anti-dsDNA, anti-Sm, anti-snRNP, anti-SSB, anti-Jo1, and anti-SCL-70). Indicators of vasculitis (p-ANCA, C-ANCA, MPO, and PR3) were all negative except for p-ANCA. Anticyclic citrullinated peptide antibody was 10.33 RU/ml (normal: 0-25 RU/ml)Rheumatic changes). C3 was in the normal range, and C4 was elevated to 0.62 g/L (reference range: 0.18-0.4 g/L). No abnormalities were found in procalcitonin (pct 0.076 ug/L; reference range: 0-0.5 ug/L).
A bone marrow puncture showed increased myelocytes and late promyelocytes (14.5%, 17%). Liver, renal, and thyroid functions were normal.
Chest imaging of the proband's brother and father showed interstitial changes.
# Mutation analysis
Whole genome sequencing of samples from the index patient was performed, and genetic analysis by Sanger sequencing confirmed a heterozygous mutation (c.842G>A p.Arg281Gln substitution of arginine to glutamine) in the exon 7 of STING1 (NM_198282.3), e.g., the gene that encodes the stimulator of interferon genes, consistent with SAVI.
Mutation analysis revealed that the father and younger brother of the proband both carry the same p.Arg281Gln variant, and his mother is a WT, so the mutation was inherited from the father. The family pedigree is shown in.
The proband's aunt, grandmother, and grandfather are healthy, and no DNA samples were available.
## Diagnosis
The diagnosis of SAVI was made due to the presence of interstitial lung disease, systemic inflammation, and genetic analysis results.
## Treatment
After 13 days of non-glucocorticoid therapy for phlegm (acetylcysteine was used for atomization) and 7 days of antiinfection (cefathiamidine) treatment-on the seventh day of hospitalization the patient experienced fever and increased leukocytes in the blood-the patient's symptoms improved, but mild shortness of breath was still present when physically active. However, no oxygen therapy was required.
The proband's younger brother showed no signs of discomfort, but their father could not tolerate heavy physical activity; none of them received treatment.
So far, we have followed the family for more than 3 years. There was no significant change in the proband and his brother with no signs of hypoxia in their normal activity, but his father died of sudden respiratory failure in the third year, which was unanticipated.
# Discussion
Pediatric ILD is a heterogeneous group of rare pulmonary diseases presenting with chronic respiratory pathologies, including inflammatory and fibrotic changes. ILD detected in infancy is often related to genetic mutations. At present, mutations on STING1 and other novel genes, such as COPA, MIM612374, or MIM601924, have been shown to cause systemic autoinflammatory diseases involving the lungs.
STING1 encodes the STING protein, which is an adaptor molecule linking the sensing of foreign DNA (viral and bacterial) to the production of type 1 IFNs. Its mutations are responsible for type-I IFN overproduction, which has been recently identified as a new cause for interferonopathy, such as SAVI. SAVI is characterized by systemic inflammation (fever, high ESR and CRP levels, IgG and IgA hypergammaglobulinemia) associated with cutaneous vasculitis and tissue damage in addition to interstitial lung disease. To date, 13 identified mutations in the STING1 gene (V147L, V147M, F153V, N154S, V155M, G166E, C206Y, C206G, G207E, F279L, R281Q, R284G, R284S; https:// infevers.umai-montpellier.fr/web/search.php#ancre1768) (1, 3-7) have been described. However, this case study is the first report of STING1 point mutation (c.842G>A p.Arg281Gln) in Chinese patients.
Lung involvement is reported in up to 90% of previous cases of SAVI. In our report, the proband and his brother were found to have imaging changes in interstitial pneumonia starting in infancy, and their father had dyspnea after physical activity from the age of 18 and had the same changes in imaging. In the study by Liu et al., the authors found that five of six patients had evidence of interstitial lung disease on CT though three of them had no respiratory symptoms. STING protein was expressed in bronchial epithelium, alveolar macrophages, and alveolar type II pneumocytes (1). STING mutation results in constitutive activation of the STING-interferon pathway and upregulated type 1 IFN level, resulting in an inflammatory vasoocclusive process as well as pulmonary lesions, possibly through the activation of alveolar macrophages or pneumocytes. In addition, lung toxicity has been reported in patients with multiple sclerosis treated with exogenous type 1 IFN (10), which is consistent with the specific lung pathology seen in SAVI. In this report, usually, they have no fever or other symptoms of infection though the proband was admitted with severe pulmonary fibrosis and no obvious signs of infection. The fiberoptic bronchoscopy suggested the presence of inflammation in the mucosa, but bacterial cultures and cytological classification of a high-magnification smear of alveolar lavage fluid did not support infection although they were consistent with noninfectious inflammation-induced tissue damage, and the BALF showed neutrophilic alveolitis (the percentage of neutrophils in the cell classification was 79%). This change was considered to be caused by mutations in STING1 that led to a constitutive production of high levels of type I IFNs without infectious triggers (1).
Our results show that all patients carried the same c.842G>A (p.Arg281Gln) mutation as previously reported by Melki in a 7 year-old, white, European ethnicity girl in 2016. The authors used a luciferase-based IFN-β transcription reporter assay and detected robust IFN-β activation in vitro with p.Arg281Gln substitution in the STING1 gene. The variant led to a constitutively active STING, and the type I IFN signature was positive in peripheral blood, which was responsible for the disabling autoinflammation in the patient. The authors also confirmed that this variant lies within a novel functional cluster in STING.
Common laboratory features of SAVI include raised IgG and IgA and positive autoantibodies-not all of which were observed in our patients. In the index case, IgA was raised, but IgM and IgG were normal, and autoantibodies were negative. SAVI is characterized by systemic inflammation with elevations of CRP, hs-CRP, and ESR levels, which were also characteristic of our patients.
As one of the type I interferonopathies, SAVI has clinical manifestations that are variable even within familial cluster. A literature review reveals that the age of clinical presentation is typically in infancy between birth and 6 months (17/26). Clinical onset in teenage years is rare (3/26). Initial presentations include tachypnea, chronic cough (13), telangiectatic erythema, fever and pustular rash, failure to thrive, ischemic acral lesions, and even partially necrotizing lesions. Among previously described cases, lung involvement occurs in up to 79% of cases and usually starts in the neonate period. Not all patients were symptomatic, and some interstitial lung diseases were identified by imaging. Our youngest patient was in accordance with this phenotype and was not overtly symptomatic. Respiratory manifestations range from wheezing to tachypnea, chronic cough, and so on.
However, genotype-phenotype correlations may vary. Compared with this disease as previously reported and the same mutation reported by , the symptoms of the patients in our study have significant deviations. The proband and his brother showed a rather mild clinical presentation, and even in the lung CT, the fibrosis was obvious, but their father's breathing was more pronounced during physical activity. The patient in the Melki report presented with feeding difficulties, telangiectatic skin lesions, and damage of liver function. In our cases, the initial and main presentations were interstitial pneumonia; they had few extrapulmonary manifestations and no skin lesions. Considering that other variations in the STING1 gene may result in a decrease in the expression or function of the haplotype, resulting in a mild clinical phenotype, we further checked the exome sequencing of variants INFAR1 and 2, JAK-STAT, TBK1, and IRF3 to see if the variants were affecting the pathway, and the results were negative. The other interesting finding is the arthritis of the proband. It is increasingly recognized that type 1 interferonopathies can manifest as musculoskeletal disease. A recent study performed DNA sequencing in 100 patients diagnosed with juvenile idiopathic arthritis (JIA) with analysis of all coding exons and flanking introns, including STING1. They found that these JIA patients had a relatively frequent incidence of autoinflammatory syndromes. In the SAVI literature, arthritis and arthralgia have previously been described as minor features [heterozygous mutation of c.463G> A; p.Val155Met], but for mutation of c.842G>A (p.Arg281Gln) in STING1, this is the first report. Type I interferonopathies are a clinically heterogenic group of diseases with a constitutive activation of the type I interferon generate pathway that might present as atypical, severe, early onset rheumatic diseases with which interstitial lung diseases are common. It is suggested that the differential diagnosis of such clinical cases should include type I interferonopathies, such as SAVI.
Currently, therapeutic options for SAVI are limited, and traditional immunosuppressive medications and biologic therapies have disappointing efficacy. No standard immunosuppressive treatment approach is able to control disease progression.
Fremond et al.described the efficacy of a selective oral JAK1/2 inhibitor, ruxolitinib, in three children aged between 5 and 12 years with STING1-activating mutations; two of them carried a p.V155M mutation, and the other carried a p.V147M mutation. The authors found that JAK inhibition had a marked positive effect on all aspects of the symptoms in all three children, including major improvement in pulmonary function, and was also well-tolerated. However, in another study, the author reported a patient with SAVI (p.ser102pro and p.phe279leu, two variants of STING1) who was treated with another Janus kinase inhibitor, tofacitinib, and the patients' skin lesions improved but the pulmonary defects remained unchanged. In another report, involving treatment with a JAK1/2 inhibitor, ruxolitinib, in a patient with severe pulmonary involvement of SAVI (c.842G>A p.Arg281Gln mutation in TMEM173), at 18 months of therapy, a CT scan revealed a worsening of the interstitial disease. JAK inhibition may be worth considering as a therapeutic approach for some subtypes of SAVI.
Our patients were not treated with a Jak1/2 inhibitor due to limited conditions. On admission, the index required short-term anti-infective treatment and oxygen intake during bronchiolitis, but with no need for ventilator support, he was treated with no immune-modifying agents.
Other therapies for SAVI include TBK1 antagonists, and ER exit-blocking agents may be an option. Further study needs to be done to get its effect on SAVI.
From published reports, we can see that some cases of SAVI, especially in patients with severe lung involvement, the prognosis is poor, and in our report, the proband's father died at 38 years old of sudden respiratory failure. Therefore, it is emphasized that such diseases should be treated early even in asymptomatic patients.
Limitations of this study include that we have not been able to perform a cytokine dosage in the patients' plasma or sera.
In summary, we describe three individuals of two generations from the same family with SAVI. It is the first report of this specific variant causing familial SAVI in China. The initial and major clinical manifestations are largely identified in the lungs with ILD, and extrapulmonary symptoms were minimal except for the proband diagnosed with arthritis 8 years after onset. This phenotype has been reported rarely.
# Data availability statement
This article contains previously unpublished data. The name of the repository and accession number(s) are not available.
# Ethics statement
Written informed consent was obtained from the individual(s) and/or minor(s) ′ legal guardian/next of kin for the publication of any potentially identifiable images or data included in this article.
# Author contributions
JL wrote the manuscript under the direction of SA and ZD and his classmate modified the language. We thank them for their efforts. All the authors designed the study and analyzed experimental results together. |
The role of salt intake and salt sensitivity in the management of hypertension in South Asian people with chronic kidney disease: a randomised controlled trial
# Introduction
Chronic kidney disease (CKD) prevalence is three to five times higher in the Bangladeshi population compared with other ethnic groups in the UK. [bib_ref] Population need for renal replacement therapy in Thames regions: ethnic dimension, Roderick [/bib_ref] Hypertension is the cause and consequence of CKD 2 and extracellular volume expansion is an important factor in the development and maintenance of hypertension in patients with CKD. [bib_ref] Volume status and blood pressure during long-term hemodialysis: role of ventricular stiffness, Chen [/bib_ref] The British Bangladeshi population has an unusually high salt intake, twice that of the general population, which might contribute to CKD and hypertension. [bib_ref] Dietary salt intake of Bangladeshi patients with kidney disease in East London:..., Brito-Ashurst [/bib_ref] An intake of less than 6 g salt/d (∼100 mmol Na/d) [bib_ref] Why 6g? A summary of the scientific evidence for the salt intake..., Agency [/bib_ref] is advised in the UK and Europe based on the association of salt intake with blood pressure (BP) and cardiovascular disease rather than the much lower intake required to prevent sodium deficiency. However, even greater restrictions to below 60 mmol Na/d have been proposed in NICE guidelines. Nevertheless, a recent systematic review suggested an adverse effect of intakes as low as 3.5 g/d. [bib_ref] Reduced dietary salt for the prevention of cardiovascular disease, Taylor [/bib_ref] Meta-analyses of salt restriction studies show a greater BP lowering effect of salt restriction in hypertensive compared with normotensive subjects; under optimal conditions reductions in systolic/diastolic BP of 7/4 mm Hg 7 can be obtained with salt intake reduced to 100 mmol/d. Patients with resistant hypertension, defined as BP that remains above target despite the use of three or more antihypertensive medications, [bib_ref] Resistant hypertension: diagnosis, evaluation, and treatment: a scientific statement from the American..., Calhoun [/bib_ref] are thought to be salt sensitive. [bib_ref] Effects of dietary sodium reduction on blood pressure in subjects with resistant..., Pimenta [/bib_ref] However, adherence to dietary advice is often poor resulting in smaller changes (<1mm Hg) in BP of dubious clinical value. In previous work we investigated barriers to dietary salt restriction in Bangladeshi patients with CKD in East London 10 and developed a strategy intended to reduce salt intake in these patients. This present study set out to test the hypothesis that a low-salt educational intervention tailored to meet the needs, customs and practices of this population in addition to standard hypertension medication management would result in greater BP reduction than standard care alone: a first such intervention with this population group.
# Methods
A parallel-group randomised trial design was selected, and conducted between June 2008 and July 2009. Ethical approval was obtained from the relevant research ethics committee.
Participants were patients with established moderate-severe CKD of Bangladeshi origin residing in East London, UK. Inclusion criteria were estimated glomerular filtration rate (eGFR) <60 mL/min and mean BP >130/80 mm Hg on at least two clinic visits or taking antihypertensive medication. Patients on dialysis, those with a body mass index <20 or >35 kg/m 2 , urinary incontinence, or
Open Access Scan to access more free content cognitive impairment or mental problems impairing their ability to participate were excluded. Participants were recruited at the predialysis clinic of a tertiary renal unit in London by the researcher. Randomisation to treatment was conducted by the study statistician using computer-generated random blocks with block sizes between four and eight and the group assignment given to the researcher. This was a dietary behaviour intervention, thus, neither participants nor the dietitian administering the intervention could be blinded to treatment allocation. Data analysis was conducted by the study statistician who was blinded to treatment allocation.
## Intervention
The intervention group was initially advised by the study dietitian at the hospital clinic followed by practical cooking and educational sessions in the community facilitated by Bengali workers and attended by the researcher (see online supplementary file). Community cooking sessions were delivered in conjunction with Community Kitchen UK.In the community sessions, intervention participants cooked two versions of their traditional meals: one followed their usual recipe, the other had salt reduced by 50%. Fortnightly telephone calls from a Bengali worker followed, to reinforce advice and set new targets. The control group received usual care from the renal clinic in the form of a low sodium general dietary advice sheet sent by post with the physician's letter. This had not been specifically adapted for Bangladeshi diets.
## Data collection
Data collected at enrolment included age, sex, medication use and comorbidities, including diabetes mellitus. Data collection for the primary outcome was by ambulatory BP measured using TM-2430-13 devices (A&D Medical, Milpitas, California, USA; graded A/A by the British Hypertension Society) and Doctor Pro software, in accordance with recommendations. [bib_ref] European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure..., O'brien [/bib_ref] Daytime measurements were taken at 30 min intervals, night-time measurements every 60 min. Height, weight and body composition (total body water) were measured using the Fresenius Medical Care D GmbH Body Composition Monitor; blood samples were obtained for glycosylated haemoglobin (HbA1c). Physical activity levels were recorded using the YamaxDigi-Walker SW-200 (Yamax Corporation, Tokyo, Japan) pedometer, shown to have an overall mean absolute error of 3% for outdoor normal walking. [bib_ref] The validity and reliability of a novel activity monitor as a measure..., Ryan [/bib_ref] The accuracy of the pedometer on each participant was checked by a 20-step test at the outset, with an acceptance criterion of ±2 steps. [bib_ref] The validity and reliability of a novel activity monitor as a measure..., Ryan [/bib_ref] Data were collected at two time points, at baseline and at end of study-6 months later.
## Outcomes
The primary outcome was reduction in systolic BP (SBP) determined by 24 h ambulatory monitoring. Secondary outcomes were changes in diastolic BP and reduction in eGFR. Measurement of 24 h urinary sodium, potassium and creatine were undertaken using routine methods at baseline and follow-up as indices of adherence to the intervention and determined by assessors blinded to treatment allocation.
# Statistical analysis
Sample size calculations were based on a sample of 25 participants per group giving 80% power to detect a significant reduction in the mean SBP of 8 mm Hg at p<0.05 between the two groups (which was regarded as clinically relevant difference), assuming a SD of 10 mm Hg. 14 Sample size was increased to 26 per group to allow for non-compliance or dropout. Analyses were conducted on an intention-to-treat basis. Changes within groups between baseline and follow-up at 6 months were compared using analysis of covariance and results are expressed as mean values with 95% CIs using Stata V.10 (StataCorp LP, Texas, USA).
# Results
Participant recruitment and progress through the trial is shown in [fig_ref] Figure 1: CONSORT flow chart of participant recruitment, allocation and assessment [/fig_ref]. Of the 56 participants recruited six withdrew; three cited the inconvenience of 24 h urine collection, two were unwilling to undergo ambulatory BP monitoring and one was unwilling to attend the community cooking activity. One intervention group participant died; one control group member relocated to Bangladesh. Data were available for 48 participants. Details are shown in table 1; groups were well-matched including for antihypertensive medication, with most receiving ACE inhibitor or angiotensin-receptor blocking medicines and diuretics.
## Adherence to the dietary intervention
All participants attended the initial briefing session with the study dietitian. Male participants attended with their wives, daughters or sisters while female participants attended with their daughters or daughters-in-law. Participants were split into four groups of six or seven to attend the community cooking sessions; each group was to attend two weekly consecutive sessions. Male participants chose not to attend but sent a female representative; a wife, daughter or sister for single men. The first weekly session was attended by 88% (23/25) of the participants or representatives; the second and final session was attended by 84% . Overall, all participants attended at least one cooking session.
Adherence to dietary salt recommendations was indicated by urinary sodium excretion. At baseline urinary sodium excretion was approximately 260 mmol/24 h in both groups (figure 2). After 6 months, this had reduced by 122 mmol/24 h (95% CI −140 to −105, p<0.001) in the intervention group, and by 13 mmol/24 h (95% CI −18 to −8, p<0.001) in the control group. At follow-up sodium excretion differed significantly between groups, by 103 mmol (95% CI −131 to −76, p<0.001).
## Primary and secondary outcomes
Systolic BP was elevated in both groups at baseline but fell by 8 mm Hg (95% CI 5 to 11, p=0.0003) on tailored intervention compared with the usual care group. [fig_ref] Figure 3: Changes in daytime and night-time systolic blood pressure [/fig_ref] shows the significant ( p<0.001) falls in daytime and night systolic and diastolic BP in the intervention group compared with the control group. Non-dipping, that is loss of the normal nocturnal reduction in night-time SBP, with a difference >10 mmHg between nighttime and daytime SBP, was observed in 60% (15/25) of the intervention group and 56% (13/23) of the control group at baseline. At follow-up this reduced by 40% (6/25; p=0.02) in the intervention group but remained unchanged in the control group.
The observed changes in eGFR from baseline to follow-up were similar for both groups. An eGFR decline of 3.0 (95% CI 0.1 to 6.0) and 3.4 (95% CI 1.0 to 5.7) mL/min per 1.73 m 2 were observed in the intervention and control groups respectively.
## Covariate findings
Potassium excretion was low (40 mmol/d) in both groups and unchanged at follow-up. Physical activity levels were low in both groups and remained unchanged during the study. Glycosylated haemoglobin concentrations remained elevated at >8.0%, indicating poor but unchanging diabetic control in both groups. Body weight did not change in either group, but there was a modest but statistically significant reduction in mean total body water in the intervention group (0.50 L, p<0.01) compared with no change in the control group (0.26 L, not statistically significant.).
# Discussion
Dietary advice to lower salt intake is routinely given to patients with CKD in the form of an information sheet; this study suggests this is ineffective at changing behaviour. By contrast, the dietitian-led intervention which identified the sources of salt in the Bangladeshi diet and developed strategies to lower intake, achieved a reduction in dietary salt intake of over 100 mmol/d. While mean urinary sodium excretion still remained well above the UK target of 100 mmol/day, 15 postintervention group results 24 h urinary sodium excretion for intervention and control groups is shown as the mean difference with 95% CI in parenthesis. All differences are p<0.001. Control n=23, Intervention n=25.
were more similar to those seen in the UK white population. This reduction in salt intake led to a highly significant fall of 8/ 3 mm Hg in BP; very close to the figure predicted by meta-analyses. [bib_ref] Effect of modest salt reduction on blood pressure: a meta-analysis of randomized..., He [/bib_ref] The dietary approach used has also been shown to be effective in BP reduction with other ethnic groups with high dietary salt intakes. [bib_ref] Salt and blood pressure: a community trial, Forte [/bib_ref] Behavioural intervention studies have previously demonstrated that knowledge is a key contributing factor to adherence to low-salt diet [bib_ref] Effects of comprehensive lifestyle modification on blood pressure control: main results of..., Appel [/bib_ref] and that lack of knowledge is a key barrier in dietary modification and adherence. However, knowledge of the need to reduce dietary salt intake is not always enough to ensure dietary modification and adherence. This study shows that knowledge tailored to recipients' needs and contexts, delivered in a practical and acceptable manner, can effect behavioural change and achieve health benefits.
Our previous studies showed that Bangladeshi patients with CKD have much higher intakes of salt than the general population, with much of the salt being added during home preparation of food rather than during processing, as is the case in the general population. Consequently, routine advice for salt reduction designed for predominantly white European populations was not appropriate for this group of patients, for whom the engagement of family members was crucial. This was particularly relevant for these study participants, who almost exclusively ate home-prepared meals in family groups.
We noted that over half of the Bangladeshi patients with CKD had raised BP throughout the day and night: the 'nondipping' effect. Previously 'non-dipping' prevalence has been found to increase with worsening CKD, with 15% of normal subjects affected increasing to 75% in those with stage five CKD. [bib_ref] Relationship between arterial hypertension and renal damage in chronic kidney disease: insights..., Paoletti [/bib_ref] Non-dipping has been associated with increased target-organ damage (heart, brain, kidney), 20-22 raised frequency of stroke and myocardial infarction, [bib_ref] Ambulatory blood pressure. An independent predictor of prognosis in essential hypertension, Verdecchia [/bib_ref] and higher cardiovascular mortality. [bib_ref] Daytime and nighttime blood pressure as predictors of death and cause-specific cardiovascular..., Fagard [/bib_ref] Decreased salt intake and urinary sodium excretion led to greater reduction in night-time SBP compared with daytime, and the restoration of the normal physiological night-time dip in BP in many patients. In black salt-sensitive patients with hypertension salt restriction improved the circadian rhythm of BP with a return to dipping pattern. [bib_ref] Influence of salt intake on the daytime-nighttime blood pressure variation in normotensive..., Damasceno [/bib_ref] Our study confirmed that salt reduction can change the pattern from 'non-dipper' to 'dipper', and a recent review has concluded that the South Asian population is salt sensitive. [bib_ref] Salt sensitivity, insulin resistance, and public health in India, Ganda [/bib_ref] A return to a 'dipping' BP pattern may lead to a significant reduction in the risk of vascular events for this patient group but long-term follow-up is required to demonstrate this.
## Comparison with other studies
Data supporting our findings were reported by MacGregor et al, [bib_ref] Double-blind study of three sodium intakes and long-term effects of sodium restriction..., Macgregor [/bib_ref] who demonstrated that when urinary sodium excretion dropped by 100 mmol/24 h, subjects supine BP declined by 8/5 mm Hg ( p<0.01); a sodium reduction of 150 mmol/24 h led to a larger decline of 16/9 mm Hg. Similarly, a modest salt reduction of 50 mmol/24 h (from a much lower baseline than in our participants: an average of 177 mmol/24 h) resulted in 7/3 mm Hg drop in BP in a randomised trial in older people. [bib_ref] Double-blind randomised trial of modest salt restriction in older people, Cappuccio [/bib_ref] A recent meta-analysis of 17 trials in hypertensive individuals over ≥4 weeks supports the approximate magnitude of this effect. 7 A recent study of modest dietary sodium restriction in patients receiving ACE medicines showed 11mm Hg reduction in SBP in non-diabetic nephropathy. [bib_ref] Moderate dietary sodium restriction added to angiotensin converting enzyme inhibition compared with..., Slagman [/bib_ref] Our study confirms this magnitude of association between reduction in sodium excretion and BP values. Moreover, our study is the first study to our knowledge of a population with CKD with traditionally high salt intake and hypertension with a follow-up period of 6 months suggestive of sustained benefit in the tailored intervention approach.
# Strengths and limitations
The strength of this study is that it delivered an effective salt reduction dietary intervention for this group of patients, and demonstrated participants' adherence to dietary advice through 24 h urinary sodium excretion. However, only single 24 h urine collections and 24 h ambulatory BP recordings were made. Further, treatment allocation could not be blinded. Therefore, ambulatory BP readings were analysed centrally in the hypertension unit by a statistician who was blinded to the treatment allocation. Hence the results were relatively less likely subject to bias. It remains uncertain whether reducing BP may translate into slowing of disease progression or reduction in cardiovascular events.
## Applicability and generalisability
The Bangladeshi population and indeed the South Asian group are known for a high dietary salt intake, are at a high risk of CKD and thus, hypertension development. Dietary salt reduction can safely and usefully be extended to other family members who may in time also be at risk of developing hypertension. Other ethnic groups with a high prevalence of CKD and hypertension, such as black African and Afro-Caribbean populations, may also benefit from tailored dietary interventions to reduce salt intake.
# Conclusion
This study demonstrates the importance of tailoring dietary advice to patients' contexts, cultures and needs, particularly for minority and high-risk groups. Healthcare professionals need education and training in methods to enable them to translate generic principles of healthy living and health promotion in such a way as to successfully deliver education and promote its application in the daily lives of their patients. Policy makers need to recognise the importance of resourcing complementary approaches to medication for effective BP control. An integrated approach, drawing on multiple successful approaches to hypertension reduction, offers the best option for BP management in patients with CKD. Acknowledgements This is an original work that has not been published before and is not under consideration in any other journal. IdeB-A would also like to thank Nadia Pendleton, food writer director of community kitchen UK, for the training and support on conducting the practical educational sessions in the community. Finally, this study was funded by Barts and The London Charitable Foundation. IdeB-A, the recipient of this Fellowship award, would like to thank this organisation for their financial support.
Contributors IdeB-A, researcher, designed the study and data collection tools, conducted educational intervention, collected data for the whole trial, drafted and revised the paper. MV, statistician, wrote the statistical analysis plan, analysed the data, drafted and revised the paper. MMY, LP, HD, TABS and JET were supervisors. The supervisors contributed to the design of the study and data collection tools, monitored data collection for the whole trial, drafted and revised the paper.
Funding This research was funded by a PhD fellowship grant made to IdeB-A from the trustees of Barts and The London Charitable Foundation. The analysis, interpretation of data, generation of the manuscript and decision to submit for publication were carried out independently of the funding body.
Competing interests All authors declare that for this submitted work: IdeB-A received salary support from a grant made to Barts and The London NHS Trust by the trustees of Barts and The London Charitable Foundation.
## Ethics approval east london & the city research ethics committee 3.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The original article was prepared from a database established by the corresponding author and it is accessible to all listed authors. This database remains the intellectual property of Barts and The London NHS Trust, but confidential data sharing agreements may be entered into to address any questions in relation to the trial. All data relevant to this author are presented here.
Open Access This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 3.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/ licenses/by-nc/3.0/
[fig] Figure 1: CONSORT flow chart of participant recruitment, allocation and assessment. [/fig]
[fig] Figure 3: Changes in daytime and night-time systolic blood pressure (SBP) and diastolic blood pressure (DBP). Mean values and changes with 95% CI. All differences are p<0.001. Control n=23, Intervention n=25. [/fig]
[table] Table 1: Baseline characteristics for the intervention and control groups [/table]
|
An ERP Analysis of Recognition and Categorization Decisions in a Prototype-Distortion Task
Background: Theories of categorization make different predictions about the underlying processes used to represent categories. Episodic theories suggest that categories are represented in memory by storing previously encountered exemplars in memory. Prototype theories suggest that categories are represented in the form of a prototype independently of memory. A number of studies that show dissociations between categorization and recognition are often cited as evidence for the prototype account. These dissociations have compared recognition judgements made to one set of items to categorization judgements to a different set of items making a clear interpretation difficult. Instead of using different stimuli for different tests this experiment compares the processes by which participants make decisions about category membership in a prototype-distortion task and with recognition decisions about the same set of stimuli by examining the Event Related Potentials (ERPs) associated with them.Method: Sixty-three participants were asked to make categorization or recognition decisions about stimuli that either formed an artificial category or that were category non-members. We examined the ERP components associated with both kinds of decision for pre-exposed and control participants.Conclusion:In contrast to studies using different items we observed no behavioural differences between the two kinds of decision; participants were equally able to distinguish category members from non-members, regardless of whether they were performing a recognition or categorisation judgement. Interestingly, this did not interact with prior-exposure. However, the ERP data demonstrated that the early visual evoked response that discriminated category members from nonmembers was modulated by which judgement participants performed and whether they had been pre-exposed to category members. We conclude from this that any differences between categorization and recognition reflect differences in the information that participants focus on in the stimuli to make the judgements at test, rather than any differences in encoding or process.
# Introduction
A fundamental aspect of human cognition is the ability to acquire knowledge of categories. This enables us to assign properties to an object that we have learned are common to other members of that category. This has clear survival value, for instance we may be reluctant to eat a plant with milky sap that we have not encountered before if we have learned that other plants with milky sap are poisonous. We should infer that the new plant is also likely to fall into the category of poisonous plants.
Precisely how the mind represents categories has received a substantial amount of attention and recently theories of categorization have been informed by studies involving amnesic patients and functional imaging. The present research is concerned with two classes of theory of categorization in particular: episodic models and prototype models. Both prototype and episodic models assume that categorization decisions are based on similarity. According to episodic models we memorize each instance of a category [bib_ref] Context theory of classification learning, Medin [/bib_ref]. When asked to decide whether novel items are category members or not, the decision is based on a comparison of the item with each stored exemplar. In effect categorization is little more than a form of episodic memory. By contrast, prototype models assume that the categorization decision is based on the similarity of the item to a prototype, rather than to stored exemplars [bib_ref] Distinguishing prototype-based and exemplar-based processes in dot-pattern category learning, Smith [/bib_ref]. A prototype is usually defined as an abstraction of the central tendency or an average of previously encountered exemplars. The exemplars themselves need not be stored in memory giving prototype theory an economical advantage. Exemplar theory has the advantage of computational simplicity. There are a number of different candidate models within each class. We refer to episodic models as any model that describes categorization as essentially a memory based process as distinct from models that assume some form of abstraction occurs during learning as is described by prototype models. We discuss in more detail one episodic model that is based on exemplar similarity [bib_ref] Similarity scaling and cognitive process models, Nosofsky [/bib_ref] , although it is our intention to compare episodic models generally with prototype models rather than any one specific episodic model. A method that is frequently used to decide between these two classes of model is to identify dissociations between categorization and recognition. That is, if different patterns of data are observed when experimental participants are asked to make recognition decisions for category members that they have previously encountered compared to when they are asked to make decisions about the category membership of novel items, the conclusion that is often made is that these two decisions recruit different processes [bib_ref] Memory processes in classification learning: An investigation of amnesic performance in categorization..., Kolodny [/bib_ref] [bib_ref] The learning of categories: Parallel brain systems for item memory and category..., Knowlton [/bib_ref] [bib_ref] Dissociating explicit and implicit category knowledge with fMRI, Reber [/bib_ref]. The same logic is often used in studies that attempt to understand many cognitive processes that appear to involve separate processes, such as between implicit and explicit learning, re, priming and recognition, and recollection and familiarity based memory. Typically however, studies of this kind compare responses to different stimuli. In many respects it seems sensible to use stimuli in a recognition test that have previously been memorized and to compare this to a categorization test using novel stimuli. However, it is inevitably unclear whether any observed differences in behaviour are due to the differences in the stimuli (old items are by definition more familiar than novel items), rather than differences in the processes used to make the decisions. That is, when dissociations between categorization and recognition are based on different stimuli it is difficult to determine if the reported differences are due to the underlying processes, rather than some difference in the stimuli. A convincing dissociation would be apparent when it is observed in different decisions about the same stimuli. The principle aim of this paper is to compare recognition and categorization using the same set of stimuli. If differences in behaviour are observed in this case then we can conclude that these two kinds of decision do indeed recruit different processes. Because many previous studies have used neuropsychological methods to dissociate processes, and because similar behaviour can arise from different underlying processes, we examined both ERP activity and behavioural responses for recognition and categorization.
Prototype-distortion tasks have been influential in developing our understanding of how knowledge of categories is acquired [bib_ref] Perceived distance and the classification of distorted patterns, Posner [/bib_ref]. This particular paradigm is useful, because it permits the study of how participants learn information, whether by memory or abstraction, that is unlikely to be influenced by prior knowledge. In this paradigm a prototype stimulus is formed by generating a random pattern of nine dots, and additional category members are created by distorting the coordinates of each dot of the prototype (see [fig_ref] Figure 1: Example Stimuli [/fig_ref]. In the standard preparation participants are first shown a set of category members but not the prototype stimulus itself. In a subsequent test participants are shown a set of previously unseen patterns that consists of category members that vary in their similarity to the prototype, along with the prototype item itself, and category non-members. Numerous studies report that participants are more likely to endorse items that are similar to the prototype, including the prototype, as category members than dissimilar items [bib_ref] Human category learning, Ashby [/bib_ref]. This pattern of results is often interpreted as evidence that participants abstract a representation of the category that closely matches the prototype even though this is not present in the study period. This contrasts with an alternative model that assumes categories are represented by storing previously encountered instances in episodic memory [bib_ref] Exemplar-based accounts of ''multiplesystem'' phenomena in perceptual categorization, Nosofsky [/bib_ref].
Evidence from studies of amnesic patients and brain imaging support the prototype abstraction model of categorization by showing dissociations between categorization and recognition of study items. These suggest that categorization is predicated on a set of neural processes different from memory of the study items. [bib_ref] Memory processes in classification learning: An investigation of amnesic performance in categorization..., Kolodny [/bib_ref] [bib_ref] The learning of categories: Parallel brain systems for item memory and category..., Knowlton [/bib_ref] [bib_ref] Learning about categories in the absence of memory, Squire [/bib_ref]. The rationale for this is that if the prototype is abstracted during the study episode then episodic memory would not be required to store the study exemplars. It follows that knowledge of the category could be acquired by patients with organic amnesia. Indeed, at least three studies have found similar patterns of categorization in amnesic patients and healthy controls [bib_ref] Memory processes in classification learning: An investigation of amnesic performance in categorization..., Kolodny [/bib_ref] [bib_ref] The learning of categories: Parallel brain systems for item memory and category..., Knowlton [/bib_ref] [bib_ref] Learning about categories in the absence of memory, Squire [/bib_ref]. However, the amnesic patients performed at chance in a subsequent recognition test of the study exemplars. By contrast the control participants performed much better in the recognition test. The conclusion from prototype-distortion studies in amnesic patients is that category knowledge can be acquired in the absence of memory for study exemplars.
Data from a number of fMRI studies also lend support to this model. These tend to be concerned with activity that occurs when participants are asked to make decisions at test, rather than activity that might result from prototype abstraction during the study episode. The first study of this kind reported decreased activity in regions of the posterior occipital cortex for category members relative to category non-members [bib_ref] Cortical areas supporting category learning identified using functional MRI, Reber [/bib_ref]. One possibility is that category members are processed more fluently than non-members. An increase in activity was observed in frontal cortical areas that may be related to conscious deliberation of whether an item is a category member or not [bib_ref] Cortical areas supporting category learning identified using functional MRI, Reber [/bib_ref]. A related study [bib_ref] Contrasting cortical activity associated with category memory and recognition memory, Reber [/bib_ref] replicated the finding that the posterior occipital cortex shows a decrease in activation for category members relative to category nonmembers. Moreover, a separate recognition task revealed increased activation in the frontal and temporal lobes and, importantly, that the posterior occipital cortex showed increased activation. This finding appears to show a dissociation in the kind of activation resulting from categorization and recognition decisions. A possible interpretation may be that categorization relies on processes akin to perceptual priming and perhaps familiarity based memory [bib_ref] Cortical areas supporting category learning identified using functional MRI, Reber [/bib_ref]. However, the activation of different neural regions may also be influenced by how the participants are instructed to learn the category. For example, different patterns of activity have been observed when participants categorize test items following either incidental or intentional learning instructions during the study episode [bib_ref] Dissociating explicit and implicit category knowledge with fMRI, Reber [/bib_ref] : intentional learning results in activation of the hippocampus; by contrast incidental learning results in deactivation of the posterior occipital cortex. Other studies indicate that explicit memory might be involved in the early stages of learning as shown by hippocampal activation, but this declines as knowledge of the category is acquired [bib_ref] Event-related fMRI of category learning: Differences in classification and feedback networks, Little [/bib_ref] [bib_ref] Prototype-distortion category learning: A twophase learning process across a distributed network, Little [/bib_ref].
Despite the evidence in favour of prototype abstraction an alternative episodic model proposes that categorization is based merely on exemplar similarity. According to this model [bib_ref] Exemplar-based accounts of ''multiplesystem'' phenomena in perceptual categorization, Nosofsky [/bib_ref] [bib_ref] Similarity-scaling studies of dot-pattern classification and recognition, Shin [/bib_ref] participants make categorization judgements on the basis of the similarity of the test items to an episodic representation of the study items. Dissociations between categorization and recognition arise from a more liberal criterion for accepting test items as category members than for accepting test items as previously encountered. This model has had some success in accounting for behavioural data in healthy participants. How then can this model account for the preserved capacity to form categories in amnesic patients? The model does this by assuming that episodic memory is impaired but not entirely eliminated by organic amnesia [bib_ref] Dissociations between categorization and recognition in amnesic and normal individuals: An exemplar-based..., Nosofsky [/bib_ref]. In this way a liberal response criterion results in preserved categorization. It is not clear however, how this model can account for the data obtained from fMRI studies, which show qualitatively different patterns of activity for recognition and categorization, unless these effects result from the use of different stimuli in the categorization and recognition tests.
Event Related Potentials (ERP) can also provide potentially useful information about the neural correlates of category learning [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref] , but this technique has not previously been used in this specific paradigm. This method has the advantage over fMRI in that categorization and recognition can be disambiguated by differences in both timing and region. One previous experiment that used different materials (blobs rather than patterns of dots) found different ERPs for categorization and recognition [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref]. Early visual potentials (N1, 156-200 msecs) were associated with category membership. The amplitude was significantly more negative for category members than for non-members. These data are consistent with fMRI studies that implicate the posterior occipital cortex in categorization [bib_ref] Cortical areas supporting category learning identified using functional MRI, Reber [/bib_ref] , and support the view that this region (and categorization) is predicated on largely visual processes [bib_ref] The visual N1 component as an index of a discrimination process, Vogel [/bib_ref]. Middle latency components (FN400, 300-500 msecs) were associated with both category membership and with recognition. This component is interesting because it is thought to underlie familiarity based processing in dual-process theories of recognition memory [bib_ref] Dissociation of the neural correlates of implicit and explicit memory, Rugg [/bib_ref] [bib_ref] Brain potentials of recollection and familiarity, Curran [/bib_ref]. Later potentials in parietal regions (400-msecs) were associated with recognition only. This component is related to explicit recognition (i.e. recollection) of information from previous study episodes [bib_ref] Brain potentials of recollection and familiarity, Curran [/bib_ref] , and is thought to result from deeper hippocampal and medial temporal lobe activity because it is absent in patients with lesions of the hippocampus [bib_ref] Brain activity evidence for recognition without recollection after early hippocampal damage, Düzel [/bib_ref]. Because the timings of these effects are so brief differences between categorization and recognition are not likely to be detected using fMRI. If there are differences in the processes recruited by categorization and recognition we believe that they are most likely to be found using ERPs.
In the experiment that follows we compare the ERPs of categorization and recognition in the prototype-distortion task. The experiment is necessary because all of the previous neuropsychological investigations of this task have been conducted using fMRI, amnesic patients, or both. Moreover, given that prototype abstraction is typically related to early visual processing, ERPs, with their high-temporal resolution, ought to provide the ideal technique to study these processes [bib_ref] The visual N1 component as an index of a discrimination process, Vogel [/bib_ref]. The previously mentioned examination of ERPs in categorization was conducted using a different kind of stimuli [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref] and it is necessary to determine if the same processes and neural mechanisms are involved in this task. The experiment is also important because it introduces a methodological advance over the previous ones. All of the studies mentioned previously have compared recognition and categorization tasks using different items, because the categorization tasks require previously unstudied category members and recognition tasks require that at least half of the test items have been studied before. Because these previous studies have used different items in their categorization and recognition tests it is conceivable that the reported dissociations have occurred because of differences in the stimuli rather than purely differences in the processes underlying the decisions. In the experiment that follows two groups of participants made either recognition or categorization decisions about the same set of test items. Crucially, for the recognition group the category members used in the test were also used during the study period as category exemplars. The categorization group differs in that a different set of category exemplars was used for the study items. In this way we can ensure that any differences between categorization and recognition are due to the underlying cognitive process and not to perceptual differences between the items used in the two tests. This control of stimulus equivalence is also crucial for the ERP comparison. There were also two control groups who made either recognition or categorization decisions about the same sets of items but who saw no exemplars in the study period. Two previous comparisons of categorization and recognition in prototype-distortion tasks found that in terms of behavioural accuracy recognition was superior to categorization [bib_ref] Learning about categories in the absence of memory, Squire [/bib_ref] [bib_ref] Comparing the brain areas supporting nondeclarative categorization and recognition memory, Reber [/bib_ref]. These studies did use different items for each test, but if as the authors claim there is a process difference between the two kinds of decision we should obtain similar results even when, as in the experiment that follows, the stimuli are identical. On the other hand, if the two decisions involve the same underlying process, and if the previous results are due to differences in the items, then there should be no difference in the accuracy of recognition and categorization behaviour.
Predications about the precise ERP components that we might observe are necessarily speculative. However, generally the prototype model would predict different ERP components for categorization than for recognition. For instance Early visual potentials (N1, 156-200 msecs) [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref] for categorization and latter parietal potentials (400-msecs) for recognition [bib_ref] Brain potentials of recollection and familiarity, Curran [/bib_ref]. Models such as the exemplar or episodic accounts that claim that categorization and categorization are predicated on the same underlying processes are likely to predict similar components.
# Methods
# Ethics statement
This study was approved by the ethical review board at the School of Psychology, University of Nottingham, UK. Written consent was obtained from all the participants who were free to withdraw from the study at any time.
## Participants
Sixty-three right-handed volunteers took part in the experiment. Their mean age was 26 years (sd = 5.35); 40 were male and 23 were female. The participants were paid £20 (approx. J27, US$40).
## Stimuli
The stimuli consisted of dot-patterns constructed using the method described by Posner, Goldsmith and Welton [bib_ref] Perceived distance and the classification of distorted patterns, Posner [/bib_ref]. Using this method we first created a prototype pattern and then we created three lists each with 40 items. Distorting the coordinates of the prototype pattern created List 1 and List 3 items. List 2 items were pseudo-random patterns (see [fig_ref] Figure 1: Example Stimuli [/fig_ref].
## Design and procedure
This was a 26262 mixed model design with Exposure (Preexposed vs. Control) and Instructions (Recognition vs. Categorization) as between-subjects factors and List (1 vs. 2) as a withinsubject factor. In each case these lists were composed of the same items.
The experiment consisted of a study period and a test period. During the study period the pre-exposed participants were presented with either List 1 items in the Recognition Condition, or List 3 items in the Categorization condition. The participants were told that the study was an experiment on visual attention and were asked to look for the dot closest to the centre of the screen but were not given any instructions about the presence of a category or how to encode the items. Study trials consisted of a 3000 msec white fixation cross, followed by a study item that appeared for 5000 msecs with a white frame. The control participants were informed that they were taking part in an experiment on subliminal perception and visual attention. Control 'study' trials consisted of a 3000 msec white fixation cross. After this a black screen was displayed for 1000 msec, then the screen flashed white for 50 msecs, followed by a black screen with an empty white box visible for 50 msecs, followed by another white screen for 50 msecs before an empty black screen returned for 4000 msecs. The control participants were also asked to try and identify the central dot in each pattern but that they would be presented very briefly and be difficult to see. After the end of the study period there was a short break during which the electrodes were checked. No EEG was recorded during the study period.
Prior to the test the participants in the Categorization conditions were told that all of the items that they had just seen were instances of a category and that they would now see some new items, some of which belonged to the category and some did not. Each test trial consisted of a 3000 msec fixation cross. Each test item appeared for 5000 msecs followed by a prompt to indicate whether the item was category member or not. The participants in the recognition conditions were told that they would be given a recognition test for the items that they had just studied.
We presented the same 40 category members and 40 category non-members to the four groups during the test phase. The two Pre-exposed groups had already been presented with category members during the study phase. For the Pre-exposed Recognition group these were the same 40 category members as were subsequently used in the testing phase (i.e. List 1). For the Preexposed Categorization group the study items were different category members to those we subsequently used for the test phase (i.e. List 3). By changing the study items in each condition, List 1 items were 'old' for the recognition group because they had also appeared as study item. For the Categorization group the List 1 items were new in the sense that they had not appeared in the study period but belonged to the same category as the study items.
## Erp recording and erp formation
EEG was recorded throughout each block in the test-phase using a 128-channel electrical geodesic net (Electrical Geodesics, Inc.: EGI) [bib_ref] Spatio-temporal analysis of brain electrical fields, Tucker [/bib_ref] , digitised at 250 Hz. The recording was performed with a hardware bandpass filter of 0.01 Hz to 100 Hz. Before recording, impedance on each of the 128 electrodes was reduced to ,50 kV. Due to amplification techniques, the EGI system provides an excellent signal-to-noise ratio, despite these relatively high electrode impedances [bib_ref] Scalp electrode impedance, infection risk, and EEG data quality, Ferree [/bib_ref]. The vertex was used as an acquisition reference.
Stimulus-locked epochs were created, time-locked to each test item, with each epoch starting 100 ms before stimulus onset and ending 1000 ms afterwards. Segments were rejected if contaminated by eye-blinks/movements (indicated by EOG activity greater that 70 mV). This process was also checked manually. Trials containing voltage amplitudes greater than 200 mv or a change greater than 100 mv were also removed. We did not reject 'error' trials, as control participants who have never seen the items before cannot correctly 'recognize' them. Instead we objectively classified stimuli as to whether they were List 1 or List 2 items. ERPs elicited by these two types of stimuli were compared across the four groups of subjects. The average waveform for each stimulus type, for each subject, comprised at least 25 individual trials.
## Waveform comparisons
Segments were average-referenced to a standard adult 128electrode montage. Epochs were baseline-corrected for the first 100 ms before the onset of the stimulus. We formed clusters of electrodes, as means of data reduction. By using clusters rather than individual electrodes we were able to cover a large portion of the scalp and still include 'electrode position' in our ANOVAs, alongside the other within-and between-subjects factors, without the ANOVAs become uninterpretable. Our clusters covered 71 electrode sites, and were organised into 12 clusters, one around each of the following electrodes: F3, Fz, F4, C3, Cz, C4, P3, Pz, P4, PO3, Oz and PO4. The specific clusters that we used can be seen in [fig_ref] Figure 2: Electrode Montage [/fig_ref].
We created four time-bins, based upon the preceding literature, and compared the mean amplitude values across each bin. These are labelled as the early bin (160-200 ms, post item onset), the mid-latency bin (320-480 ms) and the late-latency bin (480-800 ms). In each bin we used a mixed-design ANOVA. This always comprised the between-subjects factors of Exposure (two levels, Pre-exposed versus un-exposed Control groups), and Instructions (two levels, Recognition versus Categorisation groups).
The ANOVA also always included the within-subjects factor of List (two levels, List 1 vs. List 2). This enabled us to test whether List 1 and List 2 items elicited different ERPs, and whether the evoked response to the items was influenced by the participants' prior experience and judgement type. It also enabled us to test whether or not these factors interacted with one another.
In addition to these three experimental factors we included electrode cluster location in our ANOVAs, to test whether the distribution of any of the above effects differed across the conditions. Electrode cluster location was entered as two factors: cluster position along the left-to-right lateral axis (three levels, subsequently labelled electrode [L-R]) and cluster position along the anteroposterior axis (four levels, subsequently labelled electrode [A-P]).
All of our analyses were initially conducted on unscaled data. However, where this revealed an interaction between any of the experimental factors and any of the electrode factors we recalculated the ANOVA using data scaled according to McCarthy and Wood's [bib_ref] Scalp distributions of event-related potentials: an ambiguity associated with analysis of variance..., Mccarthy [/bib_ref] rescaling technique. The logic behind this was as follows: with un-scaled data the interaction between the experimental and electrode factors is necessarily ambiguous; an interaction could arise from a genuine difference in the distribution of the effects across the two conditions, or simply from the main effect having a multiplicative effect across the electrodes. Scaling results in data normalisation; with the main effect removed, one can then test for a genuine interaction between that experimental factor and electrode position. It is, however, worth noting that this technique is not perfect: Urbach and Kutas [bib_ref] The intractability of scaling scalp distributions to inter neuroelectric sources, Urbach [/bib_ref] demonstrated that this approach can fail to properly take account of the main effect, resulting in incorrectly reporting a significant interaction with electrode location; in some cases this normalisation may produce the opposite effect, masking genuine topographical differences. However, this is the most recognised means of disambiguating interactions involving electrode location, and as such we applied it to our data where necessary. That said, any topographical differences between experimental factors established using an ANOVA, either reported here or elsewhere, should be interpreted with caution. All of the results that we report are corrected using the Greenhouse-Geisser correction, to account for the potential nonsphericity of EEG data [bib_ref] The e-adjustment procedure for repeatedmeasures analyses of variance, Jennings [/bib_ref].
# Results
# Behavioural results
Responses to items that were identified as category members were treated as endorsements in the categorization condition, and responses to items that were identified as 'old' were treated as endorsements in the recognition condition. The mean proportions of endorsements for each List and Condition are shown for each condition in [fig_ref] Figure 3: Behavioural Results [/fig_ref]. These data were entered into a 26262 mixed model ANOVA with Exposure (Pre-exposed vs. Control) and Instructions (Recognition vs. Categorization) as betweensubjects factors and List (1 vs. 2) as a within-subject factor. This revealed a main effect of List (F 1, 59 = 109.08, MSE = .02, p,.01, g 2 p ,.65) indicating reliable discrimination between items. A marginal effect of Exposure (F 1, 59 = 2.87, MSE = .02, p,.09, g 2 p ,.05) and a reliable interaction between Exposure and List indicated that discrimination was higher in the Pre-exposed conditions than in the Control conditions (F 1, 59 = 26.41, MSE = .02, p,.01, g 2 p ,.31). These results clearly show that discrimination between category members and non-members, and between old and new items is greater following pre-exposure than in control participants.
However, there was no effect of Instructions (F 1, 59 = 0.18, MSE = .02, p = .67, g 2 p ,.01), and no interaction between either Instructions and List (F 1, 59 = 1.82, MSE = .02, p = .18, g 2 p = .03), or between Instructions and Exposure (F 1, 59 = 0.26, MSE = .02, p = .61, g 2 p ,.01). There was no 3-way interaction (F 1, 59 = 0.08, MSE = .02, p = .78, g 2 p ,.01). This aspect of the results is consistent with the view that when items to be recognized are the same as items to be categorized similar patterns of discrimination performance result. This is indicative of the same or similar processes being utilized to make different decisions. That is, categorization is a form of recognition.
To more closely examine discrimination in the four conditions, and to permit a power analysis, we next computed the sensitivity index d' by treating endorsements to List 1 items as hits, and endorsements to List 2 items as False Alarms. A comparison of the two pre-exposed groups showed that instructions to recognize or categorize items did not result in a reliable difference in discrimination (d' = 1.30, vs. 1.20 respectively, t 31 = 0.39, p = .70). So that we may be confident that our experiment was sufficiently powerful to detect a possible difference between recognition and categorization we estimated the effect sizes of data from two previous reported comparisons. Both of these made within subject comparisons of these decisions but used different items in each test. The first reported d9 values of 7.23 and 0.72 for recognition and categorization respectively, from a sample of 4 participants who were acting as controls against an amnesic patient [bib_ref] Learning about categories in the absence of memory, Squire [/bib_ref]. From the figures that they report we first estimated the pooled standard deviation (s 2 p = 0.80) and used this to estimate the effect size (Cohen's d = 7.27). We then calculated the sample size that we would need to find an effect of this magnitude in a between subjects design. The result was 2 participants in each group. Our sample size of 31 easily exceeds this. The second study reported a mean percentage of correct recognition decisions to be 86.0% versus, 64.2% correct categorization decisions, with a sample of 10 and 9 respectively (due to a recording error) [bib_ref] Comparing the brain areas supporting nondeclarative categorization and recognition memory, Reber [/bib_ref]. As before we estimated the pooled standard deviation from their reported figures (s 2 p = 11.81). The resulting effect size smaller than the other study but is nonetheless large (Cohen's d = 1.85). The sample size needed to find an effect of this magnitude in a between subjects design is 8 participants in each group. Our sample size also easily exceeds this. The average weighted effect size of both of these studies is Cohen's d = 3.45, and requires a total sample size of just 6 participants. We are therefore confident that had there been a difference in categorization and recognition judgements it would have been detected in our sample of 31 participants. We conclude from this that previous dissociations between categorization and recognition might reflect differences in the test items rather than differences in process.
We also examined whether the control participants were able to discriminate between items by comparing their performance against a chance value of d' = 0. There was some indication of above chance performance following recognition instructions (d' = 0.47, t 14 = 3.09, sd = .59, p,.01), but not following categorization instructions (d' = 0.32, t 14 = 1.76, sd = .71, p = .10). This suggests that some abstraction of the category structure can occur during the testing period, in the sense that untrained controls were able to make accurate decisions without any exposure to the study items [bib_ref] Learning about categories in the absence of training: Profound amnesia and the..., Palmeri [/bib_ref] , but this is insufficient to account for the substantially greater number of correct decisions made by the pre-exposed participants.
# Erp results
Although the behavioural performance was equivalent across the Recognition and Categorisation groups we were keen to Early effects (160-200 ms). Our first comparison was centred on the window most reliably associated with the N1 potential. This had previously distinguished between category members and non-members [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref]. In our data, the effect elicited by the items appeared to be more negative in the Categorisation groups than in either of the Recognition groups; more negative in the un-exposed Control groups than in the Pre-exposed groups and more negative for List 1 items than for List 2 items. Furthermore, it appeared that these factors interacted: whilst both the Pre-exposed Categorisation and Recognition groups showed a more negative N1 component for List 1 relative to List 2, of the un-exposed Control groups, only the Categorisation group showed a greater negativity for List 1 relative to List 2.
Our ANOVA revealed main effects of Instruction with the amplitudes in the Categorisation groups being more negative than in the Recognition groups (F 1, 56 = 7.40, p,.01), Exposure with the Exposed groups being more negative than the Control groups (F 1, 56 = 4.03, p,.05), and List with List 1 eliciting a greater negativity than List 2 (F 1, 56 = 5.93, p,.02). There was also a marginally significant interaction between these three factors (F 1, 56 = 3.98, p,.06). This resulted from an interaction between List and Instructions in the un-exposed Control groups only (F 1, 28 = 5.04, p = .03): there was a relative negativity for List 1 relative to List 2 in the Control Categorisation group (F 1, 14 = 4.84, p..05), but not in the Control Recognition group. By contrast, in the Pre-exposed groups there was no interaction between List and Instructions, there was just a simple main effect of List (F 1, 28 = 4.65, p = .04), with List 1 items eliciting a greater negativity relative to List 2 items.
We also observed various interactions between these experimental factors and one or both of the electrode factors. Given the main effects of the experimental factors, we scaled the data and recalculated the ANOVA. An interaction between List and electrode[A-P] survived the normalisation (F 2.53, 141.85 = 3.02, p = .04). This resulted from List 1 items being significantly more negative than List 2 items over the parietal (F 1, 59 = 5.47, p = .02) and occipital electrode clusters (F 1, 59 = 4.35, p = .04), though not over the other electrode clusters. An interaction between List, instruction and electrode[L-R] also survived the normalisation (F 1.95, 109.54 = 3.73, p,.03. This resulted from an interaction between List and instruction over the midline electrodes (F 1, 58 = 4.83, p = .03), though not over the left or right-hemisphere electrodes. This in turn resulted from a significant negativity for List 1 items relative to List 2 items in the Categorisation groups (F 1, 29 = 7.34, p = .01), though not for the Recognition groups.
Mid-latency effects (320-480 ms). There were no main effects of any of the experimental factors within this time window. However, there were two interactions between the experimental and electrode factors: prior exposure interacted significantly with electrode[L-R] (F 1.72, 96.50 = 3.87, p = .03). This was not driven by any one cluster significantly, though List 1 items elicited a marginally more negative amplitude over the midline electrodes in the Exposed groups relative to the Control groups (F 1, 56 = 3.43, p = .06). We also observed a marginally significant interaction between List, electrode[L-R] and electrode[A-P] (F 3.53, 197.72 = 2.45, p = .05). This was because List 1 items elicited a greater positivity than List 2 items over the right-hemisphere central cluster (F 1, 56 = 12.73, p,.01), but not over any other clusters.
Late-latency effects (480-800 ms). There were no main effects or interactions between any of the experimental factors. There were two significant interactions between experimental and electrode factors: Instruction, Exposure and electrode[L-R] interacted significantly (F 1.55, 86.83 = 3.67, p = .04). There was no clear effect driving this interaction, though the closest to reaching significance was an interaction between Exposure and electrode[L-R] in the Categorisation groups (F 1.37, 38.33 = 3.44, p = .06), which had not been present in the Recognition groups. This marginal effect was, in turn, driven by relatively more negative amplitudes for the Exposed, relative to the un-exposed Control group, over the left-hemisphere clusters (F 1, 28 = 3.96, p = .05). We also noted a significant interaction between List, electrode[L-R] and electrode[A-P] (F 3.94, 220.87 = 5.62, p,.01). This was the result of a significant negativity over the left frontal cluster (F 1, 58 = 5.45, p = .02), and a significant positivity over the right frontal (F 1, 58 = 7.24, p,.01) and central (F 1, 58 = 7.31, p,.01) clusters, for List 1 items relative to List 2 items.
# Discussion
The aim of this experiment was to examine the ERPs of categorization and recognition in the well-known prototypedistortion task [bib_ref] Perceived distance and the classification of distorted patterns, Posner [/bib_ref]. Previous studies have used either amnesic patients [bib_ref] The learning of categories: Parallel brain systems for item memory and category..., Knowlton [/bib_ref] or fMRI [bib_ref] Dissociating explicit and implicit category knowledge with fMRI, Reber [/bib_ref] to dissociate categorization from recognition. To our knowledge this is the first to use ERP to do so, although one previous study has used a less well known paradigm to examine the same processes [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref]. These previous studies have compared recognition judgements made to one set of items to categorization judgements made to a different set of items. The resulting differences are frequently cited as evidence that prototype knowledge is used to make categorization decisions using a separate process than episodic memory of the study items that is used to make recognition judgements. This particular experimental preparation makes the interpretation of dissociations involving amnesic patients or fMRI difficult because they may be due, as is claimed, to the processes involved, or to differences in the stimuli themselves. In this experiment we sought to resolve this problem by asking participants to make recognition and categorization judgements to the same set of stimuli. The participants were allocated to four groups were presented with the same category members and non-members (termed 'List 1' and 'List 2' items, respectively). A key finding was that participants were equally able to distinguish List 1 from List 2 items, regardless of whether they had been asked to attempt to categorise or recognise them. This is in contrast to previous studies that have compared these decisions to different items [bib_ref] Learning about categories in the absence of memory, Squire [/bib_ref] [bib_ref] Comparing the brain areas supporting nondeclarative categorization and recognition memory, Reber [/bib_ref]. We did not replicate these effects despite sufficient experimental power to do and so conclude that the dissociations reported previously could be due to differences in the stimuli rather than difference in the processes used to judge them. Nonetheless, previous studies have reported activity in different brain regions dependent on the decision that participants are asked to make, and although we observed no differences in the decisions that the participants made this does not preclude differences in the neural processes used to make those decisions. Indeed the ERP data from our experiment demonstrated that the early visual evoked response that distinguished category members and non-members was modulated by whether participants had been asked to categorise or recognise them, and, although not to the same extent, by whether participants had previously been exposed to those items.
Influence of instruction and prior exposure on early visual processing of items. In the window usually associated with the N1 potential, List 1 elicited a significantly more negative potential than did List 2 items. This was most prominent over the parietal and occipital electrodes. However, unlike previous research [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref] , we found that the N1 amplitude was modulated by more than just category membership. Whilst the early effect of List was not affected by judgement for both the Pre-exposed Recognition and Categorisation groups, the effect was only present for the Control Categorisation group. The Control Recognition group showed no such early differentiation between List 1 and List 2 items. This result overlapped with another result: taken together the Categorisation groups showed an increased negativity for List 1 relative to List 2 items over the midline electrodes, whereas the Recognition groups did not, presumably because whilst both Categorisation groups showed the effect, only the Pre-exposed Recognition group did. Effects driven primarily by changes to the N1 amplitude are typically ascribed to early visual processing in the extrastriate cortex [bib_ref] The visual N1 component as an index of a discrimination process, Vogel [/bib_ref]. Indeed, recent fMRI studies have linked early visual processing with prototype abstraction/application [bib_ref] Dissociating explicit and implicit category knowledge with fMRI, Reber [/bib_ref] [bib_ref] Contrasting cortical activity associated with category memory and recognition memory, Reber [/bib_ref]. Interestingly, we observed these effects for both Categorisation groups, regardless of whether or not they had previously been exposed to the category members. Possibly, when the participants' task is to categorise items, prototype abstraction/application can occur within the test phase. This would explain why even the un-exposed Control Categorisation group showed the N1 amplitude differentiation for List 1 and List 2 items. By contrast, this might not be the case when the participants' task is to attempt to recognise the items. In this case, participants only engage in prototype abstraction/application when they have already had some experience of category members. This account makes intuitive sense: when the task is to categorise abstract items based upon arbitrary but consistent perceptual characteristics, the participant will pay close attention to those consistent characteristics that distinguish category members from nonmembers -this is the case even these have not been encountered previously. By contrast, when the task is to 'recognise' abstract items, the participant might only proceed with prototype abstraction/ application when it becomes apparent that those membershipdefining characteristics discriminate between the items that were studied earlier form those that were not. That is, when attempting to recognise the items, participants will only engage in this process when they have already been exposed to the category members. The Control group might simply not realise the prototype abstraction/ application is beneficial for their 'recognition' task.
When one observes an effect that can occur in both Pre-exposed Categorisation and Recognition groups, it might be tempting to take this as evidence that both tasks tap some common recognition-like mechanism. However, in the case of this early N1 effect, this would not explain why the Control Categorisation group also show the effect: they cannot be recognising items that they have not seen before. By contrast, both the Pre-exposed Categorisation and Recognition groups could be employing a categorisation-like strategy. With our stimuli and procedure such a strategy would be successful for both of the Pre-exposed groups, but only for the Control Categorisation group. The Control Recognition group, in contrast to the Pre-exposed Recognition group, would have no category for ''old'' items and therefore would not be able to use this strategy. Interestingly, this is precisely the pattern of N1-like effects that we observed.
Influence of instruction and prior exposure on later potentials. The effects elicited by List 1 and List 2 items also differed later in the epoch: there was a greater positivity for List 1 relative to List 2 items over the right-hemisphere central electrode cluster, between 320 and 480 ms. Again, later in the epoch, between 480 and 800 ms, List 1 items elicited a negativity over the left-hemisphere and positivity over the right-hemisphere, relative to List 2 items. Whilst the time window corresponds to the differences reported in previous papers [bib_ref] An electophysiological comparison of visual categorization and recognition memory, Curran [/bib_ref] , the effects themselves are quite different in terms of distribution and amplitude in our data. Unlike the early N1-like effects, the later effects were primarily driven by differences between List 1 and List 2 items, regardless of which judgement participants performed and their prior exposure to category members. At least with regard to these later effects, they appear to reflect some process common to categorisation and recognition, a conclusion perhaps supported by the similarity in behavioural performance between the two judgement types.
# Conclusions
Participants were equally good at distinguishing category members from non-members, regardless of whether they were performing a categorisation or recognition judgement. This result contrasts shapely with previous studies that have reported differences between categorization and recognition judgements. However, the ERPs suggested that participants' early visual potentials (160-200 ms), often associated with prototype abstraction/application, distinguished category members from nonmembers in both the Pre-exposed Recognition and Categorisation groups. By contrast, in the un-exposed Control groups, only the participants explicitly asked to categorise the items showed this early visual differentiation of members and non-members -the un-exposed Control Recognition group did not. One possible interpretation of these data is that prototype abstraction/ application occurs on both categorisation and recognition tasks, but only when participants have actually been pre-exposed to category members. If they have not been pre-exposed then prototype abstraction/application will only occur in a categorisation task. The data suggest that both categorization and recognition in prototype distortion tasks appear to rely on the same underlying process.
[fig] Figure 1: Example Stimuli. doi:10.1371/journal.pone.0010116.g001 [/fig]
[fig] Figure 2: Electrode Montage. doi:10.1371/journal.pone.0010116.g002 [/fig]
[fig] Figure 3: Behavioural Results. Showing the proportions of endorsements for each list for recognition and categorization conditions by exposure. Error bars are +/2 SEM. doi:10.1371/journal.pone.0010116.g003 establish whether the ERP data shown in Figures 4 and 5 could distinguish between categorisation and recognition judgements, with and without prior exposure. [/fig]
[fig] Figure 4: ERPs potentials in the Recognition conditions. doi:10.1371/journal.pone.0010116.g004 [/fig]
[fig] Figure 5: ERPs potentials in the Categorization conditions. doi:10.1371/journal.pone.0010116.g005 [/fig]
|
TripNet: A Method for Constructing Rooted Phylogenetic Networks from Rooted Triplets
The problem of constructing an optimal rooted phylogenetic network from an arbitrary set of rooted triplets is an NP-hard problem. In this paper, we present a heuristic algorithm called TripNet, which tries to construct a rooted phylogenetic network with the minimum number of reticulation nodes from an arbitrary set of rooted triplets. Despite of current methods that work for dense set of rooted triplets, a key innovation is the applicability of TripNet to non-dense set of rooted triplets. We prove some theorems to clarify the performance of the algorithm. To demonstrate the efficiency of TripNet, we compared TripNet with SIMPLISTIC. It is the only available software which has the ability to return some rooted phylogenetic network consistent with a given dense set of rooted triplets. But the results show that for complex networks with high levels, the SIMPLISTIC running time increased abruptly. However in all cases TripNet outputs an appropriate rooted phylogenetic network in an acceptable time. Also we tetsed TripNet on the Yeast data. The results show that Both TripNet and optimal networks have the same clustering and TripNet produced a level-3 network which contains only one more reticulation node than the optimal network.
# Introduction
Phylogenetic networks are a generalization of phylogenetic trees that permit the representation of non-tree-like underlying histories. A rooted phylogenetic network is a rooted directed acyclic graph in which no node has indegree greater than 2 and the outdegree of each node with indegree 2 is 1. Such nodes are called reticulation nodes. In rooted phylogenetic networks the nodes with indegree 1 and outdegree 0 are called leaves and are distinctly labeled by a set of given taxa. Mathematicians are interested in developing methods that infer a phylogenetic tree or network from basic building blocks. In the computation of a rooted tree or network, one group of the basic building blocks are rooted triplets, the rooted binary trees on three taxa.
In 1981, Aho et al., studied the problem of constructing a rooted tree from a set of rooted triplets [bib_ref] Inferring a tree from lowest common ancestors with an application to the..., Aho [/bib_ref]. They proposed an algorithm called BUILD algorithm which shows that, given a set of rooted triplets, it is possible to construct in polynomial time a rooted tree that all the input triplets are contained in it or decide that no such tree exists.
When there is no tree for a given set of triplets one may try to produce an optimal phylogenetic network. In this context, the goal is to compute an optimal rooted phylogenetic network that contains all the rooted triplets. One possible optimality criterion is to minimize the level of the network, which is defined as the maximum number of reticulation nodes contained in any biconnected component of the network. The other optimality criterion is to minimize the number of reticulation nodes. In [bib_ref] Algorithms For combining rooted triplets into a galled phylogenetic network, Jansson [/bib_ref] and [bib_ref] Inferring a Level-1 Phylogenetic Network from a Dense Set of Rooted Triplets, Jansson [/bib_ref] the authors considered the problem of deciding whether, given a set of rooted triplets as input, is it possible to construct a level-1 rooted phylogenetic network that contains all the input triplets? They showed that, in general, this problem is NP-hard. However, in [bib_ref] Inferring a Level-1 Phylogenetic Network from a Dense Set of Rooted Triplets, Jansson [/bib_ref] the authors showed that when the set of rooted triplets is dense, which means that for each set of three taxa there is at least one rooted triplet in the input set, the problem can be solved in polynomial time. After their results, all research in this new area has up to this point focused on constructing rooted phylogenetic networks from dense rooted triplet sets.
LEV1ATHAN is an algorithm for generating a level-1 rooted phylogenetic network from a set of rooted triplets [bib_ref] A Practical Algorithm for Reconstructing Level-1 Phylogenetic Networks, Huber [/bib_ref]. Specifically, it attempts to find a level-1 rooted phylogenetic network that contains as many of the input rooted triplets as possible. This problem is an NP-hard problem [bib_ref] A Practical Algorithm for Reconstructing Level-1 Phylogenetic Networks, Huber [/bib_ref]. The algorithm by [bib_ref] Constructing the simplest possible phylogenetic network from triplets, Van Iersel [/bib_ref] can be used to find a level-1 or a level-2 rooted phylogenetic network which minimizes the number of reticulation nodes, if such a network exists. In [bib_ref] Constructing the simplest possible phylogenetic network from triplets, Van Iersel [/bib_ref] the authors also showed that for a dense set of rooted triplets t, if t is precisely equal to the set of rooted triplets that are contained in some rooted phylogenetic network, then they can construct such a rooted phylogenetic network with smallest possible level in time O(|t| k+1 ), where k is a fixed upper bound on the level of the network. In addition based on the ideas described in [bib_ref] Constructing the simplest possible phylogenetic network from triplets, Van Iersel [/bib_ref] , for a given dense set of rooted triplets t, the authors proposed the SIMPLISTIC algorithm which always returns some rooted phylogenetic network that contains t. But it does not give any minimality guarantees.
In [bib_ref] Level-k Phylogenetic Networks are Constructable from a Dense Triplet Set in Polynomial..., To [/bib_ref] the authors showed that given a dense set of rooted triplets t and a fixed number k, it is possible to construct in time O(|t | k+1 ) a level-k rooted phylogenetic network that contains t or decides that no such network exists.
In this paper we present a heuristic algorithm called TripNet for constructing rooted phylogenetic networks with the minimum number of reticulation nodes from an arbitrary set of rooted triplets. Despite of current methods that work for dense set of rooted triplets, a key innovation is the applicability of TripNet to non-dense set of rooted triplets.
In ''unpublished data'' the authors applied TripNet on both real and simulated data. Here TripNet algorithm is described in details, some theorems are proved, and one simulation is performed to show the accuracy of TripNet. Also TripNet is tested on the Yeast data. This paper is organized as follows. In section 2, first some definitions and notation are presented. Then we describe BUILD algorithm. Finally a new method called TCD, is introduced for constructing rooted triplets from (biological) sequences. In section 3 we compare TripNet with SIMPLISTIC on the triplets sets that are obtained from TCD method. Then we test TripNet on the Yeast data. In section 4 we discuss the performance of TripNet. In the last section the directed graph Gt related to a set of triplets t is introduced. Then we show that if either a set of triplets is obtained from a set of sequences using TCD method or a set of triplets is consistent with a tree, then G t is a DAG. This property has a key role in solving the Integer Programming system which is introduced in the remaining, in polynomial time. Then the concept of the height function of a rooted phylogenetic network is introduced, and an efficient method for obtaining a height function h t for a given set of rooted triplets t is explained. It is shown that the condition of consistency of a rooted phylogenetic network N with the height function h t can be a good alternative for the condition of consistency of N with t. To show this, firstly we define the Integer Programming system in such a way that its constraints intuitively force the consistency of N with t. Secondly, we show that if t is consistent with a tree T, then T is consistent with h t and T can be constructed using this height function. In the last section we present TripNet algorithm.
## Preliminaries
Here first we present some definitions and notation. Then we describe BUILD algorithm. Finally a new method called TCD, is introduced for constructing rooted triplets from a set of sequences.
## Definitions and notation
Let X be a set of taxa. A rooted phylogenetic tree (tree for short) on X is a rooted unordered leaf labeled tree whose leaves are distinctly labeled by X and every node which is not a leaf has at least outdegree two. A directed acyclic graph (DAG) is a directed graph that is free of directed cycles. A DAG G is connected if there is an undirected path between any two nodes of G. It is biconnected if it contains no node whose removal disconnects G. A biconnected component of a graph G is a maximal biconnected subgraph of G. A rooted phylogenetic network (network for short) on X is a rooted DAG in which the root has indegree 0 and outdegree 2 and every node except the root satisfies one of the following conditions:
a) It has indegree 2 and outdegree 1. These nodes are called reticulation nodes. b) It has indegree 1 and outdegree 2. c) It has indegree 1 and outdegree 0. These nodes are called leaves and are distinctly labeled by X.
A reticulation leaf is a leaf whose parent is a reticulation node. A network is said to be a level-k network if each of its biconnected components contains at most k reticulation nodes. A tree can be considered as a level-0 network.
A rooted triplet (triplet for short) is a rooted binary unordered tree with three leaves. We use ij|k to denote a triplet with taxa i and j on one side and k on the other side of the root [fig_ref] Figure 1: A triplet and a network consistent with it [/fig_ref]. A set of triplets t is called dense if for each subset of three taxa, there is at least one triplet in t. A triplet ij|k is consistent with a network N or equivalently N is consistent with ij|k if the leaf set of ij|k is a subset of the leaf set of N, and N contains a subdivision of ij|k, i.e. if N contains distinct nodes u and v and pairwise internally nodedisjoint paths u R i, u R j, v R u and v R k. [fig_ref] Figure 1: A triplet and a network consistent with it [/fig_ref] shows an example of a network consistent with ij|k. A set t of triplets is consistent with a network N if all the triplets in t are consistent with N. We use the symbols t(N) and L N to represent the set of all triplets that are consistent with N and the set of labels of its leaves respectively. For any set t of triplets define L(t) = | t[t L t . The set t is called a set of triplets on X if L(t) = X.
## Build algorithm
Let t be a set of triplets. BUILD is a top-down algorithm, constructs a tree consistent with t if such a tree exists. The algorithm is guided by the Aho graph. Definition 1. (Aho graph) Let X be a set of taxa and t be a set of triples on X. The Aho graph AG(t) = (V,E) associated with t has node set V = X and any two nodes i and j are connected by an edge in E if and only if there exists a triplet ij|k M t.
BUILD algorithm: Given a non-empty set of rooted triples t on X, the aim is to construct a rooted phylogenetic tree T on X that is consistent with t, if one exists. If AG(t) has only one connected component, then the algorithm reports fail. Else, for each node set U of a connected component of AG(t), determine the set t| U which denotes the set of all triplets in t whose leaves are in U and recursively compute the rooted phylogenetic subtree T(t| U ) which denotes the tree constructed with BUILD algortihm consistent with t| U . Finally, create a root node r and combine all computed subtrees by connecting r to the root of each of them. For an example see [fig_ref] Figure 2: An example of BUILD algorithm for the given set {bc |a, ac... [/fig_ref].
# Triplets construction method
There exist different methods like Maximum Parsimony or Maximum Likelihood for constructing triplets from (biological) sequences [bib_ref] Constructing the simplest possible phylogenetic network from triplets, Van Iersel [/bib_ref]. In this section a method for constructing triplets is presented. Suppose that X is a set of n taxa, and D = [D ij ] be an n6n distance matrix on X. For each three taxa i, j, and k M X, and the entries D ij , D ik , and D jk , we assign the triplet ij|k if D ij , min {D ik , D jk }. We name this method Triplets Construction with Distance; TCD for short. In this paper we use TCD method for constructing triplets.
# Results
In this section to show the performance of TripNet on the triplets sets which are obtained from TCD, we compare TripNet with SIMPLISTIC. Also we test TripNet on the Yeast data. It is the only published triplets data that are obtained from biological data.
## Comparing simplistic and tripnet
SIMPLISTIC is the only available software which has the ability to return some rooted phylogenetic network consistent with a given dense set of rooted triplets. But it does not give any minimality guarantees [bib_ref] Constructing the simplest possible phylogenetic network from triplets, Van Iersel [/bib_ref].
SplitsTree is a valuable tool for constructing an special kind of unrooted phylogenetic networks from different types of data as input. This program converts a given set of sequences X into a distance matrix D X to compute the resulting network. The distance matrix D X is reported as one of the output of SplitsTree [bib_ref] SplitsTree: analyzing and visualizing evolutionary data, Huson [/bib_ref].
Let t DX be the set of triplets that is obtained from D X using TCD, and consider it as the input for TripNet. Note that t DX is not necessarily dense, since for some three taxa i, j, and k we might have D Xij = D X jk ,D X ik . In this case one of the triplets ij|k or jk|i is assigned to i, j, and k to obtain a dense set of triplets t X dense as the input of SIMPLISTIC. Also if D Xij = D Xjk = D Xik , then randomly one of the three possible triplets related to i, j and k is assigned to them.
To perform the simulation we generate 160 different sets of sequences are generated using TREEVOLVE. TREVOLVE is a software which simulate the evolution of DNA sequences under a coalescent model. TREEVOLVE contains many input parameters which one can adjust them. In this study we adjust the Number of samples, the Number of sequences, and the Length of sequence, and for the other parameters the default values are adjusted. In this study the Number of sequences is 10, 20, 30, and 40. For each input parameter the Number of sequences the Length of sequence is 100, 200, 300, and 400. For each case the Number of samples is set to [bib_ref] Constructing level-2 phylogenetic networks from triplets, Van Iersel [/bib_ref].
In this study we run both methods on a PC with an Intel DuallCore processor running at 1.80 GHz. We set the running time restriction 6 hours for methods. Let N finite be the set of networks for which the running time is less than 6 hours.
The results of the comparison between TripNet and SIMPLIS-TIC on the three most important parameters i.e. running time of both methods, number of the reticulation nodes and the level of the final networks, are shown in [fig_ref] Table 1: SIMPLISTIC and TripNet network results [/fig_ref].
The results show that when the number of input taxa is 10, both methods always return a network in at most one second. For the number of input 20, in 5% of cases SIMPLISTIC returns no results in less than 6 hours. For the remaining 95% of the cases, the SIMPLISTIC running time is on average 306 seconds, while in all cases on average the TripNet running time is at most 2 seconds. But by increasing this parameter to 30, in 67.5% of the cases, SIMPLISTIC has not the ability to return a network in less Moreover when this parameter is set to 40, in all cases SIMPLISTIC fails to return any network in less than 6 hours, while on average TripNet outputs a network in 775 seconds. Totally for all 160 input triplets sets on average TripNet outputs a network in less than 250 seconds, while on average in 57% of the SIMPLISTIC networks which belong to N finite , the running time is near to 750 seconds.
Also the results show that in all cases the number of the reticulation nodes and the level of TripNet networks are less than SIMPLISTIC networks. Note that for the number of input 40, on average the number of the reticulation nodes and the level of the TripNet networks are 15.825 and 15.25, while for these data SIMPLISTIC can not return any network in less than 6 hours.
## Yeast data
The Yeast data is a dense set of triplets generated using real yeast data, obtained from the Fungal Biodiversity Center in Utrecht. This data set which contains information about 21 species is available online from (http://skelk.sdf-eu.org/level2triplets. html). Based on the algorithm developed in [bib_ref] Constructing level-2 phylogenetic networks from triplets, Van Iersel [/bib_ref]. Steven Kelk has developed a software application, called LEVEL2, for constructing level-2 networks from dense sets of triplets. LEVEL2 is not applicable to general triplet sets and it produces a network only if there exists a level-2 network consistent with the input triplets. However, LEVEL2 has the advantage that it always produces the best possible network which also minimizes the number of reticulation nodes. LEVEL2 network for the Yeast data is a 21-leaf level-2 network which is given in [fig_ref] Figure 3: Resulting networks from Yeast triplets [/fig_ref] [bib_ref] Constructing level-2 phylogenetic networks from triplets, Van Iersel [/bib_ref]. As our only chance for comparing TripNet networks with the best possible networks we repeated the analysis of Yeast data using TripNet. The TripNet network for the Yeast dataset is given in [fig_ref] Figure 3: Resulting networks from Yeast triplets [/fig_ref]. As one can see, TripNet produced a level-3 network which contains only one more reticulation node than the network obtained by LEVEL2. The running time of both algorithms is nearly one second.
# Discussion
In this paper we introduced TripNet which is the software that has the ability to return some network consistent with an arbitrary given set of triplets.TripNet and supplementary files are freely available for download at (www.bioinf.cs.ipm.ir/software/tripnet). Unlike previous methods which only work on dense triplet sets, our method works on any set of triplets. Some theorems were proved to clarify the rationale behind the steps of TripNet. In this paper the TCD method was introduced for constructing triplets. In order to study the performance of TripNet on the triplets that are obtained from TCD method we performed a simulation on 160 different sets of triplets, and compared TripNet with SIMPLISTIC.
The results showed that in all 160 cases TripNet outputs an appropriate network in an acceptable time, while just in 57.5% of these cases SIMPLISTIC has the ability to return some network in less than 6 hours. Also on average in all cases TripNet outperforms SIMPLISTIC on the number of the reticulation nodes, and the level of the output network.
Also by increasing the number of input taxa, the running time of SIMPLISTIC exceeds abruptly, such that for the input taxa 40, it could not return any network in less than 6 hours.
These results showed that for large size input data that are obtained from TCD method, SIMPLISTIC is not a practical method for constructing networks, while TripNet works well in all cases.
To establish the performance of TripNet on real datasets, we tested TripNet on Yeast data, and compared our results with those of LEVEL2. For Yeast data TripNet produced a level-3 network which contains only one more reticulation node than the optimal network obtained by LEVEL2. Both networks have the same clustering and represent the same evolutionary relationship between taxa. While TripNet has been designed for general triplet sets (not necessarily dense or consistent with a restricted The steps i to n shows that l is the reticulation leaf. In these steps criterion III is applied. doi:10.1371/journal.pone.0106531.g008 . Steps of TripNet for input triplets: jk |i, li |j, mj |i, jn|i, kl |i, ik |m, ik |n, lm|i, ln|i, mn|i, kl |j, km|j, jn|k, lm|j, jl |n, mn|j, kl |m, kl |n, mn|k, mn|l }. doi:10.1371/journal.pone.0106531.g009
TripNet: A Method for Constructing Networks from Triplets PLOS ONE | www.plosone.org level network), this example shows that the network produced by TripNet is very close to the best possible solution.
# Materials and methods
In this section we prove some theorems to clarify the rationale behind the steps of TripNet. Then TripNet is presented in nine steps.
## The directed graph related to a set of triplets and height function
Throughout this subsection we denote i, j by ij for short. Let t be a set of triplets. Define G t , the directed graph related to t, by
[formula] V(G t ) = {ij: i,j M L(t), i ? j} and E(G t ) = {(ij,ik): ij|k M t} < {(ij,jk): ij|k M t}. [/formula]
In the following we present some basic properties of G t .
In what follows the height function of a tree is introduced. Let
[formula] ( X 2 [/formula]
) denotes the set of all subsets of X of size 2.
Definition 2. Let X be an arbitrary finite set. A function h:
[formula] ( X 2 ) R is called a height function on X. [/formula]
Let T be a rooted tree with the root r, c ij be the lowest common ancestor of the leaves i and j, and l T denotes the length of a longest path starting at r. Definition 3. The height function of T, h T is defined as h T (i,j) = l T -d T (r,c ij ) where i and j are two distinct leaves of T (d T (r,c ij ) denotes the length of the path between r and c ij ).
Let T be a tree. The definition above implies that a triplet ij|k is consistent with T if and only if h T (i, j),h T (i, k) or h T (i, j),h T (j, k).
Let X = {x 1 , x 2 , …, x m } be a finite set, D be a distance matrix on X, and t be the set of triplets on X that are obtained from TCD method using D. Let G t contains a cycle x 1 x 2 R x 2 x 3 R … R x n21 x n R x 1 x 2 . Then D x1x2 vD x2x3 v:::vD xn{1xn vD x1x2 , which is a contradiction. So G t is a DAG.
Moreover if t is a triplet set consistent with a tree T, then G t is a DAG. This is so because if G t contains a cycle
[formula] x 1 x 2 R x 2 x 3 R … R x n21 x n R x 1 x 2 , then h T (x 1 ,x 2 ) , h T (x 2 ,x 3 ) , … , h T (x n21 ,x n ) [/formula]
, h T (x 1 ,x 2 ), which is a contradiction.
The height function of a DAG is introduced as what follows. Let t be a set of triplets, G t be a DAG and l Gt denotes the length of the longest path in G t . Since G t is a DAG, the set of nodes with outdegree zero is nonempty. Assign l Gt +1 to the nodes with outdegree zero and remove them from G t . Assign l Gt to the nodes with outdegree zero in the resulting graph and continue this procedure until all nodes are removed.
Definition 4. For any two distinct i, j M L(t), define h Gt (i,j) as the value that is assigned by the above procedure to the node ij and call it the height function related to G t .
Let t be a set of triplets that is consistent with a tree, and T t denotes the unique tree that is produced by BUILD algorithm. Then G t is a DAG and h Gt is well-defined. The following theorem represents an upper bound for h Tt based on h Gt . Theorem 1. Let t be a set of triplets that is consistent with a tree. Then h Tt ƒ h Gt .
Proof. The proof proceeds by induction on DL Tt D. It is trivial when DL Tt D = 3. Assume that theorem holds when DL Tt Dƒk. Let DL Tt D = k+1 and T 1 , T 2 , …, T m be m subtrees which are obtained from T t by removing its root. For each i, 1ƒiƒm, let t i~t D LT i , and r i be the root of T i . By the induction assumption for each i, 1ƒiƒm,h Tt i #h Gt i . Moreover we conclude from BUILD algorithm that T i~Tti , for 1ƒiƒm. Thus h Ti ƒh Gt i , for 1ƒiƒm. Also for i, 1ƒiƒm , the maximum length of the longest path in T i is l Tt {1. It means that for i, 1ƒiƒm, the maximum length of the longest path in G ti is at least l Tt {2. Therefore the length of the longest path in G t is at least l Tt {1. Let a, b[L Tt . We have two cases. Case 1. For some i and j, 1ƒivjƒm, a[L Ti and b[L Tj . Since the outdegree of ab in G t is zero and c ab = r, then h Tt (a,b)~l Tt # h Gt (a,b).
Case 2. For some i, 1ƒiƒm, a,b [L Ti . By the induction assumption h Tt i (a,b)#h Gt i (a,b) for i, 1ƒiƒm. Therefore h Tt (a,b) . The last inequality is obtained by construction of G t from G ti for i, 1ƒiƒm.
[formula] l Tt {d Tt (r,c ab ) = l Tt {(d Tt i (r i ,c ab )z1) = (l Tt {l Tt i {1)z(l Tt i { d Tt i (r i ,c ab )) = (l Tt {l Tt i {1)zh Tt i (a,b)#(l Tt {l Tt i {1)z h Gt i (a,b) #h [/formula]
So for each a,b [L Tt , h Tt (a,b)#h Gt (a,b) and the proof is complete. Now we describe an algorithm similar to BUILD algorithm, using height functions. We refer to this algorithm by HBUILD. Let h be a height function on X. Define a weighted complete graph (G,h) where V(G) = X and edge {i, j} has weight h(i,j). Remove the edges with maximum weight from G. If removing these edges results in a connected graph the algorithm stops. Otherwise, the process of removing the edges with maximum weight is continued in each connected component until each connected component contains only one node. At the end of this procedure one can reconstruct the tree by reversing the steps of the algorithm similar to BUILD algorithm (see [fig_ref] Figure 4: The steps of constructing T t from the given set t =... [/fig_ref]. The algorithm above decides in polynomial time whether a tree with height function h exists.
So if t is a set of triplets which is consistent with a tree, then G t is a DAG and h Tt (a,b)#h Gt (a,b) = h and HBUILD algorithm constructsa tree consistent with t. Note that based on theorem 1 the tree that is produced by HBUILD is exactly T t.
The HBUILD tree is not necessarily a binary tree. To obtain a binary tree consistent with a set of triplets, we do the following procedure.
Let T be a tree and x be a node of T with x 1 , x 2 , …, x k , k §3 as its children. Consider a new node y. Construct T 0 by removing the edges (x, x 1 ), (x, x 2 ), …, (x, x k-1 ) from T and adding the edges (x, y), (y, x 1 ), (y, x 2 ), …, (y, x k-1 ) to T. Continuing the same method for each node with outdegree more than 2 a binary tree is obtained, and call it a binarization of T (see [fig_ref] Figure 5: An example of binarization [/fig_ref]. Obviously, one can obtain different binarization of T. Let t be a set of triplets that is consistent with a tree T 1 , and T 2 be a binarization of T 1 . Then t is consistent with T 2 .
In the remaining of this section we generalize the concept of height function from trees to networks. This generalization is not straightforward because the concept of (lowest) common ancestor of two leaves of a network is not well-defined. Let N be a network with the root r and l N be the length of a longest directed path from r to the leaves. For each node u consider d(r,u) as the length of the longest directed path from r to u. For any two nodes u and v, we call u an ancestor of v, if there is a directed path from u to v. If u is an ancestor of v then we say that v is lower than u. Let i and j be two leaves of N. c is called a lowest common ancestor of i and j in N, if c is a common ancestor of i and j and there is no common ancestor of i and j lower than c. For any two leaves i and j, let C ij denote the set of all lowest common ancestors of i and j.
Definition 5. For each pair of leaves i and j, define h N (i,j) = min{l N -d(r,c): cMC ij } and call it the height function of N.
Obviously, every network N indicates a unique height function h N . But two different networks may have the same height function (see [fig_ref] Figure 6: Two different networks with the same height function [/fig_ref].
In the following proposition we prove that for a given height function h there is a network N such that h N = h+1. Proposition 1. Let X be an arbitrary finite set and h be a height function on X. Then there exists a network N not necessarily binary, such that its leaves are distinctly labeled by X and h N = h+1.
Proof. Let X = {x 1 , x 2 , …, x n } and h max = max{h(x i , x j ): 1ƒi,jƒn}. Let r be the root of N,, and X9 = {x9 1 , x9 2 , …, x9 n }. Consider n nodes that are distinctly labeled by X9 members. For each pair of nodes x i and x j with h(x i , x j ) = h max , connect x9 i and x9 j to r by two paths of length h max which just are common in the root. For each pair of nodes x i and x j with h(x i , x j ) , h max , consider a new node and connect x9 i and x9 j to this new node and connect this node to r by a path of length h max -h(x i , x j ). For each node which is labeled by x9 i , consider a new node as its child and label it by x i . The resulting network in which its leaves are distinctly labeled by X satisfies the condition h N = h+1.
Note that the network N which is constructed in the proof of Proposition 1 is not necessarily a rooted phylogenetic network. To construct a rooted phylogenetic network N9 from N in such a way that if a triplet is consistent with N then it is consistent with N9, do the following procedure. Replace each path in which all its inner nodes have indegree and outdegree one, with a path of length one. The method of constructing N shows that If there is a node v with indegree d §2, then it has just one child as a leaf. Let this child is labeled by x, d §3 and its d parents are labeled by x 1 , x 2 , …, x d . Replace the edge which is connected to x with a path of length d-2 in such a way that its d-2 inner nodes from v to x are labeled with 1 to d-2. For each i, 1ƒiƒd{2 remove the edge x i v and connect x i to i. Do the binarization on the root. The resulting network N9 is consistent with all triplets which are consistent with N.
The following theorem shows relation between the height function of a network and a triplet consistent with it. Theorem 2. Let N be a network, i, j, and k be its three distinct
[formula] leaves. If h N (i, j) , h N (i, k) or h N (i, j) , h N (j, k) then ij|k is consistent with N. [/formula]
Proof. Suppose that h N (i, j) , h N (i, k). Let v ij $ and v ik be common ancestors of i, j and i, k respectively, such that h N (i, j) = l N -d(v ij , r) and h N (i, k) = l N -d(r,v ik ). Let l i and l j be two distinct paths from v ij to i and j, respectively. Let l k be an arbitrary path from v ik to k. If l i \l k =1 then it follows that h N (i,j) §h N (i,k) which is a contradiction. So ij|k is consistent with N.
The reverse of the above theorem is not necessarily true. For example, consider the network of [fig_ref] Figure 7: A counter example for the reverse of Theorem 2 [/fig_ref]. The triplet ij|k is consistent with it, but h(i,j) = h(i,k) = 3 and h(j,k) = 2.
The basic idea of TripNet algorithm is to find a height function as an intermediate computational step that yields the minimum amount of information required to construct the network from a set of triplets. So it is important to find a way for computing h N from a set of triplets. In the rest of this section we introduce a computational method for computing h N using Integer Programming. Let t be a set of triplets with |L(t)| = n. Inspired from the two inequalities that are the consequence of Definition 3 and Theorem 2, for each triplet ij|k M t, define two inequalities h(i,k){h(i,j) §1 and h(j,k){h(i,j) §1. Since the number of variables in such inequalities are at most D( L(t) 2 )D, we obtain the following system of inequalities from t.
[formula] h(i,k){h(i,j) §1 ij D k[t, h(j,k){h(i,j) §1 ij D k[t, [/formula]
0vh(i,j)ƒ D( L(t)
2 )D 1ƒi,jƒn:
Let s be an integer. Define the following Integer Programming and call it IP(t,s).
Maximize P 1ƒi,jƒn h(i,j),
[formula] Subject to : h(i,k){h(i,j) §1 ijDk[t, h(j,k){h(i,j) §1 ijDk[t, [/formula]
0vh(i,j)ƒs 1ƒi, jƒn:
Intuitively if IP(t,s) has a feasible solution, we expect that the optimal solution to this integer programming is an approximation of the height function of an optimal network N consistent with t.
The following theorems support this intuition. Theorem 3. Let t be a set of triplets. Then G t is a DAG if and only if for some integer s, the IP(t,s) has a feasible solution. In this case the minimum number s, for which IP(t,s) has a feasible solution, is l Gt +1.
Proof. Let G t be a DAG. Without loss of generality assume that G t is connected.
The proof proceeds by induction on l Gt . If l Gt = 1 then obviously for s = 1, IP(t,s) has no feasible solution and for each s §2, IP(t,s) has a feasible solution. Assume that the theorem holds for l Gt ƒk. Suppose that t is a set of triplets with l Gt = k+1. Let A be the set of the terminal nodes of all longest paths in G t . For each ij M A there is some x M L(t) such that ix|j M t. Let B be the set of all such triplets and t9 = t\B. Apparently, B?w and the length of the longest path in G t 9 is k. By the induction assumption the minimum number s for which IP(t9,s) has a feasible solution, is l G t 0 +1 = l Gt . Consider IP(t,l Gt +1). Define h(i, j) = l Gt +1, for each ij M A and h(t,l) = h9(t,l), for each tl = [ A. h is a feasible solution to IP(t,l Gt +1). Now if s is a solution for IP(t,s) then s-1 is a solution for IP(t9,s-1). So l Gt +1 is the minimum solution for IP(t,s). Now suppose that t is a set of triplets and for some integer s, IP(t,s) has a feasible solution h. Assume that G t has a cycle i 1 j 1 ?i 2 j 2 ? . . . ?i m j m ?i 1 j 1 . Corresponds to C we have inequalities h(i 1 j 1 )vh(i 2 ,j 2 )v . . . vh(i m ,j m )vh(i 1 ,j 1 )which is a contradiction and the proof is complete.
Let t be a set of triplets that is consistent with a tree or constructed from a given set of taxa, using TCD method. It was shown that G t is a DAG and by Theorem 3, h Tt is a feasible solution to IP(t,l Gt +1).
Theorem 4. Let t be a set of triplets consistent with a tree. Then h Tt is the unique optimal solution to IP(t,l Gt +1).
Proof. The graph G t is a DAG, since t is consistent with a tree. So l Gt is well efined.
The proof proceeds by induction on l Gt . Without loss of generality assume that G t is connected. The theorem is trivial when l Gt = 1. Let for each set of triplets consistent with a tree, h Tt be the unique optimal solution to IP(t,l Gt +1) where l Gt = k §1. Suppose that t is a set of triplets consistent with a tree and l Gt = k+1. Let t9 be the set of triplets which is introduced in the proof of Theorem 3. By the induction assumption h T t 0 is the unique optimal solution to IP(t9, l G t 0 +1). By Theorem 3 the minimum s for which IP(t, s) has a feasible solution is l Gt +1. Also l G t 0 +1 = l Gt . It follows that h Tt is the unique optimal solution to the IP(t,l Gt +1) and the proof is complete.
It is important to point out that the introduced target function of the above IP can be replaced with other appropriate target functions. But we use this special target function because it can be easily possible to find a solution for this IP in polynomial time when the input triplets are obtained from TCD method. Secondly using this target function, enable us to prove those above theorems which show the consistency of the result of the TripNet algorithm with a tree when there is a tree consistent with given triplets.
## Tripnet algorithm
Now we describe the TripNet algorithm in nine steps. In this algorithm the input is a set of triplets t and the output is a network consistent with t. Also if t is consistent with a tree the algorithm constructs a binarization of T t .
Step 1. In this step we find a height function h on L(t). If G t is a DAG we set G9 t = G t . If G t is not a DAG we remove some edges from G t in such a way that the resulting graph G9 t is a DAG. Set h = h G 0 t . If t is obtained from a set of taxa using TCD method, then G t is a DAG. Removing minimum number of edges from a directed graph to make it a DAG is known as the minimum Feedback Arc Set problem which is NP-hard [bib_ref] Reducibility among combinatorial problems, Karp [/bib_ref]. Thus we use the following heuristic method and try to remove as minimum number of edges as possible from G t in order to lose minimum information. First a cycle C is selected randomly. Let C max denote the set of nodes in C with the maximum degree. Remove an edge of C which one of its ends belongs to C max . This process continues until the resulting graph is a DAG. However, any such missing information will be recaptured in Step 9.
Step 2. In this step TripNet first apply HBUILD on h. If the result is a tree, TripNet constructs a binarization of this tree. Otherwise TripNet goes to Note that if t is consistent with a tree, TripNet constructs a binarization of T t .
Step 3. Remove all the maximum-weight edges from G. The process of removing all the maximum-weight edges from the graph continues until the resulting graph is disconnected.
In [bib_ref] Algorithms For combining rooted triplets into a galled phylogenetic network, Jansson [/bib_ref] and [bib_ref] Inferring a Level-1 Phylogenetic Network from a Dense Set of Rooted Triplets, Jansson [/bib_ref] the authors introduced the concept of SN-sets for a set of triplets t. A subset S of L(t) is an SN-set if there is no triplet ij|k M t such that i = [S and j, k M S. In [bib_ref] Inferring a Level-1 Phylogenetic Network from a Dense Set of Rooted Triplets, Jansson [/bib_ref] it is shown that if t is dense then the maximal SN-sets partition L(t) and can be found in polynomial time. By contracting each of the SN-set to a single node and assuming a common ancestor for all of these leaves, the size of the problem is reduced. In these papers, for finding the maximal SN-sets in polynomial time, the authors use the high density of the input triplet sets. TripNet algorithm uses the concept of height function as an auxiliary tool to obtain SN-sets instead of the high density assumption.
Step 4. For each connected component obtained in Step 3 which is not an SN-set, we apply Step 3. This process continues until all of the resulting components are SN-sets. Let {S 1 , S 2 , …, S k } be the set of resulting SN-sets. If each SN-set contains only one node, HBUILD is applied and if the result is a tree TripNet constructs a binary tree and goes to Step 6. Otherwise TripNet goes to Step 5. If for some i, |S i |.1, contract each S i to a single node s i and set S = {s 1 , s 2 , …, s k }. Update the set of triplets by defining t S = {s i s j |s k : if ' xy|z M t, x M S i , y M S j and z M S k }. Constructs a weighted complete graph (G S , w S ) with V(G S ) = S and w S (s i , s j ) = min {h(x, y): x M S i and y M S j }. Set (G, w) = (G S , w S ) and TripNet goes to Step 3.
The following theorem is a consequence of the definition SN-set for (G S , w S ).
Theorem 5. Applying Steps 3 and 4 on (G S , w S ) and t S , each resulting SN-set has one member.
Proof. Suppose that S = {s 1 , s 2 , s 3 , …, s r } is an SN-set in (G S , w S ). Now assume that in the procedure of Step 3 by removing the edges with weight l, S 1 separates from S 2 . Thus there exists k . l such that by removing the edges with weight at least k in (G S , w S ), the connected component S separates from other components of G S . It means that by removing the edges with weight at least k in G, we obtain the SN-set S 1 |S 2 | . . . |S r which is a contradiction.
In the next step the reticulation leaves are recognized using the following three criteria:
Criterion I. Let m i and M i be the minimum and maximum weight of the edges in (G,h) with exactly one end in S i . Choose the node with minimum m i and if there is more than one node with minimum m i then choose among them the nodes which has minimum M i . Let R 1 denotes the set of such nodes.
Criterion II. Let w min = min {w(s i ,s j ): 1ƒi,jƒk}. In G S consider the induced subgraph on the edges with the weight w min . Choose the nodes of R 1 with the maximum degree in this induced subgraph. Let R 2 denotes the set of such nodes.
Criterion III. For each node s M R 2 , remove it from GS and find SN-sets for this new graph using Steps 3 and 4. Let n s be the number of SN-sets of this new graph with cardinality greater than one. Choose the nodes in R 2 with maximum n s . Let R 3 denotes the set of such nodes.
We state an example to show the idea behind these three criteria.
Let t = {ij|l, jk|i, kl|j, kl|i, no|m, lo|k, jl|o, mn|l, mn|j, no|k, mo|i, jk|n, ij|o, ik|m, il|n}.
t is not consistent with a tree but it is consistent with the network N shown in [fig_ref] Figure 8: An example to show how TripNet works to find a reticulation leaf... [/fig_ref]. Obviously, N is an optimal network consistent with t. In order to find SN-sets we construct G9 t and (G, h), and find SN-sets from (G, h) using Steps 3 and 4 [fig_ref] Figure 8: An example to show how TripNet works to find a reticulation leaf... [/fig_ref] [fig_ref] Figure 8: An example to show how TripNet works to find a reticulation leaf... [/fig_ref]. we expect that the reticulation is in R 1 . In this example both k and l are in R 1 . Also we expect that if there is a reticulation leaf, it belongs to R 2 which again both k and l are in R 2 . Now just l belongs to R 3 . Thus we consider l as the reticulation leaf [fig_ref] Figure 8: An example to show how TripNet works to find a reticulation leaf... [/fig_ref]. Remove triplets from t S which contain l and denote the new set of triplets by t9 S . Obviously t9 S is consistent with a tree. We add this reticulation leaf to a binarization of T t 0 S such that the resulting network is consistent with t S . Note that if we consider each node except than l as the reticulation leaf then final network consistent with t S has at least two reticulation leaves.
Step 5. In this step the reticulation leaf is recognized using three criteria. Do the criterion I. If |R 1 | = 1 then choose the node x M R 1 as the reticulation node. Otherwise if |R 1 |.1 do the criterion II. If |R 2 | = 1 then choose the node x M R 2 as the reticulation node. Otherwise if |R 2 |.1 do the criterion III. If |R 3 | = 1 then choose the node x M R 3 as the reticulation node. Otherwise if |R 3 |.1 then by the speed options we choose the reticulation node as follows.
Slow. Each node in R 3 is examined as the reticulation leaf. Normal. Two nodes in R 3 are selected randomly and each of these two nodes is examined as the reticulation leaf.
Fast. One node in R 3 is selected randomly as the reticulation leaf.
Let x be a node which is considered as a reticulation leaf. Remove x from G S and all of the triplets which contain x from t S . Define G = G \ {x} and go to Note that for the Fast option the running time of the algorithm is polynomial.
For biological data almost always the criteria I and II find a unique reticulation leaf.
So on real data the running time of TripNet is almost always polynomial.
Step 6. Let x 1 , x 2 , …, x m be m reticulation leaves which are obtained in Step 5 with this order and T be the tree that is constructed in Step 4. Now add these m nodes in the reverse order to T as what follows. Let e 1 and e 2 be two edges of T. Consider two new nodes y 1 and y 2 in the middle of e 1 and e 2 . Connect y 1 and y 2 to a new node y 3 and connect the reticulation leaf x m to y 3 . Do this procedure for all pairs of edges and choose a pair such that the resulting network is consistent with maximum number of triplets in t. Continue this procedure until all the reticulation nodes are added.
Step 7. For each SN-set S i and the set t Si of triplets we run the algorithm again.
Step 8. Replace each SN-set in the network of Step 6 with its related network constructed in Step 7 to obtain a network N9.
Let t9 M t be the set of the triplets which are not consistent with N9. For each pair of leaves a and b assume that t9 ab is the set of triplets in t9 which are of the form ab|c. Consider the pair of leaves i and j such that t9 ij has the maximum cardinality. Assume that p i and p j are the parents of i and j, respectively.
Step 9. Create two new nodes in the middle of the edges p i i and p j j and connect them with a new edge. This new edge creates a reticulation node and all of the triplets in t9 ij will be consistent with the new network. All consistent triplets with the new network are removed from t9 and this procedure will continue until t9 becomes empty. presents an example of the algorithm with all of its Steps.
[fig] Figure 1: A triplet and a network consistent with it. (a) The triplet $ij|k$, (b) The triplet ij|k is consistent with the given network. doi:10.1371/journal.pone.0106531.g001 [/fig]
[fig] Figure 2: An example of BUILD algorithm for the given set {bc |a, ac |d, de |b} of triplets. doi:10.1371/journal.pone.0106531.g002 [/fig]
[fig] Figure 3: Resulting networks from Yeast triplets. (a) LEVEL2 algorithm result. (b) TripNet algorithm result. doi:10.1371/journal.pone.0106531.g003 [/fig]
[fig] Figure 4: The steps of constructing T t from the given set t = {kl |j, kl |i, jk |i, jl |i}, using HBUILD.(a) The graph G t , (b) The graph (G,h), (c) Removing maximum weights from the graph (G,h), (d) Constructing T t using step c. doi:10.1371/journal.pone.0106531.g004 [/fig]
[fig] Figure 5: An example of binarization. The binary tree is a binarization of the non-binary tree. doi:10.1371/journal.pone.0106531.g005 [/fig]
[fig] Figure 6: Two different networks with the same height function. For the given network N and tree T, h N = h T = h. h(j,k) = 1, h(i,j) = h(i,k) = 2 and h(i,l) = h(j,l) = h(k,l) = 3. doi:10.1371/journal.pone.0106531.g006 than 6 hours. For the remaining 22.5% of the cases on average SIMPLISTIC outputs a network in 2675 seconds, while in all cases the TripNet running time is on average 200 seconds. [/fig]
[fig] Figure 7: A counter example for the reverse of Theorem 2. ij|k is consistent with the given network, but h(i,j) = h(i,k) = 3 and h(j,k) = 2. doi:10.1371/journal.pone.0106531.g007 [/fig]
[fig] Figure 8: An example to show how TripNet works to find a reticulation leaf by applying step 5. Edges with weight 6 are shown by dotted lines. (a) t = {ij|l, jk|i, kl|j, kl|i, no|m, lo|k, jl|o, mn|l, mn|j, no|k, mo|i, jk|n, ij|o, ik|m, il|n} is not consistent with a tree and is consistent with the given level-1 network, (b) G9 t is obtained from G t by removing the dotted line,(c) Computing (G, h), (d) Remove edges with weights 6 and 5 from (G, h) to obtain SN-sets {n, o} and {m}, (e) Remove edges with weights 4 and 3 from the remaining graph to obtain SN-set {i}, (f) Remove edges with weights 2 from the remaining graph to obtain SN-set {j}, (g) Remove edges with weights 1 from the remaining graph to obtain SN-sets {k} and {l}, (h) Compute G S . both SN-sets {k} and {l} satisfy Criteria I and II, (i) Remove {l} from G S , (j) Remove edges with weights 6 from the graph of previous step to obtain SNsets {i, j, k} and {m, n, o}, (k) Remove {k} from G S , (l) Remove edges with weights 6 and 5 from the graph of previous step to obtain SN-sets {n, o} and {m}, (m) Remove edges with weights 4 from the remaining graph to obtain SN-set {l}, (n) Remove edges with weights 3 from the remaining graph to obtain SN-sets {i} and {j}. [/fig]
[table] Table 1: SIMPLISTIC and TripNet network results. [/table]
|
Use of a Prescribed Ephedrine/Caffeine Combination and the Risk of Serious Cardiovascular Events: A Registry-based Case-Crossover Study
Ephedrine and herbal ephedra preparations have been shown to induce a small-to-moderate weight loss. Owing to reports on serious cardiovascular events, they were banned from the US market in 2004. There have been no large controlled studies on the possible association between prescribed ephedrine/caffeine and cardiovascular events in general. The authors linked data from four different sources within Statistics Denmark, using data on 257,364 users of prescribed ephedrine/caffeine for the period 1995-2002. The data were analyzed using a casecrossover technique with a composite endpoint: death outside of a hospital, myocardial infarction, or stroke. To account for effects of chronic exposure and effects in naïve users, the authors performed a secondary case-control study nested within the cohort of ephedrine/caffeine ever users. Among 2,316 case subjects, 282 (12.2%) were current users of ephedrine/caffeine. The case-crossover analysis yielded an odds ratio of 0.84 (95% confidence interval: 0.71, 1.00); after adjustment for trends in ephedrine/caffeine use, it was 0.95 (95% confidence interval: 0.79, 1.16). Subgroup analyses revealed no strata with significantly elevated risk. In the case-control substudy, there was no increased risk among naïve users or users with large cumulative doses. Prescribed ephedrine/ caffeine was not associated with a substantially increased risk of adverse cardiovascular outcomes in this study. ). Cardiovascular Effects of Ephedrine 967 Am J Epidemiol 2008;168:966-973 c Adjusted for discordant use of aspirin, statins, antihypertensive agents, antidiabetic agents, and nonsteroidal antiinflammatory drugs and for trend in ephedrine/caffeine prescriptions. d Death outside of a hospital, myocardial infarction, or stroke. Cardiovascular Effects of Ephedrine 969 Am J Epidemiol 2008;168:966-973 Cardiovascular Effects of Ephedrine 971 Am J Epidemiol 2008;168:966-973 Cardiovascular Effects of Ephedrine 973 Am J Epidemiol 2008;168:966-973
Ephedrine is a sympathomimetic compound with lipolytic and appetite-inhibiting properties [bib_ref] Efficacy and safety of ephedra and ephedrine for weight loss and athletic..., Shekelle [/bib_ref]. It is an alkaloid extract from ephedra shrubs growing in Asia, Europe, and America. The Asian species (Ephedra sinica) has the highest alkaloid content, and it has been used in traditional folk remedies (ma huang), particularly in China [bib_ref] Cardiovascular effects of ephedra alkaloids: a comprehensive review, Andraws [/bib_ref]. In Western countries, ephedrine has been used to promote weight loss and to enhance athletic performance [bib_ref] Efficacy and safety of ephedra and ephedrine for weight loss and athletic..., Shekelle [/bib_ref]. During the 1990s, it was found that nearly one-third of young, obese women had used a weightloss supplement containing ephedra [bib_ref] Adverse cardiovascular and central nervous system events associated with dietary supplements containing..., Haller [/bib_ref]. In 1999 alone, 12 million persons in the United States used 3 billion doses of ephedra alkaloids [bib_ref] Adverse cardiovascular and central nervous system events associated with dietary supplements containing..., Haller [/bib_ref]. The US Food and Drug Administration banned the sale of ephedrine in 2004 because of a considerable number of reports of adverse vascular complications.
The suspicion of cardiovascular toxicity of ephedrine was mainly based on spontaneous reporting [bib_ref] Adverse cardiovascular and central nervous system events associated with dietary supplements containing..., Haller [/bib_ref] [bib_ref] Safety of ephedra: lessons learned, Soni [/bib_ref]. There were no large-scale controlled observational studies conducted to broadly assess the cardiovascular toxicity of ephedrine. Until 2002, a prescribed ephedrine preparation, Letigen (Nycomed International Management GmbH, Zurich, Switzerland), had a dominant position in the Danish market for weight-loss products. The existence of a national prescription database with comprehensive recording of all prescriptions filled in Denmark since 1995, with the possibility of linkage to a wide array of other health databases [bib_ref] The Danish prescription registries, Gaist [/bib_ref] , offered us a unique opportunity to address the question of ephedrine's cardiovascular toxicity in a controlled observational study. only by prescription. Its recommended dose was 1-3 tablets per day, depending on the user's tolerance. It was approved for sale in Denmark in 1990. During the peak of its use in 1999, some 110,000 persons, corresponding to 2% of the Danish population, were treated. In 2002, the marketing license was suspended, after a number of reports had suggested a safety problem.
## Data sources
We used data from Statistics Denmark, a governmental institution that collects and maintains electronic records for statistical and scientific purposes. Four comprehensive national data sources were used: the Danish National Registry of Patients, the Prescription Database of the Danish Medicines Agency, the Danish Register of Death, and the Danish Person Registry. All data sources were linked by use of a mutual identifier.
Since 1977, data on all individual hospital discharges have been stored in the Danish National Registry of Patients. Each record contains an identifier as well as selected medical data, including discharge diagnoses and operative procedures. Diagnoses were recorded according to the International Classification of Diseases, Eighth Revision (ICD-8), from 1977 through 1993 and have been recorded according to the International Classification of Diseases, Tenth Revision (ICD-10), since 1994. Virtually all inpatient medical care in Denmark during the study period was furnished by the public health authorities; thus, this data resource allows true population-based studies, covering all 5.4 million inhabitants of Denmark [bib_ref] When an entire country is a cohort, Frank [/bib_ref].
The Prescription Database of the Danish Medicines Agency contains data on all prescriptions redeemed by Danish citizens since 1995, independently of whether or not the cost of the medication was reimbursed. Among the data included are a person identifier, the date of purchase, a prescriber code, the substance, the brand name, and the quantity dispensed. Dosing instructions and the indication for prescribing the medication are not recorded. Drugs are categorized according to the Anatomic-Therapeutic-Chemical classification.
The Danish Register of Death contains information on date of death, cause of death, and mode of death (natural cause, accident, suicide, or homicide) and a code indicating the location where the death occurred (in a hospital, at another health institution, at home, or elsewhere). The Danish Person Registry contains data on all migrations into and out of Denmark, which allowed us to keep track of all ephedrine/caffeine users.
The base population for this study was all persons who had filled prescriptions for ephedrine/caffeine during the period January 1, 1995-December 31, 2001. For this cohort of 298,848 persons, we extracted data on all prescriptions for cardiovascular, antithrombotic, antidiabetic, antirheumatic, antiasthmatic, and appetite-suppressant medications filled during the same period, and from the other registers, we obtained all available information on the subjects.
# Analysis
Design. For our primary analysis, we used the casecrossover design. It is based on the case-base paradigm, but instead of the use of matched controls, the past experience of the case serves as the case's own reference. Thereby, confounders that are stable over time cancel each other out. This even extends to stable confounders that cannot be measured or are unknown [bib_ref] The case-crossover design: a method for studying transient effects on the risk..., Maclure [/bib_ref] [bib_ref] Should we use a case-crossover design?, Maclure [/bib_ref] [bib_ref] Case-crossover and case-time-control designs as alternatives in pharmacoepidemiologic research, Schneeweiss [/bib_ref]. In the present context, use of ephedrine/caffeine could have been related to smoking and obesity [bib_ref] Use of nonprescription weight-loss products: results from a multistate survey, Blanck [/bib_ref]. Case-crossover designs are particularly suitable when the exposure is intermittent, the effect on risk is immediate and transient, and the outcome is abrupt [bib_ref] The case-crossover design: a method for studying transient effects on the risk..., Maclure [/bib_ref]. The available reports suggested that the potential cardiovascular toxicity would indeed be immediate and transient and that the outcome would be abrupt [bib_ref] Adverse cardiovascular and central nervous system events associated with dietary supplements containing..., Haller [/bib_ref] [bib_ref] Safety of ephedra: lessons learned, Soni [/bib_ref].
Certain aspects of our research question were not amenable to a case-crossover study. Some spontaneous reports had suggested a particular risk in naïve, first-time users [bib_ref] Adverse cardiovascular and central nervous system events associated with dietary supplements containing..., Haller [/bib_ref]. Since subjects could not be naïve users both on their case date and on their control date, this would violate the conditions for a crossover study. We also wanted to study the effect of chronic exposure, for which the crossover type of study is inefficient [bib_ref] Should we use a case-crossover design?, Maclure [/bib_ref]. Finally, we wanted to explore a possible cumulative dose effect, for which the crossover design would not work (highest cumulative dose on the reference date would be impossible).
To address these issues, we conducted a supplementary casecontrol study. To minimize potential confounding by indication, we nested the case-control study within a cohort of ever users of ephedrine/caffeine as described below.
Study cohort definition. Both the case-crossover study and the case-control study were nested within a cohort of ever users of ephedrine/caffeine. Persons entered the cohort and were eligible to become cases when all of the following events had occurred: 1) the start of the study period on July 1, 1996; 2) January 1 of the year of the subject's 18th birthday; 3) the filling of the first recorded ephedrine/caffeine prescription; and 4) residence in Denmark for at least 18 months. Thus, we required all subjects to have been observable for any medication dispensation for at least 18 months and to have filled at least one prescription for ephedrine/caffeine.
The subjects left the cohort at the first occurrence of one of the following: 1) any case-defining event; 2) death or emigration; 3) January 1 of the year of the subject's 70th birthday; 4) any diagnosis of malignancy, excluding nonmelanoma skin cancer (ICD-10 codes C00-C97, excluding code C44; ICD-8 codes 140-207, excluding code 173); and 5) the end of study period on December 31, 2001.
We excluded persons who underwent one of the cohortterminating events (e.g., a cancer diagnosis or emigration) before their potential cohort entry and persons who had more than one migration event.
Case definition. Since the literature on adverse effects of ephedrine is ambiguous about what particular cardiovascular effects to expect (2-5), we employed a broad, composite primary endpoint and performed secondary analyses on each subset of this endpoint.
Our primary endpoint was defined by the occurrence of any of the following: Death coded as due to natural causes, occurring outside of a hospital or nursing home. Myocardial infarction, defined by hospitalization with a discharge diagnosis of myocardial infarction (ICD-10 codes I21-I22). We also required that the hospital stay be at least 3 days in duration, unless the patient died within the first 3 days of admission. To account for hospital transfers for invasive procedures or other specialist treatment, serial admissions with no discharge interval between them were considered a single admission. Fatal myocardial infarction was defined as the patient's dying within 30 days after the index hospitalization. Stroke, defined by hospitalization with an ICD-10 discharge code of I61, I63, or I64, excluding codes I63.1 and I63.4 (cerebral embolism). Duration-of-stay and case-fatality criteria similar to those for the myocardial infarction cases were applied.
Exposure definition. We computed the frequency distributions of intervals between presentations of prescriptions for ephedrine/caffeine. On the basis of these analyses, we estimated the modal intake as slightly less than 2 tablets per day [bib_ref] Templates for analysis of individual-level prescription data, Hallas [/bib_ref]. Very few users had taken 4 or more tablets per day for extended periods of time. In our main analysis, we assigned a period of exposure for each prescription, starting on the day of prescription redemption and assuming a daily intake of 2 tablets, until all tablets were taken. Subjects were considered exposed on all days within a prescription's supply period and unexposed on all other days. As a sensitivity analysis, we repeated all analyses with an assumed daily intake of 1 or 3 tablets and with a fixed 90-day exposure period assigned to each prescription.
Our exploratory analyses confirmed the episodic nature of ephedrine/caffeine therapy. Using a waiting-time-distribution technique [bib_ref] Templates for analysis of individual-level prescription data, Hallas [/bib_ref] , we arrived at an average duration varying between 8.8 months and 13 months for the period 1995-2001. For other medications included as time-dependent covariates, we employed a fixed 90-day window to define exposure; that is, a person was considered exposed to aspirin if he or she had filled a prescription for aspirin within the past 90 days.
Case-crossover study. In the case-crossover design, a patient's exposure status at the time of disease is compared with the same patient's individual exposure at an earlier point in time. For each case subject, 10 control dates were assigned randomly within the period 9-15 months before the case date for the same individual. For cases that arose shortly after cohort entry, this entailed sampling control dates before cohort entry. We chose these control dates because of the seasonality in ephedrine/caffeine use. For example, there were approximately twice as many new users in January as in December and a small excess of use in the early summer months.
Time-dependent potential confounders were controlled by conditional logistic regression [bib_ref] The case-crossover design: a method for studying transient effects on the risk..., Maclure [/bib_ref]. We included use of low-dose aspirin, statins, antihypertensive agents, antidiabetic agents, and nonsteroidal antiinflammatory drugs as time-dependent covariates in all analyses. We also conducted analyses without inclusion of these covariates, but results are not presented, since they differed very little from those of the main analysis. We calculated odds ratios for our main endpoint, for each subset of endpoints, and for a number of prespecified patient subgroups as detailed in the tables. Adjustment for trends in exposure. There was a moderate decline in ephedrine/caffeine use during the last part of the study period, and the prevalence of ephedrine/caffeine use declined rapidly with age for persons aged 55 years or older (7). This would convey a spurious protective effect of ephedrine/caffeine; cases were less likely to use ephedrine/ caffeine on their case date than on their control date, since they were approximately 1 year older and possibly had entered a period of lower ephedrine/caffeine use in general.
To adjust for this, Suissa [bib_ref] The case/time/control design, Suissa [/bib_ref] proposed the case-timecontrol design. A matched control group is extracted, and their exposure attributes on the index day and on a control date in the past are established in exactly the same ways as for the cases. Control subjects are then used as a reference group for the odds ratio derived from the case-crossover study [bib_ref] The case/time/control design, Suissa [/bib_ref]. For each subject, we used 4 random controls selected as described below.
Case-control study. In the case-control substudy, we used the same cases and the same exposure definition as in the case-crossover study. For each case, we randomly selected 10 control subjects matched to the case with respect to exact birth year and sex among the persons available in the cohort on the index date. Controls were assigned an index date identical to that of the case date of the corresponding case. Since we sampled controls from our study cohort, the reference category for comparison of current ephedrine/caffeine exposure was remote use of the medication.
In the case-control study, we placed a special focus on potential dose-response and duration-response effects. We used the number of tablets purchased within the last 90 days before the index date as a measure of current dose and the cumulative amount of ephedrine/caffeine ever recorded up to the index date as a measure of cumulative dose. For the duration-response effect, we categorized subjects according to the time of the first recorded prescription relative to the index date.
The following variables were included as potential confounders: 1) ever diagnosis of obesity (ICD-10 code E66; ICD-8 code 27799); 2) ever use of antidiabetic agents or a diagnosis of diabetes (ICD-10 codes E10-E14; ICD-8 code 250); 3) current use of low-dose acetylsalicylic acid; 4) ever diagnosis of chronic obstructive pulmonary disease (ICD-10 code J44; ICD-8 codes 490-491) or ever use of systemic beta-agonists or inhaled anticholinergic agents; 5) ever diagnosis of hypertension (ICD-10 code I10; ICD-8 code 40) or ever use of antihypertensive agents (i.e., thiazides, beta-blockers, calcium channel blockers, and medications acting on the renin-angiotensin system); 6) ever use of statins; 7) ever diagnosis of ischemic heart disease or myocardial infarction (ICD-10 codes I20-I25; ICD-8 codes 412-414); and 8) ever diagnosis of cerebral ischemia or stroke (ICD-10 codes I61, I63, and I64, excluding codes I631 and 641; ICD-8 codes 431 and 433-435). Confounders were controlled by conditional logistic regression.
For all estimates, we report 95% confidence intervals. The study was approved by Statistics Denmark's scientific board. Approval from an ethics committee was not required according to Danish law. Data were analyzed using Stata, version 8.0 (Stata Corporation, College Station, Texas).
# Results
## Study cohort
Of the 298,848 persons registered as using ephedrine/ caffeine, 41,484 did not enter the study cohort, for the following reasons: 1) the subject was outside the allowed age range of 18-70 years throughout the study period (n ¼ 11,347); 2) a cancer diagnosis or other cohort-terminating event preceded potential cohort entry (n ¼ 25,395); and 3) more than one migration event was recorded (n ¼ 4,445) or the subject immigrated too late to be observable for 18 months (n ¼ 297). In all, 257,364 users of ephedrine/caffeine (51,974 men and 205,390 women) entered the cohort. They contributed 1,023,297 person-years of observation, during which 2,316 case-defining events occurred. Their clinical characteristics are detailed in [fig_ref] Table 1: Characteristics of all users and of those experiencing a case-defining event a... [/fig_ref].
## Case-crossover study
Of the 2,316 cases, 531 were deaths that occurred outside of a hospital, 839 were myocardial infarctions, and 946 were strokes. The characteristics of the case subjects are listed in [fig_ref] Table 1: Characteristics of all users and of those experiencing a case-defining event a... [/fig_ref]. Their median age was 55 years, and 1,019 (44%) Abbreviations: AOR, adjusted odds ratio; CI, confidence interval. a The analysis was based on the assumption of an intake of 2 tablets per day starting from the day on which the prescription was filled. For each case subject, 10 random control days were sampled within the interval of 9-15 months prior to the index date. b Adjusted for discordant use of aspirin, statins, antihypertensive agents, antidiabetic agents, and nonsteroidal antiinflammatory drugs by conditional logistic regression.
were men. As expected, cardiovascular risk factors were more prevalent among cases than among users of ephedrine in general; 50.9% had a history of antihypertensive use, 15.0% of acetylsalicylic acid use, and 12.2% of antidiabetic use. Of the cases, 282 (12.2%) were current users of ephedrine/caffeine, the remaining being past users.
Among the 2,316 cases, 824 (36%) showed some degree of discordant exposure-that is, they were either exposed on the case date and unexposed on one or more of the reference dates or they were unexposed on the case date and exposed on at least one reference date. The odds ratio associating ephedrine/caffeine use with an adverse cardiovascular outcome, adjusted for discordant use of other cardiovascular medications and antidiabetic agents, was 0.84 (95% confidence interval (CI): 0.71, 1.00). Further adjustment for trend in ephedrine prescriptions yielded an odds ratio of 0.95 (95% CI: 0.79, 1.16). The odds ratios for sub-endpoints are shown in [fig_ref] Table 2: Main results from a case-crossover analysis of the association between use of... [/fig_ref]. For death occurring outside of a hospital, there was an inverse association with current ephedrine use (adjusted odds ratio (OR) ¼ 0.54, 95% CI: 0.35, 0.84).
The explorative analyses revealed no subjects whose odds ratios were substantially elevated above the main group [fig_ref] Table 3: Results from subgroup analysis in a case-crossover study of the association between... [/fig_ref]. For women and for users of statins, an inverse association was observed (OR ¼ 0.76 (95% CI: 0.59, 0.99) and OR ¼ 0.42 (95% CI: 0.18, 0.93), respectively). The 95% confidence intervals for all other estimates spanned the null value.
The sensitivity analyses assuming a daily intake of 1 or 3 tablets or assigning a fixed 90-day window to each ephedrine prescription produced odds ratios of the same magnitude as those seen in the main analysis. The odds ratios for the main estimate were 1.20 (95% CI: 1.00, 1.43), 0.90 (95% CI: 0.73, 1.10), and 1.08 (95% CI: 0.90, 1.29), respectively.
## Case-control study
In the case-control substudy, an odds ratio of 1.23 (95% CI: 0.67, 2.27) was found for the subgroup whose first ephedrine prescription was filled 0-10 days before the index Abbreviations: AOR, adjusted odds ratio; CI, confidence interval. a The analysis was based on the assumption of an intake of 2 tablets per day starting from the day on which the prescription was filled. For each case subject, 10 random control days were sampled within the interval of 9-15 months prior to the index date. The main composite endpoint was employed for all analyses (see [fig_ref] Table 2: Main results from a case-crossover analysis of the association between use of... [/fig_ref]. b Adjusted for discordant use of aspirin, statins, antihypertensive agents, antidiabetic agents, and nonsteroidal antiinflammatory drugs by conditional logistic regression. c Adjusted for discordant use of aspirin, statins, antihypertensive agents, antidiabetic agents, and nonsteroidal antiinflammatory drugs and for trend in ephedrine/caffeine prescriptions. d No prior cardiovascular diagnoses or use of cardiovascular, antidiabetic, antithrombotic, or antihypertensive medications. date [fig_ref] Table 4: Results from explorative dose-response and dose-duration analyses in a case-control study of... [/fig_ref]. The odds ratio decreased slightly with increasing duration since the first ephedrine/caffeine prescription (for each successive step, OR ¼ 0.89, 95% CI: 0.82, 0.97). We found no indication of a trend with lifetime cumulative dose (per step, OR ¼ 0.93, 95% CI: 0.85, 1.01) or cumulative dose within the past 90 days (per step, OR ¼ 0.95, 95% CI: 0.81, 1.10).
In the case-control substudy, the adjusted odds ratios for the main composite endpoint, for death outside a hospital, for myocardial infarction, and for stroke were 0.88 (95% CI: 0.77, 1.01), 0.76 (95% CI: 0.54, 1.07), 0.95 (95% CI: 0.76, 1.18), and 0.92 (95% CI: 0.75, 1.13), respectively (data not shown).
# Discussion
Our main finding was that prescription of an ephedrine/ caffeine product was not associated with adverse cardiovascular outcomes. This was found across a wide range of patient subgroups, different cardiovascular outcomes, dif-ferent assumptions about exposure, and different utilization patterns.
The strength of our study is the comprehensive recording of clinical details for a very large population with virtually no loss to follow-up. One limitation of our study is that it concerned a pharmaceutical ephedrine preparation, produced under strict control. Findings may not be applicable to herbal products whose content of ephedra alkaloids varies substantially [bib_ref] Ephedrine-type alkaloid content of nutritional supplements containing Ephedra sinica (ma-huang) as determined..., Gurley [/bib_ref]. In addition, herbal products may be used under less appropriate instruction or surveillance.
The evidence linking ephedrine to cardiovascular morbidity is based mainly on spontaneous reporting. However, with the very large number of users (3) and their possible adverse health behavior [bib_ref] Use of nonprescription weight-loss products: results from a multistate survey, Blanck [/bib_ref] , coincidental cardiovascular events probably occur in large numbers.
There have been only a few controlled observational studies on ephedra alkaloids and cardiovascular outcomes. In the Hemorrhagic Stroke Project, an association was found between phenylpropanolamine and hemorrhagic stroke [bib_ref] Phenylpropanolamine and the risk of hemorrhagic stroke, Kernan [/bib_ref]. Phenylpropanolamine is a minor metabolite of ephedrine which, in its synthetic form, is used as a cold remedy and Abbreviations: CI, confidence interval; OR, odds ratio. a The main composite endpoint was used in all analyses (see [fig_ref] Table 2: Main results from a case-crossover analysis of the association between use of... [/fig_ref]. b Adjusted for a prior diagnosis of obesity, diabetes, hypertension, ischemic heart disease, chronic obstructive pulmonary disease, cerebral ischemia, or stroke and for ever use of antidiabetic agents, thiazides, beta-blockers, calcium channel blockers, medications acting on the reninangiotensin system, inhaled anticholinergic agents, systemic beta-agonists, and statins. (See text for details.) dieting aid. The association depended strongly on whether the substance was used for colds or for weight loss (OR ¼ 1.2 vs. OR ¼ 15.9), which may suggest some confounding by obesity. In a substudy exclusively evaluating exposure to genuine ephedra alkaloids, an overall null finding was reported (OR ¼ 1.0), with a possible detrimental effect with large doses (OR ¼ 3.6, 95% CI: 0.7, 18) [bib_ref] Use of ephedra-containing products and risk for hemorrhagic stroke, Morgenstern [/bib_ref]. In a recent study from South Korea, Yoon et al. [bib_ref] Phenylpropanolamine contained in cold remedies and risk of hemorrhagic stroke, Yoon [/bib_ref] reported an association between phenylpronalolamine and stroke, even when it was used against cold. In a meta-analysis of randomized studies of ephedra alkaloids, no serious cardiovascular events were observed among the subjects, and the authors were able to confidently exclude event rates above 1 per 1,000 treated persons (1).
A major prerequisite for the case-crossover design is that the unmeasured confounders are stable over time. Obviously, smoking behavior or body mass index may change over a year for an individual. However, stable tobacco abstinence or weight loss is an exception rather a rule [bib_ref] Smoking cessation guidelines for health professionals: an update, West [/bib_ref] [bib_ref] Can anyone successfully control their weight? Findings of a three year communitybased..., Crawford [/bib_ref] , and smokers continue to have a high cardiovascular risk for some time after having stopped smoking [bib_ref] Decline in the risk of myocardial infarction among women who stop smoking, Rosenberg [/bib_ref] , as do persons who have lost weight [bib_ref] What are the longterm benefits of weight reducing diets in adults? A..., Avenell [/bib_ref]. Thus, within the time frames of our study, the individual health effects of smoking and overweight were reasonably stable and were unlikely to have confounded our estimates materially.
For one endpoint, death occurring outside of a health institution, we observed odds ratios below unity. This should not be taken too literally as a protective effect. One possible explanation is that some of these subjects died at home from chronic nonmalignant diseases that had not resulted in secondary care contacts. These subjects were obviously very unlikely to have used ephedrine/caffeine shortly before their deaths. In addition, some subjects with impending cardiovascular events could have been warned by subtle symptoms and could have chosen to discontinue use of ephedrine for fear of its claimed toxicity. We performed a subanalysis in subjects with no prior cardiovascular diagnoses and no prior use of cardiovascular, antidiabetic, antithrombotic, or antihypertensive agents [fig_ref] Table 3: Results from subgroup analysis in a case-crossover study of the association between... [/fig_ref]. The estimate for this subgroup differed very little from the main estimate (OR ¼ 0.81, 95% CI: 0.61, 1.08; after adjustment for trend, OR ¼ 0.84, 95% CI: 0.62, 1.14). These ''confounding-bycontraindication'' effects are difficult to manage in observational studies, insofar as the warning symptoms are not always captured by available data sources, and we cannot rule out the possibility that the odds ratios for the main estimates may have been biased downward.
Another limitation is that in our main analysis, we assumed an immediate effect of ephedrine/caffeine. If an adverse effect of ephedrine/caffeine had delayed onset (e.g., if it were mediated through a hypertensive effect), we might not have captured it by our crossover analysis. However, there was nothing in our case-control analysis to suggest a delayed effect with continuous exposure.
Regarding the possibility of misclassification of case or exposure status, recent validations in our setting have shown positive predictive values on the order of 95% for both myocardial infarction and stroke diagnoses [bib_ref] Antibiotics active against Chlamydia do not reduce the risk of myocardial infarction, Bjerrum [/bib_ref] [bib_ref] Selective serotonin reuptake inhibitors and the risk of stroke: a population based..., Bak [/bib_ref]. The prescription data in our data sources have a high level of accuracy [bib_ref] A population-based Danish data resource with possible high validity in pharmacoepidemiological research, Sørensen [/bib_ref]. The main uncertainty regarding exposure was the exact timing of ephedrine intake relative to the filling of the prescriptions. Our sensitivity analyses showed the findings to be robust within reasonable assumptions about daily intake.
Our results have two major implications. First, we should not blindly trust the findings of spontaneous reporting schemes. Such systems are vulnerable to a ''snowball'' effect, whereby products that have acquired a poor reputation generate new adverse reports. The few controlled clinical studies that have been conducted-including ours-have failed to demonstrate any cardiovascular toxicity of ephedrine [bib_ref] Use of ephedra-containing products and risk for hemorrhagic stroke, Morgenstern [/bib_ref]. Second, although ephedrine has been banned from the US market, it is probably available through illicit channels, through the Internet or as a nondeclared dietary supplement, just as similar products are in free trade in other parts of the world. Thus, the safety of these compounds is still relevant.
[table] Table 1: Characteristics of all users and of those experiencing a case-defining event a among 257,364 users of a fixed-dose caffeine/ephedrine product,Denmark, 1995Denmark, -2001 [/table]
[table] Table 2: Main results from a case-crossover analysis of the association between use of prescribed ephedrine/caffeine and cardiovascular morbidity, Denmark, 1995-2001 a [/table]
[table] Table 3: Results from subgroup analysis in a case-crossover study of the association between use of prescribed ephedrine/caffeine and cardiovascular morbidity, Denmark, 1995-2001 a [/table]
[table] Table 4: Results from explorative dose-response and dose-duration analyses in a case-control study of the association between use of prescribed ephedrine/caffeine and cardiovascular morbidity, Denmark, 1995-2001 a [/table]
|
Adventures in the Enormous: A 1.8 Million Clone BAC Library for the 21.7 Gb Genome of Loblolly Pine
Loblolly pine (LP; Pinus taeda L.) is the most economically important tree in the U.S. and a cornerstone species in southeastern forests. However, genomics research on LP and other conifers has lagged behind studies on flowering plants due, in part, to the large size of conifer genomes. As a means to accelerate conifer genome research, we constructed a BAC library for the LP genotype 7-56. The LP BAC library consists of 1,824,768 individually-archived clones making it the largest single BAC library constructed to date, has a mean insert size of 96 kb, and affords 7.6X coverage of the 21.7 Gb LP genome. To demonstrate the efficacy of the library in gene isolation, we screened macroarrays with overgos designed from a pine EST anchored on LP chromosome 10. A positive BAC was sequenced and found to contain the expected full-length target gene, several gene-like regions, and both known and novel repeats. Macroarray analysis using the retrotransposon IFG-7 (the most abundant repeat in the sequenced BAC) as a probe indicates that IFG-7 is found in roughly 210,557 copies and constitutes about 5.8% or 1.26 Gb of LP nuclear DNA; this DNA quantity is eight times the Arabidopsis genome. In addition to its use in genome characterization and gene isolation as demonstrated herein, the BAC library should hasten whole genome sequencing of LP via next-generation sequencing strategies/technologies and facilitate improvement of trees through molecular breeding and genetic engineering. The library and associated products are distributed by the Clemson University Genomics Institute (www.genome.clemson.edu).This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose.
# Introduction
Loblolly pine (LP; Pinus taeda L.) is an organism of tremendous economic and ecological importance and a key representative of the conifers, an ancient lineage of plants that dominates many of the world's temperate and boreal ecosystems. LP's fast growth, amenability to intensive silviculture, and high-quality lumber/pulp have made it the cornerstone of the U.S. forest products industry and the most commonly planted tree species in Americaapproximately 75% of all seedlings planted each year are LPs [bib_ref] Deployment of genetically improved loblolly and slash pines in the South, Mckeand [/bib_ref]. Its ability to efficiently convert CO 2 into biomass and its widespread use as a plantation tree have also made LP a cost-effective feedstock for lignocellulosic ethanol production [bib_ref] Production of ethanol from carbohydrates from loblolly pine: A technical and economic..., Frederick [/bib_ref] and a promising tool in efforts to curb greenhouse gas levels via carbon sequestration [bib_ref] Belowground carbon dynamics in loblolly pine (Pinus taeda) immediately following diammonium phosphate..., Gough [/bib_ref].
Despite the importance of LP and other conifers, genomic sequence information for this taxon is extremely limited. Like other conifers, LP has a relatively huge genome -its 1C DNA content is reported at 21.7 Gb. Its long generation time, approximately eight years to sexual maturity, also poses an obstacle to tree improvement through traditional breeding techniques. Though molecular resources such as genetic maps [bib_ref] A consensus map for loblolly pine (Pinus taeda L.). I. Construction and..., Sewell [/bib_ref] [bib_ref] Identification of quantitative trait loci influencing wood property traits in loblolly pine..., Brown [/bib_ref] [bib_ref] Comparative genome and QTL mapping between maritime and loblolly pines, Chagne [/bib_ref] , a FISH-based karyotype [bib_ref] Reference karyotype and cytomolecular map for loblolly pine (Pinus taeda L.), Islam-Faridi [/bib_ref] , EST sequences [bib_ref] Anchored reference loci in loblolly pine (Pinus taeda L.) for integrating pine..., Brown [/bib_ref] [bib_ref] Water stressresponsive genes in loblolly pine (Pinus taeda) roots identified by analyses..., Lorenz [/bib_ref] [bib_ref] Expressed sequence tags from loblolly pine embryos reveal similarities with angiosperm embryogenesis, Cairney [/bib_ref] [bib_ref] Characterization of EST-SSRs in loblolly pine and spruce, Bérubé [/bib_ref] , and QTL maps are available [bib_ref] A consensus map for loblolly pine (Pinus taeda L.). I. Construction and..., Sewell [/bib_ref] [bib_ref] Identification of quantitative trait loci influencing wood property traits in loblolly pine..., Brown [/bib_ref] [bib_ref] Comparative genome and QTL mapping between maritime and loblolly pines, Chagne [/bib_ref] for LP, efficient tree improvement will ultimately require integration of EST, sequence polymorphism, gene expression, and genetic data with actual genomic sequence including noncoding regulatory regions missed by EST approaches.
To accelerate pine genomics, we constructed and initiated characterization of a bacterial artificial chromosome (BAC) library for the LP tree ''7-56,'' a valuable and widely used parent selection in various loblolly pine breeding programs. The utility of the library for gene isolation and genome characterization was verified by macroarray analysis and DNA sequencing. The 7-56 BAC library is a high quality resource that will expedite research on pine and conifers in general.
# Results and discussion
## Library construction and characterization
The completed library consists of 1,824,768 clones archived in 4752 384-well microtiter plates -to our knowledge this is the single largest BAC library ever made (see [fig_ref] Figure 1: Clone numbers [/fig_ref]. Two sets of replicate libraries were prepared and stored in separate 280uC freezer banks at the Mississippi Genome Exploration Laboratory (MGEL; www.mgel.msstate.edu) while the original was sent to the Clemson University Genomics Institute (CUGI; www.genome. clemson.edu) for distribution and remote storage. For distribution, the library has been gridded onto macroarrays using a 565 format in which 27,648 clones are double-spotted on each 22 cm 2 membrane. A complete set of macroarrays consists of 66 filters. However, screening of the library at MGEL was performed using macroarrays with a 464 gridding pattern (i.e., 18,432 doublespotted clones on each 22 cm 2 macroarray).
Unlike other plant and animal species for which we typically obtain clones with mean insert sizes in excess of 100 kb with only modest optimization, application of standard BAC library construction protocols resulted in LP clones with mean insert sizes ,75 kb. We used a variety of techniques to increase insert size including varying tissue sources, the enzymes for partial restriction digestion, the cloning vectors, the vector to insert ratio used in ligation, and the ligase concentration. Many of these steps led to minor increases in mean insert size, but ultimately were not sufficient to provide mean insert lengths $100 kb. The breakthrough that permitted realization of the .100 kb mean insert size goal came with the discovery and adoption of the ''preelectrophoresis'' procedure of Osoegawa et al. [bib_ref] An improved approach for the construction of bacterial artificial chromosome libraries, Osoegawa [/bib_ref]. In preelectrophoresis, agarose plugs containing DNA are placed in a dialysis tube, the tube is positioned in the center of a pulsed-field gel electrophoresis (PFGE) chamber, and the tube is exposed to a voltage that permits charged low molecular weight molecules to migrate out of the plugs. Pre-electrophoresis has been used to remove residual proteins from DNA plugs to help prevent inhibition of downstream processes such as ligation [bib_ref] An improved approach for the construction of bacterial artificial chromosome libraries, Osoegawa [/bib_ref]. In our experience, it appears to elute much of the low molecular weight DNA from the plugs which, in turn, appears to enhance restriction enzyme digestion and fragment separation via PFGE. Clones from the first 2650 384-well plates had an average insert size of 87 kb based on NotI digestion and PFGE analysis. However, with the addition of the pre-electrophoresis step, mean insert size was increased to 110 kb for plates 2651-4752. The insert size distributions of clones are shown in [fig_ref] Figure 2: Inserts of LP 7-56 BAC clones [/fig_ref]. The average insert size of the entire library is estimated at 96 kb. A typical gel containing NotI-digested clones from the latter half of the library is shown in [fig_ref] Figure 2: Inserts of LP 7-56 BAC clones [/fig_ref]. Probing of a Southern blot of NotI-digested clones with LP 7-56 genomic DNA was used to confirm that library inserts were indeed derived from pine (data not shown).
Of note, LP DNA appears to contain relatively few NotI sites as more than 80% of the NotI-digested clones examined yield a single insert band [fig_ref] Figure 2: Inserts of LP 7-56 BAC clones [/fig_ref]. In this regard, the LP clone NotI digestion pattern is similar to those observed for BAC libraries from dicotyledonous plants; monocots typically possess much higher densities of NotI sites [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref]. Based on examination of 95 gels (7626 clones), we estimate that the LP genome contains an average of one NotI site per 1000 kb of sequence.
As is standard in evaluating plant BAC libraries, we estimated the fraction of clones that lack inserts (i.e., false positives) and the fraction that contain chloroplast DNA. PFGE results indicate that roughly 5.7% of clones appear to be false positives. Hybridization of a 464 macroarray with pine chloroplast DNA probes revealed that about 0.6% of the LP clones contain chloroplast DNA, a midlow level of chloroplast contamination compared to other plant BAC libraries (range: 0.02-2.78%; see [bib_ref] Melon bacterial artificial chromosome (BAC) library construction using improved methods and identification..., Luo [/bib_ref] [bib_ref] Construction and characterization of a bacterial artificial chromosome (BAC) library of hexaploid..., Nilmalgoda [/bib_ref] [bib_ref] Construction of a BAC library of Rosa rugosa Thunb. and assembly of..., Kaufmann [/bib_ref] [bib_ref] Construction of a bacterial artificial chromosome library from the spikemoss Selaginella moellendorffii:..., Wang [/bib_ref] [bib_ref] Construction and characterization of two bacterial artificial chromosome libraries of pea (Pisum..., Coyne [/bib_ref] [bib_ref] BAC libraries construction from the ancestral diploid genomes of the allotetraploid cultivated..., Guimarães [/bib_ref] [bib_ref] The construction and characteristics of a BAC library for Cucumis sativus L, Gutman [/bib_ref] [bib_ref] Development of genomic resources for Citrus clementina: Characterization of three deep-coverage BAC..., Terol [/bib_ref] [bib_ref] Construction and characterization of two Citrus BAC libraries and identification of clones..., Baig [/bib_ref] [bib_ref] Characterization of a deep-coverage carrot (Daucus carota L.) BAC library and initial..., Cavagnaro [/bib_ref] [bib_ref] A BAC library of Beta vulgaris L. for the targeted isolation of..., Jacobs [/bib_ref] [bib_ref] Physical analysis of the complex rye (Secale cereale L.) Alt4 aluminium (aluminum)..., Shi [/bib_ref]. While macroarray screening with mitochondrial DNA was not performed, automated analysis of Sanger and 454 sequence reads [bib_ref] An automated, high-throughput sequence read classification pipeline for preliminary genome characterization, Chouvarine [/bib_ref] prepared from LP 7-56 nuclear DNA using our nuclear isolation protocol revealed that mitochondrial DNA contamination is 10 to 100 times less frequent than contamination from chloroplast DNA (unpublished data).
Assuming that 0.057 of clones are false positives and 0.006 of clones contain chloroplast DNA, the number of clones containing pine nuclear DNA is approximately [i.e., (1-(0.057+0.006))* 1,824,768 = ] 1,709,808. Since the LP genome is 21.7 Gb, a library containing 1,709,808 pine nuclear DNA-containing clones with 96 kb inserts affords coverage of roughly 7.6 genome equivalents [i.e., (1,709,808 clones N 96,000 bp) 4 21.7610 9 bp = 7.6].
A 7.6X library affords a 99.93% probability that any locus of interest will be found in the library at least once.
Identification, sequencing, and annotation of a genecontaining BAC A major use of BAC libraries is in the isolation of intact genes including non-coding regions missed by cDNA/EST sequencing approaches [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref] [bib_ref] BAC as tools for genome sequencing, Zhang [/bib_ref]. To demonstrate the utility of the library for this task, we selected an EST (GenBank AA739884) that has been mapped to LP chromosome 10 [10] and displays significant homology (S' = 392) with a Picea glauca late embryogenesis abundant (LEA) protein. Overgos designed from the marker were used to screen the first two 464 macroarrays of the library [fig_ref] Figure 3: Screening LP 7-56 BAC macroarrays [/fig_ref] , and PCR was used to check for the presence of the marker in clones exhibiting probe hybridization. A positive clone, PT_7Ba_00066 J18, which has an insert size of 86.5 kb, was sheared, bar coded, and added to a solution containing numerous differentially bar coded chloroplast genomes. From relatively lowcoverage (20X) Illumina sequencing, the BAC was assembled into 158 contigs (not including vector contigs) with a combined length of 85,504 bp, i.e., roughly 98.8% of the estimated insert size. The sequences that resulted from this BAC were deposited to GenBank (Accession Number HQ141589). A 4048 bp contig in the BAC contains the target full-length LEA gene (see below for more information). [fig_ref] Figure 4: BLAST-based analysis of the LP 7-56 clone PT_7Ba_00066 J18 sequence [/fig_ref] summarizes the results of an initial sequence analysis of the BAC. The majority of the BAC sequence exhibits no recognizable homology to any annotated plant sequences in GenBank. With regard to retroelements, our results indicate that long-terminal repeat (LTR) retrotransposons account for at least 18.8% of the total BAC length with the majority of such sequences showing their most significant (S' .50) homology to the previously described Gypsy subfamily LTR elements IFG-7 [bib_ref] IFG, a gypsy-like retrotransposon in Pinus (Pinaceae), has an extensive history in..., Kossack [/bib_ref] , PpRT1, [bib_ref] PpRT1, the first complete gypsy-like retrotransposon isolated in Pinus pinaster, Rocheta [/bib_ref] , and/or Corky (GenBank Accession No. EU862277.1). No recognizable Copia subfamily LTR element was found in the BAC, and indeed our analysis of this BAC coupled with characterization of random sequences we have generated from pine via 454 and capillary sequencing indicates that the LP genome contains far more LTR Gypsy elements than LTR Copia elements (unpublished data). DNA transposons were not identified by homology (BLAST; [bib_ref] Gapped BLAST and PSI-BLAST: A new generation of protein database search programs, Altschul [/bib_ref] searches. However, using the program FINDMITE [bib_ref] Eight novel families of miniature inverted repeat transposable elements in the African..., Tu [/bib_ref] we identified 122 potential miniature inverted-repeat transposable elements (MITEs) in the BAC. MITEs are non-autonomous DNA transposons characterized by terminal inverted repeats, target site duplications, and no coding sequence [bib_ref] Identification of miniature inverted-repeat transposable elements (MITEs) and biogenesis of their siRNAs..., Kuang [/bib_ref]. Three of the putative MITEs appear to be portions of retroelements. The other putative MITEs are currently being further investigated, though it is probable that most of these sequences do not actually represent true MITE families. However, four instances were found where a MITE recognized by FINDMITE exhibited .80% sequence identity to another region in the BAC not recognized by the FINDMITE program. These instances may represent four different MITE families where duplicated copies have undergone moderate divergence, and indeed these sequences are priorities in our MITE investigations.
Putative genes, i.e., sequences exhibiting significant alignment (S' $50) to known genes and/or cDNA sequences, constitute approximately 4.1% of the BAC sequence (including the putative LEA gene -see below). However, only the LEA gene appears to have a complete coding sequence.
## Annotation of the lea gene
The targeted LEA gene sequenced in this study (i.e., LBAC) was found within a 4026 bp contig. A BLAST (blastn) comparison of the contig with the NCBI non-redundant (nr) database produced a top hit (S' = 675) to the complete coding sequence of a Pinus halepensis LEA mRNA (GenBank Accession No. AY705798.1). Examination of the aligned portions of LBAC with the P. halepensis LEA mRNA suggests that LBAC is composed of two exons and one intron (see [fig_ref] Figure 1: Clone numbers [/fig_ref]. The coding sequences from the two species are the same length but contain six interspecific single nucleotide differences [fig_ref] Figure 1: Clone numbers [/fig_ref] which are predicted to result in four amino acid differences between the predicted P. taeda and P. halepensis proteins [fig_ref] Figure 5: Alignment of the predicted amino acid sequences of the Pinus taeda 7-56... [/fig_ref].
BLASTX analysis of coding sequence of the LP LEA gene indicates that a 50 amino acid region (positions 43 through 92) exhibits significant homology (S' = 68) to pfam03242 [bib_ref] The Pfam protein families database, Bateman [/bib_ref] , the LEA 3 family of proteins [fig_ref] Figure 2: Inserts of LP 7-56 BAC clones [/fig_ref]. The LEA proteins, including the LEA 3 family, have been implicated in response to water stress, though the exact function of these proteins is not clear [bib_ref] A repeating 11-mer amino acid motif and plant desiccation, Dure [/bib_ref] [bib_ref] Plant responses to water deficit, Bray [/bib_ref] [bib_ref] Highly hydrophilic proteins in prokaryotes and eukaryotes are common during conditions of..., Garay-Arroyo [/bib_ref] [bib_ref] Induction of desiccation tolerance in plant somatic embryos: how exclusive is the..., Hoekstra [/bib_ref].
## Comparison of the lbac gene and ests
The LBAC gene was also aligned with Pinus taeda sequences in the NCBI non-human, non-mouse EST database (est_others). The top 250 blastn hits (S' = 719 to 1700) produced via blastn comparison fall into two structurally distinct groups.
Variant 1. Exactly 239 of the 250 top hits have exon sequences that are 100% identical to the LBAC exons. Of these, six (i.e., 2.5%) contain a putative intron in their sequences; each of these introns is identical to the LBAC intron. Consequently, it is probable that the 239 variant 1 ESTs are products of the LBAC locus. Moreover, the presence of variant 1 transcripts with and without an intron indicates that there is a certain level of alternative and/or inefficient splicing of LBAC/variant 1 transcripts. A consensus sequence including the intron (deemed Consensus Sequence, Variant 1 or CSV1; see [fig_ref] Figure 1: Clone numbers [/fig_ref] was A B Note that IFG-7 is found in many, but not all BAC clones. Also note that some clones appear to contain higher densities of the retroelement. Eleven of the EST sequences exhibited consistent sequence differences from LBAC and CSV1. As shown in [fig_ref] Figure 1: Clone numbers [/fig_ref] , the consensus sequence for these variant 2 sequences (i.e., Consensus Sequence, Variant 2 or CSV2) possesses 20 single nucleotide differences when compared to CSV1. Moreover, all of the variant 2 sequences possess the region marked as an intron in CSV1/LBAC. Of note, this region of variant 2 ESTs contains a 12 nt deletion which may account for the apparent 100% retention of the intron in variant 2 transcripts.
While we speculate that the variant 1 sequences are products of the LBAC gene, it is unclear whether variant 1 and variant 2 transcripts represent products of different alleles of the same gene or products of paralogous genes. Of note, the sequenced Pinus halepensis LEA mRNA (GenBank Accession No. AY705798.1) lacks the putative intron (as with most variant 1 sequences) but shares a 13 nt insertion immediately after the stop codon with variant 2 ESTs [fig_ref] Figure 1: Clone numbers [/fig_ref].
## Use of the lp bac library to characterize the ifg-7 retrotransposon
A BAC library is an excellent resource for the initial characterization of a genome, especially in cases when sequence information is limited [bib_ref] Integration of cot analysis, DNA cloning, and high-throughput sequencing facilitates genome characterization..., Peterson [/bib_ref]. All plant genomes studied thus far contain numerous transposable elements, and the proportion of these elements in genomes tends to increase with genome size [bib_ref] Repeated big bangs and the expanding universe: Directionality in plant genome size..., Hawkins [/bib_ref]. To begin genome characterization of LP, we chose to look at the distribution of IFG-7 [bib_ref] IFG, a gypsy-like retrotransposon in Pinus (Pinaceae), has an extensive history in..., Kossack [/bib_ref] , the most abundant retroelement in the BAC we sequenced (i.e., PT_7Ba_00066 J18). A 568 bp pine sequence exhibiting 99% sequence identity to the Pinus radiata IFG-7 retrotransposon was used to probe a 464 LP macroarray. The differences in hybridization intensities between positive clones suggest that higher intensity clones likely harbor multiple copies of the retrotransposon [fig_ref] Figure 3: Screening LP 7-56 BAC macroarrays [/fig_ref]. Using the densitometry method of Peterson et al. [bib_ref] Integration of cot analysis, DNA cloning, and high-throughput sequencing facilitates genome characterization..., Peterson [/bib_ref] with minor modification (see Materials and Methods), we calculated that there are approximately 210,557 copies of IFG-7 in the LP genome which collectively account for about 5.8% (i.e., 21.7 Gb N 0.058 = 1.26 Gb) of pine nuclear DNA. This amount of DNA is roughly equivalent to about eight Arabidopsis thaliana (1C = 157 Mb;genomes.
An initial glance at IFG-7 hybridization to macroarrays suggests that the element is found in clusters, i.e., it is not distributed randomly throughout the genome [fig_ref] Figure 3: Screening LP 7-56 BAC macroarrays [/fig_ref]. To test this hypothesis, we used the probabilistic ''urn model'' method applied in Shan et al [bib_ref] A bacterial artificial chromosome library for the Australian saltwater crocodile (Crocodylus porosus)..., Shan [/bib_ref] and described in detail in Holst [bib_ref] Limit theorems for some occupancy and sequential occupancy problems, Holst [/bib_ref]. Using our estimate of the number of copies of IFG-7 in the LP genome (i.e., 210,557), and an average insert size of 96 kb, each macroarray, after adjustment for false positives and chloroplast-containing clones, represents about 0.076X coverage of the LP genome [i.e., Interspecific single nucleotide differences highlighted in light gray do not result in a change in overall polarity and/or charge. However, differences highlighted in black result in a polarity and/or charge difference (e.g., A = nonpolar, neutral; T = polar, neutral; E = polar, negative). doi:10.1371/journal.pone.0016214.g005 (18,432 clones/macroarray N 0.937 nuclear DNA-containing clones N 96 kb)/21.7 Gb = 0.076] and should contain 16,002 copies of IFG-7 (i.e., 210,557 copies N 0.076 = 16,002). If the distribution of IFG-7 were indeed random, we would expect that the distribution of clones lacking IFG-7 elements (i.e., lacking hybridization signal) would approximate normality; in such cases, the mean number of clones expected to lack an IFG-7 element and the theoretical standard deviation (SD) can be estimated using Theorem 2 of Holst [bib_ref] Limit theorems for some occupancy and sequential occupancy problems, Holst [/bib_ref]. Specifically, the expected standard deviation for a normal distribution, strongly reinforcing our hypothesis that the distribution of IFG-7 is not random. Of note, non-random distributions of transposable elements has been reported for many plant species [bib_ref] Non-random distribution of transposable elements in the nuclear genome of plants, Capel [/bib_ref] [bib_ref] Nested retrotransposons in the intergenic regions of the maize genome, Sanmiguel [/bib_ref] [bib_ref] Single integration and spread of a Copia-like sequence nested in rDNA intergenic..., Chester [/bib_ref] [bib_ref] Cytological characterization of sunflower by in situ hybridization using homologous rDNA sequences..., Talia [/bib_ref].
## Utilization of the bac library in genome sequencing of lp
The U.S. Department of Agriculture recently announced plans to fund draft sequencing of the LP genome. While it is not publicly known how this decision was reached, it is likely that the success of former and current investments by the National Science Foundation, Department of Energy, and USDA in LP genome research, including NSF funding of production of the 7-56 BAC library, created the scientific framework on which such an effort could be justified. If the LP 7-56 BAC library is used in sequencing the LP genome, it is unlikely that traditional BAC-based physical mapping approaches (including BAC end sequencing and BAC fingerprinting) will be employed to a large extent as the LP genome is simply too big for a clone-by-clone, physical map-based sequencing approach to be cost effective. However, the organized nature of a BAC library (specifically the storage of individual clones in indexed plates/wells) affords a mechanism that can be utilized in simplifying sequence assembly. In short, one can sequence pools of BACs in lieu of (or preferably, in addition to) random genomic DNA. Each pool contains a specific number of BACs (e.g., 1000) and hence represents a fairly small portion of the genome. The probability of two homologous or paralogous loci being represented within a pool is small, thus limiting assembly problems associated with diploidy/polyploidy and large gene families, respectively. Though a repeat sequence may be found millions of times within a genome, its representation in a BAC pool is likewise greatly reduced as are its effects on assembly of sequences in that pool. Of particular importance, the clones in each pool are archived allowing the pool to be reconstructed if necessary. Moreover, one can further refine the assembly process by using multiplexing strategies to produce partially overlapping BAC pools and/or bar coding individual BACs or BAC subpools (e.g., [bib_ref] De novo 454 sequencing of barcoded BAC pools for comprehensive gene survey..., Steuernagel [/bib_ref] [bib_ref] Highly-multiplexed barcode sequencing: an efficient method for parallel analysis of pooled samples, Smith [/bib_ref].
# Conclusion
To accelerate genomics research in pine, we constructed a pine HindIII BAC library that affords roughly 8.1X coverage of the LP genome. This resource should allow isolation and sequencing of most pine loci and represents a means of facilitating physical mapping, gene isolation, and genome sequencing. It is anticipated that the BAC library will be a key resource utilized in sequencing the loblolly pine genome.
# Materials and methods
# Plant material
Loblolly pine genotype 7-56 (origin: Williamsburg County, South Carolina, original tree is deceased) needles were provided by International Paper from a single ramet growing at their Southlands Forest near Bainbridge, Georgia. Prior to selecting this ramet, six short simple repeat (SSR) marker loci -specifically PtTX2123, PtTX4058, PtTX4093, PtTX4181, PtTX3013 and PtTX3052-and the CAD-null marker [bib_ref] Differences in wood density and growth of fertilized and non-fertilized loblolly pine..., Yu [/bib_ref] were used to genotype ten 7-56 and two non-7-56 ramets in a double blind experiment. Protocols for SSR genotyping are given in Gonzalez-Martinez et al. [bib_ref] Association genetics in Pinus taeda L. I. Wood property traits, González-Martínez [/bib_ref]. All 7-56 ramets, including the one selected as a tissue donor, were found to have the same multi-locus SSR genotype, and this genotype differed from the negative controls. In addition, these SSR genotype data matched our data from previous independent sample collections of 7-56 indicating that these samples were indeed genotype 7-56. Upon harvest needles were wrapped in moist paper towels, placed in large sealable plastic bags, and shipped on ice via overnight courier. Bags of needles were stored at 4uC until use.
## Library construction and storage
Construction of the library was performed according to Peterson et al. [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref] with the following modifications:
(1) Pre-electrophoresis. Prior to size selections, agarose plugs containing LP genomic DNA were suspended in 0.5X TBE buffer and sealed inside a Spectra/Por MWCO 12-14,000 dialysis tube (Spectrum Laboratories) that was placed in the center of the hexagonal electrode array in a Bio-Rad DRIII CHEF pulse-field gel electrophoresis (PFGE) apparatus. The PFGE unit was run for 8 hours with pulse ramping of 1 to 4 sec, an included angle of 120u, and a voltage gradient of 6 V/cm in 0.5X TBE buffer at 14uC.
(2) Size selections. Gold Agarose (Seakem) in 0.25X TBE was used in all size selection steps, along with 0.25X TBE as running buffer. After the first selection step as described in Peterson et al. [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref] , the gel was allowed to run for an addition 9 h using a 3 s start switch time and a 5 s end switch time. This additional electrophoresis period increases the resolution of separation. Agarose containing DNA molecules between 120 to 220 kb was excised and loaded into a second gel as described [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref] except the total run time was 14 rather than 18 h. At the end of the run, only agarose containing DNA in the range of 120 to 220 kb was excised for electroelution.
(3) Electroelution. The DNA elution procedure of Peterson et al. [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref] was followed except that the process was performed for 2 h, and buffer in the upper chamber of the instrument was replaced every 30 min.
(4) Ligation and transformation. The eluted DNA was quantified with a NanoDrop ND-1000 (Thermo Fisher Scientific) spectrophotometer. The ligation and transformation steps were carried out as illustrated in Peterson et al. [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref] except that the vector used was HindIII-ready pIndigoBAC5 (Epicentre Technologies). Ligation was carried out in a reaction containing 30 ng dephosphorylated vector DNA, 600 ng size-selected insert DNA, 15 mL of 10X ligase buffer, 2 mL of 2,000 units/mL T4 DNA ligase (New England Biolabs), and deionized water to produce a solution with a final volume of 150 mL.
Transformations, picking of colonies, plate replication, and storage of plates at 280uC were performed as previously described [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref]. All microtiter plates containing clones were affixed with labels containing alphanumeric and bar code identifiers. The library was given the name PT_7Ba in accordance with MGEL and CUGI naming conventions (see http://www.mgel.msstate. edu/dna_libs.htm).
## Sampling and analysis of clones
The average molecular weight of the inserts and the percentage of vectors without inserts were estimated based on sampling of 82 clones from every 50 th plate. In brief, for each plate a manual 96pin plate replicator (V & P Scientific) was used to transfer bacteria from 96 of the wells (offset A) into media in two AutoGen 96-well plates. The wells in the AutoGen plates were filled with 1.2 mL Terrific Broth (Difco) supplemented with 30 mg/mL of chloramphenicol. The plates were incubated at 37uC for 18-24 hours with shaking at 250 rpm. Bacterial cells from duplicate plates were pooled and the BACs were isolated from the cells using an AutoGen Prep 960 (AutoGen) robot. After air drying, recovered DNA pellets each were dissolved in 15 mL of a solution of 1.5 mL of 10 units/mL NotI (New England Biolabs), 0.1 mL of 10 mg/mL BSA, 1.5 mL of 10X Buffer 3 (New England Biolabs; 500 mM Tris-HCl, 1000 mM NaCl, 100 mM MgCl 2 10 mM dithiothreitol), and 12.9 mL double-distilled water). Digestion was allowed to proceed at 37uC for 5-14 hours. The digested BACs were run on a CHEF gel as previously described [bib_ref] Construction of plant bacterial artificial chromosome libraries: An illustrated guide, Peterson [/bib_ref]. The New England Biolabs Lambda PFGE ladder was used as a standard when estimating the sizes of inserts.
## Gridding and hybridization of high density filters
Macroarrays were prepared using a Genetix QPixII robot. After the clones were spotted onto membranes, they were placed on LB agar trays (clone side up) and allowed to grow overnight at 37uC. Membranes were fixed by incubation in Solution 1 (0.5 N NaOH, 1.5 M NaCl) and Solution 2 (1.5 M NaCl, 0.5M Tris Cl) for 7 min each. The membranes were allowed to dry for 1 h, washed in 0.4 N NaOH for 20 min, and washed for 7 min in aqueous 750 mM NaCl, 50 mM NaH 2 PO 4 , and 50 mM Na 2 EDTA. The membranes were pre-hybridized for at least 3 h, or overnight if they had not been hybridized before, in hybridization buffer (0.25 M Na 2 HPO 4 , pH 7.2; 7% w/v SDS; 1 mM EDTA; and 1% w/v BSA) in a hybridization oven (SciGene, Model 400) using a rotation setting of 4. The membranes were separated by nylon mesh sheets (Fisher Scientific, Pittsburgh, PA) and rolled to fit into hybridization bottles. A maximum of five membranes were placed in each bottle, along with 50 mL of temperature-equilibrated hybridization buffer. Hybridization was carried out at 55uC for overgos and 65uC for longer probes (see below for probe labeling and concentrations used in hybridization experiments).
To identify a BAC clone containing the LEA gene, two macroarrays representing the first 96 microtiter plates of the library were screened with overgo probes designed from an EST marker found on pine chromosome 10 [bib_ref] Anchored reference loci in loblolly pine (Pinus taeda L.) for integrating pine..., Brown [/bib_ref] ; GenBank Accession No. AA739884). The design and preparation of overgo sequences is described in [fig_ref] Table S1 Additional: Methods. (DOCX) [/fig_ref]. The program MacroArray Reader, developed at MGEL (manuscript in preparation), was used to identify the locations of positive clones on the high density membranes. PCR was used to verify the presence of the LEA gene in the PT_7Ba_0006 J18 BAC using EST-specific primers (sequences provided in [fig_ref] Table S1 Additional: Methods. (DOCX) [/fig_ref].
Pine genomic DNA and chloroplast clones were labeled using the Megaprime DNA Labeling System (GE Healthcare). The genomic DNA was digested with HindIII for 2 h, precipitated with ethanol, and dissolved in double-distilled water prior to labeling. The chloroplast probes were obtained from clones, in our possession (unpublished data) that align with nucleotides 27939-28367, 60489-61592, 79999-81133, and 117813-118274 of the 119,707 bp Pinus thunbergii chloroplast genome (GenBank Accession No. NC_001631). Twenty five nanograms of each chloroplast probe and 100 ng of genomic DNA were labeled with 32 P-dCTP using a random primer labeling technique [bib_ref] A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity, Feinberg [/bib_ref] [bib_ref] A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity''...., Feinberg [/bib_ref]. Labeling was performed at 37uC for 1-3 h and the unincorporated nucleotides were removed using the QiaQuick Nucleotide Removal kit.
Hybridization of probes to macroarrays, membrane washing, and visualization of positive hybridization sites were performed as described in [fig_ref] Table S1 Additional: Methods. (DOCX) [/fig_ref].
## Southern blot
To further verify that the BAC library we constructed contained pine genomic DNA inserts, one of the gels used for insert size determination was transferred to a nylon membrane and probed with labeled pine genomic DNA as described previously.
## Sequencing of lp bac clone pt_7ba_00066 well j18
Approximately 1 mg of LP BAC Clone PT_7Ba_0006 J18 was prepared for sequencing on an Illumina GAII using custom barcoding adapters that enable multiplex template sequencing [bib_ref] Multiplex sequencing of plant chloroplast genomes using Solexa sequencing-by-synthesis technology, Cronn [/bib_ref]. This specific LP BAC was barcoded with ''AGCT'' and represented ,1/10 of the sample pool. Standard Illumina chemistry was used for cluster generation. Specifically, 5 pM of the multiplex library was subjected to 36-amplification cycles to give single-end sequencing reads [bib_ref] Accurate whole human genome sequencing using reversible terminator chemistry, Bentley [/bib_ref]. The sequencing was performed at the Oregon State University Center for Gene Research and Biocomputing. A total of 6.62 million clusters passed purity filtering, with 387,855 clusters attributable to this LP BAC. Barcodes were removed in silico from the 59 ends of microreads, and the remaining 32 bp microreads were assembled de novo with Velvet [bib_ref] Velvet: Algorithms for de novo short read assembly using de Bruijn graphs, Zerbino [/bib_ref] using the following parameters: cov_cutoff = 10; min_contig_lgth = 75. Velvet produced 159 contigs $75 bp in length with an N50 of 2.97 kbp. Excluding vector sequence, these contigs had a cumulative length of 85.4 kb.
# Bac sequence analysis
The contigs from LP BAC clone PT_7Ba_00066 J18 were used as queries in BLAST (blastn) searches against plant sequences in the GenBank non-redundant (nr) and non-human, non-mouse EST (est_others) databases. The location of each hit with a bit score of 50 or greater was aligned with the contig sequence. Regions of the contig were manually classified based upon their top GenBank hits. To identify putative MITES, we used the contig sequences as queries for FINDMITE [bib_ref] Eight novel families of miniature inverted repeat transposable elements in the African..., Tu [/bib_ref] with the TIR (terminal inverted repeat) length set at 11 and a tolerance of up to two bases mismatches per TIR.
## Characterization of the sequenced lea gene
The LEA gene sequenced as part of this work (LBAC) was compared (blastn) with P. taeda ESTs in the non-human, nonmouse EST database (est_others). The EST sequences representing the top 250 hits were extracted and aligned with the LBAC using MUSCLE [bib_ref] MUSCLE: multiple sequence alignment with high accuracy and high throughput, Edgar [/bib_ref]. The NCBI BLASTX tool was used to compare the predicted amino acid sequence of the LBAC gene product with previously characterized proteins.
## Determination of copy number of repeat families
Repeat copy numbers were estimated from macroarrays as described in Peterson et al. [bib_ref] Integration of cot analysis, DNA cloning, and high-throughput sequencing facilitates genome characterization..., Peterson [/bib_ref] with modifications to account for false positives and clones containing chloroplast DNA (see [fig_ref] Table S2: Calculating the copy number and genome percentage of IFG-7 based on densitometric... [/fig_ref] for calculations). [fig_ref] Figure 1: Clone numbers [/fig_ref] Comparison of LBAC, the consensus sequences of the two major EST variants discovered through BLAST alignment (i.e., CSV1 and CSV2), and the P. halepensis LEA EST sequence (PHLE). For each sequence, the start codon is highlighted in light blue, the stop codon in pink, and the intron (if any) in light orange. The exons in CSV1 are identical to those in the LBAC gene. 2.5% of the ESTs used to create CSV1 contained a putative intron with 100% sequence identity to the intron in LBAC (light orange highlight). CSV2 was derived from 11 sequences that showed significant and consistent differences from CSV1/LBAC. The region believed to represent an intron in CSV1/LBAC was present in all transcripts used in generating CSV2, and indeed it may be that all mature sequences produced from this locus/allele contain the ''intronic'' region (hence this region is not highlighted as an intron in CSV2). Compared to CSV1, CSV2 contains a deletion in the putative intron region (bases 225-236), which may account for improper splicing of the CSV2 transcript, and a 13 nt insertion immediately after the stop codon (bases 455-467). The 13 nt insertion is also observed in PHLE. Single nucleotide differences between a particular sequence and the LBAC sequence are highlighted in yellow.
## Supporting information
(TIF) [fig_ref] Figure 2: Inserts of LP 7-56 BAC clones [/fig_ref] The LP LEA protein shows similarity to the LEA 3 family of proteins (pfam03242). (TIF)
[fig] Figure 1: Clone numbers (in millions of clones) for the largest BAC libraries. The Pinus taeda BAC library is three times larger than the second largest library. Bar colors represent the center/institute at which the library was made -red represents the Arizona Genomics Institute (AGI; www.genome.arizona.edu), blue the Children's Hospital of Oakland (CHORI; http://bacpac.chori.org), purple the Clemson University Genomics Institute (CUGI; www.genome.clemson.edu), and green the Mississippi Genome Exploration Laboratory (MGEL; www.mgel.msstate.edu). doi:10.1371/journal.pone.0016214.g001 [/fig]
[fig] Figure 2: Inserts of LP 7-56 BAC clones. Insert size distribution of clones from (A) plates 1-2650, (B) plates 2651-4752, and (C) the library as a whole. (D) A typical agarose pulsed-field gel showing NotI digests of clones from the latter half of the library. The New England Biolabs PFGE Lambda Ladder is in the lane at the far left. A 7.5 kb vector band is visible at the bottom of each sample lane. doi:10.1371/journal.pone.0016214.g002 [/fig]
[fig] Figure 3: Screening LP 7-56 BAC macroarrays. (A) Use of the BAC library in gene isolation. The autoradiogram shows hybridization of an overgo probe linked to the LEA gene on LP chromosome 10 to a double-spotted BAC clone potentially containing a full length LEA gene. The positive BAC clone, PT_7Ba_00066 J18, was sequenced and indeed found to contain an intact LEA gene. (B) Hybridization of a 464 membrane with the IFG-7 retroelement. [/fig]
[fig] Figure 4: BLAST-based analysis of the LP 7-56 clone PT_7Ba_00066 J18 sequence. doi:10.1371/journal.pone.0016214.g004 [/fig]
[fig] Figure 5: Alignment of the predicted amino acid sequences of the Pinus taeda 7-56 LEA gene (top) and the Pinus halepensis 15r LEA mRNA (bottom; GenBank Accession No. AY705798.1) suggest that the proteins differ at four amino acids (highlighted residues). [/fig]
[table] Table S1 Additional: Methods. (DOCX) [/table]
[table] Table S2: Calculating the copy number and genome percentage of IFG-7 based on densitometric analysis of a macroarray. Based on Peterson et al. [42] Supplementary Documents. Aqua shaded cells contain data generated in the current study. Violet shaded cells contain data from the literature. (TIF) [/table]
|
The Impact of the SARS-CoV-2 Pandemic on Healthcare Provision in Italy to non-COVID Patients: a Systematic Review
Background: Italy has been one of the countries most affected by the SARS-CoV-2 pandemic, and the regional healthcare system has had to quickly adapt its organization to meet the needs of infected patients. This has led to a drastic change in the routine management of noncommunicable diseases with a potential long-term impact on patient health care. Therefore, we investigated the management of non-COVID-19 patients across all medical specialities in Italy. Methods: A PRISMA guideline-based systematic review of the literature was performed using PubMed, Embase, and Scopus, restricting the search to the main outbreak period in Italy (from February 20 to June 25 2020). We selected articles in English or Italian that detailed changes in the Italian hospital care for non-COVID-19 patients due to the pandemic. Our keywords included all medical specialities combined with our geographical focus (Italy) and COVID-19. Results: Of the 4643 potentially eligible studies identified by the search, 247 were included. A decrease in the management of emergencies in non-COVID patients was found together with an increase in mortality. Similarly, non-deferrable conditions met a tendency toward decreased diagnosis. All specialities have been affected by the re-organization of healthcare provision in the hub-and-spoke system and have benefited from telemedicine. Conclusions: Our work highlights the changes in the Italian public healthcare system to tackle the developing health crisis due to the COVID-19 pandemic. The findings of our review may be useful to analyse future directions for the healthcare system in the case of new pandemic scenarios.Pag. 2 / 37 Citation: Lugli G., Ottaviani M. M., Botta A., Ascione G., Bruschi A., Cagnazzo F., Zammarchi L., Romagnani P., Portaluri T. The impact of the SARS-CoV-2 pandemic on healthcare provision in Italy to non-COVID patients: a systematic review. Mediterr J Hematol Infect Dis 2022, 14(1): e2022012, http://dx.
# Introduction.
Since the first case of the novel coronavirus was reported in Wuhan, China, in December 2019, a viral infection spread at an alarming rate worldwide. On January 30, 2020, the World Health Organization (WHO) described COVID-19 as a Public Health Emergency of International Concern, and by , it was officially declared a pandemic.Italy was the first European country to be affected by , with the first case being diagnosed on February 20 in a man living in the province of Lodi (NorthWest Italy). [bib_ref] Coronavirus Disease 2019 (COVID-19) in Italy, Livingston [/bib_ref] The epidemic went on to affect all regions in Italy, with higher incidence rates in the north. The peak of the COVID-19 epidemic in Italy was reached in the last week of March, with over 5500 new cases per day.
Since then, there has been a gradual decline due to strict containment measures that shaped the Italian lockdown phase. However, especially during the first phase of the epidemic, the outbreak put the Italian National Health System (Servizio Sanitario Nazionale, SSN) under unprecedented pressure.
In an attempt to direct the available resources at counteracting and limiting the effects of the pandemic, deferrable and non-urgent medical activities were suspended. On the other hand, patients with lifethreatening conditions, such as myocardial infarction and stroke, or chronic conditions, such as diabetes, retained the right to their medical needs being met.
In these circumstances, several medical domains have been constrained by different resource allocations with unpredictable long-term consequences on patient health care. [bib_ref] Telemedicine in the Era of COVID-19, Portnoy [/bib_ref] [bib_ref] The definition and identification of need for health care, Acheson [/bib_ref] [bib_ref] COVID-19 and Italy: what next?, Remuzzi [/bib_ref] Here, we present a systematic review of the literature, which illustrates the direct and indirect effects of the COVID-19 pandemic on the management of non-COVID patients across all medical specialities.
Methods. This systematic review was performed in accordance with PRISMA guidelines. [bib_ref] Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement, Moher [/bib_ref] The search was conducted on June 25 2020, on three databases: PubMed, Embase, and Scopus, without any date restriction. All the keywords were investigated within the title and abstract in both "AND" and "OR" combinations. Our keywords included all medical specialities (and potential synonyms) combined with our geographical focus (Italy or Italian) and COVID-19. The full search strategy is reported in Supplementary Material [fig_ref] Table 1: Management of deferrable conditions and telemedicine during COVID-19 pandemic in Italy [/fig_ref]. Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.
Selection of the Studies. The literature search returned original papers published between 1979 and 2020especially for the keyword "coronavirus." Since our focus was the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and the first positive case in Italy was detected on February 20, the literature search was restricted to the period from February 20 to June 25 2020. The databases were queried via an R script on their respective APIs, checked and cleaned for duplicates (via title, DOI and/or database id), and exported into Excel.
In the second stage, studies were selected based on their titles and abstracts: each study was independently evaluated by three different raters (AnB, GA, GL). When there was a lack of agreement among the screeners, ensemble majority voting was used for the final decision. The full texts of the selected papers were thus analyzed by two reviewers in terms of relevance and inclusion/exclusion criteria (MMO and GA). When these reviewers disagreed over the inclusion or exclusion of a paper, a third reviewer was responsible for the final decision (GL). In addition, the reference lists of selected papers were reviewed in order to find pertinent studies not identified during the initial search.
Inclusion and Exclusion Criteria. The simultaneous cooccurrence of the following characteristics was considered for the inclusion of articles: (i) articles focusing on the SARS-CoV-2 infection/COVID-19 disease; (ii) articles focusing on the impact on patients based in Italy or on the Italian hospital organization; (iii) articles detailing COVID-19-associated changes in the Italian hospital care for non-COVID-19 patients. All the investigated articles were published in English or Italian.
Type of Studies. Original papers, editorials, comments, research letters, case series, and studies focusing on non-COVID patients in Italy were included.
Data Extraction and Quality Assessment. Data were extracted from the papers by one of the investigators (TP) and were subsequently checked for accuracy by other reviewers (GL, AnB). Disagreements regarding data extraction among reviewers were solved by consensus. Extracted data included: type of medical specialty involved (surgical, medical, or public health), geographical location (north, south, centre or nationwide), type of patients (COVID/non-COVID), type of study (article or research letter/comment/editorial). Unfortunately, no quality assessment was possible as over 32% of entries were not articles but consisted of comments, research letters, opinions or editorialsfor which no quality guidelines are available.
Investigated Outcomes. This systematic review investigated the impact of the COVID-19 pandemic on patients' healthcare provision and hospital organization in Italy since the day the country was put in lockdown.Our primary goal was to identify potential short-term and long-term effects on the health of non-COVID patients. Our secondary goals were to identify: (i) organizational and/or clinical settings and decisions that were particularly effective (or counterproductive) during the pandemic; and (ii) similarities and differences across medical specialities and regional areas.
# Results.
The results are shown in. After searching the databases, we identified 4643 papers from three different databases. Database merges and the removal of duplicates resulted in 1262 records, of which 100 were immediately removed as they were not related to COVID-19 (articles published before the pandemic in Italy). A total of 1162 records were then screened: 166 were removed as not relevant to Italy; 534 were removed as they referred to COVID-19 patients rather than non-COVID-19 patients. A total of 247 were deemed eligible, of which 81 consisted of comments/letters/opinions/editorials.
Oncology and radiotherapy were the most represented category , followed by surgery (24 for general surgery, 9 for neurosurgery, 2 for cardiosurgery, 2 for vascular surgery, and 3 other types), cardiology , and dermatology 14. There was one paper each for rheumatology and microbiology.
Overall, 133 papers were related to clinical disciplines, 89 to surgery, and 24 to services. In terms of geographical distribution, many papers provided general recommendations without a specific geographical identification [bib_ref] COVID-19) outbreak: what the department of endoscopy should know, Repici [/bib_ref]. Lombardy was the most represented region [bib_ref] Pediatric endoscopic procedures during the COVID-19 pandemic: an Italian center experience, Duci [/bib_ref] , followed by Lazio [bib_ref] Unpredictable Fall of Severe Emergent Cardiovascular Diseases Hospital Admissions During the COVID-19..., Mauro [/bib_ref] and Emilia Romagna [bib_ref] Critical Organizational Issues for Cardiologists in the COVID-19 Outbreak: A Frontline Experience..., Stefanini [/bib_ref]. In the south, Campania was the most represented region [bib_ref] Acute stroke management pathway during Coronavirus-19 pandemic, Baracchini [/bib_ref] , followed by Puglia [bib_ref] The definition and identification of need for health care, Acheson [/bib_ref]. Marche, Piedmont and Friuli Venezia Giulia had 7 papers each, Tuscany and Veneto 9 each. Overall, 73 were general/nationwide, 105 pertained to northern regions, 54 to central regions, and 15 to southern regions and islands. The studies included are reported in Supplementary Material .
## Management of emergencies.
In general, non-COVID patients admitted to emergency departments (ED) decreased and remained well below the standard levels. [bib_ref] How Large Was the Mortality Increase Directly and Indirectly Caused by the..., Magnani [/bib_ref] [bib_ref] Out-of-Hospital Cardiac Arrest during the Covid-19 Outbreak in Italy, Baldi [/bib_ref] [bib_ref] Reduced Rate of Hospital Admissions for ACS during Covid-19 Outbreak in Northern..., De Filippo [/bib_ref] [bib_ref] Acute stroke management pathway during Coronavirus-19 pandemic, Baracchini [/bib_ref] The youngest age classes declined dramatically, while the oldest age classes progressively increased, remaining considerably above the standard rate of the local ED. 12 [fig_ref] Table 1: Management of deferrable conditions and telemedicine during COVID-19 pandemic in Italy [/fig_ref] shows the most relevant data regarding cardiovascular emergencies, including stroke. In northern Italy, the emergency gradually took over most cardiology intensive care units (ICUs). [bib_ref] COVID-19 experience in Bergamo, Italy, Senni [/bib_ref] As expected, the net effect of this re-organization was a significant reduction in sites and staff committed to the treatment of cardiovascular diseases. [bib_ref] An In-hospital Pathway for Acute Coronary Syndrome Patients During the COVID-19 Outbreak:..., Cosentino [/bib_ref] [bib_ref] Critical Organizational Issues for Cardiologists in the COVID-19 Outbreak: A Frontline Experience..., Stefanini [/bib_ref] [bib_ref] Cardiac patient care during a pandemic: how to reorganise a heart failure..., Agostoni [/bib_ref] [bib_ref] A dialogue between the editor-in-chief and a deputy editor of a cardiology..., Piepoli [/bib_ref] [bib_ref] The COVID-19 challenge to cardiac electrophysiologists: optimizing resources at a referral center, Mazzone [/bib_ref] Comparing out-of-hospital cardiac arrests in the same period of the previous year, Baldi et al. found that the median arrival time of the emergency medical service was three minutes longer in 2020, and the proportion of patients who received cardiopulmonary resuscitation from bystanders was 15.6% lower. Among patients in whom resuscitation was attempted by emergency medical services, the incidence of out-of-hospital death was 14.9% higher in 2020 than in 2019. [bib_ref] Out-of-Hospital Cardiac Arrest during the Covid-19 Outbreak in Italy, Baldi [/bib_ref] This finding was confirmed by additional studies highlighting an unpredictable decrease in acute coronary syndrome-related hospitalization in highvolume centres [bib_ref] Reduced Rate of Hospital Admissions for ACS during Covid-19 Outbreak in Northern..., De Filippo [/bib_ref] [bib_ref] Impact of COVID-19 epidemic on coronary care unit accesses for acute coronary..., Franco [/bib_ref] [bib_ref] Unpredictable Fall of Severe Emergent Cardiovascular Diseases Hospital Admissions During the COVID-19..., Mauro [/bib_ref] and pacemaker implantation during the weeks following the COVID-19 outbreak. [bib_ref] Unpredictable Fall of Severe Emergent Cardiovascular Diseases Hospital Admissions During the COVID-19..., Mauro [/bib_ref] [bib_ref] Urgent Pacemaker Implantation Rates in the Veneto Region of Italy After the..., Federico [/bib_ref] [bib_ref] Urgent Pacemaker Implantation Rates in the Veneto Region of Italy After the..., Migliore [/bib_ref] [bib_ref] Reduction of hospitalizations for myocardial infarction in Italy in the COVID-19 era, Rosa [/bib_ref] Of note, one study showed a 49% reduction in acute heart failure admission to the hospital. [bib_ref] The Impact of the Coronavirus Disease-2019 Pandemic and Italian Lockdown Measures on..., Colivicchi [/bib_ref] The patients admitted had longer door-to-balloon and symptoms to PCI times, higher hs-cTnI levels at presentation, lower residual left ventricular function at discharge, and higher predicted late cardiovascular mortality on the GRACE score. Two studies [bib_ref] Acute stroke management pathway during Coronavirus-19 pandemic, Baracchini [/bib_ref] [bib_ref] Has COVID-19 played an unexpected "stroke" on the chain of survival?, Naccarato [/bib_ref] also reported fewer patients with minor strokes and transient ischemic attacks (TIAs), longer onset-to-door and door-to-treatment times for major strokes, and reduced transfers from spokes centres. As a result, the number of patients who underwent intravenous thrombolysis or bridging therapy (combined intravenous and thrombectomy) decreased by − 26% and −30%, respectively.
In addition, as a consequence of the reduction in the patient eligibility for bridging therapy, the number of primary thrombectomies (performed with all the necessary personal protective equipment owing to the risk of infection) [bib_ref] Mechanical Thrombectomy in the Era of the COVID-19 Pandemic: Emergency Preparedness for..., Nguyen [/bib_ref] increased by 41%. Most of these patients had very serious strokes that would have benefited from early diagnoses. [bib_ref] Acute stroke management pathway during Coronavirus-19 pandemic, Baracchini [/bib_ref] The decrease in hospital admissions, confirmed by a survey across multiple countries, including Italy, [bib_ref] Stroke care during the Covid-19 pandemic: Experience from three large European countries, Bersano [/bib_ref] resulted in increased door-to-needle times and missed therapeutic windows for patients suffering from severe strokes. [bib_ref] COVID-19 and stroke-A global World Stroke Organization perspective, Markus [/bib_ref] The reduction in available Intensive Care Units (ICU) beds, massively dedicated to COVID-19 patients with acute respiratory failure, and the fear of infection resulted in the shrinkage of surgical activities in all fields [bib_ref] What happened to surgical emergencies in the era of COVID-19 outbreak? Considerations..., Patriti [/bib_ref] [bib_ref] Emergency Endoscopy During the SARS-CoV-2 Pandemic in the North of Italy: Experience..., Lauro [/bib_ref] [bib_ref] Minimally Invasive Surgery and the Novel Coronavirus Outbreak: Lessons Learned in China..., Zheng [/bib_ref] [bib_ref] Emergency general surgery in Italy during the COVID-19 outbreak: first survey from..., Patriti [/bib_ref] and a reduction in urgent endoscopic procedures in COVID-free hospitals. [bib_ref] A snapshot of urgent upper gastrointestinal endoscopy care during the www.mjhid.org, D'ovidio [/bib_ref] Each sub-speciality defined various non-deferrable surgical procedures that had to be guaranteed, causing a drop in consultations requested by emergency departments, as in the case of urgent urology. [bib_ref] The impact of COVID-19 pandemic on urological emergencies: a single-center experience, Motterle [/bib_ref] The surgical community also faced a shortage of blood components derived from fewer donations due to lockdown and fear of infection. [bib_ref] Blood supply and transfusion support in southern Italy: findings during the first..., Grandone [/bib_ref] To compensate for the initial fall (−10%) in blood donations in the first week of March, the government promoted a national media campaign on the importance and safety of blood donation as a priority to maintain basic healthcare services for non-COVID patients. [bib_ref] The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic and Transfusion Medicine:..., Mascaretti [/bib_ref] No blood-transmitted SARS-CoV-2 infection has been reported to date. [bib_ref] The impact of the SARS-CoV-2 outbreak on the safety and availability of..., Franchini [/bib_ref] Children's emergency departments also showed a substantial decrease in visits, [bib_ref] A COVID-19 outbreak's lesson: Best use of the paediatric emergency department, Pata [/bib_ref] and that might reflect the scarcity of resources or the reluctance of parents and health care workers to expose children to the risk of viral infection in a healthcare setting, in addition to lower rates of acute infections and trauma. [bib_ref] An Italian paediatric department at the time of Coronavirus: a resident's point..., Cognigni [/bib_ref] However, this phenomenon has been detrimental to the health of non-COVID child patients: 12 cases of delayed access to hospital care were reported during the week of March 23-27 across five hospitals of an Italian Children's Hospital Research Network. Half of the children were admitted to an ICU, and four died, highlighting the high risk of delaying access to hospital care. [bib_ref] Delayed access or provision of care in Italy resulting from fear of..., Lazzerini [/bib_ref] As a result, life- No deterioration of glucose control was observed but an improvement (more stable rhythms of life and more regular meals time)
The hybrid closed loop (HCL) system allowed physicians to download detailed data during the teleconsulting through a web-based program.
## 244-246
Dermatology Chronic inflammatory skin disease (psoriasis, threatening conditions (i.e. abdominal pain, severe ketoacidosis) seemed to be more frequent, requiring, in some cases, an aggressive approach. [bib_ref] An Italian paediatric department at the time of Coronavirus: a resident's point..., Cognigni [/bib_ref] The same phenomenon affected dermatology. [bib_ref] Which are the "emergent" dermatologic practices during COVID-19 pandemic?, Giacalone [/bib_ref] Tartari et al. compared two different weeks, before and after the outbreak of the COVID-19 pandemic in Italy, showing a decrease in unjustified referrals (93% reduction) in dermatological emergency services. [bib_ref] Changes in emergency service access after spread of COVID-19 across Italy, Tartari [/bib_ref] Despite medical care for emergencies and urgent treatments being continuously provided throughout the pandemic, the lack of personnel, resources, and ICUs beds and the patients' fear of being infected in hospital affected patient management and substantially delayed the provision of ordinary medical activities. These initial data seem to show a decrease in emergencies and an increase in mortality.
## Management of non-deferrable conditions: the huband-spoke system.
To tackle the massive impact of the overflow of SARS-CoV-2 infected patients, hospitals in Italy had to undergo a significant re-organization. [bib_ref] 2019-ncov's epidemic in middle province of northern Italy: Impact, logistic & strategy..., Gagliano [/bib_ref] [bib_ref] Executive management summary and short report of outcome, Ferrazzi [/bib_ref] [bib_ref] Management of healthcare areas for the prevention of COVID-19 emergency in an..., Baggiani [/bib_ref] [bib_ref] Hospital surge capacity in a tertiary emergency referral centre during the COVID-19..., Carenzo [/bib_ref] [bib_ref] Reorganization of a large academic hospital to face COVID-19 outbreak: The model..., Meschi [/bib_ref] [bib_ref] Surgical Strategy During the COVID-19 Pandemic in a University Metropolitan Hospital in..., Mariani [/bib_ref] In order to manage conditions needing non-deferrable treatment while avoiding the risk of infection, hub-andspoke centres were created and widely used throughout the country. [bib_ref] Executive management summary and short report of outcome, Ferrazzi [/bib_ref] [bib_ref] La pandemia COVID-19: Riorganizzazione della Cardiologia in un ospedale della Regione Lombardia, Ghio [/bib_ref] [bib_ref] COVID-19 e coronaropatia: Uso selettivo e collaborativo delle risorse durante le crisi..., Villa [/bib_ref] [bib_ref] The Italian USL Toscana Centro model of a vascular hub responding to..., Chisci [/bib_ref] [bib_ref] Fast reshaping of intensive care unit facilities in a large metropolitan, Zangrillo [/bib_ref] In the hub-and-spoke model, the main campus or hub supplies the most intensive medical services, while satellite campuses or spokes offer more limited services at sites distributed across the neighbouring area. [bib_ref] The hub-and-spoke organization design revisited: a lifeline for rural hospitals, Elrod [/bib_ref] Neurological surgery was particularly affected by the ICU re-organization, as it often requires a period of intensive monitoring in ICUs. [bib_ref] Neurosurgery during the COVID-19 pandemic: update from Lombardy, northern Italy, Zoia [/bib_ref] [bib_ref] Neurosurgical Practice During the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Pandemic:..., Fontanella [/bib_ref] [bib_ref] Neurosurgical activity during COVID-19 pandemic: an expert opinion from China, Fontanella [/bib_ref] All the cases of elective neurological surgery were deferred, while urgent neurosurgical pathologies (above all traumas) and nondeferrable tumour cases were transferred to hubs. [bib_ref] Has COVID-19 played an unexpected "stroke" on the chain of survival?, Naccarato [/bib_ref] [bib_ref] Stroke care during the Covid-19 pandemic: Experience from three large European countries, Bersano [/bib_ref] [bib_ref] COVID-19 and stroke-A global World Stroke Organization perspective, Markus [/bib_ref] [bib_ref] What happened to surgical emergencies in the era of COVID-19 outbreak? Considerations..., Patriti [/bib_ref] [bib_ref] Emergency Endoscopy During the SARS-CoV-2 Pandemic in the North of Italy: Experience..., Lauro [/bib_ref] [bib_ref] Effects of the COVID-19 Outbreak in Northern Italy: Perspectives from the Bergamo..., Bernucci [/bib_ref] [bib_ref] Letter to the editor by Dobran Mauro, Paracino Riccardo, and Iacoangeli Maurizio..., Dobran [/bib_ref] [bib_ref] The management of emergency spinal surgery during the COVID-19 pandemic in Italy, Giorgi [/bib_ref] Some minimal activities were still performed at spoke centres, only for critical cases or when specific tools were required (i.e. gamma knife treatment of neoplastic lesions). [bib_ref] Skull-base surgery during the COVID-19 pandemic: the Italian Skull Base Society recommendations, Castelnuovo [/bib_ref] [bib_ref] Remodulation of neurosurgical activities in an Italian region (Emilia-Romagna) under COVID-19 emergency:..., Mazzatenta [/bib_ref] [bib_ref] Maintaining stereotactic radiosurgical treatments during Covid-19 outbreak: the case of the Gamma..., Franzin [/bib_ref] This provided an unprecedented opportunity for transversal collaboration among different teams, representing real innovation in such a competitive setting. [bib_ref] Fast reshaping of intensive care unit facilities in a large metropolitan, Zangrillo [/bib_ref] [bib_ref] Letter to the editor by Dobran Mauro, Paracino Riccardo, and Iacoangeli Maurizio..., Dobran [/bib_ref] A hub-and-spoke system was also organized for vascular surgery and cardiac surgery units. All elective surgery was reduced, and urgent surgery (including aortic aneurysms, valvular diseases or severe coronary diseases) was performed only in hub centres, preferring the endovascular to the open surgical approach whenever possible. [bib_ref] Routine in an Italian High-Volume Vascular Surgery Unit during the COVID-19 Era:..., Mangialardi [/bib_ref] [bib_ref] The COVID-19 outbreak and its impact on hospitals in Italy: the model..., Bonalumi [/bib_ref] A transcatheter approach was generally preferred, as it usually does not require an ICU bed or a ventilator. [bib_ref] The COVID-19 outbreak and its impact on hospitals in Italy: the model..., Bonalumi [/bib_ref] The limitation of all non-urgent surgical activities also applied to general surgery [bib_ref] COVID-19 outbreak in Northern Italy: Viewpoint of the Milan area surgical community, Kurihara [/bib_ref] [bib_ref] Pediatric endoscopic procedures during the COVID-19 pandemic: an Italian center experience, Duci [/bib_ref] [bib_ref] A Snapshot of Elective Oncological Surgery in Italy During COVID-19 Emergency: Pearls,..., Torzilli [/bib_ref] [bib_ref] Bariatric surgery and the COVID-19 pandemic: SICOB recommendations on how to perform..., Navarra [/bib_ref] [bib_ref] COVID-19) outbreak: what the department of endoscopy should know, Repici [/bib_ref] [bib_ref] Preparing an obstetric unit in the heart of the epidemic strike of..., Capanna [/bib_ref] and obstetrics. [bib_ref] Preparing an obstetric unit in the heart of the epidemic strike of..., Capanna [/bib_ref] [bib_ref] Obstetric network reorganization during the COVID-19 pandemic: Suggestions from an Italian regional..., Giannubilo [/bib_ref] In highly-infected areas (such as Lombardy), hub centres were created 78 to treat only advanced symptomatic tumours, [bib_ref] Coronavirus pandemic and colorectal surgery: practical advice based on the Italian experience, Saverio [/bib_ref] [bib_ref] Laparoscopy at all costs? Not now during COVID-19 outbreak and not for..., Saverio [/bib_ref] [bib_ref] COVID-19: The European institute of oncology as a 'hub' centre for breast..., Vicini [/bib_ref] while elective oncological surgery procedures continued to be performed in less-affected regions. [bib_ref] Coronavirus pandemic and colorectal surgery: practical advice based on the Italian experience, Saverio [/bib_ref] [bib_ref] Italian society of colorectal surgery recommendations for good clinical practice in colorectal..., Gallo [/bib_ref] This was an important issue for oncological patients, especially the older ones. [bib_ref] Coronavirus Disease 2019 Emergency and Cancer in the South of Italy: What's..., Ingenito [/bib_ref] [bib_ref] Coronavirus: Older Persons With Cancer in Italy in the COVID-19 Pandemic, Fratino [/bib_ref] [bib_ref] Cancer care under the outbreak of COVID-19: A perspective from Italian tertiary..., Marano [/bib_ref] [bib_ref] Endocrine surgery during COVID-19 pandemic: do we need an update of indications..., Lombardi [/bib_ref] Many possible ways of minimizing the risks were proposed: to postpone treatments or elective surgery for stable cancer in endemic areas, provide patients with greater personal protection, and offer more intensive surveillance or treatment. [bib_ref] Cancer care during the spread of coronavirus disease 2019 (COVID-19) in Italy:..., Lambertini [/bib_ref] [bib_ref] Metastatic melanoma treatment with checkpoint inhibitors in the COVID-19 era: experience from..., Quaglino [/bib_ref] [bib_ref] Caring for Patients With Cancer During the COVID-19 Outbreak in Italy, Pietrantonio [/bib_ref] For example, neoadjuvant treatments were recommended or increased to defer surgical admission for as long as possible. [bib_ref] Coronavirus pandemic and colorectal surgery: practical advice based on the Italian experience, Saverio [/bib_ref] [bib_ref] Italian society of colorectal surgery recommendations for good clinical practice in colorectal..., Gallo [/bib_ref] [bib_ref] Management of ovarian cancer during the COVID-19 pandemic, Mandato [/bib_ref] [bib_ref] COVID-19 global pandemic: options for management of gynecologic cancers. The experience in..., Perrone [/bib_ref] For other medical conditions requiring surgery under particular circumstances, such as relapse of inflammatory bowel disease, dedicated hubs were identified. [bib_ref] Challenges in the Care of IBD Patients During the CoViD-19 Pandemic: Report..., Occhipinti [/bib_ref] [bib_ref] Impact of the COVID-19 pandemic on Gastroenterology Divisions in Italy: a national..., Maida [/bib_ref] Interestingly, a tendency toward treatments to reduce hospitalization was also found in medical oncology. [bib_ref] Metronomic oral vinorelbine and lung cancer therapy during the COVID 19 pandemic:..., Rossi [/bib_ref] [bib_ref] Results of Multilevel Containment Measures to Better Protect Lung Cancer Patients From..., De Marinis [/bib_ref] Some regions such as Tuscany created home care protocols to avoid exposure to hospital settings. [bib_ref] Home Care for Cancer Patients During COVID-19 Pandemic: The Double Triage Protocol, Porzio [/bib_ref] Oncological care delivery and cancer diagnosis [bib_ref] Cancer diagnostic rates during the 2020 'lockdown', due to COVID-19 pandemic, compared..., Vincentiis [/bib_ref] [bib_ref] CHRONIC LYMPHOCYTIC LEUKEMIA MANAGEMENT IN ITALY DURING THE COVID-19 PANDEMIC. A CAMPUS..., Cuneo [/bib_ref] [bib_ref] Children with cancer in the time of COVID-19: An 8-week report from..., Ferrari [/bib_ref] were dramatically reduced by the SARS-CoV-2 outbreak, even though suboptimal care and treatments may result in worse cancer-related outcomes. Oncologists were thus asked to preserve patients' continuum of care while adopting mitigation strategies to reduce the likelihood of infection in all cancer patients. [bib_ref] Comment on 'Reorganisation of medical oncology departments during the novel coronavirus disease-19..., Bongiovanni [/bib_ref] [bib_ref] Fighting cancer in coronavirus disease era: organization of work in medical oncology..., Brandes [/bib_ref] [bib_ref] A Guide for Oncologic Patient Management during Covid-19 Pandemic: The Initial Experience..., Mistretta [/bib_ref] [bib_ref] Cancer treatment during the coronavirus disease 2019 pandemic: Do not postpone, do..., Omarini [/bib_ref] [bib_ref] Systemic Treatment of Patients With Gastrointestinal Cancers During the COVID-19 Outbreak: COVID-19-adapted..., Pietrantonio [/bib_ref] [bib_ref] How coronavirus disease 2019 outbreak is impacting colorectal cancer patients in Italy:..., Pellino [/bib_ref] [bib_ref] The Shifting Landscape of Genitourinary Oncology During the COVID-19 Pandemic and how..., Marandino [/bib_ref] [bib_ref] Reorganisation of medical oncology departments during the novel coronavirus disease-19 pandemic: a..., Indini [/bib_ref] [bib_ref] COVID-19 Pandemic and the Crisis of Health Systems: The Experience of the..., Silvestris [/bib_ref] Arduino et al. described a worrying delay in diagnosing oral cancer in northwest Italy during the Covid pandemic. [bib_ref] The outbreak of Novel Coronavirus disease (COVID-19) caused a worrying delay in..., Arduino [/bib_ref] Moreover, the cessation of elective activities, screening programs, [bib_ref] BRCA testing in a genomic diagnostics referral center during the COVID-19 pandemic, Minucci [/bib_ref] and the drastic reduction in services regarding breast cancer restricted evaluations to only clinical observations of palpable lesions with the elevated risk of missing new diagnoses. [bib_ref] Breast Cancer Diagnosis in Coronavirus-Era: Alert From Italy, Vanni [/bib_ref] Although not requiring a structural reorganization, palliative care was forced to find a new balance between family member visits and patients' needs. [bib_ref] Palliative Care in the Time of COVID-19, Mercadante [/bib_ref] [bib_ref] The Role and Response of Palliative Care and Hospice Services in Epidemics..., Etkind [/bib_ref] [bib_ref] Response and role of palliative care during the COVID-19 pandemic: A national..., Costantini [/bib_ref] [bib_ref] Management strategies adopted by a pediatric palliative care network in northern Italy..., Lazzarin [/bib_ref] In orthopaedics, the re-organization led to the identification of poly-specialist major trauma centres and specialistic referral centres for minor trauma or nondeferrable orthopaedic surgeries (i.e. septic arthritis or malignant tumours). [bib_ref] Changes of clinical activities in an orthopaedic institute in North Italy during..., Zagra [/bib_ref] [bib_ref] Management of orthopaedic and traumatology patients during the Coronavirus disease (COVID-19) pandemic..., Randelli [/bib_ref] [bib_ref] Operational strategies of a trauma hub in early coronavirus disease 2019 pandemic, Casiraghi [/bib_ref] [bib_ref] CoViD-19 and ortho and trauma surgery: The Italian experience, Placella [/bib_ref] [bib_ref] Trauma service reorganization in Bologna (Italy) during COVID-19 pandemic, Martino [/bib_ref] [bib_ref] The deep impact of novel CoVID-19 infection in an Orthopedics and Traumatology..., Maniscalco [/bib_ref] There was a reduction in the number of proximal femur fractures in two centres, as well as a reduction in hip and knee arthroplasties. A similar re-organization was also carried out for plastic surgery: only post-traumatic, oncological and burn treatments were guaranteed. 127-129 A new approach based on enzymatic debridement was proposed for burns to reduce the need for burn surgery. In urology, only urgent, non-deferrable procedures (colicky flank pain, gross hematuria and acute urinary retention) were authorized after careful multidisciplinary evaluation, which led to a drop in urological surgical activities. Whenever possible, alternative treatments not requiring general anaesthesia (i.e. radiotherapy for genitourinary cancers) were suggested as preferable. Oral and maxillofacial surgery, otolaryngology, and ophthalmology also suspended all non-urgent treatments, especially considering healthcare workers' high risk of infection while manipulating the upper airways and eyes. Only the treatment of trauma, malignant neoplasms, and severe infections was guaranteed. In the context of radiotherapy, all follow-up visits involved a phone call in advance in order to postpone non-urgent cases. The initial consultations of patients needing treatment for malignant tumours were conducted as normal, as were certain treatments such as bone metastases radiotherapy. 154-157 Specific approaches, such as short fractionated radiotherapy, were suggested. All non-urgent and deferrable radiation treatments were delayed, while therapies for patients with better prognoses (benign and functional diseases) were postponed. Dermatology departments were also involved in an extensive re-organization. 161-162 Dermatological antineoplastic treatments were provided in the dermatology clinics of many centres, such as Bologna, Naples, Modena and Ancona, which also maintained urgent dermatological procedures and consultations required by other hospital wards. As awareness of the severity of the COVID-19 increased, some patients were concerned about continuing their medications; however, all centres followed specific recommendations and advised patients not to suspend these drugs without consultation. Lastly, microbiology labs underwent unprecedentedly high workloads with the increasing number of samples (swabs or serological tests) to analyze for the identification of Sar-Cov-2 infection. An extensive re-organization of the microbiology lab activities thus also occurred. In a large teaching hospital in Rome, the introduction of night shifts and the creation of a dedicated team significantly improved the number of samples processed without interfering with the daily laboratory routines. 171
## Replacement therapies: dialysis and transplantation.
Dialysis units experienced a profound change in their management with the introduction of COVID-19 isolation rooms and the identification of dedicated healthcare professionals. 172,173 Rombolà et al. proposed three actions to be taken in order to dialyze non-COVID patients safely: hygiene measures, the use of PPE to protect patients and the healthcare team, and the protection of the dialysis ward with an isolated area for testing patients suspected of infection. In general, all transplant programs were profoundly affected by the pandemic. First of all, the wall-towall screening of donors and recipients was established to identify positive patients who would not be able to donate or receive blood, in view of the high mortality rates COVID-19 in immunocompromised patients. Secondly, the widespread reduction in available ICU beds led to an estimated 15% drop in transplants compared with the last five years' average, such as liver transplantations. 177,181 transplantation was thus suggested only in true end-stage organ failure, preferring conservative treatments (maximizing pharmacological therapy) in all other patients. Management of Deferrable Conditions and Telemedicine. The management of chronic conditions also suffered. 185 Cesari et al. found that the integration of care services collapsed: admissions to post-acute/long-term care facilities were reduced, and several person-tailored interventions were suspendede.g., physical therapists for mobilization. 183 Lasevoli et al. reinforced the view that the current pandemic has had dramatic consequences for the mental health of serious psychiatric patients. All this inevitably led to a drastic reduction and a substantial re-organization of the clinical activity in many specialities [fig_ref] Table 1: Management of deferrable conditions and telemedicine during COVID-19 pandemic in Italy [/fig_ref] , postponing elective treatments and switching to telemedicine (TM) for consultation or not to leave vulnerable high-need patients without proper follow-up. [bib_ref] Global Telemedicine Implementation and Integration Within Health Systems to Fight the COVID-19..., Ohannessian [/bib_ref] The implementation of TM occurred in different ways and to varying degrees depending on the specific centre and specialty. An online questionnaire administered to the 176 Directors of Italian Radiation Oncology Departments revealed that to guarantee the continuity of care, in 78 centres (62.4%) activated telematic consultations for RT treatments. 152 A similar survey for RT centres in the Lombardy region revealed that 84% of RT facilities cancelled out-patient follow-up visits, 68% activated telematic consultation and 30% adopted working from home solutions. 153 Another survey administered to 122 medical oncology departments homogeneously distributed on the national territory revealed that in 72% of cases, alternative ways to get in touch with patients had been used, like telephonic interviews with the interpretation of laboratory and radiologic examination reports. According to Pietrantonio et al., WhatsApp turned out to be adequate to give a rapid answer to most queries from oncologic patients. [bib_ref] Systemic Treatment of Patients With Gastrointestinal Cancers During the COVID-19 Outbreak: COVID-19-adapted..., Pietrantonio [/bib_ref] Brunasso et al. started a teledermatology service in smart working using phone calls and e-mails by which they could monitor almost 94% of their patients. [bib_ref] Teledermatologic monitoring for chronic cutaneous autoimmune diseases with smartworking during Covid-19 emergency..., Brunasso [/bib_ref] In a Department of Urology in Northern Italy, 55% of cases were screened undergoing telephone consultation. [bib_ref] Teleurology in the Time of Covid-19 Pandemic: Here to Stay?, Luciani [/bib_ref] TM has been shown to have beneficial impacts on heart failure outcomes in a comparative analysis between 2020 and 2019 by Finally, TM positively impacted patients' life as documented by a survey in which 85% of patients were satisfied with the remote interview modality and the reduction of economic and time costs related to going to the clinic. Most of those subjects (90%) expressed their willingness to continue to be included in remote evaluation programs. [bib_ref] Telemedicine is a useful tool to deliver care to patients with Amyotrophic..., Capozzo [/bib_ref] Almost all specialities benefited from TM during the pandemic. The results are summarized in [fig_ref] Table 1: Management of deferrable conditions and telemedicine during COVID-19 pandemic in Italy [/fig_ref].
Discussion. On March 11 2020, the World Health Organisation (WHO) declared the COVID-19 pandemic. 1 However, Italy was already in lockdown, with decrees limiting mobility and strengthening the National Health System. On March 9, 2020, most outpatient services were temporarily suspended, except for a few treatments that were considered urgent and non-deferrable.Clinical support for early isolation, treatment, and, where needed, intensive care of COVID-19 patients (or suspect cases) became the priority, with a massive allocation of dedicated resources.
A large increase in all-cause mortality was revealed during the epidemic, greater than the number of deaths attributed to COVID-19 cases. The possible causes of this increase include a large number of severe undiagnosed COVID-19 cases, reduced access to health services due to the disruption of normal working processes, or the fear of contamination of sick patients affected by other diseases and possibly other factors. [bib_ref] How Large Was the Mortality Increase Directly and Indirectly Caused by the..., Magnani [/bib_ref] We provided a snapshot, across all medical specialities, of how the provision of treatments to non-COVID patients in Italy has been impacted by the shortage of resources imposed by the pandemic.
Measures put in place to mitigate the outbreak, such as social distancing and confinement, contributed to discouraging access to the emergency department (E.D.) all over the country and those conditions requiring urgent care. As a result, there was a significant decrease in overall E.D. admissions and a substantial reduction in all-speciality surgical consultations. [bib_ref] Out-of-Hospital Cardiac Arrest during the Covid-19 Outbreak in Italy, Baldi [/bib_ref] [bib_ref] Reduced Rate of Hospital Admissions for ACS during Covid-19 Outbreak in Northern..., De Filippo [/bib_ref] [bib_ref] Acute stroke management pathway during Coronavirus-19 pandemic, Baracchini [/bib_ref] More studies are needed to confirm these data and to evaluate the impact on death rates; however, recent reports from other countries seem to confirm this trend. [bib_ref] Impact of COVID-19 pandemic on ST-elevation myocardial infarction in a non-COVID-19 epicenter, Hammad [/bib_ref] [bib_ref] COVID-19: Stroke Admissions, Emergency Department Visits, and Prevention Clinic Referrals, Bullrich [/bib_ref] Cardiovascular emergencies paid significant tolls with a significant delay in time-sensitive emergency operations. [bib_ref] Out-of-Hospital Cardiac Arrest during the Covid-19 Outbreak in Italy, Baldi [/bib_ref] [bib_ref] Reduced Rate of Hospital Admissions for ACS during Covid-19 Outbreak in Northern..., De Filippo [/bib_ref] [bib_ref] Acute stroke management pathway during Coronavirus-19 pandemic, Baracchini [/bib_ref] [bib_ref] Impact of the COVID-19 epidemic on census, organization and activity of a..., Comelli [/bib_ref] [bib_ref] COVID-19 experience in Bergamo, Italy, Senni [/bib_ref] [bib_ref] An In-hospital Pathway for Acute Coronary Syndrome Patients During the COVID-19 Outbreak:..., Cosentino [/bib_ref] [bib_ref] Critical Organizational Issues for Cardiologists in the COVID-19 Outbreak: A Frontline Experience..., Stefanini [/bib_ref] [bib_ref] Cardiac patient care during a pandemic: how to reorganise a heart failure..., Agostoni [/bib_ref] [bib_ref] A dialogue between the editor-in-chief and a deputy editor of a cardiology..., Piepoli [/bib_ref] [bib_ref] The COVID-19 challenge to cardiac electrophysiologists: optimizing resources at a referral center, Mazzone [/bib_ref] More recent evidence, consistent with our results, showed a significant decrease in the mean number of endovascular therapies per hospital performed before and after COVID-19 confinement along with a significant increase in mean stroke onset-to groin puncture time. [bib_ref] Acute Stroke Management During the COVID-19 Pandemic: Does Confinement Impact Eligibility for..., Hajdu [/bib_ref] A delayed presentation of STEMI patients that may lead to worsened prognosis and unnecessary deaths has also been observed. [bib_ref] Impact on hospital admission of STelevation myocardial infarction patients during coronavirus disease..., Liberto [/bib_ref] Moreover, an additional study confirmed that more in-hospital cardiovascular deaths occurred in March 2020 compared with March 2019, a finding due to late hospital presentations and consequent greater disease severity that affected eligibility and outcome of cardiovascular procedures [bib_ref] Increased cardiovascular death rates in a COVID-19 low prevalence area, Pinto [/bib_ref] As stated before, the hospitalisation system was remodelled to allocate appropriate resources to manage patients with COVID-19; consequently, hubs were identified for specialised medical activities. Hub-andspoke centres represented an important change in care provision, especially in the most affected regions, involving almost all specialities. However, data on the efficacy of this reorganisation, measured in terms of health outcomes (such as mortality), are lacking. To date, only a few reports suggest a tendency toward a decrease in diagnosis for non-deferrable conditions despite the hub-and-spoke organisation. [bib_ref] CHRONIC LYMPHOCYTIC LEUKEMIA MANAGEMENT IN ITALY DURING THE COVID-19 PANDEMIC. A CAMPUS..., Cuneo [/bib_ref] [bib_ref] Comment on 'Reorganisation of medical oncology departments during the novel coronavirus disease-19..., Bongiovanni [/bib_ref] [bib_ref] BRCA testing in a genomic diagnostics referral center during the COVID-19 pandemic, Minucci [/bib_ref] Cautious and evidence-based studies are needed to properly assess the overall impact of this model on measurable outcomes. Nevertheless, the hub-and-spoke system seems to be a valid model, at least, in the management of ischemic emergencies. [bib_ref] Emergency management of the COVID-19 pandemic in a vascular surgery department of..., Chiesa [/bib_ref] A pandemic is a dynamic scenario, requiring reorganisation and flexibility of healthcare delivery. T.M., which consists of distributing health-related services and information via telecommunication technologies, proved a pragmatic approach to managing deferrable conditions during the COVID-19 pandemic in Italy. Moreover, TM allows for more flexibility on the side of both the clinician and the patient, as consultations can easily be rescheduled, and meetings can be held from home. [bib_ref] Teledermatologic monitoring for chronic cutaneous autoimmune diseases with smartworking during Covid-19 emergency..., Brunasso [/bib_ref] Notably, when looking at the world scenario, T.M. is thought to play an important role in delivering digital health to the general population. [bib_ref] World Heart Day 2021: COVID-19, digital health, and tackling cardiovascular disease, Pinto [/bib_ref] Almost all specialities benefited from T.M. during the pandemic, with short-term results particularly encouraging in some cases. The pandemic has generally demonstrated that information technologies should be more promoted independently from this specific context. 153 However, a longer follow-up is needed to assess the efficacy of these measures on common health outcomes. [bib_ref] Teledermatologic monitoring for chronic cutaneous autoimmune diseases with smartworking during Covid-19 emergency..., Brunasso [/bib_ref] [bib_ref] World Heart Day 2021: COVID-19, digital health, and tackling cardiovascular disease, Pinto [/bib_ref] [bib_ref] Teleurology in the Time of Covid-19 Pandemic: Here to Stay?, Luciani [/bib_ref] Avoiding face-to-face contact via T.M. has been one of the most effective measures to limit the spread of SARS-Cov-2 infection, although many issues have been raised, such as privacy management and the lack of clear guidelines.. [bib_ref] Community Mental Health Care in the COVID-19 Response: An Italian Example, Caroppo [/bib_ref] [bib_ref] COVID-19 risk contagion: Organization and procedures in a South Italy geriatric oncology..., Gambardella [/bib_ref] [bib_ref] COVID-19 disease emergency operational instructions for Mental Health Departments issued by the..., Starace [/bib_ref] [bib_ref] Telemedicine During The COVID-19 in Italy: A Missed Opportunity?, Omboni [/bib_ref] [bib_ref] A Frail Health Care System for an Old Population: Lesson form the..., Volpato [/bib_ref] We strongly encourage overcoming these limitations to promote further the multiple opportunities of T.M. in tune with its pivotal role during the second phase of the COVID-19 pandemic in Italy. [bib_ref] The Italian Fight Against the COVID-19 Pandemic in the Second Phase: The..., Giansanti [/bib_ref] T.M. has deeply influenced non-COVID patient care, enabling the remote diagnosis and monitoring of patients and allowing clinical data sharing between patients and physicians. [bib_ref] Telemedicine during Covid-19 pandemic: Advantage or critical issue?, Perrone [/bib_ref] [bib_ref] Telemedicine and the 2019 Coronavirus (SARS-CoV-2), Sossai [/bib_ref] One limitation of this systematic review is the heterogeneity in publication type, which prevented the execution of a meta-analysis to summarise the findings together with a quality assessment. Another important issue is the potential underreporting, although the studies covered the whole country's experiences. Finally, while also providing a snapshot of the first modification during the pandemic, this work does not include more recent Italian studies.
Our work thus suggests that a public health crisis has resulted from the pandemic, a concern raised in other countries too, such as France and Denmark; indeed, a danish study demonstrated that mortality rates for patients admitted to hospitals with non-covid-19 diseases (such as respiratory disease, cancer, pneumonia and sepsis) were higher. [bib_ref] What if the worst consequences of COVID-19 concerned non-COVID patients?, Goulabchand [/bib_ref] [bib_ref] First wave of COVID-19 hospital admissions in Denmark: a Nationwide population-based cohort..., Holler [/bib_ref] More detailed, nationwide population-based cohort studies are needed to assess whether emergency management benefited from the reorganisation adopted and evaluate hospital admissions and mortality rates for non-covid patients. Nevertheless, the system and telemedicine undoubtedly playedand continue to playa crucial role in dealing with nondeferrable and deferrable conditions, respectively. ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf and declare: the work was supported by the Center for Excellence and Transdisciplinary Studies (CEST) and C.R.T. Foundation, Turin. The funder of the study had no role in the study design, data collection, data analysis, data interpretation, or writing the report; no financial relationships with any organisations that might have an interest in the submitted work; no other relationships or activities that could appear to have influenced the submitted work. All authors had full access to all study data and had final responsibility for the decision to submit for publication. The authors are thankful to Professor Paolo Vineis for his critical assistance in writing the paper.
[table] Table 1: Management of deferrable conditions and telemedicine during COVID-19 pandemic in Italy. [/table]
|
The opportunity costs of birth in Australia: Hospital resource savings for a post–COVID‐19 era
## | 275
CALLANDER Et AL.
To some extent, the changes made to maternity services were related to limitations that existed pre-COVID-19. Many countries adopted or strengthened community-based antenatal and postnatal care and/or telehealth services to reduce face-to-face contact for routine care.Before COVID-19, women with additional or complex needs generally attended hospital for antenatal and postnatal care. However, the pandemic saw many of these services devolve out into community settings in combination with telehealth. Birth practices also changed, including the use of additional personal protective equipment, strict protocols about birth room occupation (eg, limiting the presence of partners and other attendants), [bib_ref] Pregnancy, birth and the COVID-19 pandemic in the United States, Davis-Floyd [/bib_ref] and even discouraging skin to skin contact at birth. [bib_ref] Should infants be separated from mothers with COVID-19? First, do no harm, Stuebe [/bib_ref] Some services looked to medical intervention to control the timing of births in an attempt to meet resource availability. [bib_ref] Labor and delivery guidance for COVID-19, Boelig [/bib_ref] [bib_ref] Safe delivery for pregnancies affected by COVID-19, Qi [/bib_ref] In the early stages of the pandemic, some of the changes to maternity service delivery were driven by women who expressed concern about traveling to and attending hospitals for care during pregnancy and/or for birth. Consequently, giving birth outside the hospital environment gained more consideration as a viable option (particularly for women without identified pregnancy concerns).
Choice of birth setting is at the heart of woman-centered care, and the demand for birth outside the hospital setting was further emphasized during the COVID-19 pandemic. Midwifery-led birth centers (which can be within or alongside hospitals, or separate) are designed for women whose pregnancy is deemed to be "low-risk." They provide a "homely" environment and support birth with minimal medical intervention but with links to referral services if required.Growing evidence suggests that hospitals are not optimized to support low-risk birth, [bib_ref] Drivers of maternity care in high-income countries: can health systems support woman-centred..., Shaw [/bib_ref] and several studies and reviews have demonstrated the safety of home and birth centers as acceptable places for birth, particularly for low-risk women. [bib_ref] Meta-analysis of the safety of home birth, Olsen [/bib_ref] [bib_ref] Perinatal or neonatal mortality among women who intend at the onset of..., Hutton [/bib_ref] Notably, research suggests that home births pose no greater mortality or morbidity risks to either the woman or baby than giving birth in a hospital, can reduce the odds of neonatal intensive care unit (NICU) admission, and result in the use of fewer medical interventions during birth (eg, reduced rates of instrumental birth). [bib_ref] Perinatal and maternal outcomes by planned place of birth for healthy women..., Brocklehurst [/bib_ref] Previously, birth centers and home births were viewed as optional services, advocated to promote women's choice about how and where they would like to birth. However, the post-COVID-19 era might see these services become a necessary component of maternity service delivery to reserve hospital resources for those who truly need them.
The current study sought to quantify health resource savings in the acute inpatient public hospital setting if all lowrisk women routinely gave birth at home or in a birth center. Although there is some research to suggest that nonhospital birth can be cost-saving, [bib_ref] Costing alternative birth settings for women at low risk of complications: a..., Scarf [/bib_ref] [bib_ref] Economic implications of home births and birth centers: a structured review, Henderson [/bib_ref] [bib_ref] Cost effectiveness of alternative planned places of birth in woman at low..., Schroeder [/bib_ref] this research has been minimal and not considered in the context of the post-COVID-19 era. We sought to quantify the number of inpatient bed days, women's intensive care unit (ICU) bed hours, infant's special care nursery (SCN) days, and infant's neonatal intensive care unit (NICU) days associated with birth.
# | methods
We created a static microsimulation model of woman and infant health service use associated with birth, using a population-based linked administrative data set. Microsimulation models use individual-level data to estimate the impact of change before it occurs, and have traditionally been used to model income and tax policy changes. By drawing on real-world data, microsimulation models can estimate the actual relationships between events that are currently observed within a population (eg, between gestation at birth and an infant's admission to SCN).
## |
CALLANDER Et AL.
Our model was designed to represent the Australian population of women giving birth in 2017. This was the most recent year with benchmarking data available. Microsimulation models are comprised of two parts: the base model containing details of the status quo and the counterfactual model that estimates change under hypothetical scenarios, in this case, low-risk birth occurring at home or in a birth center. The methods we undertook for completing these two parts are outlined below and illustrated in Figure 1.
## | underlying data set
Our microsimulation model is based on a preexisting data set created from a whole-of-population administrative data linkage. The data set contains all women who gave birth between 01/07/2012 and 30/06/2015 in Queensland (QLD), Australia, and both woman and infant clinical and health service use records from conception to the time the infant was two years of age. The data set contains 186 789 women and 189 809 infants. [bib_ref] What are the costs associated with child and maternal healthcare within Australia?..., Callander [/bib_ref] The preexisting data set was linked with Perinatal Data Collection (PDC) data to identify women and their infants for inclusion. The PDC contains the details of all births regardless of location (private hospital, public hospital), information on maternal demographics, maternal clinical characteristics, medical interventions preformed in pregnancy and childbirth, and infant outcomes. It also records length of hospital stay, a woman's time in an ICU, and an infant's time in a SCN or NICU at time of birth.
## | weighting to produce national estimates for 2017
We reweighted the data of women giving birth between 01/07/2014 and 30/06/2015 (n = 61 801) to reflect the Australian population of women giving birth between 01/01/2017 and 31/12/2017. Reweighting was conducted using GREGWT, a generalized regression reweighting algorithm developed by the Australian Bureau of Statistics (ABS).Weighting was conducted using national benchmarks for a woman's age based on Indigenous identification, a woman's age by parity, remoteness, and sector of birth (public or private) using data from the Australian Institute of Health and Welfare's (AIHW) Mothers and Babies 2017 report.
## | the base model
The base data set was limited to women who gave birth in a public hospital (n = 44 498), as we were interested in public hospital decision-making about location of birth. To quantify the health resources used in current standard care, the number of inpatient bed days, ICU hours, SCN days, and NICU days was summed based on actual health resource use.
## | estimating the impacts of change (the "counterfactual model")
## | scenario 1-home births
The first simulation estimated the health resources that would be used in a hypothetical scenario where all low-risk women gave birth at home. Women who had a higher risk of pregnancy complications were defined as having any of the following characteristics at the start of labor: a multiple pregnancy, being more than 41 weeks of gestation, a noncephalic presentation, were classified as obese (BMI > 30), had a prior cesarean or previous uterine surgery, grand multiparity (≥ five previous births), or any maternal medical condition deemed to affect pregnancy. [bib_ref] Planned private homebirth in Victoria 2000-2015: a retrospective cohort study of Victorian..., Davies-Tuck [/bib_ref] Women at higher risk will be referred to as "women with risk factors" throughout this paper. Women without any of these characteristics were classified as having a low-risk pregnancy for this study. Two sub-data sets were created from the base data set: the first contains women with risk factors, whose health resource use remained as it was recorded on the base data set (n = 26 596); and the second contains low-risk women (n = 17 902). Actual rates of unplanned cesarean, vaginal birth with vacuum, vaginal birth with forceps, and unassisted vaginal birth (without vacuum or forceps) were then identified for these low-risk women in our population. The relative risk reduction in each of these events produced by home birth and birth center births was identified from odds ratios reported in a recently published population-based retrospective study of outcomes for women who had planned home births or birth center births. [bib_ref] Maternal and perinatal outcomes by planned place of birth in Australia 2000-2012:..., Homer [/bib_ref] These relative risk reduction values were applied to the observed probability of unplanned cesarean, vaginal birth with forceps, and vaginal birth with vacuum to give a counterfactual probability of each of these birth types occurring . Monte Carlo simulation 25 was then used to randomly assign the low-risk women in the second subdata set to each of these birth types.
Still using the second sub-data set, subsequent health resource use for the records of low-risk women (the "recipient" records) was then imputed by matching to similar "donor" records who had the same demographic and clinical characteristics, and mode of birth. The donor records were drawn from the complete linked data set of low-risk women (covering births from 01/07/2012 and 30/06/2015). Recipient records were then assigned the health resource use trajectories of the donor records, thus representing the counterfactual scenarios. For example, if a recipient record of a woman was assigned in the Monte Carlo simulation step to have a vaginal birth with forceps, that record was given the subsequent health resource use of a similar donor record who actually had a vaginal birth with forceps. This recreates the actual observed dynamics in health states and resource use captured in the real-world data. [bib_ref] The redistributive features of the Italian pension system: The importance of being..., Leombruni [/bib_ref] Radius matching was used because of its performance with real-world data. [bib_ref] The performance of estimators based on the propensity score, Huber [/bib_ref] Matching scores were based on a woman's age, BMI score, if it was the woman's first pregnancy, smoking status before 20 weeks' gestation, Indigenous identification, socioeconomic status, and rurality of residence. These variables were chosen as they have previously been shown to be associated with total health care costs 28 -an outcome of primary importance 29 -but not influenced by mode of birth. Recipient and donor records were matched if their score fell within 0.02 standard deviations of the logit of the matching score, using the greedy matching technique.
For the home birth simulation, it was assumed that where women gave birth vaginally with no forceps or vacuum, there was no inpatient bed use for either the woman or baby at time of birth. However, ICU, SCN, and NICU use was included from the matched donor records. Those with an unplanned cesarean, vaginal birth with vacuum, or vaginal birth with forceps were assumed to be transferred to a public hospital with inpatient, SCN, NICU, and ICU use based on the counterfactual modeling. As this study focused on acute inpatient resource use, we did not consider the ambulance transfer resource use that may be required.
## | scenario 2-birth centers
The second simulation of the study estimated the health service resources used if all low-risk women gave birth at birth centers. Low risk was defined as above, and the aforementioned counterfactual methodology was repeated. For the birth center simulation, it was also assumed that where women gave birth vaginally with no forceps or vacuum, there was no inpatient bed use for either the woman or baby at time of birth, and both remained in the birth center until they were discharged home. However, ICU, SCN, and NICU use was included. Those with an unplanned cesarean, vaginal birth with vacuum, or vaginal birth with forceps were assumed to be transferred to a public hospital with inpatient, SCN, NICU, and ICU use based on the counterfactual modeling.
## | generation of results
After the simulation, the two sub-data sets of women with risk factors and low-risk women were recombined, and the number of inpatient bed days, ICU hours, SCN days, and NICU days was compared with that in the base data set (representing current standard care). This was repeated for the two counterfactual simulations representing the home birth and birth center scenarios. The mean number of health resources used per woman, and the health resource savings that could be made per 1,000 public hospital births were presented.
# | results
There were 44 498 records of women in the base data set, which once weighted represented 215 615 women giving birth in Australian public hospitals in 2017. Of these women, 43.9% were considered low-risk and 56.1% were considered higher risk. From our base model (standard care), 58.2% of women had a vaginal birth without forceps or vacuum, and 29.1% had an unplanned cesarean. Of the low-risk women, a higher proportion (70.8%) had a vaginal birth without forceps or vacuum, and 13.6% had an unplanned cesarean .
The results of our counterfactual model show that if all low-risk women had a home birth, then 93.5% would have a vaginal birth with no forceps or vacuum, 2.1% would have a vaginal birth with vacuum, 0.7% would have a vaginal birth with forceps, and 3.7% would have an unplanned cesarean . This would have increased the population-level percentage of women having a vaginal birth without forceps or vacuum to 68.6%, and the percentage of women having an unplanned cesarean would have reduced to 24.5%.
If all low-risk women had given birth in a birth center, 85.9% would have had a vaginal birth with no forceps or vacuum, 5.1% would have had a vaginal birth with vacuum, 2.3% would have had a vaginal birth with forceps, and 6.7% would have had an unplanned cesarean . The populationlevel percentage of women having a vaginal birth without forceps or vacuum would have increased to 65.4%, and the percentage of women having an unplanned cesarean would reduce to 25.9%.
Under current standard care, the mean number of inpatient bed days per birth is 2.6 . The mean number of ICU hours for women is 0.1, and the mean number of days infants spend in SCN and NICU is 1.1 and 0.4, respectively.
If all low-risk women gave birth at home, the mean number of inpatient bed days would reduce to 1.7 per birth. The total number of bed days used in 2017 would have reduced from 539 953 to 356 828 if all low-risk women gave birth at home. The total number of hours women spent in the ICU would have reduced from 27 737 to an estimated 24 896 . Similarly, if all low-risk women gave birth in a birth center, the mean number of inpatient bed days per birth would be 1.8. The total number of inpatient bed days would have reduced to 374 453, and the total ICU hours for women would have reduced to 25 764.
If home birth was available to all low-risk women, then for every 1000 births, an estimated 860 inpatient bed days, 11.5 infant SCN days, 2.1 NICU days, and 10.1 ICU hours for women would be saved . If birth in a birth center was available to all low-risk women, then for every 1,000 births, an estimated 768.2 inpatient bed days, 3.9 SCN days, 1 NICU days, and 5.6 ICU hours for women would be saved .
# | discussion
Our analysis indicates that enabling all low-risk women to routinely give birth at home or in birth centers would substantially increase the rate of spontaneous vaginal birth (ie, without assistance) and reduce the rates of unplanned cesarean by more than half. Importantly, substantial resource savings would arise from reduced inpatient bed days and hours spent in the ICU by women if all low-risk births moved out of the acute inpatient public hospital setting. This study contributes to the current dearth of literature about resource savings associated with giving birth at home or in birth centers. Furthermore, it is the first study that we are aware of, to consider the discussion of resource savings for childbirth in the post-COVID-19 era. This research is crucial for informing discussions about freeing up acute inpatient hospital resources for those with the greatest need, and preparing for the next, inevitable global pandemic. The findings of this study depict that rates of unplanned cesarean would have reduced from 13.4% per 1000 women to less than 4% in 2017 if all low-risk women gave birth at home, or less than 7% if they gave birth in a birth center. This has significant implications for the Australian public health care system in terms of resource savings. Rates of cesarean are particularly high in low-risk nulliparous women. [bib_ref] Systematic review: elective induction of labor versus expectant management of pregnancy, Caughey [/bib_ref] [bib_ref] Elective induction of labor and the risk of cesarean section in low-risk..., Jonsson [/bib_ref] Among nulliparous women in Australia, the rate of cesarean increased by 4% between 2004 and 2017.Cesarean in nulliparous women can also increase the risk of complications in subsequent births for women and babies, and increases the chances of undergoing additional cesarean, which exacerbates the risks and costs associated with birth. [bib_ref] Delivery after previous cesarean: long-term maternal outcomes, Silver [/bib_ref] [bib_ref] Risk of complications in a second pregnancy following caesarean section in the..., Taylor [/bib_ref] Unplanned cesarean in particular are the most costly birth procedure in Australia. [bib_ref] Cumulative economic implications of initial method of delivery, Allen [/bib_ref] Recent research demonstrates that the cost of births by cesarean in Australian public hospitals was AU$31 939 in comparison with AU$18 521 for vaginal births with no instruments, including costs for women and children up to two years postpartum. [bib_ref] Financing maternity and early childhood healthcare in the Australian healthcare system: Costs..., Callander [/bib_ref] Though not as disparate, evidence from other countries also indicates that there is a significant difference between the procedural cost of vaginal birth and cesarean. [bib_ref] Drivers of maternity care in high-income countries: can health systems support woman-centred..., Shaw [/bib_ref] [bib_ref] Costs analysis of hospital care for vaginal delivery and elective caesarean section..., Entringer [/bib_ref] Consequently, the results of the current study suggest that shifts toward home and birth center births for low-risk women may result in health care system savings in the tens to hundreds of millions of dollars by reducing the proportion of unplanned cesarean undertaken, and significant improvements in women's and children's health outcomes.
Our findings also depict that although shifting all lowrisk women's births to birth centers and home births may only modestly reduce infant NICU and SCN days, we could see major reductions in the number of days women spent as inpatients, and hours spent in the ICU. A 2012 Australian study comparing women giving birth in midwifery-led birth centers and those experiencing usual care (birth in a hospital labor ward) evidenced similar resource savings. [bib_ref] A non-randomised trial investigating the cost-effectiveness of midwifery group practice compared with..., Toohill [/bib_ref] The authors concluded that there was a statistically significant difference between the cost of receiving care in the birth center versus the hospital, which equated to roughly AU$1000 saved per birth for the hospital and an additional AU$1000 saved by the government. [bib_ref] A non-randomised trial investigating the cost-effectiveness of midwifery group practice compared with..., Toohill [/bib_ref] However, despite these resource savings, few women have access to alternate birth settings.
In 2017, 97% of births in Australia occurred in hospitals, 74% in public hospitals, and 26% in private hospitals.Only 0.3% of births took place in the home.Low rates of home births are underpinned by several contentious issues. First, there are very few publicly funded home birth programs in Australia, with different reports suggesting between 14 and 19 known programs. 39,40 Low rates of home birth are also partly because of professional indemnity insurance restrictions placed on both public and private practicing midwives in the early 2000s.Most available programs cater to women in metropolitan areas, thereby propagating inequity of access for regionally and remotely located women. Moreover, neither Queensland nor Tasmania offer publicly funded home birth programs.Women's access to home birth can also be financially restricted because of a lack of public funding. Women who want to give birth at home with a private midwife can expect to pay between $3500 and $6000 out of pocket.Even women eligible for publicly funded home births can still expect to pay up to $1500 out of pocket depending on the additional tests, scans, and support sought.Thus, both geographical and financial limitations impede Australian women's access to suitable support for home birth.
There are more birth centers than home birth programs in Australia, but only marginally. In 2016, there were at least 10 birth centers in NSW-five colocated on hospital premises or adjacent to hospital labor wards, and five midwifery-led, free-standing birth centers. [bib_ref] Mapping the trajectories for women and their babies from births planned at..., Scarf [/bib_ref] Pregnancy Birth and Beyond Pty Ltd lists 24 birth centers across Australia on their website,catering to the 2.4% of Australian births.By comparison, The Netherlands has a reported 23 birth centers (as of 2017), [bib_ref] Defining and describing birth centres in the Netherlands: a component study of..., Hermus [/bib_ref] and a much higher birth center birth rate of 11.4%. [bib_ref] Drivers of maternity care in high-income countries: can health systems support woman-centred..., Shaw [/bib_ref] Generally, there is a paucity of literature on Australian women's preferences for place of birth. Stoll and colleagues examined 760 Western Australian (WA) university students' (>75% of whom were female) preferences for place of birth. [bib_ref] Home or hospital? Midwife or physician? Preferences for maternity care provider and..., Stoll [/bib_ref] Close to half of the participants preferred to give birth in a hospital under obstetrician-led care, roughly 36% preferred a hospital birth under midwifery-led care, and 10% and 1.8% respectively preferred to give birth in birth centers or at home. [bib_ref] Home or hospital? Midwife or physician? Preferences for maternity care provider and..., Stoll [/bib_ref] This does not show an overwhelming preference for home birth or birth center births. Instead, this may be a reflection of what participants know to be available, have previously been exposed to, or view as accepted. However, while not extreme, the results do indicate a degree of disparity between current rates of home births and birth center births, and the proportions of women that would prefer these options. An overwhelming majority of the submissions made to the Australian Maternity Services Review in 2009 (a review of maternity services in Australia undertaken by the Government Department for Health and Ageing, eliciting a range of perspectives on gaps in the system at the time) were from women who wanted to be able to give birth in their home, but were unable to do so because of limited access and exorbitant out-of-pocket fees associated with the available services. [bib_ref] Birth centres and the national maternity services review: Response to consumer demand..., Dahlen [/bib_ref] [bib_ref] Home birth and the national Australian maternity services review: too hot to..., Dahlen [/bib_ref] The recent COVID-19 pandemic has also led some women to reconsider giving birth in hospitals because of perceived high risks of infection,and restrictions on the presence of family or other support persons. Alarmingly, the pandemic may have also seen a rise in free births (home births unattended by any health professional) as a consequence of women's preferences for home birth not being met. Consequently, future research should aim to formally and comprehensively examine Australian women's preferences for birthing at home and in birth centers in order to appropriately inform birthing service-related decisionmaking. In particular, preference-based methods, such as those that have been used in the United Kingdom, 50 may provide a comprehensive backdrop to inform such decisions.
## | recommendations
In the wake of COVID-19, there have been calls to make birthing options outside the hospital more accessible to pregnant women. Although there is some evidence to suggest that Australian women would prefer these alternative birthing models of care, current public and private funding arrangements and limited accessibility to nonhospital birthing services pose significant restrictions to women's ability to choose where and how they give birth. Thus, the current study proposes three key recommendations. First, greater examination of women's preferences for birth center and home births in Australia is crucial to understand the demand for these services. It may also provide us with an indication of the types of funding models that could feasibly support equitable access to these services. Second, few Australian women can access and use home births or birth centers because of the limited number of these services nationally, and their tendency to operate in metropolitan cities. Thus, to promote equity in choice of birthplace-irrespective of geographical location-future research should attempt to understand the cost-effectiveness of providing birth center and home birth models of maternity care in regional and remote Australia. Third, future research should consider modeling at the individual hospital jurisdiction level based on local population characteristics and risk factors to identify the actual proportion of women who could use home birth or birth center birth services.
# | limitations
A key limitation of the study is that resource savings represent those that could have been made in 2017 instead of the current year (2020), as this is the latest national benchmarking data available. However, this was done so that the sample weighting undertaken was representative of the Australian population. Assumptions in our modeling, which are potential limitations, include the following: transfer into hospital for care such as epidural where a spontaneous vaginal birth was still achieved, other factors that influence suitability for different birth settings such as women's preferences, and additional risk factors not included in our modeling. In addition, the results represent absolute resource savings, either all low-risk births occurring in birth centers or all low-risk births occurring at home. It may be unreasonable to assume that all women giving birth would choose the same option, or that no low-risk women would prefer to give birth in public or private hospitals.
There are also recognized differences in birth location recommendations for women giving birth for the first time compared with subsequent pregnancies, and differences between maternity practitioner practices. However, the current analysis aimed to illustrate potential resource savings and generate discussion about nonhospital birthing options that are currently inaccessible to many women giving birth in Australia, particularly in light of the COVID-19 global pandemic. Finally, our modeling did not include women's antenatal model of care because this was not available in the data set. It is recognized that the model of antenatal care has significant impacts on birth outcomes, [bib_ref] Midwifeled continuity models versus other models of care for childbearing women, Sandall [/bib_ref] and is worthy of further research.
# | conclusions
The recent COVID-19 pandemic has highlighted the importance of women's ability to choose to give birth at home and in birth centers, and the availability of these services. The current study demonstrates that if all low-risk women gave birth at home or in birth centers, there would be significant reductions in women's health service resource use, and modest reductions in that of their baby. However, we currently have a limited understanding of Australian women's preferences for these services. Future research should aim to understand women's preferences for, and willingness and ability to pay for these services to inform suitable government funding schemes. This research has the potential to address current inequities underpinning maternity care in Australia, and promote greater capacity in woman-centered maternity care globally. |
Low hydroxychloroquine blood levels in patients who have had gastric bypass surgery
Gastric bypass surgery, also called Roux-en-Y gastric bypass (RYGB), can result in the malabsorption of medications, requiring the use of higher than usual doses in order to achieve a therapeutic effect. We describe the results of hydroxychloroquine (HCQ) blood levels in three patients with systemic autoimmune disease taking standard HCQ doses and their associated disease activity levels. This is a retrospective review of all patients who had undergone RYGB and were taking HCQ in a rheumatology community-based practice. Two patients with SLE and one patient with primary Sjogren's syndrome had previously undergone RYGB. All three had subtherapeutic HCQ blood levels and active disease. Increasing their HCQ doses above the recommended 400 mg a day dosing resulted in therapeutic HCQ levels in all three patients and better disease control in two of the three patients. RYGB patients may not absorb HCQ adequately, resulting in subtherapeutic HCQ blood levels and inadequate disease control. Patients who have undergone RYGB and are taking HCQ should have drug levels monitored. RYGB patients may require higher than recommended doses of HCQ in order to achieve better disease control and avoid unneeded additional immunosuppressive agents.
# Abstract
Gastric bypass surgery, also called Roux-en-Y gastric bypass (RYGB), can result in the malabsorption of medications, requiring the use of higher than usual doses in order to achieve a therapeutic effect. We describe the results of hydroxychloroquine (HCQ) blood levels in three patients with systemic autoimmune disease taking standard HCQ doses and their associated disease activity levels. This is a retrospective review of all patients who had undergone RYGB and were taking HCQ in a rheumatology community-based practice. Two patients with SLE and one patient with primary Sjogren's syndrome had previously undergone RYGB. All three had subtherapeutic HCQ blood levels and active disease. Increasing their HCQ doses above the recommended 400 mg a day dosing resulted in therapeutic HCQ levels in all three patients and better disease control in two of the three patients. RYGB patients may not absorb HCQ adequately, resulting in subtherapeutic HCQ blood levels and inadequate disease control. Patients who have undergone RYGB and are taking HCQ should have drug levels monitored. RYGB patients may require higher than recommended doses of HCQ in order to achieve better disease control and avoid unneeded additional immunosuppressive agents.
# Introduction
Obesity, commonly defined as having a body mass index (BMI) ≥30 kg/m 2 , is a common problem in society. In SLE, some studies show that obesity is associated with higher disease activity and worse outcomes.Bariatric surgery is the most effective treatment for morbid obesity and Roux-en-Y gastric bypass (RYGB), also called gastric bypass surgery, is one of the most common weight loss procedures performed.Hydroxychloroquine (HCQ) and other antimalarial medications are the cornerstone immunomodulatory treatment for SLE.Most recently, the American Academy of Ophthalmology (AAO) recommended that HCQ should not be dosed above 5.0 mg/kg of actual body weight.This dosing recommendation was based on the results of a large retrospective study of a Kaiser Permanente Northern California database.However, these recommendations have been controversial.The ability to measure blood HCQ levels in clinical practice has been helpful in identifying patients who are non-adherent to their therapeutic regimens, 7 often leading to significantly better disease control when confronted with the results.RYGB surgery can result in the malabsorption of some medications due to changes in the anatomy of the gastrointestinal tract and its effects on gastric and intestinal pH, gastric emptying time, intestinal transit time, surface area for absorption and first pass metabolism.As bariatric surgery is more commonly performed, the prevalence is increasing among our patients with systemic autoimmune disease, and we need to be cognizant of how it may affect the management of their disease. The currently recommended maximum daily maintenance dose of HCQ is 400 mg/day. However, the oral bioavailability and efficacy of this maximum dose has not been studied in patients who have undergone RYGB surgery.
Here we report two patients with SLE and one patient with primary Sjögren's syndrome who had undergone RYGB surgery, had active disease and low HCQ blood levels on 400 mg daily of HCQ. Considering malabsorption as a potential cause of lower than desired HCQ levels, each patient had their dose raised above the recommended dose, achieving therapeutic levels; two patients had corresponding better disease control as well.
## Cases case 1
The patient is a 33-year-old, 182-pound (83 kg) Caucasian female (weight range 162-183 pounds; BMI 24.6-27.4 kg/m 2 ) who presented to us in 2017 after presenting with partial complex seizures, posterior cyclitis, inflammatory polyarthritis, photosensitive rash, painful nasal and oral ulcers, malar rash, recurrent pleuritic chest pain, Raynaud's phenomenon,
## Lupus science & medicine
103°F (39°C) fever, severe fatigue, thrombocytopenia, lymphopenia, non-nephrotic range proteinuria and positive ANA 1:320 homogeneous pattern. Further rheumatological workup disclosed positive chromatin antibody; renal biopsy showed non-specific immunoglobulin M mesangial deposits, and a previous mesenteric lymph node biopsy was consistent with reactive adenitis. Dermatological evaluation agreed with a diagnosis of acute cutaneous lupus lesions by physical examination. Initial treatment consisted of HCQ (two 200 mg tablets taken once a day for a total of 4.8 mg/kg/day), mycophenolate mofetil and intermittent intramuscular corticosteroid injections; she was already on antiepileptic therapy by her neurologist and her 25-hydroxy (OH) vitamin D level was greater than 40 nmol/L while taking ergocalciferol 50 000 IU weekly. Her flares decreased in severity and frequency with this regimen but incompletely. Recurrent steroid-responsive severe headaches with her flares ended up requiring moderate doses of prednisone to provide relief. After several months, subcutaneous belimumab was added 200 mg a week. She continued to have flares with cutaneous, joint and constitutional symptoms. A blood HCQ level was obtained via liquid chromatography coupled with mass spectrometry (testing by Exagen Diagnostics) showing a subtherapeutic level of 593 ng/ mL (a therapeutic level is considered ≥1000 ng/mL per Exagen). The patient verbally insisted on close to 100% adherence to taking her HCQ, and a verbal conversation between her rheumatologist and pharmacist disclosed regular on-time refills from her pharmacy suggesting excellent adherence. It was noted that she had previously had RYGB surgery in 2009 for the treatment of morbid obesity, so the possibility of malabsorption of HCQ was considered, and her dose was increased to 200 mg three times daily dosing (for a total of 7.4 mg/kg/day). A repeat level 3 months later showed a significantly improved level of 1684 ng/mL while taking 200 mg three times daily. However, she developed central nervous system (CNS) lupus complications at this visit requiring changing her immunosuppressant therapy. Due to having a high HCQ blood level on 200 mg three times daily, her dose was again adjusted to 200 mg twice daily. A repeat HCQ blood level 3 months later was again lower at 520 ng/mL after which we increased her HCQ dose to 500 mg daily (6.7 mg/kg/ day) subsequently resulting in a blood level of 1032 ng/ mL 3 months later. During all HCQ blood measurements, she had normal renal function with an estimated glomerular filtration rate (eGFR) range of 64-118 mL/min/1.73 m 2 (normal is >59 mL/min118 mL/min/1.73 m 2 ) and she did not have protein malnutrition from her RYGB (her serum albumin and protein levels were normal).
## Case 2
The patient is a 57-year-old, 269-pound (122 kg) Caucasian female (range 220-272 pounds; BMI 37.6-46.1 kg/ m 2 ) who was diagnosed with SLE in 1998 manifested by biopsy-proven lupus profundus, lower extremity ulcerations, nonpalpable purpura, photosensitive rash, inflammatory polyarthritis, hepatosplenomegaly, positive ANA 1:640 nucleolar pattern and a history of antiphospholipid syndrome with three miscarriages, right deep venous thrombosis and positive antiphospholipid antibodies. We started seeing the patient in late 2014 when her treatment was 300 mg/day of HCQ and ergocalciferol 50 000 IU weekly. She had been having intermittent flares of her lupus panniculitis, polyarthritis and severe fatigue. Her labs showed no active serology and her 25-OH vitamin D level was above target at 70.3 nmol/L. She was initially treated with intramuscular corticosteroids and oral methotrexate (MTX) 20 mg a week while also increasing her HCQ to two 200 mg tablets taken together once daily (3.4 mg/kg/day). She had a prompt response with resolution of her polyarthritis and panniculitis. However, over the next few years, she had recurrent, milder flares of her cutaneous, joint and constitutional manifestations; these were treated with intramuscular corticosteroids and increasing doses of MTX (advanced over time to subcutaneous doses). In June 2018, during another flare, her HCQ drug level, while on a dose of 400 mg/day (3.2 mg/ kg/day for her body weight at the time), was first measured and was 525 ng/mL. She verbally insisted on close to 100% adherence, and a phone call to her pharmacist confirmed on-time refills of her medications. She had had an RYGB surgery in 2010. The possibility of malabsorption was considered, and her HCQ dose was increased to 200 mg three times daily (4.9 mg/kg/day). A repeat HCQ level 3 months later was much better at 960 ng/mL. She was in remission on that clinical evaluation as well as the subsequent 9 months. During all HCQ blood measurements, she had normal renal function with an eGFR range of 70-104 mL/min/1.73 m 2 , and she did not have signs of protein malnutrition from her RYGB.
## Case 3
The patient is a 64-year-old, 259-pound (117 kg) African-American female (range 210-282 pounds; BMI 31.9-44.1 kg/m 2 ) diagnosed with primary Sjögren's syndrome (pSS) by her previous rheumatologist based on a positive ANA, positive rheumatoid factor and keratoconjunctivitis sicca (per her ophthalmologist); she was taking HCQ 400 mg once a day. We began to see the patient in 2010, and because of initially negative autoantibodies, we performed a minor salivary gland biopsy which confirmed the diagnosis with a Chisholm-Mason grade 3. Over the years, she had gradually lost teeth due to xerostomia even while on cevimeline, use of xylitol gum and high-dose fluoride toothpaste. She also had progressive peripheral neuropathy and intermittent bouts of inflammatory arthritis which responded well to intramuscular corticosteroids. In July 2018, during a bout of episcleritis and inflammatory oligoarthritis (shoulder and proximal interphalangeal joint), a HCQ level was checked as a screening test for adherence. Her level was subtherapeutic at 582 ng/mL on 400 mg once a day (3.4 mg/kg/ day). She verbally insisted on close to 100% adherence, and her pharmacist confirmed excellent on-time refills as well. Having had had an RYGB surgery in 2004, the possibility of malabsorption was entertained, and her dose was increased to 200 mg three times daily (5.1 mg/kg/day). A repeat level 3 months later was 829 ng/mL. She had no active arthritis or episcleritis on that occasion nor 9 months later. During all HCQ blood measurements, she had normal renal function with an eGFR range of 60-83 mL/min/1.73 m 2 , and she did not have signs of protein malnutrition from her RYGB.
# Discussion
It is recommended that all patients with SLE be treated with an antimalarial medication, most commonly HCQ.Out of all the possible side effects, antimalarial-induced toxic maculopathy is the most important potential side effect of HCQ.Fortunately, retinopathy does not occur in most patients who take HCQ. When caught early using the AAO recommended yearly screening tests, mild retinopathy is asymptomatic and does not progress. If HCQ is stopped, the progression of the retinopathy can be halted. Dosing recommendations for the use of HCQ in the treatment of lupus have fluctuated over time. The Plaquenil brand of HCQ manufacturer's insert recommends prescribing no more than 400 mg/day for lupus, while up to 600 mg/day can be used initially for the treatment of rheumatoid arthritis.Melles and Marmor evaluated the records of 2361 Kaiser Permanente HCQ long-term users retrospectively.They identified 163 cases of retinal toxicity using Humphrey central visual field (HVF 10-2) testing and spectral domain optical coherence tomography (SD-OCT) and found that the prevalence of retinal toxicity was least common among patients with an actual body weight dose of less than 5.0 mg/ kg. The findings from the Kaiser study led to the AAO recommending that HCQ be dosed at less than 5 mg/kg of actual body weight.However, this dosing guideline was fundamentally flawed because the Kaiser study was based on the doses the pharmacy dispensed rather than the dose the provider prescribed. A high percentage of patients with chronic illnesses do not pick up some of their prescribed medications at the pharmacy, 14 so the 5 mg/kg dosing that was calculated using the pharmacy-dispensed dose was probably lower than the actual dose the providers prescribed per body weight. We do not know the differences between the pharmacy-dispensed amounts and the provider-prescribed doses, but we know that it is likely that the provider-prescribed dose was somewhat higher. In other words, the actual maximum dose of HCQ to decrease the risk for maculopathy may be 5.5-6.5 mg/kg of actual body weight rather than the reported 5.0 mg/kg guideline recommendation.
Since the AAO used these results to formulate their dosing guidelines, these recommendations are certainly open to question. A further problem with the AAO guidelines is that they only addressed this potential side effect but did not address therapeutic efficacy. It is important when giving dosing recommendations that both efficacy and potential adverse events be considered together, especially considering how safe and essential HCQ is overall.
Recently, the ability to check blood levels of HCQ has become available in clinical practice. Costedeau-Chalumeau and colleagues showed that patients with SLE with active disease had a significantly lower mean wholeblood HCQ concentration (694±448 ng/mL, similar levels as our three patients) compared with patients with inactive disease (1079±526 ng/mL).Although serum and plasma HCQ levels may also be checked, it has been shown that it is better to check whole-blood levels which are more precise, have higher concentrations, and because of HCQ's preferred binding to albumin, platelets and leucocytes.Measuring HCQ blood levels have also been shown to be an excellent way to check for patient adherence to therapy. Durcan et al showed that when their patients with SLE were confronted with low HCQ levels due to non-adherence, not only did adherence significantly improve but lupus disease activity also significantly decreased.Then, Costedoat-Chalumeau et al showed that HCQ blood level measurements identified patients with SLE who were having disease flares related to severe non-adherence. There was poor correlation between actual blood levels and physician assessment of adherence as well as to adherence determinations via patient self-administered questionnaires (the Medication Adherence Self-Reported Inventory). With 68.9% of the flaring patients requiring increased doses of steroids and the poor identification of adherence rates by the physicians (without the use of HCQ levels), the authors recommended the routine use of HCQ level measurements.So far, no HCQ dosing guidelines have taken into account the use of HCQ blood levels.
Other than using actual body weight to determine HCQ dosing, few other parameters have been recommended in helping to determine proper dosing. The absorption and metabolism of HCQ in various patient populations have not thus far been extensively studied. Although HCQ has relatively high absorption from the gastrointestinal tract, we do not know if there are ethnic, genetic or digestive-issue differences in HCQ absorption, metabolism and efficacy. Whole blood concentrations of HCQ vary widely, even after similar dosing in patients.Initial observations from the Plaquenil Lupus Systemic (PLUS) study group showed no association between ethnicity, smoking and antacid or cytochrome P450 enzyme affecting medication interactions with HCQ whole-blood levels.However, high BMI, high estimated creatinine clearance and increased time between the last tablet taken and the measurement of blood levels were associated with lower HCQ concentrations. Patients with chronic kidney disease tended to have higher blood levels.
Currently, the optimal whole-blood level of HCQ to achieve efficacy in the treatment of systemic autoimmune diseases such as SLE is not known. The PLUS study confirmed that lower HCQ whole-blood levels are Lupus Science & Medicine associated with higher SLE disease activity. Patients who maintained HCQ levels ≥1000 ng/mL tended to have fewer flares over time than those with lower levels.However, this persistent level was obtained in a relatively small number of patients.
In our small series of patients (two with SLE and one with pSS), we have identified three patients who had previously had RYGB surgery. All three were taking HCQ at a dose of two 200 mg tablets once a day (total of 400 mg daily), had active inflammatory disease and had HCQ levels similar to the patients with active SLE disease activity in Costedoat-Chalumeau's 2006 study; these levels were below our laboratory's therapeutic dose recommendation. Proper adherence to regularly taking their HCQ was confirmed in all three cases by verbal affirmation from the patient and confirmation by each patient's pharmacist stating that the patient was picking up her HCQ prescriptions regularly and on time. Each patient had their dose of HCQ increased to 200 mg three times daily. Three months after each dose increase, each patient had their HCQ level repeated, and all three had significantly improved drug levels. One of the patients with SLE and the one patient with pSS had significantly improved disease control at that time (remission in both); the patient with SLE who had CNS involvement (case 1) did not.
RYGB surgery is one of the most commonly used surgical procedures to treat morbid obesity.RYGB is performed by stapling and dividing the proximal stomach creating a small gastric pouch which the surgeon connects to the jejunum (bypassing the body and antrum of the stomach and the duodenum). The effect is that vitamins, nutrients and medications have decreased exposure to gastric acid and no exposure to the absorptive mucosa of the duodenum or the proximal jejunum. It results in a large amount of excess body fat reduction in most patients as well as decreases many of the comorbid conditions. However, chronic malabsorption, especially of micronutrients and fat-soluble vitamins, is an expected consequence requiring lifelong supplementation of these nutrients.
The oral bioavailability of medications can also be affected due to changes in the anatomy of the gastrointestinal tract and its effects on gastric and intestinal pH, gastric emptying time, intestinal transit time, surface area for absorption and first-pass metabolism changes.Some medications have been shown to have lower absorption after RYGB surgery as demonstrated by lower blood levels,and physicians may need to adjust drug doses due to the possibility of post-RYGB malabsorption.To our knowledge, and after a search of the published medical literature, the oral bioavailability of HCQ has not been studied in patients who have undergone RYGB surgery. RYGB surgery increases gastric pH which can reduce the absorption of medications that are weak bases, such as HCQ.Also, it is a common practice among bariatric surgeons to prescribe proton pump inhibitors (PPI) to prevent gastrojejunal anastomotic ulcerations.All three patients were taking a PPI which would increase gastric pH further, which could further reduce HCQ absorption. Although malnutrition is evident in a small percentage of RYGB patients,this could be an additional reason for decreased absorption in some RYGB patients.The low levels of HCQ in these three RYGB patients suggest that HCQ may not be adequately absorbed resulting in lower HCQ levels and increasing the potential for inadequate systemic autoimmune disease activity control. Such patients (post-RYGB surgery) may then require the addition of steroids or immunosuppressant medications, increasing the potential risk for side effects. RYGB patients may require adjustments to their HCQ dosing by using higher than recommended doses. The optimal range of HCQ whole-blood levels to achieve the best efficacy balanced with decreased adverse events, such as retinopathy, is unknown. A preliminary study has shown that levels less than 1195 ng/mL are associated with lower retinopathy risk.A new paradigm is needed for determining appropriate HCQ blood levels to achieve good efficacy while also avoiding retinal toxicity.
Some caution does need to be used when considering the use of higher than recommended doses of HCQ (>400 mg/day) in patients who have undergone RYGB surgery. A study using 800 mg/day of HCQ (averaging 11.5 mg/ kg/day) in patients with graft versus host disease (GVHD) showed that three out of the 12 patients studied developed macular changes suggestive of possible early HCQ retinopathy using SD-OCT, HVF 10-2 and multifocal electroretinography.However, these results are probably not applicable to the RYGB population in that they did not have malabsorption, did not have blood levels of HCQ monitored in the study and GVHD itself has been associated with microvascular retinopathy.These three patients with systemic autoimmune disease who had undergone RYGB surgery had lower than expected HCQ levels while taking two 200 mg HCQ tablets (400 mg total) at the same time once daily and had active disease activity. All three had significantly improved drug levels after an increase of their HCQ to 200 mg three times daily and two of the three had much better disease control after this dose adjustment. These findings suggest that further study should be done in a larger group of RYGB patients, assessing HCQ drug levels at various dosing regimens while also assessing disease activity and adverse events. Also, two of the three patients had very high BMIs which may have contributed to their low HCQ whole blood levels.The real possibility of malabsorption of HCQ from RYGB surgery deserves to be considered and studied. Up to this time, we may potentially have been undertreating our patients who have undergone RYGB surgery by using subtherapeutic doses of HCQ, resulting in some patients being exposed to additional medications carrying more dangerous side effects. These results need to be assessed and confirmed in larger studies.
Contributors JL is an ophthalmologist who was one of the authors of the American Academy of Ophthalmology guidelines for hydroxychloroquine dosing and retinopathy screening. He assisted in supplying the comments and literature regarding the background for these recommendations and their potential shortcomings and how it pertains to the manuscript. BG is a bariatric surgeon. He assisted in supplying the comments and literature regarding the potential for the malabsorption of drugs in people who have undergone gastric bypass surgery. MP has written several papers regarding the utility of using hydroxychloroquine drug levels in the management of patients with lupus. She contributed comments and articles regarding the proper use of drug levels in these patients and how it could affect proper management of their connective tissue diseases.
Funding Exagen Diagnostics paid for part of the publication fee, otherwise, the authors received no financial support for the research, nor authorship.
Competing interests DET is on the Speakers' Bureau for Exagen Diagnostics. Exagen Diagnostics performs laboratory testing for hydroxychloroquine drug levels. Provenance and peer review Not commissioned; externally peer reviewed. data availability statement No additional data are available.
open access This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http:// creativecommons. org/ licenses/ by-nc/ 4. 0/. |
Prevention of Hypothermia in the Aftermath of Natural Disasters in Areas at Risk of Avalanches, Earthquakes, Tsunamis and Floods
# Introduction
'Only human beings can recognize catastrophes, provided they survive them; nature recognizes no catastrophes.' Max Frisch, Man in the Holocene. 1979.
Natural disasters have plagued humankind since the beginning of our species. Since modern humans spread out of Africa 60,000-70,000 years ago, they had to protect against cold with shelter, clothing, and fire. Natural disasters can destroy all personal belongings and infrastructure and can deprive humans of the possibility of protecting themselves from harsh environments. One consequence may be death from accidental hypothermia. This article gives an overview of the history, epidemiology, mechanisms, pathophysiology, rescue, and treatment of accidental hypothermia caused by natural disasters such as avalanches, earthquakes, tsunamis, and floods. We provide an approach to the prevention of accidental hypothermia. We also discuss the management of multiple casualties, which is often necessary after natural disasters. A multi-casualty incident is an event in which the number of victims exceeds or substantially stretches the technical and medical resources of the local rescue system [bib_ref] Management of Multi-Casualty Incidents in Mountain Rescue: Evidence-Based Guidelines of the International..., Blancher [/bib_ref]. We conducted a non-systematic review of the causes and means of preventing accidental hypothermia after natural disasters from avalanches, earthquakes, tsunamis, and floods.
## Mechanisms of avalanche burial
Over the last several decades in Europe, the number of avalanche fatalities on roads and in villages has steadily declined because of preventive measures. Currently, about 90% of the avalanches affecting humans in Europe or North America are triggered by recreationists [bib_ref] Correlation between Avalanche Emergencies and Avalanche Danger Forecast in the Alpine Region..., Rainer [/bib_ref] [bib_ref] Characteristics of Human-Triggered Avalanches, Schweizer [/bib_ref] [bib_ref] Snow Avalanche Hazard in Canada-A Review, Stethem [/bib_ref]. The number of people venturing into avalanche terrain is unknown even in Europe and North America. Between 1983 and 2015, avalanches claimed over 5000 lives in Europe and North America, with an average annual toll of about 130 in Europe and 36 in North America [bib_ref] Wilderness Medical Society Practice Guidelines for Prevention and Management of Avalanche and..., Van Tilburg [/bib_ref]. Recreationists account for more than 90% of avalanche fatalities in the European Alps [bib_ref] Avalanche Fatalities in the European Alps: Long-Term Trends and Statistics, Techel [/bib_ref] and in North America [bib_ref] Canadian Avalanche Association, Jamieson [/bib_ref].
Worldwide, outside North America and Europe, mortality statistics are only rough estimates. Avalanche fatalities are not systematically reported in the highest mountain regions, including the Andes, the Karakoram, and the Himalayas, where even densely populated valleys and villages are not protected from avalanches. Occasionally, avalanche disasters claim many lives. With global warming, the risk of heavy, often wet, snowfall events, and consequently severe avalanche cycles, may increase in some mountain ranges. Safety precautions may have to be re-evaluated and increased [bib_ref] Effects of Climate Change on Avalanche Accidents and Survival, Strapazzon [/bib_ref]. Mixed avalanches, with snow, ice, and rock, may occur more frequently, sometimes causing glacial lake outburst floods (GLOFs) or forming dams that block rivers. These cascading hazards can lead to flooding, with far-reaching effects downstream [bib_ref] Disaster Chain Analysis of Avalanche and Landslide and the River Blocking Dam..., Jia [/bib_ref].
## Prevention and mitigation of avalanche burial
Avalanche disasters during World War I contributed to the founding of the first Commission for Snow and Avalanche Research in Switzerland in 1931. In 1942, the Institute for Snow and Avalanche Research was established in Davos, Switzerland. Three years later, the first civil avalanche warning service in the world became operational. Other warning services were established over time in several other European countries. Especially after 1951, public funds were invested in engineering works, such as for supporting structures above villages, snow sheds, and dams, and for planting protective forests. These measures have resulted in the almost complete elimination of avalanche disasters in the European Alps.
In general, strategies for dealing with avalanche hazards include preventing the release of avalanches, influencing their paths, and avoiding hazardous areas [bib_ref] Integrative Risk Management: The Example of Snow Avalanches, Bründl [/bib_ref]. Accidents can be prevented by controlling avalanches, by regulating human presence, and by placing structures outside avalanche paths. These measures require avalanche forecasting, public warnings, and hazard mapping for land use planning. These are relatively inexpensive means of mitigation. On the other hand, engineering works are expensive and require significant resources [bib_ref] The Avalanche Handbook, Mcclung [/bib_ref]. If avalanche hazards are evaluated daily, temporary road closures and evacuation of buildings can also be effective mitigation measures. The duration of these disruptions can be limited by intentionally triggering avalanches using explosives.
In the mountain areas where residents and recreationists are threatened by avalanches and where avalanche mitigation is still in its infancy, the first step should be to establish weather forecast and avalanche warning services. Weather forecasts, as well as current avalanche warnings, should be publicly available and widely distributed. In populated areas with a high risk of avalanches, emergency response and evacuation plans are necessary. Local authorities should be specifically trained in avalanche risk management [bib_ref] IFKIS-A Basis for Managing Avalanche Risk in Settlements and on Roads in..., Bründl [/bib_ref].
For recreationists, who account for most avalanche victims in Europe and North America, training in avalanche risk assessment is critical. Proper trip planning, adequate rescue equipment, and safe travel habits complement avalanche prevention during winter backcountry activities in avalanche terrain.
## Pathophysiology of avalanche burial with a focus on accidental hypothermia
A detailed description of the pathophysiology of avalanche burial and discussion of the avalanche survival curve are beyond the scope of this article. They can be found elsewhere [bib_ref] Comparison of Avalanche Survival Patterns in Canada and Switzerland, Haegeli [/bib_ref] [bib_ref] Burial Duration, Depth and Air Pocket Explain Avalanche Survival Patterns in Austria..., Procter [/bib_ref]. Although avalanche injuries may not be severe enough to cause death, the depressed level of consciousness caused by trauma may decrease the probability of survival by predisposing to asphyxiation, limiting shivering, and accelerating cooling. Relatively mild maritime climates with high snow density are associated with a significantly earlier decrease in survival compared to continental climates, with low snow density and cold temperatures [bib_ref] Comparison of Avalanche Survival Patterns in Canada and Switzerland, Haegeli [/bib_ref]. The longest survival of a completely buried avalanche victim in the open was 44 h [bib_ref] Accidental Hypothermia-an Update: The Content of This Review Is Endorsed by the..., Paal [/bib_ref]. The longest survival in a structure was 37 days [bib_ref] Post-Traumatic Stress Disorder: Supportive Evidence from an Eighteenth Century Natural Disaster, Parry-Jones [/bib_ref]. Survival longer than 35 min has been reported in up to 25% of completely buried avalanche victims who were able to breathe under snow. The term 'triple H syndrome', referring to the combined effects of hypoxia, hypercapnia, and hypothermia, was coined in 2003 [bib_ref] Hypoxia and Hypercapnia during Respiration into an Artificial Air Pocket in Snow:..., Brugger [/bib_ref]. Avalanche victims breathing into an air pocket in low-density avalanche debris or into a large or open air pocket may be able to survive for hours. Accidental hypothermia has been reported as the sole cause of death in only about 1% of completely buried avalanche victims. Trauma can increase the risk of hypothermia. A higher injury severity score is associated with a higher incidence and severity of hypothermia [bib_ref] Epidemiology of Accidental Hypothermia in Polytrauma Patients: An Analysis of 15,230 Patients..., Weuster [/bib_ref] [bib_ref] Admission Hypothermia and Outcome after Major Trauma, Wang [/bib_ref]. Hypothermia should be suspected in an avalanche victim presenting with an open (not obstructed) airway at extrication after burial for >60 min [bib_ref] Prognostic Factors in Avalanche Resuscitation: A Systematic Review, Boyd [/bib_ref] [bib_ref] European Resuscitation Council Guidelines 2021: Cardiac Arrest in Special Circumstances, Lott [/bib_ref]. Individual cooling rates during snow burial vary widely, from 0.1 - C/h [bib_ref] Hypercapnia Increases Core Temperature Cooling Rate during Snow Burial, Grissom [/bib_ref] to 9 - C/h [bib_ref] Full Recovery of an Avalanche Victim with Profound Hypothermia and Prolonged Cardiac..., Oberhammer [/bib_ref] [bib_ref] Extreme Cooling Rates in Avalanche Victims: Case Report and Narrative Review, Mittermair [/bib_ref]. It usually takes at least 1 h after avalanche burial for a victim to reach a core temperature < 30 - C [bib_ref] European Resuscitation Council Guidelines 2021: Cardiac Arrest in Special Circumstances, Lott [/bib_ref]. During rescue operations there is an additional risk of post-extrication cooling when avalanche victims are exposed to cold air on the snow surface after being buried in snow. Post-extrication cooling can be more rapid than cooling in the snow because of wind and lower ambient temperatures [bib_ref] Extreme Cooling Rates in Avalanche Victims: Case Report and Narrative Review, Mittermair [/bib_ref]. Moving air on the skin causes more rapid cooling than still air [bib_ref] The Wind-Chill Index, Lankford [/bib_ref]. Cooling rates have not been reported for people buried in houses or vehicles. Anecdotal and statistical evidence suggests that survivors in structures tend to survive longer than buried subjects in open terrain, at least partly because they are not in constant, direct contact with snow [fig_ref] Figure 1: Survival probability for avalanche victims completely buried under the snow in Switzerland... [/fig_ref] [bib_ref] Post-Traumatic Stress Disorder: Supportive Evidence from an Eighteenth Century Natural Disaster, Parry-Jones [/bib_ref] [bib_ref] Field Management of Avalanche Victims, Brugger [/bib_ref]. evidence suggests that survivors in structures tend to survive longer than buried subjects in open terrain, at least partly because they are not in constant, direct contact with snow [fig_ref] Figure 1: Survival probability for avalanche victims completely buried under the snow in Switzerland... [/fig_ref] [bib_ref] Post-Traumatic Stress Disorder: Supportive Evidence from an Eighteenth Century Natural Disaster, Parry-Jones [/bib_ref] [bib_ref] Field Management of Avalanche Victims, Brugger [/bib_ref].
## Rescue and treatment
In disasters, burial by avalanches in buildings or vehicles differs significantly from burial in the open. Victims buried in buildings or vehicles usually lack personal protective avalanche equipment [bib_ref] The Impact of Avalanche Transceivers on Mortality from Avalanche Accidents, Hohlrieder [/bib_ref] [bib_ref] The Impact of Avalanche Rescue Devices on Survival, Brugger [/bib_ref] [bib_ref] The Effectiveness of Avalanche Airbags, Haegeli [/bib_ref] [bib_ref] Respiration during Snow Burial Using an Artificial Air Pocket, Grissom [/bib_ref]. Searching with avalanche transceivers or RECCO ® is not feasible [bib_ref] Soteras, I. Avalanche Survival after Rescue with the RECCO Rescue System: A..., Grasegger [/bib_ref]. For the first several hours or days, the first responders are often local lay rescuers who have survived the avalanche. As soon as professional rescuers arrive at the scene, disaster medicine with triage can begin [bib_ref] Management of Multi-Casualty Incidents in Mountain Rescue: Evidence-Based Guidelines of the International..., Blancher [/bib_ref]. The International Commission for Mountain Emergency Medicine (ICAR MedCom) has published guidelines for rescue and medical care in multi-casualty avalanche accidents [bib_ref] Management of Multi-Casualty Incidents in Mountain Rescue: Evidence-Based Guidelines of the International..., Blancher [/bib_ref]. The number of casualties defining a multi-casualty avalanche may be lower than in an urban environment because of limited resources. The principles of care for multi-casualty incidents (MCIs) in mountain areas are given in . . Principles and recommendations for the management of a multi-casualty incident (MCI) in a mountain area, adapted from [bib_ref] Management of Multi-Casualty Incidents in Mountain Rescue: Evidence-Based Guidelines of the International..., Blancher [/bib_ref].
## General principles and recommendations in mci management
Declare an MCI. An MCI should be identified and the appropriate rescue organisations and hospitals should be alerted as soon as possible. Assess safety. Safety of the rescuers is the highest priority. Rescuers should not access an area if the risk to themselves is considered to be too high. Initial response. The initial response should focus on setting up a command-and-control structure, triage, and rapid life-or limb-saving interventions.
## Rescue and treatment
In disasters, burial by avalanches in buildings or vehicles differs significantly from burial in the open. Victims buried in buildings or vehicles usually lack personal protective avalanche equipment [bib_ref] The Impact of Avalanche Transceivers on Mortality from Avalanche Accidents, Hohlrieder [/bib_ref] [bib_ref] The Impact of Avalanche Rescue Devices on Survival, Brugger [/bib_ref] [bib_ref] The Effectiveness of Avalanche Airbags, Haegeli [/bib_ref] [bib_ref] Respiration during Snow Burial Using an Artificial Air Pocket, Grissom [/bib_ref]. Searching with avalanche transceivers or RECCO ® is not feasible [bib_ref] Soteras, I. Avalanche Survival after Rescue with the RECCO Rescue System: A..., Grasegger [/bib_ref]. For the first several hours or days, the first responders are often local lay rescuers who have survived the avalanche. As soon as professional rescuers arrive at the scene, disaster medicine with triage can begin [bib_ref] Management of Multi-Casualty Incidents in Mountain Rescue: Evidence-Based Guidelines of the International..., Blancher [/bib_ref]. The International Commission for Mountain Emergency Medicine (ICAR MedCom) has published guidelines for rescue and medical care in multi-casualty avalanche accidents [bib_ref] Management of Multi-Casualty Incidents in Mountain Rescue: Evidence-Based Guidelines of the International..., Blancher [/bib_ref]. The number of casualties defining a multi-casualty avalanche may be lower than in an urban environment because of limited resources. The principles of care for multi-casualty incidents (MCIs) in mountain areas are given in .
An avalanche accident should prompt a helicopter rescue, if possible. Helicopters can decrease the response times of and risks to the rescue teams by transporting rescuers safely above potentially hazardous terrain. 'Safety first' should be the guiding principle for rescue operations. The risks to rescue teams must be weighed against the potential benefits for the victims. Avalanche dogs and probing may be valuable tools for timely location of buried subjects [bib_ref] Rescue Missions for Totally Buried Avalanche Victims: Conclusions from 12 Years of..., Hohlrieder [/bib_ref]. Technical rescue may require heavy machinery and equipment. Special precautions may be required when accessing vehicles or entering buried structures. For a mass avalanche incident with insufficient medical resources, a rescue strategy was developed using the Monte Carlo method [bib_ref] A Concept for Optimizing Avalanche Rescue Strategies Using a Monte Carlo Simulation..., Reiweger [/bib_ref].
Once found, the treatment of victims buried in buildings or vehicles is similar to that of victims in open terrain. Unfortunately, no data have been published on methods of accessing victims buried in buildings or in vehicles. An extrication time of 7 min for the first vertical meter of snow burial has been reported in an experimental study, but this is not likely to apply in a real-world situation with hard avalanche debris [bib_ref] Extrication Times During Avalanche Companion Rescue: A Randomized Single-Blinded Manikin Study, Wallner [/bib_ref]. The more deeply a victim is buried, the longer the extrication time, because a larger volume of snow must be removed. . Principles and recommendations for the management of a multi-casualty incident (MCI) in a mountain area, Reprinted with permission form ref. [bib_ref] Management of Multi-Casualty Incidents in Mountain Rescue: Evidence-Based Guidelines of the International..., Blancher [/bib_ref]. Copyright 2022 Copyright Clearance Center.
## General principles and recommendations in mci management
Declare an MCI. An MCI should be identified and the appropriate rescue organisations and hospitals should be alerted as soon as possible. Assess safety. Safety of the rescuers is the highest priority. Rescuers should not access an area if the risk to themselves is considered to be too high. Initial response. The initial response should focus on setting up a command-and-control structure, triage, and rapid life-or limb-saving interventions. Leadership and command. The medical commander should be trained in disaster medicine and in mountain rescue. The medical commander and leaders of the rescue services should all be at the same location on site to optimise cooperation. All should be easily identifiable. Ensuring effective communications. An effective communication structure should be implemented to support command and control. Triage. Effective triage tools adapted to mountain injuries should be implemented. Evacuations. Casualties should be evacuated to a safe area, then transferred to medical facilities appropriate to their medical needs. Identification and tracking. Tools that enable clear identification and tracking of casualties should be available for MCIs in mountains. Learning from experience. MCIs in mountain areas should be analysed afterwards and necessary improvements in prevention and management should be identified and implemented. Planning and training. Standard operating procedures should be available, familiar, and implemented with regular training exercises involving emergency services. After extrication, a victim should be insulated from cold using available materials, such as parkas, wind shells, hats, gloves, bivouac sacks, and sleeping bags, and with vapour barriers such as aluminium blankets. Medical treatment is beyond the scope of this article and is discussed in detail elsewhere [bib_ref] Wilderness Medical Society Practice Guidelines for Prevention and Management of Avalanche and..., Van Tilburg [/bib_ref] [bib_ref] European Resuscitation Council Guidelines 2021: Cardiac Arrest in Special Circumstances, Lott [/bib_ref] [bib_ref] On-Site Treatment of Avalanche Victims ICAR-MEDCOM-Recommendation, Brugger [/bib_ref] ].
## Outcome
In avalanche disasters affecting structures and vehicles, the rate of long-term survival may be higher than in avalanches in open areas, partly because air pockets in structures are likely to be larger than air pockets in the open [fig_ref] Figure 1: Survival probability for avalanche victims completely buried under the snow in Switzerland... [/fig_ref] [bib_ref] Post-Traumatic Stress Disorder: Supportive Evidence from an Eighteenth Century Natural Disaster, Parry-Jones [/bib_ref] [bib_ref] Field Management of Avalanche Victims, Brugger [/bib_ref]. The extrication times may be substantially longer in avalanches affecting structures and vehicles.
## Hypothermia caused by earthquakes and tsunamis
## Key challenges of large earthquakes
Earthquakes with associated tsunamis have killed more people than all other disasters combined. Millions of earthquakes occur every year [bib_ref] Earthquake Disasters-Lessons to be Learned, Peleg [/bib_ref]. Most are small and cause no damage. Over 80% of the fatalities from earthquakes have been in China, Japan, Pakistan, Turkey, countries of the former USSR, Peru, Chile, and Italy. Earthquakes and tsunamis are rapid-onset disasters. From 1994 to 2013, earthquakes were responsible for an estimated 1.35 million deaths and displacement of an estimated 218 million people. In addition to tsunamis, earthquakes can cause snow or ice avalanches, as well as GLOFs. All these disasters have the potential to cause accidental hypothermia in the people who initially survive an earthquake. Most people who suffer from hypothermia in earthquakes associated with tsunamis become hypothermic by being caught in the tsunami. Data regarding hypothermia caused by earthquakes and associated disasters are very limited.
## Irpinia earthquake 1980
One report of accidental hypothermia after an earthquake concerned the 1980 Irpinia earthquake in southern Italy [bib_ref] Cold injuries in children. Experiences of the earthquake of November 1980 (author's..., Risolo [/bib_ref]. On 23 November 1980, this magnitude 6.9 earthquake killed almost 2500 people, injured at least 7700, and left 250,000 people homeless. Nine children were buried for 7 to 27 h before being rescued alive. All were wearing light pyjamas or flannels. The lowest temperature overnight was 7.7 - C. Seven of the nine children were buried and immobilised for longer than 12 h and were mildly hypothermic on arrival to the hospital. All were rewarmed without incident using radiant infant warmers.
## Kashmir earthquake 2005
The magnitude 9.0 South Asian earthquake struck Kashmir on 8 October 2005. The official death toll was over 73,000 in the part of Kashmir administered by Pakistan and almost 1400 in the part administered by India [bib_ref] Profile of Injuries Arising from the 2005 Kashmir Earthquake: The First 72..., Mulvey [/bib_ref]. An estimated 3.3 million people were left homeless. Three months after the quake, most of the people left homeless were still living in tents. Night-time temperatures were below freezing. In mountain villages, temperatures fell to as low as −13 - C. Few of the tents were winterised or designed for arctic-type cold. Relief efforts were thwarted by landslides caused by heavy snow and rain. Supplies and personnel could not be airlifted because of fog. Although there was enough food, there was a shortage of warm clothing and blankets. Many tents collapsed under snow. A news report on 28 November 2005 estimated that more than 100 people had been brought to hospitals with hypothermia and respiratory diseases.
## Great east japan earthquake 2011
Probably the best studied tsunami followed the Great East Japan earthquake on 11 March 2011. There were about 16,000 fatalities. Over 90% were caused by drowning. The ratio of tsunami-related deaths to earthquake-related deaths was 9:1. The ratio of injuries to deaths was 1:3.9. If missing persons are included, the ratio was 1:3.3. This contrasts with a ratio of 1:0.15 injuries to deaths after the Great Hanshin earthquake of 1995. The occurrence of a tsunami increased the number of fatalities. Causes of death were not investigated in any detail. Autopsies were not conducted. As with most natural disasters, corpses were examined primarily to identify the victims. Interviews with medical examiners suggested that diagnoses of 'death by drowning' may have included cases of death by hypothermia and cases in which victims developed hypothermia before drowning. Hypothermia accounted for 0.2% of deaths reported in the three hardest hit prefectures.
Victims of hypothermia caused by earthquakes and tsunamis can be broadly divided into those who developed hypothermia after gradual exposure to cold, as a result of their homes collapsing or damage to infrastructure, and those who were caught in the tsunami. The average temperature during the 24 h after the earthquake was 0.8 - C. The lowest temperature was −1.3 - C. The average wind speed was 5.8 m/s, with a maximum wind speed of 9.7 m/s. During early March, the sea surface temperature is typically between 5.4 - C and 8.3 - C. The likelihood that hypothermia will develop increases as ambient temperature decreases and is increased by convective heat loss caused by wind and evaporative heat loss from wet clothing. The risk of hypothermia increases during immersion in water colder than 18 - C [bib_ref] Survival Time Prediction in Marine Environments, Xu [/bib_ref] [bib_ref] A New Look at Survival Times during Cold Water Immersion, Xu [/bib_ref]. According to a survey of 134 patients in the hardest hit prefecture who received medical care at a hospital within 72 h after the earthquake, hypothermia caused by the tsunami accounted for about three-fourths (45 of 59) of all admissions. Most patients arrived at the hospital within 24 h. The number of hypothermia cases occurring indoors increased with time, especially in victims with underlying conditions and in victims requiring assistance for activities of daily living who were living in their own homes or in shelters after evacuating their homes. Seventy-seven of 91 (85%) patients were cold stressed at presentation (core temperature - C-32 of 91) or mildly hypothermic (core temperature 32-35 - C-45 of 91). The predominance of patients with cold stress or mild hypothermia likely reflects the limited number of victims taken to hospitals during the early phase of the disaster. Land approaches were blocked when roads collapsed and were submerged. Air rescue was the primary means of reaching survivors. Air rescue was limited during the night. In a large-scale disaster, there is very little chance that hypothermic victims without vital signs will be resuscitated or transferred to a hospital [bib_ref] Accidental Hypothermia-an Update: The Content of This Review Is Endorsed by the..., Paal [/bib_ref] [bib_ref] European Resuscitation Council Guidelines 2021: Cardiac Arrest in Special Circumstances, Lott [/bib_ref]. When it is unclear if the cause of cardiac arrest is drowning or hypothermia, it is difficult to identify victims who might be candidates for extracorporeal life support (ECLS) rewarming [bib_ref] Extracorporeal Life Support in Accidental Hypothermia with Cardiac Arrest-A Narrative Review, Swol [/bib_ref]. Even if hypothermic victims are transferred, ECLS treatment may be impossible because of power outages. In the Great East Japan earthquake, about 80% of hospitals and 30% of clinics were damaged. Almost all medical facilities in coastal areas had limited capabilities. Water and power outages, including interruptions of natural gas pipelines, were widespread. It took several days before most medical facilities were able to operate normally.
Only 4 of 91 (4%) hypothermic patients arriving at hospital within 72 h of the earthquake died. The low mortality likely reflects the high proportion of cold stressed and mildly hypothermic patients. Patients without vital signs were not transferred to hospitals and were not represented in this in-hospital study. Hospital treatment is limited when infrastructure is compromised because of limited rewarming capabilities, problems in identifying patients with underlying conditions and injuries, and the difficulties of allocating available resources. Most medical facilities also had little experience treating hypothermic patients.
The situation is different during a small-scale disaster with few casualties, in which the response capabilities of the area are not overwhelmed. A wider range of victims can access medical resources. Victims in hypothermic cardiac arrest might receive medical treatment with ECLS rewarming. The allocation of ECLS rewarming should be guided by outcome stratification using a validated tool, such as the Hypothermia Outcome Prediction after Extracorporeal Life Support (HOPE) score [bib_ref] Hypothermia Outcome Prediction after Extracorporeal Life Support for Hypothermic Cardiac Arrest Patients:..., Pasquier [/bib_ref].
## Hypothermia caused by flooding
Worldwide, floods are the most common disasters. Climate change increases the risk of floods by increasing precipitation [bib_ref] Climate changes, floods, and health consequences. Recent, Michelozzi [/bib_ref]. Areas at special risk are low-lying areas near oceans or below sea level, such as parts of Bangladesh and the Netherlands, along rivers, and in or below mountain ranges in places where water can accumulate. The Netherlands is one of the countries with the highest risk of floods, because almost one-third of the country is below sea level. The Netherlands are protected by a sophisticated system of dams and water management.
In Europe, floods are the most common natural disasters. The most severe European flood of the last 100 years was the 1953 North Sea flood, caused by a combination of wind, high tides, and low atmospheric pressure. It flooded land up to 5.6 m above mean sea level. In the Netherlands, the 1953 North Sea flood claimed 2551 lives, flooded 9% of the Dutch farmland, drowned 30,000 animals, damaged 47,300 structures, and destroyed 10,000 buildings. There were 28 fatalities in Belgium, 307 in England, and 19 in Scotland [bib_ref] The Storm of 31, Hickey [/bib_ref]. Another 230 people died on boats. In the aftermath, England and the Netherlands reinforced their coastal defences with storm surge barriers. England built the Thames and Humber Estuaries. The Netherlands developed the Delta Works. In Italy, 68% of municipalities are at high risk of hydrological disasters. In 2021, heavy rainfall in Belgium, Germany, and the Netherlands again caused severe destruction and disruption.
North American and Asian countries have been struck by floods, often when dams were breached after hurricanes or typhoons. Hurricane Katrina claimed 1464 lives and destroyed vast low-lying sections of New Orleans. The immediate health impacts of floods include drowning and other injuries, heart attacks, and accidental hypothermia. Indirect and delayed effects include displacement, starvation, unemployment, mental health problems (including post-traumatic stress disorder), respiratory diseases, allergic reactions, and water-borne infectious diseases [bib_ref] Climate changes, floods, and health consequences. Recent, Michelozzi [/bib_ref].
In the US, between 2006 and 2010, 2000 US residents died because of weather events. Six percent of deaths were attributed to floods, storms, and lightning [bib_ref] Deaths Attributed to Heat, Cold, and other Weather Events in the United..., Berko [/bib_ref]. In the present climate crisis, rising seawater levels and a higher incidence of extreme precipitation are increasing the likelihood of floods along coastlines and riverbeds.
## Treatment of hypothermia after a natural disaster
Treatment strategies should be established with the assumption that infrastructure will be partly or completely disrupted over a large area. Precautions should be taken so that as few victims as possible become cold stressed or hypothermic. In the Great East Japan earthquake, 500,000 blankets were distributed. Patients with cold stress or mild hypothermia were passively rewarmed with blankets in hospitals. Hospital treatment should be limited initially to hypothermic patients with underlying conditions or associated injuries and to victims with moderate or severe hypothermia. Medical professionals experienced in treating hypothermia should be deployed to evacuation shelters to help victims with mild symptoms self-treat and to identify victims requiring care at medical facilities. Wet clothing should be removed as soon as possible. If a change of clothing is not available [bib_ref] Accidental Hypothermia-an Update: The Content of This Review Is Endorsed by the..., Paal [/bib_ref] [bib_ref] European Resuscitation Council Guidelines 2021: Cardiac Arrest in Special Circumstances, Lott [/bib_ref] , victims should be insulated with blankets [bib_ref] Accidental Hypothermia-an Update: The Content of This Review Is Endorsed by the..., Paal [/bib_ref]. Heating systems will likely not be available at evacuation shelters. Without electricity, forced air warming devices will not function. Chemical or battery-powered heat packs that do not require an external source of electricity or hot water bottles should be stockpiled for active external rewarming. A review of the diagnosis and treatment of accidental hypothermia can be found in this special issue.
## Prevention and mitigation
## Earthquakes
There is usually little or no warning before an earthquake. Earthquakes, tsunamis, and floods that occur at night when people are indoors and sleeping usually cause more injuries and deaths than disasters that occur during the day. The Great Earthquake in Nepal on 25 April 2015 took place at about noon on a Saturday, the one day of the week when schools and businesses are closed in Nepal. Children were not in school. Most people in rural areas were outdoors, working in the fields. About 8000 people died. The toll would have been far higher if the earthquake had occurred on a normal working day or at night. Earthquakes also tend to have a higher mortality in winter, because more people are indoors during the day.
Earthquake early warning systems typically give less than a minute notice of an impending earthquake, but have saved many lives. Education about safety during an earthquake and about avoiding tsunamis is critical to preventing injuries and deaths [fig_ref] Int [/fig_ref]. . Strategies to prevent and mitigate hypothermia after natural disasters.
## Areas susceptible to tsunamis should be identified.
Future housing and critical infrastructure should be built in safe places. Existing structures in areas at risk should be moved or protected.
Early warning and protection systems should cover areas at risk.
Adequate engineering of buildings, roads, and embankments is required.
Education about safety and how to avoid natural disasters should be implemented for citizens in areas at risk.
Temporary shelters and sanitation should be established on elevated land with adequate food and water and should be supplied with rewarming equipment.
Essential non-food items, e.g., clothing and blankets, should be stockpiled.
Global positioning system (GPS)-capable communication systems should be available.
Damage to buildings is a major cause of injuries and deaths. Adequate engineering of buildings, roads, and embankments can prevent injuries and deaths. Needs in humanitarian crises, such as earthquakes and tsunamis, include shelter and security, food and water, and sanitation. To prevent hypothermia, adequate shelter should be prepared in advance. Food and essential non-food items should be stockpiled. Essential non-food items include clothing, bedding, cookware, plates, utensils, and soap. Sanitation is important to prevent diseases that could contribute to hypothermia as well as morbidity or mortality.
## Tsunamis
To prevent hypothermia caused by tsunamis, susceptible areas should be identified. Tsunamis are rare events, occurring about once in a hundred years at any given location. Areas along the Pacific coasts of Canada, Alaska, Russia, and northern Japan, and along the coast of the Japan Sea, including areas of Japan, Russia, and Korea, have conditions in which hypothermia could occur after tsunamis [fig_ref] Figure 2: Areas in which victims may be vulnerable to hypothermia after tsunamis [/fig_ref]
## Floods
Early warning systems should be established for heavy rainfall and storms. Warning systems and coastline protections have been installed in many developed countries, and, to a lesser extent in developing countries, such as Bangladesh. Unfortunately, many early warning and coastline protections are rudimentary, requiring improvements to save more lives. Devastating floods in July 2021 along the Ahr and adjacent rivers in western Germany and eastern Belgium claimed more than 200 lives. The floods caused over 7 billion euros in damages. The Ahr River had previously had a severe flood in 1910 that killed 200 people. In the future, housing and critical infrastructure should be built on high ground well above past flood levels.
# Conclusions
Throughout history, accidental hypothermia has accompanied natural disasters in In the cold areas that are vulnerable to large tsunamis, temporary shelters should be established on elevated land and should be supplied with rewarming equipment and adequate food. It is difficult to raise the temperature inside a large evacuation shelter such as a gymnasium. Evacuees, especially children and people with underlying conditions, may not be able to tolerate long stays. Evacuees should only stay in evacuation shelters for a short period of time before being transferred to accommodations with better infrastructure, such as hotels in unaffected areas. People who are not independent and who are living in their own homes or in nursing facilities are likely to remain where they live, unless an earthquake makes these structures uninhabitable. Food and water should be stockpiled, as should equipment to heat surviving structures if electricity and gas lines are disrupted. To prepare for temporary disruptions to communications, systems, such as global-positioningsystem-enabled mobile phones, should be put in place to help victims summon help.
## Floods
Early warning systems should be established for heavy rainfall and storms. Warning systems and coastline protections have been installed in many developed countries, and, to a lesser extent in developing countries, such as Bangladesh. Unfortunately, many early warning and coastline protections are rudimentary, requiring improvements to save more lives. Devastating floods in July 2021 along the Ahr and adjacent rivers in western Germany and eastern Belgium claimed more than 200 lives. The floods caused over 7 billion euros in damages. The Ahr River had previously had a severe flood in 1910 that killed 200 people. In the future, housing and critical infrastructure should be built on high ground well above past flood levels.
# Conclusions
Throughout history, accidental hypothermia has accompanied natural disasters in cold, temperate, and even subtropical regions. We have explored the causes and means of preventing accidental hypothermia after natural disasters caused by avalanches, earthquakes, tsunamis, and floods. Before a disaster occurs, preventive measures are required, such as accurate disaster risk analysis for given areas, hazard mapping and warning, protecting existing structures within hazard zones to the greatest extent possible, building structures outside hazard zones, and organising rapid and effective rescue. After the event, post hoc analyses of failures and implementation of corrective actions will reduce the risk of accidental hypothermia in future disasters.
[fig] Figure 1: Survival probability for avalanche victims completely buried under the snow in Switzerland from 1981 to 1998 (n = 735) in relation to time (minutes), comparing victims buried in open areas (black curve, n = 638) with victims buried in buildings or vehicles (grey curve, n = 97) [43]. Median extrication times were 37 min in open areas and 56 min from buildings or vehicles (p = 0.17, Mann-Whitney U-test). In open areas, only 17% of all survivors were extricated after 35 min of burial, compared with 33% in buildings and on roads (p = 0.008; Pearson's chi-square). The dotted curve represents the survival function for completely buried avalanche victims in open areas (n = 422) based on the Swiss data for 1981 to 1991, calculated by Falk et al.[44]. [/fig]
[fig] Figure 2: Areas in which victims may be vulnerable to hypothermia after tsunamis. Blue circles are locations in which victims are vulnerable to hypothermia after a tsunami, where monthly average minimum temperatures in 2020 were <0 • C in at least one month. Green shaded areas are source areas of historical earthquakes that caused tsunamis with maximum wave heights > 10 m. Green stars are epicentres of the 1983 Japan Sea and the 1993 Hokkaido Nansei-oki earthquakes (maximum wave heights > 10 m). Red stars are epicentres of the 1896 Sanriku and the 1946 Aleutian earthquakes with tsunamis (maximum wave heights > 30 m). [/fig]
[fig] Author: Contributions: Concept: K.O., Y.T., J.S., K.Z., H.B. and P.P.; methodology and literature search: K.O., Y.T., J.S., K.Z., H.B. and P.P.; writing-original draft preparation: K.O., Y.T., J.S., K.Z., H.B. and P.P.; writing-review and editing: K.O., Y.T., J.S., K.Z., H.B. and P.P.; graphics: K.O., Y.T., J.S., K.Z., H.B. and P.P.; funding acquisition: K.O. and P.P. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. [/fig]
|
Transcriptome sequencing of Crucihimalaya himalaica (Brassicaceae) reveals how Arabidopsis close relative adapt to the Qinghai-Tibet Plateau
The extreme environment of the Qinghai-Tibet Plateau (QTP) provides an ideal natural laboratory for studies on adaptive evolution. Few genome/transcriptome based studies have been conducted on how plants adapt to the environments of QTP compared to numerous studies on vertebrates.Crucihimalaya himalaica is a close relative of Arabidopsis with typical QTP distribution, and is hoped to be a new model system to study speciation and ecological adaptation in extreme environment. In this study, we de novo generated a transcriptome sequence of C. himalaica, with a total of 49,438 unigenes. Compared to five relatives, 10,487 orthogroups were shared by all six species, and 4,286 orthogroups contain putative single copy gene. Further analysis identified 487 extremely significantly positively selected genes (PSGs) in C. himalaica transcriptome. Theses PSGs were enriched in functions related to specific adaptation traits, such as response to radiation, DNA repair, nitrogen metabolism, and stabilization of membrane. These functions are responsible for the adaptation of C. himalaica to the high radiation, soil depletion and low temperature environments on QTP. Our findings indicate that C. himalaica has evolved complex strategies for adapting to the extreme environments on QTP and provide novel insights into genetic mechanisms of highland adaptation in plants.Extreme environments provide an ideal natural laboratory for studies on adaptive evolution. The Qinghai-Tibet Plateau (QTP) is the highest (above 4,000 meters of average elevation) and largest young plateau in the world, which has low temperature, low oxygen, poor soils, and strong ultraviolet (UV) radiation environments. Understanding how organisms adapt to these various extreme environments can make a significant contribution to evolutionary ecology. Recently, research on the adaptive genetic mechanism of non-model organisms based on genome-wide scale can be conducted using the next generation sequencing technology 1 . Previous genome-wide studies on adaptive evolution on this region have focused mainly on humans and vertebrates 2-8 . These studies revealed that genes involved in hypoxia response, energy metabolism, and skeletal development were under positive selection and rapidly evolving. However, little genome/transcriptome-based research has been devoted to plants in this region.Relatives of Arabidopsis thaliana are important model systems to study evolutionary ecology and comparative genomics 9,10 . Several studies have been conducted on the evolution of mating systems in Capsella 11,12 , life history form in A. lyrata and Capsella rubella 13 , as well as adaptation to serpentinite soils in A. lyrata 14 and external climates in A. halleri 15 . Notably, there are other close relatives of Arabidopsis, which not only occupy distinctive niche, but also are of ecological importance. Crucihimalaya himalaica (Edgeworth) Al-Shehbaz has been known as A. rupestris Edgeworth, A. himalaica (Edgeworth) O. E. Schulz, Arabis brevicaulis Jafri, and A. brevicaulis (Jafri) Jafri 16 . This species is a diploid (2n = 16) with a relatively small genome size of 319 Mb 17 . It usually grows in rocky hillsides, sandy slopes, alpine meadows, and screes in areas around Himalayas, e.g. QTP.Previous reviews suggested that Crucihimalaya is a specialized group in Himalayas derived from the Arabidopsis genus in very late geological history18,19. However, recent molecular phylogenetic result showed that Crucihimalaya, Capsella, Arabidopsis and several genera constitute a paraphyly 20 . Dated phylogenies also indicated that the genus Crucihimalaya origin in about 5.2 Mya, and that C. himalaica split from C. lasiocarpa in about 3.56 Mya 20 , which is in accordance with the time of QTP rapid uplift, 3.6 Mya 21 . Conceivably, this species must have undergone significant genetic changes to adapt to stress factors following the rapid uplift of QTP. Therefore, C. himalaica is hoped to be a new model system to study speciation and ecological adaptation in extreme environments.Transcriptome analysis (RNA-seq) provides a rapid and effective approach to obtain massive protein-coding genes, which can be used for understanding ecological, comparative and evolutionary genomics questions for non-model organisms22,23. In this study, we performed RNA-seq to obtain most transcript sequences of C. himalaica. Subsequently, positively selected genes (PSGs) in C. himalaica related to environmental adaptation were identified by comparative genomics with closely related species whose genome have been sequenced. We aim to reveal how this Arabidopsis relative adapts to the complicated extreme environments on QTP at genome/ transcriptome level.
# Results
Transcriptome assembly and annotation. In this study, we generated 132.29 million clean reads and [bib_ref] Generic Placement of Species Excluded from Arabidopsis (Brassicaceae), Al-Shehbaz [/bib_ref].52 Gb of RNA-seq data after quality filtering (Supplementary . The clean data were submitted to the NCBI Sequence Reads Archive (SRA) database (no. SRR3138110). De novo assembly of these high-quality reads generated 66,084 transcripts and 49,438 unigenes (the longest transcript in one gene) . The unigenes are 896 bp on average with an N50 length of 1,641 bp.
There are 29,760 (60.19%) and 33,667 (68.09%) unigenes identified homologs in Nr and Nt databases on the basis of similarity, respectively. A total of 39,189 (79.26%) unigenes were successfully annotated using at least one database (a total of seven databases, see methods) with a significant match (e-value < 10 −5 ) (Supplementary [fig_ref] Table 2: List of PSGs in DNA repair pathway in C [/fig_ref]. The top species classification hits for C. himalaica in the Nr database are Capsella rubella, A. lyrata, and A. thaliana [fig_ref] Figure 1: Species classification [/fig_ref] , which all belong to Brassicaceae. The e-values are very significant, which mostly close to zero (32.1%) and 0∼ 1e-100 (16.5%) [fig_ref] Figure 1: Species classification [/fig_ref] , suggesting that most unigenes of C. himalaica have very similar homologs in above relatives.
Orthogroups identified and positive selection analysis. In this study, a total of 22,572 orthogroups were detected, including 152,769 genes. Among these, 10,487 orthogroups were shared by all six species, and 4,286 orthogroups contained putative one-to-one single copy genes. To further confirm the phylogenetic position of C. himalaica, we generated a phylogenetic tree based on 1,506,379 amino acid sequences of the trimmed and concatenated 4,286 single copy gene alignments from six species [fig_ref] Figure 2: Phylogenetic relationships and dN/dS ratios distribution of C [/fig_ref]. The phylogenic tree showed that C. himalaica was closely related to Capsella rubella, and then clustered with the clade of A. lyrata and A. thaliana. The placement of L. stylosa was outside of these two clades with B. rapa as outgroup.
We also estimated the substitution rates for each orthogroups by free-ratio model in PAML, which allows an independent dN/dS ratio for each branch [bib_ref] PAML 4: phylogenetic analysis by maximum likelihood, Yang [/bib_ref]. For each species, the dN/dS ratio major concentrated in 0.1-0.2 [fig_ref] Figure 2: Phylogenetic relationships and dN/dS ratios distribution of C [/fig_ref]. The median of dN/dS ratio in C. himalaica (0.1873) across all the orthologs was significantly [fig_ref] Table 2: List of PSGs in DNA repair pathway in C [/fig_ref]. Statistics of assembled data. larger than that of other relatives (0.1442-0.1572) [fig_ref] Figure 2: Phylogenetic relationships and dN/dS ratios distribution of C [/fig_ref] , Wilcoxon rank sum test, p < 0.0001). The frequency distribution of dN/dS ratios evidently showed that C. himalaica (1172 genes) has more genes with elevated dN/dS ratios (dN/dS > 0.3) than others (592-815 genes) [fig_ref] Figure 2: Phylogenetic relationships and dN/dS ratios distribution of C [/fig_ref]. For 4,286 single copy orthogroups, the branch-site model of the PAML 4 package 24 was used to detect genes with signals of positive selection. As a result, 1,444 genes possibly under positive selection were identified in the C. himalaica genome (ω > 1); of these genes, 598 showed significant evidence of positive selection (P-value < 0.05), and 487 were undergo extremely significant positive selection (P-value < 0.01) (Supplementary . functional classification for these PSGs (P-value < 0.01) in C. himalaica. The distribution of KEGG classification of PSGs showed that in all annotated categories, metabolic processes have the most hits, such as amino acid metabolism (10 PSGs), cofactors and vitamins metabolism (7 PSGs), lipid metabolism (7 PSGs), and energy metabolism (5 PSGs) [fig_ref] Figure 3: Distribution of KEGG classification of PSGs in C [/fig_ref]. Beyond that, another predominant hit was DNA repair and recombination, which included 21 PSGs. It contains several DNA repair pathways, such as mismatch repair (6 PSGs), nucleotide excision repair (5 PSGs), base excision repair (3 PSGs) and homologous recombination (2 PSGs) (Supplementary .
In addition, GO annotation also showed that many PSGs are related to specific adaptation traits, including nitrogen compound metabolic process (150 PSGs), regulation of nitrogen metabolic (73 PSGs), response to radiation (48 PSGs), DNA repair (21 PSGs), and photoperiodism, flowering (13 PSGs) [fig_ref] Figure 4: GO enrichment of PSGs related to ecological adaptation in C [/fig_ref]. In the following paragraphs, we will discuss the important roles of these PSGs in the adaptation of C. himalaica to the various extreme environments in QTP, respectively.
# Discussion
Our phylogenetic analysis based on genomic level highly supported that C. himalaica was most closed to Capsella rather than Arabidopsis. The relationship among these species agrees with previous study [bib_ref] Dated Molecular Phylogenies Indicate A Miocene Origin For Arabidopsis Thaliana, Beilstein [/bib_ref]. Although previous reviews suggested that Crucihimalaya is a specialized group in Himalayas derived from the Arabidopsis genus. Our result indicates that it should be derived from the common ancestor of the Arabidopsis lineage and Crucihimalaya lineage. In fact, C. himalaica grouped with C. rubella, but not A. thaliana, is conflict with morphological data, especially the principal character for taxonomy (dehiscent siliques, linear fruits in Crucihimalaya & Arabidopsis vs. dehiscent silicles, obdeltoid fruits in Capsella). Maybe that is why Crucihimalaya was classified into Arabidopsis in previous taxonomic system. This conflict between the morphological and molecular result highlights the need for further studies on the intrinsic mechanism of speciation.
Orthologs are genes that have evolved from a common ancestral gene via speciation [bib_ref] Distinguishing homologous from analogous proteins, Fitch [/bib_ref]. To investigate the selective pressures at the branch level in C. himalaica and its relatives, we estimated the substitution rates for each orthogroups. The median of dN/dS ratio in C. himalaica was significantly larger than that of other relatives, which strongly supported the accelerated evolution in C. himalaica after splitting from its ancestor lineage. Furthermore, the frequency distribution of dN/dS ratios evidently showed that C. himalaica has more genes with elevated dN/ dS ratios than other five species. These accelerated evolution of genes is often driven by positive selection or relaxed selection pressure. Combined with results of branch-site model, which revealed that many functional genes in C. himalaica undergo positive selection in extreme environment of QTP, we speculated that the evolutionary rate increased in C. himalaica are due to positive selection rather than relaxation of select.
There are very high light radiation on QTP, especially the UV-B radiation during the summer (approximately 65 kJ/m 2 ) is among the highest worldwide [bib_ref] Ground-based measurements and modeling of solar UV-B radiation in Lhasa, Norsang [/bib_ref]. The light radiation can influence plant growth and development, such as photoperiodism, flowering, and DNA damage. Our results showed that as many as 43 PSGs in C. himalaica were annotated as response to light stimulus according to GO category, among which, 13 PSGs in the C. himalaica genome were involved in photoperiodism, flowering process [fig_ref] Figure 4: GO enrichment of PSGs related to ecological adaptation in C [/fig_ref]. Among these, the gene encoding early flowering 6 protein (lysine-specific demethylase, ELF6) is a repressor in the photoperiodic flowering pathway and its loss-of-function mutation causes early flowering [bib_ref] Modulation of brassinosteroid-regulated gene expression by jumonji domain-containing proteins ELF6 and REF6..., Yu [/bib_ref]. Similarly, another gene encoding sensitive to freezing 6 protein (mediator of RNA polymerase II transcription subunit 16), plays an important role in regulation of the circadian clock and in the control of flowering time [bib_ref] Sensitive to freezing6 integrates cellular and environmental inputs to the plant circadian..., Knight [/bib_ref]. In the short growing season condition of QTP, flowering time is particularly critical and affect both the life cycles and reproductive success of alpine flora [bib_ref] Flowering and Fruiting Phenology of 24 Plant Species on the North Slope..., Zhang [/bib_ref] [bib_ref] Flowering phenology in the central highland of Iceland and implications for climatic..., Thórhallsdóttir [/bib_ref]. C. himalaica starts flowering relatively early (April), and has a very long duration of the flowering period (from April to September). This pattern suggests that C. himalaica has evolved specific reproductive strategies via positively selected more than a dozen genes associated with photoperiodism and flowering process as adaptation to the short growing season in the harsh environments of QTP.
In addition, highly energetic UV radiation causes direct damage to DNA, RNA, and proteins [bib_ref] Ultraviolet-B Radiation-Mediated Responses in Plants. Balancing Damage and Protection, Frohnmeyer [/bib_ref]. Our results showed that 25 PSGs were involved in response to DNA damage stimulus [fig_ref] Figure 4: GO enrichment of PSGs related to ecological adaptation in C [/fig_ref]. This result was consistent with 21 PSGs enriched in DNA repair pathway [fig_ref] Figure 2: Phylogenetic relationships and dN/dS ratios distribution of C [/fig_ref]. As an essential system for correcting UV-induced DNA damage, nucleotide excision repair (NER) play more important roles in UV-resistance for plants [bib_ref] Understanding DNA repair and recombination in higher plant genome: information from genome-wide..., Singh [/bib_ref]. It is worth to mention that several essential genes participating in NER process showed positive selection. Such as gene encoding DNA excision repair protein ERCC-1 (or ultraviolet hypersensitive 7), which has been reported to function as a part of a structure-specific endonuclease that cleaves 5′ to UV photoproducts in DNA. And gene encoding DNA damage-binding protein 2, which forms a complex with DDB1 to recognize damaged DNA and initiation of NER process after exposure to UV light [bib_ref] The DDB1a interacting proteins ATCSA-1 and DDB2 are critical factors for UV-B..., Biedermann [/bib_ref]. Another gene, encoding TFIIH protein, is an essential transcription initiation factor that is also pivotal for NER [bib_ref] Slowly progressing nucleotide excision repair in trichothiodystrophy group A patient fibroblasts, Theil [/bib_ref]. Besides NER, other DNA repair mechanism, such as base excision repair (BER), mismatch excision repair (MMR) and homologous recombination (HR) also participated in response to light radiation and other external environment stimulation in QTP [fig_ref] Table 2: List of PSGs in DNA repair pathway in C [/fig_ref]. Accordingly, we found PSGs encoding DNA repair protein XRCC4, DNA mismatch repair protein MUTS2, MUTS protein homolog 2/5 (MSH2/5), MUTL protein homolog 3 (MLH3), DNA repair and recombination protein RAD54, and DNA damage repair/toleration protein DRT111 . KEGG enrichment showed that these genes covered almost all aspects of the DNA repair mechanism (Supplementary , suggesting that C. himalaica has evolved an integrated DNA repair mechanism to adapt to the harsh habitat following the uplift of QTP.
Nitrogen compound metabolism is the basic physiological processes, which can generate components of cells (e.g. nucleic acids and proteins) as well as convert sugars and proteins into energy. Moreover, nitrogen compound metabolism is even more important for C. himalaica, because nutrient deficiency, especially nitrogen deficiency is a typical feature of the sandy soil habitats for its living [bib_ref] Soil nutrient availability determines the facilitative effects of cushion plants on other..., Chen [/bib_ref] [bib_ref] Plants Adapt to Long-Term Potassium Deficiency by Accumulation of Membrane Lipids in..., Wang [/bib_ref]. It means nitrogen assimilation under low-nitrogen condition is also a challenge for C. himalaica living in QTP. From our results, it was evident that as many as 150 PSGs involved in nitrogen compound metabolism in C. himalaica, which included GO categories associated with DNA metabolic process (41genes) and regulation of nitrogen compound metabolic process (73 genes) [fig_ref] Figure 4: GO enrichment of PSGs related to ecological adaptation in C [/fig_ref] , Supplementary . The latter included many transcription factors involved in the nitrogen assimilation process, including the primary assimilation of ammonia to carbon skeletons to biosynthesize amino acids and other organic compounds. Remarkably, three members of DOF (DNA BINDING WITH ONE FINGER) family, a class of zinc finger domain TFs, showed significant positive selection. DOF factors could play an important role in nitrogen regulation, which was supported by the fact that Dof1 over-expressed in Arabidopsis promote the nitrogen assimilation, thus improved plant growth under low-nitrogen conditions [bib_ref] Metabolic engineering with Dof1 transcription factor in plants: Improved nitrogen assimilation and..., Shuichi [/bib_ref]. Moreover, a PSG encoding GATA zinc finger protein homolog NTL10, also has been reported activates the expression of nitrogen-catabolic enzymes during conditions of nitrogen limitation [bib_ref] A tobacco cDNA clone encoding a GATA-1 zinc finger protein homologous to..., Daniel-Vedele [/bib_ref]. The functions of these PSGs suggested that C. himalaica may have adaptively sped up evolutionary rates of genes involved in nitrogen metabolic regulation to better adapt to the nitrogen limitation in its habitat, which seems to be the driving force behind rapid evolution of these genes.
Low temperatures and rapid changes of temperature are also prevalent features of QTP. Organisms that live in this alpine ecosystem face various growth-related challenges from low temperatures, such as reduced fluidity of lipid membranes [bib_ref] Cold-adapted archaea, Cavicchioli [/bib_ref]. The lipid bilayers transmit external signals to its interior as well as protect the integrality of various organelle membranes in unfavorable environments. Our GO category results indicated that 22 genes related to lipid biosynthesis were under significant positive selection, including 14 PSGs genes involved in phospholipid biosynthetic process and 5 PSGs in glycerolipid biosynthetic process . Among these, gene encoding lipid-A-disaccharide synthase with function of condense the lipid A, could play a structural role by stabilizing the plasma membrane lipid bilayer in plants, and involved in signal transduction or plant defense responses [bib_ref] Pathway for lipid A biosynthesis in Arabidopsis thaliana resembling that of Escherichia..., Chijun [/bib_ref]. Other PSGs, such as gene encoding Phosphatidyl-N-methylethanolamine N-methyltransferase, was reported to produce the abundant membrane lipid phosphatidylcholine [bib_ref] Functional Characterization of Phospholipid N -Methyltransferases from Arabidopsis and Soybean, Keogh [/bib_ref]. Previous findings also showed that the degree of unsaturated lipids remains unchanged in C. himalaica under rapid temperature changes [bib_ref] Plant adaptation to frequent alterations between high and low temperatures: remodelling of..., Zheng [/bib_ref] , suggesting that the remodeling of membrane lipids might protect membranes against frequent temperature changes. Based on above results, we speculate that these membranes lipids and proteins might play major roles in maintaining membrane integrity under low temperatures and rapid changes of temperature conditions.
# Conclusion
Organisms that live in QTP face a variety of external stress from their harsh environments. As such, it is often believed that these organisms have undergone a series of adaptive evolutionary changes. Low oxygen was reported as the most challenge for animals living in high-altitude, thus several genes response to hypoxia showed signature of positive selection in different species [bib_ref] Genomic insights into adaptation to high-altitude environments, Cheviron [/bib_ref] [bib_ref] The yak genome and adaptation to life at high altitude, Qiu [/bib_ref] [bib_ref] Ground tit genome reveals avian adaptation to living at high altitudes in..., Qu [/bib_ref] [bib_ref] Genomic analyses identify distinct patterns of selection in domesticated pigs and Tibetan..., Li [/bib_ref] [bib_ref] Draft genome sequence of the Tibetan antelope, Ge [/bib_ref] [bib_ref] Whole-genome sequencing of six dog breeds from continuous altitudes reveals adaptation to..., Gou [/bib_ref] [bib_ref] Population variation revealed high-altitude adaptation of Tibetan mastiffs, Li [/bib_ref]. Unlike animals, plants as sessile organisms are constantly exposed to these stresses in QTP. It is conceivable that they have developed more sophisticate mechanisms to protect themselves. Our results of transcriptome sequence of C. himalaica and evolutionary genomic analysis provided evidence for this belief. In this study, we identified 487 significantly PSGs in transcriptome sequences of C. himalaica. Functions of these PSGs elucidate that they potentially possess specific traits of adaptive significance, such as response to radiation, DNA repair, membrane stabilization and organic metabolism. These functions are likely responsible for the adaptation of C. himalaica to the high radiation, low temperature and soil depletion environments on QTP. Due to the RNA-seq technology is difficult to obtain the entire and full length transcripts, further studies are required. Nevertheless, our findings indicate that sophisticated genetic mechanisms have evolved in C. himalaica to survive the harsh conditions in QTP.
# Materials and methods
Sample collection and transcriptome sequencing. The original seedlings of C. himalaica were sampled in August 24, 2013 in Xiangcheng County (alt. 4,104 m, 29°02′ 47″ N, 99°42′ 55″ E) of QTP. These seedlings from the same individual were cultivated in greenhouse in Kunming Institute of Botany. In order to obtain more expression transcripts, two developmental stages (15 days, 30 days) of different organs (leaves, stems, and roots) were sampled and stored at − 80 °C.
High quality total RNA was extracted using the TRIZOL reagent (Sigma Aldrich) following the manufacturer's instructions. A total amount of 3 μ g RNA per developmental stage was used as input material for the RNA sample preparations. Sequencing libraries were generated using NEBNext Ultra ™ RNA Library Prep Kit for Illumina Inc. (NEB, USA) following manufacturer's recommendations. The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumina Inc.). After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2500 platform and paired-end reads were generated. The whole step of library construction and Illumina sequencing was performed at Novogene Bioinformatics Technology Co., Ltd (Beijing, China).
De novo assembly and functional annotation. Clean data (clean reads) were obtained by removing reads containing adapter, reads containing ploy-N and low quality reads from raw data. At the same time, Q20, Q30, GC-content and sequence duplication level of the clean data were calculated. The sequenced left files (read1 files) from two samples were pooled into one big left.fq file, and right files (read2 files) into one big right.fq file. Transcriptome assembly was accomplished based on the left.fq and right.fq using Trinity program (trinityrnaseq_ r20140413) with minimum k-mer coverage set to 2 and all other parameters set by default [bib_ref] Full-length transcriptome assembly from RNA-Seq data without a reference genome, Grabherr [/bib_ref].
Functional annotations of all assembled unigenes were conducted by searching against the following databases: NCBI non-redundant protein (Nr), NCBI non-redundant nucleotide (Nt), Protein family (Pfam), Clusters of Orthologous Groups of proteins (KOG), Swiss-Prot protein (Swiss-Prot), KEGG Ortholog database (KO), and Gene Ontology database (GO).
## Orthologous genes identified and phylogenetic analysis. based on previous studies in
Brassicaceae [bib_ref] Arabidopsis thaliana and its wild relatives: a model system for ecology and..., Mitchell-Olds [/bib_ref] [bib_ref] Dated Molecular Phylogenies Indicate A Miocene Origin For Arabidopsis Thaliana, Beilstein [/bib_ref] , we selected genomes of five relatives (A. thaliana, A. lyrata, Capsella rubella, Leavenworthia alabamica and Brassica rapa) and C. himalaica to identify orthologs. Furthermore, to define a set of conserved genes for cross-taxa comparison, we used OrthoMCL software [bib_ref] OrthoMCL: identification of ortholog groups for eukaryotic genomes, Li [/bib_ref] to identify homologous gene clusters (orthogroups) among the six genomes. Genes with lengths less than 50 amino acids were excluded. OrthoMCL was run with an e-value cut-off of 1e-15 and an inflation parameter of 2.0 due to the close genetic relationship between six relatives. Orthologroups with only single copy genes (one-to-one orthologs) that were shared by all six genomes were retained for further analysis. Each orthogroups was aligned using MUSCLE v3.8.31 [bib_ref] MUSCLE: multiple sequence alignment with high accuracy and high throughput, Edgar [/bib_ref] with default parameters. The poorly aligned regions were further strictly trimmed by using the trimAl v1.4 software 47 with the parameter "-gt 0.8 -st 0.001". Alignments of all orthogroups were concatenated by our python script. Then maximum likelihood (ML) trees were generated using RAxML v7.0.4 48 with PROTCATJTT model, the maximum likelihood criteria.
Positive selection analysis. In the positive selection analysis, only above single copy genes were considered.
To calculate the nonsynonymous (Ka) and synonymous (Ks) substitution rates between pairs of orthogroups, above amino acid alignments were reverse-translated to the corresponding codon-based nucleotide alignments by PAL2NAL [bib_ref] PAL2NAL: robust conversion of protein sequence alignments into the corresponding codon alignments, Suyama [/bib_ref]. For each alignment, a gene tree was constructed by RAxML 48 using GTR + GAMMA model.
To estimate lineage-specific evolutionary rates for each branch of the six species, the codeml program in the PAML 4 package 24 with the free-ratio model (model = 1) was run on each orthogroups. We conducted the boxplot analysis using the dN/dS ratio derived from free-ratio model results and filtered dS > 3 or dN/dS > 3. Significances of the deviations from the median dN/dS ratio between six species branches were detected using Wilcoxon rank sum test. We also established frequency distribution plots of all dN/dS ratios of six species.
To increase the power of our tests for positive selection, we applied the improved branch-site model [bib_ref] Evaluation of an improved branch-site likelihood method for detecting positive selection at..., Zhang [/bib_ref] implemented in codeml program [bib_ref] PAML 4: phylogenetic analysis by maximum likelihood, Yang [/bib_ref] to estimate the dN/dS substitution rates (ω value). We also deleted all gaps (clean data = 1) from the alignments to lower the effect of ambiguous bases on the inference of positive selection. A foreground branch was specified as the clade of C. himalaica. A significant likelihood ratio test (LRT) was conducted to determine whether positive selection is operating in the foreground branch. In this study, the extremely significant positively selected genes (PSGs) were inferred if the P-value was less than 0.01.
For each PSG in C. himalaica, functional information was inferred based on its ortholog in A. thaliana. Gene Ontology (GO) enrichment analyses of PSGs were conducted using web-based agriGO (bioinfo.cau.edu. cn/agriGO) [bib_ref] agriGO: a GO analysis toolkit for the agricultural community, Du [/bib_ref] with singular enrichment analysis (SEA) method and TAIR10 database. The KOBAS software [bib_ref] KOBAS 2.0: a web server for annotation and identification of enriched pathways..., Xie [/bib_ref] was also used to test the statistical enrichment of PSGs in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways [bib_ref] KEGG: kyoto encyclopedia of genes and genomes, Kanehisa [/bib_ref].
[fig] Figure 1: Species classification (A) and e-value distribution (B) of the unigenes of C. himalaica annotated to NCBI Nr database. [/fig]
[fig] Figure 2: Phylogenetic relationships and dN/dS ratios distribution of C. himalaica and its relatives. (A) Phylogenetic tree derived from concatenated all orthologs (1,506,379 amino acids) of six species. (B) Boxplots of dN/dS ratios for each species. The median dN/dS ratio and significances of the deviations using Wilcoxon rank sum test are also showed in the boxplots. (C) Number of orthologs with given dN/dS ratios for the six species. [/fig]
[fig] Figure 3: Distribution of KEGG classification of PSGs in C. himalaica. (A) Cellular Processes; (B) Environmental Information Processing; (C) Genetic Information Processing; (D) Metabolism; (E) Organismal Systems.We also conducted KEGG(Fig. 3, Supplementary Table S4) and GO(Fig. 4, Supplementary [/fig]
[fig] Figure 4: GO enrichment of PSGs related to ecological adaptation in C. himalaica. The Arabic numbers show the enriched number of PSGs in each term. [/fig]
[table] Table 2: List of PSGs in DNA repair pathway in C. himalaica. [/table]
|
Serum hyaluronic acid levels in cancer.
[bib_ref] Immunoenzymoassay of the hyaluronic acid-hyaluronectin interaction. Application to the detection of hyaluronic..., Delpech [/bib_ref] [bib_ref] The properties and turnover of hyaluronan, Laurent [/bib_ref] [bib_ref] Immunoenzymoassay of the hyaluronic acid-hyaluronectin interaction. Application to the detection of hyaluronic..., Delpech [/bib_ref] [bib_ref] Immunoenzymoassay of the hyaluronic acid-hyaluronectin interaction. Application to the detection of hyaluronic..., Delpech [/bib_ref] [bib_ref] Serum hyaluronate in malignant pleural mesothelioma, Frebourg [/bib_ref] [bib_ref] Serum hyaluronic acid in patients with disseminated neoplasm, Manley [/bib_ref] [bib_ref] Determination of hyaluronate in biological samples by a specific radioassay, Laurent [/bib_ref] [fig_ref] Figure I: Distribution of serum HA levels in advanced cancer [/fig_ref] [fig_ref] Figure I: Distribution of serum HA levels in advanced cancer [/fig_ref] [fig_ref] Table I: Serum HA levels in Wilms' tumour and neuroblastoma [/fig_ref] [bib_ref] Serum hyaluronic acid in patients with disseminated neoplasm, Manley [/bib_ref] [bib_ref] Hyperviscosity caused by hyaluronic acid in the serum in a case of..., Wu [/bib_ref] [bib_ref] Platelet dysfunction associated with Wilms' tumour and hyaluronic acid, Bracey [/bib_ref] [bib_ref] Platelet dysfunction associated with Wilms' tumour and hyaluronic acid, Bracey [/bib_ref]
[fig] Figure I: Distribution of serum HA levels in advanced cancer. [/fig]
[fig] Figure 2: Distribution of immunoreactive HA in serum from a patient with Wilms' tumour (serum HA > 20,000 ng 1-l) fractionated on a Superose 6 gel filtration column. (200 ,l serum applied, flow rate 0.5mlmin-1): indicating the majority of the HA is in the exclusion volume (exclusion limit > 106 D), markers at 440 and 25kD. [/fig]
[table] Table I: Serum HA levels in Wilms' tumour and neuroblastoma [/table]
|
The effectiveness of a nurse‐led intervention to support family caregivers in end‐of‐life care: Study protocol for a cluster randomized controlled trial
This is an open access article under the terms of the Creat ive Commo ns Attri bution-NonCo mmerc ial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.AbstractAim: To evaluate the feasibility of a structured nurse-led supportive intervention and its effects on family caregivers in end-of-life care at home.
## | introduc ti on
Patients in the last phase of life mostly want to be cared for and die at home [bib_ref] Priorities for treatment, care and information if faced with serious illness: A..., Higginson [/bib_ref]. Without the help of family caregivers that would be impossible for many patients.
In the broad definition of, family caregivers are defined as: 'individuals who provide any physical, emotional and instrumental support or assistance to individuals with a life-limiting illness that they view as family members'. Family caregivers are often intensively involved with personal and emotional care, and with household tasks and the coordination of care. For instance, in the United Kingdom, family caregivers provide care a median of 69 hr each week in the final three months of life [bib_ref] The contributions of family care-givers at end of life: A national post-bereavement..., Rowland [/bib_ref].
The important work that family caregivers provide may result in both positive and negative experiences. Nurses are in a unique position in primary health care to support family caregivers. However, little is known about which nursing interventions are effective on family caregivers' well-being [bib_ref] Nursing interventions to support family caregivers in end-of-life care at home: A..., Becqué [/bib_ref] [bib_ref] A systematic review of psychosocial interventions for family carers of palliative care..., Hudson [/bib_ref].
## | backg rou n d
Caring can be rewarding and meaningful [bib_ref] Burden for family carers at the end of life; a mixed-method study..., De Korte-Verhoef [/bib_ref] [bib_ref] White Paper on improving support for family carers in palliative care: Part..., Payne [/bib_ref]. An Italian study [bib_ref] End-of-life care in Italy: Personal experience of family caregivers. A content analysis..., Morasso [/bib_ref] reported that family caregivers felt personal growth, sense of self-efficacy and improved family relationships as positive aspects of caring.
However, many family caregivers who provide end-of-life care also experience a heavy to severe burden due to the physical, emotional and financial responsibilities associated with caregiving (De [bib_ref] Burden for family carers at the end of life; a mixed-method study..., De Korte-Verhoef [/bib_ref]. They have to cope with the impending loss of a family member and providing care in itself can be a source of stress.
Approximately 25% of family caregivers experience emotional suffering related to the patient's death or their caregiving experience [bib_ref] End-of-life care in Italy: Personal experience of family caregivers. A content analysis..., Morasso [/bib_ref]. Burden of family caregivers may be a key cause for (acute) hospital admissions of the patient [bib_ref] Dying at home: Community nurses' views on the impact of informal carers..., Jack [/bib_ref]. Family caregivers who feel less burdened have been shown to be more able to provide end-of-life care at home until the time of death [bib_ref] Dying at home: Community nurses' views on the impact of informal carers..., Jack [/bib_ref]. Some family caregivers feel insufficiently prepared for the many demands they might face and experienceand feel inadequately supported by the healthcare staff [bib_ref] End-of-life care in Italy: Personal experience of family caregivers. A content analysis..., Morasso [/bib_ref]. Healthcare professionals such as nurses usually focus on the patients and their needs for care and treatment and involve family caregivers only when needed in patient care [bib_ref] How to deal with relatives of patients dying in the hospital? Qualitative..., Witkamp [/bib_ref].
Ideally however, the position of the caregiver should not only been seen as 'co-caregiver' but also as 'co-client' [bib_ref] What is the best way to help caregivers in cancer and palliative..., Harding [/bib_ref].
Nurses have a unique position to assess the needs of family caregivers and to provide supportive interventions aimed at reducing caregivers' burden and improving preparedness [bib_ref] District nursing support for patients with cancer requiring palliative care, Kennedy [/bib_ref]. In the Netherlands, since 2015, nurses in the primary healthcare setting are responsible for the decision which and how much home care is needed and the allocation of nursing home care. This responsibility is laid down in the Healthcare Insurance Act. Together with the patient and their family, the nurse defines goals and plans for care, organizes care and evaluates processes and outcomes of care. Family caregiver needs may be included in this process [bib_ref] White Paper on improving support for family carers in palliative care: Part..., Payne [/bib_ref] , but it is unknown how and to what extent nurses allocate supportive care for caregivers. Additionally, few nursing interventions are available to support family caregivers in end-of-life care at home [bib_ref] Nursing interventions to support family caregivers in end-of-life care at home: A..., Becqué [/bib_ref] [bib_ref] A systematic review of psychosocial interventions for family carers of palliative care..., Hudson [/bib_ref].
## | the s tudy
## | aim
This study aims to evaluate the effects of a structured nurse-led supportive intervention on family caregivers in end-of-life care at home and to evaluate the feasibility of this intervention.
## | design
We will conduct a cluster randomized controlled trial. In this trial, we will evaluate the effects of a new nurse-led supportive intervention on the well-being of family caregivers at home and the feasibility of this intervention. This trial was registered at the Netherlands Trial Register (NL7702). The full form can be accessed online at https:// www.trial regis ter.nl/trial/ 7702.
## | setting
Twelve home care services in the southwest region of the Netherlands were invited to participate as research clusters. They will be randomly assigned to the intervention group or the control group by using a random number generator on the computer.
## | study population
The study population consists of family caregivers of terminally ill patients (e.g. people with advanced cancer or advanced organ failure) receiving home care. We define the family caregiver as 'family member or friend who is mostly involved with the care, or the person who is the first contact person of the patient'. Participating patients can have one or more family caregivers.
## | inclusion and exclusion criteria
Family caregivers caring for patients with a life expectancy up to six months will be included. Nurses will identify these patients by answering 'no' to the adapted surprise question: 'Would you be surprised if this patient would die within six months?' [bib_ref] How robust is the 'surprise question' in predicting short-term mortality risk in..., Gane [/bib_ref] [bib_ref] Predicting one-year mortality in heart failure using the 'Surprise Question': A prospective..., Straw [/bib_ref]. Family caregivers of patients who are estimated (by the nurse or general practitioner) to have a life expectancy of at most 2 weeks will be excluded. Other inclusion criteria are that the family caregiver must be 18 years or older, able to provide written informed consent and able to complete a Dutch questionnaire.
## | intervention
To identify effective nursing interventions on family caregiver support and prevention of burden, we conducted a systematic narrative review [bib_ref] Nursing interventions to support family caregivers in end-of-life care at home: A..., Becqué [/bib_ref]. We identified four intervention components: psychoeducation, needs-assessment, practical support with caregiving and peer support. This review showed that multicomponent nursing interventions were the most successful, whereas needs-assessment seems to be one of the most effective single components [bib_ref] Nursing interventions to support family caregivers in end-of-life care at home: A..., Becqué [/bib_ref].
One of the studies included in the review found effects of an intervention with a needs-assessment [bib_ref] The impact of the carer support needs assessment tool (CSNAT) in community..., Aoun [/bib_ref] , the Carer Support Needs Assessment Tool (CSNAT) which is a valid tool to measure the support needs of family caregivers in palliative home care. [bib_ref] The impact of the carer support needs assessment tool (CSNAT) in community..., Aoun [/bib_ref] [bib_ref] Developing a person-centred approach to carer assessment and support, Ewing [/bib_ref]. Our intervention is based on this CSNAT. The CSNAT comprises 14 domains where family caregivers commonly say they require support to: (a) enable them to care for the patient at home; or (b) increase their own health and well-being.
Completion of the CSNAT is the start of a process, consisting of five steps. First, the CSNAT tool is introduced to the family caregiver (step 1). Then the family caregiver uses the tool to identify domains where they need more support (step 2). A conversation between the family caregiver and the nurse determines needs and priorities (step 3) and this will lead to a shared action plan (step 4). Finally, family caregivers' needs will be regularly reviewed (step 5) [bib_ref] Development of a carer support needs assessment tool (CSNAT) for end-of-life care..., Ewing [/bib_ref]. The CSNAT is thus integrated in a person-centred process, led by the family caregiver, the so-called 'CSNAT-Approach'.
The CSNAT tool was translated into the Dutch language following scientific translation standards and subsequently validated for palliative care practice in the Netherlands .
## | training
Nurses in the intervention group will be trained to use the CSNAT approach. The training is built on the CSNAT training programme developed in Cambridge, with two added themes: (a) how to address varying support needs of family caregivers and (b) clinical reasoning skills. Nurses will learn how they can justify the indicated end-of-life care and support using nursing diagnoses.
The training programme consists of an e-learning, two plenary group sessions and two intervision sessions (totally 23 hr). The e-learning focuses on raising awareness of different types of family caregivers with their specific needs. Information, reflective questions and films are used in this e-learning. In the group sessions the nurses are trained to use the CSNAT approach including clinical reasoning, mainly by role plays. In the intervision sessions nurses reflect on how they apply their new skills in clinical practice. The nurses working at home care organizations in the control group will not receive the training; they will provide care as usual.
## | primary outcomes
## | burden
The primary outcome in this study is burden measured by the Self-Rated Burden Scale (SRB), published in Dutch [bib_ref] Instruments for assessing the burden of informal caregiving for stroke patients in..., Exel [/bib_ref]. The SRB is a feasible and valid instrument for assessing the burden of informal caregiving. It was developed and evaluated among informal caregivers of stroke patients and consists of a single question: 'how burdensome do you feel caring for or accompanying your partner/family member is at the moment?'.
## Ta b l e 1 overview of the study outcomes, measurement instruments and timepoints
## Outcomes
## Measurement instruments timepoints
## Primary outcome
Caregivers' burden Self-Rated Burden Scale (SRB), 1 item, visual analogue scale 0-100 [bib_ref] Instruments for assessing the burden of informal caregiving for stroke patients in..., Exel [/bib_ref] T0 (at baseline) and T1 (one month after T0, postintervention)
## Secondary outcomes
Caregivers' burden Self-Rated Burden Scale (SRB), 1 item, visual analogue scale 0-100 [bib_ref] Instruments for assessing the burden of informal caregiving for stroke patients in..., Exel [/bib_ref] Caregiver Reaction Assessment (CRA), 24 items, 5-point Likert scale (1-5). [bib_ref] The caregiver reaction assessment (CRA) for caregivers to persons with chronic physical..., Given [/bib_ref] [bib_ref] Measuring both negative and positive reactions to giving care to cancer patients:..., Nijboer [/bib_ref] T1-T3 T0-T3
Caregivers' preparedness for caregiving Preparedness for Caregiving Scale (PCS), 8 items, 5-point Likert scale (0-4) [bib_ref] Measuring the psychosocial characteristics of family caregivers of palliative care patients: psychometric..., Hudson [/bib_ref] T0-T3
The incidence of acute admissions of the patient Nursing files 4-6 weeks following patients' death or hospital/hospice admission Responses can be marked on a visual analogue scale ranging from 0 ('not at all straining') to 100 ('much too straining'). shows a summary of the study outcomes, measurement instruments and timepoints.
## | secondary outcomes
## | caregiver reaction
The caregiver reaction assessment (CRA) will be used to further specify the nature of burden. With the CRA the negative and positive response of family caregivers caring for a person in the final phase of life will be assessed. The CRA is a self-report questionnaire, consisting of 24 items and comprising four negative subscales (disrupted schedule, financial problems, lack of family support, health problems) and one positive subscale (self-esteem) [bib_ref] The caregiver reaction assessment (CRA) for caregivers to persons with chronic physical..., Given [/bib_ref]. Answers will be scored on a 5-point Likert scale ranging from 1 to 5. For each subscale, a total mean score will be calculated with a range between 1 and 5. The higher the score how stronger the impact of the attribute (either negative or positive).
The CRA was developed and tested in the study conducted by [bib_ref] The caregiver reaction assessment (CRA) for caregivers to persons with chronic physical..., Given [/bib_ref] This study showed that the CRA is a feasible, reliable and valid instrument to assess the reactions of giving care.
## | preparedness for caregiving
Preparedness for caregiving is defined as perceived readiness for multiple domains of caregiving role, such as providing physical care, providing emotional care, dealing with stress of caregiving and responding to/handling emergency situations [bib_ref] Mutuality and preparedness as predictors of caregiver role strain, Archbold [/bib_ref]. Preparedness in these domains will be measured by the Preparedness for Caregiving Scale (PCS) [bib_ref] Mutuality and preparedness as predictors of caregiver role strain, Archbold [/bib_ref]. The PCS is a self-rated questionnaire consisting of eight items, investigating how well-prepared family caregivers believe they are for multiple domains of caregiving.
Answers are rated on a 5-point scale ranging from 0 ('not at all prepared') -4 ('very well prepared'). The scale is scored by calculating the mean of all items answered with a score range of 0-4.
A higher score represents better preparedness. The psychometric
properties of the PCS have been tested in caregivers of patients in palliative care and demonstrated it is a valid and reliable instrument [bib_ref] Psychometric properties of the caregiver preparedness scale in caregivers of stroke survivors, Pucciarelli [/bib_ref]. We (two researchers) translated the original English version of the PCS into Dutch independently and it was translated backward by two native English speakers.
Differences in the forward and the backward translation were discussed until consensus was reached. The pre-final version was submitted to an advisory panel of family caregivers for feedback; they provided minor comments which were addressed in the final version.
## | acute admissions
Acute admissions and place of death will be retrieved from patients' healthcare records. We define acute admission as an unplanned admission (known less than 36 hr in advance). Halfway and at the end of our study, the researcher will check patient records and check the occurrence of acute admissions and, if applicable, place of death.
## | feasibility
To investigate the feasibility of the intervention and its adaptation to home care practice, seven nurses of the intervention group will be interviewed. The researcher will check in patient records (in the intervention group) what is documented by the nurse about the nursing process: family caregivers' supportive needs, nursing diagnoses and supportive interventions. Data will be collected by using a data extraction form. Furthermore, family caregiver participants will be asked to send the CSNAT forms to the researcher to investigate how the CSNAT is used and which items are completed.
## | recruitment, consent and data collection
Family caregivers will be approached by the nurses for participation in the study. If they are interested, the nurse will inform the researcher and pass on their contact details with permission of the family caregivers. The researcher will contact the family caregiver and provide oral information on the study purpose, the intervention. Family caregivers who agree to participate are asked to provide written consent. We also ask patients to provide written consent to check their records. Participants may withdraw from the study at any time without effect to their care. They will not receive any incentive for participation.
After informed consent, family caregivers in both the intervention and control group will be invited to complete the questionnaires (SRB, CRA, PCS) at 2-4 timepoints, depending of the illness trajectory of the patient:
- at baseline (T0),
- one month after baseline (T1),
- one month after T1 (T2),
- 4-6 weeks following the patient's death (T3).
The questionnaires will take around 15-20 min to complete.
## | statistical analyses
Analyses will be performed following the intention-to-treat principle. Descriptive statistics will be used to compare characteristics of family caregivers , relation to patient, living with patient, diagnose patient and the intensity of informal care
[hours per week]) at baseline (T0) between the intervention and control group, using independent sample t tests and chi-square tests.
Multilevel/multivariate analyses will be used to examine outcomes in the intervention and control group on T1, taking into account potential variation in age of the family caregivers and intensity of the informal care provided. Repeated measures analysis of variance will be conducted to assess the development of outcomes over time.
All statistics will be two-sided and considered significant if p < .05.
If necessary, imputation will be used to handle missing data (max. 20%).
## | other outcomes
The incidence of acute admissions of patients in the intervention and control group will be calculated. Experiences of nurses using the CSNAT approach will be qualitatively analysed.
## | sample size calculation
This study will consist of a sample size of 92 family caregivers in each group (184 in total). Assuming that we will evaluate 184 family caregivers, using a significant level of 5% and an intra-cluster correlation coefficient of 5%, we will have a power of 75% to demonstrate a difference of 15 points on the SRB.
## | validity and reliability
The included questionnaires (SRB, CRA, PCS) have been tested for validity and reliability for various settings and countries [bib_ref] Mutuality and preparedness as predictors of caregiver role strain, Archbold [/bib_ref] [bib_ref] The caregiver reaction assessment (CRA) for caregivers to persons with chronic physical..., Given [/bib_ref] [bib_ref] Measuring both negative and positive reactions to giving care to cancer patients:..., Nijboer [/bib_ref] [bib_ref] Psychometric properties of the caregiver preparedness scale in caregivers of stroke survivors, Pucciarelli [/bib_ref] [bib_ref] Instruments for assessing the burden of informal caregiving for stroke patients in..., Exel [/bib_ref]. Furthermore, we will reduce the biases in estimating the effects of the intervention. In this nursing trial, it is impossible to blind the participants and nurses for the intervention.
But blinding of the analysis will be achieved by withholding information from the analytic researcher about how the intervention and control groups are coded.
## | e thic al cons ider ations
The research protocol was reviewed and approved by the Medical Ethics Review Committee in April 2019 (NL68453.078.18). All participants will receive oral and written information on the study and sign an informed consent before entering the study. Family caregivers of patients who are cared for by a home care organization in the intervention group may receive the CSNAT intervention, even if they decline to participate in the study.
## | d iscuss i on
Few studies on nursing supportive interventions to support family caregivers in end-of-life care with a robust design are available [bib_ref] Nursing interventions to support family caregivers in end-of-life care at home: A..., Becqué [/bib_ref]. Findings of this cluster randomized controlled trial will contribute to the scientific and practical knowledge of nursing interventions to support family caregivers caring for patients in the last phase of life. Additionally, this study will yield a training programme and implementation manual to implement this new intervention in nursing practice.
# | limitations
A potential limitation of our study is that inviting and informing home care organizations and nurses about the study, may have influenced the awareness of nurses in the control group. In addition, only home care organizations who were positive about our research responded to our invitation to participate. It is possible that these organizations already had a high priority for palliative care.
We believe that the CSNAT approach will not be a burden for family caregivers but completing 2-4 questionnaires may be a barrier for participation. We therefore developed a questionnaire that is as short as possible. However, attritions are a common problem in trials in end-of-life care because of the vulnerability of family caregivers and the unpredictable illness trajectory of the patient [bib_ref] Attrition rates, reasons and predictive factors in supportive care and palliative oncology..., Hui [/bib_ref].
Finally, we realize that the intervention implementation process (training of the nurses, time, resources and adaptation) is crucial for the impact of the intervention [bib_ref] Process evaluation of complex interventions: Medical Research Council guidance, Moore [/bib_ref]. In our study we will evaluate the implementation process by interviewing nurses and screening patient records. This will offer us the opportunity to interpret our findings in the light of the implementation process.
## | con clus ion
This study will evaluate the effects of a structured nurse-led supportive intervention on family caregivers in end-of-life care. The evidence gained from this study will provide nurses insight into how to support family caregivers and to address their needs. This may improve the well-being of family caregivers.
## Ack n owled g em ents
|
What matters in development and sustainment of community dementia friendly initiatives and why? A realist multiple case study
Background Dementia friendly communities (DFCs) are seen as key to participation of people with dementia and carers. Dementia-friendly initiatives (DFI) are important building blocks for the growth of DFCs. Therefore, it is essential to understand how DFIs are developed and sustained to secure the growth of DFCs. This study identifies contextual factors and mechanisms that influence the development and sustainment of Dutch DFIs. It also explains how these contextual factors and mechanisms are interrelated and the outcomes to which they lead.Methods Mixed methods, namely interviews, observations, documentation and focus groups, were used for this realist multiple case study. Participants were professionals (n = 46), volunteers (n = 20), people with dementia (n = 1) and carers (n = 2) who were involved in development and sustainment of DFIs in four Dutch DFCs.ResultsThis study revealed three middle-range program theories as final outcomes: development of a support base, collaboration, and participation in DFIs by people with dementia and carers. These theories address institutional, organisational, interpersonal and individual levels in the community that are essential in development and sustainment of DFIs.ConclusionsThe development and sustainment of DFIs requires the development of a support base, collaboration, and participation in DFIs by people with dementia and their carers. data. All audio-and videotapes and transcripts were stored on a save directory at the Radboud University medical centre.Consent for publicationConsent for publication was obtained from all of the study's participants. Participants were provided with information letters and were also verbally informed of the fact that any information they shared would be treated completely confidentialy and anonymously (e.g. without any personidentifiable information). This is in accordance with Dutch national guidelines.
# Background
Growing recognition of dementia as an urgent global health issue has led to an increase in dementia-friendly communities. Dementia-friendly communities (DFCs) share a common goal of ensuring that people with dementia and their carers continue to participate and be valued as citizens [bib_ref] Developing a dementia-friendly Christchurch: perspectives of people with dementia, Smith [/bib_ref]. DFCs can be characterised by their location, such as a city or neighbourhood. These are called 'location-based DFCs' [bib_ref] Dementia friendly communities in England: a scoping study, Buckner [/bib_ref]. DFCs can also be organisations or entities with a specific focus (for example, airports). These are referred as 'communities of interests' [bib_ref] Dementia friendly communities in England: a scoping study, Buckner [/bib_ref]. The need for building DFCs is recognised by 90 % of Organization for Economic Co-operation and Development (OECD) countries, including the Netherlands. The need and ambition to build Dutch DFC is captured in a national strategy called Deltaplan Dementie.
Although this need to nurture DFCs is widely recognised, the development of DFCs is influenced by a broad range of stakeholders and organisational factors that make this task complex. Previous research has shown that DFCs require both top-down input by the (local) government, such as policy, facilitation and finances, and bottom-up (local) resources and initiatives, such as initiatives focusing on awareness about dementia and related social interaction [bib_ref] Dementia-friendly communities: a review of current literature and reflections on implementation, Novak [/bib_ref]. Such dementiafriendly initiatives (DFIs) are the 'building blocks' in the development of DFCs [bib_ref] Dementia friendly initiatives: a state of the science review, Hebert [/bib_ref]. DFIs are initiatives and activities that aim to promote dignity, empowerment, engagement and autonomy to enhance the wellbeing of people with dementia and their carers, and to address the needs of carers throughout the dementia trajectory [bib_ref] Dementia friendly initiatives: a state of the science review, Hebert [/bib_ref]. The specific goal of DFIs is to bring about changes to the social and/or physical environment to create DFC. DFIs therefore work towards a community that includes and empowers people with dementia and their carers. The term, "dementia-friendly community" refers to the kind of community to strive for, where everyone, including people with dementia, has a place. By "dementiafriendly initiatives" we mean the activities being undertaken to make communities more inclusive of people with dementia and their carers. As such, a DFC can evolve from a collection of DFIs. For example, a neighbourhood DFC can have different DFIs, such as education about dementia and related social interaction for supermarkets and/or adaptations to the physical environment to improve recognition in that environment.
In the Netherlands, decentralisation of health and care policies and services makes municipalities responsible for providing appropriate dementia care and support, as stated by the Social Support Act 2015Therefore, building DFC is a responsibility of municipalities. This has automatically led to an increase of local DFIs. Local DFIs may vary significantly in format, structure, content and outcomes, based on local resources and needs, the partners engaged and population groups targeted [bib_ref] How do community based dementia friendly initiatives work for people with dementia..., Thijssen [/bib_ref] [bib_ref] The place for dementia-friendly communities in England and its relationship with epidemiological..., Woodward [/bib_ref]. That means that one municipality decides to initiate a Meeting Centre as a DFI while another municipality focuses on integrating dementia in existing activities such as a choir or community garden.
The development of DFIs is a complex local process. It requires commitment of key actors in a local context, ranging from municipalities and healthcare and social organisations to businesses and voluntary-sector organisations [bib_ref] Dementia-friendly community initiatives: an integrative review, Shannon [/bib_ref]. They operate within the local processes and structures of municipal, health and social systems, with their resources and restrictions [bib_ref] Dementia-friendly community initiatives: an integrative review, Shannon [/bib_ref]. Because of the local nature of the development of Dutch DFIs, more contextual in depth knowledge is needed how to develop and sustain DFIs, which previous research about DFCs have not explicated. No previous research has investigated how key people from different backgrounds come and work together to develop and sustain a DFI. No research has clarified which contextual factors are critical to set up a DFI, for whom they are needed, and how these factors add to the development and sustainment of a DFI. Clearly, more in depth knowledge is needed about how DFIs are developed and sustained in order to secure the evolution of DFCs.
The aim of this study was to identify contextual factors and mechanisms that influence the development and sustainment of Dutch DFIs while building a DFC, and explain how they are interrelated. To this end, a multiple case study was carried out using a realist approach. The main goal of the realist approach is to illuminate how complex social interventions work, using contextmechanisms-outcome configurations from one or more realist program theories, to answer the questions: 'what works, for whom, in what contexts and why' by describing causal relationships to explain outcomes [bib_ref] Realist review--a new method of systematic review designed for complex policy interventions, Pawson [/bib_ref]. The focus of the realist approach on identifying mechanisms and exploring how they operate in different contexts also provides valuable insights into how and why intervention programmes lead to change. The realist approach can add value by enhancing the clarity, depth, and portability of findings, helping professionals and researchers deal with context and complexity in pragmatic ways. .Furthermore, the realist approach enables exploration of causal processes within and across multiple levels of a social system, as occurs in the development and sustainment of DFIs [bib_ref] Developing complexity-consistent theory in a realist investigation, Westhorp [/bib_ref]. These characterizations of the realist approach are particularly useful when (future) DFC wish to learn how to apply lessons from a local DFC elsewhere.
The research question(s) were:
- Which mechanisms are important in developing and sustaining DFIs, what outcomes do they have, and why? - Which contextual aspects affect these mechanisms?
This research was the second phase of the Mentality Project (November 2017-October 2022), which studies success factors in DFIs using the realist approach. The research was guided by the research team and an advisory panel consisting of experts in the field of dementia and public health, representatives of people with dementia and their caregivers and stakeholders from four Dutch municipalities seeking to become dementia-friendly. More information about Mentality can be found at www. Menta lity. space.
# Methods
A multiple case study design using the realist approach was chosen because it allowed in-depth study of development and sustainment of DFIs within the real-life context of Dutch location-based DFCs. Multiple methods of data collection were performed, including semi-structured interviews, participating observations, available documents and focus groups.
## Case selection
Selected cases were DFCs that were officially recognised as such according to Dutch Alzheimer Association's criteria. Within the cases of Dutch DFCs, a DFI was our unit of analyses, which means that the DFI was the case study topic to be studied. Purposeful sampling, as a method to select information rich cases [bib_ref] Needs, issues, and expectations on dementia care at home across Europe to..., Bressan [/bib_ref] was used based on maximum variation of DFCs in terms of rurality (urban vs. rural sites), the geographic scatter in the Netherlands, duration of being a DFC and characteristics of DFIs within a DFC. This sequential sampling method was established together with the advisory panel of Mentality. A multiple recruitment strategy was adopted by identifying cases through the cooperation of the Dutch Alzheimer's association, the advisory panel of Mentality, and the network of the research team at national and/or regional level.
Twenty-five potential DFCs were identified and were invited by the research coordinator (JP) to take part. Nine DFCs were willing to participate, and were considered in terms of the maximum variation and the opportunity for identifying contextual factors, mechanisms and outcomes. Four DFCs were selected by the research team and advisory panel as best possible sites.
After the initial selection of DFCs, additional information was sent to key informants, such as policy officers.
The key informants informed and consulted other stakeholders involved in the development and sustainment of DFIs, such as social and health professionals and/or volunteers, before making definite commitments. After this step, a letter of commitment was signed by the local policy officer for each case. [fig_ref] Table 1: Included Cases and Their Characteristics [/fig_ref] provides an overview of included cases and their characteristics.
## Recruitment of participants within the dfcs
Recruitment of participants within DFCs was organized around each DFI as unit of analysis. The DFIs were developed to include people with dementia, sometimes in existing activities such as the intergenerational gardening, dementia-friendly museum, dementia-friendly choir and shared living room for older people, including people with dementia and sometimes by specific dementia friendly activities such as adapting the physical environment, Alzheimer Café and Odense housing. Participants were stakeholders who were actively involved in the development and sustainment of one or more DFIs. The aim was to include a variety of stakeholders who were involved at both the municipal and implementation level of one or more DFIs. Given time constraints, we set a target of 7-12 participants per DFC for project manageability.
Participants were recruited through first purposeful sampling [bib_ref] Needs, issues, and expectations on dementia care at home across Europe to..., Bressan [/bib_ref] by selecting key informants in each DFC and then snowball samplingwhich means that key informants were asked to identify other information-rich informants who were active and engaged in the development and implementation of DFIs. As such, participants assisted in the identification of other eligible participants.
## Data collection
Following the realist and case study methodology, multiple methods of data collection were used. These were semi-structured interviews, participating observations, available documents and focus groups. In all cases, data collection started with semi-structured interviews, interspersed with observations and documentation. Data collection was divided among four researchers. One pair (MT and JP) collected data in A and B. Another pair (ML and RJ) collected data in C and D. All data were collected between January 2019 and July 2019. See [fig_ref] Table 2: Overview of Participants by Case and Data Collection Method [/fig_ref] for an overview of data collection.
## Semi-structured interviews
Semi-structured interviews were conducted using a topic guide following the realist principles. They included open-ended and exploratory questions for developing program theories [bib_ref] The craft of interviewing in realist evaluation, Manzano [/bib_ref]. The purpose was to gain insight into the nature of the DFI; the involvement of the interviewee in the DFI; and their experiences with and perspectives on how the DFI was developed, implemented and sustained. Special attention was given to contextual aspects and mechanisms leading to outcomes. All interviews were audio recorded with the permission of the participant(s). Each interview lasted approximately 40-60 minutes, and took place face to face, on location, chosen by the participants. These locations were mostly at their homes or at work. In total, 29 professionals, 10 volunteers and 1 person with dementia were interviewed. The topic guide is available in Additional File 1.
## Participating observations
The purpose of the participating observations was to gain insight into the implementation of DFI and to collect information regarding the context, such as the social and physical environment. Open notes were made during and after the observation sessions, and were used for probing contexts, mechanisms or outcomes during interviews and focus groups. Each observation lasted approximately 40-60 minutes and took place at the location of the DFI. In total, 11 participating observations of DFI were performed.
## Documentation
Documents related to organisational policy papers, evaluations of and proposals for DFIs were studied. These documents, collected and delivered by the interview participants, aimed to contextualise and supplement data from interviews and observations. In total, 17 documents were collected.
## Focus groups
In each case, interim findings from the interviews, observations and documents were presented as summary reports during a focus group meeting with the participants from interviews and observations, as well as other relevant participants who could not be interviewed individually. This was a basis for member checking, discussion to highlight unanswered questions and deepen the findings for theory development. Each focus group lasted approximately 60-90 minutes and took place at a location, chosen by the participants. These locations were at the community centre (case A) and city hall (case B,C and D). In total, 17 professionals, 10 volunteers and 2 carers participated in the focus groups. [fig_ref] Table 2: Overview of Participants by Case and Data Collection Method [/fig_ref] shows the overview of participants, specified per case and data collection method.
## Data extraction
Data extraction started per case after data from interviews, participating observations and documentation were collected. To ensure consistency and transparency, definitions of 'contexts' , 'mechanism resources' , 'mechanism response' and 'outcomes' were used [bib_ref] Achieving successful community engagement: a rapid realist review, Weger [/bib_ref] [bib_ref] Assessing the outcomes of participatory research: protocol for identifying, selecting, appraising and..., Jagosh [/bib_ref] [bib_ref] What's in a mechanism? Development of a key concept in realist evaluation, Dalkin [/bib_ref]. See for definitions. Following the realist approach, the researchers (MT and AOB) constructed context-mechanism-outcome configurations (CMOc) from each interview transcript in order to examine what worked for whom, under what circumstances, and why and how. Interviews were thus analysed using CMOc, which were drafted and placed (by AOB/ MT) into a data extraction form in Excel, which is available in Additional File 2. Next, data from field notes and documents was extracted and supplemented to the CMOc. Corresponding CMOc were triangulated, noting the data source and corresponding fieldnotes or documentation. If new or rival data were found, they were extracted in new CMOc.
Labelling CMOC on outcomes and levels of change.
In the last steps of data extraction, outcomes of each CMOc were labelled according to whether they were observed, anticipated or implied by the data source, following the reasoning that any insight into relevant contextual factors or mechanisms must relate to an outcome. Each CMOc was also labelled on levels of change as defined by Puntonto deepen the data-extraction on mechanisms, since change occurs at different social strata, consistent with the development of DFI [bib_ref] A realist evaluation of community-based participatory research: partnership synergy, trust building and..., Jagosh [/bib_ref] [bib_ref] Using complexity-consistent theory for evaluating complex systems, Westhorp [/bib_ref]. Thus, researchers (AOB/MT) labelled each CMOC according to the level on which change took place -the individual, interpersonal, organisational or institutional.
## Data synthesis
Data synthesis used an inductive, sequential and iterative approach [bib_ref] Using realistic evaluation to evaluate a practice-level intervention to improve primary healthcare..., Byng [/bib_ref]. Sythesis of information within each case was followed by a cross-case synthesis and finally overall synthesis. All authors participated in each step.
## Within-case synthesis
For this, we used retroduction, which is a central inference-making method in realist research. Retroductive theorizing involves starting with a program's effects Definitions for Data Extraction Context pertains to the backdrop of an intervention [bib_ref] A realist evaluation of community-based participatory research: partnership synergy, trust building and..., Jagosh [/bib_ref]. Context includes the pre-existing organisational structures; the cultural norms and history of the community; the nature and scope of pre-existing networks; and geographic location effects, such as social and physical environment or previous experience with dementia-friendly initiatives [bib_ref] A realist evaluation of community-based participatory research: partnership synergy, trust building and..., Jagosh [/bib_ref]. Mechanisms are not interventions. They are the -often invisible -forces, powers, processes or interactions that lead to (or inhibit) change. They can be found in the choices, reasoning and decisions that people make as a result of the resources; the interactions between individuals or groups; and the powers and liabilities that things, people or institutions have as a result of their position in a group or society [bib_ref] Understanding mechanisms in realist evaluation, Westhorp [/bib_ref]. Mechanisms are 'triggered' when (program) resources (e.g. information, money, expertise) interact with specific features of the context (individual, interpersonal, organizational, or institutional). Mechanisms resources refer to what is triggered among the context of participants/stakeholders [bib_ref] What's in a mechanism? Development of a key concept in realist evaluation, Dalkin [/bib_ref] [bib_ref] A realist evaluation of community-based participatory research: partnership synergy, trust building and..., Jagosh [/bib_ref]. Mechanisms response refers to the responses of the participants, all that suggests a change in people's minds and actions [bib_ref] What's in a mechanism? Development of a key concept in realist evaluation, Dalkin [/bib_ref] [bib_ref] A realist evaluation of community-based participatory research: partnership synergy, trust building and..., Jagosh [/bib_ref]. Outcomes are the results of how people react to the mechanisms. Outcomes are either intended or unintended and can be proximal, intermediate and final. Outcomes can be labelled as:
- observed (the participant stated during interviews or observations that it had happened); - anticipated (it had not happened yet but the participant expected it to); or - implied (no explicit mention of the outcome was made but the data enabled the research team to infer, tentatively, that the participant had observed or anticipated it).
## Labelling levels of change in mechanisms:
- Individual changes include individuals' skills and knowledge relating to dementia and DFIs, as well as the motivation, attitudes, commitment and values that affect individual behaviour. - Interpersonal and network change refers to the relationships and networks between individuals and groups that influence development or sustainability of DFIs. - Organisational change refers to the systems, policies and procedures, practices, culture and norms within an organisation that affect the access and resources needed to develop and sustain DFIs. - Institutional change relates to the wider environment within which individuals, networks and organisations operate. This includes the political system, civil society and the media, political and economic factors, and broader social factors (culture, norms, collective beliefs) that influence the development and sustainability of DFIs. and working backward to think about the conditions of reality necessary for such effects to manifest [bib_ref] Retroductive theorizing in Pawson and Tilley's applied scientific realism, Jagosh [/bib_ref] [bib_ref] • fast, convenient online submission • thorough peer review by experienced researchers..., Mukumbang [/bib_ref] Within case synthesis started by identifying concrete outcomes ('what were changes so that DFIs were developed and sustained?') and then working backwards to mechanisms ('how and why did these changes occur?') and then identifying contexts ('under what circumstances will these mechanisms lead to these changes?') [bib_ref] Retroductive theorizing in Pawson and Tilley's applied scientific realism, Jagosh [/bib_ref] [bib_ref] • fast, convenient online submission • thorough peer review by experienced researchers..., Mukumbang [/bib_ref] [bib_ref] Data analysis and synthesis within a realist evaluation: toward more transparent methodological..., Gilmore [/bib_ref]. As such, similar outcomes, such as an improved perception of the importance of a DFI or taking initiative, were clustered per case. Second, commonalities of mechanisms and contexts were also clustered, such as mechanisms referring to feeling important or contextual aspects such as stakeholders from various organizations. Third, based on the clusters, patterns in outcomes, mechanisms and contexts were outlined. An example of such an outline was contextual factors, such as stakeholders having a personal affinity with people with dementia and carers, leading to mechanisms such as feelings of being important or connected, which then themselves led to outcomes such as taking initiative or purposefulness in collaboration. The levels on which change occurred remained distinct. These outlines were compared with corresponding configurations and quotes from data sources, to check for consistency and explanatory power [bib_ref] Data analysis and synthesis within a realist evaluation: toward more transparent methodological..., Gilmore [/bib_ref].
## Cross-case synthesis
Cross case synthesis started by following the same three steps as within-case synthesis, namely clustering of the outcomes, followed by clustering of commonalities of mechanisms and contexts, and finally developing outlines. This resulted in ten outlines, including intermediate outcomes in which mechanisms remained distinct on levels of change. The ten outlines are available in Additional File 3. These outlines, including the accompanying levels of change, were discussed with all authors. This discussion confirmed the synthesis and made suggestions for wording.
## Configuring middle-range program theories
Next, the ten outlines were configurated into initial middle-range program theories (MRPTs) by clustering outcomes and selecting the most prevalent mechanisms, including their contextual factors, by identifying demiregularities through retroductive reasoning [bib_ref] Retroductive theorizing in Pawson and Tilley's applied scientific realism, Jagosh [/bib_ref] [bib_ref] • fast, convenient online submission • thorough peer review by experienced researchers..., Mukumbang [/bib_ref] [bib_ref] Data analysis and synthesis within a realist evaluation: toward more transparent methodological..., Gilmore [/bib_ref] This was carried out by three researchers independently (MT, WK and LD) and overlap and differences were discussed until consensus was reached. After this step, the initial MRPTs were presented to all authors and the advisory panel for their feedback. They discussed the MRPTs using their field expertise and confirmed those MRPTs' usefulness and applicability. Suggestions for wording led to the final MRPTs.
Formulating realist program theories at a midrange level, such as middle-range program theories (MRPT), enabled both the specification of contexts, resources responses leading to outcomes and the conceptualization and explanation of those outcomes. Such MRPTs constituted more granular hypotheses about specific causal links and processes to communicate the findings as concretely as possible in practice compared to the more abstract middle-range theories.
# Results
The following section describes three final MRPTs. Each MRPT is represented by a logic model complemented by a narrative as this allows a description of essential features of both actions and change [bib_ref] How do community based dementia friendly initiatives work for people with dementia..., Thijssen [/bib_ref]. Each MRPT describes a final outcome in the title and explains how these outcomes are built by contextual factors, mechanisms and lead to intermediate outcomes. Special attention is given to the contextual factors, resources and the levels of change.
## Mrpt 1: development of baseline support for a dementia friendly initiative (dfi)
This section describes the MRPT of the development of baseline support for a dementia friendly initiative. For development of baseline support for DFIs, professionals and volunteers from municipalities and health and welfare organisations had insight in current issues in the community. These included needs related to dementia and/or from people with dementia and their carers. Other contextual features were the visibility and approachability of the professionals, volunteers, persons with dementia and carers, so that community members knew their background and could more easily approach them, for example to ask questions. Actions such as organising information meetings about dementia or making plans for a DFI for community members were conducted by the professionals and volunteers sequentially, preferably together with people with dementia and carers. During those actions, resources from organizations were commonly used, such as the physical location of an organization or PR supplies from an Alzheimer's association. Those actions addressed the interests of both people with dementia and their carers as well as other community members, such as parents and their children, leading to a reciprocity of interests. On the organizational level, feelings of importance and making a difference arose among professionals and volunteers, because they felt that the interest of their organisations or background resonated with interest of other stakeholders and that their resources, such as a physical location or network, were important for a mutual interest, such as a DFI. On the institutional level, community members from the neighbourhood also felt that their interests resonated with interests of other stakeholders and that their resources, such as time or network, had the same importance for a mutual interest, such as a DFI. Feeling important and making a difference arose through reciprocity of interests and recognition that individual interests also mattered for a greater purpose. Therefore, it resonated with feelings to make a positive difference for others. Such responses changed collective beliefs about needs and possibilities for a DFI and therefore brought about changes on an institutional level. Intermediate outcomes of building support for a DFI were having a (more) positive image of dementia and the importance of DFCs and DFIs, and having a (more) positive idea how to contribute to a DFI and develop concrete intentions and plans.
## Mrpt 2: collaboration for developing and sustaining dfis
This section describes the MRPT of collaboration for a DFI. [fig_ref] Figure 2: Middle-range Program Theory [/fig_ref] outlines the characteristics of this MRPT.
For collaboration, professionals and volunteers from municipalities, health and welfare organisations came together, preferably complemented by people with dementia and their carers. They all shared a personal affinity with dementia or the purpose of a DFC or DFI from various experiences. For example, there was a policy officer who understood the impact of dementia from personal experience as a carer and a volunteer who used to work in elderly care. Other contextual features were transparency about manpower and available budget. As a result, the professionals and volunteers involved knew the preconditions to come together. Follow-up actions were organising regular meetings about the DFIs to be developed or sustained. At these meetings, professionals and volunteers took roles that best suited their personal and professional experiences. During those meetings, relevant information was shared, such as information about history and/or developments in the community or which funds to apply for. Accordingly, meetings were characterised by sharing, and subsequently by joint decision making. These contextual features and resources on the interpersonal level led to responses of both having a grip on and overview of collaboration, and of feeling connected with each other. Such responses led to mutual network building and therefore brought changes on the interpersonal level. Intermediate outcomes of collaboration were taking initiative and/or continuing commitment to a DFI and experiencing purposefulness, as well as satisfaction and fun during the collaboration.
## Mrpt 3: participation in dfi by people with dementia and their carers
This section describes the MRPT of participation in DFIs by people with dementia and their carers. [fig_ref] Figure 3: Middle-range Program Theory [/fig_ref] outlines the characteristics of this MRPT.
To enable participation in DFIs by people with dementia and their carers, an overview was developed by dementia trained and competent professionals and/or volunteers from municipalities and health and welfare organizations. This overview outlined the needs of people with dementia and their carers in terms of activities and information meetings, by whom these were delivered and at what locations. To avoid stigmatisation and support inclusion, the overview purposely did not include terms associated with dementia. Therefore, the term 'dementia-friendly initiative' was not used. The locations, such as a community garden or a community centre, were also chosen to be inclusive and not associated with dementia. Sequentially, actions were undertaken to communicate this overview. These actions included for example publication in a local newspaper and word of mouth information sharing by professionals and volunteers. During these actions, resources such as the personal and professional networks of professionals and volunteers were used. Moreover, professionals and volunteers from these networks became more engaged in the DFIs and their purpose as they were activated in promoting these initiatives. These contextual features and resources on the interpersonal level led to responses of feelings of encouragement among people with dementia and carers. For example, they identified with activities or information meetings that related to their needs. Another response was that people with dementia and their carers felt respected, for example by recognition of their needs in the overview or by deciding for themselves whether and when they would participate in which activities. Such responses increased personal motivation and commitment to participate, and therefore brought changes on the individual level. Intermediate outcomes of participation in DFIs by people with dementia and their carers were using or taking part in activities and taking joint action to partake in activities with other community members. In doing so, people with dementia and their carers, as well as community members, became more open to the presence of (people with) dementia. [fig_ref] Table 4: Overview of the Middle-range Program Theories and Associated Levels of Change [/fig_ref] shows an overview of the MRPTs and associated levels of change.
# Discussion
## Implications of the findings in context of the existing research
We aimed to explain which mechanisms are important in developing and sustaining DFIs in a Dutch DFC, which outcomes they produced and which contextual factors affected these mechanisms. To our knowledge, this is the first study that has addressed the development and sustainment of DFIs including the different levels on which changes occur. Our analysis revealed three MRPTs in developing and sustaining DFIs. These theories relate to the development of baseline support, collaboration, and participation in DFIs by people with dementia and their carers. Moreover, our results further clarify how each MRPT brought about changes on different levels in the community. Development of baseline support caused changes on the organizational and institutional level, collaboration caused changes on the interpersonal level, and participation in DFIs by people with dementia and their carers caused changes on the interpersonal and individual levels. Our results support previous research about the development and sustainment of complex interventions in dementia community care, which confirms the importance of a support base [bib_ref] Dementia friendly communities in England: a scoping study, Buckner [/bib_ref] [bib_ref] Sustaining community-based interventions for people affected by dementia long term: the SCI-Dem..., Morton [/bib_ref] , collaboration by diverse partners [bib_ref] Sustaining community-based interventions for people affected by dementia long term: the SCI-Dem..., Morton [/bib_ref] [bib_ref] Evaluating the social fitness Programme for older people with cognitive problems and..., Donkers [/bib_ref] and an understanding of pre-requisites for participating in initiatives [bib_ref] Sustaining community-based interventions for people affected by dementia long term: the SCI-Dem..., Morton [/bib_ref] [bib_ref] Evaluating the social fitness Programme for older people with cognitive problems and..., Donkers [/bib_ref]. Moreover, our results highlight how contextual aspects such as diversity among partners and insight into issues in the community, as well as mechanisms such as sharing of resources and reciprocity of interests, provoke feelings of importance, connection and encouragement among all community members. For example, in our study, the diversity of professionals and volunteers supported the sharing of each other's resources, such as spaces, information and expertise. Likewise, reciprocity arose through the connection of personal interests with needs and initiatives in the community, and generated significance, engagement and activity. For example, when a neighbourhood was informed about widening the sidewalks and landscaping for people with dementia, the benefits of these initiatives for people with dementia and carers were connected with benefits for parents with strollers and children's' play and educational options. Consequently, people became committed to each other and engaged with the purpose of the DFI. Of most interest is that these mechanisms of sharing, reciprocity, significance, connection and engagement rely much more on what connects people together and on positive aspects, rather than what a separate and/or vulnerable group needs. This marks a shift in setting up DFIs in which it is not so much the negative consequences of dementia that act as the impetus for change, but rather a cause that connects residents in the community. Such impetus for change differs from existing toolkitsand is known as asset based community development (ABCD), an approach which focusses on assets and strengths in individuals and communities, rather than on their problems and deficits [bib_ref] Assets-based community development, Kretzmann [/bib_ref]. The ABCD approach is used for empowering communities in addressing health inequalities and health promotion in groups from adolescents [bib_ref] Does an asset-based community development project promote health and wellbeing?, Van De Venter [/bib_ref] [bib_ref] Can asset-based community development with children and youth enhance the level of..., Agdal [/bib_ref] [bib_ref] Harnessing the social capital of rural communities for youth mental health: an..., Boyd [/bib_ref] [bib_ref] Systems of care as asset-building communities: implementing strengths-based planning and positive youth..., Mccammon [/bib_ref] to older people [bib_ref] Age-friendly cities and communities: a review and future directions, Torku [/bib_ref] [bib_ref] Asset based community development to promote healthy aging in a rural context..., Kobayashi [/bib_ref] [bib_ref] Exploring assets of people with memory problems and dementia in public space:..., Sturge [/bib_ref]. The asset-based approach has recently been promoted as an alternative policy for DFC [bib_ref] Exploring assets of people with memory problems and dementia in public space:..., Sturge [/bib_ref] [bib_ref] Illuminating community services and assets towards better dementia care, Chadborn [/bib_ref] [bib_ref] Assets-based approaches and dementia-friendly communities, Rahman [/bib_ref]. Moreover, the importance of moving away from deficits and focus on abilities was also highlighted to explain the outcomes of community DFIs for people with dementia and their carers [bib_ref] How do community based dementia friendly initiatives work for people with dementia..., Thijssen [/bib_ref]. Our current results explain how a focus on the abilities and wishes of people with dementia, carers and other community members during development and sustainment of DFIs can be the basis for positive responses such as sharing, reciprocity, significance, connection and engagement. As such, our results aligns with the new proposed policy for DFCs and offers new insights into positive responses that are needed for the process in development and sustainment of DFIs. It enables reflection on how and why the process was successful, or not. Within the context of Dutch DFCs, our results describe how institutional, organizational, interpersonal and individual levels of change are important in the development and sustainment of DFIs. In particular, they explain how each MRPT affects one or two different levels in the community. As such, our results resonate with findings about development of DFCs that address the importance of both top-down and bottom-up approaches [bib_ref] Dementia-friendly communities: challenges and strategies for achieving stakeholder involvement, Heward [/bib_ref]. However, our results deepen these findings by focusing and clarifying which contextual aspects and resources are needed and why, and how they trigger mechanisms in different levels of change to build DFIs as building blocks for a DFC. Next, our results suggest an interdependence between the mechanisms, intermediate outcomes of MRPTs and associated levels of change. For example, the positivity from the intermediate outcomes of a support base is likely to generate commitment and willingness to participate in a collaboration. As such, it can start a collaboration or intensify it by moving others to join or spending more effort in it. Consequently, the activity from the intermediate outcomes of collaboration may lead competent professionals and volunteers to support participation in a DFI. Additionally, the positive intermediate outcomes of participation in DFIs, such as openness about dementia and joint participation in activities, can also contribute to a positive idea of a person's own contribution to a DFI, and thus contribute to a support base. Such examples of interdependencies explain how the intermediate outcomes of one MRPT may become (a part of ) the context in a new MRPT [bib_ref] How do community based dementia friendly initiatives work for people with dementia..., Thijssen [/bib_ref] [bib_ref] A realist evaluation of community-based participatory research: partnership synergy, trust building and..., Jagosh [/bib_ref] without suggesting that there is only one possible direction in which interdependencies can exist. Our results regarding levels of change provide evidence and insights that encourage reflection on the development and sustainment of DFIs.
## Strengths and limitations of the study
Since the development of Dutch DFC and DFIs is locally bound, we included cases from different regions in the Netherland so diversity in contexts could be explored in depth. Additionally, a mixed method approach and a large number of participants (n = 69) ensured a complementary, rich and strong array of data. Lastly, input from multiple researchers and a focus on multiple levels deepened the analysis, and thus were important strengths of this study. It should be noted that almost all participants were professionalsand volunteers, rather than people with dementia (1) and/or carers (2), despite our intention to recruit all relevant stakeholders. Our observations revealed that most of the DFI development is 'organizationally led' -that is, initiated by organizations with strong connections with the municipalities. Most professionals and volunteers from those organizations are still searching for ways to reach out to people with dementia and their carers, which could explain the low number of people with dementia and carers in our study. These limitations, are known from research about disengagement of people with dementia and their carers in development of DFCs [bib_ref] Dementia-friendly communities: challenges and strategies for achieving stakeholder involvement, Heward [/bib_ref] [bib_ref] Dementia friendly care: methods to improve stakeholder engagement and decision making, Innes [/bib_ref] [bib_ref] Rationales and practices for dynamic stakeholder engagement and disengagement. Evidence from dementia-friendly..., Pascale [/bib_ref]. Involvement of people with dementia and their carers during development and sustainment of DFI is lacking [bib_ref] Dementia-friendly communities: challenges and strategies for achieving stakeholder involvement, Heward [/bib_ref]. Instead, interest groups are involved as advocates [bib_ref] Innovative methods for involving people with dementia and carers in the policymaking..., Keogh [/bib_ref]. People with dementia and carers are mainly questioned as 'users' of DFC about their priorities, experiences and outcomes of a DFC [bib_ref] Dementia-friendly communities: a review of current literature and reflections on implementation, Novak [/bib_ref] but are not involved in the decision making process [bib_ref] Dementia friendly care: methods to improve stakeholder engagement and decision making, Innes [/bib_ref]. We mitigated the low number of people with dementia and carers in our study partly by conducting participating observations at DFIs so that fieldnotes about experiences of people with dementia and carers, could be incorporated in interviews and focus groups. People with dementia and their carers were invited to, and two carers took part in, the focus group to discuss the summary reports.
## Practical implications and future studies
The three MRPTs highlight the importance of building a support base, collaboration, and participation in DFIs by people with dementia and their carers during development and sustainment of DFIs. Our results should encourage stakeholders in practice to reflect on intermediate outcomes in order to monitor progress and follow up on all levels in the community. Moreover, our mechanisms provide more depth in this process by explaining the importance of a positive narrative and assets, and how events and intermediate outcomes on one level affect other levels in the community. In practice, such information is important in managing the process and understanding how and why things occur. Our results give a deeper insight in the process and succesfactors and can be used next to, for example, toolboxes that offer practical materialor describe planning during implementation of DFis. Moreover, our results explain what important contextual conditions are during such a process and why -that is, which mechanisms arise and to which intermediate outcomes they lead on possible different levels of the community. Such insights will improve overview and reinforce the grip of stakeholders on the process of development and sustainment of DFIs. Future studies will be needed to further test and refine our MRPTs in other cases and contexts. Special attention is needed for the involvement of people with dementia and their carers during development and sustainment of DFIs. Insights from their perspectives will improve the understanding of building a support base, collaboration and participation in DFIs. We recommend a new study that can deepen the understanding of the perspectives of stakeholders, including people with dementia and carers regarding their involvement in the development and sustainment of DFIs. Next, future studies should address how people with dementia and their carers could be involved during the process of development and sustainment of DFIs. This will improve the management of the process.
# Conclusions
This study examined which contextual factors and mechanisms affected the development and sustainment of DFIs in a Dutch DFC, and the outcomes they produced. Two main conclusions can be drawn from the study. First, the results provide evidence about the importance of building a support base, collaboration and participation in DFIs by people with dementia and their carers. The accompanying mechanisms underpin the process of development and sustainment of DFIs and the importance of a positive narrative. Second, our results clearly suggest the interdependence between multiple levels in the community and how they impact the development and sustainment of DFIs. The results of our study can support practices of reflecting on intermediate outcomes and possible ripple effects. Accordingly, they can help stakeholders monitor their process of development and sustainment of DFIs. Our MRPTs may be used to support reflections on the process of development and sustainment of DFIs in a theory-based way.
This study provides transparency about the development and sustainment of DFIs, and is a reference point for future studies in which these theories can be tested and refined.
[fig] Figure 1: Middle-range Program Theory: Development of a Support Base [/fig]
[fig] Figure 2: Middle-range Program Theory: Collaboration to Develop and Sustain DFIs [/fig]
[fig] Figure 3: Middle-range Program Theory: Participation in DFIs by People with Dementia and their Carers [/fig]
[table] Table 1: Included Cases and Their Characteristics [/table]
[table] Table 2: Overview of Participants by Case and Data Collection Method [/table]
[table] Table 4: Overview of the Middle-range Program Theories and Associated Levels of Change [/table]
|
Collagen Can Selectively Trigger a Platelet Secretory Phenotype via Glycoprotein VI
Platelets are not only central actors of hemostasis and thrombosis but also of other processes including inflammation, angiogenesis, and tissue regeneration. Accumulating evidence indicates that these ''non classical'' functions of platelets do not necessarily rely on their well-known ability to form thrombi upon activation. This suggests the existence of nonthrombotic alternative states of platelets activation. We investigated this possibility through dose-response analysis of thrombin-and collagen-induced changes in platelet phenotype, with regards to morphological and functional markers of platelet activation including shape change, aggregation, P-selectin and phosphatidylserine surface expression, integrin activation, and release of soluble factors. We show that collagen at low dose (0.25 mg/mL) selectively triggers a platelet secretory phenotype characterized by the release of dense-and alpha granule-derived soluble factors without causing any of the other major platelet changes that usually accompany thrombus formation. Using a blocking antibody to glycoprotein VI (GPVI), we further show that this response is mediated by GPVI. Taken together, our results show that platelet activation goes beyond the mechanisms leading to platelet aggregation and also includes alternative platelet phenotypes that might contribute to their thrombus-independent functions.
# Introduction
Platelets are most commonly known for their central role in hemostasis and thrombosis, both of which rely on the so-called mechanism of platelet activation. More precisely, current knowledge of platelet activation describes it as the transition from a functionally resting state to a procoagulant and prothrombotic platelet phenotype. This multistep process is initially evoked by interactions of platelets with adhesive components of the subendothelial extracellular matrix at sites of vascular injury or by soluble platelet agonists. Such stimulated platelets go through morphological changes but above all, they undergo functional changes, with activation and surface expression of integrins and other adhesion molecules, exposure of procoagulant phosphatidylserine, and secretion of thrombogenic substances from their storage granules. Altogether, these changes result in the formation of either the primary hemostatic plug or a pathologic thrombus [bib_ref] Platelets at work in primary hemostasis, Broos [/bib_ref] [bib_ref] Blood platelet biochemistry, Broos [/bib_ref] [bib_ref] New fundamentals in hemostasis, Versteeg [/bib_ref]. This pattern of platelet activation has been extensively studied and is now widely accepted as the mechanism supporting platelet contribution to primary hemostasis and thrombosis. For this reason, when talking about platelet activation, one usually refers to this stereotypic adhesive, procoagulant and prothrombo-tic platelet state. Nevertheless, evidence that platelets can present various levels of activation has been clearly provided by in vitro and in vivo experiments showing that the acquisition of their aggregative function is a sequential and gradual process, with reversible and irreversible steps [bib_ref] A key role of adenosine diphosphate in the irreversible platelet aggregation induced..., Trumel [/bib_ref] [bib_ref] ADP induces partial platelet aggregation without shape change and potentiates collagen-induced aggregation..., Ohlmann [/bib_ref] [bib_ref] Rho-kinase is involved in the sustained phosphorylation of myosin and the irreversible..., Missy [/bib_ref] [bib_ref] Identification of a 2-stage platelet aggregation process mediating sheardependent thrombus formation, Maxwell [/bib_ref]. The concept of differential platelet activation is further supported by recent results showing that hemostatic plugs are heterogeneous in composition, with regional differences in the extent of platelet activation [bib_ref] Hierarchical organization in the hemostatic response and its relationship to the platelet-signaling..., Stalker [/bib_ref]. Moreover, it has been shown in vitro that platelets can differentially release cytokines [bib_ref] Toll-like receptor 4 ligand can differentially modulate the release of cytokines by..., Cognasse [/bib_ref] and angiogenic factors [bib_ref] Angiogenesis is regulated by a novel mechanism: pro-and antiangiogenic proteins are organized..., Italiano [/bib_ref] [bib_ref] The effect of P2Y-mediated platelet activation on the release of VEGF and..., Bambace [/bib_ref] [bib_ref] Release of angiogenesis regulatory proteins from platelet alpha granules: modulation of physiologic..., Battinelli [/bib_ref] in an agonist dependent-manner.
The relevance of alternative states of platelet activation becomes very likely if one considers that currently platelets are not only recognized as central actors of hemostasis and thrombosis, but also as regulators of many other pathophysiological processes including innate and adaptive immune responses [bib_ref] Platelets and the immune continuum, Semple [/bib_ref] [bib_ref] Crosstalk between Platelets and the Immune System: Old Systems with New Discoveries, Li [/bib_ref] , angiogenesis [bib_ref] Differential role of platelet granular mediators in angiogenesis, Brill [/bib_ref] [bib_ref] Platelets and platelet adhesion support angiogenesis while preventing excessive hemorrhage, Kisucka [/bib_ref] , or wound healing [bib_ref] Platelets, inflammation and tissue regeneration, Nurden [/bib_ref]. Although the exact mechanisms underlying these ''non classical'' functions of platelets have not been fully elucidated, there is substantial evidence that they do not necessarily rely on the classically described activation state of platelets. For example, platelets have been shown to exert a vasculoprotective action in various inflamed organs including the skin, lungs, kidneys, and solid tumors, and this, before any signs of thrombosis are evident [bib_ref] Inflammation induces hemorrhage in thrombocytopenia, Goerge [/bib_ref] [bib_ref] Platelet granule secretion continuously prevents intratumor hemorrhage, Ho-Tin-Noe [/bib_ref] [bib_ref] Mac-1 (CD11b/CD18) links inflammation and thrombosis after glomerular injury, Hirahashi [/bib_ref] [bib_ref] Platelet ITAM signaling is critical for vascular integrity in inflammation, Boulaftali [/bib_ref]. The fact that thrombus formation is dispensable for this beneficial action of platelets suggests the existence of alternative states of platelet activation that might be uncoupled from their procoagulant and/or prothrombotic activities. Here, we investigated this possibility through dose-response analysis of thrombin-and collagen-induced changes in platelet phenotype, with regards to morphological and functional markers of platelet activation including shape change, aggregation, integrin activation, P-selectin surface expression, and secretion of soluble compounds. Our results show that collagen at low dose triggers a non-thrombotic platelet secretory phenotype characterized by the release of various soluble platelet factors in the absence of the classical activation-associated changes.
# Methods
# Ethics statement
All blood donors were volunteers who gave their free and informed written consent to this research study, which conforms to the ethical standards of the Declaration of Helsinki. Legal and ethical authorization for research use of collected blood was obtained through a national convention between the French National Institute of Health and Medical Research (Inserm) and the French Blood Institute (EFS) (convention number I/DAJ/ C2675).
## Washed platelet preparation and stimulation
Blood from healthy volunteers who had taken no medication during the previous two weeks, was drawn into 15% (v/v) trisodium citrate acid-citric-dextrose (ACD-A, Vacutainer system; Beckton Dickinson, Le Pont-de-Clais, France). Washed platelets were prepared from isolated platelet rich plasma as previously reported [bib_ref] Adhesion and activation of human platelets induced by convulxin involve glycoprotein VI..., Jandrot-Perrus [/bib_ref] and resuspended at a final concentration of 2.10 8 / mL in platelet reaction buffer (Hepes 5 mM, NaHCO 3 12 mM, NaCl 137 mM, KCl 2 mM, CaCl 2 2 mM, NaH 2 PO 4 0.3 mM, MgCl 2 1 mM, glucose 5.5 mM, pH 7.4). At the end of the platelet preparation and before adding the agonists, platelets were allowed to recover from PGE1 and apyrase treatments for 20 min at 37uC. Washed platelets were then incubated with increasing doses of thrombin (purified human a-thrombin, 0.1 NIH unit/nM [bib_ref] Interaction of human alphathrombin and gamma-thrombin with antithrombin III, protein C and..., Bezeaud [/bib_ref] or fibrillar collagen from equine tendon (Horm collagen, Nycomed, Munich, Germany), either in static or stirring (500 rpm) conditions for 15 min at 37uC. In an alternative set of experiments, platelets (400 mL at 2.10 8 /mL) were stimulated in Millicell culture inserts (0.4 mm pore size, polycarbonate membrane, Millipore, Billerica, MA, USA) placed into 24-well plates containing 800 mL of reaction buffer, and incubated for 30 min at 37uC on an orbital shaker.
## Analysis of platelet morphology and p-selectin immunostaining
At the end of the incubation, 50 mL of resting or stimulated washed platelets were fixed in 950 mL of 3.7% paraformaldehyde (PFA) and centrifuged onto glass slides (15 min, 1200 g) for analysis by differential interference contrast (DIC) microscopy. For P-selectin immunostaining, platelets were permeabilized on ice with 0.1% Triton X-100 in 0.1 M citrate buffer, pH 6.0 and subjected to saturation with PBS at 3% bovine serum albumin (BSA), prior to incubation with a mouse monoclonal antibody to human P-selectin (BD Biosciences, San Jose, CA) at 5 mg/mL in PBS 1% BSA and 0.1% Tween 20. An Alexa Fluor 488conjugated rabbit polyclonal antibody to mouse IgG (Life Technologies, Saint Aubin, France) at 2 mg/mL was used as a secondary antibody. The slides were mounted with fluorescence mounting medium (Dako, Carpinteria, CA). Observation and acquisition of DIC and fluorescence microscopy images were made using a Zeiss Axio Observer microscope (Carl Zeiss, Le Pecq, France).
## Flow cytometry
Resting and stimulated platelets (5 mL) were incubated with a FITC-conjugated anti-P-selectin antibody or corresponding isotype-matched control (Beckman Coulter Immunotech, Marseille, France) in a final volume of 50 mL PBS. Integrin activation was assessed in a similar manner using the FITC-conjugated PAC1 antibody to activated glycoprotein IIb/IIIa. Phosphatidylserine exposure was quantified by annexin V labeling using Cy5-coupled annexin V (BD Pharmingen, San Diego, CA). A cell sample to which 5 mM EDTA was added just prior to staining with Cy5coupled annexin V was used as a negative control. Samples were analyzed by flow cytometry using a LSRII apparatus (BD Biosciences, Le Pont de Claix, France).
For measurement of soluble PF4 directly in platelet suspensions, we used a fluorescent microsphere-based flow cytometric immunoassay. Briefly, COOH-functionalized green fluorescent microspheres (Estapor F1XC-200, 2.45 mm diameter, Millipore) were coated with a mouse anti-human PF4 antibody (9 mg/cm 2 bead surface area, R&D systems, Minneapolis, MN) according to the manufacturer's instructions. At the end of the incubation of stirred platelets, microspheres were added to 50 mL platelet suspension aliquots at a final concentration of 20 000 microspheres/mL, together with biotinylated goat IgG to human PF4 (R&D systems) and Alexa Fluor 647 streptavidin (Life Technologies). Microspheres in platelet suspensions were identified in flow cytometry by their green fluorescence and light-scattering characteristics, and their mean Alexa 647 fluorescence intensity was measured. Mean Alexa 647 fluorescence intensities were then converted into PF4 concentration according to a standard curve determined by incubating microspheres with known amounts of recombinant human PF4 (R&D systems).
## Platelet aggregation
Platelet aggregation was measured by turbidimetry at 37uC under stirring conditions in the absence of exogenous fibrinogen. Washed platelets (2.10 8 /mL) in platelet reaction buffer were activated by addition of the agonists and aggregation was followed for 8 min using an APACT-4004 aggregometer (LABiTec, Ahrensburg, Germany).
## Calcium signaling
Washed platelets (3.10 8 /mL) in 5 mM Hepes, 137 mM NaCl, 2 mM KCl, 0.3 mM NaH 2 PO 4 , 1 mM MgCl 2 , 5.5 mM glucose, pH 7.4 (assay buffer) were incubated at 37uC for 30 min with 5 mM Oregon Green 488 BAPTA-1AM (Life Technologies, St Aubin, France). ACD-A (1/10 vol), apyrase (0.30 U/mL), PGE-1 (0.1 mM) and five volumes of wash buffer (36 mM citric acid, 5 mM glucose, 5 mM KCl, 1 mM MgCl 2 , 103 mM NaCl, 0.3 U/ mL apyrase, 0.1 mM PGE-1, 0.1% BSA, pH 6.5) were then added to platelets before a 12 min centrifugation step at 1200 g. Platelets were resuspended at 2.10 8 /mL in assay buffer supplemented with 2 mM CaCl 2, and distributed in 96-well flat-bottomed plates. Basal fluorescence of Oregon Green 488 BAPTA-1AM-loaded platelets was followed for 2 min and calcium-dependent changes in fluorescence intensity were then recorded for 5 min after adding the agonists using a 96-well plate fluorimeter (Fluoroskan Ascent FL, Thermo Labsystems, Dreieich, Germany).
## Elisa analysis of platelet releasates
At the end of the incubation of platelets with or without the agonists and after taking platelet samples for analysis in flow cytometry and microscopy, supernatants of the remaining platelets were collected after two centrifugation steps: 12 min, 1200 g at RT and 3 min, 13000 g at RT. Aliquots of supernatants were kept frozen at 280uC until analysis. The following ELISAs were used in accordance with the manufacturer's instructions: serotonin (GenWay, San Diego, CA, USA), platelet factor 4 (PF4) (RayBiotech, Norcross, GA, USA), TGF-b1 (Enzo Life Science, Villeurbanne, France) and angiopoietin-1 (Abcam, Paris, France).
## Measurement of atp levels
ATP levels were measured by chemiluminescence using the ATPLite Assay System (PerkinElmer, Boston, MA). ATPLite reagent was added directly to platelet suspensions or supernatants at a 1:10 dilution and luminescence was read using a SpectraFluor Plus plate reader (Tecan Austria Gmbh, Grodïg, Austria).
## Statistics
Data are expressed as means 6 standard error of the mean (SEM) and were compared by the non-parametric Mann-Whitney test using Kaleidagraph software (Synergy Software, Reading, PA). P-values,0.05 were regarded as statistically significant.
# Results
## Dose-response analysis of thrombin-and collageninduced platelet aggregation and shape change
To assess platelet activation and triggering of their thrombogenic phenotype by thrombin and collagen, we first analyzed aggregation and shape changes of human washed platelets in response to increasing concentrations of these two primary platelet agonists. Aggregation of stirred platelets was induced by thrombin at the highest concentration tested (5 nM), but not at the intermediate (0.5 nM) or low concentration (0.05 nM) (Figures 1A, C). To determine individual platelet morphological response, human washed platelets were incubated with the agonists without stirring and analyzed by DIC microscopy. All three thrombin concentrations provoked obvious changes in platelet shape, with membrane protrusions being already visible at the lowest thrombin concentration [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref]. In addition to intensifying the formation of membrane protrusions, increasing thrombin concentration further led to platelet swelling and emission of microvesicles [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref]. Thus, at low and intermediate doses of thrombin, despite the absence of an aggregative response, platelets already showed signs of activation at the individual level.
In response to collagen, platelet aggregation occurred at both the intermediate (5 mg/mL) and high concentration (25 mg/mL), but not at the low concentration (0.25 mg/mL) [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref]. Regarding platelet shape, collagen led to much more discrete changes than thrombin, with the sole induction of membrane protrusions at both intermediate and high concentrations, and no changes at the low concentration [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref].
## Phosphatidylserine exposure
Clot formation does not only rely on the adhesive and aggregative properties of activated platelets but also on their ability to support coagulation, notably through the exposure of phosphatidylserine at their surface [bib_ref] The nature of the binding for prothrombinase at the platelet surface as..., Bevers [/bib_ref]. Therefore, to further evaluate the thrombogenic phenotype and activity of human platelets exposed to various concentrations of collagen or thrombin, we compared their level of surface phosphatidylserine to that of resting platelets. Annexin V binding to platelets, a well established marker of phosphatidylserine surface exposure [bib_ref] Annexin V as a probe of aminophospholipid exposure and platelet membrane vesiculation:..., Dachary-Prigent [/bib_ref] , was not modified by low (0.05 nM) concentration of thrombin [fig_ref] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets [/fig_ref]. However, at the highest concentration of thrombin (5 nM), which caused platelet aggregation (Figures 1A, C) and important shape changes [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref] , a significant increase in annexin V binding to platelets was observed, thus indicating the induction of phosphatidylserine externalisation [fig_ref] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets [/fig_ref]. A modest but statistically significant increase in annexin V binding was also observed in response to the intermediate dose of thrombin (0.5 nM). Regarding collagen-stimulated platelets, annexin V binding to platelet was also observed at concentrations inducing aggregation (5 and 25 mg/mL) but not at 0.25 mg/mL (Figures 2A, B).
## P-selectin surface expression
P-selectin, a transmembrane protein anchored in the membrane of platelet alpha-granules, is translocated to the platelet surface during activation-associated alpha-granule secretion [bib_ref] PADGEM protein: a receptor that mediates the interaction of activated platelets with..., Larsen [/bib_ref] [bib_ref] GMP-140 mediates adhesion of stimulated platelets to neutrophils, Hamburger [/bib_ref]. For this reason, measurement of P-selectin surface expression by flow cytometry is widely used as a marker of platelet activation and degranulation, and this, both in clinical and experimental settings [bib_ref] Properties of GMP-140, an inducible granule membrane protein of platelets and endothelium, Mcever [/bib_ref] [bib_ref] PADGEM, a leukocyte receptor on activated platelets. Biology and application to in..., Furie [/bib_ref] [bib_ref] Quantification of platelet activation status by analyzing P-selectin expression, Leytin [/bib_ref]. When assessing platelet activation status and alphagranule secretion by this method, we observed that P-selectin surface expression was induced by thrombin at the highest dose of 5 nM, and also, though much more slightly, at the intermediate dose of 0.5 nM . Collagen significantly increased Pselectin exposure only at the highest dose (25 mg/mL) (Figures 3A-B). These results were confirmed by immunostaining. A granule-like punctiform staining was observed on control resting platelets . The same punctiform staining was obtained for platelets treated with 0.25 and 5 mg/mL collagen . In contrast, platelets that had been activated by 5 nM thrombin showed a clear peripheral distribution of P-selectin. An ''intermediate'' staining pattern was observed for platelets activated by 0.5 nM thrombin or 25 mg/mL collagen, with a more diffuse staining and a loss of granular pattern .
## Secretion of dense and alpha granule-derived bioactive agents
In addition to provoking platelet shape change, aggregation and procoagulant activity, platelet activation is also known to induce the secretion of a multitude of soluble bioactive agents from platelet storage granules. Therefore, in order to further characterize the activation status of platelets exposed to various doses of thrombin and collagen, we measured and compared the release of soluble factors by control and agonist-stimulated platelets. Supernatants recovered from the very same platelet samples that were analyzed in flow cytometry and in DIC and fluorescence microscopy were then analyzed for the presence of platelet-derived soluble factors. For this analysis, we focused on factors for which platelets represent the main source in blood: serotonin and ATP, which are stored in platelet dense granules [bib_ref] Platelet granule secretion continuously prevents intratumor hemorrhage, Ho-Tin-Noe [/bib_ref] [bib_ref] Platelet dense granules: structure, function and implications for haemostasis, Mcnicol [/bib_ref] , and PF4, TGF-b1, and angiopoietin-1, all of which are abundantly stored in platelet alpha-granules [bib_ref] Platelet granule secretion continuously prevents intratumor hemorrhage, Ho-Tin-Noe [/bib_ref] [bib_ref] Transforming growth factor-beta in human platelets. Identification of a major storage site,..., Assoian [/bib_ref] [bib_ref] Thrombin induces the release of angiopoietin-1 from platelets, Li [/bib_ref] [bib_ref] Interactions of platelet factor 4 with the vessel wall, Sachais [/bib_ref] [bib_ref] Platelet TGF-beta1 contributions to plasma TGF-beta1, cardiac fibrosis, and systolic dysfunction in..., Meyer [/bib_ref]. The release of serotonin by platelets was increased in a dose-dependent manner by thrombin with statistical significance being reached already at the lowest concentration tested (0.05 nM) [fig_ref] Figure 4: Dose-response relationship between agonist concentration and secretion of dense and alpha-granule-derived bioactive... [/fig_ref]. In contrast, there was no real dose-response effect for collagen. In fact, collagen at the lowest dose (0.25 mg/mL) significantly and substantially stimulated serotonin secretion with no notable further increase at higher doses [fig_ref] Figure 4: Dose-response relationship between agonist concentration and secretion of dense and alpha-granule-derived bioactive... [/fig_ref]. Notably, the response profiles of platelets to thrombin and collagen were confirmed when dense granule secretion was assessed by measurement of ATP levels [fig_ref] Figure 4: Dose-response relationship between agonist concentration and secretion of dense and alpha-granule-derived bioactive... [/fig_ref]. For PF4, TGF-b1, and angiopoietin-1, stimulation of platelets with thrombin and collagen led to secretion patterns comparable to those obtained for serotonin [fig_ref] Figure 4: Dose-response relationship between agonist concentration and secretion of dense and alpha-granule-derived bioactive... [/fig_ref] , with a clear doseresponse effect for thrombin and an almost maximum effect of collagen from the lowest dose used.
Taken together, these results suggest that collagen at low dose can efficiently induce the release of dense-and alpha granulederived soluble factors without triggering aggregation [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref] , phosphatidylserine exposure [fig_ref] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets [/fig_ref] , or P-selectin surface expression . However, because centrifugations were required for the preparation of platelet supernatants while all other parameters were measured in the absence or before centrifugation, we investigated whether the release of soluble factors by platelets in response to low-dose collagen in the absence of any other notable changes could be linked to this technical issue. To recover platelet releasates without the need for centrifugation, experiments using transwell systems (0.4 mm pore size) were performed. Platelets were placed and stimulated into the upper chamber (transwell insert) and, at the end of the incubation, the cell-free medium containing diffusible secretion products was directly collected from the lower chamber while platelets were recovered from the upper chamber for flow cytometric analysis. In these conditions, the results from flow cytometry analysis of Pselectin and phosphatidylserine expression were identical to those presented in [fig_ref] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets [/fig_ref] and 3 (and therefore not shown). As for the release of soluble factors, while the response pattern of platelets to thrombin remained unchanged (not shown), that to collagen, on the other hand, was appreciably modified. Indeed, although in these conditions collagen did induce a statistically significant Representative flow cytometry histograms of annexin V binding to non-stimulated resting (control) platelets or to platelets incubated with various concentration of thrombin or collagen, as indicated above the charts. The histogram obtained for control platelets was superimposed in grey to that of stimulated platelets for better visualization of fluorescence shifts. B. Bar graph representing the mean annexin V binding to non-treated control platelets (CTL) and to thrombin-or collagen-stimulated platelets. Results are expressed as percent relative to the mean fluorescence intensity of CTL platelets. n = 4 different blood donors, # indicates a significant statistical difference (p,0.05) from CTL platelets. doi:10.1371/journal.pone.0104712.g002 increase in the release of dense and alpha granule-derived soluble factors from the lowest concentration tested (0.25 mg/mL), a dosedependent profile was observed [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref] , contrasting with the maximal effect already obtained at the lowest collagen dose when analysing centrifugation supernatants [fig_ref] Figure 4: Dose-response relationship between agonist concentration and secretion of dense and alpha-granule-derived bioactive... [/fig_ref]. This dosedependent profile was further confirmed by experiments in which platelets were stimulated under stirring conditions. In these experiments, the release of soluble factors from dense and alpha granules was assessed directly in platelet suspensions. Dense granule secretion was estimated by measuring ATP levels using a luminescence-based assay and for quantification of alpha granule secretion, we developed a microspheres-based flow cytometric immunoassay. Stimulation of stirred platelets with collagen induced a dose-dependent increase in the release of ATP and PF4, with statistical significance being reached from the lowest collagen dose . Again, the release of these soluble factors in response to 0.25 mg/mL collagen occurred in the absence of significant P-selectin or phosphatidylserine surface exposure, and of integrin activation as assessed by PAC-1 binding to platelets . Such a secretory response was not obtained for collagen at 0.125 mg/mL (not shown). Notably, and in contrast to what was observed on platelets stimulated in static conditions [fig_ref] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets [/fig_ref] , a significant increase in P-selectin surface expression was induced by collagen at 0.5 mg/mL , a concentration that induced platelet aggregation [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref]. Taken together, these results show that collagen at low dose can efficiently induce the release of dense-and alpha granule-derived soluble factors without triggering aggregation, phosphatidylserine exposure, or P-selectin surface expression. They further show that this secretory response is exaggerated by centrifugation of platelets.
## Role of calcium and gpvi signaling in low-dose collageninduced platelet secretion
Activation of platelets by most stimulatory agents leads to a rise in the concentration of cytosolic Ca 2+ which can elicit various platelet responses including shape change, spreading, adhesion, aggregation, procoagulant activity, and granule secretion [bib_ref] Novel molecules in calcium signaling in platelets, Bergmeier [/bib_ref]. Therefore, we verified whether or not stimulation of platelets with low-dose collagen (0.25 mg/mL), which led to secretion of platelet soluble factors in the absence of other notable changes [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref] , was associated with an increase in cytosolic Ca 2+ . Collagen at intermediate and high concentration triggered an initial continuous rise followed by oscillations in intraplatelet Ca 2+ while thrombin at intermediate and high concentration provoked a steep and transient calcium peak [fig_ref] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets [/fig_ref]. In contrast, no change in intraplatelet Ca 2+ levels occurred after addition of collagen at 0.25 mg/mL . Similar absence of calcium was also observed when stimulating platelets with 0.05 nM thrombin [fig_ref] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets [/fig_ref] , a concentration for which the only notable changes observed were a slight shape change and a discrete release of serotonin [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref]. These results show that low-dose collagen-induced platelet activation is not associated with a substantial rise in intracellular Ca 2+ .
Since GPVI is the major collagen receptor on platelets and mediates collagen-induced platelet activation in regard to platelet aggregation and procoagulant activity, we assessed the contribution of GPVI-signalling to low-dose collagen-induced secretion of platelet soluble factors. For this, platelets were stimulated with 0.25 mg/mL collagen in the presence or absence of a GPVI blocking Fab fragment antibody (Fab 9O12, 50 mg/mL) and the levels of serotonin, TGF-b1, and angiopoietin-1 were then measured in platelet supernatants. For all three factors, the secretory response of platelets to low-dose collagen was completely inhibited by blocking of GPVI , thus demonstrating its dependency on GPVI signalling.
# Discussion
In the present study, we investigated whether or not platelets could display state(s) of activation different from that known to support their role in primary hemostasis and thrombosis. To answer this question, we analyzed and compared the changes in platelet phenotype caused by various doses of thrombin and collagen which represent the two strongest physiological primary agonists of platelets. We found that collagen at low dose (0.25 mg/ mL) selectively triggers a platelet secretory response without causing any of the other major changes that usually accompany platelet activation and that were observed at higher collagen concentrations. More precisely, stimulation of platelets with lowdose collagen led to significant secretion of dense-and alpha granule-derived soluble factors, and this without causing platelet aggregation or surface exposure of phosphatidylserine and Pselectin. Notably, such a selective secretory response of platelets was not observed when thrombin was used as an agonist. In fact, thrombin at low dose only slightly increased serotonin secretion by platelets and did not lead to secretion of alpha granule-derived PF4, TGF-b1, or angiopoietin-1. Furthermore, this secretory response was associated with a clear induction of platelet shape change. Although at higher concentrations thrombin did induce a significant release of both dense and alpha granule-derived soluble factors, this response was associated with one or more additional platelet activation-associated changes. Taken together, these results show that as previously described for P-selectin and phosphatidylserine exposure [bib_ref] Annexin V as a probe of aminophospholipid exposure and platelet membrane vesiculation:..., Dachary-Prigent [/bib_ref] [bib_ref] Collagen but not fibrinogen surfaces induce bleb formation, exposure of phosphatidylserine, and..., Heemskerk [/bib_ref] , selective triggering of the platelet secretory response is dependent on both the agonist used and its concentration. In addition to support the idea that the response pattern of activated platelets is context-related, our data indicates that platelet activation includes situations in which the secretory activity of platelets is dissociated from their ability to form thrombi.
Interestingly, we found that the release of various alpha granulederived soluble factors by platelets exposed to low-dose collagen was not associated with an increase in the surface expression of Pselectin, a transmembrane protein widely known to reside in the alpha granule membrane. This discrepancy between the secretion pattern of membrane P-selectin and alpha granule-derived soluble factors might have several implications. First, it might reflect differences in the kinetics of secretion between soluble and transmembrane factors that would however be released from similar granules. Results from a recent study by Jonnalagadda et al. have shown that molecules stored in presumably similar granules could display differences in their rate of secretion due to . Dose-response analysis of thrombin-and collagen-induced P-selectin surface expression. A. Representative flow cytometry histograms of P-selectin surface expression of non-treated (control) and thrombin-or collagen-treated platelets, as indicated. The histogram obtained for control platelets was superimposed in grey to that of stimulated platelets for better visualization of fluorescence shifts. B. Bar graph representing the mean percentage of P-selectin-positive cells of non-treated control (CTL) and thrombin-or collagen-stimulated platelets. n = 8 independent experiments with different blood donors, # indicates a significant statistical difference (p,0.05) from CTL. C. Immunolocalization of P-selectin (green) in permeabilized control and agonist-stimulated platelets. Merged images of DIC and green fluorescence are shown. Bar = 5 mm. doi:10.1371/journal.pone.0104712.g003 possible differences in solubility, granule shape, and/or granuleplasma membrane fusion routes [bib_ref] Platelet secretion is kinetically heterogeneous in an agonist-responsive manner, Jonnalagadda [/bib_ref]. One can then imagine that it might be easier for a soluble factor to diffuse out of the open canalicular system (OCS) into the extracellular medium than for a transmembrane protein to migrate from the OCS to the surface within the membrane lipid bilayer. Moreover, immunolocalisation of P-selectin in low-dose collagen-treated platelets revealed a granular staining pattern similar to that of untreated platelets. Therefore, the absence of surface P-selectin on low-dose collagenstimulated platelets might just reflect the need for platelet contraction and shape change for P-selectin to be externalized together with the OCS. Also, another possibility is the existence of various subtypes of and/or subdomains in alpha granules whose cargo molecule secretion would be differentially regulated in a dose-and agonist-dependent manner. Several studies have shown that differential sorting of alpha granule cargo in distinct subtypes or domains of alpha granules can result in their differential release [bib_ref] Angiogenesis is regulated by a novel mechanism: pro-and antiangiogenic proteins are organized..., Italiano [/bib_ref] [bib_ref] Evidence that differential packaging of the major platelet granule proteins von Willebrand..., Sehgal [/bib_ref] [bib_ref] Selective sorting of alpha-granule proteins, Italiano [/bib_ref] [bib_ref] Quantitative immunofluorescence mapping reveals little functional coclustering of proteins within platelet alpha-granules, Kamykowski [/bib_ref] [bib_ref] Distinct platelet packaging, release, and surface expression of proangiogenic and antiangiogenic factors..., Chatterjee [/bib_ref]. In this context, it is noteworthy that the maximum effect of thrombin on serotonin and angiopoietin-1 secretion was higher than that of collagen, showing that thrombin can specifically increase the secretion of these factors. This could be achieved by complete versus partial emptying of similar alpha granules, or by mobilization of additional and/or different alpha granules by thrombin as compared to collagen. All in all, these differences in the release of diffusible factors in response to various stimuli that all cause maximal aggregation reinforces the idea that platelet activation is far from being univocal [bib_ref] Hierarchical organization in the hemostatic response and its relationship to the platelet-signaling..., Stalker [/bib_ref] [bib_ref] Platelet secretion is kinetically heterogeneous in an agonist-responsive manner, Jonnalagadda [/bib_ref].
Considering the fact that platelets have been shown to exert thrombus-independent regulatory actions in various inflamed organs and solid tumors [bib_ref] Inflammation induces hemorrhage in thrombocytopenia, Goerge [/bib_ref] [bib_ref] Platelet granule secretion continuously prevents intratumor hemorrhage, Ho-Tin-Noe [/bib_ref] [bib_ref] Innate immune cells induce hemorrhage in tumors during thrombocytopenia, Ho-Tin-Noe [/bib_ref] , our findings bring new mechanistic insights that might partly explain these effects. Previous studies have suggested the implication of platelet granule content in the prevention of tumor hemorrhage by platelets, possibly through regulation of vascular permeability and immune cell activities [bib_ref] Platelet granule secretion continuously prevents intratumor hemorrhage, Ho-Tin-Noe [/bib_ref] [bib_ref] Innate immune cells induce hemorrhage in tumors during thrombocytopenia, Ho-Tin-Noe [/bib_ref] [bib_ref] Platelets: guardians of tumor vasculature, Ho-Tin-Noe [/bib_ref]. Also, platelet-derived serotonin has been shown to mediate liver regeneration [bib_ref] Platelet-derived serotonin mediates liver regeneration, Lesurtel [/bib_ref]. We show here that collagen at low dose leads platelets to release serotonin, PF4, angiopoietin-1, and TGF-b1, all of which have immunomodula-tory properties and/or the ability to regulate endothelial permeability. Moreover, it was shown recently that the vasculoprotective action of platelets in inflamed organs is dependent on ITAM receptors including GPVI [bib_ref] Platelet ITAM signaling is critical for vascular integrity in inflammation, Boulaftali [/bib_ref] , a receptor which we found to mediate the secretory response of platelets to low-dose collagen.
Furthermore, we observed that this selective secretory response of platelets to low-dose collagen was not associated with a sustained rise in intracellular calcium. In a similar manner, the prevention of inflammatory bleeding by platelets was shown previously to be independent of the calcium and diacylglycerolregulated guanine nucleotide exchange factor 1 (CalDAG-GEF1) [bib_ref] Inflammation induces hemorrhage in thrombocytopenia, Goerge [/bib_ref] [bib_ref] Platelet granule secretion continuously prevents intratumor hemorrhage, Ho-Tin-Noe [/bib_ref] [bib_ref] Platelet ITAM signaling is critical for vascular integrity in inflammation, Boulaftali [/bib_ref] , a molecule which is central to calcium signalling in platelets [bib_ref] CalDAG-GEFI is at the nexus of calcium-dependent platelet activation, Stefanini [/bib_ref]. Whether or not the collagen-induced platelet secretory phenotype we described here does actually contribute to the prevention of inflammatory bleeding by platelets however remains to be verified. The absence of substantial rise in . Comparison of soluble factor release and surface activation marker expression by stirred platelets exposed to collagen. A. ATP levels in suspensions of stirred platelets as determined using a luminescence-based assay. B. PF4 levels in suspensions of stirred platelets as determined using a fluorescent microspheres-based flow cytometric immunoassay. C-E. Surface exposure of P-selectin (C) and phosphatidylserine (D), as well as binding of the PAC-1 antibody to activated GPIIb/IIIa (E) on stirred platelets were measured by flow cytometry. All bar graphs represent means calculated from a minimum of 4 independent experiments performed with different blood donors. # indicates a significant statistical difference (p,0.05) from CTL. doi:10.1371/journal.pone.0104712.g005 intracellular calcium might also indicate that basal calcium levels are sufficient to promote the secretory response induced by lowdose collagen. Studies of secretory cells such as adrenal chromaffin cells and neurons have shown that secretion can occur at low intracellular calcium levels through a mechanism of incomplete and transient fusion of secretory vesicles with the plasma membrane. The shift from this so-called ''kiss-and-run'' mechanism to full exocytosis occurs when cell stimulation is accompanied by high intracellular calcium levels [bib_ref] Double patch clamp reveals that transient fusion (kiss-and-run) is a major mechanism..., Elhamdani [/bib_ref] [bib_ref] The debate on the kiss-and-run fusion at synapses, He [/bib_ref]. A kiss-and-run mechanism, for which evidence of its existence in platelets has been brought by single-cell amperometry studies [bib_ref] Platelet granule exocytosis: a comparison with chromaffin cells, Fitch-Tewfik [/bib_ref] [bib_ref] Quantal regulation and exocytosis of platelet dense-body granules, Ge [/bib_ref] , could thus explain the secretory response of platelets to low-dose collagen while full collapse fusion with externalization of P-selectin would occur at higher agonist concentrations associated with a marked increase in intracellular calcium.
In conclusion, our results show that platelet activation goes beyond the mechanisms leading to thrombus formation and also includes alternative platelet phenotypes that might contribute to their thrombus-independent functions. Also, the fact that GPVI can mediate the release of potent immunomodulators and regulators of vascular permeability without evoking platelet aggregation or procoagulant activity provides a new conceptual view on GPVI/collagen interactions that reconciles GPVI with its immune receptor nature. [fig_ref] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change [/fig_ref] Secretory response of platelets to collagen in static conditions. To assess the release of soluble factors by platelets in response to collagen independently of mechanical agitation, platelets were placed and stimulated into transwell inserts (pore diameter 0.4 mm), and the cell-free medium containing diffusible secretion products was collected from the lower chamber. Levels of serotonin (A), ATP (B), and platelet factor 4 (PF4) (C) in platelet releasates were measured as reflects of dense and alpha granule secretion, respectively. Results are expressed as percent relative to the mean levels found in releasates of unstimulated control platelets. n = at least 6 different blood donors, # indicates a significant statistical difference (p,0.05) from unstimulated control platelets. Raw Data S1
## Supporting information
(XLS) . Low dose collagen-induced platelet secretion requires GPVI signalling. A. Intracellular calcium levels in collagenstimulated platelets as assessed by measurement of calcium-dependent fluorescence of Oregon Green 488 BAPTA-1AM-loaded platelets. The calcium curves shown are representative of four independent experiments using different blood donors. B-D. To assess the contribution of GPVI-signalling to low-dose collagen-induced secretion of platelet soluble factors, platelets were stimulated with 0.25 mg/mL collagen in the presence or absence of a GPVI blocking Fab fragment antibody (Fab 9O12, 50 mg/mL) and the levels of serotonin (B), TGF-b1 (C), and angiopoietin-1 (D) were then measured in platelet supernatants. All bar graphs represent the mean levels from a minimum of 4 independent experiments. NS: non-statistically significant. doi:10.1371/journal.pone.0104712.g006
[fig] Figure 1: Dose-response analysis of thrombin-and collagen-induced platelet aggregation and shape change. The aggregation of human washed platelets in response to various concentrations of thrombin or collagen was analyzed by turbidimetry under stirring conditions in the absence of exogenous fibrinogen. In parallel, the individual morphological changes induced by the same concentrations of agonists were analyzed by DIC microscopy. A-B. Diagrams showing representative dose-response aggregation tracings obtained for thrombin (A) and collagen (B). CTL corresponds to non-treated resting platelets. C. Bar graph showing the mean maximum platelet aggregation in response to various doses of thrombin or collagen. # indicates a significant statistical difference (p,0.05) from non-stimulated resting human washed platelets (CTL). n = 5 different blood donors. D-E. DIC microscopy images of the morphological aspect of platelets after incubation with or without the indicated concentrations of thrombin (D) or collagen (E). The images shown are representative of 6 independent experiments performed with different blood donors. Scale bars = 5 mm. White arrowheads indicate platelet membrane protrusions, black arrows indicate microvesicles. doi:10.1371/journal.pone.0104712.g001 [/fig]
[fig] Figure 2: Dose-response analysis of thrombin-and collagen-induced phosphatidylserine exposure in human platelets. Phosphatidylserine exposure by human washed platelets was quantified by measurement of fluorescent annexin V binding in flow cytometry. A. [/fig]
[fig] Figure 4: Dose-response relationship between agonist concentration and secretion of dense and alpha-granule-derived bioactive agents. The release of dense and alpha-granule components by human platelets in response to various concentrations of thrombin and collagen was assessed and compared by measuring serotonin (A), ATP (B), platelet factor 4 (PF4) (C), TGF-b1 (D), and angiopoietin-1 (E) levels in platelet supernatants. All bar graphs represent the mean levels calculated from 6 independent experiments performed with different blood donors. # indicates a significant statistical difference (p,0.05) from CTL. doi:10.1371/journal.pone.0104712.g004 [/fig]
[fig] Figure: S2 Intracellular calcium levels in thrombinstimulated platelets. Calcium levels of Oregon Green 488 BAPTA-1AM-loaded platelets treated with thrombin were measured using a fluorescence microplate reader. The calcium curves shown are representative of four independent experiments using different blood donors. (TIF) [/fig]
|
Recent Progress in Microfluidic Models of the Blood-Brain Barrier
The blood-brain barrier (BBB) is a critical physical and chemical barrier that maintains brain homeostasis. Researchers in academia and industry are highly motivated to develop experimental models that can accurately mimic the physiological characteristics of the BBB. Microfluidic systems, which manipulate fluids at the micrometer scale, are ideal tools for simulating the BBB microenvironment. In this review, we summarized the progress in the design and evaluation of microfluidic in vitro BBB models, including advances in chip materials, porous membranes, the use of endothelial cells, the importance of shear stress, the detection specific markers to monitor tight junction formation and integrity, measurements of TEER and permeability. We also pointed out several shortcomings of the current microfluidic models. The purpose of this paper is to let the readers understand the characteristics of different types of model design, and select appropriate design parameters according to the research needs, so as to obtain the best experimental results. We believe that the microfluidics BBB models will play an important role in neuroscience and pharmaceutical research.
# Introduction
The blood-brain barrier (BBB) is a highly selective barrier that regulates passive and active transport between the brain parenchyma and peripheral blood. The BBB plays a vital role in maintaining the physical and chemical homeostasis of the central nervous system (CNS), and protects the CNS from harmful molecules and pathogens in the blood. The BBB is a complex dynamic physiological structure network, the molecular basis that underlie BBB function depend on close interactions between adjacent brain microvascular endothelial cells (BMECs). These cells form a layer that is tightly sealed by a junctional complex composed of tight junctions and adherens junctions. Tight junctions, which play an essential role in maintaining BBB integrity, are structures formed by at least three different types of transmembrane proteins, such as occludin, claudin and junctional adhesive molecule. BMECs, along with neurons, pericytes, astrocytes, microglial cells, and extracellular matrix, constitute a functional network known as the neurovascular unit,show the cellular constituents and transport pathways of the BBB. Due to the tight barrier, a number of drugs are unable to penetrate the BBB and produce therapeutic effects in the CNS. Disruption or dysfunction of the BBB has been linked to a wide range of neurological disorders, including brain tumors, epilepsy, ischemic stroke, Alzheimer's disease, and multiple sclerosis.
## Figure 1.
Cellular constituents of the blood-brain barrier (BBB). The BBB is formed by brain microvascular endothelial cells (BMECs), which are connected by tight junctions. The endothelium, together with the basal lamina, pericytes, and astrocytic end-feet forms the neurovascular unit. Some substances diffuse freely into and out of the brain parenchyma, others such as nutrients need specific transporters, while molecules such as insulin, leptin and transferrin are transported by receptor-mediated transcytosis. (Reprinted from Referencein accordance with the Creative Commons Attribution).
In order to gain in-depth understanding of the BBB structural and functional properties, to access the permeability of different drugs, chemicals and compounds through the BBB, studies of the BBB have become a high priority for pharmaceutical companies and biological researchers. Due to the complex structures, it is very important to establish suitable models in research. In vivo animal models, in vitro cell-based models, and computer models are used to explore the BBB structures and functions. In vivo animal models can reflect the real state of the microenvironment, and are considered as the golden standard in research, but animal experiments are time-consuming and costly. Transwell chambers are the most widely used in vitro cell-based models, transwell models are easy to operate and reproducible, but they are static culture models and cannot reflect fluid characteristics of the BBB.
As one of the most advanced technologies, microfluidic chips use microfabrication technology to produce all kinds of micron-scale chips. In the microfluidic chips, microchannels networks are formed. By regulating the flow of liquid in microchannel, biological, and chemical analysis are performed in the chips. In recent years, many researchers use microfluidic chip to construct the BBB models and achieved good experimental results. In this review manuscript, we discussed the technical and operational details in model design, fabrication, evaluation and application, so that we can choose appropriate models and parameters according to the research purposes in future studies.
## Current experimental blood-brain barrier (bbb) models
There is great interest in developing experimental models that mimic the well-organized and unique properties of the BBB. Ideal BBB models should have the same cell types and distribution compared with in vivo structure, can well express enzymes, receptors and transporters, can well simulate the material transport process and pathway, have high selective permeability to different substances, and have high trans-epithelial electric resistance (TEER) values. Here, we will introduce various models in the BBB research. . Cellular constituents of the blood-brain barrier (BBB). The BBB is formed by brain microvascular endothelial cells (BMECs), which are connected by tight junctions. The endothelium, together with the basal lamina, pericytes, and astrocytic end-feet forms the neurovascular unit. Some substances diffuse freely into and out of the brain parenchyma, others such as nutrients need specific transporters, while molecules such as insulin, leptin and transferrin are transported by receptor-mediated transcytosis. (Reprinted from Referencein accordance with the Creative Commons Attribution).
In order to gain in-depth understanding of the BBB structural and functional properties, to access the permeability of different drugs, chemicals and compounds through the BBB, studies of the BBB have become a high priority for pharmaceutical companies and biological researchers. Due to the complex structures, it is very important to establish suitable models in research. In vivo animal models, in vitro cell-based models, and computer models are used to explore the BBB structures and functions. In vivo animal models can reflect the real state of the microenvironment, and are considered as the golden standard in research, but animal experiments are time-consuming and costly. Transwell chambers are the most widely used in vitro cell-based models, transwell models are easy to operate and reproducible, but they are static culture models and cannot reflect fluid characteristics of the BBB.
As one of the most advanced technologies, microfluidic chips use microfabrication technology to produce all kinds of micron-scale chips. In the microfluidic chips, microchannels networks are formed. By regulating the flow of liquid in microchannel, biological, and chemical analysis are performed in the chips. In recent years, many researchers use microfluidic chip to construct the BBB models and achieved good experimental results. In this review manuscript, we discussed the technical and operational details in model design, fabrication, evaluation and application, so that we can choose appropriate models and parameters according to the research purposes in future studies.
## Current experimental blood-brain barrier (bbb) models
There is great interest in developing experimental models that mimic the well-organized and unique properties of the BBB. Ideal BBB models should have the same cell types and distribution compared with in vivo structure, can well express enzymes, receptors and transporters, can well simulate the material transport process and pathway, have high selective permeability to different substances, and have high trans-epithelial electric resistance (TEER) values. Here, we will introduce various models in the BBB research.
## In vivo models
Researchers first used in vivo BBB models to perform experiments in living organisms in a normal and intact state. In these models, function is studied using methods such as intravenous injection, brain perfusion, positron emission tomography, and microdialysis sampling. The main advantage of in vivo models is that they provide an experimental environment that closely mimics the complexities of human physiology. The entire experiment occurs under natural conditions and can generate large amounts of reliable data. However, no animal model can faithfully reproduce all of the manifestations of human disease, and thus these models must be interpreted as an approximation of human biology, limited to particular regions or other features. The most important disadvantage of in vivo models is the difficulty in translating the results they generate to the human context. More than 80% of results obtained from animal models do not correspond in a straightforward way to human responses. Animal-to-animal variability is another problem. In addition, in order to observe the whole process of disease development, separate animals must be used at different stages; making experiments expensive due to labor and animal costs. Finally, during in vivo experiments, high doses of chemicals are often used. These doses are not suitable in high throughput screens for drug discovery.
## In vitro models
In vitro models include cell culture models, brain slice models, fiber-based dynamic in vitro BBB (DIV-BBB) models, and microfluidic models.
In vitro cell culture models based on cultured cells have been used for decades to study various mechanisms that support the BBB physiology. By using the cells such as BMECs, astrocytes, and pericytes, experiments can be conducted under carefully controlled conditions. Cell culture technology is simple, with good reproducibility, and suitable for high-throughput screening. As a result, cell culture models have become important tools in the BBB research.
The most common in vitro models are the transwell models, in which one or more cell types are cultured on semi-permeable microporous inserts. According to the cell types used in the experiment, it can be divided into single endothelial cell monolayer model and co-culture model. The endothelial cells monolayer model is the simplest transwell model, but it lacks the interaction with other cell types, which is crucial for the properties and functions of the BBB. In co-culture experimental models, BMECs and other cells are cultured on different sides of the membrane. Usually endothelial cells are cultured on the upper side to form luminal layer. Other cells, such as astrocytes and/or pericytes, are cultured on the lower compartment to form abluminal layer.
From a practical point of view, transwell models are user friendly and cost-effective. They make it possible to easily manipulate experimental conditions, such as temperature and compounds concentration. The systems can reduce experimental animal numbers, increase test speeds, cut down reagent and chemical consumption, thus facilitating moderate-throughput drug permeability screening. However, transwell models are usually unable to replicate key characteristics of the BBB. First, the endothelial cells are not subjected to dynamic mechanical stimuli that can cause subtle differences in cell morphology and barrier permeability when compared with in vivo models. Second, transwell models are simplified co-culture systems that do not capture the complex architecture of the BBB.
Brain slice models are also used in the BBB studies. In these models, organotypic hippocampal slices are cultured on a membrane surface and used to study the BBB functions under different physiological and pathological conditions. All cell types and interactions are present in brain slices, so these systems provide complete structures that are excellent tools for biological and pharmacological research. However, neither fluorescent immunostaining of biomarkers, nor measurement of TEER, is convenient in slice models.
Transwell models and slice models are static models because there is no fluid flow in these systems. Fluid flow is thought to subject endothelial cells to shear forces, and these are critical for proper endothelial polarization and tight junction formation. Without dynamic flows, the integrity of the barrier is affected. In static models, the TEER is much lower than in vivo models. In vivo resistance across the BBB is about 1500-8000 Ω·cm 2 , while the TEER of static in vitro models typically reaches only 150-200 Ω·cm 2. These models also exhibit high permeability to marker molecules, and suffer from low expression and impaired functionality of P-glycoprotein efflux pump transporters, which dramatically limits their utility for drug screening.
Recently, DIV-BBB models have been developed to mimic dynamic flow. In these models, endothelial cells and other cells are cultured on the inner and outer walls of hollow fibers. Culture medium is pulsated through the tube, generating shear stress on the surface of the endothelial cells. Due to the shear stress, tight junction protein expression is increased. In these models higher TEER values and lower permeability are obtained, and the systems can better imitate the BBB environment. In spite of these advantages, the DIV-BBB model has not yet been widely used in experiments due to several shortcomings. First, nine to 12 days are required to reach the maximum TEER value. Second, it is difficult to observe changes in the endothelial cells during the course of an experiment, and can only be inferred by consumption of glucose or production of lactic acid. Third, the thickness of the fiber wall (150 µm) is much larger than that of the porous membrane (10 µm), reducing the contact between endothelial cells and other cells.
In recent years, microfluidic chips have been used to construct the BBB structures, and have become increasingly popular. We will discuss the progress of these highly promising in vitro models later.
## Computer models
With computer aided drug design and delivery as a starting point, computer models have also been developed in the BBB research. Computer models typically build quantitative structure-activity relationships for the BBB permeation, and predict drug permeability according to the physical and chemical properties of compounds, such as van der Waals volume, topological polar surface area, lipophilicity, hydrogen-bond donator and acceptor. Computer models are generally applied to small molecular compounds and are excellent for high throughput drug studies. However, they are applicable only to permeability prediction based on drug structure, and cannot provide information about the effectiveness of a drug or whether the drug itself affects permeability of the BBB. Nearly all results obtained using computer simulation must be verified by in vivo experiments.
## Microfluidic in vitro models
Recent advances in micro-electro-mechanical systems (MEMS) technology make it possible to create microfluidic devices to mimic biological microenvironments in vitro. Microfluidics is considered both a science and a technology that focuses on the behavior and manipulation of fluids at the micrometer scale. Microfluidic chips require micro-scale engineering technologies to construct channels, chambers, and valves on silicon, glass, quartz, or macromolecule polymer material to form sub-micrometer sized mechanical channel structures. We can perform precise and complicated operations with sub-millimeter scale fluids as desired in these microstructures using micropumps or microvalves. Various functional units can also be integrated to carry out purification, separation, and detection, making it possible to perform a series of experiments and analyses on a single chip. A microfluidic chip with these features is also called a micro-total-analytical system or a lab-on-a-chip device. Microfluidic chips have potential applications in chemistry, physics, biology, medicine, and other disciplines.
The channels in microfluidic chips have similar width scales in comparison to biological systems, such as cells, macromolecules, and small organisms. Through simple or sophisticated channel design, it is possible to control the flow in microfluidic channels conveniently and accurately. Microfluidic technologies have recently been used to create three-dimensional (3D) cell co-culture devices and organ-on-a-chip systems, setting the stage for simulating the activities and responses of entire organs and organ systems under physiological or pathological conditions. In 3D cell culture microfluidic chips, perfusion-based media are used to supply nutrients to the cells and remove metabolic wastes. By precisely controlling the physical and chemical microenvironment around the cells, the systems are well suited for studying biological interactions down to the cell and molecular levels, and have great potential for investigating cell-cell interactions. In recent years, with notable developments having been made in areas of bio-printing, researchers have increasingly applied bio-printing technology to the processing of microfluidic chips. As a digital manufacturing technology, bio-printing uses computer-aided design (CAD) software to accurately construct complex objects, and adds materials to 3D model data layer by layer to create microfluidic chips. Compared with conventional microfluidic chip fabrication methods, bio-printing can create more complex, uniform and reproducible architectures. Microfluidic devices are evolving rapidly, and it is now possible to design microphysiologic organ-on-a chip models for guts, lungs, kidney, and other organs.
Since the microfluidic platform offers precise control of fluid transport at microscale dimensions and makes multifunctional integration possible, the microfluidic platform has been used to simulate the BBB microenvironment in various experiments. Microfluidic models clearly provide a promising platform for studying the mechanism of the BBB and evaluating CNS drugs.
## Design of microfluidic models
Most microfluidic models use porous membrane segmentation to form sandwich structures in the chip that are similar to those used in transwell systems. Endothelial cells and other cells are cultured on two sides of a membrane that is placed at the interface of two microchannels to form a neural chamber and a vascular chamber. Other researchers also used micro-gaps, trapezoidal structures or porous tubular structures to separate epithelial cells from other cells. Designs for microfluidic BBB models vary greatly.illustrates typical design of the models, andshows the main characteristics for selected microfluidic models. These features will be discussed in detail below.
# Chip material
Since microfluidic chips can be manufactured by different materials, making appropriate choices can not only reduce experimental costs, but also help improve the stability, sensitivity, and accuracy. Silicon, the first-generation microfluidic device material, can be processed by MEMS technology to form microfluidic structures. However, the microfabrication process is usually expensive, time-consuming, and requires special microfabrication conditions.
Most microfluidic models now in use are polydimethylsiloxane (PDMS)-based systems because they have low cost and are easily microfabricated. PDMS is a polymeric organosilicon compound that is optically transparent, non-toxic, non-flammable, gas-and water-permeable. Optical transparency is convenient for observation and photography during an experiment, while gas and water permeability are important for cell culturing in the microfluidic chip environment. Critically, PDMS chips can be prepared using a mask by the replica molding process. The replication step allows mass-production of chips from one mold. PDMS can tightly bind to glass, PDMS, and other materials after a simple plasma treatment. It is therefore straightforward to design and fabricate various PDMS microfluidic chips to meet specific experimental requirements. Although PDMS devices are widely used in research, they have some limitations in the BBB studies. Native PDMS is hydrophobic and incompatible with most organic solvents, and hydrophobic molecules are easily adsorbed on the chip surface. Furthermore, the untreated PDMS surface has poor affinity for living cells. To reduce hydrophobicity and enhance cell adhesion during an experiment, it is important to modify the PDMS surface by plasma treatment or protein coating before initiating cell culture. It has also been reported that uncrosslinked free PDMS monomers can leach out into the culture medium and affect cellular behavior. Multi-layered channel structure made from patterned PDMS substrate with dynamic flows, co-cultured cells and two sets of electrodes (Reprinted from Referencewith permission). (c) The chip is consisted of four channels, two central gel regions for co-culturing astrocytes and neurons, two side channel for hosting endothelial cells and medium. There are 3 μm pores to allow diffusion of media and tracer between the central and outer compartments, astrocytes and HUVECs were cultured in the central and outer compartment (Reprinted from Referencein accordance with the Creative Commons Attribution). (d) In this chip, neuron and astrocytes are co-cultured in a vascular network, a system of two separate media microchannels is employed to independently emulate highly localized internal and external vascular microenvironments (Reprinted from Referencein accordance with the Creative Commons Attribution). (e) This is a typical two-chamber system divided by a porous polycarbonate membrane. (Reprinted from Referencein accordance with the Creative Commons Attribution). (f) A polycarbonate membrane is used to separate the top channel from the bottom channel, and Pt wire gently slide in the top and bottom groove (Reprinted from Referencewith permission).in accordance with the Creative Commons Attribution). (b) Multi-layered channel structure made from patterned PDMS substrate with dynamic flows, co-cultured cells and two sets of electrodes (Reprinted from Referencewith permission). (c) The chip is consisted of four channels, two central gel regions for co-culturing astrocytes and neurons, two side channel for hosting endothelial cells and medium. There are 3 µm pores to allow diffusion of media and tracer between the central and outer compartments, astrocytes and HUVECs were cultured in the central and outer compartment (Reprinted from Referencein accordance with the Creative Commons Attribution). (d) In this chip, neuron and astrocytes are co-cultured in a vascular network, a system of two separate media microchannels is employed to independently emulate highly localized internal and external vascular microenvironments (Reprinted from Referencein accordance with the Creative Commons Attribution). (e) This is a typical two-chamber system divided by a porous polycarbonate membrane. (Reprinted from Referencein accordance with the Creative Commons Attribution).
(f) A polycarbonate membrane is used to separate the top channel from the bottom channel, and Pt wire gently slide in the top and bottom groove (Reprinted from Referencewith permission).
To overcome the limitations of PDMS and make microfluidic chips more suitable for commercial applications, thermoplastics, such as polystyrene, cyclic olefin copolymer, and poly(methyl methacrylate) are used for microfluidic device fabrication. Thermoplastics have excellent light transmission performance and are more chemical stable, more compatible with organic solvents, and more acceptable to cell biologists. Using these materials, processed by lamination, embossing, and injection molding, microfluidic chip mass production is possible. However, it is difficult to generate very complex microstructures in a thermoplastics-based microfluidic chip.
## Porous membranes
In microfluidic models, porous membranes are used to separate the luminal and abluminal layers. The membranes provide the platform for co-culturing endothelial cells and other cells, and make permeability testing possible for models. The materials used for porous membranes include polycarbonate (PC), polyester (PE), polyethylene terephthalate, and polytetrafluoroethylene (PTFE).
The selection of membrane materials should prioritize optical transparency and their capacity for cell adhesion under shear stress. Pore density, pore size, and membrane thickness may affect signal transduction between cells on different sides of the membrane. Most membranes used in experiments are about 10 µm thick and have 0.2 or 0.4 µm pores, yielding pore densities of 10 8 /mm 2 . PC membranes are the most widely used in microfluidic models, but are optically translucent, compromising the ability to visualize cells by light microscopy during an experiment. PE and PTFE membranes are transparent and can be used to monitor the formation of monolayer with phase contrast microscopy. However, since PE membranes have low adhesive strength, they tolerate shear stress for a limited period of time. Sellgren found that PTFE membranes were more suitable than PE membranes for supporting cell attachment at realistic levels of shear stress. Although endothelial cells and other cells can be co-cultured through porous membranes, the thickness of these artificial membranes restricts cell interaction and is relatively high compared with the thickness of the basement membrane in vivo.
## Endothelial cells
Endothelial cells are the most important cells in the BBB. Initially, in vitro models used bovine, porcine, and rat primary BMECs, but the isolation of these cells is methodologically difficult and labor intensive. To decrease the time necessary to isolate primary cells, immortalized cell lines from diverse origins have been developed, but only a few of them exhibit the required barrier properties and functions. Lines that have been used in the BBB research include bEnd.3, a murine brain endothelial cell line, RBE4, a rat brain endothelial cell line, MDCK-MDR1, a line based on Mardin-Darby canine kidney cells transfected with the multidrug resistance gene MDR1, and hCMEC/D3, a human adult cerebral microvascular endothelial cell line. Compared with primary cells, immortalized cell lines can decrease the workload and reduce the time required to reach confluence. But some researchers report that immortalized cell lines cannot form complete tight junctions, resulting in a leaky barrier.
Most endothelial cells used in studies are from animal sources, which have species differences with human cells. Recently, human origin cells have been gradually used in model research. Adriani used human umbilical vein endothelial cells (HUVECs), co-cultured with rat astrocytes and neurons to form the BBB structure in microfluidic chip, and showed an intact monolayer and intercellular junctions. In Yang's research, HUVECs cell line ECV304 monoculture achieved higher TEER and lower permeability than bEnd3, but ECV304 monolayers lack the sufficient tightness and the evident of tight junction protein immunostaining, and when co-cultured with rat glioma C6 cells, barrier integrity is compromised. Wang et al. derived BMECs from human induced pluripotent stem cells and co-cultured them with rat primary astrocytes. The TEER levels of this model peaked above 4000 Ω·cm 2 and were sustained above 2000 Ω·cm 2 for 10 days, which are the highest values reported in microfluidic models. This model is closer to human physiology, not limited by cell availability, has the potential to be patient-specific, and holds great promise for drug permeability studies.
## Shear stress
Flow in a microfluidic chip generates shear stress, as is the case for flow in natural environments. The shear stress in the microchannel has effect on cell growth, morphology, and cell function, and regulates gene expression and functional phenotypes in endothelial cells, and epithelial cells of different origins may exhibit different phenotypes under shear stress. Exposure to fluid flow at physiological levels in a microfluidic chip increases the expression of tight junction proteins, enhances barrier integrity, and results in better barrier function. While low pathophysiological levels of fluid shear stress can stimulate the disintegration of tight junction. In the physiological state, shear stress is 10-20 dynes·cm −2 in a 10 µm-diameter capillary. In microfluidic chips, shear stress is mainly associated with fluid viscosity, flow rates through microchannels, channel geometry, and the flow profile. Due to the presence of cells and proteins, the viscosity of blood is greater than that of water. Flow rates in microfluidic models vary from 0.01 µL/min to 120 µL/min. Therefore, it can be very important to analyze and predict shear stress in a microfluidic model using simulation software.
## Assessment of microfluidic models
Using microfluidic models, it is possible to investigate cellular and molecular interactions. However, when implementing a new model, it is critical to assess the model's performance. Most commonly, this involves testing for the presence of specific markers of tight junctions, measuring TEER, and observing the permeability of specific substances.
## Determining specific tight junction markers
In the BBB, cells have extensive tight junctions and adhesion molecules at interendothelial cell-cell junctions to maintain the barrier integrity. Immunofluorescence or western blots can be used to observe the expression of specific markers, such as zona occludens-1 (ZO-1), claudin-5, and occludin. The P-glycoprotein efflux pump is another type of membrane transporter, and plays a vital function by preventing hydrophobic molecules from penetrating the BBB. Expression of P-glycoprotein is also used to evaluate BBB characteristics in microfluidic models.
## Trans-epithelial electric resistance (teer) measurement
TEER is a widely used parameter to monitor and evaluate barrier integrity and tightness. As epithelial cells are packed more tightly, fewer gaps exist in the barrier, which reduce the motion of ions and charged species and results in higher resistance. TEER is a quantitative measurement of the resistance over cell layers and cultured cell membranes. If a microfluidic chip contains Ag/AgCl pellet electrodes, platinum electrodes, or some other device that can perform as an electrode, TEER measurement can be performed in real time. The TEER value of an in vitro model should be as close as possible to the TEER in vivo, typically in the range of 1800 to 2000 Ω·cm 2, the in vitro values of most reported models are well below this range, 150 to 200 Ω·cm 2 is considered the lowest acceptable TEER value in functional models.
TEER offers a fast, label-free, and real-time assessment of barrier tightness, but is not sufficient to judge barrier selectivity. TEER measurements on a chip are subject to several confounding factors. First, cell origin and the extent of cell confluence may affect resistance. Second, the distribution of current across the membrane interface in a chip may not be uniform, leading to overestimates of TEER. Third, TEER is sensitive to temperature and ionic composition of the culture medium. Therefore, these parameters need to be kept constant during the measurement process.
## Permeability assessment
The BBB is a highly selective barrier that permits the passage of very few molecules. Small ions such as K + and Cl − can cross the BBB through ion channels. Small lipophilic molecules such as ethanol and nicotine can be transported passively across the barrier. Small polar molecules such as glucose, lactate, and pyruvate can cross the BBB by carrier-mediated transport. Finally, large molecules such as insulin, transferrin, leptin, albumin, and tumor necrosis factor alpha (TNFα) are passed through the BBB by receptor-mediated transport, adsorption-mediated transcytosis, and active efflux transporters.
To assess barrier function in a BBB model, permeability must be evaluated. A high quality model should have permeability characteristics similar to those found in vivo. When assessing permeability, it is important to select a suitable molecule. A good marker should be inert and have no effect on the physiology and function of the BBB. Fluorescein isothiocyanate (FITC) labeled dextrans, are widely used to evaluate the permeability. Commercially available FITC-dextrans can be obtained with molecular weights of 4, 10, 40, and 70 kDa. 14 C-D-mannitol (182 Da) and 14 C-urea (60 Da) are also used in permeability experiments. In most permeability analyses, only one marker is used. However, since different molecules traverse the barrier via different mechanisms, both hydrophilic and lipophilic molecules should be tested for permeability.
## Application of microfluidic models
Microfluidic chip models exhibit unique advantages in the BBB research. First, the chips are easy to design and fabricate, and can be customized to meet specific experimental requirements. Microchannel sizes are similar to microvascular structures in vivo, and it is easy to achieve precise fluid control. Second, multi-dimensional network structures generate relatively independent and closed environments in the chip, and thus resemble the microenvironment in vivo. In microfluidic models, the cell behavior changes greatly from 2D to 3D, and can obtain more physiological information and prediction data. Third, in microfluidic chips it is straightforward to combine and integrate various functional units. Consequently, the models can be used to observe cell morphology, image live cells for permeability studies, and monitor TEER in real time for barrier tightness using an integrated electrode.
Several microfluidic models have been developed in order to answer specific research questions. These models have been used to study the BBB function, screen drug candidates, and predict the clearance of pharmaceuticals by the BBB.
First, microfluidic models provide a new reliable platform to observe the influence of one or more molecules on the BBB characteristics. For example, Brown used the microfluidic model to study how the BBB responds to inflammatory stimuli, such as lipopolysaccharide or a cytokine cocktail. They found that inflammatory stimulation increases permeability, reduces the numbers of tight junctions, and alters the metabolomics profile. Stimulation by a cytokine mix comprising TNF, interleukin-1b, interferon c, and LPS was reported to result in the loss of VE-cadherin and ZO-1 expression, indicating disruption of the endothelial barrier. A model stimulated by 1 ng/mL TNF-α exhibited a 10-fold decrease in TEER. Exposure to histamine caused an instantaneous transient drop in TEER, while permeability coefficients increased significantly at higher pH (>10). Perfusion of an astrocyte-conditioned medium improved the BBB function, increased the expression of tight junction molecules, and decreased permeability.
Microfluidic models can also be used to assess the permeability of different compounds, including free forms, binding to nanomaterials, or functionalized. Falanga used a microfluidic model to evaluate a new nano drug delivery vector for the CNS, and found that nanoparticles pass more easily through the BBB barrier when combined with the membranotropic peptide GH625. Wevers evaluated the permeability of therapeutic antibodies by receptor-mediated endocytosis, antibody of human transferrin receptor (MEM-189) is easier to pass through the BBB compared with the control antibody. Based on microfluidic models, Bonakdar also found sub-electroporation pulsed electric field can disrupt the integrity of the BBB and increase permeability. In Papademetriou's research, he found that shear flow impacted the binding and internalization of angiopep-2 coupled liposomes nanoparticles by brain endothelial cells in microfluidic models.
# Conclusions and perspective
As the interface between the blood and the CNS, the BBB is critical for maintaining a steady state environment in the CNS. There is great interest in developing experimental models to understand the structural, physiological, and biochemical functions of the BBB, and to explore the mechanisms that change these properties under pathological conditions. Recently, microfluidic in vitro models have been established and evaluated. Microfluidic models are easy to design and fabricate, can simulate physiological fluid flow and shear stress, and can better mimic the BBB microenvironment compared with other models, and show great potential to further BBB research.
Although great progress has been made, microfluidic models are still in the early stages of development. It can be difficult to find a suitable model to meet all kinds of experimental requirements. Before these models are widely applied to the BBB-related research, several challenges must be overcome.
First, microfluidic models are not ready for commercial use. Microfluidic devices have the ability to control cell microenvironments and can be designed to meet specific requirements. However, simulation of the BBB microenvironment will require intricate systems, including complex microchannel networks with multiple elements such as pumps, valves, mixers, and detectors. As complexity increases, fabrication becomes more difficult. At the same time, minor changes and modifications to the models can produce chips with drastically different properties. Currently, most academic research use PDMS as the chip material, which is unfavorable for large-scale production for use in commercial applications. Thermoplastics may be superior for industrial purposes, but it is more difficult to obtain complicated and meticulous microstructures using this material.
Second, even with advances in chip design and fabrication, there remains a large gap between the characteristics of microfluidic models and the in vivo environment. For example, in most experiments a porous membrane is used to separate epithelial cells and other cells, but the thickness of the membrane is 10 µm, which is over 300 times thicker than the naturally occurring basement membrane in endothelial cells. This large gap makes difficult for different types of cells to establish direct contact. To enable direct contact between co-cultured cells, microchannels can be filled with hydrogels as an extracellular matrix, but rigid extracellular matrix substrates have stiffness values orders of magnitude higher than those observed in living brain microvessels. In addition, while capillaries have an average diameter of 6-9 µm, modeled blood vessels are far larger due to the constraints imposed by processing technology. The BBB is also a complex and integrated system, while current microfluidic models are based mainly on only two or three kind of cells. In a co-culture system, it is difficult to distinguish paracrine and cell-cell contact mediated effects. Moreover, the performance of a co-culture model is affected by many factors, such as cell type and origin, cell number and cell ratio, flow rates, and shear stress. All of these factors will need to be optimized.
Third, cell manipulation on a microfluidic chip requires substantial labor on the part of a researcher. There are six orders of magnitude differences between microfluidic chips and conventional laboratory equipment. Many small operation details can affect model properties and experimental reproducibility, and unintended variations may result in widely different results. For example, it is difficult to seed cells at a particular place in a sealed chamber because bubbles in the channel can interfere with the process. In order to improve experimental reproducibility, much more attention will be required to standardize details such as those involved in chip fabrication, cell seeding, cell localization, and other design factors. In order to compare the performance of different models, or compare data between in vitro models and in vivo experiments, it will also be necessary to standardize the measurement of parameters.
This review has presented recent progress in the design and evaluation of microfluidic in vitro BBB models, including advances in chip materials, porous membranes, the use of endothelial cells, the importance of shear stress, the detection specific markers to monitor tight junction formation and integrity, and measurements of TEER and permeability. However, like other models, microfluidic models cannot yet reproduce all the features of the BBB, and have limitations that must be addressed. By understanding and comparing the advantages and disadvantages of the design of different microfluidic models, we can choose the most suitable microfluidic models and experimental methods to adjust the data, so as to promote the BBB research. More distant goals are to optimize microfluidic models as screening platforms for candidate drugs and to apply them to study the pathogenesis in neurological diseases. |
Localized nasopharyngeal amyloidosis mimicking malignancy
Rationale: Nasopharyngeal amyloidosis is a benign, slowly progressive disease that is characterized by extracellular eosinophilic deposition.Patient Concerns: We report a rare case of localized nasopharyngeal amyloidosis.Diagnoses: The initial chief complaint of this patient was frequent epistaxis and right aural fullness. The initial diagnosis was nasopharyngeal tumor.Interventions: There is no universally effective medical treatment for nasopharyngeal amyloidosis but surgery can be an option. We performed careful observation with regular follow-up by nasopharyngoscopy and radiologic study.Outcomes: The patient reported no further complaints at 1-year follow-up and the lesion from nasopharyngeal amyloidosis was still present.Lessons: Although it is rare, nasopharyngeal amyloidosis should be considered in the differential diagnosis of epistaxis, nasal obstruction, and otitis media with effusion, which are the main symptoms of nasopharyngeal carcinoma. In the absence of systemic disease, localized nasopharyngeal amyloidosis may be treated conservatively.Abbreviation: MRI = magnetic resonance imaging.
# Introduction
Amyloidosis is a benign, slowly progressive disease that is characterized by extracellular eosinophilic deposition of insoluble polymeric fibrillar proteins in tissues and organs. [bib_ref] Surgical management of localized amyloidosis, Kennedy [/bib_ref] It is currently classified into 2 main forms, systemic and localized, with marked differences in prognosis between the 2 forms. The prognosis of systemic amyloidosis is worse than that for the localized form. [bib_ref] Amyloidosis of the larynx: a clinicopathologic study of 11 cases, Thompson [/bib_ref] Currently, the majority of types of systemic amyloidosis have no universally effective treatment. [bib_ref] Natural history and outcome in systemic AA amyloidosis, Lachmann [/bib_ref] Nasopharyngeal amyloidosis, a type of localized amyloidosis, is a very rare condition with few cases reported in the English scientific literature.
We describe a patient with primary localized nasopharyngeal amyloidosis whose endoscopic finding mimicked a malignant mass on both torus tubarius.
This study was approved by the institutional review board of Chonbuk National University Hospital. Informed consent was given by the patient.
## Case report
A 73-year-old man presented to our otolaryngology clinic with the chief complaint of frequent epistaxis and right aural fullness. Subsequent head and neck examination revealed middle ear effusion in the right ear. Nasopharyngoscopy revealed a nasopharyngeal mass on both torus tubarius with no evidence of complete obstruction of the Eustachian tube [fig_ref] Figure 1: Smooth-surfaced exophytic masses with visible blood vessels on both torus tubarius were... [/fig_ref]. Pure tone audiometry showed conductive hearing loss on the right side. Otherwise, there were no remarkable findings in other parts of the body.
Biopsy of this lesion demonstrated amorphous eosinophilic depositswhich showed apple-green birefringence on Congo red staining. Axial, sagittal, and coronal cuts from magnetic resonance imaging (MRI) , B-D) demonstrated a localized lesion in the nasopharynx at the torus tubarius with no evidence of destructive or invasive behavior. Serum immunofixation electrophoresis and urine protein electrophoresis showed no evidence of monoclonal gammopathy or proteinuria. Biopsy of abdominal fat and rectal samples showed no evidence of amyloid infiltration. Bone marrow biopsy showed no evidence of infiltration or tumor.
The patient was diagnosed as having localized nasopharyngeal amyloidosis. He underwent a myringotomy on the right side and was treated with prednisolone (0.3 mg/kg/day) and cefpodoxime (3 mg/kg/day) for 2 weeks and the symptoms subsided. The patient reported no further complaints at 1-year follow-up and the lesion from nasopharyngeal amyloidosis was still present.
The Ethical committee approval was acquired from the institutional ethical review board of Chonbuk National University Hospital, Korea (IRB number: 2016-01-018).
# Discussion
Localized amyloidosis is extremely rare, frequently involves the head and neck region without systemic manifestations, and has an excellent prognosis. The larynx is the most affected area (61%), followed by the oropharynx, trachea, orbit, and nasopharynx. [bib_ref] Amyloid deposits and amyloidosis: the beta-fibrilloses, Glenner [/bib_ref] Nasopharyngeal amyloidosis has previously been described in the literature in 13 different patients, all of whom had localized amyloidosis. [bib_ref] Localized amyloidosis masquerading as nasopharyngeal tumor: a review, Panda [/bib_ref] Although it is usually asymptomatic, nasopharyngeal amyloidosis may cause epistaxis, aural fullness, middle ear effusion, and conductive hearing loss as in this case. Our patient had a representative endoscopic image of hematoma due to vessel wall invasion. This indicates the importance of being alert to the major symptoms of nasopharyngeal amyloidosis; otherwise, it may be missed.
The differential diagnosis of nasopharyngeal amyloidosis includes nasopharyngeal carcinoma and lymphoma. The phenotypic examination is usually nonspecific; however, several authors have reported the appearance of nasopharyngeal amyloidosis. The endoscopic appearance of nasopharyngeal amyloidosis is of a heterogeneously reddish, smooth-surfaced exophytic tumor in the nasopharynx. [bib_ref] Localized nasopharyngeal amyloidosis, Chen [/bib_ref] In our case, a smoothsurfaced exophytic mass with visible blood vessels on the torus tubarius was characteristically noted. A blood clot was also noted in the right torus tubarius, which was the cause of nasal bleeding.
The "gold standard" for the diagnosis of amyloidosis is a tissue biopsy. Histology of the biopsied specimen using hematoxylin and eosin staining shows amyloid as an extracellular, amorphous, acellular eosinophilic deposit, sparing the overlying epithelium. Congo red staining shows characteristic apple-green birefringence and dichroism under a polarized light microscope. [bib_ref] Amyloidosis of the larynx: a clinicopathologic study of 11 cases, Thompson [/bib_ref] Staining with methyl violet or crystal violet reveals metachromatic pink-violet staining. [bib_ref] Amyloidosis of the larynx: a clinicopathologic study of 11 cases, Thompson [/bib_ref] A negative result for abdominal fat and rectal biopsy is necessary for diagnosis of the localized form, not systemic amyloidosis.
There is no universally effective medical treatment for nasopharyngeal amyloidosis but surgery can be an option. [bib_ref] Amyloidosis of the nasopharynx, Sánchez Legaza [/bib_ref] The surgical excision of nasopharyngeal amyloidosis is usually conservative, not radical, because relapse is common. Complete removal of nasopharyngeal amyloidosis can impair normal physiological functions. There have been no reports showing progression of localized form to systemic amyloidosis. In addition, a survival benefit from surgical excision of nasopharyngeal amyloidosis has not been demonstrated. For these reasons, careful observation with regular follow-up by nasopharyngoscopy and radiologic study is a reasonable therapeutic option, as in our case. [bib_ref] Localized nasopharyngeal amyloidosis, Chen [/bib_ref]
# Conclusion
Although it is rare, nasopharyngeal amyloidosis should be considered in the differential diagnosis of epistaxis, nasal obstruction, and otitis media with effusion, which are the main symptoms of nasopharyngeal carcinoma. In the absence of systemic disease, localized nasopharyngeal amyloidosis may be treated conservatively.
[fig] Figure 1: Smooth-surfaced exophytic masses with visible blood vessels on both torus tubarius were noted on endoscopy. (A) Right side, (B) left side. Note the blood clot in the right torus tubarius. [/fig]
[fig] Figure 2: (A) Hematoxylin and eosin staining showed deposits of amyloid under epithelial layers. (B) Amyloid demonstrating apple-green birefringence with polarized light after Congo red staining. [/fig]
|
An enhanced electrochemical and cycling properties of novel boronic Ionic liquid based ternary gel polymer electrolytes for rechargeable Li/LiCoO2 cells
A new generation of boronic ionic liquid namely 1-ethyl-3-methylimidazolium difluoro(oxalate) borate (EMImDFOB) was synthesized by metathesis reaction between 1-ethyl-3-methylimiazolium bromide and lithium difluoro(oxalate)borate (LiDFOB). Ternary gel polymer electrolyte membraneswere prepared using electrolyte mixture EMImDFOB/LiDFOB with poly vinylidenefluoride-cohexafluoropropylene (PVdF-co-HFP) as a host matrix by facile solvent-casting method and plausibly demonstrated its feasibility to use in lithium ion batteries. Amongst ternary gel electrolyte membrane, DFOB-GPE3, which contained 80 wt% of EMImDFOB/LiDFOB and 20 wt% PVdF-co-HFP, showed excellent electrochemical and cycling behaviors. The highest ionic conductivity was found to be 10 −3 Scm −1 at 378 K. Charge-discharge profile of Li/DFOB-GPE3/LiCoO 2 coin cell displayed a maximum discharge capacity of 148.4 mAhg −1 at C/10 rate with impressive capacity retention capability and columbic efficiency at 298 K.Published: xx xx xxxx OPEN www.nature.com/scientificreports/ 2 ScIEntIfIc RePORtS | 7: 11103 | Usually organic solvents are having meager thermal and electrochemical properties and relatively narrow electrochemical potential window 17 . To address the aforementioned issues, scientific communities have been developed various types of additives to enhance electrochemical and cycling stability greater to organic solvents. On the other hand, room temperature ionic liquids (ILs) which consists of anions and cations have some exceptional properties such as negligible vapor pressure, non flammability, better thermal stability, great chemical and electrochemical stability, inherent long life, and high ionic conductivity are considered to be a promising electrolyte salt for polymer in salt system[18][19][20][21][22]. GPEs combines with ionic liquids which are phenomenally called ionic liquid gel polymer electrolytes (ILGPEs) possess the merits of high ionic conductivity, great electrochemical stability window and good charge-discharge performance by the way of preventing the dendrite formation on lithium metal electrodes. Henceforth, it is believed to be prospective electrolytes for LIBs. Remarkable efforts have been devoted to enhance the ionic conductivity and performance of GPEs for LIBs. Interestingly, different types of ionic liquids such as TFSI − , CF 3 SO 3 − , BF 4 − , C 4 F 9 SO 3 − impregnated with lithium salts as electrolyte mixtures were successfully employed as GPEs for lithium-ion battery and the performance of those electrolytes and their batteries have extensively explored[18][19][20][21][22][23].The present work is focused on preparing a new generation of GPEs which comprised of boron anionic ionic liquid and lithium salt so as to fabricate advanced GPE with excellent mechanical strength, enlarged electrochemical window and high ionic conductivity. Here, we report an innovative ternary gel polymer electrolyte (TGPEs) membranes which are prepared by incorporation of diflurooxalato borate (DFOB) anion based ionic liquid [1-ethyl-3-methylimidazolium diflurooxalato borate (EMImDFOB)] and lithium salt (LiDFOB) into a PVdF-co-HFP matrix. Difluorooxalato borate (DFOB) is of quite interest in the present investigation due to its improved solubility compared with previously studied anions 24 . Also, in DFOB, the presence of more electron-withdrawing fluorine atom results into more delocalized charge which gives the anion less affinity for Li + and EMIm + , causing better conductivity 25 . Moreover, it possesses a lower lowest unoccupied molecular orbital (LUMO) and a higher electrochemical stability for wide electrochemical window towards batteries with high energy. The detailed investigation is carried out to synthesis and calibration of EMImDFOB with LiDFOB for LIBs. The synergistic effect of EMImDFOB/LiDFOB on the electrochemical, thermal, mechanical and cycling performance of difluorooxalato borate based gel polymer electrolytes (denoted as DFOB-GPEs) have been demonstrated plausibly. To the best of our knowledge, there is no report available based on PVdF-co-HFP-EMImDFOB/ LiDFOB as electrochemically and mechanically stable gel polymer electrolytes for LIBs applications.
In the recent times, rechargeable lithium ion batteries are widely considered as one of the efficient electro-chemical energy storage system with high energy densities and therefore has been employed in many applications such as portable electronic devices, electric vehicles and smart grid storage systems [bib_ref] Challenges Facing Lithium Batteries and Electrical Double-Layer Capacitors Angewandte, Choi [/bib_ref] [bib_ref] Development and challenges of LiFePO 4 cathode material for lithium-ion batteries, Yuan [/bib_ref] [bib_ref] Building better batteries, Armand [/bib_ref] [bib_ref] Lithium Batteries Nanomaterials for Rechargeable Lithium Batteries ** Angewandte, Bruce [/bib_ref] [bib_ref] Battery materials for ultrafast charging and discharging, Kang [/bib_ref] [bib_ref] Erratum: Lithium insertion in graphite from ternary ionic liquid-lithium salt electrolytes I...., Appetecchi [/bib_ref] [bib_ref] Development of low temperature Li-ion electrolytes for NASA and DoD applications, Plichta [/bib_ref]. However, several shortcomings which not only includes volatility and combustion of organic liquid electrolytes but also associated narrow operational temperature range restricts the frequent use of organic liquid electrolytes based lithium ion batteries [bib_ref] Novel polymer electrolytes based on thermoplastic polyurethane and ionic liquid/lithium bis (trifluoromethanesulfonyl)..., Lavall [/bib_ref] [bib_ref] Polymerized ionic liquids with guanidinium cations as host for gel polymer electrolytes..., Li [/bib_ref]. Thus, there is an urgent need to address all the above mentioned shortcomings so as to improve their operational stability and using them in hybrid electric vehicles [bib_ref] Lithium Batteries Nanomaterials for Rechargeable Lithium Batteries ** Angewandte, Bruce [/bib_ref] [bib_ref] Battery materials for ultrafast charging and discharging, Kang [/bib_ref]. An ideal way to improve operational safety without compromising on the energy density of the batteries is to use gel polymer electrolytes (GPEs) [bib_ref] Compatibility of polymer electrolyte based on N-methyl-N-propylpiperidinium bis (tri fl uoromethanesulphonyl) imide..., Swiderska-Mocek [/bib_ref] [bib_ref] PVDF-based composite microporous gel polymer electrolytes containing a novelsingle ionic conductor SiO2(Li+), Li [/bib_ref] [bib_ref] Recent development of ionic liquid membranes, Wang [/bib_ref]. Previous studies have demonstrated that GPEs have higher acceptable ionic conductivity than solid electrolytes at ambient temperature and higher thermal and mechanical stability than liquid electrolytes, making them a potential alternatives to all solid-state polymer electrolytes and traditional liquid electrolytes [bib_ref] Polymerized ionic liquids with guanidinium cations as host for gel polymer electrolytes..., Li [/bib_ref]. Till date various polymers such as poly (ethyleneoxide) (PEO), poly (propyleneoxide) (PPO), poly (methylmethacrylate) (PMMA), poly (acrylonitrile) (PAN), poly (vinylidene fluoride) (PVdF) and poly (vinylidene fluoride-co-hexafluoropropylene) (PVdF-co-HFP) have been introduced as hosts for GPEs preparations [bib_ref] UV cross-linked, lithium-conducting ternary polymer electrolytes containing ionic liquids, Kim [/bib_ref] [bib_ref] Nanocomposite PEO-based polymer electrolyte using a highly porous, super acid zirconia filler, Derrien [/bib_ref] [bib_ref] Solid polymer electrolytes based on polyethylene oxide and lithium trifluoro-Methane sulfonate (PEO-LiCF..., Karan [/bib_ref]. Amongst them, PVdF-co-HFP is considered as a better polymer host because of its high electrochemical stability and ability to dissolve lithium salts.
# Results and discussions
The synthetic route of ionic liquid is schematically represented in(supplementary information). To demonstrate the purity of our synthesized ionic liquid EMImDFOB, [bib_ref] Challenges Facing Lithium Batteries and Electrical Double-Layer Capacitors Angewandte, Choi [/bib_ref] [bib_ref] UV-cured polymer electrolytes encompassing hydrophobic room temperature ionic liquid for lithium batteries, Gerbaldi [/bib_ref] F-NMR (using D 2 O as solvent) spectrum depicts δ peak at 151.68 (2 F, t # , J = 4.5 Hz, BF 2 ) as shown inof supplementary information. The exhibited water content of as synthesized EMImDFOB is less than 35 ppm which is inferred from Karl-Fischer titration and this value is quite acceptable for practical lithium ion battery applications. The viscosity of EMImDFOB is found to be 123 ± 1 cP as shown in [fig_ref] Figure 3: FT-IR spectra of DFOB-GPEs in the wavenumber regions [/fig_ref].
The innovative TGPEs membranes were prepared by incorporation of synthesized EMImDFOB and lithium salt LiDFOB into a PVdF-co-HFP matrix. The utilization of EMImDFOB/LiDFOB based ternary GPEs in LIBs are schematically represented in. XRD measurements were performed for pure PVdF-co-HFP, DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3 to examine its phase purity and crystalline behavior. The XRD patterns of prepared DFOB-GPEs with different ratios are given in. For the comparison, pure PVdF-co-HFP XRD pattern is provided in. Two kind of characteristic diffraction peaks located at 2θ = 20.38° and 40.26° corresponds to (020) and (021) reflection planes, respectively, which revealed the semi-crystalline PVdF is presented in the complexes [bib_ref] Studies on PVdF-based gel polymer electrolytes, Periasamy [/bib_ref] [bib_ref] Preparation of LAGP/P(VDF-HFP) polymer electrolytes for Li-ion batteries, Zhang [/bib_ref]. Due to addition of EMImDFOB/LiDFOB into the PVdF-co-HFP matrix, it has been observed that the intensity of PVdF-co-HFP characteristic peak is decreased thereby confirming the decrease in crystalline behavior in the as prepared membranes. Furthermore, no other characteristic peaks are observed related to EMImDFOB/LiDFOB which affirms the complete dissolution of electrolyte mixture into the polymer matrix. The observed amorphous nature of XRD patternscertifies the formation of DFOB-GPEs through the addition of EMImDFOB/LiDFOB into PVdF-co-HFP matrix system. In addition, XRD pattern reveals the absence of characteristic peak of PVdF-co-HFP for DFOB-GPE3 electrolyte, which is in agreement with the result by Qing Zhang et al. [bib_ref] Preparation of LAGP/P(VDF-HFP) polymer electrolytes for Li-ion batteries, Zhang [/bib_ref].
Differential scanning calorimetry (DSC) analysis was performed to validate the XRD results as well as to identify the influence of EMImFOB/LiDFOB on T g and T m in PVdF-co-HFP matrix.shows the DSC thermograms of DFOB-GPE systems such as DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3 with the temperature range between −70 and 200 °C. The DSC thermogram of pure EMImDFOB and EMImDFOB/LiDFOB are represented in. From the inset thermogram, an endothermic peak is appeared at around 142.25 °C corresponding to the melting of crystalline phase and the glass transition temperature (T g ) peak is observed at −37.45 °C for PVdF-co-HFP. Whereas Tg values of pure EMImDFOB and EMImDFOB/LiDFOB are appeared around 186 and 197 °C respectively. Enlarged spectra of insetfor region −45 to −20 °C revealed that the DFOB-GPEs system T g zone is shifted from −23.77 to −14.40 °C with the increase in the EMImDFOB/ LiDFOB concentration from 60 to 80 wt%, respectively. The incorporation of EMImDFOB/LiDFOB with polymer matrix elevates the T g of DFOB-GPEs as well as widens the T g zone. The T g zone widening can be directly related to enhancement of flexibility in the polymer chains of ionic-polymer complex. Also, the melting temperature (T m ) of our prepared DFOB-GPEs membranes is shifted towards lower temperature region by the addition of EMImDFOB/LiDFOB into PVdF-co-HFP matrix. This can be due to plasticization effect i.e., the presence of EMImDFOB/LiDFOB mixture in PVdF-co-HFP matrix might have weakened the interactive bonds between the chains within the gel electrolytes thereby causing the reduction in the amount of energy needed to break the bond. Amongst the prepared DFOB-GPEs, the gel electrolyte with 80% EMImDFOB/LiDFOB shows the lowest value of T m1 and T m2 as can be seen in(b and e) that the T m of pure host (T m1 and T m2 @ 142.00, 170.88 °C) is decreased to 125.33 and 159.26 °C for DFOB-GPE3. The observed results are in concurrence with other ILGPEs using PVdF-co-HFP as host matrix [bib_ref] Influence of 2,6 (N-pyrazolyl)isonicotinic acid on the photovoltaic properties of a dye-sensitized..., Ganesan [/bib_ref] [bib_ref] Development of ion conducting polymer gel electrolyte membranes based on polymer PVdF-HFP,..., Shalu [/bib_ref]. The observed T g , T m , ΔH m and percentage of crystallinity (%α) values are tabulated in . The %α of DFOB-GPEs is calculated by following equation, The heat of enthalpy (∆H m100% ) value is observed at 104.5 J/g for 100% crystalline PVdF-co-HFP [bib_ref] UV cross-linked, lithium-conducting ternary polymer electrolytes containing ionic liquids, Kim [/bib_ref]. The value of %α is decreased from 17.23 to 12.12 with increase of EMImDFOB/LiDFOB content in DFOB-GPEs which revealed the flexibility and mobility enhancement in the polymer chain segment of gel electrolytes.
To explore the significance of EMImDFOB/LiDFOB with PVdF-co-HFP system, the attenuated total reflection (ATR) -Fourier transform-infrared (FTIR) spectroscopy analysis was performed for our DFOB-GPEs membranes. [fig_ref] Figure 3: FT-IR spectra of DFOB-GPEs in the wavenumber regions [/fig_ref] represents the ATR-FTIR spectra of DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3 polymer membranes. The FTIR peaks appeared at 1120 cm −1 and 1040 cm −1 belongs to asymmetrical stretch of CF 2 group and bending of CF 3 group of pure PVdF-co-HFP, respectively [bib_ref] PVDF-based composite microporous gel polymer electrolytes containing a novelsingle ionic conductor SiO2(Li+), Li [/bib_ref] [bib_ref] Solid-state electric double layer capacitors fabricated with plastic crystal based flexible gel..., Suleman [/bib_ref] [bib_ref] Ionic liquid based sodium ion conducting gel polymer electrolytes, Kumar [/bib_ref]. For pure EMImDFOB/LiDFOB mixture ( The addition of EMImDFOB/LiDFOB in PVdF-co-HFP matrix facilitates the interaction with the free electron pairs of fluorine atom (CF 2 and CF 3 group) of host matrix as can be confirmed by low intensity vibrational peaks in the region of 1100-1175 cm −1 as represented in [fig_ref] Figure 3: FT-IR spectra of DFOB-GPEs in the wavenumber regions [/fig_ref]. Also, the shift of ν a (CF 2 ) and δ(CF 3 ) vibrational modes towards higher wavenumber reveals the cations (Li + and EMIm + ) interaction with the polymer host. Moreover, the formation of complex of EMImDFOB/LiDFOB with PVdF-co-HFP matrix is confirmed by important conformational mode variations in two regions namely region I (1740-1620 cm −1 ) and region II (1570-1480 cm −1 ) as highlighted in [fig_ref] Figure 3: FT-IR spectra of DFOB-GPEs in the wavenumber regions [/fig_ref]. Further, the formation of complex of EMImDFOB/LiDFOB with PVdF-co-HFP is evident by the shifting of ν CH and ν NH peaks towards right in the wavenumber region 3300-3000 cm −1 as shown in [fig_ref] Figure 3: FT-IR spectra of DFOB-GPEs in the wavenumber regions [/fig_ref]. The particular attentiveness for these two regions is due to some crucial changes in the peak positions which are given in following. (a) Region I -Symmetrical ν C=O of DFOB: The deconvoluted spectra of DFOB-GPEs for region-I (1740-1620 cm −1 ) is shown in. For DFOB-GPE1, the symmetrical stretching ν C=O bands were appeared at 1702 and 1687 cm −1 which can be assigned to ion pairs and free ions, respectively with low peak broadening. The broadened ν C=O bands were observed at 1705 and 1649 cm -1 for DFOB-GPE2 and at 1705 and 1644 cm −1 for DFOB-GPE3 which are related to the ion pairs and free ions, respectively. The observed results are consistent with the earlier report [bib_ref] Effects of 1-butyl-3-methyl imidazolium trifluoromethanesulfonate ionic liquid in poly(ethyl methacrylate)/poly(vinylidenefluoride-co-hexafluoropropylene) blend based..., Sim [/bib_ref] [bib_ref] Lithium ion conduction and ion -polymer interaction in PVdF-HFP based gel polymer..., Isa [/bib_ref] [bib_ref] Effect of gamma irradiation on poly(vinyledene difluoride)-lithium bis(oxalato)borate electrolyte, Abdul Rahaman [/bib_ref]. (b) Region II -δ CH and ν CH of imidazolium ring: In the region II of [fig_ref] Figure 3: FT-IR spectra of DFOB-GPEs in the wavenumber regions [/fig_ref] (1570~1480 cm −1 ) and 3c (3300-3000 cm −1 ), all the DFOB-GPEs exhibits the vibrational bands which belongs to δ CH and ν CH of imidazolium ring. The characteristic vibrational peak of DFOB-GPE became prominent for the higher concentration of EMImFOB/LiDFOB and also vibrational peak of pure polymer host tends to DFOB-GPEs disappear which indicates that presence of EMImFOB/LiDFOB is crucial in the gel electrolyte matrix system and it has increased the amorphicity [bib_ref] Development of ion conducting polymer gel electrolyte membranes based on polymer PVdF-HFP,..., Shalu [/bib_ref] [bib_ref] Solvate Structures and Computational/Spectroscopic Characterization of Lithium Di fl uoro (oxalato) borate..., Han [/bib_ref]. TGA analysis was performed to verify the thermal decomposition of prepared DFOB-GPEs and to analyze its dimensional stability at elevated temperatures. The TGA plots of DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3 in the temperature range of 25-800 °C are shown in. The TGA plot of pristine PVdF-co-HFP is given in inset ofwhereas the TGA plot for pure EMImDFOB and EMImDFOB/LiDFOB are shown in . It is observed that the decomposition of all DFOB-GPEs is single step and decomposition temperatures of various DFOB-GPEs are found to be 273, 272 and 248 °C respectively. From, it can be found that for the higher content of EMImDFOB/LiDFOB based DFOB-GPEs (DFOB-GPE3), the thermal stability decreases slightly (273 °C to 248 °C) but is still suitable for practical application. The decrease in thermal stability at high electrolyte mixture (EMImDFOB/LiDFOB) content is explained as follows: After incorporation of EMImFOB/LiDFOB into PVdF-co-HFP, the onset of thermal decomposition temperature for DFOB-GPEs decreases slightly, which reveals that the existence of interaction between PVDF-co-HFP and EMImFOB/LiDFOB. These interactions results from intermolecular hydrogen bonds between fluorine atoms and the hydrogen atoms connected with carbon atom in imidazole ring of EMImDFOB 38 as represented pictorially in. Further, as it can be seen from the thermogram that the drastic weight losses of 34.79, 29.67 and 25.95 wt% were observed for DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3, respectively. Amongst them, DFOB-GPE3 possesses remarkably low percentage of weight loss than that of other two electrolytes. To explain the quality of our result, it is worth mentioning that the previous studies by Tang et al. have reported weight loss of 75 wt% of PVdF-co-HFP based gel electrolyte systems which is higher than our gel electrolyte membrane [bib_ref] Electrochemical and cycling performances of novel nona fl uorobutanesulfonate (nona fl ate)..., Karuppasamy [/bib_ref] [bib_ref] An efficient way to achieve high ionic conductivity and electrochemical stability of..., Karuppasamy [/bib_ref].
[formula] T m (°C) T g (°C) ΔH m (J/cal) % α σ 298K (Scm −1 ) E a (eV) Cut-off voltage (V) Mechanical Strength (MPa) T m1 T m2 T g1 T g2 DFOB-GPE1 [/formula]
Stress-strain behavior of gel polymer electrolytes plays a key role to determine the exact mechanical strength of prepared electrolytes as well as to prevent the short circuit in lithium ion batteries application. The typical stress-strain curves of pristine PVdF-co HFP, DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3 are shown in. It can be seen fromthat the pristine PVdF-co-HFP shows a one-step break (@6.09 MPa) mechanism resulting in non-linear elastic behavior with the strain of 5.77%. In the case of DFOB-GPEs, the two discrete regions are observed such as linear region for elastic characteristic and nonlinear region for plastic deformation. When increase the ratio of EMImDFOB/LiDFOB in DFOB-GPEs membranes, the mechanical strength is drastically decreased and flexibility is linearly enhanced. In Comparison with DFOB-GPE2 and DFOB-GPE3, the DFOB-GPE1 membrane which is containing 60% of ionic liquid EMImDFOB/LiDFOB has low elongation break (29.00%) and high tensile strength (2.93 MPa). For DFOB-GPE3, decrease in the tensile strength (2.22 MPa) and increase in strain at 34.72% is observed. From the above results of DFOB-GPEs, it can be inferred that DFOB-GPE3 possesses low mechanical strength compared to other electrolytes but it possess appreciable robustness (2.22 MPa), self-standing with no electrolyte flow, which can meet the requirement for practical lithium ion battery applications. This is in agreement with the earlier reports [bib_ref] Taichi-inspired rigid-flexible coupling cellulose-supported solid polymer electrolyte for high-performance lithium batteries, Zhang [/bib_ref] [bib_ref] Study of the effect of a novel high-performance gel polymer electrolyte based..., Tang [/bib_ref] [bib_ref] Rigid-flexible coupling high ionic conductivity polymer electrolyte for an enhanced performance of..., Hu [/bib_ref] [bib_ref] Tough Nanocomposite Ionogel-based Actuator Exhibits Robust Performance, Liu [/bib_ref] [bib_ref] Fabrication and properties of polybutadiene rubber-interpenetrating cross-linking poly(propylene carbonate) network as gel..., Huang [/bib_ref].
In order to validate the electrochemical properties of electrolytes, we have performed electrochemical impedance spectroscopy (EIS) analysis for DFOB-GPEs in the temperature range between 298 and 398 K. The ionic conductivity as a function of temperature for all the DFOB-GPEs is presented in. It has revealed that the ionic conductivity increases with increase of temperature for all the DFOB-GPEs samples. Also, our observed result obeys Arrhenius law of conduction relation as follows
[formula] σ σ = − e(2)E RT 0 ( / ) a [/formula]
where σ is the ionic conductivity, T is the absolute temperature, E a is the apparent activation energy for ionic transport. R is the gas constant (8.314 J/mol.K). The calculated activation energy (E a ) and pre-exponential factor (σ 0 ) values are listed in . The values of E a and σ 0 of gel electrolytes decreases with increase in EMImDFOB/ LiDFOB weight percentage (wt. %) in DFOB-GPEs as shown in. Also using the above equation (2) . The observed high ionic conductivity of DFOB-GPE3 can be attributed to the faster migration of charge carriers in polymer matrix with high flexibility when compared to other two electrolytes. Previous report by Stepniak et al. [bib_ref] Characterization and application of N-methyl-N-propylpiperidinium bis(trifluoromethanesulfonyl)imide ionic liquid-based gel polymer electrolyte prepared..., Stepniak [/bib_ref] have demonstrated that conductivity of gel electrolytes is ~0.64 × 10 -4 Scm −1 which is lower than observed results for all the DFOB-GPEs. Also, they have stated that the ionic conductivity of electrolyte membrane depends on the number of charge carriers as it act as a medium between anode and cathode as well as served as a host for the whole system. Amongst all prepared DFOB-GPEs, DFOB-GPE3 provides highest ionic conductivities of the order of 10 −4 Scm −1 and 10 −3 Scm −1 at room temperature and 398 K, respectively. The resultant conductivity values are quite higher than that of previously studied PVdF-co-HFP based gel electrolyte system 11, 44-46 which makes our prepared DFOB-GPEs as a potential candidate for lithium ion batteries with wide range of operating temperatures.
To explore the role of lithium ions as well as transfer rate of lithium ions in the electrolyte, we have demonstrated cationic transference number using chronoamperometry and EIS analyses. The symmetrical Li/ DFOB-GPE3/Li was conducted at room temperature and the resulting polarization curve of DFOB-GPE3 is shown in. We have used DFOB-GPE3 electrolyte for analyses because it possesses the maximum ionic where, the subscripts 0 and ss represents initial values and steady state values, respectively, R b is the bulk resistance, R is the passive film resistance, and their values can be evaluated from the nyquist curves of the DFOB-GPEs before and after the experiment, V ∆ is the applied voltage and I is the current. As evident from the, the chronoamperometric curve tends to decrease linearly with time for symmetric cell Li/DFOB-GPE3/Li. The value of t Li + for DFOB-GPE3 is found to be at 0.37 which is quite high compared to other PVdF-co-HFP based systems [bib_ref] Preparation and characterization of gel polymer electrolytes containing N-butyl-N-methylpyrrolidinium bis(trifluoromethanesulfonyl) imide ionic..., Libo [/bib_ref]. The higher value of t Li + is due to the faster migration of lithium ions in the gel matrix which can be explained by following facile mechanism: The cationic sites such as Li + and EMIm + can interact with the electron donor site of PVdF-co-HFP (CF 2 ) and DFOB − and therefore can possibly weakened the polymer backbone chain thereby increasing the amorphocity of gel electrolyte to yield high electrical properties. Due to the amorphous nature of polymer host matrix, the transport of charge carrier Li + becomes faster thereby improving the cationic transport number. The observed results explains the increase in the ionic conductivity in DFOB-GPEs membranes as well as transference number which is due to the inclusion of EMImDFOB/LiDFOB in PVdF-co-HFP matrix system [bib_ref] Review on gel polymer electrolytes for lithium batteries, Stephan [/bib_ref].
Linear sweep voltammetry (LSV) was performed in order to determine the operating potential and the electrochemical stability window of prepared electrolytes for batteries with the potential range between −5 and + 6 V. The LSV scans were performed using Pt/DFOB-GPEs/Li cell with a scan rate of 20 mV/s at 298 K. LSV curves of DFOB-GPEs membranes are shown in. From the spectra, LSV current (I) is increased with the applied voltage (V) for all the DFOB-GPEs membrane. The breakdown voltage of DFOB-GPEs increases with increase of EMImDFOB/LiDFOB contents and it is observed at 4.04, 4.24 and 4.46 V vs. Li/Li + for DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3, respectively . This obvious improvement of the breakdown voltage demonstrates that our DFOB-GPE3 may have potential advantage for applications in practical high performance lithium ion batteries [bib_ref] Ionic liquids to the rescue? Overcoming the ionic conductivity limitations of polymer..., Shin [/bib_ref] [bib_ref] PVDF-HFP/ether-modified polysiloxane membranes obtained via airbrush spraying as active separators for application..., Seidel [/bib_ref].
Further, to validate LSV results as well as to confirm lithium ion conduction into the GPEs, cyclic voltammetric (CV) analysis was carried out using a Li/DFOB-GPE3/LiCoO 2 coin cell. Among all the DFOB-GPE membrane, DFOB-GPE3 is of significant interest due to its high ionic conductivity in ambient temperature and thermal stability than other two electrolytes.shows the multiple cycle CV curves of DFOB-GPE3 membrane comprised coin cell. From the 1 st cycle of CV curve, the wide electrochemical stability voltage windows (5 V) is observed and it has displayed lithium stripping at 2.6 V and lithium plating at 2.7 V. Hence, anodic stripping and cathodic deposition are facile at electrode-electrolyte interface. The lithium stripping peak is observed at 2.6 V and the lithium plating peak appears at 2.7 V during anodic and cathodic scans which clearly indicates that the DFOB-GPE3 is electrochemically active. In order to confirm the red-ox behavior of DFOB-GPE3 system, five consecutive cycles of CVsis recorded and their peak stability is established. The red-ox peaks current values are increased with increase of number of cycles which may be due to faster ionic transport rate in electrode-electrolyte interface [bib_ref] Electrochemical and cycling performances of novel nona fl uorobutanesulfonate (nona fl ate)..., Karuppasamy [/bib_ref] [bib_ref] Cheap glass fiber mats as a matrix of gel polymer electrolytes for..., Zhu [/bib_ref] [bib_ref] Immobilization of Anions on Polymer Matrices for Gel Electrolytes with High Conductivity..., Wang [/bib_ref].
The scan rate dependence of the redox peak currents and the peak-to-peak separation (E p ) of DFOB-GPE3 were also analyzed, and the results are depicted in. The scan rate was varied from 5 to 50 mV/s. It was noticed that with the increase of the scan rate, the oxidation peaks shifts towards higher potential while its counterpart peak shifts towards lower potential. The value of E p is plotted versus the log of scan rate and is presented in inset of. According to Laviron's theory [bib_ref] An efficient way to achieve high ionic conductivity and electrochemical stability of..., Karuppasamy [/bib_ref] , with the increase of scan rate, the oxidation and reduction peak currents of DFOB-GPE3 increases significantly. The observed results are strongly paved a way to utilize DFOB-GPE3 as an active electrolyte for lithium ion batteries.
With the consideration of mechanical and thermal stabilities, high ionic conductivity and transference number and better CV performance, the DFOB-GPE3 electrolyte was used to assemble a LiCoO 2 /Li coin cell. The discharge profiles are obtained at different current rates at room temperature as shown in. It seems that the discharge profile exhibits obvious stable voltage profiles at all the different C-rates which may be due to the better electrochemical stability and mechanical integrity of DFOB-GPE3. The discharge capacity slightly diminishes with number of cycles for different C-rates. The cycling performance at C/10 rate, with a discharge capacity of 148.4 mAh g −1 , is apparently good. However, the discharge capacity faintly decreases to 130 and 118 mAh g −1 at C/2 and 1 C rates, respectively which is from 84 to 36% of the theoretical capacity for active material LiCoO 2 [bib_ref] Cross-Linked Polymer Electrolytes for Li-Based Batteries: From Solid to Gel Electrolytes, Chaudoy [/bib_ref] [bib_ref] Liquid-in-Solid' and 'solid-in-Liquid' Electrolytes with High Rate Capacity and Long Cycling Life..., Wu [/bib_ref]. The reversibility of the electrolyte is at 87, 54 and 40% for C/10, C/5 and 1 C rates, respectively. Its corresponding rate capability curve is displayed in. The outstanding performance of DFOB-GPE3 with different rates is ascribed to satisfactory performance of ionic conductivity and favorable electrochemical properties between the electrodes and the electrolyte in the cell.represents the specific capacity as a function of number of cycles for LiCoO 2 /DFOB-GPE3/Li cell at C/10 rate. The cell delivers an initial discharge capacity of 148.4 mAhg −1 at first cycle performance. A slight decay in capacity occurs with increase of number of cycles and it delivers a discharge capacity of 120 mAhg −1 after 100 th cycle. The slight decrease in capacity during the first few cycles may be due to formation of passive layers over the lithium electrode surface during the cycling [bib_ref] Ionic liquid incorporated nanocomposite polymer electrolytes for rechargeable lithium ion battery: A..., Karuppasamy [/bib_ref] [bib_ref] Electrochemical Properties and Thermal Stability of Li a Ni 1Ϫx Co x..., Cho [/bib_ref] [bib_ref] Investigation on modified natural rubber gel polymer electrolytes for lithium polymer battery, Ali [/bib_ref] [bib_ref] Capacity Fade Mechanisms and Side Reactions in Lithium-Ion Batteries, Arora [/bib_ref]. Moreover, no obvious changes are observed in charge capacity after few cycles which infers from the capacity retention curve as shown in. The columbic efficiency of the first cycle is merely 93.8% which increases gradually with number of cycles and observed at 98.2% for 100 th cycle. These findings imply that the DFOB-GPE3 possesses high specific capacity with very good capacity retention at ambient temperature and would be highly suitable with lithium anode and LiCoO 2 cathode for lithium ion battery applications.
# Conclusions
In this paper, we have demonstrated the design and fabrication of novel ternary gel polymer electrolytes comprising of borate anion (DFOB) based ionic liquid and lithium salt with high electrochemical and cycling stability. The interactions between the ions and polymer were demonstrated by IR shifting of the ν C=O mode for electrolyte mixture as well as the ν a (CF 2 ) mode for PVdF-co-HFP. As obtained DFOB-GPEs have very high thermal stability with appropriate mechanical strength. The DFOB-GPE3 which comprises 80% LiDFOB/EMImDFOB and 20% PVdF-co-HFP has high ionic conductivity (10 −4 Scm −1 ) and ionic transference number (0.37) at room temperature. The prepared LiCoO 2 /DFOBGPE3/Li cell possesses very good capacity retention and it provides a maximum of 148.4 mAhg −1 at C/10 rate. These outstanding properties of DFOB-GPEs pave a way to utilize as a potential candidate for application in future large scale rechargeable lithium ion batteries.
# Materials and methods
Source materials. Poly (vinylidinefluoride-co-hexafluoropropyline) (PVdF-co-HFP pellets, average molecular weight, M w 4 × 10 5 ), 1-ethyl 3-methyl imidazolium bromide (EMImBr, ≥99.1% pure, M w 169.25), anhydrous solvents such as acetonitrile and dichloromethane were gifted from Sigma Aldrich. Lithium difluoro(oxalate) borate (LiDFOB, MW = 143.77 g/mol) were purchased from Suzhou Fosai New Material Co., Ltd., China. The lithium salt and other precursors were kept under vacuum at 50 and 70 °C, respectively for 12 h prior to use.
## Synthesis of 1-ethyl-3-methylimidazolium difluoro(oxalate) borate (emimdfob). the
EMImDFOB was synthesized by metathesis reaction between 1-ethyl-3-methylimidazolium bromide (EMImBr) and LiDFOB. In brief, 50 ml solution of EMImBr (0.4 M) was instinctively stirred for 4 h under nitrogen (N 2 ) atmosphere at 60 °C. Then, 0.43 M solution of LiDFOB was added drop wise into stirred EMImBr solution. The obtained homogeneous solution was continuously stirred for 12 h at 60 °C under N 2 atmosphere. The final product was extracted by rotatory with dichloromethane (DCM). After the DCM evaporation, it was further vacuum dried at 80 °C for 24 h and collected the yield of 58%. In the whole reaction process, double deionized water was used as a solvent. The as-prepared ionic liquid of EMImDFOB was stored in a glove box for further characterization. The excellent purity of the sample was further confirmed by 1 H and 19 F.
## Preparation of difluoro(oxalate)borate gel polymer electrolytes (dfob-gpes).
Firstly, to prepare electrolyte mixture of EMImDFOB/LiDFOB, 0.3 M of LiDFOB was dissolved in neat ionic liquid (EMImDFOB) and then mechanically stirred for 8 h at 60 °C under N 2 atmosphere. The resultant solution mixture was degassed for about 15 min. The various ratios of polymer and electrolyte mixture (as shown in were dissolved in anhydrous acetonitrile and constantly mixed for 8 h to get homogeneous viscous solution. Then different composition mixtures of EMImDFOB/LiDFOB and PVdF-co-HFP were prepared by solvent-casting method as reported earlier [bib_ref] Electrochemical and cycling performances of novel nona fl uorobutanesulfonate (nona fl ate)..., Karuppasamy [/bib_ref]. The whole process was carried out in an inert atmosphere. Surface of the obtained DFOB-GPEs films was constructed with copolymer network which consisting of EMImDFOB/LiDFOB as shown in . In order to avoid moisture effects, the prepared DFOB-GPEs films were stored in the glove box for further application and characterization analyses. [bib_ref] UV-cured polymer electrolytes encompassing hydrophobic room temperature ionic liquid for lithium batteries, Gerbaldi [/bib_ref] F-NMR spectra were taken on a Bruker avance 400 spectrometer. The water content of synthesized ionic liquid was determined by Karl-Fisher coulometric moisture titrator using a Mettler 831 KF autotitrator (Metrohm Co.,). Viscosity (η) measurement was performed using Modular compact rheometer (Physica MCR 301, Anton Paar).
## Characterization techniques. 1 h-nmr and
The crystalline nature of synthesized gel electrolytes were examined by D-MAX 2500, Rigaku, X-ray diffractometer instrument equipped with Cu-Kα (λ = 1.5406 A°) radiation source, with a scan speed of 2° per minute, ranging from 0° to 100°. The differential scanning calorimetry (DSC) measurement of pure polymer host and DFOB-GPEs were made on TA instruments, (Model 2920) thermal analyzer at a heating rate of 10 °C min −1 under nitrogen atmosphere in the temperature range of −50 to 200 °C. The thermogravimetry (TGA) analyses of DFOB-GPEs were carried out using TGA 2950 (Hi-Res, TA instruments). The TG profiles were recorded under N 2 flow (20 mL min −1 ), in a temperature range between 25 and 800 °C. A temperature ramp rate of 20 °C min −1 was used. FTIR spectra were recorded with the help of a Nicolet 380 FT-IR spectrometer (Thermo Electron) in the region 4000-400 cm −1 at a signal resolution of 1 cm −1 . The mechanical stability of the gel electrolytes were measured using Instron Tester 6025 with computer evaluation. The ionic conductivities of prepared DFOB-GPEs were analyzed by electrochemical impedance spectroscopy (EIS). The electrolyte samples were performed in blocking type cells where the DFOB-GPEs were sandwiched between two stainless steel electrodes. The ionic conductivity of the DFOB-GPEs having an area of 1.7665 cm 2 were measured using ac-impedance spectroscopy (Ivium Technologies, Netherlands) in the frequency range (1-10 6 ) Hz with amplitude signal of 10 mV. The temperature dependence of ionic conductivity was performed in the temperature range between 298 and 398 K. The ionic conductivity of DFOB-GPEs was calculated using following equation,
[formula] σ = t R A / (4) b [/formula]
where t is the thickness of the DFOB-GPEs, R b is the bulk resistance and A is the area of electrode-electrolyte contact.
The total lithium transference number (t Li + ) of prepared DFOB-GPEs was measured using AC impedance and DC polarization method. A step voltage of 10 mV was applied across the symmetrical Li/DFOB-GPE3/Li cell and the resulting current was measured as a function of time (chronoamperometry) at 333 K.
The linear sweep voltammetry and cyclic voltammetry analyses of DFOB-GPEs were examined using Pt/ DFOB-GPEs/Li cell at 298 K. These analyses were carried out using Ivium Technologies electrochemical workstation. For LSV, the potential range fixed from −5 to +6 V at a scan rate of 20 mV/s. CV analysis was performed at different scan rates in the potential limit between 0 and +5 V. The cycling performance of DFOB-GPE3 was observed in galvanostatic mode using Arbin battery cycler. Charge-discharge analysis was carried out for the 2032 coin cell assembled by sandwiching the DFOB-GPE3 between the commercially available LiCoO 2 cathode foil and lithium metal anode. The coin cell was cycled between the cut-off voltage of 3.0 and 4.2 V at three different C-rates at room temperature.
[fig] Figure 1: (a) The working module of LIBs using synthesized EMImDFOB/LiDFOB gel electrolytes. [/fig]
[fig] Figure 2: (a) XRD patterns of DFOB-GPEs; (b) DSC thermogram curves of DFOB-GPEs and inset figure shows (c) DSC thermogram of DFOBGPEs in the temperature region between −50 and 20 °C (d) DSC thermogram curve of pristine PVdF-co-HFP and (e) DSC thermogram of DFOBGPEs in the temperature region between 150 and 200 °C. ScIEntIfIc RePORtS | 7: 11103 | DOI:10.1038/s41598-017 [/fig]
[fig] Figure 3: FT-IR spectra of DFOB-GPEs in the wavenumber regions (a) 2250-1200 cm −1 (The highlighted regions I and II represents symmetrical ν C=O and vibrational δ CH of imidazolium ring (region II) for DFOB-GPEs (b) 1160-1100 cm −1 and (c) 3300-3000 cm −1 (The highlighted regions I and II represents and symmetrical ν N-H (region I) and ν C-H of imidazolium ring (region II) for DFOB-GPEs. [/fig]
[fig] Figure 4: (a) TGA thermogram of DFOB-GPEs (Inset -TGA curve for pristine PVdF-co-HFP); (b) Possible intermolecular mechanism between PVdF-co-HFP and EMImDFOB/LiDFOB in DFOB-GPEs (c) Stress strain curves of pristine PVdF-co-HFP and DFOB-GPEs. [/fig]
[fig] Figure 5: (a) Temperature dependence of Ionic conductivity for DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3; (b) Comparison of activation energy and log σ RT as function of electrolyte mixture content; (c) Nyquist impedance plot of DFOB-GPE3 and its corresponding equivalent circuit (inset); (d) Chronoamperometry curve of DFOB-GPE3. ScIEntIfIc RePORtS | 7: 11103 | DOI:10.1038/s41598-017-11614-1 conductivity than that of presented other gel electrolyte systems. The lithium transference number (t Li + ) of DFOB-GPEs was obtained by equation ( [/fig]
[fig] Figure 6: (a) linear sweep voltammogram of DFOB-GPE1, DFOB-GPE2 and DFOB-GPE3; (b) Cyclic voltammogram of high conducting DFOB-GPE3 at different cycles; (c) Cyclic voltammogram of high conducting DFOB-GPE3 at different scan rates (inset-Peak potential (E p ) as a function of scan rates). ScIEntIfIc RePORtS | 7: 11103 | DOI:10.1038/s41598-017-11614-1 [/fig]
[fig] Figure 7: (a) Charge-discharge plateau of LiCoO 2 /DFOB-GPE3/Li cell at different C-rates; (b) Rate capability curve of LiCoO 2 /DFOB-GPE3/Li cell at different C-rates; (c) Charge-discharge capacities of LiCoO 2 /DFOB-GPE3/Li cell for (1 st , 50 th and 100 th ) for 100 cycles at C/10 rate; (d) Columbic efficiency plot of LiCoO 2 /DFOB-GPE3/Li cell at C/10 C-rate. ScIEntIfIc RePORtS | 7: 11103 | DOI:10.1038/s41598-017-11614-1 [/fig]
|
Deep Hypothermic Circulatory Arrest with Lung Perfusion/Ventilation in a Patient with Acute Type A Aortic Dissection
A 50-year-old black male presented with acute type A aortic dissection. Surgical repair was performed under deep hypothermic circulatory arrest (DHCA) with lung perfusion/ventilation throughout the procedure. Details of the lung perfusion technique and its potential benefits and drawbacks are discussed.
# Introduction
Acute type A aortic dissection, involving the ascending aorta, has an incidence of 5 to 30 cases per million people per year in the United States [bib_ref] The International Registry of Acute Aortic Dissection (IRAD): new insights into an..., Hagan [/bib_ref]. Untreated, death usually occurs within 48 hours in 68% of patients (1%-2% deaths per hour) most commonly due to rupture of the proximal aorta into the pericardial cavity causing cardiac tamponade [bib_ref] Epidemiology and clinicopathology of aortic dissection: a population-based Figure 3: ETCO 2..., Mészáros [/bib_ref] [bib_ref] Aortic dissection: new frontiers in diagnosis and management. Part I: from etiology..., Nienaber [/bib_ref]. In the event of a type A dissection, immediate surgery is indicated. The procedure usually involves replacement of the ascending aorta and reconstruction of the aortic wall by suturing the true and false lumens together, placement of a Dacron graft, and resuspension of the native aortic valve. Surgical mortality ranges from 10% to 25%, but as noted, without surgery, death is likely [bib_ref] Advances in the treatment of acute type A dissection: an integrated approach, Bavaria [/bib_ref] [bib_ref] Surgery for acute type A aortic dissection: is advanced age a contraindication?, Chiappini [/bib_ref] [bib_ref] Contemporary results of surgery in acute type a aortic dissection: the International..., Trimarchi [/bib_ref] [bib_ref] Paradigm change in the management of patients with acute type A aortic..., Hassan [/bib_ref]. The main cause of death in those patients having a surgical repair is usually related to respiratory problems, such as ARDS with failure, pulmonary embolism, and multiple organ failure. Even in those patients who survive surgical repair, morbidity can be high with once again respiratory failure predominating [bib_ref] Open distal anastomosis in aortic root replacement using axillary cannulation and moderate..., Takayama [/bib_ref] [bib_ref] Surgical treatment of type A aortic dissections: results with profound Case Reports..., Ehrlich [/bib_ref].
DHCA has become a standard of care for patients undergoing type A repairs as it facilitates the surgical procedure.
However, DHCA in itself is associated with a longer duration of cardiopulmonary bypass (CPB), coagulopathy, and postoperative bleeding [bib_ref] Repair of thoracic aortic aneurysm and dissection using deep-hypothermic circulatory arrest, Manapat [/bib_ref]. A major cause of morbidity and mortality associate with DHCA, as with the surgical repair of the aorta itself, is respiratory failure, which in turn leads to prolonged mechanical ventilation [bib_ref] Open distal anastomosis in aortic root replacement using axillary cannulation and moderate..., Takayama [/bib_ref].
During CPB, with or without DHCA, lung collapse is considered a standard part of the technique. It is believed that by allowing the lungs to collapse, the overall incidence of pulmonary complications is increased. It has been proposed that there may be an advantage to ventilating the lungs of patients undergoing CPB and in particular those under DHCA.
Herein, we report the use of DHCA with lung perfusion/ ventilation during the repair of an acute type A dissection. This technique's potential benefits and drawbacks are also discussed.
## Case report
A 50-year-old black male, with a history of drug addiction, was admitted with shortness of breath and peripheral edema. He denied prior medical problems and/or a history of acute chest or back pain. He had severe jugular venous distention, a 4/6 continuous diastolic murmur throughout the precordium, +3 pitting edema in the legs up to the knees, bibasilar crackles, and pulsatile hepatomegaly. Chest X-ray showed an enlarged heart with no evidence of pulmonary edema. Transthoracic echocardiography demonstrated type A aortic dissection, severe aortic and tricuspid insufficiency, and no pericardial effusion. CT angiography confirmed the diagnosis of a type A aortic dissection, beginning 2 cm from the left coronary ostium and extending to both iliac arteries. The false lumen was noted to extend into the left common carotid artery, and the left subclavian artery was occluded at its origin. The patient gave consent for the publication of this case report. The requirement for written informed consent was waived by the local IRB.
## Operative technique.
At the time of surgery, the central venous pressure (CVP) was noted to be 47 mmHg, and the PaO 2 was 122 mmHg on an FiO 2 of 100%. Transesophageal echocardiogram (TEE) confirmed the angiographic finding of a type A dissection, severe aortic and tricuspid insufficiency, and mild mitral regurgitation. Further findings demonstrated that the right atrium was notably enlarged and, the right ventricular (RV) function was decreased. The left ventricular ejection fraction was noted at 40%.
Arterial inflow was via the right subclavian artery (no. 8 mm Hemashield graft sewn to the side of the subclavian artery) and bicaval venous cannulation. CPB was instituted, and systemic cooling was initiated. The right atrium was opened, and the coronary sinus was cannulated. From the arterial cannula, a no. 8 mm polystan cannula was used to perfuse the lungs via the main pulmonary artery at flows of 300 mL/min (this is a current internal standard based on normal pulmonary artery flow rates of approximately 300-400 mL/min). The lungs were ventilated with a tidal volume of 400 mL (calculated at 4-6 mL/kg), at a rate of 8-10 breaths/min, and an FiO 2 of 21-40% was utilized throughout the procedure. Ventilation was adjusted, if needed, based on arterial blood gases results and presence of ETCO 2 . A vent was inserted into the left atrium via the fossa ovale. The aorta was cross-clamped when ventricular fibrillation occurred at approximately 28 degrees Celsius. Cold continuous blood cardioplegia (or cold blood, once arrest occurred) was administered via the coronary sinus (pressure 50 mmHg, flows 250-300 mL/min) and simultaneously into the left coronary ostium once the aorta was opened. A tear was identified in the midportion of the ascending aorta. The false and true lumens were sewn together using a twofelt suture technique with resuspension of the aortic valve. DHCA was induced at 18 degrees Celsius. Cerebral perfusion (pressure 50 mmHg, flow 800 mL) was achieved by clamping the innominate artery. The arch was repaired using a similar technique as for the proximal aorta. A no. 33 Hemashield graft was interposed. After deairing, systemic flow was resumed, and the patient was slowly rewarmed. During this period, the tricuspid valve was repaired using a DeVega plasty. The remainder of the procedure was uneventful. The patient required minimal ionotropic support to separate from CPB. Postprocedure TEE revealed mild AI and mild TR. Blood gas measurements, immediately after weaning from CPB, showed improvement in PaO 2 /FiO 2 ratio [fig_ref] Table 1: Arterial blood gases analysis [/fig_ref] , with a PaO 2 of 579 mmHg. The patient was extubated 12 hrs postoperatively. He made an uneventful recovery and was discharged home on day 8.
# Discussion
CPB-induced lung injury remains as an important cause of morbidity and mortality after cardiac surgery. The pathophysiology of post-CPB lung dysfunction is multifactorial, and our understanding remains incomplete despite many years of research into this phenomenon. During CPB, pulmonary artery blood flow is discontinued, and nonpulsatile flow is established, resulting in a low mean systemic perfusion pressure. Decreased bronchial artery flow is one of the etiological factors in ischemia-reperfusion injury of the lungs [bib_ref] Bronchial artery perfusion during cardiopulmonary bypass does not prevent ischemia of the..., Schlensak [/bib_ref] [bib_ref] Pulmonary protection during cardiac surgery: systematic literature review, Carvalho [/bib_ref]. Furthermore, lung collapse, from discontinuation of mechanical ventilation, aortic manipulation, and hypothermia, further contributes to pulmonary dysfunction, manifested by atelectasis, intrapulmonary shunting, poor gas exchange, pulmonary edema, and the need for prolonged artificial ventilation. Pulmonary artery perfusion, with or without ventilation, has been reported as a potential intervention aimed at preventing lung ischemia during CPB [fig_ref] Figure 1: ETCO 2 before CPB. [/fig_ref] , and end-tidal CO 2 has been used as an indication of pulmonary blood flow during CPB (18).
Lung perfusion/ventilation has been routinely used in valve surgery at our institution (13, 19) for the past 5 years; however, it has not been used during DHCA. To the best of our knowledge, this is the first case report in which lung perfusion/ventilation was utilized during DHCA in humans. Lung perfusion/ventilation during DHCA has been investigated in animal models [bib_ref] Bronchial artery perfusion during cardiopulmonary bypass does not prevent ischemia of the..., Schlensak [/bib_ref] , demonstrating beneficial effects in terms of gas function and morphology. In our case, during the period of DHCA with ventilation and perfusion, the lungs continued to produce end-tidal CO 2 , indicating pulmonary perfusion and alveolar gas exchange [fig_ref] Figure 1: ETCO 2 before CPB. [/fig_ref]. Furthermore, as shown in [fig_ref] Table 1: Arterial blood gases analysis [/fig_ref] , blood gases immediately after CPB showed dramatic improvement in PaO 2 compared to baseline measurements. Although this improvement may have been related to surgical correction of the aortic and tricuspid insufficiency, better lung protection is also a possible explanation. Our clinical experience with lung perfusion/ventilation during valve surgery has shown no adverse effects, and no complications associated directly with this technique. However, whether this technique provides clinical benefits remains controversial. Much remains to be determined, such as what is the optimal pulmonary artery flow rates and whether excessive perfusion of the lungs results in pulmonary edema/hemorrhage. Matching perfusion to ventilation is another critical goal to be achieved. Ventilation/perfusion mismatch during CPB with lung perfusion/ventilation may potentially lead to dead space ventilation, shunting, barotrauma, and alveolar edema. Additionally, most of the cardiac surgeons have limited or no experience in utilizing lung perfusion/ventilation during CPB and may have issue with blood in the operative field. We hope that this paper will lead others to pursue investigation into lung perfusion/ventilation either experimentally or clinically. We continue to use this technique in all cardiac procedures in which CPB is used.
[fig] Figure 1: ETCO 2 before CPB. [/fig]
[table] Table 1: Arterial blood gases analysis.Figure 2: ETCO 2 during CPB. [/table]
|
Lack of Transmission of Zika Virus Infection to Breastfed Infant
[bib_ref] Isolations and serological specificity, Dick [/bib_ref] [bib_ref] Zika virus outbreak on Yap Islands, federated states of Micronesia, Duffy [/bib_ref] [fig_ref] Figure 1: Timeline of symptoms and progression for mother and child [/fig_ref] [fig_ref] Figure 2: RT-PCR results for mother and child [/fig_ref]
## Lack of transmission of zika virus infection to breastfed infant
There has been limited evidence to confirm ZIKV transmission through breastfeeding and early post-partum cases of infants infected with ZIKV were most likely via transplacental or during delivery. [bib_ref] Transmission of Zika virus through breast milk and other breastfeeding-related bodily fluids:..., Colt [/bib_ref] [bib_ref] Evidence for mother-to-child transmission of Zika virus through breast milk, Blohm [/bib_ref] [bib_ref] Evidence of perinatal transmission of Zika virus, Besnard [/bib_ref] Our study demonstrates that breast milk contains the ZIKV RNA by RT-PCR consistent with other previous case reports. [bib_ref] Infectious Zika viral particles in breastmilk, Dupont-Rouzeyrol [/bib_ref] [bib_ref] Persistence of Zika virus in breast milk after infection in late stage..., Sotelo [/bib_ref] In our study, breast milk does not appear to be infectious to the baby through breastfeeding, even though it was daily fed to baby for 11 days. In cases where infants were infected with ZIKV, no developmental sequelae or permanent deficit demonstrated at 30 months. [bib_ref] Outcomes for 2 children after peripartum acquisition of Zika virus infection, Besnard [/bib_ref] Therefore, breastfeeding should be encouraged despite suspicion or confirmed maternal infection. Variables which may affect viral transmission through breastfeeding such as immune status, milk conditioning, viral load, and the presence of viable ZIKV need to be investigated to ensure validity of the recommendation. As of now, ZIKV transmission via the gastrointestinal system is not observed. Limitations to this study include the fact that baby's serum was not tested by RT-PCR; however, we believe that serological assays and urine RT-PCR provide adequate sensitivity for detection of ZIKV. 11
# Author contributions
PH interpreted the data and wrote the paper, SW conceived and designed the analysis, RB collected the data, SP performed the analysis, SB performed the analysis, CR performed the analysis, PR collected the data and TH performed critical revision.
[fig] Figure 1: Timeline of symptoms and progression for mother and child; breastfeeding stopped on day 12(December 12, 2017).RT-PCR, real-time polymerase chain reaction. [/fig]
[fig] Figure 2: RT-PCR results for mother and child; dates correlate withFigure 1.RT-PCR, real-time polymerase chain reaction. milk with detectable ZIKV RNA. Blood samples for Zika IgG and IgM serology testing on January 5 and 30, 2018 were negative. Parents were also tested for anti-Zika IgG and IgM drawn on December 26, 2017, [/fig]
|
Biogenic Synthesis of Copper-Based Nanomaterials Using Plant Extracts and Their Applications: Current and Future Directions
# Introduction
Apart from food, plants have traditionally been used intensively in textile, cosmetics, and medicine. Beyond traditional Chinese medicine and Ayurveda, which have developed over thousands of years of interest in the prevention and treatment of diseases, the biomedical applications of plants have broadened even further due to the advancement of technology and of time [bib_ref] Translational Potential of Ayurveda Prakriti: Concepts in the Area of Personalized Medicine, Coffin [/bib_ref] [bib_ref] The Evolution of Traditional Chinese Medicine as a Disciplinary Concept and Its..., Liu [/bib_ref] [bib_ref] What Has Traditional Chinese Medicine Delivered for Modern Medicine?, Wang [/bib_ref] [bib_ref] Techniques for Extraction and Isolation of Natural Products: A Comprehensive Review, Zhang [/bib_ref]. These biomedical applications are mainly due to the phytochemicals within plants [bib_ref] The Medicinal Uses of the Genus Bletilla in Traditional Chinese Medicine: A..., Jiang [/bib_ref] [bib_ref] Potential Implications of Ayurveda in Psoriasis: A Clinical Case Study, Nille [/bib_ref] [bib_ref] Current Understanding of Phytochemicals from Chinese Herbal Medicines for Ferroptosis-Mediated Cancer Prevention..., Hu [/bib_ref] , which are among the most fascinating aspects of plants due to their having activities such as antimicrobial, antitumour, antiaging, and others [bib_ref] Significance of Artemisia vulgaris L. (Common Mugwort) in the History of Medicine..., Ekiert [/bib_ref]. Awareness of such properties has driven researchers to discover still more applications of phytochemicals. In 1959, Richard Feynman illustrated the controlling of single atoms and molecules under the topic of "There's plenty of room at the bottom", which first shed light Cymbopogon jwarancusa Shade drying at room temperature Grinding Boiling
Step 1: 100
Step 2: 37
Step
# Plant extraction method
The first step of plant extraction is the cleaning process, which mainly aims to remove debris or dust with water so as to avoid any form of contamination that might affect the subsequent synthesis process. The second step consists of drying and downsizing. Drying is necessary to avoid the deterioration of phytochemicals that results from enzymatic and microbial activities due to the presence of water moisture [bib_ref] Review of Leaf Drying: Mechanism and Influencing Parameters, Drying Methods, Nutrient Preservation,..., Babu [/bib_ref].
## Drying
Typically, drying is performed via air drying, shade drying, oven drying, drying in a dehydrator, vacuum drying, sun drying, or on filter paper; plant materials can also be acquired in the dry form [fig_ref] Table 1: Parameters and extraction method utilised for extraction of different plant components [/fig_ref] and [fig_ref] Figure 1: General steps in plant extraction [/fig_ref].
Each of the abovementioned drying methods is able to successfully yield plant extracts with phytochemicals. Shade and air drying are considered among the best methods as they allow the greatest preservation of nutrients, such as proximate and ascorbic acid, and do so with lower financial cost as compared to mechanical drying methods such as oven drying, vacuum drying, or using a food dryer [bib_ref] Review of Leaf Drying: Mechanism and Influencing Parameters, Drying Methods, Nutrient Preservation,..., Babu [/bib_ref]. As a case in point, tangerine peel was shade dried at 27 ± 2 - C for the synthesis of iron oxide nanoparticles [bib_ref] Cadmium Removal from Aqueous Solution by Green Synthesis Iron Oxide Nanoparticles with..., Ehrampoush [/bib_ref]. However, due to being carried out at a lower temperature, shade and air drying require a longer period of time than other drying methods, which might reduce their applicability in the industrial plant-mediated synthesis of Cu-NMs [bib_ref] Review of Leaf Drying: Mechanism and Influencing Parameters, Drying Methods, Nutrient Preservation,..., Babu [/bib_ref]. For instance, in preparation for Au nanoparticle synthesis, Nepeta leucophylla root was shade dried at room temperature (24-32 - C) for 30 days [bib_ref] Novel Green Synthesis and Characterization of the Antioxidant Activity of Silver Nanoparticles..., Singh [/bib_ref]. Sun drying was also used in drying Chromolaena odorata for the synthesis of Fe 3 O 4 nanoparticles from phenolic components of the extract [bib_ref] Synthesis and Characterization of Magnetite (Fe 3 O 4 ) Nanoparticles Using..., Nnadozie [/bib_ref]. While sun drying can reduce the cost of drying just as can shade drying, it is not recommended for industrial synthesis due to high labour demand, low efficiency, hygiene issues, and more precautions being required to avoid contamination of samples [bib_ref] Review of Leaf Drying: Mechanism and Influencing Parameters, Drying Methods, Nutrient Preservation,..., Babu [/bib_ref].
The temperature of the drying process also plays a major role in preserving the phytochemicals within a plant. Specifically, drying temperatures in the range of 40-60 - C are reported to support the minimal loss of phytochemicals in plant components [bib_ref] Review of Leaf Drying: Mechanism and Influencing Parameters, Drying Methods, Nutrient Preservation,..., Babu [/bib_ref]. In prior studies, neem leaves (Azadirachta indica) were oven dried for 15 min at 50 - C [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref] , and Garcinia mangostana peelings for 10 min at 40 - C [bib_ref] Green Synthesis of Gold Nanoparticles Using Aqueous Extract of Garcinia mangostana Fruit..., Xin Lee [/bib_ref]. Although the range of 40-60 - C is recommended, the final decision on which temperature is most suitable for the drying process should be based on the characteristics of the plant material being dried. For example, a study oven dried Arachis hypogaea at 70 - C for 30 min due to its anthocyanin content, which is highly preserved under those drying parameters [bib_ref] Green Synthesis of Iron Nanoparticles Using Red Peanut Skin Extract: Synthesis Mechanism,..., Pan [/bib_ref]. Nonetheless, drying at room/ambient temperature remains the most used method owing to the low cost requirement being beneficial to industrialization. For instance, Irum et al. [bib_ref] Biogenic Iron Oxide Nanoparticles Enhance Callogenesis and Regeneration Pattern of Recalcitrant Cicer..., Irum [/bib_ref] shade dried C. jwarancusa at room temperature while Elgorban et al. [bib_ref] Green Synthesis of Cu/Fe 3 O 4 nanocomposite Using Calendula Extract and..., Elgorban [/bib_ref] dried calendula flowers at room temperature to acquire phytochemicals. This is despite the time requirement being much higher; for instance, when drying at room temperature, Yulizar et al. [bib_ref] ZnO/CuO Nanocomposite Prepared in One-Pot Green Synthesis Using Seed Bark Extract of..., Yulizar [/bib_ref] took a week to dry Theobroma cacao seeds and Rautela et al. [bib_ref] Debnath (Das), M. Green Synthesis of Silver Nanoparticles from Tectona Grandis Seeds..., Rautela [/bib_ref] 3-4 days for Tectona grandis seeds in preparation for nanoparticle synthesis. On the other hand, Pan et al. [bib_ref] Green Synthesis of Iron Nanoparticles Using Red Peanut Skin Extract: Synthesis Mechanism,..., Pan [/bib_ref] only need 30 min to dry Arachis hypogaea with an oven at 70 - C in plant-mediated iron nanoparticle synthesis, while Doan Thi et al. [bib_ref] Synthesis of ZnO Nanoparticles Using Orange Fruit Peel Extract for Antibacterial Activities, Doan Thi [/bib_ref] took 12 h to dry orange peels for ZnO nanoparticle production.
In addition to the abovementioned drying methods, some plant components simply are not subjected to any drying process, mainly in the interest of cost saving and because certain components have high water contents that will increase the cost if a drying process is applied. Such plant components can include fruits, flowers, seeds, roots, and rhizomes. In one example, Jahan et al. [bib_ref] Isildak, I. Facile Microwave-Mediated Green Synthesis of Non-Toxic Copper Nanoparticles Using Citrus..., Jahan [/bib_ref] squeezed the juice from Citrus sinensis fruits to acquire reducing sugars, amino acids, proteins, and metabolites such as flavanones and terpenoids for the synthesis of Cu nanoparticles. The same squeezing method was also applied to Zingiber officinale root by Velmurugan et al. [bib_ref] Green Synthesis of Silver and Gold Nanoparticles Using Zingiber officinale Root Extract..., Velmurugan [/bib_ref] to acquire alkaloids and flavonoids for the synthesis of Au and Au nanoparticle. Crushing is another technique for acquiring plant extracts; for example, Kumari et al. [bib_ref] Green Synthesis and Applications of Au-Ag Bimetallic Nanoparticles, Meena Kumari [/bib_ref] crushed pomegranate seeds to obtain flavonoids and terpenoids for the synthesis of Au-Ag bimetallic nanoparticles. Moreover, some methods forgo any drying treatment, such as when Patra et al. [bib_ref] Synthesis and Differential Antibacterial Activity of Bioconjugated Bimetallic Nanoparticles, Patra [/bib_ref] directly extracted Muntingia calabura flowers to acquire phytochemicals for nanoparticle synthesis and Al-Radadi [bib_ref] Facile One-Step Green Synthesis of Gold Nanoparticles (AuNp) Using Licorice Root Extract:..., Al-Radadi [/bib_ref] used licorice root without drying to obtain glycosides, organic acids, phenolic compounds, and flavonoids for the synthesis of Au nanoparticles. Ultimately, the characteristics of the plant component being used and the potential cost are important factors informing the best drying method and parameters by which to obtain the most phytochemicals from plant components for Cu-NM synthesis for either research or industrial purposes.
## Downsizing
Regarding the downsizing step, its primary purpose is to reduce the size of the plant components and increase their surface area, leading to better diffusivity and mass transfer in order to extract the greatest yields of phytochemicals such as polyphenolic compounds, phenolic acids, and tannins [bib_ref] Techniques and Modeling of Polyphenol Extraction from Food: A Review, Sridhar [/bib_ref]. There are various routes for achieving this objective, presented in [fig_ref] Table 1: Parameters and extraction method utilised for extraction of different plant components [/fig_ref] and [fig_ref] Figure 1: General steps in plant extraction [/fig_ref]. Interestingly, miniscule deviations of plant component size can cause significant alterations in overall phytochemical yield [bib_ref] Techniques and Modeling of Polyphenol Extraction from Food: A Review, Sridhar [/bib_ref]. Therefore, it is necessary to consider carefully the most suitable methods and cost requirements so as to acquire the smallest plant components with the highest phytochemical yields for Cu-NM synthesis. Just as with the drying process, there are some plant components that do not undergo any downsizing, such as those with high water content; for example, Crataegus pentagyna fruits were extracted by Ebrahimzadeh et al. [bib_ref] Eco-Friendly Green Synthesis and Characterization of Novel Fe 3 O 4 /SiO..., Ebrahimzadeh [/bib_ref] without any downsizing.
# Plant extraction methods
Plant extraction methods are mainly based on boiling and heating [fig_ref] Table 1: Parameters and extraction method utilised for extraction of different plant components [/fig_ref]. Mani et al. [bib_ref] Studies on the Spectrometric Analysis of Metallic Silver Nanoparticles (Ag NPs) Using..., Mani [/bib_ref] conducted an extraction from dried, ground, and pulverised Basella alba leaves by mixing them with DI-H 2 O and boiling them in a water bath at 60 - C for 20 min. Nnadozie and Ajibade [bib_ref] Synthesis and Characterization of Magnetite (Fe 3 O 4 ) Nanoparticles Using..., Nnadozie [/bib_ref] similarly heated crushed Chromolaena odorata at 85 - C for 2 h in DI-H 2 O, and Abisharani et al. [bib_ref] Green Synthesis of TiO2 Nanoparticles Using Cucurbita pepo Seeds Extract. Mater, Abisharani [/bib_ref] heated Cucurbita pepo seeds with DS-H 2 O at 90 - C for 2 h.
Interestingly, some alternative methods have been introduced and successfully used to extract phytochemical products from plants [fig_ref] Table 1: Parameters and extraction method utilised for extraction of different plant components [/fig_ref]. For example, Siddiqui et al. [bib_ref] Green Synthesis of Copper Oxide (CuO) Nanoparticles by Punica Granatum Peel Extract...., Siddiqui [/bib_ref] boiled powdered Punica granatum peels in sterile DI-H 2 O at 55 - C for 30 min on a Soxhlet apparatus, and Singh and Dhaliwal similarly performed Soxhlet extraction on powdered Nepeta leucophylla roots with methanol held at boiling for 8 h [bib_ref] Novel Green Synthesis and Characterization of the Antioxidant Activity of Silver Nanoparticles..., Singh [/bib_ref]. Sonication has also been used in plant extractions; for instance, one study removed the coats of Caesalpina bonducella seeds and then sonicated the ground kernels for 30 min [bib_ref] Green-Synthesized Rice-Shaped Copper Oxide Nanoparticles Using Caesalpinia bonducella Seed Extract and Their..., Sukumar [/bib_ref]. Likewise, reflux extraction has been used with various plant components. Beheshtkhoo et al. [bib_ref] Synthesis of Iron Oxide Nanoparticles by Aqueous Leaf Extract of Daphne Mezereum..., Beheshtkhoo [/bib_ref] extracted Daphne mezereum leaves by refluxing the dried leaves with a 5% (w/v) mixture in DI-H 2 O for 15 min. Microwave irradiation has also been used in extraction, such as in a study that irradiated cut Jasminum sambac leaves in DS-H 2 O for 200 s to extract phytochemicals for the synthesis of Au, Ag, and Au−Ag alloy nanoparticles [bib_ref] Phytosynthesis of Stable Au, Ag and Au-Ag Alloy Nanoparticles Using J. Sambac..., Yallappa [/bib_ref]. Maceration has also been used by many researchers, mainly due to its low cost and eco-friendliness; for instance, ground Solanum mammosum fruits were macerated with DI-H 2 O at room temperature and constant agitation for 1 h [bib_ref] Green Synthesis of Silver Nanoparticles Using Solanum mammosum L. (Solanaceae) Fruit Extract..., Pilaquinga [/bib_ref] and Crataegus pentagyna fruits with methanol at room temperature for the synthesis of Fe 3 O 4 -SiO 2 -Cu 2 O-Ag nanocomposites [bib_ref] Eco-Friendly Green Synthesis and Characterization of Novel Fe 3 O 4 /SiO..., Ebrahimzadeh [/bib_ref]. In addition to the above, autoclaving was carried out on dried and ground roots of Scutellaria baicalensis with DS-H 2 O for 30 min at 100 - C in preparation for the synthesis of ZnO nanoparticles [bib_ref] Green Synthesis of Zinc Oxide Nanoparticles from Root Extract of Scutellaria baicalensis..., Chen [/bib_ref].
Aside from single extraction methods, combinations of methods have also been applied to acquire extracts from various plants. For example, Nava et al. [bib_ref] Fruit Peel Extract Mediated Green Synthesis of Zinc Oxide Nanoparticles, Nava [/bib_ref] macerated the peels of Citrus aurantifolia, Citrus paradisi, Citrus sinensis and Lycopersicon esculentum for 3 h with stirring, then heated the mixture at 60 - C for 60 min. For plant components with high water content, a squeezing method may be introduced. For example, in the preparation of Zingiber officinale root extract by Velmurugan et al. [bib_ref] Green Synthesis of Silver and Gold Nanoparticles Using Zingiber officinale Root Extract..., Velmurugan [/bib_ref] , the downsized roots were squeezed via muslin cloth.
Every extraction method has its pros and cons, summarised in [fig_ref] Table 2: Pros and cons of various plant extraction methods [/fig_ref] and [fig_ref] Figure 1: General steps in plant extraction [/fig_ref] [bib_ref] Techniques for Extraction and Isolation of Natural Products: A Comprehensive Review, Zhang [/bib_ref] [bib_ref] Techniques and Modeling of Polyphenol Extraction from Food: A Review, Sridhar [/bib_ref] [bib_ref] Recent Advances in Extraction of Nutraceuticals from Plants, Wang [/bib_ref]. The most suitable method for any given use case depends on the types of plant components as well as the requirements and restriction posed by the actual environment, such as a need to reduce financial and labour costs for industrial purposes as well as a requirement for eco-friendliness.
## Solvents in plant extraction
In addition to the extraction method used, solvent, energy consumption, time required, and other parameters are also critical to the extraction of phytochemicals [bib_ref] Techniques and Modeling of Polyphenol Extraction from Food: A Review, Sridhar [/bib_ref] [bib_ref] Effects of Temperature, Time, and Solvent Ratio on the Extraction of Phenolic..., Sulaiman [/bib_ref]. Extraction solvents can be divided into two types. i.e., water (distilled, double distilled, Milli-Q, ultra-pure, and deionised) and alcoholic solvents (ethanol and methanol), as presented in [fig_ref] Table 1: Parameters and extraction method utilised for extraction of different plant components [/fig_ref]. Water (DS-H 2 O) was used to extract dried and ground Quercus coccifera leaves with boiling for 30 min at 90 - C [bib_ref] Kermes Oak (Quercus coccifera L.) Extract for a Biogenic and Eco-Benign Synthesis..., Kocadag Kocazorbaz [/bib_ref]. Conversely, Boruah et al. [bib_ref] Green Synthesis of Gold Nanoparticles Using an Antiepileptic Plant Extract: In Vitrobiological..., Boruah [/bib_ref] produced Moringa oleifera leaf extract by Soxhlet extraction with methanol as the solvent, incubating the dried and powdered leaves at 35-45 - C for 10 h. Some phytochemicals, such as polyphenolic compounds, anthocyanins, and polyphenols, can be obtained at higher yields when an alcoholic solvent is involved. Conversely, Do et al. [bib_ref] Effect of Extraction Solvent on Total Phenol Content, Total Flavonoid Content, and..., Do [/bib_ref] found that the phytochemical extraction yield from Limnophila aromatica improves as the solvent polarity increases; in particular, methanol could extract more phytochemicals than ethanol. There are also some cases that benefit from extraction solvents combining both water and an alcoholic solvent. For example, Piper longum fruits were dried, powdered and extracted with 30% methanolic solution at 70 - C for 30 min [bib_ref] Facile and Surfactant-Free Synthesis of Pd Nanoparticles by the Extract of the..., Nasrollahzadeh [/bib_ref] , and Zarei et al. [bib_ref] Green Synthesis of Ag Nanoparticles on the Modified Graphene Oxide Using Capparis..., Zarei [/bib_ref] used ethanol and water at a 1:1 ratio with boiling for 30 min. Remarkably, such combinations of alcoholic solvents with water can achieve the highest yields due to allowing for greater solubility of plant components [bib_ref] Techniques and Modeling of Polyphenol Extraction from Food: A Review, Sridhar [/bib_ref] [bib_ref] Effect of Extraction Solvent on Total Phenol Content, Total Flavonoid Content, and..., Do [/bib_ref] [bib_ref] Comparison of Extracts Prepared from Plant By-Products Using Different Solvents and Extraction..., Lapornik [/bib_ref]. Therefore, it could be concluded that for plant phytochemical extraction, an alcoholic aqueous solvent is generally the most suitable. Nonetheless, the specific characteristics of the plant and phytochemicals should be considered before applying a particular type of solvent. As a case in point, Maurya et al. [bib_ref] Green Synthesis of TiO 2 Nanoparticles Using Bixa orellana Seed Extract and..., Maurya [/bib_ref] produced Bixa orellana seed extract using ethanol mainly due to the primary phytochemical cis-bixin being water insoluble.
## Temperature in plant extraction
The temperature applied is also a crucial factor in the plant extraction process as it can greatly impact the yield and quality of phytochemicals and, thus, affect the nanoparticle synthesised. As listed in [fig_ref] Table 1: Parameters and extraction method utilised for extraction of different plant components [/fig_ref] , the temperature for extraction may range from room temperature to 100 - C. As an example of room-temperature extraction, Pilaquinga et al. [bib_ref] Green Synthesis of Silver Nanoparticles Using Solanum mammosum L. (Solanaceae) Fruit Extract..., Pilaquinga [/bib_ref] subjected pre-washed, oven-dried, and ground Solanum mammosum fruits to maceration with DI-H 2 O at room temperature with constant agitation for an hour, while, as an example of the highest temperature, Hu et al. [bib_ref] Rhodiola RoseaRhizome Extract-Mediated Green Synthesis of Silver Nanoparticles and Evaluation of Their..., Hu [/bib_ref] extracted Rhodiola rosea rhizome powder by heating with DI-H 2 O at 100 - C for 30 min. The temperature applied has a directly proportional relationship with solubility and diffusion. Nevertheless, when the temperature surpasses a particular threshold, it might lead to several problems such as solvent loss, introduction of impurities in the produced extract, and decomposition of thermolabile phytochemicals. For instance, when synthesizing Ag and Au nanoparticles from Impinella anisum seeds extracted at temperatures ranging from 25 to 60 - C, high surface plasmon resonance (SPR) peak intensities accompanied the raising of temperature due to the increased diffusion rate of the solvent, which destroyed the plant cell structure. However, when temperatures in the range of 60 to 85 - C were used, reduction in SPR was observed due to the decomposition of some thermolabile phytochemicals.
## Extraction time in plant extraction
Extraction time is another synergic factor that can greatly affect the phytochemicals extracted. Durations reported in the literature range from 200 s to a week; in addition, it can also be observed that the higher the temperature applied, the lower the extraction duration, and vice versa [fig_ref] Table 1: Parameters and extraction method utilised for extraction of different plant components [/fig_ref]. At the short end, Yallappa et al. [bib_ref] Phytosynthesis of Stable Au, Ag and Au-Ag Alloy Nanoparticles Using J. Sambac..., Yallappa [/bib_ref] conducted an extraction of Jasminum sambac leaves in DS-H 2 O assisted by microwave irradiation for 200 s. Meanwhile, for the longest duration, Yulizar et al. [bib_ref] ZnO/CuO Nanocomposite Prepared in One-Pot Green Synthesis Using Seed Bark Extract of..., Yulizar [/bib_ref] macerated Theobroma cacao seed bark powder in methanol with stirring for a week. Extending the extraction duration can improve extraction efficiency as the mass transfer coefficient between plant components and solvent increases; accordingly, longer extractions can boost the quantities of extracted phytochemicals and so enhance the formation of subsequently synthesised nanoparticles. However, such phenomena are restricted to within a certain time range, as when equilibrium has been reached inside and outside of the plant components, the extraction efficiency will not be further improved and could even worsen if the extraction period is excessively prolonged [bib_ref] Techniques for Extraction and Isolation of Natural Products: A Comprehensive Review, Zhang [/bib_ref]. For example, extraction of Impinella anisum seeds for 60 min results in the greatest band intensity for subsequently produced nanoparticles, and band intensity then declines as the extraction duration increases due to the oxidation and thermal decomposition of phytochemicals. Therefore, attentive consideration should be made regarding the duration of, and temperature during, phytochemical extraction.
## Filtration and preservation
After extraction, the next step is filtration, in which solid components are removed from the plant extract. There are many filtration techniques in use, as illustrated in [fig_ref] Figure 1: General steps in plant extraction [/fig_ref].
Following filtration, the obtained extracts are preserved for nanomaterial-synthesis research. Preservation is mainly achieved via refrigeration, directly using the extract for nanoparticle synthesis, or storing the extract in a container/environment with or without special conditions such as airtightness and light exclusion so as to avoid any manner of the oxidation or photodegradation of the phytochemicals. The temperature of refrigeration is mainly 4 - C as it was found that this temperature can best preserve the quality of Ananas comosus juice; moreover, increasing storage duration and temperature can greatly reduce the phytochemicals within the obtained plant extract [bib_ref] Storage Quality of Pineapple Juice Non-Thermally Pasteurized and Clarified by Microfiltration, Laorko [/bib_ref]. Therefore, in the green synthesis of Cu-NMs, the freshness of the plant extract is very significant. Once a plant extract is produced, it should be utilised for nanoparticle synthesis as soon as possible and, in the interim, stored at low temperature.
Finally, the obtained plant extract is prepared for the synthesis of nanomaterials; for example, Nasrollahzadeh et al. [bib_ref] Green Synthesis of CuO Nanoparticles Using Aqueous Extract of Thymus vulgaris L...., Nasrollahzadeh [/bib_ref] [bib_ref] Green Synthesis of the Copper Nanoparticles Supported on Bentonite and Investigation of..., Issaabadi [/bib_ref] produced Thymus vulgaris leaf extract and used it to synthesise CuO and Cu nanoparticles, as shown in.
improve extraction efficiency as the mass transfer coefficient between plant components and solvent increases; accordingly, longer extractions can boost the quantities of extracted phytochemicals and so enhance the formation of subsequently synthesised nanoparticles. However, such phenomena are restricted to within a certain time range, as when equilibrium has been reached inside and outside of the plant components, the extraction efficiency will not be further improved and could even worsen if the extraction period is excessively prolonged [bib_ref] Techniques for Extraction and Isolation of Natural Products: A Comprehensive Review, Zhang [/bib_ref]. For example, extraction of Impinella anisum seeds for 60 min results in the greatest band intensity for subsequently produced nanoparticles, and band intensity then declines as the extraction duration increases due to the oxidation and thermal decomposition of phytochemicals. Therefore, attentive consideration should be made regarding the duration of, and temperature during, phytochemical extraction.
## Filtration and preservation
After extraction, the next step is filtration, in which solid components are removed from the plant extract. There are many filtration techniques in use, as illustrated in [fig_ref] Figure 1: General steps in plant extraction [/fig_ref].
Following filtration, the obtained extracts are preserved for nanomaterial-synthesis research. Preservation is mainly achieved via refrigeration, directly using the extract for nanoparticle synthesis, or storing the extract in a container/environment with or without special conditions such as airtightness and light exclusion so as to avoid any manner of the oxidation or photodegradation of the phytochemicals. The temperature of refrigeration is mainly 4 °C as it was found that this temperature can best preserve the quality of Ananas comosus juice; moreover, increasing storage duration and temperature can greatly reduce the phytochemicals within the obtained plant extract [bib_ref] Storage Quality of Pineapple Juice Non-Thermally Pasteurized and Clarified by Microfiltration, Laorko [/bib_ref]. Therefore, in the green synthesis of Cu-NMs, the freshness of the plant extract is very significant. Once a plant extract is produced, it should be utilised for nanoparticle synthesis as soon as possible and, in the interim, stored at low temperature.
Finally, the obtained plant extract is prepared for the synthesis of nanomaterials; for example, Nasrollahzadeh et al. [bib_ref] Green Synthesis of CuO Nanoparticles Using Aqueous Extract of Thymus vulgaris L...., Nasrollahzadeh [/bib_ref] [bib_ref] Green Synthesis of the Copper Nanoparticles Supported on Bentonite and Investigation of..., Issaabadi [/bib_ref] produced Thymus vulgaris leaf extract and used it to synthesise CuO and Cu nanoparticles, as shown in. It is worth knowing that, although the production of other type of nanomaterials via green synthesis methods are referenced in this review, the plant extraction methods mentioned above are compatible in Cu-NMs synthesis.
Although the above paragraphs generalised the parameters and methods for plant extraction, there is no one best universal extraction method and parameter set for extracting all phytochemicals from all plant components. The final selections should depend on the type of plant, the plant component, and any industrial requirements.
Next, this review covers the synthesis of Cu-NMs using plant extracts. There are several factors that need to be taken into account to ensure the successful production of nanomaterials, including reaction time, temperature, pH, and the extract/precursor used; these will all influence the size and geometry of the nanomaterial produced. [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref] summarises previously reported works on the synthesis of Cu nanomaterials using plant extracts. . Representative TEM images of (a) spherical CuO nanoparticles synthesised using Annona squamosa seed extract. Reproduced from [bib_ref] Tunable Electrochemistry and Efficient Antibacterial Activity of Plant-Mediated Copper Oxide Nanoparticles Synthesized..., Singh [/bib_ref]. 2021with permission from the Royal Society of Chemistry, (b) tentacle-like bimetallic Ag-Cu nanoparticles synthesised using Carica papaya extract. Adapted with permission from Ref. [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref]. 2017, Elsevier, (c) cubical Cu nanoparticles synthesised from Azadirachta indica leaf extract Adapted with permission from Ref. [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref]. 2018, Elsevier; (d) SEM image of spherical Cu-Pt core shell nanoparticles synthesised using Agrimoniae herba extract. Adapted with permission from Ref.. 2018, Elsevier; and (e) TEM image of Cu-Co-Ni trimetallic nanoalloy nanoflakes synthesised using Origanum vulgare leaf extract. Adapted with permission from Ref.. 2020, MDPI.
[formula] - Cu(NO 3 ) 2 ·3H 2 O - Ni(NO 3 ) 2 ·6H 2 O - Co(NO 3 ) 2 ·6H 2 O- FeCl 3 ·6H 2 O - FeCl 2 ·4H 2 O - Ag(NO 3 ) - Cu(NO 3 ) 2 ·3H 2 O - tetra ethyl orthosilicate - Room temperature - - Fe 3 O-SiO 2 - Cu 2 O- [/formula]
# Cu-nms synthesis method
## Production of high tunable cu-nms
All extracts from plant components are composed of various types of phytochemicals such as flavonoids, phenolics, alizarin, quercetin, terpenoids, terpenes, alkaloids, carotenoids, and others. These phytochemicals can potentially be used for the synthesis of various types of Cu-based nanomaterials such as pure Cu nanoparticles, CuO nanoparticles, Cu-based nanocomposites, core-shell nanoparticles, and nanoalloys. For example, Ituen et al.used Citrus reticulata peel extract to produce spherical Cu nanoparticles with sizes of 54 and 72 nm. Similarly, Azadirachta indica flower extract has been used to produce pure spherical 5 nm Cu nanoparticles [bib_ref] Neem Flower Extract Assisted Green Synthesis of Copper Nanoparticles-Optimisation, Gopalakrishnan [/bib_ref] , and Thymus vulgaris leaf extract to produce CuO nanoparticles with spherical morphology and a particle size less than 30 nm [bib_ref] Green Synthesis of CuO Nanoparticles Using Aqueous Extract of Thymus vulgaris L...., Nasrollahzadeh [/bib_ref]. In addition, Dobrucka and Dlugaszewska synthesised 30 nm spherical bimetallic Pt-Cu coreshell particles with Cu as the core and Pt as the shell from the ethanolic extract of Agrimoniae herba leaves. As an example of root-extract-mediated nanoparticle synthesis, Pallela et al. [bib_ref] Antibacterial Activity Assessment and Characterization of Green Synthesized CuO Nano Rods Using..., Pallela [/bib_ref] successfully produced rod-shaped CuO nanoparticles with diameters of 50-100 nm and lengths of 400-500 nm from Asparagus racemosus root extract. Regarding nanoparticle synthesis from fruits, Ebrahimzadeh et al. [bib_ref] Eco-Friendly Green Synthesis and Characterization of Novel Fe 3 O 4 /SiO..., Ebrahimzadeh [/bib_ref] produced spherical . Representative TEM images of (a) spherical CuO nanoparticles synthesised using Annona squamosa seed extract. Reproduced from [bib_ref] Tunable Electrochemistry and Efficient Antibacterial Activity of Plant-Mediated Copper Oxide Nanoparticles Synthesized..., Singh [/bib_ref]. 2021with permission from the Royal Society of Chemistry, (b) tentacle-like bimetallic Ag-Cu nanoparticles synthesised using Carica papaya extract. Adapted with permission from Ref. [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref]. 2017, Elsevier, (c) cubical Cu nanoparticles synthesised from Azadirachta indica leaf extract Adapted with permission from Ref. [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref]. 2018, Elsevier; (d) SEM image of spherical Cu-Pt core shell nanoparticles synthesised using Agrimoniae herba extract. Adapted with permission from Ref.. 2018, Elsevier; and (e) TEM image of Cu-Co-Ni trimetallic nanoalloy nanoflakes synthesised using Origanum vulgare leaf extract. Adapted with permission from Ref.. 2020, MDPI.
# Cu-nms synthesis method
## Production of high tunable cu-nms
All extracts from plant components are composed of various types of phytochemicals such as flavonoids, phenolics, alizarin, quercetin, terpenoids, terpenes, alkaloids, carotenoids, and others. These phytochemicals can potentially be used for the synthesis of various types of Cu-based nanomaterials such as pure Cu nanoparticles, CuO nanoparticles, Cu-based nanocomposites, core-shell nanoparticles, and nanoalloys. For example, Ituen et al.used Citrus reticulata peel extract to produce spherical Cu nanoparticles with sizes of 54 and 72 nm. Similarly, Azadirachta indica flower extract has been used to produce pure spherical 5 nm Cu nanoparticles [bib_ref] Neem Flower Extract Assisted Green Synthesis of Copper Nanoparticles-Optimisation, Gopalakrishnan [/bib_ref] , and Thymus vulgaris leaf extract to produce CuO nanoparticles with spherical morphology and a particle size less than 30 nm [bib_ref] Green Synthesis of CuO Nanoparticles Using Aqueous Extract of Thymus vulgaris L...., Nasrollahzadeh [/bib_ref]. In addition, Dobrucka and Dlugaszewska synthesised 30 nm spherical bimetallic Pt-Cu core-shell particles with Cu as the core and Pt as the shell from the ethanolic extract of Agrimoniae herba leaves. As an example of root-extract-mediated nanoparticle synthesis, Pallela et al. [bib_ref] Antibacterial Activity Assessment and Characterization of Green Synthesized CuO Nano Rods Using..., Pallela [/bib_ref] successfully produced rod-shaped CuO nanoparticles with diameters of 50-100 nm and lengths of 400-500 nm from Asparagus racemosus root extract. Regarding nanoparticle synthesis from fruits, Ebrahimzadeh et al. [bib_ref] Eco-Friendly Green Synthesis and Characterization of Novel Fe 3 O 4 /SiO..., Ebrahimzadeh [/bib_ref] produced spherical Fe 3 O 4 /SiO 2 /Cu 2 O-Ag nanocomposites with diameters of 55 and 75 nm from Crataegus pentagyna fruit extract. Meanwhile, Sajadi et al. [bib_ref] Aqueous Extract from Seeds of Silybum marianum L. as a Green Material..., Sajadi [/bib_ref] used Silybum marianum seed extract to produce agglomerated Cu/Fe 3 O 4 nanoparticles with sizes of 8.5-60 nm and magnetic properties.
As indicated in [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref] , the sizes of nanoparticles produced using plant-mediated synthesis ranges from 2 to 150 nm. On the other hand, the morphology of the produced nanoparticles is predominantly spherical [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref] [bib_ref] Green Synthesised Copper and Copper Oxide Based Nanomaterials Using Plant Extracts and..., Bhavyasree [/bib_ref] with some having other shapes such as hexagonal, cubical [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref] , ellipsoidal [bib_ref] Green Synthesis of Ag-Cu Nanoalloys Using Opuntia ficus-indica, Rocha-Rocha [/bib_ref] , tentacle-like [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref] , or nanoflake [bib_ref] Green Synthesis of Ag-Mo/CuO Nanoparticles Using Azadirachta indica Leaf Extracts to Study..., Rajendaran [/bib_ref] [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref] and . Several characterization methods are used in determining the size and morphology of nanoparticles, with the preeminent being scanning electron microscopy (SEM), transmission electron microscopy (TEM), and dynamic light scattering (DLS). It should be noted that the findings elucidated by each characterization method might have some discrepancies. For instance, Rosbero and Camacho observed synthesised Ag/Cu nanoparticles to have a size of 420.70 nm according to DLS, but a size range of 90-150 nm by TEM [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref]. This discrepancy is attributable to the presence of solvent molecules on the nanoparticle surface and DLS only determining the hydrodynamic size of the particles rather than the core diameter. That is to say, when there is a hydration layer surrounding the nanoparticles, only the solvated particle size is indicated by a particle's diffusional characteristics [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref]. Hence, to obtain the most accurate results and perform effective quality control of nanomaterials, multiple characterization methods should be employed.
Overall, it can be observed that Cu-NMs produced via plant-mediated synthesis feature size and morphology tunability comparable to those obtained with chemical synthesis methods. Specifically, tunability can be achieved via altering parameters such as the precursor concentration, plant extract, reaction time, and the temperature applied during nanoparticle synthesis.
## Precursor
As listed in [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref] , most studies to date have utilised CuSO 4 [bib_ref] Facile Green Synthesis of Ag-Cu Decorated ZnO Nanocomposite for Effective Removal of..., Manjari [/bib_ref] [bib_ref] Green Synthesis of Pd/CuO Nanoparticles by Theobroma cacao L. Seeds Extract and..., Nasrollahzadeh [/bib_ref] , Cu(NO 3 ) 2 [bib_ref] Biosynthesis of Copper Nanoparticles Using Flower Extract of Aloe vera, Karimi [/bib_ref] [bib_ref] Biosynthesis of Cu Nanoparticles Using Polyalthia longifolia Roots Extracts for Antibacterial, Maulana [/bib_ref] , and CuCl 2 [bib_ref] Green Synthesis of Copper Nanoparticles Using Plantago asiatica Leaf Extract and Their..., Nasrollahzadeh [/bib_ref] as the Cu precursor, while some used copper acetate (Cu(OAc) 2 ) [bib_ref] Biosynthesis of CuO Nanoparticles Using Rosa canina Fruit Extract as a Recyclable..., Hemmati [/bib_ref]. Interestingly, for some multi-metallic nanoparticles (nanoalloys, core-shell particles, and nanocomposites), multiple precursors have been utilised and the type of nanoparticle formed depends on the methodology. For example, Cu-Co-Ni trimetallic nanoalloy was synthesised using Origanum vulgare leaf extract and the precursors of Cu(NO 3 ) 2 ·3H 2 O, Ni(NO 3 ) 2 ·6H 2 O, and Co(NO 3 ) 2 ·6H 2 O, while bimetallic Pt-Cu core-shell structures with Cu as core and Pt as shell were synthesised from Agrimoniae herba leaf ethanolic extract with K 2 PtCl 6 and CuSO 4 as the precursors. Basically, multiple metallic precursors are mixed with a plant extract and stirring and heat applied to yield multi-metallic nanoparticles. Generally, a combination of two metals will lead to the synthesis of alloy or core-shell nanoparticles. Which form of Cu-NM is produced can be determined based on SPR from UV-visible analysis: if a single SPR is found, an alloy was formed, while if two independent and continuous peaks are evident, a core-shell-type structure resulted [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref]. For instance, Ag/Cu nanoparticles produced via bio-reduction with Carica papaya extract exhibited a single peak at 776 nm as the maximum absorption, suggesting an alloyed structure [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref]. Cu-based nanocomposites are also able to be synthesised via plant-mediated methods; Ebrahimzadeh et al. [bib_ref] Eco-Friendly Green Synthesis and Characterization of Novel Fe 3 O 4 /SiO..., Ebrahimzadeh [/bib_ref] In addition to precursor choice, the concentration of precursor applied in the reaction also plays an important role in determining the Cu-NMs synthesised. Lee et al. [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref] used Magnolia kobus leaf extract to produce spherical Cu nanoparticles, mixing it with CuSO 4 ·5H 2 O at 0.5, 1, and 2 mmol/L and reducing the Cu ions to atoms. Given constant temperature and plant extract concentration, the time required to achieve a conversion rate of more than 90% was 1600, 1400, and 200 min for the concentrations of 0.5, 1, and 2 mmol/L, respectively. Therefore, it could be concluded that a higher precursor concentration can accelerate nanoparticle formation. In addition, a study that performed precursor optimization for the green synthesis of Cu nanoparticles from Senna didymobotrya root extract utilised CuSO 4 ·5H 2 O at concentrations of 0.0125, 0.03125, and 0.05 M [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. This study revealed that the higher the precursor concentration, the higher the nanoparticle size. The authors noted this could be due to a low concentration of Cu ions reducing the chance of Cu-Cu interactions and, hence, reducing agglomeration [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. Thus, the formation rate and size of synthesised nanoparticles can be controlled via altering the precursor concentration. However, the balance between conversion rate (nanoparticle formation) and nanoparticle size should be taken into account when carrying out green-synthesis research to ensure the desired nanoparticle is produced while also achieving a highly productive and efficient synthesis process.
## Plant extract
The plant extract utilised is also another major factor that should be considered in plant-mediated Cu-NM synthesis. A variety of plant extracts have demonstrated great impact on the synthesis of Cu-NMs [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref]. In a study examining the effect of extract concentration on nanoparticle synthesis rate and characteristics, Magnolia kobus leaf extract at a range of concentrations (5-20%) was used to produce spherical Cu nanoparticles [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref]. The highest synthesis rate was obtained with an extract concentration of 20%, while high average particle sizes were obtained for both the lowest (5%) and highest (20%) extract concentrations, with diameters of 91 and 82 nm, respectively. Meanwhile, an extract concentration of 15% produced nanoparticles with diameter 37 nm, which was the smallest among all the results [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref]. The reason for the production of large nanoparticles from high extract concentrations is due to the excessive abundance of capping materials promoting the aggregation of Cu particles owing to the interaction between nanoparticles that are surrounded with proteins and metabolites (reducing sugar, terpenoid, and other metabolites) [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref]. Therefore, it can be concluded that if a high yield of a small Cu nanoparticle is required, the leaf extract concentration should be optimised before conducting the plantmediated synthesis process at scale, whereas if the most rapid production is required, a high concentration of plant extract should be applied.
## Temperature
Aside from material inputs, the temperature applied during plant-mediated Cu-NM synthesis is also an essential parameter to be investigated. According to [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref] , the temperatures utilised in existing reports range between 25 and 100 - C. For example, temperatures of 25 ± 2, 30, 40, and 50 - C have been used for Cu nanoparticle production from Citrus reticulata peel extract, and 100 - C for CuO nanoparticle synthesis from Rosa canina fruit extract [bib_ref] Biosynthesis of CuO Nanoparticles Using Rosa canina Fruit Extract as a Recyclable..., Hemmati [/bib_ref]. Notably, the use of different temperatures can greatly impact the Cu-NM synthesis process. For example, in the abovementioned study using Citrus reticulata peel extract combined with CuSO 4 .5H 2 O, successful bio-reduction and nanoparticle production was indicated by a colour change to brown with absorption at 442 nm [fig_ref] Figure 4: Colour change over time during the reaction between Citrus reticulata peel extract... [/fig_ref]. At reaction temperatures of 25, 30, 40, and 50 - C with a constant pH, achieving this endpoint required 72 h, 60 h, 10 h, and 105 min, respectively. Therefore, plant-mediated synthesis of Cu-NMs is a temperature-dependent process with a positive proportional relationship: the higher the temperature, the higher the rate of conversion from Cu ion to Cu metal. Nanomaterials 2022, 12, x FOR PEER REVIEW 28 of 49 Interestingly, nanoparticle synthesis rate and size behave differently under a given reaction temperature increment, as the conversion rate increases whereas nanoparticle size decreases with increasing reaction temperatures. For instance, as mentioned above, Lee et al. [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref] synthesised Cu nanoparticles using Magnolia kobus leaf extract and observed a size reduction from 110 nm at low temperature (25 °C, conversion rate 70%) to 37 nm at high temperature (95 °C, conversion rate ~80-100%). The rationale behind such phenomena is that the increasing temperature improves the reaction rate. When the reaction rate is increased, Cu ions in the reaction solution are only able to be consumed for the formation of nuclei; the secondary reduction process on the nuclei is avoided. Thus, larger nanoparticles cannot be produced at higher temperatures [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref] [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. Consequently, it can be concluded that nanomaterials synthesis is temperature-dependent, but the size of nanomaterials has a negative proportional relationship with temperature such that producing nanoparticles with a larger size necessitates utilising a lower temperature.
There have been studies conducted on further calcination of metal oxides at temperatures ranging from 400 to 500 °C after the synthesis process [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref] [bib_ref] Muntingia Calabura Leaves Mediated Green Synthesis of Cuo Nanorods: Exploiting Phytochemicals for..., Selvanathan [/bib_ref]. One of the purposes of calcination is to produce stable metal oxides or metal oxide nanocomposites through oxidation [bib_ref] Biosynthesis of Mixed Nanocrystalline Zn-Mg-Cu Oxide Nanocomposites and Their Antimicrobial Behavior, Suresh [/bib_ref]. For example, Suresh et al. [bib_ref] Biosynthesis of Mixed Nanocrystalline Zn-Mg-Cu Oxide Nanocomposites and Their Antimicrobial Behavior, Suresh [/bib_ref] used Pisonia grandis leaf extract to synthesise Zn-Mg-Cu oxide nanocomposites, then calcinated them at 450 °C to obtain mixed metal oxide nanocomposites. In addition to producing stable oxides, increasing calcination temperature can boost the size of Cu-NMs and produce black precipitates of agglomerated cubical nanomaterials, where uncalcinated nanomaterials have elongated morphology.
## Ph
Solution pH is also a very significant factor in plant-mediated nanomaterial synthesis as it can affect the synthesis rate and products. Mechanistically, the importance of pH is due to the reducing and stabilizing agents being greatly dependent on the phytochemicals within the plant extract, which might be readily affected by pH. Generally, the best pH values for plant-mediated nanomaterial synthesis are in the range of pH 7-9, and varying the pH will alter the nanoparticle synthesised [bib_ref] A Comprehensive Review of Green Synthesis of Copper Oxide Nanoparticles and Their..., Waris [/bib_ref]. Nagar and Devra utilised Azadirachta indica leaves for Cu nanoparticle synthesis at various pH values [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref]. They found that nanoparticle synthesis is more effective at higher pH and abolished in an extreme acidic environment, such as pH 4.7. A solution with pH 6 produced small-sized nanoparticles of 56 nm, while in an alkaline environment of pH 9.3, the nanoparticles produced were of size 73 nm. In an acidic environment, the phytochemicals in the plant extract might be inactivated [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref]. In addition, lower pH can cause nanoparticles to experience high electrostatic repulsion which reduces the chances of agglomeration and, thus, yields nanomaterials of smaller size [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. Conversely, in a more alkaline condition such as pH 10, the low electrostatic forces of the nanoparticles allow further particle growth and agglomeration, which produces larger nanomaterials [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref] [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. It is worth noting that during the plantmediated synthesis process, the pH of the medium will drop as the Cu 2+ ions cause Interestingly, nanoparticle synthesis rate and size behave differently under a given reaction temperature increment, as the conversion rate increases whereas nanoparticle size decreases with increasing reaction temperatures. For instance, as mentioned above, Lee et al. [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref] synthesised Cu nanoparticles using Magnolia kobus leaf extract and observed a size reduction from 110 nm at low temperature (25 - C, conversion rate 70%) to 37 nm at high temperature (95 - C, conversion rate~80-100%). The rationale behind such phenomena is that the increasing temperature improves the reaction rate. When the reaction rate is increased, Cu ions in the reaction solution are only able to be consumed for the formation of nuclei; the secondary reduction process on the nuclei is avoided. Thus, larger nanoparticles cannot be produced at higher temperatures [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref] [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. Consequently, it can be concluded that nanomaterials synthesis is temperature-dependent, but the size of nanomaterials has a negative proportional relationship with temperature such that producing nanoparticles with a larger size necessitates utilising a lower temperature.
There have been studies conducted on further calcination of metal oxides at temperatures ranging from 400 to 500 - C after the synthesis process [fig_ref] Table 3: Summary of plant-mediated Cu nanomaterial synthesis [/fig_ref] [bib_ref] Muntingia Calabura Leaves Mediated Green Synthesis of Cuo Nanorods: Exploiting Phytochemicals for..., Selvanathan [/bib_ref]. One of the purposes of calcination is to produce stable metal oxides or metal oxide nanocomposites through oxidation [bib_ref] Biosynthesis of Mixed Nanocrystalline Zn-Mg-Cu Oxide Nanocomposites and Their Antimicrobial Behavior, Suresh [/bib_ref]. For example, Suresh et al. [bib_ref] Biosynthesis of Mixed Nanocrystalline Zn-Mg-Cu Oxide Nanocomposites and Their Antimicrobial Behavior, Suresh [/bib_ref] used Pisonia grandis leaf extract to synthesise Zn-Mg-Cu oxide nanocomposites, then calcinated them at 450 - C to obtain mixed metal oxide nanocomposites. In addition to producing stable oxides, increasing calcination temperature can boost the size of Cu-NMs and produce black precipitates of agglomerated cubical nanomaterials, where uncalcinated nanomaterials have elongated morphology.
## Ph
Solution pH is also a very significant factor in plant-mediated nanomaterial synthesis as it can affect the synthesis rate and products. Mechanistically, the importance of pH is due to the reducing and stabilizing agents being greatly dependent on the phytochemicals within the plant extract, which might be readily affected by pH. Generally, the best pH values for plant-mediated nanomaterial synthesis are in the range of pH 7-9, and varying the pH will alter the nanoparticle synthesised [bib_ref] A Comprehensive Review of Green Synthesis of Copper Oxide Nanoparticles and Their..., Waris [/bib_ref]. Nagar and Devra utilised Azadirachta indica leaves for Cu nanoparticle synthesis at various pH values [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref]. They found that nanoparticle synthesis is more effective at higher pH and abolished in an extreme acidic environment, such as pH 4.7. A solution with pH 6 produced small-sized nanoparticles of 56 nm, while in an alkaline environment of pH 9.3, the nanoparticles produced were of size 73 nm. In an acidic environment, the phytochemicals in the plant extract might be inactivated [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref]. In addition, lower pH can cause nanoparticles to experience high electrostatic repulsion which reduces the chances of agglomeration and, thus, yields nanomaterials of smaller size [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. Conversely, in a more alkaline condition such as pH 10, the low electrostatic forces of the nanoparticles allow further particle growth and agglomeration, which produces larger nanomaterials [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref] [bib_ref] Optimization, Characterization, and Antibacterial Activity of Copper Nanoparticles Synthesized Using Senna didymobotrya..., Sadia [/bib_ref]. It is worth noting that during the plant-mediated synthesis process, the pH of the medium will drop as the Cu 2+ ions cause oxidation of the plant extract, leading to the release of H + ions and, hence, acidification; this is also another important aspect to be considered by researchers carrying out plant-mediated nanomaterial synthesis [bib_ref] Green Synthesis and Characterization of Copper Nanoparticles Using Azadirachta indica Leaves, Nagar [/bib_ref]. Therefore, it is necessary to achieve a balance in producing nanoparticles with a desired size while maintaining high nanomaterial productivity.
## Reaction time
In terms of duration, it can be seen that the range of reaction times in the literature is relatively large, ranging from as low as 5 min to as long as 72 h. The duration is not as impactful compared to the other parameters mentioned above. Although a long synthesis duration allows improvement in the nanomaterial nucleation rate, the reaction rate will not continue to increase after the optimum time has been reached. In some cases, prolonging the incubation might even cause nanomaterial aggregation [bib_ref] Green Synthesised Copper and Copper Oxide Based Nanomaterials Using Plant Extracts and..., Bhavyasree [/bib_ref]. In fact, other parameters such as temperature, precursor, and type of plant extract can greatly impact the time needed to achieve complete conversion from metallic ions (Cu and other ions depending on the type of nanomaterial being produced) to metallic atoms and, finally, nanomaterials, as was mentioned previously in relation to other parameters. However, compared with other materials such as Au, Ag, and Pt, Cu forms nanoparticles relatively slowly as the initiation of Cu nucleus formation is much more difficult [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref]. Hence, the green synthesis of Cu nanoparticles necessitates longer reaction times in order to achieve 100% conversion.
## Indication of cu-nms production
During the production of Cu-NMs, a successful reaction is indicated by the colour alteration of the reaction solution. For example, Thymus vulgaris leaf extract mixed with CuCl 2 .2H 2 O with constant stirring at 60 - C undergoes a change in colour from yellow to dark brown, as shown in. In addition, calcination will also change the colour of nanomaterials produced. Carica papaya peel extract combined with Cu(NO 3 ) 2 .3H 2 O and heated at 70-80 - C changes colour from greenish-blue to green and finally produces a dark green paste. Upon calcination, a fine black-coloured powder was obtained and harvested as CuO nanoparticles.
## Applications of cu-nms from plant-mediated synthesis
After the synthesis reaction is completed, the obtained nanoparticles are washed, dried, and employed in applications. Cu-NMs synthesised using plant extracts have been utilised in two major areas, namely, biomedical and environmental remediation [bib_ref] Noble Metal Nanoparticles: Plant-Mediated Synthesis, Dauthal [/bib_ref].
## Biomedical
Plant-mediated Cu-NMs have demonstrated antimicrobial, antioxidant, and anticancer activities, and have potential as nano-sensors and in various medical applications. In this section, details and mechanisms pertaining to this area will be discussed.
## Antimicrobial
Firstly, antibacterial activity has been observed for plant-mediated Cu-NMsand can be attributed to several putative pathways. Bhavyasree and Xavier suggested that Cu-NMs, including both Cu and CuO nanoparticles produced via plant-mediated synthesis, can carry out antibacterial activity through a chemisorption-based mechanism [bib_ref] Green Synthesised Copper and Copper Oxide Based Nanomaterials Using Plant Extracts and..., Bhavyasree [/bib_ref]. This mechanism involves microbial adsorption to the nanoparticle surface, which has been bio-functionalised by phytochemicals during the plant-mediated synthesis process. The adsorption is mainly due to chemisorption via non-electrostatic forces (Van der Waals force and hydrogen bonding), which causes the destruction of the microbial cell wall and subsequent cell membrane damage, DNA breakage, and eventually cell death, as illustrated in [fig_ref] Figure 5: Diagram of the chemisorption-based mechanism of Cu-based nanomaterials' antimicrobial activity [/fig_ref]. Another antibacterial mechanism is mediated by reactive oxygen species (ROS) and the release of Cu 2+ ions. First, the CuO nanoparticles are much smaller (being of a nanometre scale) than the micrometre-scale pores of bacterial cells, which allows them to easily penetrate the cells. In addition, Cu 2+ ions are attracted toward bacterial cells due to the abundance of carboxyl and amine groups on the cell surface; this is another factor in antibacterial ability. However, the antibacterial interactions are different for Gram-positive and Gram-negative bacteria, as described in. After bypassing the cell wall, Cu 2+ ions relocate intracellularly to the cytosol due to the internalization of CuO nanoparticles and Cu 2+ ions, where they cause ROS to accumulate. Consequently, DNA and mitochondria damage occur. Cu 2+ ions within a bacterium may also stimulate cellular responses that lead to bactericidal activity. For example, radicals produced by CuO nanoparticles, such as superoxide and hydroxyl radicals, can have synergic effects in causing bacterial membrane destruction, DNA damage, attachment to ribosomes, oxidative injury, and protein and proton efflux pump damage; they can also prevent biofilm production.
Cu-NMs exert antibacterial activity through mechanisms similar to those of Cu and CuO nanoparticles. Generally, the antibacterial activity of nanomaterials is mainly owed to the induction of oxidative stress, such as through the production of free radicals and ROS. Notably, nano-sized particles will feature a smaller surface-to-volume ratio, harbour more surface defects due to oxygen vacancies, and feature greater electrostatic attraction and release of Cu 2+ ions and generate more oxidative stress within the bacterial cells [bib_ref] Antimicrobial Cellulose Nanocomposite Films with In Situ Generations of Bimetallic (Ag and..., Mamatha [/bib_ref]. In addition, bimetallic nanoparticles can demonstrate a synergic effect with improved antibacterial ability. For example, bimetallic Ag and Cu nanoparticles produced by Vitex negundo-mediated synthesis demonstrate antibacterial activity when applied in a cellulose matrix via the disc method against both Gram-positive (Escherichia coli, Pseudomonas, Klebsiella) and Gram-negative (Staphylococcus, Bacillus) species [bib_ref] Antimicrobial Cellulose Nanocomposite Films with In Situ Generations of Bimetallic (Ag and..., Mamatha [/bib_ref]. Particles having an equal ratio of Ag and Cu (2.5 mM each) exhibited the greatest antibacterial ability, with a 9 mm zone of inhibition for all the tested species.
Secondly, Cu-NMs have also been demonstrated to possess antifungal activity. A number of fungi can cause infections in humans with severe symptoms, such as Candida albicans which can cause mucosal infections (oropharyngeal or vulvovaginal candidiasis), or Trichophyton mentagrophytes which can cause dermatophytosis [bib_ref] Epidemiological and Clinical Aspects of Trichophyton Mentagrophytes/Trichophyton Interdigitale Infections in the Zurich..., Klinger [/bib_ref] [bib_ref] The Impact of the Fungus-Host-Microbiota Interplay upon Candida albicans Infections: Current Knowledge..., D'enfert [/bib_ref]. Antifungal activity is more challenging to realise than antibacterial as a fungus cell has several layers of lipids within its cell wall which impede the penetration and internalization of Cu nanomaterials. Although fewer publications exist regarding the antifungal testing of Cu nanomaterials synthesised by green methods, their hypothesised antifungal mechanism is based on altering the structure and function of fungal cell components [bib_ref] Green Synthesis of Copper Nanoparticles Using Celastrus paniculatus Willd. Leaf Extract and..., Mali [/bib_ref]. That is, the nanoparticles first distort the cell wall and become internalised by the fungus. After internalization, the same process of ROS generation and subsequent process disruption ensues as in bacteria, impacting DNA, mitochondria, replication, protein synthesis and other essential elements, eventually leading to cell death [bib_ref] Green Synthesis of Copper Nanoparticles Using Celastrus paniculatus Willd. Leaf Extract and..., Mali [/bib_ref].
In one report of Cu-NM antifungal activity, Mali et al. [bib_ref] Green Synthesis of Copper Nanoparticles Using Celastrus paniculatus Willd. Leaf Extract and..., Mali [/bib_ref] tested the efficacy of Cu nanoparticles derived from Celastrus paniculatus leaf extract against Fusarium oxysporum. Concentrations of 0.12%, 0.18% and 0.24% (w/v) Cu nanoparticles were found able to inhibit mycelial growth by 76.29 ± 1.52%, 73.70 ± 1.52%, and 59.25 ± 0.57%, respectively, calculated via the following formula:
[formula] (% inhibition rate) = (Mc − Mt) Mc × 100(1) [/formula]
where Mc represents mycelial growth in the control (with water) while Mt is mycelial growth under the Cu nanoparticle treatment. The inhibition rate was found to be dosagedependent: the higher the Cu nanoparticle dosage, the higher the degree of inhibition. Dobrucka and Dlugaszewska similarly studied the antibacterial and antifungal activities of Cu-Pt core-shell nanoparticles synthesised using Agrimoniae herba extract. The nanoparticles were applied via the well-diffusion method to three species of bacteria, including Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa, and three of fungi: Candida albicans, Trichophyton mentagrophytes, and Aspergillus fumigatus; the authors then determined the minimal inhibitory concentration (MIC), minimal bactericidal concentration (MBC), and minimal fungicidal concentration (MFC). The Cu-Pt nanoparticles exhibited good inhibitory function on all tested bacteria and Trichophyton Mentagrophytes. The overall best antibacterial and antifungal performances were obtained on Staphylococcus aureus (MIC of 16.7 and MBC of 33.3) and Trichophyton mentagrophytes (MIC and MFC of 26.7).
From the above reports, it can be concluded that plant-mediated Cu-NMs are suitable as antibacterial (for both Gram-positive and -negative) and antifungal agents. Such characteristics are useful in further broadening the application of Cu-NMs in the pharmaceutical and medical sectors.
In addition to direct antimicrobial effects, many plant-mediated Cu-NMs have also demonstrated antioxidant properties which also contribute to antibacterial and antifungal activities as a synergic factor. Multiple mechanisms contribute to antioxidant ability, which are: (1) binding of transition metal ion catalysts, (2) reductive capacity, (3) radical scavenging activity, (4) decomposition of peroxides, (5) prevention of continued hydrogen abstraction, and (6) prevention of chain initiation.
Interestingly, plant selection has been shown to impact the antioxidant ability of Cu-NMs. For example, Rehana et al. [bib_ref] Evaluation of Antioxidant and Anticancer Activity of Copper Oxide Nanoparticles Synthesized Using..., Rehana [/bib_ref] synthesised nanoparticles using extracts of Azadirachta indica, Hibiscus rosa-sinensis, Murraya koenigii, Moringa oleifera, and Tamarindus indica, then tested their antioxidant capabilities with ABTS, DPPH, and hydrogen peroxide assays. Tamarindus indica-mediated nanoparticles were found to have the highest antioxidant activity, and Moringa oleifera the lowest, though still superior to CuO nanoparticles produced via a chemical method. Therefore, plant-mediated nanomaterials have much higher antioxidant ability as compared to chemical-mediated materials, and the plant used is an essential consideration for antioxidant purposes.
## Nano-sensor
Plant-extract-mediated Cu-NMs have also been utilised in the preparation of nano-sensors. Cu nanomaterials, such as CuO nanoparticles, are suitable for nano-sensor production owing to their characteristic high electron-transfer rate, superior catalytic activity, large surface area, high glucose selectivity in heterogenous samples (such as blood or urine), chlorine poisoning resistance, and corrosion resistance. For example, Ag-CuO core-shell nanoparticles produced using Ocimum tenuiflorum extract have been used for non-enzymatic glucose sensing with a screen-printed electrode [bib_ref] Biosynthesis of Ag@CuO Core-Shell Nanostructures for Non-Enzymatic Glucose Sensing Using Screen-Printed Electrode, Dayakar [/bib_ref]. The synthesised electrode provided good glucose-sensing performance with a sensitivity of 3763.44 µAmM −1 cm −2 , linear range of 1 to 9.2 mM, detection limit of 0.006 mM (S/N = 3), and response time of less than 1 s. Moreover, the CuO-Ag coreshell-modified bio-nano-sensors demonstrated exceptional adhesion and structural strength along with great long-term stability for up to 60 days, exhibiting 99.2% of the initial value after one month with excellent repeatability and reproducibility.
The mechanism by which these nanoparticles sense glucose is based on electron transfer from the screen-printed electrode to the CuO nanoparticle core via the conduction band electrons of the Ag shell. This electron transfer occurs because the work function of CuO is bigger than that of Ag, and equalization of Fermi levels ensues after the materials come into electrical contact and the mobility of electrons is improved. The progression of current-induced charge carriers can boost electrocatalytic efficiency through a charge transfer mechanism; therefore, the Ag-CuO core-shell nanoparticles are electro-catalytically active and can induce electron-transfer reactions. The energy of a nanoparticle is dependent on the charge distribution within the energy levels of its component metal. Ultimately, the additional electrons can be discharged when glucose is introduced into the system as an electron acceptor.
This glucose oxidation mechanism can be summarised as: (1) deprotonation of glucose that causes oxidation, (2) isomerization and enediol formation, and, finally, (3) adsorption to the electrode surface, which leads to the oxidation of Cu(II)/Cu(III):
[formula] CuO + OH − → CuOOH + e −(2)CuOOH + glucose + e − → CuO + OH − + gluconic acid(3) [/formula]
In the core-shell nanoparticle, Cu(II) was oxidised to Cu(III) and this catalysed glucose oxidation to produce gluconolactone, which was further oxidised to gluconic acid as presented in Equations (2) and (3).
## Anticancer
Lastly, plant extract-mediated Cu-NMs have been studied for their anticancer properties. Generally, this activity can be realised through multiple routes including ROS generation, antioxidant activity, cell cycle arrest, apoptosis, and autophagy. One study produced CuO nanoparticles using extracts of Azadirachta indica, Hibiscus rosa-sinensis, Murraya koenigii, Moringa oleifera, and Tamarindus indica and used MTT assays to test their activity against four cancer cell lines, i.e., human breast, cervical, epithelioma, and lung cancer cells, along with one normal human dermal fibroblast (NHDF) cell line [bib_ref] Evaluation of Antioxidant and Anticancer Activity of Copper Oxide Nanoparticles Synthesized Using..., Rehana [/bib_ref]. All CuO nanoparticles exhibited anticancer ability towards all cancer cell types in a dosedependent manner: higher concentrations of CuO nanoparticles resulted in lower cancer cell viability. Interestingly, the type of plant utilised also affected anti-cancer ability, with Tamarindus indica-mediated CuO nanoparticles exhibiting greater cytotoxicity over the others; this indicates that the phytochemicals in the plant extract used for nanoparticle synthesis impact the resulting particles' anti-cancer activity.
The toxicity of Cu-NMs is one of the limitations that hinder their application biomedically. However, it has been reported that plant-mediated Cu nanomaterials have less toxicity to normal human cell lines. Therefore, such Cu nanomaterials may be more safely applied in biomedical applications. For example, CuO nanoparticles synthesised using extracts of Azadirachta indica, Hibiscus rosa-sinensis, Murraya koenigii, Moringa oleifera, and Tamarindus indica exhibited lower toxicity in NHDF cells, which suggests these to be promising anticancer agents for use in the pharmaceutical industry [bib_ref] Evaluation of Antioxidant and Anticancer Activity of Copper Oxide Nanoparticles Synthesized Using..., Rehana [/bib_ref].
## Environmental remediation
The usage of Cu-NMs in environmental applications is mainly focused on the remediation of dyes and toxic compounds, with mechanisms primarily based on photocatalysis or catalysis.
The mechanism of photocatalysis by nanomaterials is as follows: when the nanomaterials are deposited into an aqueous sample containing compounds that are desired to be degraded, such as dye, and exposed to light, an interaction occurs in which a photogenerated electron is converted from the valence band (VB) to the conduction band (CB) in the nanomaterial. A hole in the VB then results, producing an electron (e − )-hole (h + ) pair. The holes react with OH ions in the water molecules to yield OH radicals via oxidation, while the electrons react with dissolved O 2 to generate O 2 radicals via reduction. Those radicals are then responsible for the degradation of the dye into non-toxic degraded products [bib_ref] Green Synthesis of Ag-Mo/CuO Nanoparticles Using Azadirachta indica Leaf Extracts to Study..., Rajendaran [/bib_ref]. Alshehri and Malik investigated the ability of Origanum vulgare extractmediated Cu-Co-Ni trimetallic nanoparticles to photocatalyse the degradation of methylene blue. They observed degradation efficiency of more than 50% and 92.67% after 50 and 100 min, respectively. The rate of degradation could be increased via increasing nanoparticle concentration, but after a certain threshold was surpassed, the photocatalytic efficiency could be enhanced no further due to the aggregation of the nanomaterials.
With regard to catalysis mechanisms, nanoparticles can catalyse reactions by borylation, clock reactions, oxidative coupling, A3 coupling, click chemistry, tandem and multicomponent reactions, C-H functionalization, cross-coupling, reduction and oxidation reactions, and other mixed reactions [bib_ref] Cu and Cu-Based Nanoparticles: Synthesis and Applications in Catalysis, Gawande [/bib_ref]. Successful catalysis via plant-mediated Cu-NMs has been achieved, such as when Suvarna et al.studied the degradation of methyl green dye using bimetallic spherical Fe-Cu nanoparticles produced using Cyclea peltata extract, and achieved a degradation efficacy of 82% within 105 min. The Fe-Cu nanoparticles promoted hydrolysis and deprotonation reactions on the dye molecules, resulting in the demineralization of the dye molecules into simpler structures. In another example, Rosbero and Camacho utilised bimetallic (Ag and Cu) alloy nanoparticles produced via Carica papaya leaf extract to degrade the pesticide chlorpyrifos in water [bib_ref] Green Preparation and Characterization of Tentacle-like Silver/Copper Nanoparticles for Catalytic Degradation of..., Rosbero [/bib_ref]. The degradation was observed for 24 h, and yielded the products 3,5,6-trichloropyridinol (TCP) and diethylthiophosphate (DETP), of which the former is less toxic than chlorpyrifos and not mutagenic.
# Future research directions
Although plant-mediated Cu-NMs have numerous benefits and applications, they also have considerable potential yet to be discovered along with disadvantages that are unavoidable and need to be addressed to ensure realization of the applicability of these nanoparticles toward industrial production with wider applications. This section suggests areas of future research to increase the potential of Cu-NMs and propel the applicability of their production at a larger scale via eradicating current limitations; specifically, it discusses: (i) solutions by which to overcome limitations, (ii) potential new applications, and (iii) new research directions regarding Cu-NM synthesis.
# Limitations and solutions
This section illustrates the limitations of green Cu nanomaterial production and associated solutions. There are several that need attention in this respect, mainly with regard to biomass obtainability, the complexity of plant systems, the underlying synthesis process, nanomaterial quality, and low productivity.
Concerning bioresource accessibility, most research to date has focused on the use of local plant species that are not widely available throughout the globe; notably, variation in plant species and also geographical cultivation areas affect the phytochemicals within the plant extract produced [bib_ref] Green Synthesis of Nanoparticles: Current Developments and Limitations, Ying [/bib_ref]. Moreover, the inherent complexity of plants is another hindrance to the industrial production of plant-mediated Cu-NMs. That is, the phytochemicals within a plant are greatly affected by external factors such as abiotic environmental factors, cultivar, and mutagenesis [bib_ref] The Effects of Cultivar and Climate Zone on Phytochemical Components of Walnut, Wu [/bib_ref] [bib_ref] Influence of Environmental Abiotic Factors on the Content of Saponins in Plants, Szakiel [/bib_ref] [bib_ref] Phytochemical Profiles and in Vitro Anti-Inflammatory Properties of Perilla Frutescens Cv. Chookyoupjaso..., Park [/bib_ref]. These will cause batch-to-batch variation among raw materials, which might adversely affect the homogeneity and reproducibility of nanomaterial synthesis. On top of those considerations, another drawback to this method is the diversity of phytochemicals in a plant system. This can be addressed by applying molecular science techniques such as genetic engineering to maximize the most relevant phytochemicals in the target plant. Combining these techniques with plant-tissue culture methods such as cloning can allow the quality of a plant (target phytochemical composition and content) to be preserved and controlled and, thus, avoid batch-to-batch variation and mutagenic factors that might affect the phytochemical profile. Plant-tissue culture techniques can also minimize the time, cost, and labour force needed for the planting of bioresources and overcome geographical limitations [bib_ref] Improvement of Phytochemical Production by Plant Cells and Organ Culture and by..., Nielsen [/bib_ref].
At present, most research into plant-mediated nanoparticles is carried out in low quantities, and, hence, with low productivity. Although the quantity of nanoparticles required for characterization or application research purposes is not high, mass production via the green synthesis method is little-studied and needs to be researched in order to produce nanoparticles in a large quantity. Bioprocess methods can be used to produce and maximize specific phytochemicals for nanomaterial production and allow large-scale industrial production. With these solutions and more research, the global industrial production of uniform plant-mediated Cu nanomaterial products could be realised.
However, given the limited determinations of phytochemical profiles, it is not feasible to elucidate the mechanism of nanomaterial synthesis. This will affect the possibility of producing nanomaterials with good homogeneity in terms of size, shape, and crystal structure [bib_ref] Biological Synthesis of Copper Nanoparticles Using Magnolia kobus Leaf Extract and Their..., Lee [/bib_ref] [bib_ref] Green Synthesis of Copper Nanoparticles Using Plantago asiatica Leaf Extract and Their..., Nasrollahzadeh [/bib_ref] [bib_ref] Biogenic Copper Oxide Nanoparticles Synthesis Using Tabernaemontana divaricate Leaf Extract and Its..., Sivaraj [/bib_ref] [bib_ref] Green Synthesis: In-Vitro Anticancer Activity of Copper Oxide Nanoparticles against Human Cervical..., Nagajyothi [/bib_ref]. When it comes to investigating those phytochemicals that are responsible for stabilizing and reducing the ions within a precursor material, Fourier-transform infrared spectroscopy (FTIR) is the current technique of choice. This characterisation method mainly examines the functional groups that are deposited on the nanoparticle surface [bib_ref] Green Synthesis of Copper Nanoparticles Using Plantago asiatica Leaf Extract and Their..., Nasrollahzadeh [/bib_ref] [bib_ref] Biogenic Copper Oxide Nanoparticles Synthesis Using Tabernaemontana divaricate Leaf Extract and Its..., Sivaraj [/bib_ref] [bib_ref] Biosynthesis of Mixed Nanocrystalline Zn-Mg-Cu Oxide Nanocomposites and Their Antimicrobial Behavior, Suresh [/bib_ref]. Although FTIR can identify the functional groups that act as stabilizing and reducing agents, it has difficulty determining which specific phytoconstituents of a complex plant extract they originated with. Further research employing other characterization methods such as liquid chromatography-mass spectrometry and nuclear magnetic resonance can be carried out to identified the chemical structures of the contributing phytochemicals [bib_ref] Identification of Eschweilenol C in Derivative of Terminalia fagifolia Mart. and Green..., De Araujo [/bib_ref].
## Potential new applications
At present, most Cu-NMs produced via plant-mediated methods are synthesised using leaf extracts. There is plenty of room for future research into the exploitation of other plant components (peelings, roots and rhizomes, fruits, flowers, and seeds) for the synthesis of Cu alloy, core shell, and nanoparticles. In addition, there remains a knowledge gap regarding the effect of method parameters on the morphology of the synthesised particles. From an application perspective, most uses of plant-mediated Cu nanomaterials are focused on biomedical and environmental remediation. However, there are more applications that have yet to be discovered. For example, in the biological sector, research into the use of plant-mediated Cu-NMs mainly concerns their antibacterial ability and lesser antifungal ability. The relatively lower antifungal performance is owed to fungal cells less readily adsorbing nanoparticles at low concentrations as compared to bacterial cells. In addition, the binding of nanoparticles to the bacterial surface blocks bacterial respiration, whereas for fungal or eukaryotic cells, respiration occurs in the mitochondrial membrane and so is less susceptible to direct inhibition by nanoparticles [bib_ref] Noble Metal Nanoparticles: Plant-Mediated Synthesis, Dauthal [/bib_ref]. There has also been limited research on the antiviral and antiparasitic abilities of plant-mediated Cu nanoparticles; further investigation in this area may expand their antimicrobial capabilities. Finally, other noble-metal (Au and Ag) nanomaterials produced via plant-mediated synthesis have been employed in other applications such as electrochemistry, detection, surfaceenhanced Raman scattering, phase transfer, transmetallation, and modified glassy carbon electrodes; therefore, research can be carried out to expand the usage of Cu-NMs to these applications [bib_ref] Noble Metal Nanoparticles: Plant-Mediated Synthesis, Dauthal [/bib_ref].
## New research directions for synthesis
One area of future research for the application of Cu-NMs and their industrial-scale green production is leveraging machine learning. Specifically, machine-learning algorithms can be used in two respects, synthetic outcome prediction and experiment planning [bib_ref] Nanoparticle Synthesis Assisted by Machine Learning, Tao [/bib_ref]. For the first, an algorithm mathematically learns the relationship between nanomaterial properties and experimental conditions, then predicts from an example synthetic parameters dataset and the results of past experiments the characteristics of the nanoparticles that will be produced. Meanwhile, experiment-planning algorithms aim to suggest the best reaction conditions for achieving desired nanomaterial properties [bib_ref] Nanoparticle Synthesis Assisted by Machine Learning, Tao [/bib_ref]. This can aid in reducing the time and research effort required to obtain a desired outcome, such as the uniformity of the produced particles. Most machine-learning studies to date have concentrated on chemical-based nanomaterial synthesis; only a limited number of publications have concerned green synthesis; hence, there remains a large gap in this area [bib_ref] Intelligent Control of Nanoparticle Synthesis through Machine Learning, Lv [/bib_ref] [bib_ref] Application of Artificial Neural Network as a Nonhazardous Alternative on Kinetic Analysis..., Devaraj [/bib_ref]. Addressing this gap can help in making the process of Cu nanomaterial synthesis become less labour intensive, more cost effective, less time consuming, more productive, and able to yield higher quality nanoparticles, all of which are important from the industrial perspective.
# Conclusions
This literature review focused on the green synthesis of Cu nanomaterials. Compared to chemical or physical synthesis methods, green synthesis and especially plant-mediated synthesis is more environment-friendly, less toxic, and safe throughout the whole production process. The production methodology was discussed with further focus on plant-mediated nanomaterial synthesis, including the plant extraction method and Cu-NM (pure metal, metal oxide, alloy, core shell, and nanoparticles) synthesis. Leaf-extract-mediated Cu nanomaterials comprise the majority produced to date, with few synthesised using other types of plant components. The review also considered the biological and environmental applications of plant-mediated Cu-NMs. With regard to biological applications, antiviral and antiparasitic activities have received less focus than antibacterial. There also remain many research gaps regarding the application of green synthesis Cu-NMs in other sectors. Finally, current limitations and solutions with potential future research targets were described. Biomass obtainability, complexity of plant systems, underlying synthesis process, nano-material quality, and low productivity are the future challenges that need to be addressed in order to further broaden the application of plant-mediated nanomaterial synthesis.
In short, plant-mediated nanomaterial synthesis is eco-friendly, has low toxicity, and avoids using hazardous chemicals. The process can be separated into two parts, the plant extraction and the nanomaterial production. Different plant extracts with different parameters can produce nanomaterials of different sizes and geometries. As such, plant source accessibility, diversity of phytochemicals in extracts, knowledge of the synthesis mechanism, and nanomaterial quality are the limitations that presently hinder the future industrial production and application of plant-mediated nanomaterials. More research is needed in areas of the biotechnological sector such as phytochemical profiling, molecular science, tissue culture, and bioprocesses to overcome these issues. Separately, machine learning can also be adopted as a new research topic to further improve the green synthesis of Cu-NMs with better industrial applicability. Once these problems and research directions are resolved and fulfilled, respectively, the potential of plant-mediated nanomaterial synthesis could be fully unleashed in myriad applications, providing processes and materials with better sustainability and friendliness toward the environment.
[fig] Figure 1: General steps in plant extraction. [/fig]
[fig] Figure 2: (a) Thymus vulgaris leaf extract and (b) solution after green synthesis of CuO nanoparticles. Adapted with permission from Ref.[85]. 2016, Elsevier. [/fig]
[fig] Figure 4: Colour change over time during the reaction between Citrus reticulata peel extract and CuSO4.5H2O at (a) 0 min, (b) 60 min and (c) 105 min. Adapted with permission from Ref.[186]. 2020, Elsevier. [/fig]
[fig] Figure 5: Diagram of the chemisorption-based mechanism of Cu-based nanomaterials' antimicrobial activity. [/fig]
[fig] Figure 6, Figure 6: Cont(a) Diagram of the respective mechanisms of CuO nanoparticle antibacterial activity in Gram-positive and Gram-negative bacteria and (b) diagram of the mechanism of CuO nanoparticle antifungal activity. Adapted with permission from Ref.[205]. 2020, MDPI. [/fig]
[fig] Figure 6: (a) Diagram of the respective mechanisms of CuO nanoparticle antibacterial activity in Gram-positive and Gram-negative bacteria and (b) diagram of the mechanism of CuO nanoparticle antifungal activity. Adapted with permission from Ref.[205]. 2020, MDPI. [/fig]
[fig] Author: Contributions: Investigation, writing-original draft, J.V.; Conceptualization, writingreview and editing, K.S.L.; Supervision, visualization, Y.C.-Y.E.; Funding acquisition, validation, S.X.C.; Supervision, validation, M.S.; Conceptualization, supervision, writing-review and editing, C.H.C. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: This research was funded by the Center for Research and Instrumentation Management (CRIM), Universiti Kebangsaan Malaysia, grant number FRGS/1/2019/STG01/UKM/02/11. Data Availability Statement: Not applicable. [/fig]
[table] Table 1: Parameters and extraction method utilised for extraction of different plant components. [/table]
[table] Table 2: Pros and cons of various plant extraction methods. [/table]
[table] Table 3: Summary of plant-mediated Cu nanomaterial synthesis: plant extract type, key compounds, Cu precursors, synthesis time and temperature, reaction completion colour, and the Cu nanomaterial product, geometry, and size. [/table]
[bib_ref] Etlingera Elatior-Mediated Synthesis of Gold Nanoparticles and Their Application as Electrochemical Current..., Azri [/bib_ref] |
Heart Disease and Pregnancy
Cardiac disease remains a major cause of morbidity and mortality in pregnant and post-partum women, although progress has been made, with specialist joint obstetric-cardiology clinics providing an integrated, safe and personalised service to these women. As a result, fewer non-specialist cardiologists are managing women in pregnancy with cardiovascular disease. The aim of this review is to provide a brief overview of current knowledge and practice in the field, with an emphasis on the major physiological changes which occur during pregnancy, focussing on progress through the trimesters, clinical assessment in pregnancy, management of delivery (concentrating on managed vaginal delivery), drug treatment, key conditions and risk assessment. The latter factor is particularly important in terms of being able to identify high-risk women earlier and to counsel them appropriately. Pregnant women with cardiovascular conditions can, with appropriate knowledge and counselling, be managed safely in specialist multidisciplinary services, but there is a need for cardiologists to understand the key changes and risks involved in pregnancy, delivery and the post-partum period.
# Introduction
With improved maternal medical care and fertility treatments, an increasing proportion of women with congenital cardiac disease and acquired heart disease are becoming pregnant and delivering safely [bib_ref] Medical and obstetric outcomes among pregnant women with congenital heart disease, Thompson [/bib_ref]. However, pregnancy has a profound effect on the cardiovascular system, and many conditions are associated with a significant risk of foetal and maternal morbidity and mortality.
It has been reported that 0.2-0.4% of all pregnancies are complicated by cardiovascular disease [bib_ref] ESC Guidelines on the management of cardiovascular diseases during pregnancy. The task..., Regitz-Zagrosek [/bib_ref] , and although death is rare, cardiovascular disease is the biggest indirect cause of maternal death worldwide, with an attributable rate of two deaths per 100,000in the UK and a similar rate in other countries [bib_ref] Incidence and causes of maternal mortality in the USA, Kuriya [/bib_ref] [bib_ref] MBRRACE-UK-the new home for the confidential enquiries into maternal deaths-reports for the..., Bamber [/bib_ref]. The epidemiology of cardiovascular disease in pregnancy varies significantly depending the location of the mother, with much variation in disease rates and processes. Worldwide, hypertensive disease in pregnancy is by far the most prevalent cardiovascular disorder, complicating 2-8% of all pregnancies in the Western world, predominantly in Latin America and the Caribbean, where it causes one-quarter of all maternal deaths [bib_ref] Incidence and prevalence of pregnancy-related heart disease, Sliwa [/bib_ref]. Rheumatic heart disease is common in developing countries but are now rare in the Western world [bib_ref] The worldwide epidemiology of acute rheumatic fever and rheumatic heart disease, Seckeler [/bib_ref]. A similar variation is observed in peripartum cardiomyopathy (PPCM), with a rising incidence in Western society, estimated most recently at 1:2229 [bib_ref] Peripartum cardiomyopathy: current management and future perspectives, Hilfiker-Kleiner [/bib_ref] , but an estimated rate of 1:1000 in Africa and 1:300 in Haiti [bib_ref] Incidence and prevalence of pregnancy-related heart disease, Sliwa [/bib_ref]. An exact prevalence of women with congenital cardiac disorders who are pregnant is difficult to estimate largely due to the methodology involved in the reported studies, but in the USA nine per 10,000 delivery hospitalisations were for women with congenital heart disease [bib_ref] Medical and obstetric outcomes among pregnant women with congenital heart disease, Thompson [/bib_ref].
In most Western countries, most women seen in cardiac-obstetric services have congenital heart disease. Death is rare, but when it does occur it is most often due to myocardial infarction (often coronary dissection), ascending aortic dissection, cardiomyopathy and sudden adult cardiac death [bib_ref] Saving mothers' lives. Reviewing maternal deaths to make motherhood safer, Cantwell [/bib_ref].
The first area addressed in this review is the physiology of the cardiovascular system during pregnancy. In this section we focus on the key changes that cardiologists need to be aware of.
This article does not contain any new studies with human or animal subjects performed by any of the authors.
## Physiological changes in pregnancy
Pregnancy has a dramatic effect on the cardiovascular system, and the effects are sustained into the post-partum period.
Before there is even placentation, there is systemic vasodilatation at around 5 weeks gestation. Systemic vascular resistance (SVR) is progressively reduced (by 35-40%) until the middle of the second trimester, when it plateaus before beginning to increase late in the third trimester [bib_ref] Cardiac output and related haemodynamics during pregnancy: a series of meta-analyses, Meah [/bib_ref]. This variation is associated with a drop in mean arterial blood pressure in the first two trimesters prior to its recovery in the third trimester to close to pre-pregnancy levels. The changes in SVR are thought to be due to the vasodilatory properties of oestrogen, progesterone and relaxin, which are all increased during pregnancy, and to reduced vasopressor receptor sensitivity. In addition, there also appears to be an increased amount of vascular nitric oxide production which contributes to the vasodilatation and reduced SVR [bib_ref] Nitric oxide and pregnancy, Sladek [/bib_ref].
Structurally there is a progressive increase in left ventricular end-diastolic volume and mass [bib_ref] Cardiac output and related haemodynamics during pregnancy: a series of meta-analyses, Meah [/bib_ref] [bib_ref] Cardiovascular magnetic resonance in pregnancy: insights from the cardiac hemodynamic imaging and..., Ducas [/bib_ref] into the third trimester, but there is a sharp drop in left ventricular mass late in the third trimester [bib_ref] Cardiovascular magnetic resonance in pregnancy: insights from the cardiac hemodynamic imaging and..., Ducas [/bib_ref]. From the first trimester onwards, there is an increase in maternal heart rate which continues throughout pregnancy [bib_ref] A longitudinal study of maternal cardiovascular function from preconception to the postpartum..., Mahendru [/bib_ref] and is on average 30 beats per minute higher than the baseline maternal heart rate by the end of the pregnancy. These adaptations are likely to be required to cope with the extra circulating blood volume, which has increased by an average of 40%, reaching a maximum value at 24 weeks.
As expected, given the physiological demands of the utero-placental circulation and the developing foetus, there is an increase in cardiac output, with the biggest increase of up to 45% from baseline occurring during the first trimester [bib_ref] Serial study of factors influencing changes in cardiac output during human pregnancy, Robson [/bib_ref] [bib_ref] Adaptation of the maternal heart in pregnancy, Hunter [/bib_ref]. The increase in cardiac output slows late in the second trimester and drops slightly late in the third trimester [bib_ref] Cardiac output and related haemodynamics during pregnancy: a series of meta-analyses, Meah [/bib_ref] (but still stays above pre-pregnancy levels). [fig_ref] Table 1: Summary of the key cardiovascular physiological changes which occur during pregnancy [/fig_ref] summarises the main cardiovascular physiology changes during pregnancy by trimeste, and summarises the main cardiovascular physiology changes during pregnancy by trimester and percentage change.
It should be noted that these haemodynamic changes most often regress to pre-pregnancy levels, but some studies have suggested that the changes in left ventricular mass and vascular resistance do not fully return to pre-pregnancy levels.
As well as the circulatory changes described above, there are also adaptive changes that occur in the great vessels and blood that are important to women with heart disease. Expression of oestrogen receptors in the aorta causes fragmentation of reticulin fibres, reduced amount of acid mucopolysaccharides and loss of the normal arrangement of elastin fibers, predisposing women to aortic dissection, particularly if they have an aortopathy [bib_ref] Arterial dissections associated with pregnancy, Nolte [/bib_ref]. Additionally, pregnancy is a hypercoagulable state, designed to reduce the risk of post-partum haemorrhage [bib_ref] Thrombophilia and pregnancy, Kupferminc [/bib_ref]. This results in an increased risk of clotting, most commonly venous thromboembolism, but women who require anticoagulation for heart disease are at increased risk. For those with mechanical heart valves, pregnancy is a high-risk undertaking.
## Pre-conception management
In this section we discuss those aspects of cardiac-pregnancy management which most physicians feel to be of the greatest importance, and the identification of women at risk and their pre-pregnancy counselling of their own individual risk involved in pregnancy and delivery.
## Pre-pregnancy counselling and risk assessment
The most important aspect of assessment of reproductive-age women with cardiac disease is pre-conception counselling. Any assessment needs to cover the risks of pregnancy to the mother and foetus. Risks to the mother include whether she can tolerate the expected haemodynamic changes that occur in pregnancy, the need for highly medicalised antenatal care and delivery, possibly a premature delivery, that may be far from home and the long-term effects on the heart condition of a pregnancy. In the case of cardiomyopathies [bib_ref] Pregnancy in patients with pre-existing cardiomyopathies, Stergiopoulos [/bib_ref] and systemic right ventricles [bib_ref] Long-term outcome following pregnancy in women with a systemic right ventricle: is..., Bowater [/bib_ref] , there is evidence that pregnancy can have a deleterious long-term effect on ventricular function. In women with complex cardiovascular disorders that adversely affect their life expectancy, frank discussions need to be had about the long-term commitment of parenting and the importance of family or other social support. Cardiopulmonary exercise testing can be useful in estimating the likelihood of complications [bib_ref] ESC Guidelines on the management of cardiovascular diseases during pregnancy. The task..., Regitz-Zagrosek [/bib_ref]. Drugs that are contraindicated in pregnancy need to be stopped, and if cardiac function is dependent on these drugs, such as angiotensin-converting enzyme inhibitors in left ventricular dysfunction, it is important to assess the woman again when they have been drug free for several months to ensure that ventricular function does not deteriorate.
Assisted reproduction involves a wide range of techniques and medications, some of which can be invasive and lead to significant cardiac strain, with some evidence of increased complications compared to unassisted reproduction [bib_ref] Assisted reproductive technology and pregnancy-related hypertensive complications: a systematic review, Thomopoulos [/bib_ref]. These risks should be discussed, not only with the patient before she embarks on any assisted reproduction programme but also with the cardiology team looking after the patient so as to best manage the risks. As multiple foetuses are a more common occurrence in assisted fertility programmes, the extra haemodynamic stress and likelihood of premature delivery/increased rates of hypertension also need to be discussed with the patient before she starts the programme [bib_ref] Pregnancy-related complications and adverse pregnancy outcomes in multiple pregnancies resulting from assisted..., Qin [/bib_ref].
Risks to the foetus that need to be discussed are the effects of drugs that may need to be continued and the possibility of miscarriage, prematurity, intra-uterine growth restriction and low birth weight and its implications. Detailed pre-conception counselling is especially important in the case of anticoagulation (this aspect will be discussed in more detail further in the text).
It is also important to consider the likelihood of recurrence of the disease in the case of inherited cardiac conditions and congenital heart disease. In autosomal dominant conditions, such as Marfan syndrome, many patients are aware of the inheritance risk as family members are commonly affected. However, many patients are unaware of the inheritance risk of congenital heart disease and often assume it to be higher than it is. Recurrence risk varies from 0 to 20%, depending on the lesion. The most heritable lesions are left-sided obstructive conditions, such as bicuspid aortic valve disease and coarctation, and the least heritable are the transposition complexes [bib_ref] Assisted reproductive technology and pregnancy-related hypertensive complications: a systematic review, Thomopoulos [/bib_ref] [bib_ref] Pregnancy-related complications and adverse pregnancy outcomes in multiple pregnancies resulting from assisted..., Qin [/bib_ref] [bib_ref] Maternal folic acid supplementation and the risk of congenital heart defects in..., Feng [/bib_ref] [bib_ref] Prenatal multivitamin supplementation and rates of congenital anomalies: a meta-analysis, Goh [/bib_ref] [bib_ref] Cardiac disease in pregnancy I: prospective multicenter study of pregnancy outcomes in..., Siu [/bib_ref] [bib_ref] Predictors of pregnancy complications in women with congenital heart disease, Drenthen [/bib_ref].
There is some evidence that pre-conception multivitamins and folic acid supplementation reduces the risk of heritable congenital heart disease in the foetus [bib_ref] Maternal folic acid supplementation and the risk of congenital heart defects in..., Feng [/bib_ref] [bib_ref] Prenatal multivitamin supplementation and rates of congenital anomalies: a meta-analysis, Goh [/bib_ref].
## Risk assessment for individual cardiovascular disorders
Assessment of maternal risk prior to conception has been studied by several groups and refined in risk assessment scores, such as CARPREG [bib_ref] Cardiac disease in pregnancy I: prospective multicenter study of pregnancy outcomes in..., Siu [/bib_ref] and ZAHARA [bib_ref] Predictors of pregnancy complications in women with congenital heart disease, Drenthen [/bib_ref] , which are summarised in [fig_ref] Table 2: CARPREG and ZAHARA scoring systems for estimating the risk of a cardiac... [/fig_ref]. Factors that increase the risk of maternal cardiac events include left heart obstructive lesions (aortic or mitral stenosis), cardiac symptoms, cyanosis, systemic ventricular impairment and previous cardiac events. Atrioventricular valve regurgitation (mitral or tricuspid) also confers an increased risk. However, the most high-risk situation is the presence of a mechanical valve and the associated risk of full anticoagulation.
The current European Society of Cardiology guidelines suggest using the modified World Health Organisation (WHO) system [bib_ref] Risks of contraception and pregnancy in heart disease, Thorne [/bib_ref] when assessing women prior to pregnancy, and this forms the basis for the small number of conditions that are thought to be a contraindication to pregnancy (shown in [fig_ref] Table 3: Modified World Health Organisation system for risk assessment of cardiac conditions in... [/fig_ref].
## Pregnancy and pre-delivery
After the key phase of pre-assessment, and care focusses on the assessment during pregnancy and the management of specific conditions.
## Clinical assessment and investigations during pregnancy
Cardiac assessment of the pregnant patient can be difficult as common symptoms of pregnancy, such as breathlessness and fatigue, can mimic cardiac symptoms. Also, clinical signs, such as mild dependent oedema, a minimally raised jugular venous pressure, collapsing pulses and an ejection systolic murmur, are common in pregnancy and can result in difficult assessments [bib_ref] Pregnancy and delivery in cardiac disease, Ruys [/bib_ref]. Signs and symptoms which are abnormal in pregnancy include extreme breathlessness, marked oedema, a fourth heart sound, diastolic murmurs, jugular venous pressure of [2 cm and a persistent tachycardia of [100 beats per minute; any one of these should prompt further evaluation [bib_ref] Heart disease in pregnancy, Emmanuel [/bib_ref]. Electrocardiography (ECG) is a common and useful diagnostic tool throughout pregnancy for complaints such as chest pain/arrhythmias. Subtle changes in the ECG are common in later pregnancy and include left axis deviation, inverted T waves and inferior Q waves due to diaphragmatic elevation [bib_ref] Electrocradiographic Qrs Axis, Q wave and T-wave changes in 2nd and 3rd..., Sunitha [/bib_ref]. Chest radiography is a safe investigation modality in pregnancy and should be performed readily as required [bib_ref] Doctor, will that x-ray harm my unborn child?, Ratnapalan [/bib_ref].
After the ECG, the most common investigation that most cardiologists will utilise is the echocardiogram, which is based on ultrasound and therefore safe throughout the pregnancy. As would be predicted from the changes in maternal physiology, there is an observable increase in left ventricular diameter measurements on echocardiography, such as the left ventricular end diastolic and systolic dimensions [bib_ref] Structural and functional changes in maternal left ventricle during pregnancy: a three-dimensional..., Cong [/bib_ref]. Assessment of ventricular function is made utilising identical techniques to those used in the non-pregnant women, with ejection fraction, tissue Doppler and m-mode measurements all useful in serial monitoring. There have been some inconsistencies in the reporting [bib_ref] Structural and functional changes in maternal left ventricle during pregnancy: a three-dimensional..., Cong [/bib_ref] [bib_ref] Maternal left ventricular diastolic and systolic long-axis function during normal pregnancy, Bamfo [/bib_ref]. When assessing left ventricular function there does seem to be agreement across several studies that later in pregnancy the left ventricle becomes more globular in shape, accompanied by a drop in left ventricular longitudinal function and strain [bib_ref] Structural and functional changes in maternal left ventricle during pregnancy: a three-dimensional..., Cong [/bib_ref] [bib_ref] Morphological and functional adaptation of the maternal heart during pregnancy, Savu [/bib_ref] [bib_ref] Altered maternal left ventricular contractility and function during normal pregnancy, Estensen [/bib_ref] ; this would be in keeping with the late rise in afterload caused by the slight increase in SVR [bib_ref] Morphological and functional adaptation of the maternal heart during pregnancy, Savu [/bib_ref].
In the assessment of valvular lesions using echocardiography, pregnancy-related changes can often affect the severity of a valve lesion or the measurement made by echocardiography. In the context of stenotic valvular lesions, the extra volume load and heart rate increase seen in pregnancy can lead to an increase in the gradient measured across a valve without any observable change in the valve area measured by, for example, the continuity equation [bib_ref] Inaccurate noninvasive mitral valve area calculation during pregnancy, Rokey [/bib_ref]. When measuring valve regurgitation, it is important to note that the extra volume load in pregnancy leads to an increase in observable tricuspid regurgitation without necessarily any change in the valve function; in contrast, mitral regurgitation, despite the extra volume load, often appears less during pregnancy due to the drop in SVR. It is important for cardiologists to undertake serial scans and look at trends in valve pathology, and these changes need to be accompanied by clinical assessment.
Cross-sectional imaging can be extremely valuable in diagnosis and surveillance during pregnancy. Cardiac magnetic resonance imaging is safe after the first trimester although gadolinium injection is not used due to a lack of data on its safety [bib_ref] A review of the current use of magnetic resonance imaging in pregnancy..., De Wilde [/bib_ref]. Computed tomography involves ionising radiation but may be required in spite of this drawback in life-threatening scenarios, such as a concern about aortic dissection or pulmonary embolus. Cardiac catheterisation should be avoided in pregnancy but if required (for example in emergency pacing) the radiation dose should be minimised as much as possible and lead shielding used across the abdomen to reduce foetal exposure [bib_ref] Use of lead shielding on pregnant patients undergoing CT scans: results of..., Iball [/bib_ref].
## Common and important cardiovascular disorders encountered during pregnancy
Specific guidance on some of the more commonly encountered conditions and important high-risk conditions are discussed in this section. [bib_ref] Outcome of pregnancy in patients with structural or ischaemic heart disease: results..., Roos-Hesselink [/bib_ref].The normal physiological changes of pregnancy of increased heart rate, cardiac output and circulating volume are particularly problematic for women with impaired systemic ventricular function. These women often experience a worsening of symptoms, particularly from around 16 weeks of gestation. Some women who are not diagnosed prior to pregnancy may be erroneously diagnosed with asthma or respiratory tract infections before the diagnosis of cardiomyopathy is made. Ventricular dysfunction due to ischaemic heart disease is very uncommon; most women have familial or idiopathic dilated cardiomyopathy. Some may have had previous chemotherapy or previous peripartum cardiomyopathy. Diagnosis is made clinically and most commonly in conjunction with echocardiography, with dysfunction often seen as a worsening on serial echocardiography. More recently, medical units have begun to incorporate serial B-type natriuretic peptide (BNP) as a method for monitoring women with ventricular dysfunction; a worsening in their condition can be then identified earlier, and BNP is also a good negative predictive marker if \100 [bib_ref] B-type natriuretic peptide in pregnant women with heart disease, Tanous [/bib_ref].
Heart failure symptoms in pregnancy should be managed with rest and, as angiotensin-converting enzyme inhibitors and angiotensin receptor blockers are contraindicated, a combination of hydralazine and nitrate can be used to reduce cardiac afterload. Some women may need diuretics, although due to the pregnancy-related increase in glomerular filtration rate, fewer diuretics than normal may be needed. Beta-blockers can also be a useful adjunct therapy, and cardo-selective beta-blockers, such as bisoprolol and metoprolol, are more preferred. Regular foetal growth scans are needed. Careful assessment of the likelihood of the pregnancy reaching viability needs to be assessed, and occasionally the decision may be made that continuing the pregnancy represents too high a risk to the mother, with advice that the pregnancy be terminated on medical grounds.
Women may require in-patient treatment and rest, and early delivery may be needed. The maternal mortality recorded in the ROPAC for cardiomyopathies is 2.4%, but the need for hospital admission is 33% [bib_ref] Outcome of pregnancy in patients with structural or ischaemic heart disease: results..., Roos-Hesselink [/bib_ref]. Prognosis in pregnancy for ventricular dysfunction as illustrated earlier based on data in the WHO scoring system and others is dependent on the level of ventricular dysfunction and symptom severity [bib_ref] Risks of contraception and pregnancy in heart disease, Thorne [/bib_ref] [bib_ref] Sorensen S. Prospective multicenter study of pregnancy outcomes in women with heart..., Siu [/bib_ref].
Peripartum cardiomyopathy is diagnosed when acute left ventricular impairment occurs between the last 4 weeks of pregnancy and 5 months post-partum [bib_ref] Current state of knowledge on aetiology, diagnosis, management, and therapy of peripartum..., Sliwa [/bib_ref]. The aetiology is unknown, and there is a wide geographical variation in frequency, with countries such as Haiti having a particularly high prevalence. Recent experimental work suggests a potential pathogenic role for prolactin [bib_ref] Prolactin: a new therapeutic target in peripartum cardiomyopathy, Yamac [/bib_ref] and an overlap with hypertensive disorders and pre-eclampsia [bib_ref] The relationship between preeclampsia and peripartum cardiomyopathy: a systematic review and meta-analysis, Bello [/bib_ref]. Treatment should be as for other types of heart failure. Bromocriptine, a prolactin-blocking drug, has been used on a case-by-case basis [bib_ref] Recovery from postpartum cardiomyopathy in 2 patients by blocking prolactin release with..., Hilfiker-Kleiner [/bib_ref] and in a small trial [bib_ref] Evaluation of bromocriptine in the treatment of acute severe peripartum cardiomyopathy. A..., Sliwa [/bib_ref] , with some positive results and is currently undergoing testing in a larger randomised control trial. The drug has the disadvantage of precluding breast-feeding [bib_ref] Evaluation of bromocriptine in the treatment of acute severe peripartum cardiomyopathy. A..., Sliwa [/bib_ref] , and there maybe some evidence that breast-feeding improves recovery in PPCM [bib_ref] Clinical outcomes for peripartum cardiomyopathy in North America: results of the IPAC..., Mcnamara [/bib_ref]. Prompt echocardiography is essential for early diagnosis, and a low index of suspicion for the condition is needed. Response to treatment varies widely, with some women recovering fully and some requiring heart transplantation or even dying. Positive predictive factors are early diagnosis, having an ejection fraction of [30% or a left ventricular end diastolic dimension of \6.0 cm at the time of diagnosis [bib_ref] Peripartum cardiomyopathy: prognostic factors for long-term maternal outcome, Habli [/bib_ref]. Approximately 50% of women recover fully at 6 months and the condition can recur in 25% women, with those with residual left ventricular function faring worse in subsequent pregnancies [bib_ref] Maternal and fetal outcomes of subsequent pregnancies in women with peripartum cardiomyopathy, Elkayam [/bib_ref]. Left ventricular recovery and survival differs between ethnic groups, with African-American women doing worse than Caucasians despite comparable treatment [bib_ref] Poor outcome of indigent patients with peripartum cardiomyopathy in the United States, Modi [/bib_ref].
## Left heart obstruction
In the ROPAC registry, aortic stenosis and mitral stenosis are the two most common left-sided obstructive lesions, accounting for 31% of the valvular heart disease patients seen in the registry [bib_ref] Outcome of pregnancy in patients with structural or ischaemic heart disease: results..., Roos-Hesselink [/bib_ref]. To date there are insufficient data available worldwide on the incidence of other obstructive lesions, such subvalvular or supravalvular lesions or hypertrophic cardiomyopathy with an obstructive gradient. Due to the extra volume load and inability to increase cardiac output across a fixed obstruction associated with both aortic and mitral stenosis, these conditions are significantly problematic during pregnancy and carry a high risk of morbidity and mortality [bib_ref] Cardiac risk in pregnant women with rheumatic mitral stenosis, Silversides [/bib_ref] [bib_ref] Early and intermediate-term outcomes of pregnancy with congenital aortic stenosis, Silversides [/bib_ref]. Symptoms and signs often include increasing breathlessness, fatigue and oedema, and diagnosis if unknown prior to conception is usually confirmed with echocardiography.
For mitral stenosis, the mainstay of treatment is rest, beta-blockade and diuretics for heart failure symptoms. Percutaneous balloon valvuloplasty has been widely used in rheumatic mitral stenosis and should be performed readily if the woman develops symptoms or foetal growth is compromised [bib_ref] Management of severe mitral stenosis during pregnancy, Norrad [/bib_ref]. Aortic stenosis is better tolerated but often more problematic as the options for treatment are limited. It is often due to bicuspid valve disease, and the results from balloon valvuloplasty are not so predictable or successful. Diuresis is not helpful, nor is beta-blockade. In both conditions, there is an increased risk of prematurity and intra-uterine growth restriction [bib_ref] The effect of valvular heart disease on maternal and fetal outcome of..., Hameed [/bib_ref]. In moderate mitral stenosis and severe aortic stenosis, decisions around delivery often need to be made on a week by week basis as pregnancy progresses.
## Arrhythmias
Many women experience palpitations during pregnancy that are not of clinical concern, but a small number of women with pre-existing or new diagnosed rhythm disorders may need treatment during pregnancy. Initial management should be as for any other patient with rhythm disorder: prompt diagnosis based on ECG and appropriate medical treatment [bib_ref] Managing palpitations and arrhythmias during pregnancy, Adamson [/bib_ref]. Safe therapies for tachyarrhythmias during pregnancy include adenosine, digoxin, flecainide and verapamil. Beta-blockers can cause intra-uterine growth restriction, but they are an effective therapy against most tachyarrhythmias in pregnancy. In the context of clinical compromise and haemodynamic instability, electrical cardioversion should be used as normal, and no detrimental effects have been reported on the foetus due to electrical cardioversion, although foetal assessment afterwards is recommended [bib_ref] Electrical cardioversion during pregnancy: safe or not?, Tromp [/bib_ref].
Atrial fibrillation should be managed as in the non-pregnant patient, with attention paid to the need, or not, for anti-coagulation prior to or following a cardioversion.
Catheter ablation is reserved for drug-resistant arrhythmias leading to haemodynamic compromise or refractory symptoms and if needed can be performed without fluroscopy [bib_ref] Catheter ablation of atrial fibrillation using zero-fluoroscopy technique: a randomized trial, Bulava [/bib_ref].
Ventricular arrhythmias (VAs) are much less common in pregnancy than supraventricular arrhythmias and occur most commonly in women with known heart disease at a frequency of approximately 1% [bib_ref] Cardiac disease in pregnancy I: prospective multicenter study of pregnancy outcomes in..., Siu [/bib_ref]. VAs in women with heart disease occur most commonly late in pregnancy, usually in the third trimester with symptoms of heart failure [bib_ref] Ventricular tachyarrhythmia during pregnancy in women with heart disease: data from the..., Ertekin [/bib_ref]. Management, as stated earlier, should be as for the non-pregnant woman, with medications (amiodarone avoided) and defibrillator [implanted cardioverter defibrillator (ICD)] therapy as needed or emergency cardioversion. Pregnancy itself does not seem to lead to an excess of ICD therapies [bib_ref] Implantable cardioverter-defibrillators and pregnancy. A safe combination?, Natale [/bib_ref].
Congenital long QT syndromes are associated with a risk of syncope and ventricular arrhythmias and a small risk of sudden death. Pregnancy is associated with a reduced risk of events during pregnancy but an increased risk of events in the first 9 months post-partum [bib_ref] Long QT syndrome and pregnancy, Seth [/bib_ref] , with the risk being most marked for long QT2 [bib_ref] Long QT syndrome and pregnancy, Seth [/bib_ref]. It is safe to go through pregnancy with an ICD and important not to switch the device off during delivery. If Caesarean section is required unipolar diathermy should be avoided.
Bradyarrhythmias are much less common in pregnancy, and permanent pacing is rarely required [bib_ref] Bradyarrhythmias in pregnancy: a case report and review of management, Adekanye [/bib_ref] ; however, if required, radiation exposure should be minimised. Some units have suggested isoproterenol can be used safely if needed in an emergency situation [bib_ref] Isoproterenol infusion for treatment of refractory symptomatic bradycardia in parturients with congenital..., Herman [/bib_ref].
## Coronary artery disease
Coronary artery dissection, normally a rare occurrence, is significantly more common in pregnancy, with [20% of all cases of spontaneous dissection involving pregnant women [bib_ref] Acute myocardial infarction associated with pregnancy, Roth [/bib_ref]. However, atherosclerotic coronary artery disease, which is the major burden of acquired cardiovascular disease worldwide, is becoming an increasing problem in pregnancy as women with traditional cardiovascular risk factors are becoming pregnant at an older age [bib_ref] Acute myocardial infarction in pregnancy. A United States population-based study, James [/bib_ref].
The rate of maternal mortality due to an acute coronary syndrome is thought to be 5-10% and, as with aortic dissection, peaks in the peripartum period. Pregnant women presenting with chest pain should be assessed in the usual way with an ECG and appropriate biomarkers, such as troponin. Traditional risk factors do determine the risk in the pregnant woman in the same way as in the normal population, and these should be used in the risk assessment.
Treatment of an acute myocardial infarction should proceed as normal, with emergency angioplasty used per guidelines with bare metal stents as these have the widest historical evidence base in the pregnant woman [bib_ref] ESC Guidelines on the management of cardiovascular diseases during pregnancy. The task..., Regitz-Zagrosek [/bib_ref]. Thrombolysis should not be performed as coronary artery dissection is so often the cause. Clopidogrel as a dual anti-platelet may be added to the therapeutic regimen for use in as short a course as possible following stenting; there is currently no evidence available for prasugrel or ticagrelor. If clopidogrel is prescribed, then the administration of an epidural is not safe. Although the above caveats need to be borne in mind, the most important part of management is to treat the mother in the most effective way possible to save her life.
## Congenital heart disease
In developed countries, most women seen in cardiac-obstetric clinics have congenital heart disease [bib_ref] Current trends in the management of heart disease in pregnancy, Curtis [/bib_ref]. Due to advances in cardiac surgery and paediatric intensive care almost all children born with congenital heart disease now survive to adulthood, with increasingly complex conditions. Many of these children want themselves to have children. Given the wide spectrum of lesions and repairs, women with congenital heart disease should be managed jointly between obstetricians and specialist cardiologists. This spectrum of disorders in pregnancy is reviewed in detail elsewhere [bib_ref] American Heart Association Council on Cardiovascular and Stroke Nursing; Council on Clinical..., Canobbio [/bib_ref].
## Aortopathies
Aortopathies, such as Marfan syndrome, present a significant risk in pregnancy, with increased blood volume and vascular histological changes leading to increased risks of dissection. Dissection most commonly occurs in the third trimester and in the post-partum period, and those patients with dilated aortas and a family history of dissection are at greatest risk. Type B dissection is unpredictable. Frequent surveillance by ECG during pregnancy is important, and the dissection risk can be reduced by beta-blockade [bib_ref] Pregnancy and marfan syndrome: an ongoing discussion, Mulder [/bib_ref]. Delivery should be highly medicalised with a combined spinal-epidural and passive second stage. Very high-risk cases may need Caesarean section in cardiac theatres. Women should remain in hospital for 1 week after delivery.
Women with dilated aortas in the context of a bicuspid aortic valve are at much less risk of dissection [bib_ref] Frequency of cardiovascular events in women with a congenitally bicuspid aortic valve..., Mckellar [/bib_ref] , although regular surveillance by ECG during pregnancy is still important, and beta-blockade and a medicalised delivery may be recommended if the aorta is particularly dilated.
Turner's syndrome deserves a special mention as some mosaic patients do become pregnant. Around one-half of women with Turner's syndrome will have a cardiac defect (the most common being a bicuspid aortic valve and coarctation of the aorta). A large number are hypertensive. Moreover, Turner's syndrome is an independent risk factor for aortic dissection, and these patients should be managed as those with Marfan syndrome. It is important to take account of body surface area when measuring aortic size; these women are of small stature and the absolute aortic measurements may be normal.
## Management of pharmacological treatment in pregnancy and post-partum
Women with existing cardiac disorders were possibly on a variety of medications prior to trying to become pregnant, and the decision to continue these medications during pregnancy is based on balancing the risk to the mother of stopping them, the availability of a safe alternative and the risk to the foetus. Due to the ethical issue of performing clinical trials in pregnant women the only data on drugs in pregnancy are from observational studies and based on evidence from animal trials. Whenever possible, prior to conception drugs considered to be unsafe during pregnancy should be stopped. However, many drugs need to be continued in pregnancy, and close liaison with foetal medicine specialists is important.
A summary of the risk of common cardiac drugs during pregnancy are listed below in [fig_ref] Table 4: Common cardiac drugs and their risks in pregnancy [/fig_ref] [bib_ref] ESC Guidelines on the management of cardiovascular diseases during pregnancy. The task..., Regitz-Zagrosek [/bib_ref].
Anti-coagulation is a source of problems in pregnancy. Warfarin is associated with foetal [bib_ref] Use of lead shielding on pregnant patients undergoing CT scans: results of..., Iball [/bib_ref]. Where possible it should be substituted for low-molecular-weight heparin (LMWH), which does not cross the placenta. However in the case of mechanical valves warfarin results in significantly less valve thrombosis than unfractionated heparin or LMWH [bib_ref] Pregnancy with prosthetic heart valves-30 years nationwide experience in Denmark, Sillesen [/bib_ref] [bib_ref] Hall R, and on behalf of the ROPAC investigators and the EORP..., Van Hagen [/bib_ref]. Some women choose to substitute warfarin with LMWH for the period of embryopathy risk (6-12 weeks of gestation) and then resume taking warfarin for most of the pregnancy. Because warfarin crosses the placenta, vaginal delivery is unsafe for the foetus, and so warfarin is stopped and substituted with LMWH at 36 weeks. If labour begins prematurely when the woman is still taking warfarin, a caesarean section must be performed. Some studies have suggested that doses of [5 mg are associated with more complications that lower doses [bib_ref] Dose-dependent fetal complications of warfarin in pregnant women with mechanical heart valves, Vitale [/bib_ref] [bib_ref] Anti-coagulation during pregnancy in women with mechanical heart valves: a prospective study, Khamoushi [/bib_ref] [bib_ref] The effect of warfarin dosage on maternal and fetal outcomes in pregnant..., Cotrufo [/bib_ref] , but this finding is not universal
[formula] [76]. [/formula]
Currently there is no universally agreed upon strategy for the monitoring of enoxaparin use, with some centres using solely peak or trough levels of factor Xa, or both [bib_ref] Anticoagulation during pregnancy: evolving strategies with a focus on mechanical valves, Alshawabkeh [/bib_ref] , and the strategy decided upon may be particularly important in the first trimester due to larger proportional changes in plasma volume and creatinine clearance [bib_ref] Population pharmacokinetics of enoxaparin during the antenatal period. Clinical perspective, Patel [/bib_ref].
## Management of labour and delivery in women with cardiovascular disorders
The follow-up and delivery of high-risk pregnancy patients fitting into WHO class 3 [bib_ref] The worldwide epidemiology of acute rheumatic fever and rheumatic heart disease, Seckeler [/bib_ref] [bib_ref] Sorensen S. Prospective multicenter study of pregnancy outcomes in women with heart..., Siu [/bib_ref] should occur in experienced centres where combined experienced anaesthetic, obstetric, cardiology and foetal care can be provided. During follow-up visits and assessments in specialist cardiac-obstetric clinic, an individualised birth plan is drawn up jointly between the cardiac and obstetric teams, and this formal plan is placed in each woman's medical notes so that accepting teams at non-specialist centres have the appropriate information for a safe delivery in an emergency. In many cases labour can be awaited naturally, although in some cases in which there is a desire to deliver early due to maternal complications, or for geographical reasons to allow for delivery in a specialist centre, induction of labour may be recommended .
Women with cardiovascular disorders have for many years been subject to a higher rate of Caesarean sections compared to other women, and it has been suspected that this is due to a clinician's perception of the risk involved in changing to an emergency Caesarean section [bib_ref] Is a planned caesarean section in women with cardiac disease beneficial?, Ruys [/bib_ref]. Caesarean section is associated with more profound and sudden haemodynamic changes, greater blood loss, increased risk of infection and dramatically increased risk of venous thromboembolism [bib_ref] Cesarean delivery and postpartum maternal mortality: a population-based case control study in..., Esteves-Pereira [/bib_ref]. Furthermore, a Caesarean section will significantly affect the mode of future deliveries. The combination of a carefully titrated combined spinal-epidural, passive second-stage and assisted delivery with forceps allow a ''pain free, pushing free labour'' that is more physiological and safer for the mother [bib_ref] Pregnancy and delivery in cardiac disease, Ruys [/bib_ref] [bib_ref] Investigators obotR: Is a planned caesarean section in women with cardiac disease..., Ruys [/bib_ref]. However, it must be done in experienced hands by a specialist team. Caesarean section is reserved for obstetric indications; in rare cases, if there is a very high risk of aortic dissection, a planned delivery in theatres with a cardiac surgeon on stand-by is recommended.
The third stage of labour is usually facilitated by the administration of syntometrine, which causes uterine contractions and minimises post-partum haemorrhage. However, this drug also has a pronounced hypertensive and vasoconstrictive effect and so is avoided in most women with heart disease in favour of a slow infusion of syntocinon (an oxytocin analogue) [bib_ref] The management of the third stage of labour in women with heart..., Cauldwell [/bib_ref]. Misoprostol may be used for post-partum haemorrhage, with syntometrine reserved only for severe maternal haemorrhage where the benefit outweighs the risk [bib_ref] The management of the third stage of labour in women with heart..., Cauldwell [/bib_ref].
## Post-partum evaluation and management of women with cardiovascular disorders
Whilst pre-pregnancy assessment and surveillance during pregnancy are very important, post-partum follow-up is also important to assess for deterioration following pregnancy [bib_ref] Long-term outcome following pregnancy in women with a systemic right ventricle: is..., Bowater [/bib_ref] and to monitor for peripartum cardiomyopathy, which can occur late in the post-partum period. Some conditions, such as long QT and Marfan syndrome, seem to be associated with more complications in the post-partum period, and advising women accordingly is important prior to discharge. Post-partum follow-up also allows for the reintroduction of important cardiac medications whose use may have been altered during pregnancy, such as angiotensin-converting enzyme inhibitors or warfarin.
Post-partum is a key opportunity to discuss contraception and the options available to women to reduce the risk of an unplanned pregnancy, as well as to allow recovery from the most recent pregnancy, especially in the context of peripartum cardiomyopathy.
# Conclusion
Pregnancy presents the heart with a unique physiological challenge. An increasing number of women with cardiac disease are going through pregnancy successfully. The range of conditions encountered is wide, and many patients have congenital heart disease. There are multiple management challenges to achieve a successful outcome for mother and child. Pre-pregnancy counselling is vital, and specialist multi-disciplinary care by a specialist team is essential to ensure a successful outcome for mother and child. authorship for this manuscript, take responsibility for the integrity of the work and have given final approval for the version to be published.
Disclosures. Reza Ashrafi and Stephanie L Curtis have nothing to disclose.
Compliance with Ethics Guidelines. This article does not contain any new studies with human or animal subjects performed by any of the authors.
Open Access. This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/ by-nc/4.0/), which permits any noncommercial use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
[table] Table 1: Summary of the key cardiovascular physiological changes which occur during pregnancy [/table]
[table] Table 2: CARPREG and ZAHARA scoring systems for estimating the risk of a cardiac complication during pregnancy Risk factor in CARPREG scoring (1 point for each factor) Total points Risk of a cardiac complication (%) CARPREG CARdiac Disease in PREGnancy, ZAHARA Zwangerschap bij Aangeboren HARt Afwijkingen I, AV atrioventricular, NYHA New York Heart Association [/table]
[table] Table 3: Modified World Health Organisation system for risk assessment of cardiac conditions in pregnancy Aortic dilatation 45-50 mm in bicuspid aortic valve disease of alterations in measurements during the pregnancy period, with some studies reporting no change and others reporting a small drop late in pregnancy [/table]
[table] Table 4: Common cardiac drugs and their risks in pregnancy [/table]
|
LnCompare: gene set feature analysis for human long non-coding RNAs
Interest in the biological roles of long noncoding RNAs (lncRNAs) has resulted in growing numbers of studies that produce large sets of candidate genes, for example, differentially expressed between two conditions. For sets of protein-coding genes, ontology and pathway analyses are powerful tools for generating new insights from statistical enrichment of gene features. Here we present the LnCompare web server, an equivalent resource for studying the properties of lncRNA gene sets. The Gene Set Feature Comparison mode tests for enrichment amongst a panel of quantitative and categorical features, spanning gene structure, evolutionary conservation, expression, subcellular localization, repetitive sequences and disease association. Moreover, in Similar Gene Identification mode, users may identify other lncRNAs by similarity across a defined range of features. Comprehensive results may be downloaded in tabular and graphical formats, in addition to the entire feature resource. LnCompare will empower researchers to extract useful hypotheses and candidates from lncRNA gene sets.
# Introduction
Long non-coding RNAs (lncRNAs) are a numerous yet poorly understood class of genes with growing biological and biomedical interest. Their regulatory roles (1) and tissue specificity (2,3) make them promising biomarkers and therapeutic targets [bib_ref] From discovery to function: the expanding roles of long noncoding RNAs in..., Sun [/bib_ref]. High-throughput studies on disease or biological systems routinely produce sets of tens to thousands of lncRNA candidates [bib_ref] CRISPRi-based genome-scale identification of functional long noncoding RNA loci in human cells, Liu [/bib_ref]. Some examples of such sets are lncRNAs exhibiting differential expression be-tween conditions (7), association with a disease [bib_ref] Lnc2Cancer v2.0: updated database of experimentally supported long non-coding RNAs in human..., Gao [/bib_ref] or whose perturbation by CRISPR-Cas9 leads to phenotypic changes [bib_ref] Discovery and functional analysis of lncRNAs: Methodologies to investigate an uncharacterized transcriptome, Kashi [/bib_ref]. Important bottlenecks arise in assessing the quality of such sets, and generating functional and mechanistic hypotheses from them [bib_ref] Genome-scale deletion screening of human long non-coding RNAs using a paired-guide RNA..., Zhu [/bib_ref] [bib_ref] PAN-cancer analysis of S-phase enriched lncRNAs identifies oncogenic drivers and biomarkers, Ali [/bib_ref].
Controlled ontologies describing gene functions or their products' characteristics, most notably Gene Ontology and Kyoto Encyclopedia of Genes and Genomes [bib_ref] Data, information, knowledge and principle: back to metabolism in KEGG, Kanehisa [/bib_ref] [bib_ref] Gene Ontology: tool for the unification of biology, Ashburner [/bib_ref] , are powerful and widely used tools for inspecting sets of protein-coding genes (PCGs) [bib_ref] PANTHER version 11: expanded annotation data from Gene Ontology and Reactome pathways,..., Mi [/bib_ref] [bib_ref] AmiGO and Web Presence Working Group (2009) AmiGO: online access to ontology..., Carbon [/bib_ref] [bib_ref] GOrilla: a tool for discovery and visualization of enriched GO terms in..., Eden [/bib_ref] [bib_ref] ) g:Profiler--a web server for functional interpretation of gene lists (2016 update), Reimand [/bib_ref]. Unfortunately, functional labels have not yet been directly assigned to lncRNAs, making them inaccessible to ontology analyses. As a result, lncRNA candidate sets cannot be mined for biological insights to the degree which we have come to expect for PCGs. In a similar way, there are a range of tools designed to identify other PCGs with various degrees of similarity to a gene of interest [bib_ref] G-SESAME: web tools for GO-term-based gene similarity analysis and knowledge discovery, Du [/bib_ref] [bib_ref] simDEF: definition-based semantic similarity measure of gene ontology terms for functional similarity..., Pesaranghader [/bib_ref] [bib_ref] Gene functional similarity search tool (GFSST), Zhang [/bib_ref] [bib_ref] An Efficient Online Tool to Search Top-N Genes with Similar Biological Functions..., Wang [/bib_ref] , but none are available for lncRNAs. In summary, there is a need for tools to analyse lncRNA gene lists and functionally prioritize candidates for further study.
To date, most methods dedicated to revealing functional insights from lncRNA sets rely on the biological properties of PCGs with correlated expression across tissues [bib_ref] From discovery to function: the expanding roles of long noncoding RNAs in..., Sun [/bib_ref]. For example, LncRNA2Function and Co-LncRNA web servers perform functional enrichment analysis on the co-expressed coding genes of the input lncRNAs [bib_ref] LncRNA2Function: a comprehensive resource for functional investigation of human lncRNAs based on..., Jiang [/bib_ref] [bib_ref] Co-LncRNA: investigating the lncRNA combinatorial effects in GO annotations and KEGG pathways..., Zhao [/bib_ref]. Lnc-GFP and LncRNAs2Pathways (23,25) follow a similar strategy but introducing more sophisticated graph theory algorithms on co-expression networks. Finally, FARNA (26) considers transcription factor (TF)-lncRNA associations to predict lncRNA functions. However, none of these resources directly explores the features of lncRNA genes and products.
To address this need, we have developed LnCompare, a web server that compares lncRNA genes across a range of features. LnCompare is based on a comprehensive feature set with more than 100 attributes covering diverse aspects, including gene structure, nucleotide composition, evolutionary conservation, cell and tissue expression, subcellular localization, tissue specificity, repetitive sequence content and phenotypic association. Based on these features, LnCompare has two main functionalities. First, Gene Set Feature Comparison identifies statistically-enriched features of lncRNA sets, in a similar way as is presently done for PCGs [bib_ref] GOrilla: a tool for discovery and visualization of enriched GO terms in..., Eden [/bib_ref] [bib_ref] ) g:Profiler--a web server for functional interpretation of gene lists (2016 update), Reimand [/bib_ref] [bib_ref] GOstat: find statistically overrepresented Gene Ontologies within a group of genes, Beißbarth [/bib_ref] [bib_ref] GOEAST: a web-based software toolkit for Gene Ontology enrichment analysis, Zheng [/bib_ref]. Second, Similar Gene Discovery functionality seeks to identify other similar genes for a given gene-of-interest, based on user-defined features. LnCompare is freely available at http://www.rnanut.net/ lncompare/
# Materials and methods
## Compilation of lncrna features
We collected and processed various lncRNA datasets from public databases and in-house computational analysis. Altogether these comprise 109 gene/transcript attributes for the GENCODE v24 human lncRNAs annotation (15 941 genes). These features can be classified into six main classes (see [fig_ref] Table 1: Classification and summary of lncRNA features included in LnCompare database [/fig_ref] :
All data has been compiled at the level of lncRNA genes, not transcripts. For certain features, we utilized an exonic projection of all annotated transcripts from each gene, and estimated the corresponding feature accordingly--for example, GC content or phastCons overlap. Detailed information on every feature and its source are available in Supplementary [fig_ref] Table 1: Classification and summary of lncRNA features included in LnCompare database [/fig_ref] , and provided as additional information at the web server. A comprehensive list of features may be found in [fig_ref] Table 1: Classification and summary of lncRNA features included in LnCompare database [/fig_ref] , and the entire table of lncRNA features for the GENCODE v24 annotation can be downloaded in the 'Download' tab in LnCompare.
# Statistical analysis
In Gene Set Feature Comparison mode, LnCompare considers both quantitative and categorical features [fig_ref] Figure 1: Gene Set Feature Comparison module workflow [/fig_ref]. By default, the background set is defined as the entire GEN-CODE annotation. The Wilcoxon test is used to compare quantitative features between the input gene set and the background set [fig_ref] Figure 1: Gene Set Feature Comparison module workflow [/fig_ref]. For categorical features, the hypergeometric test is applied, and the detailed formula was described in our previous work (29) [fig_ref] Figure 1: Gene Set Feature Comparison module workflow [/fig_ref]. By default, the quantitative features are sorted by the absolute logarithm ratio of the average feature values between the input versus background, in order to highlight the features where the input lncRNA list and the background show the most prominent divergence. Similarly, the odds ratio is used to rank most enriched categorical features.
In Similar Gene Discovery, LnCompare performs similarity calculation between two lncRNAs based on their features [fig_ref] Figure 2: Similar Gene Discovery module workflow [/fig_ref]. After comparing several methods (e.g. Pearson correlation, Euclidian distance etc.), the cosine similarity was finally adopted:
[formula] Similarity = cos (θ ) = A · B A B [/formula]
where A and B are the feature vectors of two LncRNAs, with N/A values dropped. By definition, it is the point multiplication of A and B, divided by the product of the norm of the two vectors. To enable a flexible similarity calculation, users can either use all features, or else a defined subset of features [fig_ref] Table 1: Classification and summary of lncRNA features included in LnCompare database [/fig_ref]. Higher cosine similarity indicates greater similarity. In addition, for more robust results, we employed mutual rank, which has been successfully applied to establish gene co-expression networks (30):
[formula] Rank = Rank(Simiarity) A in B · Rank(Similarity) B in A [/formula]
Where the Rank(Similarity) A in B denotes the rank of similarity between A and B among the similarities of B to all other lncRNAs, and Rank(Similarity) B in A is defined in similar fashion. Lower mutual rank values indicate greater similarity.
## Server implementation
The web server is built on a Linux server using the Apache+MySQL+PHP framework. All graphical visual-
## Description of web server
LnCompare web server performs gene set or single gene comparisons of lncRNAs based on diverse features. It is based on the GENCODE version 24 human annotation [bib_ref] GENCODE reference annotation for the human and mouse genomes, Frankish [/bib_ref] , and therefore only genes with Ensembl 'ENSG. . . ' identifiers belonging to this annotation are assessed. In cases where a supported non-GENCODE identifiers are provided, LnCompare will attempt to map it to GEN-CODE. Supported identifiers comprise Gene symbols, Ref-Seq IDs and Ensembl transcript IDs, whose mappings to GENCODE are based on the Ensembl ID mapping file. When successfully mapped, this gene will be included in subsequent analyses. When not successful, unrecognized IDs, including out-of-date Ensembl entries, are ignored in analyses. The number of successfully found IDs is reported in the results page. For all analyses, users can populate forms with three different sets of example data using buttons. These are: the 'Simple Example' list of six randomly-selected lncRNAs; the 'CLC' list of 122 cancer-related lncRNAs (32); the 'Cell Cycle Example' of 117 lncRNAs that are differentially expressed between G1S and G2M cell-cycle stages in HeLa cells [bib_ref] Cyclin-dependent kinase 1 (CDK1) and CDK2 have opposing roles in regulating interactions..., Murthy [/bib_ref].
LnCompare has two modules, described below. A complete tutorial for both modules can be found in the 'Help' tab.
## Gene set feature comparison
Gene Set Feature Comparison module aims to identify features that characterize a user-provided gene set. This input gene set is compared to a defined background gene set, across each feature [fig_ref] Figure 1: Gene Set Feature Comparison module workflow [/fig_ref]. By default, background is the entire GENCODE annotation, although the user can provide alternative background sets.
The feature-comparison analysis runs differently for quantitative and categorical features. For quantitative features, statistical significance is assessed by Wilcoxon test, while for categorical features the hypergeometric test is used [fig_ref] Figure 1: Gene Set Feature Comparison module workflow [/fig_ref]. Results are displayed separately.
For quantitative features, LnCompare returns a summary plot with logarithm ratios of the mean feature values for the input and the background sets. For categorical features, equivalent plots display the odds ratio [fig_ref] Figure 1: Gene Set Feature Comparison module workflow [/fig_ref] and C). In addition, corresponding P-values, Benjamini-Hochberg false discovery rates (FDR) (34) and a link to a boxplot (or barplot for categorical features) can be found in tabular format (for an example see [fig_ref] Figure 3: Results from Gene Set Feature Comparison analysis of CLC [/fig_ref]. Additional information on each feature can be accessed from the '?' button in the table. For both quantitative and categorical features, the user can apply several different cutoffs to the data displayed: the top ten features, ranked by mean ratio/odds ratio (for quantitative and categorical features, respectively), features with P < 0.05, features with FDR < 0.05, or all possible comparisons). Graphical and tabular results can be directly downloaded from the website.
## Similar gene discovery
LnCompare offers two approaches to compare similarity of lncRNA genes using cosine similarity method: (i) 1-to-N comparison: computes the similarity of one user-provided lncRNA to all remaining GENCODE lncRNAs; (ii) M-to-N comparison: computes the similarity of every lncRNA from list M to every one in list N [fig_ref] Figure 2: Similar Gene Discovery module workflow [/fig_ref]. The user must provide two lists of Ensembl gene IDs (up to a maximum of 100 in each). After specifying the type of analysis desired (1-to-N or M-to-N) and entering the gene IDs, the user can choose which subsets of lncRNA features to be used for similarity analysis (from the feature classes described above (see 'Materials and Methods' section).
For both types of similarity analysis, graphical and tabular outputs display the top 10, 20 or 50 cosine scores (specified by the user) together with the corresponding gene IDs. The table also contains the relative rank the score represents among the partners for lncRNA1 and lncRNA2 lists, respectively (i.e. first number indicates the ranking number for that pair among all possible partners of lncRNA from lncRNA1 list, and the second number indicates the same for lncRNA from lncRNA2 list).
When all the feature classes are selected, LnCompare also provides a mutual rank similarity results section. In this case, graphical and tabular formats show how reciprocal the similarity is between the two genes, with a mutual rank score (see 'Materials and Methods' section for more details). Again, the user can select how many output comparisons should be displayed (top 10, 20 or 50 scores). All the tables from this module are available for download.
## Evaluation of web server; a case study us-ing cancer-related genes
We tested the performance of LnCompare using a set of 122 experimentally validated cancer lncRNAs from the Cancer LncRNA Census (CLC) [bib_ref] Unique genomic features and deeply-conserved functions of long non-coding RNAs in the..., Carlevaro-Fita [/bib_ref]. CLC genes are curated based on experimentally validated functional roles in tumorigenesis or cancer-related cellular phenotypes, and hence is a useful positive control set of lncRNA genes. The CLC genes make a good test case, since they are known to be characterized by a range of features such as high expression in tumours, spliced length and evolutionary conservation [bib_ref] Unique genomic features and deeply-conserved functions of long non-coding RNAs in the..., Carlevaro-Fita [/bib_ref]. Assessing CLC we want to represent two possible scenarios: (i) the user has no prior knowledge of a gene set, and wishes to assess their potential functionality; (ii) The user is aware that this is a set of important lncRNAs, and wishes to study their particular features.
Running the Gene Set Feature Comparison module, we searched for specific features of CLC genes compared to background (all other lncRNAs). Using a cutoff of FDR < 0.05, we observe several quantitative and qualitative traits to be significantly enriched in CLC lncRNAs [fig_ref] Figure 3: Results from Gene Set Feature Comparison analysis of CLC [/fig_ref]. These include high average expression across numerous human cell lines and tissues. CLC genes, on average, also show high exon and promoter conservation across mammals and vertebrates [fig_ref] Figure 3: Results from Gene Set Feature Comparison analysis of CLC [/fig_ref]. Moreover, the CLC set is significantly enriched with lncRNAs from functional and disease databases [fig_ref] Figure 3: Results from Gene Set Feature Comparison analysis of CLC [/fig_ref]. Together, these attributes are consistent with the input gene set being enriched with bona fide functional lncRNAs [bib_ref] Unique genomic features and deeply-conserved functions of long non-coding RNAs in the..., Carlevaro-Fita [/bib_ref] , and points to features (e.g. high expression) that are shared by both cancer-related lncRNAs and PCGs [bib_ref] Discovery of cancer driver long noncoding RNAs across 1112 Tumour Genomes: new..., Lanzós [/bib_ref] [bib_ref] Structural and functional properties of genes involved in human cancer, Furney [/bib_ref].
Interestingly, the Gene Set Feature Comparison also reports CLC genes to be on average closer in genomic distance to PCGs, and significantly more likely to be divergently transcribed from protein coding genes [fig_ref] Figure 3: Results from Gene Set Feature Comparison analysis of CLC [/fig_ref]. This may reflect a common molecular mechanism among CLC lncRNAs to be further studied. In contrast, it may result from a bias in literature to focus on lncRNAs that lie close to PCGs. Such ascertainment biases are an important confounding factor that should be borne in mind when interpreting these results. At last, we also observe that the W528 Nucleic Acids Research, 2019, Vol. 47, Web Server issue CLC set tends to be less tissue specific, more cytoplasmic (their nuclear/cytoplasmic ratio is significantly lower) [fig_ref] Figure 3: Results from Gene Set Feature Comparison analysis of CLC [/fig_ref] , and enriched with hits from proliferation CRISPR screens [fig_ref] Figure 3: Results from Gene Set Feature Comparison analysis of CLC [/fig_ref].
Similar Gene Discovery 'M-to-N' functionality may complement the above analysis to interrogate unknown lists of genes and help to prioritize candidates based on similarity to known lncRNAs. For example, to select potential cancerassociated lncRNAs for experimental validation from a list of novel candidates, one can assess their similarity to CLC genes.
In addition, Similar Gene Discovery '1-to-N' functionality makes it possible to search for the most similar genes to a given lncRNA, in order to discover new, functionallyrelated lncRNAs. For example, in searching for lncR-NAs similar to the X-inactive specific transcript, XIST [bib_ref] The human XIST gene: Analysis of a 17 kb inactive X-specific RNA..., Brown [/bib_ref] , LnCompare reports the maternally-expressed gene 8 (MEG8) to be the most similar (cosine similarity 0.9). Interestingly, both genes are associated with imprinting, are expressed during early development and have nuclearrestricted localization [bib_ref] Identification of a new imprinted gene, Rian, on mouse chromosome 12 by..., Hatada [/bib_ref] [bib_ref] Human-ovine comparative sequencing of a 250-kb imprinted domain encompassing the callipyge (clpg)..., Charlier [/bib_ref] [bib_ref] Genomic imprinting of Xist by maternal H3K27me3, Inoue [/bib_ref]. Moreover, MEG8 has been reported to interact with chromatin-binding proteins and repressor complexes [bib_ref] MEG8 long noncoding RNA contributes to epigenetic progression of the epithelial-mesenchymal transition..., Terashima [/bib_ref] [bib_ref] ) lincRNAs act in the circuitry controlling pluripotency and differentiation, Guttman [/bib_ref]. An important caveat is the fact that some lncRNAs are present in the functional database lncRNAdb, and this will influence the similarity analysis results without necessarily representing biological similarity. In order to eliminate this possible confounder, we removed the phenotype-associated feature class and repeated the analysis. This analysis now identifies ENSG00000272872 as the most similar to XIST. This lncRNA has been linked to various cancers in the lncRNADisease v2.0 database (43), making it an interesting candidate to study.
# Discussion
In recent years there has been a dramatic acceleration in the volumes of lncRNA gene candidates emerging from genomic studies. However, we remain broadly ignorant about molecular mechanisms and biological roles of these genes. Classical methods to describe newly discovered genes or to assess gene sets are inefficient for lncRNAs. This has created a need for tools to study the properties of lncRNA sets, in order to formulate new hypotheses from or gauge the success of high-throughput experiments.
To meet this need, we have curated a comprehensive feature set covering diverge quantitative and categorical aspects of lncRNAs from the GENCODE annotation. Using these features, LnCompare searches for those that are significantly over-or under-represented in an input set compared to background. In a set of lncRNAs with known roles in cancer [bib_ref] Discovery of cancer driver long noncoding RNAs across 1112 Tumour Genomes: new..., Lanzós [/bib_ref] , LnCompare identifies a range of characteristic features, including elevated expression and evolutionary conservation. Moreover, it also identifies other attributes including higher cytoplasmic localization and ubiquitous expression. These features may guide researchers to focus on potential molecular activities related to cytoplasmic processes, in contrast to most studies that concentrate on lncRNAs' roles in chromatin regulation [bib_ref] Structure and function of long noncoding RNAs in epigenetic regulation, Mercer [/bib_ref] [bib_ref] Altered nuclear retention of mRNAs containing inverted repeats in human embryonic stem..., Chen [/bib_ref] [bib_ref] Many human large intergenic noncoding RNAs associate with chromatin-modifying complexes and affect..., Guttman [/bib_ref]. Moreover, LnCompare assesses similarity between genes, which can be a powerful strategy to identify new genes playing similar roles to known examples. Conversely, such similarity analysis could be used to predict the roles of novel lncRNAs by similarity to known genes. We anticipate that LnCompare will be useful to the many colleagues who presently study lncRNAs at the global scale and wish to extract more biological insights from their data.
## Supplementary data
Supplementary Data are available at NAR Online.
[fig] Figure 1: Gene Set Feature Comparison module workflow. (A) LnCompare is based on a dataset of quantitative and categorical features of lncRNAs. For a given analysis, LnCompare divides input lncRNAs (yellow) and background lncRNAs (green). (B) For each quantitative feature, the distribution of input and background lists are compared by Wilcoxon test. Below are shown the results, displayed in the web server as a barplot. Each bar represents a feature, and length in the x-axis denotes the ratio of the mean input value and the mean background value. (C) For every categorical feature, LnCompare performs a hypergeometric test (upper panel), which is summarized in a bubble plot (lower panel). Features are distributed across the y-axis, while the x-axis displays the odds ratio obtained from the hypergeometric test. Circle radius reflects the P-value from the hypergeometric test. izations are enabled by the open source G2 package (https: //antv.alipay.com/), whilst the display and download of tabular results was established with the JavaScript plugin vis (http://visjs.org/). [/fig]
[fig] Figure 2: Similar Gene Discovery module workflow. The module compares either one gene or several input genes versus a list of background genes ('1-to-N' and 'M-to-N', respectively). The yellow box represents the input genes; the green box represents the background genes. Similarity comparisons are based on the distance between the two gene feature vectors, calculated by two different methods: cosine similarity and mutual rank. [/fig]
[fig] Figure 3: Results from Gene Set Feature Comparison analysis of CLC (A) Graphical results displaying features that are significantly different between CLC and background lncRNAs (FDR < 0.05). Feature labels are shown on the y-axis and grey boxes on the left summarize their content. The x-axis indicates the ratio between the mean of CLC genes (input) and background genes for each feature. (B) Table obtained from the same analysis for categorical features. 'Feature' and 'Name' indicate and describe the feature tested, 'List' and 'background' show the number of CLC and background genes associated with the feature, respectively. 'P-value', 'FDR' and 'Odds Ratio' from hypergeometric test are also shown in the table. [/fig]
[table] Table 1: Classification and summary of lncRNA features included in LnCompare database [/table]
|
Response to article by Matthew Wasserman et al. (2018): “Modeling the sustained use of the 13-valent pneumococcal conjugate vaccine compared to switching to the 10-valent vaccine in Mexico”
In a recent article, Wasserman et al. estimated and forecasted the health and economic impact of switching from the 13-valent (PCV-13) to the 10-valent (PHiD-CV) pneumococcal conjugate vaccine in Mexico's national immunization program. In this response letter, we highlight various methodological inconsistencies and model input considerations that potentially bias the results and further recommendations made by the authors.ARTICLE HISTORY
## Dear editor,
In a recent article, Wasserman et al. estimated and forecasted the health and economic impact of switching from the 13-valent (PCV-13) to the 10-valent (PHiD-CV) pneumococcal conjugate vaccine in Mexico's national immunization program. [bib_ref] Modeling the sustained use of the 13-valent pneumococcal conjugate vaccine compared to..., Wasserman [/bib_ref] In this response letter, we highlight various methodological inconsistencies and model input considerations that potentially bias the results and further recommendations made by the authors.
The model used by Wasserman et al., previously applied in the context of Canada, 2 forecasted future pneumococcal disease trends based on historical serotype behaviors for each PCV-13 serotype and non-vaccine serotypes. Throughout the manuscript, there appears to be no recognition regarding the time-period PHiD-CV was used in the Mexican Social Security Institute (IMSS), as it only mentions its registration and availability through the private market. They retrospectively analyzed the cost-effectiveness of pneumococcal conjugated vaccine (PCV) introduction in Mexico and described the existence of PCV-7 and PCV-13 between 2006 and 2014. Their one-sided and highly limited focus narrowly addresses the sequence from PCV-7 to PCV-13 and they improperly assume that all the health benefits observed in this time period are associated with those vaccines. Noticeably absent is any attribution or acknowledgement that PHiD-CV was used in the IMSS between January 2010 and December 2011 with approximately 2.2 million doses distributed.
Wasserman et al. used a simple liner and logistic regression model to simulate and forecast the complex behavior of pneumococcal serotypes prior to and post vaccine introduction in Mexico. To effectuate a reliable and valid health economic analysis, there are rubrics regarding the biological process that should be enclosed in the models to reflect how infections, demographic mortality, protection against infection, costs and use of healthcare resources occur over time. [bib_ref] Modeling good research practices overview: a report of the ISPOR-SMDM modeling good..., Caro [/bib_ref] In the analytical framework used by Wasserman et al., comparisons between the model's estimations and past data observations are not clearly presented due to limited explanation of how well trend regressions could predict historical data (see in Wasserman et al.). [bib_ref] Modeling the sustained use of the 13-valent pneumococcal conjugate vaccine compared to..., Wasserman [/bib_ref] It is important to emphasize that investigators affiliated to recognized public health entities such as the Pan-American Health Organization (PAHO), the International Vaccine Access Center (IVAC), and the World Health Organization (WHO) have each conducted independent systematics reviews which concluded that there was no superiority of one vaccine over the other. 4-6 Therefore, the conclusion of the analysis of Wasserman et al. is inconsistent with the evidence already generated. Additionally, PAHO and WHO support the interchangeability of these vaccines in certain cases of programmatic and resources issues.While the results of Wasserman et al. suggest that continued use of PCV-13 instead of switching to PHiD-CV would likely save 34 Billion MXN over the next 10 years, the analysis appears significantly biased considering an invasive pneumococcal disease (IPD) serotype-specific approach. The evidence of cross protection against serotype 19A provided by PHiD-CV and the heterogeneity of effectiveness/impact results on serotype 3 observed with PCV-13 appear not to have been considered, 7-9 misinterpreting the outcomes and conclusions. It is important to note the serotype content of PCVs may not automatically translate into disease protection against these serotypes and the absence of a certain serotype will not automatically translate into the absence of an effect due to cross protection. [bib_ref] Effectiveness of three pneumococcal conjugate vaccines to prevent invasive pneumococcal disease in..., Deceuninck [/bib_ref] [bib_ref] Effectiveness of ten-valent pneumococcal conjugate vaccine against invasive pneumococcal disease in Brazil:..., Domingues [/bib_ref] [bib_ref] Impact of ten-valent pneumococcal conjugate vaccination on invasive pneumococcal disease in Finnish..., Jokinen [/bib_ref] The serotype-specific regression analysis is designed to forecast future disease trends based on historical serotype behaviors. Due to the uncertainty surrounding serotype replacement and to avoid under-estimation or over-estimation of vaccine impact, tests on several scenarios were conducted by varying trend lines based on historical surveillance data from the UK and USA to reflect PCV-13 infant vaccination. Nevertheless, it is not clear how the prevalence of serotype 19A was considered under the Mexican scenario after PCV-13 introduction. According to SIREVA Laboratory Surveillance Network and National Institute of Public Health (INSP) data,serotype 19A prevalence tends to reduce after PCV-13 use, yet it still circulates following a secular pattern and it remains the most important invasive serotype in the Mexican setting.
Finally, there are certain weaknesses with the data inputs and assumptions made for clinical outcomes. First, the model presented by Wasserman et al. relies upon the historic demonstration of serotype-specific vaccine effect against IPD. Surprisingly, the authors fail to consider the evidence of PCVs efficacy against pneumonia and acute otitis media (AOM). Contrarily, the authors improperly estimate the prospective change in pneumonia and AOM cases based on the forecasted change for IPD cases.
We recognize the efforts of the authors to develop tools that evaluate the epidemiological scenarios and management costs for pneumococcal diseases, but these models should incorporate all available evidence showing the impact of PCV on overall pneumococcal disease instead of focusing solely on an arbitrary selection of evidence and assumptions. Vaccine impact estimations and costeffectiveness data are crucial to inform policy-makers on PCV use. Considering all these general concerns, the Wasserman et al. study should be interpreted with caution.
AcknowledgmentsThe authors would like to thank the Business & Decision Life Sciences platform for editorial assistance, writing assistance and publication coordination, on behalf of GSK. Pierre-Paul Prévot coordinated the letter development.Disclosure of potential conflicts of interest |
A systematic review of biomarkers for disease progression in Parkinson’s disease
Background: Using surrogate biomarkers for disease progression as endpoints in neuroprotective clinical trials may help differentiate symptomatic effects of potential neuroprotective agents from true disease-modifying effects. A systematic review was undertaken to determine what biomarkers for disease progression in Parkinson's disease (PD) exist. Methods: MEDLINE and EMBASE (1950-2010) were searched using five search strategies. Abstracts were assessed to identify papers meriting review in full. Studies of participants with idiopathic PD diagnosed by formal criteria or clearly described clinical means were included. We made no restriction on age, disease duration, drug treatment, or study design. We included studies which attempted to draw associations between any tests used to investigate disease progression and any clinical measures of disease progression. The electronic search was validated by hand-searching the two journals from which most included articles came. Results: 183 studies were included: 163 (89%) cross-sectional, 20 (11%) longitudinal. The electronic search strategy had a sensitivity of 71.4% (95% CI 51.1-86.0) and a specificity of 97.1% (95% CI 96.5-97.7). In longitudinal studies median follow-up was 2.0 years (IQR 1.1-3.5). Included studies were generally poor quality -cross-sectional with small numbers of participants, applying excessive inclusion/exclusion criteria, with flawed methodologies and simplistic statistical analyses. Conclusion: We found insufficient evidence to recommend the use of any biomarker for disease progression in PD clinical trials, which may simply reflect the poor quality of research in this area. We therefore present a provisional 'roadmap' for conducting future disease progression biomarker studies, and recommend new quality criteria by which future studies may be judged.
# Background
Parkinson's disease (PD) is the second most common progressive neurodegenerative disorder, with an annual incidence of around 16-19 cases per 100,000 [bib_ref] Systematic review of incidence studies of Parkinson's disease, Twelves [/bib_ref]. As drug therapy in PD is currently symptomatic in nature, a key aim of PD research is the development of drugs which slow or even halt neurodegeneration and, therefore, clinical progression. However, clinical trials in neurodegenerative disorders have struggled to separate out symptomatic effects of potential therapeutic agents (e.g. due to increased striatal dopamine) from true diseasemodifying effects. In PD, it is currently not possible to directly measure the number of remaining dopaminergic neurons in vivo and, therefore, alternative approaches are required. Clinical assessments in PD using scales to measure motor impairment, disability, quality of life, or disease stage scales are affected by symptomatic effects of therapy and are unable to differentiate this effect from diseasemodification, at least in the short-term.
Various clinical trial designs have been developed to try to adjust for symptomatic effects of putative neuroprotective agents and, therefore, allow clinical rating scales to be used as endpoints. These include long-term follow up studies of placebo-treated and active-agent treated patients looking for ongoing divergence, measuring outcomes following a wash-out period [bib_ref] A "cure" for Parkinson's disease: can neuroprotection be proven with current trial..., Clarke [/bib_ref] [bib_ref] Disease-modification" trials in Parkinson disease: target populations, endpoints and study design, Rascol [/bib_ref] , and delayed start trial design [bib_ref] Are delayed-start design trials to show neuroprotection in Parkinson's disease fundamentally flawed?, Clarke [/bib_ref]. However, limitations and flaws in these trial designs have as yet restricted their use. An alternative approach, the focus of much primary research, is the use of a surrogate outcome biomarker as an endpoint in neuroprotective clinical trials.
Surrogate outcome biomarkers are objectively measured characteristics of a disease, which act as indicators of the underlying pathogenic process responsible for disease progression, including the change in that process following a therapeutic intervention. To allow their use in clinical trials surrogate outcome biomarkers must have a strong association with a clinical endpoint or outcome known to measure the effect of a therapeutic intervention on disease progression, for which the biomarker can act as a substitute. Surrogate biomarkers for disease progression in PD could shorten the duration of phase III trials and thereby reduce the cost and time required to get a drug to market. Unfortunately at present there is not a single accepted surrogate outcome biomarker for any neurodegenerative disorder. outlines the ideal characteristics of a surrogate biomarker for disease progression in PD.
Much has been written about what features biomarkers for disease progression in Parkinson's disease should possess [bib_ref] Assessment of neuroimaging techniques as biomarkers of the progression of Parkinson's disease, Brooks [/bib_ref]. The ideal surrogate biomarker should:
(1) Change with neurodegeneration (i.e. degeneration of the nigrostriatal dopaminergic system); (2) Show an association with the clinical phenotype arising secondary to this degenerative process; (3) Have a direct association with disease progression, without intermediate variables; [bib_ref] Are delayed-start design trials to show neuroprotection in Parkinson's disease fundamentally flawed?, Clarke [/bib_ref] Have robust longitudinal data linking it to disease progression; (5) Not be influenced by symptomatic treatment, but only by a true change in the neurodegenerative process; (6) Predict long-term changes in disease progression by short-term changes in the biomarker;
(7) Be generalisable to people with differing characteristics (e.g. age, gender, race); (8) Be continually variable (ideally linearly for simplicity); (9) Be sensitive, reflecting small changes in disease progression; (10) Be quick and cheap to measure, and amenable to blinded assessment; [bib_ref] Criteria for diagnosing Parkinson's disease, Calne [/bib_ref] Be suitable for measurement reliably across different centres;Be suitable for repeated measurement in the same patient; (13) Be safe and tolerable to the patient.
PD is a complex neurodegenerative disorder in which many different pathophysiological processes have been identified in the nigrostriatal pathways and beyond, such as protein aggregation, oxidative damage, mitochondrial dysfunction, lysosomal dysfunction and inflammation. It is, therefore, not surprising that many different candidate biomarkers for disease progression in PD have been studied. However, the literature in this area has never been brought together systematically. We, therefore, aimed to undertake a systematic review to assess what potential surrogate biomarkers for disease progression (motor and non-motor) in PD exist, whether any meet the criteria for use in clinical trials, and if not which looks most promising. We did not aim to review the literature for diagnostic biomarkers (i.e. those designed to aid early diagnosis in the pre-motor or motor phase) or prognostic biomarkers (i.e. those aimed at identifying patients who progress at different rates). In addition, we aimed to critique data from identified disease progression biomarker studies relating to study design, participant characteristics, and statistical analyses undertaken, in order to produce guidelines for future studies. Study quality questionnaire based on the assay methods and study design sections of the REMARK reporting recommendations for prognostic tumour markers Question Yes = 1 No = 0
(1) Was the study prospective?
The study reports that patients and the performed test result were collected before the development of an outcome.
No report or clearly retrospective (e.g. patients with poor prognosis collected before biomarker measurement).
(2) Was evaluation of prognostic marker blinded to patient outcome?
The study reports an attempt to blind the person measuring the biomarker to patient outcome.
There is no such report, or assessor clearly not blinded.
(3) Was there a defined time period during which patients were enrolled?
Study defines time period, end of follow-up period, and median follow-up time.
Does not define criteria.
(4) Were there precisely defined clinical outcomes at the start of the study?
Study defines which clinical end points are to be measured.
No such definition.
(5) Were the methods for measuring the prognostic marker adequately described and referenced?
Clearly described and referenced Not clearly described and referenced (6) Cases unselected and unbiased? No attempt to select patients with exclusion criteria Only a subset of patients enter the study
# Methods
Following the development of a review protocol equivalent to the methodology described below, literature searches were conducted in the databases MEDLINE (1950 to August 2010) and Embase (1980 to August 2010), using the OVID search interface. Five separate search strategies, developed by an experienced information scientist, were run in each database. The first four were based on free-text words identified through background reading of relevant review articles. These searches included potential (1) blood, (2) urine or cerebrospinal fluid (CSF), (3) imaging and (4) neurophysiological biomarkers. A fifth search using generic terms for biomarkers based on index headings was also run in both databases. For details of the search strategy please see Additional file 1. The searches were limited to human studies. Only English language articles were included, due to lack of resources for translation. Reference lists of included articles and relevant review articles were checked to identify any studies which the electronic search may have missed.
## Validation of the electronic search strategy
The electronic search strategy was validated by hand searching five years of the two journals from which most of the included articles came: Movement Disorders and Journal of the Neurological Sciences . The number of relevant and irrelevant articles identified by hand searching and by the electronic search, was used to calculate the sensitivity and specificity for the electronic search strategy.
## Study selection
A single reviewer examined abstracts retrieved by the electronic search to identify articles meriting review in full. Full length articles were then reviewed before data were extracted from relevant papers. In both stages the inclusion and exclusion criteria detailed below were applied.
Only studies of participants with idiopathic Parkinson's disease diagnosed by formal criteria [bib_ref] Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of..., Hughes [/bib_ref] [bib_ref] Diagnostic criteria for Parkinson disease, Gelb [/bib_ref] [bib_ref] Criteria for diagnosing Parkinson's disease, Calne [/bib_ref] , or clearly described clinical means (the presence of at least two out of four of the cardinal clinical signs of PD and an attempt to exclude atypical syndromes), were included. No restriction was made on the grounds of participant's age, disease duration, or drug treatment.
Studies which investigated the efficacy of using a biomarker, including (but not restricted to) imaging, blood tests, tests of CSF and neurophysiological tests, to investigate disease progression in Parkinson's disease were included. To qualify for inclusion there must have been an attempt to examine for an association between the biomarker and a clinical measure of disease progression. Acceptable measures included measures of motor or cognitive impairment, disability, handicap, quality of life, and duration of survival.
Only studies exploring associations between a biomarker and the total score from a clinical rating scale, rather than its subsections, were included. The subsections of most clinical measures would never be acceptable outcome measures for neuroprotective trials and, therefore, developing surrogate biomarkers for these was felt not to be relevant. The exception was the Unified Parkinson's disease rating scale (UPDRS) [bib_ref] UPDRS Development Committee: The Unified Parkinson's Disease Rating Scale, Fahn [/bib_ref] , where studies examining for relationships to its main constituent parts were included. Studies only investigated the relationship between a biomarker and individual symptoms (e.g. bradykinesia or rigidity), or olfactory function were excluded for similar reasons.
Only studies examining for associations between putative biomarkers and global measures of cognition (e.g. Minimental state examination (MMSE) [bib_ref] Mini-mental state". A practical method for grading the cognitive state of patients..., Folstein [/bib_ref] , Cambridge cognitive examination (CAMCOG [bib_ref] A standardised instrument for the diagnosis of mental disorder in the elderly..., Roth [/bib_ref] ), rather than individual neuropsychological tests, were included. It is unlikely that improvement in a single neuropsychological test would be a suitable outcome measure for a neuroprotective trial. As depression in PD is not linked to overall disease progression [bib_ref] Relationship of motor symptoms, intellectual impairment, and depression in Parkinson's disease, Huber [/bib_ref] and may be commoner at the time of diagnosis [bib_ref] Depression in Parkinson's disease: its prevalence, diagnosis, and neurochemical background, Yamamoto [/bib_ref] , studies only investigating the relationship between a biomarker and depression were excluded.
Studies examining the relationship between a biomarker and treatment status, the presence or severity of complications related to therapy, or duration of illness were excluded. Studies of static predictive biomarkers (e.g. genetic markers) which try to anticipate the future rate of disease progression were excluded. A useful biomarker for clinical trials needs to be dynamic -changing with disease progression. Therefore, these studies were not relevant.
As we aimed to produce a comprehensive review and detect any evidence of the utility of a putative biomarker, we set no study quality threshold. We, therefore, included small cross-sectional studies, in which an association between a biomarker and clinical measures of disease progression were analysed at a single time point across groups of patients with different disease severities.
## Data extraction
Study methods and results were extracted by a single reviewer, and for accuracy this was performed twice. Data were extracted using a data extraction sheet (Additional file 2) relating to the following: (1) study design including restrictiveness of criteria for entry into the study;
(2) setting; (3) study population, including number of participants, gender ratio, disease duration at baseline, baseline measures of disease severity and baseline treatment status; (4) specific biomarkers investigated; (5) statistical analyses performed; (6) results of statistical analyses of the associations between the biomarkers and clinical measures of disease severity and how completely these results were reported; (7) analysis of the effect of drug treatment on the biomarker; (8) economic analysis of using the biomarker; (9) measures of suitability and acceptability of the test to patients.
The restrictiveness of the inclusion and exclusion criteria applied in each study was graded as: none, explicit statement that only criteria to exclude atypical parkinsonian syndromes were applied; mild, ≤ 3 criteria applied (except those described under moderate); moderate, 4-5 criteria applied or evidence of an attempt to limit by age, gender, cognitive state, disease stage, drug therapy for PD (e.g. all de-novo); severe, ≥ 6 criteria applied; not detailed, no mention of whether criteria were applied.
# Methodological quality
No validated tool to measure the quality of studies investigating surrogate biomarkers as outcome measures exists. An attempt was, therefore, made to assess study quality using an adapted quality questionnaire, illustrated in , based on the assay methods and study design sections of the Reporting recommendations for tumor MARKer prognostic studies (REMARK) reporting recommendations for prognostic tumour markers . This measure of study quality was also used to assess whether there was any bias in terms of the quality of studies included.
## Data synthesis
Given the likelihood that included studies would examine the relationship of multiple different putative biomarkers with multiple different clinical measures of disease severity, we were aware that any data synthesis would be qualitative in nature.
# Results
As shown in [fig_ref] Figure 1: Flow diagram outlying the selection procedure to identify 183 articles included in... [/fig_ref] , the electronic search identified 4080 records. After removing duplicates, 2435 unique records identified by the electronic search were screened, in addition to a further 66 records identified whilst performing the hand search or on reviewing reference lists of relevant review articles and included articles. The full-text articles of 409 records were then assessed for eligibility, and of those 226 articles were excluded. Finally data were extracted from a total of 183 articles.
## Hand searching
Hand searching to validate the electronic search strategy revealed a sensitivity of 71.4% (95% CI 51.1-86.0) and a specificity of 97.1% (95% CI 96.5-97.7).
## Characteristics of included articles
The majority of included articles (n=156, 85%) had a cross-sectional study design. The remaining 27 articles had a longitudinal design, but only 20 (11%) of these attempted to examine for relevant longitudinal relationships between a change in a biomarker and a change in a clinical measure of disease severity. The other seven (4%) simply detailed cross-sectional analyses, and results from these have been included with data from crosssectional studies.
The UK Parkinson's Disease Society (UKPDS) Brain Bank Criteria [bib_ref] Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of..., Hughes [/bib_ref] was most commonly used to make the diagnosis of PD [fig_ref] Table 2: Study characteristics of the 183 included articles [/fig_ref] , although a large proportion of studies simply used the presence of the cardinal clinical features of PD. Over half of included articles did not describe the setting of their study, but the majority of those who did were based in outpatient departments. Similarly, a third of studies failed to mention whether inclusion and exclusion criteria were applied. Of those providing this information over half applied moderately to severely restrictive inclusion/exclusion criteria. No significant differences in study characteristics were noted between longitudinal and cross-sectional studies.
All of the included studies used an impairment or disability scale as the clinical measure of disease progression used to test for an association with a biomarker. None of the studies used duration of survival, or measures of quality of life or handicap as a clinical outcome measure.
## Characteristics of study participants
As illustrated in [fig_ref] Table 3: Baseline characteristics of study participants in the 183 included articlesMeans are presented... [/fig_ref] the median number of study participants was low at 32 (interquartile range (IQR) 21 to 53). With regards to representativeness of all those with PD, those included were fairly young, with a mean age of 63.5 (standard deviation (SD) 5.6) years of age, particularly considering that the median duration of disease at study entry was 5.7 (IQR 3.6 to 7.4) years. In keeping with this fairly long duration of disease most included patients were already on treatment at baseline, and had moderate disease severity scores. Cognition was only mildly impaired, with many studies excluding patients with dementia.
Patients included in longitudinal studies had a significantly shorter disease duration at baseline (P = 0.005) and were less likely to be on treatment (P = 0.005) than those included in cross-sectional studies. In keeping with this, they also had milder disease at baseline, with significantly better UPDRS (II) scores and lower Hoehn and Yahr stages than those in cross-sectional studies.
## Quality criteria
The median total score produced by applying the adapted quality questionnaire to each of the included studies was 4.0 (IQR 3.0 to 4.0) out of a possible six. There was no significant difference in the total score achieved by longitudinal (median 4.0 (IQR 3.0 to 4.8)) and cross-sectional (median 4.0 (IQR 3.0 to 4.0) studies (P = 0.125, Mann Whitney test). Unfortunately three of the six questions (questions one, four and five) had almost no discriminating value, being rated in the affirmative for almost every included article.
## Types of biomarkers investigated
A broad spectrum of different biomarker modalities were investigated in the studies included in this systematic review [fig_ref] Table 4: Types of putative biomarkers for disease progression investigated in the 183 included... [/fig_ref]. Brain single-photon emission tomography (SPECT) and positron emission tomography (PET) imaging, along with tests of blood and its constituents, and cerebrospinal fluid formed the majority of biomarkers investigated. Longitudinal studies mainly examined the potential of PET and SPECT imaging as a biomarker of disease progression.
## Longitudinal studies
In the 20 studies which examined for relevant longitudinal associations, the mean number of longitudinal time points (including the baseline time point) was 2.2 (SD 0.4), with a median time period from baseline to the final time point of 2.0 (IQR 1.1 to 3.5) years.
Additional file 3 details the biomarkers examined in the 20 longitudinal studies and their relationship with clinical measures of disease severity. Nine undertook brain SPECT imaging, eight brain PET imaging, one brain MRI, one electrophysiological studies, and one acoustical analysis following a four sentence reading task. Among other parameters the imaging studies examined the relationship between clinical measures (total UPDRS, UPDRS (II), UPDRS (III), Hoehn and Yahr stage, MMSE and CAMCOG) and uptake in the caudate, putamen and striatum. A variety of different PET and SPECT ligands were used. These included ligands which provide a measure of cerebral glucose metabolism ( Full-text articles excluded (n = 226)
No diagnostic criteria stated (119) No relevant associations described (34) Foreign language paper (15) Only associations with individual symptoms or parts of cognition (12) Post-mortem study [bib_ref] Criteria for diagnosing Parkinson's disease, Calne [/bib_ref] Other (35) Full-text articles assessed for eligibility It is clear from the data presented in Additional file 3 that there is no evidence to support the use of acoustic analysis or electrophysiological tests as biomarkers for disease progression in PD. The majority of the imaging studies found no relationship between imaging parameters The P value column relates to association between study type (longitudinal and cross-sectional studies) and 1) Diagnostic criteria, 2) Setting, 3) Inclusion/exclusion criteria using Pearson's chi squared tests. and clinical measures of disease severity, as illustrated in. Those which did report a significant relationship were single studies, examining specific brain regions, with results which have not been replicated by another group. In addition they involved small numbers of participants.
No studies reported an economic analysis of using the biomarker in question, and nor did any report on the acceptability of the test to individual patients. Whilst eight (40%) longitudinal studies did not detail the treatment status of patients at baseline, three studies [bib_ref] Longitudinal change in 99mTcHMPAO cerebral perfusion SPECT in Parkinson's disease over one..., Firbank [/bib_ref] [bib_ref] Changes in network activity with the progression of Parkinson's disease, Huang [/bib_ref] [bib_ref] Wolters EC: [123I]FP-CIT SPECT is a useful method to monitor the rate..., Winogrodzka [/bib_ref] did look at the effect of drug therapy on the biomarker under examination.
In several longitudinal imaging studies it was impossible to tell from the text whether the values derived from specific brain regions which were used to draw associations with clinical measures, were mean values (values from left and right hemispheric structures summed and divided by two) or total values (values from left and right hemispheric structures simply added together). This issue was however far more prolific when examining data from cross-sectional studies.
## Cross-sectional studies
Additional file 4 demonstrates the vast array of putative biomarkers which have been examined in cross-sectional studies, and the large number of clinical measures used to try to draw associations with them. Putative blood and CSF biomarkers examined include structural proteins, neurotransmitters, free-radicals and markers of oxidative stress, amino acids and their metabolites, catecholamines, metals and their transfer proteins, vitamins and their carrier proteins, and cytokines. This heterogeneity made quantitative synthesis of the extracted data impossible. Looking at the data however, no single biomarker investigated stands out amongst the others as a good candidate for further investigation in a longitudinal study.
Similar to the longitudinal studies, in many crosssectional SPECT and PET studies it was impossible to determine whether the uptake values used to draw associations with clinical measures, were average or total values for structures present in both hemispheres. Unfortunately none of the 163 studies treated as crosssectional in this review included an analysis of the acceptability of the test in question to the patients. Regarding the influence of anti-parkinsonian drug therapy on the biomarker examined, 25 cross-sectional studies did examine this issue to some extent.
## Statistics
Correlation analysis, a basic statistical method which can be used to examine for a relationship between two variables, was solely used in 75% (15/20) of longitudinal studies. Only one study [bib_ref] Dopamine transporter brain imaging to assess the effects of pramipexole vs levodopa..., Parkinson Study Group [/bib_ref] used a relevant regression analysis. This study was not designed as a biomarker study, but rather as a drug study to compare the rate of dopamine neuron degeneration after initial treatment with either pramipexole or levodopa in early PD using [ 123 I]β-CIT SPECT. The association between change from baseline in UPDRS score (dependent variable) and the percentage change from baseline in [ 123 I]β-CIT uptake (independent variable) was examined using a multiple regression model that adjusted for initial treatment (pramipexole, levodopa) and baseline UPDRS score. The paper states that the analysis was performed separately for each time point [bib_ref] Wolters EC: [123I]FP-CIT SPECT is a useful method to monitor the rate..., Winogrodzka [/bib_ref] and 46 months) at which patients had a follow-up scan. However, despite stating that this occurred, details of the outcome are not given in the paper; correlations are simply described with a correlation coefficient (r value) and the level of significance (P value). There was no analysis that used data from all time points in a single statistical analysis, rather than analysing the change from baseline to each time point separately.
Similarly the majority of cross-sectional studies simply used basic correlation analysis rather than more complex statistical techniques.
Overall the standard of reporting results from statistical analyses in the included longitudinal and cross-sectional studies was poor. 83.1% of longitudinal correlations (examined by Pearson's or Spearman's correlation analyses), and 58.8% of cross-sectional correlations, described neither a correlation coefficient nor the direction of association. Furthermore in 81.4% of longitudinal correlations and 61.0% of cross-sectional correlations a P value was not given. No study included in this review detailed confidence intervals for a given correlation coefficient.
# Discussion
We found insufficient evidence to support the use of any biomarker to measure motor or non-motor disease progression in PD clinical trials. It has also proven difficult to determine which of the current candidate biomarkers merits selection over the others for further investigation.
Whilst some of the putative imaging biomarkers examined longitudinally (e.g. FDG PET) may look promising given the evidence presented in, unfortunately this is not really the case. In the case of FDG PET two papers examined this biomarker and both found a significant association between the change in a FDG PET measure and a change in the motor UPDRS. However, both studies only included 15 patients, which is a relatively small number of participants from which to draw conclusions. They also both examined different FDG expression patterns and only found that some (four out of six) of these patterns showed a significant association with a change in the motor UPDRS. Given the differences between the studies it was not possible to meta-analyse their results in a meaningful way. As with other imaging modalities we, therefore, felt there was insufficient evidence to single out FDG PET for further investigation.
It is possible that the lack of a current biomarker is because no suitable biomarker exists or, at least, no single biomarker given the complexity of the disease. However, at present, it probably also reflects the very poor quality of studies which have investigated biomarkers for disease progression in PD. In order to improve future studies, we therefore suggest a provisional 'roadmap' for conducting biomarker studies in this area, detailed in . In addition, we recommend new quality criteria supported by evidence from this review, detailed in , by which future studies may be judged.
The starting point for any biomarker study must be a valid reason for selecting the specific biomarker for investigation, based on the pathophysiology of PD. In many studies included in this systematic review the reasons for selecting putative biomarkers for investigation were rather tenuous. Just because a substance can be measured does not mean an attempt should be made to find an association between that substance and clinical measures of disease severity. This approach, as demonstrated by this systematic review, is likely to be fruitless. Moreover, the identification of a biomarker was not the primary aim of many studies included in this review but rather a byproduct of another study. Disease progression biomarker studies need carefully planned specific designs as highlighted by our 'roadmap'.
Secondly, the reliability of a putative biomarker must be established by demonstrating the reproducibility of its measurement in a single centre by different personnel, and between different centres. This can be a particular issue for CSF proteins and for imaging biomarkers, where a change in a small area of the brain is often measured. Before proceeding it must be confirmed that such specific imaging biomarkers can be measured with a reliable degree of consistency across different centres, which may have different imaging equipment and software, otherwise the putative biomarker is likely to be of limited use.
Thirdly, an evaluation of the effect of confounding factors on the biomarker (e.g. age, gender, smoking status or being on symptomatic treatment for PD) should be undertaken. An understanding of the influence of confounding factors on the biomarker will aid sample size calculations, and allow a rigorous analysis of the final study results by adjusting for these factors. This hopefully should avoid erroneous results which reflect a confounding factor rather than a true relationship between a biomarker and a clinical measure of disease progression.
Ideally to be a useful biomarker for neuroprotective clinical trials the biomarker should not be influenced by symptomatic treatment. This would simplify trial design meaning patients on and off treatment could be recruited and would mean those going on to treatment would not have to be censored. However, where symptomatic treatment has a limited effect on a given biomarker then knowledge of this may still allow inclusion of patients on treatment in clinical trials with appropriate adjustment of their results. Nonetheless the measurement of biomarkers which will be significantly influenced by symptomatic PD therapy (e.g. plasma dopamine or FDOPA PET uptake) is unlikely to be helpful, as the biomarker will have very limited clinical utility.
In parallel to this pre-study 'work-up' of the biomarker, the validity, reliability, and responsiveness, including to clinical change, of the selected criterion (e.g. clinical rating scale) against which a biomarker will be examined, must be explored. Much work has been undertaken in assessing the validity and reliability of psychometric instruments Flow diagram outlying a provisional 'roadmap' for conducting a study to determine whether a given biomarker is a suitable surrogate for a clinical measure of disease progression.
[24], and we would suggest a similar approach here. Maximising the scientific rigor of the selected criterion is central to improving the chance of coming to the correct conclusion about the efficacy of a biomarker for disease progression, and will have implications for biomarker study sample size calculations.
Following completion of all these preliminary steps it should then be possible to undertake a power calculation to determine an appropriate sample size for the biomarker study. Unfortunately no studies included in this review performed a power calculation, and the small number of participants (median 32 (IQR 21 to 53)) in these studies is of concern. As studies get smaller it becomes increasingly likely that potentially significant associations will not be detected, and limits the number of variables that can be included in multivariate analyses without significantly increasing the risk of spurious findings. In a heterogeneous disease such as PD, small sample sizes also mean the cohort is unlikely to be representative of all patients with PD.
Whilst cross-sectional studies formed the majority of articles included in this systematic review, a crosssectional design is not suitable to examine for a relationship between a change in a clinical measure and the change in a biomarker over time within individual patients with PD. The longitudinal studies included in this review had a median follow-up duration of 2.0 (IQR 1.1 to 3.5) years. There is currently no evidence to suggest what the minimum duration of a surrogate outcome biomarker study should be. Undoubtedly it needs to be long enough for a clinically significant change in the criterion, used to draw associations with the putative biomarker, to be observed. However, if a short-term change in a biomarker is to be associated with a long-term change in a clinical outcome measure then clearly a longer period of follow-up is required. In the included longitudinal studies the biomarker and clinical measures were generally only measured twice (mean 2.2 (SD 0.4) time points), which is clearly insufficient to allow the differentiation of a linear from a non-linear association. Future biomarker studies must be longitudinal, and measure the biomarker and clinical measures at several time points (at least three) over a sufficient follow-up period, which is more likely to be measured in years than months, as only this design will provide sufficient evidence of a biomarkers potential validity.
Parkinson's disease is a clinically heterogeneous disorder, with patients varying significantly in terms of age of onset, presentation and progression of motor and non-motor features. Ideally, biomarker studies should be large enough and use broad enough inclusion criteria to capture this heterogeneity and identify the utility of the biomarker in different subgroups of patients. The use of moderately to severely restrictive inclusion and exclusion criteria in most of the studies included in this review has influenced the characteristics of the study participants. The mean age of participants in included studies was only 63.5 (SD 5.6) years. Given that the median duration of disease at study entry was 5.7 (IQR 3.6 to 7.4) years, these patients have fairly early onset disease. As PD is New quality criteria to assess studies examining surrogate biomarkers for disease progression
## Question yes no
(1) Was the primary aim of the study to validate a biomarker for disease progression?
(2) Did the study detail a scientifically valid reason for choosing the given biomarker for investigation?
(3) Has the reproducibility of measuring the biomarker in the same centre by different trained personnel, and between centres, been evaluated? [bib_ref] Are delayed-start design trials to show neuroprotection in Parkinson's disease fundamentally flawed?, Clarke [/bib_ref] Has an assessment of the effect of likely confounding factors (e.g. age, gender, smoking status, and being on symptomatic PD treatment) on the measurement of the biomarker been made?Has an assessment of the validity and reliability of the criterion (e.g. clinical rating scale) used been made?
(6a) Was a power calculation undertaken to determine the required number of participants?
(6b) If a power calculation was undertaken, was the number of participants included appropriate?
Was the study longitudinal? [bib_ref] Assessment of neuroimaging techniques as biomarkers of the progression of Parkinson's disease, Brooks [/bib_ref] Was the study prospective? [bib_ref] Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of..., Hughes [/bib_ref] Was there a sufficient period of follow-up? [bib_ref] Diagnostic criteria for Parkinson disease, Gelb [/bib_ref] Were the biomarker and clinical measures of disease severity measured on ≥3 occasions? [bib_ref] Criteria for diagnosing Parkinson's disease, Calne [/bib_ref] Was measurement of the biomarker blind to participant characteristics?Did ≥ 75% of participants entering the study complete the full follow-up period? [bib_ref] UPDRS Development Committee: The Unified Parkinson's Disease Rating Scale, Fahn [/bib_ref] Were cases unselected/unbiased (no exclusion criteria)? [bib_ref] Mini-mental state". A practical method for grading the cognitive state of patients..., Folstein [/bib_ref] Were associations between the biomarker and clinical measures of disease severity examined for using appropriate statistical modelling (e.g. linear mixed modelling) with adjustment for confounding factors, rather than simply correlation analysis? [bib_ref] A standardised instrument for the diagnosis of mental disorder in the elderly..., Roth [/bib_ref] Were results of statistical analyses reported in sufficient detail to allow the inclusion of the study results in a meta-analysis? mainly a disorder of the elderly [bib_ref] Pilot study of the incidence and prognosis of degenerative Parkinsonian disorders in..., Taylor [/bib_ref] these cohorts of patients may, therefore, not be representative of the majority of patients with PD. Future studies should justify their inclusion and exclusion criteria, and try to minimise the latter to avoid limiting the generalisability of their results or highlight their limited generalisability. The general reporting of results in the included studies was poor, making data extraction and interpretation difficult. This was particularly true of many imaging studies, where there was often ambiguity as to whether stated values were mean values or total values for structures represented in both hemispheres. Similarly there was often a failure to report important methodological features, such as the study setting, or whether inclusion and exclusion criteria were applied.
The reporting of the statistical analyses was also inadequate. In both correlation and regression analyses, hypothesis testing can be undertaken to determine whether a relationship exists in the population as a whole, and confidence intervals calculated to indicate the strength of that relationship. Whilst significance testing was undertaken in most included studies, they failed to report confidence intervals, thereby limiting the ease of interpretation of analyses. Many studies also omitted to report precise significance values, instead simply giving results descriptive in the text. If due to pressure of space the actual results of statistical analyses cannot be fully included in a journal article then data should be provided as an additional online resource. Disappointingly several studies even failed to detail what statistical techniques they used. Without clear reporting of the study methodology, results, and the outcome of statistical analyses, investigators devalue their study and risk it being excluded from future systemic reviews or meta-analyses.
The statistical techniques applied in the included studies were in many cases inappropriate, and more often than not too simplistic. Unfortunately it appeared that many of the studies included in this review had been conducted without input from an experienced statistician. Whilst a detailed review of the appropriate statistical techniques to use in disease progression biomarker studies is beyond the scope of this review, we wish to make a few key points.
The majority of included cross-sectional studies used correlation analysis to examine for a relationship between clinical outcome measures and a given biomarker. However, correlation is a limited technique to use for this purpose as it only indicates the strength and direction of a relationship between two variables [bib_ref] Statistics review 7: Correlation and regression, Bewick [/bib_ref]. Furthermore correlation does not allow more complex assessment of the influence of possible confounders, and does not allow the value of one variable to be predicted when the other is known.
Several included studies measured a large number of variables and then calculated multiple individual correlation coefficients and significance values. This sequential testing of multiple variables makes it more likely that a significant correlation will be found simply by chance. Multivariate analysis is the appropriate technique to use when the relationship between a single dependent variable and several explanatory variables is examined at the same time, to avoid the pitfalls of multiple testing. It also has the advantage of allowing the inclusion of factors which may affect the relationship between the variables of interest, and adjusts for these potential confounders. Therefore, whilst correlation analysis can be viewed as a basic statistical technique which can in certain circumstances be used to generate hypotheses, higher levels of statistical analysis should be used to test hypotheses [bib_ref] Relation between two continuous variables, Altman [/bib_ref].
Use of correlation analysis in longitudinal studies, where repeated measurements are performed, is inappropriate as this takes no account of the possible covariance between measurements recorded in the same individual. However the use of regression is limited to situations where variables are measured twice in the same individual. In these cases regression can be used either by including baseline measurements as a covariant in the model, or by converting the pairs of measurements recorded in each individual into the actual change or percentage change from baseline, before using these calculated values in the regression model.
Where variables are measured three or more times in individuals in longitudinal studies, it is possible to undertake separate analyses to compare variables at baseline and each subsequent time point individually. However, this again introduces repeated testing, increasing the likelihood of finding a significant association by chance. It also takes no account of missing data, common in longitudinal studies due to deaths, loss to follow-up or withdrawal from the study. It is therefore not an approach we would recommend.
A potential solution to these problems is the use higher levels of statistical modelling, for example linear mixed models which can be used to test for linear relationships [bib_ref] Repeated measures data. In Applied mixed models in medicine, Brown [/bib_ref]. In the analysis of repeated measurements from longitudinal studies, these types of models take into account patients who drop-out early when estimates for later time points are made. However, the models also ensure that those with incomplete data do not influence the results as greatly as those with complete data. Using these models, a time effect on the repeated measure, independent from disease progression, is allowed for. In addition as disease progression may differ across time points these types of model can allow for this by introducing a 'disease progression-by -time interaction'. The non-independence of the repeated measures can be accounted for by simply assuming there is a constant correlation for all pairs of measurements in the same individual or by, for example, varying this correlation as measurements become more widely separated by time.
Therefore, we would strongly recommend that future biomarker studies incorporate a range of analyses, rather than simply correlation, in order to explore the validity of more advanced statistical methods. The use of appropriate statistical techniques should reduce the chance of type I and type II errors, and thereby allow sensible conclusions to be drawn about the efficacy of biomarkers for disease progression. Such analyses should be conducted by experienced statisticians, who therefore, need to be involved in the planning and conduct of future studies.
We attempted to group blood and CSF biomarkers into categories with shared characteristics (e.g. structural proteins, free radicals and markers of oxidative damage, antioxidants, pro-inflammatory cytokines, metals, amino acids, and vitamins) to see if this would allow any sensible sub-grouping of the results. However, the biomarkers did not easily fit into a simple categorical structure and hence we abandoned this approach.
It could be argued that the failure of included longitudinal studies to detect associations between the change in clinical measures of disease progression and the change in a given biomarker reflects a lack of sensitivity to change in the clinical markers of disease progression used, rather than a failure of the biomarker to change with disease progression. However, in the majority of included longitudinal studies there was a clear change in the clinical measures used between time points, making this hypothesis unlikely.
One potential criticism of our systematic review is that we included studies whose primary aim was not to develop a biomarker for disease progression in PD. We did so to ensure that our review was as inclusive as possible and did not miss potential biomarkers. It does, however, raise the question of the appropriateness of studies with an alternative primary aim (for example, drug development) of undertaking these additional, usually cross-sectional, analyses to produce information regarding such associations. As we have highlighted the development of biomarkers for disease progression in PD requires studies with a primary objective of biomarker validation, exemplified by a careful study design. Such studies could be run alongside other studies (e.g. long-term prognostic studies or clinical trials) but only if appropriately planned and funded. Conflicts of interest and issues around the subsequent ownership and availability of biomarkers in the public domain may arise if biomarkers are developed in clinical trials funded by pharmaceutical companies and these issues need to be considered beforehand.
Some of the lessons of this systematic review, in particular the necessity for longitudinal studies measuring putative biomarkers and clinical measures several times over several years, have begun to be realised by some researchers and have started to be put into practice. In PD the Parkinson's Progression Markers Initiative (PPMI)and the Parkinson's Disease Biomarkers Program (PDBP)both aim to measure various putative CSF, blood, and imaging biomarkers over several years. Their work, like that of the longitudinal Alzheimer's Disease Neuroimaging Initiative (ADNI)in Alzheimer's disease, should mark a major shift in the quality of studies of biomarkers for disease progression, and hopefully lead to advances in this important field.
[fig] [ 18 F: ]-2-fluoro-2-deoxyglucose [FDG]) and cerebral blood flow (N-isopropyl-P [ 123 I]-iodoamphetamine; [ 99m Tc]-hexamethylpropylene amine oxidase [HMPAO]), although the majority were ligands used to measure pre-synaptic dopaminergic function, either relating to fluorodopamine synthesis and storage (]6-fluoro-L-3,4-dihydroxyphenylalanine) Studies included in final analysis (n = 183) [/fig]
[fig] Figure 1: Flow diagram outlying the selection procedure to identify 183 articles included in the systematic review of biomarkers for disease progression in PD. Note that of the 58 articles identified by reviewing reference lists, 41 were excluded, one article could not be located and 16 were included in the final qualitative synthesis. All eight articles identified by hand searching (all cross-sectional) were included in the final qualitative synthesis.[FDOPA]) or dopamine active transporter (DAT) function ([ 18 F]-2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl)-N-tropane [[ 18 F]FP-CIT]; [ 123 I]-2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl)-N-tropane [[ 123 I]FP-CIT]; [ 123 I]-2β-carbomethoxy-3β-(4-iodophenyl tropane) [[ 123 I]β-CIT]; 2β-carbomethoxy-3β-(4-[ 18 F])fluorophenyl) tropane [[ 18 F]CFT]). This marked heterogeneity in the studies unfortunately made quantitative synthesis of the extracted data impossible. [/fig]
[table] Table 2: Study characteristics of the 183 included articles [/table]
[table] Table 3: Baseline characteristics of study participants in the 183 included articlesMeans are presented with standard deviations, and medians with interquartile ranges (IQR). The P value column relates to comparisons made between longitudinal and cross-sectional studies using the Mann-Whitney test. [/table]
[table] Table 4: Types of putative biomarkers for disease progression investigated in the 183 included articles [/table]
|
Addressing Atropisomerism in the Development of Sotorasib, a Covalent Inhibitor of KRAS G12C: Structural, Analytical, and Synthetic Considerations
CONSPECTUS: Nearly a century after its first description, configurationally stable axial chirality remains a rare feature in marketed drugs. In the development of the KRAS G12C inhibitor sotorasib (LUMAKRAS/LUMYKRAS), an axially chiral biaryl moiety proved a critical structural element in engaging a "cryptic" protein binding pocket and enhancing inhibitor potency. Restricted rotation about this axis of chirality gave rise to configurationally stable atropisomers that demonstrated a 10-fold difference in potency. The decision to develop sotorasib as a single-atropisomer drug gave rise to a range of analytical and synthetic challenges, whose resolution we review here. Assessing the configurational stability of differentially substituted biaryl units in early inhibitor candidates represented the first challenge to be overcome, as differing atropisomer stability profiles called for differing development strategies (e.g., as rapidly equilibrating rotamers vs as single atropisomers). We relied on a range of NMR, HPLC, and computational methods to assess atropisomer stability. Here, we describe the various variable-temperature NMR, time-course NMR, and chiral HPLC approaches used to assess the configurational stability of axially chiral bonds displaying a range of rotational barriers. As optimal engagement of the "cryptic" pocket of KRAS G12C was ultimately achieved with a configurationally stable atropisomeric linkage, the second challenge to be overcome entailed preparing the preferred (M)-atropisomer of sotorasib on industrial scale. This synthetic challenge centered on the large-scale synthesis of an atropisomerically pure building block comprising the central azaquinazolinone and pyridine rings of sotorasib. We examined a range of strategies to prepare this compound as a single atropisomer: asymmetric catalysis, chiral chromatographic purification, and classical resolution. Although chiral liquid and simulated moving bed chromatography provided expedient access to initial multikilo supplies of this key intermediate, a classical resolution process was ultimately developed that proved significantly more efficient on metric-ton scale. To avoid discarding half of the material from this resolution, this process was subsequently refined to enable thermal recycling of the undesired atropisomer, providing an continued...
even more efficient commercial process that proved both robust and green. While the preparation of sotorasib as a single atropisomer significantly increased both the analytical and synthetic complexity of its development, the axially chiral biaryl linkage that gave rise to the atropisomerism of sotorasib proved a key design element in optimizing sotorasib's binding to KRAS . It is hoped that this review will help in outlining the range of analytical techniques and synthetic strategies that can be brought to bear in addressing the challenges posed by such axially chiral compounds and that this account may provide helpful guidelines for future efforts aimed at the development of such single atropisomer, axially chiral pharmaceutical agents.
## Ligand design: axial chirality as a structural tool
Although the topic of atropisomerismin the pharmaceutical sciences has received growing attention over the past decade, the presence of configurationally stable, axially chiral stereochemical elements in marketed drugs remains a rarity a century after the first report of axial atropisomerism. [bib_ref] �The molecular configurations of polynuclear aromatic compounds. Part I. The resolution of..., Christie [/bib_ref] To date, only four United States Food and Drug Administration (FDA)-approved drugs incorporate such a stereochemical feature: the natural products vancomycin and colchicine, whose axial chirality is enforced as a result of conventional stereocenters in rings bridging their biaryl motifs; [bib_ref] Atropisomerism in medicinal chemistry: challenges and opportunities, Toenjes [/bib_ref] lesinurad, whose configurational stability was only recognized postlaunch (through separation of the racemic marketed product); [bib_ref] Discovery and Assessment of Atropisomers of (±)-Lesinurad, Wang [/bib_ref] and sotorasib (LUMAKRAS/ LUMYKRAS), which represents the first FDA-approved therapy to be manufactured and marketed as a configurationally stable, atropisomerically pure compound [fig_ref] Figure 1: FDA-approved drugs containing axially chiral atropisomeric bonds [/fig_ref]. [bib_ref] Discovery of a Covalent Inhibitor of KRAS(G12C) (AMG 510) for the Treatment..., Lanman [/bib_ref] In this work, we describe the factors that led to the decision to incorporate an axially chiral linkage in the structure of sotorasib as well as the analytical and synthetic techniques used to address the consequences of that decision. The decision to incorporate an axially chiral linkage in sotorasib arose from a specific structural need: to provide a structural motif capable of accessing a "cryptic" pocket 13 on the surface of the target protein, KRAS G12C , that had been discovered through our earlier screening efforts. 14 As shown in [fig_ref] Figure 2: Axial chirality as a structural element in accessing a novel pocket in... [/fig_ref] , the biaryl linkage 15 of quinazolinone lead 2 provided a synthetically accessible, readily diversified structural element capable of positioning a substituent in the H95/Y96/Q99 cryptic pocket (magenta), thereby mimicking the role played by the tetrahydroisoquinoline amide of indole lead 1 (which similarly accessed this cryptic pocket via an atropisomeric linkage (red bond)�albeit a nonconfigurationally stable one).The decision to retain the atropisomeric linkage of quinazolinone 2 in the final structure of sotorasib was not taken lightly. Axial chirality differs from more conventional point chirality in that the racemization of atropisomers does not depend on bond-breaking processes, but only upon simple rotational interconversion. Therefore, the energetic barriers to atropisomer isomerization are typically significantly lower than those of point stereocenter racemization processes, posing considerable added analytical and synthetic challenges for the development of axially chiral compounds. [bib_ref] Assessing atropisomer axial chirality in drug discovery and development, Laplante [/bib_ref] [bib_ref] Stereochemistry and Recent Applications of Axially Chiral Organic Molecules, Mancinelli [/bib_ref] In the following sections, we survey the range of analytical techniques we used to characterize the configurational stability of the diverse atropisomeric analogs prepared in the optimization of sotorasib as well as the various synthetic strategies employed in accessing atropisomerically pure, axially chiral compounds efficiently, economically, and on commercial scale. [bib_ref] 2022, 26, 710−729. (19) Both chemical manufacturing and control (CMC)-related (isomerization during..., Nikitidis [/bib_ref] As such efforts were central to the registration of the first atropisomerically pure, axially chiral pharmaceutical agent, it is hoped that this overview may provide helpful guidelines for future efforts seeking to bring single-atropisomer compounds to market.
## Conformational analysis: interconversion barriers and conformational stability
Efforts to optimize quinazolinone scaffold 2 initially focused on modification of the ortho substituent of the biaryl moiety (green ball, [fig_ref] Figure 2: Axial chirality as a structural element in accessing a novel pocket in... [/fig_ref] to optimally engage the cryptic pocket of KRAS G12C as well as modification of the monocyclic arene ring. Both modifications contributed to changes in the configurational stability of the axially chiral biaryl system [fig_ref] Figure 3: Interconversion of biaryl atropisomers [/fig_ref]. The configurational stability of an atropisomeric linkage, such as that present in quinazolinone 2, can have significant implications for the way in which an atropisomeric compound is developed for pharmaceutical use (e.g., as a mixture or single atropisomer; [fig_ref] Table 1: Favored Development Strategy As a Function of Atropisomer Interconversion Barrier [/fig_ref] , as has been expertly reviewed. [bib_ref] The challenge of atropisomerism in drug discovery, Clayden [/bib_ref] [bib_ref] Assessing atropisomer axial chirality in drug discovery and development, Laplante [/bib_ref] Conformational stability is typically assessed by estimation of the free-energy of activation for atropisomer interconversion (ΔG ⧧ ) [fig_ref] Figure 3: Interconversion of biaryl atropisomers [/fig_ref]. During research efforts toward the discovery of sotorasib, we sought to optimize the biaryl substituent of scaffold 2 (e.g., ring-size and substitution pattern) to maximize productive interactions with the H95/Y96/Q99 pocket of KRAS G12C while avoiding the generation of analogs with low-tomoderate atropisomeric stability (ΔG ⧧ of 20−30 kcal/mol), 1 as the configurational instability of such compounds was expected to pose significant regulatory challenges. To assess atropisomer configurational stability, we relied on a range of analytical techniques (see [fig_ref] Table 2: Choosing an Experimental Technique for Configurational Stability Assessment [/fig_ref] , each of which proved useful for characterizing conformational dynamics occurring on differing time scales and at differing interconversion energies.
NMR techniques proved one of the most versatile means of detecting and analyzing atropisomer mixtures. In the case of quinazolinone analogs such as 2, the presence of a second stereocenter in a molecule (i.e., the (S)-methyl piperazine) rendered the two (M)-and (P)-atropisomers 21 diastereomeric. Diastereotopic NMR resonances arising from the two atropisomers therefore provided a first indication of the presence of restricted rotation arising from the axial chirality in compound 2.In the case of scaffold 2, variable-temperature NMR (VT-NMR) studies proved a highly useful tool in assessing the activation energy barrier (ΔG ⧧ ) for atropisomer interconversion. For compounds with low interconversion barriers (ΔG ⧧ ), spectral line broadening (and/or merging) at elevated temperatures allowed for the direct calculation of ΔG ⧧ , ΔH ⧧ , and ΔS ⧧ from the exchange rate constants (k) or half-lives (t 1/2 ) of the interconversion processes. For compounds with slower exchange rate constants (i.e., higher interconversion energy barriers), a range of additional NMR techniques also proved useful: 23
- When interconversion rates were short on an NMR timescale (i.e., t 1/2 < 1 s), line-shape analysis could be used to determine interconversion rate constants (k) by simulating exchange processes using dynamic NMR models. [bib_ref] Iterative computer analysis of complex exchange-broadened NMR bandshapes, Stephenson [/bib_ref] [bib_ref] Monte Carlo simulation of DNMR spectra of coupled spin systems, Szalay [/bib_ref] - In cases where an intermediate exchange regime [bib_ref] An introduction to NMR-based approaches for measuring protein dynamics, Kleckner [/bib_ref] was observed (i.e., k ≈ frequency difference (Δν 0 ) of monitored nuclei), rate constants (k) could be calculated from Δν 0 of the diastereotopic nuclei at the coalescence temperature (T c ) of the interconverting atropisomers. [bib_ref] NMR determination of the rotational barrier in N, N-dimethylacetamide. A physical chemistry..., Gasparro [/bib_ref] - In the slow-to-intermediate exchange regime [bib_ref] An introduction to NMR-based approaches for measuring protein dynamics, Kleckner [/bib_ref] (k < Δν 0 ), 2D exchange spectroscopy (EXSY) could be used to calculate k with increased sensitivity by measuring the intensity of chemical-exchange cross peaks as a function of mixing time. [bib_ref] 90, 935−967. (32) This approach could also be used when an exchange..., Perrin [/bib_ref] ,32 - For systems with higher rotational barriers, kinetic analysis (i.e., time-course NMR) could be used to monitor the isomerization of purified atropisomers as a function of time and temperature, deriving the exchange rate (k) and reaction half-life (t 1/2 ) from first-order reaction kinetics.Atropisomer configurational stability analysis also took place against a background of other rotational and conformationally dynamic processes in the sotorasib scaffold, all of which additionally contributed to the generation of diastereotopic 1 H NMR resonances on the NMR-time scale [fig_ref] Figure 4: Conformationally/chemically dynamic features of the sotorasib scaffold [/fig_ref]. Conformationally dynamic features of the sotorasib scaffold included:
A. The axially chiral biaryl linkage (source of atropisomerism in most sotorasib analogs) B. A second axially chiral bond (not typically a source of configurationally stable rotamers due to low interconversion barriers at ambient temperature) C. A conformationally flexible piperazine ring/acrylamide "warhead", giving rise to E-/Z-acrylamide isomers, piperazine conformers, and rotational isomers about the piperazine/quinazolinone bond (not typically a source of configurationally stable isomers at ambient temperature) D. Proton exchange between the phenolic hydroxyl group and basic sites on the molecule/solvent.
As an illustration of the use of VT-NMR line-shape analysis in determining the interconversion energy barriers for several of these dynamic processes, below, we describe two processes within the sotorasib scaffold (bond A and B rotation) that have sufficiently different activation energy barriers to allow for their concurrent analysis using simple two-site exchange models [fig_ref] Figure 5: VT-NMR analysis [/fig_ref].
In [fig_ref] Figure 5: VT-NMR analysis [/fig_ref] , proton-decoupled 19 F VT-NMR was used to determine the barrier for fluorophenyl ring rotamer inter- Interconversion barrier classification and development strategies taken from refs 6 and 20. b Interconversion rapid versus pharmacological and manufacturing processes. c Interconversion on the same time scale as pharmacological and manufacturing processes.
As the forward and reverse rates of interconversion between the (M)-and (P)-fluorophenol rotamers are very similar (k +1 ≈ k −1 ), nearly equal concentrations of the two rotamers are observed at equilibrium (k +1 /k −1 = [M] eq /[P] eq ≈ 1), as is reflected by the similar integrals for the two 19 F signals.
At elevated temperatures, the 19 F NMR signals for (M)-3 gradually broaden (reflecting increasingly rapid rotation about bond B), ultimately reaching coalescence at ∼327 K. At the coalescence temperature (T c ), the exchange rate (k) between the two rotamers can be calculated from the frequency difference Δv 0 between these 19 F peaks in the limit of slow exchange: 23,30
[formula] = k 2 0 (2) [/formula]
The activation free energy for this rotational exchange process (ΔG ⧧ ) can then be calculated from the Eyring equation: 23,37,38
[formula] = = ‡ ‡ ‡ G H T S RT k k T ln B(3) [/formula]
By this methodology, given a frequency difference (Δv 0 ) of 97 ± 2 Hz in the absence of exchange, the exchange rate at T c = 327 ± 2 K can be calculated as k = 215 ± 4 s −1 , corresponding to a ΔG ⧧ of 15.7 ± 0.1 kcal mol −1 . [fig_ref] Figure 5: VT-NMR analysis [/fig_ref] illustrates a similar VT-NMR analysis of atropisomer interconversion rates, this time using the diastereotopic 1 H NMR resonances of the quinazolinone H5 hydrogen atom to monitor the interconversion of biaryl ("bond A") atropisomers. This figure additionally illustrates the complex NMR spectra that can result from the concomitant contributions of multiple rotationally mobile bonds. At ambient temperature, H5 shows not only two distinct 1 H NMR resonances arising from the two "bond A" atropisomers, but also significant broadening/splitting of these resonances resulting from the presence of two "bond B" (fluorophenyl ring) atropisomers (cf. [fig_ref] Figure 4: Conformationally/chemically dynamic features of the sotorasib scaffold [/fig_ref].
At temperatures >340 K, fluorophenyl rotation rates enter the fast-exchange regime, resulting in the sharpening of the H5 resonances into two narrow peaks of similar intensity, corresponding to the two "bond A" atropisomers. With further increases in temperature, these two peaks broaden and begin to coalesce, with a T c ≈ 380 ± 2 K. Given a peak separation of Δv 0 = 27 ± 1 Hz for the "bond A" atropisomers (i.e., at 343 K), this results in a calculated exchange rate of k = 60 ± 2 s −1 at T c and a ΔG ⧧ = 19.3 ± 0.1 kcal mol −1 .
As outlined by LaPlante [fig_ref] Table 1: Favored Development Strategy As a Function of Atropisomer Interconversion Barrier [/fig_ref] , 6 the low interconversion barriers for both of the atropisomer interconversion processes depicted in [fig_ref] Figure 5: VT-NMR analysis [/fig_ref] make these processes of little concern from a development perspective: atropisomer conversion about "bond A" of compound 4 and "bond B" of (M)-3 is rapid on a pharmacological and manufacturing time scale, and so no effort was made to separately isolate and characterize these rotational isomers. Unfortunately, neither of these compounds ultimately proved suitable for further development, and optimization efforts continued with more highly conformationally restricted biaryl atropisomers. [bib_ref] Discovery of a Covalent Inhibitor of KRAS(G12C) (AMG 510) for the Treatment..., Lanman [/bib_ref] A key limitation of VT-NMR methods is the requirement that rapid atropisomer interconversion rates be obtained at temperatures compatible with NMR analysis. To assess KRAS inhibitors possessing higher atropisomer interconversion barriers (i.e., up to 32−33 kcal/mol), we turned to time-course NMR experiments, where we monitored interconversion dynamics over extended periods at elevated temperatures.
## Accounts of chemical research
pubs.acs.org/accounts Article [fig_ref] Figure 6: Time-course NMR analysis [/fig_ref] illustrates the use of time-course NMR studies to assess the atropisomer ("bond A") interconversion barrier of a more highly conformationally restricted inhibitor, compound 5. In this experiment, an (M)-atropisomer-enriched sample of 5 was incubated at 325 K for ∼19 h, and the diastereotopic quinazolinone F6 19 F NMR resonances [fig_ref] Figure 6: Time-course NMR analysis [/fig_ref] ; corresponding to the two "bond A" atropisomers) were integrated at successive time-points to provide a time-course profile of evolving (M)-and (P)-5 concentrations [fig_ref] Figure 6: Time-course NMR analysis [/fig_ref].
For a two-site exchange process (eq 1), the rate of (M)atropisomer change over time is given by
[formula] [ ] = [ ] + [ ] + d M dt k M k P 1 1(4) [/formula]
where k +1 and k −1 are the rate constants for forward and reverse reaction, respectively. [fig_ref] Figure 6: Time-course NMR analysis [/fig_ref] to eq 5 revealed rate constants of k +1 ≈ k −1 = 3.46 ± 0.05 × 10 −5 s −1 and a corresponding half-life of t 1/2 = 334 ± 5 min, which (using eq 3) equated to an activation energy of ΔG ⧧ = 25.7 ± 0.1 kcal mol −1 at 325 ± 1 K.
This represented a significantly higher barrier to atropisomer interconversion than for the previously described compounds, and one expected to pose a significant development challenge, as atropisomer interconversion of compound 5 would be expected to occur on the same time-scale as that of pharmacological and manufacturing processes. Accordingly, although compounds such as quinazolinone 5 proved potent KRAS G12C inhibitors, further optimization was pursued to identify similarly potent inhibitors with atropisomer interconversion barriers <20 kcal/ mol or >30 kcal/mol (cf. [fig_ref] Table 1: Favored Development Strategy As a Function of Atropisomer Interconversion Barrier [/fig_ref]
## Guidelines).
Such optimization efforts, more fully described in ref 1, ultimately led to the identification of sotorasib [fig_ref] Figure 1: FDA-approved drugs containing axially chiral atropisomeric bonds [/fig_ref]. Using similar time-course NMR studies, the atropisomer interconversion barrier for sotorasib was determined to be >33.5 kcal mol −1 , corresponding to an atropisomer interconversion half-life of >1000 h at 373 K. Slow decomposition of sotorasib in DMSO at T > 380 K prevented more precise determination of the interconversion barrier; however, experiments with synthetic precursors (see below) have helped to more precisely define the interconversion barrier for the pyridine/quinazolinone system of sotorasib.
As previously noted, direct NMR analysis of atropisomer mixtures relies on the presence of a second stereocenter in the analyte molecule to render the atropisomers diastereotopic. Although enantiomeric atropisomers may be studied spectroscopically by using chemical shift reagents or anisotropic media, 39,40 a potentially more straightforward approach employs chiral chromatographic analysis. In the case of sotorasib, the synthetic precursor rac-6 [fig_ref] Figure 7: Chiral chromatographic analysis [/fig_ref] possessed a single chiral axis (red) and no additional stereocenters. Restricted rotation about this axis resulted in the formation of two enantiomeric atropisomers (M)-and (P)-6, which could not be distinguished using standard NMR approaches.
To assess the atropisomer interconversion barrier of compound 6, a chiral high-performance liquid chromatography (HPLC) method was developed to separate (M)-and (P)-6, enabling time-course studies of the thermally induced racemization of (P)-6.In such studies, a solution of (P)-6 in diphenyl ether was superheated to 280°C (boiling point = 258°C
) and samples were collected over ∼2 h to determine the degree of racemization at each time point [fig_ref] Figure 7: Chiral chromatographic analysis [/fig_ref]. Fitting the exponential decay of (P)-6 concentrations over time allowed for the calculation of the racemization rate constant, k (2.21 × 10 −4 s −1 ), and corresponding interconversion free energy, ΔG ⧧ (42.3 kcal/mol), resulting in a calculated racemization half-life of >10 billion years (at 25°C), confirming the exceptional configurational stability of this biaryl linkage.
Although the prior discussion focused on experimental methods to establish atropisomer interconversion barriers, computational (quantum-mechanical) methods of estimating such barriers have also become quite reliable [bib_ref] Assessing atropisomer axial chirality in drug discovery and development, Laplante [/bib_ref] [bib_ref] Revealing atropisomer axial chirality in drug discovery, Laplante [/bib_ref] and can serve as a convenient method for assessing rotational restriction about an axially chiral bond [and for predicting the most appropriate experimental approach to use in assessing atropisomer interconversion barriers (cf. [fig_ref] Table 2: Choosing an Experimental Technique for Configurational Stability Assessment [/fig_ref] ]. [fig_ref] Figure 8: Computational analysis [/fig_ref] illustrates the application of density functional theory (DFT) calculations at the B3LYP/6-31+G(d,p) level of theory to estimate the interconversion barrier of rac-6, which is predicted to be 40.7 kcal/mol, in good correspondence with the experimentally determined barrier of 42.3 kcal/mol.
## Synthetic considerations: preparing single atropisomers on scale
Having identified the atropisomeric biaryl bond of sotorasib as a key structural element in the design of potent KRAS G12C inhibitors and established the exceptional configurational stability of the specific biaryl system present in sotorasib through careful analytical work, our efforts subsequently turned to the practical question of how best to prepare the (M)atropisomer of sotorasib on industrial scale. During discovery efforts, the (M)-atropisomer was isolated by chiral chromatographic separation of the two atropisomers obtained upon completion of the synthesis. [bib_ref] Discovery of a Covalent Inhibitor of KRAS(G12C) (AMG 510) for the Treatment..., Lanman [/bib_ref] Given the low throughput, significant solvent waste, and poor scalability of such a chromatographic separation, finding ways to install the (M)atropisomeric axis without resorting to late-stage chiral separation became a key focus at the outset of process development efforts. The earliest atropisomeric intermediate in the synthesis of sotorasib, rac-6, was formed in just two steps from raw materials nicotinamide 7 and aniline 8, proceeding through nonisolated intermediate 9 [fig_ref] Figure 9: Preparation of rac-6 and approaches to prepare [/fig_ref]. 43 Several strategies were envisioned for accessing atropisomerically pure (M)-6 while retaining 7 and 8 as raw materials: (1) asymmetric ring closure of intermediate 9, (2) chiral chromatographic separation of rac-6, and (3) classical resolution of rac-6.
Asymmetric synthesis was initially envisioned as being the most efficient means of accessing (M)-6, and initial efforts toward this goal focused on the identification of suitable catalysts for the asymmetric ring closure of intermediate 9, given established supply chains for starting materials 7 and 8, which promised to expedite future scale-up. Although multiple approaches 44−46 were pursued�including asymmetric intramolecular cross-coupling, various organocatalytic methods (hydrogen bonding, chiral Brønsted acids, and nucleophilic catalysis), and stochiometric methods (chiral bases)�ultimately, none of these approaches met the required yield or selectivity requirements for the program. Given an aggressive timeline to deliver kilogram quantities of (M)-6 to support early development efforts, it was determined
## Accounts of chemical research pubs.acs.org/accounts
Article that chromatographic separation of rac-6 would be a pragmatic, phase-appropriate approach to provide material for Investigational New Drug-enabling toxicology studies and first-in-human clinical trials. Indeed, the first delivery of sotorasib for clinical studies relied on traditional chiral chromatography and delivered ∼2 kg of sotorasib drug substance. For a subsequent 10 kg delivery of drug substance, simulated moving bed chromatography was utilized for chiral separation, which reduced both processing times and solvent use when compared to traditional chromatographic separation. While preparative chromatography was key to establishing material supplies for early clinical development, it was not considered a preferred option for commercialization. In addition to the previously mentioned shortcomings (i.e., solvent waste, low throughput, etc.), such an approach also complicated supply chains and increased cycle time through the need to enlist third parties for chromatographic separation. This prompted the team to pursue the development of a classical resolution 48 of rac-6, which would both reduce cycle times and allow for implementation in standard processing equipment, eliminating the need for specialist third-party partners.
In developing a classical resolution of rac-6, the team leveraged high-throughput experimentation (HTE) to identify a suitable resolving agent and solvent system to enable the selective crystallization of the (M)-6 isomer.Two rounds of screening experiments were performed, both focusing on readily available chiral acid resolving agents and a variety of solvent systems. While an initial screen of >250 conditions failed to produce promising leads, a follow-up screen of >100 conditions�primarily focused on the use of orthogonal solvent systems�resulted in several conditions that produced complexes between 6 and several chiral acids [fig_ref] Table 3: Chiral Acids That Form Complexes with Compound 6 [/fig_ref].
Although (1S)-(−)-camphanic acid and D-(+)-malic acid produced crystalline complexes with 6 as determined by X-ray powder diffraction, neither acid provided any resolution when the resulting complexes were analyzed by chiral HPLC. (+)-2,3-Dibenzoyl-D-tartaric acid (DBTA) in a mixture of 2-methyltetrahydrofuran (2-MeTHF) and heptane, however, fortuitously generated a diastereomeric complex with (M)-6 that provided significant enantioenrichment (M/P ratio of 81:19; 62% de). Interestingly, analysis by single crystal X-ray diffraction revealed that the resolved (M)-6/DBTA complex was not a salt but rather a cocrystalline 2-MeTHF solvate with a 2:1 ratio of (M)-6 to DBTA. [bib_ref] A Proposal for a Feasible Classification System, Grothe [/bib_ref] Having achieved proof-of-principle, the process development team subsequently sought to optimize these conditions to provide a commercially viable process. Through optimization of the solvent ratio and by increasing the resolving agent stoichiometry, the team was able to both optimize (M)-6 recovery and maximize process robustness, ultimately delivering a process that consistently provided (M)-6 with an M/P ratio of >2000:1 (>99.9% de) in 42% yield (theoretical maximum yield = 50%) on >500 kg scale per batch [fig_ref] Figure 1: FDA-approved drugs containing axially chiral atropisomeric bonds [/fig_ref]. The development of a classical resolution to deliver atropisomerically pure (M)-6 was not only crucial in supporting accelerated commercialization timelines (by ensuring clinical and commercial supplies of sotorasib drug substance), but also significantly improved the overall "greenness" of the manufacturing process by substantially decreasing organic solvent use and reducing processing cycle times [fig_ref] Table 4: Process Efficiency Comparison [/fig_ref].
Despite these significant improvements in the efficient preparation of (M)-6, classical resolutions nonetheless have inherent inefficiencies, such as the need to discard the undesired stereoisomer as waste. To circumvent this shortcoming, we
## Accounts of chemical research
pubs.acs.org/accounts Article envisioned a process wherein the recovered undesired atropisomer (P)-6 could be recycled to serve as an alternative source of rac-6 through high temperature racemization, 53−55 substantially improving the process greenness metrics for this resolution (Scheme 1).As noted above, the atropisomer interconversion barrier for (P)-6 was known to be quite substantial (42.3 kcal/mol), and accordingly, it was expected that temperatures approaching 300°C or higher would likely be required to achieve reasonable racemization rates. Initial attempts to racemize (P)-6 (>99.5% ee) in a range of polar solvents led to significant decomposition; however, high-temperature racemization in nonpolar solvents was found to proceed with minimal decomposition [fig_ref] Table 5: Initial Screening Experiments to Racemize [/fig_ref]. Heating a solution of (P)-6 in anisole at 315°C in a plug flow reactor resulted in near complete racemization in only 20 min (mean residence time) while maintaining high overall purity. This process was subsequently executed on kilogram scale, affording >1.4 kg of racemized 6 in 77% isolated yield with a P/ M ratio of 50.5:49.5 following direct crystallization from the reaction mixture.
Implementing this procedure on a production scale required a means of recovering (P)-6 from the waste stream of the (+)-DBTA-mediated resolution. This was achieved by treating the 2-MeTHF/heptane waste stream with potassium carbonate to extract the resulting potassium DBTA salt into the aqueous phase, followed by crystallization of (P)-6, which was isolated in 90% yield with a P/M ratio of 88:12 (76% ee). Acidification of the aqueous extract allowed for recovery of (+)-DBTA (93% yield, crystallized) and subsequent reuse in the resolution without a reduction in efficiency. A depiction of the resolution/ recycling processes is shown in Scheme 2.Successful implementation of this resolution process allowed for the preparation of multimetric ton quantities of (M)-6 while avoiding the drawbacks initially posed by the requirement for large-scale chiral chromatography. Through optimization of the kilo-scale (P)-6/(+)-DBTA recovery and recycling process, we were additionally able to further enhance the efficiency of the recovery process, providing a greener approach to rac-6 that reduced its process mass intensity (PMI) contribution to the drug substance manufacturing process from 336 to 141 (58% reduction).
As this brief case history illustrates, while large-scale chiral chromatography played a key role in the delivery of atropisomerically pure sotorasib for early development activities, chromatographic approaches to atropisomer separation posed a significant hurdle to efficient drug synthesis. Classical resolution ultimately proved a highly selective, scalable, and efficient approach to the preparation of atropisomerically pure drug substance, particularly when coupled with process refinements that allowed for recycling of the undesired atropisomer.
## Concluding remarks
An axially chiral biaryl linkage played a central architectural role in the design and optimization of the KRAS G12C inhibitor sotorasib (LUMAKRAS/LUMYKRAS), providing stereodefined access to a novel cryptic pocket on the surface of the KRAS protein. Restricted rotation about this axially chiral linkage gave rise to configurationally stable atropisomers, which demonstrated a 10-fold difference in potency. The decision to develop sotorasib as a wholly synthetic, axially chiral singleatropisomer drug presented several analytical and synthetic challenges to be overcome, particularly as sotorasib has now become the first such molecule to receive marketing authorization.
In this account, we provide an overview of the analytical techniques (NMR-and HPLC-based) that were used to assess the configurational stability of the atropisomeric intermediates prepared during the optimization of sotorasib, along with descriptive guidelines for the selection of appropriate techniques to characterize the configurational stability of axially chiral compounds with differing degrees of rotational restriction.
We additionally survey some of the practical synthetic considerations that impacted our strategy in supporting the early clinical development of sotorasib as well as its later commercial synthesis. Whereas large-scale chromatography Data based on a clinical chromatography process not optimized for commercial production. b Includes solvent contributions from freebasing/crystallizing free (M)-6. c Process mass intensity relative to sotorasib drug substance. d Total solvent used per kg of (M)-6 produced. e Yield after crystallization (chromatography) or freebasing/crystallization (resolution).
## Scheme 1. (p)-6 recycling via thermal racemization
## Accounts of chemical research
pubs.acs.org/accounts Article provided efficient access to early supplies of atropisomerically pure sotorasib, commercial synthesis ultimately relied upon classical resolution to provide the operational simplicity and robustness required for synthesis on the metric ton scale. Further process refinements, allowing for recycling of the undesired isomer from the classical resolution, ultimately led to dramatic efficiency improvements in this resolution process, delivering a commercial process that was both robust and green. While the incorporation of an atropisomerically pure axis of chirality in sotorasib significantly increased the synthetic and analytical complexity of its development, this review illustrates the useful structural role that axial chirality can play in the design of efficient small molecule inhibitors, the range of analytical techniques that can be brought to bear in analyzing their configurational stability, and the synthetic strategies that can be used to enable the practical commercial-scale synthesis of atropisomerically pure pharmaceutical agents. It is hoped that this account may provide helpful guidelines for future efforts aimed at the development of such single atropisomer, axially chiral compounds.
[fig] •: Lanman, B. A.; Allen, J. R.; Allen, J. G.; Amegadzie, A. K.; Ashton, K. S.; Booker, S. K.; Chen, J. J.; Chen, N.; Frohn, M. J.; Goodman, G.; Kopecky, D. J.; Liu, L.; Lopez, P.; Low, J. D.; Ma, V.; Minatti, A. E.; Nguyen, T. T.; Nishimura, N.; Pickrell, A. J.; Reed, A. B.; Shin, Y.; Siegmund, A. C.; Tamayo, N. A.; Tegley, C. M.; Walton, M. C.; Wang, H. L.; Wurz, R. P.; Xue, M.; Yang, K. C.; Achanta, P.; Bartberger, M. D.; Canon, J.; Hollis, L. S.; McCarter, J. D.; Mohr, C.; Rex, K.; Saiki, A. Y.; San Miguel, T.; Volak, L. P.; Wang, K. H.; Whittington, D. A.; Zech, S. G.; Lipford, J. R.; Cee, V. J.Discovery of a Covalent Inhibitor of KRAS(G12C) (AMG 510) for the Treatment of Solid Tumors. J. Med. Chem.2020, 63, 52− 65. 1 Details the structure-based design of sotorasib (AMG 510), motivation for introducing an axially chiral biaryl linkage, and ef forts to address the conf igurational instability of early biaryl atropisomers. Additionally describes the biopharmaceutical optimization and pharmacological characterization of sotorasib. Parsons, A. T.; Caille, S.; Caporini, M. A.; Griffin, D. J.; Lovette, M. A.; Powazinik IV, W.; St-Pierre, G.Axial chirality in the sotorasib drug substance, Part 1: Development of a classical resolution to prepare an atropisomerically-pure sotorasib intermediate. Org. Process Res. Dev.2022, 26, 2629−2635. 2 Reports the discovery of (+)-2,3-dibenzoyl-D-tartaric acid [(+)-DBTA] as a resolving agent that enables the classical resolution of a key atropisomeric intermediate in the synthesis of sotorasib. Development of a commercial process that provides >2000:1 selectivity on >500 kg scale is described. Beaver, M. G.; Brown, D. B.; Campbell, K.; Fang, Y.-Q.; Ford, D. D.; Mardirossian, N.; Nagy, K. D.; Rotheli, A. R.; Sheeran, J. W.; Telmesani, R.; Parsons, A. T.Axial chirality in the sotorasib drug substance, Part 2: Leveraging a hightemperature thermal racemization to recycle the classical resolution waste stream. Org. Process Res. Dev.2022, 26, 2636−2645. 3 Describes the development of a continuousf low racemization process to recycle the off-enantiomer of an atropisomeric sotorasib intermediate obtained f rom the classical resolution. Additionally reports the demonstration of this process on kilogram scale. [/fig]
[fig] Figure 1: FDA-approved drugs containing axially chiral atropisomeric bonds. [/fig]
[fig] Figure 2: Axial chirality as a structural element in accessing a novel pocket in the design of sotorasib. An axis of chirality (red bond) in indole lead 1 facilitated occupancy of the H95/Y96/Q99 "cryptic" pocket (magenta) of KRAS G12C by a tetrahydroisoquinoline substituent (green), enhancing inhibit or potency. The metabolic instability of compound 1 prevented its further development; however, analogous use of an axially chiral linkage in quinazolinone lead 2 enabled similar occupancy of the H95/Y96/Q99 pocket while affording KRAS G12C inhibitors with significantly enhanced metabolic stability. Optimization of scaffold 2 led to the invention of sotorasib. Data from ref 1 and ref 14. [/fig]
[fig] Figure 3: Interconversion of biaryl atropisomers. Rotation about the axially chiral biaryl bond (red) allows for (M)-and (P)-atropisomer interconversion. Intramolecular interactions (e.g., steric, electronic) in the interconversion transition state lead to varying atropisomer interconversion barriers (ΔG ⧧ ), whose magnitudes depend on the identity and substitution pattern of the rings of the biaryl system. [/fig]
[fig] Figure 4: Conformationally/chemically dynamic features of the sotorasib scaffold (A−D), each demonstrating distinct activation energy barriers and NMR exchange kinetics. Restricted rotation about bond A (red) gives rise to configurationally stable (M)-and (P)-atropisomers. [/fig]
[fig] Figure 5: VT-NMR analysis: (A) Structure of compound (M)-3 and temperature dependence of the proton-decoupled 19 F NMR spectrum, monitoring fluorophenyl ring rotation from slow-to fast-exchange regimes. (B) Structure of compound 4 (mixture of atropisomers) and temperature dependence of 1 H NMR spectra of the azaquinazolinone H5 proton (blue), monitoring (M)-and (P)-atropisomer interconversion. Peak separations in the slow-exchange limit are indicated by Δv 0 . Both samples were dissolved at ∼5 mg/mL in DSMO-d 6 . NMR measurements were performed on a Bruker Avance III 500 MHz spectrometer with a BBFO probe. [/fig]
[fig] Figure 6: Time-course NMR analysis: (A) Structure of compound 5 (mixture of atropisomers). Fluorine atom (F6, magenta) was used to monitor (M)-and (P)-atropisomer interconversion (note: F6 shows a splitting of 5 J F−F = 8.6 Hz resulting from the proximal fluorophenol ring). (B) (M)-and (P)-atropisomer interconversion kinetics (T = 325 K). Atropisomer ratios (colored data points) were obtained by integration, normalizing the sum of the integrals to 100%. Solid lines show a fit to first-order kinetics with a half-life (t 1/2 ) of 334 min. (C) Proton-decoupled 19 F NMR spectra (Bruker AVIII-500, ∼5 mg/mL in DMSO-d 6 ) of F6 collected at 325 K (time points as indicated). [/fig]
[fig] Figure 7: Chiral chromatographic analysis: (A) Thermal racemization of (P)-6. (B) HPLC chromatograms illustrating the racemization (P)-6 as a function of time. (C) Plot of (P)-6 enantiomeric excess (derived from HPLC peak integration) as a function of time, following incubation at 280°C in diphenyl ether. [/fig]
[fig] Figure 8: Computational analysis: DFT analysis (B3LYP/6-31+G-(d,p)) of the torsional energy profile of the atropisomeric axis of compound 6. [/fig]
[fig] Figure 9: Preparation of rac-6 and approaches to prepare (M)-6. [/fig]
[table] Table 1: Favored Development Strategy As a Function of Atropisomer Interconversion Barrier (ΔG ⧧ ) a [/table]
[table] Table 2: Choosing an Experimental Technique for Configurational Stability Assessment [/table]
[table] Table 3: Chiral Acids That Form Complexes with Compound 6 [/table]
[table] Table 4: Process Efficiency Comparison (Chromatography vs Classical Resolution) [/table]
[table] Table 5: Initial Screening Experiments to Racemize (P)-6 a Experiments performed at 100 mg scale. b Chiral purity at t 0 = 99.6% ee. c Slurry-to-slurry transformation. d Experiment was performed in a plug-flow reactor. Adapted with permission from ref 3. Copyright 2022 The American Chemical Society. [/table]
[table] ■ AUTHOR INFORMATION: Corresponding Author Brian A. Lanman − Department of Medicinal Chemistry, Amgen Inc., Thousand Oaks, California 91320, United States; orcid.org/0000-0002-8768-7188; Email: blanman@ amgen.com Andrew T. Parsons − Department of Process Development, Drug Substance Technologies, Amgen Inc., Cambridge, Massachusetts 02142, United States; orcid.org/0000-0002-0320-0919 Stephan G. Zech − Lead Discovery and Characterization, Amgen Inc., Thousand Oaks, California 91320, United States Complete contact information is available at: https://pubs.acs.org/10.1021/acs.accounts.2c00479 [/table]
|
Safety, prosthesis wearing time and health-related quality of life of lower extremity bone-anchored prostheses using a press-fit titanium osseointegration implant: A prospective one-year follow-up cohort study
Item No. Recommendation Page No. Title and abstract 1 (a) Indicate the study's design with a commonly used term in the title or the abstract 1-2 (b) Provide in the abstract an informative and balanced summary of what was done and what was found 2
# Introduction
Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 3-4
Objectives 3 State specific objectives, including any prespecified hypotheses 4
# Methods
## Study design 4
Present key elements of study design early in the paper 4
Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, followup, and data collection 4-5
Participants 6 (a) Cohort study-Give the eligibility criteria, and the sources and methods of selection of participants.
## Describe methods of follow-up
Case-control study-Give the eligibility criteria, and the sources and methods of case ascertainment and control selection. Give the rationale for the choice of cases and controls
Cross-sectional study-Give the eligibility criteria, and the sources and methods of selection of participants *Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in cohort and cross-sectional studies.
Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at www.strobe-statement.org. |
miR-23b-3p Modulating Cytoprotective Autophagy and Glutamine Addiction in Sorafenib Resistant HepG2, a Hepatocellular Carcinoma Cell Line
# Introduction
Hepatocellular carcinoma (HCC) is the second most common malignancy increasing cancer-associated mortality worldwide [bib_ref] Global cancer statistics, Torre [/bib_ref] [bib_ref] Global trends and predictions in hepatocellular carcinoma mortality, Bertuccio [/bib_ref]. An age-standardized mortality rate of 6.8 for men and 5.1 for women per 100,000 people annually was reported in India [bib_ref] Epidemiology of hepatocellular carcinoma in India, Acharya [/bib_ref]. Among several etiologic factors, including infection from hepatitis B and C viruses, alcohol, and aflatoxin, HBV infection is predominant in HCC pathogenesis in India [bib_ref] Profile of hepatocellular carcinoma in India: An insight into the possible etiologic..., Sarin [/bib_ref]. Sorafenib is an FDA-approved first-line systemic therapy for advanced HCC. Sorafenib, a multikinase inhibitor for inhibiting Raf family kinases, vascular endothelial growth factor receptor (VEGFR), and platelet-derived growth factor receptor (PDGFR), resulting in enhanced angiogenesis, tumor proliferation, and apoptosis [bib_ref] Nexavar), a dual-action inhibitor that targets RAF/MEK/ERK pathway in tumor cells and..., Adnane [/bib_ref]. Most HCC patients show primary sorafenib resistance due to their genetic heterogeneity [bib_ref] Drug resistance in cancer-Searching for mechanisms, markers and therapeutic agents, O'connor [/bib_ref] or develop acquired sorafenib resistance from sustained drug exposure [bib_ref] Sorafenib blocks the RAF/MEK/ERK pathway, inhibits tumor angiogenesis, and induces tumor cell..., Liu [/bib_ref]. In acquired sorafenib resistance, several mechanisms are involved, including PI3K/Akt and JAK-STAT as compensatory pathways, hypoxia-inducible factors, transcription factors such as c-myc, and epithelial to mesenchymal transition factors including vimentin [bib_ref] Vimentin as a potential therapeutic target in sorafenib resistant HepG2, a HCC..., Makol [/bib_ref] , autophagy, microRNAs (such as miR-21; miR-122), tumor microenvironments, or tumor metabolism [bib_ref] New knowledge of the mechanisms of sorafenib resistance in liver cancer, Zhu [/bib_ref].
The death-promoting effect of Sorafenib-induced autophagy has been reported by recent studies [bib_ref] Mcl-1-dependent activation of Beclin 1 mediates autophagic cell death induced by sorafenib..., Tai [/bib_ref]. Although several studies have shown that autophagy has a protective role in HCC cells, the combination of autophagy inhibitors, including chloroquine, hydroxychloroquine, or autophagy-associated microRNAs, enhances the cytotoxic effects of sorafenib in HCC cells [bib_ref] Inhibition of autophagy potentiates the antitumor effect of the multikinase inhibitor sorafenib..., Shimizu [/bib_ref] [bib_ref] Inhibition of autophagy potentiates the proliferation inhibition activity of microRNA-7 in human..., Wang [/bib_ref]. However, contrasting studies also show the cytotoxic role of autophagy [bib_ref] Inhibition of Akt reverses the acquired resistance to sorafenib by switching protective..., Zhai [/bib_ref]. Thus, the exact role of autophagy in sorafenib-resistant HCC is still controversial, and understanding its molecular mechanism is required to curb resistance.
Moreover, emerging evidence demonstrates that sorafenib is also involved in mediating the metabolic reprogramming (enhanced glycolysis and glutamine addiction) of HCC cells [bib_ref] Hypoxia-mediated sorafenib resistance can be overcome by EF24 through Von Hippel-Lindau tumor..., Liang [/bib_ref] , which may develop drug resistance [bib_ref] Metabolic reprogramming: The emerging concept and associated therapeutic strategies, Yoshida [/bib_ref]. Exogenous glutamine is vital for cancer cells' survival; hence, cancer cells show "glutamine addiction" [bib_ref] Nutrition needs of mammalian cells in tissue culture, Eagle [/bib_ref]. In some human tumors, the upregulated expression of GLS1 (glutaminase) is associated with advanced disease stages [bib_ref] Kidney-type glutaminase (GLS1) is a biomarker for pathologic diagnosis and prognosis of..., Yu [/bib_ref]. BPTES (bis-2-(5-phenylacetamido1,3,4-thiadiazol-2-yl)ethyl sulfide), the chemical inhibitor of GLS1, is well known for its tumor suppressive role in different preclinical models. However, the role of glutaminase enhancing glutamine addiction in sorafenibresistant HCC is yet to be defined. Hence, understanding the molecular mechanism of glutamine addiction in sorafenib-resistant HCC may also improve HCC treatment.
MicroRNAs (miRNAs), small (18-25 nucleotides) non-coding RNA molecules that regulate their targeted genes at the post-transcriptional level, are also involved in sorafenib resistance. Although miR-122 is the highest expressed miRNA in the liver, there are only a few reports on miRNAs targeting signaling pathways to influence sorafenib resistance and its related pathways [bib_ref] Sorafenib response in hepatocellular carcinoma: MicroRNAs as tuning forks, Kanthaje [/bib_ref]. These circulatory miRNAs are known as specific regulators that mediate any signaling pathway by targeting their specific genes. However, the influence of differentially expressed miRNAs in the modulation of autophagy and glutamine addiction to mediate sorafenib resistance in HCC is yet to be evaluated.
Here, we explored the role of autophagy and glutamine addiction in sorafenib-resistant HepG2 cells, the cellular model of HCC. Furthermore, we determined the significance of differentially expressed miRNAs associated with autophagy and glutamine addiction to curb sorafenib resistance and, thus, introduced new approaches to clinical HCC management.
# Materials and methodology
## Chemicals
Minimum Essential Media (MEM) for cell culture and antibiotics were purchased from HiMedia (Chandigarh, India), Thermo Fisher Scientific (Waltham, MA, USA), and Sigma-Aldrich (St. Louis, MO, USA). Sorafenib was purchased from SignalChem Lifesciences Corporation (Canada). Chloroquine and Rapamycin were in lyophilized form and provided by the CYTO-ID ® Autophagy Detection Kit purchased from Enzo Life Sciences (Farmingdale, NY, USA). BPTES was purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). Sorafenib (stock 1 mM), rapamycin (stock 10 µM), and BPTES (stock 20 µM) were prepared in DMSO, whereas chloroquine (stock 5 mM) was prepared in deionized water.
## Cell lines
Hepatocellular carcinoma (HCC) is the primary cancer of the liver. The HepG2 cell line is a liver cancer cell line derived from well-differentiated HCC patients and carries the wild-type p53 expression. Therefore, it more closely mimics HCC. HepG2 cell lines were cultured in MEM media in the presence of required conditions. Sorafenib-resistant HepG2 cells were already established in our lab [bib_ref] Vimentin as a potential therapeutic target in sorafenib resistant HepG2, a HCC..., Makol [/bib_ref]. HepG2 parental cells were treated with minimal dosage inhibitory concentration (IC-20) of sorafenib. Then, the dosage of sorafenib was increased continuously so that HepG2 cells could survive or adapt to low to high dosages of sorafenib until IC-50 achieved more than twofold resistant cells compared to parental cells [bib_ref] Vimentin as a potential therapeutic target in sorafenib resistant HepG2, a HCC..., Makol [/bib_ref].
## -
Fold resistance = IC-50 value of resistant cells/IC-50 value of parental cells Cultured cells were routinely monitored under an inverted light microscope (Olympus CKX41, Tokyo, Japan) and imaged using Olympus cellSens software for morphological characteristics.
2.3. RNA Isolation and Its Quantification, cDNA Synthesis, and qRT-PCR Total RNA was extracted from parental and sorafenib-resistant HepG2 cells with Trizol reagent, as discussed in our previous study [bib_ref] Vimentin as a potential therapeutic target in sorafenib resistant HepG2, a HCC..., Makol [/bib_ref]. The relative mRNA level was calculated with the formula 2-∆∆Ct, where ∆Ct = (Ct target gene − Ct internal control) and ∆∆Ct = (Ct sorafenib resistant cells − Ct parental cells). GAPDH was used for normalization [fig_ref] Table 1: Primer sequences of genes [/fig_ref].
## Mtt assay
Both parental and sorafenib-resistant HepG2 cells were cultured with 5000 cells per well in a 96-well plate (Corning Inc., Corning, NY, USA) to check cell viability. After 24 h incubation, the required concentration of treatments was added to fresh media. After 24 h, sterile 20 µL MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazolium Bromide) was added to wells, and the plates were incubated for 4 h in a cell culture incubator at 37 - C with 5% CO 2 . Media along with MTT was removed, DMSO (200 µL) was added, and the plates were incubated for 15 min at RT in the dark to dissolve formazan crystals. The absorbance was measured at 570 nm with a microplate reader.
A dose response curve was plotted with % cell viability in the Y-axis and concentration of sorafenib or other chemical in the X-axis. Percentage of cell viability was measured by the formula:
- % Cell viability = ((Absorbance of test − Absorbance of Blank)/(Absorbance of control − Absorbance of Blank)) * 100 IC-50 values were then calculated from the dose response curve.
## Annexin v/propidium iodide (pi)
Parental and sorafenib-resistant cells were treated with the required concentration of chemicals for 24 h and analyzed for apoptosis with a commercial Apoptosis Kit (Invitrogen, Waltham, MA, USA) [bib_ref] A novel assay for apoptosis. Flow cytometric detection of phosphatidylserine expression on..., Vermes [/bib_ref]. Briefly, 1 × 10 6 cells were trypsinized, washed with PBS, and centrifuged. Then, 100 µL of annexin binding buffer (1×) and 0.2 µL of Annexin V-FITC followed by 0.1 µL of propidium iodide (PI) were added to each tube except for the Annexin and PI tubes. The samples were incubated at room temperature for 15 min in the dark and supplemented with 150 µL of annexin binding buffer (1×) prior to analysis. The cells were checked by a flow cytometer (FACS Canto, Becton Dickinson, CA, USA) under 488 nm for fluorescein detection and a filter >600 nm for PI detection.
## Green autophagy dye assay
CYTO-ID ® Autophagy Detection Kit (Enzo Life sciences) was used to detect the level of autophagy by flow cytometry in both parental and sorafenib-resistant HepG2 cells. The assay provides a rapid, quantitative approach to monitoring autophagic activity in vitro. The green fluorescent detection reagent becomes brightly fluorescent in autophagy vesicles. Flow cytometry calculated the level of green-emitting fluorescent probe corresponding to the number of green autophagy-stained cells.
## Next generation sequencing (ngs)
The total RNA isolated from parental and resistant cells of HepG2 was used for microRNA (miRNA) analysis by next generation sequencing (NGS) in Medgenome Labs Pvt. Ltd., Bangalore, India. Illumina HiSeq2500 was used to perform miRNA NGS, and the data were mapped to the HG19 human genome database and miRBase 20 to obtain a list of deregulated miRNAs.
# In silico analysis
In silico analysis was performed to determine microRNAs (miRNAs) as specific regulators targeting sorafenib-resistant pathways. Genes and their miRNA(s) targets were examined using three different databases, including miDRB (http://mirdb.org/, accessed on 16 May 2022) and miRWalk 2.0 predicted (http://zmf.umm.uni-heidelberg.de/apps/ zmf/mirwalk2/generetsys-self.html, accessed on 16 May 2022) and validated (http://zmf. umm.uni-heidelberg.de/apps/zmf/mirwalk2/miRpub.html, accessed on 16 May 2022) databases. Additionally, we employed a set of four databases (miRWalk, miRanda, RNA22, and TargetScan) from the miRwalk predicted database and only considered targeted miRs predicted by these four databases [bib_ref] Characterization of microRNA transcriptome in lung cancer by next-generation deep sequencing, Ma [/bib_ref]. The validated targeted miRNAs from miRwalk validated were considered [bib_ref] Practical Aspects of microRNA Target Prediction, Witkos [/bib_ref]. Moreover, these miRNA(s) were reverse-analyzed as specific regulators of their targeted genes.
## Transfection of mirna mimics and antimirs
Mimics are commercially synthesized small double-stranded RNAs that upregulate miRNA activity, whereas antimiRs are commercially synthesized small single-stranded RNA molecules that specifically bind to endogenous miRNA molecules and downregulate microRNA activity. MirVana miRNA mimics, antimiRs, and scrambled miRNA (negative control) were purchased from Thermo Fisher Scientific Inc, Waltham, MA, USA. Transfection was performed using Lipofectamine RNAiMAX (Thermo Fisher Scientific Inc., Waltham, MA, USA) according to the manufacturer's protocol. Briefly, Lipofectamine RNAiMAX reagent (3 µL) was added to 50 µL of MEM (without antibiotics). Simultaneously, 25 nM of the mimic/antimiR (siRNA) was added to 50 µL of MEM (without antibiotics). Lipofectamine RNAiMAX reagent and siRNA were added in a 1:1 ratio and incubated at RT for 5 min. For transfection, 2 × 10 5 cells were seeded into a 24-well plate to obtain 60-80% confluence, and then 50 µL of the prepared mixture was added. Transfection was confirmed by qRT-PCR.
## Mirna isolation; polyadenylation, cdna synthesis poly a tailed mirna, qrt-pcr for mirnas
The extracellular (as miRNAs secreted in cell culture media) miRNAs were isolated by the Trizol reagent, as discussed in our previous study [bib_ref] Vimentin as a potential therapeutic target in sorafenib resistant HepG2, a HCC..., Makol [/bib_ref]. Isolated miRNA quantified by an infinite M200PRO microplate reader (Tecan, Männedorf city, Switzerland). The absorbance 260/280 of >1.8 was considered an acceptable value. Each miRNA molecule was polyadenylated using the miRNA cDNA synthesis kit (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer's instructions. Finally, the prepared cDNA was stored at −20 - C until further use. Then, the extracellular (as secreted miRNAs in cell culture media) miRNA expression levels were quantified by qRT-PCR performed by Lightcycler ® 480 (Roche, Switzerland) using SYBR Green I master mix (Thermo Fisher Scientific Inc., Waltham, MA, USA). qRT-PCR was performed by Lightcycler ® 480 (Roche, Switzerland) using SYBR Green I master mix (Thermo Fisher Scientific Inc., Waltham, MA, USA). The forward primer of miRNA was the miRNA sequence and synthesized by Sigma-Aldrich. The reverse primer is the universal reverse primer (URP) provided by the kit (Agilent Technologies, Santa Clara, CA, USA). The miRNA expression was normalized using U6, a small nuclear RNA, acting as the internal control. A total of 1 µg of total RNA was used for intracellular miRNA assessments and 150 ng of miRNA for extracellular. The reaction mixture contained cDNA (1 µL), 0.3125 µM primer (forward and reverse), 1X SYBR Green Mix, and nuclease-free water for volume makeup (up to 10 µL). In all cases, initial denaturation occurred at 95 - C, for 5 min, and final melting (5 - C + annealing temperature; 1 min).
The expression levels of miRNA in test vs. control were determined by fold change using the formula (2 −∆∆Ct ) [bib_ref] Characterization of microRNA transcriptome in lung cancer by next-generation deep sequencing, Ma [/bib_ref].
Primer sequence of miRNAs; FP: Forward Primer; RP: Reverse Primer [fig_ref] Table 2: Primer Sequences of microRNA [/fig_ref].
# Statistical analysis
The required experiments were conducted three times independently. The statistical calculations were performed using GraphPad Prism 6.01 (GraphPad Software Inc., San Diego, CA, USA) software for Windows. ANOVA (for multiple comparisons), Student's t-test, and Mann-Whitney U test (two-group comparison) were used as required. The experimental data were shown as mean ± SEM, and p < 0.05 was considered statistically significant.
# Results
## Autophagy and glutamine addiction as cytoprotective in sorafenib-resistant hepg2 cells
Both HepG2 parental HepG2 (P) and HepG2 sorafenib resistant HepG2 (R) cells [bib_ref] Vimentin as a potential therapeutic target in sorafenib resistant HepG2, a HCC..., Makol [/bib_ref] were stored at −80 - C and were cultured for required experiments (Supplementary [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. The autophagy-related genes (ATGs) expression was checked; in HepG2 (R) cells, the expression of ATG7 was upregulated with fold change 1.513 ± 0.2684, whereas the levels of Beclin1, LC3II, and ATG12 significantly increased with fold change 1.678 ± 0.05773; 1.733 ± 0.08665; 1.686 ± 0.06741, respectively, compared with HepG2 (P) cells [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref] , Supplementary [fig_ref] Figure 2: MTT assay showing [/fig_ref]. Both HepG2 parental HepG2 (P) and HepG2 sorafenib resistant HepG2 (R) cells [bib_ref] Vimentin as a potential therapeutic target in sorafenib resistant HepG2, a HCC..., Makol [/bib_ref] were stored at −80 °C and were cultured for required experiments (Supplementary [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. The autophagy-related genes (ATGs) expression was checked; in HepG2 (R) cells, the expression of ATG7 was upregulated with fold change 1.513 ± 0.2684, whereas the levels of Beclin1, LC3II, and ATG12 significantly increased with fold change 1.678 ± 0.05773; 1.733 ± 0.08665; 1.686 ± 0.06741, respectively, compared with HepG2 (P) cells [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. Furthermore, with increasing concentrations (50, 80, and 100 nM) of rapamycin (autophagy initiator), the cell viability of HepG2 (R) was high. By contrast, increasing concentrations (50, 80, and 100 µM) of chloroquine (autophagy inhibitor) significantly decreased (p = 0.0034 at 80 µM (IC50) the cell viability of resistant cells. Both sorafenib (4 µM; IC50) and chloroquine (80 µM; IC50) also significantly (p = 0.0007) reduced the cell viability of HepG2 (R) cells [fig_ref] Figure 2: MTT assay showing [/fig_ref]. Afterward, the presence of chloroquine showed an increased percentage of apoptosis 50.6 ± 0.9815 in HepG2 (R) [fig_ref] Figure 2: MTT assay showing [/fig_ref] , suggesting the cytoprotective role of autophagy. Furthermore, with increasing concentrations (50, 80, and 100 nM) of rapamycin (autophagy initiator), the cell viability of HepG2 (R) was high. By contrast, increasing concentrations (50, 80, and 100 µM) of chloroquine (autophagy inhibitor) significantly decreased (p = 0.0034 at 80 µM (IC 50 ) the cell viability of resistant cells. Both sorafenib (4 µM; IC 50 ) and chloroquine (80 µM; IC 50 ) also significantly (p = 0.0007) reduced the cell viability of HepG2 (R) cells [fig_ref] Figure 2: MTT assay showing [/fig_ref]. Afterward, the presence of chloroquine showed an increased percentage of apoptosis 50.6 ± 0.9815 in HepG2 (R) [fig_ref] Figure 2: MTT assay showing [/fig_ref] , suggesting the cytoprotective role of autophagy.
Initially, we checked the expression of the oncogenic gene c-myc, which was significantly (p = 0.0011) higher in HepG2 (R) cells than in HepG2 (P) cells with a fold change of 2.533 ± 0.1822 (Supplementary [fig_ref] Figure 3: GLS1 [/fig_ref]. The increase in sorafenib concentrations (2 µM and 4 µM) significantly (p = 0.0015) enhanced the expression of c-myc with 2.267 ± 0.2942 and 4.8 ± 0.2309-fold changes respectively (Supplementary [fig_ref] Figure 3: GLS1 [/fig_ref]. The expression of c-myc with glutaminolysis through targeting its vital enzyme glutaminase (GLS1) correlated with HepG2 (R) cells. The GLS1 expression was significantly (p = 0.0008) higher in HepG2 (R) than HepG2 (P), with a fold change of 2.797 ± 0.196 [fig_ref] Figure 3: GLS1 [/fig_ref]. cantly (p = 0.0011) higher in HepG2 (R) cells than in HepG2 (P) cells with a fold change of 2.533 ± 0.1822 (Supplementary [fig_ref] Figure 3: GLS1 [/fig_ref]. The increase in sorafenib concentrations (2 µM and 4 µM) significantly (p = 0.0015) enhanced the expression of c-myc with 2.267 ± 0.2942 and 4.8 ± 0.2309-fold changes respectively (Supplementary [fig_ref] Figure 3: GLS1 [/fig_ref]. The expression of c-myc with glutaminolysis through targeting its vital enzyme glutaminase (GLS1) correlated with HepG2 (R) cells. The GLS1 expression was significantly (p = 0.0008) higher in HepG2 (R) than HepG2 (P), with a fold change of 2.797 ± 0.196 [fig_ref] Figure 3: GLS1 [/fig_ref]. Both HepG2 (P) and HepG2 (R) cells were cultured in MEM media without exogenous glutamine. Inverted microscopy has shown elongated morphology with a decreased proliferation rate in both HepG2 cells. However, the proliferation rate was lower in HepG2 (R) cells than in HepG2 (P) cells [fig_ref] Figure 4: MTT assay showing cell viability [/fig_ref]. The MTT assay also showed resistant cells without exogenous glutamine [R (-glut)] that exhibited significantly (p < 0.0001) reduced cell viability (17.52 ± 1.613) compared to resistant cells with glutamine (47.27 ± 1.295) [fig_ref] Figure 4: MTT assay showing cell viability [/fig_ref]. Both HepG2 (P) and HepG2 (R) cells were cultured in MEM media without exogenous glutamine. Inverted microscopy has shown elongated morphology with a decreased proliferation rate in both HepG2 cells. However, the proliferation rate was lower in HepG2 (R) cells than in HepG2 (P) cells [fig_ref] Figure 4: MTT assay showing cell viability [/fig_ref]. The MTT assay also showed resistant cells without exogenous glutamine [R (-glut)] that exhibited significantly (p < 0.0001) reduced cell viability (17.52 ± 1.613) compared to resistant cells with glutamine (47.27 ± 1.295) [fig_ref] Figure 4: MTT assay showing cell viability [/fig_ref].
The green autophagy dye-stained vesicles (markers of autolysosomes and autophagic compartments) were also examined using a CYTO-ID Autophagy detection kit. The increased autophagic flux was directly proportional to the increased number of green autophagy dye-stained vesicles detected through green emitting fluorescence using flow cytometry. The number of green autophagy dye-stained vesicles was significantly (p = 0.0153) higher in resistant cells than in parental cells. However, in the absence of exogenous glutamine sorafenib-resistant cells [R (-glut)], the number of green autophagy dye0stained vesicles significantly (p = 0.0321) increased with a difference of 224 ± 69.42 compared to the presence of glutamine.
We performed an MTT assay using BPTES (glutaminase inhibitor), where the cell viability of BPTES-treated sorafenib-resistant cells was significantly (p < 0.0001) decreased compared to non-treated resistant cells . The BPTES (300 nM; IC 50 ) significantly (p = 0.0010) increased cell apoptosis (10.57 ± 1.216-fold higher) of resistant cells than in non-treated HepG2 cells .
## Autophagy and glutamine addiction-specific micrornas and their target genes in sorafenib-resistant hcc cell lines
The NGS data revealed 157 upregulated and 86 downregulated miRNAs in sorafenibresistant compared to parental cells [fig_ref] Figure 7: Next generation sequencing showing a heat map of differentially expressed microRNAs [/fig_ref]. The in silico analysis using miRDB, the miRNAs target prediction online database, revealed the miR-23b-3p as autophagy-specific miRNA with a target score of 82 for ATG12, among other autophagy genes (ATG7, Beclin1, LC3II, and ATG 12) (Supplementary Material [fig_ref] Figure 4: MTT assay showing cell viability [/fig_ref]. The higher target score specifies the real prediction of the selected miRNA (http://mirdb.org/faq.html, accessed on 16 May 2022). The miR-23b-3p was also observed to target GLS1 (the vital gene of glutamine addiction) with a target score of 90 (Supplementary Material. Furthermore, miRWalk, the online database including both predicted and validated targeted miRNAs to genes and combined with other databases (miRWalk, miRanda, RNA22, and Targetscan), also confirmed miR-23b-3p as miRNA targeting ATG12 (Supplementary Material . Similarly, for glutamine addiction using a combination of different databases (miRWalk, miRanda, RNA22, and Targetscan), miR-23b-3p was selected as miRNA targeting GLS1 [(Supplementary Material [fig_ref] Figure 7: Next generation sequencing showing a heat map of differentially expressed microRNAs [/fig_ref]. Furthermore, in silico analysis using miRDB, miR-Walk predicted and validated also revealed miR-23b-3p as a common miRNA targeting autophagy through ATG12 and glutamine addiction through GLS1 [fig_ref] Figure 7: Next generation sequencing showing a heat map of differentially expressed microRNAs [/fig_ref]. The green autophagy dye-stained vesicles (markers of autolysosomes and autophagic compartments) were also examined using a CYTO-ID Autophagy detection kit. The increased autophagic flux was directly proportional to the increased number of green autophagy dye-stained vesicles detected through green emitting fluorescence using flow cytometry. The number of green autophagy dye-stained vesicles was significantly (p = 0.0153) higher in resistant cells than in parental cells. However, in the absence of exogenous glutamine sorafenib-resistant cells [R (-glut)], the number of green autophagy dye0stained vesicles significantly (p = 0.0321) increased with a difference of 224 ± 69.42 compared to the presence of glutamine. The expression analysis of ATG12 and GLS1 was also evaluated, where the presence of increasing sorafenib concentrations ranging from 2 to 6 µM and the expression of ATG12 significantly (p = 0.0226) increased with a fold change of 4.59 ± 0.6566. Similarly, the expression of GLS1 also significantly (p = 0.0014) increased with a fold change of 6.13 ± 0.4131.
The extracellular expression of miR-23b-3p (miR-23) was evaluated in the presence of increasing concentrations of sorafenib (4 and 6 µM). The increasing concentration of sorafenib decreased the miR-23 extracellular expression. Higher concentrations of sorafenib (6 µM) significantly (p = 0.0013) decreased the extracellular expression of miR-23 with a fold change of 0.3027 ± 0.05773 compared to resistant cells without sorafenib treatment [fig_ref] Figure 9: The extracellular expression analysis of miR-23b-3p [/fig_ref].
## Functional analysis of mir-23b-3p
Mimics, as well as antimiRs of miR-23b-3p, were purchased from Thermo Fisher Scientific, Waltham, MA, USA. The transfection of both mimics and antimiRs of miR-23b-3p was executed in HepG2 cells to analyze the functional perspective of microRNA (Supplementary Material [fig_ref] Figure 9: The extracellular expression analysis of miR-23b-3p [/fig_ref]. Hence, we selected mimic and antimiR concentrations of 30 nM for their transfection assays. We performed an MTT assay using BPTES (glutaminase inhibitor), where the cell viability of BPTES-treated sorafenib-resistant cells was significantly (p < 0.0001) decreased compared to non-treated resistant cells . The BPTES (300 nM; IC50) significantly (p = 0.0010) increased cell apoptosis (10.57 ± 1.216-fold higher) of resistant cells than in non-treated HepG2 cells . The autophagy and glutamine-specific miR-23b-3p expression was evaluated in the presence of its enhancer (mimic) or inhibitor (antimiR) in HepG2 (P) cells where scrambled miRNA was used as the negative control (NC). The transfection of mimic (miR-23m) significantly (p = 0.0127) increased miR-23b-3p (miR-23) expression compared to the negative control (NC). However, the transfection of antimiR (miR-23a) significantly (p = 0.0234) decreased miR-23b-3p expression compared to the negative control (Supplementary Material [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. Furthermore, the stability of the transfected mimic (30 nM) and antimiR was checked at periods ranging from 24 to 48 to 72 h. The transfection of both mimic and antimiR was found to be significantly (p = 0.0282) stable up to 72 h (Supplementary [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. The expression analysis of ATG12 and GLS1 was also evaluated, where the presence of increasing sorafenib concentrations ranging from 2 to 6 µM and the expression of ATG12 significantly (p = 0.0226) increased with a fold change of 4.59 ± 0.6566. Similarly, the expression of GLS1 also significantly (p = 0.0014) increased with a fold change of 6.13 ± 0.4131. The effects of mimics and antimiRs on the IC-50 value of sorafenib were evaluated by the transfection of both miR-23m and miR-23a in HepG2 (R) cells. The transfection of miR-23a enhanced the proliferation rate, whereas the presence of miR-23m decreased the proliferation rate in HepG2 (R) cells compared to non-treated sorafenib-resistant cells [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. The significance of miR-23b-3p mimic (miR-23m) and antimiR of miR-23b-3p (miR-23a) was validated in both parental and sorafenib-resistant HepG2 cells by the Annexin/PI assay. The presence of miR-23b-3p mimic (miR-23m) in sorafenib-resistant HepG2 cells significantly (p = 0.0224) increased cell apoptosis percentage with a mean difference of 6.367 ± 1.761 compared to sorafenib-resistant cells without mimic transfection [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. The extracellular expression of miR-23b-3p (miR-23) was evaluated in the presence of increasing concentrations of sorafenib (4 and 6 µM). The increasing concentration of sorafenib decreased the miR-23 extracellular expression. Higher concentrations of sorafenib (6 µM) significantly (p = 0.0013) decreased the extracellular expression of miR-23 with a fold change of 0.3027 ± 0.05773 compared to resistant cells without sorafenib treatment [fig_ref] Figure 9: The extracellular expression analysis of miR-23b-3p [/fig_ref].
## Functional analysis of mir-23b-3p
Mimics, as well as antimiRs of miR-23b-3p, were purchased from Thermo Fisher Scientific, Waltham, MA, USA. The transfection of both mimics and antimiRs of miR-23b-3p was executed in HepG2 cells to analyze the functional perspective of microRNA (Supplementary Material. To standardize the maximum efficient transfection concen- The antimiR (miR-23a) increased ATG12 expression, whereas the presence of miR-23b-3p mimic (miR-23m) (30 nM) significantly (p = 0.0031) decreased the expression of ATG12 in resistant HepG2 cells compared to non-transfected resistant cells [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. Similarly, the antimiR (miR-23a) also increased GLS1 expression, whereas GLS1 expression was also significantly (p = 0.0342) decreased in sorafenib-resistant HepG2 cells transfected with miR-23b-3p mimic (miR-23m) (30 nM) compared to resistant cells without miR-23m transfection [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. The effects of mimics and antimiRs on the IC-50 value of sorafenib w by the transfection of both miR-23m and miR-23a in HepG2 (R) cells. The t miR-23a enhanced the proliferation rate, whereas the presence of miR-23m proliferation rate in HepG2 (R) cells compared to non-treated sorafenib- [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref]. The significance of miR-23b-3p mimic (miR-23m) and antimi 3p (miR-23a) was validated in both parental and sorafenib-resistant HepG Annexin/PI assay. The presence of miR-23b-3p mimic (miR-23m) in soraf
# Discussion
HCC is one of the predominant causes of cancer-associated deaths [bib_ref] Global trends and predictions in hepatocellular carcinoma mortality, Bertuccio [/bib_ref]. The FDA-approved oral drug sorafenib, a multikinase inhibitor, is administered for advanced HCC
# Discussion
HCC is one of the predominant causes of cancer-associated deaths [bib_ref] Global trends and predictions in hepatocellular carcinoma mortality, Bertuccio [/bib_ref]. The FDAapproved oral drug sorafenib, a multikinase inhibitor, is administered for advanced HCC [bib_ref] Sorafenib in advanced hepatocellular carcinoma, Llovet [/bib_ref] [bib_ref] Efficacy and safety of sorafenib in patients in the Asia-Pacific region with..., Cheng [/bib_ref]. In India, the sorafenib treatment is easily tolerated and increases overall survival rate by three months [bib_ref] Advanced hepatocellular carcinoma: A regional cancer center experience of 48 cases, Chaudhuri [/bib_ref]. However, acquired resistance limits the efficiency of sorafenib in HCC patients [bib_ref] Sorafenib blocks the RAF/MEK/ERK pathway, inhibits tumor angiogenesis, and induces tumor cell..., Liu [/bib_ref]. The main factors for acquired resistance against sorafenib are autophagy, altered metabolic reprogramming, and differentially expressed microR-NAs [bib_ref] New knowledge of the mechanisms of sorafenib resistance in liver cancer, Zhu [/bib_ref]. HepG2 cells, the cellular model of HCC, are used to find cytotoxicity levels and cellular and drug metabolism. Initially, we investigated the significance of autophagy in sorafenib-resistant HepG2 cells. The higher expression of autophagy-related genes (ATGs): ATG7, ATG12, Beclin1, and LC3II in HepG2 (R) cells compared to HepG2 (P) cells indicated the protective role of autophagy in sorafenib-resistant HepG2 cells. Moreover, decreased cell proliferation and increased cell apoptosis of sorafenib-resistant HepG2 cells in the presence of chloroquine (the autophagy inhibitor) also confirmed the protective role of autophagy in sorafenib-resistant HCC. Similarly, studies performed by Manov et al. [bib_ref] Inhibition of doxorubicin-induced autophagy in hepatocellular carcinoma Hep3B cells by sorafenib-the role..., Manov [/bib_ref] and Shimizu et al. [bib_ref] Inhibition of autophagy potentiates the antitumor effect of the multikinase inhibitor sorafenib..., Shimizu [/bib_ref] also discussed the protective role of autophagy in HCC. By contrast, the cytotoxic role of autophagy was discussed in HCC (only) by Tai et al. [bib_ref] Mcl-1-dependent activation of Beclin 1 mediates autophagic cell death induced by sorafenib..., Tai [/bib_ref] and Zhai et al. [bib_ref] Inhibition of Akt reverses the acquired resistance to sorafenib by switching protective..., Zhai [/bib_ref]. However, our study is the first to demonstrate the cytoprotective role of autophagy in HepG2 (R) cells.
Altered tumor metabolism, including high glycolysis and glutaminolysis, is also crucial in cancer resistance. We demonstrated a higher glutaminase expression (GLS1) in HepG2 (R) cells. In ovarian cancer progression, the higher expression of GLS1, the vital gene of glutaminolysis, has also been discussed [bib_ref] miR-145 inhibits glutamine metabolism through c-myc/GLS1 pathways in ovarian cancer cells, Li [/bib_ref]. In the present study, the absence of exogenous glutamine or the presence of BPTES (glutaminase inhibitor) reduced cell viability and increased cell apoptosis of sorafenib-resistant cells, validating the cytoprotective significance of glutamine addiction in HepG2 (R) cells. The CYTO-ID ® Autophagy Detection Kit evaluated higher numbers of autophagy-stained vesicles, suggesting that detrimental glutamine might enhance autophagy pathways in sorafenib-resistant HCC, which is similar to a study performed by Zhu et al. [bib_ref] L-Glutamine deprivation induces autophagy and alters the mTOR and MAPK signaling pathways..., Zhu [/bib_ref].
We also correlated the importance of differentially expressed microRNAs (miRNAs) in association with autophagy and glutamine addiction in sorafenib-resistant HCC. A recent study showed that the SNGH16 regulates autophagy through miR-23b-3p and enhances sorafenib resistance [bib_ref] Gong, L. SNGH16 regulates cell autophagy to promote Sorafenib Resistance through suppressing..., Jing [/bib_ref]. Our next generation sequencing (NGS) and in silico analyses evaluated miR-23b-3p as specific miRNA targeting ATG12 (autophagy) and GLS1 (glutamine addiction). The association between miR-23b-3p and ATG12 has been discussed in gastric cancer cells [bib_ref] miR-23b-3p regulates the chemoresistance of gastric cancer cells by targeting ATG12 and..., An [/bib_ref] ; metabolic reprogramming has also been described in osteosarcoma [bib_ref] miR-23b-3p suppressing PGC1α promotes proliferation through reprogramming metabolism in osteosarcoma, Zhu [/bib_ref]. However, the influence of miR-23b-3p's association with ATG12 and autophagy or GLS1 and glutamine addiction on sorafenib resistance in HCC has not been discussed. Our NGS data and extracellular expression demonstrated downregulated levels of miR-23b-3p in HepG2 (R) cells, suggesting the tumor-suppressing role of miR-23b-3p. Similarly, the downregulated expression of miR-23b-3p as a predictor has been discussed in HCC progression [bib_ref] Downregulated miR-23b-3p expression acts as a predictor of hepatocellular carcinoma progression: A..., He [/bib_ref]. Furthermore, the association between sorafenib treatments in HCC patients and the longitudinal variation of miR-23b-3p expression has been discussed in recent research. However, for the first time, our study highlights the significance of tumorsuppressing miR-23b-3p in modulating cytoprotective autophagy through ATG12 and glutamine addiction through GLS1 in sorafenib-resistant HCC.
# Conclusions
Our study determined the cytoprotective role of both autophagy and glutamine addiction in HepG2 (R) cells. The NGS and in silico analysis confirmed miR-23b-3p as differentially expressed miRNA targeting cytoprotective autophagy through ATG12 and glutamine addiction through GLS1 in sorafenib-resistant HepG2 cells. Transfection assays validated the tumor-suppressing effect of miR-23b-3p and modulated the expression of both ATG12 and GLS1 in sorafenib-resistant HepG2 cells, curtailing sorafenib resistance in HCC.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/genes13081375/s1, [fig_ref] Figure 1: Bar graph representation of expression analysis of autophagy genes [/fig_ref] : [fig_ref] Figure 4: MTT assay showing cell viability [/fig_ref] : In-silico analysis using miRDB database identifying miRNAs as target of ATG12 (Autophagy),: In-silico analysis using miRDB database identifying miRNAs as target of GLS (Glutamine addiction), : In-silico analysis using miRwalk database identifying miRNAs as target of ATG12 (Autophagy), [fig_ref] Figure 7: Next generation sequencing showing a heat map of differentially expressed microRNAs [/fig_ref] : In-silico analysis using miRwalk database identifying miRNAs as target of GLS (Glutamine addiction),: Representation of miR-23b-3p (miR-23) transfection assay in HepG2 cells, Author Contributions: R.K. and S.K. participated in the study concept and design, acquisition of data, analysis, and interpretation of data, and drafting of the manuscript. S.T. and R.K.D. critically revised the manuscript. A.C. participated in the study supervision, data interpretation, and critical review of the manuscript for important intellectual content. All authors have read and agreed to the published version of the manuscript.
Funding: ICMR, New Delhi.
## Institutional review board statement: not applicable.
Informed Consent Statement: Our study did not involve any samples from laboratory animals and did not require the use of radiochemicals or hazardous micro-organisms. We further declare that our study was performed in accordance with standard ethical guidelines published by competent authorities, which are available at http://icmr.nic.in/ethical_guidelines.pdf, accessed on 16 May 2022. Hence, our study was ethically justified. Data Availability Statement: Not applicable.
[fig] Figure 1: Bar graph representation of expression analysis of autophagy genes (Parental: Parental HepG2 cells; Resistant: Sorafenib resistant HepG2 cells) (ns: non-significant, p > 0.05; ** p ≤ 0.01; *** p ≤ 0.001). [/fig]
[fig] Figure 2: MTT assay showing (A) Percentage of cell viability in sorafenib resistant treated with different concentrations [(Rapa: Rapamycin (nM); Chl: Chloroquine (µM)] (B) Percentage of cell viability in presence of combination of sorafenib (Sof; IC50: 4 µM) and chloroquine (Chl; IC50: 80 µM) (C) Annexin V/PI assay showing cell apoptosis (i) Parental HepG2 cells (ii) sorafenib resistant [/fig]
[fig] Figure 3: GLS1 (glutaminase) expression analysis in parental and sorafenib resistant HepG2 cells (*** p ≤ 0.001). [/fig]
[fig] Genes 2022 ,: 13, x FOR PEER REVIEW 9 of 22 [/fig]
[fig] Figure 4: MTT assay showing cell viability (A) Morphological analysis (B) Bar graph representation of cell viability percentage in Resistant: sorafenib resistant HepG2 cells; R(-glut): sorafenib resistant HepG2 cells without exogenous glutamine (*** p ≤ 0.001). [/fig]
[fig] Figure 5: (A) Flow cytometry analysis of autophagy dye stained vesicles in parental and sorafenib resistant HepG2 cells (B) Bar graph representation of number of autophagy stained vesicles in HepG2 cells [R(-glut); sorafenib resistant HepG2 cells without exogenous glutamine] (* p ≤ 0.05). [/fig]
[fig] Figure 6 22, Figure 7, Figure 7: The cell viability (%) in parental and sorafenib resistant HepG2 cells by MTT assay in presence of increasing concentration of BPTES (nM) (A) 96-well plate (B) Bar graph representation of cell viability (C) Flow cytometry showing cell apoptosis (%) by Annexin V/PI analysis. (D) Bar graph representation of percentage of apoptotic cells in presence of BPTES (300 nM) in sorafenib resistant HepG2 cells (*** p ≤ 0.001). (A) The cell viability (%) in parental and sorafenib-resistant HepG2 cells by MTT assay in the presence of increasing concentrations of BPTES (nM). (B) Bar graph representation of cell viability. (C) Flow cytometry showing cell apoptosis (%) by Annexin V/PI analysis. (D) Bar graph representation of apoptotic cell percentage in the presence of BPTES (300 nM) in sorafenib-resistant HepG2 cells. Genes 2022, 13, x FOR PEER REVIEW 13 of ContNext generation sequencing showing a heat map of differentially expressed microRNAs (miRNAs) in (A) Parental HepG2 cells; (B) sorafenib-resistant HepG2 cells; (C) differentially expressed miR-23b-3p in HepG2 (R) compared to HepG2 (P) by NGS; (D) Venn diagram representing commonly identified miR-23b-3p from three different databases targeting both autophagy and glutamine addiction, i.e., autophagy and glutamine addiction-specific microRNA (miR-23b-3p). [/fig]
[fig] Figure 7: Next generation sequencing showing a heat map of differentially expressed microRNAs (miRNAs) in (A) Parental HepG2 cells; (B) sorafenib-resistant HepG2 cells; (C) differentially expressed miR-23b-3p in HepG2 (R) compared to HepG2 (P) by NGS; (D) Venn diagram representing commonly identified miR-23b-3p from three different databases targeting both autophagy and glutamine addiction, i.e., autophagy and glutamine addiction-specific microRNA (miR-23b-3p). [/fig]
[fig] Figure 8: (A) ATG12 expression analysis (B) GLS1 expression analysis in parental and sorafenib resistant HepG2 cells [R; sorafenib resistant HepG2 cells; Sof: sorafenib with 2 µM; 4 µM; 6 µM] (* p ≤ 0.05); (** p ≤ 0.01). [/fig]
[fig] Figure 9: The extracellular expression analysis of miR-23b-3p (miR-23) in parental and sorafenib resistant HepG2 cells (** p ≤ 0.01). [/fig]
[table] Table 1: Primer sequences of genes. [/table]
[table] Table 2: Primer Sequences of microRNA. [/table]
|
Use of noninvasive positive pressure ventilation during pregnancy: Case series
Mechanical ventilation is commonly required in critically ill pregnant patients, requiring ICU admission, with higher morbidity and mortality related to airway management. Alternatively, noninvasive positive pressure ventilation (NIPPV) is increasingly used to treat nonpregnant patients. Pregnancy has been a contraindication to its use. We would like to report a case series of successful use of NIPPV in pregnancy.NIPPV is increasingly used to treat hypoxemic respiratory failure. It has rarely been used during pregnancy. On the other hand, acute respiratory failure (ARF) remains a leading cause of ICU admission in obstetric patients. The use of NIPPV in managing ARF in pregnant patients was not investigated. We report the outcome of treatment with NIPPV of four sickle cell disease pregnant patients with ARF caused by acute chest syndrome. Median APACHE II score for the four cases was 27. Intubation was avoided in all cases. None had aspiration. Mean duration of NIPPV was 40 h with ICU discharge after a mean of 4 days.
A Case Report cute respiratory failure (ARF) occurs more frequently in pregnant than nonpregnant patients, requiring admission to intensive care unit (ICU), with higher morbidity and mortality. [bib_ref] Acute respiratory distress syndrome in pregnancy and the puerperium: Causes, courses and..., Catanzarite [/bib_ref] [bib_ref] Critical illness in pregnancy: An analysis of 20 patients admitted to a..., Collop [/bib_ref] Failed intubation in this population carries mortality risk eight times higher than in nonpregnant patients. [bib_ref] Failed tracheal intubation, King [/bib_ref] Several factors have been identified that may contribute to the higher morbidity and mortality associated with airway management in the pregnant patient. [bib_ref] Toward reducing maternal mortality: The problem airway in obstetrics, Rassmussen [/bib_ref] To the best of our knowledge, noninvasive positive pressure ventilation (NIPPV) has only been used during pregnancy in non-ICU setting. [bib_ref] Normal pregnancy in primary alveolar hypoventilation treated with nocturnal nasal intermittent positive..., Pieters [/bib_ref] [bib_ref] Obstructive sleep apnea during pregnancy resulting in pulmonary hypertension, Lewis [/bib_ref] It has been increasingly used to treat ARF in a wide variety of conditions with better outcome compared to invasive ventilation. [bib_ref] A randomized, prospective evaluation of noninvasive ventilation for acute respiratory failure, Martin [/bib_ref] Some have suggested, based on experience, that moderate cases of acute chest syndrome (ACS), 81% of whom will require mechanical ventilation, may respond to noninvasive positive pressure ventilation. [bib_ref] Causes and outcomes of the acute chest syndrome in sickle cell disease...., Vichinsky [/bib_ref] Because of the frequency of SCD in our area and the frequent admission of pregnant SCD patients with ACS to our ICU, we decided to look at the use of NIPPV in these patients. Hospital research committee approved our proposal and waived the requirement of consent. In this case series, we report the initial four cases of pregnant patients with ACS causing ARF who were successfully and safely managed by NIPPV in the ICU.
## Patients
Four consecutive pregnant patients with ACS and ARF admitted to the ICU from February 2003 until May 2004 were included. All patients were young adults with only medical history of sickle cell disease. None had history suggestive of obstructive sleep apnea. ARF was defined by severe hypoxemia (i.e., PaO 2 / fraction of inspired oxygen [FiO 2 ] ratio less than 200) with respiratory distress (respiratory rate 35 breaths/ min or more). The diagnosis of ACS was based on the presence of a new infiltrate on chest Xray associated with one or more of new symptoms like fever, cough, sputum production, dyspnea or hypoxia. [bib_ref] Pulmonary manifestations of sickle cell disease, Siddiqui [/bib_ref] NIPPV was the first line mechanical ventilatory support to be offered if FiO 2 supplement through nonrebreather face mask failed to maintain oxygen saturation above 92% on the monitor in an awake and cooperative patient.
## Treatment
A uniform standardized treatment protocol was applied for all patients. Symptomatic treatment included IV hydration with hypotonic fluids to achieve euvolemic state, increasing oxygen administration to keep O 2 saturation above 95% on the monitor, patient-controlled analgesia and folic acid. Antibiotic (a combination of a macrolide and a betalactam) was started immediately after blood and urine samples were taken for culture. All patients received early exchange transfusion, before getting initial Hb S level upon admission. Target was to keep Hb S below 30%. Deep venous thrombosis prophylaxis was started in all patients. All included patients were kept on NIPPV through face mask using CPAP with pressure support to keep tidal volume at around 300-350 ml and a respiratory rate below 35/min with an oxygen saturation above 92% throughout the NIPPV trial. None of the patients had an ECHO done as the clinical impression was of ACS and none had history suggestive of other diagnostic possibility. All patients were closely followed up by an obstetrician on regular basis. Indication for intubation was a poor ABG in the first hour after the start of NIPPV. NIPPV was weaned off when the respiratory rate and oxygen requirement decreased, along with the achievement of ability to maintain oxygen saturation above 95% on maximum oxygen requirement of 8 liters/min through face mask no less than 6-8 h after the start of NIPPV. [fig_ref] Table 1: Base line characteristics of the four cases treated using noninvasive positive pressure... [/fig_ref] shows the base line characteristics of the four treated cases. In all cases, NIPPV had to be instituted within 2-3 h of ICU admission. During this time, the standardized treatment protocol was applied. There were no problems with mask fit or maintenance of the positive pressure. None of the patients subsequently had positive blood cultures. Summary of the outcome of NIPPV use in the four treated patients is shown in.
# Results
Retrospective review of three comparable cases, managed in our unit with invasive positive pressure ventilation prior to this trial of NIPPV use, revealed a potential outcome benefit with the use of NIPPV [ [fig_ref] Table 3: Comparison of outcome between the NIPPV group and comparable previously treated patients... [/fig_ref] ].
# Discussion
NIPPV has had limited application in pregnant patients due to the perceived risk of aspiration. The current case series has its own limitations in providing definitive recommendations for the indications for and use of NIPPV in the pregnant patient with ARF. A follow-up well-conducted study would be of great use in determining when and how to use this management approach in pregnant patients with sickle cell disease presenting with ACS.
Obstetric patients with preexisting medical problems are more likely to require intensive-care support than those without preexisting medical conditions. [bib_ref] Case-control study of risk factors for obstetric patients' admission to intensive care..., Bouvier-Colle [/bib_ref] Sickle cell disease is the most common major underlying chronic medical condition in obstetric patients admitted to our ICU. ACS is believed to be a specific form of acute lung injury that can progress to acute respiratory distress syndrome causing ARF. [bib_ref] Pathogenesis of lung disease in sickle hemoglobinopathies, Weil [/bib_ref] Because there are no absolute contraindications to the use of NIPPV and the boundaries for its use continue to expand, we decided to look at the use of NIPPV in pregnant patients with ARF caused by ACS.
Pregnant patients are at risk of aspiration. It is a serious complication of pregnancy and is a frequent cause of indirect obstetric death. All the patients in this series were kept nil per mouth and treated with placement of distal gastric tubes for decompression. None developed aspiration pneumonia. Young adults have a lower incidence of ACS, but it tends to be more severe and is often fatal. [bib_ref] The acute chest syndrome of sickle cell disease, Quinn [/bib_ref] Pregnant patients with ACS causing ARF admitted to ICU need aggressive and timely intervention from a multidisciplinary team of physicians to minimize the morbidity and mortality associated with this devastating complication of sickle cell disease. Moreover, acute respiratory failure contributes substantially to maternal morbidity and mortality; it can also harm the fetus by compromising fetal oxygen delivery. Upon follow-up, only one patient had LSCS. Both patients and the babies were alive and healthy at 6 months after delivery.
A significant incidence of right ventricular dysfunction and pulmonary hypertension in asymptomatic patients with sickle cell anemia has been reported. [bib_ref] Pulmonary hypertension in sickle cell disease, Sutton [/bib_ref] In the setting of severe, acute lung disease, this quiescent pulmonary hypertension can cause right ventricular dysfunction that is severe enough to cause circulatory compromise. Hypotension is therefore possible. Two of our patients required short course of inotropic support. However, none of the four deteriorated to the extent of requiring invasive mechanical ventilation.
Critical illness in pregnant women poses special challenges. Physiologic changes that affect cardiovascular and respiratory function are normal during pregnancy. Through interaction with preexisting or new co-morbidities, those changes can give rise to life-threatening complications. Always present, too, is the need to consider the effects of both disease and its treatment on fetal development and outcome. In this small case series, the use of NIPPV was found to be safe and successful with good maternal and fetal outcome in pregnant patients with ARF caused by ACS.
Currently there is not enough evidence to support safe use of NIPPV in ARF in a pregnant patient. The current case series provides the best available evidence to support the use of NIPPV in ARF during pregnancy. In closely monitored pregnant patients with ARF, NIPPV seems to have the potential to shorten ICU and hospital stay. A well-conducted randomized controlled clinical trial is required to confirm this finding.
[fig] Y: : es, N: No, SVD: Spontaneous vaginal delivery, LSCS: Lower segment cesarean section, NIPPV -Noninvasive positive pressure ventilation [/fig]
[table] Table 3: Comparison of outcome between the NIPPV group and comparable previously treated patients using invasive ventilation [/table]
[table] Table 1: Base line characteristics of the four cases treated using noninvasive positive pressure ventilation [/table]
|
Unraveling of Advances in 3D-Printed Polymer-Based Bone Scaffolds
# Introduction
Bone is capable of self-healing, but cannot regenerate in special cases such as large defects due to the lack of growth and differentiation platform of bone repair-related cells [bib_ref] The improvement of mechanical and thermal properties of polyamide 12 3D printed..., Tuan Rahim [/bib_ref]. At present, the options of treatments for critical-sized bone defects include autologous bone graft [bib_ref] Modular ceramic scaffolds for individual implants, Biggemann [/bib_ref] [bib_ref] Autologous Bone Graft: Properties and Techniques, Pape [/bib_ref] , allogenic bone graft [bib_ref] Living bone allotransplants survive by surgical angiogenesis alone: Development of a novel..., Larsen [/bib_ref] [bib_ref] The role of orthobiologics in foot and ankle surgery, Wee [/bib_ref] [bib_ref] The Short-term Follow-up Results of Open Wedge High Tibial Osteotomy with Using..., Lee [/bib_ref] , and artificial bone graft substitutes (i.e., bone tissue engineering scaffolds, BTES) [bib_ref] Bone healing in 2016. Clin. Cases Miner, Buza [/bib_ref] [bib_ref] Osteoinductive ceramics as a synthetic alternative to autologous bone grafting, Yuan [/bib_ref]. However, auto-transplantation and allotransplantation will lead to secondary injury, poor size-matching, immune response, and other risks, limiting their clinical application [bib_ref] A tissue engineering solution for segmental defect regeneration in load-bearing long bones, Reichert [/bib_ref] [bib_ref] In situ bone tissue engineering via ultrasound-mediated gene delivery to endogenous progenitor..., Bez [/bib_ref] [bib_ref] Remodeling of cortical bone allografts mediated by adherent rAAV-RANKL and VEGF gene..., Ito [/bib_ref]. To solve the above-mentioned problems of clinical treatment process, BTES with optimal biocompatibility and strong osteoinduction ability have been widely studied, especially the polymer-based composite bone scaffolds fabricated by 3D printing according to the clinical needs of patients, 3D printing technology can prepare accurately-controlled personalized-implants upon composition and structure, realizing the superior structure-function relationship and highly bioactive BTES, this cannot be accomplished by traditional processing strategy.
Recently, increasing attention has been paid to the research of polymer-based BTES, especially through the combination of natural polymers and synthetic polymer materials to create new tissue engineering scaffolds [bib_ref] Recent Progress on 3D-Printed Polylactic Acid and Its Applications in Bone Repair, Chen [/bib_ref] [bib_ref] Synthesis and characterization of gelatin-PVP polymer composite scaffold for potential application in..., Mishra [/bib_ref] [bib_ref] Preparation and characterization of collagen/PLA, chitosan/PLA, and collagen/chitosan/PLA hybrid scaffolds for cartilage..., Haaparanta [/bib_ref]. Bone tissue is composed of water, organic matter, and inorganic salt. The ideal BTES needs to simulate the original bone tissue structure. Natural polymers are biocompatible and biodegradable, but their mechanical strength and thermal stability are poor. The synthetic polymers possess ideal mechanical properties, but the hydrophobic surface leads to poor osseointegration [bib_ref] Additive Manufacturing of Material Scaffolds for Bone Regeneration: Toward Application in the..., Garot [/bib_ref]. For example, gelatin can simulate the biological properties of natural bone matrix protein, but it shows weak mechanical strength. The tissue scaffolds prepared by electrospinning combined with polylactic acid (PLA) have good biological and mechanical properties [bib_ref] Study of the electrospun PLA/silk fibroin-gelatin composite nanofibrous scaffold for tissue engineering, Yin [/bib_ref]. This paper reviews the research progress of polymer materials and tissue engineering scaffolds prepared by 3D printing technology. The properties of ideal BTES are summarized by introducing the process of bone healing, with emphasis on the polymer materials commonly used in 3D printing technology, the design optimization of scaffold structure, and its application in clinical medicine. In addition, it will describe the technical challenges arising from the current research and the potential prospects of 3D printed polymer BTES.
## Process, advantages, and disadvantages of 3d printing technology
Natural bone consists of highly dense outer cortical bone and relatively loose inner cancellous bone. There is a Havers tube rich in blood vessels and nerves longitudinally in the cortical bone, and the vessels in the Havers tube are connected to each other in a transversely oriented channel, the Volkmann tube [bib_ref] Biomimetic Approaches for Bone Tissue Engineering, Ng [/bib_ref]. Cancellous bone is a reticular structure composed of plate or rod structure about 200 microns thick. 80% of the bone remodeling process occurs in cancellous bone [bib_ref] The nanocomposite nature of bone drives its strength and damage resistance, Tertuliano [/bib_ref]. Both types of bone undergo dy-namic remodeling, maturation, differentiation and absorption, which are controlled by the interaction between osteoblasts, osteoblasts and osteoclasts [bib_ref] Bioinspired structural materials, Wegst [/bib_ref]. At the same time, extracellular matrix (ECM) provides mechanical support and appropriate environment for cell attachment, proliferation and differentiation. BTES are usually designed to simulate ECM to promote tissue regeneration. At present, a variety of manufacturing technologies are used to construct BTES, such as Solvent Cast-ing/Particulate Leaching [bib_ref] Osteogenic differentiation of hMSCs on semi-interpenetrating polymer networks of polyurethane/poly(2-hydroxyethyl methacrylate)/cellulose nanowhisker..., Shahrousvand [/bib_ref] , Gas Foaming [bib_ref] Converting 2D Nanofiber Membranes to 3D Hierarchical Assemblies with Structural and Compositional..., Chen [/bib_ref] , Freeze-Drying [bib_ref] Tissue-specific extracellular matrix scaffolds for the regeneration of spatially complex musculoskeletal tissues, Cunniffe [/bib_ref] , Phase Separation [bib_ref] Highly porous polycaprolactone scaffolds doped with calcium silicate and dicalcium phosphate dihydrate..., Gandolfi [/bib_ref] [bib_ref] Preparation, characterization and in vitro test of composites poly-lactic acid/hydroxyapatite scaffolds for..., Pavia [/bib_ref] , Electrospinning [bib_ref] Controlled Co-delivery of Growth Factors through Layer-by-Layer Assembly of Core-Shell Nanofibers for..., Cheng [/bib_ref] [bib_ref] Spatially Regulated Multiphenotypic Differentiation of Stem Cells in 3D via Engineered Mechanical..., Horner [/bib_ref]. Sometimes the two techniques are used in combina-tion [bib_ref] 3D bioactive composite scaffolds for bone tissue engineering, Turnbull [/bib_ref] [bib_ref] Combination of fused deposition modeling and gas foaming technique to fabricated hierarchical..., Zhou [/bib_ref] [bib_ref] Porous stable poly(lactic acid)/ethyl cellulose/hydroxyapatite composite scaffolds prepared by a combined method..., Mao [/bib_ref]. Among them, BTES fabricated by computeraided design (CAD) modeling 3D printing technology possess the highest accuracy and repeatability as well as high spatial control ability scaffold upon microstructure, therefore it is one of the most ideal scaffolds for clinical application.
According to the material and manufacturing process, the typical printing methods of BTES can be divided into laser-assisted printing and non-laser-assisted printing. Laserassisted printing technologies include stereo-lithography (SLA), selective laser sintering (SLS), laser-assisted Bioprinting (LAB). On the other hand, non-laser-assisted printing technologies, such as extrusion bioprinting, inkjet bioprinting and fused deposition modeling, have also been frequently reported in the literatures. This overview explains 3D printing methods in accordance with the above categories.
## Laser-assisted printing
Given that there are many types of 3D printing processes, understanding the advantages and disadvantages of 3D printing processes and related materials will enable designers to make better decisions when choosing 3D printing processes to design and manufacture the best bone engineering scaffolds. Here are three types of printing grouped together.Because they all involve laser-assisted printing. Instead, the other three are classified as non-laser-assisted printing methods.
## Sla
As a rapid prototyping process, SLA is one of the earliest 3D printing technologies used in BTES, which can be traced back to 1986. SLA solidifies the 3D scaffold through a polymerization process. Usually, a polymerization chain reaction is initiated on a layer or monomer solution by ultraviolet light or an electron beam. Once the first layer is fully cured, the platform is lowered a short distance in the vertical direction (Z-direction) and continues to connect to the new layer. These steps are repeated until a new model is completed [fig_ref] Figure 1: 3D printing process diagram [/fig_ref]. The opposite approach (layer-below-layer) is the most common approach for SLA, as it allows to finely control the stratification thickness.The platform is immersed in a photopolymer liquid and exposed to the focused light according to the The main advantage of SLS for bone tissue engineering is the 3D printing of biometals (e.g., titanium) and ceramics (e.g., hydroxyapatite). However, polymer materials such as polyether ether ketone (PEEK) with superior mechanical properties and biocompatibility are increasingly being used to manufacture bone scaffolds through SLS. Because of the high energy used in SLS to fuse the powder particles into a solid structure similar to natural cancellous bone, the scaffold shows high osteoblast attachment and bone differentiation [bib_ref] Poly ether ether ketone and its composite powder prepared by thermally induced..., Wang [/bib_ref]. The main limitation of SLS lies in the selection of polymer materials, the high temperature involved in the printing of scaffolds limits the use of cells and biological materials like FDM. The SLS printing method's disadvantages are the energy consumption and high cost in the sintering process of SLS manufacturing bracket.
## Lab
LAB is a derivative application of direct-write technology and laser-induced forward transfer technology. A typical laser-assisted biological printer consists of a laser pulse, a material donor, and a receiving substrate, as shown in [fig_ref] Figure 1: 3D printing process diagram [/fig_ref]. The donor has a ribbon structure consisting of an energy-absorbing layer at the top (such as titanium or gold), a bio-ink layer at the bottom (such as cells [bib_ref] Advancing Frontiers in Bone Bioprinting, Ashammakhi [/bib_ref] and polymers), and a Donor Layer in the middle. During the printing process, the energy absorbing layer generates pressure by receiving energy from the laser, then transmitted to the Donor Layer. The pressure is passed to the Donor Layer, which then uses the pressure to push the cell-containing material from the initial print material to the receiving substrate.
LAB has better cell printing resolution and accuracy than other bioprinting technologies and is one of the most attractive tools for in-situ printing of bone substitutes [bib_ref] Laser-assisted bioprinting for bone repair, Hakobyan [/bib_ref]. In addition, Lab does not face the problems associated with nozzles. First, when the diameter The most prominent advantage of SLA is the ability to print bone scaffolds of complex shapes at extremely high resolution, allowing the design of complex structures ranging in size from submicrons to decimeters. Compared with the minimum resolution of 0.005-0.010 mm for FDM and 0.060 mm-0.150 mm for SLA, the minimum resolution of SLA can reach 0.002 mm [bib_ref] Current status and challenges of Additive manufacturing in orthopaedics: An overview, Javaid [/bib_ref]. Although advanced scaffolds can be created using SLA, we acknowledge that there are some limitations in the application of SLA in scaffold preparation. First, because the polymer solution required for this type of printing is very sensitive to ultraviolet light, the choice of printing materials is minimal. Although the number of available polymer materials continues to increase, the technology is still limited to using only one polymer material at a time [bib_ref] A review on stereolithography and its applications in biomedical engineering, Melchels [/bib_ref]. Second, uncured resins and residual photoinitiators may be cytotoxic and have poor biocompatibility [bib_ref] 3D printing of polymer matrix composites: A review and prospective, Wang [/bib_ref]. Also, the manufacturing speed of SLA is relatively slow compared to other printing technologies, and the cost is relatively high.
## Sls
SLS is a powder-based additive manufacturing technology that creates 3D printed BTES with complex shapes by curing powdered materials layer by layer. Solidification is achieved by selective fusion or sintering of specified areas in each layer using the heat energy of the focused laser radiation system, as shown in [fig_ref] Figure 1: 3D printing process diagram [/fig_ref] (a 2 ). First, a layer of powder is deposited into the construction chamber. According to the cross-section data of the 3D CAD model, the laser beam sinters or melts the selective region on the powder layer to form a solid layer. Once the initial layer is complete, the construction platform is lowered by 100 to 200 microns, and a new layer of powder is deposited evenly and tightly onto the platform through a barrel rolled in the powder transfer system. A new layer of powder is then laser-sintered and bonded to the previous layer. This process is repeated until a complete 3D bone tissue engineering scaffold is created. Similar to SLS processing powders, Selective laser melting (SLM) and electron beam melting (EBM) are relatively new and rapid technologies.
The main advantage of SLS for bone tissue engineering is the 3D printing of biometals (e.g., titanium) and ceramics (e.g., hydroxyapatite). However, polymer materials such as polyether ether ketone (PEEK) with superior mechanical properties and biocompatibility are increasingly being used to manufacture bone scaffolds through SLS. Because of the high energy used in SLS to fuse the powder particles into a solid structure similar to natural cancellous bone, the scaffold shows high osteoblast attachment and bone differentiation [bib_ref] Poly ether ether ketone and its composite powder prepared by thermally induced..., Wang [/bib_ref]. The main limitation of SLS lies in the selection of polymer materials, the high temperature involved in the printing of scaffolds limits the use of cells and biological materials like FDM. The SLS printing method's disadvantages are the energy consumption and high cost in the sintering process of SLS manufacturing bracket.
## Lab
LAB is a derivative application of direct-write technology and laser-induced forward transfer technology. A typical laser-assisted biological printer consists of a laser pulse, a material donor, and a receiving substrate, as shown in [fig_ref] Figure 1: 3D printing process diagram [/fig_ref]. The donor has a ribbon structure consisting of an energy-absorbing layer at the top (such as titanium or gold), a bio-ink layer at the bottom (such as cells [bib_ref] Advancing Frontiers in Bone Bioprinting, Ashammakhi [/bib_ref] and polymers), and a Donor Layer in the middle. During the printing process, the energy absorbing layer generates pressure by receiving energy from the laser, then transmitted to the Donor Layer. The pressure is passed to the Donor Layer, which then uses the pressure to push the cell-containing material from the initial print material to the receiving substrate.
LAB has better cell printing resolution and accuracy than other bioprinting technologies and is one of the most attractive tools for in-situ printing of bone substitutes [bib_ref] Laser-assisted bioprinting for bone repair, Hakobyan [/bib_ref]. In addition, Lab does not face the problems associated with nozzles. First, when the diameter of the nozzle is minimal or the extrusion pressure is considerable, the shear stress of the nozzle will not cause the problem of inducing cell damage and death. Second, when the viscosity of biological ink is larger or the concentration of cells increases, it will not cause nozzle blockage.
## Non-laser-assisted printing
## Extrusion bioprinting
Compared with inkjet bioprinting, continuous bio-ink lines can give a better wholebody interface to the generated 3D structure. The two main distribution mechanisms of bioprinting are pneumatic-based and mechanical-based. As shown in [fig_ref] Figure 1: 3D printing process diagram [/fig_ref] (b 1 ). In the printing process, a continuous force driven by either pneumatic pressure or piston or screw pressure extrudes an uninterrupted line of bio-ink through a micro nozzle, rather than a droplet. The extrusion material solidifies on the substrate and acts as a supporting structure. Next, the platform is lowered horizontally and another layer of bio-ink is added until the complete 3D structure is formed.
One of the most significant advantages of this technology over thermal-based or piezoelectric-based inkjet bioprinting is that it does not involve heating or piezoelectric processes, so it is easy to combine cells with bioactive substances. Second, with a few tweaks to the technology, The platform can squeeze bio-inks and continuously deposit various bio-inks. Switching between different containers to quickly manufacture complex structures [bib_ref] Rapid Continuous Multimaterial Extrusion Bioprinting, Liu [/bib_ref]. However, the nozzle's moving speed and the bio-ink viscosity's nonlinearity affect the resolution of the extruded bioprinting [bib_ref] Effective bioprinting resolution in tissue model fabrication, Miri [/bib_ref]. Extruding bioprinting inevitably causes the shear stress of the print needle on the cells, which can also cause cell damage and death [bib_ref] Engineering considerations on extrusion-based bioprinting: Interactions of material behaviour, mechanical forces and..., Emmermacher [/bib_ref]. In the process of extruding bio-printing, the greater the extrusion pressure on the needle, the greater the shear stress on the cells in the bio-ink, and the lower the activity of the cells in the printing model [bib_ref] Advances on Bone Substitutes through 3D Bioprinting, Genova [/bib_ref]. Shao et al. Use ultrasound to assist the bioprinting process, which can alleviate the problem of reduced cell survival caused by blocked nozzles [bib_ref] Ultrasonic manipulation of cells for alleviating the clogging of extrusion-based bioprinting nozzles, Shao [/bib_ref].
## Inkjet bioprinting
Inkjet bioprinting is one of the most traditional methods used in nanomaterials. Although it works on the same principle as a traditional drip on-demand printer, biomaterial such as cells [bib_ref] Advancing Frontiers in Bone Bioprinting, Ashammakhi [/bib_ref] and polymers replace the ink inside the cartridge, while a computercontrolled receiving substrate replaces the paper. Inkjet bioprinting is a common bioprinting technology based on the traditional inkjet printing method, which uses the desktop inkjet printer for bioprinting. This is a 3D printing method based on non-contact liquid drops, which uses thermal or compressed power to spray biological ink droplets from the nozzle of the printing head onto the hydrogel matrix or petri dish under the control of the computer, as shown in [fig_ref] Figure 1: 3D printing process diagram [/fig_ref] (b 2 ). In thermal-sensitive technology, air bubbles generated by local heating in an inkjet printer create a pulse of pressure that forces a drop of bio-ink from the print head to the substrate. In piezoelectric technology, In piezoelectric technology, a piezoelectric transducer generates pulses to generate enough pressure to eject droplets from the nozzle. Thermological or piezoelectric pressures are the two typical extrusions of the print head, which are basically controlled by custom CAD files to load the ink material onto the receiving substrate accurately. The 3D positioning system consists of a receiving substrate and two degrees of freedom (X and Y axis) print head.
The advantage of inkjet bioprinting is that cells and other growth factors (bio-ink) can be added to the printed polymer material, improving bone induction and bone conduction to overcome the limitations of conventional treatment options [bib_ref] Advances on Bone Substitutes through 3D Bioprinting, Genova [/bib_ref] [bib_ref] High-precision 3D inkjet technology for live cell bioprinting, Takagi [/bib_ref]. Although the shear and thermal stress of the nozzle on the cells during printing can affect the viability of the cells in the bio-ink, it has been shown in the literature that various cell types using inkjet bio-printing can produce 80-95% cell viability [bib_ref] Inkjet printing biomaterials for tissue engineering: Bioprinting, Saunders [/bib_ref]. Secondly, its printing process is a non-contact process in the form of liquid droplets, from the nozzle to the receiving plate, the nozzle is completely separated from the printing material and the printing bracket. Compared with extrusion-based bio-printing, this printing technology can prevent the movement of the nozzle on the X-Y axis from interfering with the printing process. However, inkjet bio-printing also has some shortcomings and shortcomings. The biggest drawback is the clogging of the nozzle, which makes it difficult to achieve smooth printing. There are two main reasons for the blockage of the nozzle: one is the viscosity of the polymer solution is too high, the other is that there are too many cells in the polymer solution [bib_ref] Inkjet printing biomaterials for tissue engineering: Bioprinting, Saunders [/bib_ref]. Of course, in order to reduce the impact of this limitation, Gao et al. in the process of printing PEG peptide scaffold, the photopolymerization of acrylate esterified peptide and acrylate esterified polyethylene glycol (PEG) hydrogel co-printing method can minimize the problem of nozzle blockage to the greatest extent [bib_ref] Inkjet-bioprinted acrylated peptides and PEG hydrogel with human mesenchymal stem cells promote..., Gao [/bib_ref]. Furthermore, cell density, printing speed, and nozzle size in bio-inks are some known factors that affect the resolution and mechanical properties of inkjet bioprinting structures.
## Fdm
As shown in [fig_ref] Figure 1: 3D printing process diagram [/fig_ref] (b 3 ), the continuous filament of thermoplastic material is heated in the heating element, melted into a semi-liquid state, and then extruded onto the building platform or the previously printed layer when the platform is lowered vertically. The process takes place layer by layer, with each layer deposited and fused together. After that, the printed 3D structure solidifies at room temperature, creating complex three-dimensional geometry. The complex three-dimensional geometry produced therein can be precisely managed by the motion of the manufacturing platform (in the Z direction) and the nozzle (in the X-Y direction) controlled by a CAD data file.
Compared with other printing techniques, FDM has many advantages in manufacturing functional parts of complex structures. By optimizing printing parameters including printing temperature, printing speed and printing layer thickness, Wang et al. adjusted the mechanical properties, surface quality and microstructure of the printing model to meet practical needs [bib_ref] Effect of chemical treatment on tensile strength and surface roughness of 3D-printed..., Jayanth [/bib_ref] [bib_ref] Effects of printing parameters of fused deposition modeling on mechanical properties, surface..., Wang [/bib_ref]. Although FDM has the advantages of low cost, fast speed, and easy process operation, FDM printing technology has limitations [bib_ref] Progress in Additive Manufacturing and Rapid Prototyping, Kruth [/bib_ref]. Biologics and natural polymers have limited applications in FDM due to the high temperature required for melting during FDM printing. The viscosity of the molten polymers limits the resolution of the printing that can be achieved, so you can only produce biological scaffolds with a fixed shape and relatively regular structure. In our view, much effort is needed to overcome these limitations.
## The process of bone healing
Although fracture repair may not require support materials other than implants, bone grafts or bone substitutes may be required for bone defects exceeding the critical size. 3D printed biomaterial scaffolds with suitable interconnected porous structures play an essential role in bone tissue engineering. Especially for bone defects exceeding the critical size, they can not regenerate themselves. A matrix is needed as a scaffold to guide the activities of bone regeneration cells. The healing process of bone defects exceeding the critical size is a complex physiological process. Such bone defects of unstable long bone fractures are usually repaired by intramembrane ossification (IO) and endochondral ossification (EO) [bib_ref] New insights on the reparative cells in bone regeneration and repair, Huang [/bib_ref]. Repair initiates after the inflammatory reaction and includes a soft and hard callus formation stage. In the stage of inflammation, the rupture of blood vessels around the fracture end, the outflow of pulp through the rupture of bone marrow cavity, and the dilatation of capillaries of local inflammatory reaction resulted in the formation of hematomas around the fracture, and the osteocytes, periosteal cells and soft tissue cells around the broken end were necrotic [fig_ref] Figure 2: Repair processes of bone defeat [/fig_ref]. The important sources of hematomas are neutrophils, macrophages, and platelets, which are involved in initiating inflammatory responses. Inflammatory cells are activated, triggering a cascade of reactions triggered by the hematoma, which, characterized by hypoxia and low pH, acts as a temporary scaffold for active invasion by local tissue macrophages and polymorphonuclear neutrophilsBone defects above the critical size lack similar temporary scaffolds, and bone implants (autologous, allogeneic, tissue-engineered scaffolds, etc.) are required to perform this role.
Polymers 2022, 13, x FOR PEER REVIEW 7 of 32 scaffold for active invasion by local tissue macrophages and polymorphonuclear neutro-philsBone defects above the critical size lack similar temporary scaffolds, and bone implants (autologous, allogeneic, tissue-engineered scaffolds, etc.) are required to perform this role. After the hematoma was mechanized into granulation tissue, there were a lot of new blood vessels at the fracture end, and osteoclasts continued to remove the remaining dead bone. The initial stage of repair starts from the distal site of the stable and good peripheral vascular cortical bone fracture, from the release of the inner layer of periosteum cells continuously concentrated, proliferated, and differentiated into braided bone. Chondroblasts began to appear in the periosteum near the fracture end and formed chondroid tissue to replace the granulation tissue, leading to the formation of the bone collar through IO (Fig- After the hematoma was mechanized into granulation tissue, there were a lot of new blood vessels at the fracture end, and osteoclasts continued to remove the remaining dead bone. The initial stage of repair starts from the distal site of the stable and good peripheral Polymers 2022, 14, 566 7 of 31 vascular cortical bone fracture, from the release of the inner layer of periosteum cells continuously concentrated, proliferated, and differentiated into braided bone. Chondroblasts began to appear in the periosteum near the fracture end and formed chondroid tissue to replace the granulation tissue, leading to the formation of the bone collar through IO [fig_ref] Figure 2: Repair processes of bone defeat [/fig_ref]. At this stage, early vascularization plays an important role in healing bone defects. Because early vascularization is the basis of blood supply to the bone defect site, slow or incomplete vascularization will lead to insufficient oxygen and nutrient supply, resulting in hypoxia and cell death of the implanted material cells, or preventing the growth of host bone cells. As Ronald e. Unger, bone tissue engineering scaffold need not only bone cell growth, proliferation and differentiation in the biological material, and must happen quickly the new vascularization and graft internal blood flow [bib_ref] Improving vascularization of engineered bone through the generation of pro-angiogenic effects in..., Unger [/bib_ref]. In the stage of cartilage eschar formation, around the fracture space, the mechanical strength of the fracture is unstable and the local microenvironment hypoxia caused by the interruption of blood supply leads to the formation of cartilage mass [fig_ref] Figure 2: Repair processes of bone defeat [/fig_ref]. The chondrocytes gradually mature into hypertrophic chondrocytes (HCs). HCs secret collagen X in the transition zone where cartilage will be replaced by bone tissue and expresses a variety of active molecules, promoting mineralization [fig_ref] Figure 2: Repair processes of bone defeat [/fig_ref] and vascular invasion into the distal fracture site of the callus and reorganizing the outer boundary of the periosteum of the callus. At this stage, the arrangement of bone trabeculae is loose and irregular.
In the stage of hard eschar formation, the whole cartilage was constantly replaced by woven bone, the irregular new bone trabeculae in the original callus were gradually thickened, the arrangement began to be regular and dense, and the Harvard system of cortical bone was re-established. Osteoclasts and osteoblasts invaded the dead bone at the fracture end to complete the crawling replacement process of dead bone clearance and new bone formation. The original eschar was replaced by lamellar bone, so that the fracture site formed a strong bone connection. Then the pulp cavity recommunicates and finally forms the bone frame through the fracture space. In forming new bone links by self stress, bone scaffolds made of metal [bib_ref] Titanium Fiber Plates for Bone Tissue Repair, Takizawa [/bib_ref] and ceramics [bib_ref] 3D-printed Bioceramic Scaffolds: From Bone Tissue Engineering to Tumor Therapy, Ma [/bib_ref] [bib_ref] Fabrication of ceramic bone scaffolds by solvent jetting 3D printing and sintering:..., Lee [/bib_ref] often encounter a problem-stress barrier. Because metals and ceramics have high elastic modulus, when they are implanted into bone defects as bone scaffolds, their long-term contact with bone will reduce the physical load on the surrounding bone, resulting in the decrease of bone mineral density around the bone scaffolds and even secondary damage.
In contrast, the elastic modulus of polymer-based composite scaffolds can be adjusted according to their own ratio [bib_ref] A novel chitosan-tussah silk fibroin/nano-hydroxyapatite composite bone scaffold platform with tunable mechanical..., Ran [/bib_ref] [bib_ref] Development of gelatin-chitosan-hydroxyapatite based bioactive bone scaffold with controlled pore size and..., Maji [/bib_ref]. The bone scaffolds are sufficient to provide mechanical support for weight-bearing bone construction before newly formed bone synthesis and can not be too high to avoid stress shielding. Different cells and signaling pathways participate in the process of bone repair, and various cells and signaling pathways play a coordinator role. During the differentiation of osteoblasts, the activation of transcription factors at a specific time provides necessary clues to specify the function of bone progenitor cells when cells transform into osteoblasts. Ankit salhotra et al. Discussed the interaction between transcription factors (such as SOX9, Runx2, OSX and activated transcription factor 4 (ATF4)) necessary for osteoblast differentiation [bib_ref] Mechanisms of bone development and repair, Salhotra [/bib_ref]. More and more researchers load these transcription factors and primordial progenitor cells into bone tissue engineering scaffolds, so that when bone tissue engineering scaffolds are implanted in vivo, they can quickly activate the proliferation and differentiation of primordial bone cells. For example, it has been confirmed that the application of bioactive proteins such as bone morphogenetic protein (BMP)-2 and basic fibroblast growth factor (bFGF) is an effective way to improve the bone inducibility of bone scaffolds. Ren et al. delivered BMP-2 and bFGF to poly (L-lactic-co-glycolic)/graphene oxide/hydroxyapatite nanofibrous scaffolds. It was found that BMP-2 can induce the differentiation of bone cells, BFGF promotes the proliferation of bone cells, which makes the nanofiber scaffold have excellent bone induction rate and regeneration activity [bib_ref] Synergistic delivery of bFGF and BMP-2 from poly(l-lactic-coglycolic acid)/graphene oxide/hydroxyapatite nanofibre scaffolds..., Ren [/bib_ref]. Because of the excellent performance of BMP-2 in inducing osteocyte differentiation, Zhang et al. coupled derived bone morphogenetic protein 2 derived from bone anoligopeptide (ssvpt, Ser ser ser Val Pro THR) with dopamine coating on 3D printed polylactic acid (PLA) scaffolds. He also established a rat skull defect model to verify that the scaffold has high bone conductivity for the adhesion and proliferation of rat bone marrow mesenchymal stem cells (MSCs) [bib_ref] Immobilization of BMP-2-derived peptides on 3D-printed porous scaffolds for enhanced osteogenesis, Zhang [/bib_ref].
3D printing technology can adjust the pore diameter, porosity and microporous structure of bone scaffold according to the parameters in the printing process, and customize the shape of a matched bone defect. However, 3D printing structures made of single biomaterials (such as PLA, PLGA, etc.) lack the biological activity required by the ideal requirements for inducing bone formation. It is necessary to add additives to promote bone induction and bone transmission in scaffolds. These additives are more than natural factors and cell involvement in bone healing, so exploring the mechanism of bone healing plays a vital role in the preparation of 3D printed bone tissue engineering scaffolds.
## Properties of ideal bone scaffolds
BTES used to repair bone defects should possess several important characteristics: biocompatibility, biodegradability, mechanical properties, controllable structure. Zhang et al. demonstrated BETS prepared by natural and synthetic polymers as a promising method for creating novel tissue-engineered scaffolds. Because the scaffolds combine the advantages of the two materials and meet various requirements, including biological activity, mechanical properties, controllable degradability, and other properties [bib_ref] Three-dimensional (3D) printed scaffold and material selection for bone repair, Zhang [/bib_ref].
## Biocompatibility
Biocompatibility is the primary criterion for all tissue-engineered scaffolds. First, BTES should not inhibit the activity of bone tissue cells, that is, allow cell adhesion, migration and proliferation. Second, BETS should neither show any significant cytotoxicity during or after transplantation nor induce a positive immune response to prevent severe inflammation. Printable and biocompatible polymer materials are the optimum material in the applications of 3D printing. Because they possess high adjustability and complexity and provide a prefer bionic environment for living cells [bib_ref] Recent Progress in 3D Printing of Elastic and High-Strength Hydrogels for the..., Dai [/bib_ref]. Therefore, more and more people have begun to commit to the research of biological materials needed for 3D printing in recent years, especially in polymer-based composite print materials.
## Biodegradability
Controllable biodegradability is also a special property realized by scaffolds. Biodegradation of implanted materials will inevitably affect the mechanical support of growing bone tissue. They provide a biological and mechanical framework for the growth and differentiation of cells. It is eventually replaced by regenerated tissue designed to match the mechanical properties of natural bone. Ideally, the rate of scaffold degradation synchronizes with the rate of mineralized tissue deposition. The gradual reduction of mechanical support provided by the degradable scaffold is compensated by the gradual increase of mechanical support provided by the new tissue. If the scaffold degrades too quickly, the scaffold will not be able to provide mechanical support while new bone is formed, which is likely to exceed the load and may lead to fracture. Conversely, if the scaffold does not degrade quickly, it can trigger an inflammatory response to foreign substances in the scaffold, thus preventing tissue regeneration. In order to prepare scaffolds that can both enhance the activity of osteoblasts and display appropriate degradation rates, there have been many studies on the interaction between degradation and osteoblasts in vitro [bib_ref] Fabrication of biocompatible and bioabsorbable polycaprolactone/magnesium hydroxide 3D printed scaffolds: Degradation and..., Abdal-Hay [/bib_ref]. The biodegradability of polymers is an attractive property that can be controlled by the molecular design of polymers.
## Mechanical properties
As the essential properties of bone grafts, mechanical properties (i.e., elastic modulus and compressive strength) are the main challenges for applying 3D printing technology to porous scaffolds for bone tissue engineering. Many factors affect the mechanical properties of BTES, such as the properties of biomaterials and pore structure. On the one hand, the mechanical properties of most polymer BTES are low, but the optimized post-treatment method and component modification can improve the mechanical properties of polymer scaffolds. On the other hand, the pore density and size of the scaffold significantly affect the growth and adhesion of cells. However, the porosity is inversely proportional to the mechanical properties, so the mechanical properties of multi-pore BTES sometimes cannot meet the stress of natural bone. In addition, the mechanical properties of bone tissue engineering materials had better match that of natural bone to avoid the stress shielding phenomenon. This phenomenon usually occurs in the application of traditional metal bone scaffolds. Suppose the elastic modulus of the scaffold is much greater than the elastic modulus of the surrounding tissue. In that case, most of the stress will be borne by the scaffold (mostly metal scaffold) rather than the surrounding bone.
Over time, mechanical stimulation of bone cells around the scaffold is reduced, ultimately leading to decreased bone density (osteopenia) in healthy bone tissue near the scaffold [bib_ref] Materials design for bone-tissue engineering, Koons [/bib_ref] , accelerated absorption of surrounding bone, and even secondary infection. In general, the mechanical properties of the scaffold should be sufficient to provide mechanical support for weight-bearing bone construction before the synthesis of newly formed bone, but not too high to avoid stress shielding.
## Microstructures
The microstructure of scaffolds is also critical in promoting cell viability and tissue growth. In the absence of an engineered blood supply, the interconnected pore structure allows for the inward diffusion of oxygen and nutrients and the outward diffusion of waste from the scaffold. Besides, porosity also supports cell migration into the scaffold and increases the available surface for cells to bind to the scaffold and interact with the surrounding tissue. In the design stage of BTES, the microstructure of the scaffold must be precisely designed with parameters favorable to cells and tissues, including the porosity of the scaffold, the pore size and the interconnected pore structure, etc.
## Porosity
It is well known that the skeletal structure of adult bones is not uniform and radially graded, consisting of two distinct structural regions. The outer high-density areas are called cortical bone with a porosity of 5% to 30% (mostly in the 5% to 10% range), while the inner areas are called cancellous bone with a porosity of 50% to 90%. Considering the gradient structural characteristics associated with natural bone porosity, the design of regenerative bone scaffolds can simulate the porosity changes between cortical and cancellous bone to promote regional cell differentiation. Given the radiating gradients of natural bone, Andrea et al. have designed a method that can be used in combination with human mesenchymal stem cells (HMSC). The porosity of each scaffold area is similar to the gradient found in natural bone, with 29.6% ± 5% porosity in the outer ring and 50.8% ± 8.1% and 77.6% ± 3.2% porosity in the central and inner regions, respectively. They also matched the size of the scaffold to the structural characteristics of the natural bone. This porous scaffold with a gradient structure of 500 µm in the outer ring, 750 µm in the middle, and 1000 µm in the inner. Cell differentiation was confirmed by up-regulated gene expression of Runx2 and bone sialoprotein markers. The experiment showed that optimizing the porosity and pore size of BTES by imitating the natural bone structure was beneficial to the differentiation of HMSC and the mineralization of bone tissue [bib_ref] Toward mimicking the bone structure: Design of novel hierarchical scaffolds with a..., Di Luca [/bib_ref].
## Pore size
The size of the individual pore size within the scaffold is an important consideration. Small pores (a few microns to tens of microns) promote cell adhesion, intracellular signal transmission, cell proliferation and migration, while large pores (a few hundred microns) facilitate angiogenesis, ECM aggregation, and tissue formation. The pore size of a BTES depends mainly on the anatomical location and type of bone tissue (e.g., cortical or trabecular) in the human body. Diao et al. showed that the pore size design of bone tissue engineering porous scaffolds should consider the type of bone defect. He studied three β-tricalcium phosphate scaffolds with different pore sizes (100, 250, and 400 µm). Then, he found that 100 µm pore size scaffolds enhanced osteoblast differentiation during intramembranous ossification. It is more effective in inducing bone formation and is most suitable for repairing flat bone defects. The 400 µm scaffold can accelerate the formation of cartilage template and ossification center in endochondral ossification, showing the best ability of bone formation to repair extended bone defects. Numerous studies have demonstrated the importance of the pore size of scaffolds in bone engineering and indicated that the pore size of scaffolds should generally be between 100 and 300 µm to allow cell penetration, migration, and growth, and to achieve optimal tissue vascularization. However, there is no clear consensus on the effect of optimal pore size on optimal mechanical and osteogenic properties [bib_ref] Three-dimensional (3D) printed scaffold and material selection for bone repair, Zhang [/bib_ref].
## Pore structure
In addition to the porosity and pore size affecting the microstructure of the scaffold, the interconnected pore structure within the scaffold also plays a crucial role in influencing its function. As more and more people study the pore structure, the pore structure of different scaffolds tends to be diversified.
Piotr et al. studied 3D-printed porous bone tissue scaffolds based on shape memory polymer composites (SMPC). The microstructure of the scaffold is based on the observation and analysis of the bone trabecular structure of the lotus root. Four different pore shapes (circular holes, polygonal holes, randomly oriented holes) are designed and supported [fig_ref] Figure 3: Pore structure diversity of 3D printed bone scaffolds [/fig_ref]. The reliability of the structures was demonstrated by mechanical experiments and micromechanical theoretical studies, while biological experiments verified the biological activity and osteogenic effect of the scaffolds. The results showed that the round hole and polygonal hole scaffolds were more beneficial to promote early adhesion, proliferation, and osteogenesis. Although the scaffolds with random orientations and directional orientations showed good cell adhesion, no significant cell proliferation was observed early. In the later observation species, the scaffolds with random orientations could promote cell proliferation and osteogenesis [bib_ref] Porous bone tissue scaffold concept based on shape memory PLA/Fe 3 O..., Zhao [/bib_ref].
Piotr et al. selected five pore geometries (triangular prism with circular and flat contours, cube, octagonal prism, Sphere) and seven porosity (up to 80%), and 70 models were constructed for analysis [fig_ref] Figure 3: Pore structure diversity of 3D printed bone scaffolds [/fig_ref] , to select the appropriate pore geometry and scaffold porosity for orthopedic regenerative medicine. On the one hand, the researchers placed the scaffold in the flow channel to estimate the growth media velocity and wall shear stress. On the other hand, the researchers placed scaffolds in the bone to assess osteoblastic proliferation. The results of this study evaluated the effects of different pore shapes and porosity of scaffolds on bone regeneration. They provided a basis for the appropriate selection of pore geometry and porosity of scaffolds in orthopedic regenerative medicine [bib_ref] Numerical Analysis of the Influence of Porosity and Pore Geometry on Functionality..., Prochor [/bib_ref].
In order to solve the problems of poor mechanical properties and limited osteogenic activity of scaffolds in the past, Liu et al. developed a novel 3D printed composite scaffold consisting of poly-lactide (PLLA) matrix, surface-grafted MgO whiskers (g-MgOs) and Helosite nanotubes (g-HNTs). The scaffold not only combines the printability of PLLA, the excellent osteogenic activity of g-MgOs, and the excellent enhancement and toughening effect of g-HNTs, but also optimizes the microstructure of the scaffold. Based on the consideration of cell proliferation and migration, the researchers designed a scaffold combined with a large (510 ± 20 µm) and a small (210 ± 15 µm) honeycomb [fig_ref] Figure 3: Pore structure diversity of 3D printed bone scaffolds [/fig_ref] , with a porosity of 74.15 ± 5.32%. The experimental results show that the interconnecting pore structure in the scaffold makes the media have favorable fluidity, which is conducive to the entry of nutrients, the proliferation and migration of cells [bib_ref] The design, fabrication and evaluation of 3D printed gHNTs/gMgO whiskers/PLLA composite scaffold..., Liu [/bib_ref].
periments and micromechanical theoretical studies, while biological experiments verified the biological activity and osteogenic effect of the scaffolds. The results showed that the round hole and polygonal hole scaffolds were more beneficial to promote early adhesion, proliferation, and osteogenesis. Although the scaffolds with random orientations and directional orientations showed good cell adhesion, no significant cell proliferation was observed early. In the later observation species, the scaffolds with random orientations could promote cell proliferation and osteogenesis [bib_ref] Porous bone tissue scaffold concept based on shape memory PLA/Fe 3 O..., Zhao [/bib_ref]. Piotr et al. selected five pore geometries (triangular prism with circular and flat contours, cube, octagonal prism, Sphere) and seven porosity (up to 80%), and 70 models were constructed for analysis [fig_ref] Figure 3: Pore structure diversity of 3D printed bone scaffolds [/fig_ref] , to select the appropriate pore geometry and scaffold porosity for orthopedic regenerative medicine. On the one hand, the researchers placed the scaffold in the flow channel to estimate the growth media velocity and wall shear stress. On the other hand, the researchers placed scaffolds in the bone to assess osteoblastic proliferation. The results of this study evaluated the effects of different pore shapes and Optimizing the pore structure of scaffolds can influence the regeneration of bone defects of critical size and the remodeling process and regeneration of osteochondral defects. Considering the influence of different pore structures on the regeneration of osteochondral defects in vivo, Feng et al. prepared silk fibroin/collagen (SF/COL) composite scaffolds with three pore structures (random pore, radial pore and axial pore) [fig_ref] Figure 3: Pore structure diversity of 3D printed bone scaffolds [/fig_ref] to evaluate the effect of pore structure on the remodeling process and regeneration of osteochondral defect tissue. The results showed that the scaffolds with radial and axial pore arrangement had better regeneration ability and endogenous repairability than those with random pore arrangement, especially those with radial pore arrangement [bib_ref] Influence of pore architectures of silk fibroin/collagen composite scaffolds on the regeneration..., Feng [/bib_ref].
Vascular formation in bone scaffolds plays an important role in bone regeneration, but the vascularization of bone regeneration is slow in the critical size of bone defects. It is well known that blood vessels are initially formed by endothelial cells organized into microtubules. According to previous studies, it was found that a microfluidic system composed of a group of microchannels can be used to induce endothelial cells to form incomplete blood vessels in vitro [bib_ref] Fabrication of 3D Biomimetic Microfluidic Networks in Hydrogels, Heintz [/bib_ref]. Zhang et al. designed a kind of silicate bioceramics (BRT-H) bone scaffold with hollow tube structure and bioactive ions by optimizing the microstructure parameters of the 3D-printed bone scaffold [fig_ref] Figure 3: Pore structure diversity of 3D printed bone scaffolds [/fig_ref]. The scaffold utilizes the synergistic effect of the conduit structure of the BRT scaffold and the bioactive ion composition to solve the problem of slow vascularization of large bone regeneration. The results showed that the hollow tube structure facilitated the transfer of stem cells and growth factors and possessed a synergistic effect with bioactive ion products in enhancing the regeneration of vascularized bone [bib_ref] 3D-printed scaffolds with synergistic effect of hollow-pipe structure and bioactive ions for..., Zhang [/bib_ref]. Considering that microchannel structure can induce endothelial cells to form a basic vascular system, Feng et al. were inspired by the microstructure of the natural plant lotus root. A bionic scaffold with a multi-channel structure [fig_ref] Figure 3: Pore structure diversity of 3D printed bone scaffolds [/fig_ref] was successfully prepared by optimizing the microstructure parameters of the 3D-printed bone scaffold. The results showed that the scaffold improved the porosity and specific surface area and significantly improved the adhesion and proliferation of BMSCs in vitro, as well as osteogenesis and angiogenesis in vivo compared with traditional 3D printing materials [bib_ref] 3D Printing of Lotus Root-Like Biomimetic Materials for Cell Delivery and Tissue..., Feng [/bib_ref]. Both of the above two papers have demonstrated that optimizing the microstructure of bone scaffolds is conducive to cell transport and regeneration (blood vessels and bone) within the scaffold. Although many 3D-printed scaffolds with different pore structures have been developed, the optimal pore structures affecting bone growth are still being studied.
In conclusion, despite these latest studies, there is still much to be learned about the effect of designed scaffolds on bone regeneration in future in vivo studies. Thus, it remains a challenge to successfully balance scaffold properties conducive to cell function, cell viability, and mechanical integrity under load-bearing conditions.
## Polymer material of 3d-printed bone scaffolds
Polymers offer greater design flexibility than metals and ceramics. Biopolymers have received significant attention in material development due to their extensive availability, low toxicity/non-toxicity, biodegradability, biocompatibility, chemical versatility, and inherent functionality. The biocompatibility of biological materials affects the functional properties of 3D-printed tissues and organs. Cells need to be attached to the surface of the implanted biomaterial to maintain its activity and proliferation to promote tissue regeneration. Therefore, the selection of biocompatible materials is critical to the formulation design of biological inks. In this section, we present the fundamental aspects of these different biopolymers that can be linked to their processing and material applications; below is the summary table .
## Natural polymers and mixtures based on natural polymers
Natural polymers are considered as a class of macromolecules with multiple monomers or compound monomers linked together to form long chains. They are usually identical or very similar macromolecules that are the building blocks of extracellular tissue. As a result, there are few problems with foreign body reactions when they are implanted as implants. In addition, they usually perform biological functions at the molecular level. Most commonly used in biomaterials, natural biopolymers include chitin and chitosan, alginic acid, collagen, gelatin, hyaluronic acid, and fibrinogen.
## Chitin and chitosan
Chitin is the second most abundant carbohydrate after cellulose. It is a polysaccharide found in crustaceans (exoskeleton of crab and shrimp shells and cell walls of fungi and yeast). Its derivative, chitosan, is obtained by deacetylation of chitin. Chitin and its deacetylated derivative chitosan are natural polymers composed of randomly distributed β-(1-4)-linked d-glucosamine (deacetylated unit) and n-acetyl-d-glucosamine (acetylated unit) [bib_ref] Chitin and Chitosan: Structure, Properties and Applications in Biomedical Engineering, Islam [/bib_ref]. N-acetylglucosamine exists in the composition of chitin and chitosan, so a common degradation product is n-glucosamine. However, n-glucosamine, a substance naturally present in the extracellular matrix of eukaryotic cells, is non-toxic to the human body and biocompatible when applied to bone scaffolds. N-Acetylglucosamine (GlcNAc) is a monosaccharide that is typically polymerized linearly through the (1,4)-β bond. It is also because the components of chitin and chitosan both have n-acetylglucosamine, which can accelerate tissue repair and prevent the formation of scar tissue. However, in actual application, when the traditional chitin/chitosan lacks the necessary mechanical strength, the mechanical properties of the composite scaffold can be optimized by blending chitin/chitosan with other materials [bib_ref] Chitin and chitosan composites for bone tissue regeneration, Venkatesan [/bib_ref]. Deepthi et al. reviewed that many studies have adopted different methods to incorporate hydroxyapatite (HAP), bioglass ceramic (BGC), silicon dioxide (SiO 2 ), titanium dioxide (TiO 2 ), and zirconium oxide (ZrO 2 ) into chitin or chitosan scaffolds to enhance the mechanical properties of the scaffolds themselves [bib_ref] An overview of chitin or chitosan/nano ceramic composite scaffolds for bone tissue..., Deepthi [/bib_ref]. One of the most important manifestations of the biocompatibility of chitin/chitosan scaffolds is their ability to promote cell adhesion, proliferation, and differentiation [bib_ref] Applications of chitin and chitosan nanofibers in bone regenerative engineering, Tao [/bib_ref]. Studies have shown that chitin/chitosan composite scaffolds have a good affinity for bone marrow mesenchymal stem cells and osteogenic induction capacity, and have broad application prospects in bone regeneration. The research results showed that the content of chitosan in the composite nanofibers could affect the growth of bone marrow mesenchymal stem cells. He placed rat bone marrow mesenchymal stem cells (rBMSCs) on nanofiber membranes with three different concentrations of chitosan. After seven days of culture in vitro, the nanofiber membrane with the mass ratio of chitosan to PCL of 30:70 showed the strongest proliferation effect on BMSCs compared with the nanofiber membrane with the mass ratio of chitosan to PCL of 50:50 and pure PCL nanofiber membrane. On the 14th day, the expression levels of osteogenic genes Runx2, ALP, and OCN reached the highest in the group with a chitosan/PCL mass ratio of 50:50 nanofiber membrane. On day 21, the ratio played the most significant role in promoting calcium deposition. The results showed that chitosan could promote the proliferation and osteogenic differentiation of rBMSCs, and its activity on the chitosan nanofiber scaffold was affected by such factors as composition, composition ratio, and action time [bib_ref] Osteogenic induction of bone marrow mesenchymal cells on electrospun polycaprolactone/chitosan nanofibrous membrane, He [/bib_ref].
Although chitin and chitosan have good biocompatibility, biodegradability, antibacterial activity, non-antigenicity, and high adsorption performance, there are still many problems and challenges in the clinical application of chitin/chitosan nanofibers. We can improve the preparation process of the chitin/chitosan bone scaffold to make it more suitable for the biomedical application of bone regeneration.
## Alginate
Sodium alginate is a natural biopolymer consisting of two monosaccharide units, namely a-L-mannuronic acid (M unit) and b-D-guluronic acid (G unit) linked by 1,4glycosidic bond and composed of different GGGMM segments. Alginate tends to gel with divalent cations under normal physiological conditions, one of the most important properties apart from biocompatibility and biodegradability. Alginate is widely used in the formulation of extrudable mixtures for 3D printing due to its ability to increase the viscosity of polymer solutions.
Although interpenetrating polymer network (IPN) hydrogels composed of gelatin and hydroxypropyl cellulose (HPC) have been prepared by continuous enzymatic and chemical crosslinking methods, such hydrogels have better mechanical properties than networks composed of two or more interpenetrating polymers [bib_ref] Interpenetrating network hydrogels with high strength and transparency for potential use as..., Wang [/bib_ref]. Bone tissue engineering scaffold requires high mechanical properties, and interpenetrating hydrogel will inevitably reduce polymer concentration and increase the water content. Therefore, Luo et al. prepared an alginate/gelatin interpenetrating scaffold with uniform nano-apatite coating through 3D printing and in-situ mineralization [bib_ref] Concentrated gelatin/alginate composites for fabrication of predesigned scaffolds with a favorable cell..., Luo [/bib_ref]. The scaffold realizes double crosslinking of alginate and gelatin through CaCl2 and EDC solution so that the scaffold has good mechanical property and good biological activity. In addition, the uniformly distributed nano-apatite coating was prepared by 3D printing and in-situ mineralization technology on this scaffold. The uniformly distributed nano-apatite coating enhanced the adsorption of protein on the surface of the scaffold and significantly promoted the proliferation and osteogenic differentiation of rBMSCs [bib_ref] 3D printing of concentrated alginate/gelatin scaffolds with homogeneous nano apatite coating for..., Luo [/bib_ref].
## Collagen
Bone is composed of cells embedded in the ECM, which is an ordered network composed of two major nano-phases: collagen fibers composed of type I collagen molecules and hydroxyapatite nanocrystals distributed along the collagen fibers [bib_ref] Bioinspired structural materials, Wegst [/bib_ref]. A collagen is a group of at least 29 different polymeric proteins that are the most abundant protein component of the ECM. ECM provides a special physiological microenvironment for cells, protects cells from harmful mechanical effects, and mediates mechanically-induced signal transmission. Actually, the bone scaffold that promotes bone regeneration functions as an ECM. All collagen proteins are in a triple helix structure consisting of hydrogen bonds, and the main amino acid groups include glycine, proline, and hydroxyproline. The fibrils formed by these triple helix structures are strong and highly flexible and can be further crosslinked to improve the mechanical properties of different types of collagen [bib_ref] Additive manufacturing of natural biopolymers and composites for bone tissue engineering, Bose [/bib_ref]. However, different types of collagen have rather complex and diverse structures, splice variants, additional non-helical domains, assemblies, and functions. Among them, collagen fibers of type I and type V are involved in the structural skeleton of the bone, while collagen fibers of type II and type XI are mainly involved in the fibrous matrix of articular cartilage [bib_ref] Collagens-Structure, function, and biosynthesis, Gelse [/bib_ref]. Their torsional stability and tensile strength determine the stability and integrity of these structures. Therefore, bone scaffolds based on collagen have been widely used to treat bone/cartilage defects and bone/cartilage regeneration. Bone scaffolds composed of type I collagen usually rely on the sclerosis of type I collagen through calcium phosphate mineralization and cross-linking with substances such as HAP.
## Gelatin
Gelatin is a mixture of polypeptide and protein obtained by partial hydrolysis of collagen extracted from animal skin and bone. It widely exists in mammals and has excellent biological activity, such as cell adhesion, biocompatibility, and biodegradability [bib_ref] Tough biodegradable chitosan-gelatin hydrogels via in situ precipitation for potential cartilage tissue..., Shen [/bib_ref]. Moreover, it does not have antigenicity under a Physiological environment [bib_ref] The pro-angiogenic characteristics of a cross-linked gelatin matrix, Dreesmann [/bib_ref]. It is widely used in various biomedical applications, and currently, it is often used in clinical wound dressings and adhesives. In bone tissue engineering, gelatin in injectable form can be used as a delivery carrier for cells. There are also a large number of studies on the mixing of gelatin with different polymers. Because of its suitability and biocompatibility, gelatin has been extensively studied in the field of bone regeneration by mixing with nano-HAP [bib_ref] Gelatin sponges (Gelfoam ® ) as a scaffold for osteoblasts, Rohanizadeh [/bib_ref] , β-tricalcium phosphate [bib_ref] Fabrication and properties of βTCP/Zeolite/Gelatin scaffold as developed scaffold in bone regeneration:..., Yazdanian [/bib_ref] , and chitosan [bib_ref] Preparation and characterization of macroporous chitosan-gelatin/β-tricalcium phosphate composite scaffolds for bone tissue..., Yin [/bib_ref].
However, gelatin can form a thermoreversible gel with water. Gelatin shows a gel state at low temperatures (below 25 - C) and a dissolved state at high temperatures (above 35 - C). Gelatin is very sensitive to process temperature due to its sol-gel transformation. Therefore, it is challenging to use FDM technology for the 3D printing of gelatin. It is necessary to modify the surface of the gelatin. There are amine groups, carboxyl groups, and other pendant groups in the gelatin molecule, which can be used as active sites for gelatin modification. Wang et al. mixed gelatin with bacterial cellulose (BC) treated with tempo-medicated oxidation (TO-BC) and maleic acid (MA-BC). The carboxyl group in cellulose can be cross-linked with the amide in gelatin through the hydrogen bond interaction between N-Hydroxysuccinimide (NHS) and 1-(3-dimethyl aminopropyl)-3-(ethyl carbodiimide hydrochloride) (EDC) to enhance the mechanical strength and gel stability of gelatin, which is more suitable for printing of tissue scaffolds.
To determine the mechanical strength difference of gelatin, studies have found that glutaraldehyde (GTA) reacts with the base of the polypeptide chain to form Schiff bases, which can stabilize the gelatin structure through further reaction with other glutaraldehyde molecules during the formation of crosslinking [bib_ref] Glutar-aldehyde as a crosslinking agent for collagen-based biomaterials, Olde Damink [/bib_ref]. Kathleen et al. found that gelatin scaffolds crosslinked with GTA showed enhanced mechanical properties and stability. Meanwhile, the cytotoxicity of glutaraldehyde to cells can be neutralized by lysine [bib_ref] Architectural and Mechanical Cues Direct Mesenchymal Stem Cell Interactions with Crosslinked Gelatin..., Mcandrews [/bib_ref].
## Hyaluronic acid
Hyaluronic acid (HA) is an acidic, non-sulfated glycosaminoglycan consisting of D-glucuronide and N-acetylglucosamine disaccharide units. As a component of the extracellular matrix, HA can retain water in tissues, support cellular structures and act as a lubricant because its hydroxyl groups can tightly bind water molecules to chains through hydrogen bonds. HA is a major component of the ECM of articular cartilage (AC). HA maintains the normal homeostasis of cartilage sites by regulating cell function, including promoting the phenotype of chondrogenic genes, as well as the production and retention of matrix components [bib_ref] Hyaluronan and CD44: Strategic players for cell-matrix interactions during chondrogenesis and matrix..., Knudson [/bib_ref].
HA interacts with cells through the surface receptor CD44 and regulates cell movement and adhesion [bib_ref] Alginate Microcapsules Incorporating Hyaluronic Acid Recreate Closer in Vivo Environment for Mesenchymal..., Canibano-Hernandez [/bib_ref]. HA also helps to reduce the immunogenicity of the embedded cells as this biocompatible material reduces the adsorption of protein [bib_ref] Immobilized hyaluronic acid containing model silicone hydrogels reduce protein adsorption, Van Beek [/bib_ref]. Therefore, it has attracted significant attention in the biomedical fields, such as bone regeneration therapy, wound healing, and drug delivery. Despite these excellent biocompatibility properties, HA is an ideal biomaterial for cartilage tissue engineering, but it lacks the mechanical properties required for application to three-dimensional extrusion-based bioprinting (EBB). Limited by this reason, the uniform distribution of cells cannot be ensured. HA also lacks the gelling ability necessary to maintain the 3D structure after the printing process. Cristina Antich et al. performed 3D printing with HA and PLA and the resulting scaffold showed good mechanical properties, including printability, good gel strength, and degradation [bib_ref] Bio-inspired hydrogel composed of hyaluronic acid and alginate as a potential bioink..., Antich [/bib_ref].
## Cellulose
Cellulose is composed of β-D-glucopyranose covalently linked by C1 (carbon atom) and C4 (β-1,4-glycosidic bond) through acetal interaction [bib_ref] Zhu, Q. 3D printing with cellulose materials, Wang [/bib_ref]. Cellulose has the advantage of low cost and high elasticity [bib_ref] Natural Cellulose Fiber as Substrate for Supercapacitor, Gui [/bib_ref]. As a major component of plants, cellulose is a sustainable and nearly inexhaustible polymeric feedstock [bib_ref] Cellulose modification by polymer grafting: A review, Roy [/bib_ref] , and cellulose has many derivatives, including cellulose ethers/esters, micro/nano-sized cellulose products, etc. Many cellulose derivatives such as cellulose ether and microcrystalline cellulose (MCC) have been important commercial products for many years [bib_ref] Cellulose microfibres produced from banana plant wastes: Isolation and characterization, Elanthikkal [/bib_ref]. Cellulose can be used as a matrix with a natural layered network and porous structure and contain abundant functional groups (especially -OH group) on the surface of cellulose fiber, which can be combined with other materials.
Cellulose and its derivatives are materials suitable for 3D printing. Since cellulose cannot be melt processed, identification/development of a good solvent for cellulose is essential for its utilization. However, cellulose is insoluble in water and ordinary organic solvents due to the formation of intramolecular and intermolecular hydrogen bonds [bib_ref] Direct Dissolution of Cellulose: Background, Means and Applications, Olsson [/bib_ref]. To date, only a few solvent systems have been able to dissolve cellulose. Generally speaking, cellulose solvents can be divided into derivative solvents and non-derivative solvents [bib_ref] Competing forces during cellulose dissolution: From solvents to mechanisms, Medronho [/bib_ref]. Under the action of derivative solvent, cellulose hydroxyl undergoes yellowing, esterification, etherification, and other functionalization reactions. All these will destroy the intramolecular and intermolecular hydrogen bonds between cellulose molecules and lead to the dissolution of cellulose. Non-derivatizing solvents, such as ionic liquids, can dissolve cellulose through physical molecular interactions without prior derivatization.
## Synthetic polymers and mixtures based on synthetic polymers
Natural polymers have good biocompatibility and biodegradability. However, their compression modulus did not reach the normal level for cancellous bone (>100 MPa) [bib_ref] Competing forces during cellulose dissolution: From solvents to mechanisms, Medronho [/bib_ref] [bib_ref] A high-strength mineralized collagen bone scaffold for large-sized cranial bone defect repair..., Wang [/bib_ref] [bib_ref] Silk fibroin/gelatin microcarriers as scaffolds for bone tissue engineering, Luetchford [/bib_ref]. So they are mechanically weak and cannot withstand the forces exerted on the bone [bib_ref] Current state of fabrication technologies and materials for bone tissue engineering, Wubneh [/bib_ref]. Therefore, they are mainly used as additives or composites because their bone-inducing properties and the ability to enhance cell and protein adhesion benefit from good bionic action [bib_ref] Applications of 3D printing on craniofacial bone repair: A systematic review, Maroulakos [/bib_ref]. Synthetic polymers are often used as 3D bone tissue engineering scaffold materials or composited with other materials in 3D printing technology due to their good mechanical properties. At present, the commonly used synthetic polymers suitable for 3D printing BTES mainly include PLA, polycaprolactone (PCL), and polycarbonate (PC), PEEK, Polypropylene (PP), Polyamide (PA), etc. Acrylonitrile butadiene styrene (ABS) can perfectly simulate natural bone in vitro and is often used in clinical model teach-ing aids or surgical models.Researchers can choose appropriate 3D printing technology to meet actual printing needs based on the properties of synthetic polymer materials and their advantages and disadvantages.
## Acrylonitrile butadiene styrene
Acrylonitrile butadiene styrene (ABS) is an economical and efficient polymer, simple to manufacture, it has good impact resistance, chemical resistance, processability, and suitable strength and characteristics [bib_ref] Is 3D printing safe? Analysis of the thermal treatment of thermoplastics: ABS,..., Wojtyła [/bib_ref]. ABS is a thermoplastic and amorphous polymer made of polymerized styrene, acrylonitrile, and polybutadiene. The melting temperature of ABS is 105 - C, making ABS a common candidate for FDM and SLA systems. In recent years, it has been found that the addition of montmorillonite (OMMT) can effectively improve the tensile strength of 3D printing ABS, which makes the composite polymer more excellent in mechanical properties and thermal stability, thus becoming a promising printing material in FDM [bib_ref] Mechanical and thermal properties of ABS/montmorillonite nano-composites for fused deposition modeling 3D..., Weng [/bib_ref]. ABS tissue scaffold can provide an appropriate environment for cell growth and matrix regeneration for cartilage and intervertebral disc repair, and it is usually used for bone tissue engineering such as cartilage [bib_ref] 3D-Printed ABS and PLA Scaffolds for Cartilage and Nucleus Pulposus Tissue Regeneration, Rosenzweig [/bib_ref]. most of the ABS is not biodegradable, but ABS M30i plastic has good degradability. However, because the ABS material cannot pass through the autoclave [bib_ref] Description of the CAD-AM process for 3D bone printing: The case study..., Napolitano [/bib_ref] , which is an important Preoperative procedure before implantation in the human body, the ABS tissue scaffold cannot be used in the human body. However, it can perfectly simulate natural bone in vitro and is often used in clinical model teaching aids or surgical models [bib_ref] Design and 3D-printing of titanium bone implants: Brief review of approach and..., Popov [/bib_ref] [bib_ref] 3D printing anatomical models of head bones, Bartikian [/bib_ref] [bib_ref] 3D Printed Surgical Simulation Models as educational tool by maxillofacial surgeons, Werz [/bib_ref] to make detailed surgical plans and considerably shorten the operation time.
## Polylactic acid
PLA refers to a series of polymers: Pure poly-L-Lactic acid (L-PLA), Pure poly-D-Lactic acid (D-PLA), and poly-D, L-lactic acid (DL-PLA), homopolymers of L-PLA are mostly used in clinical practice. Each monomer can be different from other materials in the composition of copolymers suitable for 3D printing [bib_ref] Polymeric biomaterials, Griffith [/bib_ref]. Unlike ABS, PLA has the advantages of low cost, biocompatibility, and biodegradability [bib_ref] Polylactic Acid Technology, Drumright [/bib_ref] [bib_ref] A Literature Review of Poly(Lactic Acid), Garlotta [/bib_ref]. The melting temperature of PLA is between 173 and 178 - C, so it is relatively easy to be used to make scaffolds in the temperature range of 190-230 - C [bib_ref] An overview of the recent developments in polylactide (PLA) research, Madhavan Nampoothiri [/bib_ref] , which is why PLA is widely used in FDM. PLA still has good mechanical properties and has a similar compressive strength (230 mpa) as bone [bib_ref] Mechanical Properties of Optimized Diamond Lattice Structure for Bone Scaffolds Fabricated via..., Liu [/bib_ref] , making it suitable for musculoskeletal tissue engineering. However, PLA has many disadvantages. First, the tensile strength of pure PLA materials is poor, while stretchability is very important in FDM technology because they affect the extrusion of material filaments and layer stacking. Bartolomeo et al. printed the PLA and clay composite at a higher printing temperature, and the clay enhanced the thermal stability of PLA and remarkably enhanced its elastic properties [bib_ref] 3D Printing of PLA/clay Nanocomposites: Influence of Printing Temperature on, Coppola [/bib_ref]. Secondly, the release of lactic acid by-products in the degradation process need to be noticed [bib_ref] biodegradation and excretion of polylactic acid (PLA) in medical implants and theranostic..., Silva [/bib_ref]. In the early stage of degradation, lactic acid can be metabolized by itself. In the late stage of degradation, many lactic acid by-products can form a local acidic environment, leading to tissue inflammation and cell death. To address this problem, calcium phosphate may be used as a buffer to maintain the PH of the internal environment at 7.4 [bib_ref] Carbonated calcium phosphates are suitable pH-stabilising fillers for biodegradable polyesters, Schiller [/bib_ref].
## Polycaprolactone
Similar to PLA, PCL is also a kind of low-cost and degradable polyester. The melting temperature of PCL is 58-60 - C [bib_ref] Structural Evolution of PCL during Melt Extrusion 3D Printing, Liu [/bib_ref] , and it has excellent elastic properties [bib_ref] The relationship between the mechanical properties and cell behaviour on PLGA and..., Baker [/bib_ref]. The tensile strength of block PCL is 10.5~16.1 MPa, the modulus is 343.9~364.3 MPa, and the tensile yield strength is 8.20~10.1 MPa [bib_ref] Mechanical and microstructural properties of polycaprolactone scaffolds with one-dimensional, twodimensional, and three-dimensional..., Eshraghi [/bib_ref]. Compared with many polyester materials, PCL is more suitable for scaffold construction by FDM. In SLS printing, Shaun Eshraghi et al. tested the PCL samples and found that the mechanical properties would decrease sharply with the increase of porosity. However, the low-stress region could be designed as the void region by adjusting the SLS parameters, thereby improving the mechanical strength of PCL and manufacturing the tissue scaffold with controllable mechanical properties [bib_ref] Mechanical and microstructural properties of polycaprolactone scaffolds with one-dimensional, twodimensional, and three-dimensional..., Eshraghi [/bib_ref].
The PCL can degrade for a long time [bib_ref] A Novel Route To Poly(ε-caprolactone)-Based Copolymers via Anionic Derivatization, Ponsart [/bib_ref] and still exist after being placed in organisms for 12 weeks [bib_ref] Scaffolds for bone tissue engi-neering fabricated from two different materials by the..., Park [/bib_ref] , which can provide long-term protection and support for tissue healing and regeneration and does not form potentially harmful byproducts like PLA. Muwan Chen et al. embedded hyaluronic acid, methylated collagen, and terpolymer into PCL's pore-like structure, and a good cell seeding rate was achieved over a period of time [bib_ref] Self-assembled composite matrix in a hierarchical 3-D scaffold for bone tissue engineering, Chen [/bib_ref]. In summary, by selecting the appropriate printing method and combining with the long-term degradation rate of PCL, we can make scaffolds with both composite stress requirements and controllable degradation in different parts.
## Polycarbonate
Polycarbonate (PC) is a common high-strength chemical material, able to withstand temperatures below 140 - C without physical deformation. Common PC materials are synthesized by the chemical industry, while some researchers try to recover the plastic part of electronic waste products from obtaining PC materials, thus reducing carbon dioxide emissions during the process of synthesizing PC materials. After the recovered PC material was subjected to 3D printing, an intensity close to that of a commercial PC printing material was obtained [bib_ref] Transformation of E-Waste Plastics into Sustainable Filaments for 3D Printing, Gaikwad [/bib_ref]. But with the depletion of chemical fuels such as petroleum and their biological toxicity, a popular way to synthesize PC in the field of 3D printing implants in the human body is to polymerize isosorbide extracted from corn starch.
A bio-based PC, which is polymerized by isosorbide extract from corn starch, not only has excellent tensile properties suitable for 3D printing but no bio-toxicity has been observed [bib_ref] Preparation and mechanical properties of a copolycarbonate composed of bio-based isosorbide and..., Lee [/bib_ref] [bib_ref] Polymers from renewable 1,4:3,6-dianhydrohexitols (isosorbide, isomannide and isoidide): A review, Fenouillot [/bib_ref].
Seong et al. produced four samples of ABS, PLA, chemically synthesized PC, and biobased PC using 3D printing technology, respectively. The bio-based PC showed the highest tensile strength, elongation at break, and elastic modulus among the four materials. It was also determined that the bio-based PC had good rheological properties at the optimum temperature in the range of 240 - C to 270 - C. However, the PC easily absorbed moisture from the air. Many small defects were formed on the surface of the material when the moisture evaporated, which affected the performance and printability. Seong et al. overcame this defect by performing at 80 - C for 4 h.
## Polyetheretherketone
Polyetheretherketone (PEEK) is a member of the polyaryletherketone polymer family, and its chemical structure can maintain resonance stability, making it extremely inactive [bib_ref] Chemical and radiation stability of PEEK, Kurtz [/bib_ref]. PEEK has low water absorption [bib_ref] Characterization and exposure of polyetheretherketone (PEEK) to fluid environments, Stober [/bib_ref] [bib_ref] The influence of thermal history on the dynamic mechanical and dielectric studies..., Boinard [/bib_ref] and can be maintained in an aqueous environment for long periods. It has been found in the early research that the rigidity of PEEK did not change significantly after multiple continuous pressurized steam cycles [bib_ref] Carbon fiber reinforced PEEK (APC-2/AS-4) composites for orthopaedic implants, Kwarteng [/bib_ref]. Such high-temperature resistance enables high-temperature steam sterilization of PEEK material without affecting the material performance. Of course, due to its high-temperature resistance, it needs to be adjusted to a higher temperature for extrusion when 3D printing is performed.
PEEK has excellent mechanical properties, with its Young's modulus of 3.6 GPa and tensile strength of 100 MPa [bib_ref] Current understanding and challenges in high temperature additive manufacturing of engineering thermoplastic..., Das [/bib_ref]. Studies have made composite printing of PEEK with different contents with other materials, which can improve the mechanical strength of the printing scaffold and realize the specific customization of specific tissues in the human body [bib_ref] Printability and mechanical performance of biomedical PDMS-PEEK composites developed for material extrusion, Smith [/bib_ref]. In the past, metal materials were often used for tissue replacement due to the high-intensity movement of the mandible. However, due to the stress shielding effect, it was not conducive to the patient's recovery. Because PEEK has similar mechanical strength to human bones, desirable results were observed in mandibular replacements [bib_ref] Topological optimization of 3D printed bone analog with PEKK for surgical mandibular..., Cheng [/bib_ref] [bib_ref] Design of porous titanium scaffold for complete mandibular reconstruction: The influence of..., Dutta [/bib_ref].
In addition, it is worth mentioning that PEEK has a high transmittance, which enables doctors to more clearly observe the local reactions and tumor recurrence after surgery through CT or MRI, giving it unique advantages over traditional metal implants in the field of bone tumors (such as spinal tumors) [bib_ref] Are Carbon-fiber Implants More Efficacious Than Traditional Metallic Implants for Spine Tumor..., Mcdonnell [/bib_ref].
## Polypropylene
Polypropylene (PP) is a crystalline (crystallinity 60-70%) thermoplastic polymer with the advantages of low cost, strong versatility, excellent physical, mechanical, and thermal properties, and its melting point is 165 - C [bib_ref] Effect of reprocessing cycles on the degradation of PP/PBATthermoplastic starch blends, Oliveira [/bib_ref].
PP is one of the lightest polymers in common printing materials (density: 0.908 g/cm − 3 ). The yield strength (32 MPa) and stiffness of PP are slightly lower than those of ABS or PLA, but with longer service life [bib_ref] PP in 3D Printing-Technical and Economic Aspects, Savu [/bib_ref]. These two characteristics make PP material suitable for replacing tissue scaffolds which require lightweight and low-stress requirements. The mechanical strength of PP can be enhanced by mixing PP with other materials. However, although PP has a wide range of chemical compatibility, its low surface energy results in that most materials cannot be easily adhered to its surface [bib_ref] 3D printing of versatile reactionware for chemical synthesis, Kitson [/bib_ref]. At the same time, studies have found that PP can shrink and curl due to crystallization in the printing process, which is not conducive to printing [bib_ref] Spray-dried cellulose nanofibril-reinforced polypropylene composites for extrusion-based additive manufacturing: Nonisothermal crystallization kinetics..., Wang [/bib_ref]. Because of the above two points, Darsani et al. grafted PP with maleic anhydride (MA) and found that it could improve the surface adhesion property of PP. At the same time, the blending of Palm oil fuel ash (POFA) and PP increased the adhesion of polymer materials and improved their shrinkage during printing [bib_ref] Capability Behaviour of POFA Composite Filament for 3D Printing User, Darsani [/bib_ref].
## Polyamide
Polyamide (PA) is one of the commonly used polymers for tissue engineering applications. PA has good biocompatibility, mechanical properties, excellent wear resistance, and chemical stability. PAs are therefore commonly used in SLS technology for accurate printing and replacement of bone tissue, but can also be used in FDM technology [bib_ref] The improvement of mechanical and thermal properties of polyamide 12 3D printed..., Tuan Rahim [/bib_ref].
In that SLS process, it is required that the material have a relatively low melt viscosity to achieve the desired melt flow rate without inputting too much energy while reducing the loss of the driving device [bib_ref] Effects of processing on microstructure and properties of SLS Nylon 12, Zarringhalam [/bib_ref] , PA is a semi-crystalline polymer with good sintering properties and relatively low melt viscosity, in particular PA12, which is a very suitable material for SLS technology [bib_ref] Polymer powders for selective laser sintering (SLS), Schmid [/bib_ref].
In recent years, there have been many studies on the preparation of composite materials with bioactive materials such as PA and HAP to replace bone tissue scaffolds [bib_ref] Computer-Aided Design and Manufacturing and Rapid Prototyped Nanoscale Hydroxyapatite/Polyamide (n-HA/PA) Construction for..., Li [/bib_ref]. The addition of bioactive materials such as HA increases the composite material's tensile strength, stiffness, and thermal stability. Although the addition of these materials increases the friction of the composite material and reduces the content of PA, the study has found that the melt flow rate of the composite material is not much different from the common ABS material [bib_ref] Mechanical and cytotoxicity properties of hybrid ceramics filled polyamide 12 filament feedstock..., Abdullah [/bib_ref] , and can still be used as one of the appropriate materials for 3D printing.
## 3d-printing for medical applications
3D printing technology can print the bone model of real patients and print bone substitutes and scaffolds to promote bone repair.
## Bone model
The internal organ model made by 3D printing technology [bib_ref] Use of 3D printing for medical ed-ucation models in transplantation medicine: A..., O'brien [/bib_ref] has the same anatomical and geometric characteristics as the natural organ. This transparency allows the surgeon to see anatomical structures such as blood vessels and lymphatics associated with parenchyma [bib_ref] Three-dimensional print of a liver for preoperative planning in living donor liver..., Zein [/bib_ref]. Because of the accuracy of the anatomy of the model, this is equivalent to providing a "surgical road map". The accurate visualization of the complex relationship between anatomical structures can more easily identify and plan the operation [bib_ref] Pediatric cardiac transplantation: Three-dimensional printing of anatomic models for surgical planning of..., Sodian [/bib_ref] , which is very important in the discussion of preoperative planning. Detailed preoperative planning can improve the success of the operation and reduce the overall risk to surrounding tissues and organs, as well as to patients. At the same time, it can also reduce the operation time, and may even improve the surgical results. This also means shorter hospitalization and lower medical costs.
The Medical 3D Image Printing Center of the Affiliated Hospital of Hebei University successfully completed the research and development of the full-color preoperative analysis model of C6 vertebral bone tumors designed with CAD assistance in 2017. It is the first time in the field of 3D printing of cervical vertebrae that the technique of avoiding errors in the second phase of the arteriovenous phase is introduced. Operators can complete surgical navigation in the operating room through a tablet computer or mobile phone. Meanwhile, the posterior spinal canal structure can be printed as a model after CAD design [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref]. It can completely open and observe the posterior structure of the vertebral body, which helps the operator to understand the lesion and surrounding situation more conveniently and intuitively. In 2020, the most complex model of comminuted fracture and the innovation of surgical planning was realized, such as the comminuted fracture model of knee joint and pelvis in [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] , which enabled doctors to use the reconstructed model to talk and discuss with the family members of the patient before operation, so that the communication with the patient and his family was clearer, and the simple language description was no longer the same as in the past, which made the family members of the patient confused.
## Bone replacement
Trauma and tumors of bone tissue can lead to large bone defects. The mechanical strength of natural bone was restored by in-situ filling with tissue engineering scaffolds.
Patients with bony organ defects face great psychological stress because the absence of bony tissue such as the nose, mandible and scapula can make a face or body image less acceptable to the general public [bib_ref] Assessment of health-related quality of life in Turkish patients with facial prostheses...., Atay [/bib_ref] [bib_ref] Quality of life of patients with implant-retained max-illofacial prostheses: A prospective and..., Nemli [/bib_ref]. Therefore, the printing of bone substitutes is very necessary.
Reconstruction of facial bones is particularly challenging in the field of in vitro bone implantation because of the important functional and aesthetic considerations of facial bones [bib_ref] Developments in reconstruction of midface and maxilla, Futran [/bib_ref]. A mandibular defect can adversely affect the patient's appearance, speech, chewing and social activities. Therefore, joint, occlusal, facial symmetry, donor area, vascular pedicle, soft tissue area and thickness should be considered in mandibular reconstruction [bib_ref] Clinical consid-erations in face transplantation, Pushpakumar [/bib_ref]. Successful reconstruction requires the restoration of symmetrical appear-
## Bone replacement
Trauma and tumors of bone tissue can lead to large bone defects. The mechanical strength of natural bone was restored by in-situ filling with tissue engineering scaffolds.
Patients with bony organ defects face great psychological stress because the absence of bony tissue such as the nose, mandible and scapula can make a face or body image less acceptable to the general public [bib_ref] Assessment of health-related quality of life in Turkish patients with facial prostheses...., Atay [/bib_ref] [bib_ref] Quality of life of patients with implant-retained max-illofacial prostheses: A prospective and..., Nemli [/bib_ref]. Therefore, the printing of bone substitutes is very necessary.
Reconstruction of facial bones is particularly challenging in the field of in vitro bone implantation because of the important functional and aesthetic considerations of facial bones [bib_ref] Developments in reconstruction of midface and maxilla, Futran [/bib_ref]. A mandibular defect can adversely affect the patient's appearance, speech, chewing and social activities. Therefore, joint, occlusal, facial symmetry, donor area, vascular pedicle, soft tissue area and thickness should be considered in mandibular reconstruction [bib_ref] Clinical consid-erations in face transplantation, Pushpakumar [/bib_ref]. Successful reconstruction requires the restoration of symmetrical appearance, sufficient masticatory space and correct joint position as much as possible, and the operation and control of bone position are required to be very high. The usual methods for mandibular defect repair usually include autograft or allograft [bib_ref] The rationale of mandible reconstruction in advanced oral cancer: Alloplastic material versus..., Chen [/bib_ref] [bib_ref] Limitations of autograft and allograft: New synthetic solutions, Betz [/bib_ref] and metal graft [bib_ref] High-strength porous bio-materials for bone replacement: A strategy to assess the interplay..., Arabnejad [/bib_ref]. In autologous and allogenic bone transplantation, most of the donor bone segments are straight (fibula, scapula, composite radius) or slightly curved (ilium crest) [bib_ref] Special considerations in virtual surgical planning for secondary accurate maxillary reconstruction with..., Shen [/bib_ref] [bib_ref] Current strategies in reconstruction of maxillectomy defects, Andrades [/bib_ref]. It is often necessary to have osteotomy on the donor bone, which causes secondary injury to the donor area and cannot fit the anatomical shape of the jaw perfectly [bib_ref] Comparative study of three kinds of fibula cutting guides in reshaping fibula..., Dong [/bib_ref]. Large differences in elastic modulus and relative density between metallic solids and human bones may lead to osteoporosis caused by stress shielding [bib_ref] High-strength porous bio-materials for bone replacement: A strategy to assess the interplay..., Arabnejad [/bib_ref]. This problem can be solved by implanting the mandibular substitute shown in [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] by 3D printing technology, and the density of the material can be increased in the high-stress area to match the human body. In the low-stress area, the density of the material is reduced, the weight of the implantation is reduced, and the waste of the material is reduced [bib_ref] 3D printing of titanium-coated gradient composite lattices for lightweight mandibular prosthesis, Xiao [/bib_ref]. Kang et al. developed a BTES for mandibular defects combined with a 3D-printed polyetheretherketone (PEEK) implant and the free vascularized fibula graft [bib_ref] 3D-printed PEEK implant for mandibular defects repair-A new method, Kang [/bib_ref]. Improved the stress barrier caused by the metallic materials commonly used in the past, and at the same time have sufficient mechanical strength to cope with the masticatory motion, as the picture shows, the bone tissue scaffold restores the patient's mandibular anatomy to a nearly normal state. 3D printing technology also has a good application in the bone repair of large joints. A benign fibrohistiocytic tumor of bone (BFH) is an invasive primary bone tumor. When the local resection is incomplete, the risk of recurrence is very high, so it needs to be completely resected. The main challenge for clinicians is bone reconstruction after tumor resection. 3D printing technology can help doctors solve this problem. Liu et al. report that PEEK replacement [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] was performed in patients with benign fibrohistiocytoma of the scapula by 3D printing. The imaging examination of the patients after the operation showed that the substitute was in a good anatomical position. The patient's shoulder joint activity was tested: 120 - on the lift, 90 - adduction, 50 - on the external rotation, and 70 - on internal rotation. the results of surgery are satisfactory [bib_ref] Application of 3D-printed PEEK scapula prosthesis in the treatment of scapular benign..., Liu [/bib_ref].
A common tissue replacement for the face is also a nose prosthesis. The loss of a nose can be devastating for the affected person, as changes in facial features can be accompanied by a variety of psychological, functional and cosmetic difficulties [bib_ref] Computed tomography for constructing custom nasal septal buttons, Price [/bib_ref]. Surgical reconstruction of such defects is usually limited by insufficient soft tissue, hard tissue residue, and vascular damage. Therefore, removable facial prostheses are an attractive and feasible alternative to help patients reintegrate into the community and resume their daily lives. Amjad et al. reported a case of nasal repair in which the nose was completely lost due to traffic accidents. Polyglycolic acid fibers were prepared into accurate human nasal alar shapes by three-dimensional printing technology [bib_ref] Direct 3D Printing of Flexible Nasal Prosthesis: Optimized Digital Workflow from Scan..., Nuseir [/bib_ref] [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref]. In addition, by coating with polylactic acid, the scaffold can obtain sufficient mechanical strength to maintain its original shape during cell culture until the ala cartilage shown in [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] is finally formed. After testing, the tissue engineering scaffold has a faster bone formation rate and better rigidity than natural cartilage tissue (Young modulus = 8.60 ± 2.19 MPa) [bib_ref] Tissue Engineering of Human Nasal Alar Cartilage Precisely by Using Three-Dimensional Printing, Xu [/bib_ref].
lives. Amjad et al. reported a case of nasal repair in which the nose was completely lost due to traffic accidents. Polyglycolic acid fibers were prepared into accurate human nasal alar shapes by three-dimensional printing technology [bib_ref] Direct 3D Printing of Flexible Nasal Prosthesis: Optimized Digital Workflow from Scan..., Nuseir [/bib_ref] [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref]. In addition, by coating with polylactic acid, the scaffold can obtain sufficient mechanical strength to maintain its original shape during cell culture until the ala cartilage shown in [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] is finally formed. After testing, the tissue engineering scaffold has a faster bone formation rate and better rigidity than natural cartilage tissue (Young modulus = 8.60 ± 2.19 MPa) [bib_ref] Tissue Engineering of Human Nasal Alar Cartilage Precisely by Using Three-Dimensional Printing, Xu [/bib_ref].
## Bone repair
BTES can load drugs, cytokines, and cells to promote in situ regeneration, so as to restore the normal anatomical structure of the human body.
In Arvind et al. research, the fibers with a diameter of about 1.6 mm were produced by melt treatment of the prepared biological composites by using a twin-screw micro extruder, and successfully used in 3D printing of representative middle finger bone [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] [bib_ref] Multifunctional nanohydroxyapatitepromoted toughened high-molecular-weight stereocomplex poly (lactic acid)-based bionano-composite for both 3D-printed..., Gupta [/bib_ref].
Jin et al. developed a 3D scaffold with embedded drug delivery microspheres to improve the enhancing ability of honeybee cells in bone regeneration. He used BMP-2 growth factor as the releasable drug in the microspheres and successfully observed good regeneration of bone tissue, as shown in Micro-CT result [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref]. It is also confirmed that the pore interconnection of bone tissue instead of scaffold is a critical factor in bone regeneration [bib_ref] Bone regeneration using a microstereo-lithography-produced customized poly(propylene fumarate)/diethyl fumarate photopolymer 3D scaffold..., Lee [/bib_ref].
When the Large bone defect is caused by bone tissue trauma, it is believed that inhibiting the inflammation of the wound site is beneficial to accelerate the regeneration of the wounded tissue. Li et al. prepared a biodegradable three-dimensional scaffold composed of the polymer matrix (polylactic acid and polyethylene glycol), ceramics (nanohydroxyapatite) and anti-inflammatory drugs (Dex). After the test, it was found that the scaffold has the ability to induce bone formation and anti-inflammatory ability, but because it is not loaded with growth factors or stem cells, its bone formation rate is slow [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] , and the material degradation rate is not ideal. However, this study proved that PLA composite scaffold can accelerate bone regeneration and can be used for future tissue engineering applications [bib_ref] Composite PLA/PEG/nHA/Dexamethasone Scaffold Prepared by 3D Printing for Bone Regeneration, Li [/bib_ref].We drew a rough flow chart to describe the experiment.
Cartilage tissue has no blood vessels inside, and its nutrient supply depends on vascular transport in the perichondrium, which leads to its limited self-repair ability. The use of tissue engineering, that is, the technology of implanting polymer scaffolds of progenitor cells to produce new tissue equivalents, has been considered. It is a potential means to produce replacement cartilage. Cartilage cells or their precursors can generate new cartilage tissue, so cellular components are commonly used [bib_ref] Cell and Biomimetic Scaffold-Based Approaches for Cartilage Regeneration, Sun [/bib_ref]. However, cartilage is a highly specialized connective tissue. Due to the nature of its composition and structure, its selfrepairing ability is limited. Circumstances can lead to degeneration after injury. Sun et al. reported a study, they used 3D printing technology to make a hydrogel loaded with mesenchymal stem cells, using rabbits as a model for evaluation, and found that they achieved better cartilage production and anti-inflammatory effects than the natural environment in the living body. We drew a rough flow chart to describe the experiment. Zhou et al. reported the first clinical case of 3D printing material successfully implanted in the human body to treat microtia. They inoculated autologous auricle chondrocytes on a multi-component (PCL/PGA/PLA) scaffold for 3D printing [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] , which not only perfectly restored the external anatomy shown, but also observed the successful regeneration of the patient's ear cartilage after the operation. regeneration [bib_ref] Bone regeneration using a microstereo-lithography-produced customized poly(propylene fumarate)/diethyl fumarate photopolymer 3D scaffold..., Lee [/bib_ref].
When the Large bone defect is caused by bone tissue trauma, it is believed that inhibiting the inflammation of the wound site is beneficial to accelerate the regeneration of the wounded tissue. Li et al. prepared a biodegradable three-dimensional scaffold composed of the polymer matrix (polylactic acid and polyethylene glycol), ceramics (nanohydroxyapatite) and anti-inflammatory drugs (Dex). After the test, it was found that the scaffold has the ability to induce bone formation and anti-inflammatory ability, but because it is not loaded with growth factors or stem cells, its bone formation rate is slow [fig_ref] Figure 4: 3D-printed for Medical Applications [/fig_ref] , and the material degradation rate is not ideal. However, this study proved that PLA composite scaffold can accelerate bone regeneration and can be used for future tissue engineering applications [bib_ref] Composite PLA/PEG/nHA/Dexamethasone Scaffold Prepared by 3D Printing for Bone Regeneration, Li [/bib_ref].We drew a rough flow chart to describe the experiment. Cartilage tissue has no blood vessels inside, and its nutrient supply depends on vascular transport in the perichondrium, which leads to its limited self-repair ability. The use of tissue engineering, that is, the technology of implanting polymer scaffolds of progenitor cells to produce new tissue equivalents, has been considered. It is a potential means to produce replacement cartilage. Cartilage cells or their precursors can generate new cartilage tissue, so cellular components are commonly used [bib_ref] Cell and Biomimetic Scaffold-Based Approaches for Cartilage Regeneration, Sun [/bib_ref]. However, cartilage is a highly specialized connective tissue. Due to the nature of its composition and structure, its selfrepairing ability is limited. Circumstances can lead to degeneration after injury. Sun et al.
# Conclusions
Three-dimensional (3D) printing makes it possible to manufacture any object highly precisely without causing any assembly or material waste. It is lighter than objects made by traditional methods and has the ability to meet the stress of different bone tissues in the human body. These implants or objects (extended into 3D objects) produced by 3D imaging are designed to match the patient's specific anatomical structure accurately, and at the same time, they can also function as cells and drug carriers. However, there are still many problems to be solved in applying 3D printing technology in the human body.
1. In vitro drug release experiments cannot fully simulate the physiological conditions in vivo. There is no guarantee that the same drug release curve as in vitro experiments will be achieved in humans, and only certain time points can be selected for drug release concentration detection. Therefore, medical devices with the ability to monitor drug delivery efficiency in real-time should be valued by researchers. This device can be mounted on a 3D printed tissue engineering scaffold, enter the body together, release the drug through external equipment, and always check whether the drug concentration is maintained within the treatment range.
2. When there is a large area defect in bone tissue, the strength of the synthetic scaffold is mostly unable to meet the normal stress requirements in the human body. Therefore, many researchers strengthen the function of the scaffold by loading growth factors or stem cells in the tissue scaffold. However, heterologous cells and proteins may cause an immune response and cause health problems. Therefore, the 3D printed scaffold should be biocompatible, matching the natural bone stress without triggering an immune response. At the same time, stem cell therapy is more expensive, and cost reduction is necessary. In addition, the ethical issue of implantation in the body is another limitation in the whole process.
3. There are different types of 3D printing technologies on the market, each with its advantages and limitations. Traditional manufacturing methods print materials with rough surface texture and low overall strength. 3D printing uses accurate CAD models to achieve precise printing, and post-processing of implant stents manufactured by 3D technology can solve this problem, but it will increase the cost. So this technology is only suitable for customization, not mass production systems. The cost of the machine and the requirements of professional and technical personnel are also the main limitations of this technology. From this, it can be seen that how to make 3D printing maintain accuracy while carrying out assembly line production should be one of the focuses of future research. [bib_ref] Autologous Bone Graft: Properties and Techniques, Pape [/bib_ref]. Although the 3D printed scaffold has micro/nano-sized pores on the surface, it can be used as a carrier for cells, cytokines, drugs, and other substances. It can also play a role in transporting nutrients and supporting cell growth, but it cannot be controlled. The pore size and spatial distribution limit the application of 3D printed stents, which is something we should pay attention to.
[fig] Figure 1: 3D printing process diagram: (a1) SLS; (a2) SLA; (a3) LAD; (b1) Extrusion Bioprinting; (b2) Inkjet Bioprinting; (b3) FDM. [/fig]
[fig] Figure 2: Repair processes of bone defeat. (a) Inflammatory stage; (b) Soft callus stage; (c) Hard callus stage; (d) Bone shaping stage. [/fig]
[fig] Figure 3: Pore structure diversity of 3D printed bone scaffolds. (a) Based on the observation and analysis of lotus root structure and bone trabecular structure, four kinds of porous scaffolds were proposed; reproduced with permission from Zhipeng, H. et al.[66], published by Composites Science and Technology 2020. (b) Five pore geometries were selected (triangular prism with a rounded and a flat profile, cube, octagonal prism, sphere) and seven porosities (up to 80%), on the basis of which 70 models were created for finite element analyses;reproduced with permission from Piotr, P. et al.[68], published by Materials 2020; (c) The porous g-HNTs/g-MgOs/PLLA composite scaffolds with large and small pores and honeycomb structure; reproduced with permission from Kun, L. et al.[69], published by Composites Part B: Engineering 2020; (d) SF/COL composite scaffolds with random pores, radially aligned pores or axially aligned pores; reproduced with permission from Xue, F. et al.[70], published by Journal of Materials Chemistry B 2020; (e,f) Hollow channels structure; reproduced with permission from Wenjie, Z. et al.[71], published by Biomaterials 2017; reproduced with permission from Chun, F. et al.[72], published by Advanced Science 2017. [/fig]
[fig] Polymers 2022 ,: 13, x FOR PEER REVIEW 21 of 32 operation, so that the communication with the patient and his family was clearer, and the simple language description was no longer the same as in the past, which made the family members of the patient confused. [/fig]
[fig] Figure 4: 3D-printed for Medical Applications. (a) Preoperative full-color analysis model of a patient with C6 vertebral bone tumor; (b) Preoperative modeling of comminuted fractures in the knee and pelvis; (c) Gradient lattice mandibular prosthesis prepared by metal coated polymer; reproduced with permission from Xiao, R. [161], published by Composites Part B: Engineering 2020, 193, 108057; (d) Scapular prosthesis made with PEEK; reproduced with permission from Liu, D. [162], published by J Bone Oncol 2018, 12, 78-82; (e) Middle phalangeal bone implant prepared by sPLA/N-HAP composite; reproduced with permission from Nuseir, A. [78], published by J Prosthodont 2019, 28, 10-14; (f) Micro-ct showing BMP-2 loaded SFF stent promoting bone healing;reproduced with permission from Gupta, A. [163], published by ACS omega 2017, 2, 4039-4052; (g) human ear-shaped cartilage reconstruction based on tissue engineering scaffolds; reproduced with permission from Zhou, G. [164], published by EBioMedicine 2018, 28, 287-302; (h) Micro-CT images of bone regeneration in the defects of BMP-2-loaded SFF scaffold; reproduced with permission from Lee, J.W. [79], published by Biomaterials 2011, 32, 744-752; (i) Micro-CT images of bone regeneration in the defects of Biodegradable 3D scaffolds; reproduced with permission from Li, X. [165], published by Macromol Biosci 2018, 18, e1800068. [/fig]
[fig] Figure 5: (a) The clinical implantation and post-operative effect of 3D-printed PEEK implant; reproduced with permission from Kang, J [86], published by J Mech Behav Biomed Mater 2021, 116, 104335; (b) Composite tissue scaffold loaded with cartilage cells implanted in the human body; reproduced with permission from Zhou, G. [178], published by EBioMedicine 2018, 28, 287-302. [/fig]
[fig] Figure 6: (a) regeneration of skull defect in rats; (b) articular cartilage regeneration in rabbits. [/fig]
[fig] Author: Contributions: Y.W. provided the direction of the review and gave guidance. J.L. provided technical advice and guidance; S.C. reviewed the documents and provided technical advice; Y.X. completed all the pictures and wrote four chapters of the review; F.Z. wrote three chapters of the review; W.Z. assisted in collating references. All authors have read and agreed to the published version of the manuscript. Funding: This work is supported by The Project National United Engineering Laboratory for Advanced Bearing Tribology (No. 202105), Henan University of Science and Technology, and Hebei Naturel Scientists Fundation (No. H2021201005). [/fig]
[table] Table 1 Table 1: Application of polymer materials, applicable processing technology, advantages and disadvantages. Cont. [/table]
[bib_ref] Recent Progress on 3D-Printed Polylactic Acid and Its Applications in Bone Repair, Chen [/bib_ref] |
Structural alterations of the intestinal epithelial barrier in Parkinson’s disease
Functional and morphological alterations of the intestinal epithelial barrier (IEB) have been consistently reported in digestive disorders such as irritable bowel syndrome and inflammatory bowel disease. There is mounting evidence that Parkinson's disease (PD) is not only a brain disease but also a digestive disorder. Gastrointestinal involvement is a frequent and early event in the course of PD, and it may be critically involved in the early development of the disease. We therefore undertook the present survey to investigate whether changes in the IEB function and/or morphology occur in PD. Colonic biopsies were performed in 31 PD patients and 11 age-matched healthy controls. The para-and transcellular permeability were evaluated by measuring sulfonic acid and horseradish peroxidase flux respectively, in colonic biopsies mounted in Ussing chambers. The expression and localization of the two tight junctions proteins ZO-1 and occludin were analyzed by Western blot and immunofluorescence, respectively. The para-and transcellular permeability were not different between PD patients and controls. The expression of occludin, but not ZO-1, was significantly lower in colonic samples from PD patients as compared to controls and the cellular distribution of both proteins was altered in colonic mucosal specimens from PD patients. Our findings provide evidence that the IEB is morphologically altered in PD and further reinforce the potential role of the gastrointestinal tract in the initiation and/or the progression of the disease.
# Introduction
The intestinal epithelium forms a regulated barrier, known as intestinal epithelial barrier (IEB), between the blood circulation and the contents of the intestinal lumen [bib_ref] Epithelial barriers in homeostasis and disease, Marchiando [/bib_ref]. It prevents the passage of noxious contents while allowing the absorption and secretion of nutrients [bib_ref] Epithelial barriers in homeostasis and disease, Marchiando [/bib_ref]. Penetration of this barrier occurs via two routes, either between epithelial cells via the paracellular pathway, or through epithelial cell via the transcellular pathway [bib_ref] Epithelial barriers in homeostasis and disease, Marchiando [/bib_ref]. Among the most important structures of the intestinal barrier are the epithelial tight junctions (TJs) that connect adjacent enterocytes together to determine paracellular permeability through the lateral intercellular space [bib_ref] Regulation of intestinal epithelial permeability by tight junctions, Suzuki [/bib_ref]. They are formed by transmembrane proteins such as claudins and occludins connected to the actin cytoskeleton via high molecular weight proteins called zona occludens (ZO-1, ZO-2 and 3) [bib_ref] Regulation of intestinal epithelial permeability by tight junctions, Suzuki [/bib_ref]. Increased permeability of the IEB along with changes in the expression levels of TJs proteins have been consistently reported in several digestive disorders such as inflammatory bowel disease [bib_ref] Increased permeability of macroscopically normal small bowel in Crohn's disease, Peeters [/bib_ref] [bib_ref] Intestinal permeability in patients with Crohn's disease and their healthy relatives, Katz [/bib_ref] and irritable bowel syndrome [bib_ref] Impaired intestinal barrier integrity in the colon of patients with irritable bowel..., Piche [/bib_ref] [bib_ref] The expression and the cellular distribution of the tight junction proteins are..., Bertiaux-Vandaële [/bib_ref].
It has become evident over the last 20 years that PD is a gut disorder (reviewed in [bib_ref] Parkinson disease: the enteric nervous system spills its guts, Derkinderen [/bib_ref]. Gastrointestinal symptoms occur in almost every PD patient at some point and are among the most debilitating non-motor features of the disease [bib_ref] Gastrointestinal features of Parkinson's disease, Cloud [/bib_ref]. These clinical data have been supported by post mortem studies that demonstrated the presence of Lewy bodies and neurites in the enteric neurons in nearly every case examined pathologically [bib_ref] Parkinson's disease: the presence of Lewy bodies in Auerbach's and Meissner's plexuses, Wakabayashi [/bib_ref] [bib_ref] Multi-organ distribution of phosphorylated alpha-synuclein histopathology in subjects with Lewy body disorders, Beach [/bib_ref]. The German pathologist Heiko Braak suggested that the appearance of Lewy pathology in enteric neurons develop early in the course of disease, prior to the involvement of the central nervous system [bib_ref] Gastric alpha-synuclein immunoreactive inclusions in Meissner's and Auerbach's plexuses in cases staged..., Braak [/bib_ref]. This led him to suggest that the gastrointestinal tract might be a portal of entry for a putative pathogen that would breach the IEB to induce the formation of Lewy bodies and neurites in the enteric neurons [bib_ref] Gastric alpha-synuclein immunoreactive inclusions in Meissner's and Auerbach's plexuses in cases staged..., Braak [/bib_ref].
The high prevalence of gastrointestinal symptoms and pathology in PD and the possible derangement of gastrointestinal permeability in the pathogenesis of the disease prompted several groups to investigate IEB permeability in parkinsonian patients. The three studies, which have been carried out to date have all used absorption of sugar probes as a means to investigate non-invasively the paracellular permeability [bib_ref] Intestinal permeability, leaky gut, and intestinal disorders, Hollander [/bib_ref]. These studies, which have included a small number of patients, led to conflicting results. Two studies found a pattern of sugar absorption reminiscent of small intestine hyperpermeability in a subset of patients [bib_ref] Increased intestinal permeability and Parkinson disease patients: chicken or egg?, Salat-Foix [/bib_ref] [bib_ref] Intestinal permeability and orocaecal transit time in elderly patients with Parkinson's disease, Davies [/bib_ref] while the third one showed an increase in sucralose excretion without changes in the lactulose/ mannitol ratio, a pattern consistent with increased colonic permeability [bib_ref] Increased intestinal permeability correlates with sigmoid mucosa alpha-synuclein staining and endotoxin exposure..., Forsyth [/bib_ref]. We therefore undertook the present research to analyze in more details the IEB in PD. To this end, a functional and structural characterization of the IEB was performed in colonic biopsies from PD patients.
# Materials and methods
## Subjects
A total of 42 subjects participated in this study, 31 PD patients and 11 healthy controls. PD patients aged 43-74 years were recruited from the movement disorder clinic at Nantes University Hospital, France. Diagnosis of PD was made according to criteria provided by the United Kingdom Parkinson's Disease Survey Brain Bank [bib_ref] Improved accuracy of clinical diagnosis of Lewy body Parkinson's disease, Hughes [/bib_ref]. Collected demographic data included gender, age at onset and disease duration, as well as age at colonoscopy. Complete drug history was obtained, and an approximation of the cumulative dose of L-dopa was made based on the equation developed by Parkkinen and collaborators [bib_ref] Does levodopa accelerate the pathologic process in Parkinson disease brain, Parkkinen [/bib_ref]. Control subjects were healthy subjects who had a normal colonoscopy performed for colorectal cancer screening. All controls subjects underwent a detailed neurological examination to rule out PD symptoms and cognitive deficiency. Controls and PD patients were excluded if they suffered from irritable bowel syndrome and/or anorectal dysfunction. The study protocol was approved by the local Committee on Ethics and Human Research (Comité de Protection des Personnes Ouest VI) and registered on Clin-icalTrials.gov (identifier NCT01748409). Written informed consent was obtained from each patient and from each normal volunteer according to the principles of Helsinki.
## Endoscopic procedure and colonic biopsies
For each subject, nine biopsies were taken in the sigmoid/ descending colon during the course of a rectosigmoidoscopy for PD patients and during a colonoscopy for control subjects. Five biopsies were immersed in 4°C Hank's Balanced Salt Solution (Life Technologies, Saint Aubin, France): three of these biopsies were immediately processed for the assessment of para-and transcellular permeability in Ussing chambers while the two other biopsies were used for immunohistochemistry experiments. Two biopsies were stored at −80°C in lysis buffer RA1 (Macherey-Nagel, Hoerdt, France) with 1% (v/v) β-mercaptoethanol (Sigma, Saint Quentin Fallavier, France) for further analysis by immunoblotting. The two remaining biopsies were snap frozen in liquid nitrogen at the time of collection and kept at −80°C.
## Para-and transcellular permeability of colonic biopsies in ussing chambers
Three biopsies were mounted in Ussing chambers (World Precision Instruments; WPI, Hertfordshire, UK) exposing a surface of 0.011 cm 2 . Tissues were bathed on each side with 3 ml of F12 supplemented Dulbecco's Modified Eagle medium (Invitrogen, France) containing 0.1% (v/v) fetal bovine serum, 200 mM Glutamine and 45 g/L of NaHCO 3 . The medium was continuously oxygenated and maintained at 37°C by a gas flow (95% O 2 /5% CO 2 ). After a 30 min baseline period, 275 μL of apical medium was replaced with 200 μL of media containing 1 mg/mL of fluorescein-5,6-sulfonic acid (molecular weight: 400 Da) (Life Technologies) for a final concentration of 0.1 mg/mL to assess paracellular permeability. Seventy-five microliters of media with 10 mg/mL of Horse Radish Peroxydase (HRP) (Sigma) were also added to the basolateral chamber for a final concentration of 0.375 mg/mL to measure transcellular permeability in a subset of PD patients and control subjects. The fluorescence level of basolateral aliquots of 150 μl, reflecting paracellular transit from the luminal surface was measured every 30 min over a 3-hour period using a fluorimeter (Varioskan®, ThermoFisher Scientific, Cillebon sur Yvette, France). HRP quantities in the basolateral chamber, reflecting transcellular transit from the apical surface, was measured using an enzymatic activity assay with 3,3' ,5,5'-tetramethylbenzidine reagent (BD Bioscience, Le Pont de Claix, France). Paracellular and transcellular permeabilities were determined by averaging the gradient of change in fluorescence intensity over time in the three biopsies that were analyzed per patient, using a linear regression fit model (GraphPad Prism 5, La Jolla, USA).
## Western blot
For the analysis of ZO-1 expression, total proteins from the 2 biopsies stored in RA1 buffer were precipitated and prepared for Polyacrylamide Gel Electrophoresis (PAGE) using protein precipitator and resuspension buffer (Protein solving buffer and (tris(2-carboxyethyl) phosphine) TCEP reducing agent, PSB/TCEP) from NucleoSpin Triprep Kit (Macherey-Nagel, Hoerdt, France) according to the manufacturer's instructions. For experiments on the transmembrane protein occludin, the two dry frozen biopsies stored at −80°C were lysed in UTC buffer (7 M Urea, 2 M Thiourea and 4% CHAPS) containing a protease inhibitor cocktail (Complete™, Roche, Meylan, France) using the "Precellys 24" tissue homogenizer (Bertin technologies, Saint Quentin-en-Yvelines, France) and followed by sonication with "vibracell 75 186" device (Sonics, Newton CT, USA). Equal amounts of lysate were separated using the Invitrogen NuPage Novex 3-8% Tris-Acetate Midi Protein Gels™ for ZO-1 or NuPage Novex 4-12% Bis-Tris MidiGels™ for occludin before electrophoretic transfer to nitrocellulose membranes with the iBlot™ Dry Blotting System also from Invitrogen. Membranes were processed for immunoblotting using rabbit polyclonal anti-ZO-1 (1:500, Life Technologies) and rabbit anti-occludin (1:250, Abcam, Paris, France) antibodies and the relevant immunoreactive bands were quantified as previously described [bib_ref] Enteric GFAP expression and phosphorylation in Parkinson's disease, Clairembault [/bib_ref].
## Microdissection and immunohistochemistry
Microdissection was performed as previously described [bib_ref] Routine colonic biopsies as a new tool to study the enteric nervous..., Lebouvier [/bib_ref] in two out of the nine biopsies taken per patient. Each whole-mount preparation of submucosa obtained from a single biopsy was permeabilized for 3 hours in phosphate buffered saline (PBS)/NaN 3 containing 1% (v/v) Triton X-100 and 10% (v/v) horse serum and then incubated with antibodies against phosphorylated alpha-synuclein (1:5000, WAKO, Osaka, Japan) and PGP9.5 (1:10,000; Ultraclone Limited, UK). Each whole-mount preparation of mucosa was treated for 24 h with Scale A2 solution composed of 4 M urea, 10% (w/v) glycerol and 0.1% (v/v) Triton X100 [bib_ref] Scale: a chemical approach for fluorescence imaging and reconstruction of transparent mouse..., Hama [/bib_ref] then incubated with rabbit polyclonal antibodies to ZO-1 (1:100, Life Technologies) and occludin (1:100, Abcam). Suitable secondary antibodies conjugated to Alexa Fluor 488 and 594 were used (Invitrogen, Cergy-Pontoise, France). Following incubation with the secondary antibodies, the mucosa samples were treated for 10 minutes with a solution of 0.3% (w/v) of Sudan Black B powder (Sigma) dissolved in 70% (v/v) ethanol, then washed extensively with PBS. Whole specimen of submucosa and mucosa were viewed under an Axio Zoom.V16 stereomicroscope (Zeiss, Marly Le Roi, France). All samples were deidentified and studied in a blinded manner. For the analysis of ZO-1 and occludin immunofluorescence, the percentage of morphologically normal crypts per biopsy was calculated and the following classification was used: 'normal' , more than 2/3 of morphologically normal crypts; 'mild disruption': between 1/3 and 2/3 of morphologically normal crypts; 'disrupted': less than 1/3 of morphologically normal crypts.
## Statistics
All data are given as the mean ± standard error of the mean (SEM). For comparisons of means between groups, a Mann-Whitney test was performed. Differences were deemed statistically significant if p < 0.05.
# Results
A total of 31 PD patients and 11 controls were included. [fig_ref] Table 1: Main clinical characteristics of PD patients [/fig_ref] shows the main clinical and demographic features of all patients. Age and sex did not differ significantly between PD patients and control subjects (mean age was 64.2 ± 2.1 for PD and 60.6 ± 1.4 for controls, p = 0.25; 22/31 male in PD group and 6/11 male in control group, p = 0.13).
## Para-and transcellular permeability are unaffected in pd
In a first set of experiments, we evaluated whether IEB is functionally altered in PD patients. The para-and transcellular permeability of colonic biopsies were measured in Ussing chambers in both PD patients and control subjects using sulfonic acid and HRP, respectively. No difference in the sulfonic acid flux was observed between PD patients and control subjects (n = 31 and 11, respectively; p = 0.65) [fig_ref] Figure 1: Comparison of para-and transcellular permeability in PD patients and healthy controls [/fig_ref]. HRP flux was also comparable between PD subjects and healthy controls (n = 21 and 9, respectively; p = 0.39) [fig_ref] Figure 1: Comparison of para-and transcellular permeability in PD patients and healthy controls [/fig_ref]. Although not statistically different from controls, the sulfonic acid and HRP flux values were heterogeneous between PD patients [fig_ref] Figure 1: Comparison of para-and transcellular permeability in PD patients and healthy controls [/fig_ref]. We thus investigated if the main clinical features of the disease had any influence on IEB permeability. We did not observe any correlation between age, disease duration or lifetime cumulative dose of L-DOPA and the values of sulfonic acid or HRP flux [fig_ref] Table 2: Spearman's correlation with sulfonic acid [/fig_ref]. There is a growing body of evidence supporting a key role for submucosal enteric neurons in the regulation of IEB functions [bib_ref] Changes in chemical coding of myenteric neurones in ulcerative colitis, Neunlist [/bib_ref] [bib_ref] Human submucosal neurones regulate intestinal epithelial cell proliferation: evidence from a novel..., Toumi [/bib_ref] [bib_ref] Muscarinic acetylcholine receptor activation increases transcellular transport of macromolecules across mouse and..., Cameron [/bib_ref]. This prompted us to study if the flux values of sulfonic acid and HRP were related to the presence of Lewy bodies and Lewy neurites in the submucosa. To this end, two biopsies per patient were immunohistochemically assessed for the presence of Lewy pathology using antibodies against phosphorylated alpha-synuclein and PGP9.5 [fig_ref] Figure 2: Sulfonic acid and HRP flux in PD in patients with and without... [/fig_ref]. A biopsy was deemed positive when containing at least one inclusion immunoreactive for both phosphorylated alphasynuclein and PGP9.5 [fig_ref] Figure 2: Sulfonic acid and HRP flux in PD in patients with and without... [/fig_ref]. A patient was noted as positive when at least one of the two biopsies displayed inclusion(s). In accordance with our previous reports [bib_ref] Colonic biopsies to assess the neuropathology of Parkinson's disease and its relationship..., Lebouvier [/bib_ref] [bib_ref] A comparison between rectal and colonic biopsies to detect Lewy pathology in..., Pouclet [/bib_ref] , the intraneuronal inclusions found in the submucosal plexus were chiefly observed in the neuronal processes and thus reminiscent of Lewy neurites [fig_ref] Figure 2: Sulfonic acid and HRP flux in PD in patients with and without... [/fig_ref]. Twenty-three out of 31 PD patients were positive for phosphorylated alpha-synuclein inclusions. All control subjects were devoid of inclusions. The values of sulfonic acid and HRP flux were not different between PD patients with or without inclusions [fig_ref] Figure 2: Sulfonic acid and HRP flux in PD in patients with and without... [/fig_ref].
## Expression of the tight junction protein occludin is decreased in pd
We next investigated whether structural changes in the IEB occur in PD. The expression levels of the TJs proteins ZO-1 and occludin were analyzed by Western blot in colonic biopsies from PD subjects and healthy controls. A significant decrease in the expression of occludin was observed in colonic samples of PD patients as compared to controls [fig_ref] Figure 3: Expression of TJs proteins in colonic biopsies from patients with Parkinson's disease [/fig_ref]. A doublet band of approximately 220 kDa was observed on Western blot with antibodies against ZO-1 [fig_ref] Figure 3: Expression of TJs proteins in colonic biopsies from patients with Parkinson's disease [/fig_ref]. As previously shown, these two bands most likely represent the two ZO-1 isoforms [bib_ref] A dynamic ratio of the alpha + and alpha-isoforms of the tight..., Ciana [/bib_ref] [bib_ref] Localization and differential expression of two isoforms of the tight junction protein..., Willott [/bib_ref]. By contrast to occludin, no change in ZO-1 expression levels was observed in PD whether the two bands were quantified together [fig_ref] Figure 3: Expression of TJs proteins in colonic biopsies from patients with Parkinson's disease [/fig_ref] or separately (data not shown).
## Cellular distribution of the tjs proteins is altered in pd
The cellular distribution of occludin and ZO-1 was further investigated by immunofluorescence in 8 controls and 31 PD subjects. Samples from 3 controls were excluded because the mucosa was too small and/or too damaged to allow a reliable analysis of the TJs morphology. A mean of 96.4 crypts per biopsy were analyzed. We observed differences in the cellular distribution of both ZO-1 and occludin between PD patients and controls [fig_ref] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls [/fig_ref]. A normal and typical reticular pattern of occludin and ZO-1 staining was observed in the colonic samples of 6 out of 8 controls [fig_ref] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls [/fig_ref] and C, Additional file 1) and in only 9/31 PD patients [fig_ref] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls [/fig_ref] and D, Additional file 1). TJs morphology was disrupted and irregularly distributed in the mucosa of 1 out of 8 controls [fig_ref] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls [/fig_ref] and C, Additional file 1) and in 14/ 31 PD patients [fig_ref] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls [/fig_ref] and D, Additional file 1). An occasional and mild disruption of TJs morphology was observed in the remaining control subject and in 8/31 PD samples [fig_ref] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls [/fig_ref] and Additional file 1). An increased staining of occludin in the cytoplasm of colonic enterocytes, suggestive of protein internalization, was observed in PD samples as opposed to the healthy group where occludin was mostly located in the TJs [fig_ref] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls [/fig_ref] and D). Worthy of note was the presence of moderate to severe TJs disorganization in the 5 patients who had never received levodopa, suggesting that the altered TJs morphology was not related to chronic levodopa intake (Additional file 2 and [fig_ref] Table 1: Main clinical characteristics of PD patients [/fig_ref].
# Discussion
The TJs are intercellular protein complex located at the apical portions of the lateral membranes of epithelial cells which play a key role in the regulation of IEB paracellular permeability. They are composed of transmembrane proteins, such as claudins and occludin and a wide spectrum of cytosolic proteins among which is ZO-1 [bib_ref] Epithelial barriers in homeostasis and disease, Marchiando [/bib_ref]. By showing a decrease in occludin expression along with TJs disorganization, our study is the first to provide evidence that the IEB is structurally altered in PD. Previous studies have shown intestinal tissue expression and distribution of occludin to be markedly decreased in patients with intestinal permeability disorders, including inflammatory bowel disease [bib_ref] Inflammatory bowel disease is associated with changes of enterocytic junctions, Gassler [/bib_ref] and irritable bowel syndrome [bib_ref] The expression and the cellular distribution of the tight junction proteins are..., Bertiaux-Vandaële [/bib_ref]. Recent data from genetic and epidemiological studies provided support for an association between diseases of the gastrointestinal tract and the susceptibility to developing PD. The CARD15 gene known to be associated with Crohn's disease is over-represented in patients with PD [bib_ref] CARD15 variants in patients with sporadic Parkinson's disease, Bialecka [/bib_ref] ; vice versa, the Leucine-rich repeat kinase 2 (LRRK2) gene, a causative PD mutation, was recently identified as a major susceptibility gene for Crohn's disease by genome-wide association studies [bib_ref] Genome-wide association defines more than 30 distinct susceptibility loci for Crohn's disease, Barrett [/bib_ref]. Moreover, patients with irritable bowel syndrome are almost 50% more likely to develop PD than people who are free of this gastrointestinal disorder [bib_ref] Irritable bowel syndrome correlates with increased risk of Parkinson's disease in Taiwan, Lai [/bib_ref]. Our results show that the two disorders also share similarities at the molecular levels further supporting the assumption that irritable bowel syndrome may actually belong to early signs of gastrointestinal involvement in PD [bib_ref] Gastrointestinal manifestations in Parkinson's disease: prevalence and occurrence before motor symptoms, Cersosimo [/bib_ref]. They also support the hypothesis that the brain gut-axis might be critically involved in the pathophysiology of both disorders [bib_ref] Parkinson disease: the enteric nervous system spills its guts, Derkinderen [/bib_ref] [bib_ref] Brain and gut interactions in irritable bowel syndrome: new paradigms and new..., Coss-Adame [/bib_ref]. Studies of intestinal permeability in humans have mainly been carried out with in vivo techniques, usually with oral ingestion of various sugar probes and measurement of urinary excretion [bib_ref] Intestinal permeability, leaky gut, and intestinal disorders, Hollander [/bib_ref]. To date, the three studies that attempted to evaluate intestinal permeability in PD using this technique have provided only preliminary and conflicting results. Two of these studies focused on the lactulose/mannitol ratio, which evaluate small intestinal permeability. As a group, the 15 PD patients studied by Davies and collaborators had a significant increase in the lactulose/mannitol ratio when compared to age and sex matched controls, but individual results in both groups were highly overlapping [bib_ref] Intestinal permeability and orocaecal transit time in elderly patients with Parkinson's disease, Davies [/bib_ref]. Salat-Foix et al. showed that the lactulose/mannitol ratio was only marginally higher in 3 out of 12 PD patients [bib_ref] Increased intestinal permeability and Parkinson disease patients: chicken or egg?, Salat-Foix [/bib_ref]. In addition to lactulose/mannitol ratio, Forsyth et al. also used sucralose absorption for the assessment of colon permeability in 9 PD patients and 10 controls. They did not observe any difference in the lactulose/mannitol ratio between the two groups but found a significantly greater permeability to sucralose in PD subjects [bib_ref] Increased intestinal permeability correlates with sigmoid mucosa alpha-synuclein staining and endotoxin exposure..., Forsyth [/bib_ref]. The inconsistent results on intestinal permeability in PD obtained with sugar probes prompted us to measure IEB permeability by another technique, namely Ussing chambers, in a larger sample size. Although less commonly used than sugar absorption, the Ussing chambers has proven to be a reliable and effective tool to measure IEB permeability of gastrointestinal biopsies either paracellularly or transcellularly over a 3 hour period [bib_ref] Endoscopic biopsies in Ussing chambers evaluated for studies of macromolecular permeability in..., Wallon [/bib_ref]. Using this approach, we showed that there were no significant differences in para-and transcellular permeabilities between PD subjects and controls. Nevertheless, the values of paracellular permeability as assessed by the sulfonic acid flux in Ussing chamber were highly heterogeneous between PD patients, some displaying a level comparable to controls while others had a more than a 2.5 fold increase in sulfonic acid flux. These data suggest that increased colonic permeability may be a feature for a subset of PD patients, as already reported when sugar probes were used [bib_ref] Increased intestinal permeability correlates with sigmoid mucosa alpha-synuclein staining and endotoxin exposure..., Forsyth [/bib_ref]. The factors responsible for this heterogeneity still remain to be determined, as we did not observe any correlation between age, cumulative dose of L-Dopa disease duration and the severity of altered permeability.
On the surface, our results may seem contradictory, as the decreased expression of occludin observed in PD patients was not accompanied by changes in paracellular permeability. Occludin is a tetraspan protein with two extracellular loops, which homophilically interact with the adjacent cells [bib_ref] Regulation of intestinal epithelial permeability by tight junctions, Suzuki [/bib_ref]. Its role on IEB has been debated since its initial discovery in 1993 [bib_ref] Occludin: a novel integral membrane protein localizing at tight junctions, Furuse [/bib_ref]. Initial studies strongly suggested that occludin was not required for the TJs formation or the maintenance of barrier function as occludin knockout mice lacked any noticeable defect in intestinal TJs morphology or barrier function [bib_ref] Complex phenotype of mice lacking occludin, a component of tight junction strands, Saitou [/bib_ref]. This has been recently challenged in an elegant study published by Al-Sadi et al. [bib_ref] Occludin regulates macromolecule flux across the intestinal epithelial tight junction barrier, Al-Sadi [/bib_ref]. The purpose of their research was to better delineate the involvement of occludin in IEB by studying the transepithelial flux of various-sized probes after knocking down occludin both in vitro and in vivo. They showed that the occludin knock down caused a marked increase in the flux rates of macromolecules above 5 kDa such as inulin and dextran but had only modest effect on flux of smaller-sized probes under 200 Da such as mannitol and urea [bib_ref] Complex phenotype of mice lacking occludin, a component of tight junction strands, Saitou [/bib_ref]. Fluorescein-5,6-sulfonic acid, which was used for the assessment of paracellular permeability in our study has a molecular weight of 400 Da, likely to be too small for detecting defects in IEB permeability induced by a mere down regulation of occludin. This may explain the lack of significant changes in IEB permeability observed in PD patients in our study in spite of the occurrence of structural changes.
The question arises as to what might be the clinical relevance of our experimental findings. A current theory, the so-called Braak's theory, assumes that PD originates in the gastrointestinal tract [bib_ref] Gastric alpha-synuclein immunoreactive inclusions in Meissner's and Auerbach's plexuses in cases staged..., Braak [/bib_ref]. Braak and co-workers suggested that the appearance of Lewy pathology occurs in the earliest stage of PD in both the enteric nervous system and the dorsal motor nucleus of the vagus [bib_ref] Gastric alpha-synuclein immunoreactive inclusions in Meissner's and Auerbach's plexuses in cases staged..., Braak [/bib_ref] [bib_ref] Staging of brain pathology related to sporadic Parkinson's disease, Braak [/bib_ref]. This led Braak to postulate that a pathogen may breach the IEB to trigger Lewy pathology in the terminal axons of the enteric neurons, further spreading to the central nervous system via the vagal preganglionic innervation of the gut [bib_ref] Gastric alpha-synuclein immunoreactive inclusions in Meissner's and Auerbach's plexuses in cases staged..., Braak [/bib_ref] [bib_ref] Idiopathic Parkinson's disease: possible routes by which vulnerable neuronal types may be..., Braak [/bib_ref]. In light of these considerations, our results demonstrating altered intestinal TJs structure in PD gain in importance as the down regulation of occludin may favor the entry of a putative pathogen. This must be however balanced, as the stomach in contrast to the colon appears to be the most suitable target for the pathologic insult to occur in Braak's scenario. Several studies have indeed described that Lewy pathology is distributed following a rostro-caudal gradient in PD, with the lower esophagus and stomach having the greatest involvement and the colon and rectum the lowest [bib_ref] Parkinson's disease: the presence of Lewy bodies in Auerbach's and Meissner's plexuses, Wakabayashi [/bib_ref] [bib_ref] Multi-organ distribution of phosphorylated alpha-synuclein histopathology in subjects with Lewy body disorders, Beach [/bib_ref] , a distribution that parallels the vagal innervation of the gastrointestinal tract [bib_ref] Vagal efferent projections: viscerotopy, neurochemistry and effects of vagotomy, Hopkins [/bib_ref]. Further studies are therefore warranted to analyze the mucosal barrier permeability and morphology in gastric and duodenal samples from PD patients.
# Conclusions
In conclusion, we provide evidence for the first time that morphological changes in the IEB occur in PD patients. Our results further reinforce the possible role of the gastrointestinal tract in the pathophysiology of PD. Further work is needed to determine if occludin down regulation in the gut might facilitate the spreading of PD pathology in the enteric nervous system and in the brain.
[fig] Figure 1: Comparison of para-and transcellular permeability in PD patients and healthy controls. (A) For the evaluation of paracellular permeability, the flux of sulfonic acid (SA flux) was measured in colonic biopsies mounted in Ussing chambers, expressed in arbitrary units (AU), in PD patients (n = 31) and controls (CTRL, n = 11). No significant changes were observed between the two groups (p = 0.65). (B) For the evaluation of paracellular permeability, the flux of horseradish peroxidase (HRP flux) was measured in colonic biopsies mounted in Ussing chambers, expressed in ng/mL/min, in PD patients (n = 21) and controls (CTRL, n = 9). No significant changes were observed between the two groups (p = 0.39). [/fig]
[fig] Figure 2: Sulfonic acid and HRP flux in PD in patients with and without enteric Lewy pathology. Representative photomicrograph depicting multiple Lewy neurites in a whole-mount of submucosa immunoreactive for PGP9.5 (A) that was also positive for phosphorylated alpha-synuclein (B and C) Scale bar: 200 μm. (D) The values of sulfonic acid flux was not different between PD patients with (PS+) or without (PS-) phospho-synuclein immunoreactive neurites (p = 0.93). (E) The values of HRP flux was not different between PD patients with (PS+) or without (PS-) phospho-synuclein immunoreactive neurites (p = 0.46). [/fig]
[fig] Figure 3: Expression of TJs proteins in colonic biopsies from patients with Parkinson's disease (PD) and control subjects (CTRL). Biopsies lysates (20 μg of protein per sample) were subjected to immunoblot analysis using antibodies against occludin and ZO-1 (A). Beta-actin was used as a loading control. The optical densities of occludin (B) and ZO-1 (C) immunoreactive bands were measured, normalized to the optical densities of beta-actin immunoreactive bands in the same samples and expressed as percentages of controls. Data correspond to mean ± SEM of 11 samples for control subjects (CTRL) and 31 samples for Parkinson's disease (PD) patients. Patients versus control, *: p < 0.05. [/fig]
[fig] Figure 4: Localization of TJs proteins in the colonic mucosa of healthy controls (CTRL) and patients with Parkinson's disease (PD). Representative photomicrographs of the colonic mucosa labeled with antibodies against ZO-1 (A, B) and occludin (C, D) in the colonic mucosa of control and PD patients; scale bar: 100 μm. High-magnification image of each area marked by red square; scale bar: 10 μm. [/fig]
[table] Table 1: Main clinical characteristics of PD patients [/table]
[table] Table 2: Spearman's correlation with sulfonic acid (SA) and HRP permeability in PD patients [/table]
|
Dynamic Distribution of Linker Histone H1.5 in Cellular Differentiation
Linker histones are essential components of chromatin, but the distributions and functions of many during cellular differentiation are not well understood. Here, we show that H1.5 binds to genic and intergenic regions, forming blocks of enrichment, in differentiated human cells from all three embryonic germ layers but not in embryonic stem cells. In differentiated cells, H1.5, but not H1.3, binds preferentially to genes that encode membrane and membrane-related proteins. Strikingly, 37% of H1.5 target genes belong to gene family clusters, groups of homologous genes that are located in proximity to each other on chromosomes. H1.5 binding is associated with gene repression and is required for SIRT1 binding, H3K9me2 enrichment, and chromatin compaction. Depletion of H1.5 results in loss of SIRT1 and H3K9me2, increased chromatin accessibility, deregulation of gene expression, and decreased cell growth. Our data reveal for the first time a specific and novel function for linker histone subtype H1.5 in maintenance of condensed chromatin at defined gene families in differentiated human cells.
# Introduction
In humans, there are eleven subtypes of linker histones that stabilize higher order chromatin structure and are generally associated with repressed genes [bib_ref] MSX1 cooperates with histone H1b for inhibition of transcription and myogenesis, Lee [/bib_ref] [bib_ref] Histone H1 and its isoforms: contribution to chromatin structure and function, Happel [/bib_ref] [bib_ref] Action of micrococcal nuclease on chromatin and the location of histone H1, Noll [/bib_ref] [bib_ref] The structure of histone H1 and its location in chromatin, Allan [/bib_ref] [bib_ref] Position and orientation of the globular domain of linker histone H5 on..., Zhou [/bib_ref]. Depletion of mouse H1c, H1d and H1e leads to less compact packaging of chromatin, changes in core histone modifications, and reduced DNA methylation at certain loci [bib_ref] Histone H1 depletion in mammals alters global chromatin structure but causes specific..., Fan [/bib_ref]. Binding of H1 and poly (ADPribose) polymerase-1 at 758 RNA polymerase II (Pol II)transcribed promoters is mutually exclusive at actively transcribed genes [bib_ref] Reciprocal binding of PARP-1 and histone H1 at promoters specifies transcriptional outcomes, Krishnakumar [/bib_ref]. In human cancer, linker histones exhibit altered expression with at least one linker histone gene, namely H1.5, being mutated in colon cancer [bib_ref] The consensus coding sequences of human breast and colorectal cancers, Sjoblom [/bib_ref]. Linker histones are, therefore, important participants in normal biological as well as disease processes. However, while some functional differences have been reported for certain linker histones [bib_ref] Depletion of human histone H1 variants uncovers specific roles in gene expression..., Sancho [/bib_ref] , our knowledge of global distribution or function of each linker histone remains rudimentary.
Gene families are groups of homologous genes that are likely to have highly similar functions. While some gene family members are dispersed throughout the genome (e.g., solute carrier protein genes or SLCs), others are located in close physical proximity to each other, forming clusters of functionally related genes on human chromosomes. These gene family clusters include the olfactory receptor (OR), late cornified envelope (LCE), histone (HIST) and homeobox (HOX) genes. Current data indicate that different gene families have distinct chromatin features. For instance, the chromatin regions of OR and certain other gene family clusters lack histone modifications such as histone H3 lysine 4 methylation (H3K4me) and H3K27me that are found in the HOX clusters [bib_ref] Role of histone H3 lysine 27 methylation in Polycomb-group silencing, Cao [/bib_ref] [bib_ref] A chromatin landmark and transcription initiation at most promoters in human cells, Guenther [/bib_ref]. Considering the diversity of gene families in the human genome, it is not expected a priori that they would share similar chromatin characteristics or regulatory mechanisms.
Here we show for the first time that human linker histone H1.5 (HIST1H1B) binds to families of genes that are enriched for those encoding membrane or membrane-related proteins in terminally differentiated cell types representing all three embryonic germ layers. Little or no H1.5 enrichment was detected at the majority of the gene families in undifferentiated human embryonic stem cells (hESCs). H1.5 interacts with SIRT1 histone deacetylase which, along with H3K9me2, a repressive histone modification, were also enriched at H1.5 targets. Furthermore, H1.5 bound regions were mutually exclusive of DNase I sensitive regions. H1.5 depletion in fibroblasts resulted in disturbed SIRT1 and H3K9me2 distribution, and decreased chromatin compaction specifically at target genes. H1.5 knockdown cells showed extensive global deregulation of gene expression, with derepression of certain H1.5 target genes. Together, our findings reveal an unexpected but widespread function of histone H1.5 in chromatin compaction and gene expression in differentiated human cells.
# Results
## Histone h1.5 is differentially distributed in hescs and fibroblasts
To determine whether the genomic distribution of H1.5 is different in hESCs versus fibroblasts, we used chromatin immunoprecipitation combined with sequencing (ChIP-seq) of linker histone H1.5 in H1 hESCs and human lung IMR90 fibroblasts. We defined a significant peak as enrichment of ChIP over input DNA within a 100-bp window at a Poisson pvalue,0.001. We detected 8115 and 61349 significant peaks of H1.5 in H1 hESCs and IMR90, respectively, with only 171 peaks shared between the two cell lines [fig_ref] Figure 1: Histone H1 [/fig_ref]. In H1 hESCs, the peaks of H1.5 were distributed between genic regions (63 kb of genes) and distal intergenic regions which are at least 3 kb away from any gene. In IMR90 cells, the majority of H1.5 peaks were in distal intergenic regions with the rest within genic regions [fig_ref] Figure 1: Histone H1 [/fig_ref].
To understand the genome-wide distribution of H1.5, we calculated H1.5 peak density (number of peaks per kb of genomic DNA) in genic regions, which we defined as 23 kb from annotated transcription start sites (TSS) to +3 kb from annotated transcriptional stop sites (TTS), as well as intergenic regions. Regions with at least one significant peak (see Text S1) per kb were defined as H1.5 target loci. Genes that were bound by H1.5 in their genic regions or were located upstream or downstream of intergenic regions bound by H1.5 were defined as H1.5 target genes. In IMR90 fibroblasts, we detected 4535 H1.5 target genes with 1204 genes directly bound by H1.5, 2294 genes next to H1.5 target intergenic regions, and 1037 genes bound by H1.5 in both genic and neighbouring intergenic regions [fig_ref] Figure 1: Histone H1 [/fig_ref]. In hESCs, only 189 H1.5 target genes were detected [fig_ref] Figure 1: Histone H1 [/fig_ref]. Gene ontology analysis showed no significant enrichment for hESCs but enrichment of H1.5 target genes in membrane associated receptor and signalling genes in fibroblasts [fig_ref] Table 1: Gene ontology of H1 [/fig_ref]. Among the 4535 genes bound by H1.5 in IMR90, we noticed that some of them belonged to gene families, such as olfactory receptor family and other G-protein coupled receptors, and solute carrier family. This prompted us to systematically determine the enrichment of H1.5 target genes within the HGNC (HUGO Gene Nomenclature Committee) gene family database. Of the 4535 H1.5 target genes, 1659 (37%) genes were members of gene families, whereas all gene family members only accounted for 28% of total genes in human genome (binomial p-value = 4.53E-36), indicating a significant enrichment of gene families within H1.5 targets [fig_ref] Figure 1: Histone H1 [/fig_ref]. The genes from the same family can be either scattered throughout the genome or clustered in close physical proximity. When considering at least three genes from the same family that are located side by side as a 'cluster', ,23% of genes from HGNC gene families form clusters throughout the genome. Interestingly, the percentage of clustered genes increased to 37% within H1.5 target genes [fig_ref] Figure 1: Histone H1 [/fig_ref] ; binomial p-value = 2.29E-40). In addition, the level of H1.5 binding at clustered genes is significantly higher than that in non-clustered genes [fig_ref] Figure 1: Histone H1 [/fig_ref]. [fig_ref] Figure 1: Histone H1 [/fig_ref] shows an example of H1.5 binding to a block of ,1.5 Mbp on chromosome 1 covering the S100 calcium binding protein A (S100A), LCE, and small proline-rich (SPRR) gene family clusters in IMR90 cells, but not in H1 hESCs. This H1.5 binding block also includes fifteen genes that do not belong to a gene family but are located in between the three gene family clusters. Interestingly, five of these fifteen genes are involved in ''keratinocytes differentiation'' (binomial p-value = 2.2E-06) and two are peptidoglycan recognition proteins, suggesting functional relatedness to the nearby gene families.
Blocks of H1.5 enrichment were also found in intergenic regions in IMR90 cells but not in hESCs [fig_ref] Figure 1: Histone H1 [/fig_ref]. To determine if this block pattern is a genome-wide feature of H1.5 binding in fibroblasts, we calculated the average significance of H1.5 peaks within 5 kb windows and plotted the number of windows as a function of average enrichment p-value [fig_ref] Figure 1: Histone H1 [/fig_ref]. Compared to a random set of peaks (grey line) or the H1.5 peaks in hESCs (blue line), in IMR90 cells there were many windows with no significant enrichment of H1.5 but also many more windows with highly significant enrichment (red line). This distribution indicates that H1.5 forms blocks of enrichment over both genic and intergenic regions in IMR90 cells but not in hESCs. Remarkably, H1.5 genic blocks of enrichment occurs preferentially at a subset of gene families.
To confirm the H1.5 enrichment patterns in IMR90 fibroblasts, we used an Agilent promoter microarray containing probes for ,17,000 gene promoters, tiling an 8-kb region from 25.5 to +2.5 kb of the annotated transcriptional start sites (TSS) which we divided computationally into 16 fragments of 500 bp each [fig_ref] Figure 1: Histone H1 [/fig_ref]. We first verified the specificity of the H1.5 ChIP signal by ChIP-chip analysis of H1.5 after knockdown (KD) of H1.5 in IMR90 cells, which showed loss of H1.5 signal and no preferential enrichment compared to the control KD [fig_ref] Figure 1: Histone H1 [/fig_ref]. Consistent with ChIP-seq data, ChIP-chip of H1.5 showed significantly more enrichment in fibroblasts [fig_ref] Figure 2: H1 [/fig_ref] ; gene promoters are ranked from highest to lowest H1.5 enrichment in fibroblasts; all other ChIP-chip heat maps are in the same order). Gene ontology analysis of fibroblast data revealed that 59% of genes bound by H1.5 on the array were members of HGNC gene families (binomial p-value = 1.4E-08). When we sorted the genes based on their locations on each chromosome, we found that H1.5 was enriched in blocks of consecutive promoters [fig_ref] Figure 1: Histone H1 [/fig_ref]. For instance, on chromosome 1, the promoter blocks comprised families of highly homologous genes including LCE, SPRR, Fc receptor-like (FCRL) and the OR genes. Similarly, the promoter blocks on chromosome 11 corresponded to several OR gene clusters. This binding pattern was specific to H1.5 as H1.3 (HIST1H1D) did not show preferential binding to the gene families in IMR90 cells [fig_ref] Figure 1: Histone H1 [/fig_ref]. ChIP-quantitative PCR
## Author summary
In human cells, there are eleven subtypes of linker histones, five (H1.1-H1.5) of which are ubiquitously expressed in somatic cells. Somatic linker histones have been thought of as a group of similar proteins with redundant functions with few known differences among them. Our work uncovers for the first time a novel and unique role for the linker histone H1.5 (HIST1H1B). We found that H1.5, but not H1.3 (HIST1H1D), forms blocks of chromatin binding in genic and intergenic regions in differentiated human cells from all germ layers but not in embryonic stem cells. In genic regions, H1.5 binds to a large fraction of gene families that encode membrane associated proteins and are transcriptionally silent in a tissue-specific manner. H1.5 binding is associated with other repressive chromatin elements such as SIRT1 binding and H3K9me2 enrichment, and it negatively correlates with Pol II distribution. SIRT1 and H3K9me2 binding is dependent on H1.5, but not vice versa. H1.5 depletion in fibroblasts leads to increased chromatin accessibility at its target loci, altered cell cycle, and deregulation of gene expression. Our findings show that H1.5 has a dynamic distribution during human cell differentiation and is required for maintenance of proper gene expression in differentiated cells.
(qPCR) of selected genes confirmed the preferential binding of H1.5 to the gene families in IMR90 cells [fig_ref] Figure 2: H1 [/fig_ref]. Moreover, binding of H1.5 is not related to increased nucleosome density as histone H3 ChIP-chip did not show significant enrichment at H1.5 target genes in hESCs or fibroblasts [fig_ref] Figure 1: Histone H1 [/fig_ref].
## H1.5 binding is dependent on cellular differentiation state
Since hESCs and fibroblasts represent the extremes of cellular differentiation states, we sought to determine the differentiation stage at which the binding pattern of H1.5 is established. We examined H1.5 distribution in two cellular differentiation systems. First, we specified HSF1 hESCs to neural progenitor cells (NPCs) and then used standard growth factor withdrawal to drive differentiation towards neurons and astrocytes [bib_ref] Directed differentiation of human-induced pluripotent stem cells generates active motor neurons, Karumbayaram [/bib_ref] [fig_ref] Figure 2: H1 [/fig_ref] ; see Methods). Interestingly, H1.5 binding was established in terminally differentiated neural cells with the NPCs showing an intermediate pattern of H1.5 binding. Second, we obtained primary keratinocytes from skin biopsies and induced them to further differentiate in vitro using calcium (Ca 2+ ) which promotes primary keratinocytes to exit cell cycle and form stratified layers in culture [bib_ref] Calcium regulation of growth and differentiation of mouse epidermal cells in culture, Hennings [/bib_ref]. Like the neural differentiation, H1.5 binding was more significant in more differentiated, Ca 2+ -treated keratinocytes [fig_ref] Figure 2: H1 [/fig_ref]. Finally, we also examined H1.5 binding pattern in primary hepatocytes [fig_ref] Figure 3: H1 [/fig_ref] , which are derived from endoderm, and found similar H1.5 binding pattern [fig_ref] Figure 2: H1 [/fig_ref] as in fibroblasts and neural cells which are derived from mesoderm and ectoderm, respectively. Altogether, these data indicate that the H1.5 binding pattern is established progressively as cells acquire a more differentiated phenotype and occurs in fully differentiated cells derived from all three embryonic germ layers.
## H1.5 binding pattern is tissue specific
Despite the similarity of H1.5 binding patterns in different cell types, we noticed some degree of tissue specificity [fig_ref] Figure 3: H1 [/fig_ref]. For instance, LCE and SPRR genes form an ''epidermal gene cluster'', which together with the keratin gene cluster [fig_ref] Figure 2: H1 [/fig_ref] scheme), are highly expressed in all keratinocytes (with or without Ca 2+ treatment) compared to IMR90 fibroblasts [fig_ref] Figure 2: H1 [/fig_ref] bar charts). The enrichment of H1.5 specifically at the LCE, SPRR and KRT gene clusters in keratinocytes is significantly lower than that of IMR90 fibroblasts [fig_ref] Figure 2: H1 [/fig_ref] line chart) or other cell types [fig_ref] Figure 3: H1 [/fig_ref] , see bleow). These data suggest that the histone H1.5 binding pattern in differentiated cells is tissue specific, with H1.5 being depleted from gene families that are expressed appropriately in certain cell types.
## H1.5 binds to specific gene families
To systematically study H1.5 enrichment in gene families, we generated a matrix containing enrichment values of H1.5 (as well as RNA Pol II, SIRT1, and H3K9me2; see below) in 188 HGNC gene families across all cell types analysed in this study. As shown in [fig_ref] Figure 3: H1 [/fig_ref] , the enrichment values were clustered hierarchically across cell types/experiments (columns) and gene families (rows). Interestingly, the main branch point in the columns separated the differentiated cells with a dynamic pattern of H1.5 enrichment [fig_ref] Figure 3: H1 [/fig_ref] , Lanes 7-13) from those with little or no preferential enrichment of H1.5 [fig_ref] Figure 3: H1 [/fig_ref] , Lanes 1-6). The gene families (rows) were grouped into two main sub-clusters. Cluster 1 included gene families that were bound by H1.5 in at least two of the differentiated cell lines. Gene ontology analysis of genes in cluster 1 families revealed highly significant enrichment for membraneassociated proteins and sensory perception [fig_ref] Figure 3: H1 [/fig_ref]. In contrast, cluster 2 gene families that were depleted of H1.5 were significantly enriched for ribosome associated proteins and those involved in transcription regulation [fig_ref] Figure 3: H1 [/fig_ref]. By classifying the 2181 genes in cluster 1 gene families based on their molecular function or biological process [bib_ref] WebGestalt: an integrated system for exploring gene sets in various biological contexts, Zhang [/bib_ref] , we detected 729 genes (p = 1.17E-256) as signal transducers, 657 of which have receptor activity (p = 2.71E-264). 696 genes are involved in stimulus response (p = 6.81E-94) and 912 genes are involved in biological regulation (p = 1.68E-13), in which 666 genes play roles in cell surface receptor linked signal transduction (p = 2.96E-222) [fig_ref] Figure 3: H1 [/fig_ref]. These data indicate that H1.5 preferentially binds to a defined subset of membrane and membrane-associated gene families in differentiated cells.
## Binding of h1.5 is associated with repressed genes
To determine whether H1.5 is associated with transcriptional repression, we examined the global gene expression profile in IMR90 cells, hepatocytes, HK Ca+ and hESCs. The expression level of H1.5 target genes were significantly lower than that of a randomly-selected, similarly-sized group of genes in the three differentiated cell types but not in hESCs [fig_ref] Figure 4: H1 [/fig_ref] and [fig_ref] Figure 4: H1 [/fig_ref] -S4C). To further characterize the association between H1.5 binding and transcription, we sequenced messenger RNAs (mRNAs) from IMR90 fibroblasts (transfected with non-targeting siRNAs which will be later used as control for H1.5 knockdown cells), and compared the expression to H1.5 binding. As shown in [fig_ref] Figure 4: H1 [/fig_ref] , when we sorted all RefSeq genes by H1.5 enrichment and divided these genes into 11 groups (2,000 genes per group), the average gene expression level in each group decreased with increasing H1.5 binding (r = 20.39). Interestingly, H1.5 binding level in intergenic regions was also negatively correlated with the expression of neighbouring genes (r = 20.17 for 59 genes and r = 20.14 for 39 genes, [fig_ref] Figure 4: H1 [/fig_ref]. Genes that were bound by H1.5 in their genic and intergenic regions were more significantly repressed than those that were bound at either region [fig_ref] Figure 1: Histone H1 [/fig_ref]. Within the H1.5 target genes, both those belonging to families and non-families were equally repressed [fig_ref] Figure 4: H1 [/fig_ref]. In addition, we also examined Pol II binding which was negatively correlated with H1.5 (r = 20.19; [fig_ref] Figure 4: H1 [/fig_ref]. Pol II binding at the gene families also showed an opposite pattern to that of H1.5 [fig_ref] Figure 3: H1 [/fig_ref] , compare lanes 4 and 12). Taken together, we conclude that the binding of linker histone H1.5 is correlated with depletion of Pol II and repression of target genes in differentiated cells.
## Sirt1 and h3k9me2 bind to h1.5 target genes
Vaquero et al. reported previously that human SIRT1 interacts with linker histone subtype HIST1H1E [bib_ref] Human SirT1 interacts with histone H1 and promotes formation of facultative heterochromatin, Vaquero [/bib_ref]. Considering the high amino acid sequence conservation between HIST1H1E and H1.5, we asked whether H1.5 also interacts with SIRT1, and if so, whether the genomic distributions of SIRT1 and H1.5 overlap. Reciprocal co-immunoprecipitation experiments from IMR90 and hESC nuclear extracts revealed a direct or indirect H1.5-SIRT1 interaction in IMR90 cells, but not in hESCs [fig_ref] Figure 4: H1 [/fig_ref]. Consistently, the SIRT1 binding pattern at promoter regions in fibroblasts was highly similar to that of H1.5 [fig_ref] Figure 4: H1 [/fig_ref] , r = 0.58). Furthermore, SIRT1 deacetylates H3K9 which then can serve as a substrate for methylation. H3K9 methylation has been shown to be enriched at repressed regions [bib_ref] Large histone H3 lysine 9 dimethylated chromatin blocks distinguish differentiated from embryonic..., Wen [/bib_ref]. Thus, we examined H3K9me2 distribution at promoter regions in fibroblasts which was also very similar to H1.5 (r = 0.56) and SIRT1 (r = 0.67) binding [fig_ref] Figure 4: H1 [/fig_ref]. Like H1.5, H3K9me2 and SIRT1 were also enriched in gene families involved in sensory perception, and clustered together with H1.5 enrichment in differentiated cells [fig_ref] Figure 3: H1 [/fig_ref] , Lanes 9 and 10). These data suggest that H1.5, SIRT1 and H3K9me2 associate with defined gene sets that are normally repressed. Analyses of published data on distributions of other histone modifications including H3K4me1, H3K4me2, H3K4me3, H3K9me3 and H3K27me3 in IMR90 cells revealed little overlap with H1.5 binding at representative target gene cluster [fig_ref] Figure 4: H1 [/fig_ref] ; left panel) and intergenic regions [fig_ref] Figure 4: H1 [/fig_ref] ; right panel) or globally [fig_ref] Figure 5: H1 [/fig_ref].
## H1.5 is required for sirt1 and h3k9me2 enrichment
To determine whether SIRT1 and H1.5 regulate chromosomal distribution of each other, we transfected IMR90 cells with siRNAs to knockdown ( KD ) H1.5 or SIRT1 and mapped the binding of the other. Knockdown of H1.5 or SIRT1 did not affect the expression levels of other linker histone subtypes [fig_ref] Figure 6: H1 [/fig_ref]. In H1.5 KD IMR90 cells [fig_ref] Figure 5: H1 [/fig_ref] , lane 2), SIRT1 expression was down-regulated, and its distribution was globally disrupted [fig_ref] Figure 5: H1 [/fig_ref] , r = 0.012). In contrast, H1.5 expression was not changed significantly by SIRT1 knockdown [fig_ref] Figure 3: H1 [/fig_ref] , and its distribution was only partially affected [fig_ref] Figure 5: H1 [/fig_ref] , r = 0.47), indicating that H1.5 binding is less dependent on SIRT1. Knockdown of either protein resulted in lower H3K9me2 levels [fig_ref] Figure 5: H1 [/fig_ref] and loss of H3K9me2 enrichment [fig_ref] Figure 5: H1 [/fig_ref] , suggesting that both proteins are required for establishment of this repressive histone mark.
## H1.5 knockdown leads to global deregulation of gene expression
To determine if depletion of H1.5 affects gene expression, we sequenced mRNAs from H1.5 KD and control KD IMR90 fibroblasts. We detected 2367 genes with at least 1.5 fold change in expression in H1.5 KD versus control KD cells, 2022 (85%) of which were up-regulated in H1.5 KD cells. Among the genes with deregulated gene expression, 371 genes were H1.5 targets, and 345 (93%) of them were up-regulated. Interestingly, many more genes were up-regulated if they were bound by H1.5 in either genic or intergenic regions than in both [fig_ref] Figure 5: H1 [/fig_ref] , indicating that genes located in larger H1.5 binding blocks were less affected by H1.5 depletion (which could be due to incomplete KD of H1.5). Consistently, we found that over 70% of the up-regulated H1.5 target genes do not belong to gene families [fig_ref] Figure 5: H1 [/fig_ref] , which have higher levels of H1.5 binding [fig_ref] Figure 1: Histone H1 [/fig_ref]. Finally, over 90% of up-regulated genes in H1.5 targets are non-clustered, which is significantly higher than the percentage of non-clustered genes in all H1.5 targets [fig_ref] Figure 5: H1 [/fig_ref]. The expression level of non-clustered H1.5 target genes was significantly increased in H1.5 KD , while clustered genes were not affected [fig_ref] Figure 5: H1 [/fig_ref]. These data indicate that the singleton H1.5 target genes are more readily de-repressed in H1.5 KD cells. The lack of de-repression of H1.5-target clustered genes may be due to incomplete knockdown of H1.5, lack of appropriate transcriptional activators in fibroblasts or additional but unknown layers of gene regulation.
Overall, the up-regulated genes in H1.5 KD cells were enriched in cell death and apoptosis, whereas the down-regulated genes were enriched in DNA replication and cell cycle process [fig_ref] Figure 6: H1 [/fig_ref]. Consistent with these changes, we found that knockdown of H1.5 significantly decreased the growth of cells as well as the proportion of cells in S and G2/M phases of the cell cycle [fig_ref] Figure 6: H1 [/fig_ref] , suggesting that H1.5 is required for normal cell growth. Altogether, these data suggest that disruption of H1.5 affects the expression of .10% of all genes, contributing to altered cell cycle and growth of fibroblasts.
## H1.5 binding is required for chromatin compaction
The formation of H1.5 enrichment blocks in IMR90 cells prompted us to ask whether H1.5 functions to compact chromatin at its target regions. We performed micrococcal nuclease (MNase) assays in control KD , SIRT1 KD and H1.5 KD IMR90 fibroblasts and H1 hESCs. [fig_ref] Figure 7: H1 [/fig_ref] and S7B show the ethidium bromide staining of the MNase digested DNA from the indicated conditions. To determine the accessibility at regions targeted by H1.5, we performed Southern blotting using a fragment of one H1.5 target gene OR5AS1 [fig_ref] Figure 7: H1 [/fig_ref] on chromosome 11 as probe. The lanes corresponding to the highest concentrations of MNase were quantitated and visualized as line charts. As shown in [fig_ref] Figure 6: H1 [/fig_ref] , in H1.5 KD IMR90 cells nucleosomal DNA repeat length at the OR5AS1 gene locus appeared earlier with increased intensity than control cells, indicating greater accessibility to MNase. Interestingly, SIRT1 KD cells showed an intermediate level of accessibility with greater digestion than control but less than H1.5 KD cells. A similar result was also detected when using LCE4A gene, another H1.5 target gene [fig_ref] Figure 7: H1 [/fig_ref] , as a probe [fig_ref] Figure 6: H1 [/fig_ref]. In contrast, H1.5 KD or SIRT1 KD IMR90 cells did not show MNase accessibility differences at a histone gene (HIST2H2AA3) locus that is not bound by H1.5 [fig_ref] Figure 6: H1 [/fig_ref]. In H1.5 KD or SIRT1 KD H1 hESCs, we did not see significant differences in MNase digestion pattern of the OR5AS1 gene locus compared to control KD [fig_ref] Figure 6: H1 [/fig_ref]. Therefore, H1.5 contributes to compaction of chromatin at its target loci.
Finally, we determined the relationship between H1.5 enrichment and chromatin accessibility by comparing our H1.5 ChIPseq data with published DNase-seq data from IMR90 fibroblasts (GSM530665). Remarkably, H1.5 enriched regions were clearly excluded from DNase I sensitive regions with only 0.26% of H1.5 peaks overlapping with DNase I sensitive regions [fig_ref] Figure 7: H1 [/fig_ref] ; also see [fig_ref] Figure 4: H1 [/fig_ref] green lane). To determine if H1.5 knockdown increases DNase I sensitivity at target loci, we treated cell nuclei from control KD , SIRT1 KD , and H1.5 KD IMR90 cells with increasing amount of DNase I followed by quantitative amplification of two H1.5 target genes (OR5AS1 and LCE4A) and two non-target genes [fig_ref] Figure 1: Histone H1 [/fig_ref]. In H1.5 KD cells, more digestion was detected at OR5AS1 and LCE4A gene loci [fig_ref] Figure 7: H1 [/fig_ref] and 7C) compared to control KD and SIRT1 KD cells. At HIST2H2AA3, a potentially euchromatic locus, we did not observe significant differences in DNase I sensitivity between H1.5 KD and control KD cells [fig_ref] Figure 7: H1 [/fig_ref]. Importantly, HOXC11 which is not a target of H1.5 but enriched for H3K27me3 [fig_ref] Figure 7: H1 [/fig_ref] also did not exhibit sensitivity to DNase I digestion [fig_ref] Figure 7: H1 [/fig_ref]. These data indicate that H1.5 target regions are less accessible and knockdown of H1.5 primarily affects the chromatin compaction at its target regions.
# Discussion
Mammalian cells at different stages of differentiation are generally thought to have dissimilar chromatin structures [bib_ref] Global transcription in pluripotent embryonic stem cells, Efroni [/bib_ref] [bib_ref] Epigenetic modifications of stem cells: a paradigm for the control of cardiac..., Zhou [/bib_ref]. Our data indicate that the linker histone subtype H1.5 contributes to dynamic formation of compact blocks of chromatin in differentiated cells of all embryonic germ layers. These blocks were found in intergenic regions as well as at the transcriptionally inactive gene loci. The H1.5-bound intergenic regions did not overlap with the defined enhancer elements [bib_ref] Histone modifications at human enhancers reflect global cell-type-specific gene expression, Heintzman [/bib_ref] in IMR90 cells or with the CCCTC-binding factor (CTCF) binding sites (GSM935404) (data not shown). The H1.5-bound genes function mainly in cell-cell communication and/or response to the environment. Bulk of these genes is expressed in a tissue-specific manner and has evolved in multicellular organisms [bib_ref] Cell-cell communication in carcinogenesis, Trosko [/bib_ref] [bib_ref] The Rh protein family: gene evolution, membrane biology, and disease association, Huang [/bib_ref] [bib_ref] The ancestry and cumulative evolution of immune reactions, Dzik [/bib_ref] [bib_ref] The evolution, complex structures and function of septin proteins, Cao [/bib_ref]. Binding of H1.5 to these loci is established very late in the cellular differentiation process, suggesting that H1.5 may contribute to a terminally differentiated phenotype. H1.5 is an integral member of a larger chromatin regulation system that involves SIRT1 and H3K9me2. This system could establish a stable chromatin state resembling condensed heterochromatin in terminally differentiat-ed cells. H1.5 is, in fact, preferentially located in heterochromatic regions of the genome by immunofluorescence of whole nuclei and has a longer residence time in chromatin compared to other linker histone subtypes [bib_ref] H1 family histones in the nucleus. Control of binding and localization by..., Th'ng [/bib_ref]. Thus, H1.5 may contribute to the 'more closed' chromatin structure in differentiated cells compared to ESCs [bib_ref] Global transcription in pluripotent embryonic stem cells, Efroni [/bib_ref]. Consistent with this, depletion of H1.5 resulted in less chromatin compaction at a target gene family locus. In particular, H1.5 knockdown in fibroblasts resulted in decreased cell growth, which revealed that the function of H1.5 could not be replaced by other linker histone subtypes.
The mechanism by which H1.5 recognizes its target regions at a specific developmental stage is an important question that remains to be answered. While the H1.5 and SIRT1 binding patterns are drastically different in hESCs versus fibroblasts, the levels of both proteins are comparable in the two cell types , suggesting that mechanisms other than control of expression contribute to H1.5 and SIRT1 binding at gene family loci. These mechanisms may include developmentally-restricted interactions with other chromatin-binding proteins and/or post-translational modifications [bib_ref] H1 histone subtype constitution and phosphorylation state of the ageing cell system..., Happel [/bib_ref] [bib_ref] Characterization of phosphorylation sites on histone H1 isoforms by tandem mass spectrometry, Garcia [/bib_ref] [bib_ref] Multifunctionality of the linker histones: an emerging role for protein-protein interactions, Mcbryant [/bib_ref]. Certain DNA sequence elements and/or chromatin features of the gene family regions may also contribute to H1.5 binding. Considering the tissue specific binding of H1.5 to certain gene families, it is likely that more than one mechanism regulates its genomic distribution.
The binding of H1.5 to the OR genes constitutes a potential mechanism by which the expression of this important gene family cluster could possibly be regulated in olfactory neurons. In each olfactory neuron, only one of the several hundred human OR genes is expressed [bib_ref] Visualizing an olfactory sensory map, Mombaerts [/bib_ref] , which may not only involve deliberate activation of a single OR promoter but also active repression of all other OR genes. A recent study reported that OR genes are marked in a highly dynamic fashion by activating and repressive histone modifications in the mouse olfactory epithelium [bib_ref] An epigenetic signature for monoallelic olfactory receptor expression, Magklara [/bib_ref]. It will be interesting to study if and how H1.5 may also contribute to the regulation of OR gene expression in olfactory neurons.
Experimental evidence from cancer, ESCs, induced pluripotent cells (iPS) and virally-transformed cells suggest that chromatin states are dynamic and may be perturbed in disease conditions [bib_ref] Direct reprogramming of genetically unmodified fibroblasts into pluripotent stem cells, Meissner [/bib_ref] [bib_ref] Sequential expression of pluripotency markers during direct reprogramming of mouse somatic cells, Brambrink [/bib_ref] [bib_ref] Directly reprogrammed fibroblasts show global epigenetic remodeling and widespread tissue contribution, Maherali [/bib_ref]. Consistent with this, H1.5 expression is down-regulated during cellular transformation by a viral oncoprotein [bib_ref] Epigenetic reprogramming by adenovirus e1a, Ferrari [/bib_ref] and in many cancer cell types . In addition, one nonsense (K27*) and one point mutation (G86A) in H1.5 have been reported in colon cancer [bib_ref] The consensus coding sequences of human breast and colorectal cancers, Sjoblom [/bib_ref]. The nonsense mutation occurs early in the N-terminus of the protein, essentially eliminating the protein. The G86A mutation is located in the third helix of the conserved globular domain that participates in binding one side of the DNA approximately one helical turn away from the nucleosome dyad [bib_ref] Mapping the interaction surface of linker histone H1(0) with the nucleosome of..., Brown [/bib_ref]. The G86A mutation may thus change the local hydrophilicity and affect the interaction between histone H1.5 and DNA. These data suggest that H1.5 may be downregulated and/or redistributed during processes that reverse the terminally differentiated state.
# Materials and methods
Cell culturing, purification, and differentiation H1 hESCs were plated on Matrigel (BD Biosciences)-coated plates, and maintained in mTeSR (StemCell). Before purification, cells were trypsinized to single cells and TRA-1-60 expressing cells were isolated by using MACS cell separation columns (Miltenyi Biotec). Isolated cells were tested by flow cytometry, and samples with .99% purity were used. IMR90 human primary lung embryo fibroblasts (ATCC) were cultured in Dulbecco's modified Eagle's medium (DMEM) plus 10% FBS (Hyclone), 100 U/ml penicillin (Gibco), and 100 mg/ml streptomycin (Gibco) at 37uC in 5% CO 2 . Growing cells at 50,70% confluence were used for further analysis. Human primary hepatocytes (Zen-bio #HP-F) were grown in Hepatocyte Maintenance Medium (Zen-Bio #HM-2) and used at passage four. Human ESCs (HSF1) were differentiated to neural progenitor cells (NPCs) in DMEM:F12 (Gibco) plus B27 (Gibco), N2-supplement (Gibco), 20 ng/ml bFGF (R and D systems), 1 mM Retinoic Acid (Sigma), and 1 mM Smoothened Agonist (Calbiochem). NPCs were mechanically isolated from culture based on rosette morphology as described [bib_ref] Human ES cell-derived neural rosettes reveal a functionally distinct early neural stem..., Elkabetz [/bib_ref] and expanded in DMEM:F12 plus B27, N2supplement, 20 ng/ml bFGF, and 50 ng/ml EGF (Gibco). NPCs were further differentiated to neurons and glia by withdrawal of the maintenance factors (bFGF and EGF) for 10 days. Human keratinocytes were cultured per manufacturer's protocol in KSFM (Invitrogen). To induce differentiation, calcium chloride was added to 1.5 mM for 48 hours [bib_ref] Defining the impact of beta-catenin/Tcf transactivation on epithelial stem cells, Lowry [/bib_ref].
## Chip-chip assay with agilent promoter array and data analysis
ChIP was performed with ,50 million cells as described [bib_ref] Epigenetic reprogramming by adenovirus e1a, Ferrari [/bib_ref]. Agilent two-color microarray data processing is described in Text S1. H1.5 ChIP-chip in IMR90 cells, H1 hESCs, SIRT1 KD cells, neural progenitor cells, SIRT1 ChIP-chip in IMR90 cells, H1.5 KD cells, and H3K9me2 in IMR90 cells were performed twice. Antibodies against human histone H1.3 (ab24174), H1.5 (ab24175) and SIRT1 (ab32441) were purchased from Abcam and for H3K9me2 from Millipore (07-441).
## Chip-sequencing assay
,20 ng of ChIP and input DNA were end-repaired, added an 'A' base to the 39 end, and ligated to adaptors by using Illumina ChIP-seq DNA Sample Preparation Kit Box 1. DNA fragments (150-300 bp) were selected and purified by agarose gel extraction, and amplified by PCR using Phusion polymerase (Illumina ChIPseq DNA Sample Prep. Kit Box 1) according to manufacturer's instructions. Amplified DNA were purified by gel extraction and quantified by Qubit dsDNA BR assay (Invitrogen). DNA sequencing was performed by Illumina GA-IIx sequencer with read length of 76 bp as per manufacturer's protocol. Raw reads were generated by the software SCS2.6. Further information on data analysis is available via Text S1.
Whole-genome expression profiling (Affymetrix array) and data analysis ,200 ng of total RNA were extracted from ,50 million cells by using Trizol (Invitrogen) and purified by RNeasy Plus Mini Kit (QiaGen). Purified RNA was submitted to UCLA Clinical Microarray Core to perform gene expression profiling by using Affymetrix Human U133Plus2.0 Arrays. Gene expression profiling of IMR90 cells, H1 cells, control KD and H1.5 KD cells were performed twice. Probe intensities from different samples were normalized by MAS5.0 provided by Affymetrix. mRNA-seq assay 5 mg of total RNA from control KD and H1.5 KD IMR90 fibroblasts were used to prepare libraries for mRNA-seq by using mRNA-seq Sample Preparation Kit (Illumina). Sequencing was performed by Illumina HiSeq2000 sequencer with read length of 100 bp as per manufacturer's protocol. Raw reads were aligned to reference human genome (hg19) using TopHat, and expression levels of each gene were calculated by in house software. Detailed data processing information is described in Text S1.
## Chip-quantitative pcr
Real-time PCR was performed on ChIP and input DNA using SYBR Green Real-time PCR Master Mix (Roche). For each primer pair, an amplification standard curve was established by gradient amount of input DNA. Specific targets were amplified from 1/10 of ChIP DNA, and relevant template DNA amount was calculated by comparing the Ct values of ChIP and input samples to the standard curve.
Co-immunuprecipitation (co-IP) and Western assay ,100 million cells were harvested and co-IP assay was performed by using nuclear complex co-IP kit (Active Motif) according to manufacturer's instructions. The precipitates were separated in 4-20% gradient SDS-PAGE gel, and visualized by standard Western blotting assay.
RNAi assay siRNAs targeting H1.5 (MU-012049-00) or SIRT1 (MU-003540-01) were purchased from Dharmacon. 1.5 mg of siRNAs were transfected to 2 million cells by using Lipofectamine RNAiMAX (Invitrogen Cat # 13778075). Cells were collected 48 hours after transfection for ChIP-chip, expression microarray, and Western blotting assay.
Micrococcal nuclease assay 5 million IMR90 cells after control KD , H1.5 KD or SIRT1 KD were trypsinized, pelleted at 4uC for 10 minutes, and washed twice with DPBS. MNase digestion was performed as described [bib_ref] Sequence-specific positioning of nucleosomes over the steroid-inducible MMTV promoter, Richard-Foy [/bib_ref] [bib_ref] Simian virus 40-mediated cis induction of the Xenopus beta-globin DNase I hypersensitive..., Enver [/bib_ref]. Digested DNA was purified by QiaQuick PCR purification Kit (QiaGen) and quantified by Qubit dsDNA BR assay (Invitrogen). Same amount of DNA was loaded onto a 1.5% agarose gel and run at 50 V overnight at 4uC followed by Ethidium Bromide staining. The gel image was processed by MatLab Image Processing Toolbox.
## Southern blotting
OR5AS1 probe was prepared by amplifying OR5AS1 gene DNA with primers 59-ATGGCTTATGACCGCTATGC and 59-TTGACGATATTGGAGCCACA from IMR90 genomic DNA. Primers for LCE4A probe are 59-TGTCCCTCAAAGTGTG-CATC and 59-TTCGCCCACTAATTCCTTTG, and primers for HIST2H2AA3 probe are 59-ATTGCCTGGGGTAGT-GAGTG and 59-GCCTTCGTCTTTGAGACTGG. The expected PCR product was gel purified. Biotin labeling was performed by using BrightStar Psoralen-Biotin Nonisotopic Labeling Kit (Ambion AM1480) according to manufacturer's instructions. MNase digested genomic DNA were separated in 1.5% agarose gel, transferred to nylon membrane (Amersham), and cross-linked by UV light. Hybridization was performed by incubating the membrane with labeled probes in Express Hyb Hybridization Solution (Clontech #636831) overnight at 42uC, and signals were detected by using BrightStar BioDetect Kit (Ambion AM1930).
DNase I assay 5 million IMR90 cells after control KD , H1.5 KD or SIRT1 KD were trypsinized, pelleted at 4uC for 10 minutes, and washed twice with DPBS. DNase I (Roche 04716728001) digestion was performed as described. Digested DNA was purified by phenol/chloroform extraction followed by ethanol precipitation. DNA pellet was air dried and re-suspended in 16 TE buffer. Quantitative PCR solution (20 mL) was prepared by mixing 4 ng of DNA, 20 pmol and 10 mL of FastStart Universal SYBR Green Master (Roche 04913850001), and the reaction was performed by STRATAGENE Mx3000P Real-time PCR machine. Text S1 Supporting information including additional experimental procedures for data processing, list of gene families in cluster 1 shown in [fig_ref] Figure 3: H1 [/fig_ref] , and list of primers used in [fig_ref] Figure 2: H1 [/fig_ref].
## Supporting information
[fig] Figure 1: Histone H1.5 is differentially distributed in fibroblasts and embryonic stem cells. (A) Venn diagram of significant peaks of H1.5 binding in H1 hESCs and IMR90 fibroblasts by ChIP-seq. (B) Pie chart of distribution of H1.5 relative to gene structure in H1 hESCs and IMR90 fibroblasts. (C) Venn diagram of number of H1.5 target genes in H1 hESCs and IMR90 fibroblasts. (D) Enrichment of HGNC gene family members in H1.5 target genes. Height of bars represents percentage of gene family members in all RefSeq genes (left) or H1.5 target genes (right). (E) Enrichment of clustered gene family members in H1.5 target genes. Height of bars represents percentage of clustered gene family members in all RefSeq genes (left) or H1.5 target genes (right). (F) Box plot of H1.5 enrichment levels of non-clustered gene family members (left) and clustered gene family members (right). (G) H1.5 enrichment block at the LCE/SPRR/S100A gene clusters. Each dot represents 2Log10 of Poisson p-value of ChIPed DNA versus input DNA in a 100-bp window. Lines represent average values. LCE, SPRR and S100A genes are highlighted in orange, green and pink, respectively. (H) An H1.5 enrichment block in an intergenic region of chromosome 2. (I) Histogram of average significance of H1.5 enrichment (x-axis) in 5 kb windows versus the number of windows (y-axis) in H1 hESCs and IMR90 fibroblasts. doi:10.1371/journal.pgen.1002879.g001 [/fig]
[fig] Figure 2: H1.5 distribution is established during cellular differentiation. Heat maps show the genome wide promoter distribution of H1.5 in (A) H1 hESCs and IMR90 fibroblasts, (B) HSF1 hESCs, neural progenitor cells (NPC), neural cells (Neu), and (C) primary keratinocytes (HK Ca2), calciumtreated keratinocytes (HK Ca+), and (D) primary human hepatocytes (Hepa). Each row represents the promoter of a gene in 500-bp intervals from 25.5 to +2.5 kb of the transcription start sites (TSS) which is indicated by the arrows. The gene promoters for all heat maps are ordered based on the highest to lowest level of H1.5 enrichment in fibroblasts. (E) Average H1.5 enrichment and expression of the epidermal (LCE, SPRR) and KRT gene clusters in HK Ca+ and IMR90 fibroblasts are shown as line charts and bar graphs, respectively. The relative position of genes in each cluster is illustrated schematically. doi:10.1371/journal.pgen.1002879.g002 [/fig]
[fig] Figure 4: H1.5 is associated with repressed genes. (A) Boxplots of expression levels (Affymetrix array) of randomly selected genes (left) and H1.5 target genes (right) in IMR90 fibroblasts. (B-C) Line charts show gene expression levels (mRNA-seq) as a function of H1.5 binding at (B) genic regions or (C) intergenic regions in control knockdown IMR90 fibroblasts. All RefSeq genes were sorted by H1.5 binding levels; each data point represents the average expression value of 2000 genes. Genes located upstream or downstream of an intergenic region were denoted as 59 gene (red line) and 39 gene (blue line), respectively. (D) Average expression level of genes bound by H1.5 at either genic (left), intergenic (middle), or both (right) regions in IMR90 fibroblasts. Wilcoxon rank sum test p-values are indicated. (E) Average expression level of genes that belong ('family') or do not belong ('non-family') to HGNC gene families that were bound (purple bars) or not bound (yellow bars) by H1.5 in IMR90 fibroblasts. (F) Genomewide promoter binding of Pol II, H1.5, SIRT1 and H3K9me2 in IMR90 fibroblasts. (G) Distributions of H3K4me1, H3K4me2, H3K4me3, DNase I sensitive sites, H3K9me3, H3K27me3, and H1.5 peaks at LCE/SPRR/S100A gene cluster (left panel) and an intergenic region in chromosome 2 (right panel). The scale of DNase I hypersensitive sites represent z-score of counts in each 100-bp window. Scales of H1.5 and other histone modifications represent the Poisson p-values of enrichment at each 100-bp window. doi:10.1371/journal.pgen.1002879.g004 [/fig]
[fig] Figure 3: H1.5 enrichment in HGNC gene families. (A) Hierarchical clustering of H1.5 enrichment levels in HGNC gene families is shown as a heat map. Enrichment z-score of each gene family was calculated by averaging the intensities of probes within a gene family region corrected for number of probes. Each row represents a gene family, and each column represents an experiment. The main sub-clusters of gene families are highlighted on the left and the most enriched gene ontology terms for each cluster are shown on the right. (B) Ontology classification of genes in cluster 1 families. Number of genes in each category is indicated. doi:10.1371/journal.pgen.1002879.g003 [/fig]
[fig] Figure 5: H1.5 is required for enrichment of SIRT1 and H3K9me2 at H1.5 target loci and normal gene expression pattern. (A) Western blotting of H1.5, SIRT1, H3K9me2, and b-Actin in control KD , H1.5 KD , and SIRT1 KD IMR90 cells. (B-D) Genome wide promoter binding of the indicated factors and experimental conditions is shown as heat maps. The genes are ordered as in Figure 2A. (E) Percentage of up-regulated genes with the indicated H1.5 binding pattern. The binomial p-values are indicated. (F) Stacked bar chart of percentage of family (yellow) and non-family (purple) genes in all H1.5 targets (left bar) and up-regulated H1.5 targets (right bar). The binomial p-value is indicated. (G) Stacked bar chart of percentage of clustered (orange) and non-clustered (burlywood) in all H1.5 targets (left bar) and up-regulated H1.5 targets (right bar). The binomial p-value is indicated. (H) Expression levels (mRNA-seq) of clustered and non-clustered H1.5 target genes in control KD and H1.5 KD cells are shown as a bar chart. doi:10.1371/journal.pgen.1002879.g005 [/fig]
[fig] Figure 6: H1.5 knockdown increases micrococcal nuclease accessibility at target regions. (A-D) Southern blots of MNase digested genomic DNA form the indicated cells are shown. The probe used for each blot is also indicated. Quantitated data from lanes 5, 10, and 15 are shown as line chart. Y axis represents the pixel position in the images, and x axis shows the band intensity. doi:10.1371/journal.pgen.1002879.g006 [/fig]
[fig] Figure 7: H1.5 knockdown increases DNase I sensitivity at target regions. (A) Venn diagram of the overlap between significant H1.5 peaks and DNase I hypersensitive sites. The p-value for exclusivity of these two sets of peaks is indicated (Text S1). (B-E) Quantitative PCR of DNA fragments at indicated genes from genomic DNA treated with increasing amount of DNase I. Data points with t-test p-value,0.05 are labelled with *. doi:10.1371/journal.pgen.1002879.g007 [/fig]
[fig] Figure S1, Figure S6: Data representation and antibody specificity in ChIP. (A) Pie chart of H1.5 target genes classified by H1.5 enrichment pattern. (B) Design of scaling windows of each gene in ChIP-chip data analysis. Each row in the heat map represents the promoter of a gene in 500-bp intervals from 25.5 to +2.5 kb of the predicted transcriptional start site (TSS). The genes are sorted in descending order based on the average H1.5 promoter enrichment in IMR90 cells. All subsequent heat maps are in the same order. (C) Genomewide promoter distribution of H1.5 in control KD and H1.5 KD IMR90 fibroblasts. (D) Localization of H1.3 and H1.5 on gene promoters along chromosomes 1 and 11 in IMR90 fibroblasts. (E) Genomewide promoter distribution of histone H3 in H1 hESCs and IMR90 fibroblasts. (TIF) Figure S2 Validation of ChIP-chip data by quantitative ChIP-PCR. ChIP-qPCR of H1.5 and SIRT1 at the LCE4A, LCE1C, SPRR2A, OR5W2, OR5AS1, and HIST3H2A genes at promoter (PRO), transcription start site (TSS), and open reading frame (ORF) regions. Error bars represent the standard deviation of three independent ChIP-qPCR experiments. H1.5 and SIRT1 enrichment were higher in IMR90 (red lines) compared to H1 hESCs (blue lines) at its target genes (LCE4A, LCE1C, SPRR2A, OR5W2, OR5AS1) but not at HIST3H2A which is a gene family member that is not targeted by H1.5. (TIF) Figure S3 Expression of hepatocyte specific genes in primary hepatocytes. Relative expression of hepatocyte specific genes in human hepatocytes to IMR90 fibroblasts was calculated from Agilent expression array data. Bars represent the logarithm of the ratio expression in hepatocytes versus IMR90 fibroblasts. (TIF) Figure S4 H1.5 binding is associated with gene repression. (A-C) Boxplots of expression levels of randomly selected genes (left) and H1.5 target genes (right) in H1 hESCs, hepatocytes, and calcium induced keratinocytes (HK Ca+). (D) Reciprocal coimmunoprecipitation of SIRT1 and H1.5 from nuclear extracts in IMR90 fibroblasts but not from hESCs. (TIF) Figure S5 H1.5 binding is not associated with common histone modifications. (A) Venn diagram of overlaping peaks between H1.5 and indicated histone modifications. (B) Average binding profiles of indicated histone modifications across H1.H1.5 is required for normal cell growth. (A) Expression of linker histone subtype H1.1-H1.5 in knockdown cells by Western blotting. (B) Gene ontology of up-and downregulated genes in H1.5 knockdown cells. (C) Morphology (left panel), growth curve (line chart), and cell cycle distribution (stacked bar chart) of control KD , H1.5 KD , and SIRT1 KD IMR90 cells. (TIF) Figure S7 Micrococcal nuclease (MNase) digestion of chromatin. Ethidium bromide staining of MNase treated genomic DNA in control KD (Ctrl), SIRT1 KD and H1.5 KD IMR90 fibroblasts (A) or H1 hESCs (B). Quantitated data from lanes 5, 10 and 15 (highest MNase concentration) are shown as line chart. Y axis represents the pixel position in the images; x axis shows the band intensity. (C) Patterns of DNase I hypersensitive sites, H3K9me3, H3K27me3 and H1.5 enrichments at representative genes. The scale of DNase I hypersensitive sites represent z-score of counts in each 100-bp window. Scales of H3K9me3, H3K27me3 and H1.5 represent the Poisson p-values of enrichment at each 100-bp window. (TIF) Figure S8 H1.5 is generally down-regulated in cancer cells. (A) Expression levels of H1.5 and SIRT1 proteins are similar in hESC and IMR90 fibroblasts as determined by Western blotting. (B) mRNA levels of H1.5 in 173 normal cell types and 744 cancer cell lines from NextBio database [40] are represented as boxplots. (TIF) [/fig]
[table] Table 1: Gene ontology of H1.5 target genes in IMR90 fibroblasts. [/table]
|
Hallucinations and Other Psychotic Symptoms in Patients with Borderline Personality Disorder
Background: Psychotic symptoms in BPD are not uncommon, and they are diverse and phenomenologically similar to those in schizophrenia spectrum disorders. Despite their prevalence in BPD patients, knowledge about the characteristics and severity of hallucinations is limited, especially in modalities other than auditory. Aim: This review summarises the causes, phenomenology, severity, and treatment options of hallucinations and other psychotic symptoms in BPD. Methods: The PubMed database was used with the following key terms: "borderline personality disorder" and 'hallucinations' and "psychotic symptoms". Articles were selected between January 1990 and May 2021. The primary keyword search yielded a total of 545 papers, of which 102 articles met the inclusion criteria and were fully screened. Papers from the primary source reference lists were also screened, assessed for eligibility, and then added to the primary documents where appropriate (n = 143). After the relevance assessment, 102 papers were included in the review. We included adult and adolescent studies to gather more recent reviews on this topic. Results: Hallucinations are significantly prevalent in BPD, mainly auditory, similar to schizophrenia spectrum disorders. The relationship between hallucinations and depression, anxiety, suicidality, schizotypy, and loneliness in BPD has been discovered but requires more research. Studies for treatment options for hallucinations in BPD are lacking. Conclusion: Recognition of psychotic symptoms in patients with BPD as distinguished psychopathological phenomena instead of diminishing and overlooking them is essential in the clinical assessment and can be useful in predicting complications during treatment. More focused research in this area is needed.Prevalence of Psychotic Symptoms in BPDOver the past three decades, studies have shown that the prevalence of psychotic symptoms ranges from 26% to 54%.[13][14][15][16][17]24Auditory verbal hallucinations are the most common form of psychotic symptoms in patients with BPD.15,25,26For instance, https://doi.
# Introduction
Borderline personality disorder (BPD) is characterized by chronic instability of emotions and self-image leading to selfdestructive behaviour and relationship difficulties. [bib_ref] Suicidal behavior in borderline personality disorder: prevalence risk factors, prediction and prevention, Black [/bib_ref] One of BPD criteria is transient stress-related paranoid thoughts.In addition, the clinical picture includes cognitive-perceptual symptoms such as suspiciousness, ideations of reference, paranoid thoughts, delusions, derealization, depersonalization, and hallucination-like symptoms.Historically, the term "a borderline state" or "a borderline patient" described clinical features that were both psychotic and neurotic in nature. [bib_ref] Borderline states, Knight [/bib_ref] For a long time, the categorization of BPD patients was inconsistent -some considered "borderline" as either a form of schizophrenia (ie, latent, pseudoneurotic, or borderline schizophrenia) or as a personality disorder (ie, Kernberg's borderline personality organization). [bib_ref] Borderline personality organization, Kernberg [/bib_ref] [bib_ref] A recurrent question: what is borderline?, Zandersen [/bib_ref] With the introduction of the DSM-III,BPD was separated from a schizotypal personality disorder (which was then considered a psychotic disorder), and BPD criteria entered the psychiatric classification system. [bib_ref] A recurrent question: what is borderline?, Zandersen [/bib_ref] However, psychotic symptoms were, despite their presence, not included.It was not until the fourth revision of the DSM (DSM-IV) that psychotic symptoms like "severe dissociative symptoms" or "transient stressrelated paranoid ideations" were added as a criterion to the diagnosis of BPD.The DSM-5 BPD criteria remain unchanged in Section II.In Part III of the DSM-5, the state of "psychoticism" (ie, cognitive and perceptual dysregulation) is included as a possible, non-essential feature of BPD.As for the International Classification of Diseases and Related Health Problems (ICD), the BPD diagnosis did not enter until 1992 in the 10th edition as a borderline subtype of the emotionally unstable personality disorder.Psychotic symptoms were not present among the criteria. The impending 11th edition of the ICD will adopt a multi-dimensional assessment of borderline personality disorder, including "transient dissociative symptoms or psychotic signs in situations of high affective arousal".Psychotic symptoms in BPD have been usually described as short-lived, less severe, and qualitatively different from those in psychotic disorders such as schizophrenia, despite the historical inclusion of "borderline" in psychotic disorders. The terms "quasi-psychotic" (limited, brief, and non-bizarre), "transient", "atypical", "pseudo-", or even "factitious" have been often used. [bib_ref] Cognitive features of borderline personality disorder, Zanarini [/bib_ref] [bib_ref] The validity of DSM-III borderline personality disorder: a phenomenologic, family history, treatment..., Pope [/bib_ref] However, recent research shows that psychotic symptoms in people with BPD (auditory verbal hallucinations in particular) show more similarities with psychotic disorders than differences. Psychotic symptoms in BPD are also markers of more severe psychopathology and worse prognosis (specifically more frequent hospitalizations and suicidality). [bib_ref] Psychotic symptoms in borderline personality disorder, Chopra [/bib_ref] [bib_ref] Characteristics of psychosis in borderline personality disorder, Links [/bib_ref] [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] [bib_ref] Comparing the experience of voices in borderline personality disorder with the experience..., Merrett [/bib_ref] However, there is no common consensus on their aetiology, phenomenology, severity, or management. [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] [bib_ref] Psychotic symptoms in posttraumatic stress disorder, Lindley [/bib_ref] [bib_ref] The characteristic features of auditory verbal hallucinations in clinical and non-clinical groups:..., Larøi [/bib_ref] [bib_ref] Auditory verbal hallucinations in patients with borderline personality disorder are similar to..., Slotema [/bib_ref] This review aimed to summarise the causes, phenomenology, severity, and treatment of hallucinations and other psychotic symptoms in BPD. We also assessed evidence against the long-standing concept that psychotic signs in BPD are different and less severe than in psychotic disorders. We included adult and adolescent studies to gather more recent reviews on this topic. [bib_ref] Comparing the experience of voices in borderline personality disorder with the experience..., Merrett [/bib_ref] [bib_ref] Psychotic symptoms in patients with borderline personality disorder: prevalence and clinical management, Schroeder [/bib_ref] [bib_ref] Avoiding misdiagnosis when auditory verbal hallucinations are present in borderline personality disorder, Beatson [/bib_ref] [bib_ref] Psychotic symptoms in borderline personality disorder: an update, D'agostino [/bib_ref] [bib_ref] Comorbid diagnosis of psychotic disorders in borderline personality disorder: prevalence and influence..., Slotema [/bib_ref] Several research questions have been formulated:
(1) How common are hallucinations and other psychotic phenomena in BPD?
(2) What factors influence psychotic symptoms in BPD?
(3) Are psychotic symptoms in BPD linked with the severity of the disorder? (4) Are hallucinations and other psychotic symptoms in BDP different from other disorders? [bib_ref] A recurrent question: what is borderline?, Zandersen [/bib_ref] What are the treatment options for hallucinations in BPD?
# Method
A narrative review was conducted using the PubMed database to search for articles published between January 1990 and May 2021, using the following key terms: "borderline personality disorder) and (hallucinations or psychotic symptoms)" or hospitalizations or suicidality. Inclusion criteria were (1) published in a peer-reviewed journal: (2) studies in humans only; or (3) reviews on a related topic; (4) English language. The exclusion criteria were: (1) abstracts from conferences;
(2) thesis or dissertation papers; (3) textbooks; (4) popular articles; (5) editorial commentaries; (6) animal model studies. Additional texts were retrieved by searching the reference lists of the primary sources. Texts were collected and systematized according to relevance. [fig_ref] Figure 1: Diagram of the literature selection process [/fig_ref] describes the flow chart of the source literature selected for the review article. The primary keyword search yielded 545 articles, of which 102 papers met the inclusion criteria and were thoroughly reviewed. Secondary papers from the reference lists of the primary selected articles were examined, assessed for eligibility, and added to the list of primary papers (n = 143). A total of 102 papers were used for review [fig_ref] Figure 1: Diagram of the literature selection process [/fig_ref].
# Results
## Selection of studies
## (1) how common are hallucinations and other psychotic phenomena in bpd?
Yee et al found in their initial research that out of 171 patients with BPD, 50 of them reported hallucinations. Other psychotic phenomena can also be present. They selected ten patients for case reports who experienced auditory hallucinations. These patients reported a high occurrence of other psychotic phenomena too. Nearly all (90-100%) experienced a thought insertion and thought blocking. Although none of the patients thought that somebody else was causing this. The feeling of being under the external influence was reported in 7 out of 10 patients. [bib_ref] Persistent hallucinosis in borderline personality disorder, Yee [/bib_ref] Another study with 30 participants with BPD reported psychotic symptoms unrelated to affective disorder or drug abuse to be present in 60% of patients. Hallucinations were the most prevalent symptom, and other symptoms included ideas of reference of non-psychotic nature (27%) and delusions (20%). Half of the patients with delusions (10%) met the criteria for previously undiagnosed schizophrenia. [bib_ref] A study of psychotic symptoms in borderline personality disorder, Pearse [/bib_ref] In the following text, we separated information on hallucinations and other psychotic phenomena, due to the significant difference in evidence on these topics.
Prevalence and Quality of Hallucinations in BPD Auditory verbal hallucinations have been studied most frequently in patients with BPD. Studies in adults with BPD have shown that 29% -50% of patients report auditory verbal hallucinations. [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] [bib_ref] Comparing the experience of voices in borderline personality disorder with the experience..., Merrett [/bib_ref] [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] Kingdon et al compared auditory verbal hallucinations in patients with BPD to those with schizophrenia and those with comorbid BPD and schizophrenia. Auditory verbal hallucinations were found in 46% of patients with BPD, 66% of patients with schizophrenia, and 90% of patients with comorbid BPD and schizophrenia. [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] Slotema et al thoroughly described hallucinations in a sample of 60 BPD patients (37 in the hallucination group). Auditory, visual, olfactory, and tactile hallucinations were present at least once a month in 65%, 51%, 31%, and 47%, respectively. Eighty percent of respondents experienced hallucinations in more than one modality. The frequency of the
## Dovepress
## 789
## Dovepress
Belohradova Minarikova et al hallucinations ranged from once a week to once a day, and the average duration lasted one to several minutes. Fifty percent of the hallucinations had malicious content. Furthermore, 23-47% of the patients related their hallucinations to previously experienced traumatic events. For most patients perceived their auditory hallucinations as repetitive words or whole phrases. Visual hallucinations were complex images but without a specific pattern of appearance during the day. Patients´belief that the hallucinations were real ranged from probable to quite convincing. The participants sometimes interacted with their hallucinations and sometimes obeyed benign and malevolent commands, and visual commands were rarely present. Stress and discomfort caused by hallucinations ranged from mild to substantial. Interference with daily routine also varied from one specific activity to several activities. [bib_ref] Hallucinations in patients with borderline personality disorder: characteristics, severity, and relationship with..., Slotema [/bib_ref] Other hallucinatory modalities may also occur. For example, a recent study examining 324 patients with BPD found that the prevalence of different types of hallucinations was 27% for auditory hallucinations (including auditory-verbal and nonverbal), followed by 17% for olfactory, 15% for tactile, 11% for visual, and 8% for gustatory hallucinations. [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] From these studies, it can be concluded that psychotic symptoms in BPD influence a significant number of patients with BPD and clinically relevant treatment issues.
(2) What Factors Influence Psychotic Symptoms in BPD?
Despite the well-reported occurrence of psychotic symptoms and hallucinations in BPD, little is known about their aetiology and characteristics, especially in other forms than auditory hallucinations. [bib_ref] Psychotic symptoms in patients with borderline personality disorder: prevalence and clinical management, Schroeder [/bib_ref] [bib_ref] Comorbid diagnosis of psychotic disorders in borderline personality disorder: prevalence and influence..., Slotema [/bib_ref] Some studies focused solely on hallucinations; others included other psychotic symptoms.
## Hallucinations and stress
Stress is a reaction with many definitions but generally involves emotional or physical tension. [bib_ref] The Stress Reaction: a Historical Perspective, Rom [/bib_ref] Psychosocial stress is, among others, connected with dopamine dysregulation, which can lead to psychotic symptoms. [bib_ref] Schizophrenia: an integrated sociodevelopmental-cognitive model, Howes [/bib_ref] The exact pathophysiology of this process in BPD is unknown; however, some clinical data point to a possible connection.
Two studies examined the effect of stress on psychotic symptoms in BPD. Niemantsverdriet et al found a positive correlation between the severity of hallucinations and the number of current life stressors. [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] Glaser et al discovered that patients with BPD reported the strongest psychotic reactivity to stress in everyday life, compared to patients with psychotic disorder, other personality disorders and healthy controls. This included various psychotic symptoms (such as stress-related paranoid thoughts), including hallucinations. They hypothesised that this would be concerning mainly "greater" stressors such as interpersonal problems; however, even on "smaller" daily stressors, patients with BPD reported overall higher reactivity. [bib_ref] Psychotic reactivity in borderline personality disorder, Glaser [/bib_ref]
## Hallucinations, childhood trauma and dissociation
Hallucinations in BPD are also associated with posttraumatic stress disorder and a history of childhood trauma, especially emotional abuse, 17 but sexual and physical. [bib_ref] Childhood trauma and auditory verbal hallucinations, Daalman [/bib_ref] Patients with hallucinations experienced abuse more likely than patients without hallucinations. It is worth mentioning that a similar relationship was found in hallucinating patients with other disorders. [bib_ref] Childhood trauma and auditory verbal hallucinations, Daalman [/bib_ref] [bib_ref] Dissociation mediates the relationship between childhood trauma and hallucination-proneness, Varese [/bib_ref] [bib_ref] Adverse childhood experiences and hallucinations, Whitfield [/bib_ref] Beatson stated that auditory verbal hallucination is highly correlated with elevated levels of dissociation and a history of childhood trauma. [bib_ref] Avoiding misdiagnosis when auditory verbal hallucinations are present in borderline personality disorder, Beatson [/bib_ref] Tschoeke et al compared BPD and schizophrenia patients according to verbal hallucinations and dissociation. Dissociation was more frequent in BPD, but the hallucinations were deemed indistinguishable, according to diagnostic criteria, and more exploration of dissociative pathology is probably needed. [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref]
## Hallucinations and loneliness
Loneliness is a subjective experience of poor social relationships and can lead to various psychiatric disorders and somatic illnesses. 35, [bib_ref] Relationship between loneliness, psychiatric disorders and physical health? A review on the..., Mushtaq [/bib_ref] Hoffman described its influence on psychotic symptoms through "social deafferentation". This hypothesis states that social isolation of vulnerable individuals can cause the social "programs" in our brain to create false social connections in the form of complex, emotionally impactful hallucinations and delusions. [bib_ref] A social deafferentation hypothesis for induction of active schizophrenia, Hoffman [/bib_ref] [bib_ref] Auditory verbal hallucinations, speech perception neurocircuitry, and the social deafferentation hypothesis, Hoffman [/bib_ref] The social deafferentation hypothesis is similar to the theory involving "anthropomorphism", as primarily motivated to maintain social connections with other individuals in their environment. If they cannot do this, they may compensate for loneliness by perceiving human representation in non-human objects, increasing the likelihood of hearing voices. [bib_ref] Creating social connection through inferential reproduction: loneliness and perceived agency in gadgets,..., Epley [/bib_ref] These hypotheses partially come from the fact that schizophrenia is strongly associated with previous social isolation. [bib_ref] First-episode psychosis in the military: a comparative study of prodromal symptoms, Tan [/bib_ref] Forty percent of patients with schizophrenia described their voices as familiar or recognisable voices. [bib_ref] The auditory hallucination: a phenomenological survey, Nayani [/bib_ref] A positive correlation has been found between loneliness and psychotic disorders. [bib_ref] Loneliness in psychosis: a meta-analytic review, Da Rocha [/bib_ref] [bib_ref] Feelings of loneliness among adults with mental disorder, Meltzer [/bib_ref] In the Australian National Psychosis Survey (n = 1825), up to 80% of individuals with a psychotic disorder reported feeling lonely. [bib_ref] Understanding the social costs of psychosis: the experience of adults affected by..., Stain [/bib_ref] "Chronic feelings of emptiness" and "a pattern of unstable and intense interpersonal relationships" present two of the BPD criteria. Emptiness is closely linked to feelings of hopelessness, loneliness, and isolation. [bib_ref] What is emptiness? Clarifying the 7th criterion for borderline personality disorder, Klonsky [/bib_ref] Patients with BPD report more and stronger feelings of loneliness than healthy controls, [bib_ref] Loneliness, social networks, and social functioning in borderline personality disorder, Liebke [/bib_ref] and chronic loneliness is one of the most persistent aspects of the disorder. [bib_ref] The subsyndromal phenomenology of borderline personality disorder: a 10-year follow-up study, Zanarini [/bib_ref] Interestingly, the loneliness was met by 81% of BPD patients with hallucinations and 48% without hallucinations. Medians of social loneliness score and total loneliness score were higher in BPD patients with hallucinations than BPD patients. The severity of both schizotypy and loneliness was also associated with hallucinations. [bib_ref] Hallucinations in patients with borderline personality disorder: characteristics, severity, and relationship with..., Slotema [/bib_ref] Although patients with BPD and hallucinations possibly experience more painful feelings of loneliness than those without hallucinations, this relationship needs to be explored more deeply. No direct relationship between hallucinations and current relational problems has been described. Although it is worth mentioning that feelings of emptiness, loneliness and social isolation were found to reflect the level of psychiatric stress as well. [bib_ref] What is emptiness? Clarifying the 7th criterion for borderline personality disorder, Klonsky [/bib_ref] Psychotic Symptoms and Schizotypal Personality Historically, borderline personality disorder was associated with schizophrenia and schizotypal personality. [bib_ref] A recurrent question: what is borderline?, Zandersen [/bib_ref] Schizotypy is a theory of a particular personality organisation based on behavioural and neurobiological findings. It presumes that psychosis is not a qualitative state but rather lies at the end of a continuum of psychiatric symptoms. Certain personalities tend to react to psychological pressure by psychosis. [bib_ref] The concept of schizotypy -a computational anatomy perspective, Modenato [/bib_ref] A strong association between BPD and schizotypy has been found, which could also explain the high prevalence of hallucinations in BPD patients. [bib_ref] Schizotypal symptoms in patients with borderline personality disorders, George [/bib_ref] [bib_ref] Overlap between borderline and schizotypal personality disorders, Kavoussi [/bib_ref] [bib_ref] Borderline or schizotypal? Differential psychodynamic assessment in severe personality disorders, Van Riel [/bib_ref] Schizotypal Personality Questionnaire (SPQ) is one method that explores this vulnerability. The questionnaire measures nine dimensions of schizotypy, arranged into clusters -the cognitive-perceptual cluster ("ideas of reference", "social anxiety", "odd beliefs/magical thinking", "unusual perceptual experiences", and "suspiciousness/paranoid thoughts"), the interpersonal cluster ("no close friends," "constricted affect," and again "social anxiety" "suspiciousness/paranoid thoughts"), and the cluster of disorganization ("eccentric/odd behaviour and appearance", "odd speech"). [bib_ref] The SPQ: a Scale for the Assessment of Schizotypal Personality Based on..., Raine [/bib_ref] High SPQ scores have been previously associated with an increased vulnerability to psychosis. [bib_ref] Does the Schizotypal Personality Questionnaire reflect the biological-genetic vulnerability to schizophrenia?, Vollema [/bib_ref] Slotema et al linked the severity of schizotypal characteristics (using SPQ) with hallucinations in BPD. The three SPQ clusters were significantly higher in patients with hallucinations. For the schizotypal subgroups, all cognitiveperceptual and disorganization clusters scores were significantly higher in the group experiencing hallucinations. Only the suspiciousness/paranoid thoughts group significantly differed in patients with hallucinations when examining the interpersonal cluster scores. Schizotypy severity positively correlated with hallucination severity. The significant correlation between schizotypy and hallucination severity persisted even after omitting the "unusual perceptions" subgroup from the total schizotypy scores. [bib_ref] Hallucinations in patients with borderline personality disorder: characteristics, severity, and relationship with..., Slotema [/bib_ref] Another study also found higher scores for the cognitive-perceptual, interpersonal, and disorganization clusters in patients with BPD and hallucinations than those without them. [bib_ref] Schizotypal symptoms in patients with borderline personality disorders, George [/bib_ref] [bib_ref] Borderline or schizotypal? Differential psychodynamic assessment in severe personality disorders, Van Riel [/bib_ref] The link between schizotypy and hallucinations is consistent with the findings of Sommer et al, who investigated a population with auditory verbal hallucinations without psychiatric diagnoses. [bib_ref] Healthy individuals with auditory verbal hallucinations; who are they? Psychiatric assessments of..., Sommer [/bib_ref] Association with schizotypy has also been found in patients with other personality disorders, especially schizotypal personality disorder. [bib_ref] Schizotypal symptoms in patients with borderline personality disorders, George [/bib_ref] [bib_ref] Borderline or schizotypal? Differential psychodynamic assessment in severe personality disorders, Van Riel [/bib_ref] [bib_ref] Some suggestions for the DSM-5 schizotypal personality disorder construct, Hummelen [/bib_ref] In 23 patients with BPD (no difference between presence/absence of hallucinations) and 12 patients with schizotypal personality disorder, no differences were found between scores on the three clusters of schizotypy. [bib_ref] Borderline or schizotypal? Differential psychodynamic assessment in severe personality disorders, Van Riel [/bib_ref] A comorbid schizotypal personality disorder was be found in 7% of the BPD patients. [bib_ref] Axis II comorbidity of borderline personality disorder, Zanarini [/bib_ref] Psychotic Symptoms in Young Adults Adolescence and young adulthood is the period when usually start manifesting both BPD [bib_ref] Prevention and early intervention for borderline personality disorder: current status and recent..., Chanen [/bib_ref] It can be difficult to distinguish whether the psychotic symptoms stem from BPD or signify a vulnerability for developing a psychotic disorder. BPD symptoms in adolescence have been associated with psychotic symptoms and depressive and hypomanic symptoms. [bib_ref] Psychopathological outcomes of adolescent borderline personality disorder symptoms, Winsper [/bib_ref] While most studies have examined the co-occurrence of psychotic symptoms and BPD in adults, there is limited research examining associations of BPD and early stage of psychotic disorders in young adulthood. Several investigations have inspected whether the BPD diagnosis predicts a higher risk of transitioning from predisposition to psychotic disorder in vulnerable adults. These studies concluded that BPD does not seem to increase or decrease the transition risk to a psychotic disorder. For those who did develop a psychotic disorder, a BPD diagnosis or BPD symptoms were not related to any specific psychotic disorder diagnosis. [bib_ref] Borderline personality pathology in an at-risk mental state sample, Paust [/bib_ref] [bib_ref] Borderline personality pathology in young people at ultra-high risk of developing a..., Ryan [/bib_ref] [bib_ref] Personality disorders and accentuations in at-risk persons with and without conversion to..., Schultze-Lutter [/bib_ref] [bib_ref] UHR) population: a case control study: borderline personality and psychosis risk, Thompson [/bib_ref] Caspi et al, who studied young adults, proposed that one general underlying dimension, the "p" factor, summarizes an individual's vulnerability to developing any form of psychopathology, and those psychotic symptoms are at the top of the p factor. Any individual with an intense vulnerability to general psychopathology may experience psychotic symptoms, regardless of the current diagnosis. [bib_ref] The p factor: one general psychopathology factor in the structure of psychiatric..., Caspi [/bib_ref] Incorporating this model into personality pathology, Sharp & Wall and Sharp et al suggested that BPD symptoms in adolescents can be understood as a manifestation of internalizing and externalizing psychopathology severe enough to lead to subsequent mental disorders, even psychotic disorders. [bib_ref] Personality pathology grows up: adolescence as a sensitive period, Sharp [/bib_ref] [bib_ref] Adolescence as a sensitive period for the development of personality disorder, Sharp [/bib_ref] Based on this assumption, Calvetti et al suggested that BPD should be viewed as a severity factor, similar to psychosis. When they occur together, they act synergistically in determining prognosis.A recent study of 15 to 18-year-olds found a positive correlation between psychotic symptoms, as defined by the Youth Self Report (YSR), and BPD severity, defined as the number of criteria met in DSM-IV classification, after adjustment for other psychopathology and functional impairment. [bib_ref] Psychotic symptoms in adolescents with borderline personality disorder features, Thompson [/bib_ref] In another study of 15 to 25-year-olds with BPD, those who experienced auditory verbal hallucinations showed higher psychopathology (ie, self-harm, paranoid thoughts, dissociation, anxiety, and general distress) than those who did not hallucinate. [bib_ref] Exploratory comparison of auditory verbal hallucinations and other psychotic symptoms among youth..., Cavelti [/bib_ref] Zandersen & Parnas concluded that young people (aged 15-25 years) with BPD and psychotic symptoms risk many future mental health struggles, including non-psychotic and psychotic disorders. These outcomes add to the findings in adult patients and signify that patients with BPD and psychotic symptoms should be considered more at risk than patients without ones. Young people who have both BPD and psychotic symptoms tend to experience more severe psychopathology and show a higher risk of poorer prognosis, including the development of another severe mental disorder (without being specific to any partial disorder), as well as adverse outcomes beyond the binding concept of diagnosis (eg, suicidality, severe and enduring functional impairment). [bib_ref] Identity disturbance, feelings of emptiness, and the boundaries of the schizophrenia spectrum, Zandersen [/bib_ref]
## (3) are psychotic symptoms in bpd linked with the severity of the disorder?
Several studies suggested that psychotic symptoms in BPD indicate disorder severity and poor treatment outcomes. [bib_ref] Psychotic symptoms in borderline personality disorder, Chopra [/bib_ref] [bib_ref] Characteristics of psychosis in borderline personality disorder, Links [/bib_ref] [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] [bib_ref] Comparing the experience of voices in borderline personality disorder with the experience..., Merrett [/bib_ref] [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] Some studies explored hallucinations, in particular, others considered a wide variety of psychotic symptoms.
## Psychotic symptoms and suicidality
Suicidal ideations and behaviour are major topics in psychiatric disorders in general. Suicides and suicide attempts of individuals with mental illness are almost ten times more frequent than those without illness. Both BPD and psychotic disorders are linked with increased suicidality. Up to 9% of patients with BPD and 14% of patients with schizophrenia end their life by committing suicide. [bib_ref] Epidemiology of Suicide and the Psychiatric Perspective, Bachmann [/bib_ref] Some studies investigated the specific link between psychotic symptoms, suicidal plans, and suicidal attempts in patients with BPD.
Generally, auditory verbal hallucinations were linked with increased suicide plans and attempts, more frequent hospitalization [bib_ref] Comorbid diagnosis of psychotic disorders in borderline personality disorder: prevalence and influence..., Slotema [/bib_ref] and a higher level of suicidality. [bib_ref] Avoiding misdiagnosis when auditory verbal hallucinations are present in borderline personality disorder, Beatson [/bib_ref] Kelleher et al found a 2.23-fold increase in suicide attempts in patients with visual and auditory hallucinations. [bib_ref] Psychotic experiences and suicide attempt risk in common mental disorders and borderline..., Kelleher [/bib_ref] Psychotic symptoms in BPD were linked with a higher incidence of suicide plans and attempts during the month before the study. Also, all subscales of the psychotic symptoms questionnaire PSYRATS correlated positively with suicide plans (as well as emotional and phenomenological subscales). [bib_ref] Suicidality and hospitalisation in patients with borderline personality disorder who experience auditory..., Slotema [/bib_ref] Cavelti et al revealed in their review that co-occurrence of BPD and psychotic symptoms is a marker of severe pathologies, such as the increase in suicidality.symptoms and suicidality. The higher number of suicide attempts was associated with hallucinations and delusions and the severity of comorbid PTSD and depression. Interestingly, the age of initial hospitalization showed fewer and number of hospitalizations no associations at all with suicidal attempts. [bib_ref] Prevalence and phenomenology of psychotic-like symptoms in borderline personality disorders -associations with..., Schroeder [/bib_ref]
## Psychotic symptoms and psychiatric hospitalizations
The severity of an illness may also be assessed by the number of hospitalizations and readmissions to the hospital. Scarce information is available on this topic.
Psychotic symptoms in BPD predict more hospitalizations in a two-year follow-up and shorter intervals until another hospitalization. [bib_ref] Suicidality and hospitalisation in patients with borderline personality disorder who experience auditory..., Slotema [/bib_ref] Both hallucinations and delusions in BPD patients predict faster readmission to an acute psychiatric ward after discharge. [bib_ref] Personality disorder: what predicts acute psychiatric readmissions?, Furnes [/bib_ref] A co-occurring psychotic episode in patients with BPD is a significant predictor of referral to a specialized psychiatric unit for severe mental disorders after treatment in an outpatient facility. [bib_ref] Auditory verbal hallucinations in borderline personality disorder and the efficacy of antipsychotics:..., Slotema [/bib_ref] Cavelti et al also found an increase in hospitalizations for patients with BPD and hallucinations.Psychotic Symptoms and Other Comorbidities Adults with BPD and auditory verbal hallucinations meet more BPD criteria and have more comorbid mental disorders. [bib_ref] Auditory verbal hallucinations in patients with borderline personality disorder are similar to..., Slotema [/bib_ref] The previously mentioned study by Slotema et al compared BPD patients with and without hallucinations. Psychotic symptoms were linked with more severe depression, anxiety scores, schizotypy, and more pronounced feelings of loneliness. However, it is unknown whether BPD patients with more severe hallucinations experience more severe symptoms of depression and anxiety. Also, no significant difference was found between the two groups with respect to their global assessment of functioning. [bib_ref] Hallucinations in patients with borderline personality disorder: characteristics, severity, and relationship with..., Slotema [/bib_ref] In three other studies, hallucinations and other psychotic symptoms in BPD were associated with affective disorders. [bib_ref] Characteristics of psychosis in borderline personality disorder, Links [/bib_ref] [bib_ref] An empirical study of psychosis in borderline personality disorder, Pope [/bib_ref] [bib_ref] Psychotic symptoms in depression and borderline personality disorder, Nishizono-Maher [/bib_ref] [bib_ref] An empirical study of psychosis in borderline personality disorder, Pope [/bib_ref]. However, in two other studies, the presence of affective disorders failed to predict any subsequent psychotic symptoms for BPD patients. The difference between psychotic symptoms in BPD patients with and without mood disorder across the lifespan was not established. [bib_ref] Auditory verbal hallucinations in patients with borderline personality disorder are similar to..., Slotema [/bib_ref] [bib_ref] Psychotic symptoms in patients with borderline personality disorder and concurrent axis I..., Miller [/bib_ref] [bib_ref] Psychotic features in borderline patients: is there a connection to mood dysregulation?, Benvenuti [/bib_ref] Distinguishing between BPD and schizophrenia may also be complicated because these two conditions may co-occur. In the study of Kingdon et al of 111 patients, 19 of them suffered both from BPD and schizophrenia. [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] Slotema et al found in their sample a prevalence of comorbid non-specified psychotic disorder at 20%, namely schizophrenia in 2%. [bib_ref] Comorbid diagnosis of psychotic disorders in borderline personality disorder: prevalence and influence..., Slotema [/bib_ref] This knowledge may be important because there is also evidence that young people with both first-episode psychosis and co-occurring BPD have insufficient access to standard treatments, including the recommended prescription of antipsychotic medication. 81
## (4) are hallucinations and other psychotic symptoms in bpd different from other disorders?
Studies comparing psychotic symptoms in individuals with schizophrenia or BPD found more similarities than differences. [bib_ref] Psychotic symptoms in borderline personality disorder, Chopra [/bib_ref] [bib_ref] Characteristics of psychosis in borderline personality disorder, Links [/bib_ref] [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] [bib_ref] Comparing the experience of voices in borderline personality disorder with the experience..., Merrett [/bib_ref] [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] More distinct evidence has been found concerning hallucinations; other psychotic symptoms were less frequently mentioned.
## Hallucinations in bpd and schizophrenia
In studies comparing BPD and schizophrenia, verbal auditory hallucinations specifically bear no significant differences in frequency, duration, location (ie, inside or outside the head), loudness, distress, or persuasiveness. [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] [bib_ref] Auditory verbal hallucinations in patients with borderline personality disorder are similar to..., Slotema [/bib_ref] [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref] [bib_ref] Persistent hallucinosis in borderline personality disorder, Yee [/bib_ref] As for differences between patients with BPD or schizophrenia, hallucinations in BPD seems generally less disruptive in everyday life. [bib_ref] Auditory verbal hallucinations in patients with borderline personality disorder are similar to..., Slotema [/bib_ref] Tschoeke et al did not find any significant difference in the prevalence of commentary voices, whereas dialoguing voices were more common in patients with schizophrenia (71%) compared to patients with BPD (40%). [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref] Kingdon et al compared auditory verbal hallucinations in patients with BPD to those with schizophrenia or with comorbid BPD and schizophrenia. [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] Patients with BPD had higher scores in Psychotic Symptoms Rating Scales (PSYRATS) in the amount and degree of malicious content and distress from auditory verbal hallucinations, compared to patients with schizophrenia. BPD patients also felt more controlled by their voices. [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref] However, other studies suggested that BPD patients had more emotional resistance to distress from hallucinations. [bib_ref] Auditory hallucinations: a comparison of beliefs about voices in individuals with schizophrenia..., Hepworth [/bib_ref] Neuropsychiatric Disease and Treatment 2022: In a study by Slotema et al, auditory verbal hallucinations (also assessed by PSYRATS) were compared in 3 groups: patients with BPD, patients with schizophrenia/schizoaffective disorder, and individuals without a psychiatric diagnosis. In BPD patients, the average frequency of auditory verbal hallucinations was at least once a day for several minutes or longer, and in most patients, auditory verbal hallucinations were noted in the head region. Most patients believed that their voices were from an internal cause and thought they had no control over their voices most of the time. Further questions on the internal cause responses revealed that 29% stated that the voices were like their own thoughts, and 33% thought they were experiencing hallucinations due to their mental disorder. The voices were female in 65% of patients and male in 76%. The voice owner was known to the patient in 30% of patients. The most prominent voice was speaking several phrases in most patients. Fear caused by auditory verbal hallucinations was higher in the BPD group. All items related to the characteristics of auditory verbal hallucinations and subsequent distress did not differ between BPD and schizophrenia/schizoaffective disorder patients, except for the item "life disruption," which was higher in the schizophrenia/schizoaffective disorder group. Compared to the controls, patients with BPD had much higher in almost all items, except for "location" and "loudness". These studies suggest that auditory verbal hallucinations in BPD are frequent, severe, and phenomenologically similar to those in schizophrenia. [bib_ref] Auditory verbal hallucinations in patients with borderline personality disorder are similar to..., Slotema [/bib_ref] Also, auditory verbal hallucinations in BPD patients have been found to occur at a younger age (adolescence) than in schizophrenia patients (early adulthood). [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref] Other Psychotic Symptoms in BPD and Schizophrenia Compared to patients with schizophrenia, patients with BPD report fewer delusions, conceptual disorganization, and negative symptoms (such as blunted affect or social withdrawal). [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref] In another study by Niemantsverdriet et al, hallucinations in patients with BPD were associated with delusions, but not negative symptoms or disorganization. [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] Although this could be explained if schizotypal features, especially cognitive-perceptual symptoms, were predominant in this subgroup -this has not been investigated.
## (5) what are the treatment options for hallucinations in bpd?
Since information on the treatment of hallucinations in BDP is scarce, studies evaluating the efficacy of antipsychotics, cognitive-behavioural therapy, or non-invasive brain stimulation methods are needed.
Antipsychotic medications have been studied in patients with BPD for cognitive-perceptual symptoms, ie, suspiciousness, ideas of reference, paranoid thoughts, delusions, derealization, depersonalization, and hallucination-like symptoms [bib_ref] Effectiveness of pharmacotherapy for severe personality disorders: meta-analyses of randomized controlled trials, Ingenhoven [/bib_ref] or "psychotic symptoms." Three meta-analyses found small to medium effect sizes. [bib_ref] Effectiveness of pharmacotherapy for severe personality disorders: meta-analyses of randomized controlled trials, Ingenhoven [/bib_ref] [bib_ref] Differential effectiveness of antipsychotics in borderline personality disorder: meta-analyses of placebo-controlled, randomized..., Ingenhoven [/bib_ref] [bib_ref] Pharmacotherapy for borderline personality disorder: Cochrane systematic review of randomised trials, Lieb [/bib_ref] In a systematic review of studies reporting treatments for psychotic symptoms, both typical and atypical antipsychotics tended to have positive effects on psychotic features in the context of BPD. [bib_ref] Suicidality and hospitalisation in patients with borderline personality disorder who experience auditory..., Slotema [/bib_ref] Since loneliness is a contributing factor in hallucinations, according to Lim et al, factors that may positively influence it are good social support, good quality of life, less stigmatization and discrimination, and good self-esteem. [bib_ref] Loneliness in psychosis: a systematic review, Lim [/bib_ref] However, research on BPD patients in this area is lacking.
As for the timing of therapeutic interventions, Cavelti et al suggest that the period of adolescence and young adulthood, when BPD and psychotic features usually first appear, represents a critical window of opportunity for early treatment intervention to prevent the progress of severe mental disorders in the future. 67
## Discussion and future directions
This review aimed to summarise the causes, phenomenology, severity, and treatment of hallucinations and other psychotic symptoms in BPD. We also assessed evidence against the long-standing concept that psychotic signs in BPD are different and less severe than in psychotic disorders.
Firstly, it was established that a significant number of BPD patients experiences various kinds of hallucinationsauditory, visual hallucinations, even olfactory and tactile. Other psychotic symptoms (such as ideas of reference, delusions, thought insertion) were also reported. The inclusion of psychotic symptoms in BPD classifications seems therefore highly relevant. In many cases, they are pervasive and intensify under pressure, the same way that positive symptoms in psychotic disorders increase. Thus, the term "hallucination-like experiences" as defined by the DSM (2013) could be replaced simply by "hallucinations". 2 Labels such as "pseudo-hallucinations" or "quasi-psychotic" may add to the stigma already experienced by BPD patients and should be avoided. Qualitative analysis of auditory hallucinations could also help understand the nature of these symptoms and might help differentiate the groups of patients with specific treatment needs. This has already been attempted in the study of Yee et al, but whether this differentiation correlates with different clinical approaches and treatment strategies is yet to be seen. [bib_ref] Persistent hallucinosis in borderline personality disorder, Yee [/bib_ref] There is much to be explored other in psychotic symptoms connected to BPD. Given their expected interrelatedness, the role of childhood adversity and dissociation in the development of psychotic symptoms in BPD needs to be clarified. While there is robust evidence of an indirect relationship between childhood adversity and delusions and hallucinations in psychotic disorder, [bib_ref] The relationship between dissociation and symptoms of psychosis: a meta-analysis, Longden [/bib_ref] [bib_ref] Childhood trauma and psychosis, Stanton [/bib_ref] findings in BPD discovered co-occurrence of higher dissociation and hallucination; however, the exact relationship is still unclear even less so when other psychotic symptoms are concerned. [bib_ref] Schizophrenia and borderline personality disorder: similarities and differences in the experience of..., Kingdon [/bib_ref] [bib_ref] Hallucinations in borderline personality disorder: prevalence, characteristics and associations with comorbid symptoms..., Niemantsverdriet [/bib_ref] [bib_ref] Psychotic symptoms in borderline personality disorder: an update, D'agostino [/bib_ref] [bib_ref] Comorbid diagnosis of psychotic disorders in borderline personality disorder: prevalence and influence..., Slotema [/bib_ref] [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref] [bib_ref] Psychotic symptoms in adolescents with borderline personality disorder features, Thompson [/bib_ref] As for contributing factors to hallucinations in BPD, loneliness was associated with hallucinations in many psychiatric disorders, with an exceptionally high prevalence in hallucinating patients with BPD. Therefore, therapy aimed to reduce loneliness (such as family therapy) should also positively impact hallucinations. However, not enough studies have been conducted in this area. Even though the social deafferentation study was described in schizophrenia, another study suggests that the mechanisms may be similar in other disorders. [bib_ref] Hallucinations, loneliness, and social isolation in Alzheimer's disease, El Haj [/bib_ref] Considering the schizotypal concept, SPQ scores for the cognitive-perceptual, interpersonal, and disorganization clusters were higher in patients with BPD and hallucinations, and results for the cognitive-perceptual subgroups were particularly prominent. BPD and schizotypal personality disorder may be more similar than previously thought. Thus, for this group of patients, treatment of schizotypy tendency should begin with exploring interventions related to psychotic disorders, such as antipsychotic medications, cognitive behavioral therapy, and non-invasive brain stimulation. [bib_ref] The treatment of hallucinations in schizophrenia spectrum disorders, Sommer [/bib_ref] Still, no amiable results are currently available for this research area.
It is also apparent that hallucination in BPD patients has a significant negative impact on patients´well-being. They were also associated with higher depression and anxiety scores and more severe psychopathology. Evidence suggests that psychotic symptoms could be considered a severity indicator of BPD. However, since a significant number of patients experiences this and yet, the clinical picture of BPD varies, more precise criteria on psychotic symptoms as an indicator should go hand-in-hand with their recognition. For this, however, the evidence is not robust enough. When planning treatment, clinicians should consider that individuals with BPD and psychotic symptoms are at greater risk for adverse outcomes, such as more frequent hospitalizations, self-harm, or suicide attempts.
BPD patients can experience hallucinations or even delusions similar to schizophrenia. Patients with BPD had higher scores in Psychotic Symptoms Rating Scales (PSYRATS) in the amount and degree of malicious content and distress from auditory verbal hallucinations, compared to patients with schizophrenia. [bib_ref] Similarities and differences in borderline personality disorder and schizophrenia with voice hearing, Tschoeke [/bib_ref] Whether this happens due to more emotional or childhood trauma, or simply because many schizophrenia patients are already treated by antipsychotic medication in antipsychotic doses, that is difficult to determine.
The amount of evidence comparing psychotic symptoms in BPD and schizophrenia is still lacking in some areas. Most findings describe auditory verbal hallucination as the most common psychotic symptom, both in BPD and schizophrenia. [bib_ref] Auditory verbal hallucinations in patients with borderline personality disorder are similar to..., Slotema [/bib_ref] Complex visual-tactile and olfactory hallucinations were also experienced in patients with BPD, [bib_ref] Hallucinations in patients with borderline personality disorder: characteristics, severity, and relationship with..., Slotema [/bib_ref] but their comparison to psychosis is unclear. These hallucinations are less commonly experienced by psychotic patients as well (for example, the lifetime prevalence of hallucination was 64-80% auditory, 23-31% visual, 9-19% tactile, and 6-10% olfactory. Past month prevalence was 23-27% auditory, 5-8% visual, 4-7% tactile, and 2% olfactory according to a study by McCarthy et al) [bib_ref] Occurrence and co-occurrence of hallucinations by modality in schizophrenia-spectrum disorders, Mccarthy-Jones [/bib_ref] and can be present in other diseases -neurological, ophthalmological, etc. [bib_ref] Hallucinations: a Systematic Review of Points of Similarity and Difference Across Diagnostic..., Waters [/bib_ref] Considering comorbidities (esp. dissociative states, epilepsy, substance abuse etc.) might also help in identifying physiology of hallucinations in general. However, this does not oppose the fact that hallucinations in BPD and psychotic disorders share more similarities than differences.
Since psychotic symptoms also occur in BPD and its onset may happen at the same age as the onset of schizophrenia, 57-59 careful diagnosis should be conducted to avoid serious mistakes. However, it may be a false dichotomy to assume that these patients have either BPD or psychosis, and they may have both. As with depression and bipolar disorder, psychotic symptoms may indicate a more severe disorder, especially if negative symptoms and disorganized thoughts are present (less common in patients with BPD). Diminishing psychotic symptoms in patients with BPD may also lead to their stigmatization and inaccessibility of proper care. Evidence that young people with both first-
## Dovepress
## 795
## Dovepress
Belohradova Minarikova et al episode psychosis and co-occurring BPD have insufficient access to standard treatments, including the recommended prescription of antipsychotic medication, has already been presented. [bib_ref] Does co-occurring borderline personality disorder influences acute-phase treatment for first-episode psychosis?, Francey [/bib_ref] Few studies examined the neural correlations of psychotic symptoms in BPD. [bib_ref] Sensory processing deficiencies in patients with borderline personality disorder who experience auditory..., Niemantsverdriet [/bib_ref] [bib_ref] A systematic review of structural MRI investigations within borderline personality disorder: identification..., Davies [/bib_ref] To this date, no randomized controlled trial has examined the efficacy of antipsychotic medication or psychological interventions for auditory verbal hallucinations or any other psychotic symptom in BPD. [bib_ref] Auditory verbal hallucinations in borderline personality disorder and the efficacy of antipsychotics:..., Slotema [/bib_ref] The first randomized controlled trial on aripiprazole in young people with BPD and auditory verbal hallucinations is underway. [bib_ref] Aripiprazole compared with placebo for auditory verbal hallucinations in youth with borderline..., Chanen [/bib_ref] Since negative beliefs about hallucinated voices [bib_ref] Negative beliefs about voices in patients with borderline personality disorder are associated..., Slotema [/bib_ref] [bib_ref] Preliminary evidence for the cognitive model of auditory verbal hallucinations in youth..., Cavelti [/bib_ref] and self-image 98 contribute to emotional distress in BPD, hallucinations themselves could become targets for psychological interventions such as cognitive behavioural therapy or narrative-based self-experience interventions (Thomas et al 2014, [fig_ref] Figure 1: Diagram of the literature selection process [/fig_ref]. [bib_ref] Psychological therapies for auditory hallucinations (Voices): current status and key directions for..., Thomas [/bib_ref] [bib_ref] Thomas N: bringing the "self" into focus: conceptualising the role of selfexperience..., Fielding-Smith [/bib_ref] Future research should not be concerned with such spurious categorical distinctions. Instead, studies should recognize both the dimensional and dynamic nature of psychopathology and the evolving phenotypes throughout the transition from childhood to adulthood by adopting a Clinical staging approach. Such an approach needs to include measures of personality pathology to address the etiological factors and treatment options for psychotic symptoms in BPD. [bib_ref] Integrating early intervention for borderline personality disorder and mood disorders, Chanen [/bib_ref] [bib_ref] Managing borderline personality disorder from a life course perspective: clinical staging and..., Hutsebaut [/bib_ref] Furthermore, more focused comparative studies concerning specific phenomena on both BPD and schizophrenia could help better understand both conditions.
# Conclusions
Psychotic symptoms in BPD are common and phenomenologically similar to those experienced by schizophrenia spectrum disorders. Despite their prevalence in BPD patients, knowledge of the characteristics and severity of hallucinations is limited, especially in modalities other than auditory. The relationship between hallucinations and depression, anxiety, suicidality, schizotypy, and loneliness has been discovered but requires more exploration. Recognition of hallucinations and psychotic symptoms in general in patients with BPD instead of diminishing and overlooking them is essential in the clinical assessment and can be useful in predicting complications during treatment.
[fig] Figure 1: Diagram of the literature selection process. Neuropsychiatric Disease and Treatment 2022:18 https://doi.org/10.2147/NDT.S360013 [/fig]
|
The Efficacy and Safety of Ginkgo Terpene Lactone Preparations in the Treatment of Ischemic Stroke: A Systematic Review and Meta-Analysis of Randomized Clinical Trials
Background: This meta-analysis aimed to assess the efficacy and safety of ginkgo terpene lactone preparations including ginkgo diterpene lactone meglumine injection, ginkgolide injection, and ginkgolide B injection for ischemic stroke (IS).Methods:We searched the randomized controlled trials (RCTs) with publication date earlier than 31 August 2021 in PubMed, China National Knowledge Infrastructure (CNKI), Chinese Science and Technology Journals Database (VIP), Chinese Biomedical Literature Database (CBM), Wanfang Database, Embase, and the Cochrane Library. RevMan 5.3 software was applied to analyze the data and generate the forest plot and funnel plot. Meanwhile, publication bias was also assessed by Egger's test with STATA 12 software.Results: A total of 28 RCTs were eligible for inclusion. Among them, 23 RCTs were used to evaluate the efficacy of ginkgo terpene lactone preparations as the main treatment intervention for IS. To be specific, ginkgo diterpene lactone meglumine injection was superior to control drug in improving clinical efficacy [RR = 1.18, 95% CI (1.12, 1.24), Z = 6.36, p < 0.001] and neurological function [MD = −1.42, 95% CI (−1.91, −0.93), Z = 5.66, p < 0.001]. However, the effectiveness of the ginkgolide B injection group was equivalent to that of the control group. Additionally, ginkgolide injection achieved better clinical efficacy [RR = 1.10, 95% CI (1.02, 1.18), Z = 2.36, p = 0.02], but the changes of neurological function deficit was not obviously different between two groups [MD = −0.43, 95% CI (−4.32, 3.46), Z = 0.22, p = 0.83]. Furthermore, meta-analysis of five trials on ginkgo diterpene lactone meglumine injection combined with recombinant tissue plasminogen activator (rt-PA) thrombolytic therapy for acute IS showed that combination therapy was better in improving clinical efficacy [OR = 1.91, 95% CI (1.13, 3.22), Z = 2.41, p = 0.02] and neurological function [MD = −3.31, 95% Cl (−3.64,−2.98), Z = 19.63, p < 0.001]. Importantly, no serious adverse drug reactions/adverse drug events (ADRs/ADEs) were reported.Conclusion: Ginkgo terpene lactone preparations have good therapeutic effects on patients with IS. For acute IS, ginkgo diterpene lactone meglumine injection can be used as a complementary therapy to improve the clinical efficacy of rt-PA.
# Introduction
Ischemic stroke (IS) is the most common type of cerebrovascular disease, which refers to ischemic necrosis and softening of brain tissue due to brain blood circulation impairment, ischemia, and hypoxia. The clinical symptoms present disturbance of consciousness, hemiplegia, aphasia, dysphagia, and blindness . Currently, the main approaches to treat IS include platelet aggregation therapy, thrombolytic therapy, nutritive therapy, and reducing intracranial pressure, blood pressure, and blood sugar and lipids [bib_ref] Current Advances in Ischemic Stroke Research and Therapies, Barthels [/bib_ref]. However, the efficacy and prognosis were not satisfied. Thus, exploring new treatments to improve poststroke recovery is urgently needed.
Traditional Chinese medicine therapy is widely used in treatment for multiple diseases including stroke [bib_ref] Application of Chinese Medicine in Acute and Critical Medical Conditions, Luo [/bib_ref]. Ginkgo terpene lactones are a class of compounds extracted from Ginkgo biloba leaf, mainly including ginkgolides and bilobalide. Ginkgolides belong to diterpene lactones, which mainly include ginkgolides A, B, C, K, J, L, M, N, P, and Q [bib_ref] Biosynthesis Pathways of Ginkgolides, Zeng [/bib_ref]. Ginkgolides A, B, C, and K, the main active compounds, are also platelet-activating factor (PAF) receptor antagonists [bib_ref] Inhibition of Transmembrane Movement and Metabolism of Platelet Activating Factor (PAF-Acether) by..., Lachachi [/bib_ref]. Additionally, PAF exerts the vital roles on cardiovascular pathophysiological processes, which have been proved [bib_ref] Ethnopharmacology and the Development of Natural PAF Antagonists as Therapeutic Agents, Braquet [/bib_ref] [bib_ref] Role of Platelet-Activating Factor in Cardiovascular Pathophysiology, Montrucchio [/bib_ref]. Furthermore, previous studies have confirmed that its related preparations (ginkgolide injection, ginkgo diterpene lactone meglumine injection, and ginkgolide B injection, etc.) have good efficacy in treating cardiovascular and cerebrovascular diseases [bib_ref] Neuroprotective Effects of Ginkgolide B against Ischemic Stroke: A Review of Current..., Nabavi [/bib_ref] [bib_ref] Protective Effects of Ginkgo Terpene Lactones Meglumine Injection on Focal Cerebral Ischemia..., Luo [/bib_ref] [bib_ref] The Effect of Intravenous Ginkgolide on Clinical Improvement of Patients with Acute..., Dong [/bib_ref]. To be specific, ginkgolide injection contains bilobalide, ginkgolide A, ginkgolide B, and ginkgolide C. Ginkgo diterpene lactone meglumine injection is composed of ginkgolide A, ginkgolide B, and ginkgolide K. Additionally, ginkgolide B injection is the only ginkgolide monomer injection in the world (purity>99%). The aforementioned three injections are widely used for IS caused by blood stasis blocking meridians in clinical practice. In general, ginkgo diterpene lactone meglumine injection is superior to ginkgolide injection and ginkgolide B injection because it is a reasonable combination of ginkgolide A, ginkgolide B, and ginkgolide K monomers with the strongest diterpene activity .
Although ginkgo terpene lactone injections have been shown to have good efficacy on IS via several randomized controlled trials (RCTs), the main effects and corresponding add-on effects of these injections have not been systematically evaluated. The current study aimed to conduct a meta-analysis of RCTs of ginkgo terpene lactone injections in the treatment of IS and then systematically evaluate their effects and safeties. Those findings may provide guidance for clinical decision-making.
# Materials and methods
## Searching strategy
A systematic search was conducted in several databases including PubMed, China National Knowledge Infrastructure (CNKI), Chinese Science and Technology Journals Database (VIP), Chinese Biomedical Literature Database (CBM), Wanfang Database, Embase, and the Cochrane Library. The search time ranged from the establishment time of each database to 31 August 2021. The screening language was either Chinese or English. The method of combining subject words and free words was used for searching articles such as ("Ischemic stroke" OR "Cerebral infarction") AND ("Ginkgo terpene lactone injection" OR "Ginkgolides" OR "Ginkgolide A, BN52020" OR "Ginkgolide B, BN52021" OR "Ginkgolide C, BN52022" OR "Ginkgolide M, BN52023" OR "Ginkgolide J, BN52024).
## Inclusion criteria
Inclusion criteria were abided by the participants, interventions, comparison/control, outcomes, and study design (PICOS) format, which were as follows: 1) Participants in the study were adults (age ≥18 years) with IS according to diagnostic criteria, regardless of gender or ethnicity; 2) Study design: RCTs in Chinese or English; 3) Intervention measures: The experimental groups were used for ginkgo terpene lactone preparations or recombinant tissue plasminogen activator (rt-PA) thrombolytic therapy on the basis of ginkgo terpene lactone preparations; 4) The control group received only other types of drugs, and the form of drugs was not limited; 5) The outcome measures met the following primary or secondary outcomes: the primary outcomes were the clinical efficacy defined according to the nationally approved criteria (The Fourth National Conference on cerebrovascular diseases, 1996). To be specific, basic heal was indicated when the decreases of neurological function deficit were between 90 and 100%, and the level of sick was "0." The decreases of neurological function deficit were between 46 and 90%, and the level of sick was between "1" and "3," indicating significant progresses. Progress was determined when the decreases of neurological function deficit were between 18 and 45%. Ineffectiveness was determined when the decreases of neurological function deficit were <17%. The secondary outcomes were the changes of neurological function deficit evaluated by the National Institute of Health Stroke Scale (NIHSS) [bib_ref] Measurements of Acute Cerebral Infarction: A Clinical Examination Scale, Brott [/bib_ref] or Chinese clinical neurological function deficit scale (NFDS) (The Fourth National Conference on cerebrovascular diseases, 1996) and adverse drug reactions/ adverse drug events (ADRs/ADEs).
## Exclusion criteria
The exclusion criteria were 1) animal experiments, 2) repeated publications, 3) review, 4) no clear inclusion and exclusion criteria or incomplete inclusion and exclusion criteria, and 5) other diseases.
## Data extraction and quality assessment
In total, two independent reviewers conducted the literature screening and data extraction. The following data were retrieved: title, author, publication date, sample size, baseline characteristics, interventions, course of treatment, outcome evaluation indicators, and ADRs/ADEs. If any disagreement arose, it was resolved through discussion with the third evaluator.
Quality assessment was performed by the Cochrane risk-ofbias tool, which consisted of random allocation method, allocation concealment, blinding methods, incomplete outcomes, and other biases. Each study was rated "yes," "no," and "unclear". "Yes" indicated the low risk of bias. "No" indicated the high risk of bias. "Unclear" indicated the unclear risk of bias.
# Statistical analysis
For this meta-analysis, data were analyzed by RevMan 5.3 software or STATA 12.0 software. RevMan 5.3 software was utilized to analyze the outcomes including clinical efficacy, the changes of neurological function deficit, and ADRs/ADEs. Heterogeneity was evaluated using the chi-square (chi 2 ) test and I-square (I 2 ) test. When the homogeneity was high (p > 0.10, I 2 <50%), the fixed effect model was used for analysis. When p < 0.10 and I 2 ≥ 50%, the random effect model was used for analysis. Measurement data were presented as mean difference (MD), and categorical data were presented as odds ratio (OR) or relative risk (RR) with 95% confidence interval (95% CI). Additionally, publication bias was evaluated by funnel plots with RevMan 5.3 software and Egger's test with STATA 12.0 software.
# Results
## Study selection process
According to the search strategy, a total of 1,579 references were initially retrieved. After the duplicate articles were removed, the remaining 378 articles were filtered according to the inclusion and exclusion criteria, and then 327 articles were excluded. With further reading the full text, 28 studies were finally included for meta-analysis. The screening process is shown in [fig_ref] FIGURE 1 |: Flowchart of literature search and selection process [/fig_ref].
## Characteristics of included studies
The basic characteristics of ginkgo terpene lactone preparations used in 28 studies are listed in [fig_ref] TABLE 1 |: Basic characteristics of included studies [/fig_ref]. Among them, 23 RCTs were used to evaluate the effects of ginkgo terpene lactone preparations as the main treatment intervention for IS. Specifically, 23 studies involving 2,389 patients were as follows: [bib_ref] Clinical Observation of Ginkgolide Injection in the Treatment of 48 Cases of..., Ren [/bib_ref] ; [bib_ref] Clinical Study of Ginkgolide in the Treatment of Phlegm and Blood Stasis..., Gu [/bib_ref]
## Outcome publication country
Publication language . Additionally, five references [bib_ref] Cost Effectiveness of Ginkgo Diterpene Lactone Glucosamine Combined with RT-PA Intravenous Thrombolysis..., Kong [/bib_ref] [bib_ref] Efficacy Evaluation of Ginkgo Diterpene Lactone Glucosamine Injection after Intravenous Thrombolysis in..., Zhai [/bib_ref] [bib_ref] Clinical Observation of Ginkgo Diterpene Lactone Glucosamine Injection Combined with RT-PA Intravenous..., Zhang [/bib_ref] [bib_ref] Effect of Ginkgo Diterpene Lactone Glucosamine Injection Combined with Alteplase Intravenous Thrombolysis..., Zhou [/bib_ref] [bib_ref] Clinical Efficacy of Urokinase Intravenous Thrombolysis Combined with Ginkgo Diterpene Lactone Glucosamine..., Yu [/bib_ref] involving 533 patients were included, all of whom were injected with ginkgo diterpene lactone meglumine and rt-PA to evaluate the add-on effect of ginkgo terpene lactone injection on acute IS. Across all studies, the average age of the patients was approximately between 40 and 70 years. There were more male participants than female participants. In addition, the characteristics such as course of treatment in different groups were similar.
[formula] E C E C E C E C E C - - - - Guan( [/formula]
## Quality evaluation on included studies
The quality of each of the 28 RCTs is shown in [fig_ref] FIGURE 2 |: Quality evaluation of included literatures [/fig_ref]. Although all articles mentioned "random," only 12 studies described the specific randomization method [bib_ref] Clinical Study of Ginkgolide in the Treatment of Phlegm and Blood Stasis..., Gu [/bib_ref] [bib_ref] Phase III Clinical Trial of Ginkgo Diterpene Lactone Glucosamine Injection for Syndrome..., Zhao [/bib_ref] [bib_ref] Clinical Observation of Ginkgolide Injection in the Treatment of Apoplexy with Blood..., Zhang [/bib_ref] [bib_ref] Clinical Effect of Ginkgo Diterpene Lactone Glucosamine Injection on Acute Ischemic Stroke, Feng [/bib_ref] [bib_ref] Clinical Observation of Ginkgo Diterpene Lactone Glucosamine Injection Combined with RT-PA Intravenous..., Zhang [/bib_ref] [bib_ref] Effect of Ginkgo Diterpene Lactone Glucosamine Injection Combined with Alteplase Intravenous Thrombolysis..., Zhou [/bib_ref] [bib_ref] Costminimization Analysis of Ginkgolide Injection versus Butylphthalide Injection in the Treatment of..., Zhou [/bib_ref] [bib_ref] Clinical Observation of Ginkgo Diterpene Lactone Glucosamine Injection in the Treatment of..., Zhu [/bib_ref]. A total of three trials described the random allocation plan [bib_ref] Clinical Study of Ginkgolide in the Treatment of Phlegm and Blood Stasis..., Gu [/bib_ref] [bib_ref] Costminimization Analysis of Ginkgolide Injection versus Butylphthalide Injection in the Treatment of..., Zhou [/bib_ref] , and the remaining trials did not report it. There were only four trials [bib_ref] Clinical Study of Ginkgolide in the Treatment of Phlegm and Blood Stasis..., Gu [/bib_ref] [bib_ref] Phase III Clinical Trial of Ginkgo Diterpene Lactone Glucosamine Injection for Syndrome..., Zhao [/bib_ref] [bib_ref] Costminimization Analysis of Ginkgolide Injection versus Butylphthalide Injection in the Treatment of..., Zhou [/bib_ref] using the blind method to evaluate the research objects and outcomes. All articles had the complete outcome data and did not describe the other risk biases.
## Meta-analysis of ginkgo terpene lactone preparations as the main treatment intervention for is
## Analysis of clinical efficacy
A total of 20 articles were identified which compared the clinical efficacy of ginkgo terpene lactone injections with that of other drugs for further subgroup analysis. In brief, two articles reported the efficacy of ginkgolide B injection with no statistical heterogeneity (p = 0.34, I 2 = 0%), so the fixed effect model was employed for analysis. The results showed that clinical efficacy of ginkgolide B injection in the treatment of IS was similar to that of the control group [RR = 1.13, 95% CI (0.95, 1.33), Z = 1.36, p = 0.17]. A total of six trials reported the clinical efficacy of ginkgolide injection in treatment of IS, following the test of heterogeneity (p = 0.43, I 2 = 0%), so the fixed effect model was employed for analysis. The results showed that ginkgolide injection achieved better clinical efficacy [RR = 1.10, 95% CI (1.02, 1.18), Z = 2.36, p = 0.02]. Moreover, the heterogeneity of results was not statistical in 12 trials which reported the effectiveness of ginkgo diterpene lactone meglumine injection (p = 0.83, I 2 = 0%), so the fixed effect model was adopted. The results showed that the clinical efficacy of ginkgo diterpene lactone meglumine injection was superior to that of control group [RR = 1.18, 95% CI (1.12, 1.24), Z = 6.36, p < 0.001] [fig_ref] FIGURE 3 |: Forest plot of ginkgo terpene lactone preparations on clinical efficacy [/fig_ref].
## Changes of neurological function deficit
In total, 14 articles assessed the effects of ginkgo terpene lactone preparations on changes of neurological function deficit by NIHSS or NFDS. Among the articles, the heterogeneity of results was statistical (p = 0.0001, I 2 = 68%), so the random effect model was adopted. To be specific, two articles reported the changes of neurological function deficit of ginkgolide B injection, and the results showed that the effect of ginkgolide B in improving neurological function was similar to that of the control group [fig_ref] FIGURE 4 |: Forest plot of ginkgo terpene lactone preparations on the changes of neurological... [/fig_ref].
# Analysis of safety
Among the 23 included references, there were 15 references which reported the ADRs/ADEs including skin rash, nausea, fatigue, chest tightness, and palpitations. However, all ADRs/ ADEs were disappeared when the treatment was over. There were no obvious changes in the liver and kidney function before and after treatment. The rate of ADRs/ADEs was analyzed, and the results showed that it was similar between the experiment group and control group [OR = 0.95, 95% CI (0.55, 1.62), Z = 0.20, p = 0.84] [fig_ref] FIGURE 5 |: Forest plot of ginkgo terpene lactone preparations on ADRs/ADEs [/fig_ref].
## Analysis of risk of publication bias
To evaluate the risk bias of 20 articles on the clinical efficacy and 14 articles on the changes of neurological function deficit, funnel plots were used. As shown in Supplemental [fig_ref] FIGURE 1 |: Flowchart of literature search and selection process [/fig_ref] , most of the data were concentrated at the top, and the images showed basic symmetry. Furthermore, the absence of the publication bias was supported by Egger's test (p = 0.21, p = 0.32).
## Meta-analysis of ginkgo diterpene lactone meglumine injection combined with thrombolytic therapy in treatment for is
A total of five articles involving ginkgo diterpene lactone meglumine injection combined with rt-PA were included to explore its add-on effects on treatment for acute IS.
## Analysis of clinical efficacy
In total, three articles reported the clinical efficacy with no statistical heterogeneity (p = 0.21, I 2 = 36%), so the fixed effect model was employed for analysis. The results showed that the clinical efficacy of ginkgolide diterpene lactone meglumine injection combined with rt-PA was better than that of rt-PA alone [OR = 1.91, 95% CI (1.13, 3.22), Z = 2.41, p = 0.02] [fig_ref] FIGURE 6 |: Forest plot of add-on effects of ginkgo diterpene lactone meglumine injection on... [/fig_ref].
## Changes of neurological function deficit
In total, three articles reported the changes of neurological function deficit with no statistical heterogeneity (p = 0.60, I 2 = 0%), so the fixed effect model was utilized. The results showed that ginkgo diterpene lactone meglumine injection combined
# Analysis of safety
Among the five included articles, there were four articles that reported the ADRs/ADEs including nausea, fatigue, chest tightness, palpitations, and hemorrhage transformation. As presented in [fig_ref] FIGURE 8 |: Forest plot of meta-analysis of add-on effects of ginkgo diterpene lactone meglumine... [/fig_ref] , the rate of ADRs/ADEs was similar between the experiment group and control group [OR = 0.89, 95% CI (0.42, 1.88), Z = 0.32, p = 0.75].
# Discussion
In this systematic review, we found that ginkgo terpene lactone injections have good therapeutic effects on patients with IS, especially ginkgo diterpene lactone meglumine injection was more effective for IS than other types of drugs in the aspects of improving clinical efficacy and neurological function. Furthermore, the combination of ginkgo diterpene lactone meglumine injection and rt-PA was superior to only rt-PA for clinical efficacy, supporting add-on effects of ginkgo terpene lactone preparations in thrombolytic therapy on acute IS. Ginkgo terpene lactones, the main active ingredients from Ginkgo leaves [bib_ref] Research Development of Ginkgo Terpene Lactones, Geng [/bib_ref] , have the effects of blocking PAF activity [bib_ref] Human Pharmacokinetics of Ginkgo Terpene Lactones and Impact of Carboxylation in Blood..., Liu [/bib_ref] , anti-inflammatory and antioxidant activities [bib_ref] Study on the Antioxidant Activity of Ginkgo Terpene Lactone, Gong [/bib_ref] , and regulating excitatory amino acids and brain cell energy metabolism [bib_ref] Effect of Ginkgolide B on Striatal Extracellular Amino Acids in Middle Cerebral..., Yang [/bib_ref]. Initially, Ginkgo biloba extract (EGb761) has been originated by Dr Willmar Schwabe Pharmaceuticals which contains approximately 24% flavone glycosides and 6% terpene lactones (ginkgolides A, B, and C and bilobalide) [bib_ref] Current Perspectives on the Beneficial Role of Ginkgo Biloba in Neurological and..., Nash [/bib_ref]. Ginaton and Shuxuening injections are the typical representatives of EGb761, which are still widely used in the treatment of IS. However, EGb761 contains a low content of terpene lactones and multiple medicinal components with complex interactions. With the development of pharmaceutical technology, high-purity injections have been proposed to reduce the incidence of adverse reactions. Ginkgo diterpene lactone meglumine injection is one of the monomer injections Frontiers in Pharmacology | www.frontiersin.org March 2022 | Volume 13 | Article 821937 8 that have been approved for marketing, which includes effective ingredients of ginkgolides A, B, and K with advantages of high content of active ingredients and low adverse reactions [bib_ref] Post-Marketing Study on Clinical Safety of Ginkgo Diterpene Lactone Meglumine Injection in..., Zhou [/bib_ref]. Additionally, previous studies have demonstrated that ginkgolide injection and ginkgolide B can protect against IS as well [bib_ref] Neuroprotective Effects of Ginkgolide B against Ischemic Stroke: A Review of Current..., Nabavi [/bib_ref] [bib_ref] Cost-Effectiveness Analysis of Ginkgolide Injection in the Treatment of Ischemic Stroke Based..., Xiang [/bib_ref]. In this metaanalysis, we further found that clinical efficacy of ginkgo terpene lactone injections (especially ginkgo diterpene lactone meglumine injection) was superior to that of other drugs in the treatment of IS including ginkgo leaf extract and dipyridamole injection and Shuxuening injection. Additionally, none of the included studies reported serious ADRs/ADEs. Thus, ginkgo terpene lactone injections, especially ginkgo diterpene lactone meglumine injection may be deemed as better choices for treating IS in clinical practice.
For management of patient with acute IS, intravenous rt-PA was the approved therapy within 4.5 h from the symptom onset [bib_ref] Acute Stroke Intervention: a Systematic Review, Prabhakaran [/bib_ref]. However, rt-PA thrombolysis can lead to not only destroying the blood-brain barrier but also producing cytotoxic brain edema [bib_ref] Tissue-type Plasminogen Activator in the Ischemic Brain: More Than a Thrombolytic, Yepes [/bib_ref]. Therefore, how to improve the adverse effects and clinical efficacy is also a major point of this study. Previously, a meta-analysis has demonstrated that Ginkgo biloba leaf preparation combined with aspirin can help to avoid adverse reactions from high-dose aspirin and concluded that Ginkgo biloba leaf preparation can be used as a complementary therapy . Based on add-on effects, we analyzed the five trials of ginkgo diterpene lactone meglumine injection combined with rt-PA. The results showed that combination therapy was superior to rt-PA alone in improvement of clinical efficacy. Correspondingly, the previous studyhas stated that Ginkgo biloba extract can alleviate the neurotransmitter metabolism and energy and amino acid disturbances induced by rt-PA. Therefore, while rt-PA is the preferred drug for the first prevention of acute IS, ginkgo diterpene lactone meglumine injection can be used as a complementary therapy. There were some limitations in this study. In brief, although this systematic review carried out a comprehensive search, it cannot be ruled out that some of the gray literatures were not included. Additionally, most of the included trials were RCTs in a single center with small samples, so there was the possibility of bias. In future, a multicenter, large sample size, and double-blind placebo-controlled design is needed to verify the reliability of this meta-analysis.
# Conclusion
Ginkgo terpene lactone preparations have good therapeutic effects on patients with IS. For acute IS, ginkgo diterpene lactone meglumine injection can be used as a complementary therapy to improve the clinical efficacy of rt-PA. However, further large and rigorous trials should be also warranted.
# Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.
[fig] FIGURE 1 |: Flowchart of literature search and selection process. Frontiers in Pharmacology | www.frontiersin.org March 2022 | Volume 13 | Article 821937 [/fig]
[fig] FIGURE 2 |: Quality evaluation of included literatures. Frontiers in Pharmacology | www.frontiersin.org March 2022 | Volume 13 | Article 821937 with rt-PA in improving neurological function was superior to only rt-PA [MD = −3.31, 95% CI (−3.64, −2.98), Z = 19.63, p < 0.001](Figure 7). [/fig]
[fig] FIGURE 3 |: Forest plot of ginkgo terpene lactone preparations on clinical efficacy. Frontiers in Pharmacology | www.frontiersin.org March 2022 | Volume 13 | Article 821937 [/fig]
[fig] FIGURE 4 |: Forest plot of ginkgo terpene lactone preparations on the changes of neurological function deficit. [/fig]
[fig] FIGURE 5 |: Forest plot of ginkgo terpene lactone preparations on ADRs/ADEs; ADRs/ADEs, adverse drug reactions/adverse drug events. [/fig]
[fig] FIGURE 6 |: Forest plot of add-on effects of ginkgo diterpene lactone meglumine injection on clinical efficacy. [/fig]
[fig] FIGURE 7 |: Forest plot of add-on effects of ginkgo diterpene lactone meglumine injection on changes of neurological function deficit. [/fig]
[fig] FIGURE 8 |: Forest plot of meta-analysis of add-on effects of ginkgo diterpene lactone meglumine injection on ADRs/ADEs; ADRs/ADEs, adverse drug reactions/ adverse drug events.Frontiers in Pharmacology | www.frontiersin.orgMarch 2022 | Volume 13 | Article 821937 [/fig]
[table] TABLE 1 |: Basic characteristics of included studies. [/table]
|
Roles of Peroxisomes in the Rice Blast Fungus
The rice blast fungus, Magnaporthe oryzae, is a model plant pathogenic fungus and is a severe threat to global rice production. Over the past two decades, it has been found that the peroxisomes play indispensable roles during M. oryzae infection. Given the importance of the peroxisomes for virulence, we review recent advances of the peroxisomes roles during M. oryzae infection processes. We firstly introduce the molecular mechanisms and life cycles of the peroxisomes. And then, metabolic functions related to the peroxisomes are also discussed. Finally, we provide an overview of the relationship between peroxisomes and pathogenicity.
# Introduction
Peroxisomes are single membrane-bound microbodies which existed in all eukaryotic cells. In different organisms and environmental conditions, their abundance can be changed rapidly, and functions could be different. The abundance of peroxisomes is coordinated by several cellular processes, including peroxisome biogenesis, peroxisome proliferation, and peroxisome degradation [bib_ref] Peroxisomes take shape, Smith [/bib_ref]. In budding yeast, the genes involved in those cellular processes include a set of PEX genes, which encode peroxins [bib_ref] Yeast and filamentous fungi as model organisms in microbody research, Van Der Klei [/bib_ref]. Up to now, over 30 PEX genes have been found in different organisms [bib_ref] Yeast and filamentous fungi as model organisms in microbody research, Van Der Klei [/bib_ref].
There are several important metabolic processes that take place in peroxisomes, which involve fatty acidsoxidation, glyoxylate cycle, hydrogen peroxide detoxification and secondary metabolite biosynthesis, and so forth [bib_ref] Peroxisomes take shape, Smith [/bib_ref]. The roles of peroxisomes have been extensively uncovered in yeast, filamentous fungi, plant, and human [bib_ref] Peroxisomal dynamics, Platta [/bib_ref]. Some peroxisome functions are species specific, such as the methanol assimilation in yeast [bib_ref] Yeast peroxisomes: function and biogenesis of a versatile cell organelle, Van Der Klei [/bib_ref] , the glyoxylate cycle in plant seeds [bib_ref] Plant peroxisomes: biogenesis and function, Hu [/bib_ref] , and the plasmalogens biosynthesis in mammal [bib_ref] Biochemistry of mammalian peroxisomes revisited, Wanders [/bib_ref]. In filamentous fungi, they can develop a special compartment, the peroxisome-derived woronin body, to seal the septal pore when suffered cellular wounding [bib_ref] Peroxisome assembly and functional diversity in eukaryotic microorganisms, Pieuchot [/bib_ref].
The molecular mechanisms of peroxisome life cycle have been extensively studied in yeast [bib_ref] Yeast and filamentous fungi as model organisms in microbody research, Van Der Klei [/bib_ref]. Peroxisome studies in the filamentous fungi are also increased rapidly, especially in the model filamentous fungi Aspergillus nidulans [bib_ref] Genetic analysis of the role of peroxisomes in the utilization of acetate..., Hynes [/bib_ref] [bib_ref] Regulatory genes controlling fatty acid catabolism and peroxisomal functions in the filamentous..., Hynes [/bib_ref] and Neurospora crassa [bib_ref] Identification of PEX33, a novel component of the peroxisomal docking complex in..., Managadze [/bib_ref] [bib_ref] Pex7p and Pex20p of Neurospora crassa function together in PTS2-dependent protein import..., Sichting [/bib_ref] [bib_ref] The peroxin PEX14 of Neurospora crassa is essential for the biogenesis of..., Managadze [/bib_ref] , the plant pathogens Colletotrichum orbiculare [bib_ref] Atg26-mediated pexophagy is required for host invasion by the plant pathogenic fungus..., Asakura [/bib_ref] [bib_ref] Peroxisome biogenesis factor PEX13 is required for appressorium-mediated plant infection by the..., Fujihara [/bib_ref] [bib_ref] Colletotrichum orbiculare FAM1 encodes a novel Woronin body-associated Pex22 peroxin required for..., Kubo [/bib_ref] and Magnaporthe oryzae [bib_ref] MoPex19, which is essential for maintenance of peroxisomal structure and woronin bodies,..., Li [/bib_ref] [bib_ref] The PEX7-mediated peroxisomal import system is required for fungal development and pathogenicity..., Goh [/bib_ref] [bib_ref] PTS1 peroxisomal import pathway plays shared and distinct roles to PTS2 pathway..., Wang [/bib_ref] [bib_ref] The role of Snx41-based pexophagy in Magnaporthe development, Deng [/bib_ref] [bib_ref] One of three Pex11 family members is required for peroxisomal proliferation and..., Wang [/bib_ref] , and the human pathogens Candida albicans [bib_ref] The activity of the glyoxylate cycle in peroxisomes of Candida albicans depends..., Piekarska [/bib_ref] [bib_ref] A Fox2-dependent fatty acid beta-oxidation pathway coexists both in peroxisomes and mitochondria..., Gabriel [/bib_ref] , Aspergillus fumigatus [bib_ref] Fungal siderophore biosynthesis is partially localized in peroxisomes, Gründlinger [/bib_ref] [bib_ref] Characterization of the major Woronin body protein HexA of the human pathogenic..., Beck [/bib_ref] , and Cryptococcus neoformans [bib_ref] Peroxisomal and mitochondrial -oxidation pathways influence the virulence of the pathogenic fungus..., Kretschmer [/bib_ref] [bib_ref] Peroxisome function regulates growth on glucose in the basidiomycete fungus Cryptococcus neoformans, Idnurm [/bib_ref]. An important aspect is that the peroxisomes are found to play key roles in fungal pathogenicity towards their host, including plants and human. In all these pathogenic fungi, the roles of peroxisomes in M. oryzae have received extensive concern.
Over the past two decades, many components involving roles of peroxisomes have been identified in M. oryzae [fig_ref] Table 1: Peroxisome-related genes identified in M [/fig_ref]. This review focuses on recent advances in our understanding of peroxisomes in M. oryzae, including a description of the peroxisome life cycle during fungal infection, and an overview of their remarkable functions relevant to pathogenesis.
## Life cycle of the peroxisomes
The peroxisome life cycle mainly includes peroxisome biogenesis, peroxisome proliferation, and peroxisome degradation [bib_ref] The life cycle of the peroxisome, Titorenko [/bib_ref]. The peroxisomes are thought to be born originally from the endoplasmic reticulum (ER) [bib_ref] The life cycle of the peroxisome, Titorenko [/bib_ref]. Basically, the biogenesis process contains peroxisomal membrane proteins (PMPs) acquisition and peroxisomal matrix proteins import. Peroxisome proliferation can be achieved by either Peroxisomal matrix protein import [bib_ref] Host invasion during rice-blast disease requires carnitine-dependent transport of peroxisomal acetyl-CoA, Ramos-Pamplona [/bib_ref]
## Mopex7
Receptor of PTS2 peroxisomal matrix proteins [bib_ref] The PEX7-mediated peroxisomal import system is required for fungal development and pathogenicity..., Goh [/bib_ref] [bib_ref] PTS1 peroxisomal import pathway plays shared and distinct roles to PTS2 pathway..., Wang [/bib_ref]
## Mopex19
Peroxisomal membrane proteins import de novo formation from ER or/and peroxisome fission. When the peroxisomes have finished their mission, they can be degraded by pexophagy, an autophagic process [bib_ref] The life cycle of the peroxisome, Titorenko [/bib_ref]. In M. oryzae, there are several PEX genes involving peroxisome life cycle that have been characterized, including MoPEX5, MoPEX6, MoPEX7, MoPEX14, MoPEX19, and MoPEX11 family genes. MoPEX5 and MoPEX7 are involved in matrix proteins import [bib_ref] The PEX7-mediated peroxisomal import system is required for fungal development and pathogenicity..., Goh [/bib_ref] [bib_ref] PTS1 peroxisomal import pathway plays shared and distinct roles to PTS2 pathway..., Wang [/bib_ref] , MoPEX6 participates in receptors import for recycling [bib_ref] Host invasion during rice-blast disease requires carnitine-dependent transport of peroxisomal acetyl-CoA, Ramos-Pamplona [/bib_ref] , MoPex14 functions as a matrix docking protein [bib_ref] The role of Snx41-based pexophagy in Magnaporthe development, Deng [/bib_ref] , MoPEX19 functions as chaperone and receptor for importing of both matrix proteins and PMPs [bib_ref] MoPex19, which is essential for maintenance of peroxisomal structure and woronin bodies,..., Li [/bib_ref] , and MoPEX11 family genes are involved in peroxisomal fission processes [bib_ref] One of three Pex11 family members is required for peroxisomal proliferation and..., Wang [/bib_ref].
## Peroxisome biogenesis.
In yeast, during peroxisome biogenesis, peroxisome membrane proteins (PMPs) should firstly be inserted into membranes, which are mediated by PEX3, PEX16, and PEX19 [bib_ref] Saccharomyces cerevisiae Pex3p and Pex19p are required for proper localization and stability..., Hettema [/bib_ref] [bib_ref] PEX19 binds multiple peroxisomal membrane proteins, is predominantly cytoplasmic, and is required..., Sacksteder [/bib_ref] [bib_ref] The peroxin Pex3p initiates membrane assembly in peroxisome biogenesis, Ghaedi [/bib_ref]. Then the peroxisomal matrix proteins, which are synthesized in the cytoplasm, are translocated into the peroxisomes by peroxisome membrane docking complex [bib_ref] PEX3 functions as a PEX19 docking factor in the import of class..., Fang [/bib_ref]. Most of the peroxisomal matrix proteins contain either PTS1 (type I peroxisomal targeting signal) at the C-terminus or PTS2 (type II peroxisomal targeting signal) at the N-terminal and can be recognized by shuttle receptors Pex5 or Pex7-mediated complex, respectively. The cargos on the PTS1 and PTS2 receptors are accepted by the Pex13/Pex14/Pex17 docking complex, and then the receptors are recycled by the ubiquitin system. The ubiquitinated receptors can be extracted into cytoplasm by AAA+ ATPases Pex1 and Pex6 [bib_ref] The life cycle of the peroxisome, Titorenko [/bib_ref].
In M. oryzae, MoPex19 is the ortholog of yeast PMPs receptor Pex19. Pex19 protein is located in the cytoplasm and newly formed peroxisomes, which is consistent with its PMPs receptor function that shuttles the PMPs between the cytosol and peroxisomal membrane [bib_ref] PEX19 binds multiple peroxisomal membrane proteins, is predominantly cytoplasmic, and is required..., Sacksteder [/bib_ref]. Deletion of MoPEX19 will lead to PMPs mislocalization. PMP47 is a representative PMP which is normally distributed in the peroxisomes in the wild type strain, while in the MoPEX19 deletion mutants, its localization pattern is changed, which is distributed in the cytoplasm [bib_ref] MoPex19, which is essential for maintenance of peroxisomal structure and woronin bodies,..., Li [/bib_ref]. This demonstrated that the PMP47 cannot be imported into the peroxisomes. Moreover, in the Δmopex19 mutants, peroxisomal structures and peroxisomederived woronin bodies are both absent [bib_ref] MoPex19, which is essential for maintenance of peroxisomal structure and woronin bodies,..., Li [/bib_ref] , indicating that the MoPex19 is essential for biogenesis of peroxisomes and woronin bodies.
M. oryzae MoPex5 and MoPex7 are also proved to function as receptors of peroxisomal matrix proteins, which are involved in importing of the matrix proteins into peroxisomes [bib_ref] The PEX7-mediated peroxisomal import system is required for fungal development and pathogenicity..., Goh [/bib_ref] [bib_ref] PTS1 peroxisomal import pathway plays shared and distinct roles to PTS2 pathway..., Wang [/bib_ref]. Disruption of MoPEX5 and MoPEX7 will block the PTS1 and PTS2 peroxisomal import pathways, respectively. In the wild type strain, RFP-PTS1 and GFP-PTS2 are normally distributed in the punctuate peroxisomes, while in the Δmopex5 mutants, RFP-PTS1 is dispersed in cytoplasm but GFP-PTS2 is still located in peroxisomes. In contrast, in the Δmopex7 mutants, RFP-PTS1 is still located in peroxisomes but GFP-PTS2 is dispersed in cytoplasm.
These results demonstrated the MoPex5-mediated PTS1 peroxisomal import pathway and MoPex7-mediated PTS2 peroxisomal import pathway separately function in the rice blast fungus [bib_ref] The PEX7-mediated peroxisomal import system is required for fungal development and pathogenicity..., Goh [/bib_ref] [bib_ref] PTS1 peroxisomal import pathway plays shared and distinct roles to PTS2 pathway..., Wang [/bib_ref]. The thiolase MoThl1 is a candidate PTS2 protein; it is failed to be located at the peroxisomes in the Δmopex7 mutants [bib_ref] The PEX7-mediated peroxisomal import system is required for fungal development and pathogenicity..., Goh [/bib_ref] , which further supports the role of MoPex7 in the PTS2 peroxisomal protein import pathway.
The function of Pex6 ortholog in M. oryzae, MoPex6, was also evaluated [bib_ref] Host invasion during rice-blast disease requires carnitine-dependent transport of peroxisomal acetyl-CoA, Ramos-Pamplona [/bib_ref]. In MoPEX6 disruption mutants, the GFP-SRL protein is diffused in the cytoplasm, failed to be localized in the punctuate peroxisomes in mycelia, conidia, germ tubes, and appressoria [bib_ref] Host invasion during rice-blast disease requires carnitine-dependent transport of peroxisomal acetyl-CoA, Ramos-Pamplona [/bib_ref] , indicating the PTS peroxisomal import pathway is blocked. This result is consistent with the cellular function of Pex6, which is involved in recycling of matrix protein receptors (Pex5 and Pex7) during peroxisome biogenesis.
In the MoPEX14 disruption mutants, the GFP-SRL protein is also mislocalized to the cytoplasm, while when the PEX14 or PEX14 is expressed in the Δmopex14 mutants, the punctate localization of GFP-SRL can be restored [bib_ref] The role of Snx41-based pexophagy in Magnaporthe development, Deng [/bib_ref]. These data confirmed the functions of MoPex14, which act as a matrix docking protein to facilitate peroxisomal protein import and peroxisome biogenesis.
## Peroxisome proliferation.
Peroxisomes can proliferate rapidly according to suitable environment stimulation. The proliferation processes can be achieved by de novo biogenesis from the ER, or by fission from the preexisting peroxisomes. In yeast, the peroxisome fission processes mainly consist of several steps. At the beginning, the mature peroxisomes are elongated by the functions of the peroxisomal membrane protein Pex11. Then the matrix proteins can be imported into the elongated peroxisomes, and the fission machinery can also be imported into appropriate place for fission. The dynamin-like protein Dnm1 is located at the constriction sites and leads to membrane fission processes by GTP hydrolysis. At last, the daughter peroxisomes can be produced from the fission processes, which is achieved by cooperation of several proteins, including Fis1, Dnm1, and the adaptors Mdv1 or Caf4 [bib_ref] The life cycle of the peroxisome, Titorenko [/bib_ref].
The peroxisome fission process in M. oryzae is identified recently, by several independent studies. There are three members of PEX11 family genes in M. oryzae genome, named MoPEX11A, MoPEX11B, and MoPEX11C, respectively [bib_ref] One of three Pex11 family members is required for peroxisomal proliferation and..., Wang [/bib_ref]. However, it seems that only MoPEX11A plays vital role in peroxisome fission. The MoPEX11A deletion mutant exhibits decreased but enlarged peroxisomes compared to the wild type, which demonstrated the MoPEX11A is important for peroxisome elongation and proliferation. In contrast, the MoPEX11B and MoPEX11C deletion mutants are normal in both number and size of the peroxisomes [bib_ref] One of three Pex11 family members is required for peroxisomal proliferation and..., Wang [/bib_ref] , indicating they could not be key regulators during peroxisomal proliferation.
There is only one counterpart of Mdv1/Caf4 protein in M. oryzae, named PEF1 [peroxisome fission protein 1] [bib_ref] Peroxisomal fission is induced during appressorium formation and is required for full..., Chen [/bib_ref]. This gene may play dual roles of Mdv1 and Caf4, because deletion of PEF1 will lead to evident peroxisomal fission defect during the fission inducing conditions. The Δpef1 mutant forms string-linked peroxisomes, in contrast to the punctuate structures in normal cells [bib_ref] Peroxisomal fission is induced during appressorium formation and is required for full..., Chen [/bib_ref]. Similar situation can be found in the Δmdv1Δcaf4 double mutant or the Δfis1 and Δdnm1 mutants in yeast [bib_ref] Dnm1p-dependent peroxisome fission requires Caf4p, Mdv1p and Fis1p, Motley [/bib_ref]. The phenotypic defect indicates that the daughter peroxisomes cannot be cut from the elongated peroxisomes in Δpef1. Pef1 can bind to Fis1 with its N-terminal extension (NTE) region and to Dnm1 with its C-terminal WD40 repeat region. With the help of adaptors Mdv1 or Caf4, the outer membrane protein Fis1 can recruit Dnm1 to peroxisomes for fission [bib_ref] Fission and proliferation of peroxisomes, Schrader [/bib_ref]. Pef1 can be well colocalized with MoFis1, which is recently reported to play important roles in mitochondria fission in M. oryzae [bib_ref] Mitochondrial fission protein MoFis1 mediates conidiation and is required for full virulence..., Khan [/bib_ref]. This is intelligible, because in S. cerevisiae, the peroxisome fission machinery can be also used to facilitate mitochondria fission process.
## Peroxisome degradation.
Peroxisome abundance can be rapidly cleared by a selective autophagic process, which is known as the pexophagy. In yeast, there are two different peroxisome degradation modes, macropexophagy and micropexophagy [bib_ref] Regulation of the selective autophagic degradation of peroxisomes by PtdIns3P and Rab-GTPases, Reidick [/bib_ref]. The macropexophagy sequesters peroxisomes to form a pexophagosome, which leads the peroxisomes to vacuole for degradation. The micropexophagy encloses peroxisomes by vacuolar membrane protrusions and the micropexophagy specific membrane apparatus (MIPA) to vacuole for degradation [bib_ref] Regulation of the selective autophagic degradation of peroxisomes by PtdIns3P and Rab-GTPases, Reidick [/bib_ref]. There are several ATG and PEX genes reported to be involved in perophagy [bib_ref] Regulation of the selective autophagic degradation of peroxisomes by PtdIns3P and Rab-GTPases, Reidick [/bib_ref].
In Pichia pastoris, ATG30 is required for both of macropexophagy and micropexophagy [bib_ref] PpAtg30 tags peroxisomes for turnover by selective autophagy, Farré [/bib_ref] , but no ATG30 homolog gene can be found in M. oryzae. In P. pastoris and C. orbiculare, ATG26 plays key roles in pexophagy, and the CoATG26 is essential for infection process [bib_ref] Atg26-mediated pexophagy is required for host invasion by the plant pathogenic fungus..., Asakura [/bib_ref] [bib_ref] The requirement of sterol glucoside for pexophagy in yeast is dependent on..., Nazarko [/bib_ref]. However, In M. oryzae, MoATG26 is not involved in pexophagy and is dispensable for virulence [bib_ref] The role of Snx41-based pexophagy in Magnaporthe development, Deng [/bib_ref].
In S. cerevisiae, the pexophagy process can be mediated by Atg20/Snx42 [bib_ref] ATG genes involved in non-selective autophagy are conserved from yeast to man,..., Meijer [/bib_ref]. By the assistance of sorting nexins Snx41 and Atg24/Snx4, Atg20/Snx42 can be also involved in endosomal retrieval trafficking [bib_ref] ATG genes involved in non-selective autophagy are conserved from yeast to man,..., Meijer [/bib_ref]. However, only one protein [MoSnx41] with high similarity to Snx41 and Snx42/Atg20 can be found in M. oryzae. Studies have found that the MoSNX41 plays key roles in conidiation and pathogenesis. Deletion of MoSNX41 leads to pexophagy deficiency, demonstrating this gene is involved in pexophagy in M. oryzae. The yeast ScSnx42 can restore the pexophagy deficiency of the Δmosnx41 mutant but failed to recover the defects of conidiation and pathogenesis, indicating the pexophagy process is dispensable for development and pathogenicity in M. oryzae. The function of MoSnx41 in conidiation and pathogenesis is mediated by Snx41-dependent retrieval trafficking, but not pexophagy [bib_ref] The role of Snx41-based pexophagy in Magnaporthe development, Deng [/bib_ref]. As a peroxisomal membrane protein, PEX14 is also involved in pexophagy of Hansenula polymorpha [bib_ref] The Hansenula polymorpha PEX14 gene encodes a novel peroxisomal membrane protein essential..., Komori [/bib_ref]. In M. oryzae, the MoPex14 has also been proved to be essential for pexophagy, but dispensable for pathogenicity.
## Peroxisome differentiation.
Filamentous fungi can form a special structure, called woronin body (WB), which is differentiated from the peroxisomes [bib_ref] Making two organelles from one: woronin body biogenesis by peroxisomal protein sorting, Liu [/bib_ref]. It is accepted that the peroxisome-related woronin bodies are used to seal the septal pore when suffering cellular wounding [bib_ref] Occlusion of the septal pores of damaged hyphae of Neurospora crassa by..., Trinci [/bib_ref]. The formation mechanism of woronin bodies has been well studied in Neurospora crassa [bib_ref] Woronin bodies of filamentous fungi, Markham [/bib_ref]. Hexagonal peroxisome Hex1 is the major structural protein in woronin body, which can be imported into the peroxisome matrix and assembled to form a woronin body core structure. This core structure can recruit WB sorting complex protein [WSC] into the peroxisome membrane. Then the nascent WB can be budded from the mother peroxisomes.
In M. oryzae, the HEX1 homolog gene has been cloned and the roles of the woronin bodies have also been uncovered by analyses of the MoHEX1 functions. The woronin body is proved to be required for development, appressoria formation, and infection hyphae growth and therefore is essential for pathogenicity in M. oryzae [bib_ref] Woronin body function in Magnaporthe grisea is essential for efficient pathogenesis and..., Soundararajan [/bib_ref]. Ultrastructural analyses proved that the woronin bodies are located adjacent to septa of mycelia, germ tubes, and infection hyphae, but they are less observed in conidia. The HEX1 deletion mutant is lack of woronin bodies, and when the cells are damaged, the cellular materials will bleed out through the septal pores. The M. oryzae Hex1 protein contains the PTS1 peroxisome targeting signal, which can help it locate into the peroxisomes. For the woronin bodies being formed from the peroxisomes, defects in peroxisome formation cycles could lead to defect in woronin body formation. Consistent to this hypothesis, nearly all disruption of the genes involving in peroxisome life cycle can result in woronin body deficiency [bib_ref] Woronin body function in Magnaporthe grisea is essential for efficient pathogenesis and..., Soundararajan [/bib_ref].
## Regulation of peroxisome dynamics.
The peroxisomes in the cell are dynamics according to environmental conditions and development stages. Occupancy of the peroxisomes in a cell is determined by several processes, including peroxisome biogenesis, peroxisome proliferation [de novo formation and fission], and peroxisome degradation (pexophagy). De novo biogenesis and fission process are both used to increase peroxisomes numbers. However, how to coordinate de novo formation and peroxisomal fission remains unclear. It is possible that both of the ER formation and fission of the preexisting peroxisomes might exist in normal conditions. It is supposed that, upon the peroxisome fission inducing conditions, such as in fatty acids condition, the peroxisomes are needed to be largely produced in a short time and the de novo formation process may be not enough or effective. In contrast, the peroxisome fission cycles would fulfill peroxisome demands in a limited time. During infection of M. oryzae, the lipid stores should be utilized less than 12 h. The rapid lipolysis of lipid bodies produced mass fatty acids, which in turn induces peroxisome fission process (de novo biogenesis efficiency could also be elevated), and massive peroxisomes are produced in short time, thus promoting fatty acids utilization and facilitating infection. When fatty acids are utilized, the peroxisomes should be decreased in a short time, and then they can be degraded by pexophagy. In C. orbiculare, the CoAtg26-dependent pexophagy is used to recycle cellular amino acids of the appressoria for infection [bib_ref] Atg26-mediated pexophagy is required for host invasion by the plant pathogenic fungus..., Asakura [/bib_ref]. How the peroxisome fission and pexophagy are activated and suspended remains obscure.
The Snf1/AMPK pathway plays central role in response to nutrient stress in M. oryzae. A recent study demonstrated that the MoSnf1 pathway can regulate peroxisomal maintenance and lipid metabolism by responding to nutrient-free environment [bib_ref] Crosstalk between SNF1 pathway and the peroxisome-mediated lipid metabolism in Magnaporthe oryzae, Zeng [/bib_ref]. In Δmosnf1 mutant, the peroxisomes are significantly decreased during appressoria formation. Accordingly, the Δmosnf1 mutant is also defect in lipid droplet mobilization, fails to generate enormous turgor, and loses its virulence [bib_ref] Crosstalk between SNF1 pathway and the peroxisome-mediated lipid metabolism in Magnaporthe oryzae, Zeng [/bib_ref]. Other regulatory mechanisms should also be uncovered in the future.
## Metabolic functions of peroxisomes
Multiple metabolic processes occur in the peroxisomes, which makes the peroxisomes play crucial role in fungal development and pathogenesis. Peroxisomal -oxidation of fatty acids ubiquitously existed for all eukaryotes. The reactive oxygen species [ROS] homeostasis can be also mediated by peroxisomes. Peroxisomes can be involved in many other metabolic pathways, including glyoxylate cycle in plants and fungi [bib_ref] A central role for the peroxisomal membrane in glyoxylate cycle function, Kunze [/bib_ref] , penicillin biosynthesis in Penicillium chrysogenum [bib_ref] De novo peroxisome biogenesis in Penicillium chrysogenum is not dependent on the..., Opaliński [/bib_ref] , and melanin biogenesis in filamentous fungi [bib_ref] Biosynthesis of fungal melanins and their importance for human pathogenic fungi, Langfelder [/bib_ref].
## Fatty acid -oxidation.
The -oxidation metabolism is mainly used to degrade fatty acids for nutrient and energy utilization [bib_ref] Peroxisomal -oxidation-a metabolic pathway with multiple functions, Poirier [/bib_ref]. This metabolic process involves four major enzymes, acyl-CoA oxidase, 2-enoyl-CoA hydratase, 3-hydroxyacyl-CoA dehydrogenase, and 3-ketoacyl-CoA thiolase. Through a four-step pathway mediated by these enzymes, the acetyl-CoA is produced [bib_ref] Peroxisomal -oxidation-a metabolic pathway with multiple functions, Poirier [/bib_ref]. The acetyl-CoA can be fed to the glyoxylate cycle and gluconeogenesis to produce nutrient or metabolites.
In M. oryzae, the Mfp1/Fox2 protein has been shown to play an important role in fatty acid metabolism and pathogenesis [bib_ref] Functional analysis of lipid metabolism in Magnaporthe grisea reveals a requirement for..., Wang [/bib_ref]. The Δmfp1 mutants cannot grow on olive oil or oleic acid as sole carbon sources, indicating its roles in fatty acids metabolism. The expression level of MFP1 is significantly induced in the olive oil condition. The Mfp1GFP fusion protein is well colocalized with FoxA-RFP in the peroxisome-like punctuate structures [bib_ref] Functional analysis of lipid metabolism in Magnaporthe grisea reveals a requirement for..., Wang [/bib_ref]. Because the A. nidulans FoxA has been proved to be a -oxidation enzyme and can be located in the peroxisomes, the M. oryzae Mfp1 should also be located in the peroxisomes and is required for fatty acid -oxidation.
In the past, -oxidation is thought to occur exclusively in peroxisomes in the filamentous fungi, while it can occur in both peroxisomes and mitochondria in mammal. However, recent studies demonstrated that the mitochondrialoxidation is also functional and important for the infection of the fungal pathogens [bib_ref] Mitochondrial -oxidation regulates organellar integrity and is necessary for conidial germination and..., Patkar [/bib_ref]. In M. oryzae, Enoyl-CoA hydratase Ech1 is an important mitochondrial -oxidation enzyme, which is important for conidial germination, appressoria penetration, and host colonization. The Δech1 mutant cannot utilize C14 fatty acid and also cannot well utilize C16 and C18 fatty acids. Consequently, Δech1 is reduced in melanization and sensitive to oxidative stress. Generally, the short-and medium-chain fatty acids (less than 20C) can be oxidized in mitochondria, while long-chain fatty acids (over 20C) should be degraded in peroxisomes to shorter chain fatty acids for full oxidation in mitochondria [bib_ref] Mitochondrial -oxidation regulates organellar integrity and is necessary for conidial germination and..., Patkar [/bib_ref]. Thus, the mitochondrial -oxidation and peroxisomal -oxidation can Acetyl CoA is the main product of fatty acid -oxidation in the appressorium of M. oryzae, which must be transported into different cellular compartments for consumption. The acetyl CoA transportation is catalysed by carnitine acetyl transferases (CATs). In budding yeast S. cerevisiae, Cat2 is involved in transferring acetyl CoA to the cytoplasm, while Cat1 is used to transfer acetyl CoA into mitochondria for utilization by the tricarboxylic acid cycle [bib_ref] Carnitine-dependent metabolic activities in Saccharomyces cerevisiae: three carnitine acetyltransferases are essential in..., Swiegers [/bib_ref]. Two CAT genes, PTH2 and CRAT2, have been studied in M. oryzae [bib_ref] Peroxisomal carnitine acetyl transferase is required for elaboration of penetration hyphae during..., Bhambra [/bib_ref]. The results demonstrated that PTH2 plays major role in acetyl CoA transfer. Pth2-GFP protein is colocalized with the peroxisome marker protein and is abundant in appressoria. The Δpth2 is reduced in melanin deposition, defect in host penetration, and essential for pathogenesis [bib_ref] Peroxisomal carnitine acetyl transferase is required for elaboration of penetration hyphae during..., Bhambra [/bib_ref]. Further analysis found that the Δpth2 mutant cannot utilize some lipid substrates. In contrast, the Δcrat2 displays no evident defect in those mentioned phenotypes [bib_ref] Peroxisomal carnitine acetyl transferase is required for elaboration of penetration hyphae during..., Bhambra [/bib_ref]. The carnitineacylcarnitine carrier protein Crc1 functions in transferring the acetyl CoA cross the mitochondrial membrane. The M. oryzae MoCRC1 deletion mutant is severely reduced in appressorial penetration and invasive growth. MoCRC1 is also needed for utilization of olive oil [bib_ref] A carnitine-acylcarnitine carrier protein, MoCrc1, is essential for pathogenicity in Magnaporthe oryzae, Yang [/bib_ref].
## Glyoxylate cycle.
The glyoxylate cycle is a metabolic pathway which can be normally found in plants and fungi [bib_ref] A central role for the peroxisomal membrane in glyoxylate cycle function, Kunze [/bib_ref]. This pathway can assimilate acetyl CoA for gluconeogenesis and eventually generate glucose. The glyoxylate cycle is mainly induced when the fatty acids and acetate should be used [bib_ref] A central role for the peroxisomal membrane in glyoxylate cycle function, Kunze [/bib_ref]. The fatty acid -oxidation pathway produces massive acetyl CoA, which will be processed by glyoxylate cycle. In this pathway, the acetyl CoA can be converted to glyoxylate by isocitrate lyase, and then the glyoxylate can be further converted to malate by malate synthase. The malate can be further metabolized to hexoses by gluconeogenesis.
The isocitrate lyase (Icl1) and malate synthase (Mls1) are two of principal enzymes involving glyoxylate cycle [bib_ref] A central role for the peroxisomal membrane in glyoxylate cycle function, Kunze [/bib_ref]. In M. oryzae, ICL1 is highly expressed during appressoria formation and penetration stages, indicating that the glyoxylate cycle should be induced in these stages [bib_ref] The glyoxylate cycle is required for temporal regulation of virulence by the..., Wang [/bib_ref]. Defect in peroxisome biogenesis will lead to loss functions of glyoxylate cycle. For example, disruption of MoPEX19 will result in failure in acetate utilization [bib_ref] MoPex19, which is essential for maintenance of peroxisomal structure and woronin bodies,..., Li [/bib_ref].
## Redox homeostasis.
Oxidative reactions are theme of the peroxisome metabolism, which generates massive reactive oxidative species (ROS), especially the hydrogen peroxide (H 2 O 2 ) [bib_ref] Peroxisome metabolism and cellular aging, Titorenko [/bib_ref]. In order to eliminate harmful ROS, ROS scavenging becomes an important peroxisomal metabolism. To detoxification, the hydrogen peroxide can be scavenged by catalases and peroxidases, which are abundant in peroxisomes. A number of catalases and peroxidases are predicted in genome of M. oryzae, and some of which (such as CATB and CPXB) have been reported to function in host ROS detoxification [bib_ref] A novel role for catalase B in the maintenance of fungal cell-wall..., Skamnioti [/bib_ref] [bib_ref] The role of catalase-peroxidase secreted by Magnaporthe oryzae during early infection of..., Tanabe [/bib_ref]. However, no catalase or peroxidase has been revealed to take part in intracellular peroxisome ROS detoxification. Glutathione S-transferases (GSTs) and peroxiredoxins (PRXs) are other antioxidant enzymes existing in peroxisomes [bib_ref] Peroxisomes and oxidative stress, Schrader [/bib_ref]. There is also no GST or PRX protein reported to play roles in peroxisome ROS detoxification. In Alternaria brassicicola and A. fumigatus, a transmembrane protein TmpL has been identified and proved to be important for reduction of intracellular ROS and detoxification of external ROS and thus is important for fungal infection. TmpL can be located in the woronin body, and its expression level is evidently elevated in conidiation and initial invasive growth stages [bib_ref] TmpL, a transmembrane protein required for intracellular redox homeostasis and virulence in..., Kim [/bib_ref].
During peroxisome -oxidation process, acetyl-CoA formation will accompany mass formation of FADH2 and NADH. To keep peroxisome redox homeostasis, the FADH2 and NADH should be eliminated. For the NADH, it can be reoxidated to NAD + . This reaction can be catalyzed by the peroxisomal lactate dehydrogenase, which can mediate production of lactate from pyruvate. In M. oryzae, the pyruvate is generated by the alanine: glyoxylate aminotransferase 1 (AGT1), which transfers the alanine amino group to glyoxylate and results in formation of the pyruvate. AGT1GFP is colocalized with RFP-MTS1 fusion protein, demonstrating that AGT1 is located in the peroxisomes [bib_ref] Peroxisomal alanine: glyoxylate aminotransferase AGT1 is indispensable for appressorium function of the..., Bhadauria [/bib_ref]. The Δagt1 mutant cannot form appressoria on artificial inductive surface. When the NAD + and pyruvate were added during conidia germination on artificial inductive surface, the appressorium formation can be restored [bib_ref] Peroxisomal alanine: glyoxylate aminotransferase AGT1 is indispensable for appressorium function of the..., Bhadauria [/bib_ref]. Thus, the AGT-mediated pyruvate generation can function as one of factors to maintain redox homeostasis during appressoria formation.
## Melanin biosynthesis.
In filamentous fungi, peroxisomes not only are crucial for the primary metabolism, but also play important roles in the biosynthesis of secondary metabolites. Many plant pathogenic fungi can produce melanin to protect the conidia to survive in different environment and to facilitate host penetration during infection. The dihydroxy naphthalene (DHN) melanin is well studied and found to be essential for appressorial mediated penetration in M. oryzae. The fungal appressorium contains a distinct melanin layer located between the cell wall and the membrane, which can be used to generate turgor for penetration. The DHN melanin is synthesized by the polyketide pathway, through which the peroxisomes-derived acetyl-CoA can be used to produce the 1,3,6,8-tetrahydroxynaphthalene [fig_ref] Figure 1: Life cycle and functions of the peroxisomes in M [/fig_ref]. The 1,3,6,8-THN is then used to synthesize the DHN melanin, catalyzed by a series of enzymes, including Alb1, Rsy1, and Buf1 [bib_ref] Melanins in fungal pathogens, Hamilton [/bib_ref].
Because the acetyl-CoA is mainly produced in peroxisomes, defects in peroxisome formation would lead to block in melanin synthesis. Consistent with this prediction, all disruptions of the peroxisome biogenesis-related genes can result in melanin deficiency.
## Cell wall
Biosynthesis. The fungal cell wall mainly consists of chitin, -1,3-glucan, -1,6-glucan, and mannoproteins [bib_ref] The structure and synthesis of the fungal cell wall, Bowman [/bib_ref]. This rigid structure can protect fungal from extracellular stresses and is flexible for adaptation to development and environment [bib_ref] The structure and synthesis of the fungal cell wall, Bowman [/bib_ref]. It is believed that fungal cell wall chitin and glucan are derived from acetyl-CoA, and the defects in peroxisomes will lead to deficiency in cell wall integrity. Consistent with this hypothesis, the M. oryzae MoPEX5, MoPEX6, and MoPEX19 deletion mutants are all sensitive to the cell wall-perturbing agents, such as Congo Red and Calcofluor White [bib_ref] MoPex19, which is essential for maintenance of peroxisomal structure and woronin bodies,..., Li [/bib_ref] [bib_ref] PTS1 peroxisomal import pathway plays shared and distinct roles to PTS2 pathway..., Wang [/bib_ref] [bib_ref] Host invasion during rice-blast disease requires carnitine-dependent transport of peroxisomal acetyl-CoA, Ramos-Pamplona [/bib_ref].
## Peroxisome and pathogenicity
Functions of peroxisomes have been studied in kinds of organisms, including the filamentous fungi. In fungal pathogens, an intriguing feature is their roles in pathogenicity. It is reported that peroxisomes are required for virulence of almost all fungal pathogens, such as the plant pathogens C. orbiculare and M. oryzae, the insect pathogen M. robertsii, and the human pathogens C. albicans, A. fumigatus and C. neoformans.
During infection, M. oryzae can form a specialized structure known as appressorium. Along with the formation and maturation of appressoria, the lipid stores are mobilized and utilized. The lipid stores are firstly coupled to lipolysis, resulting in triglycerides and glycerol; the latter can accumulate enormous turgor pressure. Meanwhile, the triglycerides can be adopted by the peroxisomes for subsequent fatty acids -oxidation to produce acetyl CoA and ATP. The acetyl CoA can be utilized by the glyoxylate cycle and gluconeogenesis pathway for glucan and chitin biosynthesis; they can also be utilized for melanin biosynthesis. Therefore, the peroxisomes mediated cellular processes and metabolisms can provide key factors, such as melanin and cell wall integrity, and play key roles during M. oryzae infection [fig_ref] Figure 1: Life cycle and functions of the peroxisomes in M [/fig_ref]. Detailed peroxisome-related roles during M. oryzae infection will be described below.
## Peroxisome biogenesis and pathogenesis.
Defects in peroxisomal biogenesis will lead to severe disorder of peroxisomal metabolisms, including fatty acids -oxidation, glyoxylate cycle, melanin, and cell wall biosynthesis. Consequently, the fungal development and pathogenicity will be severely affected. During germination and appressorial development, the expression of MoPEX19 was evidently elevated. Deletion of this gene will lead to deficiency in glyoxylate cycle and severe defects in development and complete loss of pathogenicity [bib_ref] MoPex19, which is essential for maintenance of peroxisomal structure and woronin bodies,..., Li [/bib_ref]. The Δmopex7 mutant is reduced in utilization of short-chain fatty acids and reduced its capacity in conidiation [bib_ref] The PEX7-mediated peroxisomal import system is required for fungal development and pathogenicity..., Goh [/bib_ref]. The MoPEX5 seems to play more important role than MoPEX7, because the Δmopex5 mutant exhibits more severe defects than Δmopex7, such as failure to utilize some fatty acids, generation of less turgor, and more sensitivity to H 2 O 2 pressure. Moreover, distinct defects in developments are also detected in Δmopex5. This phenomenon demonstrated that the PEX5-mediated PTS1 peroxisomal import pathway could be more important than the PEX7-mediated PTS2 peroxisomal import pathway [bib_ref] PTS1 peroxisomal import pathway plays shared and distinct roles to PTS2 pathway..., Wang [/bib_ref]. However, both of the Δmopex5 and Δmopex7 mutants lose their pathogenicity. MoPEX6 is also required for long-chain fatty acids utilization and is essential for pathogenicity. The Δmopex6 mutant forms nonmelanized appressoria; as a result, it cannot form the penetration peg and infection hyphae. Additionally, mycelia of Δmopex6 are more sensitive to Calcofluor White, suggesting the cell wall of the mutant is defect [bib_ref] Host invasion during rice-blast disease requires carnitine-dependent transport of peroxisomal acetyl-CoA, Ramos-Pamplona [/bib_ref].
## Peroxisome fission and pathogenesis.
Block in peroxisome proliferation can result in failure to increase peroxisome number and impact the peroxisomal metabolism. In M. oryzae, deletion of MoPEX11A and PEF1 can both severely reduce the fatty acids utilization and virulence capacity [bib_ref] One of three Pex11 family members is required for peroxisomal proliferation and..., Wang [/bib_ref]. However, in contrast to totally loss of virulence in Δmopex5, Δmopex6, Δmopex7, or Δmopex19, the reduction of the virulence in Δmopex11A and Δpef1 is evidently slighter [bib_ref] One of three Pex11 family members is required for peroxisomal proliferation and..., Wang [/bib_ref] [bib_ref] Peroxisomal fission is induced during appressorium formation and is required for full..., Chen [/bib_ref]. Other phenotypic defects, such as the melanin layer formation, turgor generation, cell wall integrity, and ROS tolerance, are also slighter in Δmopex11A and Δpef1 [bib_ref] One of three Pex11 family members is required for peroxisomal proliferation and..., Wang [/bib_ref] [bib_ref] Peroxisomal fission is induced during appressorium formation and is required for full..., Chen [/bib_ref]. These indicate that the de novo formation, another way for peroxisome proliferation, can still function or compensate the defects of the peroxisome fission in the Δmopex11A and Δpef1 mutants.
## Pexophagy and pathogenesis.
In C. orbiculare, the Atg26mediated pexophagy has been proved to be essential for pathogenicity, by rapidly removing redundant peroxisomes after appressoria maturation [bib_ref] Atg26-mediated pexophagy is required for host invasion by the plant pathogenic fungus..., Asakura [/bib_ref]. The recycling of cellular components required for invasive growth could be the primary cause. However, it seems that the pexophagy process could be dispensable for pathogenicity in M oryzae. It has been proved that the Magnaporthe MoATG26 gene is not involved in pexophagy and is dispensable for virulence [bib_ref] The role of Snx41-based pexophagy in Magnaporthe development, Deng [/bib_ref]. Another gene, MoSNX41, obtains the ability to regulate pexophagy in M. oryzae, and it plays important roles in pathogenesis. However, its pathogenicity-related function is not relevant to roles in pexophagy, because the yeast ScSnx42 (homolog of MoSnx41) can restore the pexophagy deficiency of Δmosnx41, but cannot recover the defects of pathogenesis [bib_ref] The role of Snx41-based pexophagy in Magnaporthe development, Deng [/bib_ref]. The function of MoSnx41 in conidiation and pathogenesis could be related to Snx41-dependent retrieval trafficking pathway, which may function in gamma-glutamyl cycle and GSH antioxidant production.
## Woronin body and pathogenesis.
Failing to form woronin body would result in multiple phenotypic defects in M. oryzae. The Δhex1 mutant is normal in mycelial growth, conidiation, and mating processes, but it forms abnormal appressoria, delayed in host penetration and severely blocked in infection hyphae growth [bib_ref] Woronin body function in Magnaporthe grisea is essential for efficient pathogenesis and..., Soundararajan [/bib_ref]. As a result, Δhex1 is severely reduced in virulence. Besides, lack of HEX1 will also result in failure to survive in nitrogen starvation condition, which could explain why the mutant cannot survive in host cells with kinds of cellular damage. The peroxisome -oxidation is not affected in Δhex1, indicating the function of woronin body is distinct from the peroxisomes, although it is derived from the latter [bib_ref] Woronin body function in Magnaporthe grisea is essential for efficient pathogenesis and..., Soundararajan [/bib_ref].
## Fatty acid -oxidation and pathogenesis.
Peroxisomaloxidation is one of the chief catabolic processes during fungal infection, which can produce acetyl CoA and energy, as well as glycerol. The glycerol is used to form appressoria turgor, The peroxisomes are synthesized from the ER and then mature through peroxisome biogenesis process. During the fission inducing condition, the matured peroxisomes can be elongated, and then the daughter peroxisomes can be produced by the fission process, the newly formed peroxisomes will mature through the biogenesis process, and the matured peroxisomes can be elongated again for another fission cycle. When the fission inducing condition is removed, the redundant peroxisomes can be eliminated through pexophagy process. (b) Function of the peroxisomes during fungal infection. When the conidia attach the host surface, the cAMP/PKA signaling pathway is activated, leading to mobilization of the lipid stores by lipolysis, and produces the fatty acids and glycerol. The glycerol will be used to generate turgor. The fatty acids will be taken by the peroxisomes for -oxidation, which can produce mass of acetyl CoA. Acetyl CoA can be used by the glyoxylate cycle and gluconeogenesis to produce glucose for infection or be used to synthesize melanin and cell wall components. Together with the glycerol generated turgor, these products can help the fungus to penetrate the cuticle and colonize in the host cells. and the acetyl CoA can be used by the glyoxylate cycle to produce nutrient; it also can be used to synthesize melanin and cell wall contents, or other purposes. All of the mentioned products are critical for fungal infection. In M. oryzae, the Mfp1 protein involving in peroxisomal -oxidation is proved to play important roles in fatty acid metabolism and pathogenesis [bib_ref] Functional analysis of lipid metabolism in Magnaporthe grisea reveals a requirement for..., Wang [/bib_ref]. Defects in peroxisome biogenesis will severely block fatty acids -oxidation, and all peroxisome biogenesis and peroxisome fission-related genes are important for fatty acids -oxidation. For example, the PEX6 disruption mutant is defect in olive oil utilization, cell wall integrity, and appressorial melanization and is lost in penetration capacity [bib_ref] Host invasion during rice-blast disease requires carnitine-dependent transport of peroxisomal acetyl-CoA, Ramos-Pamplona [/bib_ref]. Block in acetyl CoA transportation will lead to similar defect in those mentioned cellular processes. The Δpth2 mutant produces less melanin than the wild type and fails to penetrate the host cells and thus is essential for pathogenesis [bib_ref] Peroxisomal carnitine acetyl transferase is required for elaboration of penetration hyphae during..., Bhambra [/bib_ref]. It cannot grow on lipid substrates. MoCRC1 is also required for olive oil utilization. The Δmocrc1 deletion mutant is severely reduced in penetration and invasive growth. The -oxidation can also occur in mitochondria [bib_ref] A carnitine-acylcarnitine carrier protein, MoCrc1, is essential for pathogenicity in Magnaporthe oryzae, Yang [/bib_ref]. However, it seems like that the peroxisomal -oxidation is important for appressoria-mediated penetration, while the mitochondrial -oxidation functions in conidial viability and keeping redox homeostasis during host colonization. 4.6. Glyoxylate Cycle and Pathogenesis. The peroxisomaloxidation produced acetyl CoA should be used by the glyoxylate cycle to provide a mechanism for glucose generation. Glyoxylate cycle enzymes, such as Icl1, are required for full virulence of M. oryzae. The expression of ICL1 is significantly elevated during conidial germination, appressorium formation, and penetration peg formation stages. Correspondingly, the Δicl1 mutant is delayed in conidial germination and appressorium formation and retards in cuticle penetration [bib_ref] The glyoxylate cycle is required for temporal regulation of virulence by the..., Wang [/bib_ref].
## Redox homeostasis and pathogenesis.
Accompanied with the degradation of fatty acids, redox homeostasis will be broken, where it is harmful to the fungi and must be rebalanced quickly. Failure in removing redundant oxides may lead to reducing of infection. For example, acetyl-CoA formation resulted in mass NADH, which can be eliminated by reoxidating it to NAD + . This reaction is catalyzed by peroxisomal lactate dehydrogenase and needs pyruvate. The pyruvate is generated by the alanine: glyoxylate aminotransferase 1 (Agt1) in M. oryzae. The Δagt1 mutant fails to penetration via appressoria and totally lost its pathogenicity [bib_ref] Peroxisomal alanine: glyoxylate aminotransferase AGT1 is indispensable for appressorium function of the..., Bhadauria [/bib_ref].
# Conclusions and perspective
As a model plant pathogen, the rice blast fungus M. oryzae gains more attention on role of peroxisomes than other pathogens. Considerable progress has been made for us to understand life cycle and functions of the peroxisomes in the filamentous fungi. Knowledge gained from past studies will provide comprehensive understanding in the peroxisomes and may lead to develop novel targets for new drugs against pathogenic fungi. The mechanistic details the peroxisome life cycle and functions are developing rapidly, but how these processes can be well tuned according to the developmental stages and environmental conditions is largely unknown. In the future, efforts should be done to elucidate these regulatory mechanisms.
A large number of PEX genes can be found in the genome of M. oryzae; the precise roles of these should be further characterized in the future. Another challenge is to reveal the mechanism of de novo synthesis and uncover its roles during appressorium formation. The connections between the peroxisomes and other cellular processes or structures should also be addressed. Genome-wide screening of peroxisome-related genes and global analysis of the PEX genes can help us to systematically investigate functions and mechanisms of the peroxisomes. The omics approaches can help us to establish the peroxisomal regulatory networks.
[fig] Figure 1: Life cycle and functions of the peroxisomes in M. oryzae. (a) Model of life cycle of the peroxisomes. [/fig]
[table] Table 1: Peroxisome-related genes identified in M. oryzae. [/table]
|
Influence of macrophage polarization on the effectiveness of surgical therapy of peri-implantitis
Purpose: To evaluate the influence of macrophage expression and polarization on the effectiveness of surgical therapy of peri-implantitis over a 6 month follow-up.Methods: A total of fourteen patients (n = 14 implants) diagnosed with peri-implantitis underwent access flap surgery, granulation tissue removal, implantoplasty, and augmentation at intra-bony components using a natural derived bone mineral and application of a native collagen membrane during a standardized surgical procedure. Granulation tissue biopsies were prepared for immunohistochemical characterization and macrophage polarization assessment. M1 and M2 phenotype expression was identified and quantified through immunohistochemical markers and histomorphometrical analyses. Clinical evaluation and data collection were performed initially and after a healing period of 6 months. Statistical analyses were performed to associate infiltrated area, macrophage, and M1/M2 phenotype influence on peri-implant tissue healing parameters after a 6-month follow-up.Results:Mean infiltrated compartment (ICT) values occupied a total percentage of 70.3% ± 13.0 in the analyzed granulation tissue biopsies. Macrophages occupied a mean area of 15.3% ± 7.0. M1 and M2 phenotypes were present in 7.1 ± 4.1% and 5.5 ± 3.7%, respectively. No statistically significant difference was observed between M1 and M2% expression (p = 0.16). The mean M1/ M2 ratio amounted to 1.5 ± 0.8. Surgical therapy was associated with statistically significant reductions in mean bleeding on probing (BOP), probing depth (PD) and suppuration (SUPP) scores at 6 months (p < 0.05). Linear regression analyses revealed a significant correlation between macrophage expression (CD68%) and changes in PD scores and M1 (%) expression and changes in mucosal recession (MR) scores at 6 months.Conclusions:The present data suggest that macrophages might influence peri-implant tissue healing mechanisms following surgical therapy of peri-implantitis over a short-term period. Particularly, changes in PD and MR scores were statistically significantly associated with macrophage expression and phenotype.
# Background
Emerging evidence indicates that macrophages may play a central role in the pathogenesis of peri-implant diseases. It is well documented that macrophages can be triggered by certain molecules to polarize into different phenotypes and become protagonists of either disease progression or resolution. M1 and M2 phenotypes have been deeply investigated and known to be proinflammatory or anti-inflammatory, respectively. A recent histological study revealed that M1 phenotype expression was significantly higher in peri-implantitis biopsies when compared with biopsies taken at periodontitis sites. Moreover, a previous immunohistochemical investigation pointed out that M1 phenotype expression and subsequently the ratio of M1/M2 was significantly associated with the severity of peri-implantitis. While these data suggest that macrophages could possibly play a critical role in the progression of peri-implantitis, it is currently unknown how a specific M1/M2 ratio may also affect wound healing following therapy. In fact, the available evidence indicates that nonsurgical therapy has a limited efficacy, which was particularly true for advanced peri-implantitis sites.
While various surgical approaches (i.e. open flap debridement along with either resective and/ or reconstructive techniques) were proven to improve the clinical outcomes over nonsurgical treatment procedures, the reported efficacy varied considerably.
It might be hypothesized that these variations might be due to the specific pro-inflammatory environment generated by a high M1 phenotype expression.
Therefore, the present retrospective study aimed at evaluating the influence of macrophage expression and polarization on the effectiveness of a standardized combined surgical therapy of peri-implantitis.
# Materials and methods
## Study design
This retrospective study included a total of n = 14 patients who attended the Department of Oral Surgery and Implantology, Goethe University, Frankfurt, Germany (period of recruitment: May-November 2019), underwent a standardized combined surgical therapy of peri-implantitis, and were observed over a 6-month follow-up period.
The study protocol No. 92/19 was approved by the ethics committee of Goethe-University, Frankfurt-Germany, 2019 and considered the Helsinki Declaration of 1975, as revised in 2013. The following reporting considered the checklist items of the STROBE statement.
## Case definition
Peri-implantitis was defined as the combination of "bleeding on gentle probing (BOP) with/without suppuration (SUPP), probing depths (PD) ≥ 6 mm, and radiographic marginal bone loss (MBL) (i.e. "interproximal bone levels ≥ 3 mm apical of the most coronal portion of the intraosseous part of the implant")".
## Selection and enrollment of participants inclusion criteria
For patient selection, the following inclusion criteria were considered:
(1) minimum age of 18 years old (2) partially/totally edentulous patients rehabilitated with implant-supported prosthesis that were diag-Galarraga-Vinueza et al. International Journal of Implant Dentistry (2021) nosed with peri-implantitis and underwent a combined surgical therapy.
(3) presence of granulation tissue biopsies obtained during surgery (4) plaque index < 1before surgery (5) patients that were followed-up for 6 months and presented all clinical records
## Exclusion criteria
The following exclusion criteria were considered:
(1) incomplete clinical records over a 6-month followup period (2) untreated periodontal disease (3) pregnant or lactant women (4) autoimmune or/and inflammatory diseases (5) uncontrolled diabetes (HbA1c > 7) (6) corticosteroid therapy
## Clinical examination
The following clinical parameters were evaluated before and after a healing period of 6-months at each implant site using a periodontal probe (PCV12PT Hu-Friedy Inc., Chicago, IL, USA): (1) BOP evaluated as present if bleeding was evident within 30 s after probing, (2) PD as measured from the mucosal margin to the bottom of the probable pocket, (3) plaque index (PI), (4) keratinized mucosa (KM) width, (5) mucosal recession (MR) as measured from the mucosal margin to the crown margin, and SUPP, evaluated as present if evident after probing and/or peri-implant palpation. All measurements were performed at six aspects per implant: mesio-vestibular, mid-vestibular, disto-vestibular, mesio-oral, mid-oral, and disto-oral by one calibrated examiner (K.O).
## Radiological examination
Radiological peri-implant bone loss at baseline was measured in a software program (Image J, Wisconsin, USA). MBL linear measurements were done following a previously described methodology"by drawing a vertical line from the implant shoulder to the end of the defect at distal and mesial sites". The measurement scale was set by the known implant length. Radiological assessment was done by one experienced examiner (M.E.G.).
## Surgical procedure and sample collection
All patients received a pre-operative professional supragingival tooth/implant cleaning and treated according to an established and standardized protocol for combined surgical therapy of peri-implantitis. In particular, following local anesthesia (articaine, 1:200.000), buccal and lingual mucoperiostal flaps were raised to expose the defect area. Debridement and granulation tissue removal were accomplished using conventional plastic curettes (Straumann Dental Implant System; Institut Straumann AG, Basel, Switzerland). Soft tissue biopsies were collected and conserved in 4% buffered formalin for 24 h. Implantoplasty was accomplished at supracrestally and/ or bucally exposed implant surfaces using diamond burrs. The intrabony defect components were augmented using a natural bone mineral (BioOss spongiosa granules, particle size 0.25-1 mm; Geistlich, Wolhusen, Switzerland) (NBM) and covered by a native collagen membrane (BioGide; Geistlich).
## Sample process and immunohistochemistry
Granulation tissue biopsies taken from the defect compartment were analyzed according to an established methodology. Biopsies were deposited in 70% ethanol at 4 ºC, dehydrated, and subsequently embedded in paraffin. Then, serial section (4 µm thick) were cut and mounted on glass poly-D-lysine coated slides. The paraffin-embedded sections were processed for immunohistochemical analyses. In particular, primary mouse monoclonal antibody against CD68 (1:200, NCL-L, Abcam, USA), rabbit monoclonal antibody against CD80 (1:200, EPR11572, Abcam, USA) and rabbit polyclonal mannose receptor antibody against CD206 (1:300, AB64693, Abcam, USA) were applied to detect macrophages, M1 and M2 phenotypes, respectively. Negative controls were implemented by replacing the primary antibody with non-immune serum.
## Histological examination
Each specimen was entirely scanned (40 × magnification) using a digital virtual light microscopy system (Nikon, NIS, Basic Research; Nikon E200 microscope, Japan) and evaluated according to an established methodology.
In brief, the surface area (µm 2 ) of the infiltrated compartment (ICT) was delineated and the total stained (brown pigmented) surface area was divided by the ICT surface area to determine the positive cell proportions (%) of macrophages, M1 and M2 phenotypes per sample. To confirm that the stained areas for CD80 and CD206 markers corresponded to macrophage positive cells, every analyzed region was simultaneously observed for CD80 and CD68 or CD206 and CD68 markers. If the stained areas (cells) for M1 or M2 markers did not match with the stained areas for CD68 positive cells, they were not considered for the analysis.
# Statistical analysis
Statistical analysis was done using a commercially available software program (SPSS, 19.0, Chicago, IL, USA). Mean values, standard deviations, and confidence intervals were calculated at the patient level (equivalent to implant level). The Shapiro-Wilk test was used to assess data distribution. The paired t-test was applied to analyze clinical parameter changes at baseline and at 6 months and to determine the difference between M1 and M2 expression in the analyzed specimens. Linear regression analyses were executed to assess the relationship between baseline ICT (%), CD68%, M1%, M2%, as well as M1/M2 ratio and the changes in BOP, PD, MR, and KM scores at 6 months. The alpha error was set at 0.05.
# Results
This study included a total of 14 patients (10 females and 4 males) (n = 14 implants, mean age: 66.3 years; range: 53 to 78 years). Three (21.4%) patients reported a history of periodontal disease and four patients (28.5%) reported to smoke occasionally. The 14 evaluated implants were bone level and had an internal conical connection system. The mean implant function was 10.0 ± 6.2 years and the mean MBL at baseline amounted to 3.2 ± 1.3 mm. Implant site characteristics and frequencies are summarized in.
## Immunohistochemical assessment
Mean ICT values in the analyzed granulation tissue biopsies amounted to 70.3% ± 13.0. Macrophages identified with the CD68 marker occupied a mean proportioned area of 15.3% ± 7.0. M1 and M2 phenotypes were present in 7.1 ± 4.1% and 5.5 ± 3.7%, respectively. No significant difference was observed between M1 and M2% expression (p = 0.16). The mean M1/ M2 ratio amounted to 1.5 ± 0.8.
## Clinical measurements
Postoperative wound healing was commonly uneventful and no implants were lost during follow-up. Clinical parameters at baseline and after 6 months are summarized in .
All patients revealed low PI scores at baseline with further improvements of − 0.5 ± 0.47 at 6 months (p = 0.002). Surgical therapy was associated with marked and significant reductions in mean BOP (− 57.0 ± 40.0%; p = 0.0001) and PD (− 1.30 ± 1.50; p = 0.009) scores at 6 months. This was associated with a significant decrease in the frequency of SUPP positive implant sites from 35 to 0% (p = 0.012).
Minimal and non-significant changes were noted for mean MR and KM values, amounting to 0.18 ± 0.7 (p = 0.37) mm and 0.1 ± 1.6 mm (p = 0.8), respectively .
# Regression analysis
The linear regression analysis revealed a significant correlation between macrophage expression (CD68%) and changes in PD scores (R 2 = 0.4; p = 0.014), indicating that lower CD68% scores at baseline were associated with higher reductions in PD scores at 6 months.
A statistically significant correlation was also noted between M1 (%) expression and MR scores (R 2 = 0.31; p = 0.039), indicating that lower M1% scores at baseline were associated with higher increases in MR scores at 6 months.
# Discussion
The present retrospective study aimed at assessing the influence of macrophage expression and polarization on the effectiveness of surgical therapy of peri-implantitis after a 6 month follow-up period. Basically, it was observed that the combined surgical procedure was associated with statistically significant reductions in mean BOP, PD and SUPP scores at 6 months. Furthermore, the analysis revealed that macrophage expression and changes in PD scores were statistically significantly correlated, indicating that lower CD68% scores at baseline were associated with higher reductions in PD scores. Moreover, a statistically significant correlation between M1 expression and MR scores revealed that lower M1% scores at baseline were associated with higher increases in MR scores.
To the author`s best knowledge, these are the first data on a potential correlation between macrophage polarization and changes in clinical outcomes following surgical therapy of peri-implantitis. In general, there is limited evidence on how macrophages residing in peri-implantitis lesions might affect or influence disease progression or resolution. A recent investigation revealed that macrophage polarization to M1 phenotype was statistically significantly associated with disease severity. In particular, advanced sites (i.e. MBL > 50% of implant length) had a statistically significantly higher M1% expression when compared with moderate or initial sites. Accordingly, M1 phenotype which has been reported to be proinflammatory and induce osteolytic effects in several studiescould possibly be a significant factor in pathogenesis of peri-implantitis. Furthermore, another immunohistochemical analysis revealed a statistically significantly higher M1 expression in biopsies taken at peri-implantitis when compared with periodontitis sites. The latter study claimed that peri-implantitis lesions do progress faster than periodontitis lesions due to the quantity and phenotypic expression of macrophages. Unfortunately, no further evidence regarding the potential role of macrophages in the pathogenesis or resolution of peri-implantitis lesions exists, thus limiting any comparisons of the presented results with previous literature.
However, basic research on wound healing mechanisms have indicated that macrophage's plasticity and polarization are essential to achieve an effective wound repair and that essentially the ability of the M2 phenotype to trigger angiogenic responses and endothelial growth factor expression could accelerate the resolution of inflammatory lesions and subsequently tissue repair. While the present study, for obvious ethical reasons, could not consider changes in the M1 and M2% expression following therapy, it is assumed that the noted statistically significant improvements in all clinical parameters (i.e. BOP, PD, SUPP) investigated may also be linked with a resolution of the pro-inflammatory component of the associated lesions. In this context, it must be emphasized that these clinical improvements were comparable with those reported in previous clinical studies evaluating a similar combined surgical treatment procedure. However, the latter studies also indicated that a disease resolution could not be obtained in all patients investigated, as evidenced by the presence of e.g. high residual PD scores at 6 to 12 months following therapy.
This observation is also supported by the present analysis, suggesting that higher CD68% scores at baseline were associated with lower PD reductions at 6 months. One Clinical parameters (mean, SD, confidence interval and difference values) before surgical procedure and after a 6 month follow-up period (patient level) (n = 14). *p < 0.05 potential explanation for this finding may be the fact that macrophage and M1% expression was significantly associated with higher PD values, which in turn revealed lower changes following therapy. When further analyzing the present data, there was also a significant association noted between M1% expression and changes in MR scores at 6 months. It is reasonable to assume that a higher M1 expression was associated with a more intense swelling of the peri-implant soft tissues at baseline, thus leading to a more pronounced tissue remodeling and subsequently increases in MR following therapy at respective sites. In fact, recent data have indicated that the peri-implant mucosa undergoes considerable volumetric changes after combined surgical therapy, with mean thickness changes (i.e. loss) amounting to − 0.11 and − 0.28 mm at 1 and 6 months. These changes were particularly pronounced at the marginal aspect of the peri-implant mucosa. Nevertheless, the present study has certain limitations such as a highly variable population (i.e. smokers & patients with history of periodontal disease), missing control group due to ethical considerations, and a small sample size, therefore, the aforementioned associations showed possible tendencies between macrophage expression and changes in PD and MR scores. These tendencies should be further investigated in future studies.
[formula] * [/formula]
Within its limitations, the present study suggest that macrophages might influence peri-implant tissue healing mechanisms following surgical therapy of peri-implantitis over a short-term period. Particularly, changes in PD and MR scores appear to possibly be linked to macrophage expression and phenotype. |
Citation: Rajagopal, M.; Paul, A.K.; Lee, M.-T.; Joykin, A.R.; Por, C.-S.; Mahboob, T.; Salibay, C.C.; Torres, M.S.; Guiang, M.M.M.; Rahmatullah, M.; et al. Phytochemicals and Nano-Phytopharmaceuticals Use in Skin, Urogenital and Locomotor Disorders: Are We There? Plants 2022, 11, 1265. https://doi.org/10.3390/ plants11091265 Academic Editors: Konstantinos Gardikis and Ioannis Mourtzinos
# Introduction
Physicians and patients have recognized the use of herbal medicine since ancient times [bib_ref] Bioavailability enhancers of herbal origin: An overview. Asian Pac, Kesarwani [/bib_ref]. For instance, the first-ever plant-derived painkiller, morphine which belongs to the benzylisoquinoline class of alkaloid, was isolated from Papaver somniferum L. (Papaveraceae) and authorized to be used in 1827 [bib_ref] An analysis of FDA-approved drugs: Natural products and their derivatives, Patridge [/bib_ref]. Herbal medicines are well known for their better therapeutic performance as well as lesser side effects compared to modern medicines. The demand for phytochemicals and plant products has been increasing rapidly in many areas of medicine, as in the treatment of dermal, urogenital, and locomotor disorders. Advanced phytopharmaceutical research especially with novel drug delivery systems by applying nanotechnology plays an important role in troubleshooting scientific needs with the determination of the pharmacokinetics, mechanism of action, site of action, accurate dosage, improved bioavailability, and reduced toxicity of various herbal medicines [bib_ref] Nanotechnology-based drug delivery systems and herbal medicines: A review, Bonifácio [/bib_ref] [bib_ref] Nanomedicine-based approaches for improved delivery of phytotherapeutics for cancer therapy, Hafez [/bib_ref]. Several safety concerns related to biocompatibility, possible toxicity (of unknown natural compounds), and lack of enough clinical trials on medicinal plants and herbal medicines can be resolved by the implementation of nano-based drug delivery systems [bib_ref] Precision and advanced nano-phytopharmaceuticals for therapeutic applications, Lim [/bib_ref] [bib_ref] Nano based drug delivery systems: Recent developments and future prospects, Patra [/bib_ref] [bib_ref] Natural product-based nanomedicine: Recent advances and issues, Watkins [/bib_ref]. Thus, herbal medicines can be used for the treatment of a wide range of ailments, including dermal, urogenital, and locomotor disorders.
Nanoparticles are often classified as particles of less than 100 nm in diameter. They occur extensively in nature as products of photochemical, plant, and algae activity and have also been created as by-products of combustion and food cooking for thousands of years [bib_ref] Nanoscience and nanotechnologies: Opportunities and uncertainties, Dowling [/bib_ref].
There are various kinds of nanosystems available, such as niosomes, liposomes, nanostructured lipid carriers (NLCs), and nanoemulsions. Niosomes are defined as microscopic vesicles composed of non-ionic surfactants, liposomes as microscopic spherical vesicles having one or more phospholipid bilayer membrane, NLCs as novel nano-sized pharmaceutical formulations composed of solid and liquid lipids, surfactants, and co-surfactants. Nanoemulsions as nano-sized emulsions have droplet sizes between 20 and 500 nm, respectively. Nanomedicine is the application of nanoscale materials such as nanoparticles for the diagnosis, monitoring, control, prevention, and treatment of disease. The field of nanomedicine and the application of nanoparticles has been of keen interest in several industries, such as electronics, communications, cosmetics, biology, and medicine [bib_ref] A review on herbal drug loaded into pharmaceutical carrier techniques and its..., Sandhiya [/bib_ref]. In medicine, these nanoparticles have various attractive properties that have brought them to the forefront in the search for novel drug delivery systems with most advances in the utilization of nanoparticle drug delivery for the treatment of cancer with several nanotherapies being used clinically after approval by the FDA in the United States of America [bib_ref] Diverse applications of nanomedicine, Pelaz [/bib_ref] [bib_ref] Engineering precision nanoparticles for drug delivery, Mitchell [/bib_ref]. The properties exhibited by nanoparticles include a high surface-to-volume ratio, high surface energy, unique mechanical, thermal, electrical, magnetic, and optical behaviors [bib_ref] Nanochemistry and nanomedicine for nanoparticle-based diagnostics and therapy, Chen [/bib_ref].
The term "nanotechnology" is derived from a Greek word that means dwarf, which employs the concepts of engineering and manufacturing at the molecular level [bib_ref] A review on nanotized herbal drugs, Sachan [/bib_ref]. The advantages generated by the use of nanotechnology can assure the revolutionary changes in herbal medicines along with several other multidisciplinary emerging applications in chemistry and physics. The reason behind the achievements of nanotechnology in medicine includes the possibility of working at the same scale of many biological processes, cellular mechanisms, and organic molecules. For this reason, medicine has looked at nanotechnology for the ideal solution in the treatments of several diseases. Furthermore, the methodology has drawn attention toward providing treatments in a safe and effective form [bib_ref] Nanotechnology in medicine: From inception to market domination, Morigi [/bib_ref].
From the existing literature, the increasing trend in nanoformulation using phytochemicals studies has been remarkable, particularly from the last 5 years (from 2018 to date) and it is commonly investigated against cancer-related disorders. Thus, it is timely for us to write a focused review on the current situation of the application of nanoformulations with phytochemicals and herbal medicines. This review therefore focuses on the potential of herbal medicines highlighting the successful application of nanotechnology to treat some diseases, specifically dermal, urogenital, and locomotor activities [fig_ref] Figure 1: Representation of delivery of phytopharmaceutical using nanotechnology [/fig_ref]. In addition, this review aims to understand the justification and significance of using nanotechnology- derived phytochemicals or herbal formulations (i.e., nano-phytopharmaceuticals) in the three specific disorders based on locomotion, skin, and urogenital conditions [fig_ref] Figure 1: Representation of delivery of phytopharmaceutical using nanotechnology [/fig_ref]. phytochemicals studies has been remarkable, particularly from the last 5 years (from 2018 to date) and it is commonly investigated against cancer-related disorders. Thus, it is timely for us to write a focused review on the current situation of the application of nanoformulations with phytochemicals and herbal medicines. This review therefore focuses on the potential of herbal medicines highlighting the successful application of nanotechnology to treat some diseases, specifically dermal, urogenital, and locomotor activities [fig_ref] Figure 1: Representation of delivery of phytopharmaceutical using nanotechnology [/fig_ref]. In addition, this review aims to understand the justification and significance of using nanotechnology-derived phytochemicals or herbal formulations (i.e., nano-phytopharmaceuticals) in the three specific disorders based on locomotion, skin, and urogenital conditions [fig_ref] Figure 1: Representation of delivery of phytopharmaceutical using nanotechnology [/fig_ref].
# Materials and methods
Databases such as PubMed, Medline, Scielo, Thomson Reuters ISI Web of Knowledge, and Science Direct were searched, combining the following keywords: "Nanotechnology", "plant-based medicine", "herbal nanoformulations", "phytochemical-based nanoformulations", and "nano-phytopharmaceuticals". In addition, the available scientific literature within the last decade (2011-2022) was considered in this review. The scholarly search engine "Google Scholar", a search using the keywords "nanoformulation'" and "phytochemicals" showed a total of 5420 publications (without any time limits and citations). The search results also demonstrated that this research field has been growing steadily from 2018 to date, as we found 4980 articles available since 2018. Similarly, using the keywords "nanoformulation" and "herbs" produced a total of 7180 articles (without any time limit) and especially showed an uprising trend of 5630 articles between 2018 and 2022.
## Therapeutic applications of nano-phytopharmaceuticals
## Nano-phytopharmaceuticals in dermal disorders
# Materials and methods
Databases such as PubMed, Medline, Scielo, Thomson Reuters ISI Web of Knowledge, and Science Direct were searched, combining the following keywords: "Nanotechnology", "plant-based medicine", "herbal nanoformulations", "phytochemical-based nanoformulations", and "nano-phytopharmaceuticals". In addition, the available scientific literature within the last decade (2011-2022) was considered in this review. The scholarly search engine "Google Scholar", a search using the keywords "nanoformulation'" and "phytochemicals" showed a total of 5420 publications (without any time limits and citations). The search results also demonstrated that this research field has been growing steadily from 2018 to date, as we found 4980 articles available since 2018. Similarly, using the keywords "nanoformulation" and "herbs" produced a total of 7180 articles (without any time limit) and especially showed an uprising trend of 5630 articles between 2018 and 2022.
## Therapeutic applications of nano-phytopharmaceuticals
## Nano-phytopharmaceuticals in dermal disorders
Dermatological disorders are prevalent worldwide and regarded as one of the major global burdens among various diseases [bib_ref] The global burden of skin disease in 2010: An analysis of the..., Hay [/bib_ref]. Severe skin damage from burns or wounds as well as acne (i.e., often causes erythematous papulopustular lesions such as rash consisting of papules and pustules) can also lead to trauma and further psychosocial stresses besides possible pain or other aggravations caused by the disorder itself [bib_ref] Psychosocial effect of common skin diseases, Barankin [/bib_ref] [bib_ref] The psychosocial impact of acne vulgaris. Ind, Hazarika [/bib_ref]. Dermatological disorders can be atopic dermatitis, alopecia (androgenic alopecia and alopecia areata, both indicating hair loss), hirsutism (growth of excess coarse body hair usually in women in places where hair is not supposed to grow), hyperhidrosis (excessive sweating), hidradenitis suppurativa (chronic and progressive inflammatory skin condition affecting groin, buttocks, and perineal and perianal regions), vitiligo, psoriasis, and melanoma [bib_ref] A review of the most common dermatologic conditions and their debilitating psychosocial..., Mian [/bib_ref].
Most dermatological disorders affect the outermost layer of the skin (horny layer), which is typically water repellent and dense, the latter characteristic acting as an effective barrier against rapid passage of any outward items, which may be chemicals or infectious agents. Topical therapeutic agents usually contain in combination the agent and a base-formulation, which facilitates the absorption of the agent. Drugs for dermatological disorders must cross the horny layer to get to the root of skin infection to produce their therapeutic effects. A low molecular weight of the therapeutic agent (20-300 kDa) enhances penetration of the horny layer or stratum corneum [bib_ref] Human skin penetration of hyaluronic acid of different molecular weights as probed..., Essendoubi [/bib_ref] , which is further enhanced if the agent is applied as an oleaginous ointment, emulsified ointment, cream, or gel. Nanotechnology can be an important tool for the delivery of therapeutic agents for both topical and transdermal applications through engineered nanoparticles of drugs and enabling them to better reach their target sites. Various types of nanoformulations are available such as solid nanoparticles, liposomes, secosomes, transferosomes, ethosomes, niosomes, nanoemulsions (NE), nanostructured lipid carriers (NLCs), solid lipid nanoparticles (SLNPs), and flexible nanovesicles [bib_ref] Topical and cutaneous delivery using nanosystems, Roberts [/bib_ref].
Solid nanoparticles, such as zinc oxide and titanium dioxide nanoparticles (NPs) are mainly used in sunscreens to filter out UVA and UVB radiations. Studies on keratinocytes suggest that titanium dioxide nanoparticles are safer than zinc oxide, as zinc oxide NPs can generate reactive oxygen species within cells. Both NPs have been found to produce adverse effects in human keratinocytes in vitro following long-term exposure [bib_ref] Toxicological aspects of long-term treatment of keratinocytes with ZNO and TiO 2..., Kocbek [/bib_ref]. Liposomes are usually composed of cholesterol and phospholipids that show higher biocompatibility, improved solubility, and efficacy of lipophilic and amphiphilic drugs and thus facilitate the application of topical drugs [bib_ref] Liposomes in topical drug delivery, Schaeffer [/bib_ref].
Nanomaterials such as NLCs are prepared from a combination of solid lipid (SL) and liquid lipid (LL) ingredients. The use of LL in the manufacture of NLCs permits a greater drug load. The SLs include compounds such as glyceryl monostearate and glyceryl tripalmitate; the LLs include a more diverse variety of compounds such as oleic acid and squalene. Surfactants used in the preparation of NLCs include lecithin and Tween 80 [bib_ref] Nanostructured lipid carriers for delivery of chemotherapeutics: A review, Haider [/bib_ref]. Flexible or deformable nanovesicles have greater penetrability through biological barriers but thus far have seen limited use because of their physical and chemical instabilities. However, a recent study reported that flexible nanovesicles at a low density and containing 8% lactose and trehalose at a ratio of 1:4 have a spherical shape, smooth surface morphology in the lyophilized state, a whorl-like structure, high entrapment efficiency, and deformability after reconstitution; thus confirming their stability. Importantly, the secondary structure of insulin was well protected in the insulin-phospholipid complex deformable nanovesicles [bib_ref] Stabilization of deformable nanovesicles based on insulin-phospholipid complex by freeze-drying, Xu [/bib_ref] , which further confirmed their functional ability.
From the above section(s), it is apparent that nanovesicles and nanoparticles can play an important role in the delivery of drugs to target organs especially on skin. It is important because many drugs have poor aqueous solubility; thus limiting their bio-absorption. These lipophilic drugs can be encapsulated within nanovesicles as nanoparticles and then administered through suitable routes. Various nanotechnological approaches have been and still are experimented with towards a more efficacious treatment of skin disorders. The therapeutic nanoparticles comprise conventional drugs, crude extract of plants, and phytochemicals. For example, the ethanolic extract of Ocimum sanctum L. (Lamiaceae) reportedly has anti-aging properties on skin, as demonstrated by its anti-oxidant and anti-inflammatory properties, as well as its inhibitory features against hyaluronic acid and collagen fiber degradation inhibition [bib_ref] Ocimum sanctum linn. as a natural source of skin anti-ageing compounds, Chaiyana [/bib_ref]. The encapsulation of the ethanolic extract was completed in several types of nanodelivery systems, including NLCs, NEs, liposomes, and niosomes. Among the various delivery systems containing Ocimum sanctum L. (Lamiaceae) extract nanoparticles, NLC and NE were the most stable, with NLC delivering the highest amount of extract to the skin layer [bib_ref] Dermal delivery enhancement of natural anti-ageing compounds from Ocimum sanctum linn. extract..., Chaiyana [/bib_ref]. The ethosome gel was reported to deliver quercetin to treat inflammation, and amphotericin B to treat fungal infections [bib_ref] Topical anti-inflammatory potential of quercetin in lipid-based nanosystems: In vivo and in..., Caddeo [/bib_ref] [bib_ref] Nanoethosomal formulation for skin targeting of amphotericin B: An in vitro and..., Kaur [/bib_ref]. Quercetin-loaded phospholipid vesicles containing, in addition, 5% polyethylene glycol demonstrated effectiveness in amelioration of skin inflammation induced by TPA (12-O-tetradecanoylphorbol-13-acetate). The nanoethosomal formulation exhibited a 3.5-fold higher skin deposition of amphotericin B, leading to a significant increase in anti-fungal activity against Candida albicans.
Application of various forms of nanodelivery systems for the treatment of skin disorders have been reviewed by Roberts et al. [bib_ref] Topical and cutaneous delivery using nanosystems, Roberts [/bib_ref]. These include liposome, ethosome, and deformable liposome-based delivery of ketoconazole to treat dermatological fungal infections from Candida albican; the use of nanostructured lipid carrier-based gel to deliver clobetasol propionate to treat eczema; the use of solid lipid nanoparticles for delivery of artemisone and doxorubicin for the treatment of melanoma and squamous cell carcinoma, respectively. Silver nanoparticles have been used to treat scalp-based fungal infections caused by Malassezia furfur; and gold nanoparticles are used for the treatment of psoriasis. The use of tyrospheres (tyrosine-derived nanospheres) as a delivery medium for vitamin D 3 has also proved to be effective for psoriasis treatment. It appears that there is enhanced absorption of vitamin D 3 through this nano-treatment method [bib_ref] Polymeric nanospheres for topical delivery of vitamin D3, Ramezanli [/bib_ref]. In fact, as reviewed by Petit et al., the use of biodegradable nanocarriers for delivery of vitamin D 3 or other therapeutics for psoriasis treatment includes nanospheres, nanocapsules, liposomes, ethosomes, solid lipid nanoparticles, and nanostructured lipid carriers [bib_ref] Psoriasis: From pathogenesis to pharmacological and nano-technological-based therapeutics, Petit [/bib_ref].
Curcumin, which is derived from rhizomes of Curcuma longa L. (Zingiberaceae), containing nanomaterials, including lipid-based nanoparticles such as liposomes, niosomes, solid lipid nanoparticles, and nanostructured lipid carriers are used in various dermatological disorders such as psoriasis, dermatitis, bacterial, viral and fungal infections, burns, acne, vitiligo, arthritis, and skin cancer [bib_ref] Preparation of curcumin-loaded liposomes and evaluation of their skin permeation and pharmacodynamics, Chen [/bib_ref] [bib_ref] Emerging trends in topical delivery of curcumin through lipid nanocarriers: Effectiveness in..., Waghule [/bib_ref] [bib_ref] The role of medicinal and aromatic plants against obesity and arthritis: A..., Paul [/bib_ref]. Lipid-based nanoparticles (NLCs and SLNPs) of curcumin have higher biocompatibility with skin layers, can increase their penetration into this organ and thus increase their solubility, stability, and therapeutic efficiencies [bib_ref] Emerging trends in topical delivery of curcumin through lipid nanocarriers: Effectiveness in..., Waghule [/bib_ref] [fig_ref] Table 1: Role of nano-phytopharmaceutical formulations against various locomotor, skin, and urogenital disorders [/fig_ref]. NLCs and SLNPs can also increase patient compliance by maintaining delayed and regulated release and improving their pharmacological activities [bib_ref] Uv spectrophotometric method for simultaneous estimation of betamethasone valerate and tazarotene with..., Mahmood [/bib_ref] [bib_ref] Lipid nanoparticle: State of the art, new preparation methods and challenges in..., Battaglia [/bib_ref]. NPQ improved memory and cognitive ability in rats (but no effects on locomotor activity test) [bib_ref] Effect of the oral administration of nanoencapsulated quercetin on a mouse model..., Moreno [/bib_ref] [bib_ref] Chapter two-quercetin and tryptanthrin: Two broad spectrum anticancer agents for future chemotherapeutic..., Shankar [/bib_ref] Citrus fruits, onions, apples, parsley, sage, tea, and berries.
## Quercetin nanoparticles preclinical (rats)
Quercetin nanoparticles improved memory and pathological damage induced by scopolamine [bib_ref] Quercetin nanoparticles attenuates scopolamine induced spatial memory deficits and pathological damages in..., Palle [/bib_ref] [bib_ref] Flavonoids as therapeutic agents in Alzheimer's and Parkinson's diseases: A systematic review..., De Andrade Teles [/bib_ref] Berries, currants, grapes, red to purplish blue colored leafy vegetables, grains, roots, and tubers.
Anthocyanin-loaded poly (ethylene glycol)-gold nanoparticles (PEG-AuNPs)
[formula] Preclinical (mice) [/formula]
PEG-AuNPs improved amyloid-beta (Aβ 1-42 ) induced neuronal damage and neuroinflammation [bib_ref] Anthocyanidins and anthocyanins: Colored pigments as food, pharmaceutical ingredients, and the potential..., Khoo [/bib_ref] Curcuma longa L.
(Zingiberaceae) Nano-curcumin particles Preclinical (mice) Enhanced memory, motor function, contextual fear [bib_ref] Highly stabilized curcumin nanoparticles tested in an in vitro blood-brain barrier model..., Cheng [/bib_ref] Anamirta cocculus(L.)
## Wight and arn. (menispermaceae)
A. cocculus NPs in cocc 30c, in a homeopathic formulation Preclinical Improved attention and motor functions in sleep-deprived rats [bib_ref] Plantderived nanoparticle treatment with cocc 30c ameliorates attention and motor abilities in..., Zubedat [/bib_ref] Solanum tuberosum L.
(Solanaceae)
## S. tuberosum lectin nps preclinical
Helped improved drug delivery enhanced memory and motor function [bib_ref] Intranasal nanoparticles of basic fibroblast growth factor for brain delivery to treat..., Zhang [/bib_ref] Azadirachtaindica A.Juss.
Neem oil incorporated in argan-liposomes and argan-hyalurosomes by sonicating with argan oil, soy lecithin, and water
In vitro Protected skin cells by reducing oxidative stress. Rauwolfia serpentina L.
Biologically synthe-sized gold nanopar-ticles with aqueous leaf extract of R. serpentina L.
In vitro Antibacterial effects on E. coli and S. aureusA phospholipid-based nanoformulation containing neem oil, derived from Azadirachta indica A. Juss. (Meliaceae), was incorporated in argan-liposomes and argan-hyalurosomes by sonicating with argan oil and soy lecithin in the presence of water, as described by Manca and colleagues. The formulation contained vesicles of 140 nm in diameter with negative charge, which protected skin cells from oxidative stress [fig_ref] Figure 2: The uses of nanomaterials with phytochemicals of Curcuma longa and Azadirachta indica... [/fig_ref]. and colleagues. The formulation contained vesicles of 140 nm in diameter with negative charge, which protected skin cells from oxidative stress [fig_ref] Figure 2: The uses of nanomaterials with phytochemicals of Curcuma longa and Azadirachta indica... [/fig_ref]. Apart from phospholipid-based nanoformulations of neem oil, nanostructures (NSs) have started to show efficacy in healing burns caused by fire or scalding objects. NSs are single or multidimensional nanomaterial-fortified structures within the measurement range of nanometer (10 -9 m) scale. NSs are classified into two major types, namely organic NSs, which include nanoemulsions, nanogels, liposomes, and so forth, and inorganic NSs containing nanocarbons or silver, copper, or gold nanoparticles (NPs). Nanocarbons can be fullerenes, graphene, or carbon nanotubes. Organic polymeric NPs for burns may contain curcumin, chitosan, and a variety of other substances along with poly lactic-co-glycolic acid (PLGA), the latter acting as a biodegradable and biocompatible copolymer. It has been shown that full-thickness wounds treated with epidermal growth factor (EGF)-loaded PLGA-NPs gave the fastest healing with the highest level of fibroblast production [bib_ref] Preparation of bioactive and antimicrobial PLGA membranes by magainin II/EGF functionalization, Yüksel [/bib_ref]. Considering the increasing use of nanoparticles and nanodelivery systems, it comes as no surprise that 94 patents were published in the area of nanotechnology-based delivery systems as skin penetration enhancers between 2008 and 2018 [bib_ref] Recent patents concerning the use of nanotechnology-based delivery systems as skin penetration..., Medeiros-Neves [/bib_ref].
Various research studies are on-going on herbal remedies and natural products for effective and safe therapeutics, whereby advancement of novel drug delivery systems with such candidates are in basic and clinical trials. The main requirement is to develop better systems for proper delivery of such drugs at the targeted site. Nanoparticles with use of herbal medicines will significantly increase their potential for treatment of multiple chronic Apart from phospholipid-based nanoformulations of neem oil, nanostructures (NSs) have started to show efficacy in healing burns caused by fire or scalding objects. NSs are single or multidimensional nanomaterial-fortified structures within the measurement range of nanometer (10 −9 m) scale. NSs are classified into two major types, namely organic NSs, which include nanoemulsions, nanogels, liposomes, and so forth, and inorganic NSs containing nanocarbons or silver, copper, or gold nanoparticles (NPs). Nanocarbons can be fullerenes, graphene, or carbon nanotubes. Organic polymeric NPs for burns may contain curcumin, chitosan, and a variety of other substances along with poly lactic-co-glycolic acid (PLGA), the latter acting as a biodegradable and biocompatible copolymer. It has been shown that full-thickness wounds treated with epidermal growth factor (EGF)-loaded PLGA-NPs gave the fastest healing with the highest level of fibroblast production [bib_ref] Preparation of bioactive and antimicrobial PLGA membranes by magainin II/EGF functionalization, Yüksel [/bib_ref]. Considering the increasing use of nanoparticles and nanodelivery systems, it comes as no surprise that 94 patents were published in the area of nanotechnology-based delivery systems as skin penetration enhancers between 2008 and 2018 [bib_ref] Recent patents concerning the use of nanotechnology-based delivery systems as skin penetration..., Medeiros-Neves [/bib_ref].
Various research studies are on-going on herbal remedies and natural products for effective and safe therapeutics, whereby advancement of novel drug delivery systems with such candidates are in basic and clinical trials. The main requirement is to develop better systems for proper delivery of such drugs at the targeted site. Nanoparticles with use of herbal medicines will significantly increase their potential for treatment of multiple chronic diseases. A number of successful examples with evidence have been presented in the direction of nano-research. It is predicted that beneficial relevance of herbal medicine utilized with nanotechnology will potentially strengthen existing drug delivery systems. Though nano-phytomedicines may promise extraordinary opportunities in the field of drug delivery of conventional and herbal medicines for treatment of various ailments, their safety should not be neglected. The alterations in physicochemical and structural characteristics of synthesized nano-size particles with a reduced size could be responsible for a few material interactions that could lead to toxicological properties. Despite the toxicity, the benefits of applying nanoparticles and nanodelivery systems in cutaneous disorders seem to outweigh the risks and can prove to offer greater benefits in the treatment of disorders such as burns, acne, and a host of other damaging diseases with associated psychosocial problems.
## Nano-phytopharmaceuticals in urogenital disorders
The application of nanotechnology to deliver herbal molecules permits bioactive compounds for targeted site delivery [fig_ref] Figure 3: Application of nano-phytopharmaceuticals in urogenital disorders [/fig_ref]. This application is crucial for the management of menopause as the targeted delivery can minimize the side effects of the herbal product, which contains hormone-like activity. Hormone replacement therapy (HRT) is the primary management strategy for menopause. Although the benefits of using HRT (estrogen and progesterone) for the management of moderate-to-severe menopausal symptoms outweigh the risk, the non-selective delivery of the hormones may cause increased risks of cerebrovascular diseases, such as stroke [bib_ref] Menopausal hormone therapy and health outcomes during the intervention and extended poststopping..., Manson [/bib_ref]. Herbal medicines are promising alternatives for the management of menopause. Phytoestrogen is a plant-derived compound that is structurally and/or functionally similar to estrogen. Plant compounds such as soy, red clover, hop, and other botanicals contain naturally occurring phytoestrogens [bib_ref] Efficacy of phytoestrogens for menopausal symptoms: A meta-analysis and systematic review, Chen [/bib_ref]. Genistein is a primary phytoestrogen compound of soybean which is poorly soluble in an aqueous medium. Its poor aqueous solubility and low serum concentration after administration warrant the development of a novel drug delivery system [bib_ref] Role of nanoparticles for production of smart herbal drug-An overview, Mathur [/bib_ref]. Encapsulation of genistein in Fe 3 O 4carboxymethylated chitosan nanoparticles and EudragitR E cationic copolymers improves water solubility, leading to better absorption from the gastrointestinal tract [bib_ref] Role of nanoparticles for production of smart herbal drug-An overview, Mathur [/bib_ref] [bib_ref] Eudragit nanoparticles containing genistein: Formulation, development, and bioavailability assessment, Tang [/bib_ref]. A low dose of phytoestrogen is associated with the development and progression of breast cancer in vitro and in vivo [bib_ref] Phytoestrogens and prevention of breast cancer: The contentious debate, Bilal [/bib_ref]. Activation of estrogen receptors in the breast by phytoestrogen promotes the growth of breast cancer. These limitations can be overcome with the incorporation of a nanotechnology-based drug delivery system. Encapsulating phytoestrogen in nanoparticles may help delivery of the bioactive compounds to the estrogen receptors in endothelium and vascular smooth muscle specifically. The agonist effect of estrogen receptors on vascular smooth muscle helps to relieve vasomotor symptoms (hot flash, night sweat) in menopausal women. The extended-release activity of the herbal preparation can be achieved through encapsulation into nanocarriers, such as multivesicular liposomes. This approach is valuable in delivering bioactive compounds which are intended to produce long-lasting action. Genistein nanoparticle preparation has been widely used for anticancer therapy [bib_ref] Fabrication of genistein-loaded biodegradable TPGS-b-PCL nanoparticles for improved therapeutic effects in cervical..., Huang [/bib_ref]. However, its potential as a phytoestrogen to treat menopause is not yet fully elucidated.
Herbal products such as rhizome extract of wild yam (Dioscorea villosa L. (Dioscoreaceae), root extract of Dong Quai (Angelica sinensis (Oliv.) Diels (Apiaceae)), evening primrose oil (Oenothera biennis L. (Onagraceae)), dried root of Maca (Lepidium meyenii Walp. (Brassicaceae)) are commonly used among menopausal women to relieve menopausal symptoms [bib_ref] Complementary and alternative medicine for menopause, Johnson [/bib_ref]. Black cohosh (Cimicifuga racemosa L.) Nutt. (Ranunculaceae) is one of the common herbal products that has been used among indigenous people for the management of menopausal symptoms. Several mechanisms of action of black cohosh have been proposed: selective estrogen receptor modulation, serotoninergic pathway, anti-oxidation, and anti-inflammation [bib_ref] Black cohosh: Insights into its mechanism(s) of action, Ruhlen [/bib_ref]. The blood-brain barrier is a challenge for the delivery of bioactive compounds, which act centrally. Formulating black cohosh in nanoparticles may help enhance the crossing of the bioactive compound through the blood-brain barrier. This novel formulation increases the selectivity of black cohosh bioactive compounds towards the central serotoninergic pathway in the brain [bib_ref] Nanoparticles for brain drug delivery, Masserini [/bib_ref]. Herbal products such as rhizome extract of wild yam (Dioscorea villosa L.
(Dioscoreaceae), root extract of Dong Quai (Angelica sinensis (Oliv.) Diels (Apiaceae)), evening primrose oil (Oenothera biennis L. (Onagraceae)), dried root of Maca (Lepidium meyenii Walp. (Brassicaceae)) are commonly used among menopausal women to relieve menopausal symptoms [bib_ref] Complementary and alternative medicine for menopause, Johnson [/bib_ref]. Black cohosh (Cimicifuga racemosa L.) Nutt. (Ranunculaceae) is one of the common herbal products that has been used among indigenous people for the management of menopausal symptoms. Several mechanisms of action of black cohosh have been proposed: selective estrogen receptor modulation, serotoninergic pathway, antioxidation, and anti-inflammation [bib_ref] Black cohosh: Insights into its mechanism(s) of action, Ruhlen [/bib_ref]. The blood-brain barrier is a challenge for the delivery of bioactive compounds, which act centrally. Formulating black cohosh in nanoparticles may help enhance the crossing of the bioactive compound through the bloodbrain barrier. This novel formulation increases the selectivity of black cohosh bioactive compounds towards the central serotoninergic pathway in the brain [bib_ref] Nanoparticles for brain drug delivery, Masserini [/bib_ref]. Copaiba oil-resin is obtained from Copaifera L. species. It is effective against endometrial cell growth. To facilitate the delivery of Copaiba oil-resin towards endometrial derived cells, it has been formulated into nanoparticles using organically modified sodium montmorillonite derivatives as a nanocarrier [bib_ref] Development and pharmacological evaluation of in vitro nanocarriers composed of lamellar silicates..., De Almeida Borges [/bib_ref]. Reduction in the proliferation of endometriotic cells in vitro by the nanoparticles of Copaiba oil-resin suggested the promising alternative therapy for the treatment of endometriosis [bib_ref] Development and pharmacological evaluation of in vitro nanocarriers composed of lamellar silicates..., De Almeida Borges [/bib_ref] [fig_ref] Figure 3: Application of nano-phytopharmaceuticals in urogenital disorders [/fig_ref].
When it comes to male reproductive disease, erectile dysfunction is one of the common debilitating conditions which affects aging men. Current management with phosphodiesterase type 5 (PDE5) inhibitor has side effects such as headaches and decreased blood pressure. The side effects become less prominent with increased selectivity of the PDE5 inhibitors [bib_ref] Potency, selectivity, and consequences of nonselectivity of PDE inhibition, Bischoff [/bib_ref]. The current areas where nanotechnology can be applied for the management of erectile dysfunction are: (1) topical delivery of drugs for on-demand Copaiba oil-resin is obtained from Copaifera L. species. It is effective against endometrial cell growth. To facilitate the delivery of Copaiba oil-resin towards endometrial derived cells, it has been formulated into nanoparticles using organically modified sodium montmorillonite derivatives as a nanocarrier [bib_ref] Development and pharmacological evaluation of in vitro nanocarriers composed of lamellar silicates..., De Almeida Borges [/bib_ref]. Reduction in the proliferation of endometriotic cells in vitro by the nanoparticles of Copaiba oil-resin suggested the promising alternative therapy for the treatment of endometriosis [bib_ref] Development and pharmacological evaluation of in vitro nanocarriers composed of lamellar silicates..., De Almeida Borges [/bib_ref] [fig_ref] Figure 3: Application of nano-phytopharmaceuticals in urogenital disorders [/fig_ref].
When it comes to male reproductive disease, erectile dysfunction is one of the common debilitating conditions which affects aging men. Current management with phosphodiesterase type 5 (PDE5) inhibitor has side effects such as headaches and decreased blood pressure. The side effects become less prominent with increased selectivity of the PDE5 inhibitors [bib_ref] Potency, selectivity, and consequences of nonselectivity of PDE inhibition, Bischoff [/bib_ref]. The current areas where nanotechnology can be applied for the management of erectile dysfunction are: (1) topical delivery of drugs for on-demand erection, (2) injectable gel into the penis, (3) hydrogels for neuroprotection, and (4) encapsulation of drugs to increase erectile function [bib_ref] Role of nanotechnology in erectile dysfunction treatment, Wang [/bib_ref]. Topical delivery and encapsulation are feasible approaches that can be implemented to deliver herbal molecules for erectile dysfunction. Encapsulating into nanoparticles allows transdermal delivery of the bioactive agents, which improve the safety profile and minimize the first-pass metabolism [bib_ref] Nanoparticles as a novel delivery vehicle for therapeutics targeting erectile dysfunction, Han [/bib_ref]. Panax ginseng C.A. Mey (Araliaceae) is one of the most popular herbs for the treatment of erectile dysfunction [fig_ref] Table 1: Role of nano-phytopharmaceutical formulations against various locomotor, skin, and urogenital disorders [/fig_ref] [bib_ref] Advances in nanoparticle delivery system for erectile dysfunction: An updated review, Masuku [/bib_ref] [bib_ref] Evaluation of the protective effect of Panax ginseng nanoparticles against nicotine-induced reproductive..., Linjawi [/bib_ref] [bib_ref] Pharmacological potential of ginseng and its major component ginsenosides, Ratan [/bib_ref]. Ginsenoside, a steroid glycoside from Panax ginseng C.A. Mey (Araliaceae), was reported to demonstrate a direct effect on triggering an erection, which is mediated through the release of endothelial nitric oxide (NO) [bib_ref] Effect of korean red ginseng on the rabbit corpus cavernosal smooth muscle, Choi [/bib_ref]. Formulation of cream containing nanoparticles of ginsenosides is a promising approach to provide an ondemand erection effect for patients with erectile dysfunction [fig_ref] Figure 3: Application of nano-phytopharmaceuticals in urogenital disorders [/fig_ref]. The local delivery of the bioactive compound minimizes the systemic absorption of the compound into the bloodstream.
Urological disorders, especially urinary tract infections (UTIs), are common and affect over 150 million people worldwide every year [bib_ref] Urinary tract infections: Disease panorama and challenges, Stamm [/bib_ref]. According to the World Health Organi-zation (WHO), around 50% of women experience a UTI at some point in their lives, but UTIs are present in men as well [bib_ref] Urinary tract infections, Bacheller [/bib_ref]. UTIs can be uncomplicated or complicated. Uncomplicated UTIs typically affect healthy persons with no other known disorders and these are further classified into lower (e.g., cystitis) and upper UTIs (e.g., pyelonephritis) [bib_ref] Uncomplicated urinary tract infections in women, Wagenlehner [/bib_ref]. Complicated UTIs occur when the urinary tract or host defense is compromised by disorders, including urinary retention, renal failure, or immunosuppression [bib_ref] Complicated urinary tract infections, Lichtenberger [/bib_ref].
Gram-negative, Gram-positive bacteria, and some fungi are the causative agents of UTIs. Among them, the most common agent is the uropathogenic Escherichia coli (UPEC), which can be responsible for both complicated and uncomplicated UTIs. Other microorganisms causing uncomplicated UTI are Klebsiella pneumoniae, Staphylococcus saprophyticus, Enterococcus faecalis, Proteus mirabilis, Pseudomonas aeruginosa, Staphylococcus aureus, and Candida spp. [bib_ref] Host-pathogen interactions in urinary tract infection, Nielubowicz [/bib_ref] [bib_ref] Urinary tract infection syndromes: Occurrence, recurrence, bacteriology, risk factors, and disease burden, Foxman [/bib_ref]. Besides UPEC, Enterococcus spp., K. pneumoniae, Candida spp., S. aureus, P. mirabilis, and P. aeruginosa are mostly responsible for complicated UTIs [bib_ref] Complicated catheter-associated urinary tract infections due to Escherichia coli and Proteus mirabilis, Jacobsen [/bib_ref] [bib_ref] Treatment of complicated urinary tract infections with an emphasis on drug-resistant gram-negative..., Levison [/bib_ref]. Since UTIs are commonly treated with antibiotics, this can cause increases in antibiotic resistance and alterations in gut and vaginal microflora [bib_ref] Antibiotic use and microbiome function, Ferrer [/bib_ref] [bib_ref] Antibiotic resistance in urinary infection: The never-ending story, María [/bib_ref]. On the other hand, any proper vaccines against UTI-causing microorganisms are currently absent, leaving antibiotics as the only therapeutic choice. Searches are ongoing for new antibiotics or other phytochemicals from plant sources [bib_ref] In silico screening of some phytochemicals for treating urinary tract infection (UTI)..., Sharma [/bib_ref] , but thus far without any significant development in this field.
The current situation regarding UTIs and drugs is a classic example of the scientists and researchers maybe needing to adopt the policy of "less is more". This can be done through nanotechnological approaches. These approaches can be undertaken in several fields, such as application of nanotherapeutics (application of nanodrugs may result in greater efficacy and so lesser use of drugs and thus decreasing drug resistance), nanodelivery (which can not only improve drug absorption but also enable drugs to reach only target organs), nanodiagnostics (can reduce the need for invasive procedure diagnostics), and nanocarriers (for targeted delivery, sustained release, and allowing a longer time in circulation for drugs).
Polymeric nanoparticles (NPs) are emerging as suitable agents as nanotherapeutics due to their ability to accumulate onto cell membranes and then destroy bacterial cells, therefore producing an antibiotic effect. Moreover, NPs can differ from each other according to the needs and can be loaded with different drugs [bib_ref] Antimicrobial polymers: The potential replacement of existing antibiotics?, Kamaruzzaman [/bib_ref]. The antibiotic-like effects of antimicrobial polymers would depend on their chemical structures, which may be quaternary nitrogen groups, halamines, or polylysine. Various amphiphilic polymers have been evaluated against ESKAPE pathogens, which have been reviewed by Kamaruzzaman et al. [bib_ref] Antimicrobial polymers: The potential replacement of existing antibiotics?, Kamaruzzaman [/bib_ref]. A number of organic NPs have been tested against uropathogens as reviewed by Sánchez et al. [bib_ref] Nanoparticles as potential novel therapies for urinary tract infections, Sánchez [/bib_ref]. Two of the NPs containing herbal components include polyphenol 60 and curcumin nanoemulsion-based gel for intravaginal use against UPEC, and polyphenol 60 plus cranberry nanoemulsion-based gel for intravaginal applications against E. coli [bib_ref] Development of nanoemulsion based gel loaded with phytoconstituents for the treatment of..., Kaur [/bib_ref] [bib_ref] Intravaginal delivery of polyphenon 60 and curcumin nanoemulsion gel, Kaur [/bib_ref].
The use of plants for biological synthesis of NPs containing inorganic elements or compounds (green NPs) is another rapidly developing area for the use of NPs against UTI pathogens. The leaf extract of Azadirachta indica A.Juss (Meliaceae) has been used to synthesize Ag-embedded mesoporous silica nanoparticles (mSiO 2 -AgNPs) against Candida albicans [bib_ref] Silver nanoparticles embedded mesoporous SiO 2 nanosphere: An effective anticandidal agent against..., Qasim [/bib_ref]. AgNPs, synthesized with the plant Anogeissus acuminata Wall. (Combretaceae), when tested against 11 multidrug-resistant (MDR) pathogens isolated from UTI patients (Staphylococcus aureus, Enterococcus faecalis, Acinetobacter baumannii, Citrobacter freundii, Enterobacter aerogenes, Escherichia coli, Klebsiella oxytoca, Klebsiella pneumoniae, Proteus mirabilis, Proteus vulgaris, and Pseudomonas aeruginosa) demonstrated good effects against the pathogens [bib_ref] Antibacterial activity of green silver nanoparticles synthesized from Anogeissus acuminata against multidrug..., Mishra [/bib_ref]. Green copper-based NPs using leaf extract of Camellia japonica L. (Theaceae) inhibited the growth of the uropathogens Klebsiella pneumoniae and Pseudomonas aeruginosa [bib_ref] Biologically synthesized copper oxide nanoparticles enhanced intracellular damage in ciprofloxacin resistant esbl..., Rajivgandhi [/bib_ref]. Green NPs have the advantage of not only containing antimicrobial compounds or extracts, but the plant or plant part extract used to produce the NPs may have antimicrobial activity itself.
Catheters are a major cause of UTIs. It has been found that coating catheters with NPs of essential oils (Eos) such as tea tree oil or EO components such as terpinen, cineole, and eugenol can protect against Proteus mirabilis biofilm formation [bib_ref] Screening of antimicrobial and antioxidant activity of commercial Melaleuca alternifolia (tea tree)..., Markovi [/bib_ref]. Taken together, alternative medicinal systems can be expanded to incorporate the nanotechnological application of plant-based nanomaterials for providing not only new but better treatments for UTIs, as well as producing preventive techniques against the development of UTIs in the first place. Besides the green silver and copper-based NPs mentioned above, there are other examples of new developments in this field of metal-plant-based nanoparticle therapeutics. Zinc oxide nanoparticles (ZnO-NPs) synthesized from the leaves of Berberis aristata DC. (Berberidaceae) have been found effective against Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Bacillus subtilis, Bacillus cereus, and Serratia marcescens [bib_ref] Phyto-mediated synthesis of zinc oxide nanoparticles of Berberis aristata: Characterization, antioxidant activity..., Chandra [/bib_ref]. Zinc oxide nanoparticles synthesized from leaves of Passiflora caerulea L. (Passifloraceae) demonstrated inhibitory activity against several uropathogens [bib_ref] Synthesis of zinc oxide nanoparticles using plant leaf extract against urinary tract..., Santhoshkumar [/bib_ref]. Copper nanoparticles (Cu-NPs) of Cissus vitiginea L. (Vitaceae) showed efficacy against several UTI pathogens such as E. coli, Enterococcus sp., and Klelbsiella sp. [bib_ref] Green synthesis of copper nanoparticles using Cissus vitiginea and its antioxidant and..., Wu [/bib_ref]. Silver (Ag)-NPs Mimosa pudica L. (Fabaceae) alcoholic extracts and Nigella sativa L. (Ranunculaceae) seeds have excellent antibacterial effects on UPEC and Staphylococcus aureus and other UTI pathogens [bib_ref] Comparative analysis and synthesis of silver nano-particles from selected parts of Mimosa..., Yogapiya [/bib_ref] [bib_ref] Green synthesis and characteriza-tion of silver nanoparticles from Nigella sativa and its..., Ranjan [/bib_ref] [bib_ref] Urinary tract infection -A review on its prevalence and recent advances, Badiger [/bib_ref]. Noticeably, Badiger and colleagues mentioned that cranberry extract is the only supplement effective against UTIs along with antibiotics [bib_ref] Urinary tract infection -A review on its prevalence and recent advances, Badiger [/bib_ref].
Therefore, it is quite feasible that green NPs can be the UTI therapeutics of the future. Not only can they be easily administered, but despite their metallic content, they are less toxic than conventional medicines due to the extremely low concentrations involved. Moreover, they provide an alternative to antibiotic-resistant UTI pathogens. The plant coatings of the NPs can be antibacterial by themselves, so every green NP containing a metal antibacterial or other antibacterial component(s) would contain two different antibacterial components, and as such, raise their efficacy. Despite the immense promises of nanotechnology, it is still basically a research topic and less an applied subject in the field of UTI therapeutics. It is expected that this situation will change in favor of NP application to treat UTI infections sooner rather than later. Appropriate herbal extracts in green NPs will facilitate their use against multidrug-resistant pathogens [bib_ref] Antibacterial activity of herbal extracts against multi drug resistant strains of bacteria..., Khan [/bib_ref].
## Nano-phytopharmaceuticals in locomotor disorders
Nano-phytochemicals are being used widely that are not only limited to dermatological or urological disorders, but also movement or coordination disorders. Movement disorders are neurological conditions with either an excess of movement (hyperkinesias) or a paucity of voluntary and automatic movements, unrelated to weakness or spasticity (hypokinesias) [bib_ref] Opioid analgesia and opioid-induced adverse effects: A review, Paul [/bib_ref] [bib_ref] Chapter 8-Movement disorders, Warner [/bib_ref] [bib_ref] Fatigue in hypokinetic, hyperkinetic, and functional movement disorders, Vico [/bib_ref]. Long-term opioid intake can cause motor behavioral disorders like hyperkinesias and hypokinesias, and it is more prevalent in patients with chronic pain, and who need to take opioids to relieve pain [bib_ref] Profiling the effects of repetitive morphine administration on motor behavior in rats, Paul [/bib_ref] [bib_ref] Morphine dosing strategy plays a key role in the generation and duration..., Paul [/bib_ref] [bib_ref] Age-dependent antinociception and behavioral inhibition by morphine, Paul [/bib_ref]. Abnormalities in the upper motor neurons, lower motor neurons, neuroinflammation, and/or the effector muscle tissues are known to be underlying factors. In this section, we discuss a few examples of natural products that were reported to have an ameliorative effect on animal or clinical models of movement disorders and the potential applications of nanoformulations to enhance their biomedical application. It is noteworthy that the reports on nano-phytoformulations on locomotor disorders are scarce; thus, the potential applications should be further explored.
Mucuna pruriens (L.) DC. (Fabaceae), commonly known as "Mucuna" or "velvet bean" is an annual legume of the family Fabaceae. It is known to have a variety of therapeutic effects, such as anti-oxidant, anti-inflammatory, anti-epileptic, antimicrobial, and aphrodisiac effects. Among all the bioactive substances found in Mucuna pruriens (L.) DC. (Fabaceae) seed extract, L-DOPA, the precursor of dopamine, constitutes almost 5% of its total phytochemical content [bib_ref] Estimation of levodopa in the unani drug Mucuna pruriens bak. and its..., Mohapatra [/bib_ref]. This distinctive feature of Mucuna pruriens (L.) DC. (Fabaceae) seed extract was used as the complementary therapy for Parkinson's disease (PD), in animal models and clinical settings [bib_ref] Mucuna pruriens seeds in treatment of parkinson's disease: Pharmacological review, Kasture [/bib_ref]. Clinical studies have demonstrated that a single dose [bib_ref] Mucuna pruriens in parkinson's disease: A double-blind, randomized, controlled, crossover study, Cilia [/bib_ref] , but not prolonged 16-week treatment with Mucuna pruriens (L.) DC. (Fabaceae) seed extracts caused significant motor improvement in PD patients, comparable with conventional levodopa medication [bib_ref] Daily intake of Mucuna pruriens in advanced parkinson's disease: A 16-week, noninferiority,..., Cilia [/bib_ref]. The application of prolonged Mucuna pruriens (L.) DC. (Fabaceae) seed extract was hampered by the emergence of gastrointestinal side effects [bib_ref] Daily intake of Mucuna pruriens in advanced parkinson's disease: A 16-week, noninferiority,..., Cilia [/bib_ref]. Nonetheless, in developing countries conventional levodopa medication is too costly, hence the Mucuna pruriens (L.) DC. (Fabaceae) seed extract could be regarded as a potential source of natural L-DOPA particularly by implementation of nanotechnology to reduce its potential side effects [bib_ref] Mucuna pruriens for parkinson's disease: Low-cost preparation method, laboratory measures and pharmacokinetics..., Cassani [/bib_ref]. Further study on identifying the other possible bioactive compound(s) synergizing with natural L-DOPA in Mucuna pruriens (L.) DC. (Fabaceae) seed extract, without the aforementioned gastrointestinal side effects, would benefit the future pharmacological development of this natural product. Given that another clinical study demonstrated that oral administration of Mucuna pruriens (L.) DC. (Fabaceae) seed extract exerted rapid onset of action and longer efficacy in PD patients without the increase in dyskinesias, a common side effect of conventional levodopa medication, it is suspected that other bioactive substances were playing a pivotal role in the PD-relieving effect of Mucuna pruriens (L.) DC. (Fabaceae) seed extract [bib_ref] Mucuna pruriens in parkinson's disease: A double blind clinical and pharmacological study, Katzenschlager [/bib_ref]. Indeed, such observation was also replicated in nonhuman primates, suggesting another distinct mechanism of action besides the dopaminergic supplementation via L-DOPA [bib_ref] The antiparkinsonian and antidyskinetic mechanisms of Mucuna pruriens in the MPTP-treated nonhuman..., Lieu [/bib_ref]. Another study showed that L-DOPA deprived Mucuna pruriens (L.) DC. (Fabaceae) seed extract could also significantly exert a neuroprotective effect in an in vitro model of PD [bib_ref] Levodopa-reduced Mucuna pruriens seed extract shows neuroprotective effects against parkinson's disease in..., Johnson [/bib_ref]. Taking into consideration that synergism among bioactive components in Mucuna pruriens (L.) DC. (Fabaceae) seed extract is crucial for its anti-PD effect, the methanolic extract (0.2 g'kg, intraperitoneal (i.p.)) of Mucuna pruriens (L.) DC. (Fabaceae) seeds produced 1-methyl-4phenyl-1, 2, 3, 6-tetrahydropyridine-induced neurotoxicity and motor behavioral toxicity in a mouse model of PD [bib_ref] The behavioral performance tests of Mucuna pruriens gold nanoparticles in the 1-methyl..., Arulkumar [/bib_ref]. Noticeably, a single daily gold nanoparticle incorporated in the methanolic extract of Mucuna pruriens (L.) DC. (Fabaceae) supplementation (0.5-20.0 mg/kg/day i.p., for 7 days) prevented motor behavioral neurotoxicity more efficiently than the effects of the methanolic extract of Mucuna pruriens (L.) DC. (Fabaceae) [bib_ref] The behavioral performance tests of Mucuna pruriens gold nanoparticles in the 1-methyl..., Arulkumar [/bib_ref]. In this study, motor behavioral toxicity was measured using rotarod, narrow beam walking and hang test [bib_ref] The behavioral performance tests of Mucuna pruriens gold nanoparticles in the 1-methyl..., Arulkumar [/bib_ref].
Apart from Mucuna pruriens (L.) DC. (Fabaceae), Moringa oleifera Lam. (Moringaceae) that is known as "mironga'" or "drumstick tree", is a member of the Moringaceae family, widely cultivated in tropical and subtropical regions. Its leaves and fruits are commonly consumed as food as well as herbal medicine in Ayurvedic and traditional Chinese medicine, for their anti-diabetic, anti-cancer, anti-inflammatory, and anti-oxidant properties [bib_ref] The antiparkinsonian and antidyskinetic mechanisms of Mucuna pruriens in the MPTP-treated nonhuman..., Lieu [/bib_ref] [bib_ref] Moringa oleifera: A review on nutritive importance and its medicinal application, Gopalakrishnan [/bib_ref] [bib_ref] A review of properties, nutritional and pharmaceutical applications of Moringa oleifera: Integrative..., Meireles [/bib_ref]. Furthermore, literature on the effect of Moringa oleifera Lam. (Moringaceae) leaves extract in alleviating central and peripheral movement disorder is available on preclinical models [bib_ref] Studies on neuropharmacological profile of ethanol extract of Moringa oleifera leaves in..., Bakre [/bib_ref]. In rats induced with focal ischemic stroke, the stroke-associated-motor impaired condition, e.g., hypolocomotion and stereotypic behaviors, was significantly suppressed by oral administration of Moringa oleifera Lam. (Moringaceae) leaves extract for 7, 14, and 21 days. In a Parkinson's disease mimicking model, a single oral administration of Moringa oleifera Lam. (Moringaceae) leaves extract was reported to reverse the haloperidol-induced catalepsy in mice, measured by akinesia and rigidity responses [bib_ref] Cerebroprotective effect of Moringa oleifera against focal ischemic stroke induced by middle..., Kirisattayakul [/bib_ref]. In another sub-chronic model of PD in mice induced by the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine, 10-day repeated treatment with moringin, the bioactive substance isolated from Moringa oleifera Lam. (Moringaceae) leaves extract, significantly ameliorated the PD-like motor deficits and bradykinesia. In another study that used the sciatic nerve injury-induced muscle atrophy as a peripheral model of motor impairment, mice fed with Moringa oleifera Lam. (Moringaceae) leaves extract-enriched chow for 14 days showed a significant muscle mass and motor grip force restoration, as compared with the normal chow control group [bib_ref] Acute effect of ethanolic extract of Moringa oleifera on haloperidol induced catalepsy..., Joy [/bib_ref]. However, another report showed that similar doses of Moringa oleifera Lam. (Moringaceae) leaf extract, administered via the oral route, exerted CNS depressant and muscle relaxant effects [bib_ref] Cns depressant and muscle relaxant effect of ethanolic leaf extract of Moringa..., Bhattacharya [/bib_ref] , suggesting possible opposing effects among bioactive compounds of Moringa oleifera Lam. (Moringaceae) leaves extract. Nonetheless, these data warrant further research on the varied contribution of each isolated compound in Moringa oleifera Lam. (Moringaceae) leaves extract, possibly on both hyperkinesias and hypokinesias-related movement disorders. Besides its therapeutic effect in animal models, Moringa oleifera Lam. (Moringaceae) leaves extract has been reported to be used as a reducing agent in the green biosynthesis of silver nanoparticles, owing to its strong antioxidant effect [bib_ref] Biofabrication of ag nanoparticles using Moringa oleifera leaf extract and their antimicrobial..., Prasad [/bib_ref] [bib_ref] Biosynthesis of silver nanoparticles using Moringa oleifera leaf extract and its application..., Sathyavathi [/bib_ref]. Interestingly, the nanoformulation that incorporates Moringa oleifera Lam. (Moringaceae) leaves extract with silver nanoparticles showed enhanced biological activities such as antioxidant, cytotoxic, and free radical scavenging activity [bib_ref] Evaluation of the biological activity of Moringa oleifera leaves extract after incorporating..., Shousha [/bib_ref]. The effects of nano-formulations of Moringa oleifera Lam. (Moringaceae) leaf extract or its bioactive compound, moringin, can be investigated in animal models of movement disorder and the outcome of this study can provide its suitability in future clinical use. Kolaviron, a natural flavonoid obtained from the seeds of Garcinia kola Heckel (Guttiferae) that is found in Cameroon and some other African countries, is used as a traditional medicine to treat the common cold, coughs, fever, and similar diseases [bib_ref] Kolaviron via anti-inflammatory and redox regulatory mechanisms abates multi-walled carbon nanotubes-induced neurobehavioral..., Adedara [/bib_ref]. Kolaviron showed anti-inflammatory, anti-colitis, antioxidant effects and prevented genotoxicity [bib_ref] Kolaviron via anti-inflammatory and redox regulatory mechanisms abates multi-walled carbon nanotubes-induced neurobehavioral..., Adedara [/bib_ref] [bib_ref] Kolaviron, a natural antioxidant and anti-inflammatory phytochemical prevents dextran sulphate sodium-induced colitis..., Farombi [/bib_ref]. Kolaviron prevented multi-walled carbon nanotubes (MWCNTs)-induced neurotoxicity (as shown with reduced locomotor activities) in rats [bib_ref] Kolaviron via anti-inflammatory and redox regulatory mechanisms abates multi-walled carbon nanotubes-induced neurobehavioral..., Adedara [/bib_ref]. In this study, 10-week-old male Wistar rats treated with kolaviron (100 mg/kg/day, oral) over a period of 15 days prevented neurotoxicity (defined as reduced exploratory and locomotor activities, such as total distance traveled, increased horizontal and turning behavior) induced by MWCNTs (1.0 mg/kg/day, intraperitoneal injections). Additionally, kolaviron treatment showed some neuroprotective effects as shown by their histological analysis of the brain [bib_ref] Kolaviron via anti-inflammatory and redox regulatory mechanisms abates multi-walled carbon nanotubes-induced neurobehavioral..., Adedara [/bib_ref].
# Conclusions
Nano-encapsulation of herbal extracts and phytoconstituents has been reported as an outstanding strategy to overcome current challenges associated with herbal medicines, such as lower solubility, less target specificity, less bioavailability, and shelf life. A number of studies have reported the successful use of nano-phytopharmaceuticals as therapeutic agents for locomotor disorders, dermatological, and urological disorders in vivo and in vitro. Nanotechnology has been reported to improve the physicochemical properties, efficacy, and bioavailability of the herbal medicines used in locomotor, dermatological, and urological disorders. However, available data in this research field is mostly from preclinical, in vivo, and in vitro studies over short-term observations. Physicochemical characteristics of nano-phytomedicines that modify their in vivo efficacy should be validated in future studies to knock over current impediments in this research and development of rationally designed nano-phytomedicines for clinical studies. Moreover, a focus on clinical translation studies is warranted, such as pharmacokinetics and long-term toxicity, to address the important questions regarding the clinical feasibility of nano-phytomedicine in locomotor, dermatological, and urological disorders. In the future, it is expected to have meaningful development in nanoparticles-based phytomedicines as an essential aspect of human health management.
[fig] Figure 1: Representation of delivery of phytopharmaceutical using nanotechnology. The figure was made with www.biorender.com (access date: 15 March 2022). [/fig]
[fig] Figure 2: The uses of nanomaterials with phytochemicals of Curcuma longa and Azadirachta indica oil in dermatological disorders. The figure was made with www.biorender.com (access date: 15 March 2022). [/fig]
[fig] Figure 3: Application of nano-phytopharmaceuticals in urogenital disorders. The figure was made with www.biorender.com (access date: 15 March 2022). [/fig]
[fig] Author: Contributions: Conceptualization, V.N. and M.R. (Mogana Rajagopal); methodology, K.J.; C.C.S. and A.K.P.; validation, M.R. (Mogana Rajagopal), M.-T.L., C.L.L., C.-S.P., C.C.S., M.R. (Mohammed Rahmatullah), R.J., A.K.P., M.d.L.P., P.W. and V.N.; investigation, A.K.P., A.R.J., M.S.T. and M.M.M.G.; writing-original draft preparation, M.R. (Mogana Rajagopal), A.K.P., T.M., M.-T.L., C.L.L., C.-S.P., M.M.M.G., C.C.S., A.R.J., M.S.T., K.J. and M.R. (Mohammed Rahmatullah); writingreview and editing, V.N., A.K.P., M.d.L.P., M.R. (Mohammed Rahmatullah), R.J. and P.W.; supervision, V.N., P.W. and M.R. (Mohammed Rahmatullah). All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: M.d.L.P. thanks Project CICECO-Aveiro Institute of Materials, UIDB/50011/2020, UIDP/50011/2020, and LA/P/0006/2020, financed by national funds through the FCT/MEC (PIDDAC). [/fig]
[table] Table 1: Role of nano-phytopharmaceutical formulations against various locomotor, skin, and urogenital disorders. [/table]
[bib_ref] Diverse applications of nanomedicine, Pelaz [/bib_ref] |
Change sign detection with differential MDL change statistics and its applications to COVID-19 pandemic analysis
We are concerned with the issue of detecting changes and their signs from a data stream. For example, when given time series of COVID-19 cases in a region, we may raise early warning signals of an epidemic by detecting signs of changes in the data. We propose a novel methodology to address this issue. The key idea is to employ a new information-theoretic notion, which we call the differential minimum description length change statistics (D-MDL), for measuring the scores of change sign. We first give a fundamental theory for D-MDL. We then demonstrate its effectiveness using synthetic datasets. We apply it to detecting early warning signals of the COVID-19 epidemic using time series of the cases for individual countries. We empirically demonstrate that D-MDL is able to raise early warning signals of events such as significant increase/decrease of cases. Remarkably, for about 64% of the events of significant increase of cases in studied countries, our method can detect warning signals as early as nearly six days on average before the events, buying considerably long time for making responses. We further relate the warning signals to the dynamics of the basic reproduction number R0 and the timing of social distancing. The results show that our method is a promising approach to the epidemic analysis from a data science viewpoint.Motivation. We address the issue of detecting changes and their signs in a data stream. For example, when given time series of the number of COVID-19 cases in a region, we may expect to warn the beginning of an epidemic by detecting changes and their signs.Although change detection 1-3 is a classical issue, it has remained open how signs of changes can be found. In principle the degree of change at a given time point has been evaluated in terms of the discrepancy measure (e.g.. the Kullback-Leibler (KL) divergence) between probability distributions of data before and after that time point 1,4 . It is reasonable to think that the differentials of the KL divergence may be related to signs of change. This is because the first differential of the KL divergence is a velocity of change while its second differential is an acceleration of change.The problem is here that in real cases, the KL-divergence and its differentials cannot be exactly calculated since the true distribution is unknown in advance. A question lies in how we can estimate the discrepancy measure and their differentials from data when the parameter values are unknown.The purpose of this paper is to answer the above question from an information-theoretic viewpoint based on the minimum description length (MDL) principle 5 (see also studies 6,7 for its recent advances). The MDL principle gives a strategy for evaluating the goodness of a probabilistic model in terms of codelength required for encoding the data where a shorter codelength indicates a better model. We apply this principle to change detection where a shorter codelength indicates a more significant change. Along this idea, we introduce the notion called the differential MDL change statistics (D-MDL) for the measure of change signs. We theoretically and empirically justify this notion, and then apply it to the COVID-19 pandemic analysis using open datasets. investigation, K.Y. and L.X.; resources, K.Y.; data curation, L.X. and R.Y.; writing-original draft preparation, K.Y. and L.X.; writing-review and editing, K.Y., L.X. R.Y., and S.F.; visualization, L.X. and R.Y.; supervision, K.Y. and L.X.; project administration, K.Y.; funding acquisition, K.Y.
www.nature.com/scientificreports/ hood ratio, KL-divergence. However, there is no work on relating the differential information such as the velocity of the change to change sign detection.
Most of previous studies in change detection are concerned with detecting abrupt changes. In the scenario of concept drift [bib_ref] A survey on concept drift adaptation, Gama [/bib_ref] , the issues of detecting various types of changes, including incremental changes and gradual changes, have been addressed. How to find signs of changes has been addressed in the scenarios of volatility shift detection [bib_ref] Detecting volatility shift in data streams, Huang [/bib_ref] , gradual change detection [bib_ref] Detecting gradual changes from data stream using MDL change statistics, Yamanishi [/bib_ref] and clustering change detection [bib_ref] Detecting latent structure uncertainty with structural entropy, Hirai [/bib_ref] [bib_ref] Graph-based entropy for detecting explanatory signs of changes in market, Ohsawa [/bib_ref] [bib_ref] Detecting model changes and their early warning signals using MDL change statistics, Hirai [/bib_ref]. However, the notion of differential information has never been related to change sign detection.
The MDL change statistics has been proposed as a test statistics in the hypothesis testing for change detection [bib_ref] Detecting gradual changes from data stream using MDL change statistics, Yamanishi [/bib_ref] [bib_ref] Model change detection with the MDL principle, Yamanishi [/bib_ref]. It is defined as the difference between the total codelength required for encoding data for the non-change case and that for the change case at a specific time point t. A number of data compression-based change statistics similar to it have also been proposed in data mining [bib_ref] Toward parameter-free data mining, Keogh [/bib_ref] [bib_ref] Mining itemsets that compress, Vreeken [/bib_ref] [bib_ref] Detecting change in data streams, Van Leeuwen [/bib_ref]. However, any differential variation of the compression-based change statistics has never been proposed.
Significance of this paper. The significance of this paper is summarized as follows:
(1) Proposal of D-MDL and its use for change sign detection. We introduce a novel notion of D-MDL as an approximation of KL-divergence of change and its differentials. We then propose practical algorithms for on-line detection of change signs on the basis of D-MDL. [bib_ref] Inference about the change-point in a sequence of random variables, Hinkley [/bib_ref] Theoretical and empirical justification of D-MDL. We theoretically justify D-MDL in the hypothesis testing of change detection. We consider the hypothesis tests which are equivalent with D-MDL scoring. We derive upper bounds on the error probabilities for these tests to show that they converge exponentially to zero as sample size increases. The bounds on the error probabilities are used to determine a threshold for raising an alarm with D-MDL. We also empirically justify D-MDL using synthetic datasets. We demonstrate that D-MDL outperforms existing change detection methods in terms of AUC for detecting the starting point of a gradual change. (3) Applications to COVID-19 pandemic analysis. On the basis of the theoretical and empirical advantages of D-MDL, we apply it to the COVID-19 pandemic analysis. We are mainly concerned with how early we are able to detect signs of outbreaks or the contraction of the epidemic for individual countries. The results showed that for about 64% of outbreaks in studied countries, our method can detect signs as early as about 6 days on average before the outbreaks. Considering the rapid spread, 6 days can earn us considerably long time for making responses, e.g., implementing control measures [bib_ref] Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their..., Bi [/bib_ref] [bib_ref] The effect of human mobility and control measures on the COVID-19 epidemic..., Kraemer [/bib_ref] [bib_ref] Early dynamics of transmission and control of COVID-19: A mathematical modelling study, Kucharski [/bib_ref]. The earned time is especially precious in the presence of a considerably long period of the incubation of the COVID-19 [bib_ref] Period of novel coronavirus (2019-nCoV) infections among travellers from Wuhan, China, Backer [/bib_ref] [bib_ref] novel coronavirus infections with right truncation: A statistical analysis of publicly available..., Linton [/bib_ref] [bib_ref] The incubation period of coronavirus disease 2019 (COVID-19) from publicly reported confirmed..., Lauer [/bib_ref]. Moreover, we analyze relations between the change detection results and social distancing events. One of findings is that for individual countries, an average of about four changes/change signs detected before the implementation of social distancing correlates a significant decline from the peak of daily new cases by the end of April 2020.
The change analysis is a pure data science methodology, which detects changes only using statistical models without using differential equations about the time evolution. Meanwhile, SIR (Susceptible Infected Recovered) model [bib_ref] A contribution to the mathematical theory of epidemic, Kermack [/bib_ref] is a typical simulation method which predicts the time evolution of infected population with physics model-based differential equations. Although the fitness of the SIR model or its variants to COVID-19 data was argued, the complicated situation of COVID-19 due to virus mutations [bib_ref] Tracking changes in SARS-CoV-2 spike: Evidence that D614G increases infectivity of the..., Korber [/bib_ref] [bib_ref] Covid-19: New coronavirus variant is identified in UK, Wise [/bib_ref] [bib_ref] Complete map of SARS-CoV-2 RBD mutations that escape the monoclonal antibody LY-CoV555..., Starr [/bib_ref] , international interactions, highly variable responses from authorities 34 , environmental effects 35,36 etc. does not necessarily make any simulation model perfect. Therefore, the basic reproduction number R0 37 (a term in epidemiology, representing the average number of people who will contract a contagious disease from one person with that disease) estimated from the SIR model may not be precise. We empirically demonstrate that as a byproduct, the dynamics of R0 can be monitored by our methodology which only requires the information of daily new cases. The data science approach then may form a complementary relation with the simulation approach and gives new insights into epidemic analysis. The effect of social distancing in Germany has been evaluated using the framework of change point analysis together with SIR model [bib_ref] Inferring change points in the spread of COVID-19 reveals the effectiveness of..., Dehning [/bib_ref]. However, there is no work on machine learning approaches to detecting signs of outbreak for COVID-19. The software for the experiments is available at https://github.com/IbarakikenYukishi/differential-mdlchange-statistics. An online detection system is available at https://ibarakikenyukishi.github.io/d-mdl-html/ index.html The rest of this paper is organized as follows: "Methods" introduces D-MDL and gives a theory of its use in the context of change sign detection. "Result I: experiments with synthetic data" gives empirical justification of D-MDL using synthetic datasets. "Result II: applications to COVID-19 pandemic analysis" gives applications of D-MDL to the COVID-19 pandemic analysis. "Conclusion" gives concluding remarks.
# Methods
Definitions of changes and their signs. Let X be a domain, which is either discrete or continuous. Hereafter we assume that X is discrete without loss of generality. For a random variable x ∈ X , let p(x; θ) = p θ (x) be the probability mass function (or the probability density function in the continuous case) specified by a parameter θ . Supposing that θ changes over time. In the case when θ gradually changes over time, we define the signs of change as the starting point of that change.
Let us consider the discrete time t. Let θ t be the parameter value of θ at time t. Let D(p||q) denote the Kullback-Leibler (KL) divergence between two probability mass functions p and q:
[formula] t , �(2) [/formula]
t as the velocity of change and the acceleration of change of the parameter at time t, respectively. All of them quantify the signs of change. However, the parameter values are not known in advance for general cases. The problem is how we can define the degree of changes for such cases.
Differential MDL change statistics. In the case where the true parameter values are unknown, the MDL change statistics has been proposed to measure the change degree 14,18 from a given data sequence. Below we denote x a , . . . ,
[formula] x b = x b a . [/formula]
In the case of a = 1 , we may drop off a and write it as x b . When the parameter θ is unknown, we may estimate it as θ using the maximum likelihood estimation method from a given sequence x n . I.e., θ = argmax θ p(x n ; θ). Note that the maximum likelihood function p(x n ;θ) does not form a probability distribution of x n because x n p(x n ;θ) > 1 . Thus we construct a normalized maximum likelihood (NML) distribution 40 by and consider the logarithmic loss for x n relative to this distribution by which we call the NML codelength, where log means the natural logarithm and C n is called the parametric complexity defined as It is known 39 that Eq. (1) is the optimal codelength that achieves the Shtarkov's minimax regret in the case where the parameter value is unknown. It is known 40 that under some regularity condition for the model class, C n is asymptotically expanded as follows:
where I(θ) is the Fisher information matrix defined by I(θ) = lim n→∞ (1/n)E θ [−∂ 2 log p(X n ; θ)/∂θ∂θ ⊤ ] , d is the dimensionality of θ , and lim n→∞ o(1) = 0.
According to the study [bib_ref] Detecting gradual changes from data stream using MDL change statistics, Yamanishi [/bib_ref] , the MDL change statistics at time point t is defined as follows:
The MDL change statistics is the difference between that the NML codelength of a given data sequence for nonchange and that for change at time t. It is a generalization of the likelihood ratio test [bib_ref] Continuous inspection schemes, Page [/bib_ref] [bib_ref] Optimal stopping times for detecting changes in distributions, Moustakides [/bib_ref]. Therefore, by extending the change degrees � (0)
[formula] t , � (1) t , �(2) [/formula]
t , . . . to the cases where the true parameters are unknown, we may consider the following statistics:
[formula] � (α) t corresponds to � (α) t . We call � (α) [/formula]
t the α th differential MDL change statistics, which we abbreviate as the α th D-MDL ( α = 0, 1, 2, . . . ) . The 0th D-MDL is the original MDL change statistics as in the study [bib_ref] Detecting gradual changes from data stream using MDL change statistics, Yamanishi [/bib_ref].
For example, let us consider the uni-variate Gaussian distribution:
[formula] D(p||q) = x p(x) log p(x) q(x) . � (0) t def = D(p θ t ||p θ t−1 ), � (1) t def = � (0) t+1 − � (0) t = D(p θ t+1 ||p θ t ) − D(p θ t ||p θ t−1 ), � (2) t def = � (1) t − � (1) t−1 = D(p θ t+1 ||p θ t ) − 2D(p θ t ||p θ t−1 ) + D(p θ t−1 ||p θ t−2 ). p NML (x n ) def = max θ p(x n ; θ) y n max θ p(y n ; θ) = max θ p(x n ; θ) C n (1) L NML (x n ) def = − log p NML (x n ), [/formula]
(2) C n def = x n max θ p(x n ; θ).
(3) www.nature.com/scientificreports/ where x ∈ R and θ = (µ, σ ) . We assume |µ| < µ max and σ min < σ < σ max where µ max < ∞ , 0 < σ min , σ max < ∞ are hyper parameters. The 0th D-MDL at time t is calculated as where σ 0 ,σ 1 and σ 2 denote the maximum likelihood (ML) estimators of σ calculated for x n 1 , x t 1 and x n t+1 , respectively. C n is the parametric complexity, which is calculated according to the study [bib_ref] Detecting gradual changes from data stream using MDL change statistics, Yamanishi [/bib_ref] Hypothesis testing for change detection. The 0th D-MDL test. We give rationale of D-MDL using the framework of hypothesis testing for change detection. First suppose that a change point exists at t or not. Let us consider the following hypothesis testing framework: The null hypothesis H 0 is that there is no change point while the alternative hypothesis H 1 is that t is an only change point.
[formula] C n = d 2 log n 2π + log |I(θ)|dθ + o(1),(4)� (0) t def = 1 n {L NML (x n 1 ) − (L NML (x t 1 ) + L NML (x n t+1 )). (5) � (1) t def = � (0) t+1 − � (0) t , (6) � (2) t def = � (1) t − � (1) t−1 = � (0) t+1 − 2� (0) t + � (0) t−1 , · · · (7) p(x; θ) = 1 √ 2πσ exp − (x − µ) 22σ [/formula]
where θ 0 , θ 1 , θ 2 (θ 1 = θ 2 ) are all unknown.
With the MDL principle, the test statistics is given as follows: For an accuracy parameter ǫ > 0,
where � (0) t is the 0th D-MDL as in equation (4). H 1 is accepted if h 0 (x n ; t, ǫ) > 0 , otherwise H 0 is accepted. We call this test the 0th D-MDL test.
We define Type I error probability as the probability that the test accepts H 1 although H 0 is true (false alarm rate) while Type II error probability as the one that the test accepts H 0 although H 1 is true (overlooking rate). The following theorem justifies the use of the 0th D-MDL in change detection.
## Theorem 2.1 14 type i and ii error probabilities for the 0th d-mdl test are upper bounded as follows:
where C n is the parametric complexity as in Eq. (2) and d(p, q) in Eq. [bib_ref] A survey on concept drift adaptation, Gama [/bib_ref] is the Bhattcharyya distance between p and q.
This theorem shows that Type I and II error probabilities in Eqs. [bib_ref] On-line inference for multiple change point problem, Fearnhead [/bib_ref] and (11) converge to zero exponentially in n as n increases for some appropriate ǫ when d(p NML , p θ 1 * θ 2 ) is large. We see that the error exponents are governed by the parametric complexity (2) of the model class. In this sense the 0th MDL test is effective in change point detection.
The 1st D-MDL test. Next we give a hypothesis testing setting equivalent with the 1st D-MDL scoring. We consider the situation where a change point exists at time either t or t + 1 . Let us consider the following hypotheses: The null hypothesis H 0 is that the change point is t while the alternative one H 1 is that it is t + 1.
[formula] where θ 0 , θ 1 , θ 2 , θ 3 (θ 0 = θ 1 , θ 2 = θ 3 ) are all unknown. [/formula]
We consider the following test statistics: For an accuracy parameter ǫ > 0,
[formula] (8) � (0) t = 1 n logσ n 0 σ t 1σ n−t 2 + 1 n log C n C t C n−t , log C n = 1 2 log 16|µ max | πσ 2 min + n 2 log n 2e − log Ŵ n − 1 2 . H 0 : x n 1 ∼ p(X n ; θ 0 ), H 1 : x t 1 ∼ p(X t ; θ 1 ), x n t+1 ∼ p(X n−t ; θ 2 ), (9) h 0 (x n ; t, ǫ) def = 1 n {L NML (x n 1 ) − (L NML (x t 1 ) + L NML (x n t+1 ))} − ǫ = � (0) t − ǫ,(10)Type I error prob. < exp −n ǫ − log C n n ,(11)Type II error prob. ≤ exp −n d(p NML , p θ 1 * θ 2 ) − log C t C n−t 2n − ǫ 2 ,(12)d(p, q) def = − 1 n log x n (p(x n )q(x n )) 1 2 , p NML (x n ) = max θ p(x n ; θ) y n max θ p(y n ; θ) , p θ 1 * θ 2 (x n ) = p(x t 1 ; θ 1 )p(x n t+1 ; θ 2 ) [/formula]
. www.nature.com/scientificreports/ which compares the NML codelength for H 0 with that for H 1 . We accept H 1 if h 1 (x n ; t, ǫ) > 0 , otherwise we accept H 0 . We call this test the 1st D-MDL test. We easily see where � [bib_ref] Continuous inspection schemes, Page [/bib_ref] t is the 1st D-MDL. This implies that the 1st D-MDL test is equivalent with testing whether the 1st D-MDL is larger than ǫ or not. Hence this test is also equivalent with comparison of the degree of change at time t + 1 and that at time t. Intuitively, if the degree of change increases significantly as time goes by, then H 1 is accepted. Thus the basic performance of discrimination of the 1st D-MDL can be reduced to that of the 1st D-MDL test.
[formula] H 0 : x t 1 ∼ p(X t ; θ 0 ), x n t+1 ∼ p(X n−t ; θ 1 ), H 1 : x t+1 1 ∼ p(X t+1 ; θ 2 ), x n t+2 ∼ p(X n−t−1 ; θ 3 ),(13)h 1 (x n ; t, ǫ) def = 1 n L NML (x t 1 ) + L NML (x n t+1 ) − L NML (x t+1 1 ) + L NML (x n t+2 ) − ǫ, [/formula]
The following theorem shows the basic property of the 1st D-MDL test.
## Theorem 2.2 type i and ii error probabilities for the 1st d-mdl test are upper bounded as follows:
where C n is the parametric complexity as in Eq.
(2), d is the Bhattacharyya distance as in Eq. [bib_ref] A survey on concept drift adaptation, Gama [/bib_ref] and
(The proof is in Sec. 1 of the supplementary information.) This theorem shows that for some appropriate ǫ , Type I and II error probabilities in Eqs. [bib_ref] Detecting latent structure uncertainty with structural entropy, Hirai [/bib_ref] and (16) converge to zero exponentially in n as n increases where the error exponents are related to the parametric complexities for the hypotheses as well as the Bhattacharyya distance between the null and alternative hypotheses. In this sense the 1st MDL test is effective. Type I error probability in Eq. (15) will be used for determining a threshold of the alarm.
The 2nd D-MDL test. Next we consider a hypothesis testing setting equivalent with the 2nd D-MDL scoring. Suppose that change points exist either at time t or at t − 1 and t + 1.
where θ 0 , θ 1 , θ 2 , θ 3 , θ 4 , (θ 0 = θ 1 , θ 2 = θ 3 = θ 4 ) are all unknown. H 0 is the hypothesis that a change happens at time t while H 1 is the hypothesis that two changes happen at time t − 1 and t. In H 0 , t is a single change point while in H 1 , t is a transition point between two close change points. Thus this hypothesis testing evaluates whether time t is a change point or a transition point of close changes.
The test statistics is: For an accuracy parameter ǫ > 0,
[formula] We accept H 1 if h 2 (x n ; t, ǫ) > 0 , otherwise accept H 0 . We call this test the 2nd MDL test. U n d e r t h e a s s u m p t i o n (1/n)L NML (x t+1 1 ) ≈ (1/n)(L NML (x t−1 1 ) + L NML (x t x t+1 )) a n d (1/n)L NML (x n t ) ≈ (1/n)(L NML (x t x t+1 ) + L NML (x n t+2 )), we have [/formula]
This implies that the 2nd D-MDL test is equivalent with testing whether the 2nd D-MDL is larger than 2ǫ or not. Intuitively, if the degree of two-step change exceeds significantly that of one-step change as time increases, then H 1 is accepted. Thus the basic performance of discrimination of the 2nd D-MDL can be reduced to that of the 2nd D-MDL test.
The following theorem shows the basic property of the 2nd D-MDL test. www.nature.com/scientificreports/ where C n is the parametric complexity as in Eq. (2), d is the Bhattacharyya distance as in Eq. [bib_ref] A survey on concept drift adaptation, Gama [/bib_ref] and
## Theorem 2.3 type i and ii error probabilities for the 2nd d-mdl test are upper bounded as follows:
[formula] (14) h 1 (x n ; t, ǫ) = � (1) t − ǫ = � (0) t+1 − � (0) t − ǫ, (15) Type I error prob. < exp −n ǫ − log C t C n−t n ,(16)Type II error prob. ≤ exp −n d(p NML (t) , p θ 2 * θ 3 ) − log C t+1 C n−t−1 2n − ǫ 2 , p NML (t) (x n ) = max θ p(x t 1 ; θ) y t 1 max θ p(y t 1 ; θ) · max θ p(x n t+1 ; θ) y n t+1 max θ p(y n t+1 ; θ) , p θ 2 * θ 3 (x n ) =p(x t+1 1 ; θ 2 )p(x n t+2 ; θ 3 ). H 0 : x t 1 ∼ p(X t ; θ 0 ), x n t+1 ∼ p(X n−t ; θ 1 ), H 1 : x t−1 1 ∼ p(X t−1 ; θ 2 ), x t x t+1 ∼ p(X 2 ; θ 3 ), x n t+2 ∼ p(X n−t−1 ; θ 4 ).(17)h 2 (x n ; t, ǫ) def = 1 n L NML (x t 1 ) + L NML (x n t+1 ) − L NML (x t−1 1 ) + L NML (x t x t+1 ) + L NML (x n t+2 ) − ǫ. (18) � (2) t ≈ 2h 2 (x n ; t, ǫ) + 2ǫ. (19) Type I error prob. < exp −n ǫ − log C t C n−t n ,(20)Type II error prob. ≤ exp −n d(p NML(t) , p θ 2 * θ 3 * θ 4 ) − log C t−1 C 2 C n−t+1 2n − ǫ 2 , [/formula]
This theorem can be proven similarly with Theorem 2.2 Type I probability in Eq. [bib_ref] Toward parameter-free data mining, Keogh [/bib_ref] will be used for determining the threshold in "Sequential change sign detection with D-MDL".
Sequential change sign detection with D-MDL. In previous sections, we considered how to measure the change sign score at a specific time point t. In order to detect change signs sequentially for the case where there exist multiple change points, we can conduct sequential change sign detection using D-MDL in a similar manner with the study [bib_ref] Detecting gradual changes from data stream using MDL change statistics, Yamanishi [/bib_ref]. We give two variants of the sequential algorithms. One is the sequential D-MDL algorithm with fixed windowing while the other is that with adaptive windowing. In the former, we prepare a local window of fixed size to calculate D-MDL at the center of the window. We then slide the window to obtain a sequence of D-MDL change scores as with the study 14 (see also the study 42 for local windowing). We raise an alarm when the score exceeds the predetermined threshold β . The algorithm is summarized as follows:
In the study [bib_ref] Detecting changes in streaming data with information-theoretic windowing, Kaneko [/bib_ref] , the sequential algorithm with adaptive windowing (SCAW) was proposed by combining the 0th D-MDL with ADWIN algorithm 9 (see also the study [bib_ref] Optimal detection of change points with a linear computational cost, Killick [/bib_ref] for adaptive windowing) where the window grows until the maximum of the MDL change statistics in the window exceeds a threshold. Once it exceeds the threshold, we drop the data earlier than the time point where the maximum is achieved and the window shrinks. Then the process restarts. It outputs the size of window whenever a change point is detected.
According to the study [bib_ref] Detecting changes in streaming data with information-theoretic windowing, Kaneko [/bib_ref] , for the window size w, the threshold ǫ w for w� (0) is set so that the total number of false alarms is finite. This is set as follows: For some parameter δ > 0 , when the parameter is d-dimensional, Hierarchical sequential D-MDL algorithm. Practically, we combine the algorithm with adaptive windowing for the 0th D-MDL and the algorithms with fixed windowing for the 1st and 2nd D-MDL. We call this algorithm the hierarchical sequential D-MDL algorithm. It is designed as follows. We first output not only the 0th D-MDL score but also a window size with the 0th D-MDL with adaptive windowing and raise an alarm when the window shrinks, i.e., Eq. (21) is satisfied for some time in the window. We then output the 1st and 2nd D-MDL scores using the window produced by the 0th D-MDL and raise alarms when for some time in the window, the 1st or 2nd D-MDL exceeds the threshold so as to expect the 1st and 2nd D-MDL to detect change signs before the window shrinkage. Note that the window shrinks only with the 0th D-MDL, but neither with the 1st nor 2nd D-MDL.
In this algorithm, for the window size w, the threshold for the 1st D-MDL score w� [bib_ref] Continuous inspection schemes, Page [/bib_ref] t is determined so that Type I error probability in Eq. (15) is less than the confidence parameter δ 1 . That is, from Eqs. [bib_ref] Detecting latent structure uncertainty with structural entropy, Hirai [/bib_ref] and (3), letting the threshold be ǫ The threshold ǫ [bib_ref] Inference about the change-point in a sequence of random variables, Hinkley [/bib_ref] w for the 2nd D-MDL score w� [bib_ref] Inference about the change-point in a sequence of random variables, Hinkley [/bib_ref] t can also be derived similarly with the 1st one. Note that by Eq. (18), the threshold is 2 times the accuracy parameter for the hypothesis testing. Letting δ 2 be the confidence parameter, we have www.nature.com/scientificreports/ We employ the righthand side of Eq. (23) as the threshold of an alert of the 2nd D-MDL. In practice, δ 1 and δ 2 are estimated from data (see "Data modeling"). The hierarchical sequential D-MDL algorithm is summarized as follows:
[formula] p NML (t) (x n ) = max θ p(x t 1 ; θ) y t 1 max θ p(y t 1 ; θ) · max θ p(x n t+1 ; θ) y n t+1 max θ p(y n t+1 ; θ) , p θ 2 * θ 3 * θ 4 (x n ) =p(x t−1 1 ; θ 2 )p(x t x t+1 ; θ 3 )p(x n t+2 ; θ 4 ). (21) ǫ w = (2 + d/2 + δ) log w + log(1/δ). Type I prob. < exp(−ǫ (1) w + log C t C n−t ) ≈ exp(−ǫ (1) w + (d/2) log(w/2) × 2) ≤ δ 1 .(22) [/formula]
## Result i: experiments with synthetic data
Datasets. To evaluate how well D-MDL performs for abrupt/gradual change detection, we consider two cases; multiple mean change detection and multiple variance one.
In the case of multiple mean change detection, we constructed synthetic datasets as follows: Each datum was independently drawn from the Gaussian distribution N (µ t , 1) where the mean µ t abruptly/gradually changed over time according to the following rule: In the case of abrupt changes, where H(x) is the Heaviside step function that takes 1 if x > 0 otherwise 0. In the case of gradual changes, H is replaced with the following continuous function:
In the case of multiple variance change detection, each datum was independently drawn from the Gaussian distribution N (0, σ 2 t ) where the variance σ 2 t abruptly/gradually changed over time according to the following rule: In the case of abrupt changes,
In the case of gradual changes, H is replaced with S as with the multiple mean changes.
We define a sign of a gradual change as the starting point of that change. In all the datasets, change points for abrupt changes and change signs for gradual changes were set at nine points: t = 1000 , 2000, . . . , 9000.
Evaluation metric. For any change detection algorithm that outputs change scores for all time points, letting β be a threshold parameter, we convert change-point scores {s t } into binary alarms {a t } as follows: (23) ǫ (2) w ≥ 2(d log(w/2) + log(1/δ 2 )). www.nature.com/scientificreports/ By varying β , we evaluate the change detection algorithms in terms of benefit and false alarm rate defined as follows: Let T be a maximum tolerant delay of change detection. When the change truly starts from t * , we define benefit of an alarm at time t as where t * is a change point for abrupt change, while it is a sign for gradual change.
[formula] µ t = 0.3 9 i=1 (10 − i)H(n − 1000i), S(x) = 0 (x < 0), x/300 (0 ≤ x < 300), 1 (x ≥ 300). log σ t = 0.1 9 i=1 (10 − i)H(n − 1000i [/formula]
The total benefit of alarm sequence a n−1
# Methods for comparison.
In order to conduct the sequential D-MDL algorithm, we employed the univariate Gaussian distribution whose probability density function is given by Eq. (7). We employed three sequential change detection methods for comparison:
(1) Bayesian online change point detection (BOCPD): A retrospective Bayesian online change detection method. It originally calculates the posterior of run length. We modified it to compute a change score by taking the expectation of the reciprocal of run length with respect to the posterior. (2) ChangeFinder (CF) [bib_ref] A unifying framework for detecting outliers and change-points from time series, Takeuchi [/bib_ref] : A state-of-the-art method of abrupt change detection.
(3) ADWIN2 9 : A change detection method with adaptive windowing.
We conducted the sequential D-MDL algorithms with fixed window size in order to investigate their most basic performance in terms of the AUC metric. The sequential D-MDL algorithm with adaptive windowing outputs the window size rather than the D-MDL values themselves, hence in order to evaluate the effectiveness of the magnitude of D-MDL, the sequential D-MDL with fixed windowing is a better target for the comparison. All of CF, BOCPD, and ADWIN2 had some parameters, which we determined from five training sequences drawn from the data generation mechanism so that the AUC scores were made the largest. [fig_ref] Table 1: Average AUC scores ± standard deviation on the synthetic datasets [/fig_ref]. We see that both for the datasets, in the case of abrupt changes, the 0th D-MDL performs best, while in the case of gradual changes, the 1st D-MDL performs best and the 2nd D-MDL performs worse than the 1st but better than the 0th. That matches our intuition. Because the 0th D-MDL was designed so that it could detect abrupt changes while the 1st one was designed so that it could detect starting points of gradual changes.
## Results. the performance comparison is summarized in
## Result ii: applications to covid-19 pandemic analysis
Since the beginning of 2020, many regions/countries have suffered from the epidemic of COVID-19. The purpose of our analysis is to demonstrate the importance of monitoring the dynamics of the epidemic through detecting the occurrence of drastic outbreaks and their signs. We define outbreak as a significant increase in the number www.nature.com/scientificreports/ of cases in a region/country. We note that to contain the spread of COVID-19, many countries have enacted social distancing policies, e.g., stay-at-home order, closing non-essential services, and limiting travel. We thus also relate the results of our analysis to social distancing events.
[formula] a t = 1 (s t > β), 0 (otherwise). b(t; t * ) = 1 − |t−t * | T (0 ≤ |t − t * | < T), 0 (otherwise), B(a n−1 0 ) = n−1 k=0 a k b(k; t * ). N(a n−1 0 ) = n−1 k=0 a k �(b(k; t * ) = 0). [/formula]
In particular, we are mainly concerned with the following two problems:
1. How early are the outbreak signs detected prior to outbreaks? 2. How are the outbreaks/outbreak signs related to the social distancing events?
As a byproduct, the analysis of the dynamics of the basic reproduction number R0 37 is conducted, which can serve as supplementary information to the particular value estimated from the SIR model 45 .
Data source. We studied the data provided by European Centre for Disease Prevention and Control (ECDC) which can be accessed through the link https://www.ecdc.europa.eu/en/publications-data/downloadtodays-data-geographic-distribution-covid-19-cases-worldwide. In this paper, we focused on the first wave because various factors made the situations very complicated in later waves, e.g., virus mutations [bib_ref] Tracking changes in SARS-CoV-2 spike: Evidence that D614G increases infectivity of the..., Korber [/bib_ref] [bib_ref] Covid-19: New coronavirus variant is identified in UK, Wise [/bib_ref] [bib_ref] Complete map of SARS-CoV-2 RBD mutations that escape the monoclonal antibody LY-CoV555..., Starr [/bib_ref] , people being tired of social distancing and the mixture of two waves in the transition period. In particular, we studied 37 countries with no less than 10,000 cumulative cases by Apr. 30, 2020 since some countries started to ease the social distancing around the date. More details about these countries can be found in Sec. 2 of the supplementary information. It is worth mentioning that the proposed method can be applied to any region/country where there is a COVID-19 epidemic because the input to the method is only the number of cases. In practice, we suggest starting to run our algorithm when the spread of the virus into the region of concern through local infections begins but not when the cases are just imported.
Data modeling. We studied two data models by considering the value of R0, which by definition is the product of transmissibility, the average contact rate between susceptible and infected individuals, and the duration of infectiousness [bib_ref] Notes on R0, Jones [/bib_ref]. At the initial phase of an epidemic, R0 is larger than one [bib_ref] On the definition and the computation of the basic reproduction ratio R..., Diekmann [/bib_ref]. And the cumulative cases may grow exponentially [bib_ref] Contributions to the mathematical theory of epidemics IV. Analysis of experimental epidemics..., Kermack [/bib_ref] [bib_ref] A generalized-growth model to characterize the early ascending phase of infectious disease..., Viboud [/bib_ref]. We thus employed the Malthusian growth model 50 because it is widely used for characterizing the early phase of an epidemic [bib_ref] A generalized-growth model to characterize the early ascending phase of infectious disease..., Viboud [/bib_ref]. In particular, the cumulative cases at time t, C(t), grows according to the following equation:
where C(0) is the number of cases at the start of an epidemic, and r is the growth rate of daily new cases. In the experiments, we took the logarithm of C(t) to obtain the linear regression of the logarithm growth with respect to time as follows:
We modeled the residual error of the linear regression using the univariate Gaussian. See Sec. 3 in the supplementary file for the detail of calculation of the MDL change statistics for this model. When a change is detected in the modeling of the residual error, we examine the increase/decrease in the coefficient of the linear regression, i.e., r. We expect to detect changes in the parameter of the exponential modeling to monitor the increase/decrease of R0 because R0 − 1 is proportional to r 47 . In later phases, the exponential growth pattern may not hold. For instance, when R0 < 1 , daily new cases would continue to decline and cease to exist [bib_ref] On the definition and the computation of the basic reproduction ratio R..., Diekmann [/bib_ref]. Considering the complicated real scenarios, epidemic models with certain assumptions on the growth rate or R0 may not fit an epidemic at a given time. Therefore, we employed the univariate Gaussian distribution as in Eq. (7) to directly model the number of daily new cases, without assuming any patterns of the growth. The change in the parameter of the Gaussian modeling may reveal the relation between one and R0, i.e., R0 > 1 when daily new cases increase significantly or R0 < 1 when daily new cases decrease significantly.
We conducted the hierarchical sequential D-MDL algorithm as in "Hierarchical sequential D-MDL algorithm". The confidence parameter δ for the 0th D-MDL as in Eq. (21) was set to be 0.05. Those for the 1st and 2nd D-MDL, i.e. δ 1 , δ 2 as in Eqs. [bib_ref] Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their..., Bi [/bib_ref] , [bib_ref] The effect of human mobility and control measures on the COVID-19 epidemic..., Kraemer [/bib_ref] were determined as follows: We calculated the D-MDL scores around the time when the initial warning was announced by an authority; we determined δ 1 , δ 2 so that the score was the threshold. For example, the initial warning for Japan was set on Feb. 27, when the government required closing elementary, junior high and high schools. If the resulting δ 1 , δ 2 was larger than 1, it was set to be 0.99 because of the concept of confidence parameter. More details about the implementation are provided in Sec. 4 of the supplementary information.
Case study. We present a representative case study of Japan due to space consideration. For results of all the studied countries, please refer to Sec. 5 of the supplementary information. In Japan, state of emergency as the social distancing event was issued on Apr. 7. The results are presented in for the Gaussian modeling and the exponential modeling, respectively. Change scores were normalized into the range [0, 1]. The data of Japan did not include the confirmed cases from 'Diamond Princess' .
With the Gaussian modeling, there were several alarms raised before the social distancing event. For each alarm raised by the 0th D-MDL, the interpretation can be a statistically significant increase in cases, with reference to . Hereafter, a change that was detected by the 0th D-MDL and that corresponded to the increase of cases was regarded as an outbreak, which instantiates our definition of outbreak. The outbreak detection is the classic change detection. We further relate it to R0. Around the dates of the alarms, R0 > 1 was considered www.nature.com/scientificreports/ since we can confirm that the new infections resulted from community transmission. Correspondingly, R0 was estimated around 2.5 in early March by an epidemiological study. When the 0th D-MDL raised an alarm, the window size shrank to zero. Before that, both the 1st and the 2nd D-MDL raised alarms, which are interpreted as the changes in the velocity and the acceleration of the increase of cases, respectively. We can conclude that the 1st and the 2nd D-MDL were able to detect the signs of the outbreak by examining the velocity and the acceleration of the spread. The sign detection is the new concept with which we propose to supplement the classic change detection. The 0th D-MDL raised no alarms about outbreaks after the event. We think the social distancing played a critical role in containing the spread because it can significantly suppress R0 through reducing the contact rate. The 1st D-MDL still raised alarms, which were signs of decreases in the cases. As for the exponential modeling, there were alarms raised by the 0th D-MDL both before and after the social distancing event. By looking at the growth pattern of local cumulative cases in , we can see that all the alarms were about the cessations of the exponential growth. Moreover, we checked that the alarms were associated with decreases in the coefficient of the linear regression. Therefore, we concluded that all the alarms indicated significant decreases in R0. Although the last two alarms were raised on Mar. 26 and Apr. 28, the dates as the change points were within the windows as of Mar. 26 and Apr. 28, and were identified as Mar. 12 and Apr. 18, respectively. There was an epidemiological study 51 which showed the effectiveness of the initial warning announced on Feb. 27 at reducing R0. As a result, it demonstrated that our method can effectively identify the decrease in R0 around Mar. 12. According to the result, our method identified another decrease in R0 around Apr. 18, which we think was mainly due to the social distancing event on Apr. 7. Therefore, our method based on the exponential modeling also confirmed that social distancing was very effective at containing the spread. The alarms raised by the 1st and 2nd D-MDL demonstrated the capability of the sign detection.
[formula] (24) C(t) = C(0) exp(rt),(25) [/formula]
As a comparison, the Gaussian modeling was effective at estimating the relation between one and R0 while the exponential modeling was able to monitor the change in the value of R0. The two models form a complementary relation on monitoring the dynamics of R0. For instance, for Japan, the Gaussian modeling showed that the value of R0 reminded at a value larger than one, and the exponential modeling showed that its value decreased during the studied period. Due to the difference in the modeling, the changes detected by the 0th D-MDL were at different dates between the Gaussian modeling and the exponential modeling. In terms of the sign detection, both the Gaussian modeling and the exponential modeling were effective.
Summarization on individual countries. This section summarizes several statistics about the change detection results in and presents two interesting observations. The first is about how early the signs can be detected prior to changes. For the countries studied, there were 106 and 54 changes in total detected by the Gaussian modeling and the exponential modeling, respectively. There were more changes detected by the Gaussian modeling because daily cases would significantly change with either R0 > 1 or R0 < 1 while it may take relatively longer time for significant changes in R0. The number of changes whose signs were detected by either the 1st or the 2nd D-MDL was 68 and 26 for the Gaussian modeling and the exponential modeling, respectively, representing high detection rates. For each change whose signs were detected, we measured the time difference between the earliest sign alarm and the change alarm. For the Gaussian modeling which can detect outbreaks, the time difference in terms of the number of days is 6.25 (mean) ± 6.04 (standard deviation). Considering the fast spread, six days can buy us considerably long time to prepare for an outbreak, and even to avoid a potential outbreak.
In particular, with the Gaussian modeling, the 1st D-MDL detected signs for 65 changes and the 2nd D-MDL detected signs for 27 changes. The smaller number by the 2nd D-MDL might be because the 1st D-MDL is better at detecting starting points of gradual changes, and is consistent with results on the synthetic datasets as in [fig_ref] Table 1: Average AUC scores ± standard deviation on the synthetic datasets [/fig_ref]. The number of days before which the 1st D-MDL detected signs was 6.35 ± 5.91, and the number for the 2nd D-MDL was 5.56 ± 6.50. Note that not all the changes allowed for sign detection since the 1st D-MDL . Summarization of statistics where changes represent the alarms raised by the 0th D-MDL and signs were alarms raised by either the 1st or the 2nd D-MDL.
## Measurement
Gaussian Exponential www.nature.com/scientificreports/ and the 2nd D-MDL sign detection require one more and two more data points in the window than the 0th D-MDL, respectively. The number of changes allowing for a 1st D-DML sign was 88 while the number for a 2nd D-DML sign was 81. Hence, it turned out that some changes occurred too quickly before signs can be detected. The analysis of the results obtained by the exponential modeling is similar and omitted for space consideration. Second, we observed that on average, countries responding faster in terms of a smaller number of alarms raised by the Gaussian modeling before the social distancing event saw a quicker contraction of daily cases. As of Apr. 30, the curve of daily cases in many countries had been flatten, and even started to be downward. Therefore, alarms for declines in the number of daily cases from the global peak number were raised for ten countries including Austria, China, Germany, Iran, Italy, Netherlands, South Korea, Spain, Switzerland, and Turkey. These countries are referred to as downward countries. In total, the number of all kinds of alarms raised before the event for downward countries was 4.30 ± 2.79 while it was 5.96 ± 4.22 for other countries. Therefore, if the social distancing is a viable option, it is suggested that the action should better be taken before it is late, e.g., later than four alarms. We further measured that it took an average of 30 days to suppress the spread if prompt social distancing policies were enacted. By contrast, the average number of days from the date of social distancing event to Apr. 30 was nearly 37 for non-downward countries, which was considerably more than the time used for suppressing the spread in downward countries. The results of the exponential modeling confirmed the above observation. In particular, changes and their signs which corresponded to decreases in R0 for the downward countries were more than those for the non-downward countries.
Limitations and challenges of the COVID-19 analysis. Since the proposed method only examines the number of COVID-19 cases, the analysis can only give an overall estimation of the dynamics of the pandemic which are the results of the joint effects of various kinds of physical factors including the characteristics of the virus, human mobility patterns, mask usage, vaccine coverage, environmental factors, and etc. When changes happen to any one of the physical factors, e.g., virus mutations or the entry of the virus into sewage 52 , the number of cases may change. Accordingly, the major limitation of the proposed method is that itself cannot associate the detected changes, either outbreaks or their signs, with a particular physical factor.
We were concerned with detecting signs of the first wave of COVID-19. Although we employed the Gaussian model and the exponential growth model in computing D-MDL, such models might not be necessarily most appropriate for dealing with later waves, since a number of waves are mixed in the transition periods. One of challenges is to consider more sophisticated models such as latent variable models in dealing with later waves.
# Conclusion
This paper has proposed a novel methodology for detecting signs of changes from a data stream. The key idea is to use the differential MDL change statistics (D-MDL) as a sign score. This score can be thought of as a natural extension of the differentials of the Kullback-Leibler divergence for measuring the degree of changes to the case where the true mechanism for generating data is unknown. We have theoretically justified D-MDL using the hypothesis testing framework and have empirically justified the sequential D-MDL algorithm using the synthetic data. On the basis of the theory of D-MDL, we have applied it to the COVID-19 pandemic analysis. We have observed that the 0th D-MDL found change points related to outbreaks and that the 1st and 2nd D-MDL were able to detect their signs several days earlier than them. We have further related the change points to the dynamics of the basic reproduction number R0. We have also found that the countries with no more than five changes/change signs before the implementation of social distancing tended to experience the decrease in the number of cases considerably earlier. This analysis is a new promising approach to the pandemic analysis from the view of data science.
Change detection, which aims to detect points in a sequence of random variables at which the probability distribution change, has been studied for decades and has wide applications, such as event detection, failure detection, malware detection, etc [bib_ref] A unifying framework for detecting outliers and change-points from time series, Takeuchi [/bib_ref] [bib_ref] Detecting gradual changes from data stream using MDL change statistics, Yamanishi [/bib_ref] [bib_ref] Detecting changes in streaming data with information-theoretic windowing, Kaneko [/bib_ref]. Change sign detection proposed in this paper aims to detect early warning signals of such changes by identifying the speed and acceleration of changes in the probability distribution, and therefore has the same applicability as the change detection.
Future work includes studying how we can integrate the change analysis such as our methodology with the conventional simulation studies such as SIR model. It is expected that our data science approach has a complementary relation with the simulation approach and gives new insights into epidemiology. Moreover, we plan to study later waves which are more complicated situations than the first wave. www.nature.com/scientificreports/
[fig] Figure 1, Figure 2: log C(t) = rt + log C(0). The results for Japan with the Gaussian modeling. The date on which the social distancing was implemented is marked by a solid line in black. (a) The number of daily new cases. (b) The change scores produced by the 0th D-MDL where the line in blue denotes values of scores and dashed lines in red mark alarms. (c) The window sized for the sequential D-MDL algorithm with adaptive windowing where lines in red mark the shrinkage of windows. (d) The change scores produced by the 1st D-MDL. (e) The change scores by the 2nd D-MDL. In all figures the negative scores are omitted. The results for Japan with the exponential modeling. The label "Local cumulative cases" in subfigure (a) means that the cumulative cases only accumulate daily cases from each starting date of change detection and would be set to zero after each change detected by the 0th D-MDL. The date on which the social distancing was implemented is marked by a solid line in black. (a) The number of cumulative cases. (b) The change scores produced by the 0th D-MDL where the line in blue denotes values of scores and dashed lines in red mark alarms. (c) The window sized for the sequential D-MDL algorithm with adaptive windowing where lines in red mark the shrinkage of windows. (d) The change scores produced by the 1st D-MDL. (e) The change scores produced by the 2nd D-MDL. Scientific Reports | (2021) 11:19795 | https://doi.org/10.1038/s41598-021-98781-4 [/fig]
[table] Table 1: Average AUC scores ± standard deviation on the synthetic datasets. [/table]
|
Effectiveness of a WeChat Combined Continuous Flash Glucose Monitoring System on Glycemic Control in Juvenile Type 1 Diabetes Mellitus Management: Randomized Controlled Trial
# Introduction
Type 1 diabetes mellitus (T1D) is an autoimmune disease in which immune cells attack and destroy insulin-producing pancreatic β cells. According to the latest nationwide population-based registry study, the estimated incidence of T1D per 100,000 person-years for all ages in China was 1.01. [bib_ref] Incidence of type 1 diabetes in China, 2010-13: population based study, Weng [/bib_ref] Early glycemic control is particularly important for preventing the development and progression of chronic complications of diabetes. It is well known that hypoglycemia causes cognitive impairment and mood changes, such as depression, irritation, and fear. In severe cases, changes in brain metabolism can even cause permanent sequelae and be potentially life-threatening. [bib_ref] Impaired awareness of hypoglycemia disrupts blood flow to brain regions involved in..., Nwokolo [/bib_ref] The goal of diabetes treatment in adolescents is to maintain blood glucose level within the normal range without recurrent or severe hypoglycemia; that is, a balance between hyperglycemia treatment and minimizing the frequency of hypoglycemic episodes should be maintained. [bib_ref] Diabetic emergencies -ketoacidosis, hyperglycaemic hyperosmolar state and hypoglycaemia, Umpierrez [/bib_ref] Regular self-monitoring of blood glucose (SMBG) is particularly important for patients with T1D to assess glycemic control, determine the risk of hypoglycemia, and respond promptly. [bib_ref] Frequency of SMBG correlates with HbA1c and acute complications in children and..., Ziegler [/bib_ref] However, it is also accompanied by the burden and pain associated with multiple needle pricks for blood collection and the complexity of the procedures for use. Continuous glucose monitoring systems (CGM) measure interstitial fluid glucose concentrations (ISFG) throughout the day and night; thus, they can provide much more sufficient blood glucose management information than SMBG. [bib_ref] Efficacy of continuous glucose monitoring system (CGMS) to detect postprandial hyperglycemia and..., Maia [/bib_ref] [bib_ref] Limitations of conventional methods of self-monitoring of blood glucose: lessons learned from..., Boland [/bib_ref] The newly developed Abbott FreeStyle Libre is a flash glucose monitoring system (FGM) that requires a reader to scan the subcutaneous probe to obtain glycemic values.
HbA1c is currently recognized as the key surrogate marker for developing long-term diabetes complications and has been used as the primary endpoint for many CGM studies. HbA1c reflects average glucose over the last 2-3 months, but it does not reflect glycemic variability. The use of FreeStyle Libre allows not only recurrent painless glucose sampling but also a real-time observation of glycemic excursions and daily profiles, which can impact immediate therapy decisions.
Achieving optimal glycemic control requires intensive self-management for adolescents with T1D. However, adolescents with T1D struggle with achieving therapeutic targets for their shortage of medical knowledge, variable lifestyles, and less outpatient visits. Additionally, owing to many patients with diabetes and the shortage of medical staff in China, clinicians and nurses are often too overwhelmed to provide normative and continuous management for patients with diabetes. The widespread use of smartphones, especially among the younger population, offers an opportunity to provide diabetes information and interaction with physicians. [bib_ref] The effects of dietary mobile apps on nutritional outcomes in adults with..., Khoury [/bib_ref] Given adolescents' propensity for new technology, web-based interventions may provide a feasible communication with healthcare providers and motivate them to improve self-management.
Although interest in this technology increases, the clinical value of these web-based interventions remains unclear. [bib_ref] The promise and peril of mobile health applications for diabetes and endocrinology, Eng [/bib_ref] To date, most studies were performed in adults with type 2 diabetes, rarely among adolescents with T1D. Furthermore, few of the web-based intervention provided users with personalized education, feedback, or motivation, which are emphasized in current clinical guidelines for behavior change and comprehensive selfmanagement. [bib_ref] Mobile applications for diabetes self-management: status and potential, El-Gayar [/bib_ref] [bib_ref] Features of mobile diabetes applications: review of the literature and analysis of..., Chomutare [/bib_ref] In China, WeChat is the most popular social networking platform; it is a free app characterized by high convenience and accessibility. Through the app, complex information can be presented through video and graphics and provide a feasible way to spread health information to the public. Many studies have shown that WeChat has great potential in health intervention. [bib_ref] How the public uses social media wechat to obtain health information in..., Zhang [/bib_ref] [bib_ref] Social media-promoted weight loss among an occupational population: cohort study using a..., He [/bib_ref] However, to our knowledge, no well-designed studies have focused on a WeChat-based intervention for T1D in China.
Therefore, this study aims to clarify the effect of the WeChat intervention combined with FGM management on glycemic control, hypoglycemia incidence, and life quality in T1D patients.
# Patients and methods
## Trial design
A parallel, three-group, randomized controlled trial study was carried out for 6 months. Participants were recruited from the outpatient clinic of the Department of Endocrinology and Metabolism of Shanghai Jiao Tong University Affiliated Sixth People's Hospital from January 2019 to June 2019. By using a set of random numbers, the participants were randomly assigned into three groups in a 1:1:1 ratio: group A (SMBG group), group B (FGM group), and group C (FGM combined WeChat-interactive management group). This study was registered at chictr.org.cn, number ChiCTR1900025495.
## Study participants
The study inclusion criteria were as follows: 1) adolescent patients (age 10-19 years) diagnosed with T1D with the criteria established by WHO in 1999, with diabetes duration longer than 1 year; 2) glycosylated hemoglobin A1c concentration between 7% and 10%; 3) use of multiple daily insulin (MDI) and continuous subcutaneous insulin infusion (CSII) for at least 3 months, stable diabetes submit your manuscript | www.dovepress.com
## Dovepress
Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy 2021:14 medication regimen for 3 months before study entry (change in insulin ≤20%); 4) previous documentation of blood glucose level self-monitoring regularly for 2 months (at least three times per day) and willingness to continue for at least 6 months; 5) willingness to wear CGM; 6) ability to speak, read, and write Chinese; 7) ability to use WeChat or have a family member able to use the WeChat platform.
The exclusion criteria included the following: 1) the use of CGM 3 months before study entry; 2) having severe diabetic complications such as diabetic retinopathy and diabetic nephropathy; 3) having a known allergy to medical-grade adhesives or CGM and its affiliated components; 4) being pregnant or planning pregnancy; 5) recent severe diseases like myocardial infarction, stroke, psychiatric diseases (historical/recent), malignant tumor, kidney disease (defined as eGFR <45), dermatosis, decided by the investigator; 6) current participation in another investigational study (must have completed any previous studies at least 30 days before being enrolled in this study); 7) current abuse of illicit drugs, alcohol, or prescription drugs; 8) any condition that could impact the reliability of the HbA1c measurement (eg, hemoglobinopathy, hemolytic anemia, chronic liver disease), decided by the investigator.
The study was approved by the Ethics Committee of the Shanghai Jiao Tong University Affiliated Sixth People's Hospital and conformed to the provisions of the Declaration of Helsinki (2018-011). Written informed consent was obtained from all patients before starting this enrollment. And written informed parental consent was obtained from all patients under the age of 18 before starting this enrollment.
## Grouping and intervention
Patients were randomly assigned to groups A, B, or C of conventional blood glucose monitoring, flash glucose monitoring, and flash glucose monitoring combined with WeChat monitoring, respectively. We recorded the data on demographics, laboratory findings, and questionnaire results of the patients at baseline and at the 6-month follow-up.
In Group A, a conventional home glucometer was used to monitor blood glucose ≥ three times a day, and the blood glucose monitoring values were uploaded to the Wenjuan survey platform.
In Group B, the flash blood glucose monitoring system (Libre 1, Abbott Diabetes Care Inc., Alameda, CA) was used. A specialist applied the flash glucose monitor to the back of the upper arm through a simple disposable applicator: a thin wire (flexible probe) was subcutaneously implanted, and the sensor was fixed to the application site with an adhesive film. It recorded the blood glucose value at 15-minute intervals automatically, and the blood glucose value can be determined at any time from the display.
In Group C, patients with the Abbott FreeStyle Libre monitor were asked to subscribe to a WeChat Official Account named "KongTangTianDi," which disseminates scientific diabetes-related information once a week. Furthermore, the WeChat Official Account platform was also used for real-time patient-doctor interactions. A thirdparty health manager was involved in interactive management with patients through the platform. Further, a nurse who specialized in diabetes helped analyze, evaluate, and review the glycemic monitoring data.
According to the study purpose, the functions of the WeChat program included several modules: (1) real-time registration and update of the child's health records; [bib_ref] Impaired awareness of hypoglycemia disrupts blood flow to brain regions involved in..., Nwokolo [/bib_ref] sharing of diabetes-related knowledge including information on healthy diet, proper exercise, school and travel notifications, home environment and lifestyle guidance, and other practical medical knowledge; (3) answering of frequently left messages through text, audio or video; (4) making a remote diagnosis as well as regular online consultations; (5) provision of self-emergency guidance (call for consultation in acute situations), such as the occurrence of hypoglycemia or ketoacidosis [fig_ref] Figure 1: The screen captures to show how the KongTangTianDi WeChat public account works [/fig_ref].
Basic diabetes education, including adequate exercise, diet control, and regular follow-up, was conducted to the whole population after enrollment. All participants could contact clinicians on phone calls during follow-up; those who had subscribed to the WeChat Official Account were highly encouraged to contact clinicians through the platform.
## Data collection
Data were collected at baseline and at the 6-month follow-up for all participants. The data collected included information on demographic characteristics, anthropometric measurements, biochemical indicators, and questionnaires. The age, sex, past medical history, and lifestyle behaviors were recorded. Body mass index (BMI) was calculated as weight divided by height squared (kg/m 2 ). Biochemical indicators included fasting plasma glucose (FPG), HbA1c, and blood lipid profile. Venous blood was collected in the morning after a 10-hour overnight fast. For FPG, blood samples were Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy 2021:14 submit your manuscript | www.dovepress.com DovePress collected into sodium fluoride + potassium oxalate anticoagulation tubes, and the plasma glucose was tested within 2 hours of sampling (glucose oxidase method; Shanghai Kehua Bio-Engineering Co., Ltd. kit, assayed on a Glamour2000 automated biochemical analyzer). The HbA1c (normal reference value: 4.3-6.5%) was measured through highperformance liquid chromatography (Variant II hemoglobin testing system, Bio-Rad). The Hitachi 7080 automated biochemical analyzer was used for blood lipid panel testing. The questionnaires were divided into three parts and were completed with guidance from dedicated personnel. These were Diabetes Monitoring and Treatment Satisfaction Questionnaires (DMTSQ), Diabetes Quality of Life Measure (DQOL), and the Chinese Version Hypoglycemia Fear Survey II (CHFSII).
An episode of hypoglycemia is defined as glucose value by SMBG under 3.9 mmol/L or a presentation with classic symptoms. Hypoglycemic events in patients detected by Libre were recommended to retest fingertip blood glucose since the Libre is less accurate in low glucose levels.
## Sample size
The sample size was determined by estimating the change in HbA1c values assuming a 0.5% reduction in values in the intervention group. With alpha=0.05 and beta=0.20, we calculated an ideal sample size of 14 patients per group based on this preliminary study. After a dropout rate of 20% was incorporated, a sample size of more than 18 patients per group was required.
# Statistical analysis
All data were entered into SPSS [computer program] (Version 23.0. Armonk, NY: IBM Corp; 2015) by a specialist and analyzed. Data with a normal distribution were presented as mean and standard deviations (SD), and data with a non-normal distribution were presented as median with interquartile ranges (IQR). Analyses of variance (ANOVA) and covariance were used for intergroup comparisons of normally distributed data, whereas nonparametric analysis was used for non-normally distributed data. The Chi-Square test was used to compare proportions, and the linear correlation was used for single correlation analysis. Differences with p <0.05 (two-tailed) were considered statistically significant.
# Results
## Baseline characteristics
Of the 120 participants who underwent screening in this study, 80 were enrolled and randomized to groups A (n=30), B (n=25), or C (n=25). Of these, 10 patients from group A, 5 from group B, and 5 from group C discontinued the study because they either withdrew their consent or were lost to follow-up [fig_ref] Figure 2: Study design and flow chart of participants recruitment and follow-up [/fig_ref]. Thus, a total of 60 adolescent patients (31 males, 29 females; age, mean ± SD 13.20 ± 1.68 years) with T1D were included, with 20 patients in each group (A, B, or C). The general baseline characteristics, metabolic indicators, and questionnaire scores of the patients at the time of enrollment are shown in [fig_ref] Table 1: Clinical Characteristics and Questionnaire Scores of Three Groups at Baseline Notes [/fig_ref] ; no significant inter-group difference at baseline was observed.
## Changes in glycemic control
After 6 months of study intervention, the HbA1c levels of Group A had no significant difference from the baseline (7.75% ± 1.34% vs 7.82% ± 1.54%, p=0.30). While Group B had slightly declined (7.2% ± 1.88% vs 7.43% ± 2.21%, p=0.04), and Group C decreased obviously (7.26% ± 0.89% vs 7.78% ± 1.23%, p< 0.001). The largest decrease was observed in Group C (Group B vs Group C, p =0.04). [fig_ref] Figure 3: Comparison of HbA1c decline from baseline to 6 months later among SMBG,... [/fig_ref].
## Dovepress
## Changes in episodes of hypoglycemia
After the 6-month intervention, the number of hypoglycemic episodes per month in groups B and C decreased significantly from the baseline: Group A: 8.7 ± 5.12 vs 9.25 ± 2.9, respectively (p =0.324); Group B: 5.7 ± 5.9 vs 7.4 ± 7, respectively (p =0.03); and Group C: 4.1 ± 3.46 vs 7.0 ± 5.63, respectively (p< 0.001). However, the magnitude of the decrease was higher in Group C than that observed in Group B (p =0.02) [fig_ref] Figure 4: Incidence of hypoglycemia per month in three management groups [/fig_ref].
## Changes in self-monitoring level, life quality and satisfaction rate
At the 6-month follow-up, the DMTSQ scores of all three groups increased from the baseline as follows: Group A: 57.65 ± 11.17 vs 56.70 ± 10.94, respectively (p =0.03); Group B: 57.70 ± 7.03 vs 53.90 ± 7.57, respectively (p <0.001); and Group C 62.75 ± 6.21 vs 56.00 ± 5.88, respectively (p <0.001); the most significant increase was noted in Group C (Group A vs Group B, p = 0.003; Group A vs Group C, p <0.001; Group B vs Group C, p =0.006) [fig_ref] Figure 5: Effects of SMBG, FGM, and FGM combined WeChat-interactive management methods on quality... [/fig_ref]. Similarly, the DQOL scores in groups B and C decreased from the baseline, whereas the scores of Group A did not change significantly: [fig_ref] Figure 2: Study design and flow chart of participants recruitment and follow-up [/fig_ref]. The decrease in Group C was greater than that observed in Group B (p=0.048) [fig_ref] Figure 5: Effects of SMBG, FGM, and FGM combined WeChat-interactive management methods on quality... [/fig_ref]. The CHFSII scores in all three groups at the 6th month follow-up decreased from the baseline, as follows: Group A: 31.85 ± 11.59 vs 35.55 ± 12.24, respectively (p < 0.001); Group B: 37.30 ± 16.66 vs 42.25 ± 18.27, respectively (p < 0.001); and Group C: 24.6 ± 10.67 vs 32.25 ± 11.84, respectively (p < 0.001), with significant intergroup differences in Group A vs Group C (p < 0.001) and Group
# Discussion
In this study, we found that the HbA1c levels and the number of hypoglycemic episodes per month decreased from the baseline in all three groups after 6 months of intervention, with the greatest decrease observed in the FGM combined WeChat-interactive management group. The FGM groups (Group B and Group C) showed a greater improvement in DQOL scores. Therefore, our novel integrated WeChat-Flash CGM model achieved a convenient doctor-patient interaction and further optimized the management effects. The incidence of type 1 diabetes under 15 years old increased rapidly in the past two decades in China. [bib_ref] Incidence of type 1 diabetes in China, 2010-13: population based study, Weng [/bib_ref] Chinese adolescents with type 1 diabetes struggle with considerable challenges, including lower selfmanagement, high depressive symptoms, suboptimal metabolic control, and quality of life. [bib_ref] Diabetes self-management, depressive symptoms, quality of life and metabolic control in youth..., Guo [/bib_ref] Retrospective studies demonstrated that CGM achieved great glycemic benefit in pediatric, adolescent patients. [bib_ref] Current state of type 1 diabetes treatment in the U.S.: updated data..., Miller [/bib_ref] [bib_ref] Effect of continuous glucose monitoring on glycemic control in adults with type..., Beck [/bib_ref] As CGM becomes more routine, especially given the potential benefit of early CGM initiation, 16 the decision-making process about initiating CGM may be less tangled for patients and families.
Recent studies have shown that smartphone-based apps are an efficient and effective tool for diabetes management. [bib_ref] Do mobile phone applications improve glycemic control (HbA1c) in the self-management of..., Hou [/bib_ref] [bib_ref] Mobile phone applications for diabetes management: a systematic review, Quevedo Rodríguez [/bib_ref] [bib_ref] The influence of the smart glucose manager mobile application on diabetes management, Gunawardena [/bib_ref] Kirwan et al used an app called Glucose Buddy in a 9-month follow-up randomized study, which enrolled 72 patients with T1D, and found that HbA1c decreased by an average of 1.28% (from 9.08% [SD 1.18] to 7.8% [SD 0.75]) in the intervention group and an average of 0.11% (from 8.47% [SD 0.86] to 8.58% in the control group. [bib_ref] Diabetes self-management smartphone application for adults with type 1 diabetes: randomized controlled..., Kirwan [/bib_ref] In their randomized clinical trial, Zhang et al concluded that the app interactive management group showed a greater reduction in HbA1c (2.03% vs 1.37%) when compared to studies with the control group. [bib_ref] Effectiveness of smartphone app-based interactive management on glycemic control in chinese patients..., Zhang [/bib_ref] Consistent with this previous study, our trial showed a greater improvement in HbA1c level in the WeChat interactive group after a 6-month follow-up. However, the improvement in HbA1c levels identified in this study was not as high as that observed in previous studies. [bib_ref] Diabetes self-management smartphone application for adults with type 1 diabetes: randomized controlled..., Kirwan [/bib_ref] [bib_ref] Effectiveness of smartphone app-based interactive management on glycemic control in chinese patients..., Zhang [/bib_ref] This may have been due to a much lower baseline HbA1c level in our study compared to that observed for the other studies.
Improving quality of life (QOL) for children and their families while maintaining glycemic control within targets is a substantial challenge in diabetes treatment. [bib_ref] Quality of life and technology: impact on children and families with diabetes, Hirose [/bib_ref] A previous study showed that CGM could positively influence glucose control and help adolescent patients with T1D to enhance their quality of life. [bib_ref] The impact of continuous glucose monitoring on markers of quality of life..., Polonsky [/bib_ref] DQOL domains usually refer to treatment satisfaction, family relationships, self-efficacy, lifestyle flexibility, anxiety, fear of glucose fluctuations, fear of diabetes complications, and treatment expectations. [bib_ref] The DAWN youth WebTalk study: methods, findings, and implications, Peyrot [/bib_ref] [bib_ref] Quality of life measurement in children and adolescents: issues, instruments, and applications, Wallander [/bib_ref] Hence, apart from the HbA1c level, our study included QOL as a second major outcome to ensure The design of this study was adjusted based on previous recent study outcomes on this topic. Zhang et al conducted a national web-based survey on use, perspectives, and attitudes regarding diabetes management mobile apps among patients with diabetes and diabetologists in China. [bib_ref] Use, perspectives, and attitudes regarding diabetes management mobile apps among diabetes patients..., Zhang [/bib_ref] They found that diabetes education, knowledge, patient-doctor communication, and diabetes diaries were considered the most important functions of a diabetes app. Performance expectancy and social influence are the most important determinants of the intention to use diabetes management apps. [bib_ref] Factors influencing patients' intentions to use diabetes management apps based on an..., Zhang [/bib_ref] Thus, our WeChat public account platform had these elements integrated into each module. We observed that sending diabetic health education messages via the WeChat platform to diabetes patients was regarded as the most useful module through the endpoint interview, which supports that self-care behaviors (such as appropriate diet, regular exercise, and medication) contributed to achieving effective diabetes control.
In this study, we optimized the method for blood glucose monitoring in adolescents with T1D and used the internet to improve the conventional diabetes education model to strengthen the patient's understanding of T1D. Specifically, the flash glucose monitoring system used in this study was able to record blood glucose values at any time and allowed adolescent patients to receive warnings that indicated hypoglycemia or hyperglycemia. Accordingly, the patients were able to make adjustments to their diet and medication. Furthermore, the WeChat platform is a widely used social software app in China. Patients do not have to download, install, and register on other apps since WeChat has multipurpose features. Furthermore, it could make patients receive professional suggestions from the physician online and save the roundtrip time. It can help juvenile patients gain professional knowledge of diabetes, strengthen self-management consciousness, reduce the fear of the disease, and change unhealthy habits. There have been no similar research studies in China; thus, this presents preliminary insights into a novel glycemic monitoring method.
The study has some limitations. First, the sample size of the present study was small, and further studies are needed to recruit more participants. Second, this was a single-center study with a follow-up time of only 6 months. Considering the chronic characteristics of diabetes, the study could not assess the long-term impact of FGM combined WeChat-interactive model on diabetes management. Third, though statistically non-significant, there were differences in diabetes duration by 1 year between Groups A, B, and C. It is well known that most people with T1D have a honeymoon period of 6 months after diagnosis, which can last 1-2 years. So, this difference in diabetes duration may have affected the outcome. Finally, owing to the trial being an internet-based intervention, this study required high self-consciousness.
# Conclusion
Our study suggests that the FGM combined with the WeChat-based support model can improve glycemic control, reduce the incidence of hypoglycemia, improve blood glucose fluctuations, and increase diabetes-related awareness in adolescents patients with T1D. It may be another solid clinical evidence for the feasibility and effectiveness of digital diabetes. Therefore, it might be popularized and implemented in the management of T1D, especially in the context of the current COVID-19 pandemic.
## Abbreviations
CGM, continuous glucose monitoring systems; DCCT, Diabetes Control and Complications Trial; DMTSQ, Diabetes Monitoring and Treatment Satisfaction Questionnaire; DQOL, Diabetes Specific Quality of Life; FGM, flash glucose monitoring system; FPG, fasting plasma glucose; HbA1c, glycated hemoglobin A1c; ISFG, interstitial fluid glucose concentrations; QOL, quality of life; SMBG, self-monitoring of blood glucose; T1D, type 1 diabetes mellitus; CSII, continuous subcutaneous insulin infusion; MDI, multiple daily insulin.
# Data sharing statement
After publication, the authors intend to share individual deidentified participant data, for 1 year, when asked by e-mail from the corresponding author upon reasonable request. execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
# Funding
[fig] Figure 1: The screen captures to show how the KongTangTianDi WeChat public account works. (A) The platform module: diabetic frontier information and lifestyle guidance; (B) registration and update of the child's health records; (C) real-time online consultation. submit your manuscript | www.dovepress.com DovePress Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy 2021:14 [/fig]
[fig] Figure 2: Study design and flow chart of participants recruitment and follow-up. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy 2021:14 submit your manuscript | www.dovepress.com [/fig]
[fig] Figure 3: Comparison of HbA1c decline from baseline to 6 months later among SMBG, CGM, and CGM + WeChat groups of T1D patients. [/fig]
[fig] Figure 4: Incidence of hypoglycemia per month in three management groups. Notes: *P<0.05, ***P<0.001 6-month follow-up vs baseline. submit your manuscript | www.dovepress.com DovePress Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy 2021:14 B vs Group C (p=0.018) (Figure 5C). And the comparison of the fluctuation range of each indicator is shown in the Supplementary Figure 1. [/fig]
[fig] Figure 5: Effects of SMBG, FGM, and FGM combined WeChat-interactive management methods on quality of life and self-perception of T1D patients. (A) Changes of DMTSQ Scores in three groups. (B) Changes of DQOL Scores in three groups. (C) Changes of CHFSII Scores in three groups. Notes: *P<0.05, **P<0.01 6 months later vs baseline. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy 2021:14 submit your manuscript | www.dovepress.com DovePress that developing technologies have a positive impact on children and families with diabetes. [/fig]
[table] Table 1: Clinical Characteristics and Questionnaire Scores of Three Groups at Baseline Notes: Data are expressed as mean (SD). Group A: SMBG group; Group B: FGM group; Group C: FGM combined WeChat-interactive management group. [/table]
|
Influence of Chronobiology on the Nanoparticle-Mediated Drug Uptake into the Brain
Little attention so-far has been paid to the influence of chronobiology on the processes of nanoparticle uptake and transport into the brain, even though this transport appears to be chronobiologically controlled to a significant degree. Nanoparticles with specific surface properties enable the transport across the blood-brain barrier of many drugs that normally cannot cross this barrier. A clear dependence of the central antinociceptive (analgesic) effects of a nanoparticle-bound model drug, i.e., the hexapeptide dalargin, on the time of day was observable after intravenous injection in mice. In addition to the strongly enhanced antinociceptive effect due to the binding to the nanoparticles, the minima and maxima of the pain reaction with the nanoparticle-bound drug were shifted by almost half a day compared to the normal circadian nociception: The maximum in the pain reaction after i.v. injection of the nanoparticle-bound dalargin occurred during the later rest phase of the animals whereas the normal pain reaction and that of a dalargin solution was highest during the active phase of the mice in the night. This important shift could be caused by an enhanced endo-and exocytotic particulates transport activity of the brain capillary endothelial cells or within the brain during the rest phase.
# Introduction
Drug delivery to the brain is restricted or for most drugs totally prevented by the blood-brain barrier (BBB) [bib_ref] The blood-brain barrier: Principles for targeting peptides and drugs to the central..., Begley [/bib_ref] [bib_ref] Delivery of therapeutic agents to the central nervous system: The problems and..., Begley [/bib_ref]. One possibility to overcome this barrier is a delivery of the drugs by binding OPEN ACCESS them to biocompatible and rapidly biodegradable nanoparticles [bib_ref] Drug delivery to the central nervous system by polymeric nanoparticles: What do..., Kreuter [/bib_ref]. Materials for such particles include poly(alkyl cyanoacrylates), polylactides, and crosslinked human serum albumin (HSA). The mechanism of the transport across the BBB appears to be receptor-mediated endocytosis followed by transcytosis of the drug-loaded nanoparticles into the brain or by the release of the drugs within the endothelial cells [bib_ref] Drug delivery to the central nervous system by polymeric nanoparticles: What do..., Kreuter [/bib_ref] [bib_ref] Transport of drugs across the blood-brain barrier by nanoparticles, Wohlfart [/bib_ref] [bib_ref] Mechanism of polymeric nanoparticle-based drug transport across the blood-brain barrier (BBB), Kreuter [/bib_ref] [bib_ref] Albumin nanoparticles targeted with Apo E enter the CNS by transcytosis and..., Zensi [/bib_ref]. In order to enable this receptor-mediated uptake the modification of the nanoparticle surface is necessary [bib_ref] Drug delivery to the central nervous system by polymeric nanoparticles: What do..., Kreuter [/bib_ref] [bib_ref] Transport of drugs across the blood-brain barrier by nanoparticles, Wohlfart [/bib_ref] [bib_ref] Mechanism of polymeric nanoparticle-based drug transport across the blood-brain barrier (BBB), Kreuter [/bib_ref]. This can be achieved by overcoating of the nanoparticles with certain surfactants that adsorb specific apolipoproteins from the blood after intravenous injection [bib_ref] Influence of the type of surfactant on the analgesic effects induced by..., Kreuter [/bib_ref] [bib_ref] Chemotherapy of brain tumour using doxorubicin bound to surfactant-coated poly(butyl cyanoacrylate) nano-particles:..., Petri [/bib_ref] or by the covalent attachment of such apolipoproteins [bib_ref] Covalent attachment of apolipoprotein A-I and apolipoprotein B-100 to albumin nanoparticles enables..., Kreuter [/bib_ref] , transferrin [bib_ref] Transferrin-and transferrin-receptor-antibodymodified nanoparticles enable drug delivery across the blood-brain barrier (BBB), Ulbrich [/bib_ref] , insulin [bib_ref] Targeting the insulin receptor: Nanoparticles for drug delivery across the blood-brain barrier..., Ulbrich [/bib_ref] , or other moieties to the particles that enable their interaction with the respective receptors on the brain capillary endothelial cells.
So-far, little attention has been paid to the influence of chronobiology on the processes of nanoparticle uptake and transport into the brain. However, these processes appear to be controlled by chronobiology to a significant extent [bib_ref] Circadian phase-dependent antinociceptive reaction in mice after i. v. injection of dalargin-loaded..., Ramge [/bib_ref]. Chronobiology, i.e., biological rhythms, play a major role in the quantity and quality of a lot of disease states, biological transport processes, as well as drug actions [bib_ref] Chronobiology, drug-delivery, and chronotherapeutics, Lemmer [/bib_ref]. Indeed, these rhythms cover nearly every division of time: oscillations of one per second (e.g., in the electroencephalogram), one per several seconds (e.g., respiratory rhythm, heart rate), one within 24 h (e.g., circadian rhythms), up to one per year (e.g., circannual rhythms) [bib_ref] Discoveries of rhythms in human biological functions: A historical review, Lemmer [/bib_ref]. Blood pressure, heart rate, peripheral resistance, pressure and the release/activity of vasodilating hormones all display pronounced circadian variations [bib_ref] The importance of circadian rhythms on drug response in hypertension and coronary..., Lemmer [/bib_ref]. The mammalian circadian clock is located in the neurons of suprachiasmatic nuclei (SCN) in the brain and in peripheral tissues [bib_ref] Coordination of circadian timing in mammals, Reppert [/bib_ref] [bib_ref] Clock genes, oscillators, and cellular networks in the suprachiasmatic nuclei, Hastings [/bib_ref]. At a cellular level large portions of cellular physiology -from transcription and translation to intracellular signaling cascades -can show daily variations in activity [bib_ref] New Insights and therapeutic implications, Dallmann [/bib_ref]. Numerous studies in animals, as well as clinical studies, have provided convincing evidence that the pharmacokinetics and/or the effects/side effects of drugs can be modified by the circadian time and/or the timing of drug administration within 24 h of a day [bib_ref] Chronopharmacology and controlled drug release, Lemmer [/bib_ref]. A circadian phase dependency also is known for the perception of pain [bib_ref] Chronopharmacology of nonsteroid anti-inflammatory drugs, Labrecque [/bib_ref] [bib_ref] Biological rhythms in pain and analgesia, Labrecque [/bib_ref]. Rhythmicity also has been demonstrated for the concentrations of neurotransmitters involved in pain modulation. Moreover, opiate receptor binding in various regions of the rat brain was demonstrated to vary with circadian time [bib_ref] Circadian rhythm in rat brain opiate receptor, Naber [/bib_ref].
## Circadian influence on the pharmacological activity of an analgesic hexapeptide delivered by nanoparticles across the blood-brain barrier
Ramge et al. [bib_ref] Circadian phase-dependent antinociceptive reaction in mice after i. v. injection of dalargin-loaded..., Ramge [/bib_ref] addressed the problem of the influence of chronobiology on the drug transport across the BBB and investigated the fluctuations and the day-time dependence of the pain reactivity of mice using the tail-flick as well as the hot plate test models confirming earlier data of Frederickson et al. [bib_ref] Hyperalgesia induced by naloxone follows diurnal rhythm in responsivity to painful stimuli, Frederickson [/bib_ref] : A maximal pain reaction indicated by a minimal reaction time was observed during the night time which represents the activity phase of the mice. The intravenous injection of a solution of the antinociceptive (analgesic) model drug dalargin did not change this pattern of the pain reaction [fig_ref] Figure 1: Dependence of the reaction times [/fig_ref]. Dalargin is a hexapeptide that being an artificial endorphin, has antinociceptive properties upon injection directly into the brain ventricles, but is not able to produce such effects after intravenous injection as it is unable to permeate the BBB [bib_ref] Activity profiles of dalargin and its analogues in μ-, δ-and κ-opioid receptor..., Pencheva [/bib_ref]. However, as already shown earlier [bib_ref] Passage of peptides through the blood-brain barrier with colloidal polymer particles (nanoparticles), Kreuter [/bib_ref] [bib_ref] Analgesic activity of the hexapeptide dalargin adsorbed on the surface of polysorbate..., Alyautdin [/bib_ref] , the binding of dalargin to poly(butyl cyanoacrylate) and additional overcoating with polysorbate 80 (Tween ® 80) enabled the transport of this drug across the BBB into the brain and yielded a significant dose-dependent non-nociceptive effect after intravenous injection [fig_ref] Figure 1: Dependence of the reaction times [/fig_ref] , upper curve and [fig_ref] Figure 2: Dose-response effects of dalargin-coated nanoparticles [/fig_ref]. Without the polysorbate coating the drug-loaded particles cannot interact with the respective receptors and, therefore, also are not able to traverse the BBB [bib_ref] Drug delivery to the central nervous system by polymeric nanoparticles: What do..., Kreuter [/bib_ref] [bib_ref] Transport of drugs across the blood-brain barrier by nanoparticles, Wohlfart [/bib_ref] [bib_ref] Mechanism of polymeric nanoparticle-based drug transport across the blood-brain barrier (BBB), Kreuter [/bib_ref] [bib_ref] Passage of peptides through the blood-brain barrier with colloidal polymer particles (nanoparticles), Kreuter [/bib_ref] [bib_ref] Analgesic activity of the hexapeptide dalargin adsorbed on the surface of polysorbate..., Alyautdin [/bib_ref]. However, as mentioned above, the coating with polysorbate 80 led to the adsorption of apolipoprotein E [bib_ref] Chemotherapy of brain tumour using doxorubicin bound to surfactant-coated poly(butyl cyanoacrylate) nano-particles:..., Petri [/bib_ref] which then interacted with the LRP1-receptors located on the brain capillary endothelial cells, followed by endocytosis and transcytosis of the nanoparticles across these cells [bib_ref] Drug delivery to the central nervous system by polymeric nanoparticles: What do..., Kreuter [/bib_ref] [bib_ref] Transport of drugs across the blood-brain barrier by nanoparticles, Wohlfart [/bib_ref] [bib_ref] Mechanism of polymeric nanoparticle-based drug transport across the blood-brain barrier (BBB), Kreuter [/bib_ref] [bib_ref] Albumin nanoparticles targeted with Apo E enter the CNS by transcytosis and..., Zensi [/bib_ref] [bib_ref] Apolipoprotein-mediated transport of nanoparticle-bound drugs across the blood-brain barrier, Kreuter [/bib_ref] [bib_ref] Uptake mechanism of apoE-modified nanoparticles on brain capillary endothelial cells as a..., Wagner [/bib_ref]. This effect was strongly day time-dependent [fig_ref] Figure 1: Dependence of the reaction times [/fig_ref]. As a consequence, the maximal possible effect (MPE) increased to 90% in the morning (8:00) and to only 70% in the evening (20:00) at a dalargin dose of 10 mg/kg, again showing the influence of the circadian rhythm [fig_ref] Figure 2: Dose-response effects of dalargin-coated nanoparticles [/fig_ref].
Even more interesting was the finding that whereas the pain reactions without drug and with dalargin solution were highest (shortest reaction times) during the active phase of the mice in the night, the maximum in the pain reaction after i.v. injection of the nanoparticle-bound dalargin occurred in the later day times resulting in a shift of the calculated circadian response curve of over 10 h, almost a half-day shift [fig_ref] Figure 1: Dependence of the reaction times [/fig_ref]. This result is very surprising. One would expect that most processes are more active during the activity phase of the animals. Cerebral blood flow, for instance, has been shown to also be circadian phase-dependent, being greater during the activity period of nocturnal animals [bib_ref] Diurnal variation of cerebral blood flow in rat hippocampus, Endo [/bib_ref]. As a consequence, a more efficient transport of the nanoparticles by the blood to the brain endothelium would be anticipated. Also the energy-dependent endocytosis and transcytosis processes also should be expected to be up-regulated and more operative, and hence the reaction time should be increased accordingly during this time. However, the opposite is the case, the reduction of the pain reaction time by the dalargin-loaded nanoparticles was 20% lower in the animals' activity phase than in the rest phase, indicated by the observed lower MPE [fig_ref] Figure 2: Dose-response effects of dalargin-coated nanoparticles [/fig_ref].
One possible explanation for this result could be that the quality or quantity of the interaction of dalargin on the level of the opioid receptors in the brain is somehow altered by the intrinsic properties of the nanoparticles causing a decelerated transport of the particles into or within the brain or a retarded release of dalargin. However, in all experiments with dalargin, loperamide, tubocurarine, or other drugs the response times where very short [bib_ref] Drug delivery to the central nervous system by polymeric nanoparticles: What do..., Kreuter [/bib_ref]. The antinociceptive responses displayed in [fig_ref] Figure 1: Dependence of the reaction times [/fig_ref] , for instance, were recorded 15 min after intravenous injection. For this reason this possibility appears to be unlikely. Another possibility is that the chronobiological fluctuation of the expression in the quantity of the opioid receptor in various regions of the rat brain could affect the drug effect [bib_ref] Circadian rhythm in rat brain opiate receptor, Naber [/bib_ref] , regardless of intrinsic effect of the nanoparticles. In this case the observed 10 h shift in the minima and maxima caused by binding to the nanoparticles would be difficult to interpret as this fluctuation should be similar in both cases and only could be explained by an alteration of the drug disposition in the brain by the nanoparticles compared to free drug. Again, the rapid pharmacological response times requiring a fast release speak against this. It, therefore, appears to be most likely that during the rest phase of the animals the activity of certain energy-requiring transport processes such as endocytosis and transcytosis is upregulated. This is of course a speculative hypothesis that has to be further substantiated experimentally.
The above finding has very important and further reaching implications. It has to be noted that the difference between the day-time and night time nanoparticle-caused effects is considerable, i.e., over 20%. This non-negligible difference demonstrates that the outcome of results concerning the drug transport across the BBB using nanoparticles significantly depend on the time of day of the experiments. Therefore, in comparing different experiments this fact has to be taken in consideration. Secondly, it should be kept in mind that these findings also transcend to the influence of chronobiology on the permeation process efficacy of other particulates across this barrier. This includes the transcellular lymphocyte migration through the brain endothelium [bib_ref] Mini review: Transendothelial migration of leukocytes: Through the front door or around..., Engelhardt [/bib_ref] or the BBB crossing of endogenous particulates such as antibodies or of pathogens like viruses and bacteria that also may be day time-dependent and could be more active during the rest phase. The complexity and interactivity of these processes may make it difficult to elucidate the single contribution of these processes.
# Conclusions
The pain reaction of mice was significantly influenced and altered in a dose-dependent fashion by the nanoparticle-mediated transport of an antinociceptive (analgesic) drug, dalargin, across the blood-brain barrier. The minima and maxima of the pain reaction with the nanoparticle-bound drug were shifted by almost half a day compared to the normal circadian nociception. The maximum in the pain reaction after i.v. injection of the nanoparticle-bound dalargin occurred during the later rest phase of the animals whereas the normal pain reaction was highest during the active phase of the mice in the night. This shift could be caused by an enhanced endo-and exocytotic nanoparticle transport activity of the brain capillary endothelial cells or within the brain during the rest phase.
## Conflicts of interest
The author declares no conflict of interest.
[fig] Figure 1: Dependence of the reaction times (s) in the hot-plate test 15 min after intravenous injection of 10 mg/kg of a dalargin solution (DAL/SOL) or of dalargin-coated nanoparticles (DAL/NP) on the time of day. Groups of 10-12 DAB/2 mice were tested every 2 h within 24 h; lights on from 07:00 to 19:00. Shown are mean values ± SEM; the solid line represents the cosine fit to the data. (Adapted with permission from[13]. Copyright 1999 Informa Healthcare). [/fig]
[fig] Figure 2: Dose-response effects of dalargin-coated nanoparticles. Response determined 15 min after intravenous injection (DAL/NP, 2.5-10 mg/kg) at either 08:00 or 20:00; lights on from 07:00 to 19:00. Shown are mean values ± SEM of the maximal possible effect (% MPE. Adapted with permission from[13]. Copyright 1999 Informa Healthcare). [/fig]
|
Cell-Free DNA as a Diagnostic and Prognostic Biomarker in Pediatric Rhabdomyosarcoma
PURPOSE Total cell-free DNA (cfDNA) and tumor-derived cfDNA (ctDNA) can be used to study tumor-derived genetic aberrations. We analyzed the diagnostic and prognostic potential of cfDNA and ctDNA, obtained from pediatric patients with rhabdomyosarcoma.METHODS cfDNA was isolated from diagnostic plasma samples from 57 patients enrolled in the EpSSG RMS2005 study. To study the diagnostic potential, shallow whole genome sequencing (shWGS) and cell-free reduced representation bisulphite sequencing (cfRRBS) were performed in a subset of samples and all samples were tested using droplet digital polymerase chain reaction to detect methylated RASSF1A (RASSF1A-M). Correlation with outcome was studied by combining cfDNA RASSF1A-M detection with analysis of our rhabdomyosarcoma-specific RNA panel in paired cellular blood and bone marrow fractions and survival analysis in 56 patients.RESULTSAt diagnosis, ctDNA was detected in 16 of 30 and 24 of 26 patients using shallow whole genome sequencing and cfRRBS, respectively. Furthermore, 21 of 25 samples were correctly classified as embryonal by cfRRBS. RASSF1A-M was detected in 21 of 57 patients. The presence of RASSF1A-M was significantly correlated with poor outcome (the 5-year event-free survival [EFS] rate was 46.2% for 21 RASSF1A-M-positive patients, compared with 84.9% for 36 RASSF1A-M-negative patients [P , .001]). RASSF1A-M positivity had the highest prognostic effect among patients with metastatic disease. Patients both negative for RASSF1A-M and the rhabdomyosarcoma-specific RNA panel (28 of 56 patients) had excellent outcome (5-year EFS 92.9%), while double-positive patients (11/56) had poor outcome (5-year EFS 13.6%, P , .001).CONCLUSION Analyzing ctDNA at diagnosis using various techniques is feasible in pediatric rhabdomyosarcoma and has potential for clinical use. Measuring RASSF1A-M in plasma at initial diagnosis correlated significantly with outcome, particularly when combined with paired analysis of blood and bone marrow using a rhabdomyosarcoma-specific RNA panel.JCO Precis Oncol 7:e2200113.
# Introduction
Rhabdomyosarcoma, the most common sarcoma among children and adolescents, accounts for approximately 3% of pediatric tumors.Despite considerable research regarding treatment and risk stratification, 1 of 3 patients will experience relapse. [bib_ref] Prognostic factors in metastatic rhabdomyosarcomas: Results of a pooled analysis from United..., Oberlin [/bib_ref] [bib_ref] Randomized comparison of intensified six-drug versus standard three-drug chemotherapy for high-risk nonmetastatic..., Oberlin [/bib_ref] [bib_ref] Prognostic factors after relapse in nonmetastatic rhabdomyosarcoma: A nomogram to better define..., Chisholm [/bib_ref] The use of liquid biopsies in pediatric patients is drawing growing interest. [bib_ref] The pitfalls and promise of liquid biopsies for diagnosing and treating solid..., Van Paemel [/bib_ref] [bib_ref] Circulating cell-free tumor DNA analysis in pediatric cancers, Andersson [/bib_ref] Our group reported that the presence of rhabdomyosarcoma-derived mRNA in the cellular fraction of peripheral blood (PB) and bone marrow (BM) at initial diagnosis is correlated with poor outcome and could potentially improve current risk stratification. [bib_ref] Improving risk stratification for pediatric patients with rhabdomyosarcoma by molecular detection of..., Lak [/bib_ref] Studies on other pediatric solid tumors demonstrated cell-free DNA (cfDNA) analysis from plasma to provide added value for diagnostics, prognostics, and response monitoring. [bib_ref] Hypermethylated RASSF1A as circulating tumor DNA marker for disease monitoring in neuroblastoma, Van Zogchel [/bib_ref] [bib_ref] Circulating tumor DNA analysis enables molecular characterization of pediatric renal tumors at..., Jimenez [/bib_ref] [bib_ref] Genomic copy number profiling using circulating free tumor DNA highlights heterogeneity in..., Chicard [/bib_ref] [bib_ref] Whole-exome sequencing of cell-free DNA reveals temporo-spatial heterogeneity and identifies treatmentresistant clones..., Chicard [/bib_ref] [bib_ref] Combining hypermethylated RASSF1A detection using ddPCR with miR-371a-3p testing: An improved panel..., Lobo [/bib_ref] [bib_ref] The feasibility of using liquid biopsies as a complementary assay for copy..., Van Paemel [/bib_ref] In rhabdomyosarcoma, the presence of tumor-derived cfDNA (ctDNA) has been shown to correlate to tumor burden throughout treatment in a few small case series. [bib_ref] Early detection of the PAX3-FOXO1 fusion gene in circulating tumor-derived DNA in..., Eguchi-Ishimae [/bib_ref] ctDNA can be studied using various techniques, using genetic aberrations present in rhabdomyosarcoma. The alveolar subtype has a tumordriving fusion between the PAX3 or PAX7 gene and the FOXO1 gene. Epigenetic analyses revealed distinct methylation profiles in alveolar and embryonal rhabdomyosarcoma, allowing for the classification of cases into fusion-positive versus fusion-negative tumors. [bib_ref] Distinct methylation profiles characterize fusion-positive and fusion-negative rhabdomyosarcoma, Sun [/bib_ref] [bib_ref] Integrated genetic and epigenetic analysis defines novel molecular subgroups in rhabdomyosarcoma, Seki [/bib_ref] Van Paemel et al [bib_ref] Minimally invasive classification of paediatric solid tumours using reduced representation bisulphite sequencing..., Van Paemel [/bib_ref] showed that these distinct methylation patterns can be detected in ctDNA from diagnostic plasma, using cell-free reduced representation bisulphite sequencing (cfRRBS) to correctly classify rhabdomyosarcoma as either the embryonal or alveolar subtype. Copy number aberrations (CNAs) have been found to occur in several chromosomes. [bib_ref] Classification of rhabdomyosarcoma and its molecular basis, Parham [/bib_ref] These can be analyzed in cfDNA by shallow whole genome sequencing (shWGS).Recently, Van Paemel et al [bib_ref] The feasibility of using liquid biopsies as a complementary assay for copy..., Van Paemel [/bib_ref] showed that shWGS data from cfDNA can be complementary to CNA analysis on the primary tumor.
However, cfDNA typically contains a relatively small amount of ctDNA; the remaining cfDNA is derived from healthy cells, which can cause high background noise and limit the ability to detect a tumor-derived signal. [bib_ref] Liquid biopsies come of age: Towards implementation of circulating tumour DNA, Wan [/bib_ref] To overcome this, a tumor-specific assay can be used, such as droplet digital polymerase chain reaction (ddPCR) which is highly sensitive and less expensive. [bib_ref] Performance of four platforms for KRAS mutation detection in plasma cell-free DNA:..., Vessies [/bib_ref] A target suited for analysis by ddPCR is methylation of the tumor-suppressor gene RASSF1A; this gene has been shown to be silenced by methylation in several adult [bib_ref] Clinical utility of RASSF1A methylation in human malignancies, Grawenda [/bib_ref] and pediatric [bib_ref] Inactivation of the RASSF1A in osteosarcoma, Lim [/bib_ref] [bib_ref] Frequent RASSF1A tumour suppressor gene promoter methylation in Wilms' tumour and colorectal..., Wagner [/bib_ref] [bib_ref] RASSF1A methylation indicates a poor prognosis in hepatoblastoma patients, Honda [/bib_ref] [bib_ref] RASSF1A promoter region CpG island hypermethylation in phaeochromocytomas and neuroblastoma tumours, Astuti [/bib_ref] tumors. Moreover, methylated RASSF1A (RASSF1A-M) has been detected in cfDNA in patients with neuroblastoma. [bib_ref] Hypermethylated RASSF1A as circulating tumor DNA marker for disease monitoring in neuroblastoma, Van Zogchel [/bib_ref] [bib_ref] Methylated RASSF1a is the first specific DNA marker for minimal residual disease..., Stutterheim [/bib_ref] Recently, we developed a methylation-specific enzyme-based approach involving ddPCR to detect RASSF1A-M in several pediatric solid tumors, including rhabdomyosarcoma. 14 Here, we report the detection of ctDNA in plasma of patients with rhabdomyosarcoma for diagnostic purposes, such as cfRBBS and shWGS. Furthermore, we study the prognostic potential of RASSF1A-M detection in cfDNA and measure the added value of combining RASSF1A-M ctDNA detection with our rhabdomyosarcoma-specific mRNA panel in paired BM and PB samples.
# Methods
## Patients and sample collection
Plasma samples were collected prospectively from the same cohort described in our previous article, [bib_ref] Improving risk stratification for pediatric patients with rhabdomyosarcoma by molecular detection of..., Lak [/bib_ref] consisting of all patients included in the Dutch Minimal Residual Disease add-on study within the EpSSG RMS2005 trial (EudraCT number: 2005-000217-35) from 2013 through July 2019.
Informed consent was given via the EpSSG RMS2005 trial until 2017. From 2017, consent was provided if the patients/caretakers consented to the collection of samples for biobanking. PB was collected in EDTA tubes (Becton-Dickinson, Franklin Lakes, NJ) and processed within 24 hours. Plasma was obtained by centrifuging the blood samples at 1,375×g for 10 minutes and stored at −20°C until further processing. Matched tumor material was not available.
## Cfrrbs and shwgs
We performed cfRRBS [bib_ref] Minimally invasive classification of paediatric solid tumours using reduced representation bisulphite sequencing..., Van Paemel [/bib_ref] and shWGS 16,32 on cfDNA as described and validated previously. In brief, cfDNA was isolated from 200 µL of plasma as described previously. [bib_ref] The feasibility of using liquid biopsies as a complementary assay for copy..., Van Paemel [/bib_ref] [bib_ref] Minimally invasive classification of paediatric solid tumours using reduced representation bisulphite sequencing..., Van Paemel [/bib_ref] [bib_ref] Shallow whole genome sequencing on circulating cell-free DNA allows reliable noninvasive copy-number..., Van Roy [/bib_ref] For shWGS, the modified copy number profile abnormality score was calculated in order to quantify the copy number tumor burden present in the cfDNA. [bib_ref] The feasibility of using liquid biopsies as a complementary assay for copy..., Van Paemel [/bib_ref] On the basis of 80 healthy volunteers, the level corresponding to a 1% false discovery rate was set at 0.355 for shWGS.
ddPCR Assay for Measuring RASSF1A-M For ddPCR, cfDNA was isolated from plasma samples using the Quick-cfDNA Serum & Plasma kit (Zymo Research, Irvine, CA). The RASSF1A-M ddPCR assay was performed using double digestion with the methylationsensitive restriction enzymes Hhal and Bsh1236I BstUI; Thermo Fisher Scientific, Waltham, MA) using a thermocycler T100 and QX200 reader (Bio-Rad, Hercules, CA) as described previously. 14 The sequences and concentrations of the primers and probes, cycling conditions, and analyses were performed as described previously, with the threshold for RASSF1A-M positivity per sample set at ≥ 14 copies/ml and ≥ 4 RASSF1A-M-positive droplets, as determined in 18 healthy pediatric and 22 adult control plasmas. 14 The percentage of RASSF1A-M was calculated relative to total RASSF1A. On the basis of the plasma volume available
## Context key objective
In pediatric rhabdomyosarcoma, the use of liquid biopsies can assist in generating a more comprehensive view of the molecular landscape of the tumor. We explore different methods for analysis of cell-free DNA (cfDNA) from plasma by cellfree reduced representation bisulphite sequencing, shallow whole genome sequencing, and droplet digital polymerase chain reaction for RASSF1A methylation (RASSF1-M). Furthermore, we study whether combining cfDNA analyses with detection of rhabdomyosarcoma-specific RNA in the cellular fraction of blood and bone marrow (BM) has a complementary value. Knowledge Generated Both cell-free reduced representation bisulphite sequencing and shallow whole genome sequencing have diagnostic potential, whereas the presence of RASSF1A-M at diagnosis correlates to poor survival, especially in patients testing positive for rhabdomyosarcoma-specific RNA in cells from blood and BM. Relevance Analysis of cfDNA through different molecular approaches can be of additional value to current clinical risk stratification, especially the detection of RASSF1A-M in cfDNA and rhabdomyosarcoma-specific RNA in paired blood and BM.
(ranging from 150 mL to 1 mL), different amounts of plasma were used to isolate cfDNA. To correct for variations in the amount of input plasma, cfDNA is reported in ng/mL plasma. In all ddPCR assays, total cfDNA was determined using the reference gene ACTB. Since there was no matched tumor material available, we used data on RASSF1A-M in rhabdomyosarcoma tumors from published data sets from Clay et al, [bib_ref] Methylation profiling reveals novel molecular classes of rhabdomyosarcoma, Clay [/bib_ref] Koelsche et al, [bib_ref] Sarcoma classification by DNA methylation profiling, Koelsche [/bib_ref] and specifically requested data from Seki et al. [bib_ref] Integrated genetic and epigenetic analysis defines novel molecular subgroups in rhabdomyosarcoma, Seki [/bib_ref] Data from Clay and Koelsche were analyzed in R2. We focused on hypermethylation of the promotor region of RASSF1A as this is typically hypermethylated in cancer. [bib_ref] Epigenetically inactivated RASSF1A as a tumor biomarker, Raos [/bib_ref] We calculated the mean beta value and report the range of the beta values. [bib_ref] The relationship between RASSF1A promoter methylation and thyroid carcinoma: A meta-analysis of..., Niu [/bib_ref] [bib_ref] Identification of RASSF1A promoter hypermethylation as a biomarker for hepatocellular carcinoma, Xu [/bib_ref] Detection of Rhabdomyosarcoma-Specific mRNA Using an RNA Panel
Rhabdomyosarcoma-specific mRNA was detected in the cellular fractions of matched diagnostic patient PB and BM samples using our previously reported 11-marker RNA panel. [bib_ref] Improving risk stratification for pediatric patients with rhabdomyosarcoma by molecular detection of..., Lak [/bib_ref] The RNA panel was considered positive if either PB or BM was positive.
# Statistical analysis
Statistical analyses were performed using SPSS version 23.
Figures were generated using GraphPad Prism version 8. The correlation between continuous variables was determined using Pearson's test. Continuous variables were analyzed using the nonparametric Mann-Whitney U test, and two or more groups were analyzed using the Kruskal-Wallis test. Independence between two categorical variables was determined using the nonparametric Pearson chi-square test. Event-free survival (EFS) and overall survival (OS) were estimated using the Kaplan-Meier approach, and differences in survival were analyzed using the log-rank test. Differences were considered significant at P , .05.
# Results
## Patient and sample characteristics
We collected a total of 152 plasma samples from 65 patients, treated according to the EpSSG RMS2005 protocol; diagnostic plasma samples were available for 57 patients. The patient characteristics, assigned risk group and tumor histology, are summarized in [fig_ref] TABLE 1: Patient Abbreviation [/fig_ref]. The median followup was 4.21 years (range, 0.34-10.60 years).
## Diagnostic potential of various molecular techniques for detecting ctdna
First, total cfDNA levels at diagnosis were determined by measuring ACTB using ddPCR for all samples. No significant differences in total cfDNA levels were observed between patients with respect to tumor histology, risk group, localized versus metastatic disease, tumor size, or eventfree survival (Data Supplement). Next, we examined the feasibility to detect ctDNA using cfRRBS, shWGS, and ddPCR ( Methylation profiling for diagnostic classification. As negative control, cfRRBS was performed on 31 samples from Methylated RASSF1A. Using data from Clay et al, [bib_ref] Methylation profiling reveals novel molecular classes of rhabdomyosarcoma, Clay [/bib_ref] Koelsche et al, [bib_ref] Sarcoma classification by DNA methylation profiling, Koelsche [/bib_ref] [fig_ref] FIG 1: Methylated RASSF1A [/fig_ref]. Finally, we found no difference in total cfDNA levels (ACTB) between RASSF1A-M-positive and RASSF1A-M-negative cases (P = .96; [fig_ref] FIG 1: Methylated RASSF1A [/fig_ref].
## Cell-free rassf1a-m correlates with poor outcome
We examined whether the detection of ctDNA in 57 diagnostic plasma samples was associated with patient outcome. Eleven of 21 (52.3%) RASSF1A-M-positive patients suffered from an event.
The 5-year EFS rate was 46.2% for the RASSF1A-M-positive patients, compared with 84.9% for the RASSF1A-Mnegative patients (P = .001 [fig_ref] FIG 2: Survival outcome defined by detection of cell-free methylated RASSF1A [/fig_ref] , and the 5-year OS rate was 55.7% for the RASSF1A-M-positive patients, compared with 100% for the RASSF1A-M-negative patients (P , .001; [fig_ref] FIG 2: Survival outcome defined by detection of cell-free methylated RASSF1A [/fig_ref]. The prognostic value of detecting RASSF1A-M at diagnosis was attributed almost exclusively to patients with metastasized disease [fig_ref] FIG 2: Survival outcome defined by detection of cell-free methylated RASSF1A [/fig_ref] , Data Supplement).
In 27 samples, both shWGS and RASSF1A-M were performed (Data Supplement). In six patients, shWGS was positive while RASSF1A-M was negative, and only one patient suffered from an event, while 6 of 10 double-positive patients suffered from an event, suggesting that the presence of both RASSF1A-M and ctDNA by shWGS may be more prognostic than detection of ctDNA by shWGS alone.
We next examined whether combining RASSF1A-M detection with detection of rhabdomyosarcoma-specific mRNA (on the basis of our previously published mRNA panel [bib_ref] Improving risk stratification for pediatric patients with rhabdomyosarcoma by molecular detection of..., Lak [/bib_ref] respectively). In the multivariable model, only RASSF1A-M, RNA panel, and tumor size larger than 5 cm had a significant effect on outcome. The known EpSSG RMS2005 risk group classification, metastatic disease, alveolar subtype, over 10 years of age, and lymph node involvement were not significantly associated with outcome in our multivariable model. Finally, OS could not be analyzed due to the low number of events in this cohort. and two at second relapse); no samples at initial diagnosis were available for these five patients. After initiating relapse therapy, all subsequent samples from these patients were RASSF1A-M-negative. The sample taken from the patient at progressive disease (patient RMS133) was RASSF1A-M-negative, and no previous plasma samples were available for this patient.
## Rassf1a-m during treatment and clinical follow-up
# Discussion
On the basis of our findings, we propose that each cfDNAbased technique can address a specific clinical need, ranging from assisting at initial tumor diagnosis to finetuning of risk stratification. In our cohort, cfRRBS proved its potential as a highly sensitive method for identifying rhabdomyosarcoma-derived cfDNA at initial diagnosis, and the majority was classified correctly as embryonal. Van Paemel et al 21 found that cfRRBS was also able to correctly identify alveolar ctDNA. Thus, cfRRBS can provide added value at initial diagnosis, particularly if the ability to perform a tumor biopsy is restricted by clinical features such as tumor location or the patient's condition and when the ability to distinguish between other types of pediatric solid tumors is important. [bib_ref] Minimally invasive classification of paediatric solid tumours using reduced representation bisulphite sequencing..., Van Paemel [/bib_ref] We detected CNAs in 53.3% of samples analyzed by shWGS, mostly metastatic cases. On the basis of the literature, CNAs are present in nearly all fusion-negative rhabdomyosarcomas [bib_ref] Integrated genetic and epigenetic analysis defines novel molecular subgroups in rhabdomyosarcoma, Seki [/bib_ref] and in approximately one third of all fusion-positive rhabdomyosarcomas. 39, [bib_ref] A study of alveolar rhabdomyosarcoma copy number alterations by single nucleotide polymorphism..., Lynn [/bib_ref] We detected CNAs in the cfDNA of only half of the patients with fusionnegative tumors. This relatively low rate may have been due in part to contamination of the cfDNA with genomic DNA, as the protocol for drawing and storing blood was not standardized, which can lower the sensitivity to detect CNA. [bib_ref] The feasibility of using liquid biopsies as a complementary assay for copy..., Van Paemel [/bib_ref] Van Paemel et al noted that performing shWGS on cfDNA can provide additional value with respect to analyzing CNAs in the primary tumor, resulting in a more complete overview of the patient's genetic landscape and bypassing any potential heterogeneity within the tumor and/or metastatic lesions. This is important to consider when designing further studies.
On the basis of the previous reports, demonstrating feasibility to use RASSF1A-M ddPCR as a tumor-specific marker with a high specificity due to extremely low background in plasma from healthy controls, 14,15 we studied RASSF1A-M ddPCR in cfDNA of patients with rhabdomyosarcoma. One of the limitations of this study was the absence of paired primary tumor samples. However, the presence of RASSF1A-M, as extracted from data published by several groups, [bib_ref] Integrated genetic and epigenetic analysis defines novel molecular subgroups in rhabdomyosarcoma, Seki [/bib_ref] [bib_ref] Methylation profiling reveals novel molecular classes of rhabdomyosarcoma, Clay [/bib_ref] [bib_ref] Sarcoma classification by DNA methylation profiling, Koelsche [/bib_ref] indicated the potential to detect RASSF1A-M in primary tumors, with admittedly a large variation in the level of RASSF1A-M. Still, for the patients in our cohort who were RASSF1A-M-negative, on the basis of cfDNA obtained at diagnosis, we were unable to determine whether this was due to absence of RASSF1A methylation or no detectable ctDNA. This is underlined by the 18 samples testing negative for RASSF1A-M, in which ctDNA was detected by cfRRBS and/or shWGS. Future studies should include matching tumor material to establish the contribution of different approaches for cfDNA analysis. Nonetheless, we were able to detect RASSF1A-M in cfDNA in 36% of diagnostic samples and found a strong correlation between RASSF1A-M positivity and event-free survival and OS. Importantly, this predictive value was obtained almost exclusively in the group of patients with metastatic disease. This finding might suggest that more aggressive tumors contain methylated RASSF1A and deserves further investigations in a follow-up study, including matching primary tumor material. Interestingly, in the samples that were tested by both shWGS and RASSF1A-M, the results suggest that detection of ctDNA by both methods may be more prognostic than detection of ctDNA by shWGS alone. This should be studied further in a larger cohort.
As we previously showed rhabdomyosarcoma-specific RNA detection in PB and/or BM at diagnosis to detect additional disseminated disease and to correlate with the outcome, [bib_ref] Improving risk stratification for pediatric patients with rhabdomyosarcoma by molecular detection of..., Lak [/bib_ref] we now showed that combining mRNA and ctDNA (RASSF1A-M) in paired diagnostic samples identifies patients with very good and very poor outcome. Our multivariable analysis revealed that combining the cfDNA RASSF1A-M assay with rhabdomyosarcoma-specific RNA detection in PB and BM samples provides an even better tool for discriminating between low-risk patients and patients with a poor prognosis. Given the relatively small number of patients in our cohort, however, we were unable to investigate the effect of adding both RASSF1A-M and the RNA panel to established prognostic factors, particularly in the EpSSG RMS2005 risk group; nevertheless, our results can form a starting point for future studies involving a prospective cohort.
An interesting finding from our study is the dynamics of ctDNA. Before our study, we hypothesized that ctDNA would still be present during primary treatment and decrease slowly, tracking the decrease in tumor burden.
However, in our rhabdomyosarcoma cohort, we found that most of the samples were negative for ctDNA after the first course of chemotherapy. This rapid transition to a ctDNAnegative state is consistent with the results reported by Klega et al 18 who found that most samples were negative for ctDNA before the second course of chemotherapy. Thus, an interesting question is whether performing earlier sampling and obtaining multiple samples during the first 2 weeks after the start of treatment would reveal the presence of ctDNA and-if so-would lead to the development of a prognostic marker, similar to the marker for minimal residual disease developed for use in leukemia. [bib_ref] Early responses to chemotherapy of normal and malignant hematologic cells are prognostic..., Laughton [/bib_ref] [bib_ref] Prognostic value of minimal residual disease in acute lymphoblastic leukaemia in childhood, Van Dongen [/bib_ref] In conclusion, we demonstrate the feasibility to study ctDNA in pediatric rhabdomyosarcoma by different approaches. The choice of a given technique will depend on whether the underlying question is diagnostic or prognostic. We show that the presence of methylated RASSF1A in cfDNA is associated with poor outcome and can be used to improve risk stratification at diagnosis. Furthermore, we show that combining detection of methylated RASSF1A in plasma with analysis of tumor-specific RNA in blood and bone marrow identified patients with good versus poor outcome.
## Affiliations
[fig] D: Outcome of RASSF1A-M ddPCR [/fig]
[fig] FIG 1: Methylated RASSF1A (RASSF1A-M) in diagnostic plasma samples of patients with rhabdomyosarcoma. The percentage of cell-free methylated RASSF1A (RASSF1A-M) is calculated according to total RASSF1A copies at diagnosis in patients: (A) with different subtypes, (B) with localized and metastatic disease, and (C) plotted against tumor volume at diagnosis. (D) Level of cfDNA (quantified by beta-actin [ACTB]) at diagnosis in plasma samples with detectable RASSF1A-M and with no detectable RASSF1A-M; note that the y-axis is plotted on a log scale. In this figure, each symbol represents an individual patient, and the red horizontal lines represent the median values. Tumor size was determined by MRI, CT scan, or ultrasonography. ARMS, alveolar rhabdomyosarcoma; cfDNA, cell-free DNA; CT, computed tomography; ddPCR, droplet digital polymerase chain reaction; ERMS, embryonal rhabdomyosarcoma; MRI, magnetic resonance imaging; NOS, not otherwise specified. [/fig]
[fig] FIG 2: Survival outcome defined by detection of cell-free methylated RASSF1A (RASSF1A-M) at diagnosis. (A and B) EFS and OS, respectively, of patients with no detectable methylated RASSF1A in the diagnostic plasma (RASSF1A-Mneg; n = 36) and patients with detectable methylated RASSF1A in the diagnostic plasma (RASSF1A-Mpos; n = 21). (C and D) EFS and OS of RASSF1A-M-negative patients (n = 8) and RASSF1A-Mpositive patients (n = 12) with metastatic disease. Shown below each plot is the number of patients at each time point and 5-year survival with the 95% CI. cfDNA, cell-free DNA; EFS, event-free survival; OS, overall survival. [/fig]
[fig] FIG 3: Survival outcome defined by detection of cell-free methylated RASSF1A (RASSF1A-M) from plasma and rhabdomyosarcoma-specific RNA in blood and bone marrow at diagnosis. (A and B) EFS and OS of 56 patients on the basis of the absence or presence of rhabdomyosarcoma-specific RNA (RNA-negative and RNA-positive, respectively) combined with RASSF1A-M status. Shown below each plot is the number of patients at each time point and 5-year survival with the 95% CI. OS, overall survival; EFS, event-free survival. [/fig]
[table] TABLE 1: Patient Abbreviation: IRS, intergroup rhabdomyosarcoma studies.healthy controls, all classified correctly as normal (Data Supplement). We applied cfRRBS to diagnostic samples from 24 patients with the embryonal subtype, 1 with botryoid subtype, and 1 with alveolar subtype, successfully detecting rhabdomyosarcoma DNA in 24 of these 26 samples (92.3% of cases). Twenty of these samples were correctly identified as embryonal tumors. Three cases with embryonal histology were classified as alveolar, one case of botryoid rhabdomyosarcoma was classified as embryonal, and no tumor DNA was detected in two samples (one alveolar and one embryonal).CNAs. We performed shWGS on 30 plasma samples and obtained a median copy number profile abnormality score of 0.35 (range, 0.27-3.94; Data Supplement). In three cases (two embryonal and one alveolar), the analysis failed (Table 2, Data Supplement). Twelve embryonal cases (7 of 12 metastatic) and four alveolar (all metastatic) cases had CNAs, while 10 embryonal cases and one botryoid case had no CNA. Most CNAs were detected in patients with metastatic disease, and 7 of 16 (43.8%) patients with detectable CNAs suffered from an event. [/table]
[table] TABLE 2: Overview of the Results of Different Approaches on Cell-Free [/table]
[table] TABLE 3: HRs With 95% CI Estimated With a Multivariable Cox Abbreviation: HR, hazard ratio.Indicates significance at P , .05. [/table]
|
Prospective applications of microRNAs in oral cancer
MicroRNAs (miRNAs) are non-coding RNA molecules that are generally encoded by endogenous genes and exert suppressive effects on post-transcriptional regulation of their target genes by translation repression or degradation of mRNA. This subsequently mediates activation or blocking of downstream signaling pathways associated with oral malignancies. Aberrant levels of certain miRNAs have been identified in cell experiments, clinical carcinomatous specimens, saliva, serum or plasma samples of patients with oral malignancies. miRNAs are associated with multiple aspects of oral cancer, including tumor growth, cellular proliferation, apoptosis, migration, invasion, metastasis, glycometabolism, radiosensitivity and chemosensitivity. miRNAs have the potential to be used in clinical applications as minimally invasive or non-invasive tools for early diagnosis and prognosis by the detection of serum, plasma and saliva levels, and may provide a new ancillary or additional reference index of traditional pathological grading and clinical staging. Furthermore, miRNAs may be used as prognostic biomarkers or targets for novel therapies for oral cancer.
# Introduction
Oral cancer is a common and fatal malignancy among head and neck malignant neoplasms; the number of new cases of oral cancer globally was 354,864 in 2018 [bib_ref] Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for..., Bray [/bib_ref]. At present, principal treatments of oral cancer include extensive exeresis of the primary carcinoma, with or without neck dissection, and pre-or postoperative adjuvant chemotherapy and radiotherapy [bib_ref] Cancer of the oral cavity, Montero [/bib_ref]. However, the overall 5-year survival rate of patients with oral cancer was 65%, and the overall 5-year survival rate of patients with advanced oral cancer was as low as 27% between 2007 and 2013 in the USA. Despite the application of reconstructive radical resection and postoperative radiotherapy or chemotherapy, the 5-year survival rate of patients with terminal oral cancer has not improved effectively over the past years [bib_ref] Neoadjuvant chemotherapy for locally advanced squamous carcinoma of oral cavity: A pilot..., Sadighi [/bib_ref] [bib_ref] Preoperative chemotherapy in advanced resectable OCSCC: Long-term results of a randomized phase..., Bossi [/bib_ref]. Furthermore, the dysphagia, maxillofacial malformation and dysarthria induced by the aforementioned therapies negatively affect the quality of life and psychology of patients [bib_ref] Impact of oral cancer on quality of life, Valdez [/bib_ref]. Therefore, there is a requirement to identify more effective treatment strategies to improve the survival rate and reduce complications of patients with oral cancer.
MicroRNAs (miRNAs/miRs) have attracted increasing attention over the past years as their roles in malignant tumors, where they regulate target genes and downstream signaling pathways, have been recognized [bib_ref] MicroRNA therapeutics: Towards a new era for the management of cancer and..., Rupaimoole [/bib_ref]. Non-coding RNA, composed of 18-22 nucleotides, silences corresponding target genes by binding to the 3'-untranslated regions (3'UTRs) of mRNA to mediate the biological behavior of cancer cells [bib_ref] Non-coding RNA networks in cancer, Anastasiadou [/bib_ref]. Dysregulation of miRNA in hepatoma, lymphoma and colorectal, ovarian and pancreatic cancer has been previously reported [bib_ref] Reversal of cisplatin resistance by microRNA-139-5p-independent RNF2 downregulation and MAPK inhibition in..., Chen [/bib_ref] [bib_ref] MiRNA-17 encoded by the miR-17-92 cluster increases the potential for steatosis in..., Gong [/bib_ref] [bib_ref] MicroRNA-451a regulates colorectal cancer proliferation in response to radiation, Ruhl [/bib_ref] [bib_ref] MicroRNA-221 induces autophagy through suppressing HDAC6 expression and promoting apoptosis in pancreatic..., Yang [/bib_ref] [bib_ref] Epstein-Barr virus-encoded EBNA2 alters immune checkpoint PD-L1 expression by downregulating miR-34a in..., Anastasiadou [/bib_ref] [bib_ref] The nefarious nexus of noncoding RNAs in cancer, Anastasiadou [/bib_ref]. In oral carcinoma, miRNAs are associated with oral carcinomatous cell proliferation, apoptosis, invasion, metastasis, epithelial-mesenchymal transition (EMT), chemoresistance, radioresistance and cell cycle arrest. The abnormal expression of miRNA detected in tumor, serum and saliva samples obtained from patients with oral cancer has clinical significance in prognosis prediction and the development of effective treatments [bib_ref] MicroRNA-29a upregulates MMP2 in oral squamous cell carcinoma to promote cancer invasion..., Lu [/bib_ref] [bib_ref] Downregulation of miR-153 contributes to epithelial-mesenchymal transition and tumor metastasis in human..., Xu [/bib_ref] [bib_ref] Serum, plasma and saliva biomarkers for head and neck cancer, Arantes [/bib_ref] [bib_ref] The role of long non-coding RNA ANRIL in the carcinogenesis of oral..., Chai [/bib_ref]. The present review discusses recent developments with regard to miRNAs and their potential clinical applications in oral cancer.
## Biogenesis and function of mirna
The biogenesis of miRNA involves processes in the nucleus and cytoplasm. In the nucleus, primary miRNA (pri-miRNA), a type of miRNA that contains 300-1,000 nucleotides, is generally issued from the miRNA gene and is transcribed by RNA polymerase II. Subsequently the long transcript, pri-miRNA, is cleaved by the drosha ribonuclease III (DROSHA)/DiGeorge syndrome chromosomal region 8 (DGCR8) complex into a ~70-nt structure termed precursor miRNA (pre-miRNA), which has lost a 7-methyl guanine nucleoside in the 5'-capped end and a 3'poly-(A) tail, but has a conserved stem-loop. Subsequently, the aforementioned stem-loop is sheared by RNA III enzyme Dicer and double-strand RNA-binding domain protein after the GTP-binding nuclear protein Ran/exportin-5 (RanGTP/XPO5) complex carries pre-miRNA to the cytoplasm, to form a double-strand miRNA molecule consisting of 22 nucleotides. Transactivation response element RNA binding protein (TRBP) recruits the mature miRNA to RNA-induced silencing complex (RISC) together with argonaute 2 (Ago2) that has endonuclease activity. A single strand of the double-strand miRNA is conserved in the RISC, and another one is degraded [bib_ref] Regulation of microRNA biogenesis, Ha [/bib_ref] [bib_ref] The nuclear RNase III Drosha initiates microRNA processing, Lee [/bib_ref]. Partial or complementary pairing between the RISC and the 3'-UTR of target mRNA functionally plays a role in repressing the translation of mRNA or degrading target mRNA [bib_ref] TRBP recruits the Dicer complex to Ago2 for microRNA processing and gene..., Chendrimada [/bib_ref] [bib_ref] Characterization of DGCR8/Pasha, the essential cofactor for Drosha in primary miRNA processing, Yeom [/bib_ref]. The mechanisms are presented in Figs. 1 and 2.
## Mirnas in oncogenesis, diagnosis and prognosis of oral cancer
Previous studies suggested that common potentially malignant disorders and precancerous conditions, including oral leukoplakia (OL) and oral lichen planus (OLP), were correlated with aberrant miRNA, but the mechanism of malignant transformation by miRNA remains unclear [bib_ref] Genetic variants in microRNA-146a (C>G) and microRNA-1269b (G>C) are associated with the..., Chen [/bib_ref] [bib_ref] MicroRNAs-208b-3p, 204-5p, 129-2-3p and 3065-5p as predictive markers of oral leukoplakia that..., Philipone [/bib_ref]. However, investigating unknown neoplastic transformations for precancerous lesions or conditions may provide novel insight into the mechanisms of tumorigenesis. In addition, the significant differential expression of miRNA between normal tissue, potential malignant disorders and oral cancerous specimens indicates that miRNA may be used as an independent prognostic marker [bib_ref] Evaluating the accuracy of microRNA-27b and microRNA-137 as biomarkers of activity and..., Aghbari [/bib_ref]. Ultimately, a minimally invasive or non-invasive method, such as the detection of miRNA in saliva or serum, may be applied in preoperative prediction and postoperative follow-up [bib_ref] Exploiting salivary miR-31 as a clinical biomarker of oral squamous cell carcinoma, Liu [/bib_ref] [bib_ref] Salivary microRNAs in oral cancer, Zahran [/bib_ref] [bib_ref] Prognostic significance of altered miRNA expression in whole blood of OSCC patients, Ries [/bib_ref]. Indeed, a number of recent studies demonstrated that the aberrant expression of different miRNAs, such as miR-195-5p, miR-375, miR-143, miR-26b, miR-155-5p and miR-483-5p, was associated with oral cancer [bib_ref] MiR-195-5p suppresses the proliferation, migration, and invasion of oral squamous cell carcinoma..., Wang [/bib_ref] [bib_ref] MicroRNA-375 inhibits growth and enhances radiosensitivity in oral squamous cell carcinoma by..., Zhang [/bib_ref] [bib_ref] MiR-26b is downregulated in human tongue squamous cell carcinoma and regulates cell..., Cao [/bib_ref] [bib_ref] MicroRNA-155-5p is associated with oral squamous cell carcinoma metastasis and poor prognosis, Baba [/bib_ref] [bib_ref] Serum miR-483-5p: A novel diagnostic and prognostic biomarker for patients with oral..., Xu [/bib_ref]. Therefore, miRNAs may have prognostic and diagnostic value in oral cancer and may serve as targets for novel therapeutic strategies. Latent mechanism of tumorigenesis by miRNAs. Previous studies identified that the miR-31 expression level in oral potential malignant disorder (OPMD) is higher than that in normal oral mucosa, which is correlated with higher expression of vascular endothelial growth factor and lower expression of E-cadherin in OPMD. miR-31 expression was further upregulated in patients with recurrent OPMD and malignant transformation [bib_ref] MicroRNA-31 upregulation predicts increased risk of progression of oral potentially malignant disorder, Hung [/bib_ref]. Another previous study suggested that aberrant expression of miR-200c was associated with oral submucous fibrosis (OSF). The overexpression of miR-200c inhibited collagen gel contraction and migration and invasion of fibrotic buccal mucosal fibroblasts induced by arecoline via inhibition of zinc finger E-box binding homeobox 1 (ZEB1). Additionally, reverse transcription-quantitative PCR (RT-qPCR) analysis revealed that miR-200c was downregulated in 25 OSF samples compared with 25 normal mucosae [bib_ref] MiR-200c inhibits the arecoline-associated myofibroblastic transdifferentiation in buccal mucosal fibroblasts, Lu [/bib_ref]. Brito et al [bib_ref] Relationship between microRNA expression levels and histopathological features of dysplasia in oral..., Brito [/bib_ref] concluded that higher expression of miR-21 in OL was associated with increased mitotic figures, incremental nuclear/cytoplasmic ratio and hyperchromasia. Nylander et al [bib_ref] Changes in miRNA expression in sera and correlation to duration of disease..., Nylander [/bib_ref] found that miR-21 was upregulated in 30 patients diagnosed with multifocal OLP compared with 10 healthy subjects, and in agreement, another study demonstrated that upregulation of miR-21 served a tumor-promoting role in oral cancer and upregulation of miR-21 was observed in 60 of 79 individuals with the disease (41). Aghbari et al [bib_ref] Evaluating the accuracy of microRNA-27b and microRNA-137 as biomarkers of activity and..., Aghbari [/bib_ref] identified that miR-27b and miR-137 levels were downregulated in tissue and saliva samples of patients with OLP compared with those in normal controls. Among OLP subgroups, it was demonstrated that miR-137 exhibited the lowest expression level in the erosive type, suggesting that it serve as a biomarker for monitoring potential malignant transformation. The level of miR-375 in progressive lesions was significantly downregulated compared with that in non-progressive control lesions, and miR-375 expression was significantly downregulated in tissues following the transformation of premalignant lesions (including verrucous hyperplasia and verrucopapillary hyperkeratosis) into carcinoma, by comparison of premalignant lesions and oral carcinoma in situ. While only 2-3 and 0.4-2% of patients with OL and OLP, respectively, exhibit malignant transformation [bib_ref] Recurrence patterns and management of oral cavity premalignant lesions, Arnaoutakis [/bib_ref] [bib_ref] Oral potentially malignant disorders: An overview of more than 20 entities, Mortazavi [/bib_ref] , further investigation of the roles of miR-21, miR-27b, miR-137, miR-200c and miR-375 may provide novel insights into pathways involved in the development of oral cancer.
Aberrant expression of miRNA in oral cancer tissues. Dysregulation of specific miRNAs has been previously reported in oral cancer. Wang et al [bib_ref] MiR-195-5p suppresses the proliferation, migration, and invasion of oral squamous cell carcinoma..., Wang [/bib_ref] reported that miR-195-5p was significantly downregulated in 40 oral cancerous tissues compared with non-tumor tissues. Another study reported a distinct downregulation of miR-375 in 44 cancerous tissues compared with that in normal mucosae [bib_ref] MicroRNA-375 inhibits growth and enhances radiosensitivity in oral squamous cell carcinoma by..., Zhang [/bib_ref]. Furthermore, a previous study revealed that miR-143 was downregulated in cancerous tissues compared with that in corresponding adjacent non-cancerous tissues in 81.6% (40/49) of patients. miR-802 was downregulated in 60.0% (12/20) of tongue squamous cell carcinoma (TSCC) cases compared with that in normal tissues [bib_ref] MicroRNA-802 plays a tumour suppressive role in tongue squamous cell carcinoma through..., Wu [/bib_ref]. The expression levels of miR-137 (n=25) and miR-204-5p (n=52) were downregulated in oral cancer samples compared with those in matched normal tissues [bib_ref] MicroRNA-137 suppresses tongue squamous carcinoma cell proliferation, migration and invasion, Sun [/bib_ref] [bib_ref] MiR-204-5p regulates cell proliferation and metastasis through inhibiting CXCR4 expression in OSCC, Wang [/bib_ref]. Moreover, upregulation of specific miRNAs was observed in oral cancer tissues. Upregulation of miR-183 was identified in 68.3% (41/60) of TSCC tissues compared with adjacent non-cancerous tissues [bib_ref] MiR-183 and miR-21 expression as biomarkers of progression and survival in tongue..., Supic [/bib_ref]. The miR-373-3p expression level in oral cancerous tissues (n=63) was increased compared with that in adjacent non-cancerous tissues [bib_ref] MiR-373-3p targets DKK1 to promote EMT-induced metastasis via the Wnt/β-catenin pathway in..., Weng [/bib_ref]. The miR-155 expression level was upregulated in oral squamous cell carcinoma (OSCC) tissues (n=46) compared with that in normal oral mucosa, and the expression level was increased with increasing Tumor-Node-Metastasis (TNM) stage [bib_ref] MiR-155 regulates oral squamous cell carcinoma Tca8113 cell proliferation, cycle, and apoptosis..., Fu [/bib_ref]. miR-31, miR-182, miR-200a and miR-141 were significantly upregulated in cancerous tissues in 10 patients compared with adjacent non-cancerous tissues [bib_ref] MiR-182 activates the Ras-MEK-ERK pathway in human oral cavity squamous cell carcinoma..., Wang [/bib_ref]. The expression of miR-24 was significantly increased in TSCC tissues of 84 patients compared with adjacent non-cancerous tissues (51). Liu et al [bib_ref] MiR-1275 promotes cell migration, invasion and proliferation in squamous cell carcinoma of..., Liu [/bib_ref] demonstrated that 67% (10/15) of patients with primary oral cancer had increased miR-1275 expression in tumor tissues compared with that in adjacent tissues.
## Aberrant expression of mirna in serum, plasma and saliva.
Previous studies showed that RNA in saliva is protected from degradation by binding to macromolecules such as apoptotic bodies and RISC, a mechanism also observed in plasma and serum RNAs (53-55). Park et al [bib_ref] Salivary microRNA: Discovery, characterization, and clinical utility for oral cancer detection, Park [/bib_ref] analyzed saliva by immunoblotting analysis using an antibody against Ago2, and demonstrated that Ago2 was present in saliva, where it may confer stability to miRNAs. Furthermore, Park et al [bib_ref] Salivary microRNA: Discovery, characterization, and clinical utility for oral cancer detection, Park [/bib_ref] found lower levels of miR-125a and miR-200a in the saliva of patients with oral cancer (n=12) compared with those in healthy controls (n=12), suggesting that the aforementioned miRNAs may serve as stable biomarkers of the disease. Liu et al [bib_ref] Exploiting salivary miR-31 as a clinical biomarker of oral squamous cell carcinoma, Liu [/bib_ref] reported that the level of miR-31 in the saliva of patients with OSCC (n=45) prior to surgery was significantly increased compared with that in healthy subjects (n=24). Moreover, the miR-31 level in saliva samples was higher compared with that in plasma samples. The upregulation of miR-31 was detected with high sensitivity even in very small tumors, and the ability to detect miR-31 levels in the saliva of patients with small tumors was not significantly different compared with patients with advanced tumors, suggesting that salivary miR-31 may be utilized to detect and diagnose oral cancer lesions in high-risk populations. Zahran et al [bib_ref] Salivary microRNAs in oral cancer, Zahran [/bib_ref] reported a significant upregulation in salivary miR-21 and miR-184 levels in patients with oral cancer compared with those in healthy controls. Specifically, a four-fold increase in miR-21, with 65% specificity and 65% sensitivity, and a three-fold increase in miR-184, with 75% specificity and 80% sensitivity, were observed. The expression of salivary miR-145 was significantly decreased in patients with OSCC compared with clinically healthy controls, with 70% specificity and 60% sensitivity. Ries et al [bib_ref] Prognostic significance of altered miRNA expression in whole blood of OSCC patients, Ries [/bib_ref] reported that miR-3651 and miR-494 levels were upregulated, while the miR-186 level was significantly downregulated in whole blood samples of patients with recurrent tumors compared with non-recurrent controls. Therefore, miRNAs may be promising candidates for the development of diagnostic tools for oral cancer.
miRNAs as feasible biomarkers of pathology, metastasis and prognosis. Certain miRNAs are known to be downregulated in tumor tissues compared with noncancerous tissues. Lower levels of miR-195-5p were associated with higher pathological differentiation grade [bib_ref] MiR-195-5p suppresses the proliferation, migration, and invasion of oral squamous cell carcinoma..., Wang [/bib_ref]. In comparison with patients without lymph node metastases (n=26), patients with lymph node metastases (n=18) had significantly downregulated miR-375. Furthermore, the overall survival of patients with oral cancer with low miR-375 expression (n=19) was lower than that of patients with high miR-375 expression (n=25) (32). Cao et al [bib_ref] MiR-26b is downregulated in human tongue squamous cell carcinoma and regulates cell..., Cao [/bib_ref] revealed that advanced clinical stage and large tumor size of oral cancer were associated with low miR-26b expression. The 5-year survival rates of patients with low and high miR-26b levels were 26.7 and 53.3%, respectively.
On the other hand, certain miRNAs are upregulated in tumor tissues compared with adjacent non-cancerous tissues. Upregulation of miR-183 in patients with oral cancer markedly shortened the overall survival time and increased the risk of poor prognosis by 5.666 times. Upregulation of miR-21 resulted in a higher risk of short survival time [bib_ref] MiR-183 and miR-21 expression as biomarkers of progression and survival in tongue..., Supic [/bib_ref]. The expression of miR-373-3p was higher in primary tumors with metastases compared with that in tumors with no metastases [bib_ref] MiR-373-3p targets DKK1 to promote EMT-induced metastasis via the Wnt/β-catenin pathway in..., Weng [/bib_ref]. The upregulation of miR-24 was associated with advanced clinical stage in patients with oral cancer [bib_ref] MiR-24 promotes the proliferation, migration and invasion in human tongue squamous cell..., Zhao [/bib_ref]. A positive correlation between a high miR-155-5p level and cervical lymphatic metastases was observed, and the survival analysis of carcinomatous recurrence and metastasis identified an association between high miR-155-5p expression and a poor survival rate (n=73); miRNA-155-5p may be considered as a specific factor resulting in a worse prognosis [bib_ref] MicroRNA-155-5p is associated with oral squamous cell carcinoma metastasis and poor prognosis, Baba [/bib_ref].
Serum miR-483-5p was higher in patients with oral cancer (n=101) compared with healthy controls (n=103); the survival rates of patients with high miR-483-5p serum level (n=43; >3.23-fold higher compared with healthy controls) were decreased compared with that of patients with lower miR-483-5p serum levels (n=42; <3.23-fold higher compared with healthy controls), and multivariate analyses for overall survival suggested that a high miR-483-5p serum level was an independent prognostic indicator [bib_ref] Serum miR-483-5p: A novel diagnostic and prognostic biomarker for patients with oral..., Xu [/bib_ref]. Furthermore, patient follow-up revealed that patients with higher blood miR-372 levels had more extensive primary tumors, a greater tendency of node metastases, a more terminal stage and higher mortality rates [bib_ref] MiR-372 inhibits p62 in head and neck squamous cell carcinoma in vitro..., Yeh [/bib_ref]. Additionally, miR-372 was downregulated in plasma and saliva among postoperative patients compared with preoperative patients [bib_ref] MiR-372 inhibits p62 in head and neck squamous cell carcinoma in vitro..., Yeh [/bib_ref]. By measuring salivary miR-31, it was determined that 86.4% (19/22) of patients exhibited a significant decrease in miR-31 levels following tumor resection [bib_ref] Exploiting salivary miR-31 as a clinical biomarker of oral squamous cell carcinoma, Liu [/bib_ref]. miR-372 and miR-31 may therefore serve as biomarkers for the evaluation of surgical efficacy [bib_ref] Exploiting salivary miR-31 as a clinical biomarker of oral squamous cell carcinoma, Liu [/bib_ref] [bib_ref] MiR-372 inhibits p62 in head and neck squamous cell carcinoma in vitro..., Yeh [/bib_ref]. Sun et al [bib_ref] Association of decreased expression of serum miR-9 with poor prognosis of oral..., Sun [/bib_ref] suggested that serum miR-9 is an independent prognostic factor for oral cancer, as downregulated miR-9 was associated with lymph node metastases, advanced TNM stage and poor prognosis. Patients with low serum miR-9 expression had a worse disease-free survival rate compared with patients with high miR-9 expression (26.5 and 54.6%, respectively). The overall survival rate of patients with low and high miR-9 expression was 42.9 and 67.3%, respectively. By determining levels of miRNAs in the serum or saliva of patients, miRNAs may be used in minimally or non-invasive methods to predict lymph node metastasis and assess the prognosis of patients with oral cancer. [fig_ref] Table I: Dysregulation of miRNAs associated with oral cancer detected in saliva, blood, serum... [/fig_ref] presents the aberrant levels of other miRNAs in the saliva, blood, serum and plasma (28-30,59-68).
## Partial subtypes of oral cancer and mirnas
Mucoepidermoid carcinoma. As a salivary gland-derived malignancy, mucoepidermoid carcinoma may occur in the oral cavity. Shin et al [bib_ref] High expression of microRNA-127 is involved in cell cycle arrest in MC-3..., Shin [/bib_ref] reported that the overexpression of miR-127-3p led to an increase in the number of cells in the G 1 phase, indicating that miR-127-3p resulted in G 1 /S cell cycle arrest in vitro. miR-127-3p-induced cell cycle arrest in mucoepidermoid carcinoma MC-3 cells was associated with the increase of cyclin-dependent kinase inhibitor 1A and interferon α inducible protein 27 expression via the regulation of Sp1 transcription factor (69). Binmadi et al [bib_ref] miRNA expression profile of mucoepidermoid carcinoma, Binmadi [/bib_ref] found that miR-302a was significantly increased in mucoepidermoid carcinoma tissues compared with normal tissues. Furthermore, upregulated miR-302a induced invasion of mucoepidermoid carcinoma cells in vitro.
Adenoid cystic carcinoma. Adenoid cystic carcinoma (ACC) generally occurs in the minor salivary glands and has a poor long-term prognosis due to perineural invasion and lung metastasis [bib_ref] Prognostic value of quantitative p63 immunostaining in adenoid cystic carcinoma of salivary..., Ramer [/bib_ref]. Wang et al [bib_ref] The MYB/miR-130a/NDRG2 axis modulates tumor proliferation and metastatic potential in salivary adenoid..., Wang [/bib_ref] analyzed the expression of miR-130a in 21 patients with ACC and corresponding normal salivary gland tissues. Compared with that in normal salivary gland tissues, the expression of miR-130a in ACC tissues increased by 1.58-29.1 times. In addition, the level of miR-130a was negatively correlated with N-myc downstream-regulated gene 2 in ACC tissues. Chen et al [bib_ref] Identification of microRNA profiles in salivary adenoid cystic carcinoma cells during metastatic..., Chen [/bib_ref] analyzed miRNAs during the metastasis of ACC cells and found that the expression levels of miR-4487, miR-4430 and miR-486-3p were upregulated, and the expression levels of miR-5191, miR-3131 and miR-211-3p were downregulated. Andreasen et al [bib_ref] MicroRNA dysregulation in adenoid cystic carcinoma of the salivary gland in relation..., Andreasen [/bib_ref] found that high expression levels of miR-21, miR-181a-2 and miR-152 in patients with ACC was associated with a decreased overall survival rate, and high expression of miR-374c was associated with an improved relapse-free survival rate. Wang et al [bib_ref] Effect of simvastatin and microRNA-21 inhibitor on metastasis and progression of human..., Wang [/bib_ref] found that an miR-21 inhibitor significantly reduced the resistance of lung metastatic salivary adenoid cystic carcinoma cells (SACC-LM) to simvastatin. Furthermore, the combination of simvastatin and miR-21 inhibitor decreased the proliferation of SACC-LM cells (75).
## Conceivable therapeutic value of mirnas
Tumor suppressor miRNAs. Yang et al [bib_ref] miR-381-3p suppresses the proliferation of oral squamous cell carcinoma cells by directly..., Yang [/bib_ref] found that when a recombinant lentivirus carrying the miR-381-3p gene was transduced into SCC-9 and Tca-8113 cell lines, miR-381-3p-overexpressing cells demonstrated downregulation of fibroblast growth factor receptor 2. Furthermore, the percentage of cells at the G 1 /G 0 phase was increased and the number of cells at the S phase was decreased. In addition, the apoptotic rate was significantly increased and the number of colonies was decreased in SCC-9 and Tca-8113 cells overexpressing miR-381-3p compared with those in the negative control group. EMT, an important mechanism of invasion or metastasis of cancer, which involves the downregulation of E-cadherin and increases metastasis and invasion as a consequence of loss of intercellular adhesion, may be regulated by miRNAs [bib_ref] Epithelialmesenchymal transitions in development and disease, Thiery [/bib_ref]. ZEB was reported as an EMT-related transcription factor, which was shown to directly combine with the E-cadherin promoter and inhibit its transcription [bib_ref] Epithelialmesenchymal transitions in development and disease, Thiery [/bib_ref] [bib_ref] Transcription regulation of E-cadherin by zinc finger E-box binding homeobox proteins in..., Wong [/bib_ref].
Hashiguchi et al [bib_ref] Tumor-suppressive roles of ΔNp63β-miR-205 axis in epithelial-mesenchymal transition of oral squamous cell..., Hashiguchi [/bib_ref] identified an association between miR-205 and the EMT phenotype in SQUU-B cells and demonstrated that overexpression of miR-205 downregulated ZEB1, ZEB2 and N-cadherin, and upregulated E-cadherin. Another study reported that overexpression of miR-375 significantly upregulated SCC-4 cell radiation-induced apoptosis by directly regulating the insulin-like growth factor 1 receptor (IGF1R) [bib_ref] MicroRNA-375 inhibits growth and enhances radiosensitivity in oral squamous cell carcinoma by..., Zhang [/bib_ref]. Tumor suppressor miRNAs, and their respective target genes and downstream signaling pathways, are presented in [fig_ref] Table I: Dysregulation of miRNAs associated with oral cancer detected in saliva, blood, serum... [/fig_ref]. Tumor suppressor miRNAs inhibit their respective target genes, which inhibit cellular proliferation, growth, motility, migration, invasion, metastasis, glucose metabolism, EMT, promote cell arrest and apoptosis and increase chemosensitivity and radiosensitivity. The (-) symbol indicates inhibition of downstream signaling pathways. miRNA/miR, microRNA; EMT, epithelial-mesenchymal transition; KLF5, kruppel like factor 5; SLC7A11, solute carrier family 7 member 11; AKT1, AKT serine/threonine kinase 1; YAP-1, yes associated protein 1; SEMA6A, semaphorin 6A; PIK3CA, phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit α; ABCG2, ATP-binding cassette subfamily G member 2; TRIM14, tripartite motif containing 14; CCL-2, C-C motif chemokine ligand 2; IL-8, interleukin-8; PAK1, p21 (RAC1) activated kinase 1; FOXC1, forkhead box C1; FZD7, frizzled class receptor 7; Wnt10b, wingless-type MMTV integration site family, member 10b; TIMP2, TIMP metallopeptidase inhibitor 2; ILK, integrin-linked kinase; BSG, basigin; HDAC9, histone deacetylase 9; NR4A1, nuclear receptor subfamily 4 group A member 1; FAP, fibroblast activation protein α; DENND2D, DENN domain containing 2D; CDK4/6, cyclin-dependent kinase 4/6; SOX-2, sex determining region Y box 2; SIX1, sine oculis-related homeobox 1; HOXA , homeobox A9; HOXB7, homeobox B7; MTDH, metadherin; Glut1, glucose transporter-1. Bmi1, B lymphoma Mo-MLV insertion region 1 homolog.
Tumor-promoting miRNAs. Fu et al [bib_ref] MiR-155 regulates oral squamous cell carcinoma Tca8113 cell proliferation, cycle, and apoptosis..., Fu [/bib_ref] demonstrated that miR-155 targeted the cyclin-dependent kinase inhibitor 1B (CDKN1B) 3'-UTR by a luciferase reporter assay. Furthermore, downregulation of miR-155 led to an increase in CDKN1B and inhibited cell proliferation and cell cycle progression in oral cancer Tca8113 cells [bib_ref] MiR-155 regulates oral squamous cell carcinoma Tca8113 cell proliferation, cycle, and apoptosis..., Fu [/bib_ref]. A previous study demonstrated that the mRNA and protein levels of vimentin and N-cadherin were upregulated in SCC-9 and UM1 cells transfected with miR-373-3p, respectively, while E-cadherin and CK18 were downregulated, suggesting that miR-373-3p may stimulate the EMT phenotype [bib_ref] MiR-373-3p targets DKK1 to promote EMT-induced metastasis via the Wnt/β-catenin pathway in..., Weng [/bib_ref]. The repression of the negative regulator of the Wnt signaling pathway, Dickkopf-related protein 1, and the nuclear accumulation of β-catenin promoted the Wnt signaling pathway, which facilitated proliferation of SCC-9 and UM1 cells [bib_ref] MiR-373-3p targets DKK1 to promote EMT-induced metastasis via the Wnt/β-catenin pathway in..., Weng [/bib_ref]. Dysregulation of the Wnt signaling pathway in tumors may result in chemoresistance. Protein phosphatase 2 regulatory subunit Bα (PPP2R5A) is a repressor of Wnt signaling, and in oral cancer cells, miR-218 increased cisplatin resistance via the Wnt signaling pathway by repressing PPP2R5A expression [bib_ref] MicroRNA-218 promotes cisplatin resistance in oral cancer via the PPP2R5A/Wnt signaling pathway, Zhuang [/bib_ref]. demonstrated that when UM1 and Cal27 cells were transfected with miR-218 mimics, the miR-218 overexpression resulted in increased levels of β-catenin, indicating activation of Wnt signaling, and enhanced cell viability. Furthermore, blocking the effect of miR-218 reversed chemoresistance in resistant cells. Jiang et al [bib_ref] MiR-222 targeted PUMA to improve sensitization of UM1 cells to cisplatin, Jiang [/bib_ref] revealed that low expression of miR-222 increased the chemosensitivity of oral cancer cells to cis-diaminedichloroplatinum (CDDP) and that the combination of antisense-miR-222 and CDDP may be an effective curative strategy by upregulating the expression of p53, a modulator of apoptosis. [fig_ref] Table I: Dysregulation of miRNAs associated with oral cancer detected in saliva, blood, serum... [/fig_ref] presents tumor-promoting miRNAs with their respective target genes and downstream signaling pathways [bib_ref] MicroRNA-218 promotes cisplatin resistance in oral cancer via the PPP2R5A/Wnt signaling pathway, Zhuang [/bib_ref] [bib_ref] MiR-222 targeted PUMA to improve sensitization of UM1 cells to cisplatin, Jiang [/bib_ref] [bib_ref] Oral squamous cell carcinoma cells are resistant to doxorubicin through upregulation of..., Du [/bib_ref] [bib_ref] Downregulation of miR-221/222 by a microRNA sponge promotes apoptosis in oral squamous..., Zhou [/bib_ref] [bib_ref] MicroRNA-24 induces cisplatin resistance by targeting PTEN in human tongue squamous cell..., Zheng [/bib_ref] [bib_ref] Up-regulation of miR-455-5p by the TGF-β-SMAD signalling axis promotes the proliferation of..., Cheng [/bib_ref] [bib_ref] MTSS1 gene regulated by miR-96 inhibits cell proliferation and metastasis in tongue..., Guo [/bib_ref] [bib_ref] MiR-497 enhances metastasis of oral squamous cell carcinoma through SMAD7 suppression, Hu [/bib_ref] [bib_ref] MiR-200c-3p spreads invasive capacity in human oral squamous cell carcinoma microenvironment, Kawakubo-Yasukochi [/bib_ref] [bib_ref] -5p promotes growth in oral squamous cell carcinoma by inhibiting CAMK2N1, Li [/bib_ref] [bib_ref] Up-regulation of miR-187 modulates the advances of oral carcinoma by targeting BARX2..., Lin [/bib_ref] [bib_ref] MicroRNA-92b promotes tumor growth and activation of NF-κB signaling via regulation of..., Liu [/bib_ref] [bib_ref] miR-654-5p targets GRAP to promote proliferation, metastasis, and chemoresistance of oral squamous..., Lu [/bib_ref] [bib_ref] MiR-134 targets PDCD7 to reduce E-cadherin expression and enhance oral cancer progression, Peng [/bib_ref] [bib_ref] A: microRNA-27a-3p modulates the Wnt/β-Catenin signaling pathway to promote epithelial-mesenchymal transition in..., Qiao [/bib_ref] [bib_ref] Loss of PTEN expression is associated with high microRNA-24 level and poor..., Zhao [/bib_ref] [bib_ref] MYCN-mediated miR-21 overexpression enhances chemo-resistance via targeting CADM1 in tongue cancer, Zheng [/bib_ref] [bib_ref] MicroRNA-211 enhances the oncogenicity of carcinogen-induced oral carcinoma by repressing TCF12 and..., Chen [/bib_ref].
Recent conventional experimental methods. Bioinformatics analysis may be used to predict the pairing sequences of miRNAs and target genes, which may be verified by luciferase reporter assays. Western blotting and RT-qPCR may subsequently be used to detect protein and miRNA expression, respectively. Frequently used oral cancer cell lines include SCC-9, SCC-4, Tca-8113 and Cal27 [bib_ref] MicroRNA-137 suppresses tongue squamous carcinoma cell proliferation, migration and invasion, Sun [/bib_ref] [bib_ref] MiR-373-3p targets DKK1 to promote EMT-induced metastasis via the Wnt/β-catenin pathway in..., Weng [/bib_ref] [bib_ref] MiR-155 regulates oral squamous cell carcinoma Tca8113 cell proliferation, cycle, and apoptosis..., Fu [/bib_ref] [bib_ref] MiR-24 promotes the proliferation, migration and invasion in human tongue squamous cell..., Zhao [/bib_ref]. Cell experiments demonstrated that specific miRNAs served an anticancer role such as miR-138, miR-200c, miR-15b, miR-485-5p and miR-340 [bib_ref] suppresses the proliferation of oral squamous cell carcinoma cells by targeting Yes-associated..., Xu [/bib_ref] [bib_ref] MicroRNA-200c suppresses tumor metastasis in oral squamous carcinoma by inhibiting epithelial-mesenchymal transition, Xie [/bib_ref] [bib_ref] miR-15b inhibits cancer-initiating cell phenotypes and chemoresistance of cisplatin by targeting TRIM14..., Wang [/bib_ref] [bib_ref] Hsa-miR-485-5p reverses epithelial to mesenchymal transition and promotes cisplatin-induced cell death by..., Lin [/bib_ref] [bib_ref] MicroRNA-340 mediates metabolic shift in oral squamous cell carcinoma by targeting glucose..., Xu [/bib_ref] ; whereas other miRNAs stimulated oral cancer by regulating cellular proliferation, apoptosis, invasion, EMT, metastasis, radiosensitivity, chemosensitivity and glucose metabolism, such as miR-221, miR-455-5p, miR-27a-3p, miR-21 and miR-10a [bib_ref] Oral squamous cell carcinoma cells are resistant to doxorubicin through upregulation of..., Du [/bib_ref] [bib_ref] Up-regulation of miR-455-5p by the TGF-β-SMAD signalling axis promotes the proliferation of..., Cheng [/bib_ref] [bib_ref] A: microRNA-27a-3p modulates the Wnt/β-Catenin signaling pathway to promote epithelial-mesenchymal transition in..., Qiao [/bib_ref] [bib_ref] MYCN-mediated miR-21 overexpression enhances chemo-resistance via targeting CADM1 in tongue cancer, Zheng [/bib_ref] [bib_ref] MiRNA-10a promotes cancer cell proliferation in oral squamous cell carcinoma by upregulating..., Chen [/bib_ref]. The results obtained from in vitro cell experiments may provide novel therapeutic targets for potential clinical application.
Experimental conclusion in oral carcinoma. miRNAs are promising therapeutic targets, as they serve important roles (115) and miR-24 [bib_ref] MicroRNA-24 induces cisplatin resistance by targeting PTEN in human tongue squamous cell..., Zheng [/bib_ref]. Therefore, the combination of more specific inhibitors or activators of cancer-associated genes or signaling pathways may be a suitable therapeutic strategy. With regard to controversial miRNAs, including miR-222, which served an ambivalent role in disparate experiments, more research is required to verify their roles in oral cancer [bib_ref] Deregulation of the miR-222-ABCG2 regulatory module in tongue squamous cell carcinoma contributes..., Zhao [/bib_ref] [bib_ref] MiR-222 targeted PUMA to improve sensitization of UM1 cells to cisplatin, Jiang [/bib_ref] [bib_ref] Downregulation of miR-221/222 by a microRNA sponge promotes apoptosis in oral squamous..., Zhou [/bib_ref].
# Conclusions
On the basis of current research, the aberrant expression of miRNAs has been demonstrated to be significantly associated with oral cancer. As either a tumor marker or a therapeutic target, miRNA has potential to diagnose or treat oral cancer and improve survival. miRNA serves important roles in the occurrence, development, therapy and prognosis of oral cancer, and is a promising target for clinical application. In terms of the mechanism of malignant transformation or oncogenicity, prognostic and diagnostic value, and potential as a therapeutic target, miR-31 seems to be a promising candidate for clinical application. miR-31 is differentially expressed in normal mucosa, OPMD and oral cancer, and may be detected with high sensitivity in tissue, saliva and plasma. Furthermore, miR-31 may be used to evaluate surgical efficacy. However, miR-375 and miR-203 may be superior therapeutic targets, as they target multiple genes that regulate additional factors and malignant biological properties in oral cancer.
There are a number of challenges in the experimental research and clinical application of miRNA. The transcriptional activation of miRNA and the regulation of indispensable components (containing Drosha/DGCR8, Dicer, XPO5 and TRBP) in the maturation process of oncogenic or antineoplastic miRNA through signaling pathways, and the interference of signaling pathways by mature miRNA, form a series of feedback loops, which may either contribute to tumorigenesis or be used for effective treatments. Further clinical trials that explore specific or highly sensitive miRNA closely associated with oral cancer are required to identify biomarkers with prognostic value. Combining multiple miRNAs for diagnosis and therapy is also a promising strategy that requires further examination, and investigating the association between oral cancer subtypes and miRNA may facilitate the development of targeted medicine in oral cancer. Identifying the detection threshold of different miRNAs in serum, specimen and saliva may aid in predicting the risk of malignant transformations and in evaluating the risk of tumor metastasis or relapse. In addition to further elucidating the mechanisms and anticancer strategies targeted at miRNA, the potential resistance and complications of new antitumor drugs are novel challenges to overcome in order to identify more effective treatments for oral cancer.
## Acknowledgements
Not applicable.
# Funding
The present study was supported by the Chongqing Municipal Health Bureau (grant no. 2017HBRC004) and the Natural Science Foundation Project of CQ CSTC (grant no. cstc2018jcyjAX0763).
## Availability of data and materials
Not applicable.
# Authors' contributions
YL designed the review and revised the manuscript. CF wrote the manuscript. Both authors reviewed the final version and approved it for publication.
## Ethics approval and consent to participate
Not applicable.
## Patient consent for publication
Not applicable.
[fig] Figure 1: Biogenesis of miRNA. Pri-miRNA is transcribed by RNA Pol II. Drosha/DGCR8 shears the 7-methyl guanine nucleoside (m7GpppN) and 3'poly-(A) tail (AAAAA), which forms pre-miRNA. The RanGTP/XPO5 complex functions as a carrier for nucleocytoplasmic transport of pre-miRNA. miRNA duplex is generated after Dicer cleaves the stem-loop of pre-miRNA. miRNA, microRNA; pri-miRNA, primary miRNA; RNA Pol II, RNA polymerase II; pre-miRNA, precursor miRNA; DGCR8, Drosha/DiGeorge syndrome chromosomal region 8; RanGTP/XPO5, GTP-binding nuclear protein Ran/exportin-5. [/fig]
[fig] Figure 2: Modulatory mode of miRNA and target mRNA. TRBP recruits the miRNA duplex and Ago2, forming RISC. One strand of miRNA degrades and the other remains, which targets mRNA. There are two ways to silence the target genes: (A) Translative repression by incomplete pairing and (B) degradation of target mRNA by complementary pairing. miRNA, microRNA; TRBP, transactivation response element RNA binding protein; Ago2, argonaute 2; RISC, RNA-induced silencing complex. [/fig]
[table] Table I: Dysregulation of miRNAs associated with oral cancer detected in saliva, blood, serum and plasma. [/table]
|
Plasma-Derived Fibronectin Stimulates Chondrogenic Differentiation of Human Subchondral Cortico-Spongious Progenitor Cells in Late-Stage Osteoarthritis
Migration and chondrogenesis of human subchondral cortico-spongious progenitor cells (SPCs) are the key steps in the repair of microfracture-induced articular cartilage defects. The aim of this study was to evaluate the effect of human plasma-derived fibronectin (Fn) on the chondrogenic differentiation of SPCs, which was isolated from subchondrol cortico-spongious bone of late-stage osteoarthritis (OA) patients. SPCs were isolated and cultured for three passages. Stem cell surface antigens of SPCs were analyzed by flow cytometry. The osteogenic, chondrogenic and adipogenic differentiation potential were detected by histological staining. The chondrogenesis potential of SPCs with or without stimulation of either Fn or BMP-2 were studied by immunochemical staining and gene expression analysis. Cells isolated from subchondral bone presented to be positive for CD44, CD73, CD90, and CD166, and showed high capacity of osteogenic, adipogenic and chondrogenic differentiation, which suggested this cell population to be MSC-like cells. Stimulating with Fn increased the expression of SOX-9, aggrecan, collagen II while decreased the formation of collagen I by immunochemical staining. Gene expression OPEN ACCESS analysis showed similar results. These results suggest that plasma-derived Fn can increase the chondrogenic differentiation of SPCs isolated from late-stage OA and improve cartilage repair after microfracture.
# Result
## Morphology and cell-surface antigens of spcs in late stage oa knee
Cells were grown out from subchondral bone chips and attached after 4-6 days of culture, which were spindle-shaped and fibroblast-like [fig_ref] Figure 1: Morphology and cell-surface antigens of subchondral cortico-spongious progenitor cells [/fig_ref]. They grew in monolayer, maintained a stable morphology with no signs of granulation. SPCs presented typical cell surface antigens known from mesenchymal stem cells. SPCs at passage 3 was homogenously positive for CD44 (98.69%), CD73 (97.94%), CD90 (99.88%), CD166 (91.46%). CD105 (6.13%) and CD146 (10.41%) were also detected, while the lipopolysaccharide receptor CD14 (0.24%), B-lymphocyte antigen CD19 (0.96%), haematopoietic surface antigen CD34 (0.47%), and the leukocyte common antigen CD45 (0.95%) and HLD-DR (0.21%) were not present [fig_ref] Figure 2: Cont. [/fig_ref].
## Osteogenic or adipogenic differentiation potential of spcs
After 21 days of culture, cells stimulated with osteogenic inducing medium showed alkaline phosphatase activity, and evidence of mineralized extracellular matrix components by alizarin red staining, compared with non-stimulated cells ,B,D,E). Cells stimulated with adipogenic inducing medium showed lipid droplets by oil red O staining, while non-stimulated cells did not .
## Figure 3.
Histological analysis of SPCs undergoing osteogenic or adipogenic differentiation. By day 21, osteogenic induced SPCs cells showed alkaline phosphatase activity (D) and presence of mineralized matrix components (E); compared with non-differentiated cells (A,B); Adipogenic induced SPCs showed lipid droplets (F); while non-differentiated cells did not (C) (×400).
## Chondrogenic differentiation of spcs
After incubation for 28 days, histological analysis of micro-masses showed abundant proteoglycans in the positive control group [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref] , while it was not shown in the negative control group [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref]. The chondrogenesis-related genes COL II, aggrecan, SOX-9 and COL I in micro-mass cultures of MSC were detected by immunohistochemical staining. Only COL I was detected in MSC of negative control group [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref] , which showed no existence of COL II, aggrecan or SOX-9 [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref] There was a strong staining of COL I in negative group (C); while a relatively weak staining in other stimulated groups (H,M,R). SOX-9, Aggrecan, COL II and COL I were brown immune-staining products with haematoxylin counterstain. Proteoglycans were stained blue with haematoxylin counterstain (×400).
## The effect of fn on spcs' chondrogenic differentiation
To assess effects of Fn stimulation on chondrogenic differentiation, COL II, aggrecan, SOX-9 and COL I were also detected by immunohistochemical staining, together with proteoglycans staining by alcian blue in each group at day 28. In general, SPCs stimulated by either BMP-2 or Fn formed matrix rich in COL II and proteoglycan, and pellet which seemed more dense in Fn group. SPCs showed a decreased level of COL I [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref] , and increased levels of COL II [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref] , aggrecan [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref] and Sox-9 [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref] in Fn and BMP-2 groups. SPCs stimulated by either BMP-2 or Fn showed a significantly increased staining of proteoglycans [fig_ref] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators [/fig_ref].
## Chondrogenic differentiation gene expression of spcs stimulated by fn
Genes related to chondrogenic differentiation were analyzed by RT-PCR at day 28. As for the fibrous tissue marker COL I, the mean expression level was 0.57 (positive control), 0.57 (BMP-2), and 0.46 (Fn), respectively (expression level in negative control group was designed as 1). As for the cartilage key marker gene COL II, the expression level was 88.11 (positive control), 168.97 (BMP-2), and 270.11 (Fn), respectively. The expression level of chondrogenic marker gene aggrecan was 54.04 (positive control), 53.76 (BMP-2), and 67.18 (Fn), respectively. The expression level of transcription factor Sox-9 was 41.29 (positive control), 78.09 (BMP-2), and 162.50 (Fn), respectively. Compared with negative group, all gene expression levels showed significant differences (p < 0.01). There were also significant differences between each stimulated groups [fig_ref] Figure 5: The gene expression of osteogenic marker gene COL I, as well as... [/fig_ref]. Treatment with BMP-2 or Fn stimulated chondrogenic differentiation of SPCs and the effect of Fn was stronger. The mean of each triplicate well is plotted and the error bars represent SD. * compared to negative control group, p < 0.05; ^ compared to positive control group, p < 0.05; # compared to BMP-2 group, p < 0.05.
## Proliferation and migration of spcs stimulated by fn
Proliferation of SPCs had significantly increased since day 1 compared with day 0 in positive, BMP-2 and Fn groups [fig_ref] Figure 6: Quantification of proliferation and migration of SPCs cultured under different conditions [/fig_ref].
Fn could stimulate migration of SPCs significantly [fig_ref] Figure 6: Quantification of proliferation and migration of SPCs cultured under different conditions [/fig_ref]. Few SPCs migrated through the pores of the polycarbonate membrane [fig_ref] Figure 6: Quantification of proliferation and migration of SPCs cultured under different conditions [/fig_ref] ) and adhered at the bottom site of the membrane in negative control (migration index 1 ± 0.41). Stimulation with Fn significantly increased the migration index (3.6 ± 0.95, p < 0.05) [fig_ref] Figure 6: Quantification of proliferation and migration of SPCs cultured under different conditions [/fig_ref].
# Discussion
Regeneration or repair of symptomatic articular cartilage defects has been at the forefront of regenerative medicine for decades [bib_ref] Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation, Brittberg [/bib_ref]. A number of studies showed bone marrow MSCs improved symptoms of knee OA in terms of both short-term clinical outcomes and magnetic resonance observations of cartilage repair tissue scores, although the long-term clinical outcomes remained controversial [bib_ref] Osteochondral lesions of the knee: A new one-step repair technique with bone-marrow-derived..., Buda [/bib_ref] [bib_ref] Cartilage healing after microfracture in osteoarthritic knees, Bae [/bib_ref].
In the current study, the human SPCs in late stage OA patients were confirmed to have the characteristics of MSCs by flow cytometry, displaying typical morphological features of MSCs as reported by previous studies [bib_ref] BMP2 initiates chondrogenic lineage development of adult human mesenchymal stem cells in..., Schmitt [/bib_ref] , and representing the multipotent developmental capacity of MSCs to generate a diversity of mesodermal lineages, such as bone, fat, and cartilage. Using the micromass pellet culture system, SPCs from late stage OA knees possessed chondrogenic potential. Furthermore, this potential increased by adding different stimulators. The immunohistochemistry results showed the addition of plasma-derived Fn or BMP-2 increased the proteoglycans, COL II production compared with the positive control. Also, Fn group showed a larger increase, compared with the BMP-2 group. RT-PCR studies showed that SPCs treated with Fn or BMP-2 increased mRNA expression of COL II, aggrecan and SOX-9, again the increases of transcription were higher in the Fn group. Among these indexes, both COL II and aggrecan were major constituents of cartilage matrix, and SOX-9 played an essential role in cartilage formation. Fn stimulated the migration and proliferation of SPCs more clearly than negative control as well.
The highlights of the "one-step repair technique" are how to induce recruitment and/or migration, and chondrogenesis of progenitor cells, with chondrogenesis the most important step. Many different factors have been discussed as being responsible for this progression.
Bone morphogenetic proteins (BMPs)-members of the TGF-beta family-regulate cell proliferation, apoptosis, and differentiation [bib_ref] The TGF-β family mediator Smad1 is phosphorylated directly and activated functionally by..., Kretzschmar [/bib_ref]. They have been shown to play a crucial role in chondrogenic development of mesenchymal progenitors [bib_ref] BMP2 initiates chondrogenic lineage development of adult human mesenchymal stem cells in..., Schmitt [/bib_ref] , stimulating the synthesis of cartilage matrix components by articular chondrocytes [bib_ref] Natural bovine osteogenin and recombinant human bone morphogenetic protein-2B are equipotent in..., Luyten [/bib_ref]. Here, our results were in line with these conclusions. Over the last few decades, Fn has been proven as a key factor in human serum to enhance mesenchymal progenitor cell migration and proliferation among a wide range of cell types and tissues [bib_ref] Identification of fibronectin as a major factor in human serum to recruit..., Kulawig [/bib_ref] , including subchondral mesenchymal progenitor cells from the femur head of healthy donors post-mortem [bib_ref] Fibronectin stimulates migration and proliferation, but not chondrogenic differentiation of human subchondral..., Kalkreuth [/bib_ref]. As a part of the matrix of various tissues, Fn has frequently been a topic of extracellular matrix research. It has shown significant migratory and proliferative effects on subchondral progenitor cells which relied on both chemotactic and haptotactic activity [bib_ref] Fibronectin stimulates migration and proliferation, but not chondrogenic differentiation of human subchondral..., Kalkreuth [/bib_ref] , and therefore promotes the formation of cartilage-like repair tissue containing collagen type II [bib_ref] Fibronectin and stem cell differentiation-lessons from chondrogenesis, Singh [/bib_ref]. Our results were in line with the previous conclusions. In this study, SPCs from subchondral bone of OA donors also showed increased migration and proliferation capacities by stimulation of Fn. Regarding differentiation, some previous studies observed inhibitory effects of plasma-derived Fn on chondrogenic differentiation of chicken limb bud mesenchymal cells, while others reported plasma-derived Fn had no potential effect on chondrogenic differentiation [bib_ref] Fibronectin stimulates migration and proliferation, but not chondrogenic differentiation of human subchondral..., Kalkreuth [/bib_ref]. However, we could show that Fn is associated with chondrogenic differentiation of SPCs from the subchondral bone. The difference in the differentiation profile of the progenitor cells could be induced by several reasons. Firstly, MSCs from OA donors showed reduced proliferative capacity and chondrogenic and adipogenic differentiation, compared to those in the normal population, which was not age-or site-dependent, but was associated with disease [bib_ref] Reduced chondrogenic and adipogenic activity of mesenchymal stem cells from patients with..., Murphy [/bib_ref]. SPCs here were isolated from the subchondral bone of OA patients, while subchondral mesenchymal progenitor cells from the femur head of healthy donors post-mortem were examined in a previous study [bib_ref] Fibronectin stimulates migration and proliferation, but not chondrogenic differentiation of human subchondral..., Kalkreuth [/bib_ref]. Secondly, the concentration of Fn used here was different from the previous study (25 vs. 15 μg/mL) [bib_ref] Fibronectin stimulates migration and proliferation, but not chondrogenic differentiation of human subchondral..., Kalkreuth [/bib_ref]. There was lack of knowledge regarding effects of treatment with higher doses of Fn, which needed further study Thirdly, these could also be partially deduced to various isoforms of Fn, which resulted in the determination of impact by the selected type of molecule and the proper concentration. Although highly speculative, one possible reason is that Fn plays a structural role by providing a scaffold for cell adhesion and differentiation. Another possible reason is that Fn signals through integrins to activate intracellular pathways that regulate changes in gene expression during chondrogenesis [bib_ref] The regulation of differentiation in mesenchymal stem cells, Augello [/bib_ref] [bib_ref] PI-3 kinase-Akt, and MAP kinase signaling pathways in chondrocytes, Beier [/bib_ref] , which was also reflected by our mRNA results. A third probable reason was that Fn was involved in the organization of the cartilage matrix. The Fn matrix was shown to support deposition of collagens, latent TGF-β binding proteins and other ECM proteins [bib_ref] Fibronectin and stem cell differentiation-lessons from chondrogenesis, Singh [/bib_ref]. Our results showed the increasing level of collagen II and aggrecan, which was in line with previous studies. Therefore, Fn could be a potential application for the histological outcomes of cartilage regeneration after using the "one-step repair technique". Besides, from data of gene expression analysis, we found that plasma-derived Fn is not inferior to BMP-2 in chondrogenic inducing capacity. However, the mechanisms of Fn mediated stimulation of chondrogenesis need further study, which is a limitation of this study.
In conclusion, this study confirms for the first time that plasma-derived Fn has chondrogenic differentiation impact on human SPCs, and is a potential source of autologous growth factors that is clinically applied using the "one-step repair technique" in areas such as microfracture for the treatment of focal articular cartilage defects in patients with OA.
## Experimental section
## Isolation and cultivation of spcs
Human subchondral bone chips with 3-5 mm diameters were obtained from the tibial platform of 13 donors (8 females, 5 males; age 54-79 years, average age 66.8 years) undertaking total knee arthroplasty due to late stage OA. The study was approved by the ethics committee of Peking Union Medical College Hospital. All patients who had late stage OA and underwent total knee arthroplasty provided written informed consent. MSC was then isolated and cultured as described previously [bib_ref] Chondrogenic differentiation capacity of human mesenchymal progenitor cells derived from subchondral cortico-spongious..., Neumann [/bib_ref]. In brief, subchondral bone chips were washed with phosphate buffered saline (PBS), placed in cell culture flasks (Costar, Corning Scientific, New York, NY, USA), and cultured in DMEM (Hyclone, Logan, UT, USA) containing 10% fetal bovine serum, 100 U/mL penicillin, 100 μg/mL streptomycin, 2 mM L-glutamine (all Invitrogen, San Diego, CA, USA) at 37 °C. After 4-6 days of incubation without medium change, spindle-like and adherent cells began to come up around bone chips. The media was then exchanged for the first time, and later exchanged every 2-3 days. At 80%-90% confluence, cells were harvested using trypsine/EDTA (Invitrogen, Karlsruhe, Germany) and re-plated at a density of 6000 cells/cm 2 .
# Flow cytometric analysis
## Osteogenic and adipogenic differentiation of spcs
To induce osteogenic and adipogenic differentiation, SPCs (80,000 cells, passage 3) were cultured in a 24-well plate with either osteogenic (CCM007 + CCM008; R&D Systems, Inc., USA) or adipogenic (CCM007 + CCM011; R&D Systems, Minneapolis, MN, USA) differentiation medium. Cells were cultured for 21 days and medium was changed every three days.
## Chondrogenic differentiation of spcs
Chondrogenic differentiation of SPCs (250,000, passage 3) was performed in a 96-well U-bottom plate by high-density pellet cultures as previously described. Cells were incubated in different culture as follows: basic media (CCM005; R&D Systems, Inc., USA) without chondrogenic differentiation supplement (negative control group), chondrogenic differentiation media (CCM005 + CCM006; R&D Systems, Inc., USA) (positive control group), chondrogenic differentiation media with 100 ng/mL BMP-2 (R&D Systems, Inc., USA) (BMP-2 group), chondrogenic differentiation media with 25 μg/mL human fibronectin (Sigma, Saint Louis, MO, USA) (Fn group). Chondrogenic differentiation media was basic media with chondrogenic differentiation supplement, which was used without adding additional serum. The medium was exchanged every three days and cells were maintained for up to 28 days.
## Histology and immunohistochemistry
Osteogenic differentiation (n = 9) was assessed by Alizarin red staining of mineralized matrix components and alkaline phosphatase activity (ALP kit, Nanjin-Jiancheng, Nanjin, China). The lipid droplets were stained by Oil red O (Sigma, Saint Louis, MO, USA) to assess the adipogenic differentiation (n = 9).
The cryo-sections (6 μm; n = 9) were prepared for assessing chondrogenic differentiation and stained by Alcian Blue (Sigma). Besides, the cryo-sections were fixed and incubated with the primary mouse anti-human type I collagen (COL I) antibody, type II (COL II) collagen antibody, and aggrecan (Agg) antibody (all Abcam, Cambridge, UK) for 40 min. Mouse IgG served as controls. Subsequently, cells were colormetrically detected with 3-amino-9-ethylcarbazole (EnVisionTM; Dako, Glostrup, Denmark) according to the manufacturer's instructions, followed by counterstaining with hematoxylin (Merck, Darmstadt, Germany).
## Quantitative rt-rcr
Total RNA (12 pellets per experimental condition) was isolated with Trizol reagent (Invitrogen), and 2 μg RNA was reversely transcribed with cDNA Synthesis Kit (TaKaRa, Dalian, China) according to the manufacturer's recommendations. The relative expression level of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used to normalize the samples. RT-PCR was performed with 1 μL of each cDNA sample in triplicates using 7500 Real-Time PCR System (Applied Biosystems) with SYBR green PCR Core Kit (Applied Biosystems, Foster City, CA, USA). Relative quantitation of the marker genes [fig_ref] Table 1: Sequences of the primers used in RT-PCR [/fig_ref] was performed according to the ΔΔ Ct method and is given as fold change compared to controls.
## Proliferation and migration assay
For cell proliferation assay, SPCs were seeded (2000 cell pre well) in 96-well plate and incubated for 16 h. Then, attached cells were washed twice with PBS, and incubated under different conditions, i.e., MSC medium without FBS (negative control), MSC medium with 10% FBS (positive control), 100 ng/mL BMP-2 in MSC medium containing 1% FBS, 25 μg/mL human fibronectin in MSC medium containing 1% FBS. At time point of 0, 1, 3, 5, 7 days, 10 μL CCK8 was added to the medium and incubated for an additional 2 h. The optical density at a wavelength of 450 nm was analyzed with a microliter plate reader.
For cell migration assay, SPCs were starved in basal medium for 12 h, and 2 × 10 4 SPCs were suspended in 400 μL medium with 1% FBS and placed in the upper chamber. Five hundred microliters of medium used in different groups (the same with cell proliferation assay) was placed in the lower compartment. After 6 h incubation at 37 °C, the lower side of the filter was washed with PBS and fixed with methol. For quantification, cells were stained with Giemsa. Cells migrating into the lower chamber were counted manually in 10 random microscopic fields (×200).
# Statistical analysis
The Kolmogorove-Smirnov method was applied for testing normal distribution of the data. The t-test was applied for normally distributed data, while the Mann-Whitney U-test was used for data that failed normality testing. In all tests, p values less than 0.05 were considered statistically significant.
[fig] Figure 1: Morphology and cell-surface antigens of subchondral cortico-spongious progenitor cells (SPCs) in late-stage osteoarthritis (OA) patients. (A) Cells outgrew from subchondral bone chips and attached after 4-6 days of culture (×400); (B) Cells (P1) at 90% confluence after 14 days of culture were spindle-shaped and fibroblast-like (×100). [/fig]
[fig] Figure 2: Cont. [/fig]
[fig] Figure 4: Histological and immunochemical analysis of SPCs undergoing chondrogenic differentiation with different stimulators. At day 28, SOX-9, Aggrecan, COL II and proteoglycans were evident in the positive group (F,G,I,J); BMP-2 group (K,L,N,O) and the Fn group (P,Q,S,T); negative group showed weak or absent staining (A,B,D,E); [/fig]
[fig] Figure 5: The gene expression of osteogenic marker gene COL I, as well as the chondrogenic marker genes COL II, Aggrecan, and SOX 9 stimulated by different chondrogenic factors. [/fig]
[fig] Figure 6: Quantification of proliferation and migration of SPCs cultured under different conditions. (A) SPCs migrated through the pores of the membrane and adhered at the bottom site of the polycarbonate membrane. Fenestra was the pore of the polycarbonate membrane, and SPCs were stained in purple and were spindle-shaped (×200). (1) Negative control; (2) positive control; (3) BMP-2 group; (4) Fn group; (B) The migration index in different groups. * Compared with negative control, p < 0.05; (C) Cell proliferation under different conditions. * Compared with day 0, p < 0.05. The mean of each triplicate well is plotted and the error bars represent SD. [/fig]
[table] Table 1: Sequences of the primers used in RT-PCR. [/table]
|
Large-scale deletions of the ABCA1 gene in patients with hypoalphalipoproteinemia[S]
SupplementalTable S1: Screening primers for ABCA1 copy-number variations.CNV Breakpoint Primer directionPrimer sequence (5' to 3') SupplementalFigure S1. Identification of ABCA1 CNVs using the VarSeq-CNV® caller algorithm on targeted sequencing data. Chr9:107,542,273-107,697,356 (hg19 genome build) is the region visualized in each panel, with the CNV "ratio", "Z score", and "state" available for each subject. A) Subject 1, carrier of a heterozygous deletion of exon 4. B) Subject 2, carrier of a heterozygous deletion spanning exons 8 to 31. C) Subject 3, carrier of a heterozygous deletion spanning exons 8 to 31. D) Subject 4, carrier of a heterozygous deletion of the entire ABCA1 gene. CNV = copy-number variation.
## Annealing temperature (°c)
Primer labels in and 2 P12 The primers listed were designed to flank the two breakpoints for each CNV. The "Breakpoint" listed is relative to the deletion section of the gene. The sequence orientation for P1-P8 are relative to the ABCA1 gene, while the sequence orientation for P9-P12 are relative to the full chromosome. Abbreviations: CNV = copy-number variation; FWD = forward; REV = reverse. |
Modeling the Differentiation of Embryonic Limb Chondroprogenitors by Cell Death and Cell Senescence in High Density Micromass Cultures and Their Regulation by FGF Signaling
Considering the importance of programmed cell death in the formation of the skeleton during embryonic development, the aim of the present study was to analyze whether regulated cell degeneration also accompanies the differentiation of embryonic limb skeletal progenitors in highdensity tridimensional cultures (micromass cultures). Our results show that the formation of primary cartilage nodules in the micromass culture assay involves a patterned process of cell death and cell senescence, complementary to the pattern of chondrogenesis. As occurs in vivo, the degenerative events were preceded by DNA damage detectable by γH2AX immunolabeling and proceeded via apoptosis and cell senescence. Combined treatments of the cultures with growth factors active during limb skeletogenesis, including FGF, BMP, and WNT revealed that FGF signaling modulates the response of progenitors to signaling pathways implicated in cell death. Transcriptional changes induced by FGF treatments suggested that this function is mediated by the positive regulation of the genetic machinery responsible for apoptosis and cell senescence together with hypomethylation of the Sox9 gene promoter. We propose that FGF signaling exerts a primordial function in the embryonic limb conferring chondroprogenitors with their biological properties.
# Introduction
Cell death is a constant event accompanying the differentiation, growth, and tissue remodeling of biological systems [bib_ref] Cell Deaths in Normal Vertebrate Ontogeny, Glücksmann [/bib_ref]. In embryonic systems, cell death may appear scattered in the differentiating tissues or, most often, it takes place in a massive fashion sculpting the shape and/or the structure of the organ primordia. In adult individuals, cell death is a basic mechanism of tissue turnover, but accumulating evidence indicates that it is also of great importance in stem cell and tumor biology [bib_ref] A Matter of Life and Death: Stem Cell Survival in Tissue Regeneration..., Soteriou [/bib_ref].
The formation of the vertebrate skeleton is accompanied by programmed cell death and cell senescence within or around the differentiating skeletal tissues, including the cartilaginous bone primordia and the developing joints [bib_ref] Developmental Expression of Dkk1-3 and Mmp9 and Apoptosis in Cranial Base of..., Nie [/bib_ref] [bib_ref] Cell Death in the Developing Vertebrate Limb: A Locally Regulated Mechanism Contributing..., Montero [/bib_ref]. The formation of the cartilaginous template that prefigures the future chondral bones is a critical step of skeletogenesis [bib_ref] Cellular Interactions and Signaling in Cartilage Development, Delise [/bib_ref]. The regulation of the cellular mass of the cartilage templates to establish their size and morphology might be among the functions of cell death in skeletogenesis [bib_ref] Bmpr1a and Bmpr1b Have Overlapping Functions and Are Essential for Chondrogenesis in..., Yoon [/bib_ref]. In embryonic tissues, a variable intensity of dead cells is recognizable around the zones of cartilage formation [bib_ref] Cell Death in the Developing Vertebrate Limb: A Locally Regulated Mechanism Contributing..., Montero [/bib_ref] [bib_ref] Experimental Analysis of Msx-1 and Msx-2 Gene Expression during Chick Mandibular Morphogenesis, Mina [/bib_ref] , but the contribution of cell degeneration to the early steps of skeletogenesis has not received much attention. Remarkably, dysregulation of these dying processes causes skeletal malformations [bib_ref] Embryonic Expression of Cyclooxygenase-2 Causes Malformations in Axial Skeleton, Shim [/bib_ref].
Genetic and classical embryological studies in vivo have identified complex signaling networks that influence the formation of the skeleton, including WNTs and BMPs, but FGFs appear to exert a central role in the formation of most cartilages [bib_ref] Ectodermal FGFs Induce Perinodular Inhibition of Limb Chondrogenesis in Vitro and in..., Moftah [/bib_ref] [bib_ref] Fibroblast Growth Factor and Bone Morphogenetic Protein Signaling Are Required for Specifying..., Kumar [/bib_ref] [bib_ref] Fibroblast Growth Factor Maintains Chondrogenic Potential of Limb Bud Mesenchymal Cells by..., Kumar [/bib_ref]. Most of these signals have dual opposite functions, being growth-and differentiation-or cell death-inducing signals in a stage-and regional-dependent fashion [bib_ref] Cell Death in the Developing Vertebrate Limb: A Locally Regulated Mechanism Contributing..., Montero [/bib_ref] [bib_ref] Antagonistic Effects of FGF4 on BMP Induction of Apoptosis and Chondrogenesis in..., Buckland [/bib_ref] [bib_ref] Positionally-Dependent Chondrogenesis Induced by BMP4 Is Co-Regulated by Sox9 and Msx2, Semba [/bib_ref] [bib_ref] FGF Mediated Sulf1 Regulation, Zhao [/bib_ref] [bib_ref] Molecular Signatures Identify Immature Mesenchymal Progenitors in Early Mouse Limb Buds That..., Reinhardt [/bib_ref]. The design of an in vitro model to analyze mechanisms involved in the control of programmed cell death is of great interest to understand this important biological process.
The study of the regulatory signals and the cellular basis of chondrogenesis have been largely addressed in vitro by employing a tridimensional organoid-like culture assay termed the "micromass culture". This assay mimics in vitro changes associated with skeletogenesis in vivo [bib_ref] From Skeletal Development to Tissue Engineering: Lessons from the Micromass Assay, Klumpers [/bib_ref] and has been proposed to replicate the regulatory signals involved in cartilage formation [bib_ref] Digit Patterning Is Controlled by a Bmp-Sox9-Wnt Turing Network Modulated by Morphogen..., Raspopovic [/bib_ref]. Studies in monolayer culture provided evidence for the association between apoptosis and chondrodifferentiation [bib_ref] Role of Chondrogenic Tissue in Programmed Cell Death and BMP Expression in..., Omi [/bib_ref] , but the micromass culture assay has largely been employed in studies focused on the growth and differentiation of progenitors, and less attention was paid to clarify the potential implications of cell death [bib_ref] Loss of Discordant Cells during Micro-Mass Differentiation of Embryonic Stem Cells into..., Yamashita [/bib_ref].
In this work, we have analyzed the pattern of cell death and the role of FGF signaling in primary micromass cultures of limb skeletogenic progenitors. For this study, we chose the undifferentiated autopod mesodermal cells whose fate in vivo is to form digits or to be eliminated by cell death and senescence in the interdigital regions [bib_ref] Cell Death in the Developing Vertebrate Limb: A Locally Regulated Mechanism Contributing..., Montero [/bib_ref] [bib_ref] Confluence of Cellular Degradation Pathways During Interdigital Tissue Remodeling in Embryonic, Montero [/bib_ref]. We show that the formation of primary cartilage nodules in the micromass culture assay involves a patterned process of cell death and cell senescence, complementary to the pattern of chondrogenesis. Treatments with FGF2 alone or in combination with growth factors active during limb skeletogenesis revealed that, in addition to the well-known function of FGF signaling directing the growth of the limb primordia, this pathway modulates the response of progenitors to signals implicated in the regulation of programmed cell death. Transcriptional changes induced by FGF treatments suggest that this function is mediated by positive regulation of the genetic machinery responsible for apoptosis and cell senescence. We propose that FGFs not only direct limb outgrowth but also serve a morphogenetic function tuning the response of progenitors to WNT and BMP signaling.
# Materials and methods
We employed Rhode Island chicken embryos of 4.5 days for incubation, which represent stage HH25. This stage precedes in two days the establishment of the areas of interdigital cell death.
## Micromass cultures and treatments
The leg buds were excised from the embryo and immersed in L-15 culture medium (Leibovitz medium, Lonza Group) and the undifferentiated tissue of the distal margin of the limb ("progress zone") expressing HoxA13-A') was dissected free under the dissecting microscope. To dissociate the cell components, samples were first incubated at 37 - C in 0.25% trypsin (Sigma) for 6 min and next in 0.25% collagenase (Worthington) for 12 min. Digestion was blocked by addition of 1 mL of DMEM (Dulbecco s modified eagle medium, Lonza Group) with 10% fetal bovine serum (FBS, Lonza Group) where cells were dissociated by gently pipetting. Next, 9 mL of L-15 medium were added to the dissociated tissue and filtered through a 70 µm strainer (Miltenyi Biotec) to remove clumps of undissociated tissue and ectodermal debris. The cell suspension was centrifuged at 1000 rpm for 10 min and re-suspended in DMEM/10% FBS medium. Cell density was adjusted to 3 × 10 5 cells/mL, and 10 µL drops were pipetted into each well of a 48-well plate (Thermo Scientific), allowed to be attached for 2 h in an incubator (37 - C and 5% CO2) and then 200 µL of the selected growth medium was added into each well. DMEM with or without 10% FBS containing 100 units/mL penicillin and 100 mg/mL streptomycin were employed as culture medium.
Cell death, cell senescence, cell proliferation, DNA methylation, and transcriptional changes, were analyzed in control cultures and experimental cultures treated with recombinant proteins and signaling inhibitors added to the culture medium, including: FGF2 (25 or 50 ng/mL, Peprotech); BMP7 (200 ng/mL, Peprotech); Noggin (200 ng/mL, Peprotech); human WNT5A (100 ng/mL, R&D System); DKK1 (200 ng/mL, Peprotech); SU5402 (800 ng/mL, Calbiochem); and U0126 (20 nM; Calbiochem). The medium was changed every alternate day. At least 8 experiments for each treatment were performed.
## Morphological and immunohistochemical studies
For histological analysis, cultures were fixed in 4% paraformaldehyde (PFA), washed in PBS, dehydrated in acetone, and embedded in araldite. Semithin sections (1 μm) were stained with toluidine blue.
Chondrogenic differentiation was analyzed by Alcian blue staining (0.5% at pH 1.0) in cultures fixed with alcohol/acetic. Cell senescence was analyzed using the β-galactosidase activity assay [bib_ref] Protocols to Detect Senescence-Associated Beta-Galactosidase (SA-Bgal) Activity, a Biomarker of Senescent Cells..., Debacq-Chainiaux [/bib_ref] in cultures fixed overnight in 4% glutaraldehyde at pH 6.
Immunolabeling was performed in samples grown on glass coverslip previously treated with 1 mg/mL fibronectin. At the desired stage, cultures were fixed for 3 h in 4% PFA, washed for 2 h in PBS-0.1% triton and blocked for 1.5 h in PBS-3% BSA and incubated overnight in the primary antibody. The samples were next washed for 30 min in PBS and incubated in the secondary antibody. We employed rabbit polyclonal anti-SOX9 (Merck-Millipore, AB5535) as a chondrogenic lineage marker and mouse monoclonal anti-γH2AX (Merck-Millipore, 05-636) to evaluate DNA damage. Actin labeling was performed with rhodamine-phalloidin (Sigma).
Apoptotic cell death was analyzed by TUNEL assay using the in-situ cell death detection kit (Roche) following the manufacturer recommendations. Cell death, cell senescence, cell proliferation, DNA methylation, and transcriptional changes, were analyzed in control cultures and experimental cultures treated with recombinant proteins and signaling inhibitors added to the culture medium, including: FGF2 (25 or 50 ng/mL, Peprotech); BMP7 (200 ng/mL, Peprotech); Noggin (200 ng/mL, Peprotech); human WNT5A (100 ng/mL, R&D System); DKK1 (200 ng/mL, Peprotech); SU5402 (800 ng/mL, Calbiochem); and U0126 (20 nM; Calbiochem). The medium was changed every alternate day. At least 8 experiments for each treatment were performed.
## Morphological and immunohistochemical studies
For histological analysis, cultures were fixed in 4% paraformaldehyde (PFA), washed in PBS, dehydrated in acetone, and embedded in araldite. Semithin sections (1 µm) were stained with toluidine blue.
Chondrogenic differentiation was analyzed by Alcian blue staining (0.5% at pH 1.0) in cultures fixed with alcohol/acetic.
Cell senescence was analyzed using the β-galactosidase activity assay [bib_ref] Protocols to Detect Senescence-Associated Beta-Galactosidase (SA-Bgal) Activity, a Biomarker of Senescent Cells..., Debacq-Chainiaux [/bib_ref] in cultures fixed overnight in 4% glutaraldehyde at pH 6.
Immunolabeling was performed in samples grown on glass coverslip previously treated with 1 mg/mL fibronectin. At the desired stage, cultures were fixed for 3 h in 4% PFA, washed for 2 h in PBS-0.1% triton and blocked for 1.5 h in PBS-3% BSA and incubated overnight in the primary antibody. The samples were next washed for 30 min in PBS and incubated in the secondary antibody. We employed rabbit polyclonal anti-SOX9 (Merck-Millipore, AB5535) as a chondrogenic lineage marker and mouse monoclonal anti-γH2AX (Merck-Millipore, 05-636) to evaluate DNA damage. Actin labeling was performed with rhodamine-phalloidin (Sigma).
Apoptotic cell death was analyzed by TUNEL assay using the in-situ cell death detection kit (Roche) following the manufacturer recommendations. Histological observations were made with a LSM51O laser confocal microscope (Zeiss).
## Flow cytometry
Control and treated cultures were dissociated with trypsin EDTA (Lonza) and fixed in 90% ethanol. Samples containing one million cells were incubated overnight at 4 - C with 0.1% sodium citrate, 0.01% Triton X-100, and 0.1 mg/mL Propidium iodide (Sigma). The cell suspension was subjected to flow cytometry analysis in a Becton Dickinson FacsCanto cytometer and analyzed using Cell Quest software.
## Quantification of culture growth
Culture growth was evaluated using the MTT assay. A total of 20 µL of 5 mg/mL tetrazolium salt was added to each culture well. After 2 h of further incubation the medium was replaced by dimethyl sulfoxide (DMSO) to lysate the tissue and the absorbance was evaluated by spectrophotometry at 570 nm.
## In situ hybridization
The expression of growth factors tested in the study and a panel of limb mesodermal markers was confirmed by in situ hybridization in the cultured micromasses. Cultures fixed in 4% PFA were washed in PBT and treated with 7 µg/mL proteinase K for 2-5 min at 20 - C. Hybridization with digoxigenin-labeled anti-sense RNA probes was performed at 68 - C. Reactions were developed with BM-purple AP substrate (Roche).
# Epigenetic analysis
We employed the Methylation-Sensitive Restriction Enzyme and Quantitative Polymerase Chain Reaction (MSRE-qPCR) technique to study changes in the methylation status of the Sox9 and Scleraxis promoters. Genomic DNA samples from micromass cultures were extracted using NucleoSpin Tissue (Macherey-Nagel). In order to evaluate the methylation level of the target genes, we employed the EpiJET DNA Methylation Analysis Kit (MspI/HpaII) following manufacturer's instructions. A total of 100 ng of purified genomic DNA were digested with the non-CpG-methylation-sensitive enzyme MspI or with the CpG-methylation-sensitive restriction enzyme HpaII for 6 h. Undigested Genomic DNA samples were treated as digested samples by replacing the volumes of the enzymes with DNase-free water. SYBRGreen-based qPCR was carried out in triplicates with a total volume of 20 µL per tube containing 2 µL of genomic DNA (MspI-digested, HpaII-digested, or undigested DNA), 0.4 µL of each specific primer, 10 µL of SYBR Select Master Mix (Life Technologies), and 7.2 µL of DNase-free water. Reactions were carried out in a StepOne Real Time System and analyzed by StepOne software v2.3 (Life Technologies). The relative percentage of methylated DNA was calculated according to the equation 2−∆∆Ct. Nonspecific amplification was monitored by melting curve analysis of each reaction. Specific primers for CpG islands are indicated in Supplementary Table S1.
## Real-time quantitative pcr (qpcr) for gene expression analysis
Total RNA was extracted using the NucleoSpin RNA kit (Macherey-Nagel). Firststrand cDNA was synthesized using random hexamers with the RevertAid RT Kit (Thermo Scientific). The cDNA concentration was adjusted to 0.5 µg/µL. SYBRGreen-based qPCR was performed employing SYBR Select Master Mix (Life technologies) using the CFX Connect Real-Time System (BioRad). Specificity was checked by the presence of single peaks in the dissociation curves. Gapdh was chosen as the normalizer. Mean values for gene expression fold changes were calculated relative to a calibrator according to the 2-∆∆Ct equation. qPCR chicken specific primers are indicated in Supplementary Table S1.
# Results
Limb mesoderm micromass culture is an organoid-like tridimensional culture assay that closely recapitulates the chondrogenic differentiation of limb skeletal progenitors in vivo, including their spatial pattern and molecular regulatory signals [bib_ref] From Skeletal Development to Tissue Engineering: Lessons from the Micromass Assay, Klumpers [/bib_ref] [bib_ref] Geometric Analysis of Chondrogenic Self-Organisation of Embryonic Limb Bud Cells in Micromass..., Rolfe [/bib_ref]. To avoid employing a pool of cells with heterogeneous molecular signatures [bib_ref] Molecular Signatures Identify Immature Mesenchymal Progenitors in Early Mouse Limb Buds That..., Reinhardt [/bib_ref] and differences in their pattern of differentiation [bib_ref] Role of Chondrogenic Tissue in Programmed Cell Death and BMP Expression in..., Omi [/bib_ref] [bib_ref] Morphogenetic Differences between Fore and Hind Limb Precartilage Mesenchyme: Relation to Mechanisms..., Downie [/bib_ref] , in this study, we selected mesodermal progenitors obtained from the distal margin of stage 25 leg buds that express HoxA13-A'). All these progenitors have identical potential to form digits, but in normal development, they follow three distinct fates according to their position in the autopod: (i) forming digit cartilages at the digit ray position; (ii) forming the different autopodial fibrous connective tissues in the peridigital regions; or, (iii) undergoing programmed cell death in the interdigital spaces.
## Differentiation, proliferation, and cell death in relation to the presence or absence of fbs in the culture medium
Our findings show that the micromass cultures obtained from limb bud mesoderm at stage 25 mimic, not only the chondrogenic differentiation, but also the degenerative events occurring in normal skeletal development. The difference is that cartilages in vitro take a nodular appearance rather than the radial arrangement of the digits observed in vivo, likely due to the polarized growth of the limb bud versus the uniform growth of the micromass culture.
To select the conditions of our experiments we first evaluated differences in chondrogenesis, cell death, and cell proliferation in cultures growing in medium with or without FBS. In the first three days of culture, the growing tissue undergoes chondrogenic differentiation following a geometric nodular pattern that is believed to reflect the interactive molecular cues that govern normal skeletal development in vivo [bib_ref] Geometric Analysis of Chondrogenic Self-Organisation of Embryonic Limb Bud Cells in Micromass..., Rolfe [/bib_ref] [bib_ref] Interplay between Activator-Inhibitor Coupling and Cell-Matrix Adhesion in a Cellular Automaton Model..., Kiskowski [/bib_ref]. At Day 1 of culture in DMEM, the cells showed an irregular cell distribution, with zones of higher cell concentration and regions where cells were less compacted, but staining with Alcian blue, which is a cartilage matrix-specific dye, was negative at this stage. Cultures growing in a medium containing 10% FBS followed a similar sequence of chondrogenic differentiation, but the size of the cartilage nodules were significantly larger than those present in cultures lacking FBS (not shown).
Analysis of cellular DNA contents by measuring propidium iodide uptakewas performed in micromasses lacking FBS in the medium to avoid the interference of components present in the FBS with the treatments.
## Patterned cell death and cell senescence
Tissue sections of three-day-old DMEM-only cultures revealed that degenerating cells were located around the differentiating cartilages. These alterations included characteristic apoptotic cells, identifiable by the condensed nuclear morphology, and senescent cells [bib_ref] Senescent Cells in 3D Culture Show Suppressed Senescence Signatures, Yadav [/bib_ref] displaying a foamy and enlarged aspect because of vacuolization of the cytoplasm that was positive for senescence-associated β-galactosidase (SA-β-gal) labeling.
Analysis by flow cytometry revealed that the number of dead cells was similar in cultures of the same stage, i.e., the dead intensity was stage dependent. On Day 1 of culture, 9.82% cells were dead. At Day 2, dead cells were 16.41% of the dissociated cells, and the rate of dead cells was 19.13% at Day 3. Consistent with these findings, the total tissue mass of the culture, evaluated by the MTT test, decreased during the first three days of culture. In subsequent growth stages, the cartilage nodules expanded, showing a parallel increase in the total tissue mass of the culture as evaluated by MTT. In the course of limb development in vivo, cell death and cell senescence remove the undifferentiated mesodermal progenitors of the autopod that separate the digit rays [bib_ref] Confluence of Cellular Degradation Pathways During Interdigital Tissue Remodeling in Embryonic, Montero [/bib_ref]. To assess whether the micromass cultures replicate the degenerative events occurring during skeletogenesis in vivo, the spatial pattern of cell degeneration was evaluated. SAβ-gal activity labeling was employed to detect cell senescence, and a TUNEL assay was used to detect apoptosis. At Day 1 of culture, senescent cells are scarce and predominate in the zones of lower cell density. From Day 2 of culture, the distribution of degenerating cells replicated the degenerative process occurring in vivo. As shown in,G, both SA-β-gal-positive senescent cellsand TUNEL-positive apoptotic cellsshowed a preferential arrangement in the internodular spaces.
We previously observed that interdigital cell degeneration is preceded by intense DNA damage (DNAD) [bib_ref] DNA Damage Precedes Apoptosis during the Regression of the Interdigital Tissue in..., Montero [/bib_ref]. The histone variant H2AX is a key factor of the molecular cascade that cells activate by phosphorylation at serine 139 (γH2AX) to repair damaged DNA, and is a precise and precocious marker of the interdigital cells destined to die [bib_ref] The Methylation Status of the Embryonic Limb Skeletal Progenitors Determines Their Cell..., Sanchez-Fernandez [/bib_ref]. Consistent with these observations, we found a preferential distribution of cells with intense γH2AX immunolabeling in the undifferentiated perinodular tissue where apoptotic and senescent cells were distributed. Reduced γH2AX labeling was also observed in some cells located in the core of the differentiating nodules of cartilage that were intensely positive for SOX9. This feature is consistent with the so-called noncanonical distribution of γH2AX observed during the differentiation of stem cells, which reflects chromatin epigenomic modifications associated with cell differentiation. Additional markers of cell degeneration, such as a disorganized cytoskeleton [bib_ref] Caspase Redundancy and Release of Mitochondrial Apoptotic Factors Characterize Interdigital Apoptosis, Zuzarte-Luis [/bib_ref] , confirmed the internodular distribution of the dying cells.
Overall, the above findings revealed a pattern of cell degeneration in the initial stages of differentiation of chondroprogenitors growing in micromass cultures that is reminiscent of the events associated with digit skeletogenesis in vivo.
## Fgfs sensitize chondroprogenitors to dying signals
During normal development, FGFs are considered survival signals for skeletal progenitors, but in vivo gain-of-function experiments via interdigital administration of an exogenous source of FGF showed transiently inhibited cell death [bib_ref] FGF Mediated Sulf1 Regulation, Zhao [/bib_ref] [bib_ref] In Vivo Inhibition of Programmed Cell Death by Local Administration of FGF-2..., Macias [/bib_ref] [bib_ref] Progressive Interdigital Cell Death: Regulation by the Antagonistic Interaction between Fibroblast Growth..., Hernández-Martínez [/bib_ref] ; but 24 h later, cell death increases dramatically. Considering the double effect of FGFs inhibiting and promoting cell death, we performed different protocols of gain-and loss-of-function experiments of FGF signaling.
When FGF2 (25 or 50 ng/mL) was maintained for 48 h in the medium, the rate of cell death was similar to that of the control untreated cultures. However, removal of FGF from the medium for a period of 3 or 6 h at the end of the second day of culture increased cell death. Remarkably, in contrast to the perinodular distribution of apoptotic and senescent cells in the control cultures, in FGF-treated cultures, degenerating cells were also associated with the regions of cell aggregation.
To further characterize the role of FGF in cell degeneration in the cultured progenitors, we performed pharmacological inhibition of FGF signaling with SU5402 and U0126 during the first two days of culture. Consistent with observations in the developing mouse limb [bib_ref] Progressive Interdigital Cell Death: Regulation by the Antagonistic Interaction between Fibroblast Growth..., Hernández-Martínez [/bib_ref] , cultures growing in a medium containing 800 ng/mL SU5402, which is a selective inhibitor of FGF receptor tyrosine kinase activity, increased cell death by 43%. Consistent with the anti-FGF influence of SU5402, the intensity of cell death by SU5402 was attenuated in the presence of exogenous FGF2 in the medium.
In contrast to SU5402, the addition of the MAPK inhibitor U0126 at a dose of 7.6 ng/mL (20 nM), alone or in combination with FGF, did not modify the rate or the intensity of cell death. Consistent with studies in vivo [bib_ref] Progressive Interdigital Cell Death: Regulation by the Antagonistic Interaction between Fibroblast Growth..., Hernández-Martínez [/bib_ref] , U0126 at a much higher concentration (50µM vs. 20 nM) increased cell death in the cultures (not shown). To further characterize the role of FGF in cell degeneration in the cultured proge tors, we performed pharmacological inhibition of FGF signaling with SU5402 and U01 during the first two days of culture. Consistent with observations in the dev oping mouse limb [bib_ref] Progressive Interdigital Cell Death: Regulation by the Antagonistic Interaction between Fibroblast Growth..., Hernández-Martínez [/bib_ref] , cultures growing in a medium containing 800 ng/mL SU54 which is a selective inhibitor of FGF receptor tyrosine kinase activity, increased cell dea
## Epigenetic influence of fgfs
We have previously observed an intense positive influence of FGF signaling on the limb expression of the epigenetic modulators Uhrf1 and Uhrf2. These genes encode factors that modulate transcriptional regulation and death sensitivity via chromatin modifica-tions [bib_ref] UHRF Genes Regulate Programmed Interdigital Tissue Regression and Chondrogenesis in the Embryonic..., Sanchez-Fernandez [/bib_ref]. In addition, FGFs transiently delay chondrogenic differentiation of limb skeletal progenitors as well as counteract the permanent antichondrogenic influence of WNT signaling secondary to the cytosine methylation of the Sox9 promoter [bib_ref] Fibroblast Growth Factor Maintains Chondrogenic Potential of Limb Bud Mesenchymal Cells by..., Kumar [/bib_ref]. Hence, to explore the importance of epigenetic modifications of differentiating progenitors in the outcome of the micromass cultures, we analyzed by MSRE-qPCR the cytosine methylations in the promoter of Sox9 and Scleraxis (Sclx) genes as master genes for chondrogenic and fibrogenic differentiation of the skeletal progenitors, respectively. In the control cultures, the promoter of Sclx maintained a low percentage of 5-mCs (0.24% at Day 1), and the rate increased moderately during the three days analyzed in our study. In contrast to Sclx, the promoter of Sox9 showed higher levels of methylation (3.97% of 5-mC at Day 1) that also increased in the course of culture. The addition of FGF2 (25 ng/mL) to the medium significantly decreased the methylation of the Sox9 promoterbut not that of Scleraxis. of cell death. Consistent with studies in vivo [bib_ref] Progressive Interdigital Cell Death: Regulation by the Antagonistic Interaction between Fibroblast Growth..., Hernández-Martínez [/bib_ref] , U0126 at a much higher concentration (50μM vs. 20 nM) increased cell death in the cultures (not shown).
## Epigenetic influence of fgfs
We have previously observed an intense positive influence of FGF signaling on the limb expression of the epigenetic modulators Uhrf1 and Uhrf2. These genes encode factors that modulate transcriptional regulation and death sensitivity via chromatin modifications [bib_ref] UHRF Genes Regulate Programmed Interdigital Tissue Regression and Chondrogenesis in the Embryonic..., Sanchez-Fernandez [/bib_ref]. In addition, FGFs transiently delay chondrogenic differentiation of limb skeletal progenitors as well as counteract the permanent antichondrogenic influence of WNT signaling secondary to the cytosine methylation of the Sox9 promoter [bib_ref] Fibroblast Growth Factor Maintains Chondrogenic Potential of Limb Bud Mesenchymal Cells by..., Kumar [/bib_ref]. Hence, to explore the importance of epigenetic modifications of differentiating progenitors in the outcome of the micromass cultures, we analyzed by MSRE-qPCR the cytosine methylations in the promoter of Sox9 and Scleraxis (Sclx) genes as master genes for chondrogenic and fibrogenic differentiation of the skeletal progenitors, respectively. In the control cultures, the promoter of Sclx maintained a low percentage of 5-mCs (0.24% at Day 1), and the rate increased moderately during the three days analyzed in our study. In contrast to Sclx, the promoter of Sox9 showed higher levels of methylation (3.97% of 5-mC at Day 1) that also increased in the course of culture. The addition of FGF2 (25 ng/mL) to the medium significantly decreased the methylation of the Sox9 promoterbut not that of Scleraxis.
## Transcriptional effects of fgfs
The survival role of FGFs deduced from SU5402 treatments, along with the intense degeneration caused by the removal of FGFs when the cultures were grown in a medium containing exogenous FGFs, prompted us to characterize the transcriptional modifications induced by FGFs in the skeletal progenitors by qPCR. For this purpose we selected a panel of genes associated with: chondrogenic (Sox9) and fibrogenic (Scleraxis) differentiation; mesodermal cell undifferentiation (Msx1, Msx2, Oct4); cell death (Bcl2, Bak1); cell senescence (p21, GBL1, Mmp2, Cathepsin D, Il-6); epigenetic regulators active in the embryonic limb (Dnmt1; Dnmt3A; Dnmt3B; TET3; HDAC2; HDAC3; HDAC8; and Prmt5); and components of the major signaling pathways active during digit skeletogenesis (Bmp2, Bmp4, Bmp5, Bmp7, Noggin, FgfR1, FgfR2, FgfR3, FgfR4, Wnt5a, and Dkk1). The efficiency of FGF treatments was confirmed by detecting the regulation of Sprouty1 [bib_ref] The Fibroblast Growth Factor Signaling Pathway, Ornitz [/bib_ref]. , no major changes were observed in genes associated with differentiation or undifferentiation in the FGF-treated cultures. Changes were not appreciated in Wnt5a. Minor changes were observed in FGF receptors and in the selected epigenetic regulators, in addition to significant down-regulation of FgfR2, Dnmt1, and HDAC3. As described in other experimental settings [bib_ref] Targeted Delivery of FGF2 to Subchondral Bone Enhanced the Repair of Articular..., Yang [/bib_ref] , various BMP ligands involved in programmed cell death, including Bmp2, Bmp4 and Bmp7, were intensely upregulated, while Bmp5 and the BMP antagonist Noggin were down-regulated. Dkk1 was also up-regulated. . Gene expression fold change values in FGF-treated versus control untreated, two-dayold micromass cultures. Fold change values are expressed as Mean ± SD. * p < 0.05; ** p < 0.01; *** p < 0.001. a very low expression level. Bold is regulated genes. Remarkably, the most important changes were detected in the expression of apoptotic and cell senescence factors, including members of the senescence-associated secretome (SASP). Thus, the pro-apoptotic factor Bak1 and Cyclin-Dependent Kinase Inhibitor 1 (p21), which is considered the main regulator of developmental senescence, and Cathepsin D and Il-6 were upregulated by more than twofold, while Bcl2, which is an antiapoptotic factor, was not regulated at significant levels. Beta-galactosidase (GBL1) was up-regulated but without reaching statistically significant levels.
## As shown in
## Gene
## Fgf2 modifies the response of cultured progenitors to bmp-and wnt-signaling
To analyze the crosstalk between FGF signaling and other signals active in the developing limb, we examined the effects of combined treatments with FGF2 and growth factors that regulate chondrogenic differentiation and cell death in the developing limb [bib_ref] Activation of the WNT-BMP-FGF Regulatory Network Induces the Onset of Cell Death..., Díaz-Hernández [/bib_ref]. We selected Bmp7, Wnt5a, and the WNT antagonist Dkk1 because they are highly expressed in the micromass culturesand show overlapping expression with regions of programmed cell death in vivo.
To analyze the crosstalk between FGF signaling and other signals active in the developing limb, we examined the effects of combined treatments with FGF2 and growth factors that regulate chondrogenic differentiation and cell death in the developing limb [bib_ref] Activation of the WNT-BMP-FGF Regulatory Network Induces the Onset of Cell Death..., Díaz-Hernández [/bib_ref]. We selected Bmp7, Wnt5a, and the WNT antagonist Dkk1 because they are highly expressed in the micromass culturesand show overlapping expression with regions of programmed cell death in vivo. BMPs have been identified as the apoptotic triggering signals for the undifferentiated embryonic limb mesoderm [bib_ref] BMPs Are Direct Triggers of Interdigital Programmed Cell Death, Kaltcheva [/bib_ref] but they are also growth-promoting signals for prechondrogenic aggregates [bib_ref] Antagonistic Effects of FGF4 on BMP Induction of Apoptosis and Chondrogenesis in..., Buckland [/bib_ref] [bib_ref] Activation of the WNT-BMP-FGF Regulatory Network Induces the Onset of Cell Death..., Díaz-Hernández [/bib_ref] [bib_ref] Genetic Analysis of the Roles of BMP2, BMP4, and BMP7 in Limb..., Bandyopadhyay [/bib_ref] [bib_ref] A New Role for BMP5 during Limb Development Acting through the Synergic..., Zuzarte-Luís [/bib_ref]. Our findings revealed that the proapoptotic influence of BMPs in two-day cultures was only detected in BMP7 treatments combined with FGF2 (182% double-treated vs. 108% BMP7-treated-only:. In a complementary fashion, the apoptotic-protective effect of the BMP antagonist NOGGIN also required combined treatments with FGFs. These findings suggest a permissive influence of FGF signaling on the regulation of apoptosis driven by BMP signaling.
## The fgf/wnt axis
WNT signaling exerts a critical role in limb outgrowth and ,. Members of the family are believed to exert a protective role against programmed cell death associated with the crosstalk between FGF and BMP signaling rather than a direct influence in the degenerative process [bib_ref] Activation of the WNT-BMP-FGF Regulatory Network Induces the Onset of Cell Death..., Díaz-Hernández [/bib_ref]. However, WNT5a, which signals via noncanonical pathway [bib_ref] Wnt-5a Inhibits the Canonical Wnt Pathway by Promoting GSK-3-Independent Beta-Catenin Degradation, Topol [/bib_ref] , and the WNT antagonist DKK1, which inhibits Wnt/β-catenin signaling, have been associated with cell death because they show specific expression domains in the areas of interdigital cell death [bib_ref] The Wnt Antagonist Dickkopf-1 Is Regulated by Bmp Signaling and c-Jun and..., Grotewold [/bib_ref].
In the micromass assay, endogenous expression of Wnt5a and Dkk1 was confirmed by in situ hybridizationB,C) prior to analysis of the influence of their exogenous administration to the medium. As observed in vivo [bib_ref] WNT5A-Ca2+-CaN-NFAT Signalling Plays a Permissive Role during Cartilage Differentiation in Embryonic Chick..., Farrera-Hernández [/bib_ref] , individual addition of WNT5A (100 ng/mL) or DKK1 (200 ng/mL) to the culture medium did not significantly increase the intensity of cell death. However, both treatments increased cell death when they were applied in combination with FGF2.
# Discussion
The culture of limb skeletal progenitors at high concentrations recapitulates embryonic skeletogenesis and is controlled in a fashion similar to that in vivo [bib_ref] From Skeletal Development to Tissue Engineering: Lessons from the Micromass Assay, Klumpers [/bib_ref] [bib_ref] Geometric Analysis of Chondrogenic Self-Organisation of Embryonic Limb Bud Cells in Micromass..., Rolfe [/bib_ref] [bib_ref] Patterns of Mesenchymal Condensation in a Multiscale, Christley [/bib_ref]. Our findings reveal that the differentiation of progenitors in this assay is accompanied by a patterned process of cell degeneration that delimits the initial regions of prechondrogenic aggregation. Cell death modulates tissue differentiation and morphogenesis during embryonic development of most organs, including limb skeletogenesis [bib_ref] Cell Death in the Developing Vertebrate Limb: A Locally Regulated Mechanism Contributing..., Montero [/bib_ref]. Our present results reveal that a comparable degeneration process accompanies the differentiation of chondroprogenitors in vitro. The initial formation of cartilaginous nodules in the micromass culture assay is accompanied by the elimination of progenitors located in the contour of the differentiating cartilage nodules. Remarkably, degeneration in this assay exhibits the same mechanistic features detected during in vivo tissue remodeling [bib_ref] Confluence of Cellular Degradation Pathways During Interdigital Tissue Remodeling in Embryonic, Montero [/bib_ref] [bib_ref] DNA Damage Precedes Apoptosis during the Regression of the Interdigital Tissue in..., Montero [/bib_ref] , which includes initial DNA damage, followed by the appearance of TUNEL-positive apoptosis and cell senescence positive for beta-galactosidase at pH 6. Apoptosis mediated by caspases is the major dying mechanism responsible for programmed cell death in embryonic systems. However, during the last decade cell senescence has gained interest in embryonic studies [bib_ref] Is Senescence-Associated β-Galactosidase a Reliable in Vivo Marker of Cellular Senescence During..., De Mera-Rodríguez [/bib_ref] [bib_ref] Apoptosis during Embryonic Tissue Remodeling Is Accompanied by Cell Senescence, Lorda-Diez [/bib_ref] [bib_ref] Complementary and Distinct Roles of Autophagy, Apoptosis and Senescence during Early Inner..., Varela-Nieto [/bib_ref] because it may represent a nonapoptotic cell degeneration involving the active participation of lysosomes [bib_ref] Confluence of Cellular Degradation Pathways During Interdigital Tissue Remodeling in Embryonic, Montero [/bib_ref]. The combined participation of apoptosis and cell senescence in this in vitro assay supports a similar relevance of both mechanisms for tissue remodeling in the course of tissue differentiation.
The internodular pattern of distribution of cell death and cell senescence in the control cultures is consistent with observations in monolayer cultures [bib_ref] Role of Chondrogenic Tissue in Programmed Cell Death and BMP Expression in..., Omi [/bib_ref] , suggesting that prechondrogenic aggregation serves a survival function for chondroprogenitors. A tempting explanation for the patterned dying process in the micromass culture assay is that the intense cell rearrangement involved in the formation of chondrogenic nodules might generate mechanical cues that contribute to modulating degeneration [bib_ref] Interplay between Mechanics and Signalling in Regulating Cell Fate, De Belly [/bib_ref]. It has been shown that mechanical stress generated during the initial stages of cell aggregation [bib_ref] Mechanical Feedback Defines Organizing Centers to Drive Digit Emergence, Parada [/bib_ref] together with BMP signaling [bib_ref] Visualization of Cartilage Formation: Insight into Cellular Properties of Skeletal Progenitors and..., Barna [/bib_ref] [bib_ref] BMP-Smad4 Signaling Is Required for Precartilaginous Mesenchymal Condensation Independent of Sox9 in..., Lim [/bib_ref] regulate the expression of Sox9 [bib_ref] Mechanical Loading Stimulates Chondrogenesis via the PKA/CREB-Sox9 and PP2A Pathways in Chicken..., Juhász [/bib_ref] , a master chondrogenic factor that protects chondroprogenitors from cell death [bib_ref] The Transcription Factor Sox9 Has Essential Roles in Successive Steps of the..., Akiyama [/bib_ref].
Experiments in vivo were suggestive of an antagonist effect of FGF and BMP signaling [bib_ref] Antagonistic Effects of FGF4 on BMP Induction of Apoptosis and Chondrogenesis in..., Buckland [/bib_ref]. Our findings revealed a survival influence of FGF signaling on the differentiating progenitors accompanied by the sensitization of cells to other apoptotic stimuli. Thus, cell death increased dramatically when FGF signaling was blocked by treatments with the FGF inhibitor SU5402 and also by addition and subsequent removal of FGF2 from the culture medium. Remarkably, the difference in the internodular distribution of senescent and apoptotic cells in control cultures indicates that increased apoptosis and senescence due to FGF removal take place at the expense of the peripheral cells of the prechondrogenic aggregates, which survive and form cartilage in the control untreated cultures. These observations support the role of FGF signaling in maintaining chondroprogenitor survival and proliferation, favoring either subsequent differentiation [bib_ref] Fibroblast Growth Factor and Bone Morphogenetic Protein Signaling Are Required for Specifying..., Kumar [/bib_ref] [bib_ref] Fibroblast Growth Factor Maintains Chondrogenic Potential of Limb Bud Mesenchymal Cells by..., Kumar [/bib_ref] or removal by cell death, depending on complementary signals active in specific spatial and temporal patterns. FGFs are master factors that maintain stemness in mesenchymal stem cells [bib_ref] Roles of FGF Signaling in Stem Cell Self-Renewal, Senescence and Aging, Coutu [/bib_ref] and are responsible for controlling embryonic limb outgrowth [bib_ref] The Roles of FGFs in the Early Development of Vertebrate Limbs, Martin [/bib_ref] [bib_ref] Conditional Inactivation of Fgfr1 in Mouse Defines Its Role in Limb Bud..., Verheyden [/bib_ref] [bib_ref] Fibroblast Growth Factor Receptor 2 (FGFR2)-Mediated Reciprocal Regulation Loop between FGF8 and..., Xu [/bib_ref]. Furthermore, in the developing limb there is a temporal association between the extinction of Fgf8 expression in the AER and establishment of the areas of interdigital cell death [bib_ref] Progressive Interdigital Cell Death: Regulation by the Antagonistic Interaction between Fibroblast Growth..., Hernández-Martínez [/bib_ref] [bib_ref] Morphological Diversity of the Avian Foot Is Related with the Pattern of..., Gañan [/bib_ref].
The transcriptional changes induced by FGF treatments in the micromass culture assay included the upregulation of factors that participate in the physiological degeneration of the interdigital tissue in the embryonic limb. The increased presence of these factors in the progenitors when the survival influence of FGFs is extinguished may contribute to the onset of degeneration. Among such degeneration-promoting factors upregulated by FGF-treatments are: p21, a specific marker of cell senescence [bib_ref] Apoptosis during Embryonic Tissue Remodeling Is Accompanied by Cell Senescence, Lorda-Diez [/bib_ref] , and Bak1, which is responsible for the activation of the mitochondrial apoptotic pathway [bib_ref] Progressive Interdigital Cell Death: Regulation by the Antagonistic Interaction between Fibroblast Growth..., Hernández-Martínez [/bib_ref]. The influence of these factors may be counteracted by protective factors regulated by FGF while present in the culture medium. An additional mechanism underlying the effects of FGF signaling is the changes in the epigenetic profile of progenitors. It has been shown that FGFs protect progenitors from irreversible methylation of the promoter of Sox9, favoring their subsequent differentiation into cartilage upon contact with pro-chondrogenic signals [bib_ref] Fibroblast Growth Factor Maintains Chondrogenic Potential of Limb Bud Mesenchymal Cells by..., Kumar [/bib_ref]. Our observations reveal a negative influence of FGF signaling on the expression of DNA methyltransferase 1 (Dnmt1) and histone deacetylase 3 (HDAC3), accompanied by a decrease in the methylation of the Sox9 promoter. DNA methyl transferase 3A (Dnmt3A) was also downregulated but did not show statistical significance. These changes would result in chromatin architectural modifications likely making progenitors more susceptible to DNA damage. This hypothesis is consistent with the importance of SOX9 for the survival of chondrogenic [bib_ref] The Transcription Factor Sox9 Has Essential Roles in Successive Steps of the..., Akiyama [/bib_ref] [bib_ref] Analysis of the Molecular Cascade Responsible for Mesodermal Limb Chondrogenesis: Sox Genes..., Chimal-Monroy [/bib_ref] and other cell populations [bib_ref] SOX9 Is Essential for Triple-Negative Breast Cancer Cell Survival and Metastasis, Ma [/bib_ref].
A major finding observed in the present study is the differential response of growing progenitors to BMP and WNT signaling depending on the presence in the medium of FGF2. There is compelling evidence supporting a central role of BMPs as death triggering factors for embryonic limb programmed cell death, including regression of the AER that is the source of FGFs [bib_ref] Antagonistic Effects of FGF4 on BMP Induction of Apoptosis and Chondrogenesis in..., Buckland [/bib_ref] [bib_ref] DNA Damage Precedes Apoptosis during the Regression of the Interdigital Tissue in..., Montero [/bib_ref] [bib_ref] Activation of the WNT-BMP-FGF Regulatory Network Induces the Onset of Cell Death..., Díaz-Hernández [/bib_ref] [bib_ref] BMPs Are Direct Triggers of Interdigital Programmed Cell Death, Kaltcheva [/bib_ref] [bib_ref] Genetic Analysis of the Roles of BMP2, BMP4, and BMP7 in Limb..., Bandyopadhyay [/bib_ref] [bib_ref] A New Role for BMP5 during Limb Development Acting through the Synergic..., Zuzarte-Luís [/bib_ref]. However, similar experimental approaches demonstrate an opposite role of BMP signaling in promoting the formation and growth of chondrogenic aggregates [bib_ref] Antagonistic Effects of FGF4 on BMP Induction of Apoptosis and Chondrogenesis in..., Buckland [/bib_ref] [bib_ref] BMP-Smad4 Signaling Is Required for Precartilaginous Mesenchymal Condensation Independent of Sox9 in..., Lim [/bib_ref]. These contradictory functions of BMPs generated controversy about the physiological hierarchy of the death-triggering machinery during interdigit regression [bib_ref] Interdigital Cell Death Function and Regulation: New Insights on an Old Programmed..., Hernández-Martínez [/bib_ref]. Under the experimental conditions of our study, BMP7 exerted a mild pro-apoptotic influence in the cultures that was potentiated when exogenous FGF2 was added to the culture medium. Considering that suppression of FGF signaling induces cell death and cell senescence in the cultures, our findings indicate that FGF cessation and BMP activation are two complementary pathways accounting for cell death of progenitors.
Wnt/β-catenin canonical signaling inhibits chondrogenesis via epigenetic silencing of the promoter of Sox9 [bib_ref] Fibroblast Growth Factor Maintains Chondrogenic Potential of Limb Bud Mesenchymal Cells by..., Kumar [/bib_ref] and stimulates limb outgrowth by promoting the expression of FGFs in the AER, which in turn maintain progenitor proliferation and survival [bib_ref] Comparative Analysis of the Expression and Regulation of Wnt5a, Fz4, and Frzb1..., Chimal-Monroy [/bib_ref]. The protective role of WNT signaling appears to be abolished in the areas of cell death due to the expression of the WNT antagonist Dkk1 which is expressed under the control of BMPs [bib_ref] The Wnt Antagonist Dickkopf-1 Is Regulated by Bmp Signaling and c-Jun and..., Grotewold [/bib_ref] [bib_ref] Comparative Analysis of the Expression and Regulation of Wnt5a, Fz4, and Frzb1..., Chimal-Monroy [/bib_ref] [bib_ref] Dickkopf1 Is Required for Embryonic Head Induction and Limb Morphogenesis in the..., Mukhopadhyay [/bib_ref]. However, in an opposite fashion, cell death in the developing hindbrain is inhibited by WNT antagonists via upregulation of BMP signaling [bib_ref] The WNT Antagonist CSFRP2 Modulates Programmed Cell Death in the Developing Hindbrain, Ellies [/bib_ref]. Our findings confirm the interaction between FGF and WNT signaling in the control of cell death [bib_ref] The Wnt Antagonist Dickkopf-1 Is Regulated by Bmp Signaling and c-Jun and..., Grotewold [/bib_ref]. We show that neither WNT5A nor DKK treatments modified cell death at significant levels in the absence of exogenous FGFs, but both factors induced apoptosis and cell senescence when they were combined with FGF2. It must be emphasized that WNT5A is a member of the WNT family, which signals via noncanonical inhibiting Wnt/β-catenin canonical signaling in the developing limb [bib_ref] Wnt-5a Inhibits the Canonical Wnt Pathway by Promoting GSK-3-Independent Beta-Catenin Degradation, Topol [/bib_ref]. Together, these results emphasize the importance of tuning WNT signaling to control the balance between differentiation and cell death proposed by Kumar and Lassar [bib_ref] Fibroblast Growth Factor Maintains Chondrogenic Potential of Limb Bud Mesenchymal Cells by..., Kumar [/bib_ref].
## Informed consent statement: not applicable.
# Data availability statement:
The data presented in this study are available on request from the corresponding author.
[fig] Figure 1: (A-A') In situ hybridization of the embryonic leg bud at stage 25, showing the expression of HoxA13. A' the tissue selected for the experiments is illustrated in red. (B-D) are dark field low magnification views of micromasses cultured for 24 (B), 48 (C), and 72 h (D). (B'-D') illustrate the same cultures after Alcian blue cartilage staining. Note the appearance of Alcian blue-positive nodules on Day 2 of culture. (E) 3-day-old micromass section immunolabeled for SOX9. (F) Flow cytometry histogram showing the cell cycle of 2 days cultures grown in DMEM. Scale bar in B−D'= 1 mm; scale bar in E= 100 μm. [/fig]
[fig] Figure 2: (A) Semithin section of a 3-day-old culture showing the presence of dark apoptotic (arrows) and vacuolated senescent cells (arrowheads) around a chrondrogenic nodule (Ch). (B tailed view of a micromass tissue showing senescent cells positive for SAβ-gal histochemistry counterstaining). (C) Graphic representation of the rate of cell death evaluated by flow cyto during the first 3 days of culture grown in DMEM-only. (D) Total tissue mass of micromass cu at Days 1, 3, and 7 evaluated by MTT staining. (E) Low-magnification view of 1-day culture SAβ-gal histochemistry (dark blue staining) showing the distribution of senescent cells. (F) Per ular restricted distribution of senescent cells positive for SAβ-gal in a 3-day-old micromass. (G tical section of 3-day-old micromass showing the arrangement of apoptosis (green TUNEL lab in the perinodular tissue. Note the intense positivity for SOX9 (red labeling) in chondrogenic a gates. (H) A 3-day-old micromass section immunolabeled for γH2AX (green) and SOX9 (r show the preferential distribution of γH2AX in perinodular tissue. Note the very reduced nu yH2AX labeling in the aggregated cells highly positive for SOX9 (arrows). (I-I") Detailed view SOX9-positive cell (I) with a couple of dots positive for γH2AX (I'). (I") is the merged image. Detailed view of an internodular cell with poor SOX9 labeling (J) but massive γH2AX labelin (J") is the merged image (J"). (K-L) Single channel and merged image of double labeling γH2AX(green) and phalloidin (red) showing the aligned cytoplasmic actin filaments in health genitors in contrast to the irregular aggregation of actin clumps (arrow) in the cytoplasm o highly positive for γH2AX (K). Scale bar in A and H = 20 μm; scale bar in B = 25 μm; scale ba and F = 300 μm; scale bar in G = 100 μm; scale bar I-J-K-L= 5 μm.Analysis by flow cytometry revealed that the number of dead cells was simil cultures of the same stage, i.e., the dead intensity was stage dependent (Figure 2CDay 1 of culture, 9.82% cells were dead. At Day 2, dead cells were 16.41% of the dissoc [/fig]
[fig] Figure 3: (A-A') Graphic representation of the rate of cell death evaluated by flow cytometry cultures treated for 2 days with 25 (light columns) or 50 ng/mL of FGF2 (red columns) to comp the effects of a continuous treatment with FGF (+FGF) and the effect of FGF withdrawal (-FGF) for (A) or 6 h (A'). The rate of cell death in the control untreated micromasses was considered 100 a is indicated by the dotted line. (B-B') Confocal view of the perinodular arrangement of TUNE positive cells (green labeling) in a labeled control 2-days culture. The sample is also labeled w phalloidin (red labeling). (B) is a merged image, and (B') shows only the green channel. (C-C') C focal view showing the widespread arrangement of TUNEL-positive cells (green labeling) after of FGF withdrawal in micromasses treated for 2 days with FGF2 (25 ngr/mL). Red labeling cor sponds to phalloidin. (C) is a merged image, and (C') shows only the green channel. (D,E) Detai view of the SAβ-gal arrangement in control (D) and experimental micromasses subjected to FG withdrawal (E). Note the perinodular distribution of SAβ-gal in the control in contrast with intranodular distribution in the experimental culture. (F) Graphic representation of the rate of death evaluated by flow cytometry in 2-day micromasses subjected to treatments with SU5402 ( ng/mL), SU5402 plus 25 ng/mL of FGF2, U0126 (7.6 ng/mL), and U0126 plus FGF2. Scale bar in B = 250 μm; scale bar in D−E = 50 μm. * p < 0.05 (versus control); ##, p < 0.01 (versus SU5402+FGF2) [/fig]
[fig] Figure 4: (A,B) Changes in the rate of cytosine methylation of the Scleraxis (A) and Sox9 (B) promoters during the first three days of micromass culture. (C,D) Changes in the rate of cytosine methylation of the Sox9 (C) and Scleraxis (D) promoters during the first two days of culture induced by the addition of FGF2 (25 ng/mL) to the culture medium. * p < 0.05; *** p < 0.001. [/fig]
[fig] Figure 5: (A-C) In situ hybridizations of 2-day-old micromass cultures showing the expression of Bmp7 (A), Wnt5a (B), and Dkk1 (C). (D) Graphic representation of the rate of cell death evaluated by flow cytometry in cultures treated for 2 days with BMP7 (200 ng/mL); BMP7 plus FGF2(25 ng/mL); NOGGIN (200 ng/mL); and NOGGIN plus FGF2 (25 ng/mL). The rate of cell death in untreated control cultures was considered to be 100% and indicated by the dotted line. (E) Graphic representation of the rate of cell death evaluated by flow cytometry in cultures treated for 2 days with WNT5A (100 ng/mL), WNT5A plus FGF2(25 ng/mL), DKK1 (200 ng/mL), and DKK-1 plus FGF2 (25 ng/mL). The rate of cell death in untreated control cultures was considered to be 100 and is indicated by the dotted line. Scale bar in A, B, C = 0.5 mm. *** p < 0.001. 3.6.1. The FGF/BMP Axis [/fig]
[fig] Author: Contributions: Conceptualization, J.M.H., J.A.M. and C.I.L.-D.; data curation, C.D.-O.; formal analysis, C.D.-O. and C.I.L.-D.; funding acquisition, J.A.M.; investigation, C.D.-O. and C.I.L.-D.; methodology, C.D.-O.; supervision, J.M.H., J.A.M. and C.I.L.-D.; writing-original draft, J.A.M. and C.I.L.-D.; writing-review and editing, J.M.H., J.A.M. and C.I.L.-D. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by a Grant (PID2021-125651NB-I00) from the Spanish Science and Innovation Ministry to J.A.M. C.D-O. is a recipient of a predoctoral grant from the University of Cantabria. Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, in accordance with Spanish legislation (RD53/2013) and approved by the ethics committee of the local government through the authorization corresponding to our research project (Code PI-03-18, March 2018). [/fig]
|
Effects of stretching exercises on human gait: a systematic review and meta-analysis
Background: Stretching is commonly used in physical therapy as a rehabilitation tool to improve range of motion and motor function. However, is stretching an efficient method to improve gait, and if so, for which patient category? Methods: A systematic review of randomized and non-randomized controlled trials with meta-analysis was conducted using relevant databases. Every patient category and every type of stretching programs were included without multicomponent programs. Data were meta-analysed where possible. Estimates of effect sizes (reported as standard mean difference (SMD)) with their respective 95% confidence interval (95% CI) were reported for each outcome. The PEDro scale was used for the quality assessment. Results: Twelve studies were included in the analysis. Stretching improved gait performance as assessed by walking speed and stride length only in a study with a frail elderly population, with small effect sizes (both SMD= 0.49; 95% CI: 0.03, 0.96; PEDro score: 3/10). The total distance and the continuous walking distance of the six-minute walking test were also improved only in a study in an elderly population who had symptomatic peripheral artery disease, with large effect sizes (SMD= 1.56; 95% CI: 0.66, 2.45 and SMD= 3.05; 95% CI: 1.86, 4.23, respectively; PEDro score: 5/10). The results were conflicting in healthy older adults or no benefit was found for most of the performance, spatiotemporal, kinetic and angular related variables. Only one study (PEDro score: 6/10) showed improvements in stance phase duration (SMD=-1.92; 95% CI: -3.04, -0.81), swing phase duration (SMD=1.92; 95 CI: 0.81, 3.04), double support phase duration (SMD= -1.69; 95% CI: -2.76, -0.62) and step length (SMD=1.37; 95% CI: 0.36, 2.38) with large effect sizes.
# Introduction
Gait is the medical term used to describe the human whole body movement of walking 1 . Gait involves internal and external forces that act on the body to move the center of mass (COM) across a given distance 2 . It depends on many biomechanical features that can be observed during gait analysis such as center of mass shift, joint range of motion (ROM), forces, muscle activity, joint moments, and joint powers 3 . Spatiotemporal features (e.g. velocity, step length, stride length, step with, step variability) and kinematics parameters (ROM) can be observed subjectively with functional evaluations by clinicians (e.g. the Tinetti test 4 or the timed up and go test 5 ), but, it can be further objectified with biomechanical analysis in a laboratory 2 . Kinetics variables (the forces that cause the body to move) must be collected in a laboratory environment with force plates (e.g. 6-9 for recent studies that used this technic).
Gait is a highly complex motor skill that is classically considered as an integrative measure and a predictor of health in older adults (e.g. 10; cf. 11 and also 12 for recent research topics on this matter). The loss of gait or its alteration with pathological conditions are known to be related to mortality, especially in the elderly (e.g. 13,14), stressing the importance of addressing gait disorders in physiotherapy. Gait requires body propulsion and balance control for safe progression, two "subtasks" that require the coordination of multiple skeletal muscles and the integration of sensory information arising from the vestibular, visual and somatosensory systems . As such, gait may expose populations with sensory or motor deficits to the risk of falling with serious consequences for health and autonomy. For these reasons, improving gait is a major aim in rehabilitation for most neurological/orthopaedic disorders, such as stroke or Parkinson's disease, and for frail older adults. Various therapeutic methods have been used to improve gait, such as resistance training The successful completion of numerous daily life activities is conditioned by the ability to move efficiently through a sufficient ROM 24 . Recent studies on gait initiation 25-27 and seat-to-stand task showed that the experimental restriction of postural chain ROM induced by orthosis wear in young healthy adults led to instability and lower motor performance. It is well established that ROM significantly decreases with aging 30-35 and more generally with reduced functional demand (e.g. sedentarity, immobilization, disease etc.) . Consequently, stretching has become an important part of many sport and rehabilitation programs to maintain or improve ROM, reduce stiffness and promote physical activity. This method has been applied in older adults 36,37 , patients with stroke 38 , Parkinson's disease 39 , multiple sclerosis 40 , plantar fasciitis 41 and spastic paraplegia 42 , for example. In sport programs, the influence of stretching on motor performance remains an issue of debate, although recent reviews conclude that maximal muscle performance (e.g. force, power, jump height, reaction time, etc.) is impaired primarily immediately after long durations of stretch (>90 seconds) . To date, no review has collected results on the relationship between stretching and locomotor performance in rehabilitation programs.
Hence, the purpose of this article is to analyse the effects of a stretching program on gait in each patient category by means of a systematic literature review and meta-analysis, comparing the gait outcomes of the intervention groups with the control groups. It will contribute to provide evidence-based practice from scientific data in order to integrate stretching in rehabilitation programs in a reasoned manner.
# Methods
## Design and literature screening
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology was employed in this systematic review 45 . A completed PRISMA checklist was submitted to an online repository (Reporting guidelines).
PubMed, Science Direct, Springer and Sage databases were used for a comprehensive systematic literature search for articles published prior to 28 April 2020 with no time limit. In addition, a manual search was conducted using the reference list of selected studies. The keywords used for the search strategy in PubMed were: "stretching" AND (gait OR walk). We included only articles published in English or French.
The selection procedure was conducted by two experts in rehabilitation (TV and AD). Disagreements were discussed with a third expert in a group until a mutual consensus was reached. First, a review was performed on all available titles obtained from the literature search with the selected keywords. All relevant or potentially relevant titles were included in the subsequent phase. Then, the abstracts were reviewed with all relevant or potential articles included in the following phase. Finally, full-text articles were reviewed to ensure that only relevant studies were included. In the same way, reference lists of all included articles were reviewed to possibly include articles through cross-referencing.
## Inclusion and exclusion criteria
We included randomized controlled trials (RCT) and controlled clinical trials (CCT) published in peer-reviewed journals that aimed to explore the effects of stretching on gait parameters.
## Amendments from version 1
We have added a whole paragraph at the beginning of the introduction section to specify how gait can be related to the different variables seen in the review.
No article in children with cerebral palsy fitted our inclusion criteria during the systematic search of the literature. We have added a whole paragraph in a specific section "limitations of the study" for explain it.
We have added some precisions in the discussion section (healthy older adults paragraph) to specify how we have limited the risk of bias.
We have added some precisions in the discussion section (young adults paragraph) to specify why stretching is less interesting to improve gait parameters in young healthy adults.
## Any further responses from the reviewers can be found at the end of the article
## Revised
We included all categories of subjects, all stretching techniques and different durations of treatment since standardized protocols are lacking in the purpose of the present study. Gait could be evaluated by functional tests, electromyographic (EMG) or biomechanical analysis. The following exclusion criteria were used: lack of gait assessment, non-application of muscle stretching, multimodal exercise programs, no control group, case report and review.
Data extraction and main measurements examined Data were extracted from the selected articles by one of the authors (TV). The extracted data were checked by another author (AD) and disagreements were resolved with a third (EY).
The following data were extracted for each selected article: (1) the names of the authors and the date of publication;
(2) the number of subjects involved in the experiment with their characteristics and breakdown in each group; (3) stretching training details (in the following order: number of participants, stretching technique, muscle groups stretched, number of sets, duration of stretch, frequency, protocol duration); (4) control group details; and (5) the main outcomes related to gait with the main results. When information could not be provided, it was indicated by a "?".
## Quality and risk of bias assessment
The PEDro scale was used to assess the risk of bias, and thus the methodological quality of the selected studies 46 . This scale was chosen for its ability to provide an overview of the external (criterion 1), internal (criteria 2-9) and statistical (criteria 9 and 10) validity of clinical trials. The scale is divided into 11 criteria, but the first is not calculated in the total score. The output of each criterion could be either "yes" (y), "no" (n) or "do not know" (?). A "y" was given a score of one point, while an "n" or "?" was assigned zero points. Studies with a total score of 5-10/10 (≥ 50%) were considered to be of high quality, and scores of 0-4/10 (<50%) of low quality 47 . Two evaluators independently assessed the quality of the included studies. In the event of disagreement, a group discussion was held with a third expert to reach a consensus.
# Statistical analysis
Estimates of effect sizes (comparing the intervention group and the control group) accompanied with a measure of statistical uncertainty (95% confidence interval [95% CI]) were calculated for each outcome when sufficient data were reported. Estimates of effect sizes were reported by standard mean difference (SMD) and their respective 95% CI. In this way, the magnitude of the overall effect can be quantified as trivial (<0.2), small (0.2-0.49), moderate (0.5-0.79) or large (≥0.8) . When data were lacking to calculate estimates of effect sizes, exact p values were reported.
When at least two studies used the same outcome, meta-analysis was performed, comparing the intervention groups with the control groups. When outcomes were identified in only one study, no meta-analysis could be performed but the effect of intervention was still calculated, reporting the estimate of effect size and its 95% confidence interval. Statistical analysis and figures (i.e. forest plot to facilitate the visualization of values) were produced using a random-effect model in Review Manager software (RevMan, v 5.3, Cochrane Collaboration, Oxford UK). A random-effect model was used to take into account heterogeneity between study effects. Statistical heterogeneity was calculated using the I 2 and Cochrane Q statistic tests 48 . Statistical significance was set at p<0.05.
## Level of evidence
The strength of evidence of primary outcomes was established as described by Van Tulder et al. 2003 50 based on effect size estimates with a measure of statistical uncertainty (SMD; 95% CI), statistical heterogeneity (I 2 ) when applicable (multiple studies) and risk of bias (PEDro scale). The level of evidence was considered strong with consistent findings among multiple high-quality RCT (at least two RCT with a PEDro score ≥5/10 that were statistically homogenous: I 2 p≥0.05). The level of evidence was considered moderate with consistent findings among multiple low-quality RCT and/or CCT (two trials with a PEDro score <5/10 that were statistically homogenous) and/or one high quality RCT. The level of evidence was considered limited when only one low quality RCT and/or CCT was identified. The level of evidence was conflicting when there was inconsistency among multiple trials (I 2 p < 0.05).
# Results
## Included studies
A total of 821 titles were screened in the first search stage, one more was included through cross-referencing, and 671 were excluded because they did not concern our research question. Following exclusion, 150 studies were considered for an abstract review. A further 105 were excluded in this second stage because they did not meet the inclusion criteria. Finally, 45 full-text articles were assessed for eligibility with 33 not accepted.
Thus, 12 articles were ultimately included in this systematic review. Six studies evaluated the effects of stretching in healthy older adults 23,51-55 , one in a frail elderly population 56 , one study in an elderly population with stable symptomatic peripheral artery disease 57 , one in stroke patients 58 , one study in adults with limited ankle ROM associated with a history of lower limb overuses injury 59 , one study in healthy adults with limited ankle dorsiflexion 60 and one in healthy young adults 61 . A summary of the studies selected is provided in, and their quality assessment is reported in. The results in different patient categories are reported below.
## Results in different patient categories healthy older adults description of the studies and quality assessment
Six studies examined the effects of stretching on healthy elderly subjects 23,51-55 . Regarding the characteristics of the subjects, the average sample size was 46.6±33.9 subjects (ranging from 19 23 to 96 subjects 53 ) and the mean age was 70.1±3.6 years (ranging from 65.40 52 to 75.30 years 23 ). Regarding the characteristics of the training programs, the average training duration was 8.6±2.7 weeks (ranging from 4 52 to 12 weeks 54 ), with an average frequency of 8.3±6.2 sessions per week (ranging from 2 54 to 14 sessions 51,53,55 ). The average number of sets per session was 4.5±2.8 sets (ranging from 2 55 to 10 sets 23 ), with an average stretching time of 45.0±18.9 seconds (ranging from 15 23 to 60 seconds 52,54,55 ). Static stretching was provided in all studies. The muscle groups stretched were the hip flexors 51-55 , ankle plantar flexors 23,51,52,54 , ankle dorsiflexors 54 , hip extensors 52 , knee extensors and flexors 54 . There was great heterogeneity in gait outcomes. Angular variables during gait included peak hip extension 51,53,55 , ankle plantar flexion during gait 53 , ankle ROM during gait 52 , anterior pelvis tilt 52,55 , knee ROM, pelvic rotation, lateral pelvic tilt and hip ROM 52 . Spatiotemporal variables were: gait speed 51,52,55 , stance and swing duration, double support phases, step length 52 and stride length 52,55 . Kinetic variables were hip torque and ankle plantar flexion power 53 . Finally, two functional tests were used: the 10-meter walk test (10MWT) 23 and the 6-minute walk test (6MWT) . Regarding the quality of the studies, the average PEDro score was 4.6±1.6 and one study was identified as a non-randomized trial 54 . The range of score varied from 3 54,55 to 7 23 . Healthy adults n: criterion not fulfilled; y: criterion fulfilled; ?: criterion not mentioned; total score: each item (except the first) contributes 1 point to the total score, yielding a PEDro scale score that can range from 0 to 10.
## Meta-analyses
Four meta-analysis were conducted for the following outcomes: gait speed, stride length, hip extension during gait and anterior pelvic tilt.
Gait speed: For gait speed, two studies were included in the meta-analysis 51,52 . One study was excluded because intervention and control groups were not similar at baseline 55 . Statistical analysis showed no significant difference between groups (SMD= 0.45; 95% CI: -1.15, 2.06), with heterogeneous results (I 2 =86%, p=0.007). Thus, the level of evidence was conflicting.
Stride length: For stride length, two studies were included in the meta-analysis 51,55 . Statistical analysis showed no significant difference between groups (SMD= 0.22; 95% CI: -0.44, 0.88), with consistent results (I 2 =59%, p=0.12).
Only one study was of high quality 51 , thus a moderate level of evidence supports the lack of beneficial effect of stretching to improve stride length in the elderly.
Hip extension: For hip extension during gait (kinematic data), three studies were included in the meta-analysis 51,53,55 . Statistical analysis showed no significant difference between groups (SMD= 0.20; 95% CI: -0.06, 0.47), with consistent results (I 2 =0%, p=0.99). Two studies were of high quality 51,53 , thus a strong level of evidence supports the lack of beneficial effect of stretching to improve hip ROM during gait in the elderly.
Anterior pelvic tilt: For anterior pelvic tilt, three studies were included in the meta-analysis 52,53,55 . Statistical analysis showed no significant difference between groups (SMD= -0.70; 95% CI: -1.60, 0.21), with heterogeneous results (I 2 =87%, p<0.01). Thus, the level of evidence was conflicting.
## Effects of interventions in other outcomes
For the outcomes below, no meta-analysis could be performed because only one study was identified. Nevertheless, for each outcome, effect size estimates with a measure of statistical uncertainty (95% CI) were provided.
## Frail elderly description of the study and quality assessment
The study of Watt et al. 2011 examined the effects of stretching on frail elderly subjects 56 . Regarding the characteristics of the subjects, 74 subjects were included, and the mean age was 77.0±8.0 years. Regarding the characteristics of the training programs, the stretching program lasted ten weeks, with a frequency of 14 sessions per week (two sessions per day). Participants performed two sets per session, holding the stretch for 60 seconds (static stretching), alternating the right and left limb (four minutes in total). The muscle group stretched was the hip flexors. The outcomes were cadence (steps/minute), walking speed (meters/second), stride length (meters) peak hip extension (degree) and peak anterior pelvic tilt (degree). Regarding the quality assessment, the study was identified as RCT and had an average PEDro score of 3 (low level of evidence).
## Effects of intervention
## Elderly with symptomatic peripheral artery disease description of the study and quality assessment
The study of Hotta et al. (2019) examined the effects of stretching in elderly with symptomatic peripheral artery disease 57 . Regarding the characteristics of the subjects, 13 subjects were included and the mean age was not mentioned. Regarding the characteristics of the training programs, the stretching program lasted four weeks, with a frequency of five sessions per week. Participants performed one set daily, holding the stretch for 30 minutes (static stretching with splints). The muscle group stretched was ankle plantar flexors. The gait outcome was 6MWT. Regarding the quality assessment, the study was identified as RCT and had an average PEDro score of 5 (moderate level of evidence).
## Effects of intervention
## Stroke
## Description of the study and quality assessment
The study of Kim et al. (2013) examined the effects of stretching on stroke patients 58 . Only a static muscle stretching training group and control group were included in the analysis. Regarding the characteristics of the subjects, 24 subjects were included, and the mean age was 53.3±3.1 years. Regarding the characteristics of the training programs, the stretching program lasted six weeks, with a frequency of four sessions per week. Participants performed one set per session, holding the stretch for 20 minutes (static stretching). The muscle group stretched was ankle plantar flexors. The outcome was the sway of the centre of pressure during the stance phase.
Regarding the quality assessment, the study was identified as CCT and had an average PEDro score of 3 (low level of evidence).
## Effects of intervention
The study of showed no significant difference between groups in the sway of the centre of pressure (SMD=0.75; 95% CI: -0.09, 1.58).
## Young adults with limited ankle rom and a history of lower limb overuse injury description of the study and quality assessment
The study of Johanson et al. examined the effects of stretching on healthy adults with limited passive ankledorsiflexion ROM (less than eight degrees) and a history of lower limb overuse injury 59 . Regarding the characteristics of the subjects, 19 subjects were included and the mean age was 30.3±9.8 years. Regarding the characteristics of the training programs, the stretching program lasted three weeks, with a frequency of two sessions per day. Participants performed five sets per session, holding the stretch for 30 seconds (static stretching). The muscle group stretched was ankle plantar flexors. The outcomes were ankle dorsiflexion and time-to-heel-off during the stance phase of gait. Regarding the quality assessment, the study was identified as RCT and had an average PEDro score of 5 (moderate level of evidence).
## Effects of intervention
The study of Johanson et al. showed no significant difference between groups in ankle dorsiflexion during gait in both the right and left ankle (SMD= 0.50; 95% CI: -0.42, 1.43 and SMD= 0.41; 95% CI: -0.52, 1.33 respectively). There was also no significant difference between groups for time-to-heel-off during the stance phase of gait in both the right and left ankle (SMD= -0.50; 95% CI: -1.43, 0.43 and SMD= -0.48; 95% CI: -1.41, 0.45 respectively). In the current study, 16 subjects were included, and the mean age was 27.4±8.2 years. The muscle group stretched was the ankle plantar flexors. The outcomes were maximum ankle dorsiflexion, maximum knee extension and EMG amplitude of the gastrocnemius during the stance phase of gait. Regarding the quality assessment, the study was identified as RCT and had an average PEDro score of 6 (moderate level of evidence).
## Young adults with limited ankle rom description of the study and quality assessment
## Effects of intervention
## Healthy young adults description of the study and quality assessment
The study of examined the effects of stretching on healthy young adults 61 . Only a static hip extension stretching group and control group were included in the analysis. Regarding the characteristics of the subjects, 16 subjects were included, and the mean age was 21.0±1.0 years. Regarding the characteristics of the training programs, the stretching program lasted three weeks, with a frequency of two sessions per week. Participants performed three sets per session, holding the stretch for two minutes (static stretching). The muscle group stretched was the hip flexors. The outcome was walking economy (ml/kg/min). Regarding the quality assessment, the study was identified as RCT and had an average PEDro score of 5 (moderate level of evidence).
## Effects of intervention
The study of showed no significant difference between groups in gait economy in terms of oxygen consumption (SMD= 0.83; 95% CI: -0.21, 1.87).
# Discussion
The aim of this systematic review was to determine the effects of a stretching program on human gait by means of a systematic literature review and meta-analysis. Twelves studies were identified in six different patient categories. Statistical analyses showed no strong level of evidence supporting the beneficial effect of a stretching program to improve any gait outcome. The major issue in conducting meta-analyses and establishing strong level of evidences was the great heterogeneity in gait variables. The results obtained in the different patient categories are discussed in detail below.
## Healthy older adults
The healthy older adult population was the most studied. Two muscle groups were systematically stretched in the six identified studies: hip flexors 23,51-53,55 and ankle plantar flexors 23,51-53 . Hip flexor stiffness, associated with reduced hip extension during gait has been demonstrated in the elderly and may alter gait 62,63 . In the same way, decreased calf muscle length associated with restricted dorsiflexion ROM is well documented in older adults 35,64 . A decreased ankle dorsiflexion ROM has been correlated with poorer balance test scores in the elderly 65 and may contribute to an increased risk of falls 66 . All the studies included in the present analysis showed that specific stretching programs were efficient to improve passive ROM of the targeted joints, but results are more heterogeneous regarding gait performance and dynamic ROM. This led to inconsistency in the results or the impossibility to conclude with a strong level of evidence that a stretching program improves gait in healthy older adults. Moreover, when improvement in ROM or gait performance occurred, it was not associated with a significant increase in dynamic hip extension or ankle dorsiflexion. Only trends toward increased dynamic ROM after stretching interventions were observed 51,53,55 . This observation was consistent in young adults.
When data were meta-analyzed, we ensured that the groups and the training characteristics were similar to limit the risk of bias. This explains that a limited number of studies was included in the meta-analysis. It is worth noticed that the stretching technic was the same (i.e. static stretching), but that details of interventions varied across these studies. For example, both studies selected for the meta-analysis of gait speed included hip flexors and plantar flexors stretching, but, one study included hip extensor stretching 52 whereas the other did not 51 . This difference may partially explain the heterogenous results in the meta-analysis (I 2 =86%, p=0.007). In the same way, the heterogenous results observed in the meta-analysis of anterior pelvic tilt (I 2 =87%, p<0.01) may be explained by the stretching of additional muscle groups (hip extensors and plantar flexors) in the study of Cristopoliski et al. (2009) compared to the other studies (in which only hip flexors were stretched) 53,55 . Nevertheless, heterogeneity in the results was not systematically observed between studies that used slight different protocols, as showed by the consistent results int the meta-analyses of stride length (I 2 =59%, p=0.12) and hip extension (I 2 =0%, p=0.99). Thus, we assume that we have limited the risk of bias in the meta-analyses.
## Stroke patients
In stroke patients, ankle plantar flexor stretching has been successfully used to improve ankle stiffness 67-70 . Decreased plantar flexors stiffness may have a beneficial effect on postural control during gait because triceps surae is known to play an important role during gait 71-73 and an increase in muscle stiffness might alter synergistic muscle activities during human gait. However, only one non-randomized study 58 was identified and included in the current systematic review. Other studies that used stretching in multicomponent programs 74-76 or in control groups 77,78 were identified but excluded because of the addition of resistance training or the lack of a control group. Nevertheless, it should be noted that some studies showed improvements between pre-and post-stretching conditions. Other authors showed that one week of immobilization in dorsiflexed position (casting) followed by one week of plantar flexor stretching and gait training improved gait performances in 10MWT and 6MWT 74 . Hence, these encouraging results suggest that further randomized controlled trials of good quality are needed to explore the ability of ankle stretching to improve gait parameters in stroke or in other neurological diseases exposing patients to joint stiffness, e.g. Parkinson's disease 79 .
## Young adults
In healthy adults, the interest of practicing stretching to improve gait seems limited as they are assumed to have sufficient mobility for walking. Moreover, the included study involved athletic males 61 , a population that is known to be more flexible than inactive persons 80 . Stretching should be more indicated when ROM is limited 24 . However, even in young adults with limited ankle ROM, stretching did not improve dynamic dorsiflexion during gait 59,60 . Stretching programs in apparently healthy adults should be more indicated after a prolonged period of reduced functional demand (e.g. immobilization, sedentarity), when ROM is insufficient to practice a specific activity or when high levels of flexibility are required for sport performance (e.g. gymnastics or dance) and in sports that involve stretch-shortening cycles (e.g. basketball, volleyball) 24 .
Limitations of the study Some patient categories were not included in the present review, although muscle stretching is commonly indicated in their clinical care to reduce spasticity 81 . This is for example the case for children with cerebral palsy 82 . In fact, we were able to identify studies in the literature focusing on the effects of stretching on gait in this population during the first phase of the present review, but the protocol of these studies combined stretching with another form of training (e.g. 83,84) or there was no control group (e.g. 85). These studies therefore did not fit with the inclusion criteria of the present systematic review and were consequently excluded. Now, it should be stressed that the effectiveness of static stretching to improve motor function in children with cerebral palsy is still controversial 86 , although some authors showed that functional stretching exercises may be effective to improve gait 85 . Further randomized controlled trials are needed to explore the impact of stretching on gait in this population.
To reduce the risk of bias, data were meta-analyzed when at least 2 studies with similar populations and training characteristics were found. Considering these constraints and the great heterogeneity in the gait outcomes, we only performed metaanalyses in healthy older adults for 4 outcomes. The Cochrane Qualitative and Implementation Methods Group recommends the application of Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) in the Evidence from Qualitative Reviews to assess the level of confidence in systematic review with meta-analyses 87 . However, the GRADE necessitates assessing the risk of publication bias with a funnel plot, determining its asymmetry, which can be performed with at least 10 studies 88 . In the present study, the meta-analyses included less that 10 randomized controlled trials, so, we chose to implement other guidelines described by a Cochrane collaboration group to assess the level of evidence 50 . Because this method includes fewer criteria, our confidence in the results must be taken with caution
# Conclusion
Twelve studies were identified, involving a total of 442 subjects. Despite some improvements in isolated studies, statistical analyses showed no strong level of evidence supporting the beneficial effect of using stretching alone to improve gait outcomes in rehabilitation programs. The major obstacle in conducting meta-analyses and establishing strong levels of evidence were the great heterogeneity in gait variables and the low quality of the included studies. Because the effects of stretching are not clear, further randomized controlled trials of good The purpose of this article is to analyse the effects of a stretching program on gait in each patient category by means of a systematic literature review and meta-analysis, comparing the gait outcomes of the intervention groups with the control groups. This is a very interesting research direction and it has good innovation. But different ages and diseases have different gait results and different stretching training modes and intervention methods. How to eliminate the bias caused by these differences in meta-analysis and comparison?
1.
Four meta-analysis were conducted for the following outcomes: gait speed, stride length, hip extension during gait and anterior pelvic tilt. Why choose these four dimensions? Is there any theoretical basis?
2.
Kinetic variables: The study of 44 showed no significant difference between groups for hip torque (SMD= 0.35; 95% CI: -0.06, 0.75) and ankle plantar flexion power (SMD=0.00; 95% CI: -0.40, 0.40), with a moderate level of confidence (PEDro score: 6/10).: It seems that gait analysis cannot directly measure muscle power. sEMG can only evaluate muscle recruitment signals to evaluate muscle fiber contraction. Therefore, how is ankle flexion power measured by gait analysis?
3.
Twelve studies were included in the analysis. Stretching improved gait performance as assessed by walking speed and stride length only in a study with a frail elderly population, with small effect sizes. There is no strong evidence supporting the beneficial effect of using stretching to improve gait. I think that since stretching can improve gait parameters for adults with special weakness, the effect of stretching on gait is significant, although it has little effect on healthy adults or young adults. Should it be explained? In general, this study made a meta-analysis on the effect of stretching on gait, which has good innovation. However, whether gait analysis has guiding significance for different groups of rehabilitation training is still controversial especially for healthy adults, so the research conclusion of this paper has practical significance in clinical guidance.
## 4.
## Are the rationale for, and objectives of, the systematic review clearly stated? yes
Are sufficient details of the methods and analysis provided to allow replication by others? Yes
## Is the statistical analysis and its interpretation appropriate? partly
Are the conclusions drawn adequately supported by the results presented in the review?
Partly example, in the study of Watt et al. (2011), the intervention group had a significantly higher gait speed than the control group, so the trial was excluded of the meta-analysis. Comment 4. I think that since stretching can improve gait parameters for adults with special weakness, the effect of stretching on gait is significant, although it has little effect on healthy adults or young adults. Should it be explained?
## Reply:
We add some precisions in the discussion section (young adults paragraph) to specify why stretching is less interesting to improve gait parameters in young healthy adults:
Please see L387-390: In healthy adults, the interest of practicing stretching to improve gait seems limited as they are assumed to have sufficient mobility for walking. Moreover, the included study involved athletic males 61 , a population that is known to be more flexible than inactive persons 81 . Stretching should be more indicated when range of motion is limited 24 . Overall, the methodology is correct, and the meta-analysis well described. The results show weak effects of stretching on gait parameters.
The main issue arises from the various parameters which are considered in all the studies, both during gait initiation and straight walking (spatiotemporal parameters, kinematics, joint strength and dynamics, muscle activity, etc…). Therefore, in the results and the discussion, grouping the types of parameters according to their role in the process of walking would add value to this manuscript. The effects, demonstrated or not, have not all the same meaning depending on whether functional parameters or kinematics angles or even muscle strength and activity are considered.
Gait should be defined in a better way, especially how it is related to the different measures in terms of processes, even if there is anyway no real benefit demonstrated by stretching used as a unique therapy. Surprisingly, no article in children with cerebral palsy has been uncovered, while the clinical care of these children is mostly based on muscle stretching.
## Are the rationale for, and objectives of, the systematic review clearly stated? yes
Are sufficient details of the methods and analysis provided to allow replication by others? Yes
## Is the statistical analysis and its interpretation appropriate? yes
Are the conclusions drawn adequately supported by the results presented in the review? degrees) 51 . It is worth noticed that these participants were not the same than in the study of Johanson et al. . In contrast, the characteristics of the training programs were the same as in Johanson et al. (2006)".
Comment 2. In the results and the discussion, grouping the types of parameters according to their role in the process of walking would add value to this manuscript. The effects, demonstrated or not, have not all the same meaning depending on whether functional parameters or kinematics angles or even muscle strength and activity are considered.
## Reply:
In the results and the discussion, we chose to group the participants by patient categories because differences in ages and health parameters may result in different training effects. However, in each patient category, each variable was considered separately in an organized way.
Comment 3. Gait should be defined in a better way, especially how it is related to the different measures in terms of processes, even if there is anyway no real benefit demonstrated by stretching used as a unique therapy.
Reply: We agree. We add a whole paragraph at the beginning of the introduction section to specify how gait can be related to the different variables seen in the review:
Please see L35-44: "Gait is the medical term used to describe the human whole body movement of walking 1 . Gait involves internal and external forces that act on the body to move the center of mass (COM) across a given distance 2 . It depends on many biomechanical features that can be observed during gait analysis such as center of mass shifts, joint range of motion, forces, muscle activity, joint moments, and joint powers 3 . Spatiotemporal features (e.g. velocity, step length, stride length, step with, step variability) and kinematics parameters (range of motion) can be observed subjectively with functional evaluations by clinicians(e.g. the Tinetti test 4 or the timed up and go test 5 ), but, it can be further objectified with biomechanical analysis in a laboratory 2 . Kinetics variables (the forces that cause the body to move) must be collected in a laboratory environment with force plates (e.g. 6-9 for recent studies that used this technic)".
Comment 4. Surprisingly, no article in children with cerebral palsy has been uncovered, while the clinical care of these children is mostly based on muscle stretching.
Reply: We agree. However, no article in children with cerebral palsy fitted our inclusion criteria during the systematic search of the literature. We add a whole paragraph in a specific section "limitations of the study":
Please see L397-406: Some patient categories were not included in the present review, although muscle stretching is commonly indicated in their clinical care to reduce spasticity 82 . This is for example the case for children with cerebral palsy 83 . In fact, we were able to identify studies in the literature focusing on the effects of stretching on gait in this population during the first phase of the present review, but the protocol of these studies combined stretching with another form of training (e.g. 84,85 ) or there was no control group (e.g. . These studies therefore did not fit with the inclusion criteria of the present systematic review and were consequently excluded. Now, it should be stressed that the effectiveness of static stretching to improve motor function in children with cerebral palsy is still controversial 87 , although some authors showed that functional stretching exercises may be effective to improve gait 86 . Further randomized controlled trials are needed to explore the impact of stretching on gait in this population.
Best regards Arnaud Delafontaine on behalf of all the authors
## Competing interests: none
The benefits of publishing with F1000Research:
Your article is published within days, with no editorial bias - You can publish traditional articles, null/negative results, case reports, data notes and more - The peer review process is transparent and collaborative - Your article is indexed in PubMed after passing peer review - Dedicated customer support at every stage - For pre-submission enquiries, contact [email protected] |
Impact of inactivity and exercise on the vasculature in humans
The effects of inactivity and exercise training on established and novel cardiovascular risk factors are relatively modest and do not account for the impact of inactivity and exercise on vascular risk. We examine evidence that inactivity and exercise have direct effects on both vasculature function and structure in humans. Physical deconditioning is associated with enhanced vasoconstrictor tone and has profound and rapid effects on arterial remodelling in both large and smaller arteries. Evidence for an effect of deconditioning on vasodilator function is less consistent. Studies of the impact of exercise training suggest that both functional and structural remodelling adaptations occur and that the magnitude and time-course of these changes depends upon training duration and intensity and the vessel beds involved. Inactivity and exercise have direct ''vascular deconditioning and conditioning'' effects which likely modify cardiovascular risk.
# Introduction
Physical activity, inactivity and cardiovascular risk Contemporary westerners have reached an historical pinnacle of physical inactivity and further technological change is likely to reinforce this [bib_ref] Exercise and gene expression: physiological regulation of the human genome through physical..., Booth [/bib_ref]. Physical inactivity is an independent risk factor for atherosclerosis, cardiovascular diseases and diabetes [bib_ref] Changes in physical fitness and allcause mortality. A prospective study of healthy..., Blair [/bib_ref] [bib_ref] Physical activity and diabetes prevention, Lamonte [/bib_ref] [bib_ref] A prospective study of walking as compared with vigorous exercise in the..., Manson [/bib_ref] and low cardiopulmonary fitness is a strong independent predictor of all-cause mortality [bib_ref] Relationship between low cardiorespiratory fitness and mortality in normal-weight, overweight, and obese..., Wei [/bib_ref]. Sedentary living is estimated to be responsible for approximately onethird of deaths due to coronary heart disease, colon cancer, and type 2 diabetes [bib_ref] The public health burdens of sedentary living habits: theoretical but realistic estimates, Powell [/bib_ref]. Physical inactivity is therefore a key factor in the etiology and progression of chronic diseases, including cardiovascular and metabolic diseases which are common, debilitating and costly. Given the low daily energy expenditure which is characteristic of modern living [bib_ref] Exercise and gene expression: physiological regulation of the human genome through physical..., Booth [/bib_ref] , the consequences of physical inactivity seem likely to worsen. Regular physical exercise is associated with reduction in primary [bib_ref] Effects of walking on coronary heart disease in elderly men: the Honolulu..., Hakim [/bib_ref] [bib_ref] Exercise capacity and mortality among men referred for exercise testing, Myers [/bib_ref] [bib_ref] Physical activity, all-cause mortality, and longevity of college alumni, Paffenbarger [/bib_ref] [bib_ref] Physical activity and coronary heart disease in men: the Harvard Alumni Health..., Sesso [/bib_ref] and secondary vascular events [bib_ref] Exercise-based rehabilitation for coronary heart disease, Jolliffe [/bib_ref] [bib_ref] Cardiac rehabilitation after myocardial infarction. Combined experience of randomized clinical trials, Oldridge [/bib_ref]. Meta-analyses, including those of exercise-based cardiac rehabilitation undertaken in the contemporary statin, ACE inhibitor and revascularisation era, indicate that a *30% exercise-related benefit is evident in terms of cardiac events, relative to usual care [bib_ref] Exercise and cardiovascular risk reduction: time to update the rationale for exercise?, Green [/bib_ref]. Indeed, some expert opinion suggests that the most physically active individuals demonstrate CAD rates half those of the most sedentary [bib_ref] Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular..., Thompson [/bib_ref]. This magnitude of benefit approximates or exceeds that associated with antihypertensive or lipid lowering interventions in large multicenter trials [bib_ref] Effects of different blood-pressure-lowering regimens on major cardiovascular events: results of prospectivelydesigned..., Turnbull [/bib_ref] [bib_ref] Effectiveness of statin therapy in adults with coronary heart disease, Wilt [/bib_ref]. It is well established that individuals with higher cardiopulmonary fitness exhibit lower cardiovascular disease rates than unfit individuals [bib_ref] Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular..., Thompson [/bib_ref] , with the relative risk of being unfit exceeding that associated with smoking, elevated systolic blood pressure, hypercholesterolemia and overweight/obesity [bib_ref] Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause..., Blair [/bib_ref]. These data indicate that exercise training and maintenance of physical fitness have important impacts on the prevalence and progression of cardiovascular diseases in humans.
Does risk factor modulation explain the effects of inactivity and exercise on CV risk?
Detrimental changes in cardiovascular risk factors have been suggested to explain the increased cardiovascular risk associated with chronic inactivity [bib_ref] A prospective assessment of mortality in chronic spinal cord injury, Garshick [/bib_ref] [bib_ref] The relationship between neurological level of injury and symptomatic cardiovascular disease risk..., Groah [/bib_ref] [bib_ref] Standardized indices of mortality among persons with spinal cord injury: accelerated aging..., Imai [/bib_ref]. Human models of short-term physical inactivity (4-8 weeks) demonstrate no change in blood pressure, obesity, BMI or cholesterol levels [bib_ref] WISE 2005: chronic bed rest impairs microcirculatory endothelium in women, Demiot [/bib_ref] [bib_ref] Regulation of muscle sympathetic nerve activity after bed rest deconditioning, Pawelczyk [/bib_ref] [bib_ref] A causal role for endothelin-1 in the vascular adaptation to skeletal muscle..., Thijssen [/bib_ref]. Subjects with a spinal cord injury (SCI), who are predisposed to develop cardiovascular diseases [bib_ref] A prospective assessment of mortality in chronic spinal cord injury, Garshick [/bib_ref] [bib_ref] The relationship between neurological level of injury and symptomatic cardiovascular disease risk..., Groah [/bib_ref] [bib_ref] Standardized indices of mortality among persons with spinal cord injury: accelerated aging..., Imai [/bib_ref] , have similar cholesterol, triglyceride and blood pressure levels to healthy subjects [bib_ref] Coronary artery disease: metabolic risk factors and latent disease in individuals with..., Bauman [/bib_ref] [bib_ref] Coronary risk in spinal cord injury: assessment following a multivariate approach, Cardus [/bib_ref] [bib_ref] Pressor dose responses and baroreflex sensitivity in quadriplegic spinal cord injury patients, Krum [/bib_ref] [bib_ref] Different risk factor patterns for metabolic syndrome in men with spinal cord..., Liang [/bib_ref]. Recently, prevalence rates of the metabolic syndrome and traditional cardiovascular risk factors were compared between 185 SCI subjects and age-, gender-and race-matched controls. No differences were found for traditional risk factors between both groups and SCI patients demonstrated lower levels of blood cholesterol and glucose [bib_ref] Different risk factor patterns for metabolic syndrome in men with spinal cord..., Liang [/bib_ref]. These results suggest that the strong link between physical inactivity and cardiovascular mortality and morbidity cannot be fully explained by an effect of inactivity on traditional cardiovascular risk factors.
The effects of exercise on conventional risk factors are substantially less than those achieved by pharmacological therapies and also much less than that required to explain the mortality benefits associated with exercise and fitness [bib_ref] Exercise and cardiovascular risk reduction: time to update the rationale for exercise?, Green [/bib_ref]. For example, the magnitude of decrease in LDL required to explain a 30% mortality benefit, approximates 25% (LIPID study group 1998). The impact of exercise training on LDL is typically \5% [bib_ref] Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular..., Thompson [/bib_ref]. In addition, improvement in traditional risk factors cannot solely account for the magnitude of risk reduction associated with exercise training, since the association with reduced mortality is independent of these risk factors [bib_ref] Exercise and cardiovascular health: get active to ''AKTivate'' your endothelial nitric oxide..., Dimmeler [/bib_ref] [bib_ref] Exercise as cardiovascular therapy, Shephard [/bib_ref]. Indeed, a recent analysis of 27,000 subjects in the Women's Health Study reported that differences in risk factors explained 59% of the relative cardiovascular risk reduction associated with exercise [bib_ref] Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms, Mora [/bib_ref]. The impact of hemoglobin A1c, lipid sub profiles, lipoprotein (a), apolipoprotein A1, apolipoprotein B-100, creatinine, homocysteine, hs-CRP, fibrinogen, s-ICAM-1, weight, height, BP and diabetes were taken into account. This statistical modelling suggests that at least 40% of the risk reduction associated with exercise cannot be explained by establish risk factors . The Percentage reduction in CVD events associated with physical activity that is explained by risk factors (adapted from [bib_ref] Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms, Mora [/bib_ref]. Differences in risk factors explain *59% of the relative cardiovascular risk reduction associated with exercise. This statistical modeling suggests that at least 40% of the risk reduction associated with exercise cannot be explained by established or emerging risk factors mechanisms responsible for the cardiovascular benefits of exercise in humans can therefore not be solely attributed to risk factor modulation.
In summary, traditional cardiovascular risk factors do not appear to explain the link between cardiovascular risk and physical inactivity, whilst at least 40% of the cardiovascular risk reduction attributable of exercise training remains unexplained. In this review, we examine evidence supporting the notion that inactivity and exercise have direct effects on the function and structure of the vasculature and play an important role in explaining the links between inactivity, exercise and cardiovascular risk. We propose that inactivity and exercise training have direct ''vascular deconditioning and conditioning'' effects, which modify cardiovascular risk.
## Effects of physical inactivity on the vasculature in humans
Studies of the impact of inactivity on the vasculature in humans have adopted different intervention approaches including space travel and (head-down) bed rest, unilateral lower limb immobilisation (ULLS), immobilization via casting and assessment of subjects with spinal cord injury. Each of these approaches has its limitations. For example, changes in plasma volume accompany space flight and bed rest [bib_ref] Female exposure to high G: chronic adaptations of cardiovascular functions, Convertino [/bib_ref] whilst upper extremity activity is not typically unrestricted in the bed-rest model. ULLS is only suitable to study localized deconditioning in one lower limb and there is apparently an increased risk of (deep venous) thrombosis [bib_ref] Changes in muscle function in response to 10 days of lower limb..., Berg [/bib_ref] [bib_ref] Unilateral lower limb suspension can cause deep venous thrombosis, Bleeker [/bib_ref] [bib_ref] The effect of unloading on protein synthesis in human skeletal muscle, Galbo [/bib_ref] [suggested incidence of 2.7% [bib_ref] Unilateral lower limb suspension can cause deep venous thrombosis, Bleeker [/bib_ref] ]. Casting of a limb (lower leg or forearm) [bib_ref] Effect of casting on forearm resistance vessels in young men, Green [/bib_ref] typically follows a fracture or trauma and vascular changes to ''inactivity'' may therefore be influenced by the impact of healing and inflammatory processes.
Finally, it has been suggested that loss of supraspinal sympathetic vascular tone of the legs may affect vascular function and structure independent of the effects of inactivity per se. However, sympathectomized patients, who lack sympathetic vascular innervation, but participate in normal physical activity, do not exhibit the structural or functional vascular adaptations typically observed in the SCI model [bib_ref] Effects of chronic sympathectomy on vascular function in the human forearm, Eisenach [/bib_ref] [bib_ref] History of sympathetic surgery, Hashmonai [/bib_ref]. In addition, previous studies in SCI have demonstrated that vascular adaptations are partly reversible by electrical stimulation training of the inactive paralyzed legs in SCI [bib_ref] Increased vascular resistance in paralyzed legs after spinal cord injury is reversible..., Hopman [/bib_ref] [bib_ref] A causal role for endothelin-1 in the vascular adaptation to skeletal muscle..., Thijssen [/bib_ref]. These data have led some investigators to conclude that vascular adaptations observed in the paralyzed legs in SCI may primarily result from physical inactivity.
Physiological studies have been undertaken to examine the impact of inactivity on both resistance and conduit artery function and structure. These impacts of each of the models of inactivity on the function and structure of these vessel beds will be considered separately below.
## Resistance vessel function
Plethysmographic measurement of change in limb volume has typically been employed as a method to assess peripheral resistance vessel function in humans [bib_ref] Circulation to skeletal muscle. Handbook of physiology. The cardiovascular system peripheral circulation..., Shepherd [/bib_ref]. Several limb volume techniques evolved after the initial use of plethysmography in 1905, but all rely upon a ''congesting'' cuff inflated around the upper arm or thigh and inflated to approximately 40-50 mmHg [bib_ref] From Belfast to Mayo and beyond: the use and future of plethysmography..., Joyner [/bib_ref]. Mercury-in-silastic strain-gauges are the most frequently used technique [bib_ref] The measurement of volume changes in human limbs, Whitney [/bib_ref] and were widely used in the twentieth century to examine vascular physiology and pharmacology [bib_ref] From Belfast to Mayo and beyond: the use and future of plethysmography..., Joyner [/bib_ref].
The majority of studies which have used venous occlusion plethysmography to assess resting blood flows have reported decreases after 4-120 days of head down tilt bed rest (26-48%, [bib_ref] Changes in microvascular fluid filtration capacity during 120 days of 6 degrees..., Christ [/bib_ref] [bib_ref] Changes in size and compliance of the calf after 30 days of..., Convertino [/bib_ref] [bib_ref] Changes in leg vein filling and emptying characteristics and leg volumes during..., Louisy [/bib_ref] [bib_ref] Regulation of muscle sympathetic nerve activity after bed rest deconditioning, Pawelczyk [/bib_ref] , 4-14 days space flight (41%, [bib_ref] Effects of spaceflight on human calf hemodynamics, Watenpaugh [/bib_ref] , 28 days limb immobilization (24%, [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] and chronic spinal cord injury (26-70%, [bib_ref] Vascular adaptation to deconditioning and the effect of an exercise countermeasure: results..., Bleeker [/bib_ref] [bib_ref] Increased vascular resistance in paralyzed legs after spinal cord injury is reversible..., Hopman [/bib_ref] [bib_ref] Preserved alpha-adrenergic tone in the leg vascular bed of spinal cordinjured individuals, Kooijman [/bib_ref] [bib_ref] A causal role for endothelin-1 in the vascular adaptation to skeletal muscle..., Thijssen [/bib_ref]. To gain better insight into the mechanisms responsible for these decreases in limb blood flows, Bleeker et al. examined leg vascular responses to femoral artery infusions of the nitric oxide (NO) antagonist, N G -monomethyl-L-arginine (L-NMMA), the NO vasodilator sodium nitroprusside (SNP) and angiotensin II [bib_ref] Vascular adaptation to deconditioning and the effect of an exercise countermeasure: results..., Bleeker [/bib_ref]. Studies were undertaken in SCI individuals, age-matched controls and also in healthy young men before and after 28 days of ULLS. The vasoconstrictor responses to L-NMMA and angiotensin II, and the dilator responses to SNP, were unaltered in both models of inactivity. The authors concluded that short-and long-term deconditioning preserved vasoactive responses in the lower limb skeletal muscle vascular bed [bib_ref] Vascular adaptation to deconditioning and the effect of an exercise countermeasure: results..., Bleeker [/bib_ref].
In another study, forearm blood flow responses to L-NMMA were examined immediately after forearm cast removal for treatment of scaphoid or Colles' fractures, and again 6 weeks after cast removal. Responses were compared to those in healthy uncasted controls. The response to L-NMMA was similar between groups and before and after 6 weeks recovery [bib_ref] Effect of casting on forearm resistance vessels in young men, Green [/bib_ref]. Although forearm injury and fracture may have influenced these results, this study suggests that physical inactivity does not influence basal activity of the NO dilator system in vivo in resistance vessels during short-term physical activity.
Another factor that contributes to baseline vascular tone, the a-adrenergic system, has been compared in SCI individuals and able-bodied controls [bib_ref] Preserved alpha-adrenergic tone in the leg vascular bed of spinal cordinjured individuals, Kooijman [/bib_ref]. Leg blood flow changes were assessed in response to intrafemoral infusion of incremental doses of phentolamine (a competitive antagonist of the a-adrenoceptor) during local b-adrenergic receptor blockade with propranolol. Interestingly, despite the complete spinal cord lesion in all SCI subjects, a-adrenergic tone was not exaggerated in the leg vascular bed of SCI individuals. Recently, [bib_ref] A causal role for endothelin-1 in the vascular adaptation to skeletal muscle..., Thijssen [/bib_ref] examined the contribution of endothelin-1 (ET-1), a powerful endothelium-derived vasoconstrictor, to baseline blood flow in the inactive legs of SCI individuals, using intra-femoral administration of selective ET A/Breceptor blockers. They found a larger vasodilator response to blockade of ET-1 in the legs of SCI subjects compared with age-matched able-bodied controls, suggesting that ET-1 importantly contributes to the increased vascular tone observed as a consequence of chronic physical inactivity. To further examine this hypothesis, the authors trained their SCI subjects for 6 weeks using functional electrical stimulation (FES). After this training, the SCI subjects showed reversed ET-1-mediated vascular tone, further supporting a role for ET-1 in the increased baseline vascular tone observed in this model of deconditioning.
In a final study, which assessed resistance artery endothelial function, intra-brachial administration of endothelium-dependent (acetylcholine) and -independent (sodium nitroprusside) vasodilators was undertaken in healthy volunteers before and after 13 days of bed rest [bib_ref] Modulation of endothelial and smooth muscle function by bed rest and hypoenergetic,..., Hesse [/bib_ref]. Bed rest impaired endothelium-dependent vasodilation of resistance vessels, but not smooth muscle NOmediated vasodilation. These short-term effects of bed rest were not present when bed rest was performed under energy restriction (25% reduction in fat intake) [bib_ref] Modulation of endothelial and smooth muscle function by bed rest and hypoenergetic,..., Hesse [/bib_ref]. Furthermore, the level of actual upper limb activity change associated with bed rest (subjects were allowed to do upper limb daily activities) was not controlled. These results should therefore be interpreted with caution.
## Conduit artery function
Using high resolution ultrasound, flow-mediated dilation (FMD) can be examined as a marker for conduit artery endothelial function. Assuming the occluding cuff is placed distal to the scanned artery [bib_ref] Flow-mediated dilatation following wrist and upper arm occlusion in humans: the contribution..., Doshi [/bib_ref] , that the period of ischemia does not exceed 5 min [bib_ref] Heterogenous nature of flow-mediated dilatation in human conduit arteries in vivo: relevance..., Mullen [/bib_ref] and that diameter responses are appropriately collected [bib_ref] Importance of measuring the time course of flow-mediated dilatation in humans, Black [/bib_ref] and normalised [bib_ref] Importance of measuring the time course of flow-mediated dilatation in humans, Black [/bib_ref] [bib_ref] Peak vs. total reactive hyperemia: which determines the magnitude of flow-mediated dilation?, Pyke [/bib_ref] , FMD is predominantly mediated by NO [bib_ref] Flow-mediated dilatation following wrist and upper arm occlusion in humans: the contribution..., Doshi [/bib_ref] [bib_ref] Nitric oxide is responsible for flowdependent dilatation of human peripheral conduit arteries..., Joannides [/bib_ref] [bib_ref] Flow-mediated dilatation in the superficial femoral artery is nitric oxide mediated in..., Kooijman [/bib_ref] and serves as a valid index of conduit artery endothelium-dependent NO function [bib_ref] Testing endothelial vasomotor function: nitric oxide, a multipotent molecule, Ganz [/bib_ref]. FMD also correlates with coronary endothelial function [bib_ref] Endothelium-dependent flow-mediated vasodilation in coronary and brachial arteries in suspected coronary artery..., Takase [/bib_ref] [bib_ref] Close relationship between the vasodilator response to acetylcholine in the brachial and..., Takase [/bib_ref]. This technique is non-invasive, allows direct arterial visualisation and, in contrast to plethysmography, allows measurement of absolute arterial diameters and has a good temporal resolution [bib_ref] Assessment of brachial artery blood flow across the cardiac cycle: retrograde flows..., Green [/bib_ref]. However, arterial diameters are critically dependent upon image quality [bib_ref] The importance of Doppler angle of insonation on differentiation between 50-69% and..., Logason [/bib_ref] , while sometimes manual analysis is used, which is subject to significant observer error [bib_ref] Importance of measuring the time course of flow-mediated dilatation in humans, Black [/bib_ref] [bib_ref] Non-invasive detection of endothelial dysfunction in children and adults at risk of..., Celermajer [/bib_ref] [bib_ref] Reproducibility of brachial ultrasonography and flow-mediated dilatation (FMD) for assessing endothelial function, Hardie [/bib_ref] [bib_ref] Intracardiac echocardiography: computerized detection of left ventricular borders, Sonka [/bib_ref]. Finally, it is possible that the shear stress or flow stimulus that elicits FMD may provide some independent prognostic information [bib_ref] Comparison of new measures of vascular function to flow mediated dilatation as..., Philpott [/bib_ref].
Whilst impaired endothelium-dependent NO-mediated dilator function has typically been observed in subjects with traditional cardiovascular risk factors, surprisingly, an enhanced superficial femoral artery FMD response was observed after 25 and 52 days of bed rest [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref] , 28 days of limb suspension [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] and in acute [bib_ref] Magnitude and time course of arterial vascular adaptations to inactivity in humans, De Groot [/bib_ref] (within 21-42 days post-injury) [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref] , as well as in chronic, [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref]. In contrast, a recent study reported decreased posterior tibial artery FMD in SCI compared with controls [bib_ref] Upper vs lower extremity arterial function after spinal cord injury, Stoner [/bib_ref]. However, in the latter study the occlusion cuff was placed proximal to the vessel-imaging site, an arrangement that likely results in vasodilation which is less NO-dependent [bib_ref] Flow-mediated dilatation following wrist and upper arm occlusion in humans: the contribution..., Doshi [/bib_ref] than distal cuff occlusion. This methodological issue may account for this atypical finding.
Changes in brachial artery endothelial function after shorter periods of bed rest also reveal conflicting results. While 5 days of bed rest did not alter the FMD response [bib_ref] Physical inactivity rapidly induces insulin resistance and microvascular dysfunction in healthy volunteers, Hamburg [/bib_ref] , another study reported an increased brachial artery FMD following 7 days of bed rest [bib_ref] Enhanced flow-dependent vasodilatation after bed rest, a possible mechanism for orthostatic intolerance..., Bonnin [/bib_ref]. Methodological differences, such as different approaches to the detection of peak arterial diameter [bib_ref] Physical inactivity rapidly induces insulin resistance and microvascular dysfunction in healthy volunteers, Hamburg [/bib_ref] and the question of whether the upper limbs are truly inactive during bed rest, may have impacted these results.
Recent studies regarding the interpretation of FMD data suggest the change in diameter should be normalized for the eliciting shear stress stimulus on the endothelial cell membrane. This approach is conceptually equivalent to the interpretation of drug responses in accordance with the administered dosage. When the superficial femoral artery FMD response of spinal cord-injured individuals was expressed as diameter change relative to the hyperemic peak shear rate, FMD responses were found to be preserved [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref] or slightly increased compared to able-bodied controls [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref]. A preserved normalised FMD of the superficial femoral artery was also found after 25 and 52 days of bed rest [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref] and 4 weeks of unilateral lower limb suspension [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref]. However, the appropriate normalization approach involves assessment of the area-under-the-curve (SR AUC ) for shear rate, from the time of occluding cuff deflation to that of peak diameter attainment [bib_ref] Peak vs. total reactive hyperemia: which determines the magnitude of flow-mediated dilation?, Pyke [/bib_ref] , rather than simply the peak shear rate following cuff deflation. SR AUC normalization is particularly appropriate under circumstances where diameter responses are likely to differ as a function of the shear stress that the artery is exposed to, such as the case where arteries of differing baseline diameter are compared (e.g., SCI vs. controls). Using a 0-40 s time window to calculate the AUC, a similar FMD response was observed in SCI and able-bodied controls [bib_ref] Flow-mediated dilatation in the superficial femoral artery is nitric oxide mediated in..., Kooijman [/bib_ref].
A possible explanation for the apparently preserved FMD response in deconditioned vessels may relate to upregulation of smooth muscle cell sensitivity to NO, possibly initiated as a counter-regulatory response to chronic decreases in endothelial shear stress and down-regulation of eNOS in the deconditioned vessels. Increased NO smooth muscle sensitivity to NO donors was observed after 28 days limb suspension [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] and after 52 days of bed rest [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref]. In SCI subjects, the superficial femoral artery endothelium-independent dilatation varies between 15.6 and 19.6% [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref] [bib_ref] Local vascular adaptations after hybrid training in spinal cord-injured subjects, Thijssen [/bib_ref] , while healthy young controls demonstrate a dilatation of 10.9-13.4% [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref] [bib_ref] Sympathetic nervous system contributes to the age-related impairment of flow-mediated dilation of..., Thijssen [/bib_ref]. Nevertheless, differences between these groups did not achieve statistical significance [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref].
These data go some way towards supporting the notion of enhanced smooth muscle NO responsiveness as a consequence of deconditioning. However, sublingual administration of a single dose of a NO-donor, such as nitroglycerine (GTN), has important limitations relating to between subjects differences in pharmacokinetics. Recently, therefore, smooth muscle cell NO-sensitivity was examined using incremental intra-femoral doses of sodium nitroprusside (SNP) in SCI subjects and age-matched ablebodies controls. After correcting for absolute differences in artery dimensions, both groups demonstrated similar NO sensitivity of the superficial femoral artery .
Taken together, the results of recent studies that examined acute (21-52 days) and chronic (longer than 6 months) deconditioning suggest that endotheliumdependent and -independent vasodilation are preserved in peripheral conduit arteries. A possible explanation for the preserved conduit artery endothelial function after physical inactivity may relate to compensatory changes in conduit artery dimensions. As a result of the changes in shear rate, deconditioning induces an inward remodelling of conduit arteries and a new equilibrium is set that preserves or normalizes the vascular function. These considerations are discussed further below. Currently no information is available regarding coronary vascular function after deconditioning.
## Resistance vessel remodelling
Peak reactive hyperemic blood flow assessment is an approach traditionally used to assess remodelling of resistance vessels in humans [bib_ref] A vascular abnormality in hypertension. A study of blood flow in the..., Conway [/bib_ref] [bib_ref] Adaptive structural changes of the vascular walls in hypertension and their relation..., Folkow [/bib_ref] [bib_ref] Reactive hyperaemia in the human forearm, Patterson [/bib_ref] [bib_ref] A comparison of the effects of vasodilator stimuli on peripheral resistance vessels..., Zelis [/bib_ref] [bib_ref] The fourth Volhard lecture: cardiovascular structural adaptation; its role in the initiation..., Folkow [/bib_ref] [bib_ref] Decreased vasodilator capacity of forearm resistance vessels in borderline hypertension, Takeshita [/bib_ref] [bib_ref] Enhanced maximal metabolic vasodilatation in the dominant forearms of tennis players, Sinoway [/bib_ref] [bib_ref] ) A 30-day forearm work protocol increases maximal forearm blood flow, Sinoway [/bib_ref] [bib_ref] Exercise training enhances leg vasodilatory capacity of 65-yr-old men and women, Martin [/bib_ref] [bib_ref] Leg exercise conditioning increases peak forearm blood flow, Silber [/bib_ref] [bib_ref] Reversible impairment of forearm vasodilation after forearm casting, Silber [/bib_ref]. Reactive hyperemic blood flow through the superficial femoral artery in response to a 5 min ischemic stimulus decreased by 28% after 52 days of bed rest [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref]. A 31% decrease was observed in the forearm vascular bed after 14 days of bed rest using a 10 min ischemic stimulus [bib_ref] Head-down-tilt bed rest alters forearm vasodilator and vasoconstrictor responses, Shoemaker [/bib_ref]. After 29 days of cast immobilization of the wrist as a treatment of bone fracture, minimal resistance during reactive hyperemia was higher in the casted than in the control arm. During the recovery period, minimal resistance during reactive hyperemia decreased in the casted arm but did not change in the control arm [bib_ref] Reversible impairment of forearm vasodilation after forearm casting, Silber [/bib_ref]. Similar findings were reported in the calf directly after 2-12 weeks of cast treatment for leg trauma and after a 6 week recovery period [bib_ref] The effect of inactivity on reactive hyperaemia in the human calf: a..., Kroese [/bib_ref].
Although not significant, a trend towards decrease in hyperemic flow was present after 28 days of lower limb immobilization [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] and during the first 6 weeks after a spinal cord injury . Reactive hyperemia is 40-60% lower in the legs of SCI compared with controls [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref]. In summary, reactive hyperemic blood flow, an index of resistance vessel structure or cross-sectional area, is reduced after a period of inactivity in most models of deconditioning.
## Conduit artery remodelling
Shear stress plays an important role in the regulation and adaptation of large arteries. Langille and O'Donnell established that changes in vessel structure occur secondary to chronic changes in flow, and that such changes are dependent upon the release of a labile factor from endothelial cells [bib_ref] Reductions in arterial diameter produced by chronic decreases in blood flow are..., Langille [/bib_ref]. This conclusion was later confirmed by studies demonstrating that arterial remodelling is shear stress and NO-dependent [bib_ref] Shear level influences resistance artery remodeling: wall dimensions, cell density, and eNOS..., Tuttle [/bib_ref] and acts in a manner which homeostatically regulates wall shear [bib_ref] Role of NO in flow-induced remodeling of the rabbit common carotid artery, Tronc [/bib_ref].
Baseline shear levels were found to be almost doubled in the femoral artery of chronic SCI individuals compared with able-bodied subjects [bib_ref] Time course of arterial vascular adaptations to inactivity and paralyses in humans, De Groot [/bib_ref] [bib_ref] Bruce Dill Historical lecture. Historical concepts of the athlete's heart, Thompson [/bib_ref] , while shear rate levels were also significantly increased after 28 days ULLS [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref]. Several models of physical inactivity of different duration demonstrated that these increases in shear rate associated with deconditioning were not due to increases in blood flow [bib_ref] Vascular remodeling after spinal cord injury, Olive [/bib_ref]. The principal reason for elevated shear rate in deconditioned conduit arteries is therefore related to decreased conduit artery diameter and the consequent increase in velocity. Deconditioning induced a 6% reduction in artery diameter after 7 days of leg casting [bib_ref] Reductions in basal limb blood flow and lumen diameter after short-term leg..., Sugawara [/bib_ref] , a 13% decrease after 28 days ULLS [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] , 13 and 17% reductions after 25 and 52 days of bed rest [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref] , respectively, a 25% decrease 18 days after a spinal cord lesion [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref] and 30% reduction in vessel diameter in chronic SCI [bib_ref] Time course of arterial vascular adaptations to inactivity and paralyses in humans, De Groot [/bib_ref] [bib_ref] Bruce Dill Historical lecture. Historical concepts of the athlete's heart, Thompson [/bib_ref] [bib_ref] Size and blood flow of central and peripheral arteries in highly trained..., Huonker [/bib_ref]. Hence, conduit artery diameter decreases markedly in response to physical inactivity, which coincides with a marked increase in conduit artery shear rate. In a previous cross-sectional study in chronic SCI individuals and able-bodied controls, [bib_ref] Vascular remodeling after spinal cord injury, Olive [/bib_ref] reported that the significantly smaller femoral vessel diameter in SCI was no longer different when diameter was expressed per unit muscle mass. Interestingly, when changes in vascular properties and limb volume during the first 6 weeks after a spinal cord injury were examined [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref] , femoral artery size decreases substantially, while a simultaneous decrease in limb volume was evident. Although changes in limb volume do not completely reflect changes in muscle mass, the corrected femoral artery diameter for limb volume showed no differences over the 6 week time period and values in SCI were comparable with control values [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref]. These findings suggest a link between vascular structural adaptations and muscle atrophy as a consequence of inactivity.
The above studies have reported baseline conduit artery diameter as an index of vascular structure. However, baseline diameter reflects competitive functional influences on vascular tone, in addition to artery size. The use of baseline diameter as an index of structure is therefore polluted by factors which regulate vascular function. The importance of measuring maximal conduit artery diameter as a marker of structural changes has recently been emphasized . Vasodilatation to nitroglycerin represents near maximal diameter and can therefore reflect the limits of arterial structure and size. Maximal femoral artery diameter decreased by 9% after 4 weeks of limb suspension [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] , 16% after 25 and 52 days of bed rest [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref] and by 35% in chronic SCI individuals [bib_ref] Time course of arterial vascular adaptations to inactivity and paralyses in humans, De Groot [/bib_ref] [bib_ref] Bruce Dill Historical lecture. Historical concepts of the athlete's heart, Thompson [/bib_ref]. These findings reinforce those above and indicate that inactivity is indeed associated with arterial structural remodelling and decreased conduit artery dimension.
## Time-course of vascular adaptations to inactivity
In a recent longitudinal study by [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref] , conduit artery characteristics and limb volume were examined in SCI subjects during the 6 weeks immediately following spinal cord injury. Interestingly, the femoral artery diameter decreased substantially and rapidly, with a decrease in diameter 3 week post-injury which approached vessel dimension of chronic SCI subjects [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref]. Other models of deconditioning have also been used to establish that rapid and extensive changes occur in conduit artery dimensions [bib_ref] Reductions in basal limb blood flow and lumen diameter after short-term leg..., Sugawara [/bib_ref]. These findings indicate that physical inactivity provides a strong stimulus for rapid structural remodelling of human conduit arteries. Whether these changes in conduit artery dimension parallel changes in other vascular beds, and whether functional and structural changes have a similar time-course, is discussed below.
## Resistance versus conduit vessels
Evidence suggests that resistance and conduit vessels have a different time-course of adaptation to deconditioning. For example, the rapid decrease in femoral artery dimension after SCI was not accompanied by a simultaneous rapid decrease in reactive hyperemic flow, the latter a reflection of arteriolar structural changes [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref]. Parallel to this, a large decrease in conduit artery diameter during 52 days bed rest was present in the first 4 weeks, with little change thereafter, while the reactive hyperemia response did not change significantly . This suggests that the mechanisms and time-course of structural vascular adaptation to deconditioning may differ between conduit and resistance vessels. This hypothesis was first proposed by Laughlin et al., based on animal data [bib_ref] Endothelium-mediated control of coronary vascular tone after chronic exercise training, Laughlin [/bib_ref] , and later also suggested for exercise training studies in humans [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref]. Future studies should further examine the possible differences in time-course for activity-induced changes in conduit and resistance vessels.
## Functional versus structural changes
The sparse data present suggest that physical inactivityinduced conduit artery structural and functional changes may differ in terms of time-course [fig_ref] Figure 2: Hypothesised changes in artery function and structure [/fig_ref]. As stated above, the decrease in femoral artery dimension after a spinal cord injury was nearly complete 3 weeks post-injury, while the FMD responses demonstrated large changes between 4 and 6 weeks after the lesion [bib_ref] Rapid and extensive arterial adaptations after spinal cord injury, De Groot [/bib_ref]. In parallel, during bed rest and limb suspension, the change in diameter dimension occurred predominantly in the initial phase, while femoral artery endothelial function showed a gradual change over time [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref]. Future studies that utilize more frequent functional and structural measurements within the same vascular bed during deconditioning will provide greater insight into the time-course of structural and functional adaptation in resistance and conduit vessels.
## Local versus systemic adaptations to physical inactivity
Experimental models in which the whole body is subject to physical inactivity, for example spaceflight or bed rest, are associated with vascular changes throughout the body [bib_ref] Preserved contribution of nitric oxide to baseline vascular tone in deconditioned human..., Bleeker [/bib_ref] [bib_ref] Enhanced flow-dependent vasodilatation after bed rest, a possible mechanism for orthostatic intolerance..., Bonnin [/bib_ref] [bib_ref] Changes in leg vein filling and emptying characteristics and leg volumes during..., Louisy [/bib_ref] [bib_ref] Vascular adaptation to microgravity: what have we learned?, Zhang [/bib_ref]. It must be borne in mind, however, that inactivity is also associated with changes in systemic cardiovascular parameters [bib_ref] Physical inactivity rapidly induces insulin resistance and microvascular dysfunction in healthy volunteers, Hamburg [/bib_ref]. In contrast, models which utilise localized physical inactivity, for example ULLS and cast immobilisation, are associated with vascular adaptations specific to the immobilized region or limb only [bib_ref] Reductions in basal limb blood flow and lumen diameter after short-term leg..., Sugawara [/bib_ref] , in the absence of change in systemic cardiovascular risk factors. SCI subjects demonstrate marked vascular adaptations in the inactive, paralyzed legs [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref] [bib_ref] Upper vs lower extremity arterial function after spinal cord injury, Stoner [/bib_ref] , while upper extremity vascular function and structure are comparable with able-bodied controls [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref] [bib_ref] Increased vascular resistance in paralyzed legs after spinal cord injury is reversible..., Hopman [/bib_ref] [bib_ref] Upper vs lower extremity arterial function after spinal cord injury, Stoner [/bib_ref]. This raises one final consideration, pertinent to the models of inactivity used in humans. Using ULLS or a cast to immobilize one limb forces volunteers to use crutches for locomotion. Whether, on balance, this leads to a training effect in the ''unaffected'' weight-bearing limb, or a small deconditioning because of the global reduction in levels of physical activity, is unknown. Similarly, it is likely that a different level of activity is present in the upper extremities of SCI individuals after the lesion. The dependence on the upper limbs for locomotion may theoretically result in SCI subjects exhibiting vascular adaptations, compared with ablebodied controls, yet studies consistently indicate that nonimmobilized areas do not demonstrate enhanced vascular function or structure [bib_ref] Vascular adaptation to 4 wk of deconditioning by unilateral lower limb suspension, Bleeker [/bib_ref] [bib_ref] Preserved flow-mediated dilation in the inactive legs of spinal cord-injured individuals, De Groot [/bib_ref] [bib_ref] Upper vs lower extremity arterial function after spinal cord injury, Stoner [/bib_ref] [bib_ref] Local vascular adaptations after hybrid training in spinal cord-injured subjects, Thijssen [/bib_ref]. Taken together, available evidence indicates that physical inactivity appears to lead to localized vascular adaptations in the affected inactive regions only.
## Physical inactivity, vascular adaptations and cardiovascular risk
As mentioned in the introduction, the impact of physical inactivity on traditional risk factors may not adequately explain the increased cardiovascular mortality associated with deconditioning, particularly in SCI subjects. It is therefore feasible that direct effects of inactivity on vascular structure and function, rather than the indirect effects of traditional risk factors, may explain the increased cardiovascular risk associated with a sedentary life style. However, current data, discussed above, suggests that Hypothesised changes in artery function and structure (remodelling) in response to inactivity and exercise training in humans. Studies performed in both animals and humans suggest that rapid changes occur in artery function, including nitric oxide (NO) bioavailability, in response to exercise training and that these changes are superseded by arterial remodelling and normalisation of function.
Physical inactivity is associated with rapid changes in arterial diameter, with structural remodelling occurring within weeks of, for example, spinal cord injury. There is little evidence for longer term vascular dysfunction in response to inactivity. Changes in artery function and structure occur rapidly in response to activity and inactivity Eur J Appl Physiol (2010) 108:845-875 851 deconditioning does not impair vasodilator function. It remains possible that other vascular parameters, such as intima-media thickness [bib_ref] Intima-media thickness: a new tool for diagnosis and treatment of cardiovascular risk, Simon [/bib_ref] or arterial stiffness (O'Rourke 1999) which correlate with cardiovascular risk, are abnormal in inactive subjects. To date, these parameters have only been examined using shortterm deconditioning (2-7 days) and, perhaps predictably, no changes were reported in these longer-term markers of vascular adaptation [bib_ref] Further study on the carotid baroreflex system in the cardiovascular deconditioning induced..., Pannier [/bib_ref] [bib_ref] Reductions in basal limb blood flow and lumen diameter after short-term leg..., Sugawara [/bib_ref]. SCI subjects, however, demonstrate a lower arterial compliance . Some evidence exists for enhanced vasoconstrictor function in models of inactivity, including exaggerated endothelin-1 responses [bib_ref] A causal role for endothelin-1 in the vascular adaptation to skeletal muscle..., Thijssen [/bib_ref]. Data regarding the impact of physical inactivity on sympathetic-mediated vasoconstrictor tone is scant. Despite the sparse and dispersed data, and the various models of inactivity that have been adopted, it remains likely that changes in conduit and resistance vessel structure and function may help to explain the increased cardiovascular mortality and morbidity associated with deconditioning. Future studies are necessary to further examine this interesting hypothesis.
Remarkably, despite the epidemiological evidence linking physical inactivity and increased cardiovascular risk and the general acceptance of physical inactivity as a cardiovascular risk factor , no physiological studies have ever examined the direct impact of physical inactivity on coronary artery function and structure.
## Effects of exercise or cardiopulmonary fitness on vascular events
The Health Professional's Follow-up Study (HPFS) [bib_ref] Exercise type and intensity in relation to coronary heart disease in men, Tanasescu [/bib_ref] included analysis of 44,452 men who had their leisure time physical activity assessed by questionnaire every 2 years between 1986 until 1996. The relative risk of MI, comparing the highest and lowest total exercise volume quintiles, was 0.7 (P \ 0.001). The authors concluded that a significant inverse dose relation exists between total physical activity and CHD risk, that exercise intensity is associated with risk reduction and that the addition of ''weight'' training may be beneficial in terms of further risk reduction. Although studies such as the HPFS are prospective and undertaken in large cohorts, they rely heavily on self-report of physical activity and are therefore subject to misclassification bias. The Aerobics Center Longitudinal Database (ACLD) estimated _ VO 2 max following graded exercise testing in 25,341 men and 7,080 women between 1970 and 1989 [bib_ref] Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause..., Blair [/bib_ref]. Increased relative risk of CV mortality was associated with low levels of fitness (RR = 1.70), cigarette smoking (RR = 1.57), elevated systolic blood pressure (RR = 1.34) and elevated serum cholesterol (RR = 1.65) in men. Higher levels of cardiorespiratory fitness were associated with significant protection from the impact of other risk factors. Notwithstanding the important distinction between cardiorespiratory fitness and exercise training, this prospective data provides evidence for the CV benefits of exercise for primary prevention of CV disease.
There is no evidence available from large prospective cohort trials regarding the effect of exercise in a secondary prevention setting, although Myers et al. concluded that, after adjustment for age, peak exercise capacity measured in METs is the strongest predictor of the risk of death among both normal subjects and those with established cardiovascular disease [bib_ref] Exercise capacity and mortality among men referred for exercise testing, Myers [/bib_ref]. A series of meta-analyses performed on smaller studies of exercise-based cardiac rehabilitation have estimated the benefits in terms of reduction in total and cardiac mortality at between 20 and 32%, relative to usual care ), although earlier analyses included studies performed prior to the widespread adoption of contemporary lipid-lowering, antihypertensive and interventional strategies. The most recent meta-analysis, which included 48 eligible studies to March 2003 involving 8,940 patients, attempted to address this concern. Cardiac rehabilitation was associated with significant decreases in the odds ratio for all-cause and cardiac mortality (OR = 0.74) and the authors concluded that recent trials conducted in the era of contemporary cardiovascular therapies and case mix (post-angioplasty, CABG etc.), continued to report similar or greater benefits (OR = 0.62). An interesting aspect of this analysis, which confirms previous findings [bib_ref] Multiple risk factor interventions for primary prevention of coronary heart disease, Ebrahim [/bib_ref] , was the suggestion that studies involving exercise training alone (OR = 0.76) exert a similar effect to those studies which involved ''comprehensive'' rehabilitation (OR = 0.84) consisting of behavioural strategies aimed at decreasing the impact of other CV risk factors. This suggests that exercise is an important component of cardiac rehabilitation programs and also indirectly suggests that modification of risk factors may not be the primary means by which exercise exerts its beneficial effect.
In summary, exercise or physical fitness is associated with an approximate 30% benefit in terms of decreasing the risk of cardiac events. Indeed, some expert opinion has stated that the most physically active individuals demonstrate CAD rates half those of the most sedentary and that in many studies the lower frequency of CAD is independent of other known atherosclerotic risk factors [bib_ref] Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular..., Thompson [/bib_ref]. As described above, exercise training improves vascular function in the absence of changes in lipid levels [bib_ref] Exercise training increases basal nitric oxide production from the forearm in hypercholesterolemic..., Lewis [/bib_ref] , blood pressure [bib_ref] Regular aerobic exercise augments endothelium-dependent vascular relaxation in normotensive as well as..., Higashi [/bib_ref] , glucose tolerance, or BMI [bib_ref] Exercise training normalizes vascular dysfunction and improves central adiposity in obese adolescents, Watts [/bib_ref]. In addition, a recent study reported significant improvements in both conduit and resistance vessel endothelial function, in the absence of change in plasma lipids, blood pressure, blood glucose, waist:hip or BMI, with no significant correlations between changes in any of these risk factors and the improvements in vascular function . These studies indicate that improvement in risk factors is not an obligatory requirement for improvement in vascular function as a result of exercise training. Direct effects of exercise on the vasculature, perhaps by virtue of a repetitive impact of shear stress, pulse pressure or pulsatility, may therefore contribute to the reduction in coronary events associated with exercise training [bib_ref] Exercise and cardiovascular risk reduction: time to update the rationale for exercise?, Green [/bib_ref].
## Effect of exercise training on vascular function
Different modalities of exercise exert distinct physiological and health benefits. For example, recent studies indicate that exercise involving different volumes of muscle mass or exercise intensity may be associated with different shear stress mediated impacts on vascular function [bib_ref] Effect of different intensities of exercise on endothelium-dependent vasodilation in humans: role..., Goto [/bib_ref] [bib_ref] Brachial artery blood flow responses to different modalities of lower limb exercise, Thijssen [/bib_ref]. Aerobic exercise involving large muscle groups is associated with systemic changes in pulse pressure and heart rate which generate a recurrent hemodynamic (shear) stress that may directly induce vascular function and structure adaptation (see below) [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref] , whereas small muscle group exercise is not associated with such hemodynamic signals and provides insight into localised effects of exercise independent of central regulatory or neural changes. A further consideration relates to the types of study designs adopted. Many ''exercise training'' studies have relied upon cross-sectional comparisons of subjects with high and low cardiorespiratory fitness as a model of the physiological impact of chronic exercise in humans. Despite efforts to match subject groups for numerous confounders (age, sex, body composition, risk factor profiles etc.), there remains the strong possibility that between-subject differences pollute the purported effects of ''exercise'' on vascular outcomes. Furthermore, whilst cross-sectional comparisons to some extent reflect the impact of different fitness levels on outcome measures, such differences cannot validly be ascribed to the impact of exercise training per se, since genetic cardiopulmonary capacities are included in fitness measures. In any event, the best assessment of the impact of exercise training on vascular outcome measures is derived from longitudinal training studies where within subject changes in fitness following an exercise training intervention are the focus.
## Small muscle group exercise
## Resistance vessels
Studies of handgrip exercise training were initially undertaken as a physiological model of the impact of localised muscle activity on vasomotor control, in the absence of marked changes in central hemodynamics during the training stimulus. Two of the three studies undertaken in healthy subjects failed to demonstrate improvement in forearm blood flow responses to ACh, SNP or LNMMA infusions [bib_ref] Effects of short-term forearm exercise training on resistance vessel endothelial function in..., Bank [/bib_ref] [bib_ref] Effects of intense exercise training on endothelium-dependent exercise-induced vasodilatation, Franke [/bib_ref] [bib_ref] Modification of forearm resistance vessels by exercise training in young men, Green [/bib_ref] , whilst two of the three studies in heart failure demonstrated improvement in endothelium-dependent responses [bib_ref] Effects of short-term forearm exercise training on resistance vessel endothelial function in..., Bank [/bib_ref] [bib_ref] Correction of endothelial dysfunction in chronic heart failure: additional effects of exercise..., Hambrecht [/bib_ref] [bib_ref] Training improves endothelium-dependent vasodilation in resistance vessels of patients with heart failure, Katz [/bib_ref]. The disparate findings both derived from a single study [bib_ref] Effects of short-term forearm exercise training on resistance vessel endothelial function in..., Bank [/bib_ref]. In general, the impact of localised exercise training on resistance vessel function has not been thoroughly studied in humans and recent investigations have focussed on more clinically applicable, large muscle group, training modalities.
## Conduit arteries
Studies which have examined the impact of small muscle group exercise on conduit artery function have been performed in CHF subjects [bib_ref] Correction of endothelial dysfunction in chronic heart failure: additional effects of exercise..., Hambrecht [/bib_ref] [bib_ref] Physical training improves endothelial function in patients with chronic heart failure, Hornig [/bib_ref] and hypertensives [bib_ref] Acute vascular responses to isometric handgrip exercise and effects of training in..., Mcgowan [/bib_ref]. [bib_ref] Physical training improves endothelial function in patients with chronic heart failure, Hornig [/bib_ref] performed an early study in which radial artery FMD was enhanced after 4 weeks of handgrip training and this improvement was NO mediated as indicated by its abolition using NO blockade. [bib_ref] Correction of endothelial dysfunction in chronic heart failure: additional effects of exercise..., Hambrecht [/bib_ref] also demonstrated that hand grip exercise training, particularly with the addition of L-arginine supplementation, enhanced ACh mediated radial artery diameter change. These studies have some clinical relevance in CHF, as they suggest that vascular adaptations which favour enhanced O 2 delivery and decreased peripheral resistance are possible as a consequence of small muscle group exercises which can reverse the peripheral abnormalities which may limit functional capacity in such subjects, without the hemodynamic burden associated with more systemic forms of training. Finally, McGowan recently observed enhanced brachial FMD responses after isometric hand grip exercise in patients with primary hypertension [bib_ref] Acute vascular responses to isometric handgrip exercise and effects of training in..., Mcgowan [/bib_ref] [bib_ref] The fourth Volhard lecture: cardiovascular structural adaptation; its role in the initiation..., Folkow [/bib_ref].
## Large muscle group dynamic exercise
## Resistance vessels
There have been at least 12 studies of whole body exercise programs (e.g., running, walking, cycling) in healthy control subjects either as a primary intervention group or as matched controls in studies of subjects with cardiovascular disease. The outcomes in terms of improvement in resistance vessel function are mixed. In young subjects who undertook cycle exercise training, some improvement in basal NO function was observed after brachial LNMMA infusions, but no changes in stimulated endothelial function were evident . Another study of fit young men who undertook high intensity training found a decrease in ACh and LNMMA responses [bib_ref] Intense physical training decreases circulating antioxidants and endothelium-dependent vasodilatation in vivo, Bergholm [/bib_ref]. In middle aged subjects, exercise training enhanced ACh responses in two studies [bib_ref] Regular aerobic exercise prevents and restores age-related declines in endotheliumdependent vasodilation in..., Desouza [/bib_ref] [bib_ref] Regular aerobic exercise augments endothelium-dependent vascular relaxation in normotensive as well as..., Higashi [/bib_ref] , but failed to do so in a third [bib_ref] Exercise training, vascular function, and functional capacity in middle-aged subjects, Maiorana [/bib_ref]. Finally, a recent study suggested that basal NO bioavailability is enhanced with training in middle aged and older subjects, an adaptation to some extent counteracted by sympathetic vasoconstrictor tone [bib_ref] Systemic alpha-adrenergic and nitric oxide inhibition on basal limb blood flow: effects..., Sugawara [/bib_ref].
Whilst there is no apparent consensus regarding the impact of exercise training on resistance vessel function in healthy subjects, the majority of studies performed in subjects with impaired endothelial function have documented improvements. This includes studies in patients with CHF [bib_ref] Regular physical exercise corrects endothelial dysfunction and improves exercise capacity in patients..., Hambrecht [/bib_ref] [bib_ref] Endothelial dysfunction in patients with chronic heart failure: systemic effects of lower-limb..., Linke [/bib_ref] [bib_ref] Effect of aerobic and resistance exercise training on vascular function in heart..., Maiorana [/bib_ref] [bib_ref] Augmentation of endothelial function following exercise training is associated with increased L-arginine..., Parnell [/bib_ref] , hypertension , hypercholesteremia [bib_ref] Exercise training increases basal nitric oxide production from the forearm in hypercholesterolemic..., Lewis [/bib_ref] [bib_ref] Effects of exercise training on conduit and resistance vessel function in treated..., Walsh [/bib_ref] , obesity [bib_ref] Weight loss in combination with physical activity improves endothelial dysfunction in human..., Sciacqua [/bib_ref] [bib_ref] Exercise training normalizes vascular dysfunction and improves central adiposity in obese adolescents, Watts [/bib_ref] and diabetes [bib_ref] The effect of combined aerobic and resistance exercise training on vascular function..., Maiorana [/bib_ref]. Indeed, there are no published studies in this category that we are aware of that have not reported improvements in resistance vessel vasodilator function as a result of large muscle group training.
Finally, a recent study investigated the effects of both chronic training and short term exercise in older subjects on microvascular function using microdialysis and laser Doppler flow assessments in the skin [bib_ref] Exercise prevents agerelated decline in nitric-oxide-mediated vasodilator function in cutaneous microvessels, Black [/bib_ref]. Exercise was associated with enhanced NO-mediated skin vasodilator responses to both ACh infusions and local heating stimuli in this study.
## Conduit arteries
One study of conduit artery responses in healthy subjects indicated that training of army recruits increased brachial FMD [bib_ref] Exercise training enhances endothelial function in young men, Clarkson [/bib_ref] and another suggested improvements after 4 weeks training [bib_ref] Time course of changes in endothelial function following exercise in habitually sedentary..., Pullin [/bib_ref]. However, resistance exercise training did not alter brachial responses in healthy volunteers and cycle training had no impact in older subjects [bib_ref] Low frequency regular exercise improves flow-mediated dilatation of subjects with mild hypertension, Moriguchi [/bib_ref] [bib_ref] Vascular adaptations to 8-week cycling training in older men, Thijssen [/bib_ref].
In subjects with CV disease or risk factors, the majority of studies indicate that exercise training involving a large muscle mass induced enhanced conduit artery function.
Studies undertaken in CHF [bib_ref] Moderate exercise training improves functional capacity, quality of life, and endothelium-dependent vasodilation..., Belardinelli [/bib_ref] [bib_ref] Regular physical activity improves endothelial function in patients with coronary artery disease..., Hambrecht [/bib_ref] [bib_ref] Endothelial dysfunction in patients with chronic heart failure: systemic effects of lower-limb..., Linke [/bib_ref] [bib_ref] Superior cardiovascular effect of aerobic interval training versus moderate continuous training in..., Wisloff [/bib_ref] , CAD [bib_ref] Effect of exercise training on endothelial function in men with coronary artery..., Edwards [/bib_ref] [bib_ref] Risk stratification for postoperative cardiovascular events via noninvasive assessment of endothelial function:..., Gokce [/bib_ref] [bib_ref] Exercise training improves conduit vessel function in patients with coronary artery disease, Walsh [/bib_ref] , peripheral artery disease [bib_ref] Acute impairment of the endothelial function by maximal treadmill exercise in patients..., Andreozzi [/bib_ref] , hypertensives [bib_ref] Low frequency regular exercise improves flow-mediated dilatation of subjects with mild hypertension, Moriguchi [/bib_ref] [bib_ref] Too old to benefit from sports? The cardiovascular effects of exercise training..., Westhoff [/bib_ref] , hypercholesterolemics [bib_ref] Effects of exercise training on conduit and resistance vessel function in treated..., Walsh [/bib_ref] , diabetics [bib_ref] Exercise training improves vascular endothelial function in patients with type 1 diabetes, Fuchsjager-Mayrl [/bib_ref] [bib_ref] Physical training improves flow-mediated dilation in patients with the polymetabolic syndrome, Lavrencic [/bib_ref] [bib_ref] The effect of combined aerobic and resistance exercise training on vascular function..., Maiorana [/bib_ref] [bib_ref] Regular aerobic exercise training improves endothelium-dependent arterial dilation in patients with impaired..., Xiang [/bib_ref] and subjects with obesity [bib_ref] Improvement of early vascular changes and cardiovascular risk factors in obese children..., Meyer [/bib_ref] [bib_ref] Moderate resistance training and vascular health in overweight women, Olson [/bib_ref] [bib_ref] Weight loss in combination with physical activity improves endothelial dysfunction in human..., Sciacqua [/bib_ref] [bib_ref] Effects of diet and exercise on obesity-related vascular dysfunction in children, Woo [/bib_ref] have also demonstrated enhanced FMD responses following different exercise training programs. Whilst a few studies exist which have not observed changes in conduit artery function following whole body exercise regimes [bib_ref] Effect of exercise on upper and lower extremity endothelial function in patients..., Gokce [/bib_ref] [bib_ref] Endothelial progenitor cell mobilization and increased intravascular nitric oxide in patients undergoing..., Paul [/bib_ref] , data in conduit arteries fundamentally reinforces the findings described above in resistance vessels, suggesting that vascular function is more amenable to enhancement in subjects with cardiovascular disease and risk factors who may exhibit impaired vasomotor and endothelial function a priori, than in healthy subjects with less impaired vascular function at the outset.
## Summary: conduit and resistance arteries
The use of small muscle group exercise training provides information regarding the impact of localised muscular activity, in the absence of marked change in central hemodynamics, on vascular adaptation. Conversely, larger muscle mass exercise involves both a localised stimulus in the active muscle, and systemic hemodynamic stimulus by virtue of both increases in cardiac output and peripheral changes in the sympathetic vasomotor tone sub-serving the control of blood flow and distribution. It is clear that these forms of exercise represent disparate stimuli in terms of their impact on endothelial shear stress and it is likely that some of the differences observed in terms of training effects in response to small and large muscle mass exercise reflect these differences.
No clear consensus emerges from the studies reviewed above regarding the impact of exercise training in healthy volunteers, possibly because of the disparate nature of the subjects studied or the various forms of exercise interventions applied. Conversely, most studies indicate that subjects with CV disease and risk factors benefit from exercise training. An explanation for some of the disparate findings may relate to the type or form of exercise used and in particular the intensity of exercise [bib_ref] Effect of different intensities of exercise on endothelium-dependent vasodilation in humans: role..., Goto [/bib_ref] [bib_ref] Exercise and cardiovascular risk reduction: time to update the rationale for exercise?, Green [/bib_ref] [bib_ref] Superior cardiovascular effect of aerobic interval training versus moderate continuous training in..., Wisloff [/bib_ref]. In this context, the study of Goto et al. is revealing, in that it suggests that low intensity exercise may fall below the threshold required for vascular adaptation in healthy subjects, whilst moderate intensity exercise is associated with improvements in endothelial function. Exercise at higher intensity [bib_ref] Intense physical training decreases circulating antioxidants and endothelium-dependent vasodilatation in vivo, Bergholm [/bib_ref] [bib_ref] Effect of different intensities of exercise on endothelium-dependent vasodilation in humans: role..., Goto [/bib_ref] may be associated with elevated levels of oxidative stress and inflammatory responses after exercise and this may, in turn, diminish any underlying shear and hemodynamic-mediated beneficial adaptations [bib_ref] Importance of hemodynamic forces as signals for exercise-induced changes in endothelial cell..., Laughlin [/bib_ref].
## Coronary vascular function
Animal studies have convincingly demonstrated that exercise induced flow-mediated epicardial coronary vasodilation is dependent upon the integrity of the endothelium [bib_ref] Role of vascular endothelium in exercise-induced dilation of large epicardial coronary arteries..., Berdeaux [/bib_ref] [bib_ref] Cardiovascular effects of exercise: role of endothelial shear stress, Niebauer [/bib_ref] , and experimental studies of coronary blood flows in humans indicate that NO blockade abolishes pacinginduced epicardial vessel dilation, indicating that NO contributes to vasodilation in these conduit vessels in patients free of CAD risk factors [bib_ref] Nitric oxide activity in the human coronary circulation. Impact of risk factors..., Quyyumi [/bib_ref] and in those with angiographically normal coronary arteries [bib_ref] Basal and flow-mediated nitric oxide production by atheromatous coronary arteries, Tousoulis [/bib_ref]. NO is also involved in pacing-induced hyperemia in patients with risk factors for CAD, including some with mildly irregular arteries [bib_ref] Contribution of vasodilator prostanoids and nitric oxide to resting flow, metabolic vasodilation,..., Duffy [/bib_ref] , whilst NO production increased by pacing was abolished at the sites of stenoses in patients with CAD disease [bib_ref] Basal and flow-mediated nitric oxide production by atheromatous coronary arteries, Tousoulis [/bib_ref]. These studies strongly indicate that acute exercise is associated with changes in endothelial function and, furthermore, that repeated exercise stimuli (i.e., training) which may induce chronic change in vascular function and also structure. It is important to note that animal studies suggest that coronary arteries of different caliber possess differences in eNOS content and NO-related adaptations to exercise training may be vessel caliber dependent. Larger vessels, which are exposed to higher shear stress forces, possessing greater capacity for NO production [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref] [bib_ref] Short-term training enhances endothelium-dependent dilation of coronary arteries, not arterioles, Laughlin [/bib_ref] [bib_ref] Influence of coronary artery diameter on eNOS protein content, Laughlin [/bib_ref]. Laughlin has previously expertly reviewed evidence regarding vascular adaptations to exercise training in conduit and resistance vessels of animals [bib_ref] Endothelial function and exercise training: evidence from studies using animal models, Jasperse [/bib_ref] [bib_ref] Exercise training-induced coronary vascular adaptation, Laughlin [/bib_ref].
In humans, several important studies have indicated a beneficial impact of exercise training on coronary vasodilator and endothelial function. Hambrecht et al. studied 19 stable CAD patients randomised to exercise training or control groups for a period of 4 weeks [bib_ref] Effect of exercise on coronary endothelial function in patients with coronary artery..., Hambrecht [/bib_ref]. Intra-coronary infusion of ACh and adenosine were used to assess epicardial coronary artery endothelium-dependent vasodilator function and resistance vessel function, respectively. Quantitative angiography and Doppler flow-wires assessed coronary diameter and flow. Training improved ACh and adenosine responses, indicating that coronary conduit and resistance artery endothelium-dependent vasodilator function was enhanced by exercise. In a subsequent study, the authors found that home-based exercise training sustained part of the effects of this hospital-based intervention . These authors also completed a comprehensive study which concluded that exercise training improves endothelial function in vivo by upregulating NO synthase protein expression and by increasing phosphorylation of NO synthase, effects consistent with a shear-stress mechanism for enhanced NO bioactivity with training ) (see ''Mechanisms responsible for change in vascular function and structure'').
Effect of exercise training on sympathetic nervous system and vasoconstrictor function Exercise training may alter autonomic balance in a manner which decreases cardiovascular risk. Increased parasympathetic, and decreased sympathetic, outflow to the heart would typically be cardioprotective [bib_ref] Effect of endurance exercise training on heart rate onset and heart rate..., Billman [/bib_ref]. Low heart rate variability is a prognostic index of cardiovascular mortality [bib_ref] Heart rate variability from short electrocardiographic recordings predicts mortality from all causes..., Dekker [/bib_ref] [bib_ref] Impaired autonomic function is associated with increased mortality, especially in subjects with..., Gerritsen [/bib_ref] [bib_ref] Reduced heart rate variability and mortality risk in an elderly cohort. The..., Tsuji [/bib_ref]. Some studies suggest that exercise training enhances heart rate variability [bib_ref] Heart rate variability and intensity of habitual physical activity in middle-aged persons, Buchheit [/bib_ref] [bib_ref] Elevated heart rate variability in physically active young and older adult women, Davy [/bib_ref] [bib_ref] Effect of endurance exercise training on heart rate variability at rest in..., Levy [/bib_ref] [bib_ref] Effects of moderate and vigorous physical activity on heart rate variability in..., Rennie [/bib_ref] [bib_ref] Prevalent low-frequency oscillation of heart rate: novel predictor of mortality after myocardial..., Wichterle [/bib_ref] , possibly by increasing large artery compliance [bib_ref] Does reduced vascular stiffening fully explain preserved cardiovagal baroreflex function in older,..., Hunt [/bib_ref] [bib_ref] Impact of resistance training on endurance performance, Tanaka [/bib_ref] , including that of the carotid sinus and aortic arch. This, in turn, may increase baroreceptor nerve traffic and increase parasympathetic tone. Alternatively, exercise training may lead to brain stem cardiorespiratory center remodelling which reduces sympathetic and enhances parasympathetic outflow [bib_ref] Effect of endurance exercise training on heart rate onset and heart rate..., Billman [/bib_ref] [bib_ref] Neuroplastic adaptations to exercise: neuronal remodeling in cardiorespiratory and locomotor areas, Nelson [/bib_ref].
Direct evidence for changes in sympathetic nervous system (SNS) mediated control of the vasculature after exercise training is lacking. It has been pointed out [bib_ref] Integration of cardiovascular control systems in dynamic exercise, Rowell [/bib_ref] that training is associated with large increases in cardiac output (stroke volume), whereas mean arterial pressure does not markedly change. A substantial increase in vascular conductance therefore occurs to accommodate the increase in output that accompanies training. Whilst this increase in conductance may result from enhanced vasodilator function (''Effect of exercise training on vascular function'') or arterial remodelling (''Effect of exercise on training on vascular structure''), another possibility is that SNS-mediated vasoconstrictor tone decreases with training. Ray et al. reviewed the few available studies relating to muscle sympathetic nerve activity (MSNA) and exercise and concluded that MSNA may decrease with training [bib_ref] Sympathetic neural adaptations to exercise training in humans: insights from microneurography, Ray [/bib_ref]. [bib_ref] The effect of unloading on protein synthesis in human skeletal muscle, Galbo [/bib_ref] also suggested that norepinephrine (NE) spillover decreases at matched absolute workloads following training. However, it has also been suggested that MSNA and NE spillover do Eur J Appl Physiol (2010) 108:845-875 855 not change after training when expressed in relative terms [bib_ref] Integration of cardiovascular control systems in dynamic exercise, Rowell [/bib_ref]. This suggests that SNS withdrawal is not responsible for the enhanced conductance evident following training ). In addition, Tanaka et al. recently reported enhanced vasodilation in response to phentolamine infusion following training, suggestive of increased SNS-mediated vasoconstriction [bib_ref] Systemic alpha-adrenergic and nitric oxide inhibition on basal limb blood flow: effects..., Sugawara [/bib_ref]. This, and the previous findings of [bib_ref] Coronary artery size and dilating capacity in ultradistance runners, Haskell [/bib_ref] , suggest that basal vasoconstriction may, in fact, increase following training. The impact of exercise training on locally produced and circulating vasoconstrictors such as ET-1 and angiotensin II (Ang II) has not been thoroughly researched [bib_ref] Physical (in)activity and endothelium-derived constricting factors: overlooked adaptations, Thijssen [/bib_ref]. Data derived from exercise-trained swine suggest that training improves aortic and coronary sensitivity to the vasoconstrictor ET-1 [bib_ref] Endothelin-1 sensitivity of porcine coronary arteries is reduced by exercise training and..., Jones [/bib_ref] and lowers ET-1 DNA expression [bib_ref] Exercise training attenuates coronary smooth muscle phenotypic modulation and nuclear Ca2? signaling, Wamhoff [/bib_ref]. In humans, two recent studies have reported that exercise training partly reverses increased ET-1-mediated vascular tone in the leg [bib_ref] Enhanced endothelin-1-mediated leg vascular tone in healthy older subjects, Thijssen [/bib_ref] and forearm [bib_ref] Endothelin-1 vasoconstrictor tone increases with age in healthy men but can be..., Van Guilder [/bib_ref] of older humans. Regarding Ang II, [bib_ref] Impact of regular physical activity on the NAD(P)H oxidase and angiotensin receptor..., Adams [/bib_ref] found that exercise training in patients with stable coronary artery disease leads to a 49% reduction in Ang II-induced vasoconstriction. Whilst the evidence relating to exercise training effects on vasoconstrictor pathways is far less comprehensive than that relating to vasodilator control mechanisms, the few studies performed in humans suggest that the contribution of vasoconstrictor pathways to vasomotor control may decrease after training.
## Effect of exercise training on vascular structure
Autopsy and angiographic studies performed in athletes [bib_ref] Half a century of running. Clinical, physiologic and autopsy findings in the..., Currens [/bib_ref] [bib_ref] Coronary arteries in physiological hypertrophy: echocardiographic evidence of increased proximal size in..., Pelliccia [/bib_ref] and physically fit individuals [bib_ref] Coronary flow reserve is supranormal in endurance athletes: an adenosine transthoracic echocardiographic..., Hildick-Smith [/bib_ref] [bib_ref] Atherosclerosis in the Masai, Mann [/bib_ref] [bib_ref] Myocardial infarction and the intrinsic calibre of coronary arteries, Rose [/bib_ref] suggest that physical conditioning induces an increase in arterial cross sectional area, also referred to as ''arterial remodelling''. Similarly, cross-sectional studies have consistently reported enlargement of skeletal muscle conduit [bib_ref] Arterial expansive remodeling induced by high flow rates, Ben Driss [/bib_ref] [bib_ref] Regular endurance exercise induces expansive arterial remodelling in the trained limbs of..., Dinenno [/bib_ref] [bib_ref] Structural and functional adaptations of the cardiovascular system by training, Huonker [/bib_ref] [bib_ref] Size and blood flow of central and peripheral arteries in highly trained..., Huonker [/bib_ref] [bib_ref] Diurnal pattern of vessel-wall properties of large arteries in healthy men, Kool [/bib_ref] [bib_ref] Arterial properties of the carotid and femoral artery in endurance-trained and paraplegic..., Schmidt-Trucksass [/bib_ref] [bib_ref] Vessel wall properties of large arteries in trained and sedentary subjects, Wijnen [/bib_ref] [bib_ref] Echocardiographic size of conductance vessels in athletes and sedentary people, Zeppilli [/bib_ref] and resistance [bib_ref] Endothelium-derived nitric oxide activity in forearm vessels of tennis players, Green [/bib_ref] [bib_ref] Enhanced maximal metabolic vasodilatation in the dominant forearms of tennis players, Sinoway [/bib_ref] vessels in athletes relative to matched controls, indicating that exercise training may induce arterial enlargement [bib_ref] Exercise-induced vascular remodeling, Prior [/bib_ref]. Findings from animal studies regarding epicardial and resistance coronary vasculature concur [bib_ref] Exercise and coronary vascular remodelling in the healthy heart, Brown [/bib_ref]. The studies described below have primarily utilised longitudinal designs in which subjects were followed across a training program.
## Resistance vessel remodelling
Sinoway et al. performed two of the earliest studies which specifically addressed the question of the impact of exercise training on resistance vessel ''structure''. They measured blood flow responses using stain-gauge plethysmography and utilised a metabolic stimulus (ischemia or ischemic exercise) to achieve maximal, or peak, forearm blood flow responses. By using a stimulus that induced peak localised dilation [bib_ref] Reactive hyperaemia in the human forearm, Patterson [/bib_ref] [bib_ref] Decreased vasodilator capacity of forearm resistance vessels in borderline hypertension, Takeshita [/bib_ref] , without inducing reflex changes in vasomotor control, they sought to assess the impact of exercise training on structural vascular adaptations, independent of central regulatory changes. As mentioned above, maximal or peak blood-flow responses in response to ischemic stimuli have commonly been used to assess resistance vessel structural adaptations in various settings [bib_ref] A vascular abnormality in hypertension. A study of blood flow in the..., Conway [/bib_ref] [bib_ref] Adaptive structural changes in the vascular walls in hypertension and their relation..., Folkow [/bib_ref] [bib_ref] The hemodynamic importance of structural vascular changes in essential hypertension, Sivertsson [/bib_ref] , based on the assumption that peak reactive hyperemia in response to a maximal vasodilator diminishes the impact of functional differences between subjects or following interventions [bib_ref] Reactive hyperaemia in the human forearm, Patterson [/bib_ref] [bib_ref] Decreased vasodilator capacity of forearm resistance vessels in borderline hypertension, Takeshita [/bib_ref]. Peak reactive hyperemic stimuli used in this context include 10 minutes of limb ischemia; the blood flow response to which cannot be increased by co-infusion of vasodilator agents [bib_ref] Decreased vasodilator capacity of forearm resistance vessels in borderline hypertension, Takeshita [/bib_ref]. More recently, periods of ischemia combined with ischemic exercise have been utilized to induce peak blood flow responses . While plethysmography has commonly been used for peak blood flow assessment, Doppler ultrasound methodology can also be used to directly measure blood flow through conduit arteries in humans [bib_ref] Assessment of brachial artery blood flow across the cardiac cycle: retrograde flows..., Green [/bib_ref] [bib_ref] Dependence of muscle VO 2 on blood flow dynamics at onset of..., Hughson [/bib_ref] [bib_ref] Ultrasound Doppler estimates of femoral artery blood flow during dynamic knee extensor..., Radegran [/bib_ref] at a higher temporal resolution .
Sinoway et al. demonstrated that the preferred arms of tennis players exhibit much higher peak vasodilator responses to 5 and 10 min periods of forearm ischemia ? exercise than the non-preferred limbs of these athletes or either limb of non-tennis playing control subjects [bib_ref] Enhanced maximal metabolic vasodilatation in the dominant forearms of tennis players, Sinoway [/bib_ref]. A subsequent study demonstrated that 4 weeks of hand-grip exercise training significantly enhanced the peak dilator response to a 10 min period of forearm ischemia in the trained, but not untrained, contralateral forearm [bib_ref] ) A 30-day forearm work protocol increases maximal forearm blood flow, Sinoway [/bib_ref]. The authors concluded that exercise training enhances the intrinsic ability of skeletal muscle resistance vessels to dilate [bib_ref] Central and peripheral circulatory changes after training of the arms or legs, Clausen [/bib_ref] [bib_ref] Effects of hypoxia and physical training on hemodynamic adjustments to one-legged exercise, Gleser [/bib_ref] [bib_ref] The nature of the training response; peripheral and central adaptations of one-legged..., Saltin [/bib_ref] [bib_ref] Cross transfer effects of muscular training on blood flow in the ipsilateral..., Yasuda [/bib_ref]. The findings were unlikely to result from changes in sympathetic tone, as Takeshita and Mark demonstrated that peak blood flow responses following 10 min ischemic stimuli were not altered by lower body negative pressure induced increases in sympathetic outflow [bib_ref] Enhanced maximal metabolic vasodilatation in the dominant forearms of tennis players, Sinoway [/bib_ref] [bib_ref] Decreased vasodilator capacity of forearm resistance vessels in borderline hypertension, Takeshita [/bib_ref]. In addition, [bib_ref] Central and regional circulatory adaptations to one-leg training, Klausen [/bib_ref] had previously demonstrated that, when cardiac output is not a limiting factor during leg exercise, training is associated with increased maximal leg blood flows due to enhanced vasodilator capacity.
The increases in peak vasodilator capacity observed by Sinoway were also not associated with muscle hypertrophy, as the training program did not alter maximal workload performed or forearm circumference, and because plethysmographic flow responses are normalised to dL of forearm tissue [bib_ref] Enhanced maximal metabolic vasodilatation in the dominant forearms of tennis players, Sinoway [/bib_ref] [bib_ref] ) A 30-day forearm work protocol increases maximal forearm blood flow, Sinoway [/bib_ref]. Finally, training-induced increases in skin blood flow were discounted, as an index of skin blood flow did not change with training [bib_ref] ) A 30-day forearm work protocol increases maximal forearm blood flow, Sinoway [/bib_ref].
The enhanced intrinsic vasodilator capacity of active muscle beds following training may conceivably result from the well established increase in capillary density that occurs with training [bib_ref] Capillary supply of the quadriceps femoris muscle of man: adaptive response to..., Andersen [/bib_ref]. However, as explained by [bib_ref] Maximal vascular leg conductance in trained and untrained men, Snell [/bib_ref] , muscle blood flow may not predominantly depend upon capillary density [bib_ref] Oxidative capacity, blood flow, and capillarity of skeletal muscles, Maxwell [/bib_ref]. Whilst capillaries regulate transit time and O 2 extraction, they contribute much less resistance to flow than upstream arterioles [bib_ref] Exercise and coronary vascular remodelling in the healthy heart, Brown [/bib_ref] [bib_ref] Morphometry of pig coronary arterial trees, Kassab [/bib_ref] [bib_ref] Vascular transport capacity of hindlimb muscles of exercise-trained rats, Laughlin [/bib_ref]. In addition, electrical stimulation experiments suggest that the time-course of adaptation in capillary density (rapid *4 days) [bib_ref] The effects of different patterns of muscle activity on capillary density, mechanical..., Brown [/bib_ref] is dissociated from adaptations in hyperemic flows (14-28 days) [bib_ref] The effect of long-term stimulation of fast muscles on their blood flow,..., Hudlicka [/bib_ref]. Adaptations observed in maximal blood flow or conductance responses with training therefore most likely reflect changes in the caliber or cross-sectional area of the resistance arteries, i.e. arterial remodelling, rather than increases in capillarity [bib_ref] Exercise and coronary vascular remodelling in the healthy heart, Brown [/bib_ref].
In addition to the studies of Sinoway described above, several other studies have observed increased resistance artery vasodilator capacity in response to localized small muscle group exercise training. Green et al. confirmed the Sinoway findings and added the observation that no significant change in either endothelium-dependent or -independent NO vasodilator function was apparent in trained limbs of healthy young subjects [bib_ref] Modification of forearm resistance vessels by exercise training in young men, Green [/bib_ref] [bib_ref] Endothelium-derived nitric oxide activity in forearm vessels of tennis players, Green [/bib_ref]. Martin observed enhanced maximal calf conductance following swim training in middle-aged men and women ) and also in older subjects following walking/running exercise [bib_ref] Exercise training enhances leg vasodilatory capacity of 65-yr-old men and women, Martin [/bib_ref]. Two studies performed in subjects with heart failure observed enhanced calf vasodilator capacity following lower limb exercise training [bib_ref] Exercise training in patients with severe congestive heart failure: enhancing peak aerobic..., Demopoulos [/bib_ref] [bib_ref] Effects of exercise training on limb blood flow in patients with reduced..., Dziekan [/bib_ref] , whilst other studies induced increased peak common femoral artery blood flows in response to voluntary and electrically stimulated thigh exercise [bib_ref] Local vascular adaptations after hybrid training in spinal cord-injured subjects, Thijssen [/bib_ref]. Finally, some studies have reported enhanced upper limb resistance artery vasodilator capacity following predominantly lower limb exercise training interventions [bib_ref] Exercise training, vascular function, and functional capacity in middle-aged subjects, Maiorana [/bib_ref] [bib_ref] Leg exercise conditioning increases peak forearm blood flow, Silber [/bib_ref] , although the generalisability of this vascular adaptation remains somewhat controversial [bib_ref] Exercise training does not induce vascular adaptations beyond the active muscle beds, Thijssen [/bib_ref] (''Local versus systemic adaptations to exercise'').
## Conduit artery remodelling
Several cross-sectional and longitudinal studies suggest that exercise training is associated with enlargement of skeletal muscle conduit arteries in humans. In an early study which utilized M-mode echocardiography, Zeppilli et al. observed significantly increased large artery (aorta, carotid, subclavian arteries) size in endurance trained athletes, relative to matched sedentary controls [bib_ref] Echocardiographic size of conductance vessels in athletes and sedentary people, Zeppilli [/bib_ref]. These differences persisted after correction for body surface area differences between the athletes and controls. Wheelchair athletes demonstrated enhanced dimensions in the aortic arch and subclavian artery, but lower values in the abdominal aorta and mesenteric artery. These findings essentially extended previous reports of enlargement in conduit arteries of athletes compared to control subjects [bib_ref] Vessel wall properties of large arteries in trained and sedentary subjects, Wijnen [/bib_ref]. Later findings by Huonker et al. suggested that larger elastic arteries (e.g., aorta) show less adaptability than smaller conduit arteries (subclavian, femoral) supplying peripheral limb muscle groups [bib_ref] Structural and functional adaptations of the cardiovascular system by training, Huonker [/bib_ref] [bib_ref] Size and blood flow of central and peripheral arteries in highly trained..., Huonker [/bib_ref] [bib_ref] Arterial properties of the carotid and femoral artery in endurance-trained and paraplegic..., Schmidt-Trucksass [/bib_ref]. One of these studies observed diminished femoral artery diameters in paraplegic subjects, as well as in the affected, but not unaffected, limbs of below-knee amputees, whereas cyclists possessed larger femoral arteries than these subjects and healthy controls and tennis players larger subclavian arteries in their racket arm than the contralateral limb [bib_ref] Size and blood flow of central and peripheral arteries in highly trained..., Huonker [/bib_ref]. Together, these cross-sectional data strongly suggest that chronic exercise training or detraining are associated with arterial remodelling. However, few of these studies corrected findings for between-subjects scaling factors which can potentially generate misleading interpretations [bib_ref] The athlete's heart: a contemporary appraisal of the 'Morganroth hypothesis', Naylor [/bib_ref].
One way to avoid the problems associated with between-subject comparisons is to examine the impact of an exercise training intervention on skeletal muscle conduit artery adaptations. In small sample studies of healthy young men, Miyachi et al. observed significant increases in the dimensions of the ascending and abdominal aorta following 8 weeks of cycle ergometer training [bib_ref] Effects of endurance training on the size and blood flow of the..., Miyachi [/bib_ref] and of the femoral artery in the trained, but not untrained limb, after 6 weeks of one-legged cycle exercise [bib_ref] Effects of one-legged endurance training on femoral arterial and venous size in..., Miyachi [/bib_ref]. These training effects were reversed following detraining [bib_ref] Effects of one-legged endurance training on femoral arterial and venous size in..., Miyachi [/bib_ref]. The authors concluded that regional, rather than systemic, factors are responsible for conduit artery remodelling as a result of training. Change in femoral artery diameter explained 70% of the variance in change in _ VO 2max with training, reinforcing the tight nexus that exists between conduit artery remodelling and enhanced central hemodynamic adaptation to training [bib_ref] Central and regional circulatory adaptations to one-leg training, Klausen [/bib_ref]. These data also infer that exercising limb changes in shear rate, rather than systemic changes in blood pressure or artery transmural pressure, are responsible for conduit artery adaptation [bib_ref] Importance of hemodynamic forces as signals for exercise-induced changes in endothelial cell..., Laughlin [/bib_ref]. Finally, in a study submitted concurrent with that of Miyachi [bib_ref] Effects of one-legged endurance training on femoral arterial and venous size in..., Miyachi [/bib_ref] , enhanced resting femoral artery diameter was observed following aerobic (walking) exercise training in previously sedentary men [bib_ref] Regular endurance exercise induces expansive arterial remodelling in the trained limbs of..., Dinenno [/bib_ref].
In all of the above studies, resting arterial diameter has been used as an index of arterial remodelling. However, resting diameter is dependent upon sympathetic nervous system tone, as well as paracrine and circulating hormonal modulation. As these competing vasodilator and constrictor influences impact upon resting tone, baseline artery diameter may not be an optimal index of vascular structure and remodelling following exercise training, a stimulus which modulates each of these factors . Therefore, it has been proposed that peak artery diameter may serve as a more appropriate assessment of conduit structure than resting measurements. This suggestion is reinforced by the data of [bib_ref] Coronary artery size and dilating capacity in ultradistance runners, Haskell [/bib_ref] , who observed no differences using quantitative angiography in the cross-sectional area of coronary arteries in athletes at rest, relative to matched sedentary control subjects, despite marked increases in maximal dilator capacity in response to pharmacological stimulation (GTN). These investigators suggested that, in the basal state when myocardial O 2 demand is not elevated, the coronary arteries of athletes may exhibit greater vascular tone [bib_ref] Coronary artery size and dilating capacity in ultradistance runners, Haskell [/bib_ref]. In summary, then, a strong argument can be made that, just as assessment of resistance vessel vasodilator capacity requires the application of a peak vasodilator stimulus, studies of conduit artery remodelling should attempt to assess maximal diameter or cross-sectional areas. We recently demonstrated, at least in the brachial artery, that peak conduit and resistance vessel structure can be simultaneously assessed using ultrasound Doppler approach and a combination of ischemic exercise or pharmacological stimulation . In another study, Naylor et al. observed training-induced enhancement of brachial artery diameter at rest and following a 10-min period of ischemia, thereby providing evidence for arteriogenic adaptation in response to a peak dilator stimulus [bib_ref] Decreased vasodilator capacity of forearm resistance vessels in borderline hypertension, Takeshita [/bib_ref]. The findings of this study in elite rowers suggest that chronic exercise induces marked arterial remodelling, and also that resumption of exercise training after a brief sojourn is also associated with further structural adaptation in highly trained individuals [bib_ref] Effects of training resumption on conduit arterial diameter in elite rowers, Naylor [/bib_ref].
## Coronary artery remodelling
Training-induced adaptations in large and small coronary artery diameter in animals have been expertly reviewed by [bib_ref] Exercise training-induced coronary vascular adaptation, Laughlin [/bib_ref] and [bib_ref] Exercise and coronary vascular remodelling in the healthy heart, Brown [/bib_ref]. These reviews indicate that well-conducted exercise studies suggest increased coronary flow capacity, a measure of resistance artery structural remodelling, as well as large epicardial arterial remodelling after training.
The earliest studies regarding coronary structural adaptation in humans involved autopsy analysis. In a series of necropsy experiments [bib_ref] Myocardial infarction and the intrinsic calibre of coronary arteries, Rose [/bib_ref] , blind analysis of plaque free segments of the right main coronary artery internal diameter was undertaken in subjects who died from myocardial infarction and a group of ''controls'' who had no evidence of post-mortem infarction. All measurements were undertaken at a constant distension pressure (80 mmHg). Subjects were classified according to whether they had been engaged in light, moderate or heavy physical activity occupations [bib_ref] Myocardial infarction and the intrinsic calibre of coronary arteries, Rose [/bib_ref]. The authors reported a 75% greater coronary artery diameter in control versus infarction subjects within the active and heavy physical activity occupational groups. This paper has been widely quoted in evidence that exercise is associated with coronary artery enlargement. However, whilst mean diameter in control subjects was somewhat larger (0.1 [ P [ 0.05) in those engaged in ''active'', compared to ''light'' physically active occupations (3.90 vs 4.30 mm), there was a surprising lack of effect for those in ''heavy'' occupations, compared to the light group (3.90 vs. 3.98 mm). The majority of the difference between infarction and control subjects within each occupational category was, in fact, evident in the diameter of the infarction subjects, whose diameters were lower in the more active groups (3.82, 3.22, 2.94 mm for light, active and heavy). This evidence is therefore open to the interpretation that, in those with small coronary arteries, higher activity levels are associated with greater risk of infarction.
Another widely quoted paper on the impact of exercise on coronary dimensions relates to the autopsy of Clarence De Mer [bib_ref] Half a century of running. Clinical, physiologic and autopsy findings in the..., Currens [/bib_ref] , who ran 100 marathons and 1,000 distance races and won the Boston marathon 7 times. His coronary artery dimensions were quoted as being ''two or three times the normal diameter''. However, the cardiac findings must be considered with some caution, as the post-mortem was performed after the heart was embalmed and had a trocar passed through it in several places. Unconfirmed reports also suggest that the examination may have occurred post-exhumation [bib_ref] Bruce Dill Historical lecture. Historical concepts of the athlete's heart, Thompson [/bib_ref].
A final historical report relating to the impact of exercise on coronary artery size relates to autopsy studies performed on 50 tribal Masai from Tanzania and Kenya [bib_ref] Atherosclerosis in the Masai, Mann [/bib_ref]. The coronary intimal thickness of the older Masai exceeded that of a cohort of US subjects aged 60-69 years and there was widespread presence of atheroma, similar to that expected in American men of similar age. However, symptomatic cardiovascular disease was rare in these subjects [bib_ref] Cardiovascular Disease in the Masai, Mann [/bib_ref] and the autopsies revealed that the coronary lumen displayed marked enlargement, such that the intimal thickening was not compromising. The Masai were remarkable fit [bib_ref] Physical fitness and immunity to heart-disease in Masai, Mann [/bib_ref] and the authors speculated that they were protected from the clinical manifestations of atherosclerosis by physical fitness, which caused their coronary arteries to enlarge.
The advent of trans-thoracic echocardiography encouraged an ambitious cross-sectional study of coronary artery size in 125 male athletes, selected from an initial sample of 625 subjects on the basis of acceptable image quality [bib_ref] Coronary arteries in physiological hypertrophy: echocardiographic evidence of increased proximal size in..., Pelliccia [/bib_ref]. Using a 3.25 MHz transducer to assess proximal epicardial diameter in the short axis view using M-mode echocardiography performed approximately 1 cm from the aortic ostium, these authors reported a correlation between coronary size and LV mass and wall thicknesses. A weak correlation was also evident for coronary size and _ V O 2max . The authors concluded from this correlational analysis that training-induced myocardial hypertrophy involves a proportionate increase in coronary lumen size [bib_ref] Coronary arteries in physiological hypertrophy: echocardiographic evidence of increased proximal size in..., Pelliccia [/bib_ref]. A cross-sectional study performed a decade later compared distal LAD diameter and blood flow, assessed using transthoracic echocardiography, in healthy control subjects and athletes during nitroglycerine administration and intravenous adenosine infusion. The diameter increase to nitroglycerine was greater in athletes (14.1%) than controls (8.8%) and coronary flow reserve was also greater in the athletes. At least some of the increased flow reserve was attributed to lower resting coronary flow in athletes, associated with relative bradycardia. The increase in peak flow was similar to that in controls when data were scaled for LV mass. Although the findings of this study were limited by the use of flow reserve in the absence of differences in peak flows, the limited spatial resolution of echocardiography as a technique for imaging coronary arteries and the systemic infusion of a dilator agent which may have induced different reflex coronary responses, it nonetheless suggests that epicardial coronary diameters are increased in athletes.
A number of studies have performed quantitative angiography to assess the impact of exercise training on coronary artery structure. In the study by no differences in the cross-sectional area of resting coronary arteries were observed in athletes, relative to matched sedentary control subjects, despite marked increases in maximal dilator capacity in response to nitroglycerine. The investigators concluded that the arteries of highly trained middle aged men exhibit greater dilating capacity than those who are sedentary [bib_ref] Coronary artery size and dilating capacity in ultradistance runners, Haskell [/bib_ref].
The intervention studies of Hambrecht et al. provide further insight. In the 10 stable CAD patients randomised to exercise training for a period of 4 weeks [bib_ref] Effect of exercise on coronary endothelial function in patients with coronary artery..., Hambrecht [/bib_ref] , intra-coronary infusion of ACh revealed higher epicardial diameter responses than controls, indicating that endothelium-dependent vasodilator function was enhanced (''Coronary vascular function''). Interestingly, the endothelium-independent epicardial dilator responses to GTN and adenosine were unaffected by training, suggesting no change in coronary conduit artery remodelling in response to this short period of exercise training. Resting arterial diameters, pre versus post intervention in each group, were not reported. Coronary velocity and flow reserve, measured using a Doppler flowwire, revealed increased coronary flow reserve responses to adenosine, suggesting enhanced vasodilator capacity of coronary resistance vessels. This study can therefore be interpreted as indicating that short-term exercise training in humans increases resistance vessel vasodilator capacity without inducing conduit artery remodelling. However, a follow-up study, involving a further 6 months of home based exercise training, reported enhanced adenosinemediated coronary flow reserve and coronary artery diameter, apparently evident at 4 weeks . In a further experiment involving assessment of left internal mammary artery adaptations to 4 weeks of exercise training in patients awaiting coronary bypass graft surgery, the training group exhibited enhanced adenosineinduced diameter and peak blood flow velocity responses, suggesting increased coronary conduit and resistance artery adaptation . There is therefore some inconsistency between these studies in terms of the reported conduit artery responses but, in general, evidence was provided for enhanced conduit and resistance vessel vasodilator capacity.
In common with the study of Haskell, no effects of exercise were reported by Hambrecht et al. in terms of resting resistance or conduit artery characteristics in the studies described above. This is further reinforced by a final experiment in which the effects of percutaneous coronary intervention with stenting (PCI) were compared to exercise training in a 12 month randomised trial of 101 male subjects. Relative to the PCI group, exercise training significantly enhanced cardiopulmonary fitness and cost approximately half as much due to fewer rehospitalisations or repeat procedures. At 12 months follow-up, the PCI group exhibited significantly increased lumen diameter (0.53-2.57 mm) and decreased relative stenosis diameter (80.7-11.8%), whereas exercise training had no impact on arterial stenotic size (0.66-0.69 mm, 77.9-76.5%). Despite this, there was significantly higher event-free survival in the exercise training group (88 vs. 70%). The implication of this study is that, whilst coronary interventions treat only a short segment of the diseased coronary tree, exercise training exerts beneficial effects on disease progression in the entire arterial bed.
Local versus systemic adaptations to exercise A series of studies which investigated the impact of exercise training on vascular function in humans [bib_ref] Combined aerobic and resistance exercise training improves functional capacity and strength in..., Maiorana [/bib_ref] [bib_ref] Combined aerobic and resistance exercise improves glycemic control and fitness in type..., Maiorana [/bib_ref] [bib_ref] Exercise training improves conduit vessel function in patients with coronary artery disease, Walsh [/bib_ref] [bib_ref] The fourth Volhard lecture: cardiovascular structural adaptation; its role in the initiation..., Folkow [/bib_ref] involved a similar exercise training intervention, a combination of resistance and aerobic exercise, performed under close supervision, in which upper limb exercise was avoided. In some of these studies upper limb conduit artery function [bib_ref] The effect of combined aerobic and resistance exercise training on vascular function..., Maiorana [/bib_ref] was examined, whilst in others the resistance vessel adaptations were assayed using intra-brachial infusions of acetylcholine, sodium nitroprusside and sometimes L-NMMA. All of these studies demonstrated improvement in upper limb endothelial function following lower limb training [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref] [bib_ref] Exercise and the nitric oxide vasodilator system, Maiorana [/bib_ref] , despite subjects being expressly requested to avoid hand-grip exercise. Hambrecht et al. [bib_ref] Endothelial dysfunction in patients with chronic heart failure: systemic effects of lower-limb..., Linke [/bib_ref] also observed systemic vascular effects of lower limb exercise. In addition, a review of the literature preceding these publications [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref] unearthed other studies indicating improvement in upper limb vascular function as a result of predominantly lower limb exercise [bib_ref] Exercise training enhances endothelial function in young men, Clarkson [/bib_ref] [bib_ref] Regular aerobic exercise prevents and restores age-related declines in endotheliumdependent vasodilation in..., Desouza [/bib_ref] [bib_ref] Regular physical exercise corrects endothelial dysfunction and improves exercise capacity in patients..., Hambrecht [/bib_ref] , although many of these did not report whether subjects were specifically requested to avoid incidental hand-gripping. More recent studies have largely reinforced evidence that lower limb exercise generates changes in upper limb vascular function (e.g., [bib_ref] Exercise training enhances endothelial function in young men, Clarkson [/bib_ref] [bib_ref] Regular aerobic exercise prevents and restores age-related declines in endotheliumdependent vasodilation in..., Desouza [/bib_ref] [bib_ref] Exercise training improves vascular endothelial function in patients with type 1 diabetes, Fuchsjager-Mayrl [/bib_ref] [bib_ref] Effect of different intensities of exercise on endothelium-dependent vasodilation in humans: role..., Goto [/bib_ref] [bib_ref] Superior cardiovascular effect of aerobic interval training versus moderate continuous training in..., Wisloff [/bib_ref] [bib_ref] Regular aerobic exercise augments endothelium-dependent vascular relaxation in normotensive as well as..., Higashi [/bib_ref] [bib_ref] Endothelial dysfunction in patients with chronic heart failure: systemic effects of lower-limb..., Linke [/bib_ref] [bib_ref] Combined aerobic and resistance exercise training improves functional capacity and strength in..., Maiorana [/bib_ref] [bib_ref] Exercise training, vascular function, and functional capacity in middle-aged subjects, Maiorana [/bib_ref] [bib_ref] Regular physical exercise improves endothelial function in heart transplant recipients, Schmidt [/bib_ref] (see.
In terms of generalised effects of training on vascular structure, a number of studies have investigated the impact of leg exercise training on forearm vasodilator capacity as an index of resistance artery remodelling. [bib_ref] Leg exercise conditioning increases peak forearm blood flow, Silber [/bib_ref] observed a significant increase in forearm peak reactive hyperemic flows following bicycle ergometer training during which arm exercise was minimized. They concluded that vascular responses to conditioning stimuli can involve vascular beds not specifically involved in the training stimulus. They furthermore suggested that this effect may be dependent upon the mass of muscle involved in the training stimulus, as hand-grip training studies were not associated with contra-lateral limb adaptations [bib_ref] Modification of forearm resistance vessels by exercise training in young men, Green [/bib_ref] [bib_ref] Endothelium-derived nitric oxide activity in forearm vessels of tennis players, Green [/bib_ref] [bib_ref] Enhanced maximal metabolic vasodilatation in the dominant forearms of tennis players, Sinoway [/bib_ref] [bib_ref] ) A 30-day forearm work protocol increases maximal forearm blood flow, Sinoway [/bib_ref]. Whilst some other studies involving predominantly lower limb exercise have observed evidence for upper limb resistance artery remodelling [bib_ref] Effect of aerobic and resistance exercise training on vascular function in heart..., Maiorana [/bib_ref] , others have not observed such an effect [bib_ref] Exercise training in patients with severe congestive heart failure: enhancing peak aerobic..., Demopoulos [/bib_ref] [bib_ref] Effects of exercise training on limb blood flow in patients with reduced..., Dziekan [/bib_ref] [bib_ref] Effects of exercise training on conduit and resistance vessel function in treated..., Walsh [/bib_ref]. However, some of the latter studies [bib_ref] Exercise training in patients with severe congestive heart failure: enhancing peak aerobic..., Demopoulos [/bib_ref] [bib_ref] Effects of exercise training on limb blood flow in patients with reduced..., Dziekan [/bib_ref] utilized a reactive hyperemic stimulus involving only 5 min of forearm ischemia, a stimulus which has repeatedly been demonstrated as insufficient to induce a true maximal hyperemic responses [bib_ref] Enhanced maximal metabolic vasodilatation in the dominant forearms of tennis players, Sinoway [/bib_ref] [bib_ref] Decreased vasodilator capacity of forearm resistance vessels in borderline hypertension, Takeshita [/bib_ref]. Studies of conduit arteries have not typically observed adaptation in non-exercised regions in healthy subjects [bib_ref] Regular endurance exercise induces expansive arterial remodelling in the trained limbs of..., Dinenno [/bib_ref] [bib_ref] Size and blood flow of central and peripheral arteries in highly trained..., Huonker [/bib_ref] [bib_ref] Regular aerobic exercise and the age-related increase in carotid artery intima-media thickness..., Tanaka [/bib_ref].
A mechanistic explanation for generalised changes in arterial structure and function may relate to the impact of lower limb exercise on blood flow and shear stress patterns in inactive arterial beds. Shear stress provides a major physiological signal to improvement in endothelial function [bib_ref] Crucial role of endothelium in the vasodilator response to increased flow in..., Pohl [/bib_ref] [bib_ref] Integration of cardiovascular control systems in dynamic exercise, Rowell [/bib_ref] and also adaptation in arterial size [bib_ref] Reductions in arterial diameter produced by chronic decreases in blood flow are..., Langille [/bib_ref] [bib_ref] Role of NO in flow-induced remodeling of the rabbit common carotid artery, Tronc [/bib_ref] [bib_ref] Shear level influences resistance artery remodeling: wall dimensions, cell density, and eNOS..., Tuttle [/bib_ref]. This effect is, at least in part, transduced by nitric oxide. Recent data suggests that lower limb exercise such as cycling induces a pattern of blood flow change in the brachial artery of the inactive upper limb characterized by increases in anterograde flow during systole as cardiac output increases, along with large increases in retrograde flows during diastole [bib_ref] Assessment of brachial artery blood flow across the cardiac cycle: retrograde flows..., Green [/bib_ref]. The ''amplitude'' of flow and shear rate changes increases with exercise intensity ). The shear stress sensitive endothelium is therefore not exposed to a smooth increase in laminar anterograde flows as lower limb exercise intensity increases, but rather, large oscillations in shear as blood is dragged in both directions across the cell membrane. This flow pattern in the brachial artery of the resting upper limb during lower limb exercise is associated with NO release [bib_ref] Effect of lower limb exercise on forearm vascular function: contribution of nitric..., Green [/bib_ref] , which in fact exceeds that associated with hand grip exercise, even when both types of exercise cause similar mean or average blood flows into the limb . It seems, therefore, that the mode and intensity of exercise performed has important impacts on the pattern of flow, even when bulk or mean flows are similar over time ). If endothelial phenotype is indeed sensitive to flow and shear stress patterns [bib_ref] Importance of hemodynamic forces as signals for exercise-induced changes in endothelial cell..., Laughlin [/bib_ref] , then different types of exercise may logically result in different endothelial adaptations and, consequently, different degrees of change in the health of the vessel wall and itsSummary of exercise training studies of conduit and/or resistance vessel function in healthy humans and subjects with cardiovascular disease and risk factors undertaking localised predisposition to atherogenic change [bib_ref] Effect of different intensities of exercise on endothelium-dependent vasodilation in humans: role..., Goto [/bib_ref] [bib_ref] Exercise and cardiovascular risk reduction: time to update the rationale for exercise?, Green [/bib_ref] [bib_ref] Superior cardiovascular effect of aerobic interval training versus moderate continuous training in..., Wisloff [/bib_ref].
## Time-course of vascular adaptation to exercise
There have been few studies of the time-course of arterial functional or structural adaptation to exercise training in humans. However, 2-4 week training in rats increased endothelial NO synthesis in skeletal muscle arterioles and vasodilator responses to ACh and L-arginine. SNP responses were unaltered, suggesting enhanced endothelial function, but unchanged smooth muscle cell sensitivity to NO [bib_ref] Short-term daily exercise activity enhances endothelial NO synthesis in skeletal muscle arterioles..., Sun [/bib_ref]. A study of similar duration demonstrated augmented dilator response, which were partially abolished by L-NMMA infusion in rats [bib_ref] Exercise training augments flow-dependent dilation in rat skeletal muscle arterioles. Role of..., Koller [/bib_ref].
In another study, 4 weeks of training not only enhanced ACh-induced vasodilation of the rat aorta but also increased eNOS protein levels in aortic tissue [bib_ref] Time course of enhanced endothelium-mediated dilation in aorta of trained rats, Delp [/bib_ref] [bib_ref] Exercise training alters endothelium-dependent vasoreactivity of rat abdominal aorta, Delp [/bib_ref]. Eight weeks of running in rabbits increased ACh reactivity in the aorta and pulmonary arteries, but not in the carotid artery [bib_ref] Physical conditioning can modulate endothelium-dependent vasorelaxation in rabbits, Chen [/bib_ref] , whilst 4 weeks of exercise in rats improved flowinduced dilation in skeletal muscle arteries, but not in mesenteric vessels [bib_ref] Adaptation of flowinduced dilation of arterioles to daily exercise, Sun [/bib_ref]. Improved endothelium-dependent vasodilation has been observed after as few as 7 days of endurance training in pigs in peripheral conduit vessels [bib_ref] Short-term exercise training alters responses of porcine femoral and brachial arteries, Mcallister [/bib_ref] as well as coronary conduit arteries [bib_ref] Short-term training enhances endothelium-dependent dilation of coronary arteries, not arterioles, Laughlin [/bib_ref]. These findings suggest that increased production of endothelial NO occurs rapidly in response to exercise training, particularly in arteries supplying active regions. Studies performed over a longer duration have not consistently shown augmented endothelial function. Endothelium-dependent vasodilation was unaltered after 16-20 weeks of training in pigs [bib_ref] Effects of exercise training on responses of peripheral and visceral arteries in..., Mcallister [/bib_ref] and 16 weeks in rats [bib_ref] Endothelium-mediated control of coronary vascular tone after chronic exercise training, Laughlin [/bib_ref]. There is also evidence that changes in eNOS expression are also time dependent. Expression of eNOS protein and enhanced ACh-mediated relaxation [bib_ref] Short-term exercise training increases ACh-induced relaxation and eNOS protein in porcine pulmonary..., Johnson [/bib_ref] were evident after 1 week of training in pigs, whereas changes were not present after 16 weeks [bib_ref] Chronic exercise training does not alter pulmonary vasorelaxation in normal pigs, Johnson [/bib_ref]. These data suggest that long-term training is not consistently associated with enhanced vascular function. However, it should be borne in mind that prolonged exercise training enlarges arterial diameters in animals [bib_ref] Reduction of coronary atherosclerosis by moderate conditioning exercise in monkeys on an..., Kramsch [/bib_ref] [bib_ref] Functional adaptations of rat skeletal muscle arterioles to aerobic exercise training, Lash [/bib_ref] [bib_ref] Effects of exercise and its cessation on the heart and its blood..., Leon [/bib_ref] [bib_ref] Influences of physical conditioning and deconditioning on coronary vasculature of dogs, Wyatt [/bib_ref]. As previously proposed [bib_ref] Endothelium-mediated control of coronary vascular tone after chronic exercise training, Laughlin [/bib_ref] , it is conceivable that vascular remodelling, an endothelium and NO-dependent phenomenon [bib_ref] The emerging concept of vascular remodeling, Gibbons [/bib_ref] [bib_ref] Adaptive regulation of wall shear stress to flow change in the canine..., Kamiya [/bib_ref] [bib_ref] Reductions in arterial diameter produced by chronic decreases in blood flow are..., Langille [/bib_ref] [bib_ref] Exercise-induced vascular remodeling, Prior [/bib_ref] [bib_ref] Direct evidence for the importance of endotheliumderived nitric oxide in vascular remodeling, Rudic [/bib_ref] [bib_ref] Shear stress regulation of artery lumen diameter in experimental atherogenesis, Zarins [/bib_ref] , may partly supplant the need for acutely Watts J Ped responsive vasodilator mechanisms to normalise shear stress during exercise bouts. A time-dependent change in conduit and resistance artery endothelial function is found in the coronary circulation. Early animal studies indicated that training increased vasodilator responses in dogs [bib_ref] Daily exercise enhances coronary resistance vessel sensitivity to pharmacological activation, Dicarlo [/bib_ref] and transport capacity in swine [bib_ref] Exercise training increases coronary transport reserve in miniature swine, Laughlin [/bib_ref]. were the first to report that eNOS gene expression is enhanced by exercise training in the coronary arteries. They observed increased nitrite and NO production, and eNOS gene expression, following 10 days of training in dogs [bib_ref] Chronic exercise in dogs increases coronary vascular nitric oxide production and endothelial..., Sessa [/bib_ref]. Exercise training also enhanced conduit artery NO-mediated dilation after 7-10 days of treadmill exercise training in another study of dogs [bib_ref] Chronic exercise enhances endothelium-mediated dilation of epicardial coronary artery in conscious dogs, Wang [/bib_ref]. No changes were evident in response to SNP. [bib_ref] Short-term training enhances endothelium-dependent dilation of coronary arteries, not arterioles, Laughlin [/bib_ref] observed an improvement in conduit artery endothelial function in coronary arteries after 7 days of exercise training in pigs. In contrast, in the same study [bib_ref] Short-term training enhances endothelium-dependent dilation of coronary arteries, not arterioles, Laughlin [/bib_ref] , they did not demonstrate a change in resistance artery endothelial function after short-term exercise training. Interestingly, the same group also demonstrated that 16-20 week exercise training in pigs did not increase coronary conduit artery endothelial function [bib_ref] Endothelium-mediated control of coronary vascular tone after chronic exercise training, Laughlin [/bib_ref] , while they demonstrated the presence of increased eNOS mRNA [bib_ref] Induction of nitric oxide synthase mRNA in coronary resistance arteries isolated from..., Woodman [/bib_ref] and increased bradykinin induced vasodilation [bib_ref] Vasodilator responses of coronary resistance arteries of exercise-trained pigs, Muller [/bib_ref] in coronary resistance vessels, suggesting that enhanced NO and endothelium-dependent dilation was persisting in these vessels. Finally, however, 16-22 weeks of training augmented vasodilator responses to adenosine in the large epicardial arteries of pigs, even after removal of the endothelium. This evidence suggests that changes in vasomotor response in coronary arteries with longer-term training might be due, at least in part, to adaptations within smooth muscle [bib_ref] Effects of exercise training on vasomotor reactivity of porcine coronary arteries, Oltman [/bib_ref]. As recently reviewed by Laughlin et al., the animal data suggest that coronary arteries follow a time-dependent adaptation to exercise training, which is different between conduit and resistance arteries. Whilst smaller coronary arterioles exhibit enhanced endothelium dependent vasodilation after longer term training, larger coronary arteries adapt rapidly to training and then return towards baseline levels when exercise training continues [bib_ref] Importance of hemodynamic forces as signals for exercise-induced changes in endothelial cell..., Laughlin [/bib_ref].
In summary, animal studies suggest that short-term exercise training enhances eNOS and NO production and bioactivity, producing a short term buffer to the increased shear associated with exercise. After extended training, at least in the peripheral circulation, the increased production of NO and possibly other mediators induces structural changes in the vessels resulting in an increase in lumen diameter [bib_ref] Exercise and coronary vascular remodelling in the healthy heart, Brown [/bib_ref] [bib_ref] Exercise-induced vascular remodeling, Prior [/bib_ref]. Shear stress may therefore be ''structurally'' normalised and endothelial NO activity returns towards initial levels.
Very little data exists regarding the time-course of arterial functional and structural adaptation to exercise training in humans. However, we recently completed a study in which measures of brachial and popliteal artery function and structure were collected every 2 weeks across an 8 week exercise training program in healthy young male subjects [bib_ref] Conduit artery functional adaptation is reversible and precedes structural changes to exercise..., Tinken [/bib_ref]. The results indicated that functional adaptation preceded changes in artery peak vasodilator capacity [fig_ref] Figure 2: Hypothesised changes in artery function and structure [/fig_ref]. These findings support the notion that functional adaptations may be superseded by structural changes including artery remodelling may normalise shear stress. They confirm previous reports that endothelial function rapidly adapts to training and detraining [bib_ref] Time-course of endothelial adaptation following acute and regular exercise, Haram [/bib_ref] [bib_ref] Time course of changes in endothelial function following exercise in habitually sedentary..., Pullin [/bib_ref].
## Healthy subjects versus subjects with cardiovascular disease
The detailed effects of exercise training in various populations with endothelial dysfunction are summarised in theand fully reviewed elsewhere [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref] [bib_ref] Exercise and the nitric oxide vasodilator system, Maiorana [/bib_ref]. In subjects with cardiovascular risk factors and disease, exercise training of localised muscle groups [bib_ref] Effect of exercise on coronary endothelial function in patients with coronary artery..., Hambrecht [/bib_ref] [bib_ref] Physical training improves endothelial function in patients with chronic heart failure, Hornig [/bib_ref] [bib_ref] Training improves endothelium-dependent vasodilation in resistance vessels of patients with heart failure, Katz [/bib_ref] , and whole body exercise predominantly involving the lower limbs [bib_ref] Effect of exercise on upper and lower extremity endothelial function in patients..., Gokce [/bib_ref] [bib_ref] Regular physical exercise corrects endothelial dysfunction and improves exercise capacity in patients..., Hambrecht [/bib_ref] [bib_ref] Effect of exercise on coronary endothelial function in patients with coronary artery..., Hambrecht [/bib_ref] [bib_ref] Regular physical activity improves endothelial function in patients with coronary artery disease..., Hambrecht [/bib_ref] [bib_ref] Regular aerobic exercise augments endothelium-dependent vascular relaxation in normotensive as well as..., Higashi [/bib_ref] [bib_ref] Endothelial dysfunction in patients with chronic heart failure: systemic effects of lower-limb..., Linke [/bib_ref] [bib_ref] Combined aerobic and resistance exercise training improves functional capacity and strength in..., Maiorana [/bib_ref] [bib_ref] The effect of combined aerobic and resistance exercise training on vascular function..., Maiorana [/bib_ref] [bib_ref] Regular physical exercise improves endothelial function in heart transplant recipients, Schmidt [/bib_ref] , are associated with improvement in measures of NO vasodilator function. There is consistency in the literature pertaining to exercise training mediated improvement in vascular function in groups in whom it is initially depressed. In contrast, studies of healthy subjects, with presumably normal endothelial function, are less compelling [bib_ref] Intense physical training decreases circulating antioxidants and endothelium-dependent vasodilatation in vivo, Bergholm [/bib_ref] [bib_ref] Exercise training enhances endothelial function in young men, Clarkson [/bib_ref] [bib_ref] Effects of intense exercise training on endothelium-dependent exercise-induced vasodilatation, Franke [/bib_ref] [bib_ref] Modification of forearm resistance vessels by exercise training in young men, Green [/bib_ref] [bib_ref] Exercise training, vascular function, and functional capacity in middle-aged subjects, Maiorana [/bib_ref] [bib_ref] Vascular adaptations to 8-week cycling training in older men, Thijssen [/bib_ref] and improvement may be limited to older subjects or those who undertake greater volumes of training [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref] [bib_ref] Exercise training, vascular function, and functional capacity in middle-aged subjects, Maiorana [/bib_ref]. Collectively, these findings suggest that subjects with impaired endothelial function may be more amenable to improvement in NO function as a result of training than healthy subjects.
Mechanisms responsible for change in vascular function and structure
## Vascular function
Hambrecht et al. provided an insight into the mechanisms responsible for exercise-mediated improvements in endothelial function . They studied the impact of the impact of 4 weeks of cycle exercise on the internal mammary artery of CAD subjects awaiting coronary artery bypass surgery. Training increased peak endothelium-dependent flow and flow-mediated dilation responses in the arteries of trained subjects, but not sedentary controls. After the final training session and the repeat in vivo vascular function assessments, a section of the internal mammary artery was harvested for in vitro vascular function assessment, immunohistochemistry, NO synthase mRNA isolation and protein quantification. Exercise training was associated with significantly higher NO synthase mRNA and protein expression and higher shear-stress related eNOS phosphorylation, which correlated with in vivo ACh-mediated vasodilator capacity. Exercise training therefore improves endothelial function in vivo by upregulating NO synthase protein expression and by increasing phosphorylation of this enzyme, effects consistent with a shear-stress mechanism for enhanced NO bioactivity with training.
Another explanation for improved vascular function with training was indirectly proposed by the results of [bib_ref] Effect of different intensities of exercise on endothelium-dependent vasodilation in humans: role..., Goto [/bib_ref] , who studied the effects of 12 weeks of exercise undertaken at low, moderate and high intensity and observed improvement in endothelial function in the moderate intensity group only. This occurred in the absence of changes in blood-borne measures of oxidative stress, whereas the high intensity group exhibited increased oxidative stress and no improvement in endothelial function. These findings suggest that the impact of exercise training on endothelial function may be dependent upon the intensity of exercise performed and the consequent balance between acute oxidative stress, chronic changes in antioxidant defences and their impact on NO bioavailability. Reactive oxygen species, such as superoxide anions, degrade NO before it has a vasodilator impact on vascular smooth muscle. However, repeated stimulation of NO production as a result of exercise training may reduce its degradation by free radicals [bib_ref] Regulation of the vascular extracellular superoxide dismutase by nitric oxide and exercise..., Fukai [/bib_ref] , directly decrease free radical production [bib_ref] Impact of regular physical activity on the NAD(P)H oxidase and angiotensin receptor..., Adams [/bib_ref] , or increase the expression of genes encoding for antioxidant enzymes [bib_ref] Physical training in patients with chronic heart failure enhances the expression of..., Ennezat [/bib_ref].
## Vascular remodelling
The classic study of Langille and O'Donnell established a link between changes in flow and arterial remodelling [bib_ref] Reductions in arterial diameter produced by chronic decreases in blood flow are..., Langille [/bib_ref]. They examined rabbit carotid arteries after unilateral ligation-mediated chronic decreases in flow (70% reduction, 2 weeks). The diameter of the ligated vessel was significantly smaller than the contralateral control vessel and this change was dependent upon the endothelium, inferring that flow-mediated changes in vessel structure are dependent upon the release of a substance from endothelial cells. This concurred with a similar study which found that shear stress was autoregulated after initial perturbation by an arteriovenous fistula [bib_ref] Adaptive regulation of wall shear stress to flow change in the canine..., Kamiya [/bib_ref]. Recent studies confirm that the stimulus to arterial remodelling is shear stress [bib_ref] Shear level influences resistance artery remodeling: wall dimensions, cell density, and eNOS..., Tuttle [/bib_ref] and that vessels enlarge to regulate wall shear in NO-dependent manner [bib_ref] Role of NO in flow-induced remodeling of the rabbit common carotid artery, Tronc [/bib_ref].
The above data are consistent with the evolving hypothesis that arterial shear stress is a homeostatically regulated variable [bib_ref] Flow-induced arterial remodeling relates to endothelial function in the human forearm, Vita [/bib_ref]. In this conceptual framework, shear stress mediated arterial enlargement, which is at least partly NO-dependent, acts to mitigate the increases in transmural pressure and wall stress brought about by repeated exercise bouts [bib_ref] Flow restriction of one carotid artery in juvenile rats inhibits growth of..., Guyton [/bib_ref] [bib_ref] Adaptations of carotid arteries of young and mature rabbits to reduced carotid..., Langille [/bib_ref] [bib_ref] Arteriogenesis and angiogenesis in rat ischemic hindlimb: role of nitric oxide, Lloyd [/bib_ref] [bib_ref] Exercise-induced vascular remodeling, Prior [/bib_ref] [bib_ref] Role of NO in flow-induced remodeling of the rabbit common carotid artery, Tronc [/bib_ref] [bib_ref] Tolerance to unilateral or bilateral ischemic hand exercise, Rodbard [/bib_ref] [bib_ref] Shear forces and blood vessel radii in the cardiovascular system, Zamir [/bib_ref]. The consequent ''structural'' normalisation of shear may obviate the need for ongoing and acute functional adaptations [bib_ref] Effect of exercise training on endothelium-derived nitric oxide function in humans, Green [/bib_ref] [bib_ref] Exercise and the nitric oxide vasodilator system, Maiorana [/bib_ref]. One recent study in humans provides some support for this notion, as functional adaptations were superseded by apparent changes in artery size [bib_ref] Conduit artery functional adaptation is reversible and precedes structural changes to exercise..., Tinken [/bib_ref].
## Optimal exercise training regimens
Several studies raise the possibility that different modalities or intensities of exercise may impact upon the magnitude of vascular adaptation observed. reported that 3 months of high intensity running reduced endothelium-dependent function but not endotheliumindependent function [bib_ref] Intense physical training decreases circulating antioxidants and endothelium-dependent vasodilatation in vivo, Bergholm [/bib_ref]. The degree of endothelial dysfunction following training was greatest in subjects with the largest improvements in _ V O 2max : The authors postulated that the training-induced decrease in circulating antioxidant levels may adversely affect endothelial function in the highly trained or overtrained state. [bib_ref] Effect of different intensities of exercise on endothelium-dependent vasodilation in humans: role..., Goto [/bib_ref] studied the effects low (25% _ VO 2max ), moderate (50% _ VO 2max ) and high (75% _ VO 2max ) intensity training in young men. Endothelium-dependent forearm vasodilation improved in the moderate intensity group only. This occurred in the absence of changes in oxidative stress. In the high intensity group, endothelial function did not improve, but there was evidence for increased oxidative stress. Taken together, the findings of Bergholm and Goto suggest that low intensity exercise may fall below a given threshold for improvement in endothelial function, whilst moderate intensity exercise enhances NO bioavailability. Any improvement in vascular function resulting from highintensity exercise may be abrogated by excess oxidative stress. However, this hypothesis clearly requires further testing, as it is also evident that higher intensity training may enhance antioxidant defence against oxidative stress [bib_ref] Impact of regular physical activity on the NAD(P)H oxidase and angiotensin receptor..., Adams [/bib_ref] [bib_ref] Physical training in patients with chronic heart failure enhances the expression of..., Ennezat [/bib_ref] [bib_ref] Regulation of the vascular extracellular superoxide dismutase by nitric oxide and exercise..., Fukai [/bib_ref].
In terms of the impact of exercise modality on vascular adaptation, the majority of studies indicating improvement in vascular function have utilised aerobic, or large muscle group dynamic exercise modalities, such as walking, running or cycling (see. Some studies which have combined aerobic and weight resisted exercises have also demonstrated generalised improvements in vascular function [bib_ref] Effect of aerobic and resistance exercise training on vascular function in heart..., Maiorana [/bib_ref] [bib_ref] The effect of combined aerobic and resistance exercise training on vascular function..., Maiorana [/bib_ref] [bib_ref] Exercise training normalizes vascular dysfunction and improves central adiposity in obese adolescents, Watts [/bib_ref]. Studies of resistance training have not observed changes in conduit artery function, but suggest that arterial remodelling may occur . Importantly, recent studies suggest that different modes of exercise, even when performed at similar relative intensities, generate distinct patterns of blood flow and shear stress through active and inactive vessel beds, raising the possibility of different shear mediated signals for adaptation ). Differences in blood pressure and transmural pressure between exercise modes may also contribute to differences in the adaptations evident [bib_ref] Importance of hemodynamic forces as signals for exercise-induced changes in endothelial cell..., Laughlin [/bib_ref]. Recent studies also suggest differences in endothelial adaptations to exercise of differing modalities in heart failure and healthy subjects [bib_ref] Both aerobic endurance and strength training programs improve cardiovascular health in obese..., Schjerve [/bib_ref] [bib_ref] Superior cardiovascular effect of aerobic interval training versus moderate continuous training in..., Wisloff [/bib_ref].
## Summary: exercise training and vascular adaptation
Collectively, the above data from both animal and human studies of exercise training suggest that functional and structural adaptations of the vasculature to exercise training alter with training duration and intensity and the vessel beds involved. Exercise training is associated with significant reductions in primary [bib_ref] Effects of walking on coronary heart disease in elderly men: the Honolulu..., Hakim [/bib_ref] [bib_ref] Exercise capacity and mortality among men referred for exercise testing, Myers [/bib_ref] [bib_ref] Physical activity and coronary heart disease in men: the Harvard Alumni Health..., Sesso [/bib_ref] and secondary vascular events [bib_ref] Exercise-based rehabilitation for coronary heart disease, Jolliffe [/bib_ref]. The effects of exercise on cardiovascular risk factors do not account for the magnitude of risk reduction [bib_ref] Exercise and cardiovascular risk reduction: time to update the rationale for exercise?, Green [/bib_ref] [bib_ref] Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms, Mora [/bib_ref]. Exercise exerts direct effects on the vasculature by virtue of the impact of repetitive intermittent increases in shear stress on the vascular endothelium, which responds by transducing functional and structural vascular adaptations which ultimately decrease atherosclerotic risk. Changes in transmural wall pressure may also represent a signal for chronic adaptation [bib_ref] Importance of hemodynamic forces as signals for exercise-induced changes in endothelial cell..., Laughlin [/bib_ref]. Hence, exerciseinduced improvements in vessel wall function and structure represent a ''vascular conditioning'' effect, which provides a plausible mechanistic explanation for the cardioprotective benefits of exercise, independent of the impact of exercise on traditional CV risk factors.
The clinical relevance of the vascular adaptations to exercise training was recently highlighted in a study which compared the effects of percutaneous coronary intervention with stenting (PCI) to exercise training alone in 101 male subjects [bib_ref] Percutaneous coronary angioplasty compared with exercise training in patients with stable coronary..., Hambrecht [/bib_ref]. After 12 months follow-up, the PCI group exhibited significantly increased lumen diameter (0.53-2.57 mm) and decreased relative stenosis diameter (80.7-11.8%), whereas exercise training had no impact on stenotic characteristics (0.66-0.69 mm, 77.9-76.5%). Despite this, there was significantly higher eventfree survival in the exercise training group (88 vs. 70%). This study reinforces the fact that coronary intervention treats a short segment of the diseased coronary tree, whilst exercise training exerts beneficial effects on endothelial function and disease progression in the entire arterial bed. The authors concluded that, in contrast to exercise training, interventional cardiology represents a palliative care measure with respect to the underlying atherosclerotic disease process and that exercise training should be a cornerstone of primary and secondary prevention efforts.
## Summary
Physical inactivity is considered a risk factor for cardiovascular disease in humans and exercise training is associated with a decrease in risk which is similar in magnitude (*30%) to that associated with pharmacological strategies. However, the effects of inactivity and exercise training on established and novel cardiovascular risk factors are relatively modest and cannot fully account for the impact of inactivity and exercise on vascular risk. However, both inactivity and exercise have direct effects on the vasculature and specific mechanistic pathways, for example those associated with shear stress and pulse pressure, have been identified which provide a basis for direct vascular conditioning and deconditioning effects of exercise and inactivity, respectively.
Inactivity is associated with rapid changes in arterial structure which result in inward remodelling. Vasodilator function in the remodelled resistance and conduit arteries appears to be within the normal range. It is possible that studies have not observed changes in vasodilator function because structural changes occur rapidly and allow for normalisation of function [fig_ref] Figure 2: Hypothesised changes in artery function and structure [/fig_ref]. Some evidence exists for enhanced and persistent increases in vasoconstrictor activity as a result of inactivity, even after artery remodeling has occurred.
Exercise training is typically associated with enhanced vasodilator activity which is ultimately superseded by outwards arterial remodelling and consequent re-normalisation of dilator function. These effects occur in vasculature of the active muscle beds, but also seem to be generalised to arteries supplying skeletal muscle that is not directly involved in the training stimulus. Shear stress on the artery wall appears to be a variable which is regulated by changes in vasodilator function in the first instance, and ultimately adaptive changes in artery size as the exercise stimulus persists. There is less evidence, in either conduit or resistance vessels, for diminished vasoconstrictor tone or function as an adaptation to training.
By far the majority of research has been undertaken on the relatively accessible and easily studied peripheral skeletal muscle conduit and resistance arteries. However, the limited extant evidence regarding coronary vascular adaptations suggests that the impact of exercise is consistent between the skeletal muscle and coronary circulations. Surprisingly, given the established epidemiological evidence linking inactivity and cardiovascular risk, there appear to be no studies which have directly assessed the impact of inactivity on coronary artery structure or function in humans.
Studies on the time-course of change in artery structure and function in response to inactivity and exercise are in their infancy in humans, but the limited available evidence supports the notion that functional change precedes that in structure and that both can occur rapidly across a timecourse of days and weeks, rather than months or years. The relative impacts of different durations or models of inactivity, or different exercise volumes and intensities, has not been comprehensively studied to date.
In conclusion, inactivity and exercise have direct ''vascular deconditioning and conditioning'' effects which likely modify cardiovascular risk. However, the nature and impact of inactivity and exercise on vascular structure and function suggest that inactivity and exercise are not simply the opposite ends of a linear spectrum of physiological adaptation.
[fig] Figure 2: Hypothesised changes in artery function and structure (remodelling) in response to inactivity and exercise training in humans. Studies performed in both animals and humans suggest that rapid changes occur in artery function, including nitric oxide (NO) bioavailability, in response to exercise training and that these changes are superseded by arterial remodelling and normalisation of function. [/fig]
|
Examining the Characteristics of Visuospatial Information Processing in Individuals with High-Functioning Autism
Information processing in individuals with autism is marked by a unique interplay of strengths and weaknesses that in concert distinguishes social cognition in autism from individuals with typical-functioning brains. In autism, difficulties with higher cognitive processing and enhancement of low-level visuospatial processing, such as in visual search tasks, may lead to diminished central coherence, which has the potential to hinder how an individual functions in social interactions where integration of components such as intention, emotion, and context paints the global picture necessary for social processing. A more thorough understanding of the cognitive and neural processes in autism is important for the advancement of intervention programs. The intention of this review is to discuss the implications of neuroimaging and behavioral studies that have analyzed the higher cognitive functions in individuals with high-functioning autism, with a particular emphasis on studies that have investigated visuospatial processing.
autism are often used interchangeably in the current clinical and research settings due to the historic challenges in segmenting the population because of the overlapping symptoms among the multiple Autism Spectrum Disorders. In the fourth edition of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Dis-orders (DSM-IV), ASD in fact is designated as an umbrella term under which there are five distinct pervasive developmental disorders: Rett syndrome, Childhood Disintegrative Disorder (CDD), autistic disorder, Pervasive Developmental Disorder Not Otherwise Specified (PDD-NOS), and Asperger syndrome. Rett syndrome is a disorder associated with a genetic mutation, and it mostly affects females. It is characterized by a series of events that includes normal childhood maturation and achievement of developmental milestones up until around 5 months of age, after which there is a significant regression in development that results in severe impairment of the child's psychomotor skills, social engagement, and language abilities. Features of Rett syndrome include dementia, characteristic hand movements (e.g., handwringing), and deceleration of cranial growth (in contrast to accelerated head growth seen in autism). CDD is an exceptionally rare syndrome in which children develop normally for at least 2 years and then experience a dramatic regression of the previously acquired skills, much like Rett syndrome, which leaves the children with severely affected language, social, and motor skills. However, CDD is also associated with seizures and metabolic disorders. Autistic disorder, PDD-NOS, and Asperger syndrome are phenotypically most similar and the most difficult to dissociate. Individuals with autism have marked difficulty with social interaction, certain cognitive skills, and language ability. Asperger syndrome, which is considered a milder variant, similarly affects cognitive and social functioning, but in contrast to autism, there is no generalized impairment of language development. Individuals who fall under the category of PDD-NOS display difficulties with language, social interaction, and cognition, similar to autism and Asperger syndrome, but do not share the full extent of the symptoms of either disorder. Although the causes and/or inciting factors of these various Autism Spectrum Disorders have yet to be definitively isolated, there is currently a basic understanding that the etiologies of these disorders are genetic, neurological, and cognitive in nature.
Autistic disorder, more commonly known as autism, is defined as a pervasive development disorder marked by impairments in social interaction, language, and communication and is characterized by restricted interests and repetitive behaviors. Accordingly, individuals with autism may display a narrow range of interest, often presenting at a young age with a preoccupation with particular objects or toys. Repetitive behavior can involve spending hours doing a specific task, for example, a child lining up his/her toys in order of size. A majority of individuals with autism have a developmental delay in verbal communication.
Language and communication abilities, if they develop at all, are often affected by pronoun reversal and echolalia, i.e., the involuntary repetitive imitation of vocalizations made by another individual. About 20 percent of individuals with autism display average to above average intellectual ability relative to individuals with typically functioning brains and are referred to as having high-functioning autism. For these individuals with high-functioning autism, performance of tasks involving simple language, memory, arithmetic, and rule-learning are in essence unimpaired. However, difficulty with more complex problemsolving, language, and memory and concept formation tasks impedes full-functioning in society. For example, in these high-functioning individuals, communication is hampered by the inability to respond appropriately to social cues, especially with regard to emotional context and common social expressions. Current treatment of individuals with autism centers on behavioral therapy to improve functioning in activities of daily living as well as enhancement of communication and interpersonal skills.
Although not part of the triad of symptoms, notable aspects of information processing in individuals with autism are increased reliance on the visuospatial method of processing and, correspondingly, a perceived advantage in that realm. Temple Grandin, an individual with high-functioning autism, has written in her book, Thinking in Pictures, that she attributes much of her cognitive functioning to dependence on visuospatial processing and that she, as many others with high-functioning autism have described, sees the world through pictorial representations. This greater command of visual cues over social and lingual cues in day-to-day operations may even help to explain some of the more archetypal characteristics of individuals with autism. It is still unclear, though, whether the dependence on visuospatial information processing causes the typically observed disconnect between visuospatial function and social cognition or if, instead, enhancement of visuospatial processing results as compensation for autonomous functioning of visuospatial brain regions that are disengaged from brain areas that typically perform executive functions. Regardless, the distinct pattern of behavior in autism points clearly to an atypical neural circuitry with alternate mechanisms of information processing. The distinct schemas of information processing at the cognitive and neural levels have been examined as a base from which the characteristic features of autism stem. Behavioral and neuroimaging studies have been successful in beginning to uncover the neural bases of autism.
## How does information processing in autism differ from the typically functioning brain?
A multitude of theories have been proposed to characterize information processing in autism, which is typically marked by the dichotomy of an individual exhibiting difficulties in certain higher-cognitive processing tasks along with a simultaneous enhancement of lower-level visuospatial processing. One such higher-cognitive process that is impaired in autism is described as "mindblindness". An impairment of theory of mind, mindblindness is the inability to attribute thoughts, feelings, emotions, and other mental states to oneself and others. Without the ability to understand that others are capable of formulating their own thoughts, it is impossible to interpret action and intention. This leaves individuals with autism incapable of "mind-reading," a skill that is imperative for facilitating social in-teraction among typically functioning individuals. As in a common example used to describe this phenomenon, looking down at a watch during a meeting implies an individual's desire to know the time and perhaps signals to the other parties in attendance to consider that time is running out, that the meeting may be winding down, that the watch-holder is busy and may have another meeting coming up, or is bored and waiting for the meeting to end. In order to have an effective social exchange, one must interpret the meaning behind such a simple action by getting into the mindset of another individual and attempting to predict that person's intentions. The various components in such a scenario must be integrated together and holistically understood before a contextually appropriate reaction can take place, e.g., accelerating the pace of the meeting, concluding the exchange, moving on to another more interesting topic, etc. Without the ability to see "the big picture," individuals with autism often struggle with formulating a suitable response in this type of social exchange.
One of the cognitive theories of autism, the Weak Central Coherence (WCC) theory, discusses this tendency in autism to "miss seeing the forest for the trees." The WCC theory offers two models to explain how this detail-focused bias in individuals with autism is both a disadvantage (the deficit model) and an advantage (the strength model), depending on the information processing tasks at hand. The WCC theory suggests that the impairment of social and cognitive functioning in autism is due to difficulties in putting information together to form context and hence failure to see "the big picture." This is explained as a result of an individual's inability to see the relationship between multiple components of a depictionor a detail-oriented bias. This is evident in tests, such as the homograph test, in which it has been observed that autistic individuals have difficulty distinguishing the appropriate pronunciation of homonyms when given two different contexts, e.g., there was a tear in her dress vs. there was a tear in her eye. Another example of the deficit model is with regard to social cognition and social interaction. As discussed earlier, in these instances, autistic individuals display an inability to holistically integrate the components of a social situation in order to comprehend its greater meaning. As such, it is more difficult for people with autism to distinguish emotions, respond appropriately to others' emotions, or to comprehend the relevance of emotions in social situations. On the other hand, the "strength" model of the WCC theory proposes that individuals with autism simply have a superior ability to process features, rather than an inferior ability to process a global pattern. The increased focus on details in autism has been found to be advantageous in tasks such as the Embedded Figures Task (EFT) and the Block Design (BD) component of the Wechsler IQ test. In an EFT, subjects are asked to identify a simple image concealed within a more intricate image. For example, the subject is asked to find a triangle hidden within a more complex drawing of a clock. In the BD task, the subject is given blocks that individually have different color patterns and is asked to arrange the blocks according to a given pattern. Performance in such tasks re-quires ignoring the global aspects and focusing on the details, and subjects with autism exhibited intact or superior performance when compared to their typically functioning counterparts.
The WCC theory also relates to the Gestalt principles of psychology, which emphasize the human tendency to see things in global form. However, the difference between global and Gestalt stimuli must be explained. Global stimuli are recognized without regard to the specificity of the local element, while the Gestalt stimuli are apparent by the precise orientation of particular local features. For example, a global figure might be composed of nine small triangles oriented to construct a greater triangle, while a Gestalt figure might include three circles placed in the shape of a triangle with precise pie sliceshaped exclusions that when seen in the cohesive "whole" form induces the mind to re-organize the seemingly unrelated components of the figure to perceive a triangle where one has not been drawn. A parallel can be made between this Gestalt processing and social cognition, which also requires the integration of seemingly un- connected stimuli to derive the nature of a given situation. Examination of Gestalt perception in high-functioning individuals with autism, similar to investigations of global processing, has revealed inferior Gestalt perceptual ability.
There is conflict in the literature concerning the performance of autistic individuals on tasks that require global and local visuospatial information processing compared to typically functioning individuals. Performance is compared based on accuracy and reaction time of the task completed. While some studies find superior performance by the autism group on tasks that require participants to focus on the details and some find inferior performance of individuals with autism on tasks that require attention to the global picture, other studies find that there is no significant difference in accuracy or reaction time between the two groups. The "hierarchization deficit model"provides an explanation of the differences in performance in visuospatial tasks and contrasts the WCC model by reasoning that individuals with autism are equally capable of performing global and local tasks. This theory proposes that the default way of information processing in typical individuals is global and the default in people with autism is local. Accordingly, individuals with autism are capable of performing equally to control participants on visuospatial processing tasks, but display a preference for local processing over global processing. On the other hand, the variation in results across the behavioral studies may be explained by the inconsistency in methodological and analysis techniques used. Additionally, the heterogeneity within autism itself may make it difficult to analyze the extent of the detail-focused bias. Interestingly, White et al.elucidated how studies that demonstrated a considerable difference in performance on global versus local tasks used lower-functioning groups than the studies that did not discover a significant difference in performance. The severity of ASD may be positively correlated with the intensity of the difference in central coherence, thereby supporting the case for the Weak Central Coherence theory.
Minshew et al.suggests that performance capability of individuals with autism on mental tasks, including visuospatial tasks, is dependent on the complexity of the task and proposes that autism is characterized by a disorder of complex information processing rather than a detail-oriented bias. Similarly, Bertone et al.proposes a complexity-specific account that characterizes visuospatial information processing in individuals with autism. The study by Bertone et al. suggested that high-functioning autistic participants had enhanced performance in less complex image discrimination tasks (first order or luminancedefined processing) and impaired performance in tasks that were more complex (second-order or texture-defined processing). In typically functioning participants, the less complex image discrimination task is associated with a single visual area (V1) versus in the more complex image discrimination task, which is associated with recruitment of multiple visual areas (V1+V2/V3). A proposed "superior when autonomous, inferior when synchronized" theory explains the difference in performance of more complex versus less complex visuospatial processing in autism. This theory suggests a processing pathway that is efficient and superior when recruited in isolation, as in the less complex task. In contrast, the more complex image discrimination task that requires multiple visual areas to work in sync are less efficient and perform inferiorly in high-functioning autism.
## Cortical correlates
Most of the previous studies that have investigated visuospatial information processing in high-functioning adults with autism used mainly behavioral tasks that gathered data such as accuracy and reaction time to make inferences about cognitive processing styles. There are only a few neuroimaging studies that target global and local processing in autism. Functional magnetic resonance imaging (fMRI) scanners have been a prominent tool used in gathering information about autism. fMRI studies have been used to measure brain activation in cortical gray matter regions during performance of behavioral tasks and, through group analysis techniques that subtract intensity of cortical activation, allow examiners to identify how cortical activation patterns differ between participants with autism and typically functioning controls. Data gathered from the scanner along with data collected on accuracy and reaction time help to develop a fuller story of how visuospatial processing in autism is different. The Embedded Figures Task has been used commonly in fMRI studies to evaluate cortical activation in global and local aspects of static visuospatial information processing tasks. In an fMRI study in which subjects performed the EFT, Ring et al.found that there were differences between the autism and control groups in the cortical regions recruited for executing the task and suggested that the control group's mechanism for task completion is reliant on working memory systems in contrast to the autism group, which demonstrated a different activation pattern that may suggest a greater dependence on visual systems for visuospatial reasoning. In particular, this study found a lack of frontal activation in autism and greater reliance on the inferior temporal lobe for complex visual search tasks. In a similar fMRI study utilizing the EFT, Damarla et al.identified differences in activation in several distinct cortical regions, with autism observed to have reduced activation in the left dorsolateral prefrontal and inferior parietal areas, and enhanced activation in bilateral superior parietal extending to the inferior parietal and right occipital regions when compared to typically developing controls. In an additional study utilizing the EFT, Manjaly et al.found that individuals within the ASD group had greater activation in the right primary visual cortex and bilateral extrastriate areas, and the control group of typically developing individuals recruited the left parietal and premotor regions to complete the task. The general interpretation of these findings is that during visuospatial tasks, the ASD groups display greater dependence on visual processing areas, while the control groups demonstrate a significant dependence on regions of the brain that are traditionally thought of as executive functioning regions. In another fMRI study, Lee et al.discovered that during EFT performance, control participants had activation in the left dorsolateral, medial, and dorsal premotor regions of the frontal cortex and bilateral superior parietal and occipital cortical regions. ASD participants, in contrast, activated only the dorsal premotor, left superior parietal, and right occipital regions. Lee et al. proposes that the left dorsolateral prefrontal recruitment observed in the control group suggests involvement in verbal working memory, indicating that typically functioning individuals were reliant on additional verbal strategies for working through the complex visual EFT. Generally, across these multiple fMRI studies, the ASD group had greater involvement of the visual processing areas of the cortex during performance of the EFT task compared to the control group . As discussed earlier, the behavioral analysis in many of the studies did not reveal a significant difference in accuracy or response time; however, the activation patterns indicate a greater suppression of the global bias in typically functioning individuals and a more concise pattern of involvement in ASD individuals in local processing, which is consistent with the Weak Central Coherence theory.
In contrast, the control groups demonstrated greater activation in cortical regions responsible for higher-order thought processes. According to Friston et al., in typically functioning individuals, the brain generates global inferences by making predictions of how sensory information fits in with prior knowledge and situational context and evaluates its significance based on a developed theory of mind. This inferencemaking ability is important for formulation of proper responses to stimuli in the surrounding world. When there is a mismatch between prediction developed in executive functioning regions and sensory information from visuospatial processing cortical regions, there is greater error in inferencemaking ability and disabled formation of proper responses. The activation and behavioral patterns identified in the various studies discussed earlier suggest that there is a lack of coordination between executive functioning areas and visuospatial sensory processing in autism, and as supported by table 1. cortical activation differences between autism and control groups in four major fmri studies utilizing the Embedded Figures task to examine visuospatial processing. Ring et al.Damarla et al.Manjaly et al.Lee et al.cortical Activation Areas in which the control group showed more activation than the autism group Areas in which the autism group showed more activation than the control group study Ring et al.Damarla et al.Manjaly et al.Lee et al.cortical Activation Loth et al.and Soulieres et al., this altered top-down influence may consequently produce deficient modulation of visuospatial information by context and prior knowledge. This disruption in autism in multiple level analyses may explain the inability to access the global meaning developed by integrating sensory information, prior knowledge, context, and theory of mind.
While the WCC theory explains the cognitive aspects of information processing in autism, the cortical underconnectivity theoryprovides a compelling related neural model of autism. Functional connectivity is defined as the collaboration between brain areas through synchronization. Functional underconnectivity refers to the lower levels of synchronization of these brain areas in the performance of certain cognitive tasks. The theory of functional underconnectivity helps to explain the consistent discrepancy in mechanism of task completion between autism and control groups, suggesting that individuals with autism are more reliant on posterior regions of the brain to work in autonomy rather than collaborating with the frontal areas to perform functions of visuospatial as well as cognitive tasks. Functional underconnectivity in autism has been demonstrated in several higher cognitive functions, such as sentence comprehensionand theory of mind. The postulates of underconnectivity theory are consistent with the WCC theory. While WCC theory talks about failure of integrating cognitive information to form coherence, the underconnectivity theory provides the neural bases of weak coherence in autism. In autism, the brain areas (especially frontal and posterior) do not work as a team to overcome increased computational demand that accompanies more complex tasks. In other words, such underconnectivity among brain regions might be the reason behind the difficulty faced by individuals with autism when performing highly demanding global processing tasks, such as social cognition, problem solving, inhibition, and language comprehension. Just et al.elaborates that the decreased communication between frontal and parietal areas is further demonstrated by decreased communication bandwidth, meaning the maximal rate of data transmission, between the regions. This impairment when integrating information across cortical regions is supported by the "superior when autonomous, inferior when synchronized" theory suggested as an explanation of the impairment in more-complex tasks and enhancement in less-complex tasks. Additionally, Just et al. suggests that in autism, there is a preference to complete tasks using posterior cortical regions. The study also demonstrated that in autism there is a greater dependence on the posterior cortical areas to perform executive functions independently of the frontal regions. This may lead individuals with autism to process content in a more visual or graphic context due to a decreased dependence on frontal areas. The important causal relationship between connectivity and perception is only starting to be mapped out.
# Conclusion
A comprehensive understanding of how the trends obtained from behavioral studies compliment cortical activation patterns derived from imaging studies is important for developing insight into how information processing differs in individuals with ASD. This perspective is crucial for the advancement of intervention programs and designing recommendations for health care providers on specific treatment approaches for both children and adults with autism. Additionally, this understanding of how individuals with autism process the stimuli in the world around them is important for parents and caretakers. For example, understanding that individuals with autism have a detail-focused bias and preference for pictorial representations of concepts suggests that parents could use picture books or other simple, static visual media over oral reasoning or descriptions to explain ideas to help facilitate their child's educational and social development.
Further studies need to be conducted in order to advance our understanding of the neural circuitry that supports cognitive functioning and how these circuits are differentially engaged in people with autism relative to typically functioning individuals. Perhaps an experiment that examines the detail-focused bias in Gestalt processing can help draw a parallel to the disruptions in social cognition. Continued analysis has the potential of developing valuable clues that can help in cultivating more tailored intervention programs to suit the needs of the unique cognitive styles in autism. |
Factors that influence smokers’ and ex-smokers’ use of IQOS: a qualitative study of IQOS users and ex-users in the UK
[bib_ref] Heat-not-burn tobacco product use in Japan: its prevalence, predictors and perceived symptoms..., Tabuchi [/bib_ref] [bib_ref] fourth quarter & full year results; full-year 2018 reported diluted EPS of..., Morris International [/bib_ref]
# Introduction
The IQOS battery-powered heated tobacco product (HTP) by Philip Morris International (PMI) was launched in 2014 1 2 (figure 1). PMI claims that IQOS heats rather burns tobacco sticks (branded HEETS), thus avoiding burning and emitting "95% lower levels of harmful chemicals compared with cigarettes". Available in over 40 countries, PMI estimate that 6.6 million smokers have quit smoking and switched to IQOS. [bib_ref] fourth quarter & full year results; full-year 2018 reported diluted EPS of..., Morris International [/bib_ref] Since the 2016 UK launch of IQOS, sale locations have expanded beyond official IQOS stores to other retailers. Prevalence of IQOS use in the UK is uncertain. One study estimated overall HTP use at 0.82%; however, the measure used may have been problematic. [bib_ref] Awareness and use of 'heat-not-burn' tobacco products in Great Britain, Brose [/bib_ref] IQOS advertising appears in the UK, but HEETS advertising does not. IQOS retails between £59 and £89; HEETS retailed at £8 for a packet of 20 until July 2019 when PMI reduced the price to £5, less than half that of most combustible cigarettes. [bib_ref] Overview of Tax Legislation and Rates, Hm Treasury [/bib_ref] HEETS are not subject to standardised packaging and pictorial warnings, and only require a 30% black-and-white text warning (figure 2). There are no official policies on using IQOS in public.
In April 2019, the US Food and Drug Administration (FDA) authorised IQOS and 'Heat Sticks' for marketing in the USA.Research (mostly tobacco industry funded) suggests that, compared with combustible cigarettes, there are reduced toxins in the aerosol produced by IQOS; IQOS may reduce users' exposure to some "harmful and potentially harmful constituents"; and some biomarkers of harm may be reduced when switching from smoking combustible cigarettes to IQOS. [bib_ref] Heat-not-burn tobacco products: a systematic literature review, Simonavicius [/bib_ref] [bib_ref] Carbonyl emissions from a novel heated tobacco product (IQOS): comparison with an..., Farsalinos [/bib_ref] [bib_ref] Levels of selected analytes in the emissions of "heat not burn" tobacco..., Mallock [/bib_ref] [bib_ref] Changes in biomarkers of exposure on switching from a conventional cigarette to..., Gale [/bib_ref] [bib_ref] IQOS: examination of Philip Morris International's claim of reduced exposure, St [/bib_ref] [bib_ref] Tobacco-specific nitrosamines (TSNA) in heated tobacco product IQOS, Leigh [/bib_ref] [bib_ref] Free-base and total nicotine, reactive oxygen species, and carbonyl emissions from IQOS,..., Salman [/bib_ref] [bib_ref] Reduction in exposure to selected harmful and potentially harmful constituents approaching those..., Haziza [/bib_ref] However, the public health impact of IQOS will depend on the degree to which use is less harmful than smoking, whether smokers switch completely to IQOS, and the extent of non-smoker uptake. Most research on consumer responses to IQOS has been conducted by PMI. [bib_ref] Light and mild redux: heated tobacco products' reduced exposure claims are likely..., Popova [/bib_ref] [bib_ref] IQOS labelling will mislead consumers, Mckelvey [/bib_ref] Independent survey research found that IQOS is perceived as less harmful than smoking. [bib_ref] Awareness, experience and prevalence of heated tobacco product, IQOS, among young Korean..., Kim [/bib_ref] [bib_ref] An online survey of users of tobacco vaporizers, reasons and modes of..., Queloz [/bib_ref] Focus group studies in Switzerland and Japan found IQOS was more socially acceptable, cleaner and uptake was encouraged by 'clean and chic' marketing and less of a harsh feeling on the throat compared with cigarettes. [bib_ref] Examining perceptions about IQOS heated tobacco product: consumer studies in Japan and..., Hair [/bib_ref] However, the bulkiness of the device, charging/ cleaning, cost and the odd smell discouraged use. [bib_ref] Examining perceptions about IQOS heated tobacco product: consumer studies in Japan and..., Hair [/bib_ref] Popular reasons for using IQOS identified in a Swiss study included complete smoking cessation, the reduced toxicity relative to smoking, the reduced smell and the improved taste. [bib_ref] An online survey of users of tobacco vaporizers, reasons and modes of..., Queloz [/bib_ref] Most users found IQOS weaker on the throat than combustible cigarettes and reported improved physical health, but feared becoming dependent. [bib_ref] An online survey of users of tobacco vaporizers, reasons and modes of..., Queloz [/bib_ref] Due to increased availability of IQOS, its use for harm reduction and its rapid uptake in some countries, [bib_ref] Effect of IQOS introduction on cigarette sales: evidence of decline and replacement, Stoklosa [/bib_ref] it is crucial that we develop a better understanding of smokers' perceptions and responses to IQOS. Thus, we conducted a qualitative study with IQOS users and ex-users in London to explore and understand the factors that encourage and discourage current and ex-smokers to initiate, continue, and discontinue IQOS use. We used interviews to explore individuals' experiences in detail and to obtain a greater depth of data than could be generated from focus groups.
# Methods
Eligibility criteria were adults (18+) who (1) currently smoked or quit smoking in the last 2 years, (2) use or used IQOS at least weekly for at least 1 month, and (3) lived in the UK. We recruited online (research recruitment website, classifieds, social media) and via a university-wide email. Posters were also displayed in vape shops selling IQOS across London and we approached individuals using IQOS in public. After screening, eligible individuals were selected for interview based on their demographics and smoking/IQOS histories. Between October 2018 and February 2019, CNET, a qualitative researcher and joint lead author, interviewed 30 people (28 face-to-face, 2 by telephone). A semistructured topic guide, informed by consultations with seven tobacco/nicotine experts and a panel of 12 current and ex-smokers, guided the interviews. It covered participants' smoking history, motivations and experiences of using IQOS (including reasons for continued and discontinued use), the benefits and drawbacks of IQOS, and the perceived impacts of IQOS (overall and on their smoking). Participants were offered a £20 shopping voucher. Interviews lasted 36-102 min (mean=67).
Iterative categorisation, a rigorous and transparent approach, guided data coding and analyses.The authors read transcripts to familiarise themselves with the data, discussed the content, and developed a coding frame of deductive (based on areas included in the topic guide) and inductive (based on additional areas that arose during the interviews) thematic codes. The transcripts and the coding frame were imported into MaxQDA; CNET systematically coded each transcript and assigned all text to the relevant code/s. Then, AMcN, SCH and CNET jointly reviewed the coded data relating to why participants initiated, continued, and discontinued IQOS more inductively. They consolidated and re-organised the codes into health, financial, physical, practical, psychological and social factors. During analysis, the similarities and differences between participants were explored. Key findings are illustrated using pseudonymised verbatim quotations.
## Findings sample
Participants included n=22 current and n=8 ex-IQOS users, all UK residents from a variety of nationalities (table 1). Many had smoked or used nicotine products for 20+ years; most had tried e-cigarettes (table 2 and online supplementary tables 1-3). All currently used or had used IQOS 2.4 Plus.
Although we recruited smokers and ex-smokers, the interviews highlighted that smoking, IQOS and e-cigarette status was more complex. For example, some people identified as ex-smokers but when interviewed described recent situations where they smoked combustible cigarettes. Consequently, the quotations from people who said they no longer smoked at all (ex-smokers) are labelled as 'former smoking' to reflect that they have former experience of smoking but are not necessarily ex-smokers.
## Health factors
## Smoking goal
Commonly, participants said that they experimented with IQOS because their goal was to reduce or stop smoking combustible cigarettes due to health risks. They thought that IQOS potentially *These included Bulgaria, Italy, Lithuania, Switzerland and Ukraine. †As defined by answering "How often do you currently use e-cigarettes?" as "Not at all" and answering "Have you ever previously used e-cigarettes?" as "Yes". ‡As defined by answering "Have you ever previously used e-cigarettes?" as "No". §This includes use of other tobacco and nicotine products during this period (eg, shisha and cigars). One participant did not disclose this information in a specific enough way, so exists as missing data. N.B. This also excludes large periods of time (over 1 year) in which successful quit attempts were made (ie, these years were removed from the total time spent using tobacco/nicotine products).
## Original research
provided a new and 'healthier' path away from cigarettes. Some explained this goal differently-with no intention of 'giving up' or 'quitting' altogether, they perceived IQOS offered an alternative way to carry on 'smoking'. While these individuals started IQOS to continue smoking (using it instead of combustible cigarettes), they hoped that their health would benefit by replacing combustible cigarettes with HEETS. Smoking goals also contributed to ongoing IQOS use. Although some reported that they had successfully used IQOS to stop smoking combustible cigarettes, few had quit smoking both combustible cigarettes and HEETS altogether, as they used HEETS in place of, or alongside combustible cigarettes.
## Perception of harm
Participants repeatedly reported that they tried IQOS because they believed it was 'better', 'less harmful', 'less hazardous' or 'less damaging' for their health than combustible cigarettes. This view originated from the well-documented harms of combustible cigarettes, claims participants had read on PMI marketing material about IQOS, and the HEETS packets. Participants commented that HEETS packets were 'cleaner' than combustible cigarettes as they had 'less alarming' written health warnings and did not contain graphic pictorial warnings.
I looked into IQOS and it did claim that it had… fewer harmful chemicals than a regular cigarette… That was part of the appeal. And I think that's how it was marketed to me, that it was less harmful. (Alison, 35, ex-IQOS user, former smoking) Believing IQOS was less harmful than smoking combustible cigarettes motivated continued use, often prompting participants to use up to twice as many HEETS as combustible cigarettes, and discouraging quitting IQOS.
Smoking cigarettes, I knew I was doing something really, really bad, and that I should… stop… whereas with this (IQOS), because I don't know that it's so bad for me, there really is no input or drive or motivation for me to think, 'stop!' (Francesca, 46, current IQOS user, former smoking) At the same time, participants said that their mistrust of the tobacco industry and the 'lack' of independent evidence created uncertainty about the potential harms of IQOS. These issues did not discourage participants from trying IQOS but prompted them to seek more information and to consider their ongoing use. A few participants who found reports which contradicted claims of reduced harm debated discontinuing IQOS to return to combustible cigarettes. Participants acknowledged that IQOS was unlikely to be riskfree because HEETS packets carried a warning and they had seen disclaimers on the PMI IQOS website and in IQOS stores. Furthermore, they perceived potential harms from 'inhaling something' into the lungs and from the tobacco, chemical and nicotine composition of HEETS. Nevertheless, they said that without detailed information about the ingredients and nicotine content of HEETS and in absence of longitudinal and independent research into IQOS, they tried to remain optimistic about the harms of IQOS compared with combustible cigarettes. While participants believed that they would re-evaluate their IQOS use if evidence uncovered health risks, some questioned whether knowing the risks would influence their use given that they had not quit smoking combustible cigarettes despite knowing the harms.
## Physical health indicators
Participants often reported that their physical health improved since they started using IQOS-they felt 'fitter' and more energised, had fewer breathing difficulties, noticed increased cardiovascular ability and coughed less. Such improvements reinforced their perceptions that IQOS was less harmful than combustible cigarettes and encouraged continued use. Conversely, no participants reported direct experience of adverse health consequences resulting in second thoughts about using IQOS or discontinuation of IQOS use.
## Financial factors
## Start-up and ongoing costs
Participants regularly expressed concerns that start-up costs of IQOS discouraged initiation. While free demos encouraged them to try IQOS, they felt that IQOS was 'expensive' and too costly compared with cigarettes or e-cigarettes to appeal to lowincome smokers. To overcome financial barriers, some participants received IQOS as a gift, purchased IQOS secondhand, or took advantage of promotions or more affordable prices in mainland Europe. Nevertheless, participants with fewer financial constraints considered that the price of IQOS, although 'not cheap', was 'worth it' if it was 'better' than combustible cigarettes.
I thought it was worth me spending the £80, or whatever it was for the IQOS and a few free packets (of HEETS), even if it turned out not to be something I stuck to. (Raj, 43, current IQOS user, former smoking) There were mixed views across participants on how the cost of IQOS influenced their continued use. A key driver for continuing to use IQOS was that HEETS were cheaper than combustible cigarettes, particularly when purchased in bulk via the PMI website or from countries with lower prices. Yet, participants who had previously smoked rolling tobacco or used e-cigarettes complained that HEETS were less affordable. As such, cost discouraged continued IQOS use for those with limited finances.
## Physical factors enjoyment and satisfaction
Participants gave mixed accounts regarding how enjoyable and satisfying IQOS was. Reflecting on their first experiences, some commented that the similarity of the physical 'kick' of nicotine with smoking combustible cigarettes made IQOS enjoyable to use and easy to switch to. Others were 'underwhelmed' or 'disappointed' by what they described as a 'lighter' and 'less strong' physical feeling but found that they grew accustomed to this with continued use, and rarely discontinued using IQOS because of it.
Generally, current users reported that the inhalation from IQOS felt 'smoother', 'cleaner', 'less harsh' and 'lighter' in their throats than that of combustible cigarettes. They described how this encouraged them to use IQOS and provided a 'pleasant' and 'enjoyable' experience like smoking cigarettes but with fewer negative physical feelings in the throat. Nevertheless, some remained uncertain about using IQOS. For example, they said that the experience was less satisfying than smoking combustible cigarettes on account of the lack of throat hit, the delivery of nicotine and the 'too light' flavour of the tobacco. These more variable experiences contributed to participants discontinuing IQOS or using cigarettes alongside IQOS.
## Original research
I got tired of it after a while… I wasn't getting long-term… satisfaction from it, and I don't think that my craving for cigarettes actually decreased as much as I would have liked it to… there wasn't the same fulfilment. (Hayden, 48, ex-IQOS user, daily smoking)
## Sensory experiences: sight, smell, taste and touch
The influence of IQOS on the senses of sight and smell regularly explained why participants experimented with IQOS. Participants were attracted to try IQOS due to its 'slick' and 'stylish' appearance, discreet size and high-quality finish, which differed to 'bulky', 'huge', 'ugly' tank-style e-cigarettes. The promise that IQOS created little smell also appealed to participants and stood in contrast with the 'horrible', 'disgusting' smell of combustible cigarettes and the 'weird', fruity odours of e-cigarettes.
## (iqos) is a nice piece of kit, very, very high quality, it looks very nice… it's inoffensive, it's not in anybody's faces like… a great big vape kind of battery… it was very smart, very small, very stylish.
(Neal, 47, current IQOS user, former smoking) Participants commonly claimed that the overall sensory experience of using IQOS was equivalent to, or better than smoking combustible cigarettes, which accounted for continued use. They praised aspects of the sight (eg, the visual appeal of the branded HEETS packets, the 'clearer' and reduced volume of emissions, the lack of staining on fingers and teeth, and the cleanliness of no ash), smell (eg, the odourless nature of used HEETS and the lack of residual smell on hands, breath, clothes and furnishings), taste (eg, the comparable taste with combustible cigarettes and the lack of aftertaste) and touch (eg, the tactile feel of the device and the familiar circumference and texture of HEETS in the fingers and on the lips).
## You can use iqos and not feel bad about yourself afterwards in
terms of like the aftertaste, the smell… you just feel nice. (Yulia, 19, current IQOS user, less than monthly smoking) Furthermore, participants who had smoked combustible cigarettes after switching to IQOS described how the 'dirty', 'smelly' and 'disgusting' sensory experiences contrasted with 'cleaner' and 'less messy' use of IQOS. Consequently, participants often envisaged continuing to use IQOS even if the health-related harms were identified as equivalent to, or worse than smoking combustible cigarettes due to the overall 'cleaner' experience.
If there was a headline that says that IQOS was actually more harmful than cigarettes… I probably would still use them… there's less smell, there's no ash, it's cleaner. (Sanjay, 43, current IQOS user, less than monthly smoking) Reports of less positive sensory experiences were limited to not liking the taste of HEETS and noticing an unpleasant smell when IQOS started to heat. Indeed, some described that when they inserted HEETS into the holder and turned it on, it smelt 'burnt' or like 'sewage', 'a fart' or 'manure'. These experiences did not deter initial use of IQOS but sometimes discouraged ongoing use.
## Practical factors accessibility
Until availability of HEETS expanded, early users noted that purchasing HEETS from an IQOS store was inconvenient; sometimes impeding use. Similarly, variable availability and reduced availability outside of London interfered with IQOS use when visiting other cities or when travelling overseas. When faced with issues of availability, participants temporarily discontinued IQOS and smoked combustible cigarettes instead. Although some praised the availability of HEETS online, they noted that bulk purchasing cartons of HEETS encouraged ongoing IQOS use and deterred stopping.
## Use in private and public places
Commonly, participants were encouraged to continue using IQOS as they could use it in places where they would not smoke. Prompted by the reduced smell and appearance of emissions, participants used IQOS in their homes and cars. Confident that they would be undetected, they also used IQOS in public places where smoking and vaping were prohibited, including trains, non-office workplaces, pubs, restaurants, the grounds of a psychiatric hospital and at boarding school. While beneficial on the one hand, participants sometimes disliked that they used HEETS more than combustible cigarettes and as a result, they had less incentive to quit. At the same time, participants expressed uncertainty about the rules governing IQOS use in public, relative to smoking and vaping.
## Maintenance and operation
Participants reported that the need to charge and clean IQOS was burdensome and hindered IQOS use, especially if their device was faulty or broken. Accordingly, they drew parallels between IQOS and e-cigarettes, contrasting them with the simplicity of smoking combustible cigarettes. Although having to re-charge the holder in the pocket charger between HEETS prevented participants from 'chain-smoking' with IQOS, it prompted them to smoke combustible cigarettes when they wanted to chain-smoke.
## Psychological factors ritual and routines
Participants were encouraged to continue to use IQOS as it mirrored their rituals and routines of smoking combustible cigarettes. They drew parallels between the situations when they used IQOS and identified similarities between taking a HEET out of the packet, the hand-to-mouth action of using IQOS, the amount of puffs/time a HEET lasted and the definitive end point of finishing a HEET-all of which they contrasted with using e-cigarettes. Finally, participants discussed how charging and cleaning IQOS led them to develop new habits and rituals. However, participants were divided on the extent to which these new practices helped or hindered continued use of IQOS.
## Trailblazing and trendsetting
Some participants first used IQOS as they were attracted to the 'cool' design and technological appeal. Indeed, individuals recalled their enjoyment of being an 'early adopter' or 'first mover' when few others in the UK knew of or used IQOS. Using IQOS, participants felt 'trendy', 'futuristic' and at the forefront of technological developments in smoking-perceptions that were reinforced when others complimented them on IQOS and when they visited 'high-end', 'Apple-like' IQOS stores. Accordingly, the 'status symbol' of IQOS, and the more 'exclusive' experience it offered, differed from combustible cigarettes and e-cigarettes and inspired participants to pursue IQOS. In opposition, other participants shared worries about the newness of IQOS. They raised anxieties that IQOS may not have been sufficiently tested, questioned whether IQOS fit with their image, worried about attracting unwanted attention and feared the reactions of others. Participants explained that these concerns stemmed from their mistrust of PMI and fears of having fallen 'victim' to their advertising. Although such views led participants to question their use of IQOS, they continued using IQOS.
## Social factors
## Influence of others
Participants reported that family and friends who used IQOS often promoted it to them as an alternative to smoking and urged them to try it. Other participants experimented with IQOS to appease non-smoking partners, or in response to seeing IQOS users in Japan or mainland Europe.
Nevertheless, once participants had switched to IQOS, they were sometimes tempted to smoke combustible cigarettes when among cigarette smokers.
I buy them (combustible cigarettes)… when I go party and especially with friends who smoke… It's more of a social thing… to build the rapport with the person I think it's easier to follow the lead they are doing, instead of trying to break it. (Alexei, 27, current IQOS user, monthly smoking)
## Acceptability
Participants discussed several social consequences of using IQOS which encouraged their ongoing use. First, on account of the reduced visibility, amount, smell and perceived harm of the emissions, participants said that IQOS was 'better' to use around non-smokers than combustible cigarettes or e-cigarettes.
[I] feel less guilty smoking the IQOS around other people because… as far as I know it's not as harmful to secondhand smokers as cigarettes. (Yasmina, 25, current IQOS user, weekly smoking) Furthermore, as using IQOS attracted little attention, participants felt more comfortable using it in public or when in the company of non-smokers. Participants felt less 'vilified' and experienced less stigma and negative judgement with IQOS than when they had smoked or vaped. Others attributed improvements in relationships with romantic partners and work colleagues to the absence of cigarette smoke on their breath, hair, and clothes.
Meanwhile, participants also criticised IQOS for limiting shared social experiences with other smokers. For example, users did not pass and share HEETS or IQOS in the same way as combustible cigarettes and they missed spontaneous conversations sparked by borrowing a lighter. Furthermore, IQOS users occasionally worried about being judged by cigarette smokers due to perceptions that IQOS was only affordable to the wealthier middle classes. These experiences discouraged them from using IQOS in situations where they did not want to be seen as 'different' or where they wanted to bond with cigarette smokers.
# Discussion
As far as we are aware, this is the first independent qualitative study of IQOS users/ex-users outside of Japan and Switzerland, and the first to use in-depth one-to-one interviews. Similar to previous studies, initiation was encouraged by novelty and the sleek appearance of IQOS. [bib_ref] Examining perceptions about IQOS heated tobacco product: consumer studies in Japan and..., Hair [/bib_ref] Participants commonly used IQOS to reduce the harms of smoking, also reflecting earlier findings. [bib_ref] An online survey of users of tobacco vaporizers, reasons and modes of..., Queloz [/bib_ref] Perceptions of reduced harm relative to smoking were underpinned by messages that IQOS emits 95% less chemicals than cigarettes, and were reinforced by the cleanliness of using IQOS, the lighter sensory perceptions, the simple branding of HEETS packets with a text-only warning and the lack of pictorial health warnings. Optimism helped maintain the belief that IQOS was less harmful than smoking, yet participants understood that IQOS was not risk-free. Visual and sensory similarities between HEETS and combustible cigarettes also encouraged continued use, along with the lack of smoke and lingering smell. Yet, the lack of smell also reduced thoughts about stopping IQOS, in contrast to how the lingering smell of combustible cigarette smoke sometimes served as a reminder to quit.
The high initial start-up cost was a barrier to trying IQOS, while the cheaper price of HEETS compared with combustible cigarettes (but not rolling tobacco or e-cigarettes) encouraged continued use. Perceived physical health improvements contributed to continued use, whereas reduced satisfaction deterred continued use. Practical factors, such as availability of HEETS, and wanting to bond in social interactions with cigarette smokers discouraged some situational use but not overall use.
No single factor above all else was important for IQOS use. For example, despite the importance of reduced harm, participants explained that if IQOS turned out be equally or more harmful than combustible cigarettes, they would likely still use IQOS due to other perceived benefits. Furthermore, various factors interacted with one another and influenced use. For example, the ability to use IQOS in more places combined with perceptions of reduced harm led some participants to use more HEETS than combustible cigarettes, despite initial intentions
## Original research
What this paper adds What is already known on this subject ► Smokers in Japan and Switzerland appear to be taking up IQOS use due to several reasons, including lower perceived harm and cleanliness.
What important gaps in knowledge exist on this topic ► Outside of Japan and Switzerland, independent consumer research on IQOS is mostly non-existent, particularly in countries such as the UK where strong tobacco control policies exist and the e-cigarette market is well established.
What this paper adds ► UK smokers and ex-smokers endorsed similar reasons for using IQOS as those in Switzerland and Japan, particularly reduced harm, smell and cleanliness. Perceptions of harm reduction were underpinned by industry pronouncements, the non-combustible process involved, and the appearance of IQOS-related packaging and labelling. Some advantages of IQOS over e-cigarettes were reported such as reduced emissions, smell, and clear start and end point of a bout of use.
to reduce smoking behaviours. In addition, even though practical challenges such as charging deterred use, the same finding emerged in our early qualitative e-cigarette research. [bib_ref] How and why do smokers start using ecigarettes? Qualitative study of vapers..., Wadsworth [/bib_ref] It is worth noting that e-cigarette users/the market adapted to meet these challenges, and IQOS users/manufacturers will likely do the same. Many of the influences on IQOS use were similar to those identified in e-cigarette research-harm reduction encouraged use 32 whereas practical factors discouraged use. [bib_ref] Reasons for regular vaping and for its discontinuation among smokers and recent..., Hua-Hie [/bib_ref] However, unlike previous studies, participants described the reasons that they used IQOS by comparing the benefits of IQOS with the perceived drawbacks of e-cigarettes. Compared with e-cigarettes, IQOS smelt better, was less bulky, felt more like combustible cigarettes in the hands and mouth, and HEETS had an end point.
## Policy and research implications
Our findings on how reduced harm perceptions, the packaging, price and indoor use of IQOS influenced use have important policy and research implications. Research on how tobacco/ nicotine policies, such as health warnings and taxation/price, and industry behaviour interact to influence switching from combustible cigarettes to IQOS is needed. The finding that some participants used up to twice as many HEETS as combustible cigarettes warrants follow-up considering a study comparing HTP and combustible cigarette costs 34 assumed that people smoked equal numbers of combustible cigarettes and heated tobacco sticks. As current smoke-free policies in the UK do not officially cover IQOS, the ability to use IQOS undetected also warrants further research, as does research into the content of IQOS emissions and potential harms to bystanders. Given that the FDA has now authorised IQOS for the US market, our findings may also help inform research into US policy options.
As our study took place in London, our findings may not be generalisable. Yet, many of our findings aligned with studies of IQOS users in Switzerland and Japan, suggesting that they will likely be applicable to other contexts. Given most participants were currently using IQOS, our findings may reflect more positive experiences of use. Although reasons behind individuals' discontinuation of IQOS was uncovered, additional explanations may exist given our small sample of ex-users. In addition, as most of the sample had negative experiences of e-cigarettes, we cannot speculate how those with more positive experiences of e-cigarette use may experience IQOS. Lastly, although the team regularly discussed the coding process, having a single coder could be criticised for potential bias.
Nevertheless, this is the first independent qualitative interview study of IQOS, which was deliberately informed by consultations with experts and smokers/ex-smokers to prioritise key areas of research and policy relevance, and used multiple recruitment methods to access a diverse sample with a range of views and experiences.
# Conclusion
Overall, this study found that a variety of factors influenced IQOS initiation and continued use. The perceived drawbacks of IQOS and the continuing appeal of combustible cigarettes led some people to stop using IQOS altogether, and others to stop situational use of IQOS, smoking combustible cigarettes instead. Several factors that encouraged/discouraged IQOS use may be considered by regulators if they wish to influence use, including packaging, labelling, risk communication, taxation and smokefree policies.
Correction notice This article has been corrected since it was published Online First. Section on disclaimer has now been updated.
## Twitter ann mcneill @kingsnrg
[fig] Figure 1: IQOS 2.4 Plus and HEETS tobacco sticks. [/fig]
[fig] Figure 2: HEETS tobacco sticks and combustible cigarette packaging in the UK. [/fig]
[table] Table 1: Demographic characteristics (n=30) [/table]
|
Spatially resolved TiOx phases in switched RRAM devices using soft X-ray spectromicroscopy
X-ray photoelectron spectroscopy (XPS)XPS was used to characterise the TiOx thin film. All spectra were recorded on a Thermo Scientific Theta Probe Angle-Resolved X-ray Photoelectron Spectrometer (ARXPS) system with a monochromated Al Kα X-ray source (hν = 1486.6 eV). The X-ray source was operated at 6.7 mA emission current and 15 kV anode bias and pass energies of 200 eV and 50 eV were used for survey and core level spectra, respectively. Spectra were corrected for any charge shifts by aligning them to the C 1s core level at 285.0 eV and all data were analysed using the Avantage software package.
## X-ray photoelectron spectroscopy (xps)
XPS was used to characterise the TiOx thin film. All spectra were recorded on a Thermo Scientific Theta Probe Angle-Resolved X-ray Photoelectron Spectrometer (ARXPS) system with a monochromated Al Kα X-ray source (hν = 1486.6 eV). The X-ray source was operated at 6.7 mA emission current and 15 kV anode bias and pass energies of 200 eV and 50 eV were used for survey and core level spectra, respectively. Spectra were corrected for any charge shifts by aligning them to the C 1s core level at 285.0 eV and all data were analysed using the Avantage software package. [bib_ref] In situ control of oxygen vacancies in TiO2 by atomic layer deposition..., Park [/bib_ref] The XPS survey spectrum of the TiOx thin film along with the O 1s and Ti 2p core levels are shown in , S1b and S1c, respectively. The Ti 2p core level shows charge transfer satellites S3/2 and S1/2 at higher binding energies. [bib_ref] Split-Off State Formation in the Final State of Photoemission in Ti Compounds, Okada [/bib_ref] Furthermore, a small population of Ti 3+ of the order of 4 % of the total Ti is observed . The O 1s core level shows a surface oxygen component, often referred to as non-lattice oxygen, on the higher binding energy side of the main core line. core level. (c) Ti 2p core level including higher binding energy satellites S3/2 and S1/2. (d) Fit of the Ti 2p3/2 peak.
## Txm-nexafs
The TXM set-up used in this work presents significant advantages compared with previously used geometries. In particular, it does not require replacement of the silicon wafer support with a fragile standing Si3N4 window [bib_ref] Direct identification of the conducting channels in a functioning memristive device, Strachan [/bib_ref] [bib_ref] Characterization of electroforming-free titanium dioxide memristors, Strachan [/bib_ref] (which could affect the device electrical behaviour due to strain effects and difficulty in sinking the Joule heating 5,6 ) and does not require removal of the top electrode prior to analysis (which is usually performed by a scotch tape method and could lead to unwanted removal of the thin TiO2 layer underneath critical areas 5,7 ), both necessary steps to enable the X-ray transmission if irradiating the device from the top electrode. Most importantly, our geometry allows direct visualization and chemical investigation of the cross-section of the device, as shown in (main manuscript). Ti 2p and O 1s spectra extracted from the TiOx film in PRI device case are shown in and 3d (main manuscript), respectively. The first doublet (2p3/2) (457-462 eV) of Ti 2p spectra originates from transitions to (2p3/2, 3d-t2g) and (2p3/2,3d-eg) states while the second doublet (2p1/2) (462-468 eV) originates from transitions to the corresponding 2p1/2 states. The 2p3/2 -2p1/2 splitting is due to spin-orbit coupling while the t2g-eg separation is the crystal-field splitting due to the surrounding O atoms. [bib_ref] Characterization of surface metallic states in SrTiO3 by means of aberration corrected..., Sánchez-Santolino [/bib_ref] [bib_ref] The effect of valence state and site geometry on Ti L3,2 and..., Stoyanov [/bib_ref] It has to be noted that in all spectra, the (2p3/2 , eg) peak is broader than the (2p3/2, t2g) due to the large degree of hybridization of eg orbitals with O ligand orbitals. [bib_ref] Electronic structure of titania aerogels from soft x-ray absorption spectroscopy, Kucheyev [/bib_ref] Satellite peaks at 470.5 and 476.0 eV due to polaronic transitions are also often observed. [bib_ref] Linear dichroism in ALD layers of TiO2, Das [/bib_ref] [bib_ref] Polaronic satellites in x-ray-absorption spectra, Laan [/bib_ref] The O 1s spectra
## Chemical mapping
X-ray microscopy image sequences can be analysed to provide quantitative maps of chemical components [bib_ref] NEXAFS microscopy and resonant scattering : Composition and orientation probed in real..., Ade [/bib_ref]. In this study maps of three spectrally distinct components -TiOx amorphous, TiOx reduced and a third component similar to TiOx reduced but slightly different at the O 1swere generated by fitting to linear combinations of reference spectra extracted from specific regions of the area measured. If additional TiOx components were added, the component maps from the fit contained large regions with unphysical negative coefficients, indicative of an over-determined fit.
If any one of the three reference spectra were removed from the fit, the residual of the fit increased significantly in the regions of the missing component. [fig_ref] Figure S7 |: NEXAFS Ti 2p [/fig_ref] presents the reference spectra and color coded composite maps for the 3-component fit using stack fit analysis (which also includes a constant to fit the non-Ti, non-O components such as the Pt electrodes) to the separate (Ti 2p or O 1s) and the combined (Ti 2p appended O 1s, aligned) image sequences.
In order to derive quantitative thickness scales for component maps the normal procedure would be to scale the as-extracted reference spectra to the elemental response for a material of the expected elemental stoichiometry and density 15, 16 evaluated from tabulated X-ray absorption data, in which case the grey scales of the derived component maps would give nm thicknesses.
However, because we do not know the composition or density, in this case the edge jumps in background subtracted Ti 2p and O 1s spectra were used to derived amounts of Ti and O based on the scaling to the corresponding edge jumps in pure elemental reference spectra of Ti (Δ480-450eV = 0.0023) and O (Δ 555-428eV =0.0019) at a density of 1.0 g.cm -3 .While this approach has quantitative uncertainties with respect to density, it does have the merit of giving a reliable measure of the O/Ti ratio since the density in any specific compositional region will be constant. This method was used to derive the O/Ti ratios reported in [fig_ref] Table S1: Edge jumps, absolute thickness of OD maps and O/Ti nominal ratio of... [/fig_ref]. All data manipulations were performed using aXis2000.The edge jump values measured from the extracted X-ray absorption spectra have a precision of ~20 %, based on variability of the choice of 'reasonable' background subtraction.
[fig] 2, Figure S1 |: XPS spectra of TiOx. (a) Survey spectrum of TiOx showing all core levels. (b) O 1s [/fig]
[fig] 13, Figure S2 |: main manuscript) can be divided in two regions. The doublet between 528 and 536 eV can be attributed to O 1s excitation to hybrid excited states in which the final level is a mixture of O 2p and Ti 3d orbitals. The spectral features at 531.3 and 533.4 eV are assigned to the t2g and eg orbitals, respectively. 13 This region is very sensitive to local symmetry and coordination. Peaks in the region between 536 eV and 555 eV corresponds to O 1s excited states in which the final level is a hybridization of O 2p and Ti 4sp orbitals. 14 This region is more sensitive to long-range order. Optical images of the device before (a) and after (b) switching into LRS.Morphological defect of TE is circled in green. [/fig]
[fig] Figure S3 |: Average of all images of the Ti stack showing in red the regions containing only Ti species. Scale bar = 200 nm. [/fig]
[fig] Figure S4 |: NEXAFS point Ti 2p (a) and O 1s (b) spectra extracted from the ROI_2, ROI_3, ROI_4, ROI_5 and ROI_6 circled in the X-ray image of Fig. 5c in the main manuscript. [/fig]
[fig] Figure S5 |: Ti 2p and O 1s stacks adjusted to the same spatial area and mesh size corresponding to 4.75 nm pixel size, used for combined stack of Ti 2p and O 1s. The final size for both stack is 291x93 pixels. Scale bar = 200 nm. [/fig]
[fig] Figure S6 |: (a) O 1s spectrum from the SiO2 layer support. (b) Color-coded composition map of selected components: red (SiO2), green (reduced TiOx) and blue (amorphous TiOx). Scale bar = 100 nm. [/fig]
[fig] Figure S7 |: NEXAFS Ti 2p (a) and O 1s (b) spectra of amorphous TiOx (A), reduced TiOx (B) and the Ti containing phase observed in the TE and BE (C). (c-e) Color-coded composition maps of selected components at the Ti 2p and O 1s generated independently (c, d) and combining Ti 2p and O 1s (e). Blue (amorphous TiOx), green (reduced TiOx) and red (phase in Pt electrodes). Scale bar = 50 nm. [/fig]
[fig] Figure S8 |: Alignment of images using cross correlation. Shift in X direction (green line) and Y direction (cyano line) is less than ±0.6 pixels (± 3 nm). [/fig]
[table] Table S1: Edge jumps, absolute thickness of OD maps and O/Ti nominal ratio of amorphous and reduced TiOx and third phase in TE and BE. Error were estimated from reproducibility with small variations in reference spectra. [/table]
|
First molecular and isotopic evidence of millet processing in prehistoric pottery vessels
Analysis of organic residues in pottery vessels has been successful in detecting a range of animal and plant products as indicators of food preparation and consumption in the past. However, the identification of plant remains, especially grain crops in pottery, has proved elusive. Extending the spectrum is highly desirable, not only to strengthen our understanding of the dispersal of crops from centres of domestication but also to determine modes of food processing, artefact function and the culinary significance of the crop. Here, we propose a new approach to identify millet in pottery vessels, a crop that spread throughout much of Eurasia during prehistory following its domestication, most likely in northern China. We report the successful identification of miliacin (olean-18-en-3β-ol methyl ether), a pentacyclic triterpene methyl ether that is enriched in grains of common/broomcorn millet (Panicum miliaceum), in Bronze Age pottery vessels from the Korean Peninsula and northern Europe. The presence of millet is supported by enriched carbon stable isotope values of bulk charred organic matter sampled from pottery vessel surfaces and extracted n-alkanoic acids, consistent with a C 4 plant origin. These data represent the first identification of millet in archaeological ceramic vessels, providing a means to track the introduction, spread and consumption of this important crop.Millet, a small-grained cereal with glumes of hard silica-rich tissue and a short growing season, was an important prehistoric crop that became established across Eurasia. However, the origins and dispersal of domesticated millet and the motivations for its incorporation into cropping regimes are far from clear. The most important species of millet cultivated in Eurasia are common or broomcorn millet (Panicum miliaceum) and foxtail millet (Setaria italica). By 6000 BC, based on charred grain assemblages, millets appear to have been cultivated across several northern Chinese cultures stretching some 1500 km along the Yellow River from the Xinglongwa Culture of the northeast in Manchuria, the Cishan Culture (Hebei Province), the Peiligang Cultures in northern Henan, the Houli Culture in northwest Shandong and in southeast Gangsu with the Dadiwan Culture 1-5 . Earlier dates for domestication at Cishan, based on starch grain and phytolith morphology 6,7 , have been questioned 3,4 . Where human remains are available, stable carbon isotope analysis of bone collagen has proved important in establishing the contribution of millet, a C 4 plant, to diet in the period after domestication. The current isotope data show a mixed picture with millet appearing to range in the diet from a minor food component to a major food resource 8 .From the presumed centre of origin in northern China, both millets are subsequently found over large areas of Eurasia, although the earliest finds are mostly of broomcorn millet 4 . Evidence from charred millet grain and stable isotope analysis of bone collagen from human and faunal remains has shown that broomcorn millet was
in Europe, considered to be found in much earlier contexts, has shown that they date to the second millennium BC or later [bib_ref] The early chronology of broomcorn millet (Panicum miliaceum) in Europe, Motuzaite-Matuzeviciute [/bib_ref]. In Europe, millet cultivation developed much later than the earliest agriculture at the onset of the Neolithic prompting debates that seek to answer the question 'Why move starch into areas where starch was already available?' [bib_ref] Why move starchy cereals? A review of the isotopic evidence for prehistoric..., Lightfoot [/bib_ref] [bib_ref] Food globalization in prehistory, Jones [/bib_ref] [bib_ref] Old World globalization and the Columbian exchange: comparison and contrast, Boivin [/bib_ref].
The dispersal of millet may have been facilitated by mobile hunters, foragers and pastoralists occupying temporary or seasonal settlements since there are clear practical and logistical benefits due to its short growing season and ability to grow in marginal soils with low rainfall. In agrarian societies, millet may have been integrated into flexible crop regimes enabling an extended growing season and a consequent increase in food security [bib_ref] The earliest evidence of millet as a staple crop: new light on..., Liu [/bib_ref] [bib_ref] The Third Food Revolution? Setting the Bronze Age Table: Common Trends in..., Kneisel [/bib_ref] [bib_ref] Food globalization in prehistory, Jones [/bib_ref]. In contrast Boivin and co-workers [bib_ref] Old World globalization and the Columbian exchange: comparison and contrast, Boivin [/bib_ref] emphasize the social factors involved in the adoption of millet cultivation, noting that the perceived value of millet as an exotic crop may have been important in early exchange systems, well before the crop became established as a staple food. However, missing from this debate is any knowledge of how millet was prepared and consumed. This is essential to understand its perceived value and evolving role as a foodstuff.
Evidence from material culture has rarely been deployed in discussions involving the presence and role of specific crops, such as millets, in early societies. Impressions on pottery vessels, thought to be of millet grains, have occasionally been reported. Molecular and isotopic analysis of organic residues associated with pottery containers has developed as a powerful tool for the direct identification of vessel use and insight into diverse cultural practices [bib_ref] From the inside out: Upscaling organic residue analyses of archaeological ceramics, Roffet-Salque [/bib_ref]. A range of plant products, including oils, resins and waxes has been recognised. Detecting residues of grain or seed crops in pottery vessels has proved more challenging. One exception is the presence of the C4 crop maize in vessels from North America utilising a combined lipid biomarker and compound-specific carbon isotope approach [bib_ref] How did Mississippians prepare maize? The application of compound-specific carbon isotope analysis..., Reber [/bib_ref] [bib_ref] Identification of maize in absorbed organic residues: a cautionary tale, Reber [/bib_ref]. Other approaches have been explored in order to document the occurrence of millets in the archaeological record. The most promising is molecular analysis of soils and sediments demonstrating associations with millet cultivation in pit fills or sediment cores [bib_ref] Early millet use in northern China, Yang [/bib_ref] [bib_ref] Millet cultivation history in the French Alps as evidenced by a sedimentary..., Jacob [/bib_ref] [bib_ref] Distribution of miliacin (olean-18-en-3β -ol methyl ether) and related compounds in broomcorn..., Bossard [/bib_ref] [bib_ref] Miliacin in palaeosols from an Early Iron Age in Ukraine reveal in..., Motuzaite-Matuzeviciute [/bib_ref] ; although this approach provides little evidence as to how the crop was processed or consumed.
Here, we propose a new approach for studying millet utilisation in the past, through its identification in direct association with archaeological pottery vessels using a combination of bulk isotope analysis, compound-specific carbon isotope analysis and identification of the pentacyclic triterpene methyl ether (PTME), miliacin. Miliacin is the principal (c. 99%) PTME in broomcorn millet (P. miliaceum) but is absent in many other commonly cultivated species or present in similar abundance to other PTMEs [bib_ref] Distribution of miliacin (olean-18-en-3β -ol methyl ether) and related compounds in broomcorn..., Bossard [/bib_ref]. It is abundant in other Panicum species but occurs with high abundances of amyrin methyl ethers. Miliacin is absent in Setaria italica (foxtail millet) and is not ubiquitous in Pennisetum. The strength of this approach is that only a small number of the C4-grasses of the Panicoideae subfamily synthesize miliacin as their sole PTME [bib_ref] Distribution of miliacin (olean-18-en-3β -ol methyl ether) and related compounds in broomcorn..., Bossard [/bib_ref]. The presence of high abundances of miliacin in sediment profiles indicating the locations of past millet cultivation has already been demonstrated [bib_ref] Millet cultivation history in the French Alps as evidenced by a sedimentary..., Jacob [/bib_ref]. In this study of Alpine lake sediments, P. miliaceum was deemed to be the only possible source of miliacin, given the context and since it was the only PTME detected. High abundances of miliacin have also been reported in Early Iron Age palaesols (eastern Ukraine) from pits comprising broomcorn millet crop processing waste [bib_ref] Miliacin in palaeosols from an Early Iron Age in Ukraine reveal in..., Motuzaite-Matuzeviciute [/bib_ref]. In contrast, a study of pit fills from Neolithic Cishan, China identified miliacin in authentic samples of both P. miliaceum and S. italica and suggested differences between the two species based on the presence of other PTMEs in lower abundance [bib_ref] Early millet use in northern China, Yang [/bib_ref]. The finding of miliacin in S. italica is not supported by investigations of authentic samples [bib_ref] Distribution of miliacin (olean-18-en-3β -ol methyl ether) and related compounds in broomcorn..., Bossard [/bib_ref].
Unlike previous studies we focus on the identification of millet in pottery using molecular and isotopic characterisation. Ceramic vessels are abundant on archaeological sites during the period of millet domestication and dispersal. Prehistoric Eurasian pottery has been studied by archaeologists for centuries and often benefits from well-established chronologies. This approach also provides a means of contextualizing millet within the spheres of culinary practice and artefact use, allowing further and more detailed insight into the cultural significance of the crop.
## Archaeological samples
Ceramics from two prehistoric sites, one in East Asia and one in Europe, were selected for analysis in order to demonstrate the approach [fig_ref] Figure 1: Location of sites investigated in Europe and East Asia [/fig_ref]. These sites were chosen as charred millet grains were identified amongst the botanical remains at both sites.
Bruszczewo, a settlement in Poland some 60 km south of Poznań (52°00′ 47″ N, 16°35′ 07″ E), has been subject to comprehensive investigations [bib_ref] Ausgrabungen und Forschungen in einer prähistorischen Siedlungskammer Großpolens (Dr, Müller [/bib_ref] [bib_ref] The settlement and fortification in the mineral zone of the site (Dr, Czebreszuk [/bib_ref] [bib_ref] Natural resources and economic activities of the Bronze Age people (Dr, Czebreszuk [/bib_ref]. Situated on a spur and close to a lake, two main settlement periods each with several sub-phases can be distinguished: An Early Bronze Age phase (EBA, 2100-1650 cal BC) with fortification, houses and clear settlement activities, and after an apparent gap of several hundred years, occupation dated to the Late Bronze Age/Early Iron Age (LBA/EIA, 1100-800 cal BC). The excavations uncovered several EBA layers and houses preserved in a wet area below the spur at the shore of the lake that date to ca. 1800-1650 BC cal [bib_ref] Natural resources and economic activities of the Bronze Age people (Dr, Czebreszuk [/bib_ref]. Constructed originally on dry ground at the lakeshore during the EBA, the water level rose in the Middle Bronze Age and preserved parts of the site under wet conditions. Archaeobotanical analyses revealed a diachronic differentiation in the occurrence of broomcorn millet. Only single charred grains were recovered from the EBA contexts, but they are numerous in the LBA thus indicating its relevance as food crop in this phase. Furthermore, there are impressions of millet grain in some sherds, particularly on the interior surface of the lower portion of the vessel, perhaps to facilitate the drying process prior to firing. Sixty-one visible charred deposits or 'foodcrusts' adhering to pottery vessel surfaces were sampled. The samples include those from clearly defined and dated EBA and LBA/EIA contexts as well as a mixed transgression layer from the lake. This layer includes significant EBA material and very few younger finds; hence this group is identified as 'mostly EBA' .
The Majeon-ri site (36°09′ 89″ N, 127°08′ 37″ E) is a settlement with a cemetery and paddy field system with wooden structures belonging to the Late Bronze Age and dated to 800-500 cal BC, located in the western region of the Korean peninsula. Fifteen pottery sherds excavated from the irrigation ditch and puddle features at this site were selected for analysis. Unlike the pottery sherds from Bruszczewo, no charred surface deposits were observed.
# Results
## Mobilisation and absorption of miliacin in pottery.
To test whether miliacin is mobilised from millet during processing and absorbed into the ceramic matrix, an experiment was undertaken involving the prolonged cooking of four species of millet; broomcorn millet (Panicum miliaceum), finger millet (Eleusine coracana), barnyard millet (Echinochloa esculenta) and foxtail millet (Setaria italica) in replica ceramic vessels. The results performed on solvent extracts of powdered sherds from these vessels show that a low quantity of miliacin (290 ng g −1 ceramic) is transferred to the pottery vessel wall during cooking of P. miliaceum. No miliacin was detected in the three other millet varieties, confirming previous analysis made directly on the plants [bib_ref] Distribution of miliacin (olean-18-en-3β -ol methyl ether) and related compounds in broomcorn..., Bossard [/bib_ref]. This study reports yields of 297-476 μ g miliacin g −1 P. miliaceum grain following solvent extraction and column chromatography to isolate the neutral fraction [bib_ref] Distribution of miliacin (olean-18-en-3β -ol methyl ether) and related compounds in broomcorn..., Bossard [/bib_ref]. Much lower levels of miliacin were found in the leaves, roots and stems of P. miliaceum and the compound was found to be absent in the hulls. It is therefore likely that only a small proportion of miliacin is released and transferred to the ceramic matrix during cooking of millet grain.
Bulk carbon and nitrogen isotope analysis of Bruszczewo charred surface deposits. The presence of charred surface deposits at Bruszczewo permitted the application of elemental analysis isotope ratio mass spectrometry (EA-IRMS) to determine bulk carbon (δ 13 C) and nitrogen (δ 15 N) isotope ratios. [fig_ref] Figure 2: Plot of δ 15 N versus δ 13 C values for foodcrusts... [/fig_ref] shows a marked change in carbon isotope ratios between the EBA (and 'mostly EBA') and LBA/EIA phases with indications in the latter of a wider range of foods including a possible C 4 plant input (see also . The EBA samples form a clustered group considered to represent C 3 plants and animals consuming C 3 plants, relatively depleted in 13 C. In contrast the LBA/EIA samples plot over a much wider range of carbon isotope values with 45% of the samples having δ 13 C values greater than − 20.0‰, consistent with the input of 13 C-enriched C 4 plants such as millet. Any input from marine foods is highly unlikely as a result of the distance from the coast, absence of marine fish bone in the faunal assemblage, the low δ 15 N values and the high C/N ratios. The difference in carbon isotope ratios between the EBA and LBA/EIA samples is significant (heteroscedastic two-tailed t-test, p = < 0.05). The δ 15 N values are not significantly different. The low δ 15 N values and high C:N values in the majority of samples suggest the presence of low trophic level foods expected from the charring of plant tissues in the pots [bib_ref] Aquatic resources in foodcrusts: identification and implication, Heron [/bib_ref].
In general, the LBA/EIA foodcrusts from Bruszczewo have lower δ 13 C values when compared with direct measurements on charred millet grains suggesting that mixtures of C 3 and C 4 resources were processed in some of the vessels. For example, bulk isotope data (δ 13 C = − 12.5‰; δ 15 N = 4.7‰, [fig_ref] Figure 2: Plot of δ 15 N versus δ 13 C values for foodcrusts... [/fig_ref] was obtained on a deposit comprising carbonized millet grains adhering to a potsherd from Ryugasaki A (Shiga Prefecture, Japan) and radiocarbon dated to 800-555 cal BC [bib_ref] Oldest common millet from western Japan: Carbonized material adhering to pottery from..., Miyata [/bib_ref]. Bulk carbon isotope analysis of charred broomcorn and foxtail millet grain from archaeological sites in northern China gave mean values of − 11.1 ± 0.5‰ (n = 26) and − 10.5 ± 0.6‰ (n = 27) respectively [bib_ref] Variability of the stable carbon isotope ratio in modern and archaeological millets:..., An [/bib_ref]. Direct 14 C dates (ca. 1500 BC to 500 AD) determined on charred P. miliaceum grains from central and southeastern Europe yielded δ 13 C values from − 9.6 to − 11.2‰ 20 . Data from cooking experiments in pottery vessels show relatively minor (ca. 1‰) shifts in bulk isotope values of foods during cooking, including millet grain [bib_ref] Dating and stable isotope analysis of charred residues on the Incipient Jomon..., Yoshida [/bib_ref] and this is consistent with heating experiments undertaken directly on cereal grains [bib_ref] Assessing natural variation and the effects of charring, burial and pre-treatment on..., Fraser [/bib_ref]. These data are helpful when evaluating bulk isotope data of 'foodcrusts' transformed by heating and suggest that changes due to cooking are unlikely to impact on the differences in carbon isotope values between C 3 and C 4 plants. Assessing the proportion of C 3 and C 4 (maize) foods incorporated into charred deposits has been attempted but cooking experiments and modelling of the data highlights caution in interpreting bulk carbon isotope data in this way Molecular characterisation by gas chromatography-mass spectrometry (GC-MS). Two of the 'foodcrusts' from Bruszczewo, one EBA and one LBA/EIA, were selected for GC-MS and GC-c-IRMS . They were chosen to represent widely separated carbon isotope values (− 26.0‰ for the EBA sample and − 11.8‰ for the LBA/EIA sample). The GC-MS data are summarised in . Despite markedly different bulk carbon isotope ratios, the lipid composition of the samples is very similar. Both residues are characterized by a narrow range of n-alkanoic acids (C 14:0 -C 18:0 ) with a high C 16:0 over C 18:0 abundance. The presence of C 16 and C 18 ω -(o-alkylphenyl)alkanoic acids was detected in both samples. The C 18 analogues are clearly visible in the total ion current (TIC) chromatogram of the LBA/EIA sample whereas in the EBA sample they were only detected in the extracted ion chromatograms (m/z 105 and 290). These ω -(o-alkylphenyl)alkanoic acids are thermal alteration products of unsaturated n-alkanoic acids [bib_ref] Thermally produced ω -(o-alkylphenyl)alkanoic acids provide evidence for the processing of marine..., Hansel [/bib_ref]. The distribution of the C 18 acids is consistent with those of authentic heated plant oils and not consistent with the profile given by a terrestrial animal fat and aquatic oil [bib_ref] Experimental evidence for the processing of marine animal products and other commodities..., Evershed [/bib_ref]. No trace of C 20 and C 22 analogues, commonly associated with the thermal alteration of longer-chain n-alkenoic acids in aquatic oils 41 , was detected. In summary, the lipid extracts are consistent with a plant origin but further differentiation is not possible. Epicuticular waxes, phytosterols and aquatic biomarkers are absent. Miliacin was detected by GC-MS operating in SIM mode in the LBA/EIA sample, suggesting the presence of broomcorn millet, but absent in the EBA sample.
Lipids were detected in all of the samples from Majeon-ri (n = 15). Separate solvent and acid/methanol extracts are dominated by saturated n-alkanoic acids . A range of other compounds were detected in the residues including n-alkanols, ketones and sterols/stanols. The presence of levoglucosan in MJR03 supports the cooking of plant tissues since this molecule is formed from the pyrolysis of starch and cellulose [bib_ref] Levoglucosan, a tracer for cellulose in biomass burning and atmospheric particles, Simoneit [/bib_ref]. In addition, C 18 ω -(o-alkylphenyl)alkanoic acids were detected in two samples, MJR06 and 08. Miliacin was identified in seven of the 15 sherds from Majeon-ri. In five of these cases, the molecule was identified in both acidified methanol and solvent extracts obtained on samples taken from the same sherd. In the remaining . Bulk isotope data obtained on one EBA and one LBA/EIA sample selected for GC-MS and GC-C-IRMS.
two sherds, miliacin was only identified in the acidified methanol extract, presumably due to the higher recovery of lipid using this extraction methodology. [fig_ref] Figure 3: Identification of miliacin [/fig_ref] shows the presence of miliacin in extracts from sample (MJR10). The mass spectrum [fig_ref] Figure 3: Identification of miliacin [/fig_ref] is indicative of a pentacyclic triterpene methyl ether (M + m/z 440; base peak m/z 189). The retention time of this peak matched an authentic sample of miliacin [fig_ref] Figure 3: Identification of miliacin [/fig_ref] as well as the peak corresponding to miliacin released experimentally by boiling broomcorn millet [fig_ref] Figure 3: Identification of miliacin [/fig_ref].
## Compound-specific carbon isotope analysis of n-alkanoic acids by gc-combustion-isotope ratio mass spectrometry (gc-c-irms). the most frequent compounds usually encountered in archaeologi-
cal cooking vessels are n-alkanoic acids (fatty acids). However, C 3 and C 4 plants cannot be distinguished solely on their n-alkanoic acid distributions [bib_ref] How did Mississippians prepare maize? The application of compound-specific carbon isotope analysis..., Reber [/bib_ref]. In view of this, the compound-specific carbon isotope compositions of the most abundant n-alkanoic acids in the lipid extracts were determined by GC-c-IRMS. presents the data obtained on the two samples from Bruszczewo compared with the bulk carbon isotope data. As expected, the lipids are more depleted in 13 C than the bulk samples, which may contain degraded proteins and carbohydrates relatively enriched in 13 C. However, there is a marked difference in the offset between the bulk and compound-specific carbon isotope values (Δ BULK-CSIA ) of the two samples. The magnitude of the offsets fall within the range determined from modern bulk and compound-specific carbon isotope data from C 3 plants (i.e., 3-9‰) and C 4 plants (i.e., 7-14‰) respectively [bib_ref] Variability of stable carbon isotopic compositions in individual fatty acids from combustion..., Ballentine [/bib_ref]. These data, together with the lipid molecular information, suggest that the residues are dominated by a single foodstuff -a C 3 and a C 4 plant (millet) respectively. Compound-specific carbon isotope analysis was also undertaken on seven lipid extracts from Majeon-ri where sufficient quantities of n-alkanoic acids were preserved (Supplementary . Miliacin had been identified by GC-MS in four of these samples. Three samples (MJR06, MJR09 and MJR10) contained n-alkanoic acids relatively enriched in 13 C in keeping with the processing of C 4 plants such as broomcorn and foxtail millet. However, miliacin was identified in only two of these, perhaps reflecting differential preservation or that another species of millet, such as foxtail millet, was predominantly processed in this vessel. Conversely, despite the presence of miliacin, two of the samples (MJR02, MJR08) have more mixed fatty acid δ 13 C values indicating a combination of C 3 and C 4 resources.
# Discussion
The presence of miliacin as the only PTME in the ceramic lipid extracts strongly suggests the presence of broomcorn millet processed in pottery vessels from the Korean and European Bronze Age. This represents the first molecular identification of this plant in archaeological pottery vessels opening up the potential for further research. At Bruszczewo, the high δ 13 C and low δ 15 N values of the visible surface deposits and the larger offset between the bulk and compound-specific carbon isotope values are consistent with C 4 plant signals. The presence of charred deposits and thermally altered lipids also implies that millet was heated in the preparation of the foodstuff for consumption. The presence of millet in the LBA/EIA samples is supported by archaeobotanical evidence from the site. A few charred millet grains have been recovered from EBA layers, which could point to small-scale cultivation whereas abundant charred millet grains have been recovered in LBA/EIA contexts. However, caution is required given the possibility of movement of small grains through the soil column [bib_ref] The early chronology of broomcorn millet (Panicum miliaceum) in Europe, Motuzaite-Matuzeviciute [/bib_ref]. The archaeobotanical evidence suggests that millet became a widespread and important crop during the LBA/EIA of Eastern, South-eastern, Central and northern Central Europe [bib_ref] Vegetationsgeschichtliche und archäobotanische Untersuchungen zur Landwirtschaft und Umwelt im Bereich der prähistorischen..., Kirleis [/bib_ref].
At Majeon-ri, the occurrence of miliacin in the majority of the sherds analysed (Supplementary indicates that broomcorn millet was regularly processed in these vessels. It has been suggested that human diet in the Korean Bronze Age was largely based on C 4 plants. For example, carbon isotopic analysis of human bone collagen from the contemporary Konam-ri site on the west coast of Korea produced values consistent with C 4 plant consumption [bib_ref] Dietary reconstruction by stable isotopic analysis: the Konam-ri shell midden in Korea, An [/bib_ref]. However, consumption of marine resources, also enriched in 13 C, complicates this interpretation. Whilst the identification of millet in pottery does not allow dietary quantification, it does demonstrate that millet was directly consumed in Bronze Age Korea. Further insights into how millet was consumed are also provided by the analysis. For example, variability in δ 13 C values of alkanoic acids extracted from the vessels containing miliacin shows that millet was commonly mixed, or sequentially processed, with other food products. This may indicate it was not treated . Compositional data, bulk and compound-specific δ 13 C values for one probable C 3 and one probable C 4 sample from Bruszczewo. Both samples also yielded low levels of n-alkenoic acids, odd-carbon number n-alkanoic acids and dicarboxylic acids. The Δ BULK -CSIA value expresses the difference between the bulk carbon isotope value and the mean of the compound-specific carbon isotope values obtained on the C 16:0 and C 18:0 n-alkanoic acids. These data are compared with depletions in n-alkanoic acids extracted from authentic C 3 and C 4 plants relative to bulk tissue values using literature data [bib_ref] Variability of stable carbon isotopic compositions in individual fatty acids from combustion..., Ballentine [/bib_ref]. Compound-specific carbon isotope data have been reported for miliacin extracted from lake sediments and compared with authentic samples of P. miliaceum specially or, at least, afforded dedicated ceramic vessels for its preparation. The presence of thermally-altered plant derived compounds in the potsherds, such as C 18 ω -(o-alkylphenyl)alkanoic acids and levoglucosan shows that millet was heated during its preparation. Millet is often used as gruels and porridge or ground into flour by contemporary societies [bib_ref] Millet, the late comer: on the tracks of Panicum miliaceum in prehistoric..., Valamoti [/bib_ref]. Another potential role for millet is as a base for fermented beverages. For example, Boza is a slightly alcoholic (1.5%), sour drink made without any additional malt prepared by grinding the millet grains, mixing with water, cooking and fermentation [bib_ref] An ethnobotanical perspective on traditional fermented plant foods and beverages in Eastern..., Sõukand [/bib_ref]. Distinguishing a gruel or porridge from a cooked and fermented millet-based drink is not possible from the data presented here but further experimental work, consideration of vessel typologies and analysis of vessels from ethnographic contexts may help elucidate these practices in the past.
The identification of millet associated with pottery provides a new opportunity to examine the history of this important crop. Firstly, this approach has the advantage of identifying direct consumption of millet, a distinction that is not afforded by most archaeobotanical finds and particularly pertinent since millet is often considered to be an important fodder crop [bib_ref] Millet, the late comer: on the tracks of Panicum miliaceum in prehistoric..., Valamoti [/bib_ref]. Secondly, the identification of millet in pottery provides unique information regarding its mode of preparation and consumption, which may be crucial for understanding the role and value of the crop when first introduced to a region [bib_ref] Food globalization in prehistory, Jones [/bib_ref] [bib_ref] Old World globalization and the Columbian exchange: comparison and contrast, Boivin [/bib_ref]. Given that the wild ancestor of broomcorn millet has not yet been identified, pottery residue analysis could help to identify the first intensive use of this plant. Thirdly, the abundance of pottery already recovered from archaeological sites, and held in collections, provides an easily accessible resource for undertaking a wider survey of the introduction and spread of millet cultivation and the role it played in food processing and diet across Eurasia. Methods Production of experimental vessels. Experimental vessels (h. 30 cm, d. 28 cm, 7 mm thickness) were formed by coiling, and fired at a temperature of 1,000 °C using an electric kiln. These vessels were used to cook four varieties of millet; broomcorn millet (Panicum miliaceum), finger millet (Eleusine coracana), barnyard millet (Echinochloa esculenta) and foxtail millet (Setaria italica). The seeds were either purchased or collected in Japan. After filling the lower part of the pot with the seeds, water (ca. 900 ml), was added up to the inflection point below the rim. The pot was then placed on a fire made with broadleaf wood fuel until visible carbonised deposits had formed. The cooking continued for up to one hour. The temperature measured on the outer surface of the vessel often exceeded 300 °C. After the experiment the vessels were dried at room temperature. Then 1 g of ceramic was drilled from the ceramic interior at a depth between 2-5 mm.
Lipid extraction. Lipids were extracted and methylated simultaneously using acidified methanol [bib_ref] High throughput screening of organic residues in archaeological potsherds using direct acidified..., Correa-Ascencio [/bib_ref]. Briefly, methanol was added to homogenized charred deposits (1 mL to 10-30 mg) and ceramic powders drilled (d. 2-5 mm) from the sherd surface (4 mL to 1 g). The mixture was ultrasonicated for 15 min, and then acidified with concentrated sulphuric acid (200 μ L). The acidified suspension was heated in sealed tubes for 4 h at 70 °C and then allowed to cool. Lipids were then extracted with n-hexane (3 × 2 mL) and dried under a gentle stream of N 2 . 10 μ g of internal standard (n-hexatriacontane) was added to each sample prior to analysis by GC-MS and GC-C-IRMS using standard conditions and protocols [bib_ref] Direct chemical evidence for widespread dairying in prehistoric Britain, Copley [/bib_ref] [bib_ref] Molecular and isotopic demonstration of the processing of aquatic products in Northern..., Craig [/bib_ref]. In addition, the sherds from Majeon-ri were subjected to conventional solvent extraction following established methods 52 to compare the recovery of miliacin between acidified methanol and solvent extraction. Sherd powders (1 g) were extracted by ultrasonication into a mixture of DCM:MeOH (2/1 v/v, 3 × 2 mL) to obtain the extracts which were combined and dried under a gentle stream of N 2 , followed by silylation with N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) with 1% trimethylchlorosilane (TMCS) at 70 °C for one hour, and dried again under N 2 . n-hexane was added to the derivatised samples prior to gas chromatography-mass spectrometry (GC-MS).
## Gas chromatography-mass spectrometry (gc-ms). gc-ms was carried out on all samples using an
Agilent 7890 A Series chromatograph attached to an Agilent 5975 C Inert XL mass-selective detector with a quadrupole mass analyser (Agilent technologies, Cheadle, Cheshire, UK). A splitless injector was used and maintained at 300 °C. The carrier gas used was helium, and inlet/column head-pressure was constant. The column (DB-5 ms) was coated with 5% phenyl-methylpolysiloxane column (30 m × 0.250 mm × 0.25 μ m; J&W Scientific, Folsom, CA, USA). The oven temperature was set at 50 °C for 2 min, then raised by 10 °C min −1 until 325 °C was reached, where it was held for 15 min until the end of the run. The GC column was inserted directly into the ion source of the mass spectrometer. The ionisation energy of the mass spectrometer was 70 eV and spectra were obtained in scanning mode between m/z 50 and 800. The MS was also used in selected ion monitoring (SIM) mode with the oven temperature set at 50 °C for 1 min, then raised by 20 °C min −1 until 280 °C, then raised at 10 °C min −1 until reaching 325 °C, where it was held for 10 min until the end of the run. In SIM mode, a first group of ions (m/z 189, m/z 204, m/z 231, m/z 425, m/z 440) corresponding to miliacin fragmentation were monitored. After 16 min, a second group of ions (m/z 57, m/z 71, m/z 85, m/z 478, m/z 506) were monitored to record the internal standard.
## Gas chromatography-combustion-isotope ratio mass spectrometry (gc-c-irms).
In each extract, compound specific stable carbon isotope analyses (GC-c-IRMS) of methyl palmitate (C 16:0 ) and methyl stearate (C 18:0 ) were conducted according to previous methodologies 53 . An Isoprime 100 (Isoprime, Cheadle, UK) linked to a Hewlett Packard 7890B series gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) with an Isoprime GC5 interface (Isoprime, Cheadle, UK) was used. All samples were diluted with hexane and subsequently 1 μ L of each sample was injected into a DB-5MS ultra-inert fused-silica column. The temperature was set for 0.5 min at 50 °C, and raised by 10 °C min-1 until 300 °C was reached, and held for 10 min. The carrier gas was ultra-high purity grade helium with a flow rate of 3 mL min −1 . The gases eluting from the chromatographic column were split into two streams. One of these was directed into an Agilent 5975C inert mass spectrometer detector (MSD), for sample identification and quantification, while the other was directed through the GC5 furnace held at 850 °C to oxidise all carbon species to CO 2 . A clear resolution and baseline separation of the analysed peaks was achieved. Eluted products were ionized in the mass spectrometer by electron impact. Ion intensities of m/z 44, 45, and 46 were monitored in order to automatically compute the 13 C/ 12 C ratio of each peak in the extracts. Computations were made with IonVantage and IonOS Softwares (Isoprime, Cheadle, UK) and were based on comparisons with a standard reference gas (CO 2 ) of known isotopic composition that was repeatedly measured. The results from the analysis are reported in parts per mil (‰) relative to an international standard (V-PDB). Replicate measurements of each sample and a mixture of fatty acid methyl esters (FAMEs) with δ 13 C values traceable to international standards were used to determine instrument precision (± 0.3‰) and accuracy (± 0.5‰) standard deviation. Finally a standard mixture of C 16:0 and C 18:0 fatty acids of known isotopic composition processed in each batch under identical conditions was used to correct for the methylation of the carboxyl group that occurs during acid extraction.
Elemental Analysis Isotope Ratio Mass Spectrometry. About 1 mg of crushed and homogenised charred deposit was weighed into tin capsules in duplicate. These were directly analysed by elemental analysisisotope ratio mass spectrometry (EA-IRMS) to determine their bulk stable carbon (δ 13 C) and nitrogen isotope (δ 15 N) composition, as previously reported [bib_ref] Variability of stable carbon isotopic compositions in individual fatty acids from combustion..., Ballentine [/bib_ref]. Samples yielding less than 1% N were not used and instrument precision on repeated measurements was ± 0.2‰ (s.e.m.). δ 13 C, δ 15 N = [(R sample /R standard − 1)] × 1,000, where R = 13 C/ 12 C and 15 N/ 14 N. All sample measurements are expressed in per mil relative to the δ 13 C standard (Vienna PeeDee Belemnite, V-PDB) and the δ 15 N standard (air N 2 ) respectively.
[fig] Figure 1: Location of sites investigated in Europe and East Asia (Adobe Illustrator CS4 14.0.0. http://www. adobe.com/uk/). [/fig]
[fig] Figure 2: Plot of δ 15 N versus δ 13 C values for foodcrusts adhering to the interior walls of pottery vessels from Bruszczewo. The result from a single foodcrust containing carbonized millet grain from Japan (Ryugasaki) is also plotted for comparison37 . [/fig]
[fig] EBA: (n = 24) 'Mostly ' (n = 15) LBA/EIA (n = 22) δ 13 C (‰) δ 15 N (‰) δ 13 C (‰) δ 15 N (‰) δ 13 C (‰) δ 15 N (‰) [/fig]
[fig] Figure 3: Identification of miliacin (olean-18-en-3β -ol ME) by GC-MS, (a) Partial total ion current (TIC) chromatogram of the archaeological sample from MJR10 (Majeon-ri). (b) Partial TIC chromatogram of the modern broomcorn millet cooking experimental sample. (c) Partial TIC chromatogram of the authentic miliacin. (d) Mass spectrum of miliacin. (e) Chemical structure of miliacin. Peak identities -M: miliacin, *: stigmastanol (trimethylsilyl ether) IS: internal standard (n-hexatriacontane). Scientific RepoRts | 6:38767 | DOI: 10.1038/srep38767 [/fig]
|
Efficacy and Safety of HSK3486 for Anesthesia/Sedation in Patients Undergoing Fiberoptic Bronchoscopy: A Multicenter, Double-Blind, Propofol-Controlled, Randomized, Phase 3 Study
Background Fiberoptic bronchoscopy is a complex procedure with the need for sufficient patient anesthesia/sedation while maintaining safety. This trial aimed to evaluate the efficacy, safety, and pharmacokinetics of HSK3486 during fiberoptic bronchoscopy. Methods This multicenter, double-blind, randomized, non-inferiority, parallel-group phase 3 trial was conducted in patients who underwent fiberoptic bronchoscopy. Patients randomly received HSK3486 0.4 mg/kg (N = 134) or propofol 2.0 mg/kg (N = 133). The primary efficacy endpoint was the successful rate of fiberoptic bronchoscopy, and secondary efficacy endpoints included successful induction of anesthesia/sedation, duration, time to being fully alert, and time to patient discharge. Safety assessments and drug concentrations were also measured. Results A total of 267 patients completed fiberoptic bronchoscopy, with a success rate of 100% and a 95% confidence interval of − 2.8 to 2.8% for the difference between the groups, which met the predesigned criteria of > − 8%, confirming the noninferiority of anesthesia/sedation produced by HSK3486 compared to propofol. Among the secondary efficacy endpoints, only time to full alertness (median 8.50 vs. 6.00 min, P = 0.012) and time to discharge (median 13.00 vs. 9.87 min, P = 0.002) were slightly longer in the HSK3486 group. The incidence of adverse events was significant lower in the HSK3486 group (52.6 vs. 76.5%, P < 0.001) mainly because of less pain on injection (4.4 vs. 39.4%, P < 0.001) compared to the propofol group. HSK3486 had a similar terminal elimination half-life as propofol. Conclusions HSK3486 exhibited non-inferiority anesthesia/sedation compared to propofol in patients undergoing fiberoptic bronchoscopy, and had a good safety profile with a lower incidence of pain on injection.
# Introduction
Fiberoptic bronchoscopy has gradually become an important diagnostic and therapeutic procedure for the treatment of respiratory diseases, which has been enhanced by the use of a laryngeal mask airway (LMA) method that improves airway support and stable oxygen saturation, with a convenient port of entry during fiberoptic bronchoscopy [bib_ref] The safety of laryngeal mask airwayassisted bronchoscopy versus standard nasal bronchoscopy, Alon [/bib_ref]. Sedation was initially rarely used in fiberoptic bronchoscopy because there was a concern about adverse events (AEs) [bib_ref] American College of Chest Physicians consensus statement on the use of topical..., Wahidi [/bib_ref]. As this procedure can produce great discomfort in patients and elicit
## Key points
This multicenter, double-blind, randomized, noninferiority, parallel-group phase 3 trial was conducted to evaluate the efficacy, safety, and pharmacokinetics of HSK3486 in patients undergoing fiberoptic bronchoscopy.
Patients who received HSK3486 completed fiberoptic bronchoscopy with a success rate of 100% and showed non-inferiority anesthesia/sedation characteristics compared to propofol during fiberoptic bronchoscopy.
HSK3486 exhibited a good safety profile with a lower incidence of pain on injection. a high incidence of hypoxia, sedation has gradually been developed for painless bronchoscopy [bib_ref] Sedation for flexible bronchoscopy: current and emerging evidence, José [/bib_ref] [bib_ref] Bronchoscopy: the past, the present, and the future, Becker [/bib_ref] [bib_ref] British Thoracic Society guideline for diagnostic flexible bronchoscopy in adults: accredited by..., Rand [/bib_ref]. Subsequent studies have shown that sedation leads to better tolerance [bib_ref] Should patients undergoing a bronchoscopy be sedated?, Gonzalez [/bib_ref] and higher physician satisfaction for the procedure [bib_ref] Does sedation help in fibreoptic bronchoscopy?, Hatton [/bib_ref]. Therefore, to ensure the safety and comfort of patients while meeting the operational requirements of sharing the airway is a major challenge for the successful implementation of painless fiberoptic bronchoscopy [bib_ref] British Thoracic Society guideline for diagnostic flexible bronchoscopy in adults: accredited by..., Rand [/bib_ref] [bib_ref] Anesthesia for advanced bronchoscopic procedures: state-of-the-art review, Goudra [/bib_ref].
The British Thoracic Society states that patients undergoing fiberoptic bronchoscopy should be sedated when there are no contraindicationsto facilitate patient comfort and alleviate anxiety, coughing, and dyspnea, while reducing the complications of the procedure [bib_ref] Should patients undergoing a bronchoscopy be sedated?, Gonzalez [/bib_ref] [bib_ref] Patient satisfaction with conscious sedation for bronchoscopy, Putinati [/bib_ref] [bib_ref] Sedation in outpatient bronchoscopy, Matot [/bib_ref]. Previous studies have shown that propofol is often used to induce moderate sedation or general anesthesia for a fiberoptic bronchoscopy procedure because of its rapid onset, fast recovery, high clearance rate, and good tolerance [bib_ref] Titrated sedation with propofol or midazolam for flexible bronchoscopy: a randomised trial, Clark [/bib_ref] [bib_ref] Feasibility of bispectral index-guided propofol infusion for flexible bronchoscopy sedation: a randomized..., Lo [/bib_ref] [bib_ref] Propofol versus propofol plus hydrocodone for flexible bronchoscopy: a randomised study, Schlatter [/bib_ref]. There is improved patient perception of sedation and procedure tolerance, and an overall reduction in cough and the sensation of asphyxiation when propofol is administered [bib_ref] Should patients undergoing a bronchoscopy be sedated?, Gonzalez [/bib_ref]. Propofol used as monotherapy for sedation does not result in analgesia, so it is often combined with opiates or local anesthetics to improve patient tolerance [bib_ref] Propofol versus propofol plus hydrocodone for flexible bronchoscopy: a randomised study, Schlatter [/bib_ref] [bib_ref] The effect of mixing lidocaine with propofol on the dose of propofol..., Tan [/bib_ref] , However, propofol has a number of known limitations [bib_ref] Pain on propofol injection: causes and remedies, Desousa [/bib_ref] [bib_ref] Propofol: therapeutic indications and side-effects, Marik [/bib_ref] , such as a narrow therapeutic window, high incidence of hypotension, respiratory depression, pain on injection, and a lack of availability of antagonists.
HSK3486 (Trade name: ciprofol, Haiso Pharmaceutical Group Co., Ltd, Chengdu, China), a novel 2,6-disubstituted phenol derivative, binds more effectively to the gammaaminobutyric acid-A (GABA A ) receptor than does propofol [bib_ref] Design, synthesis, and evaluation of novel 2,6-disubstituted phenol derivatives as general anesthetics, Qin [/bib_ref]. After initial evaluations of absorption, distribution, metabolism, and excretion (ADME) processes [bib_ref] Mass balance, pharmacokinetics and pharmacodynamics of intravenous HSK3486, a novel anaesthetic, administered..., Bian [/bib_ref] , the maximum concentration (C max ), time to C max (t max ), terminal elimination half-life (t 1/2 ), and mean residence time (MRT) were similar between HSK3486 and propofol, while clearance (CL), apparent volume of distribution (V d ) and apparent volume of distribution at steady state (V ss ) were different in a phase 1 trial [bib_ref] Sedation effects produced by a ciprofol initial infusion or bolus dose followed..., Hu [/bib_ref]. In addition, a study on age-related effects of HSK3486 revealed that a dose of 0.3 mg/kg was similarly efficacious in the elderly compared with 0.4 mg/kg in nonelderly patients [bib_ref] Safety, pharmacokinetics, and pharmacodynamics of a single bolus of the γ-aminobutyric acid..., Li [/bib_ref].
Phase 2 [bib_ref] Efficacy and safety of ciprofol for the sedation/anesthesia in patients undergoing colonoscopy:..., Teng [/bib_ref] and phase 3 trials (clinicaltrials.gov, NCT03674008, on submission) for gastrointestinal endoscopy procedures demonstrated the efficacy and safety of HSK3486 in patients undergoing endoscopy. The duration of anesthesia/sedation in patients undergoing fiberoptic bronchoscopy was similar to that of gastrointestinal endoscopy, which is generally 10-30 min and the drug administration methods are also similar, which is an initial intravenous (i.v.) bolus injection followed by top-up doses if required. Therefore, this multicenter, double-blind, propofol-controlled, randomized, phase 3 trial was carried out to evaluate the efficacy and safety of HSK3486 in patients undergoing fiberoptic bronchoscopy.
# Material and methods
## Study design and procedure
This multicenter, double-blind, randomized, propofol-controlled, non-inferiority prospective phase 3 trial was conducted in ten study centers in China between 6 December 2019 and 19 June 2020 in accordance with the Declaration of Helsinki. The study was approved by the ethics committee of each participating hospital. The trial was prospectively registered at clinicaltrials.gov (NCT04111159). Written informed consent was obtained from all patients before they were enrolled.
The intention of the trial was to enroll consecutively 260 eligible patients from each study center and randomly allocate them to a HSK3486 0.4 mg/kg group or a propofol 2.0 mg/kg group in a ratio of 1:1. The duration of the trial included the screening period (within 7 days before fiberoptic bronchoscopy), baseline period (the same day of fiberoptic bronchoscopy), the observation period drug postadministration and the follow-up post-procedure period (within 2-4 days post fiberoptic bronchoscopy).
The patients enrolled were aged between 18 and 80 years, male or female, with American Society of Anesthesiologists (ASA) grade I-III, and were to receive diagnostic and/or therapeutic fiberoptic bronchoscopy LMA assisted. Exclusion criteria included patients allergic to egg and soybean products, propofol, opioids, and their antidotes. Patients with contraindications to deep sedation/general anesthesia, or with a history of previous anesthesia/sedation accidents, or who had undergone endotracheal intubation and/ or mechanical ventilation prior to fiberoptic bronchoscopy, were also excluded. Detailed inclusion and exclusion criteria are presented in the Electronic Supplementary Material (ESM) [fig_ref] Table 1: Demographic and baseline characteristics of patients [/fig_ref].
Patients had baseline inclusion/exclusion criteria, weight, vital signs, blood samples, pregnancy tests (urine human chorionic gonadotropin for child-bearing aged female patients), 12-lead electrocardiogram (ECG), and Modified Observer Assessment of Alertness/Sedation (MOAA/S) scores measured on the day of the procedure. Eligible patients inhaled 2% lidocaine (10 mL) delivered by atomization within 1 h before the bolus doses of HSK3486 or propofol were i.v. injected. Supplementary oxygen was administered at a flow rate of > 6 L/min at least 3 min before the first study dose was administered and until patients were fully alert (the first time of MOAA/S = 5 for three consecutive measurements).
All the researchers were divided into a blinded group and an un-blinded group and the administrations of study drugs were obscured. The anesthesiologist gave instructions and an investigating nurse of the un-blinded group administered the study drugs. Patients aged < 65 years were given 0.2 μg/kg sufentanil i.v. in advance and the initial dose of HSK3486 or propofol was then injected i.v. after 3 min (± 5 s). Previous phase 2 [bib_ref] Efficacy and safety of ciprofol for the sedation/anesthesia in patients undergoing colonoscopy:..., Teng [/bib_ref] and phase 3 (NCT03674008, on submission) studies on sedation for gastrointestinal endoscopy procedures demonstrated that the initial dose of 0.4 mg/kg HSK3486 was comparable to 2.0 mg/kg propofol. Because of the similar duration and administration method during gastrointestinal endoscopy procedures, the HSK3486 0.4 mg/kg and positive control propofol 2.0 mg/kg doses were also selected for patients undergoing fiberoptic bronchoscopy in the present trial. The infused volumes of both drugs were similar, comprising 0.4 mg/kg with a dilution of 2.5 mg/mL = 0.16 mL/kg for HSK3486 and 2 mg/kg with a dilution of 10 mg/mL = 0.2 mL/kg for propofol. HSK3486 was initially administered i.v. at a dose of 0.4 mg/kg for patients aged < 65 years, with top-up doses of 0.15 mg/kg during the induction and maintenance phases if required. Similarly, propofol was injected i.v. at 2.0 mg/kg in patients aged < 65 years, with top-up doses of 0.75 mg/kg if required. Top-up doses were allowed when inserting the laryngeal mask and the fiberoptic bronchoscope, otherwise at intervals ≥ 2 min and at most five times during any 15-min window, with propofol being the alternative rescue sedative. If sedation was still insufficient, sufentanil (0.05-0.1 µg/kg) was given up to a maximum dose of 0.4 µg/kg. Patients aged ≥ 65 years had their doses of sufentanil and the study medication reduced by 25%.
The laryngeal mask (LM), which is routinely used in our hospitals, was inserted when the MOAA/S score was ≤ 1 and a swivel adapter was connected to its breathing circuit. Then the bronchoscope was inserted through a tunnel inside the LM and pushed into the trachea. Physical monitoring including continuous electrocardiography and pulse oximetry, and automated noninvasive blood pressure was recorded every 2 min. Vital signs and adverse events (AEs) were continuously measured and recorded during the procedure. The MOAA/S score was recorded within 30 s once the fiberoptic bronchoscope was extracted and then at 1-min intervals until the patient was fully alert, followed by the modified Aldrete score, assessed every 2 min until the score was ≥ 9 for three consecutive measurements. A satisfaction evaluation scale that ranged from 0 to 10 (sum of five items) was used for patients and anesthesiologists before discharge, with a higher score indicating better satisfaction (ESM .
## Efficacy assessments
## Primary efficacy outcomes
The success rate of fiberoptic bronchoscopy was assessed according to the following criteria: (1) completion of fiberoptic bronchoscopy; (2) no requirement for any alternative sedative/anesthetic, which means the top-up doses of the experimental drugs were administered no more than five times in any 15-min window from the initial administration to the completion of the procedure.
## Secondary efficacy outcomes
(1) Time to successful induction of anesthesia/sedation, defined as the time from the start of study drug administration to MOAA/S ≤ 1; (2) time to being fully alert, defined as the time from the fiberoptic bronchoscope extraction or/ and the time from the last drug administration to a MOAA/S score of 5 for three consecutive measurements; (3) time to discharge, defined as the time from the fiberoptic bronchoscope extraction or/and the time from the last drug administration to the initial appearance of three consecutive Aldrete measurements of 9; (4) the top-up frequency and dosage of the study medications and sufentanil; (5) anesthesia/sedation satisfaction scores of patients and anesthesiologists collected when patients were ready for discharge.
## Safety assessments
AEs were evaluated for frequency, severity, association to the study drug, relationship to the procedure, and outcomes. The severity of AEs was graded based on the Common Terminology Criteria for Adverse Events (CTCAEs, version 5.0). The causal relationship of an AE to the investigated drug (HSK3486 or propofol) was assessed by the investigator (or a medically qualified assistant) using the classifications provided in ESM [fig_ref] Table 3: Summary of secondary efficacy outcomes [/fig_ref]. AEs were considered to be severe if a patient's daily function was interfered with or their life was threatened. If cough, physical activity, bronchospasm, wheezing, or other reactions caused by operation irritation were observed in patients, they were recorded as operational reactions of fiberoptic bronchoscopy. Sedation-related AEs, including hypoxia, hypotension, and bradycardia, were evaluated from the time of administration of bolus doses of the study drug until the discharge of patients. Hypoxia was defined as pulse oxygen saturation (SpO 2 ) < 90% lasting for > 30 s; hypotension was defined as systolic blood pressure (SBP) < 90 mmHg or a 20% decrease from baseline lasting for > 2 min; bradycardia was defined as a heart rate (HR) < 50 beats/min and lasting for > 2 min. Laboratory tests (routine blood/urine and blood biochemistry), vital signs, and 12-lead ECG recordings were also assessed throughout the study.
## Blood sampling and plasma concentration
A phase 1 study conducted in Chinese healthy subjects found that the plasma concentration of HSK3486 had a threephase elimination characteristic and the t max was 2 min after administration [bib_ref] Pharmacokinetic and pharmacodynamic properties of ciprofol emulsion in Chinese subjects: a single..., Teng [/bib_ref]. Thus, four venous blood samples (3 mL) were collected within 1 h before the initiation of lidocaine atomization inhalation, 2 min ± 10 s after the initial administration of the study drug, and at approximately 10 min and within 1-6 h after extraction of the fiberoptic bronchoscope. Plasma concentrations of HSK3486 and propofol were measured by validated liquid chromatography tandem mass spectrometry (LC-MS/MS) with a low limit of quantification (LLOQ) of 5 ng/mL. A plasma concentration below LLOQ was recorded as below the quantitation limit (BQL) [bib_ref] Safety, pharmacokinetics, and pharmacodynamics of a single bolus of the γ-aminobutyric acid..., Li [/bib_ref].
# Statistical analysis
Statistical analyses were carried out using SAS ver. 9.4 (SAS Institute Inc, Cary, NC, US). The study was designed as a non-inferiority study with a non-inferiority margin of 8% with respect to the success rate [bib_ref] Safety and efficacy of remimazolam compared with placebo and midazolam for moderate..., Pastis [/bib_ref] [bib_ref] A phase 3, randomized, double-blind study to assess the efficacy and safety..., Silvestri [/bib_ref]. At a power of 80%, a one-sided type I error rate of 0.025 and assuming that the success rates of fiberoptic bronchoscopy after HSK3486 and propofol administration were both 95%, a total of 260 patients needed to be enrolled in the study, considering an anticipated dropout rate of 10%, with 130 patients in each treatment arm.
For the primary efficacy outcome, the difference of success rates of fiberoptic bronchoscopy between HSK3486 and propofol (HSK3486-propofol) and the corresponding two-sided 95% confidence intervals (CIs) were calculated. Non-inferiority between HSK3486 and propofol was concluded if the lower boundary of the 95% CI of the difference (HSK3486-propofol) of the success rate was > − 8%. The success rates of fiberoptic bronchoscopy were compared between the HSK3486 and propofol groups using Fisher's exact method adjusted for age-stratification factors. For comparison of secondary efficacy outcomes, the Cochran-Mantel-Haenszel (CMH) test or Fisher's exact probability test was used to compare the categorical variables between two groups, while a t-test or Wilcoxon rank sum test was used to compare continuous variables. The incidence of AEs and drug-related AEs between the two groups was compared using a chi-squared test or Fisher's exact probability test. Continuous variables are presented as the mean ± standard deviation (SD) or median with range (minimum, maximum), while categorical variables are given as numbers with percentages. The study was designed to prove the non-inferiority between HSK3486 and propofol, with respect to success rate (primary endpoint), and all other comparisons were only of an exploratory nature (all P-values are two-sided and were not adjusted for comparisons). The full analysis set (FAS) included all patients randomly assigned to receive at least one drug dose (HSK3486 or propofol) after randomization according to the intention to-treat (ITT) principle and at least one evaluable efficacy outcome was used for analysis of all efficacy outcomes and baseline characteristics of patients. The safety set (SS) and pharmacokinetics analysis set (PKS) included all patients based on the "as treated" principle, was used for the analysis of safety outcomes and plasma concentration measurements, respectively.
# Results
## Patients' disposition and baseline characteristics
A total of 302 patients were screened, of whom 267 met the eligible criteria, and were randomized to receive HSK3486 (n = 134) or propofol (n = 133) [fig_ref] Figure 1: Enrolment flow diagram [/fig_ref]. Thus, all efficacy outcomes were analyzed in the FAS (134 vs. 133, ITT). Due to one patient being randomized to the propofol group initially, but who inadvertently was given HSK3486, there were 135 patients in the HSK3486 group and 132 patients in the propofol group for safety and pharmacokinetic outcome (135 vs. 132, as treated) measurements. Overall, the demographic and baseline characteristics were similar between the two groups [fig_ref] Table 1: Demographic and baseline characteristics of patients [/fig_ref]. Of the patients, 49.4% were female and 89.1% were aged < 65 years. All patients had a MOAA/S score of 5 at baseline and 263 (98.5%) patients were ASA grade I-II.
## Efficacy
## Primary efficacy outcome
All patients completed the fiberoptic bronchoscopy and no patient received alternative sedative/anesthetic drugs, with success rates of 100% in both the HSK3486 and the propofol groups. The 95% CI of the difference (HSK3486-propofol) of success rates was − 2.8 to 2.8%, which met the predesigned criteria of > − 8%, indicating that the non-inferiority of HSK3486 compared to propofol with respect to the success rate was confirmed .
## Secondary efficacy outcomes
The time course of the MOAA/S scores in the two groups is shown in [fig_ref] Figure 2: The [/fig_ref]. [fig_ref] Table 3: Summary of secondary efficacy outcomes [/fig_ref] summarizes the secondary efficacy endpoints of patients in the HSK3486 and propofol groups. The median duration of fiberoptic bronchoscopy, median insertion duration, and the median time to successful induction of anesthesia/sedation were similar in the two groups (all P > 0.05). Patients in the HSK3486 group took a longer time to regain full alertness and to be discharged (all P < 0.05). More than half of the patients in both groups completed fiberoptic bronchoscopy without requiring top-up dosing, and the median times of top-up doses were nearly the same between the two groups. HSK3486 produced comparable satisfaction scores for both patients (P = 0.341) and anesthesiologists (P = 0.972) compared to the propofol group. Three patients in each group received on one occasion a top-up dose of sufentanil, due to cough and/or body movements during the procedure.
## Safety
A total of 172 (64.4%) patients experienced 286 AEs. There were no ≥ grade 3 AEs in the HSK3486 group, all being grade 1 or grade 2. Most AEs were self-curing and selflimited. No patients withdrew from the study because of AEs and no serious adverse events (SAEs) occurred. A total of 66 drug-related AEs occurred in 50 (37.0%) patients treated with HSK3486 and 123 occurred in 93 (70.5%) patients treated with propofol (P < 0.001). The higher incidence of AEs in the propofol group can be mainly attributed to the higher incidence of pain on injection (P < 0.001) [fig_ref] Table 5 Summary: to HSK3486 administered in a fivefold lower concentration [/fig_ref].
Sedation-related AEs occurred in the two groups, being 40 (29.6%) and 42 (31.8%), respectively (P = 0.698). The incidence of hypoxia was a little higher in the HSK3486 patient group than that in the propofol group, but the duration was similar in the two groups. The incidence of hypotension and mean duration in the HSK3486 group were lower than in the propofol group. Although the mean duration of bradycardia was slightly longer in the HSK3486 group, the incidence of bradycardia was similar to that in the propofol group. AEs related to operant reaction of fiberoptic bronchoscopy occurred in both groups, with three patients (one case of wheezing and two cases of bronchospasm) in the HSK3486 group and four patients (two cases of bronchospasm and two cases of body movement) in the propofol group.
Vital signs of patients in the HSK3486 and propofol groups during the procedure are presented in [fig_ref] Figure 3: Vital signs over time [/fig_ref]. The blood pressure of both groups showed a downward trend 4 min after initial drug administration, with a decrease range within 10.92%, which gradually recovered to the baseline value. HSK3486 induced a smaller decreasing trend in mean arterial blood pressure (MAP) from baseline than did propofol (P = 0.002), with maximum mean percentage decreases of − 20.00% (SD 10.99%) and − 24.00 (SD 9.91%), respectively. The respiratory rate in both groups showed a decreasing trend within 2 min of drug administration. HR in both groups fluctuated around baseline after administration, with a maximum fluctuation of 10%, while SpO 2 changed slightly after administration, within a range of 1.88%. Laboratory tests and 12-lead ECG recordings showed no clinically meaningful differences.
## Plasma concentration
HSK3486 had a similar terminal elimination half-life to propofol. The average plasma concentrations of both drugs reached maximal values at 2 min ± 10 s after initial administration, and decreased rapidly between 10 min and 6 h after extraction of the fiberoptic bronchoscope. The plasma concentration during the first 6 h followed first-order kinetics .
# Discussion
Various classes of anesthetic drugs have been used in fiberoptic bronchoscopy, such as benzodiazepines and opioids, or fospropofol and dexmedetomidine. Benzodiazepines with sedative, hypnotic, anxiolytic, anti-convulsant, and anterograde amnesia propertieshave been proven to be helpful during bronchoscopy [bib_ref] Midazolam and other benzodiazepines, Olkkola [/bib_ref]. Midazolam was commonly used in bronchoscopy because of its pharmacokinetic characteristics, but can produce long recovery times. Opioids were frequently used for bronchoscopy in combination with benzodiazepines because of their analgesic and sedative properties, and the reduction in the amount of other sedatives required [bib_ref] Sedation for flexible bronchoscopy: current and emerging evidence, José [/bib_ref]. However, this approach reduced the respiratory rate, resulting in ventilatory depression. Fospropofol is a water-soluble prodrug of propofol, which has been demonstrated to be safe and efficacious in patients undergoing bronchoscopy in a phase 3 randomized controlled trial [bib_ref] A phase 3, randomized, double-blind study to assess the efficacy and safety..., Silvestri [/bib_ref] , but fospropofol has risks similar to propofol [bib_ref] American College of Chest Physicians consensus statement on the use of topical..., Wahidi [/bib_ref]. Dexmedetomidine is a selective α 2 -agonist with sedative and analgesic properties. It has been shown to be safe and effective to use with flexible bronchoscopy [bib_ref] A Phase IIIb, randomized, double-blind, placebocontrolled, multicenter study evaluating the safety and..., Bergese [/bib_ref] without eliciting sympathomimetic vagolytic actions, which may cause bradycardia and hypotension [bib_ref] Clinical uses of alpha2-adrenergic agonists, Kamibayashi [/bib_ref].
In the present study, HSK3486 induced, similar to propofol, a sedative/anesthetic effect with a success rate of 100%. Patients treated with HSK3486 completed the fiberoptic bronchoscopy without the requirement for an alternative anesthetic/sedative and the lower boundary of 95% CI for the difference (HSK3486-propofol) of success rates was − 2.8%, which met the predesigned criteria of > − 8%. In our study, HSK3486 had a similar duration of action and a rapid onset of anesthesia/sedation compared to propofol, both groups reaching a level of deep sedation or general anesthesia (MOAA/S ≤ 1), which met the ideal sedation profiles as previously reported [bib_ref] Safety and efficacy of remimazolam compared with placebo and midazolam for moderate..., Pastis [/bib_ref] [bib_ref] A phase 3, randomized, double-blind study to assess the efficacy and safety..., Silvestri [/bib_ref]. More than half of the patients completed fiberoptic bronchoscopy without needing top-up doses in both groups. The above results indicated that the efficacy of HSK3486 0.4 mg/kg was similar to that of propofol 2.0 mg/kg. HSK3486 had similar terminal elimination half-life characteristics to propofol, but with a potency about five times greater, and HSK3486 patients exhibited slightly longer recovery durations. The reason for this is probably the very short initial distribution half-life and high clearance rate of propofol. However, though the fully alert and recovery times were significantly longer in the HSK3486 group, it was still within 10-15 min, and the clinical implications of this delay are likely to be irrelevant. Nevertheless, a comparison of the pharmacological properties of HSK3486 and propofol will require further studies. More than half of the patients completed the fiberoptic bronchoscopy without top-up dosing, which indicated that overall anesthesia/sedation and recovery potency of HSK3486 were comparable to propofol.
A total of 172 (64.4%) patients experienced 286 AEs, with 71 (52.6%) patients in the HSK3486 group and 101 (76.5%) in the propofol group. There was no grade 3 or above AEs in the HSK3486 group, and most were selfcuring or self-limited, findings similar to other anesthetics/ sedatives [bib_ref] Safety and efficacy of remimazolam compared with placebo and midazolam for moderate..., Pastis [/bib_ref] [bib_ref] A phase 3, randomized, double-blind study to assess the efficacy and safety..., Silvestri [/bib_ref] [bib_ref] Sedation for fiberoptic bronchoscopy: fewer adverse cardiovascular effects with propofol than with..., Oztürk [/bib_ref]. The types of AEs and drug-related AEs were similar in both the HSK3486 and propofol groups. The most remarkable difference in drug-related AEs between the two groups was pain on injection, where the incidence in patients in the HSK3486 group was significantly lower than in the propofol group. Injection pain is a common adverse reaction to propofol, which produces an uncomfortable experience for patients and increases their nervousness and anxiety. It can even elicit body movements that are not conducive to diagnosis and therapy using fiberoptic bronchoscopy. The rate of occurrence of injection pain with propofol has been reported to be about 70% [bib_ref] Prevention of pain on injection with propofol: a quantitative systematic review, Picard [/bib_ref]. Besides other discussed causes, the drug concentration has been proposed to be a major factor for this side effect, with reduced drug concentrations leading to less pain on injection [bib_ref] Reducing pain during propofol injection: the role of the solvent, Doenicke [/bib_ref] [bib_ref] Pain on injection of propofol: effects of concentration and diluent, Klement [/bib_ref] , which might explain the lower incidence compared Previous studies have noted that propofol is associated with an increased risk of hypertriglyceridemia [bib_ref] Hypertriglyceridemia: a potential side effect of propofol sedation in critical illness, Devaud [/bib_ref] , with an incidence of 27.9% in patients after propofol applications for a median time of 47 h [bib_ref] The incidence of propofol-induced hypertriglyceridemia and identification of associated risk factors, Corrado [/bib_ref]. However, hypertriglyceridemia after HSK3486 infusions is less likely to occur since the lipid content in essentially lower than with propofol since a 1% HSK3486 solution includes 5% soybean oil, 2.25% glycerol, and 1.2% purified egg phosphatide versus 10% soybean oil, 2.25% glycerol and 1.2% purified egg phosphatide for a 1% propofol solution [bib_ref] Propofol: a new intravenous anesthetic, Sebel [/bib_ref] , but further studies are necessary to evaluate the effects of HSK3486 on hypertriglyceridemia. Another aspect might be less environmental pollution of HSK3486 compared to propofol, since propofol has been recognized as a contributor to environmental contamination [bib_ref] Propofol wastage in anesthesia, Mankes [/bib_ref]. The most common sedation-related AEs, including hypoxia, hypotension, and bradycardia, occurred within 10 min after the initial administration of HSK3486 or propofol, with 40 (29.6%) in the HSK3486 group and 42 (31.8%) in the propofol group. HSK3486 patients exhibited a lower incidence of hypotension and bradycardia and an equivalent duration of hypoxia (180.80 s in the HSK3486 group vs. 180.70 s in the propofol group), similar to reported studies for other sedatives [bib_ref] Utility of flexible fiberoptic bronchoscopy for critically ill pediatric patients: a systematic..., Field-Ridley [/bib_ref]. Hypotension is a common AE produced by anesthesia/sedation, and the underlying mechanisms still remain to be elucidated. Studies have shown that propofol inhibits myocardial contraction by reducing the concentration of free calcium ions in cardiomyocytes through a protein kinase C-dependent pathway [bib_ref] Experimental conditions are important determinants of cardiac inotropic effects of propofol, Kanaya [/bib_ref]. It has also been suggested that part of the reason for hypotension caused by propofol during induction and maintenance of anesthesia is inhibition of vascular tone [bib_ref] The mechanisms of propofol-mediated hyperpolarization of in situ rat mesenteric vascular smooth..., Nagakawa [/bib_ref]. Vital signs and laboratory tests of patients were all stable, with no clinically significant differences between the HSK3486 and propofol groups. There was a good trend of efficacy and safety of HSK3486 in elderly patients, although the number of elderly patients in each group was small, and further studies need to be conducted to establish the routine use of HSK3486 in older patients who will undergo fiberoptic bronchoscopy. Taken together, the present non-inferiority trial showed that HSK3486 is a useful alternative sedative drug for fiberoptic bronchoscopy with less side effects, particularly injection pain, compared to propofol. The limitations of the present study are: (1) The large proportion of patients aged < 65 years, with 119 patients in each group, leading to a lack of guidelines for older patients, although the results are very promising. Prospective comparative studies with a larger cohort of patients aged > 65 years will be conducted in the near future. (2) Pre-administration of sufentanil prior to fiberoptic bronchoscopy may have had an additional effect on anesthesia/sedation. To avoid this potential problem, follow-up studies will be conducted in the absence of sufentanil administration.
# Conclusions
HSK3486 induced comparable anesthesia/sedation in patients undergoing fiberoptic bronchoscopy and exhibited non-inferiority to propofol, with a success rate of 100%. HSK3486 had a good safety profile and a lower incidence of AEs were reported, all of severity grade 1 or 2.
## Supplementary information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s40263-021-00890-1.
Conflict of interest ZL, HT, XZ, XW, WOY, XCW, XHZ, ZQZ, YLL, WNSG, HW, YPW and QLG declare that they have no conflict of interest.
Availability of data and material The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Code availability Not applicable. Mean plasma concentration-time curve linear plot. The plasma concentration during the first 6 h followed first-order kinetics Ethics approval The study was approved by the ethics committee of each participating hospital.
Consent to participate Informed consent was obtained from all individual participants included in the study. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by-nc/4. 0/.
## Consent for publication not applicable.
# Authors' contributions
[fig] Figure 1: Enrolment flow diagram. FAS full analysis set, PKS pharmacokinetics analysis set, SS safety set [/fig]
[fig] Figure 2: The [/fig]
[fig] Figure 3: Vital signs over time. A Systolic blood pressure (SBP), B diastolic blood pressure (DBP), C mean arterial pressure (MAP), D heart rate, E respiratory rate and F oxygen saturation (SpO 2 ) [/fig]
[table] Table 1: Demographic and baseline characteristics of patients (full analysis set) Data are given as the mean ± SD or numbers with percentages in parentheses ASA American Society of Anesthesiologists, BMI body mass index, MOAA/S Modified Observer Assessment of Alertness/Sedation, SD Standard deviation [/table]
[table] Table 3: Summary of secondary efficacy outcomes (full analysis set) [/table]
[table] Table 5 Summary: to HSK3486 administered in a fivefold lower concentration. [/table]
|
Mitochondrial Involvement in Migration, Invasion and Metastasis
Mitochondria in addition to be a main cellular power station, are involved in the regulation of many physiological processes, such as generation of reactive oxygen species, metabolite production and the maintenance of the intracellular Ca 2+ homeostasis. Almost 100 years ago Otto Warburg presented evidence for the role of mitochondria in the development of cancer. During the past 20 years mitochondrial involvement in programmed cell death regulation has been clarified. Moreover, it has been shown that mitochondria may act as a switchboard between various cell death modalities. Recently, accumulated data have pointed to the role of mitochondria in the metastatic dissemination of cancer cells. Here we summarize the modern knowledge concerning the contribution of mitochondria to the invasion and dissemination of tumor cells and the possible mechanisms behind that and attempts to target metastatic cancers involving mitochondria.
# Introduction
Mitochondria are intracellular organelles that produce the majority of the energy in the cells, providing synthesis of ATP by oxidative phosphorylation (OXPHOS). Beyond energy production mitochondria have multiple functions including the generation of reactive oxygen species (ROS), metabolite production, the regulation of intracellular Ca 2+ homeostasis and modulation of cell death pathways. Additionally, mitochondria contribute to the regulation of signaling pathways linked to the cell proliferation, differentiation, and many others. The multiple functions of mitochondria allow cells to adapt to the changing of environment, including the availability of nutrients and oxygen, making them perfect stress sensors. These functions also determine the crucial role of mitochondria in development and progression of cancer. Indeed, mitochondria may drive tumor progression through adaptation to changing metabolic demand, contributing to chemoresistance, and regulating cell death pathways. Furthermore, mitochondria have been shown to be linked to the metastatic dissemination of cancer cells. Importantly, mitochondrial turnover, i.e., fission/fusion, is deeply involved in the regulation of different mitochondrial functions and metastatic cascade. However, the role of mitochondrial dynamics in cancer cell invasion and metastasis remains highly controversial. Here we took an attempt to summarize the present knowledge about the functions of mitochondria that contribute to the metastatic dissemination and invasion including mitochondrial dynamics, cell death, oxidative stress, metabolism and bioenergetics, Ca 2+ signaling, and mtDNA. Additionally, we highlight the existing therapy approaches to target metastatic cancers involving mitochondria.
## Mitochondria and migration
Metastasis is one of the main cause of cancer patients' death. Metastatic dissemination is characterized by cell detachment from the primary tumor mass, further migration through blood and lymphatic vessels and colonization of different tissues. The metastatic cascade can be subdivided into different stages, including local invasion, intravasation, survival in the circulation, extravasation, survival at a second site and finally outgrowth at this site. The epithelial to mesenchymal transition (EMT) is a biological phenomenon occurring during embryonic development but also associated with cancer metastasis. During EMT cancer cells lose their epithelial features and temporally acquire mesenchymal characteristics which allow them to migrate from the original site in order to colonize different tissues.
Epithelial to mesenchymal transition is regulated by various molecular pathways including TGF-β/Smad, Wnt/β-catenin, Notch, mitogen-activated protein kinase (MAPK), IKK/NF-κB and PI3K/Akt, cytokines (e.g., TGF-β and EGF) and transcription factors Snail (Snai1), Slug (Snai2), Twist (helix-loop-helix factor), and ZEB1/2 (zinc-finger E-box-binding homeobox), etc. Overexpression of EMT transcriptional factors leads to the downregulation of epithelial markers and Tight Junction proteins, such as E-cadherin, occludin, and claudins, which in turn results in the loose of apical cell polarity. On the other hand, EMT activation provides the upregulation of mesenchymal markers: N-cadherin, vimentin, and fibronectin. In addition, EMT is accompanied by increased expression of matrix metalloproteinases (MMPs) and urokinase plasminogen activator (uPA), which contribute to the degradation of the extracellular matrix (ECM) and the basal membrane of epithelial tissue. These events lead to the loss of cell-cell and cell matrix adhesion contacts and an increase in cell motility and cell migration, which are the hallmarks of metastasis.
## Mitochondrial dynamics and migration
Mitochondria have been shown to contribute to carcinogenesis including metastatic dissemination and EMT by different mechanisms. Being extremely dynamic organelles mitochondria continuously change their morphology undergoing fission (fragmentation) and fusion (elongation). These processes are regulated by highly conserved guanosine triphosphatases (GTPases). Fission is controlled by cytosolic dynamin-related protein 1 (Drp1), which is recruited to mitochondria by adapter proteins, including mitochondrial fission factor (Mff) and mitochondrial dynamics proteins of 49 and 51 kDa (Mid49/51), where it forms oligomeric ring structures and executes mitochondrial fission. Fusion is mediated by two GTPases in the outer mitochondrial membrane (OMM), i.e., mitofusins 1 and 2 (Mfn1, 2), whereas the inner mitochondrial membrane (IMM) fusion is promoted by the cristae-shaping protein Opa1. In cancer cells, mitochondrial fission/fusion is unbalanced due to the mitochondrial dysfunction. Several studies have demonstrated that increased fission and/or reduced fusion are associated with malignant transformation in different types of cancer. Furthermore, upregulation of mitochondrial fission and increased expression of Drp1 was shown to promote cancer metastasis. For example, overexpression of Drp1 was detected in breast cancer metastatic cells compared to the non-metastatic, whereas silencing of Drp1 or overexpression of Mfn1 resulted in mitochondrial elongation and significantly suppressed the metastatic properties of breast cancer cells . Similarly, increased mitochondrial fission was observed in hepatocellular carcinoma (HCC) metastatic cells. Comparison of levels of Drp1 in tumor samples and in the normal tissues revealed its higher expression in the former, which is associated with the promotion of tumor cell survival and metastasis formation. In addition, downregulation of Drp1 inhibits glioma cells invasive properties affecting cytoskeleton remodeling through the RhoA/ROCK1 pathway. Recent data have also suggested the existence of a link between mitochondrial fission and hypoxiainduced migration. The inhibition of Drp1 by Mdivi-1 leads to the decreased migration induced by hypoxia. Taken together, these studies provide the evidence that mitochondrial fission is required for cancer cell migration and to support the metastatic potential of cancer cells. In migrating cancer cells, mitochondria localize at the leading edge along microtubules, where the energy demand is higher, providing necessary supply. Unfortunately, the mechanism involving Drp1 in regulation of the process of cancer metastasis remains not fully understood. Different studies suggest that fission is required for efficient redistribution of mitochondria, and the upregulation/activation of Drp1 is associated with the migration of cancer cells. Another study has shown that inhibition of mitochondrial fusion may abolish invasion of syntaphilindepleted prostate adenocarcinoma cells. Syntaphilin suppresses mitochondrial dynamics, cancer cell dissemination in vivo. Moreover, its downregulation correlates with poor outcome of cancer patients. Thereby, the silencing of both Mfn1, 2 and syntaphilin abolished mitochondrial trafficking and abrogated the migratory response . Thus, mitochondrial dynamics are linked to cancer cell migration, and the relative contribution of fission or fusion depends on tumor type and molecular context.
## Ros contributes to migration and metastasis
Reactive oxygen species constantly generated during the metabolic process and play a crucial role in the regulation of various cellular functions. Mitochondrial electron transport chain (ETC) is the main source of ROS. Complexes I and III are often regarded as the major sites of mitochondrial ROS (mtROS) production, but more recent studies indicate that at least ten other mitochondrial enzymes also contribute to ROS generation, including Complex II. The role of ROS in cancer remains highly controversial. First, in cancer cells a higher level of ROS is detected compared to their normal counterparts. A moderate increase of ROS level was shown to support cancer cell proliferation and migration and to activate different signaling pathways associated with cell survival, contributing to tumor growth and malignant transformation. Indeed, the level of ROS has been shown to activate the PI3K pathway. The primary known ROS target in the PI3K pathway is phosphatase and tensin homolog (PTEN). ROS promote the inactivation of the tumor suppressor PTEN by oxidizing active-site cysteine residues, causing the formation of a disulfide bond, which prevents PTEN from inactivating the PI3K pathway. Since ROS can inactivate protein tyrosine phosphatases through oxidation of cysteine residues, ROS may have many yet-to-be discovered effects on diverse, mitogen-activated pathways that are normally inhibited by phosphatases. ROS can stimulate the phosphorylation of MAPK and extracellular signal-regulated kinase (ERK), cyclin D1 expression and JUN N-terminal kinase (JNK) activation, all of which are linked to tumor cell survival and growth . ROS may activate different processes associated with metastatic dissemination and invasion. They may be involved in cytoskeleton remodeling. The cell cytoskeleton is dynamic structure composed of microtubules and filaments. Cytoskeletal rearrangements are important for driving cell migration and invasion through the formation of different types of cellular protrusions including filopodia, lamellipodia, and invadopodia. Recent studies have shown that Rac-mediated actin remodeling is attributed to increased O 2 − levels. Specifically, Rho activation leads to the filopodia formation, while induction of Rac contributes to the formation of lamellipodia. Another mechanism by which ROS may promote tumor cell invasion is by stimulation of the proteolytic degradation of ECM components such as glycosaminoglycan (GAG), contributing to metastatic dissemination.
Increased ROS levels can activate different pathways that induce morphological changes associated with the EMT. For example, increased ROS generation stimulates the acquisition of invasive properties by pancreatic cancer cells through the activation of NF-κB signaling. In turn, treatment with antioxidants leads to the suppression of EMT and attenuates metastasis. NF-κB signaling is strongly associated with the EMT process by promoting the expression of the main EMT-related transcription factors Snail, Slug, Twist1, and ZEB1/2, which is also involved in the disruption of the cell-cell junctions. Furthermore, NF-κB activation may contribute to the transcription of vimentin and MMPs such as MMP-2, MMP-9, to maintain the mesenchymal phenotype and promote tumor cell migration. Another pathway involved in EMT and regulated by ROS is the transcription factor hypoxia inducible factor 1-alpha (HIF-1α), which is induced under hypoxic conditions and can stimulate cancer cell EMT by activating EMT-inducing transcription factors such as Twist, Snail and ZEB1/2. Thus, NF-κB is activated under hypoxic conditions, and thereby, in the presence of hypoxia, may co-regulate many of EMTlinked transcription factors. Importantly, ROS accumulation leads to the stabilization of HIF-1 due to inhibition of the HIF-degrading enzyme prolyl hydroxylase.
There is a complex interplay between the level of ROS and the TGF-β signaling pathway exists, which is the one of the most important pathways involved in EMT regulation. It was reported that ROS mediate TGF-β-induced EMT in cancer. ROS may affect the activation of TGF-β downstream effector Smad, while treatment with the ROS scavenger N-acetyl cysteine (NAC) abolishes Smad phosphorylation . Additionally, ROS may regulate TGF-β activation through different signaling pathways as described above, including MAPK and NF-κB. Conversely, TGF-β can induce ROS production by many alterations in mitochondrial functioning and antioxidant protection. For example, TGF-β affects ROS levels by blocking of ETC Complex IV and upregulation of NADPH oxidase 4 (NOX4). Further, it was revealed that TGFβ increases ROS levels inhibiting ETC Complex III. TGF-β also downregulates the synthesis of the antioxidant glutathione (GSH) and several antioxidant enzymes contributing to cellular redox misbalance and EMT-related processes, such as fibrosis. Using the mitochondriatargeted antioxidant SkQ1 was also shown that oxidative stress is implicated in EMT induced by TGF-β. In cervical carcinoma SiHa cells depletion of ROS leads to increase of E-cadherin and downregulation of Snail, the main negative regulator of E-cadherin. Similarly, pretreatment with the ROS scavenger carotenoid astaxanthin (AST) leads to the suppression of EMT and the production of inflammatory cytokines by mesothelial cells. Thus, TGFβ, as inducer of EMT, is likely to affect this process through the ROS production.
Reactive oxygen species accumulation may influence migration and metastasis of cancer cells through different mechanisms affecting cytoskeleton remodeling, ECM degradation and the activation of signaling pathways. However, in conditions of strong oxidative stress, ROS suppress metastatic dissemination, due to induction of cell death or cellular senescence. Furthermore, elevated ROS levels may activate antioxidant pathways. Indeed, oncogenic K-Ras-, B-Raf-, and c-Myc-mediated pathways may downregulate ROS production through regulation of nuclear factor (erythroid-derived 2)-like 2 (Nrf2), one of the main regulators of the antioxidant response. Nrf2 provides a transcriptional activation of several genes involved in glutathione (GSH) synthesis. It promotes tumorigenesis contributing to the cancer cell protection against oxidative stress and chemotherapeutic agents (Rojo de la. Recent data have demonstrated that Nrf2 activation can stimulate cancer cell migration and metastasis and Nrf2 deletion attenuates metastatic potential breast cancer cells suppressing RhoA GTPases activity . Altogether, these observations demonstrate that the role of ROS in tumor progression and metastasis remains highly controversial. It has been suggested that tumors should maintain the ROS at a definite level in order to sustain their growth and metastasis without causing cytotoxicity. Moreover, for tumor promotion it also necessary to provide the right balance between ROS production and antioxidants.
## Mitochondrial dna mutations contribute to the migration and metastasis
It is known that mtDNA mutations can contribute to tumor initiation and progression. Variations in copy number of mtDNA are associated with tumorigenesis and depend on tumor type. Thus, decreased copy number of mtDNA was detected in breast cancer, HCC, nonsmall cell lung cancer (NSCLC) and gastric cancer. On the other hand, increased copy number of mtDNA was found in prostate, head and neck, and colorectal cancers. Mutations and variations in mtDNA content might be associated with regulation of the metastatic properties of tumor cells . Replacement of mtDNA from a highly metastatic to a poorly metastatic cell line led to an increase in the metastatic potential in the recipient cell line . mtDNA mutations are also associated with EMT of cancer cells. Indeed, EMT induced by TGF-β leads to an increase of mtDNA copy number in NSCLC cells. Conversely, knockdown of mitochondrial transcription factor A (TFAM) leads to a decrease in mtDNA copy number, upregulation of E-cadherin expression, and suppression of cell migration rate in esophageal squamous cell carcinoma. Increased mtDNA content in this type of tumor is associated with the higher energy required for EMT. Furthermore, mtDNA mutations contribute to the acquisition of an aggressive phenotype in oncocytic thyroid tumors leading to their bioenergetic crisis. As a consequence, mitochondrial dysfunction may lead to the activation of glycolysis. Thus, oxygen deprivation may provide positive selective pressure for cancer cells carrying damaging mtDNA mutations. However, another study provided the evidence that EMT could also be induced in mtDNA-depleted cells. Indeed, Frontiers in Cell and Developmental Biology | www.frontiersin.org it has been demonstrated that TGF-β-induced EMT occurs in mitochondria-depleted cell lines leading to the stimulation of invasive properties through activation of Raf/MAPK. These data are consistent with the observation that in human mammary epithelial cells (hMECs) a decrease in mtDNA copy number promotes calcineurin-mediated mitochondrial retrograde signaling, which initiates EMT. Likewise, reduction of mtDNA content by suppression of mitochondrial pyrimidine nucleotide carrier 1 (PNC1), which is responsible for mitochondrial DNA replication, leads to EMT induction in hMECs. A recent study has revealed that increased mtDNA copy number may sustain tumor progression and metastasis by upregulating OXPHOS function in cancer cells that rely on mitochondrial OXPHOS. On the other hand, in cancer cells that depend on glycolytic type of metabolism reduction of mtDNA was shown to promote proliferation and chemoresistance. Summing up, mtDNA mutations and variations of mtDNA copy number are associated with EMT, increased invasiveness and metastasis in different types of cancer. The opposite role of mtDNA content in cancer progression and metastatic dissemination depends on metabolic pattern of different types of cancer.
## Bcl-2 and metastasis
B-cell lymphoma/leukemia gene 2 (Bcl-2) family proteins are considered to be regulators of the apoptotic mitochondrial pathway. This family includes both anti-apoptotic Bcl-2, Bcl-XL, Bcl-w, Mcl-1, Bcl-B, and pro-apoptotic multidomain Bax and Bak proteins. In addition, the pro-apoptotic subfamily includes socalled BH3-only domain proteins, such as Bim, Puma, Noxa, Bad, Bid, and Bnip3. The ratio between these proteins with opposite functions determines the success of apoptosis. Upregulation of anti-apoptotic and downregulation of pro-apoptotic proteins is a hallmark of cancer, and their misbalance is contributed to the chemo-, immune-, and radioresistance of anticancer therapies. However, further evidence has demonstrated that the functions of Bcl-2 family proteins are not limited to cell death control and tumor resistance. It has been established, that Bcl-2 family proteins play crucial roles in the regulation of migration, invasion and metastasis .
Indeed, the overexpression of Bcl-2, Bcl-XL, Bcl-w, and Mcl-1 in different cancers, including glioma, neuroblastoma, melanoma, squamous carcinoma, and breast, lung, and colorectal cancer cells, leads to significant increase in their migratory and invasive properties. On the other hand, downregulation of these proteins attenuates invasiveness without affecting apoptosis or tumor growth, indicating that their prosurvival functions are not linked to regulation of cell migration and invasion. Importantly, several studies have demonstrated that overexpression of the Bcl-2 family proteins is not always sufficient to induce pro-invasive properties of cancer cells, and could require the co-expression of other proteins stimulating invasiveness, such as c-Myc, N-Myc, or Twist1. In certain cases, the exposure to hypoxic conditions may also be essential. It has been suggested that the pro-invasive activity of pro-survival Bcl-2 family members appears to vary depending on the cell type and environment.
The mechanisms involved in Bcl-2 proteins-mediated regulation of invasiveness and metastasis remain incompletely understood. Anti-apoptotic Bcl-2 family proteins have been shown to activate different signaling pathways controlling migration, invasiveness and metastasis ability in cancer. Indeed, Bcl-2 may modulate the EMT program by direct interaction with Twist1 through the helix-loop-helix DNA binding domain of Twist1 and two domains of Bcl-2 in hepatocellular and oral squamous cells. In addition, almost all anti-apoptotic Bcl-2 family members regulate the PI3K pathway involved in metastasis progression. Interplay between Bcl-2 and the p110α subunit of PI3K regulates human colorectal cancer cell migration through actin polymerization and filopodia formation. Likewise, in lung cancer cells, Bcl-XL increases PI3K and p38 MAPK activities, which subsequently stimulate MMP-2 expression via Akt. Specifically, Bcl-w was shown to affect migration and invasion pathways through regulation of PI3K, EGF, Src, MMP-2, uPA, and focal adhesion kinase (FAK). Furthermore, Bcl-w promotes migration and invasion of glioblastoma cells through β-catenin signaling via its translocation into the nucleus to act as transcription factor for MMP-2 and mesenchymal marker expression. Additionally, Mcl-1 supports breast cancer cell migration and invasion via Src family kinases (SFKs) and their targets, and also by alteration of the phosphorylation state of the cytoskeletal protein cofilin.
It has been reported that Bcl-2 family members are capable of regulating the functioning of mitochondria, during cellular respiration and of stimulating ROS generation in the form of the superoxide anion radical O 2 − and H 2 O 2. Thus, multidomain pro-apoptotic proteins Bax and Bak have been shown to bind to the ETC Complex-I resulting in decreased ROS production, whereas anti-apoptotic Bcl-w and Bcl-XL interact with Bax and Bak, and abolish their binding to the Complex-I, stimulating ROS production and PI3K-, Src-, and EGFR-dependent cell migration and invasion.
Analysis of patients' clinical samples confirmed the involvement of pro-survival Bcl-2 family members in metastasis. Bcl-2 expression is associated with lymph node metastases of bladderand gastriccancer, liver metastases of colorectal cancerand lymphovascular invasion of patients with breast cancer. Upregulation of Bcl-XL was observed in lymph node metastases and venous permeation in colorectal cancer, hematogenous metastases of osteosarcoma patientsand lymph node metastases in oral tongue squamous cell carcinoma . Overexpression of Bcl-w is associated with the infiltrative morphotypes of gastric cancerand is overexpressed in patients with FIGURE 2 | Bcl-2 family members regulate metastasis by activation/inhibition of signaling kinases, matrix-degrading enzymes, and transcriptional factors. Arrows or blunt ends indicate activation or inhibition, respectively. Colored lines corresponds to each protein. * -function depends on the tumor type. Bcl-2, B-cell lymphoma 2; Mcl-1, myeloid cell leukemia 1; Bcl-XL, B-cell lymphoma-extra large; Bcl-w, Bcl-2-like protein 2; Bnip3, BCL2/adenovirus E1B 19 kDa protein-interacting protein 3; Bax, Bcl-2-like protein 4; Bak, Bcl-2 homologous antagonist/killer; Bad, Bcl-2-associated death promoter; Puma, p53 upregulated modulator of apoptosis; Twist, class A basic helix-loop-helix protein 38; Sp1, specificity protein 1; Snail, zinc finger protein SNAI1; Slug, zinc finger protein SNAI2. For details, see text. Figure is created using BioRender. lung and breast cancers. Mcl-1 is upregulated among III-IV a stage of esophageal squamous cell carcinoma patients with lymph node metastases.
Conversely, pro-apoptotic proteins of this family may suppress cancer cell invasion and metastatic dissemination. Thus, Bax and Bad expression is associated with downregulation of MMP-2, -9, and -10. Interestingly, Bax and Bid may downregulate tumor cell invasiveness, indirectly repressing the gene expression of c-Jun, cyclin D1, β-catenin, and Sp1, which are known to stimulate invasive properties and metastasis in breast cancer. In addition, Bad, Bim, and Puma were shown to suppress EMT, inhibiting related transcription factors, including Snai1, Sp1, Snai2, and Slug, and subsequent upregulation of epithelial phenotype markers.
Another member of the Bcl-2 family implicated in cancer cell migration, invasion and metastasis is Bnip3, which is considered to be a proapoptotic Bcl-2 protein. Bnip3 also plays an important role in autophagy and mitophagy regulation. However, its role in cancer progression and metastasis remains highly controversial. Thus, in human triple-negative breast cancer (TNBC) the lack of Bnip3 results in tumor progression and metastasis via storage of dysfunctional mitochondria and subsequent ROS accumulation; the events that, as was discussed earlier, lead to expression of HIF-inducible genes including metastasis-related angiogenesis genes. Conversely, in melanoma cells Bnip3 silencing reduces the formation of lamellipodia and filopodia as well as cell migration through the downregulation of integrin-associated glycoprotein CD47, Rac1 and Cdc42. Bnip3 in HCC may suppress metastasis through the JNK/Bnip3/SERCA/CaMKII axis, leading further to the cofilin/F-actin/lamellipodia inhibition. As mentioned above, Bnip3 is involved in autophagy, the process, which is tightly linked to EMT, migration and metastasis. Specifically, Bnip3-dependent autophagy via hypoxia-induced ROS-modulated p38 MAPK and JNK activation contributes to keratinocytes migration . In NSCLC Bnip3 supports metastasis via its modulation by aryl hydrocarbon receptor (AhR), which has an impact on Bnip3 proteasomal degradation and subsequent autophagy disturbance. These events result in decreased EMT progress. Thus, Bnip3, depending on cancer cell type and hypoxic conditions, fulfills opposite functions in metastasis.
Analysis of clinical samples confirmed the participation of Bcl-2 pro-apoptotic members in metastasis formation. Increased expression of Bax, Bak and Puma is associated with a lack of vascular invasion in patients with oral squamous cell carcinoma. Loss of Bax protein was demonstrated in retinoblastoma specimens with massive choroidal invasion. Downregulation of Bnip3 is characteristic of lymph node metastases in breast cancer. Conversely, in patients with renal cell carcinomas Bnip3 expression is correlated with lymph node metastasis. Methylation of Bim and Bnip3 genes is associated with metastasis and the gene methylation rate is increased among colorectal and pancreatic cancer patients compared to healthy individuals. Interestingly, overexpression of pro-apoptotic Bcl-2 family members, except Bnip3, frequently correlates with decreased metastasis and favorable outcomes in patients with various cancer types.
Thus, anti-apoptotic members of the Bcl-2 family, including Bcl-2, Bcl-XL, Bcl-w, and Mcl-1, support the invasion and metastasis in various types of cancer. This positive influence is achieved through EMT, subsequent cytoskeleton rearrangement, overexpression of MMPs and uPA, and regulation of PI3K, p38 MAPK, Akt, and ERK. In contrast to anti-apoptotic proteins, almost all pro-apoptotic members, such as Bax, Bak, Bad, Bid, and Puma, were characterized by suppression of the metastatic potential of cancer cells. Despite Bnip3 being considered a pro-apoptotic member, its influence on metastasis remains highly controversial, likely due to its atypical BH3-domain contributing to autophagy-dependent processes. Bcl-2 family members play important role in the regulation of ROS production and activity of mitochondrial complexes leading to the activation of molecular pathways controlling invasion and metastasis. Moreover, the Bcl-2 family proteins affect cell migration of both malignant and normal tissues. Analysis of patient specimens with tumors confirms the participation of Bcl-2 family members in invasion and metastasis, which gives a reason to consider these proteins for target therapy.
## Er-mitochondria network and metastasis
The endoplasmic reticulum (ER) is a crucial cellular Ca 2+ reservoir, that coordinates Ca 2+ signaling, protein synthesis and folding and traffic of properly folded proteins to the Golgi apparatus. Accumulation of misfolded proteins in the ER lumen triggers unfolded protein response (UPR), which is an adaptive signaling pathway to restore protein homeostasis (proteostasis). If accumulation of misfolded proteins remains unresolved activation of UPR signaling may lead to the initiation of apoptotic cascades. Crosstalk between apoptosis and UPR is maintained by mitochondria functioning partly due to contact sites with ER. These so-called mitochondria-associated ER membranes (MAMs) are key for Ca 2+ transport between the ER and mitochondria to maintain cellular homeostasis and regulate ER stress. Moreover, MAMs form functional networks essential in determining pro-survival/pro-death and inflammation signaling.
As mentioned above, ER is a multifunctional organelle, the main function of which is to control protein-folding quality. Numerous factors may affect proper protein folding in the ER, including oxidative stress, hypoxia, glucose deficiency, viral infections and other physical/chemical stresses. As a result, it leads to ER stress and subsequent UPR. In mammals the UPR is carried out by three distinct ER-related transmembrane proteins, including protein kinase RNAlike ER kinase (PERK), endoribonuclease inositol-requiring enzyme 1 alpha (IRE1α/IRE1), and activating transcription factor 6 (ATF6). In unstressed cell, ER-related transmembrane proteins including PERK, IRE1α/IRE1, and ATF6 are bound to immunoglobulin heavy chain protein/glucose-regulated protein 78 (BiP/GRP78). Under ER stress BiP dissociates from these proteins to trigger signaling pathways that result in the reduction of global protein synthesis, degradation of unfolded proteins and increase of protein-folding capacity of the ER.
Endoplasmic reticulum plays an important role in mitochondrial calcium signaling via the contact sites between mitochondria and ER (MERCs). The portion of membranes involved in these interactions defines the MAMs, which, as mentioned above, provide Ca 2+ traffic between these organelles. The transport of extracellular Ca 2+ into the cytosol occurs through the voltage-gated, ligand-gated, and store-operated Ca 2+ -channels (SOCCs) including Orai and/or transient receptor potential channels (TRPC). Orai1 and TRPC are activated through their binding to the stromal interaction molecule 1 (STIM1), which is the ER Ca 2+ -sensor. Ca 2+ transport from the cytosol and its accumulation in the ER depends on the action of ATP-driven sarco/ER Ca 2+ -ATPase (SERCA). Ca 2+ is transported from the ER via 1,4,5-trisphosphate (IP3) and ryanodine receptors (IP3Rs, RyRs), after which Ca 2+ invades the mitochondria through the voltage-dependent anion channels (VDACs) on the OMM. Ca 2+ then is transferred by the mitochondrial calcium uniporter (MCU) on the IMM. Expression levels of these calciumsignaling proteins are frequently altered in numerous types of tumor cellsand most of all govern metastasis-related processes.
As mentioned above, ER is the main cellular Ca 2+ -store. Decreased Ca 2+ level in the ER results in STIM1 oligomerization and its transfer from ER to the plasma membrane where it promotes Orai1-dependent store-operated Ca 2+ entry (SOCE). A growing body of evidence indicates the existence of an interplay between mitochondria and SOCE. Indeed, SOCE activation is accompanied by TRPC-modulated increase in cytosolic Na + level that in turn promotes the activation of mitochondrial Na + /Ca 2+ exchanger (NCLX) leading to mitochondrial Na + influx and Ca 2+ efflux. Therefore, NCLX tightly regulates mitochondrial Ca 2+ level and prevents excessive Ca 2+ accumulation in mitochondria that can lead to the increase of the mtROS level and subsequent SOCE suppression via oxidation of redox-sensitive Cys195 of Orai1. Besides NCLX, the activity of SOCE-related proteins is regulated by Bcl-2. The mutations in BH1 domain of Bcl-2 protein leads to STIM1, Orai1-3, TRPC1 overexpression and SOCE enhancement. It has been established, that hyperactive SOCE induced by STIM1 and Orai1 overexpression correlates with increased metastasis in different types of cancer. Intensified SOCE supports tumor cell invasion and migration by cytoskeleton rearrangement, ECM degradation and tumor microenvironment remodeling.
Sarco/ER Ca 2+ -ATPase is a well-known regulator of Ca 2+ stores in the ER and maintains the level of Ca 2+ uptake and leak properties. Furthermore, SERCA inactivation associated with Yap deficiency has been shown to inhibit HCC metastasis through the cofilin/F-actin/lamellipodium pathway. Additionally, the downregulation of SERCA leads to a significant decrease in Ca 2+ level in migrating cells that in turn inhibits cell migration and tracheogenesis. Other proteins localized to MAMs are VDAC and IP3R, the action of which is dependent on Bcl-2 family proteins and, in addition to their role in apoptosis, partly regulate Ca 2+ signaling via their complex with MAM proteins. Anti-apoptotic family members Bcl-2, Bcl-XL, and Mcl-1 bind to VDACs and suppress mitochondrial Ca 2+ transport that in turn supports cell migration and invasion. Both Bcl-2 and Bcl-XL interact with VDAC1 through BH4 domain; however, Bcl-XL BH4 is more effective than Bcl-2-BH4 in targeting VDAC1 activity. Dissociation between these anti-apoptotic Bcl-2 family members and VDACs results in decreased migration of TNBCand NSCLC cells . Bcl-2 family proteins could also interact with IP3R suppressing Ca 2+ -release. Like VDACs, Bcl-2 binds IP3R through its BH4 domain inhibiting its activity. Besides Bcl-2, other anti-apoptotic family members, including Bcl-XL and Mcl-1, are able to influence IP3R activity and Ca 2+ signaling, but their role in mitochondria-associated ER membrane-related calcium signaling still remains controversial. Also like VDACs, IP3R may regulate cell migration separately from its complex with Bcl-2 family members. It has been revealed, that inhibition of ryanodine receptor subtype IP3R3 and subsequent decrease in Ca 2+ release results in suppression of the invasion and migration of glioblastoma cell lines and metastasis in glioblastoma mouse model . Overexpression of IP3R3, but not of IP3R1 and IP3R2, leads to stimulation of the migration properties of breast cancer cells sustaining Ca 2+ signaling. Thus, IP3R could regulate cancer cell migration and metastasis through modifying calcium ER level.
Another MAM-related protein is Sig1R (stress-activated chaperone sigma-1 receptor). When ER stress is not activated, Sig1R cooperates with MAMs chaperone BiP/GRP78, whereas under activation of IP3Rs Sig1R dissociates from chaperone BiP and binds to IP3R3, leading to its stabilization at the MAM and increasing Ca 2+ flux to the mitochondria. The expression level of MAM-associated Sig1R is increased in metastatic breast and colorectal cancer cells as compared to normal tissues. Consistently with the abovementioned MAM-related proteins, MCU also affects migration, invasion and metastasis. Silencing of this uniporter results in decreased mitochondrial Ca 2+ level and ROS production, as well as migratory and invasiveness capacities. These findings are in good agreement with in vivo experiment. MCU gene deletion reduces tumor metastasis in TNBC MDA-MB-231 xenografts via HIF-1-dependent gene expression.demonstrated similar results in breast cancer MCF-7 cells by MCU overexpression, which leads to enhanced migratory and invasiveness potential in vitro and lung metastasis mouse model in vivo. Furthermore, overexpressed MCU was found in specimens from breast cancer patients with metastases. Thus, MCU expression correlates with migration and invasion of cancer cells, as well as with tumor metastasis, which has been proved by both in vitro and in vivo studies. Besides MAM-anchored proteins, UPR regulators, including PERK, IRE1 and BiP/GRP78, all influence the migration and metastasis. PERK as a key UPR sensor, also participates in MAM signaling. A growing amount of evidence proves that UPR signaling and EMT reprogramming mutually activate each other. In gastric cancer cells knockdown of UPR-related proteins such as PERK, ATF4, and ATF6, decrease TGF-β expression and abrogates EMT under severe hypoxia. The inverse pattern in this UPR-EMT axis has been demonstrated in both in vivo and in vitro models of breast cancer: cells undergoing EMT have a branched ER structure and activated PERK-eIF2α link of the UPR, which helps cells to metastasize. Analysis of specimens from patients with breast, gastric, colon and lung metastatic tumors revealed correlations between expression of EMT and PERK-eIF2α genes. Additionally, ATF4, ATF6, another ER-transmembrane protein IRE1 and its-related X-box binding protein-1 (XBP1) play a role in metastatic progression. IRE1α regulates actin cytoskeleton rearrangement and influences the cell migration via filamin A in MEFs, fly, and zebrafish models. IRE1-XBP1 pathway is regulated by lysyl oxidase-like 2 (LOXL2) overexpression, and activates EMT via Snai1/2, ZEB2 and TCF3 transcription factors in breast carcinoma cells. Notably, XBP1 expression is significantly upregulated in tumor and lymph node metastases compared to normal tissues from patients with oral squamous cell carcinoma, and downregulated XBP1 expression results in decreased cell invasion capacity. Thus, proteins of three distinct UPR branches, including PERK, ATF4, ATF6, IRE1, XBP1, and BiP/GRP78, contribute to cancer cell invasive properties and metastatic dissemination regulating MAM signaling . FIGURE 3 | Schematic representation of the link between ER-mitochondria network and motility of cancer cells. Arrows or blunt ends indicate activation or inhibition, respectively. Blue arrows indicate direction of Ca 2+ current. Yellow circles -Ca 2+ . MAM, mitochondria-associated ER membrane; Orai1, calcium release-activated calcium channel protein 1; TRPC, transient receptor potential cation channel; SERCA, sarco/endoplasmic reticulum Ca 2+ -ATPase; STIM1, stromal interaction molecule 1; IP3R, inositol trisphosphate receptor; RyR, ryanodine receptor; IRE1, serine/threonine-protein kinase/endoribonuclease inositol-requiring enzyme 1; XBP1, X-box binding protein 1; PERK, protein kinase RNA-like endoplasmic reticulum kinase; eIF2α, eukaryotic translation initiation factor 2; ATF4, activating transcription factor 4; NCLX, mitochondrial Na + /Ca 2+ exchanger; BiP, binding immunoglobulin protein; SigR1, sigma receptor 1; Mfn2, mitofusin2; VDAC, voltage-dependent anion-selective channel; MCU, mitochondrial calcium uniporter. For details, see text. Figure is created using BioRender.
## Metabolism and metastasis
Altered metabolic activity is one of the hallmarks of cancer. Cancer cells change their metabolism in order to satisfy increasing of bioenergetic and biosynthetic demand and maintain tumor growth (DeBerardinis and Chandel, 2016). Unlike normal cells, which generate much of their ATP via mitochondrial-dependent OXPHOS, cancer cells often demonstrate upregulation of glycolysis even under conditions when oxygen concentration is not limited. This phenomenon was observed in the 1920s by Otto Warburg, who demonstrated that tumor tissues metabolize approximately ten-fold more glucose to lactate in a given time than do normal tissues, which led him to conclude that cancer cells rely on glycolysis more than do healthy cells. Enhanced aerobic glycolysis has been detected in many types of cancer and is correlated with worse clinical outcome . However, further studies have demonstrated that cancer cells may also engage mitochondrial respiration in addition to glycolysis. Indeed, breast cancer cells produce most of their ATP through mitochondrial oxidation. Similarly, glioma cell lines are strongly dependent on mitochondrial OXPHOS for ATP production. Moreover, cancer cells may display distinct metabolic characteristics depending on the tissue of origin. Thus, lung, liver and colorectal cancers, and leukemias depend on glycolysis, whereas, melanomas, lymphomas, and glioblastomas are characterized as oxidative tumors. Tumor cells can also switch from one type of metabolism to another under glucose-limiting conditions as observed in cervical cancer, breast carcinoma, hepatoma and pancreatic cancer cells. A growing number of studies provide the evidence that cancer cell migration is associated with significant metabolic alterations supporting metastatic dissemination. Thus, it was reported that increased motility of cancer cells requires the shift toward utilization of glycolytic pathways. Glycolytic genes activation has been detected in different tumors and is often associated with malignant and aggressive phenotypes. For example, expression of hexokinase 2 (HK2), the embryonic isoform of hexokinase, the enzyme which defines the start of glycolysis, is associated with increased risk of recurrence, and adverse clinical outcome for breast cancer, pancreatic cancer, and neuroblastoma patients. Further studies have demonstrated that glycolytic enzymes also contribute to the metastatic progression of cancer cells. For example, pyruvate kinase isozyme M2 (PKM2), which mediates the final rate-limiting step of glycolysis, promotes aggressive phenotype and metastasis in different types of tumors. This enzyme also acts as a transcriptional coactivator of HIF-1α in cancer cells, thus promoting glycolysis and inducing EMT. Furthermore, EMT stimulation induced by TGF-β leads to the nuclear translocation of PKM2 in colon cancer cells, where it interacts with TGIF2 and other transcription factors, promoting EMT and supporting the malignant properties of tumor cells. Phosphohexose isomerase (PHI) is another glycolytic enzyme that involved in stimulation of invasion and metastatic dissemination through extracellular autocrine motility factor (AMF). The overexpression of PHI leads to the increased invasion and metastasis of colon cancer cells. Additionally, PHI/AMF overexpression has been reported to promote the EMT activation through the NF-κB pathway and increased expression of EMT markers such as Snai1 and ZEB1/2. In keeping with these observations, high PHI levels in the serum correlate positively with metastases of colorectal and esophageal squamous cells, and lung tumors. Conversely, downregulation of glycolytic enzymes including PKM2 and PHI, inhibit the proliferation and migration of cancer cells. Inhibition of glycolysis attenuates cell motility even while mitochondrial ATP synthesis remains intact, and inhibition of mitochondrial respiration reduces cell motility only minimally compared to inhibition of glycolysis.
Glycolysis regulates different stages of metastatic dissemination, contributing to the different stages of the metastatic cascade. Thus, prostate cancer cells undergoing EMT and acquiring mesenchymal features exhibit higher glycolytic activity than their epithelial counterparts. High glycolysis rate is associated with increased cytoskeletal rearrangement and cell migration. In turn, inhibition of glycolysis suppresses the migration properties of prostate cancer cells. In addition, an interrelation between EMT induced by TGF-β, activation of the glycolytic pathway, and repression of mitochondrial function was demonstrated. In breast cancer, loss of fructose-1,6-bisphosphatase together with the loss of E-cadherin promotes cancer stem cell (CSC)-like features and cancer cell dissemination by enhancing β-catenin signaling and the EMT program. These events are concomitant with the induction of glycolysis, increase in glucose uptake, and inhibition of oxygen consumption.
Although it is known that metastasis requires activation of the glycolytic program, recent studies demonstrate an equal importance of OXPHOS for metastatic dissemination. For example, the strong correlation between expression of PGC-1a (peroxisome proliferatoractivated receptor gamma coactivator-1a), a key regulator of mitochondrial biogenesis and OXPHOS, and invasive properties was observed in breast cancer cells. PGC-1a supports migration of cancer cells, stimulating mitochondrial biogenesis and respiration, whereas downregulation of PGC-1a decreases the frequency of metastasis. Elevated OXPHOS activity is also linked to the high metastatic potential in mouse melanoma and human cervical cancer cells. In turn, prostate cancer cells exhibit a mixed phenotype, where both glycolysis and OXPHOS are required for energy metabolism at different stages of disease progression. This hybrid metabolic state, also called metabolic plasticity, can sustain tumor cell survival under different micro-environmental conditions, while at the same time supporting tumor metastasis and therapy-resistance. Thus, a hybrid metabolic phenotype, characterized by high HIF-1/AMPK activities and high glycolysis/OXPHOS (glucose oxidation and FAO) activities, allows cancer cells to acquire metabolic plasticity and utilize different types of nutrients. Furthermore, it permits the cells to produce energy efficiently through multiple metabolic pathways and meanwhile synthesize biomass for rapid proliferation using by-products from glycolysis. A hybrid metabolic phenotype maintains the cellular ROS at a moderate level so that cancer cells can benefit from ROS signaling and avoid DNA damage due to excessive ROS.
Finally, different metabolic profiles may dictate metastatic fitness to distinct organ sites. It has been shown that metastatic breast cancer cells may display different metabolic pathways depending on the site of metastasis. Hence, breast cancer cells obtained from bone and lung metastases rely on OXPHOS, whereas liver-metastatic breast cancer cells engage a glycolytic type of metabolism.
In consequence, metabolic pathways of migrating cancer cells appear to be inter-connected and characterized by plasticity depending on different factors, i.e., tumor type, microenvironment, site of metastasis formation, etc. A better understanding of this metabolic plasticity will permit the design of specific therapy approaches in order to target metastatic cancer cells more efficiently.
## Therapeutic targeting metastasis
Metastasis is associated with poor outcome of cancer patients. Existing therapeutic approaches are often ineffective or provide limited clinical benefit. Hence, mitochondria play an important role in metastatic dissemination, the targeting mitochondria might represent an attractive approach for the development of new strategies for treatment of metastatic cancers.
Metastatic tumors have been shown to reprogram their metabolism in order to successfully metastasize. Accordingly, significant efforts have been made to target cancer cell metabolism in different tumors for the prevention of metastasis progression. For example, the antidiabetic drug metformin has been shown to possess anticancer properties in different types of cancer. Metformin is a Complex I inhibitor, providing cancer metabolism suppression through downregulation of mitochondrial glycerophosphate dehydrogenase (mGPDH) and OXPHOS inhibition, leading to decreased metastasis levels in a thyroid cancer mouse model. Metformin also attenuates the growth of lung metastatic nodules in an ovarian cancer mouse model by inhibiting the mTOR signaling pathway. At low concentrations metformin inhibits breast cancer invasion and metastasis by suppressing ROS production, suggesting the use of metformin as a chemopreventive agent to block cancer cell invasiveness. Although, the precise mechanisms of action of metformin are still debated, a number of clinical studies have confirmed its antitumor properties. At present metformin is used during the treatment of different cancer types in order compound to inhibit hypoglycemia non-target effect of various chemotherapy drugs (da Veiga. Further clinical studies are required to clarify its anti-metastatic properties.
Glycolysis inhibition has been shown to suppress metastasis in several types of cancer. For example, HKII inhibitor lonidamine (TH-070), a derivative of indazole-3-carboxylic acid, provided significant effectiveness in preclinical studies when the drug was administered in combination with paclitaxel and cisplatin. However, despite promising early stage results, further phase II and phase III trials targeting lung cancer with lonidamine have shown its limited efficacy and hepatic toxicity. Importantly, a more recent study demonstrated that the use of modified lonidamine is significantly more efficacious in inhibiting mitochondrial bioenergetics in lung cancer cells, leading to suppression of lung cancer progression and metastasis. Mitochondrial-lonidamine activates the generation of ROS in lung cancer cells, which leads to the inactivation of the Akt/mTOR/p70S6K signaling pathways and autophagic cell death. Glycolysis can be targeted by 2-deoxyglucose (2-DG), a non-metabolizable glucose analog, which is also pursued in the clinic. However, dose-escalation phase I trials in patients with castrate-resistant prostate cancer and other advanced solid tumors resulted in asymptomatic QTc prolongation that limited further drug evaluation. Since tumors may shift from glycolysis to OXPHOS, or even engage hybrid metabolisms, several studies have proposed the dual inhibition of cancer metabolism using metformin and 2-DG. Indeed, combined treatment with metformin and 2-DG led to the significant suppression of tumor growth and metastasis in preclinical models.
Mitochondrial ROS have been reported to function as signaling molecules implicated in the regulation of tumor growth and metastasis. The different mechanisms by which ROS contribute to tumor growth and metastatic dissemination were discussed above. Thus, targeting mtROS seems to be an attractive approach for cancer therapy. However, contrary to the expected results, the use of antioxidants for anticancer treatment led to increased risk of cancer. Furthermore, the treatment with NAC was shown to enhance the metastatic dissemination of human melanoma cells, providing evidence that oxidative stress may, in certain circumstances, stimulate metastasis. The cause of the failure of treatment with antioxidants could be their lack of specificity. They also may regulate many different processes involved in tumor growth and metastasis. On the other hand, it has been shown that inhibition of ROS with antioxidants that target precisely mitochondrial oxidative stress may stop metastatic spread. Scavenging with MitoTempo, specific mitochondrial antioxidant, significantly reduced cancer cell invasion and prevented metastasis. Thus, ROS targeting appears to be more complex than believed before and requires further detailed investigation.
Another therapeutic agent that has shown promising results in preclinical studies is an inhibitor of mitochondrial heat shock protein 90 (Hsp90) Gamitrinib. This compound induces mitochondrial dysfunction, providing depolarization of inner membrane potential that in turn regulates the release of cytochrome c. In mouse model of prostate cancer, Gamitrinib administration inhibited tumor growth and metastasis affecting mitochondria. Furthermore, targeting Hsp90 with Gamitrinib suppresses cancer cell migration and metastasis preventing metabolic reprogramming and increasing AMPK phosphorylation, Gamitrinib is also effective in combination therapies with inhibitors of both TRAIL and PI3K.
Since the expression of Bcl-2 family proteins has been detected in metastases of different tumors, another possible approach for cancer therapy may be focused on targeting Bcl-2 family members. BH3-mimetics are promising therapeutic drugs that mimic endogenous Bcl-2 family member antagonists, thereby target some of them and abrogating their anti-apoptotic functions. Initially, BH3-mimetics displayed encouraged results in hematological malignancies including lymphoma lymphocytic leukemia, acute myeloid leukemia, small lymphocytic lymphoma and mantle-cell lymphoma. Thus, first-generation BH3-mimetics such as ABT-737 and its orally available derivative navitoclax (ABT-263), which are inhibitors of Bcl-2 and Bcl-W, have shown clinical efficacy. However, in several cases the treatment with these agents was limited by severe thrombocytopenia. Clinical studies of ABT-199 in chronic lymphocytic leukemia and non-Hodgkin's lymphoma have shown impressive antitumor efficacy, with higher response rates than navitoclax and without thrombocytopenia. Several clinical trials have also demonstrated the efficacy of BH3-mimetics in solid tumors. In particular Mcl-1 has emerged as a promising target for the treatment of melanoma. Additionally, a novel gossypol derivative and BH3-mimetic ch282-5 (2-aminoethanesulfonic acid sodiumgossypolone) induced colon cancer cell death in vitro and in vivo. Ch282-5 treatment activated mitochondria-dependent apoptotic pathway accompanied by mitophagy disruption and mTOR pathway activation. Furthermore, Ch282-5 provided suppression of colon cancer cell migration, invasion and liver metastasis . Notably, Bcl-2 family members were shown to interact with Drp1 and treatment with BH3-mimetic A-1210477 led to Drp1-dependent mitochondria fragmentation, whereas Drp1 silencing significantly reduced apoptosis induced by BH3-mimetic in lung, cervical, and breast cancer cell lines. Conversely, inhibition of Drp1 in combination with BH3-mimetic treatment significantly enhanced apoptotic response in melanoma cells. Additionally, inhibition of Drp1 by Mdivi-1 increased the cytotoxic effect of combination treatment with A-1210477 and ABT-263 in different melanoma cell lines.
Another interesting approach to target metastatic cancers is the regulation of mitochondrial K + /H + exchange. Salinomycin is an antibiotic from the polyether ionophores group widely used in agriculture. Recently, it has been revealed that it possesses anticancer properties in different types of cancer. Salinomycin may target chemoresistant tumor cells, inhibiting Wnt/β-catenin and Sonic Hedgehog signaling pathways. Furthermore, it suppresses the migration of colorectal, breast, lung and colon cancer cell lines as well the invasion of nasopharyngeal carcinoma and bladder cancer cells in vitro. Consistently, in vivo studies proved that salinomycin may reduce metastasis formation in mammary tumor mouse model, bladder tumor rat model and intravenous mouse tumor model. Importantly, salinomycin is able to suppress the late stages of autophagy contributing to the ROS generation and mitochondria dysfunction. This might explain the mechanism by which salinomycin targets mitochondrial K + /H + exchange and prevents migration, invasion and metastasis.
Summarizing, mitochondria contribute to tumor progression and metastasis through different mechanisms including redox signaling, mitochondrial biogenesis, regulating Bcl-2 family members, metabolic reprogramming and mitochondrial K + /H + exchange. The better understanding of these mechanisms and the possible interplay between them may provide new therapeutic approaches to target metastatic diseases.
# Conclusion
Mitochondria are very important and complex organelles that affect tumorigenesis and metastatic dissemination through different mechanisms including regulation of metabolism, redox status, signaling and cell death pathways. Recent evidence has demonstrated the existence of complex interplay between mitochondria-related functions and mitochondrial dynamics. Thus, dysregulated mitochondrial turnover contributes to tumorigenesis and metastases. However, the mechanisms connecting mitochondrial dynamics to the development of metastasis remain poorly understood. In addition, the flexibility of mitochondria that allow cancer cells to adapt to the changing microenvironment and stresses should be considered in order to combat cancer successfully. Consequently, a better understanding of the processes regulated by mitochondria and their complex interplay with mitochondrial biogenesis may offer new promising therapeutic strategies for cancer treatment. |
Outcome of HIV-exposed uninfected children undergoing surgery
Background: HIV-exposed uninfected (HIVe) children are a rapidly growing population that may be at an increased risk of illness compared to HIV-unexposed children (HIVn). The aim of this study was to investigate the morbidity and mortality of HIVe compared to both HIVn and HIV-infected (HIVi) children after a general surgical procedure.Methods: A prospective study of children less than 60 months of age undergoing general surgery at a paediatric referral hospital from July 2004 to July 2008 inclusive. Children underwent age-definitive HIV testing and were followed up post operatively for the development of complications, length of stay and mortality.Results: Three hundred and eighty children were enrolled; 4 died and 11 were lost to follow up prior to HIV testing, thus 365 children were included. Of these, 38(10.4%) were HIVe, 245(67.1%) were HIVn and 82(22.5%) were HIVi children. The overall mortality was low, with 2(5.2%) deaths in the HIVe group, 0 in the HIVn group and 6(7.3%) in the HIVi group (p = 0.0003). HIVe had a longer stay than HIVn children (3 (2-7) vs. 2 (1-4) days p = 0.02). There was no significant difference in length of stay between the HIVe and HIVi groups. HIVe children had a higher rate of complications compared to HIVn children, (9 (23.7%) vs. 14(5.7%) (RR 3.8(2.1-7) p < 0.0001) but a similar rate of complications compared to HIVi children 34 (41.5%) (RR = 0.6 (0.3-1.1) p = 0.06).Conclusion: HIVe children have a higher risk of developing complications and mortality after surgery compared to HIVn children. However, the risk of complications is lower than that of HIVi children.
Background HIV-exposed uninfected (HIVe) children are a rapidly growing population. Programs for the prevention of mother to child transmission (PMTCT) have reduced the transmission rate of perinatal HIV infection to approximately 2% to 5% [bib_ref] Addressing the paediatric HIV epidemic: a perspective from the Western Cape Region..., Eley [/bib_ref] [bib_ref] Single-dose perinatal nevirapine plus standard zidovudine to prevent mother-to-child transmission of HIV-1..., Lallemant [/bib_ref]. Such programs have therefore effectively reduced the number of HIV infected (HIVi)children but identified an increasing population of HIVe children [bib_ref] Severe infections in HIV-exposed uninfected infants: clinical evidence of immunodeficiency, Slogrove [/bib_ref].
HIVe children have been overlooked as a group of children who may be at an increased risk of illness compared to HIV-unexposed (HIVn) children. Recently, increased morbidity and mortality in HIVe children compared to HIVn children has been reported [bib_ref] Severe infections in HIV-exposed uninfected infants: clinical evidence of immunodeficiency, Slogrove [/bib_ref] [bib_ref] Altered immunological reactivity in HIV-1-exposed uninfected neonates, Hygino [/bib_ref] [bib_ref] Infectious disease morbidity among young HIV-1-exposed but uninfected infants in Latin American..., Mussi-Pinhata [/bib_ref] [bib_ref] Clinical profile and morbidity pattern of infants born to HIV infected mothers..., Adhikari [/bib_ref] [bib_ref] Clinical and immunologic characteristics and therapeutic interventions in children born to human..., Lin [/bib_ref] [bib_ref] Haematological parameters of HIV-1-uninfected infants born to HIV-1-infected mothers, Bunders [/bib_ref] [bib_ref] Effect of age, polymicrobial disease, and maternal HIV status on treatment response..., Mcnally [/bib_ref]. Many factors may account for this including innate deficiencies in immunity [bib_ref] Placental transfer and maternally acquired neonatal IgG immunity in human immunodeficiency virus..., De Moraes-Pinto [/bib_ref] [bib_ref] Human immunodeficiency virus (HIV)-specific cellular immune responses in newborns exposed to HIV..., Kuhn [/bib_ref] [bib_ref] Risk factors for postnatal mother-child transmission of HIV-1, Embree [/bib_ref] , feeding practices [bib_ref] Breastfeeding, HIV status and weights in South African children: a comparison of..., Patel [/bib_ref] , poor protection from maternal antibodies or environmental exposures [bib_ref] Infectious disease morbidity among young HIV-1-exposed but uninfected infants in Latin American..., Mussi-Pinhata [/bib_ref].
As PMTCT programs expand, an increasing number of HIVe children can be expected to require a routine or emergency surgical procedure [bib_ref] Perioperative management of special populations: immunocompromised host (cancer, HIV, transplantation), Desai [/bib_ref] [bib_ref] HIV/AIDS pandemic and surgical practice in a Nigerian teaching hospital, Owotade [/bib_ref]. Currently no data exist on the risk of morbidity and mortality postsurgery in such children. The aim of this study was to investigate the mortality and post-operative complications in HIVe children compared to both HIVn and HIVi children after a general surgical procedure. regional and national referral centre with approximately 2400 operations performed annually. The study was approved by the ethics committee of the Faculty of Health sciences, University of Cape Town.
The study site is a high HIV prevalence area, with an estimated HIV prevalence amongst pregnant women of approximately 16%. There is a well-developed PMTCT program with HIV transmission rates of approximately 2-4%. The Western Cape has approximately 70000 births per annum [bib_ref] Addressing the paediatric HIV epidemic: a perspective from the Western Cape Region..., Eley [/bib_ref] , of which approximately 11000 babies per year are expected to be HIVe.
Inclusion criteria were children less than 60 months of age, undergoing a general surgical procedure. The recruitment was conducted in two phases. 1) Phase 1 (pilot data), enrolled only HIV-exposed children, from July 2004 to December 2006. This was done to gather pilot data, to inform the larger phase 2 study. 2) Phase 2 enrolled all children irrespective of HIV exposure from January 2007 until July 2008 inclusive. Analysis was conducted on both phase 1 and phase 2 data.
Age definitive HIV testing with pre and post-test counselling was done in children whose HIV status was unknown. HIV infection was defined as 2 positive HIV ELISA (Determine ® Abbott, Abbott Park, Ill. USA) tests in children > 18 months; a positive ELISA test, and if positive a confirmatory PCR test in children less than 18 months. HIV exposure was defined as a positive ELISA but negative PCR in children < 18 months or in children > 18 months knowledge of an infected mother but a negative ELISA in the child.
Informed consent from a parent or legal guardian was obtained. Details of the child's demographics, clinical examination, medication, haematological tests, HIV status and treatment, surgery and complications were recorded until discharge, death or in the case of prolonged admissions 60 days following their last surgical procedure or surgical complication.
Children formed 3 groups-1. HIV-exposed (HIVe) but uninfected 2.HIV unexposed (HIVn), 3. HIV infected (HIVi). HIVe children were compared to the HIVn and the HIVi groups A procedure entailed a general surgical operation under general anaesthetic. Surgical procedures were defined as emergency (which in the opinion of the treating clinician unless performed in 24 hours of hours of admission would result in either morbidity or mortality of the child) or elective. Procedures were also defined as major (which entailed entry into a body cavity (abdomen or thorax), oncological resection, and a prolonged procedure longer than 90 min or vascular reconstruction) or minor. A contaminated case was defined as pus, infection, or gross contamination present at the site of surgery when performing the procedure. Prophylactic antibiotics were used in major cases or contaminated cases according to local microbiological guidelines. This included a cephalosporin and metronidazole or alternatively, penicillin an aminoglycoside and metronidazole, depending on the procedure.
The primary outcome was the occurrence of a complication, defined as an adverse event occurring in the intra, or post-operative period which is not expected to occur either during or following the normal course of the procedure. Secondary outcomes were length of stay (calculated in days from the day of admission until the day of discharge or death) and mortality.
# Statistical analysis
Descriptive data is presented as means +/-standard deviation, medians with an interquartile range (IQR) or proportions with a 95% confidence interval. Normality of data was tested using the Shapiro-Wilks test. Hypothesis testing, using either the chi squared or fisher exact test for dichotomous or ordinal variables was done. The student t-test was used for the comparison of means and the Wilcoxon rank-sum test for comparing medians of continuous variables. Nutrition assessed using weight for height Z scores as a continuous variable and a dichotomous variable with a weight for age Z-Score of less than -2 being malnourished. Relative risks were calculated for complications or mortality. Following the initial comparisons, age adjusted control groups were formed in the HIVn and HIVi groups due to the significantly younger age of the HIVe children. This was done by selecting appropriate matched controls within one month of age for each HIVe case. Due to the sample size, 3 age matched controls could be selected from the HIVn group, and a single age matched control from the HIVi group. Furthermore age as a confounding variable was tested on multivariate analysis in three different models, the entire cohort, HIVe and HIVn children and HIVe and HIVi children. The regression models were constructed for complications. This was done by including factors significant on univariate analysis or alternatively of strong clinical importance as a backward step regression until the best fit was obtained. Data was analysed using Stata (Version10.0, College Station, Tx USA). Significance was set at p < 0.05 and results were expressed as proportions and relative risk with 95% confidence intervals.
# Results
Three hundred and eighty children were enrolled (94 during, phase 1); 4 died and 11 were lost to follow up prior to PCR testing, thus 365 children were included. Of these, 38(10.4%) were HIVe, 245(67.1%) were HIVn and 82(22.5%) were HIVi children.
The median age of children was 11 (IQR 4.4-26.6) months. HIVe children were younger, 4.5 (IQR 0.8-11.2) months than the HIVi (11.5 months (IQR 6-24) p < 0.0001) or HIVn (12.7 months (IQR 4.4-29.4) p = < 0.0001) children. HIVe children had a higher number of major procedures, contamination at the surgical site and worse nutrition compared to HIVn children (table 1). Following age matching of HIVe and HIVn children, an increased surgical site contamination and worse nutrition persisted in HIVe children [fig_ref] Table 2: Comparison HIV exposed uninfected children and age matched HIV unexposed children HIVe... [/fig_ref]. HIVe and HIVi children were well matched for the level of procedure, surgical site contamination, urgency, nutrition and number of urgent procedures in both the age matched and unmatched groups [fig_ref] Table 4: Comparison HIV exposed uninfected and age matched HIV infected children [/fig_ref] The overall mortality (8 deaths, 2.2%) was low, with 2 (5.2%) deaths in the HIVe group, none in the HIVn group and 6(7.3%) in the HIVi group. The mortality rate was significantly higher in HIVe compared to HIVn children (p = 0.0003). This difference persisted when matching for age p = 0.01 [fig_ref] Table 2: Comparison HIV exposed uninfected children and age matched HIV unexposed children HIVe... [/fig_ref]. No significant difference in mortality was demonstrated between HIVe and HIVi children [fig_ref] Table 4: Comparison HIV exposed uninfected and age matched HIV infected children [/fig_ref].
HIVe had a longer duration of hospitalisation than HIVn children (3 (2-7) vs. 2 (1-4) days p = 0.02), but this difference did not persist with age matching of the groups ( (3 (2-7)
# Discussion
This study has shown that HIVe children undergoing surgery have an intermediate risk of complications postsurgery; higher than that of HIVn children, but lower than HIV infected HIVi children. In addition, in hospital mortality post-surgery was higher in HIVe children compared to that in HIVn children.
The increased morbidity and mortality related to HIV exposure may be due to many factors. HIVe children may have poorer growth and nutrition, increased risk of infection, impaired passive immunity from maternal antibodies, or greater exposure to potential pathogens in a household with a HIV infected adult.
HIVe children had worse nutrition compared to those who were HIVn. Reasons for poorer nutrition may include parental illness and poverty [bib_ref] Current issues in prevention of mother-to-child transmission of HIV-1, Coovadia [/bib_ref] [bib_ref] The HIV-exposed, uninfected African child, Filteau [/bib_ref] , infant feeding practices [bib_ref] Breastfeeding, HIV status and weights in South African children: a comparison of..., Patel [/bib_ref] [bib_ref] Current issues in prevention of mother-to-child transmission of HIV-1, Coovadia [/bib_ref] , lack of breast feeding, and an increase disease burden in exposed children [bib_ref] Infectious disease morbidity among young HIV-1-exposed but uninfected infants in Latin American..., Mussi-Pinhata [/bib_ref] [bib_ref] Risk factors for postnatal mother-child transmission of HIV-1, Embree [/bib_ref] [bib_ref] Breastfeeding, HIV status and weights in South African children: a comparison of..., Patel [/bib_ref] [bib_ref] The HIV-exposed, uninfected African child, Filteau [/bib_ref]. Although breast feeding rates were not recorded, it is likely that there would have been a low breast feeding Comparison of HIV exposed uninfected and HIV infected children rates in keeping with the PMTCT policy, which was formula feeding of children born to HIV-infected mothers in the Western Cape Province at the time of the study. Several cohort studies have reported HIVe children to have worse nutrition compared to HIVn children [bib_ref] The HIV-exposed, uninfected African child, Filteau [/bib_ref]. HIVe children have been reported to have an increased risk of infection compared to HIVn children [bib_ref] Severe infections in HIV-exposed uninfected infants: clinical evidence of immunodeficiency, Slogrove [/bib_ref] [bib_ref] Infectious disease morbidity among young HIV-1-exposed but uninfected infants in Latin American..., Mussi-Pinhata [/bib_ref] [bib_ref] Effect of age, polymicrobial disease, and maternal HIV status on treatment response..., Mcnally [/bib_ref] [bib_ref] The HIV-exposed, uninfected African child, Filteau [/bib_ref] [bib_ref] Patterns of postnatal growth in HIVinfected and HIV-exposed children, Isanaka [/bib_ref] [bib_ref] Strong HIV-1-specific T cell responses in HIV-1-exposed uninfected infants and neonates revealed..., Legrand [/bib_ref] [bib_ref] Survival from 9 months of age among HIV-infected and uninfected Zambian children..., Sutcliffe [/bib_ref]. Infections in HIVe also tend to be more severe than in HIVn children [bib_ref] Infectious disease morbidity among young HIV-1-exposed but uninfected infants in Latin American..., Mussi-Pinhata [/bib_ref] [bib_ref] Effect of age, polymicrobial disease, and maternal HIV status on treatment response..., Mcnally [/bib_ref] [bib_ref] The HIV-exposed, uninfected African child, Filteau [/bib_ref] [bib_ref] Pneumocystis carinii pneumonia in infants who were exposed to human immunodeficiency virus..., Heresi [/bib_ref] [bib_ref] Increased severe anemia in HIV-1-exposed and HIV-1-positive infants and children during acute..., Otieno [/bib_ref]. Many of these infections are opportunistic infections occurring in an immune-compromised host [bib_ref] Severe infections in HIV-exposed uninfected infants: clinical evidence of immunodeficiency, Slogrove [/bib_ref] [bib_ref] Effect of age, polymicrobial disease, and maternal HIV status on treatment response..., Mcnally [/bib_ref] [bib_ref] The HIV-exposed, uninfected African child, Filteau [/bib_ref] [bib_ref] Pneumocystis carinii pneumonia in infants who were exposed to human immunodeficiency virus..., Heresi [/bib_ref] ]. There appears that in addition to environmental and feeding practices HIVe children have an impaired innate immunity [bib_ref] Altered immunological reactivity in HIV-1-exposed uninfected neonates, Hygino [/bib_ref] [bib_ref] The HIV-exposed, uninfected African child, Filteau [/bib_ref]. This may increase susceptibility to infection and increase the risk for developing post-operative complications. This was consistent with our findings of a high rate of infections in HIVe children post operatively.
Major procedures are a clearly identified risk factor associated with an increased risk for the development of complicationsand risk stratification to account for complications is well described [bib_ref] Child mortality and HIV infection in Africa: a review, Newell [/bib_ref] [bib_ref] HIV and mortality of mothers and children: evidence from cohort studies in..., Zaba [/bib_ref]. Increased complexity of the procedure (thus increasing the risk of technical problems), immune dysregulation related to the stress response to surgery or impaired respiratory function in the post-operative period may all contribute to this [bib_ref] Metabolic response to surgery in infants and children, Mchoney [/bib_ref].
There are several limitations to this study. The relatively lower age of the HIVe compared with HIVi and HIVn patients may have been a confounding variable as the HIVe children may not have had a fully developed immune system. We have attempted to account for this in multivariate analysis and secondly by age matching to HIVi and HIVn children. In both instances when age was accounted for, HIV status had an important effect on post-operative outcomes with HIVe children having a risk in between that of HIVn and HIVi children. This study has a small number of HIVe children which may have influenced the results and larger studies are needed with stratification of the HIVe group based on age of the child, feeding practice and the stage of maternal HIV disease. Lastly as this was an observational cohort study HIV results were not blinded to either treating physicians or observers and this may have resulted in some reporting bias.
# Conclusion
This is the first study to report that HIVe children undergoing surgery have a higher risk of developing complications and mortality compared to HIVn children, but a lower risk than that of HIVi children. This should be considered when assessing the risk benefit ratio and age of any surgical procedure.
# Acknowledgements and funding
The funds received from the regional authority of Lombardia were a donation to assist in the care and research of children with HIV and carries no restrictions in study design, collection, analysis, interpretation or publication of data. The funds from the Colleges of Medicine of South Africa are an academic award to the first author to use for any research he
[table] Table 2: Comparison HIV exposed uninfected children and age matched HIV unexposed children HIVe and HIVi children included significant variables on univariate analysis, and other clinically important variables such as nutrition, urgency of surgery and surgical site contamination. Following backward step regression for HIVe and HIVn children it was demonstrated that HIV exposure and undergoing a major procedure were the most predictive of developing a complication(table 6). Models for HIVe and HIVi children demonstrated that HIV status, age under 12 months and undergoing a major procedure were the most predictive of developing a complication(table 7). [/table]
[table] Table 4: Comparison HIV exposed uninfected and age matched HIV infected children [/table]
[table] Table 5: Comparison of post-operative complications in HIV infected, HIV unexposed and HIV exposed uninfected children [/table]
[table] Table 6: Logistic regression for complications with HIV exposed uninfected children and HIV unexposed children [/table]
[table] Table 7: Logistic regression for complications with HIV exposed uninfected children and HIV infected children [/table]
|
Grade 3/4 Adverse Event Costs of Immuno-oncology Combination Therapies for Previously Untreated Advanced Renal Cell Carcinoma
Background: Despite 4 approved combination regimens in the first-line setting for advanced renal cell carcinoma (aRCC), adverse event (AE) costs data are lacking.Materials and Methods:A descriptive analysis on 2 AE cost comparisons was conducted using patient-level data for the nivolumab-based therapies and published data for the pembrolizumab-based therapies. First, grade 3/4 AE costs were compared between nivolumab + ipilimumab vs. nivolumab + cabozantinib vs. pembrolizumab + axitinib using data from the CheckMate 214 (median follow-up [mFU]: 13.1 months), CheckMate 9ER (mFU: 12.8 months), and KEYNOTE-426 (mFU: 12.8 months) trials, respectively. Second, grade 3/4 AE costs were compared between nivolumab + ipilimumab vs. nivolumab + cabozantinib vs. pembrolizumab + lenvatinib using data from the CheckMate 214 (mFU: 26.7 months), CheckMate 9ER (mFU: 23.5 months), and KEYNOTE-581 (mFU: 26.6 months) trials, respectively. Per-patient costs for all-cause and treatment-related grade 3/4 AEs with corresponding any-grade AE rates ≥ 20% were calculated based on the Healthcare Cost and Utilization Project database and inflated to 2020 US dollars.Results:Per-patient all-cause grade 3/4 AE costs for nivolumab + ipilimumab vs. nivolumab + cabozantinib vs. pembrolizumab + axitinib were $2703 vs. $4508 vs. $5772, and treatment-related grade 3/4 AE costs were $741 vs. $2722 vs. $4440 over ~12.8 months of FU. For nivolumab + ipilimumab vs. nivolumab + cabozantinib vs. pembrolizumab + lenvatinib, per-patient all-cause grade 3/4 AE costs were $3120 vs. $5800 vs. $9285, while treatment-related grade 3/4 AE costs were $863 vs. $3162 vs. $5030 over ~26.6 months of FU.Conclusion:Patients with aRCC treated with first-line nivolumab-based therapies had lower grade 3/4 all-cause and treatment-related AE costs than pembrolizumab-based therapies, suggesting a more favorable cost-benefit profile.There is currently limited evidence on the economic benefits and risks associated with novel immunotherapy-based combinations used to treat patients newly diagnosed with advanced or metastatic renal cell carcinoma (aRCC). The present study addresses this gap in literature by comparing all-cause and treatment-related grade 3/4 adverse event costs of nivolumab-based and pembrolizumab-based combinations as first-line treatments for patients with aRCC using data from relevant clinical trials. The study results suggest that nivolumab-based combinations have a more favorable cost-benefit profile and offer clinicians and payers a therapeutic option that may reduce the substantial clinical and economic impacts for previously untreated patients with aRCC.
# Introduction
Renal cell carcinoma (RCC) is the most common type of kidney cancer, estimated to develop in more than 76,000 new patients and lead to nearly 14,000 deaths in the US in 2021. [bib_ref] Cancer statistics, 2021, Siegel [/bib_ref] At diagnosis, approximately 25%-35% of patients have advanced or metastatic RCC (aRCC) which, until recently, was associated with a 5-year survival rate below 15%. [bib_ref] Epidemiologic and socioeconomic burden of metastatic renal cell carcinoma (mRCC): a literature..., Gupta [/bib_ref] [bib_ref] Real-world treatment patterns and adverse events in metastatic renal cell carcinoma from..., Pal [/bib_ref] Targeted therapies, such as the tyrosine kinase inhibitor (TKI) sunitinib, had been the standard of care for patients with previously untreated aRCC until recently. [bib_ref] Sunitinib versus interferon alfa in metastatic renal-cell carcinoma, Motzer [/bib_ref] In the past 5 years, novel therapeutic options, including immunotherapy-based combination therapies, emerged for this population offering better disease control and improved survival outcomes. [bib_ref] Nivolumab plus ipilimumab versus sunitinib in advanced renal-cell carcinoma, Motzer [/bib_ref] [bib_ref] Nivolumab plus cabozantinib versus sunitinib for advanced renal-cell carcinoma, Choueiri [/bib_ref] [bib_ref] Pembrolizumab plus axitinib versus sunitinib for advanced renal-cell carcinoma, Rini [/bib_ref] [bib_ref] Lenvatinib plus pembrolizumab or everolimus for advanced renal cell carcinoma, Motzer [/bib_ref] In April 2018, nivolumab plus ipilimumab was the first immunotherapy-based combination to be approved by the FDA for the first-line treatment of intermediate/poor-risk aRCC based on results of the CheckMate 214 phase III clinical trial (NCT02231749).Among all treated patients, 65% of those who received nivolumab plus ipilimumab had an all-cause grade 3/4 adverse event (AE) compared with 76% of the patients treated with sunitinib.Pembrolizumab plus axitinib was approved in April 2019 for the first-line treatment of patients with aRCC based on evidence from the KEYNOTE-426 trial (NCT02853331). In this trial, allcause grade 3 or higher AEs occurred in 76% of patients in the pembrolizumab plus axitinib arm and 71% of patients in the sunitinib arm. [bib_ref] Pembrolizumab plus axitinib versus sunitinib for advanced renal-cell carcinoma, Rini [/bib_ref] The combination of nivolumab plus cabozantinib was granted FDA approval in January 2021 for treatment-naïve patients with aRCC in any-risk group based on the results of the CheckMate 9ER trial (NCT03141177)wherein allcause grade 3 or higher AEs occurred in 75% of patients in the nivolumab plus cabozantinib arm and 71% of patients in the sunitinib arm. [bib_ref] Nivolumab plus cabozantinib versus sunitinib for advanced renal-cell carcinoma, Choueiri [/bib_ref] In August 2021, pembrolizumab plus lenvatinib was approved for the same indication based on the results of the KEYNOTE-581 trial, also known as the CLEAR trial (NCT02811861).Compared with sunitinib, treatment with pembrolizumab plus lenvatinib was associated with 82% vs. 72% all-cause grade 3 or higher AEs. [bib_ref] Lenvatinib plus pembrolizumab or everolimus for advanced renal cell carcinoma, Motzer [/bib_ref] While the efficacy and safety of these novel immunotherapy-based combinations has been evaluated, there is currently a paucity of evidence on the economic benefits and risks associated with these therapies. Given the marked impact of financial toxicity on patients with cancer in the US,this information can be valuable to healthcare decision-makers and payers. Therefore, the objective of this study was to compare descriptive analyses of all-cause and treatment-related grade 3/4 AE costs of nivolumab plus ipilimumab, nivolumab plus cabozantinib, pembrolizumab plus axitinib, and pembrolizumab plus lenvatinib as first-line treatments for patients with aRCC, using individual patient-level data (IPD) from the CheckMate 214 and CheckMate 9ER trials, as well as published data from the KEYNOTE-426 and KEYNOTE-581 trials.
# Materials and methods
## Data sources
Due to the different follow-up times from the KEYNOTE-426 and KEYNOTE-581 (CLEAR) trials for the adverse events reported in the KEYTRUDA prescribing information (median follow-up: 13.1 and 26.6 months, respectively), [bib_ref] Lenvatinib plus pembrolizumab or everolimus for advanced renal cell carcinoma, Motzer [/bib_ref] were used to obtain the all-cause and treatment-related grade 3/4 AE rates, respectively, for all treated patients.
Unit costs for grade 3/4 AEs were obtained from the US 2017 Healthcare Cost and Utilization Project (HCUP) National Inpatient Database, which are estimated from the US payer perspective. The HCUP database is a family of healthcare databases with a national information resource of encounter-level healthcare data (HCUP Partners) which is developed through a Federal-State-Industry partnership and sponsored by the Agency for Healthcare Research and Quality (AHRQ). 18-20 As patients with grade 3/4 AEs require hospitalization based on the definitions by Common Terminology Criteria for Adverse Events (CTCAE), the national level of inpatient costs from the HCUP database were used to estimate the unit costs associated with each grade 3/4 AE. All costs were inflated to 2020 US dollars (USD) using an inflation factor of 1.0793 from 2017 USD based on the Consumer Price Index (CPI) for all urban consumers in medical care service.
## Study design
The analyses focused on all treated patients with clear cell aRCC, which represented the study population for all 4 clinical trials. All-cause and treatment-related grade 3/4 AEs with corresponding any grade AEs that occurred in at least 20% of patients in the nivolumab plus ipilimumab arm (CheckMate 214), nivolumab plus cabozantinib arm (CheckMate 9ER), pembrolizumab plus axitinib arm (KEYNOTE-426), and pembrolizumab plus lenvatinib arm (KEYNOTE-581) were considered in the cost calculation. The analysis assumed that each all-cause and treatment-related grade 3/4 AE occurred only once per patient treated in each arm of each trial since the KEYTRUDA prescribing information,Rini et al 9 (for pembrolizumab plus axitinib), and Motzer et al 10 (for pembrolizumab plus lenvatinib) only reported the percentage of patients experiencing each AE.
To be consistent with the reporting criteria of the KEYNOTE-426 trial, the analysis comparing nivolumab plus ipilimumab vs. nivolumab plus cabozantinib vs. pembrolizumab plus axitinib using IPD from CheckMate 214 and Checkmate 9ER included all AEs that occurred while patients received treatment and within 30 days after the end of the trial treatment period or within 90 days after the end of the trial treatment period for serious AEs. Likewise, to be consistent with the reporting criteria of the KEYNOTE-581 trial, the analysis comparing nivolumab plus ipilimumab vs. nivolumab plus cabozantinib vs. pembrolizumab plus lenvatinib using IPD from Checkmate 214 and Checkmate 9ER considered all AEs (including serious AEs) that occurred while patients received treatment or within 30 days after the end of the trial treatment period. To ensure a fair comparison, all-cause grade 3/4 AEs from the IPD of the CheckMate 214 and CheckMate 9ER trials were recategorized according to the definitions the KEYTRUDA prescribing information (Supplementary Tables S1 and S2).
# Statistical analysis
The cost associated with each grade 3/4 AE was calculated by multiplying the AE rate by the respective unit AE cost. The total per-patient costs of all-cause and treatment-related grade 3/4 AEs were then calculated by summing the costs of all relevant AEs and were descriptively reported for each treatment arm. The top cost drivers of each treatment arm were described. All analyses were performed using SAS 9.4 (SAS Institute, Inc., 100 SAS Campus Drive, Cary, NC 27513, USA).
# Results
Nivolumab Plus Ipilimumab Vs. Nivolumab Plus Cabozantinib Vs. Pembrolizumab Plus Axitinib (Median Follow-Up of ~12.8 Months)
## All-cause grade 3/4 ae costs
The average all-cause grade 3/4 AE costs for patients with aRCC who received nivolumab plus ipilimumab vs. nivolumab plus cabozantinib vs. pembrolizumab plus axitinib are presented in [fig_ref] Figure 1: Total all-cause and treatment-related grade 3/4 AE costs a [/fig_ref] and [fig_ref] Table 1: All-cause grade 3/4 AE costs associated with nivolumab plus ipilimumab, nivolumab plus... [/fig_ref]. Patients treated with pembrolizumab plus axitinib incurred the highest allcause grade 3/4 AE costs per patient ($5772). Those who received nivolumab plus cabozantinib incurred $4508 per patient (22% lower than pembrolizumab plus axitinib), while patients who received nivolumab plus ipilimumab incurred $2703 per patient (53% lower than pembrolizumab plus axitinib).
In each treatment arm, the top 5 AEs with the highest costs contributed to the majority of the total all-cause grade 3/4 AE cost. Hepatotoxicity, hypertension, diarrhea, fatigue/asthenia, and palmar-plantar erythrodysesthesia (PPE) syndrome accounted for 88% of the total all-cause grade 3/4 AE cost related to pembrolizumab plus axitinib. In the nivolumab plus cabozantinib arm, the same AEs had the highest allcause grade 3/4 AE costs and accounted for 80% of the total cost. For nivolumab plus ipilimumab, the top 5 AE cost drivers were fatigue/asthenia, hepatoxicity, diarrhea, rash, and decreased appetite, which accounted for 86% of the total all-cause grade 3/4 AE cost [fig_ref] Table 1: All-cause grade 3/4 AE costs associated with nivolumab plus ipilimumab, nivolumab plus... [/fig_ref].
## Treatment-related grade 3/4 ae costs
The average treatment-related grade 3/4 AE costs for aRCC patients who received nivolumab plus ipilimumab vs. nivolumab plus cabozantinib vs. pembrolizumab plus axitinib are presented in [fig_ref] Figure 1: Total all-cause and treatment-related grade 3/4 AE costs a [/fig_ref] and [fig_ref] Table 2 .: Nivolumab Plus Ipilimumab Vs [/fig_ref]. Patients treated with pembrolizumab plus axitinib incurred the highest treatment-related grade 3/4 AE costs per patient ($4440). Those who received nivolumab plus cabozantinib incurred $2722 in treatment-related grade 3/4 AE costs per patient (39% lower than pembrolizumab plus axitinib), while patients who received nivolumab plus ipilimumab incurred $741 in treatment-related grade 3/4 AE costs per patient (83% lower than pembrolizumab plus axitinib).
In each treatment arm, the top AEs with the highest costs contributed to a majority of the total treatment-related grade 3/4 AE cost. Hypertension, increased alanine aminotransferase (ALT), increased aspartate aminotransferase (AST), diarrhea, and PPE syndrome accounted for 90% of the total treatment-related grade 3/4 AE cost associated with pembrolizumab plus axitinib. In the nivolumab plus cabozantinib arm, the same AEs had the highest treatment-related grade 3/4 AE costs and accounted for 86% of the total cost. For nivolumab plus ipilimumab, fatigue, diarrhea, rash, and pruritus contributed to 100% of the total treatment-related grade 3/4 AE cost (
## All-cause grade 3/4 ae costs
The average all-cause grade 3/4 AE costs for patients with aRCC who received nivolumab plus ipilimumab vs. nivolumab plus cabozantinib vs. pembrolizumab plus lenvatinib are presented in [fig_ref] Figure 1: Total all-cause and treatment-related grade 3/4 AE costs a [/fig_ref] and [fig_ref] Table 3: Notes [/fig_ref]. Patients treated with pembrolizumab plus lenvatinib incurred $9285 in all-cause grade 3/4 AE costs per patient. Those who received nivolumab plus cabozantinib incurred $5800 per patient (38% lower than pembrolizumab plus lenvatinib), while patients who received nivolumab plus ipilimumab incurred $3120 per patient (66% lower than pembrolizumab plus lenvatinib).
In each treatment arm, the top 5 AEs with the highest costs contributed to a majority of the total all-cause grade 3/4 AE cost. Hypertension, fatigue, weight loss, diarrhea, and hepatotoxicity accounted for 58% of the total all-cause grade 3/4 AE cost related to pembrolizumab plus lenvatinib. In the nivolumab plus cabozantinib arm, hypertension, hepatotoxicity, fatigue, diarrhea, and stomatitis accounted for 69% of the total all-cause grade 3/4 AE cost. For nivolumab plus ipilimumab, the main drivers were fatigue, hepatotoxicity, diarrhea, musculoskeletal disorders, and rash, which accounted for 83% of the total all-cause grade 3/4 AE cost [fig_ref] Table 3: Notes [/fig_ref].
## Treatment-related grade 3/4 ae costs
The average treatment-related grade 3/4 AE costs for patients with aRCC who received nivolumab plus ipilimumab vs. nivolumab plus cabozantinib vs. pembrolizumab plus lenvatinib are presented in [fig_ref] Figure 1: Total all-cause and treatment-related grade 3/4 AE costs a [/fig_ref] and [fig_ref] Table 4: Treatment-related grade 3/4 AE costs associated nivolumab plus ipilimumab, nivolumab plus cabozantinib,... [/fig_ref]. Patients treated with pembrolizumab plus lenvatinib incurred treatment-related grade 3/4 AE costs of $5030 per patient; those who received nivolumab plus cabozantinib incurred $3162 in treatment-related grade 3/4 AE costs per patient (37% lower than pembrolizumab plus lenvatinib) while patients who received nivolumab plus ipilimumab incurred only $863 in costs per patient (83% lower than pembrolizumab plus lenvatinib).
In each treatment arm, the top 5 AEs with the highest costs contributed to a majority or all of the total treatment-related grade 3/4 AE cost. Hypertension, diarrhea, asthenia, proteinuria, and decreased appetite accounted for 78% of the total treatment-related grade 3/4 AE cost associated with pembrolizumab plus lenvatinib. In the nivolumab plus cabozantinib arm, hypertension, diarrhea, PPE syndrome, increased ALT, and increased AST accounted for 80% of the total treatment-related grade 3/4 AE cost. For nivolumab plus ipilimumab, fatigue, diarrhea, nausea, rash, and pruritus accounted for 100% of the total treatment-related grade 3/4 AE cost [fig_ref] Table 4: Treatment-related grade 3/4 AE costs associated nivolumab plus ipilimumab, nivolumab plus cabozantinib,... [/fig_ref].
# Discussion
A solid understanding of the safety profile of available immunotherapy-based combinations is a crucial component of treatment optimization,allowing clinicians to maximize the benefit of novel first-line treatments for aRCC while properly managing the associated AEs and reducing ensuing costs. In the absence of head-to-head trials, the present study characterized the safety profiles of 4 novel immunotherapy-based combinations in patients with previously untreated aRCC and evaluated the costs associated with the all-cause and treatment-related grade 3/4 AEs. The nivolumab-based combinations were associated with lower average all-cause and treatment-related grade 3/4 AE costs per treated patient compared with the pembrolizumab-based combinations. The all-cause AE cost drivers (≥$500) of nivolumab-based combinations are fatigue/asthenia, hepatotoxicity, and diarrhea ($410 in the second comparison) for nivolumab plus ipilimumab, and hypertension, hepatotoxicity, fatigue/asthenia, diarrhea, and stomatitis/mucosal inflammation ($426 in the first comparison) for nivolumab plus cabozantinib. The all-cause AE cost drivers (≥$500) of pembrolizumab-based combinations are hepatotoxicity, hypertension, diarrhea, and fatigue/asthenia for pembrolizumab plus axitinib, and hypertension, fatigue, weight loss, diarrhea, hepatoxicity, and proteinuria for pembrolizumab plus lenvatinib.
The results show that the use of nivolumab-based combinations in first-line aRCC treatment leads to numerically lower grade 3/4 AE costs relative to pembrolizumab-based combinations. These findings align with the results of a prior cost-effectiveness analysis that compared the AE costs of nivolumab plus ipilimumab and pembrolizumab plus axitinib among patients from the all treated population in the intermediate/poor-risk subgroup, as defined by the International Metastatic Renal Cell Carcinoma Database Consortium. [bib_ref] Cost-effectiveness of pembrolizumab with axitinib as first-line treatment for advanced renal cell..., Bensimon [/bib_ref] The AE cost estimate (based on grade 3 or higher AEs reported in ≥ 5.0% of patients of the respective pivotal trials) for nivolumab plus ipilimumab ($1151) was lower than the estimate for pembrolizumab plus axitinib ($3842). Together, the results provide valuable insight regarding the potential for cost savings associated with nivolumab-based combinations over pembrolizumab-based combinations, which can guide treatment decision making.
The top AE cost drivers varied across different immunotherapy-based combinations despite similarities. It is worth noting that some AEs such as fatigue, hypertension, and diarrhea, are known class effects of VEGF-targeted TKIs, [bib_ref] Axitinib induces and aggravates hypertension regardless of prior treatment with tyrosine kinase..., Kadowaki [/bib_ref] [bib_ref] Management of adverse events associated with cabozantinib therapy in renal cell carcinoma, Schmidinger [/bib_ref] [bib_ref] Managing the adverse events associated with lenvatinib therapy in radioiodine-refractory differentiated thyroid..., Cabanillas [/bib_ref] Some others are often difficult to relate to any of the 2 classes of agents. [bib_ref] Hepatotoxicity of immune checkpoint inhibitors: an evolving picture of risk associated with..., Suzman [/bib_ref] [bib_ref] Risk of elevated transaminases in cancer patients treated with immune checkpoint inhibitors:..., Abdel-Rahman [/bib_ref] Our results show that costs related to treatment-related hypertension and increased ALT and AST and diarrhea in the pembrolizumab plus axitinib arm drive the cost difference between this treatment arm and either of the nivolumab-based combinations. This may echo the findings of a recent network meta-analysis which showed that the combination of nivolumab plus ipilimumab was associated with lower rates of serious AEs than pembrolizumab plus axitinib. [bib_ref] Systemic therapy for metastatic renal cell carcinoma in the first-line setting: a..., Mori [/bib_ref] Given that severe AEs can lead to treatment discontinuation, additional health resource utilization, poorer clinical outcomes, and impact the quality of life in this vulnerable population, 32-34 the different patterns of toxicities related to the use of immunotherapy-based combinations and the resulting clinical and economic impact for the patients, healthcare providers, and payers, must be considered when selecting the appropriate treatment strategy.
# Limitations
Results of this study should be considered along with certain limitations. First, the current analysis focused on costs associated with grade 3/4 AEs since these AEs are more prevalent, require medical treatment or hospitalization, and are more consistently reported in product prescribing information and the literature. Costs associated with grades 1, 2, and 5 AEs were not considered. As a result, the present analysis may slightly underestimate the total AE costs for the study cohorts. Further studies assessing the impact of grades 1, 2, and 5 AEs are warranted to supplement the findings from this analysis. Second, the unit costs for grade 3/4 AEs were consistently assumed to be the unit cost of one hospitalization with the corresponding diagnosis and were obtained from the 2017 US HCUP database. As some AEs, such as hypertension or PPE, may be less likely to require a hospitalization, the unit costs may not reflect the true costs incurred during the trial and may be subject to measurement errors which could result in either an underestimation or overestimation of costs. However, since this assumption was applied to all therapies, the impact on the comparative results between different therapies is limited. Third, the costs associated with other allcause and treatment-related grade 3/4 AEs with corresponding any-grade AE rates less than 20% were not included in the analysis due to lack of such data for pembrolizumab plus axitinib in the KEYTRUDA prescribing information, although notably, nivolumab plus ipilimumab combination may be associated with rare but serious AEs, such as myocarditis, [bib_ref] Myocarditis following the use of different immune checkpoint inhibitor regimens: a real-world..., Fan [/bib_ref] warranting further investigation. Fourth, indirect costs were, per definition, excluded from these analyses. Thus, further real-world studies are warranted to comprehensively assess these AE costs. Fifth, statistical significance testing was not conducted since aggregated trial-level data were used for pembrolizumab-based therapies. Finally, the study results were based on the AE profile of the patients in the CheckMate 214, CheckMate 9ER, KEYNOTE-426, and KEYNOTE-581 trials, and therefore, may not be generalizable to patients in real-world settings. Further studies with real-world data are warranted to compare the clinical and economic outcomes of these treatments.
# Conclusions
In this trial-based descriptive economic assessment of the allcause and treatment-related AE costs of novel combination therapies for first-line treatment for patients with aRCC, the combinations of nivolumab plus ipilimumab and nivolumab plus cabozantinib were associated with lower grade 3/4 AE costs than pembrolizumab plus axitinib and pembrolizumab plus lenvatinib. For nivolumab plus ipilimumab, the fact that ipilimumab is stopped after the induction phase may contribute to lower observed AE costs. These results suggest that nivolumab-based combinations have a more favorable cost-benefit profile and offer clinicians and payers a therapeutic option for previously untreated aRCC that may reduce the substantial clinical and economic impacts in this population.
[fig] Figure 1: Total all-cause and treatment-related grade 3/4 AE costs a . Abbreviations: AE: adverse event; ALT: alanine aminotransferase; AST: aspartate aminotransferase; NIVO + CABO: nivolumab plus cabozantinib; NIVO + IPI: nivolumab plus ipilimumab; PEM + AXI: pembrolizumab plus axitinib; PEM + LEN: pembrolizumab plus lenvatinib; PPE: palmar-plantar erythrodysesthesia; USD: United States dollar. Note: [a] Grade 3/4 AEs with a corresponding any grade AE rate <20% were not included in the cost assessment and are not included in the figures. [/fig]
[table] 10: 2 distinct comparisons were conducted in order to ensure comparability in the follow-up time across the CheckMate and KEYNOTE trials. For the comparison of nivolumab plus ipilimumab (N = 547) vs. nivolumab plus cabozantinib (N = 320) vs. pembrolizumab plus axitinib (N = 429), IPD from the CheckMate 214 trial (data cutoff: August 31, 2016; median follow-up: 13.1 months) and the CheckMate 9ER trial (data cutoff: November 30, 2019; median follow-up, 12.8 months) for all treated patients were used to assess all-cause and treatment-related grade 3/4 AE rates associated with nivolumab plus ipilimumab and nivolumab plus cabozantinib, respectively. To obtain the all-cause and treatment-related grade 3/4 AE rates associated with pembrolizumab plus axitinib for all treated patients, published results of the KEYNOTE-426 trial (data cutoff: August 24, 2018; median follow-up: 12.8 months) from the KEYTRUDA prescribing information17 and Rini et al 9 were used. Similarly, for the comparison of nivolumab plus ipilimumab (N = 547) vs. nivolumab plus cabozantinib (N = 320) vs. pembrolizumab plus lenvatinib (N = 355), IPD from the CheckMate 214 trial (data cutoff: November 30, 2017; median follow-up: 26.7 months) and the CheckMate 9ER trial (data cutoff: September 10, 2020; median follow-up: 23.5 months) for all treated patients were used to obtain the all-cause and treatment-related grade 3/4 AE rates associated with nivolumab plus ipilimumab and nivolumab plus cabozantinib, respectively. For pembrolizumab plus lenvatinib, published results of the KEYNOTE-581 trial (data cutoff: August 28, 2020; median follow-up: 26.6 months) from the KEYTRUDA prescribing information 17 and Motzer et al [/table]
[table] Table 2 .: Nivolumab Plus Ipilimumab Vs. Nivolumab Plus Cabozantinib Vs. Pembrolizumab Plus Lenvatinib (Median Follow-Up of ~26.6 Months) [/table]
[table] Table 1: All-cause grade 3/4 AE costs associated with nivolumab plus ipilimumab, nivolumab plus cabozantinib, and pembrolizumab plus axitinib (median follow-up of ~12.8 months). a [/table]
[table] Table 2: Treatment [/table]
[table] Table 3: Notes: [a] "-" represents an AE that was not experienced in any grade by ≥ 20% of patients treated with nivolumab plus ipilimumab in the CheckMate 214 trial, patients treated with nivolumab plus cabozantinib in the CheckMate 9ER trial, or not reported among patients treated with pembrolizumab plus lenvatinib in the KEYNOTE-581 trial per the KEYTRUDA prescribing information.[b] AE unit costs were based on AE-related hospitalization cost, which were estimated from the Healthcare Cost and Utilization Project (HCUP) 2017 National Inpatient Database based on the ICD-10 diagnosis codes associated with each AE. Cost inputs were inflated from 2017 USD to 2020 USD using an inflation factor of 1.0793 based on the Consumer Price Index (CPI) for all urban consumers in medical care service.[c] Total costs were calculated by multiplying each AE rate by the corresponding AE unit cost and reflect the rounded sum of the individual AE costs in each treatment arm. Abbreviations: AE, adverse event; NIVO + CABO, nivolumab plus cabozantinib; NIVO + IPI, nivolumab plus ipilimumab; PEM + LEN, pembrolizumab plus lenvatinib; PPE, palmar-plantar erythrodysesthesia. [/table]
[table] Table 4: Treatment-related grade 3/4 AE costs associated nivolumab plus ipilimumab, nivolumab plus cabozantinib, and pembrolizumab plus lenvatinib (median follow-up of ~26.6 months). a Notes: [a] "-" represents an AE that was not experienced in any grade by ≥ 20% of patients treated with nivolumab plus ipilimumab in the CheckMate 214 trial, nivolumab plus cabozantinib in the CheckMate 9ER trial or pembrolizumab plus lenvatinib in the KEYNOTE-426 trial, or not reported among patients treated with pembrolizumab plus axitinib in the KEYNOTE-581 trial, Motzer et al.[b] AE unit costs were based on AE-related hospitalization cost, which were estimated from the Healthcare Cost and Utilization Project (HCUP) 2017 National Inpatient Database based on the ICD-10 diagnosis codes associated with each AE. Cost inputs were inflated from 2017 USD to 2020 USD using an inflation factor of 1.0793 based on the Consumer Price Index (CPI) for all urban consumers in medical care service.[c] Total costs were calculated by multiplying each AE rate by the corresponding AE unit cost and reflect the rounded sum of the individual AE costs in each treatment arm. Abbreviations: AE, adverse event; NIVO + CABO, nivolumab plus cabozantinib; NIVO + IPI, nivolumab plus ipilimumab; PEM + LEN, pembrolizumab plus lenvatinib; PPE, palmar-plantar erythrodysesthesia. [/table]
|
Novel Mutations Mapping to the Fourth Sodium Channel Domain of Nav1.7 Result in Variable Clinical Manifestations of Primary Erythromelalgia
We identified and clinically investigated two patients with primary erythromelalgia mutations (PEM), which are the first reported to map to the fourth domain of Nav1.7 (DIV). The identified mutations (A1746G and W1538R) were cloned and transfected to cell cultures followed by electrophysiological analysis in whole-cell configuration. The investigated patients presented with PEM, while age of onset was very different (3 vs. 61 years of age). Electrophysiological characterization revealed that the early onset A1746G mutation leads to a marked hyperpolarizing shift in voltage dependence of steady-state activation, larger window currents, faster activation kinetics (time-to-peak current) and recovery from steady-state inactivation compared to wild-type Nav1.7, indicating a pronounced gain-of-function. Furthermore, we found a hyperpolarizing shift in voltage dependence of slow inactivation, which is another feature commonly found in Nav1.7 mutations associated with PEM. In silico neuron simulation revealed reduced firing thresholds and increased repetitive firing, both indicating hyperexcitability. The lateonset W1538R mutation also revealed gain-of-function properties, although to a lesser extent. Our findings demonstrate that mutations encoding for DIV of Nav1.7 can not only be linked to congenital insensitivity to pain or paroxysmal extreme pain disorder but can also be causative of PEM, if voltage dependency of channel activation is affected. This supports the view that the degree of biophysical property changes caused by a mutation may have an impact on age of clinical manifestation of PEM. In summary, these findings extent the genotype-phenotype correlation profile for SCN9A and highlight a new region of Nav1.7 that is implicated in PEM.Electronic supplementary material The online version of this article (
# Introduction
Recent studies have confirmed a pivotal role for the Nav1.7 voltage-gated sodium channel in human familial gainof-function [bib_ref] Mutations in SCN9A, encoding a sodium channel alpha subunit, in patients with..., Yang [/bib_ref] [bib_ref] SCN9A mutations in paroxysmal extreme pain disorder: Allelic variants underlie distinct channel..., Fertleman [/bib_ref] and loss-of-function pain syndromes [bib_ref] An SCN9A channelopathy causes congenital inability to experience pain, Cox [/bib_ref]. Nav1.7, encoded by SCN9A, is preferentially expressed in nociceptive dorsal root ganglia (DRG) and sympathetic ganglia [bib_ref] A novel tetrodotoxin-sensitive, voltage-gated sodium channel expressed in rat and human dorsal..., Sangameswaran [/bib_ref] and is thought to serve a threshold triggering function, enabling depolarizing stimuli to elicit action potential propagation [bib_ref] Slow closedstate inactivation: A novel mechanism underlying ramp currents in cells expressing..., Cummins [/bib_ref] [bib_ref] Multiple sodium channels and their roles in electrogenesis within dorsal root ganglion..., Rush [/bib_ref]. By identifying patients with novel mutations in SCN9A and analyzing the biophysical properties of the mutant Nav1.7 channels, we aim to provide insights into how this important channel functions and contributes to neuronal action potential firing, potentially suggesting ways in which its function could be normalized in affected individuals.
Recessive loss-of-function mutations in SCN9A result in congenital insensitivity to pain (CIP), whereas gainof-function, dominant mutations lead to sensory neuronal hyperexcitability and the development of painful phenotypes described as inherited or primary erythromelalgia (PEM or IEM) [bib_ref] Mutations in SCN9A, encoding a sodium channel alpha subunit, in patients with..., Yang [/bib_ref] [bib_ref] Gain-of-function mutation in Nav1.7 in familial erythromelalgia induces bursting of sensory neurons, Dib-Hajj [/bib_ref] and paroxysmal extreme pain disorder (PEPD) [bib_ref] SCN9A mutations in paroxysmal extreme pain disorder: Allelic variants underlie distinct channel..., Fertleman [/bib_ref] [bib_ref] Paroxysmal extreme pain disorder: A molecular lesion of peripheral neurons, Choi [/bib_ref]. PEM is classically described as a peripheral, bilateral disorder associated with erythema and severe burning pain of affected extremities that is triggered by stress, exhaustion and warmth and often relieved by cooling of the affected sites [bib_ref] Imaging the neural correlates of neuropathic pain and pleasurable relief associated with..., Segerdahl [/bib_ref]. Clinical onset has been reported previously as within 1 year after birth in some cases, while in the majority of cases, symptoms of PEM develop within the first decade of life . With no medication for Nav1.7-specific sodium channel block available, options for pharmacologic treatment are limited to systemic application of non-specific sodium channel blocking agents, for example, lidocaine [bib_ref] A Nav1.7 channel mutation associated with hereditary erythromelalgia contributes to neuronal hyperexcitability..., Sheets [/bib_ref] and mexiletine [bib_ref] Mexiletine-responsive erythromelalgia due to a new Na(v)1.7 mutation showing use-dependent current fall-off, Choi [/bib_ref] , and symptomatic approaches for neuropathic pain treatment including anticonvulsants, opioids and antidepressants [bib_ref] SCN9A-related inherited erythromelalgia, Hisama [/bib_ref]. Further elucidation of the connection between single-point mutations in SCN9A encoding for Nav1.7 gain-of-function mutations, impact on biophysical channel properties, disease onset and clinical symptoms may contribute to our understanding of this debilitating disease, and enable us to interpret results from genetic testing more rationally and to develop a causative treatment.
We report a detailed clinical characterization and biophysical analysis of two novel mutations in Nav1.7, which are associated with PEM. Interestingly, while the described mutations are the first reported PE mutations to map to the fourth domain of Nav1.7, they are associated with very different clinical characteristics with one patient having a documented onset of symptoms at 3 years of age and the other patient having an extremely late age of onset at 61 years of age. These clinical differences might be explained by the observed different biophysical properties of the two mutated Nav1.7 channels.
# Materials and methods
## Standard protocol approvals, registrations and patient consents
Informed consent was given by both investigated patients for genetic testing as well as for publication of results and clinical case histories. Patient 1 was recruited as part of the IRB-approved Pain in Neuropathy Study (National Research Ethics Service Ref. 10/H07056/35), which included quantitative sensory testing (QST) on a research basis.
Quantitative Sensory Testing QST following informed consent was performed on the dorsum of the foot and the hand according to the German DFNS protocol [bib_ref] Quantitative sensory testing in the German research network on neuropathic pain (DFNS):..., Rolke [/bib_ref]. This protocol includes 13 parameters measuring temperature detection and pain thresholds, mechanical detection and pain thresholds, mechanical pain sensitivity, and signs of allodynia and wind-up ratio among others, allowing characterization of sensory profiles. The findings in the two evaluated patients were compared with a control Caucasian population by means of Z scores in which the z score represents the result of a raw score minus the mean of the population and this is further divided by the standard deviation of the population. Z scores above or below ±2 standard deviations represent hyper-/hypo-sensitivity or hyper-/hypo-algesia depending on the evaluated parameter.
## Scn9a sequencing
Genomic DNA was isolated from blood by standard methods. All coding exons and flanking splice sites of SCN9A were bidirectionally sequenced as part of clinical diagnostic investigation (Dept. Gastroenterology, Radboud University, Nijmegen Medical Centre, The Netherlands) and compared to the human genome sequence (build 37) at http://genome. ucsc.edu. In patient 1, a heterozygous c.4612t[c variant (W1538R) was found, and in patient 2, a heterozygous c.5237C[G variant (A1746G) was found. Both variants are absent from the latest dbSNP database release (dbSNP 135). Nucleotide numbering reflects cDNA numbering with ?1 corresponding to the A of the ATG translation initiation codon in the reference sequence (Ref Seq) nm_002977.3. An in silico prediction of the functional effects of the 2 nonsynonymous variants was performed using PolyPhen-2 and SIFT (http://genetics.bwh.harvard.edu/pph2/ and http://sift. jcvi.org, respectively).
## Plasmid and site-directed mutagenesis
A previously described full-length human SCN9A cDNA sequence cloned into a modified pcDNA3 expression vector containing downstream polio IRES and DsRED2 sequences (FLRED) was used [bib_ref] An SCN9A channelopathy causes congenital inability to experience pain, Cox [/bib_ref]. The patient mutations were introduced into FLRED using the QuikChange II XL Site-Directed Mutagenesis Kit (Stratagene) according to the manufacturer's instructions. The coding sequence of both constructs was fully sequenced to verify the desired mutation and to ensure the lack of other introduced variations.
## Cell culture and transfection
Human embryonic kidney cells (HEK293A) were cultured in a humidified atmosphere containing 5 % carbon dioxide at 37°C and were grown in Dulbecco's modified Eagle's medium (DMEM; Gibco, Life Technologies, Carlsbad, CA) supplemented with 10 % heat-inactivated fetal bovine serum. Unless stated otherwise, reagents were purchased from Sigma-Aldrich (St. Louis, MO).
Cells were transiently transfected with plasmid DNA for expression of Nav1.7 human wild-type a subunits (Ref Seq NM_002977), A1746G or W1538R-mutated a subunits (SCN9A-IRES-DsRed2 in pcDNA3 vector) combined with human wild-type b1 (Ref Seq NM_001037) and b2 (Ref Seq NM_001037) subunits (SCN1B-IRES-SCN2B-IRES-eGFP in a pIRES2-AcGFP1 backbone vector) as previously described [bib_ref] An SCN9A channelopathy causes congenital inability to experience pain, Cox [/bib_ref]. In brief, transient transfection was performed with cells seeded at 80-90 % confluency in 35-mm cell culture dishes using Lipofectamine 2000 (Invitrogen, Life Technologies) according to the manufacturer's recommendations. After 6 h, transfection medium was replaced with fresh culture medium and the cells were re-seeded for electrophysiological recordings at 20-30 % confluency.
## Electrophysiological recordings
Whole-cell membrane current recordings were performed 46-78 h after transfection. All recordings were made at room temperature. Micropipettes were pulled from borosilicate glass capillaries (GC150F-10; Harvard Apparatus, Kent, UK) using a Brown-Flaming P-97 horizontal micropipette puller (Sutter Instruments, Novato, CA, USA) and then fire polished on a microforge (MF-830 Narishige Group, Tokyo, Japan). Voltage errors were minimized with correction and prediction mode of series resistance compensation both set to 50 %. Extracellular solution contained (in mmol/L) 140 NaCl, 4 KCl, 2 CaCl 2 , 1 MgCl 2 , 10 HEPES, adjusted to pH 7.4 with NaOH, osmolarity 320-325 mOsm/L with glucose. Pipettes were filled with an intracellular solution containing (in mmol/L) 140 CsCl, 5 NaCl, 5 EGTA, 2 MgCl 2 , 10 HEPES adjusted to pH 7.3 with CsOH, osmolarity 305-310 mOsm/L with glucose.
Once filled with the appropriate intracellular solution, recording electrodes had a resistance between 2.0 and 3.2 MX. A silver chloride-coated silver wire served as a reference electrode with one end connected to the ground input of the amplifier and the tip placed directly into the bath solution. Cells having a leak current after establishing whole-cell configuration of more than 10 % of the peak sodium current were discarded and those which had developed leak of this magnitude during the experiment were not used in the final analysis. The liquid junction potential between the bath and the pipette solutions was not corrected. Whole-cell membrane currents were filtered at 5 kHz and sampled at 20 kHz using an Axopatch 200B patch clamp amplifier (Molecular Devices, Foster City, CA) and Digidata 1200B A/D converter (Molecular Devices, Foster City, CA). Data were acquired on a Windows-based PC using Clampex software (Molecular Devices, Foster City, CA) and analyzed by pCLAMP (Clampfit) 9.2 software (Molecular Devices, Foster City, CA).
## Voltage-clamp protocols
To characterize the voltage dependency of steady-state channel activation, currents were evoked by voltage increments of 5 mV from -80 to ?70 mV from a holding potential of -120 mV [bib_ref] Voltage-clamp and current-clamp recordings from mammalian DRG neurons, Cummins [/bib_ref]. Persistent currents were quantified at 8 ms after onset of the depolarizing voltage step [bib_ref] SCN9A mutations in paroxysmal extreme pain disorder: Allelic variants underlie distinct channel..., Fertleman [/bib_ref] [bib_ref] Paroxysmal extreme pain disorder mutations within the D3/S4-S5 linker of Nav1.7 cause..., Jarecki [/bib_ref]. Conductance (G) values were calculated from measured peak inward currents (I) and observed reversal potential for sodium ions (E rev Na ) using the equation G = I/(V m -E rev Na ). Reversal potential was measured by extrapolating the linear portion of the I/V relationship between ?15 and ?60 mV. Resulting values for conductance were normalized to the peak inward current and fitted using the Boltzmann equation. Steady-state inactivation of wild-type Nav1.7 channels, W1538R or A1746G mutations was assessed by holding cells at potentials incrementing from -110 mV to 0 mV for 500 ms, followed by a step to -10 mV for 50 ms. Voltage dependence of slow inactivation was investigated using a 20-ms pulse to 0 mV after a 10-s pre-pulse to potentials from -100 to 0 mV followed by a 100-ms pulse to -120 mV to remove fast inactivation.
Resulting I-V curves were then fitted by using the Boltzmann equation. With a preconditioning pulse of 10 s, no complete slow inactivation was achieved; therefore, resulting values may not compare to previously published data for slow inactivation in WT Nav1.7. However, a time window of 10 s is considered to be relevant for inducing slow inactivation [bib_ref] Can robots patch-clamp as well as humans? Characterization of a novel sodium..., Estacion [/bib_ref] , and therefore, differences seen in mutated Nav1.7 channels compared to WT are representative of enhanced slow inactivation. To investigate the time to recovery from fast inactivation, first a 20-ms pulse to -10 mV was applied followed by a recovery phase between 5 and 100 ms and a second pulse to -10 mV [bib_ref] Voltage-clamp and current-clamp recordings from mammalian DRG neurons, Cummins [/bib_ref]. Recorded peak currents were then normalized and fitted using a single exponential equation (one-phase association).
# Statistical analysis
All data are expressed as mean ± SEM. Differences in means between wild-type channels and mutations were tested by two-tailed Student's t test or one-way ANOVA with the Bonferroni post-test where appropriate. P \ 0.05 was considered significant. Calculations were made using the Graph Pad Prism Software version 5.0 (GraphPad Software Inc., La Jolla, CA).
# Results
## Clinical presentation: patient 1
The first case is a 68-year-old woman who first developed bilateral foot pain at the age of 61. The pain was mild at onset, however, progressively worsened in severity and she subsequently noted erythema of the feet. She currently suffers from constant pain affecting particularly the soles and heels of her feet and extending to the level of her ankles. She typically rates this pain as 8 out of 10 on a numerical rating scale; however, this can increase to 10 out of 10 during exacerbations. The pain is described as burning, and at times, she feels that her feet are about to burst. She is never pain free. Exacerbations are triggered by heat, prolonged standing and wearing enclosed shoes. These exacerbations last a number of hours, are associated with increased foot erythema and can be relieved by cooling. She scored 7 out of 10 on the DN4 screening questionnaire for neuropathic pain (suggesting pain symptomatology consistent with a neuropathic etiology) [bib_ref] Comparison of pain syndromes associated with nervous or somatic lesions and development..., Bouhassira [/bib_ref]. She is significantly disabled by her pain and now requires a wheelchair. She notes occasional paresthesia and numbness of the feet. There is no history of foot ulceration or autonomic symptoms. There may be a relevant family history in that her mother who is deceased complained of foot pain, which limited her ability to mobilize.
## Examination and investigations: patient 1
On clinical examination, erythema of the forefoot was noted. Her gait was hesitant due to foot pain; otherwise, motor function was normal. Deep tendon reflexes, light touch, vibration sense and proprioception were normal. There was a subjective reduction in pinprick sensibility to the ankles. Nerve conduction studies of the upper and lower limbs were normal, and EMG of the peroneus tertius muscle was normal. Quantitative sensory testing of her right hand (which was asymptomatic) and right foot (symptomatic) was performed as per DFNS protocol [bib_ref] Quantitative sensory testing in the German research network on neuropathic pain (DFNS):..., Rolke [/bib_ref]. It should be noted that in the hand, findings were normal other than hyperalgesia to high threshold mechanical stimuli [fig_ref] Figure 1: Results of quantitative sensory testing in patient 1 [/fig_ref]. In the feet, which were symptomatic, there was a mixture of both loss and gain of sensory function. There was reduced sensitivity to cooling and thermal sensory limen but enhanced sensitivity to mechanical stimuli including pressure pain, mechanical pain and dynamic mechanical allodynia [fig_ref] Figure 1: Results of quantitative sensory testing in patient 1 [/fig_ref].
On testing autonomic function, there was no orthostatic hypotension. The pressor responses to mental arithmetic tasks and cutaneous cold stimuli were present, although there was a minimal response to isometric exercise. Respiratory sinus arrhythmia was present, while heart rate response to hyperventilation was minimal. A well-performed Valsalva maneuver yielded normal blood pressure profile. Overall, there was no evidence for cardiovascular autonomic dysfunction. The sympathetic skin response in the feet was normal (422 lV amplitude, 2,315 ms latency). Blood tests including full blood count, plasma protein electrophoresis, erythrocyte sedimentation rate, C-reactive protein, renal profile, liver profile, thyroid function tests, glucose, vitamin B12 levels, hepatitis B and C serology, anti-nuclear antibodies and cryoglobulins did not show any significant abnormalities. Skin biopsy of the distal leg was performed as described previously ) in order to investigate the possibility of a small fiber neuropathy. This demonstrated a normal intra-epidermal nerve fiber density of 11.93 fibers/mm as assessed by immunostaining for protein gene product 9.5 according to the guidelines of the European Federation of Neurological Societies [bib_ref] European Federation of Neurological Societies/Peripheral Nerve Society guideline on the use of..., Lauria [/bib_ref].
## Treatment: patient 1
She received partial benefit from treatment with gabapentin (900 mg; three times a day) and in addition used tramadol (50-150 mg) as rescue medication. A large number of alternative medications have not provided pain relief including carbamazepine, amitriptyline, duloxetine as well as 8 % capsaicin and 5 % lidocaine transdermal systems applied to the feet. Results of quantitative sensory testing in patient 1. Comparisons of evaluated sites are made against a Caucasian control population for each one of the evaluated sites, and tested sites differences are remarked below. Gain of sensory function is presented as a z score [2, and loss of sensory function as \2. CDT cold detection thresholds. WDT warm detection thresholds, TSL thermal sensory limen, CPT cold pain thresholds, HPT heat pain thresholds, PPT pressure pain thresholds, MPT mechanical pain thresholds, MPS mechanical pain sensitivity, WUR wind-up ratio, MDT mechanical detection thresholds, VDT vibration detection thresholds, DMA dynamic mechanical allodynia, PHS paradoxical heat sensations. Test sites: right hand and right foot. In the right foot, which was the symptomatic region, there was evidence of both gain and loss of sensory function. Specifically, we noted cold and thermal sensory limen hypoesthesia and heat pain hypoalgesia. There was evidence of mechanical hypersensitivity including pressure and mechanical pain hyperalgesia as well as mechanical pain hypersensitivity and dynamic mechanical allodynia Neuromol Med (2013) 15: Clinical Presentation: Patient 2 A White Caucasian boy aged 7 presented with a 4-year history of painful feet, which progressively worsened. It was described as episodic, burning pain affecting the soles and less so extending to the dorsal aspect of both feet. He rated the pain as 10 out of 10 on a numerical scale, typically lasting 15 min but on other occasions up to 24 h. The pain was particularly severe following exercise such as playing football or prolonged standing. It was also triggered by ambient hot temperature and wearing enclosed shoes. During exacerbations, he was unable to stand on his feet. He required cold air to be blown over his feet or had his feet hanging out of the car window, stuck on the window screen or submerged in cold water. He could walk up to 300 m. He wore open-toe shoes or more often ambulated barefoot. More recently, his hands have become affected and have complained of diffuse pains on his shoulders, knees and arms albeit infrequently (not more than once a month). The pain has never resulted in episodic stiffness, paralysis or dystonic posturing. He has never had any myoglobinuria, cramps, foot ulceration or dysautonomic symptoms. He did not suffer from allodynia or hyperalgesia. He had difficulty in concentrating at school. The patient is the third child of non-consanguineous parents. In his family history, his father is known to suffer from non-organic, overwhelming aggression and high tolerance to pain with his paternal grandmother was diagnosed with a late-onset neuropathy. There were no neonatal concerns or feeding difficulties. His early development was normal. He had a personal history of early infantile seizures, one associated with fever.
## Examination and investigations: patient 2
Examination revealed a well-grown boy with normal physical examination. He ambulated barefoot with no limitation to his motor function. Tandem gait was equivocal (slight in coordination). It was noted that he had initially subtle erythema of his feet. Deep tendon reflexes, light touch, proprioception, vibration and two-point discrimination were all normal. He had no tremor, muscle atrophy or hypertrophy or orthostatic hypotension. Nerve conduction velocities studies showed no evidence of large fiber peripheral neuropathy, and EMG studies of the tibial anterior muscle (deep fibular nerve) were normal. Thermography revealed a rapid hyperemic response to cold challenge after cooling [fig_ref] Figure 1: Results of quantitative sensory testing in patient 1 [/fig_ref]. There was no cardiovascular autonomic dysfunction, and heart echocardiogram and EKG were both normal. He has had both genetic and psychological review, the latter in view of his debilitating symptoms and lack of concentration. Further investigations including baseline bloods, autoimmune screen, plasma and urine amino acids, troponin I, creatine kinase, carnitine profile, lactate, ammonia, vitamin E, cholesterol, triglyceride, B12 and red cell folate test, urine organic acids, EEG, brain CT, MRI of the brain and whole spine were normal.
Treatment: Patient 2 Initially, he was treated with topical 0.025 % capsaicin cream on both feet with minimal improvement. He was subsequently started on mexiletine on a dose of 50 mg once daily, which was gradually increased to 100 mg three times daily (10 mg/kg/day). Treatment with mexiletine was started in association with 5 % lidocaine patches applied to the feet at night time over 12-h periods. His symptoms have gradually improved such that he now is able to walk to school and take part in sporting activities. His concentration has improved, and he is able to sleep at night without disturbance. He no longer submerges his feet in ice-cold water. The quality of life of the patient and his family has improved substantially.
## Scn9a mutations identified in patients 1 and 2
Given the phenotype of erythromelalgia seen in the patients, albeit extremely late onset in patient 1, we decided to sequence SCN9A as this was the best candidate gene for the disorder in both patients. In patient 2, DNA sequencing of genomic DNA revealed a heterozygous missense variant, c.5237C[G, which is predicted to change amino acid 1746 of Nav1.7 from an alanine to a glycine (A1746G). This variant occurred de novo as both parents were shown to not carry this variant when sequenced. In silico analysis of this non-synonymous variant using PolyPhen-2 and SIFT tools gave scores of 0.999 and 0, respectively, indicating that both algorithms predict the mutation to be probably damaging. The mutation was absent from the most recent dbSNP database and not previously reported as a disease-causing mutation.
In patient 1, DNA sequencing identified a heterozygous missense variant, c.4612T[C, which is predicted to change amino acid 1538 from a tryptophan to an arginine (W1538R). The parents of the patient were both deceased. The variant was absent from the most recent dbSNP database. In silico analyses using PolyPhen-2 and SIFT gave scores of 0.001 and 0.85, respectively, indicating that the mutation is benign. However, the same mutation was previously reported to cause a late-onset chronic non-paroxysmal neuropathic pain phenotype [bib_ref] Chronic non-paroxysmal neuropathic pain-novel phenotype of mutation in the sodium channel SCN9A..., Dabby [/bib_ref] , although no biophysical analysis was performed for the mutation (note that in the publication of Dabby et al., the mutation is annotated as W1550R due to a different cDNA sequence used as the reference sequence). Patient 1 was also a heterozygous carrier of 4 additional SCN9A variants, which are annotated as SNPs in dbSNP (rs71428908, rs12478318, rs4369876 and rs3750904). Of note, a patient with small fiber neuropathy has previously been reported who is a compound heterozygote for rs12478318 and rs4369876 [bib_ref] Gain of function Nam1.7 mutations in idiopathic small fiber neuropathy, Faber [/bib_ref].
The A1746G and W1538R mutations both map to the fourth sodium channel domain of Nav1.7 [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref] , where mutation A1746G maps to the sixth transmembrane segment, while mutation W1538R maps to the second transmembrane segment. Sequence alignment of voltage-gated sodium channels (Nav) indicated that the alanine at position 1746 is unilaterally conserved in every member of the Nav family in humans. Additionally, it is also highly evolutionarily conserved within each Nav1.7 channel across species. The late-onset mutation, W1538R, also shows high conservation across species but does not show such high conservation within the human Nav family with the mutant amino acid (arginine) found in both Nav1.1 and Nav1.3. We were therefore very keen to test the biophysical properties of these channels to determine whether each mutation was likely to be disease causing.
## Mutation of a1746g and w1538r leads to a hyperpolarizing shift in the current-voltage (i-v) relationship
The functional effect of A1746G and W1538R mutations on Nav1.7 function was investigated using the whole cell patch clamp technique on HEK cells transiently transfected with Nav1.7a (WT or mutant) ? Navb1 ? Navb2. Peak currents and whole cell capacitance in HEK293 cells transfected with Nav1.7a subunit containing the mutation A1746G (n = 18) or W1538R (n = 14) in combination with Navb1 and Navb2 subunits did not significantly differ from cells with wild-type Nav1.7 (n = 20) [fig_ref] Table 1: Cell and recording properties [/fig_ref].
We investigated the possibility of ion selectivity being affected by A1746G and W1538R mutations. This was done by calculating the reversal potential (E rev ) for every analyzed cell. On average, E rev was not changed by mutation A1746G (79.0 ± 3.2 mV) and W1538R (82.4 ± 2.9 mV) when compared to WT Nav1.7 controls (81.7 ± 2.3 mV). This result suggests that the two investigated mutations have no significant impact on the relative permeability of Na ? to Cs ? [fig_ref] Table 1: Cell and recording properties [/fig_ref].
Voltage dependence of channel activation was found to be changed by both mutations to more hyperpolarizing potentials when plotting peak inward currents (I) as a function of depolarization potential to generate I-V curves ;.
Quantification of persistent currents as a fraction of peak currents at the time point of 8 ms after the beginning of the depolarizing voltage step revealed a trend to higher values for persistent currents in mutant A1746G, although without statistical significance . When investigating timeto-peak as a measure of channel activation rate, we found a trend to shorter time-to-peak as a sign for faster channel activation in A1746G . At the voltage step that induced peak current in WT Nav1.7 transfected cells (5 mV), time-to-peak was significantly shorter in A1746Gmutated channels (0.37 ± 0.04 ms) when compared to Nav1.7 WT channels (0.53 ± 0.05 ms; P \ 0.05).
Further analysis of steady-state activation revealed a left shift of activation curves in both investigated mutations with significantly different voltages for half-maximal activation (V act ) for A1746G (-23.9 ± 0.3 mV; P \ 0.001) and W1538R (-16.9 ± 0.6 mV; P \ 0.05) when compared to wild-type Nav1.7 channels (-8.3 ± 0.5 mV), while slope factors of voltage-activation relationships were unchanged [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref].
Steady-state fast inactivation was not found to be affected by A1746G or W1538R mutations, as indicated by largely overlapping curves for voltage dependency of inactivation with similar slope factors and voltages for half-maximal inactivation (V inact ) not significantly different compared to wild type [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref].
Combined analysis of the voltage dependence of steadystate activation and fast inactivation showed a marked increase in the size of the window current for both mutations calculated from area under the curves (Controls 0.67; W1538R 1.63; A1746G 1.73) indicative of a gain-offunction mutation .
## Mutation of a1746g leads to enhanced slow inactivation and faster recovery from steady-state fast inactivation
Mutation A1746G additionally leads to a hyperpolarizing shift in the observed voltage dependency of steady-state slow inactivation of Nav1.7 (P \ 0.01) with a V inact of -40.2 ± 2.9 mV compared to -7.9 ± 5.0 mV in WT Nav1.7 [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref]. In contrast, the W1538R mutation of the Nav1.7a subunit did not change this channel property significantly [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref].
Investigating recovery from steady-state fast inactivation, we found a reduced time constant for recovery of 7.7 ± 0.5 ms for Nav1.7 channels with the A1746G mutation compared to Nav1.7 WT channels with a time constant of 11.9 ± 1.2 ms (P \ 0.05; . Recovery from steady-state fast inactivation in cells transfected with the W1538R mutation of the Nav1.7 a subunit was not significantly different from WT channels [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref].
In silico simulation of the observed biophysical changes caused by mutation A1756G and W1538G indicated a reduction in current injection thresholds for induction of a single action potential [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref]. The threshold for induction of action potential firing after simulated current injection of 225 pA in wild-type control was reduced to 115 pA (-49 %) by simulating mutation A1746G, while mutation W1538R had a threshold of 144 pA (-36 %). We found an increase in action potential firing frequency [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref]. At a simulated current injection of 250 pA, mutation A1746G increased the number of action potentials within 200 ms from 1 to 20, while mutation W1538R resulted in 13 action potentials [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref].
In summary, simulation of mutation A1746G was associated with greater changes in both investigated parameters compared to wild-type Nav1.7, while mutation W1538R induced less pronounced differences in action potential
[formula] P = 0.51 [/formula]
Data presented as mean values and SEM. WCC whole cell capacitance, E rev Reversal potential. Differences in means of cells with mutant channels compared to wild-type controls were tested by twotailed Student's t test. P \ 0.05 was considered significant Properties of Nav1.7-mediated currents in voltage-clamp configuration. a Example traces from whole-cell voltage-clamp recordings performed in HEK293 cells expressing wild-type Nav1.7 (control, left), W1538R mutant (middle) or A1746G mutant (right) channels. From a holding potential of -120 mV, currents were evoked by voltage increments of 5 mV from -80 to 40 mV. Results from one recording per group are displayed. b Normalized currentvoltage plots from recordings as described above. c Persistent current at 8 ms after onset of the depolarizing voltage step plotted as a fraction of peak currents against voltage. d Time-to-peak current as a measure for channel activation kinetics. Data are presented as mean values and SEM (n = 20 for controls, n = 14 for W1538R and n = 18 for A1746G). Wild-type Nav1.7 (Control, black squares), W1538R mutation (blue circles) and A1746G mutation (red triangles) firing frequency and current threshold. Both simulated mutations lead to increased neuronal excitability indicative of a gain-of-function mutation.
[formula] V (mV) Slope V (mV) Slope V (mV) Slope Tau (ms) K (1/ms) T( [/formula]
# Discussion
So far, all primary erythromelalgia (PE) SCN9A-related mutations that were characterized electrophysiologically have been located in the first three domains of Nav1.7 with the main cluster of mutations in the S4-S5 region of the second domain (DII) [bib_ref] Inherited erythermalgia: Limb pain from an S4 charge-neutral Na channelopathy, Choi [/bib_ref] [bib_ref] Sodium channelopathies and pain, Lampert [/bib_ref].
We report the effect of two mutations within the fourth domain of Nav1.7 on sodium channel function (W1538R in S2 of DIV and A1746G in S6 of DIV). The half-point of activation was shifted toward more hyperpolarized potentials, while steady-state fast inactivation was not different in both mutants. The hyperpolarizing shift in steady-state activation results in a greater overlap of steady-state activation and fast inactivation, which is often referred to as window currents [bib_ref] De novo mutation in the SCN5A gene associated with early onset of..., Wedekind [/bib_ref]. We have observed an enhanced slow steady-state inactivation and a shorter recovery from steady-state inactivation in channels with the A1746G mutant only, while these properties were normal in channels with W1538R mutation. These findings are overall consistent with those observed in other SCN9A mutations causative for primary erythromelalgia . Finally, in silico simulations of both mutations in a smalldiameter neuron revealed reduced thresholds for action potential firing and enhanced repetitive firing, indicating hyperexcitability caused by biophysical channel property changes by mutation W1538R and A1746G. Since the patient with the A1746G mutation showed a positive response to systemic treatment with the sodium channel blocker mexiletine, it might be useful to elucidate the effects of mexiletine on Nav1.7 channels with this mutation in vitro.
## Spectrum of inherited gain-of-function conditions
Mapped to SCN9A
The original range of inherited conditions associated with SCN9A (i.e. PE, PEPD and CIP) has recently been Voltage dependence of steady-state activation and fast inactivation of Nav1.7-mediated currents in voltage-clamp configuration. a Conductance values from HEK293 cells expressing wildtype Nav1.7 channels (Control, black squares), W1538R mutation (blue circles) or A1746G mutation (red triangles) were calculated from peak inward currents in activation protocols, normalized and fitted using the Boltzmann equation. Please refer tofor values for half-maximal activation (V) and slope factors. All data are presented as mean values and SEM (n = 20 for controls, n = 14 for W1538R and n = 18 for A1746G). b Voltage dependence of inactivation for wild-type Nav1.7 channels (Control, black squares), W1538R mutation (blue circles) or A1746G mutation (red triangles). Please refer tofor values for half-maximal inactivation (V) and slope factors. (n = 20 for controls, n = 13 for W1538R and n = 14 for A1746G). c Combined analysis of voltage dependence of steady-state activation and fast inactivation in HEK293 cells expressing wild-type Nav1.7 channels (Control, black lines), W1538R mutation (blue lines) and A1746G mutation (red lines).
Window currents with increased area under the curve (AUC, see in insert in cells expressing W1538R mutation (blue area) or A1756G mutation (red area) compared to wild-type controls (gray area). Differences in means were tested by ANOVA and Bonferroni post hoc test. P \ 0.05 was considered significant supplemented with chronic non-paroxysmal neuropathic pain [bib_ref] Chronic non-paroxysmal neuropathic pain-novel phenotype of mutation in the sodium channel SCN9A..., Dabby [/bib_ref] , small fiber neuropathies (SFN) [bib_ref] Gain of function Nam1.7 mutations in idiopathic small fiber neuropathy, Faber [/bib_ref] and partial congenital insensitivity to pain [bib_ref] Two novel mutations of SCN9A (Nav1.7) are associated with partial congenital insensitivity..., Staud [/bib_ref] [bib_ref] Two novel SCN9A gene heterozygous mutations may cause partial deletion of pain..., Yuan [/bib_ref]. It is as yet not fully understood why some gain-of-function mutations in SCN9A are found to cause small fiber neuropathy, while others cause gain-of-function but not small fiber neuropathy [bib_ref] Gain of function Nam1.7 mutations in idiopathic small fiber neuropathy, Faber [/bib_ref] [bib_ref] Nav1.7-related small fiber neuropathy: Impaired slow-inactivation and DRG neuron hyperexcitability, Han [/bib_ref] , as in the patient reported here with the W1538R mutation, in whom intraepidermal nerve fiber density was found to be normal.
## Different phenotypes attributed to the same mutation
Dabby and co-authors have attributed chronic non-paroxysmal neuropathic pain to the W1538R mutation [bib_ref] Chronic non-paroxysmal neuropathic pain-novel phenotype of mutation in the sodium channel SCN9A..., Dabby [/bib_ref] , while our patient 1 carrying the same mutation had exhibited classic clinical features of primary erythromelalgia. In a similar fashion, Estacion and colleagues have reported three different clinical presentations in patients genotyped for the same I228 M substitution in Nav1.7 [bib_ref] Intra-and interfamily phenotypic diversity in pain syndromes associated with a gain-of-function variant..., Estacion [/bib_ref]. One patient presented with severe facial pain, while the other sufferer from the same family had distal pain in hands and feet. The third and unrelated subject was complaining of discomfort in the scalp region. Complex phenotypes representing a combination of pathological conditions and the same mutations manifesting differently in the same and different families are not exclusive to SCN9A [bib_ref] Intra-and interfamily phenotypic diversity in pain syndromes associated with a gain-of-function variant..., Estacion [/bib_ref]. For example, mutations in the predominantly cardiac isoform of VGSC family, SCN5A, can be represented by phenotypes combining long QT and Brugada syndromes [bib_ref] A single Na(?) channel mutation causing both long-QT and Brugada syndromes, Bezzina [/bib_ref] , or a combination of long QT syndrome and congenital heart block [bib_ref] Homozygous SCN5A mutation in long-QT syndrome with functional two-to-one atrioventricular block, Lupoglazoff [/bib_ref].
In a wider medical context, when a disease-causing mutation is expressed, it is not necessarily expressed in all affected individuals in the same way. These variable phenotypes, such as the one described here, may be caused by a number of factors, such as gene-environment interactions, gene modifiers (genetic background possibly affecting penetrance, dominance or expressivity) and allelic variation among others [bib_ref] The van der Woude syndrome in a large kindred: Variability, penetrance, genetic..., Janku [/bib_ref] [bib_ref] Holoprosencephaly: A family showing dominant inheritance and variable expression, Collins [/bib_ref] [bib_ref] Modifier genes in mice and humans, Nadeau [/bib_ref]. Whether the observed phenotypic diversity between the two carriers of the same mutation, W1538R, is attributed to epigenetic, environmental factors or gene modifiers remains unknown [bib_ref] Intra-and interfamily phenotypic diversity in pain syndromes associated with a gain-of-function variant..., Estacion [/bib_ref]. However, the common feature of this mutation appears to be the relatively late onset of clinical manifestation.
## Coupling of activation-inactivation
Each domain of the a unit of VGSC consists of 6 transmembrane segments (S1-S6). The first four, S1 to S4, are believed to form the voltage sensing part of the channel, while the channel pore comprises S5-S6. When aligning the sequences of the four domains contributing to the a subunit, the sequence of DI is similar to DIII, while DII is Steady-state slow inactivation and recovery from inactivation of Nav1.7-mediated currents in voltage-clamp configuration. a Voltage dependence of steady-state slow inactivation in HEK293 cells expressing wild-type Nav1.7 channels (Control, black squares), W1538R mutation (blue circles) or A1746G mutation (red triangles) was investigated using a 20-ms pulse to 0 mV after a 10-s pre-pulse to potentials from -100 to 0 mV followed by a 100-ms pulse to -120 mV to remove fast inactivation. Resulting currents were then fitted by using the Boltzmann equation. Please refer tofor values for half-maximal inactivation (V) and slope factors. All data are presented as mean values and SEM (n = 16 for controls, n = 8 for W1538R and n = 7 for A1746G). b Representative overlays of example current traces recorded in HEK293 cells expressing wildtype Nav1.7 channels (Controls), W1538R mutation and A1746G mutation. Results from one recording per group are displayed. To investigate the time to recovery from fast inactivation, first, a 20-ms pulse to -10 mV was applied followed by a recovery phase of 5-100 ms and a second pulse to -10 mV. c Recorded peak currents were then normalized and fitted using a single exponential equation (one-phase association).reports values for time constant (Tau), rate constant (K) and half-time (T). All data are presented as mean values and SEM (n = 17 for controls, n = 10 for W1538R and n = 13 for A1746G). Differences in means were tested by ANOVA and Bonferroni post hoc test. P \ 0.05 was considered significant Neuromol similar to DIV. It is therefore thought that sodium channels developed evolutionary from a one domain precursor via a two-domain structures as an intermediate developmental step [bib_ref] NaChBac: The long lost sodium channel ancestor, Charalambous [/bib_ref]. The only previously described lesion mapped to DIV is mutation A1632E, which in fact has features overlapping two inherited conditions-PEM and PEPD. It is mapped to S4-S5 intracellular linker of DIV ). Both patients described here had clinical symptoms strongly consistent with those of primary erythromelalgia, but no features suggestive of PEPD. It has been proposed that based on our current understanding derived from mapping gain-of-function mutations in the SCN9A gene, the channel activation process is associated with DI and DII , while inactivation or coupling of activation-inactivation is linked to DIII and DIV conformational changes [bib_ref] Studies of alpha-helicity and intersegmental interactions in voltage-gated Na ? channels: S2D4, Ma [/bib_ref]. Intriguingly, a recent publication by Waxman and co-workers demonstrated that two different mutations, located within the DIII/S4-S5 linker region and only 10 amino acids apart from each other, lead to distinct biophysical effects (enhancement of activation or inactivation, respectively) and therefore distinct pain disorders (PEM and PEPD). This once more indicates that the effect of a particular mutation on channel function may be more relevant than its location within the channel protein . So far, W1538R and A1746G are the only mutations mapped to DIV that have been found to predominantly affect the channel property of steady-state activation, while steady-state fast inactivation is unchanged. The W1538 residue is conserved only in three of the nine known human voltage-gated sodium channels. Interestingly, an arginine (R) at the corresponding position in Nav1.1 and Nav1.3 is documented as the normal variant [fig_ref] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of... [/fig_ref].
Correlation Between Age of Onset and Phenotype with Differences Found in Biophysical Properties of Nav1.7 Mutations A correlation between the age of onset and the magnitude of the hyperpolarizing shift in voltage dependence of steadystate activation of Nav1.7 has been proposed previously [bib_ref] Early-and late-onset inherited erythromelalgia: Genotype-phenotype correlation, Han [/bib_ref] , although alternative splicing has been proposed to have an impact on age of PEM manifestation as well . In recent reviews, it has been suggested an association between the degree of steady-state activation alteration in particular with early age of onset (0-16 years) of clinical symptoms typical for primary erythromelalgia [bib_ref] Sodium channelopathies and pain, Lampert [/bib_ref]. The patient we describe here with the W1538R mutation had an extremely late onset of symptoms after the age of 61. Prior to molecular genetic testing for SCN9A mutations and given the lack of a clear family history, she would have previously been diagnosed as having secondary erythromelalgia. We have not seen any other cases with such late onset reported in the literature. Few detailed QST profiles have been reported in patients with primary erythromelalgia and confirmed SCN9A mutations. She demonstrated a mix of reduced sensitivity to thermal stimuli as well as a gain-of-function in relation to mechanical hyperalgesia and dynamic mechanical allodynia. There was no loss of intra-epidermal nerve fibers to explain the raised cooling detection thresholds and reduced thermal sensory limen. Microneurographic recordings from patients suffering from erythromelalgia demonstrate ongoing activity in primary afferent nociceptors [bib_ref] Microneurographic findings of relevance to pain in patients with erythromelalgia and patients..., Orstavik [/bib_ref]. The mechanical hypersensitivity which we observed may be due to altered transduction properties of primary afferents to mechanical stimuli or central sensitization as a consequence of ectopic activity. This QST profile is quite different to a patient we recently described with an L858F mutation [bib_ref] Imaging the neural correlates of neuropathic pain and pleasurable relief associated with..., Segerdahl [/bib_ref] , who demonstrated selective thermal hyperalgesia and further studies of such patients are needed to see whether particular QST profiles may relate to certain SCN9A mutations or whether the QST profile may change over time.
# Conclusions
In conclusion, we clinically and biophysically characterized two novel mutations within DIV of Nav1.7 encoded by SCN9A, both leading to typical phenotypes of primary erythromelalgia. While we found that mutation A1746G led to marked changes in biophysical properties of Nav1.7 and early onset of clinical symptoms, mutation W1538R resulted in less pronounced consequences on evaluated channel properties and a very late onset of disease. This finding demonstrates that mutations encoding for DIV of Nav1.7 can not only be linked to CIP or PEDP but also be causative of primary erythromelalgia.
[fig] Figure 1: Results of quantitative sensory testing in patient 1. Comparisons of evaluated sites are made against a Caucasian control population for each one of the evaluated sites, and tested sites differences are remarked below. Gain of sensory function is presented as a z score [2, and loss of sensory function as \2. CDT cold detection thresholds. WDT warm detection thresholds, TSL thermal sensory limen, CPT cold pain thresholds, HPT heat pain thresholds, PPT pressure pain thresholds, MPT mechanical pain thresholds, MPS mechanical pain sensitivity, WUR wind-up ratio, MDT mechanical [/fig]
[fig] Figure 2: Amino acid residue alignment demonstrating magnitude of inter-isoform and inter-species conservation of the affected regions. Blue color residues identify the mutant W1538R (patient 1) and red color indicates A1746G mutation (patient 2) mapping to S2 DIV and S6 DIV of human Nav1.7a subunit, respectively. Green color indicates corresponding non- [/fig]
[table] Table 1: Cell and recording properties [/table]
|
Metal-responsive regulation of enzyme catalysis using genetically encoded chemical switches
## Minor points
Introduction, last paragraph: "…used MD simulations to identify all residue pairs that undergo significant changes… during catalysis." -Rephrase. To my knowledge the actual catalysis event cannot be simulated with MDs. Introduction, last paragraph: "…to select functional candidates for analysis." -Clarify what is meant by "functional candidate".
Results, chapter "Model selection and design of POP switches", 1st paragraph -The authors talk once of "domains" and then again of "subunits". Clearly the β-propeller and the peptidase parts of POP are domains since they are located on the same protein chain. The term subunits is used for two separate protein chains. Replace "subunits" with "domains".
Results, chapter "Model selection and design of POP switches", Figure 2 -which PDB was used for the representation of POP? -in a) "Placement between domains" has not been mentioned as a criterium in the main text -in b) Is the Zn(Bby)2 Ligand correctly oriented? Further guidance including atom-wise coloring and orientation similar to the Bby-Ala in would help to understand how the 10.4 Å distance was derived.
Results, chapter "Model selection and design of POP switches", 2nd paragraph "(i) … in one enzyme conformational state," -in which state? Closed or open? "(iii) undergo suitable changes in Cβ distance to prohibit bond formation a different conformational state." -Clarify this point. I understand this criterium that metal binding should only occur in one conformation but not the other and that hence the two Bby-Ala positions need to be close in one conformation but distant in the other.
Results, chapter "HTE screening of POP switches", 1st paragraph, "… at an average yield of 1098 +/-124 µg/mL,…" -Define µg/mL. Do the authors mean the yielded concentration or do they mean 124 µg protein yielded from 1 mL expression volume?
Results, chapter "HTE screening of POP switches", 2nd paragraph, "… understanding allosteric effects on protein function." -Replace "allosteric" with "conformational". -Highlight the three variants POP<sub>167/517</sub>, POP<sub>169/510</sub> and POP<sub>159/517</sub> as in Results, chapter "HTE screening of POP switches", 3rd paragraph, "… showing the weakest degree of inhibition or activation…" -Give quantitative numbers as in in the main text and .
# Results, supplementary
Results, chapter "Functional characterization of a BbyAla POP switch", Supplementary Figure 5 -Full-length and truncated protein should be highlighted in the figure or the molecular weight given in the legend.
Results, chapter "Functional characterization of a BbyAla POP switch", -What is meant by the labels "167N517V" and "159K517V"?
Results, chapter "Functional characterization of a BbyAla POP switch",-Provide the standard error of the Michaelis-Menten fit as calculated by Origin.
Results, chapter "Extension to luciferase chemical switches", -Why does luciferase show a substrate inhibition effect? Is this known for the enzyme? Please clarify in the main text. SI, chapter "MD simulations of POP" and "MD simulations of Luciferase" p. 26 -For reproducibility reasons please provide the PDB-IDs for POP and luciferase structures used in MD simulations -Clarify in more detail how the BpyAla side chains were introduced. -What is meant by "cross-linked luciferase"? Please explain. SI, chapter "Cell extract preparation" p. 31 -pEVOL vectors contain constitutive promoters for aaRS1 and tRNA as well as an araC induction system for aaRS2. However, here IPTG was used for induction. Does pEVOL-BpyRS differ from other pEVOLs? SI, chapter "Steady-State Kinetic Assays" p.39-41 -"For steady-state kinetic assays with variants POP167/513, POP169/510, POP169/512, and POP167/516, protein samples …" Three of the variants have never been mentioned in the main text, however, POP159/517 is missing.
-Why were only POP wt, POP167, POP517, and POP167/517 purified after EDTA treatment by SEC? Could diafiltration for the other variants leave traces of EDTA, which might have inhibited POP activity in the kinetic assays? -Why was POP transferred to pure H2O. Most proteins require a buffer system for stability, why not POP? -Why does Supplementary show exemplary kinetic traces for POP167/513 which is never mentioned in the main text? -The data for POP167, POP517 and POP169/510 inor at least referred to in the main text.
-Data for POP169/512 and POP167/513 inare irrelevant because they have never been mentioned in the main text.
## Comments:
Zubi, et al demonstrated that the formation of M(bpy)2 species can endow a genetically encoded chemical switch to regulate the catalytic activities. They conducted structural analysis and MD simulation to assess the desirable positions to incorporate a pair of bpy-Ala residues. Consequently, 24 POP variants were prepared by cell-free protein expression systems, and at least 12 ones are responsive to the first-row transition divalent metal ions, such as Co, Ni, Cu, and Zn. In addition, the proof of concept was applied to luciferase, expanding the scope of systems. 1) One of the general prerequisites for any chemical switch would be not to perturb the native catalytic activity and protein stability. Otherwise, it would be no longer useful. In this regard, the catalytic activities of Pluc variants were substantially perturbed from the wild-type, yielding 25% or 50% of the Vmax even in the absence of Ni ion. Can authors speculate how such a significant loss of activities results from the incorporation of bpy-Ala residue? Perhaps, are bpy-Ala residues incorporated in the vicinity to the catalytically essential residues? Based on these results, can authors improve the designing scheme to propose the metal-dependent switches?
2) The authors carried out structural analysis and MD simulation to design 24 POP variants to create a pair of bpy-Ala variants. Although all satisfied the geometric constraints in the designing process, their response to metal ions varies significantly. To provide a better guideline, additional biochemical characterization would be necessary to speculate why some variants worked but not others. Perhaps, the authors can group 24 POP variants into four categories, one showing metaldependent activity as expected, inverse metal-dependent activity, no metal-dependence with catalytic activity, and no metal-dependence/activity. Then, the authors may characterize any representative variants from each group to explain their distinct behavior on metal ions.
3) 167/513 and 170/513 show inverse metal-dependent activity in that they exhibit even accelerated catalytic activity upon the addition of metal ions. How is it even possible? 4) To my best knowledge, POP is a dimeric protein. Therefore, the addition of bpy-Ala residue at one position of a protomer would form not only M(bpy)2 moiety but also two distantly located bpy residues. Although it would influence both structure and function of the proteins, the authors neither described nor discussed this aspect in this manuscript. Because these residues are exposed to the protein surface, the latter bpy-Ala may form intermolecular complexes, possibly generating undesirable high-ordered species. Is it why the SDS-PAGE analysis was carried out in the presence of excess EDTA? Perhaps, the formation of high-ordered species might be related to their distinct response to the metal ion. Do authors examine the oligomeric structures of the proteins upon the addition of metal ions? Perhaps, size exclusion chromatography or analytical ultracentrifugation in the presence of Ni can be carried out to quantify the fraction of catalytically active and metal-responsive species. 5) If the intermolecular complexation occurs, as shown below, what are the steady-state kinetic parameters of the 1:1 mixture of the 167 and 517 variants?
6) The authors characterized one of the POP variants by UV-Vis and CD spectroscopy, demonstrating that Fe would be a surrogate metal element to show Ni-dependent chemical switch. However, neither ferric nor ferrous ion was not applied for this system in for no apparent reason. Please provide the catalytic activities of POP variants with Fe ion. 7) Please include why the authors used a cell-free expression system instead of standard e coli heterologous expression. Is it to increase protein yields of having two bpy-Ala residues in a protomer? More detailed information in the experimental setup might be necessary for readers. 8) Among the divalent metal ions that the authors used, the cupric ion is redox-active under physiological conditions, and reduction may lead to significant changes in the geometry of M(bpy)2 and subsequent metal-dependent catalytic activities. Because the catalytic activities of POP are independent of dioxygen, the function of the metal-dependent switch can be monitored under anaerobic conditions. Can authors explore whether these systems are redox-dependent? 9) Do authors only consider the distance between two residues for MD simulation? What about the directionality of the side-chains? Now that the authors have both simulation and experimental data sets, can any conclusion or criteria in the determination of two bpy-positions can be drawn and reassessed? 10) has no information related to residues. Please provide the residue labels. In addition, add the distance between i and j residues in Supplementary Table 2. Please also include a column showing whether the pairs are responsive to metal ions as expected or not.
11) The metal-dependent chemical switch can be valuable, but it would not be applicable for every enzyme system. The limitation in the scope of the enzyme should be discussed. For example, metalloenzymes, having a metal cofactor are unlikely to be applicable for this chemical switch because it can be removed by EDTA treatment. In addition, only the enzymes having sufficient conformational changes at the region of substrate access and product release step can be applied.
12) There is no standard deviation or experimental error in. Please provide the values.
We thank the reviewer for raising these key concerns. In the text that follows and per the reviewer's suggestions, we have revised major parts of text to avoid overstating significance and to clarify data representation.
## Major points
1) The authors' core selling point for a broad audience is the "key challenge to engineering allosterically regulated enzymes" (Introduction, 2nd paragraph), which they intend to tackle in this publication. I have several major concerns with this: i) The principle of allostery is not explained at all. It is merely presented as "regulation and control mechanism to ensure that they (enzymes) can respond to environmental stimuli" (Introduction, 1st paragraph). However, "allostery" is an established term to describe that ligand binding to an allosteric site affects ligand binding or enzyme catalysis at the active site. In this regard, engineering true allostery is much more complex than engineering a control mechanism of enzyme activity in response to external stimuli, which seems to be the actual goal of this manuscript.
ii) The authors point out that control mechanisms are "appealing to chemists and synthetic biologists" and that current engineering strategies often "neglect native control capabilities" (Introduction, 1st paragraph). I do not understand this line of argument. For the majority of engineered enzymes it is not important to have a control mechanism. Only when the enzyme is used in a certain application a control mechanism is possibly required. Hence, to work out the main point of interest, this paragraph should focus on WHY control mechanisms are appealing e.g., for which fields of research.
We appreciate reviewer for raising this point of clarification. We reframed the introduction so that we avoid using the term allostery to describe our system, and instead focus on controlling conformation of proteins. We also added several examples where control mechanisms have been shown to be/could be important for biocatalysis/chemical synthesis. Both issues raised primarily concern paragraph 1. This has been rewritten as follows:
Naturally occurring enzymes have evolved to catalyze chemical reactions, in many cases, with exquisite selectivity, substrate specificity, and high catalytic rates. 1 Many enzymes also possess regulation and control mechanisms to ensure that they can respond to environmental stimuli. 2 Leveraging these remarkable properties for chemical synthesis has long appealed to chemists and synthetic biologists, 3,4 but current enzyme engineering strategies largely focus on improving catalytic properties or changing substrate specificity. 5 These engineering strategies often neglect native control capabilities like allosteric regulation, which can be inefficient for non-native substrates, 6 or lost without selection pressure as has been observed in several directed evolution approaches . Most engineered enzymes are used in highly optimized single-reaction processes where allosteric control is not needed and could even be detrimental. The need for modular, orthogonal methods to control enzyme activity in a stimulusdependent manner has become increasingly apparent, 13 however, for efforts to evolve proteins for in vitro multi-enzyme cascade catalysis 14 , diagnostic tools, and construction of novel signaling circuits 20 iii) In the 3rd paragraph of the introduction the authors minimalistically provide an overview of how protein function can be rendered sensitive to external stimuli. This paragraph is kept quite superficial and does not reflect the current state-of-the-art in this field. Many advances regarding "functionally-orthogonal methods to control enzyme activity in a stimulus-dependent manner" have been achieved over the last decades. I am especially missing a reference to control mechanisms with light, which are not only highly successful but also include manifold renowned strategies e.g., optogenetics, including allosteric switches or regulation of allostery (see various reviews on light regulation).
We agree that we should have better highlighted past studies focused on the development of functionally-orthogonal methods to control enzyme activity in a stimulus-dependent manner. Per the reviewer's suggestion, we have now added more detail to describe methods used to control protein dynamics and conformations with light and cite two recent reviews. We felt that the light dependent systems were best described in paragraph 2, which has been rewritten as follows:
A key challenge to engineering dynamically regulated enzymes is the complex interpla between catalytic activity, substrate specificity, and regulation. For example, while native allosteric proteins have been engineered to respond to new stimuli, this control typically applies to the native protein function, frequently DNA binding. Chimeric systems have also been developed to couple a naturally responsive protein to an protein of interest (POI) so that regulation of the former allows for control over the latter 22 . Optical regulation is particularly notable in this regard given that it allows for precise spatiotemporal control of protein function 23 . This approach is commonly achieved by fusing a POI to the light, oxygen, and voltage (LOV) responsive domain of plant phototropins, which can control protein function through blue-light induced conformational changes (e.g. sterically hindering accessibility to enzyme active sites or control ordered and disordered protein regions). Optogenetic dimerization has been successfully implemented in systems that utilize LOV domains and those based on Arabidopsis thalia cryptochrome 2 (CRY2) . These approaches often require that large (potentially disruptive) regulatory domains be used, and achieving efficient regulatory transduction between domains can require extensive protein engineering 30 . Finally, de novo design has been used to produce switchable proteins, 31,32 but switching has only been reported in response to peptides and proteins, and de novo design of switchable enzymes has not been reported. iv) In the 4th paragraph of the introduction, the authors explain their objective, which is to use "small molecule switches" incorporated into enzymes through amber codon suppression to control protein conformation and catalysis in a reversible manner. However, they do not demonstrate which achievements have already been made towards this goal. To my knowledge at least the field of light regulation has already achieved some major milestones. The group of Lei Wang has shown to disrupt the conformation and binding capabilities of proteins with a photoswitchable click UAA by controlling the α-helical content . The group of Peter Hamm has picked up this concept and controlled ligand binding by a factor of >100-fold with an azobenzene-switch crosslinked to an α-helix; they could even characterize the signal transmittance . Moreover, the group of Reinhard Sterner light-regulated true allostery in a reversible manner with photoswitchable UAAs at positions distant from the active site by a factor of ~10-fold To address the reviewers concern, we have now added several references with respect to innovations and milestones from modulating protein activity with light. We have added the references provided by the reviewer: We have decided to not include the work by Peter Hamm as these papers do not use ncAAs to accomplish their work but use cysteines to ligate on an azobenzene switch into the proteins of interest. We also added important context to clarify differences between these earlier studies and our results. These edits were in paragraph 3, which has been rewritten as follows:
We
## Metal ions are ligands that bind to bpy
LGs. Ligand binding occurs outside of the active site and regulates enzyme activity, so our system meets the basic criteria for allostery-specifically metalloallostery. That being said, our intent was to focus on the control that the resulting allostery provided as the reviewer suggested. To this end, we have taken the reviewer's advice regarding the entire introduction as outlined above and in the passage noted in comment (v) to focus on controlling activity rather than any specific mechanism. We instead bring this idea up only in the discussion and as more of an analogy. In response to comment (v), we clarified (in paragraph 3, see above) how our approach allows for reversible control of enzyme catalysis (unlike the previously cited examples that allow for irreversible activation) in enzymes that were not natively subject to allosteric control (unlike the new cited systems from Kneuttinger et. al.).
vi) Finally, the authors promise that their approach has "no fundamental limits on which proteins and activities could potentially be controlled" (Introduction, 4th paragraph) meaning that it "could be applied to any protein" (Results, chapter 1, 2nd paragraph). This objective is highly ambitious and in order to publish in Nature Communications it should be demonstrated (at least to some degree) that their approach is powerful enough to realize this. To this end, the authors have employed two non-allosteric enzymes, which catalyze different reactions, and which are structurally not related. However, both enzymes exhibit a well-known domain movement, which the authors control by inserting the metal-responsive Bpy-Ala UAA. While I consider this approach highly effective for proteins with such domain motions, the authors did not show that this approach also works for proteins for which such large conformational transitions have not been identified. Since a large percentage of proteins does not show domain movements, I am not convinced that their approach will achieve to control the biological activity of any protein.
Thus, in my opinion the authors' selling point should be to design a control mechanism of protein activity based on the modulation of protein conformation and not to engineer allostery. In the introduction, they should furthermore clearly lay out the achievements that have already been made towards this goal, especially in the light regulation sector. In this context they should point out the differences to their approach and the potential advantages of metals over light or other external stimuli. In the discussion, they should finally elaborate whether their approach could be used to control any protein and what needs to be improved in order to achieve this.
As outlined above, we have taken steps to mitigate much of these concerns, switching our focus to designing a modular control mechanism, as suggested by the reviewer. We now note how metal regulation could be useful in the final paragraph of the introduction and specify that our initial system relies on enzymes that possess catalytically relevant conformational dynamics in both the introduction and the second paragraph of the results section. Finally, we included additional discussion on how we envision expanding this method to regulate enzymes that do not undergo large conformational movements like those systems selected for our initial studies.
## Paragraph 4 from the introduction:
To address this difficulty, here we developed a framework to design, build, test, and analyze enzymes with genetically encoded chemical switches. Initial efforts focused on bipyridine LGs that could induce reversible activation of enzymes in the presence/absence of metal salts We thank the reviewer for highlighting that our approach is highly interesting. The utility of any reversible switch for an in vitro cascade reaction will depend on the specific enzymes used and the specific reaction conditions associated with those enzymes. The goal of our study is to report the development of an approach that users can evaluate relative to other systems. It may be that for some enzymes, the use of metals and chelating agents could be problematic, but there are also engineering approaches to address these issues (e.g. metal chelating resins that can be removed, immobilized enzymes, etc.). In our opinion, that level of engineering is beyond the scope of this report, but we hope to move beyond this to pursue such applications in due course. We agree with the reviewer that in vivo switching would be difficult for the reasons noted and changed the text to focus on in vitro cascades as outlined above. The in vivo applications we had in mind involved inhibition of enzyme activity that could be removed in vitro. For example, expression of toxic proteins that are inactivated by metals in vivo (since our proteins express in metal-bound form) could then be used in the absence of metal in vitro. Likewise, sensing applications would not require reversible activation. We have modified the final paragraph to make this clear as shown here:
Looking forward, we anticipate that genetically encoded chemical switches installed into proteins using the integrated computational/experimental approach described here will enable new strategies for protein engineering and synthetic biology. This capability could expand opportunities for controlling multi-enzyme biocatalysis in vitro 3 for systems that can tolerate added metal salts and chelators. These applications would be facilitated by immobilized chelators or enzymes to minimize accumulation of chelator or both metal and chelator, respectively, since these components could adversely affect particular enzymes. While removal of these components cannot be readily achieved in vivo, it may also be possible to use metal switching for in vivo sensing , and other applications 20 where reversible regulation is not required.
3) While I enjoyed reading the results section and congratulate the authors for their experimental success, I stumbled upon a few flaws concerning the presentation of the approach, the validity of the data as well as inconsistencies in the data, which in my opinion prohibit immediate publication. i) Introduction, last paragraph and Results, first chapter. Despite the clear presentation in , there is no point-by-point description of the approach in the text or the legend to the figure. It is, hence, hard to follow and the approach raises some questions:
-
## Why did the authors use the c β distance in md simulations
This is now clarified in the
## - what is meant by double variants
Consistent with the protein engineering literature, we use -mutant‖ to refer to an altered gene and -variant‖ to refer to an altered protein (no change was made to the manuscript). Double variants would therefore refer to protein variants that were designed to have two BpyAla residues in place of their native residues at positions that are specified.
## - how does the metal-triggered switch work and related to this why does the bpy-ala side chains need to be located in close distance to each other
This switching mechanism is shown in . The following text was added to make this explicit in the text:
Structural analysis and MD simulations were used to identify residues that (i) are close enough in space to enable covalent bond formation between LG pairs in one enzyme conformational state (e.g., the closed conformation of POP), (ii) are solvent exposed to allow the proper chemistry to take place, (iii) undergo suitable changes in C β distance to prohibit bond formation in a different conformational state (e.g., the open conformation of POP), and (iv) are placed on opposite domains of the enzyme . Because we were targeting metal regulation using genetically encoded bipyridinylalanine (BpyAla) residues, 47 the structure of a Zn(Bpy) 2 complex was analyzed to establish that a distance of approximately 10.5 Å would allow for formation of a metal bis-BpyAla complex, M(BpyAla) 2 , at the interface of the β-propeller and peptidase domains of POP .
- how are the 180,000 potential double variants derived
The 180,000 potential double variant number was an approximation derived by considering all possible double mutants of POP. In general, this is n(n-1)/2 where n is the number of residues in the protein. As POP is a 616 amino acid long protein, the number of double mutants is equal to 616(616-1)/2= 189,420 combinations. We have edited 180,000 to the exact number:
POP is a 72 kDa protein with 189,420 possible double variants.
ii) Results, chapter "HTE screening of POP switches", 1st paragraph The authors state that all POP variants were produced at full-length and without truncation products (legend to . However, supplementary does not prove this statement. In the autoradiogram in b) no protein marker was used to show that the molecular weight range is covered. In fact, some weak bands are visible for the majority of variants, which are missing for the wt POP.
We agree that analyzing the amount of full-length and truncated protein is crucial when analyzing the incorporation of noncanonical amino acids. To address this, we have edited Supplemental to feature a side-by-side comparison of the autoradiogram and the original dried SDS-PAGE gel which shows the relevant molecular weight ranges and markers (shown below). These gels and autoradiograms show that most smaller bands are common between all variants, which may suggest general protease activity within cell extracts. Bands suggesting truncation products, which generally are between 20 -25 kDa or around ~59 kDa, do not appear to be strongly present. To clarify the presence of these smaller bands, we have edited the text in the figure legend of Supplemental to read: The reviewer is correct in that the MS data as presented does not cover the range of potential truncation fragments. However, we note that no truncation products would be expected because POP variants were purified using a C-terminal StrepII tag, which would only purify full-length proteins. We have reanalyzed several of the variants as requested and show that this is the case. Only peaks above 20% relative ion counts were labelled, and no peaks of major truncation products were observed in any of the MS traces. Truncation product weights are included below. As the MS data would not be expected to show truncation products, we do not feel that this reanalyzed data would add to the manuscript and have not edited Supplemental further. However, we have reiterated our purification approach in the text by adding additional detail in the Supplemental Methods: Interestingly, as is apparent from gels of POP variants purified by IMAC using Ni-NTA resin, (e.g. samples used for data collection presented in Functional characterization of a BpyAla POP Switch), truncated protein is observed. This is different than what was seen with proteins purified for HT analysis (e.g. POP purified by Strep-Tactin XT Spin Columns), which contained no significant amount of truncated protein. We have previously found that truncated POP has affinity for Ni-NTA resin even if it doesn't contain a His 6 tag. This phenomenon has been known to our lab for several years (first discovered by those working on developing POP artificial metalloenzymes) and has been validated by several students.
## Supplementary figure 2. bpyala incorporation into pop.
iii) Results, chapter "HTE screening of POP switches", 1st paragraph A high-throughput screening system was set up to "enable screening and characterization of hundreds of enzyme variants and reactions in less than a day." However, the authors have produced 27 and not hundreds of variants to test.
We thank the reviewer for the feedback. The reviewer is correct in stating that hundreds of protein variants were not tested within this work. Our intent was to convey that we could test hundreds of candidates with the workflow if there is a suitable activity assay. Indeed, cell-free systems have been used to assess even millions of reactions (10.1038/s41467-020-15798-5). However, as hundreds of variants were not tested in this work, we have edited the sentence to read:
The HTE workflow is purification-free and could allow for analysis of hundreds of reactions in less than a day, with the potential to screen large numbers of variants with a suitable functional assay.
Moreover, with regard to the ambition to apply the approach to any protein, it will be difficult or hardly possibly for many other enzymes to set up such a HTE screening since not all reactions allow a spectrophotometric detection. It should be discussed in the discussion section how this step can be adapted for other proteins e.g., for those without the possibility to monitor activity spectrophotometrically.
We thank the reviewer for highlighting this point of clarification and agree that enzyme activity cannot always be directly measured spectrophotometrically, as was done for POP and pnitroaniline. This, of course, is a challenge for any enzyme and not specific to our method for controlling enzyme activity. That said, we note that our approach is amenable to a variety of other detection methods, including fluorescence and luminescence (as was done for Pluc), that are compatible with modern instrumentation such as plate readers. In addition, any molecule that can be measured spectrophotometrically, such as NADH or other molecules associated with enzymes, could be used to monitor activity. We now discuss these challenges and opportunities in the discussion section as the reviewer requests. Specifically, we write: We thank the reviewer for pointing out this inconsistency in how we selected variants to describe the text. We have now only highlighted variants discussed in text, referencing all else in SI. Changes include:
[formula] While [/formula]
(1) Focused results entirely on five variants, four of which were deactivated (167/517, 169/510, 159/517, and 169/512) and one of which was activated (167/513) to different extents in the main text:
We first used this workflow to screen POP variants for Ni(II) inhibition in the presence of 0 -1,000-fold molar excess of Ni(II) relative to enzyme
## We next characterized pop x/y responses to different divalent metal cations, including cu(ii), co(ii), zn(ii), and fe(ii) (figure 3b and supplementary figure 3). while wt pop did not respond to these metals, pop 169/510 , pop 159/517 , pop 167/517 , and pop 169/512 were inhibited by cu(ii), co(ii), and zn(ii) in a dose-dependent manner, but only the latter two were inhibited by fe(ii). pop 167/513 , which was activated by ni(ii), was also activated by cu(ii) and co(ii) but not zn(ii) or fe(ii) (figure 3b, supplemental figure 3). these differences in extent of activation and inhibition may reflect different preferred coordination geometries of the m(bpy) n complexes or the relative binding affinities of the bpy lg for the different metals. 51
The reversibility of POP variant inhibition/activation was examined by adding a competitive chelator, ethylenediaminetetraacetic acid (EDTA), to reactions conducted using the workflow outlined above .
(2) Highlighted and included fold improvement numbers on 169/510, 169/512, and 167/513, in and Supplemental so that all five featured variants are presented in the same way. (3) Removed mention of 156/513 from the text to focus on the five variants noted above (4) Added data for 169/512 to so that all five featured variants are presented in the same way in . (5) Because there is not room or need to include steady state kinetics for all five variants in the manuscript, we removed 159/517 from so that this figure only shows a single example of the comparison in steady state kinetics been the best deactivated variant (167/517) and WT. (6) Additionally, as we discuss more in detail below, we have added data for activity in the presence of Fe(II) in . (7) Steady state kinetics for the other four variants noted above are mentioned in the text and data for these are provided in the SI. Additionally, kinetics for the 1:1 mixture of POP 167 and POP 517 are shown in this figure and table as was suggested by reviewer 2. Supplementary was edited to reflect this, and Supplementary [fig_ref] Table 2: Additionally, we have moved kinetic parameters for variants POP 167 and POP... [/fig_ref] was added as follows:
## Figure 3. kinetic analysis of pop x/y variants in response to divalent metals. (a,b) the activity of pop x/y variants from cfps was assessed on z-ala-pro-pna. (a) reaction rates (au/min) at increasing ni(ii) concentrations (0-100 µm
## Supplementary figure 5. kinetic traces for bpyala pop variants inhibited by ni(ii).
## Steady-state kinetic assays were performed at 85 °c with purified enzymes (~10-30 nm) in either the presence of 1 mm edta (black) or 5 µm nicl 2 (green). initial rates of reaction (µm/sec) are plotted versus substrate concentration (mm) and data was fit, when appropriate, with the michaelis-menten equation (equation 2
). The fold-change between reaction rates at 1 mM substrate is shown. Each data point represents the average of 3 replicates (n=3) and error bars represent standard deviations. and edited the figure caption as follows:
## Supplementary
## Supplementary figure 6. representative sds-page gel of purified pop variants.
SDS-PAGE analysis of (top) POP .
(9) Clarified in the main text that further characterization of 167/517 (including kinetic analysis of the single-point mutants e.g. POP 167 and POP 517 ) was based on the fact that this variant that had best control factor:
Particularly notable is POP 167/517 , which displayed a 17-fold change in activity in response to Ni(II) . but Fe(II) data are missing. The data should be provided before publication.
## Vi) results, chapter "functional characterization of a bpyala pop switch", last paragraph i understand why the authors used iron to visualize the metal-bpyala linkage in the uv/vis. however, to correctly conclude that the formation of the linkage is also the cause for a drop in activity, the authors must show that inhibition of pop also works with fe(ii). they have shown successfully that inhibition works with ni(ii), cu(ii), co(ii) and zn(ii) in supplementary
We understand the reviewer's concern about missing data for kinetic inhibition of variants of interest by Fe(II) and we appreciate the opportunity to add additional data that we believe strengthens our paper. We have conducted additional experiments to screen how Fe(II) may affect enzyme behavior in the panel of enzyme variants. . We find that two of our best-performing variants, POP 167/517 and POP 169/512 meet our criteria for < 50% apo activity at the highest concentration of Fe(II) tested in this experiment. As before, these metal concentrations were not observed to inhibit the wild-type enzyme. We note that POP 167/517 was the variant that was taken for further characterization, and that it was responsive to Fe(II) in a manner similar to other metals. We find that response to Fe(II) is less common throughout the panel of variants tested. Fe(II) has a significantly weaker constant of formation for M(Bpy) 2 complexes at K ~10 8 compared to Ni, which is on the order of 10 14 . This weaker constant of formation may explain the observation that Fe(II) is less effective at inhibiting POP variants in this experiment, particularly at intermediate doses of Fe(II). . Interestingly, we found that inhibition by Fe(II) is not reversible under the same conditions used for Ni(II), Cu(II), Co(II), and Zn(II). This is supported by previous literature, in which phenanthroline was necessary for Fe(II) removal from BpyAla complexes (10.1073/pnas.1600188113).
## We have also conducted experiments to test for reversibility of fe(ii)-based inhibition of the panel of enzyme variants (supplementary
## Supplementary figure 3. high-throughput screening of pop activity in response to different divalent metals.
## High-throughput experiments for pop rates in response to a range of (a) cu, (b) co, (c) zn, and (d) fe concentrations show that m(ii) ions can inhibit and activate select pop
variants in a dose dependent manner. Data represent mean standard deviation from n=3 replicates and were replicated in two independent experiments.
## Supplementary figure 4. high-throughput screening of reversibility of metaldependent activity changes in pop. high-throughput screening of reversibility shows that addition of edta to pop enzymes that are pre-treated with (a) ni, (b) cu, (c) co,
## (d) zn, and (e)
Fe can restore apo-enzyme activity. Data represent mean standard deviation from n=3 independent replicates and were replicated in two independent experiments.
We have added the following text to address the Fe(II) data:
While WT POP did not respond to these metals, POP . These differences in extent of activation and inhibition may reflect different preferred coordination geometries of the M(Bpy) n complexes or the relative binding affinities of the Bpy LG for the different metals 52 .
## Were inhibited by cu(ii), co(ii), and zn(ii) in a dose-dependent manner, but only the latter two were inhibited by fe(ii). pop 167/513 , which was activated by ni(ii), was also activated by cu(ii) and co(ii) but not zn(ii) or fe(ii) (figure 3b, supplemental
## Recovery of activity was not observed following addition of edta to variants treated with fe(ii), perhaps due to the relatively low affinity of edta for fe(ii) relative to the other metals tested
Along with data showing inhibition of POP 167/517 by Fe(II) , these data suggest that bidentate metal coordination by two Bpy LGs to generate a M(II)(BpyAla) 2 linkage as shown in favors the closed conformation of POP 167/517 and prohibits catalytic turnover. We agree that this needed to be resolved and thank the reviewer for suggesting we repeat these experiments at lower concentrations. As such, CD were collected at 10 nM for POP variants WT, 167, 517, and 167/517 in the absence or presence of Fe(II) using the same exact protein stock solutions (stored at -80 °C) used to collect the older data. The voltage traces for these samples were more reasonable and did not indicate any overloading as was the case for data collected with 5 µM protein. Moreover, these spectra closely match previously collected CD of Pfu POP (https://www.nature.com/articles/ncomms8789) and also resemble those for a related POP from porcine brain (https://www.sciencedirect.com/science/article/pii/S2451945619302715?via%3Dihub).
## Vii) si, chapter "physical characterization of pop variants -circular
We have revised Supplementary (now with these data as shown below:
## Supplementary figure 21. cd of various dilute pop samples.
CD spectra were collected of POP (all at 10 nM) in the absence or presence of Fe(II). At the higher concentrations (50 µM) necessary to visualize certain features associated with the metalloprotein (i.e. Cotton effects near 300 nm), the signal was too high to clearly observe secondary structural elements of the protein present in the far-UV region, thus we measured spectra at lower concentrations as well. A deep peak around 210 nm, characteristic of folded POP was observed.
In the original , there were differences between signal at 220 nm. As was suggested by the reviewer, this can be attributed to differences in protein concentration or small changes in secondary structure content due to protein degradation. However, upon repeating this assay at lower concentrations (10 nM vs. 5 µM), the differences were not as large, and perhaps were exaggerated when assays were conducted beyond the linear response region at higher concentrations (i.e. 5 µM) as would be evident by large High Tension Voltage values.
## Minor points
Introduction, last paragraph: "…used MD simulations to identify all residue pairs that undergo significant changes… during catalysis." -Rephrase. To my knowledge the actual catalysis event cannot be simulated with MDs.
## Updated as requested:
Our framework used molecular dynamics (MD) simulations to identify all residue pairs that undergo significant changes in C β distance between POP conformations.
Introduction, last paragraph: "…to select functional candidates for analysis." -Clarify what is meant by "functional candidate".
## Updated as requested:
Enzyme variants containing LG pairs at candidate sites were then designed and rapidly prototyped using cell-free protein synthesis (CFPS) 44 and high-throughput experimentation (HTE) to select active enzyme variants with metal-dependent activity for analysis.
Results, chapter "Model selection and design of POP switches", 1st paragraph -The authors talk once of "domains" and then again of "subunits". Clearly the β-propeller and the peptidase parts of POP are domains since they are located on the same protein chain. The term subunits is used for two separate protein chains. Replace "subunits" with "domains".
Updated as requested.
Results, chapter -Model selection and design of POP switches‖, -which PDB was used for the representation of POP? Updated as requested:
(b) Distance and geometry requirements for Bpy LG placement were obtained by mapping the structure of a Zn(Bpy) 2 (CCDC:756656) complex onto the Pfu POP structure (PDB ID: 5T88). A distance of 10.5 ± 0.5 Å between the C-5 substituents on the two Bpy ligands and the Cβ atoms of the selected residues was used to ensure that Zn(BpyAla) 2 complex could be accommodated between the residues.
in a) "Placement between domains" has not been mentioned as a criterium in the main text Updated as requested:
Structural analysis and MD simulations were used to identify residues that (i) are close enough in space to enable covalent bond formation between LG pairs in one enzyme conformational state (e.g. the closed conformation of POP), (ii) are solvent exposed to allow the proper chemistry to take place, (iii) undergo suitable changes in C β distance to prohibit bond formation in a different conformational state (e.g. the open conformation of POP), and (iv) are placed on opposite domains of the enzyme . would help to understand how the 10.4 Å distance was derived.
## -in b) is the zn(bby)2 ligand correctly oriented? further guidance including atom-wise coloring and orientation similar to the bby-ala in
Updated as requested: was updated so that: the protein was oriented in a similar way as presented in ; the Bpy ligands on the Zn complex were colored crimson red; the Cβ atoms of the residues on which the Zn complex is mapped were colored crimson; the metal of the Zn complex was colored gray; and finally, the other cis ligands were colored gold to distinguish them from the Bpy ligands. See above for the updated figure caption.
## Results, chapter "model selection and design of pop switches", 2nd paragraph "(i) … in one enzyme conformational state," -in which state? closed or open? "(ii) undergo suitable changes in cβ distance to prohibit bond formation a different conformational state." -clarify this point. i understand this criterium that metal binding should only occur in one conformation but not the other and that hence the two bby-ala positions need to be close in one conformation but distant in the other.
We have parenthetically noted the relevant states in the case of POP. Different states may be relevant to different enzymes where -open‖ or -closed‖ are not appropriate descriptors:
## Structural analysis and md simulations were used to identify residues that (i) are close enough in space to enable covalent bond formation between lg pairs in one enzyme
It has been previously shown by others that the kinetics of luciferase can be affected by varying the concentrations of its substrate including D-luciferin and ATP. Byproducts of the reaction like oxyluciferin and L-AMP can inhibit the enzyme (https://www.sciencedirect.com/science/article/pii/S1074552110001973?via%3Dihub). This could be the reason we observe lower activity at increasing substrate concentrations. As we do not know the exact reason for this behavior, we have modified the text and now include this reference in the caption for In VMD, once the structure and connectivity of the protein and the desired side chain were settled, you can mutate the side chain as the amino acid residue.
-What is meant by "cross-linked luciferase"? Please explain.
The cross-linked luciferase referred to in the Supplementary Information comes from (https://pubs.acs.org/doi/10.1021/bi300934s), which used a chemical cross-linker to trap Pluc in a conformation that was incompetent for the adenylation reaction but competent for the oxidative reaction. We used this crystal structure to design luciferase variants that were trapped in the oxidative-competent conformation, which would render Pluc sensitive to metals by preventing the adenylation reaction from occurring. We treated cross-linked structure as another confirmation of the wild-type (by mutating the residues from cross-linked one to all wild-type one, excluding the ones that are highly conserved, while keeping the backbone structure as the cross-linked one. The initial structures of both have to go through MD simulations). We thank the reviewer for noticing this oversight in the SI. We also purified POP159/517 in the same manner as those other variants mentioned here and have edited the SI to reflect that.
We further purified these variants of interest with SEC for the purposes of intact ESI-MS and crystallization efforts (the latter of which has failed thus far) where a secondary purification step was desirable. Diafiltration was performed extensively for the other variants so that the theoretical concentration of EDTA was in the nM regime. This small amount of EDTA should not have a significant effect on the kinetic assays which were either performed with an excess of EDTA (1 mM) or Ni(II) (5 µM).
-Why was POP transferred to pure H2O. Most proteins require a buffer system for stability, why not POP?
We have extensive experience working with Pfu POP and have found that it is thermostable and tolerates storage in MQ H 2 O.
-Why does show exemplary kinetic traces for POP167/513 which is never mentioned in the main text?
We have corrected the main text (see above) so that this activating variant is referenced to, and data presentation was made consistent with other highlighted variants.
-The data for In general, reversible switching of enzyme activity almost always involves a reduction in the activity of the WT enzyme. In an earlier study, we provided simulations that suggest that the catalytically essential H592 residue hydrogen bonds with a number of residues, some of which were targeted in the current study, as it transits from its orientation in the open and closed forms of the enzyme (https://pubs.acs.org/doi/abs/10.1021/acs.biochem.9b00031). Disrupting those interactions could cause problems with catalysis. Given that this is only speculation, as the reviewer suggested, and given the complexity of those interactions and their relevance to catalysis even if the speculation were to prove accurate, we think the best approach to correcting this issue would be via directed evolution (and this would also apply for luciferase, where we saw larger decreases in apo activity upon incorporation of BpyAla). Screening for activity in the presence and absence of metal ions would allow for identification of enzymes whose activity is not compromised relative to parent in the absence of metal and improved relative to parent in the presence of metal. This is an approach we plan to pursue in due course.
2 We have grouped these variants as suggested into general categories (Active/Inactive, and Inhibited/Activated/Non-responsive to Ni) along with information about the distance between the two BpyAla residues in the Closed conformation as well as the change in distance between the residues from the Open to Closed conformations (see below in response to point number 10).
We are unable to notice any obvious trends that can explain why certain sites afford better switching activity compared to others. The only thing that was apparent from this presentation of the data was that mutations from H442 and I514 were not tolerated well and led to loss of activity. Perhaps, geometric information (e.g. orientation of the BpyAla residues) could reveal trends in performance, however, we did not include these constraints in our selection of sites for BpyAla incorporation, but are interested in pursuing in further studies with similar systems.
3) 167/513 and 170/513 show inverse metal-dependent activity in that they exhibit even accelerated catalytic activity upon the addition of metal ions. How is it even possible?
We, too, have wondered about this, and thank the reviewer for the opportunity to speculate on the nature of this unexpected finding. Indeed, we were surprised to find that POP 167/513 and POP 170/513 were activated in response to metal, rather than inhibited. This result was independently seen within two separate laboratories, confirming the activation. We conducted kinetic assays on POP 167/513 which show improvements in both K M and k cat in the presence of Ni (II) (Supplementary and Supplementary [fig_ref] Table 2: Additionally, we have moved kinetic parameters for variants POP 167 and POP... [/fig_ref]. This suggests that the Ni (II) ion may play a role in stabilizing the active site (thus increasing k cat ) and in improving binding of the substrate to the active site (thus decreasing K M ).
## 4) to my best knowledge, pop is a dimeric protein.
Therefore, the addition of bpy-Ala residue at one position of a protomer would form not only M(bpy)2 moiety but also two distantly located bpy residues. Although it would influence both structure and function of the proteins, the authors neither described nor discussed this aspect in this manuscript. Because these residues are exposed to the protein surface, the latter bpy-Ala may form intermolecular complexes, possibly generating undesirable high-ordered species. Is it why the SDS-PAGE analysis was carried out in the presence of excess EDTA? Perhaps, the formation of high-ordered species might be related to their distinct response to the metal ion. Do authors examine the oligomeric structures of the proteins upon the addition of metal ions? Perhaps, size exclusion chromatography or analytical ultracentrifugation in the presence of Ni can be carried out to quantify the fraction of catalytically active and metal-responsive species.
We thank the reviewer for this question. Pfu POP is a monomeric protein with two domains, a βpropeller domain which restricts substrate access to the active site and a peptidase domain which contains the catalytic residues. We apologize for the confusion regarding the conditions under which SDS-PAGE analysis was performed under. The SDS-PAGE data presented in the SI was obtained with protein that had been purified by SEC in the presence of EDTA; the samples were subsequently buffer-exchanged into water prior to SDS-PAGE analysis and gels were not run in the presence of metal chelator.
We have not run size exclusion chromatography in the presence of divalent metals before, but certainly, that could be informative in future studies focused on structural characterization of these BpyAla enzyme switches. We have, however, run SEC in the absence of EDTA with the Fe-bound protein isolated by NiNTA, and the chromatograms were nearly identical in nature compared to those run after treatment with EDTA/phenanthroline or those eluted in the presence of a chelator. Given that the proteins purified by SEC without chelator retained their metal, apparent by their pink color, coupled with the similarities between chromatograms from SEC run with/without chelator, would suggest similar assembly states of the apo and metalloproteins. We do not rule out the possibility that higher-ordered species could form, however to date, we have no evidence for the formation of higher-ordered species as the reviewer has suggested. We would therefore prefer not to speculate on potential effects of their formation. We thank the reviewer for this question, and as such have performed the suggested experiments. Like the kinetics of POP 167 and POP 517 , the kinetics of a 1:1 mixture of POP 167 and POP 517 showed little inhibition in the presence of Ni(II), suggesting that the metal-dependent inhibition of POP 167/517 is due to an intramolecular interaction. We have included these data in SI and SI , consistent with the proposed intra-protein cross-link rather than potential inter-protein interactions that could also result in inhibition.
## 5) if the intermolecular complexation occurs
6) The authors characterized one of the POP variants by UV-Vis and CD spectroscopy, demonstrating that Fe would be a surrogate metal element to show Ni-dependent chemical switch. However, neither ferric nor ferrous ion was not applied for this system in for no apparent reason. Please provide the catalytic activities of POP variants with Fe ion.
Reviewer 1 also raised this important concern. We have included data for activity in the presence of Fe(II) in and . Please see response above.
## 7)
Please include why the authors used a cell-free expression system instead of standard e coli heterologous expression. Is it to increase protein yields of having two bpy-Ala residues in a protomer? More detailed information in the experimental setup might be necessary for readers.
In this work, we decided to use a cell-free expression system for several reasons. Firstly, as the reviewer identifies, CFPS is volumetrically efficient and rapid for BpyAla incorporation, as ~ 1,500 μg 2TAG-sfGFP / mL of CFPS reaction can be obtained overnight. Secondly, the CFPS reaction is an open reaction environment, allowing for enzymatic assays to be conducted within the enzyme-enriched CFPS reaction by simply adding appropriate buffers and substrates without concern for factors such as membrane permeability. This importantly avoid tedious processing steps necessary for in vivo expression in E. coli for similar lysate-based enzyme assays, such as washing, centrifuging, and lysing, which must be repeated for every enzyme variant and thus becomes quickly untenable.
## 8) among the divalent metal ions that the authors used, the cupric ion is redox-active under physiological conditions, and reduction may lead to significant changes in the geometry of m(bpy)2 and subsequent metal-dependent catalytic activities.
Because the catalytic activities of POP are independent of dioxygen, the function of the metal-dependent switch can be monitored under anaerobic conditions. Can authors explore whether these systems are redox-dependent?
We agree this could be an interesting future direction. At this point, however, we would prefer to avoid this additional complexity.
## 9) do authors only consider the distance between two residues for md simulation? what about the directionality of the side-chains? now that the authors have both simulation and experimental data sets, can any conclusion or criteria in the determination of two bpy-positions can be drawn and reassessed?
This is also an excellent suggestion. We did not consider coordination geometry in our original designs, and that could certainly explain differences that we see between different divalent metals. Ongoing studies are focused on obtaining crystal structures of different LG variants with different divalent metals. We think this will be essentially to understanding the extent to which the metal ion and the protein scaffold dictate the observed coordination geometry and thus the extent to which these considerations are important for LG design. Further analysis of our simulation data could also reveal differences in the effects of side chain orientation on switch behavior. Particularly if coupled with analysis of improved variants obtained via directed evolution, this approach could provide a wealth of information regarding the possibility of rational switch improvement. has no information related to residues. Please provide the residue labels. In addition, add the distance between i and j residues in Supplementary We used excess metal elements in steady-state kinetic assays to ensure saturating concentrations were reached and that most enzyme was in a metal-bound state. As we observed minimal inhibition of the WT enzyme, there was no harm in using excess amounts for this purpose.
## 10)
We agree with you regarding the use of minimal amounts of metal in the reversibility experiment. We tried to user lower concentrations of metals when first optimizing this experiment, however at the high temperatures which were used for kinetic assays, we observed incomplete inhibition of the enzyme, so the concentration was increased until similar levels of inhibition compared to steady-state experiments were observed.
Screening data was consistent with different binding affinities among the BpyAla variants, as variable amounts of metal were required to reach full inhibition depending on the variant and metal identity (see . We are interested in characterizing these affinities more rigorously to accurately compare K d values and are currently pursuing this, however this would be featured with other experiments focused on better understanding why certain switches performed better than others and how to better improve our site-selection criteria (e.g. microscopy to look at distribution of conformations, crystallography, FRET, etc.).
## Reviewers' comments:
Reviewer #1: Remarks to the Author: I congratulate the authors to this successfully revised version of their manuscript. I like to thank the authors for their meticulous efforts to answer my questions and incorporate my suggestions.
All of my concerns have been taken care of and the updated article is nice to read, consistent and tells a really exciting story.
As a result of rereading the manuscript, I have only some further minor points for improvement: -The story suffers from a missing description of the design principle of BpyAla. Although it is shown in , it is not explained in text form neither in the main text nor in the figure caption. The authors should supplement such a description in the caption to and at least mention the design principle in the first chapter of the results in the first or at the beginning of the second paragraph. Otherwise, the reader will not understand why double variants are required for the design strategy (p. 4, l.142). Similarly, would benefit from a detailed description in the caption.
-P.1, l. 46: I believe "however" should be positioned after "… in a stimulus-dependent manner" -P.2, l.58: LOV stands for "light-oxygen-voltage" and not "light, oxygen and voltage" -P. 2, l. 63: It should be "Arabidopsis thaliana" (in italic) -P.4, l.145 "… that undergoes catalytically relevant conformation changes." This is too vague. Many different kinds of conformational changes exist and I believe not all of them can be manipulated with the authors' technique. I suggest to specify (e.g. "relevant conformational domain change").
-P.7, l. 238 : What is the authors' definition of "minimal"? Please specify using approximations to avoid misunderstanding, e.g.: o Wt ~95% pure (ok) o POP167 ~90% pure and <5% truncation product (ok) o POP517 ~85% pure and ~10% truncation product (still ok) o POP167/517 ~90% pure and <5% truncation product (ok) o POP159/517, POP169/510, POP169/512, POP167/513 ~60% pure and 30-40% truncation product ◊ in my opinion not "minimal"; they authors might want to add their explanation given in the rebuttal that truncated POP shows affinity for Ni-NTA -P.7, l. 241, Supplementaryis fine since the variants are >85% pure) ◊ or give in general vmax to stay consistent -P.7, l.243: kcat and Km could not be measured in presence of Ni(II) ◊ for quantitative comparison, the authors could compare kcat/Km as here they only need to fit the initial linear part of the Michaelis-Menten curve Reviewer #2: Remarks to the Author:
The authors addressed all essential concerns and questions and revised the manuscript accordingly. Although the strategy developed herein as "Metal-Responsive Regulation of Enzymes" can be applied for specific systems for now, and thus it needs further improvements for in vivo experiments or else, the proof of concept is noteworthy, and they demonstrated the experiments thoroughly. Therefore, I believe that it is suitable for publishing in Nature Communications.
Reviewer #1 (Remarks to the Author):
-The story suffers from a missing description of the design principle of BpyAla. Although it is shown in , it is not explained in text form neither in the main text nor in the figure caption. The authors should supplement such a description in the caption to and at least mention the design principle in the first chapter of the results in the first or at the beginning of the second paragraph. Otherwise, the reader will not understand why double variants are required for the design strategy . Similarly, would benefit from a detailed description in the caption.
The manuscript notes that the Bpy LG works via bis-bidentate metal binding on P2, Line 91:
Initial efforts focused on bipyridine (Bpy) LGs that could induce reversible activation of enzymes in the presence/absence of metal salts . Specifically, bis-bidentate metal binding by the Bpy LGs would restrict the enzyme to a closed conformation incapable of turnover while removal of the metal would allow the enzyme to access conformational changes required for turnover.
To make this more clear in , we modified the figure description as follows:
(a) Reversible metal regulation of Pfu prolyl oligopeptidase (POP) via bis-bidentate metal binding by a BpyAla LG pair.
-P.7, l. 238 The reviewer is correct that some of the variants that were lumped into this general statement had impurity levels beyond what can reasonably be considered "minimal". We have rewritten this passage to state the full range of purity observed and to specify the high purity of the key variants as follows:
Purified variants were produced in yields of approximately 50 mg/L, and BpyAla incorporation was confirmed by intact protein ESI-MS. The isolated variants included ~5-40% of POP that had been truncated at the first amber stop codon due to the inherent affinity of POP for Ni-NTA resin, though >90% purity was observed for POP 167/517 , POP 167 , and POP 517 .
The figures caption for was also updated (relevant section only):
In addition to a band corresponding to the MW of the desired full-length protein, we also observe varying amounts of the truncated protein depending on the variant (<5% -~40%), that forms because of competition between release factor 1 mediated (RF1) termination of translation and BpyAla incorporation at the amber stop codon 1 . Truncated protein was isolated along with full-length protein after IMAC due to apparent affinity of the truncated POP side products for Ni-NTA.
[fig] : C. Hoppmann et. al. J. Am. Chem. Soc. 2015, 137, 11218. A. C. Kneuttinger et. al. Cell Chem. Biol. 2019, 26, 1501-1514.e9.; A. C. Kneuttinger et all Biochemistry 2020, 59, 2729. [/fig]
[table] caption: Figure 2. Design strategy for Bpy-Ala POP variants. (a)Filtering process used to select sites for Bpy LG pairs. (b) Distance and geometry requirements for Bpy LG placement were obtained by mapping the structure of a Zn(Bpy) 2 (CCDC:756656) complex onto the Pfu POP structure(PDB ID: 5T88). A distance of 10.5 ± 0.5 Å between the C-5 substituents on the two Bpy ligands and the Cβ atoms of the selected residues was used to ensure that Zn(BpyAla) 2 complex could be accommodated between the residues. (c) Locations of the final 27 LG pairs. The spheres represent the positions of the Cβ atoms of the selected residues on the β-propeller (black) and peptidase (grey) domains and labels represent residue numbers. Lines are drawn between residues in each pair, and the color of the line represents the change in distance between the residues in the open and closed POP conformations.• why were MD simulations required when structures of both open and closed conformations of POP and luciferase existA structure of Pfu POP in the closed conformation does not exist (our structure, which was open, was the first reported), so MD simulations were required. Even though structures of Luc in different conformations do exist, we wanted to have a general protocol that would work for enzymes of interest. As noted above, we edited the sentence that clarifies this so that it specifically refers to conformationally dynamic systems:As many proteins of interest could have a similarly large number of potential LG site pairs, or more, we developed a computationally guided approach for identifying sites for LG incorporation that could be applied to any protein that undergoes catalytically relevant conformation changes. [/table]
[table] Table 2: Additionally, we have moved kinetic parameters for variants POP 167 and POP 517 toTable 1in the main text as suggested below, under Minor Points. Additionally, errors for determination of kinetic parameters from fitting with the Michaelis-Menten equation have been added to the table. [/table]
|
Peripheral arterial disease (PAD) is associated with considerable morbidity and mortality. Consensus rightly demands the incorporation of supervised exercise training (SET) into PAD treatment protocols. However, the exact role of SET particularly its relationship with intervention requires further clarification. While supervised exercise is undoubtedly an excellent tool in the conservative management of mild PAD its use in more advanced disease as an adjunct to open or endovascular intervention is not clearly defined. Indeed its use in isolation in this cohort is incompletely reported. The aim of this review is to clarify the exact role of SET in the management of symptomatic PAD and in particular to assess its role in comparison with or as an adjunct to invasive intervention. A systematic literature search revealed a total 11 randomised studies inclusive of 969 patients. All studies compared SET and intervention with monotherapy. Study results suggest that exercise is a complication-free treatment. Furthermore, it appears to offer significant improvements in patients walk distances with a combination of both SET and intervention offering a superior walking outcome to monotherapy in those requiring invasive measures.
# Introduction
Peripheral arterial disease (PAD) affects 12-16% of the population over the age of 60 years with intermittent claudication (IC), its primary symptom, proving detrimental to patient quality of life [bib_ref] The prevalence of peripheral arterial disease in a defined population, Criqui [/bib_ref] [bib_ref] Estimation of peripheral arteriosclerotic disease by ankle blood pressure measurements in a..., Schroll [/bib_ref] [bib_ref] Prevalence of and risk factors for peripheral arterial disease in the United..., Selvin [/bib_ref] [bib_ref] Correlating clinical indicators of lower-limb ischaemia with quality of life, Chetter [/bib_ref]. Typically PAD follows a stable course with management confined to conservative measures; however one in ten PAD patients will develop critical limb ischaemia (CLI) with all-cause mortality in the CLI cohort rising to 50% at 5 years [bib_ref] Natural history of arteriosclerosis of the lower extremities-hunterian lecture delivered at the..., Bloor [/bib_ref] [bib_ref] The fate of patients with critical leg ischemia, Dormandy [/bib_ref] [bib_ref] Bypass versus angioplasty in severe ischaemia of the leg (BASIL): multicentre, randomised..., Bradbury [/bib_ref]. This reduction in life expectancy is due largely to concomitant cardiovascular disease [bib_ref] Mortality over a period of 10 years in patients with peripheral arterial..., Criqui [/bib_ref] [bib_ref] Epidemiology of atherosclerotic arterial disease in the lower limbs, Fowkes [/bib_ref].
Therefore, treatment goals should focus not only on the alteration of disease progression and symptomatic relief but also on the improvement of patient long-term survival [bib_ref] Management of peripheral arterial disease (PAD), Dormandy [/bib_ref]. Approaches include the modification of risk factors through optimum medical therapy (OMT) and supervised exercise therapy (SET) with endovascular (EVR) and open surgical revascularization reserved for those failing conservative measures. Further novel therapies including kinesitherapy and electrotherapeutic procedures have also been proposed [bib_ref] The role of kinesitherapy and electrotherapeutic procedures in nonoperative management of patients..., Markovic [/bib_ref]. While endovascular treatment offers a minimally invasive revascularization option for many patients data supporting its ability to improve long-term survival is lacking. Regular exercise, on the other hand, is associated with a 50% reduction in cardiovascular mortality [bib_ref] Physical activity and health, Powell [/bib_ref]. Supervised exercise training consists of a prescribed, evidence based exercise program which is performed under the direct observation of a trained practitioner. It is now well established as an initial noninvasive option in all PAD patients with robust supporting data [bib_ref] ACC/AHA 2005 guidelines for the management of patients with peripheral arterial disease..., Hirsch [/bib_ref] [bib_ref] Inter-society consensus for the management of peripheral arterial disease (TASC II), Norgren [/bib_ref] [bib_ref] Exercise rehabilitation programs for the treatment of claudication pain. A metaanalysis, Gardner [/bib_ref].
Rationale for Review. Despite its intuitive benefits a wide variation in the use and availability of SET exists and it remains a greatly underutilized resource due to limited patient access [bib_ref] Exercise for intermittent claudication. Supervised programmes should be universally available, Stewart [/bib_ref]. The BASIL study highlights the deficiencies in the current medical optimization with few participants utilizing 2 Surgery Research and Practice clinically proven best medical therapy [bib_ref] Bypass versus angioplasty in severe ischaemia of the leg (BASIL): multicentre, randomised..., Bradbury [/bib_ref] [bib_ref] Management of peripheral arterial disease in primary care, Burns [/bib_ref]. The issue is further clouded by conflicting literature as to the optimal nonsurgical management of these patients [bib_ref] Supervised exercise versus primary stenting for claudication resulting from aortoiliac peripheral artery..., Murphy [/bib_ref] [bib_ref] Is percutaneous transluminal angioplasty better than exercise for claudication? Preliminary results from..., Creasy [/bib_ref] [bib_ref] Exercise training versus angioplasty for stable claudication. Long and medium term results..., Perkins [/bib_ref] [bib_ref] The constitutive procoagulant and hypofibrinolytic state in patients with intermittent claudication due..., Hobbs [/bib_ref] [bib_ref] Intermittent claudication: clinical effectiveness of endovascular revascularization versus supervised hospital-based exercise training-randomized..., Spronk [/bib_ref] [bib_ref] Cost-effectiveness of endovascular revascularization compared to supervised hospital-based exercise training in patients..., Spronk [/bib_ref] [bib_ref] Long-term clinical effectiveness of supervised exercise therapy versus endovascular revascularization for intermittent..., Fakhry [/bib_ref] [bib_ref] Randomized clinical trial of percutaneous transluminal angioplasty, supervised exercise and combined treatment..., Mazari [/bib_ref] [bib_ref] The adjuvant benefit of angioplasty in patients with mild to moderate intermittent..., Greenhalgh [/bib_ref] [bib_ref] Additional supervised exercise therapy after a percutaneous vascular intervention for peripheral arterial..., Kruidenier [/bib_ref] [bib_ref] Twelve-months followup of supervised exercise after percutaneous transluminal angioplasty for intermittent claudication:..., Bø [/bib_ref] [bib_ref] Treatment efficacy of intermittent claudication by surgical intervention, supervised physical exercise training..., Gelin [/bib_ref] [bib_ref] Intermittent claudication-surgical reconstruction or physical training? A prospective randomized trial of treatment..., Lundgren [/bib_ref] [bib_ref] Benefits of a supervised exercise program after lower limb bypass surgery, Badger [/bib_ref] [bib_ref] Supervised exercise, stent revascularization, or medical therapy for claudication due to aortoiliac..., Murphy [/bib_ref]. Thus while its use is strongly supported by current literature it appears that SET is underutilized in mild-to-moderate PAD while its use in more advanced disease requires further clarification. With this likely in mind the Institute of Medicine has prioritized research into the comparative efficacy of the different treatment modalities for PAD . Furthermore, the role of SET as an adjunct to or substitute for intervention remains unclear. The aim of this review was to compare the use of supervised exercise therapy to invasive measures in the management of symptomatic peripheral arterial disease thus clarifying an exact role for SET in the management of this patient cohort.
# Methods
## Study eligibility.
All randomised controlled trials (RT) assessing exercise in conjunction with or in comparison to an endovascular or open intervention in the management of peripheral arterial disease were included for review [fig_ref] Table 1: Treatment groups and disease level [/fig_ref]. All observational and review data were excluded from the results. Relevant papers were searched and evaluated independently by two assessors. Outcomes were tabulated where figures were included.
## Literature
Search. The online medical literature database PUBMED was systematically searched. All studies and relevant reviews were manually cross-referenced to identify any outstanding articles.
PubMed was last searched on September 18, 2015 [fig_ref] Figure 1: Flow diagram depicting study identification [/fig_ref]. The database was comprehensively searched without date or language restriction using the following search strategy.
[
# Results
A total of 15 papers [fig_ref] Table 1: Treatment groups and disease level [/fig_ref] report outcomes of 11 RT. These trials include a total of 969 patients and all directly compare supervised exercise with various invasive interventions. Maximum walking distance (MWD), intermittent claudication distance (ICD), and ankle brachial pressure index (ABPI) measurement form the cornerstones of vascular assessment in each study.
## Quality assessment of assessed data.
The risk of bias in each included study is summarised in [fig_ref] Table 2: Assessment of bias. [/fig_ref]. Few papers reported any participant heterogeneity with regard to baseline function, comorbidities, and smoking status. Risk assessment was performed with guidance from the Cochrane Handbook for Systematic Reviews of Interventions.
## Supervised exercise versus endovascular intervention.
Five trials including 519 patients directly compare the outcomes of EVR and SET in the management of peripheral arterial disease [fig_ref] Table 3: Supervised exercise versus endovascular revascularization [/fig_ref].
At six months Murphy et al. reported significant improvements in maximum walk times in those undergoing SET compared to those in the EVR group [bib_ref] Supervised exercise versus primary stenting for claudication resulting from aortoiliac peripheral artery..., Murphy [/bib_ref]. However, at 18-month follow-up this benefit was lost with no significant difference in walk times identified [bib_ref] Supervised exercise, stent revascularization, or medical therapy for claudication due to aortoiliac..., Murphy [/bib_ref]. ABPI were consistently higher in the EVR group. Creasy and Perkins noted significant improvements in the functionality of both groups [bib_ref] Is percutaneous transluminal angioplasty better than exercise for claudication? Preliminary results from..., Creasy [/bib_ref] [bib_ref] Exercise training versus angioplasty for stable claudication. Long and medium term results..., Perkins [/bib_ref]. Again, no significant change in ABPI was noted in any SET cohort. Improvement in mobility was most significant when the disease affected the superficial femoral artery. Hobbs et al. only noted significant improvements in walk distances in those receiving EVR [bib_ref] The constitutive procoagulant and hypofibrinolytic state in patients with intermittent claudication due..., Hobbs [/bib_ref]. No improvement was seen with SET alone. Spronk et al. reported a 1-week clinical success rate of 88% following EVR decreasing to 68% at 12 months while the SET group had an early success rate of Surgery Research and Practice 3 16% increasing to 65% by 12 months [bib_ref] Intermittent claudication: clinical effectiveness of endovascular revascularization versus supervised hospital-based exercise training-randomized..., Spronk [/bib_ref] [bib_ref] Cost-effectiveness of endovascular revascularization compared to supervised hospital-based exercise training in patients..., Spronk [/bib_ref]. Clinical success was defined as an improvement in at least one category in the Rutherford scale posttreatment. Long-term outcomes (7 years) from this RT show maintenance of the functional gains achieved at 1 year with no variance in amputation rate 10 between groups [bib_ref] Long-term clinical effectiveness of supervised exercise therapy versus endovascular revascularization for intermittent..., Fakhry [/bib_ref]. Finally, Mazari et al. identified significant functional improvements in both groups at one year; however no statistically significant difference was identified between cohorts [bib_ref] Randomized clinical trial of percutaneous transluminal angioplasty, supervised exercise and combined treatment..., Mazari [/bib_ref]. Again, only EVR was associated with improved ABPI measurements.
## Supervised exercise plus invasive measures (open surgery or evr) versus monotherapy.
In total six studies inclusive of 514 patients compare the merits of combination therapy consisting of invasive intervention and SET and single intervention alone ( alone [bib_ref] Intermittent claudication-surgical reconstruction or physical training? A prospective randomized trial of treatment..., Lundgren [/bib_ref]. This study identified improvements in walk distances in both groups; however, those undergoing combination therapy experienced significantly improved walking performance at 13 months. Similarly, Badger et al. identified significant improvements in MWD in patients undergoing peripheral arterial bypass in conjunction with SET compared to those undergoing bypass in isolation [bib_ref] Benefits of a supervised exercise program after lower limb bypass surgery, Badger [/bib_ref].
More recently, studies have focused on the use of SET as an adjunct to EVR. Mazari et al. identified that a combined treatment group achieved the greatest (but not statistically significant) improvement in MWD and ICD with a lower incidence of reintervention compared to the monotherapy groups [bib_ref] Randomized clinical trial of percutaneous transluminal angioplasty, supervised exercise and combined treatment..., Mazari [/bib_ref]. This benefit was further supported by randomised data from Greenhalgh et al. [bib_ref] The adjuvant benefit of angioplasty in patients with mild to moderate intermittent..., Greenhalgh [/bib_ref] [bib_ref] Additional supervised exercise therapy after a percutaneous vascular intervention for peripheral arterial..., Kruidenier [/bib_ref] [bib_ref] Twelve-months followup of supervised exercise after percutaneous transluminal angioplasty for intermittent claudication:..., Bø [/bib_ref]. At six months Kruidenier et al. identified significantly lengthened walk distances when SET was used as an adjunct to EVR. Furthermore, both Greenhalgh and Bø et al. examined patients with both aortoiliac and infrainguinal disease in separate trial limbs. While Greenhalgh compared combination therapy with SET alone Bø et al. contrasted dual therapy with EVR alone. Both studies identified improvements in walk distances in both trial limbs for patients undergoing combination therapy versus monotherapy; however only Greenhalgh identified significantly better ABPI in the dual intervention group.
## Invasive (evr or surgery) management versus supervised
Exercise. Gelin alone compared invasive intervention (either EVR or open surgery based on preoperative angiography) and supervised exercise [bib_ref] Treatment efficacy of intermittent claudication by surgical intervention, supervised physical exercise training..., Gelin [/bib_ref] [fig_ref] Table 5: Invasive [/fig_ref]. At one year only those randomised to invasive measures experienced any improvement in walk distance or lower limb arterial pressures.
## Open surgery versus supervised exercise. lundgren et al. included a comparison of open revascularisation and set
in the previously discussed RT [bib_ref] Intermittent claudication-surgical reconstruction or physical training? A prospective randomized trial of treatment..., Lundgren [/bib_ref]. At 13 months those undergoing open surgery had better functional performance than those undergoing SET alone. In addition, those undergoing surgery experienced a significantly higher ABPI to those in the exercise group.
# Discussion
Peripheral arterial disease is a widespread phenomenon in the elderly population [bib_ref] Femoral atherosclerosis in an older British population: prevalence and risk factors, Leng [/bib_ref]. The optimum management of claudication continues to raise debate. Exercise is a straightforward and effective conservative treatment option. Meta-analysis has shown a 122% increase in walking distances in patients undergoing exercise therapy [bib_ref] Exercise rehabilitation programs for the treatment of claudication pain. A metaanalysis, Gardner [/bib_ref]. This has been reinforced by the more recent Cochrane review examining exercise for IC [bib_ref] Exercise for intermittent claudication, Watson [/bib_ref]. However, despite its success intensive programs continue to be associated with high dropout rates [bib_ref] Supervised exercise therapy for intermittent claudication in a community-based setting is as..., Bendermacher [/bib_ref].
The exact physiological mechanism by which exercise improves performance is incompletely understood. Multiple physiological adaptions have been proposed as contributing factors. Arterial collateralization has the potential to improve peripheral blood flow in the ischaemic limb with the exercised muscle displaying increased levels of the proangiogenic vascular endothelial growth factor [bib_ref] Exercise-induced expression of angiogenesisrelated transcription and growth factors in human skeletal muscle, Gustafsson [/bib_ref]. However, improved functional performance in the trained limb is not reflective of improved ABPI measurements as seen in endovascular revascularization [bib_ref] Supervised exercise versus primary stenting for claudication resulting from aortoiliac peripheral artery..., Murphy [/bib_ref] [bib_ref] Physical training of patients with intermittent claudication: indications, methods, and results, Ekroth [/bib_ref]. Increased arterial shear stress in exercise is associated with nitric oxide (NO) release, a powerful vasoactive agent [bib_ref] Nitric oxide synthesis by cultured endothelial cells is modulated by flow conditions, Noris [/bib_ref]. This endothelial-mediated response is impaired in patients with PAD [bib_ref] Endothelial reactivity and cardiac risk factors in older patients with peripheral arterial..., Yataco [/bib_ref]. The concept of "hemorheologic fitness" suggests that a proven reduction in blood viscosity in the trained individual may result in improved peripheral metabolic efficiency [bib_ref] Influence of regular physical activity on blood rheology, Ernst [/bib_ref]. In addition, gait proficiency, reversal of acquired metabolic myopathies, and modified inflammatory responses all have the potential to improve exercise related function [bib_ref] Improved walking economy in patients with peripheral arterial occlusive disease, Womack [/bib_ref] [bib_ref] Calf muscle oxygen saturation and the effects of supervised exercise training for..., Beckitt [/bib_ref] [bib_ref] Exercise training for intermittent claudication: does it adversely affect biochemical markers of..., Tisi [/bib_ref] [bib_ref] Exercise training for claudication, Stewart [/bib_ref].
This review suggests a number of roles for supervised exercise therapy in the symptomatic PAD patient [fig_ref] Figure 2: Management of intermittent claudication incorporating supervised exercise therapy [/fig_ref]. Significantly, no study reported exercise related complications.
(i) Direct comparison of SET and endovascular measures revealed similar functional outcomes for both interventions in the medium term across a number of studies with long-term data from one study identifying comparable limb salvage between groups. However, only EVR alone resulted in significant improvements in lower limb perfusion as measured by ABPI. These data have significant applicability to symptomatic patients whose comorbidities allow them to exercise to an adequate level for SET. Supervised exercise offers this group an acceptable, effective initial step in those capable and motivated to take part. in the dual therapy group. Similarly, all four trials assessing endovascular intervention in conjunction with SET compared to monotherapy found greater improvements in the walking distances of those undergoing dual therapy. These data would strongly support the use of SET as an adjunct to any operative intervention in the management of symptomatic PAD.
(iii) Only one study directly compared open surgery and SET. This strongly supported the surgical approach in terms of both function and perfusion outcomes.
Multiple exercise modalities including treadmill walking, resistance training, and upper limb ergometry have proven benefit in PAD [bib_ref] Physical activity and health, Powell [/bib_ref] [bib_ref] Exercise for intermittent claudication, Watson [/bib_ref] [bib_ref] Treadmill exercise and resistance training in patients with peripheral arterial disease with..., Mcdermott [/bib_ref] [bib_ref] Upper-vs lower-limb aerobic exercise rehabilitation in patients with symptomatic peripheral arterial disease:..., Zwierska [/bib_ref] ]. Unfortunately, the success of SET is reliant on excellent compliance and availability. Supervised exercise training has been shown to consistently result in improved functionality with superior outcomes to a "go home and walk" approach [bib_ref] Systematic review of exercise training or percutaneous transluminal angioplasty for intermittent claudication, Frans [/bib_ref] [bib_ref] A meta-analysis of the outcome of endovascular and noninvasive therapies in the..., Ahimastos [/bib_ref] [bib_ref] Supervised exercise therapy versus non-supervised exercise therapy for intermittent claudication, Bendermacher [/bib_ref]. Compliance can further be augmented by regular exercise reminders, patient specific exercise prescriptions, target setting, and the incorporation of exercise into daily activities [bib_ref] Interventions for promoting habitual exercise in people living with and beyond cancer, Bourke [/bib_ref]. Gains in walk time and claudication onset occur rapidly over the initial 2 months and can be maintained with good compliance [bib_ref] Optimal exercise program length for patients with claudication, Gardner [/bib_ref]. An exercise prescription of at least two 30-minute sessions per week offers optimal walking outcomes [bib_ref] Optimizing supervised exercise therapy for patients with intermittent claudication, Nicolaï [/bib_ref].
In those failing or unable to partake in an initial exercise program endovascular revascularization offers low rates of periprocedural morbidity and mortality with excellent initial success rates [bib_ref] Techniques for the endovascular management of complications following lower limb percutaneous transluminal..., Papavassiliou [/bib_ref] [bib_ref] The management of severe aortoiliac occlusive disease: endovascular therapy rivals open reconstruction, Kashyap [/bib_ref] [bib_ref] Balloon dilation and stent implantation for treatment of femoropopliteal arterial disease: meta-analysis, Muradin [/bib_ref]. However the two main shortcomings of endovascular intervention include long-term durability and overall poor survival rates associated with concomitant cardiovascular disease. Supervised exercise addresses some of these concerns. By stimulating collateralization and the release of vasodilators such as nitric oxide exercise therapy should confer superior functional results in the longer-term. This is supported by a number of studies [bib_ref] Proximal superficial femoral artery occlusion, collateral vessels, and walking performance in peripheral..., Mcdermott [/bib_ref] [bib_ref] Plasma nitrite flux predicts exercise performance in peripheral arterial disease after 3..., Allen [/bib_ref]. Thus, for this patient group SET offers not only functional improvement but also significant benefits to the cardiovascular health of these patients [bib_ref] Physical activity and health, Powell [/bib_ref].
# Strengths and limitations
. This review is strengthened by the fact that all incorporated data are from well-designed randomised controlled studies. Follow-up ranged from 6 months to 12 years. Cumulatively these trials incorporate a significant number of claudicants. Individually papers provide strong data that exercise results in significant gains in walk distance. In addition, a number of studies have found that a combination of SET and intervention provides a superior outcome to monotherapy.
This review is however subject to a number of limitations. Firstly, while all patients had significant arterial symptoms a degree of heterogeneity exists in this cohort. Each included study describes slight variations in the level of disease, the symptomatic presentation, and the intervention received rendering definitive conclusion difficult. Secondly, many studies included small numbers of participants with a number reporting moderate losses to follow-up. Finally, due to the heterogeneity of outcome data meta-analysis was not performed as part of this review.
# Conclusion
Exercise is an effective, safe, and economical method of treating symptomatic PAD [bib_ref] Cost-effectiveness of endovascular revascularization compared to supervised hospital-based exercise training in patients..., Spronk [/bib_ref]. Some randomised data now suggests that it may be comparable to EVR in the management of symptomatic noncritical ischaemia. Furthermore there is emerging evidence that combining endovascular intervention following by supervised exercise achieves better long-term results than endovascular intervention alone; however greater surgeon access to and awareness of this important adjuvant therapy is essential to maximize therapeutic outcomes.
[fig] Figure 1: Flow diagram depicting study identification. [/fig]
[fig] Figure 2: Management of intermittent claudication incorporating supervised exercise therapy. [/fig]
[table] Table 1: Treatment groups and disease level. [/table]
[table] Table 2: Assessment of bias. [/table]
[table] Table 3: Supervised exercise versus endovascular revascularization. ABPI: ankle brachial pressure index; MWD: maximal walk distance; ICD: intermittent claudication distance; EVR: endovascular revascularization; SET: supervised exercise therapy. Distances in metres unless otherwise stated. min represents time in minutes, claudication onset time/maximal walk time. value represents statistical comparison of interventions except in Hobbs et al.[21] where † represents change in measurement over study period. [/table]
[table] Table 4: Supervised exercise plus invasive measures (open surgery or EVR) versus monotherapy. [/table]
[table] Table 5: Invasive (EVR or surgery) management versus supervised exercise. [/table]
|
A comprehensive evaluation of module detection methods for gene expression data
A critical step in the analysis of large genome-wide gene expression datasets is the use of module detection methods to group genes into co-expression modules. Because of limitations of classical clustering methods, numerous alternative module detection methods have been proposed, which improve upon clustering by handling co-expression in only a subset of samples, modelling the regulatory network, and/or allowing overlap between modules. In this study we use known regulatory networks to do a comprehensive and robust evaluation of these different methods. Overall, decomposition methods outperform all other strategies, while we do not find a clear advantage of biclustering and network inference-based approaches on large gene expression datasets. Using our evaluation workflow, we also investigate several practical aspects of module detection, such as parameter estimation and the use of alternative similarity measures, and conclude with recommendations for the further development of these methods.
E ver since the introduction of genome-wide gene expression profiling technologies, module detection methods have been a cornerstone in the biological interpretation of large gene expression compendia [bib_ref] Cluster analysis and display of genome-wide expression patterns, Eisen [/bib_ref] [bib_ref] How does gene expression clustering work?, D'haeseleer [/bib_ref] [bib_ref] Democratizing systems immunology with modular transcriptional repertoire analyses, Chaussabel [/bib_ref]. Modules in this context are defined as groups of genes with similar expression profiles, which also tend to be functionally related and co-regulated. Apart from allowing a more global and objective interpretation of gene expression data [bib_ref] Transcriptomic analysis of autistic brain reveals convergent molecular pathology, Voineagu [/bib_ref] [bib_ref] Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease, Jostins [/bib_ref] , co-expression modules are also frequently used to infer regulatory relationships between transcription factors and putative target genes [bib_ref] Dynamic regulatory network controlling TH17 cell differentiation, Yosef [/bib_ref] [bib_ref] Identification of transcriptional regulators in the mouse immune system, Jojic [/bib_ref] [bib_ref] Transcriptional heterogeneity and lineage commitment in myeloid progenitors, Paul [/bib_ref]. In addition, modules can improve functional genome annotation through the guilt-by-association principle [bib_ref] Targeted exploration and analysis of large cross-platform human transcriptomic compendia, Zhu [/bib_ref] and allow a better understanding of disease origin [bib_ref] A narrow repertoire of transcriptional modules responsive to pyogenic bacteria is impaired..., Alsina [/bib_ref] and progression [bib_ref] A modular analysis framework for blood genomics studies: application to systemic lupus..., Chaussabel [/bib_ref].
Numerous approaches and algorithms have been proposed for module detection in gene expression data. The most popular approach, clustering, has been used since the first gene expression datasets became available and is still the most widely used to this day [bib_ref] Dynamic regulatory network controlling TH17 cell differentiation, Yosef [/bib_ref] [bib_ref] Identification of transcriptional regulators in the mouse immune system, Jojic [/bib_ref] [bib_ref] Transcriptional heterogeneity and lineage commitment in myeloid progenitors, Paul [/bib_ref] [bib_ref] A narrow repertoire of transcriptional modules responsive to pyogenic bacteria is impaired..., Alsina [/bib_ref]. However, in the context of gene expression, clustering methods suffer from three main drawbacks. First, clustering methods only look at co-expression among all samples. As transcriptional regulation is highly context specific [bib_ref] Circuitry and dynamics of human transcription factor regulatory networks, Neph [/bib_ref] , clustering potentially misses local co-expression effects which are present in only a subset of all biological samples. Second, most clustering methods are unable to assign genes to multiple modules. The issue of overlap between modules is especially problematic given the increasing evidence that gene regulation is highly combinatorial and that gene products can participate in multiple pathways [bib_ref] Architecture of the human regulatory network derived from ENCODE data, Gerstein [/bib_ref] [bib_ref] Crosstalk in NF-κB signaling pathways, Oeckinghaus [/bib_ref]. A third limitation of clustering methods is that they ignore the regulatory relationships between genes. As the variation in target gene expression can at least be partly explained by variation in transcription factor expression [bib_ref] Wisdom of crowds for robust gene network inference, Marbach [/bib_ref] , including this information could therefore boost module detection. Several alternative module detection approaches have therefore been developed in order to alleviate these three limitations. Decomposition methods [bib_ref] Integrating genome-wide genetic variations and monocyte expression data reveals trans-regulated gene modules..., Rotival [/bib_ref] and biclustering [bib_ref] A comparative analysis of biclustering algorithms for gene expression data, Eren [/bib_ref] try to handle local coexpression and overlap. These methods differ from clustering because they allow that genes within a module do not need to be co-expressed in all biological samples, but that a sample can influence the expression of a module to a certain degree (decomposition methods) or not at all (biclustering methods). Two other alternative methods, direct network inference [bib_ref] Wisdom of crowds for robust gene network inference, Marbach [/bib_ref] and iterative NI [bib_ref] Integrated module and gene-specific regulatory inference implicates upstream signaling networks, Roy [/bib_ref] , use the expression data to additionally model the regulatory relationships between the genes.
Given the importance of module detection within the transcriptomics field and the wealth of available methods, it is critical that existing and new approaches are evaluated on objective benchmarks. However, we identified several shortcomings in past evaluation studies [bib_ref] A comparative analysis of biclustering algorithms for gene expression data, Eren [/bib_ref] [bib_ref] A systematic comparison and evaluation of biclustering methods for gene expression data, Prelić [/bib_ref] [bib_ref] Biclustering methods: biological relevance and application in gene expression analysis, Oghabian [/bib_ref] [bib_ref] Evaluation and comparison of gene clustering methods in microarray analysis, Thalamuthu [/bib_ref] [bib_ref] Methods for evaluating clustering algorithms for gene expression data using a reference..., Datta [/bib_ref] , related to the use of multiple evaluation metrics, the correct tuning of parameters, and the biological relevance of synthetic data. In this study we therefore propose a new evaluation pipeline for module detection methods for gene expression data. Central to our approach is that we use known regulatory networks to define sets of known modules, which can be used to directly assess the sensitivity and specificity of the different module detection methods on real data. Using our evaluation strategy we analyze the performance of 42 module detection methods spanning all five main approaches. We also consider several practical aspects of module detection, such as the Clustering [bib_ref] Methods for evaluating clustering algorithms for gene expression data using a reference..., Datta [/bib_ref] Biclustering (9) Decomposition (5) Direct NI [bib_ref] Transcriptomic analysis of autistic brain reveals convergent molecular pathology, Voineagu [/bib_ref] Iterative NI [bib_ref] How does gene expression clustering work?, D'haeseleer [/bib_ref] Gene contributions Sample contributions RegulonDB ± ± Overview of our evaluation methodology. a The nine different datasets used in this evaluation. b We used three different module definitions to extract known modules from known regulatory networks for the evaluation on E. coli, yeast and synthetic data. c To avoid parameter overfitting on characteristics of particular datasets, we first optimized the parameters on every dataset using a grid search, and then used the optimal parameters on one dataset (training score) to assess the performance of a method on another dataset (test score). d We evaluated a total of 42 methods, which can be classified in 5 categories: clustering, biclustering, direct network inference (NI), decomposition, and iterative NI. e For the evaluation on human data, we compared how well the targets of each regulator is enriched in at least one of the modules. f We used four different regulatory networks in our evaluation, each generated from different types of data relative data requirements of the methods, parameter estimation, and the use of alternative similarity measures for clustering. The purpose of this evaluation study is twofold. We first want to provide an overview of the characteristics and performance of current module detection methods to guide the biologist in their choice. Second, we propose a benchmark strategy, which can be used in future studies to compare novel methods with the current state of the art. For this purpose, we provide all gold standards, expression datasets, and the evaluation procedure to the community.
# Results
Evaluation workflow. Our evaluation procedure was structured as follows . We applied publicly available module detection methods on nine gene expression compendia from Escherichia coli, Module detection methods evaluated in this study Clustering: grouping genes based on a global similarity in gene expression profiles A FLAME: fuzzy clustering by selecting cluster supporting objects based on the K-nearest neighbor density estimation B K-medoids: iteratively refines the centers (which are individual genes) and the average dissimilarity within the cluster C K-medoids (see B) but with automatic module number estimation D Fuzzy c-means: similar to k-means (see F), but using fuzzy instead of crisp cluster memberships E Self-organizing maps: maps each gene on a node embedded in a two-dimensional graph structure F K-means: iteratively refines the mean expression with a cluster and the within-cluster sum of squares G MCL: simulates random walks within the co-expression graph by alternative steps of expansion and inflation H Spectral clustering: applies K-means in the subspace defined by the eigenvectors of the Pearson's correlation affinity matrix I Affinity propagation: clustering by exchange of messages between genes J Spectral clustering: applies K-means in the subspace defined by the eigenvectors of the K-nearest-neighbor graph K Transitivity clustering: tries to find the transitive co-expression graph in which the total cost of added and removed edges is minimized L WGCNA: agglomerative hierarchical clustering (see M), but using the topological overlap measure and a dynamic tree cutting algorithm to implicitly determine the number of modules M Agglomerative hierarchical clustering: generates a hierarchical structure by progressively grouping genes and clusters based on their similarity N Hybrid hierarchical clustering: combination of agglomerative and divisive hierarchical clustering O Divisive hierarchical clustering: generates a hierarchical structure by progressively splitting the genes into clusters P Agglomerative hierarchical clustering (see M), but with automatic module number estimation Q SOTA: combination of self-organizing maps and divisive hierarchical clustering R First finds cluster centers by searching for high-density regions, each gene is then assigned to the cluster of its nearest neighbor of higher density S CLICK: uses density estimation to find tight groups of similar genes, after which these are expanded into modules T DBSCAN: groups genes within core, non-core and outlier genes based on the number of neighbors U Clues: first applies a shrinking procedure which moves each gene towards nearby high-density regions, after which the genes are partitioned into an automatically determined number of clusters using the silhouette width V Mean shift: moves each gene towards nearby high density regions until convergence Decomposition: extracting the components corresponding to co-expression modules by decomposing the expression matrix in a product of smaller matrices A Independent component analysis: decomposes the expression matrix into a set of independent components using the FastICA algorithm, detects potentially overlapping modules within each source signal using false-discovery rate (FDR) estimation B Similar to A, but detects two modules per independent component depending on whether genes have positive or negative weights C Similar to A, but detects modules within each source signal using z-scores D Combination of principal component analysis and independent component analysis, uses FDR estimation to find modules E Principal component analysis: decomposes the expression matrix into a set of linearly uncorrelated components, detects potentially overlapping modules within each component using FDR estimation
Biclustering: simultaneous grouping of genes and samples in biclusters based on similar local behavior in expression A Spectral biclustering: detecting checkerboard patterns within the gene expression matrix B ISA: iteratively refines a set of genes and samples based on high or low expression in both the gene and sample dimension C QUBIC: finds biclusters in which the genes have similar high or low expression levels in a discretized expression matrix D Bi-Force: finds biclusters with over-or under-expression by solving the bicluster editing problem E FABIA: builds a multiplicative model of the expression matrix layer by layer. Every layer represents a bicluster F Plaid: builds an additive model of the expression matrix layer by layer. Every layer represents a bicluster G MSBE: finds additive biclusters starting from randomly sampled reference genes and conditions H Cheng & church: minimizes the mean squared residue within every bicluster I OPSM: searches for biclusters where the expression changes in the same direction between genes and samples Iterative network inference: iterative optimization of an inferred network and a set of clusters A MERLIN: iteratively refines a direct regulatory network and modules within a probabilistic graphical network framework B Genomica: starts from an initial hierarchical clustering and iteratively refines this clustering and an inferred module network using a model based on Bayesian regression trees Direct network inference: inference of a regulatory network based on gene expression similarity between regulators and target genes A GENIE3: predicts the expression of each target gene based on random forest regression B CLR: calculates the likelihood of mutual information estimations based on the network neighborhood C Pearson's correlation between regulator and target gene D TIGRESS: network inference using a combination of Lasso sparse regression and stability selection yeast, human, and in silico simulated regulatory networks . We scored the different methods by comparing the observed modules with a set of known modules. These known modules were extracted from known regulatory networks using three different module definitions , two requiring co-regulation by either one or all known regulators and one looking at strong interconnectedness within the gene regulatory network. To compare a set of observed modules with known modules, we considered several scores described in the literature (Supplementary Note 1) and ultimately chose four scores as follows: recovery, relevance, recall, and precision . Note that classical scores comparing clusterings could not be used because these cannot handle overlap. As all methods generally performed equally or worse than random on human datasets, due to the high number of false positives in the gold standard (Supplementary Note 1), we instead used a scoring system which looks at how well the observed modules cover the targets of regulators in the dataset . To avoid certain gold standards and module definitions from disproportionately influencing our final score, we normalized each score using random permutations of the known modules. The final score for a method ultimately represented a fold improvement of a given module detection method over the score obtained from randomly permuted known modules.
Parameter tuning is a necessary but often overlooked challenge with module detection methods. Although good performance generally depends on the correct choice of parameters, this also increases the risk of overfitting on specific characteristics of one dataset, as such parameters will lead to suboptimal results when generalizing the parameter settings to other datasets. To address both problems, we optimized the parameters for every method with a grid search (Supplementary Note 2) and used an approach akin to cross-validation where the optimal parameter settings from one dataset were used to assess the performance of a method on another dataset . For every method we give two scores: the training score represents the score at the optimal parameter settings, whereas the test score denotes the performance when parameters were estimated on an alternative dataset .
Overall performance. We evaluated a total of 42 module detection algorithms covering all 5 approaches (clustering, decomposition, biclustering, direct NI, and iterative NI) using the described methodology and Supplementary Note 2). Overall, our results indicate that decomposition methods detect the modules which best correspond to the known modular structure within the gene regulatory network . The best decomposition methods are all variations of independent component analysis (ICA) with different post-processing methods [bib_ref] Integrating genome-wide genetic variations and monocyte expression data reveals trans-regulated gene modules..., Rotival [/bib_ref] [bib_ref] Elucidating the altered transcriptional programs in breast cancer using independent component analysis, Teschendorff [/bib_ref]. Surprisingly, neither biclustering nor direct NI, nor iterative NI methods outperform clustering methods, although in theory they should offer several advantages by allowing overlap, modelling transcriptional regulation and/or looking for local coexpression effects .
Note that decomposition methods not only perform well when the gold standard modules contains overlap, in the case of minimally co-regulated modules, but also when no overlap is present in the known modules . To further investigate this, we calculated separate scores for genes within one or multiple modules. This analysis showed that both clustering and decomposition methods are better at grouping genes that are present in one module, whereas biclustering and direct NI methods are slightly biased toward genes present in more than one module . These results indicate that the higher performance of decomposition methods over clustering is not exclusively caused by their ability to detect overlapping modules, but that also other factors such as local co-expression could have a role.
We further classified clustering algorithms into four categories: graph-based clustering, representative-based clustering, hierarchical clustering, and density-based clustering. We found that graph-based, representative-based, and hierarchical clustering all Overall performance of 42 module detection methods based on the agreement between observed modules and known modules in gene regulatory networks. The methods can be divided in five categories: clustering, decomposition, biclustering, direct network inference (direct NI) and iterative network inference (iterative NI) methods. Clustering and biclustering methods were further classified in subcategories (see Methods). a Average test and training scores across datasets and module definitions. The score represents a fold improvement over permutations of the known modules. *Automatic estimation of number of modules. b Different properties of the module detection methods (see Supplementary Note 2). A+ (green background) denotes that a method can handle a certain property listed on the left. We distinguish between explicit (−), implicit (±), and automatic (+) module number estimation. Note that running times strongly depend on the implementation, hardware, dataset dimensions, and parameter settings, and are therefore only indicative. c Test scores at each of the four datasets, averaged over module definitions. d Test scores on each of the three module definitions, averaged over different datasets performed equally well, with the clustering method FLAME (Fuzzy clustering by Local Approximation of Memberships) [bib_ref] FLAME, a novel fuzzy clustering method for the analysis of DNA microarray..., Fu [/bib_ref] , one of the only clustering methods able to detect overlap, slightly outperforming other clustering methods. Among hierarchical clustering methods, agglomerative methods provide the highest performance compared with the intermediate and low scores of respectively hybrid and divisive methods. Density-based clustering methods on the other hand had much lower performance, which can be partly explained by a higher parameter sensitivity for some density-based methods. Although the overall performance of biclustering methods was low, we also made a similar categorization of these methods based on the type of biclusters they detect. Methods that detect constant or extreme biclusters generally outperformed other methods detecting more complex bicluster patterns. In fact, except for FABIA (Factor Analysis for Bicluster Acquisition), the performance of the latter methods was generally not much better or even worse than random permutations of the known modules.
[formula] + + + + + ± − + − − − ± − ± − ± ± − − − + ± ± ± ± + ± − − − − − − ± ± ± − − ± ± ± ± ± ± ± ± ± [/formula]
Although the relative ranking of the main methods is remarkably stable across datasets and [fig_ref] Figure 3: Effect of automatic parameter estimation using four different cluster validity indices and... [/fig_ref] , individual methods can still perform well in one setting even though their overall performance is poor [fig_ref] Figure 4: Influence of the number of samples on the performance of the top... [/fig_ref]. Most profoundly is the higher performance of certain biclustering methods, such as ISA (Iterative Signature Algorithm), QUBIC (Qualitative Biclustering), and FABIA, and direct NI methods, primarily GENIE3, on human and/or synthetic data, where these methods can in some cases compete with clustering and decomposition methods. Performance was generally very consistent across different module definitions and [fig_ref] Figure 3: Effect of automatic parameter estimation using four different cluster validity indices and... [/fig_ref] , despite limited similarity between the sets of known modules [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref]. We also found that the overall ranking of the methods remained similar when we used different randomization procedures to normalize our scores . The relative performance of individual methods was more variable when we compared different scores, especially for scores which can handle overlapping modules, although the overall ranking of the different module detection approaches remained stable (Supplementary Figs. 7 and 8). Together, this again highlights the importance of using multiple datasets and scoring metrics for a robust and unbiased evaluation of bioinformatics methods [bib_ref] The self-assessment trap: can we all be better than average?, Norel [/bib_ref] [bib_ref] Comparing the performance of biomedical clustering methods, Wiwie [/bib_ref].
Parameter tuning. The need to tune parameters on individual datasets varied greatly among methods, which we quantified by comparing training and test scores. Some methods, such as FLAME, WGCNA (Weighted Gene Co-expression Network Analysis), and MERLIN (Modular regulatory network learning with per gene information) were relatively insensitive to parameter tuning, despite requiring the optimization of two or more parameters . On the other hand, methods such as fuzzy c-means, self-organizing maps, and agglomerative hierarchical clustering performed very differently between test and training parameters. Nonetheless, the overall ranking of the different module detection approaches does not change drastically between training and test scores. The top decomposition methods for instance outperform clustering methods both before and after controlling for parameter overfitting and [fig_ref] Figure 4: Influence of the number of samples on the performance of the top... [/fig_ref].
The most central parameters in module detection (and unsupervised data analysis in general) are those affecting the number of clusters detected within a dataset. We distinguish three different ways a method determines the number of modules . Explicit methods, such as k-means and all decomposition methods, require that the number of modules is specified by the user. Implicit methods, such as affinity propagation, adapt the module number on each dataset based on other parameters supplied by the user. Finally, automatic methods determine the number of modules completely automatically, usually by iterating over several parameter settings and selecting the one that optimizes some criterion of cluster quality. Measures for cluster quality can range from the stability of clusters among several resamplings of the dataset 27 , the balance between cluster tightness and separateness (as measured by cluster validity indices 28 ), or the optimal functional enrichment of the modules [bib_ref] Methods for evaluating clustering algorithms for gene expression data using a reference..., Datta [/bib_ref]. Although the top method within each clustering subcategory estimate the number of modules implicitly (coinciding with a relatively low parameter sensitivity) , we found that implicit or automatic estimation of the number of modules is not a prerequisite for a high performance . Indeed, the top decomposition methods all require the number of modules to be specified beforehand. Interestingly, the performance of iterative NI methods depended only little on the initial parameters, possibly because these methods adapt the number of modules depending on the inferred regulatory network.
Apart from those parameters influencing the number of clusters, most module detection methods also have other parameters, frequently affecting the compactness of the modules, the way local co-expression is defined or the minimal number of genes within a module. As all these parameters can have significant effects on the resulting modules (and thus the performance of a method), we assessed how well automatic parameter estimation can improve method performance. Automatic parameter estimation can be seen as an alternative to determining the optimal parameters on one or more training datasets and using these parameters on a test dataset to assess performance. Based on a previous evaluation studywe chose four of the most promising cluster validity indices, and found that the benefits of cluster validity indices were mostly confined to clustering methods [fig_ref] Figure 3: Effect of automatic parameter estimation using four different cluster validity indices and... [/fig_ref]. Notably, spectral clustering, affinity propagation, and k-medoids frequently increased in test score when their parameters were automatically estimated using the average silhouette width and the Davis-Bouldin index (Supplementary . On the other hand, the performance of FLAME clustering and decomposition methods generally decreased when using cluster validity indices, usually performing even worse than randomly selecting parameters within the parameter grid . It is important to note here that all four cluster validity indices have been developed in the context of clustering methods and were therefore never really designed with overlap and local co-expression in mind, which could explain their low performance with these methods. We also analyzed two alternative measures, which try to estimate the number of clusters based on the optimal enrichment of functional terms or pathways within the modules. We found that a measure that assesses both the coverage of all functional terms as well as the strong individual enrichment of every module (F-aucodds) performed very well, in a large majority of cases performing better than using the optimal parameters of another dataset and random parameter settings . Another important parameter for most clustering methods is the distance or similarity measure for comparing gene expression profiles. The most popular metric for gene expression data is undoubtedly the Pearson's correlation coefficient, which measures the extent of the linear dependence between two expression profiles regardless of differences in absolute expression levels. Several authors have criticized this measure [bib_ref] Evaluation of gene-expression clustering via mutual information distance measure, Priness [/bib_ref] [bib_ref] Gene network interconnectedness and the generalized topological overlap measure, Yip [/bib_ref] [bib_ref] Comparison of co-expression measures: mutual information, correlation, and model based indices, Song [/bib_ref] , mainly due to three limitations: (i) it ignores inverse relations between genes , (ii) it is unable to capture non-linear relationships , and (iii) it is not robust to outliers and skewed distributions . Several alternative measures have therefore been proposed, which try to tackle some of these limitations (Supplementary Note 3).
[formula] 0 2 4 6 B A C A D H L I G B D J E K M B N A A F A O D C Q B S B R C T E D V F G I E H Davis-Bouldin index [/formula]
To investigate whether these alternatives are able to improve the module detection, we used 15 such measures as the input for four of the top clustering methods which require having a similarity or distance measure as parameter. Surprisingly, none of the alternative similarity metrics are able to improve performance of any of the four top clustering methods . When investigating this further, we found several cases where these alternative measures can indeed retrieve known coregulated genes which were ranked lower than Pearson's correlation, as illustrated with three case examples (Supplementary . However, when comparing the top 10% gene pairs between Pearson's correlation and alternative measures, more known co-regulated gene pairs are removed than there are gained .
Sensitivity to number of samples and noise. We next tested the influence of the number of samples within an expression dataset on the relative performance of the top module detection methods within every category. Although, as expected, the performance of every method declined with decreasing dataset size, the magnitude and timing of this decline varied strongly per method. Notably, ICA-based decomposition methods (decomposition methods A and B) seem to be much more sensitive to the number of samples in the dataset compared with other methods [fig_ref] Figure 4: Influence of the number of samples on the performance of the top... [/fig_ref] and [fig_ref] Figure 3: Effect of automatic parameter estimation using four different cluster validity indices and... [/fig_ref]. On the other hand, the performance of several network inference based methods, such as Genomica (iterative NI method A) and GENIE3 (direct NI method A), remained relatively stable with decreasing number of samples. Together, this indicates that despite its better performance on large datasets, current matrix decomposition methods are unable to meet the performance of clustering methods when a smaller number of biological conditions are being considered.
We also analyzed the noise sensitivity of the different methods by applying different levels of noise on the synthetic datasets. Although we saw that most methods were similarly sensitive to noise compared with their overall performance, we found that some methods, notably WGCNA and fuzzy c-means, were more sensitive [fig_ref] Figure 4: Influence of the number of samples on the performance of the top... [/fig_ref].
# Discussion
Unsupervised data analysis has the potential to provide an unbiased and global overview of biological datasets. Compared with other unsupervised clustering tasks in biology (extensively evaluated elsewhere 26 ), module detection in gene expression data is unique, because the complexity of the underlying gene regulatory network poses particular challenges, such as local coexpression and overlap. These challenges led to the development of numerous algorithms and tools specifically dedicated to gene expression data; however, so far the comparative performance of these methods was unclear. In this work we therefore introduced a general framework for evaluating module detection methods and used it to provide a first comprehensive evaluation of stateof-the-art module detection methods for gene expression data. Based on this evaluation we analyzed several practical aspects of module detection, such as the choice of methods and parameter estimation, which are summarized in [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref] and will be further discussed here. Moreover, we also provide several guidelines for further development of these methods combined with what in our view has already been accomplished, as summarized in Supplementary [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref].
Module detection in gene expression data can serve a variety of roles and different methods are better suited for particular roles [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref]. Owing to the ease of visualization and interpretation, non-overlapping clustering methods can quickly generate a global overview of the dataset, revealing the main expression and functional effects among the different biological samples in the dataset 2 . Our analysis showed that FLAME, WGCNA, Affinity Propagation, Markov clustering (MCL), and Spectral clustering are particularly suited for such an analysis, outperforming other clustering methods on most datasets. However, because clustering methods do not detect local co-expression effects, they could potentially miss relevant modules or exclude important genes from a module. In use cases where it can be desirable that all modules are discovered in a dataset, e.g., to generate signatures for disease, therapy and prevention [bib_ref] Transcriptomic analysis of autistic brain reveals convergent molecular pathology, Voineagu [/bib_ref] [bib_ref] A modular analysis framework for blood genomics studies: application to systemic lupus..., Chaussabel [/bib_ref] [bib_ref] Coherent functional modules improve transcription factor target identification, cooperativity prediction, and disease..., Karczewski [/bib_ref] , or to find a set of genes responsible for a biological function, methods that detect such local co-expression and/or overlapping modules could therefore provide a substantial advantage. Consistent with this, we found that decomposition methods based on ICA were better at recovering known modules consistently across datasets. Although a handful of studies have already shown the potential of these methods in true biological settings [bib_ref] Integrating genome-wide genetic variations and monocyte expression data reveals trans-regulated gene modules..., Rotival [/bib_ref] [bib_ref] Elucidating the altered transcriptional programs in breast cancer using independent component analysis, Teschendorff [/bib_ref] [bib_ref] Coherent functional modules improve transcription factor target identification, cooperativity prediction, and disease..., Karczewski [/bib_ref] , this was never shown in direct comparison to alternative methods and/or based on objective benchmarks. Finally, a third major application of coexpression modules is in the inference of gene regulatory networks, where modules can be used to improve the network by combining information from several genes 33 but can also improve the ease of interpretation. When we assessed the accuracy of the inferred network by combining a state-of-the-art network inference methods with different module detection methods, we found that ICA-based decomposition methods lead to the highest improvement in accuracy (primarily on yeast and synthetic datasets), closely followed by clustering and graph clustering methods . For most methods, free implementations are available either with a graphical or programming interface, of which we give an overview in Supplementary .
The choice of method influences subsequent steps of parameter estimation, visualization, and functional interpretation [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref]. For parameter estimation we found that cluster validity indices, the Davis-Bouldin and Kim-Ramakrishna indices in particular, are sufficient to estimate the parameters for most top clustering methods. However, the performance of these measures on alternative module detection methods was generally worse than randomly selecting parameters. For these methods, biclustering, decomposition, and direct NI in particular, we found that a measure based on functional enrichment provides a better alternative [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref]. The kind of visualization of the modules also heavily depends on the method. Although the results of a nonoverlapping clustering analysis can be readily visualized using heat maps or networks 34 , visualizing overlapping modules requires more complex and hierarchical visualizations (which, e.g., indicate the overlap between modules) and is still an active research field [bib_ref] Furby: fuzzy force-directed bicluster visualization, Streit [/bib_ref] [bib_ref] BicOverlapper 2.0: visual analysis for gene expression, Santamaría [/bib_ref] [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref]. In both cases, additional annotations can be added to the visualization to improve interpretation of the modules, e.g., to indicate the functional annotation of the genes, and interactivity can be used to accelerate the exploration of the modules. Finally, several tools and databases can be used to functionally interpret the modules, to analyze what biological functions and pathways are enriched within the modules or to find whether the module could be associated with certain diseases. To reduce redundancy in the results of such enrichment analysis, alternative visualization and trimming methods can be used to extract the main biological functions, pathways, or diseases associated with particular modules [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref]. We give an overview of freely available methods that can be used to interpret co-expression modules in .
Apart from those listed in , there are also several other characteristics of module detection methods, which are important to consider in practice. Non-exhaustive module detection methods, which include some clustering methods such as FLAME and WGCNA, do not necessarily assign every gene to at least one module. Although this has the advantage that the method itself detects noisy expression profiles, users should be aware that it can also remove a lot of relevant expression profiles if the parameter values are too stringent. Network inference-based methods are unique among the different approaches, because they also generate hypotheses that can explain molecularly why certain genes are grouped in a module. Despite their lower performance according to our evaluation, they could therefore still be advantageous in certain use cases. Finally, we note that some methods are stochastic, and to assure the robustness of the results users should consider re-running the methods several times in different random states. We list these different properties of a method in Supplementary Note 2, along with a brief discussion about their algorithmic approach, important parameters, and directions to freely available implementations. As most of the evaluation studies in the past focused only on a limited number of methods, a direct comparison with our evaluation study is difficult [bib_ref] A comparative analysis of biclustering algorithms for gene expression data, Eren [/bib_ref] [bib_ref] A systematic comparison and evaluation of biclustering methods for gene expression data, Prelić [/bib_ref] [bib_ref] Biclustering methods: biological relevance and application in gene expression analysis, Oghabian [/bib_ref] [bib_ref] Evaluation and comparison of gene clustering methods in microarray analysis, Thalamuthu [/bib_ref] [bib_ref] Methods for evaluating clustering algorithms for gene expression data using a reference..., Datta [/bib_ref]. Furthermore, in these evaluations major conclusions frequently rely on synthetic datasets, and although it can certainly give insights into assumptions made by certain algorithms, it cannot be used to make conclusions about the usefulness of a particular algorithm on real datasets. Still, we acknowledge two noteworthy differences as follows: most studies find that biclustering methods outperform clustering [bib_ref] A systematic comparison and evaluation of biclustering methods for gene expression data, Prelić [/bib_ref] [bib_ref] Biclustering methods: biological relevance and application in gene expression analysis, Oghabian [/bib_ref] or observe substantial performance differences between graphbased, hierarchical, and representative-based clustering methods [bib_ref] Evaluation and comparison of gene clustering methods in microarray analysis, Thalamuthu [/bib_ref] [bib_ref] Methods for evaluating clustering algorithms for gene expression data using a reference..., Datta [/bib_ref]. We relate these differences mainly to issues with parameter estimation, reliance on synthetic datasets and use of a limited number of evaluation metrics. In Supplementary Note 4 we give an overview of past evaluation studies, the methods they evaluated, and the evaluation aspects where we believe these studies are lacking. Similar to a recent study evaluating clustering methods on several biological datasets [bib_ref] Comparing the performance of biomedical clustering methods, Wiwie [/bib_ref] , we found that no single clustering is the best performer on all datasets, although certain methods are certainly better than others at retrieving the known modular structure within the data.
Nonetheless, we acknowledge that our evaluation workflow still has some limitations for particular applications. As we wanted to make sure that most of the modules present in our gold standard were also differentially expressed in the expression data, we used large expression compendia from very different biological conditions. However, this means that when expression differences are very subtle, other methods such as biclustering could perform better. Indeed, some biclustering methods such as FABIA are frequently used in drug discovery [bib_ref] Using transcriptomics to guide lead optimization in drug discovery projects: lessons learned..., Verbist [/bib_ref]. An evaluation focusing on these kind of subproblems is still a possibility for future research.
The detection of overlapping and locally co-expressed modules has been a longstanding challenge in transcriptomics research. Despite great efforts towards the development of these methods, their application on real biological data has been hampered because of several practical challenges. Foremost, the visualization and interpretation of overlapping and locally co-expressed modules is more difficult. Despite some efforts [bib_ref] Furby: fuzzy force-directed bicluster visualization, Streit [/bib_ref] [bib_ref] BicOverlapper 2.0: visual analysis for gene expression, Santamaría [/bib_ref] , current visualization tools, e.g., do not directly show why certain genes are grouped in a module, which can make the module detection methods seem like a black box with unclear biological relevance. Moreover, decomposition and biclustering methods usually have several parameters, which need to be tuned on a dataset and which can affect the biological interpretation. Although we found that external functional information can be used to estimate the parameters of these methods on most datasets, the requirement for such external information can limit their applicability on wellstudied organisms. Parameter estimation of biclustering and decomposition methods, which uses only the expression matrix itself, therefore remains an open issue. Finally, our evaluation also indicates that the top performing decomposition methods are much more sensitive to the number of samples in a dataset and are outperformed by clustering methods when the number of samples is limited. We anticipate that improvements on these points (visualization, parameter estimation, and data requirements), will allow these advanced module detection methods to gain more traction in biological research. We list some past accomplishments and points for future research in [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref].
# Methods
Regulatory networks and module definitions. For E. coli datasets, we used a regulatory network from the RegulonDB database version 8 (regulondb.ccg.unam. mx, accessed 03/06/2015), a database integrating both small-scale experimental evidence as well as genome-wide data of transcriptional regulation [bib_ref] 0: omics data sets, evolutionary conservation, regulatory phrases, cross-validated gold standards and..., Salgado [/bib_ref]. We only included interactions with at least one strong evidence type (APPH, BPP, FP, IDA, SM, TA, CHIP-SV, GEA, ROMA, and gSELEX). We did not group the regulatory interactions at operon level, as we found that this has only minimal impact on the overall ranking of the different methods . We also did not include sigma factor regulations as we found that this would have a negligible effect on performance . For the yeast datasets we used two regulatory networks. One network was generated from an integration of chromatin immunopurification-on-chip data and conserved binding motifs as described by MacIsaac et al. [bib_ref] An improved map of conserved regulatory sites for Saccharomyces cerevisiae, Macisaac [/bib_ref]. Another regulatory network was generated by combining genomewide transcription factor binding data, knockout expression data, and sequence conservation [bib_ref] De-novo learning of genome-scale regulatory networks in S. cerevisiae, Ma [/bib_ref]. We used the most stringent dataset, which required evolutionary conservation in at least two species. For the human datasets we used the 'regulatory circuits' generated by in which regulators were linked with target genes through a series of steps starting from binding motifs in active enhancers using FANTOM5 project data.
For every gold standard, we obtained sets of known modules based on three different module definitions. We defined minimally co-regulated modules as overlapping groups of genes that shared at least one regulator. Strictly co-regulated modules were defined as groups of genes known to be regulated by exactly the same set of regulators. Strongly interconnected known modules, on the other hand, were defined as groups of genes that are strongly interconnected, and this does therefore not necessarily reflect co-regulation. We used three different graph cluster algorithms (markov clustering, transitivity clustering, and affinity propagation) with in every case three different parameter settings representing different levels of cluster compactness. For the Markov Clustering Algorithm 42 we used inflation parameters 2, 10, and 50. For transitivity clustering [bib_ref] Partitioning biological data with transitivity clustering, Wittkop [/bib_ref] we used two different cutoff parameters for the fuzzy membership 0.1 and 0.9. These two parameter settings allowed the modules to overlap . In the third parameter setting for transitivity clustering, we assigned every gene to the module with the highest fuzzy membership value. For affinity propagation [bib_ref] Clustering by passing messages between data points, Frey [/bib_ref] we varied the preference value between 0.5, 2, and an automatically estimated value (see Supplementary Note 2). All known modules were then filtered for the genes present in the expression matrix. Finally, we filtered strongly overlapping known modules by merging two modules if they overlapped strongly (Jaccard coefficient > 0.8) and removed small modules by requiring at least five genes. The latter cutoff was defined based on where the average optimal performance of all methods reached a maximum.
To further validate the known modules, we assessed the extent to which the modules are co-expressed in our expression datasets. We found that all three main module definitions generate modules which are both more globally and more locally (according to extreme expression biclustering definition, see Supplementary Note 2) co-expressed compared with permuted modules . Certain module definitions, strict coregulation in particular, and datasets, E. coli, and synthetic data generate modules that are better co-expressed within the expression data, which could explain why module detection methods generally also perform better on these datasets and module definitions . We further confirmed the biological relevance of the known modules by investigating their functional enrichment. We found that on the E. coli datasets, 50-70% of all functional terms (both for Gene Ontology (GO) [bib_ref] Gene Ontology Consortium: going forward, Ontology Consortium [/bib_ref] and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways [bib_ref] KEGG as a reference resource for gene and protein annotation, Kanehisa [/bib_ref] were enriched in at least one known module, and that 60-80% of all known modules were enriched in at least one functional term . The coverage of the whole functional space was much less on the yeast data, with about 5-15% GO terms and 10-30% KEGG pathways covered . On the other hand, a substantial number of all known modules were enriched in at least one functional term, ranging from 30% to 60% on GO terms and 30% to 50% on KEGG pathways . Compared with known modules, observed modules covered the functional space in most cases a little bit better for the top methods .
Gene expression data. We used a total of nine expression datasets for the study, two from E. coli, two from Saccharomyces cerevisae, three human datasets, and two synthetic datasets. Datasets consisted of hundreds of samples in various genomic and/or environmental perturbational settings.
We obtained a first E. coli dataset from the Colombos database (version 2.0, colombos.net) [bib_ref] 0: an ever expanding collection of bacterial expression compendia, Meysman [/bib_ref]. This dataset is unique among the four because it does not contain raw expression values from one sample but instead contains log ratios between test and reference conditions, which allowed the authors to integrate across different microarray platforms and RNA-sequencing experiments. A second E. coli dataset was downloaded from the DREAM5 network inference challenge 15 website (synapse.org/#!Synapse:syn2787209/wiki/70349).
For S. cerevisiae, we aggregated an expression compendium by integrating data from 127 experiments (filtered on S. cerevisae samples) using the GPL2529 platform from Gene Expression Omnibus (ncbi.nlm.nih.gov/geo). Raw expression data were normalized using Robust Multichip Average as implemented in the Bioconductor affy package. A second yeast dataset was obtained from the DREAM5 website (synapse.org/#!Synapse:syn2787209/wiki/70349).
We obtained the human TCGA datasets from a pan-cancer study of 12 cancer types (synapse.org/#!Synapse:syn1715755) [bib_ref] Multiplatform analysis of 12 cancer types reveals molecular classification within and across..., Hoadley [/bib_ref]. The human GTEX dataset, which contains expression profiles from different organs from hundreds of donors [bib_ref] Human genomics. The Genotype-Tissue Expression (GTEx) pilot analysis: multitissue gene regulation in..., Consortium [/bib_ref] , was downloaded from the GTEX website (gtexportal.org). The SEEK GPL5175 dataset is an aggregation of public datasets using the GPL5175 microarray platform and was retrieved from seek.princeton.edu.
We generated two synthetic datasets starting from the E. coli regulondb network and yeast MacIsaac network (both described above) using GeneNetWeaver. This network simulator models the gene regulation using a detailed thermodynamic model and simulates this model using ordinary differential equations [bib_ref] GeneNetWeaver: in silico benchmark generation and performance profiling of network inference methods, Schaffter [/bib_ref]. Different experimental conditions were simulated using the 'Multifactorial Perturbations' setting, where transcription rates for a subset of genes are randomly perturbed.
For all expression datasets we filtered out the least variable genes by requiring a minimal standard deviation in expression of 0.5 (for yeast and E. coli) and 1 (for human datasets). Heatmaps for every dataset .
Each dataset has its own advantages and disadvantages. Real datasets better fit the real use case and are thus the most biologically relevant, although limited availability of gold standard can make an evaluation on real data challenging. Although our knowledge of the regulatory networks of model micro-organisms, primarily E. coli, is already substantial, it is still far from complete 51 . While evaluating on data with more complex regulatory networks such as humans is certainly necessary to ensure the broad relevance of the evaluation, the definition of gold standards on these datasets can be even more problematic because of the broad prevalence of false-positive and false-negative interactions due to a variety of reasons, such as cellular context 12 and non-functional binding [bib_ref] Transcription factors: from enhancer binding to developmental control, Spitz [/bib_ref]. We therefore also included synthetic datasets where the known regulatory network is completely given and thus estimates of both sensitivity and precision of a method can be accurately estimated. Together, we believe these datasets give complementary support to our evaluation strategy and assure its broad relevance.
Similar to a previous evaluation study of biclustering methods 53 , our datasets can contain both large differences between samples, as well as small differences, as indicated by the distribution of all log-fold changes between samples .
Module detection methods. We chose a total of 42 module detection methods based on (i) a freely available implementation, (ii) performance within previous evaluation studies [bib_ref] A comparative analysis of biclustering algorithms for gene expression data, Eren [/bib_ref] [bib_ref] A systematic comparison and evaluation of biclustering methods for gene expression data, Prelić [/bib_ref] [bib_ref] Biclustering methods: biological relevance and application in gene expression analysis, Oghabian [/bib_ref] [bib_ref] Evaluation and comparison of gene clustering methods in microarray analysis, Thalamuthu [/bib_ref] , and (iii) novelty of the algorithm. See Supplementary Note 2 for a brief overview of every method and Supplementary for an overview of the implementations used in this study and alternative implementations. We classified all module detection methods in five major categories. We acknowledge however that the boundaries between the different categories are not always clear, as certain clustering and biclustering methods, e.g., also use a matrix decomposition step within their algorithm. The common theme of clustering methods is that they group genes according to a global similarity in gene expression. Even if clustering methods can detect (after some post-processing) overlapping clusters, this overlap is detected only because a certain gene is still globally similar to both two clusters, and not necessarily because of a local co-expression. Decomposition methods try to approximate the expression matrix using a product of smaller matrices. Two of these matrices contain the individual contributions of respectively genes and samples to a particular module. As samples are allowed to contribute to a particular module only to a certain degree, decomposition methods can detect local co-expression. Related to these methods are biclustering methods, which detect groups of genes, and samples, which show some local co-expression only within the bicluster. In biclustering, samples either contribute to a particular module or not, in contrast to decomposition methods where all samples contribute to a certain extent. Modules detected by biclustering methods can therefore be easier to interpret compared with those of decomposition methods, as the exact origin of the local co-expression is better defined. In some cases, a biclustering method is simply an extension of an existing decomposition method but with an extra requirement that the contribution of a gene and sample to a module is sparse (i.e., contains lots of zeros). Direct NI methods try to generate a simple model of gene regulation, in most cases by using the expression matrix to assign a score to every regulator-gene pair [bib_ref] Wisdom of crowds for robust gene network inference, Marbach [/bib_ref]. Although their primary application is to predict novel regulatory relationships between genes, some studies have also used the resulting weighed regulatory network to detect gene modules [bib_ref] Arabidopsis ensemble reverse-engineered gene regulatory network discloses interconnected transcription factors in oxidative..., Vermeirssen [/bib_ref] [bib_ref] Clustering contextspecific gene regulatory networks, Ramesh [/bib_ref]. A list of regulators was generated for E. coli by looking for genes annotated by GO with either "transcription, DNA-templated," or "DNA binding," and for yeast and human with "sequence-specific DNA-binding RNA polymerase II transcription factor activity." The same list was also used for iterative NI methods, which start from an initial clustering, and iteratively refine this clustering and an inferred regulatory network.
We further classified clustering methods according to their "induction principle," a classification that does not use the way clusters are represented in the algorithm (the model), but rather looks at the optimization problem underlying the clustering algorithm [bib_ref] Why so many clustering algorithms: a position paper, Estivill-Castro [/bib_ref]. Graph-based clustering algorithms make use of graph-like structures, such as K-nearest-neighbor graphs and affinity graphs, and group genes if they are strongly connected within this graph-like structure. Representationbased methods iteratively refine a cluster assignment and representative (such as the centroid) of the cluster. Hierarchical clustering methods construct a hierarchy of all the genes within the expression matrix. Finally, density-based methods detect modules by looking at contiguous regions of high density. It should also be noted that some clustering methods use elements from multiple categories. FLAME (clustering method A), e.g., uses elements from graph-based, representative-based, and density-based clustering, whereas affinity propagation contains both elements from graph-based and representative-based methods. In cases like this, we ultimately classified an algorithm based on which aspect of the algorithm we believe has the major impact on the final clustering result. Biclustering methods were further classified according to the type of biclusters they detect. The expression within constant biclusters remains relatively stable, whereas the genes within extreme biclusters have a relatively high expression in a subset of conditions compared with other genes. The expression within pattern-based biclusters follow more complex models such as additive models 57 , multiplicative models 53 , and coherent evolution [bib_ref] Discovering local structure in gene expression data: the order-preserving submatrix problem, Ben-Dor [/bib_ref].
Post-processing steps were required for certain methods to get the results in a correct format for comparison with the known modules. All parameters for these post-processing steps were optimized within the grid search approach (as described in Supplementary Note 2 for every method). For fuzzy clustering methods, we obtained crisp but potentially overlapping modules by placing a cutoff on the membership values. For direct NI methods, we first used a cutoff to convert the weighed to an unweighted network, and then detected modules using the same module definitions as previously described. For decomposition methods we explored several post-processing steps in literature (see Supplementary Note 2).
As gene regulatory networks, even in these model organisms, are still very incomplete 51 , a small majority of the genes was not included in any known module [fig_ref] Figure 3: Effect of automatic parameter estimation using four different cluster validity indices and... [/fig_ref]. Although we did retain these genes in the expression matrix pbefore module detection, we removed these genes in the observed modules before scoring. This was to avoid a strong overestimation of the number of false positives in the observed modules, as most of these genes probably belong to one or more co-regulated modules, which we do not yet know. Finally, similar to the known modules, we filtered the observed modules so that each module contained at least five genes.
Similar to our analysis with known modules, we assessed the extent to which the genes detected by each of the methods are co-expressed in the datasets based on three co-expression metrics inspired by the three types of biclustering metrics [fig_ref] Figure 4: Influence of the number of samples on the performance of the top... [/fig_ref]. (1) An overall co-expression metric using the average correlation, (2) an extreme expression metric by looking at the top 5% average z-scores for every gene in the module, and (3) the root mean-squared deviation within the expression values of each module. For each metric, we compared the distribution of the real modules with permuted modules by calculating the median difference using the wilcox.test function in R. We found that every module detection method found modules, which were more strongly co-expressed than permuted modules. Compared with the co-expression of known modules, the module detection methods also produced modules that are more strongly co-expressed. Specifically for biclustering methods, we also investigated the co-expression only in those samples within each bicluster. Here we found that, except for some pattern-based biclustering methods, most biclustering methods detected the type of modules, which they are designed to detect [fig_ref] Figure 4: Influence of the number of samples on the performance of the top... [/fig_ref].
Parameter tuning. Parameter tuning is a necessary but often overlooked challenge with module detection methods. All too often, evaluation studies use default parameters which were optimized for some specific test cases by the authors. This does not correspond well with the true biological setting, where some parameter optimization is almost always necessary to make sense of the data. Therefore, to make sure an evaluation is as unbiased as possible, some parameter optimization is always required. However, one should be careful of overfitting parameters on specific characteristics of one dataset, as such parameters will lead to suboptimal results when generalizing the parameter settings to other datasets. This could again introduce a bias in the analysis, where methods with a lot of parameters would better adapt on particular datasets, but would not generalize well on other datasets. In this study we tried to address both problems as follows. We first used a grid search to explore the parameter space of every method and determine their most optimal parameters given a certain dataset and module definition, which resulted in a training scores. Next, in a process akin to cross-validation, we used the optimal parameters of one training dataset from another organism to score the performance on another test dataset, which resulted in a test scores for every training dataset. As we saw that optimal parameters were in most cases very different between synthetic and real datasets, we only used real datasets to train parameters for other real datasets and synthetic datasets for other synthetic datasets. We refer to Supplementary Note 2 for an overview of what parameters were varied for every method.
Evaluation metrics. We used four different scores to compare a set of known modules with a set of observed modules and, after normalization, combined them in one overall score. Note that classical scores comparing clusterings, such as the Rand index, the F1, or the normalized mutual information, could not be used as these scores are unable to handle overlap and/or overlap [bib_ref] A comparison of extrinsic clustering evaluation metrics based on formal constraints, Amigó [/bib_ref].
The recall and specificity within the recently proposed CICE-BCubed scoring system measure whether the number of modules containing a certain pair of genes is comparable between the known and observed modules. It is based on the Extended BCubed 59 , but reaches the perfect score of 1 only when both known and observed overlapping clusterings are equal. If G represents all genes, M a set of known modules, M' a set of observed modules, M(g) the modules that contain g, and E(g, M) the set of genes that are together with g in at least one module of M (including g itself), the precision is defined as: Recall is calculated in the same way but with M' and M switched. The recovery and relevance scores, which have been previously used in evaluation studies of biclustering methods, assess whether each observed module can be matched with a known module and vice versa [bib_ref] A comparative analysis of biclustering algorithms for gene expression data, Eren [/bib_ref] [bib_ref] A systematic comparison and evaluation of biclustering methods for gene expression data, Prelić [/bib_ref]. Relevance is defined as
[formula] Precision ¼ 1 G j j X g2G 1 E g; M′ ð Þ j j X g′2E g;M′ðRelevance ¼ 1 M′ j j X m′2M′ max m2M Jaccard m′; m ð Þ [/formula]
Recovery is calculated in the same way but with M' and M switched. Before combining scores across different datasets and module definitions, we first normalized every score by dividing it by an average score of 500 permuted versions of the known modules [fig_ref] Figure 5: Practical guidelines for module detection in gene expression data [/fig_ref]. The goal of this step was to prevent easier module structures (small modules, low number of modules, and no overlap) of certain module definitions and datasets from disproportionally influencing the final score. Permuted modules were generated by randomly mapping the genes of a dataset to a random permutation of the genesand replacing every occurrence of a gene in a known module with its mapped version. Based on this random model, module structure (size, number, and overlap) remained the same, while only the assignment changed. We also tested two alternative random models. The STICKY random model has been previously described [bib_ref] Modelling protein-protein interaction networks via a stickiness index, Pržulj [/bib_ref]. We adapted this model for directed networks by calculating the stickiness index separately for incoming and outgoing edges. For the scale-free network, we used the networkx Python package with default parameters.
We finally calculated the harmonic mean between the normalized versions of all four scores to obtain a final score representing the performance of a particular method on a given dataset and module definition.
For human data we used an alternative score that assesses the extent to which the targets of every regulator are enriched with at least one module of the dataset. As described earlier, we used the clustered version of the regulatory circuits dataset 41 , which contains weights for every regulator and target gene combination across 32 tissue and cell-type contexts. For every combination of target genes and observed module we calculated a p-value of enrichment using a right-tailed Fisher's exact test (corrected for multiple testing using the Holm-Šídák procedureand the strength of this enrichment using the odds ratio. Although we calculated these values within every cell type and tissue context separately, we retained for every regulator its minimal p-value and the corresponding odds ratio across the different contexts, as we do not know the exact cell-type and tissue context in which the genes of the observed modules are co-expressed. We then extracted for every regulator its maximal odds ratio across the observed modules where the targets of the regulators were enriched (corrected p-value < 0.1). The aucodds score was then calculated by measuring the area under the curve formed by the percentage of regulators with an odds ratio equal or larger than a particular cutoff and the log 10 odds ratios within the interval 1 and 1000-fold enrichment. To work in a cutoffindependent manner we averaged the aucodds scores over a range of weight cutoffs. Performance generally decreased with more stringent cutoffs although some biclustering methods and direct NI methods remained more stable across a wide range of cutoff values . This score was normalized in a similar way as previously described, where the initial known modules were defined using the minimal co-regulation module definition and subsequently randomly permuted by mapping the genes within the modules to a random permutation.
We reweighted the scores between datasets and module definitions using a weighted mean so that module definitions (minimal co-regulation, strict coregulation, and interconnected subgraphs) and each organism (E. coli, yeast, human, and synthetic) had equal influence on the final score.
Influence of overlap. We split the genes of every datasets based on whether they belonged to only one or multiple modules using the minimal co-regulation module definition. If G* denotes such a subset of genes in the expression matrix, we calculated a precision* score specifically for this subset using: A Recall* score was calculated similarly but with M' and M switched. A final score for a particular set of genes was obtained by taking the harmonic mean between the normalized versions of the Recall* and Precision*.
[formula] Precision à ¼ 1 G à j j X [/formula]
Automatic parameter estimation. The four cluster validity indices evaluated in this study all performed well in a recent evaluation study and are defined there. Most indices try to optimize the balance between tightness (the expression variability within a module) and separation (the expression differences between modules). For metrics requiring a distance matrix, we subtracted the absolute Pearson's correlation from one.
We also investigated two metrics to assess the functional coherence of the modules according to the GO database 45 and the KEGG pathways database [bib_ref] KEGG as a reference resource for gene and protein annotation, Kanehisa [/bib_ref]. We filtered redundant gene sets by, starting from the largest gene set, removing gene sets if they overlap too much with larger non-removed gene sets (Jaccard index > 0.7). The biological homogeneity index measures the proportion of gene pairs within a module which are also matched within a functional class [bib_ref] Methods for evaluating clustering algorithms for gene expression data using a reference..., Datta [/bib_ref]. For the Faucodds score we calculated an aucodds score as described earlier in both the dimension of the gene sets (denoting how well all functional sets are covered by the observed modules) and the dimension of the observed modules (denoting how well the modules are enriched in at least one function gene set), and combined both scores by calculating its harmonic mean.
As automatic parameter estimation performed very poorly on non-exhaustive module detection methods (which include some clustering methods, see Supplementary Note 2), we assigned every unassigned gene to the module with which the average correlation was the highest prior to calculating the cluster validity indices.
Similarity measures. For clustering methods requiring a similarity matrix as input, we used the Pearson's correlation in our initial evaluation. For methods requiring a dissimilarity matrix, we subtracted the Pearson's correlation values from two. For the comparison of different similarity measures, we selected four top clustering methods that require a similarity measure as input. We compared a total of 10 alternative measures that are briefly described in Supplementary Note 3 along with directions to implementations. We did not evaluate the distance correlation, percentage bend correlation, Hoeffding's D, and maximal information coefficient 64 , because they required excessive amounts of computational time and/or memory, which would be impractical for module detection in general use cases. To convert a similarity matrix to a dissimilarity matrix or vice versa, we subtracted the values from the maximal value between all gene pairs on a given dataset. To determine the influence of an alternative similarity measure on the performance of clustering methods, we re-ran all parameter optimization steps for every alternative measure and again calculated test scores as described earlier.
[fig] Figure 3: Effect of automatic parameter estimation using four different cluster validity indices and two measures based on functional enrichment on the performance of top module detection methods. Shown are changes in test scores after parameter estimation (either using measures based on functional enrichment in blue or cluster validity indices in red-orange), averaged over datasets and module definitions, of the top module detection methods in [/fig]
[fig] Figure 4: Influence of the number of samples on the performance of the top module detection methods. Shown are average training scores (left) and test scores (right) over all datasets and module definitions at different levels of random subsampling (five repeats) [/fig]
[fig] Figure 5: Practical guidelines for module detection in gene expression data. Module detection in gene expression data has three main applications (left; panel a). For each application, we suggest different module detection methods (b), which in turn influences the way parameters are estimated (c), how the modules can be visualized (d), and how they can be functionally interpreted (e) [/fig]
|
Rare Diseases of Larynx, Trachea and Thyroid
AbStr Ac tThis review article covers data on rare diseases of the larynx, the trachea and the thyroid. In particular, congenital malformations, rare manifestations of inflammatory laryngeal disorders, benign and malignant epithelial as well as non-epithelial tumors, laryngeal and tracheal manifestations of general diseases and, finally, thyroid disorders are discussed. The individual chapters contain an overview of the data situation in the literature, the clinical appearance of each disorder, important key points for diagnosis and therapy as well as a statement on the prognosis of the disease. Finally, the authors refer to study registries and self-help groups.contents Abstract S1
1 Introduction
The treatment of rare diseases represents a particular challenge. This article comprises a summary of rare diseases of the larynx and the trachea. Another short chapter is dedicated to rare diseases of the thyroid gland. The following overview does not claim to be complete, it will rather focus on several main aspects of the topic. Beside malformations, specific types of inflammations and laryngeal manifestations of general diseases, the manuscript illustrates in particular benign and malignant tumor diseases. This overview will focus on entities for which sufficient data material is available in the literature, mainly in form of case reports or rarely as reviews.
Most of the cited articles were published in German or English. The respective diseases will be briefly summarized in order to achieve a significant overview while keeping the word limitations of the article. For more detailed information we will refer to the cited literature. Within the article, the term of "patient" is used for affected persons, which include all genders. Patient registries and databases allow an improvement of clinical research, optimized care and treatment for patients with rare diseases. One example is the Orphanet database. The website entitled www.orpha.net provides information on registries, current research projects, and trials on rare diseases as well as contact data of self-help organizations. In addition, European networks for care and treatment of patients with rare diseases are founded currently. In this context, the homepage of the European Reference Networks (ERN) of the European Commission shall be mentioned.
## Malformations
Malformations of the larynx and trachea are congenital and belong to the group of rare diseases. Besides stenoses they also include clefts and fistulas. In the following chapter, the most important congenital malformations are illustrated.
## Laryngomalacia
Laryngomalacia is the congenital laryngo-tracheal malformation with the highest incidence and most frequent reason for connatal ▶table 1 Classification of laryngomalacia according to [bib_ref] Laryngomalacia and its treatment, Olney [/bib_ref] [bib_ref] Laryngomalacia and its treatment, Olney [/bib_ref].
Olney type 1 Mucosal prolapse of the arytenoid region/hypertrophy of the accessory laryngeal cartilage Olney type 2 Short aryepiglottic folds Olney type 3 Dorsal displacement of the entire epiglottis stridor in newborns and infants [bib_ref] Etiology of Stridor in the Neonate, Infant and Child, Holinger [/bib_ref]. Numerous publications are found on this disease. An immature, instable cartilaginous skeleton of the larynx is assumed to be responsible for laryngomalacia. This instability leads to a collapse of the supraglottis with consecutive stridor especially in context with forced inspiration. Mostly, the epiglottis is affected. Additionally, a relative hypertrophy of the arytenoid mucosa is observed. Differential diagnostics must exclude other origins of laryngo-tracheal stenoses. 45-75 % of all pediatric cases with stridor are associated with laryngomalacia [bib_ref] Chronic Pediatric Stridor -Etiology and Outcome, Friedman [/bib_ref] [bib_ref] Etiology of stridor in infants, Zoumalan [/bib_ref].
## Clinical manifestation
Typically, the symptoms already occur shortly after birth. An inspiratory stridor is inevitable especially in the context of forced inspiration. Dysphagia with aspiration have been described. In severe cases, the oxygen saturation decreases with resulting cyanosis.
## Diagnostics
In cases of laryngomalacia, transnasal flexible endoscopy under spontaneous breathing is the standard diagnostic procedure. Phonation and respiration but also swallowing can be reliably assessed this way. A well-instructed team and the involvement of the accompanying person (parent) is crucial in order to achieve high quality endoscopy despite defense reactions of the child. Attention must be paid to the risk of possible emergency situations due to acute laryngospasms. For classification (e. g. according to [bib_ref] Laryngomalacia and its treatment, Olney [/bib_ref] , see ▶table 1) and planning of further procedures, rigid endoscopy under sedation with and without spontaneous breathing is optimal. In 10-20 % of the cases, further pathologies such as subglottic stenoses or vocal fold pareses are associated with laryngomalacia [bib_ref] Supraglottoplasty outcomes in relation to age and comorbid conditions, Hoff [/bib_ref] [bib_ref] Synchronous airway lesions in laryngomalacia, Krashin [/bib_ref] [bib_ref] Synchronous Airway Lesions and Outcomes in Infants With Severe Laryngomalacia Requiring Supraglottoplasty, Schroeder [/bib_ref] [bib_ref] Synchronous airway lesions and associated anomalies in children with laryngomalacia evaluated with..., Yuen [/bib_ref].
## Therapy
The decision pro/contra surgical treatment is made based on the clinical overall impression. Crucial criteria are stridor with resting dyspnea, respiration-related nutritional problems, failure to thrive, obstructive sleep disorders, stress-related hypoxia and hypercapnia and cyanosis [bib_ref] When does surgery make sense?, Koitschev [/bib_ref] [bib_ref] Pathologies of the larynx and trachea in childhood, Sittel [/bib_ref]. The surgical therapy is orientated on the classification of laryngomalacia according to Olney. As first measure, a so-called transoral microlaryngoscopical supraglottoplasty is performed. The mucosa in the arytenoid region is partially resected with preservation of the posterior commissure and an incision of the aryepiglottic folds (most frequently) or -in rare cases -epiglottopexy or individual combination of these three measures is performed. Epiglottopexy means a fixation of the epiglottis to the base of the tongue by suture, which prevents a collapse of the epiglottis into the laryngeal aperture [bib_ref] Epiglottopexy for the treatment of severe laryngomalacia, Werner [/bib_ref]. [bib_ref] Synchronous airway lesions in laryngomalacia, Krashin [/bib_ref] Diagnostics By means of flexible endoscopy, the mobility of the vocal folds can be assessed. In newborns, the diagnosis requires an enormous experience due to the anatomical situation as well as secretory retentions which complicate endoscopy. General or comorbid disorders should be excluded by an interdisciplinary team. EMG of the larynx is controversially and very critically discussed because of the difficult performance and assessability [bib_ref] Pathologies of the larynx and trachea in childhood, Sittel [/bib_ref].
## Therapy
Intubation is necessary in most of the cases. While older publications mention tracheostomy as unique and permanent therapeutic option, more recent articles also recommend endolaryngeal surgery, e. g. partial unilateral laser chordotomy modified according to Kashima [bib_ref] Laryngeal paralysis in children: a long-term retrospective study, Cohen [/bib_ref] [bib_ref] Outcome of laryngeal paralysis in neonates: a long term retrospective study of..., De Gaudemar [/bib_ref] [bib_ref] Pediatric vocal fold paralysis: a long-term retrospective study, Daya [/bib_ref] [bib_ref] Vocal cord paralysis in children 1 year of age and younger, Gentile [/bib_ref] [bib_ref] Vocal cord palsy in pediatric practice: a review of 71 cases, Emery [/bib_ref] [bib_ref] Vocal fold paralysis in infants twelve months of age and younger, Zbar [/bib_ref]. Spontaneous regeneration of the vocal fold mobility within the first 12 months of life have been described [bib_ref] Natural history of tracheostomy-dependent idiopathic congenital bilateral vocal fold paralysis, Berkowitz [/bib_ref] so that interventions that might have irreversible consequences in the sense of scarring and dysphonia are recommended only after the 12 th month of life [bib_ref] Carbon Dioxide Laser Cordotomy as Possible Alternative for Tracheotomy in a Neonate..., Ruf [/bib_ref].
## Prognosis
According to the literature, up to 80 % of the cases require tracheostomy as initial airway management. In 29-71 % of the cases, the vocal fold mobility starts spontaneously after 6 months to 11 years. Partial chordotomy is recommended after the 12 th month of life in cases of persisting symptoms. All these procedures lead to dysphonia and impaired respiratory function. The potential morbidity of children with tracheostoma must be taken into account. Note: The therapy of congenital bilateral vocal fold palsy is an interdisciplinary challenge involving pediatricians as well as otolaryngologists/phoniatricians.
## Laryngeal clefts
In the literature, 214 publications are found dealing with diagnostics, therapy, comorbidities, and complications of laryngeal clefts. The occurrence of occult, non-symptomatic small laryngeal clefts seems to be rather high. The number of publications has continuously increased during the last years due to improved diagnostic options.
## Clinical manifestation
The incidence of laryngeal clefts amounts to 0.5-1 per 100 000 live births [bib_ref] Dissertatio médico de infanticide inartis obstetriciae: Thesis for Doctor of, Richter [/bib_ref] [bib_ref] Laryngo-Tracheo-Esophageal Cleft -Clinical-Features, Diagnosis and Therapy, Roth [/bib_ref]. Male newborns are more frequently affected than females. The development of laryngeal clefts is based on a disturbed fusion of the tracheoesophageal septum in the 5 th to 7 th week of embryogenesis [bib_ref] Congenital Larygeal Cleft, Delahunty [/bib_ref]. While laryngeal clefts appear as isolated disorder in most cases, they may also be part of a general syndromatic l disease, e. g. Opitz G/BBB syndrome, CHARGE or VACTERL association [bib_ref] Diagnosis and management of type 1 laryngeal cleft, Watters [/bib_ref] [bib_ref] The presentation and management of laryngeal cleft: a 10-year experience, Rahbar [/bib_ref]. Laryngeal clefts are classified according to Benjamin and Inglis (▶Fig. 1 and ▶table 2).
The clinical symptoms of a type 1 cleft are mostly unspecific: difficulties in drinking, intra-and postdeglutitive cough, bronchopulmonary infections up to pneumonias requiring inpatient treat-▶Fig. 1 Laryngeal clefts. type [bib_ref] Etiology of Stridor in the Neonate, Infant and Child, Holinger [/bib_ref] The laryngeal cleft only concerns the postcricoid region and ends without cartilage involvement above the level of the vocal folds. type 2 The laryngeal cleft draws into the cricoid cartilage and ends below the level of the vocal folds. type 3 There is also a cleft of the trachea, which ends above the thoracic aperture. type 4 The tracheal part of the cleft draws beyond the thoracic aperture just above the bifurcation. ment [bib_ref] Diagnosis and management of type 1 laryngeal cleft, Watters [/bib_ref] [bib_ref] The presentation and management of laryngeal cleft: a 10-year experience, Rahbar [/bib_ref] [bib_ref] Endoscopic repair of laryngeal cleft type I and type II: when and..., Rahbar [/bib_ref] [bib_ref] Minor congenital laryngeal clefts: diagnosis and classification, Benjamin [/bib_ref]. The primary diagnosis is often delayed. Whereas the symptoms of type 3 or type 4 cleft become apparent mostly within the first minutes or hours after birth, type 1 and type 2 are frequently discovered only in the further course [bib_ref] Laryngeal cleft: evaluation and management, Johnston [/bib_ref].
## Diagnostics
The gold standard in the diagnostics of laryngeal clefts is rigid endoscopy under general anesthesia with palpation of the posterior larynx [bib_ref] Laryngeal cleft: evaluation and management, Johnston [/bib_ref]. In most cases, X-ray or CT scan of the thorax are performed prior to endoscopy. Flexible endoscopy in the awake patient or functional swallowing examination may give hints on a cleft but they cannot exclude it in case of regular findings [bib_ref] Type 1 laryngeal cleft: establishing a functional diagnostic and management algorithm, Chien [/bib_ref].
## Therapy
In cases of type I clefts, wait-and-see strategy over 6 months is often sufficient with adaptation of the nutrition (up to feeding tubes non per os) as well as therapy of bronchopulmonary infections [bib_ref] The presentation and management of laryngeal cleft: a 10-year experience, Rahbar [/bib_ref] [bib_ref] Endoscopic laryngotracheal cleft repair without tracheotomy or intubation, Sandu [/bib_ref] [bib_ref] Endoscopic surgical treatment of laryngotracheal clefts: indications and limitations, Garabedian [/bib_ref]. In most cases, the normal growth leads to regression of the symptoms so that surgical intervention is only rarely required in type I clefts [bib_ref] Endoscopic repair of laryngeal cleft type I and type II: when and..., Rahbar [/bib_ref] [bib_ref] Type 1 laryngeal cleft: establishing a functional diagnostic and management algorithm, Chien [/bib_ref] [bib_ref] Type I posterior laryngeal clefts, Parsons [/bib_ref] [bib_ref] Clinical aspects of type 1 posterior laryngeal clefts: literature review and a..., Van Der Doef [/bib_ref]. However, if the symptoms are still present after 6 months, the indication of surgical treatment is given. While type I and sometimes II can be sufficiently treated by microlaryngoscopy including suture techniques or augmentation, type III clefts require an open approach, type IV even thoracotomy [bib_ref] Management and long-term follow-up of patients with types III and IV laryngotracheoesophageal..., Kawaguchi [/bib_ref] [bib_ref] Laryngotracheoesophageal cleft (type IV): management and repair of lesions beyond the carina, Ryan [/bib_ref].
## Prognosis
In type I and II laryngeal clefts, a posttherapeutic healing rate of up to 94 % is described. The revision rate amounts to less than 5 % [bib_ref] Endoscopic surgical treatment of laryngotracheal clefts: indications and limitations, Garabedian [/bib_ref]. Larger clefts of type III or IV require a challenging management beside tracheostomy and an open, median (thyrofissure) or lateral surgical approach. One critical complication is the postoperative tracheoesophageal fistula. Depending on possible congenital general diseases, the mortality in type III and IV clefts amounts to about 50 % [bib_ref] Management and long-term follow-up of patients with types III and IV laryngotracheoesophageal..., Kawaguchi [/bib_ref] [bib_ref] Strategies for managing Type IV laryngotracheoesophageal clefts at Great Ormond Street Hospital..., Mathur [/bib_ref] [bib_ref] Surgical management of type III and IV laryngotracheoesophageal clefts: the three-layered approach, Geller [/bib_ref]. Note: The therapy of laryngeal clefts depends on their severity. Smaller clefts with mild symptoms should be observed and treated symptomatically during the first 6 months. In the further course, endoscopic treatment may be necessary. Sever clefts of type III and IV usually become symptomatic immediately after birth and require an extensive management including tracheostomy.
## Tracheoesophageal fistulas and esophageal atresia
In the literature, more than 5000 publications are found on this topic. They cover case reports, original papers on diagnostics and surgical procedures as well as numerous review articles.
## Clinical manifestation
The development of tracheoesophageal fistulas (TEF) is not yet finally clarified since its first description by Gibson in 1697 [bib_ref] Defective epithelial-mesenchymal interactions dictate the organogenesis of tracheoesophageal fistula, Crisera [/bib_ref]. Missing formation of the esophageal septum as well as disturbed development of the embryonic bronchial structures are discussed [bib_ref] Defective epithelial-mesenchymal interactions dictate the organogenesis of tracheoesophageal fistula, Crisera [/bib_ref] [bib_ref] Esophageal atresia with tracheoesophageal fistula: suggested mechanism in faulty organogenesis, Crisera [/bib_ref]. In 90 % of the cases, congenital TEF are associated with esophageal atresia, their incidence is estimated to about 1 of 3500-4500 live births [bib_ref] One shall become two: Separation of the esophagus and trachea from the..., Billmyre [/bib_ref] [bib_ref] Prevalence, characteristics, and survival of children with esophageal atresia: A 32-year populationbased..., Cassina [/bib_ref]. The clinical appearance of TEF depends on the severity of the esophageal atresia and the type of fistula. The classification of esophageal atresia is based on the description of the American radiologist Edward Vogt (▶Fig. 2) [bib_ref] From Vogt to Haight and Holt to now: the history of esophageal..., Muensterer [/bib_ref]. A tracheoesophageal fistula is observed in cases of type IIIa, IIIb, IIIc, and IV, the so-called H fistula [bib_ref] Esophageal atresia with proximal tracheoesophageal fistula: A missed diagnosis, Parolini [/bib_ref]. In cases of esophageal atresia, the drainage of the amniotic fluid is impeded so that a polyhydramnios is pathognomonic [bib_ref] Tracheoesophageal fistula in utero. Twenty-two cases, Pretorius [/bib_ref]. Type IIIb is seen most frequently. Beside nutritional impairment, affected newborns also show dyspnea and stridor. Furthermore, aspiration pneumonia may appear due to regurgitation of the stomach content via the fistula into the trachea. ▶table 2 Laryngeal clefts according to [bib_ref] Minor congenital laryngeal clefts: diagnosis and classification, Benjamin [/bib_ref] In the context of a so-called H fistula between the esophagus and trachea with regular patency of the esophagus, early pulmonary symptoms such as aspiration, cough, or even suffocation may result. Smaller H fistulas can also be asymptomatic [bib_ref] The diagnosis and treatment of H-type tracheoesophageal fistula, Karnak [/bib_ref]. In more than 40 % of the cases, esophageal atresia with/without TEF is associated with other -mainly cardiac -malformations and appears also in the context of syndromes.
## Diagnostics
Before birth, the first suspicion arises due to the presence of polyhydramnios. Therefore, in rare cases the diagnosis can already be made before birth. The clinical suspicion implicates the recommendation of delivery in a center with according experience and equipment. The impossibility to completely explore the esophagus with a nasogastric tube (not more than 10-15 cm) indicates the presence of esophageal atresia. By means of X-ray, the tube can be displayed in the upper esophageal sac. In addition, air-fluid levels are typically seen. The use of barium contrast agent is obsolete since it may induce pneumonitis. Fistulas and atresias may be well examined by endoscopy, which is essential for planning an intervention. Esophageal atresia is often associated with syndromic diseases (e. g. trisomy 18) [bib_ref] Esophageal atresia in the Northern Region Congenital Anomaly Survey, 1985-1997: prenatal diagnosis..., Sparey [/bib_ref].
## Therapy
The responsibility of the therapy is with the pulmonary/pediatric surgeons. Surgeries of the thoracic area in the sense of end-to-end anastomoses can be performed via an open approach (which is associated with high morbidity) or as minimally invasive procedure. The fistulas are closed surgically [bib_ref] Thoracoscopic repair of esophageal atresia and tracheo-esophageal fistula, Rothenberg [/bib_ref] [bib_ref] Thoracoscopic repair of esophageal atresia and tracheo-esophageal fistula in neonates: evolution of..., Rothenberg [/bib_ref]. Smaller TEF can be closed with fibrin glue. Insufficiencies of the anastomoses occur frequently in 20 % of the cases and are treated conservatively.
## Prognosis
The risk of post-therapeutic esophageal stenosis amounts to 25-40 %; therapy consists of repeated dilatations. The post-therapeutic mortality of 5-10 % is influenced among others by comorbidities. Affected patients may contact the self-help organization for esophageal diseases called KEKS.
Note: Esophageal atresia is mostly associated with tracheoesophageal fistula. Therapy consists of surgery; the post-therapeutic morbidity is relatively high.
## Laryngeal web (diaphragm)
In the literature, about 200 publications are found on the management of this disease.
## Clinical manifestation
The congenital glottic web or diaphragm was first described in 1882 by Fleischmann. These webs make up about 5 % of the congenital anomalies of the larynx and are caused by missing laryngeal recanalization during organ genesis [bib_ref] Congenital webs, cysts, laryngoceles and other anomalies of the larynx, Holinger [/bib_ref] [bib_ref] The anterior laryngeal webs, Nicollas [/bib_ref]. Sometimes they are associated with syndromes such as 22q11.2 deletion syndrome [bib_ref] Association of anterior glottic webs with velocardiofacial syndrome (chromosome 22q11.2 deletion), Miyamoto [/bib_ref]. The congenital glottic web is classified according to Cohen. Depending on the severity of the glottic stenosis four types are differentiated. Type I causes glottic stenosis of less than 35 %, type II between 35 and 50 %, type III comprises 50-75 % of the glottic surface and always reaches to the anterior cricoid part in the subglottis. In type IV, a stenosis of 75-99 % is found. The complete congenital laryngeal stenosis is called "congenital high airway obstruction syndrome" or CHAOS (see tracheal agenesis) [bib_ref] Congenital high airway obstruction syndrome: natural history and management, Lim [/bib_ref] [bib_ref] Laryngeal Web in the Pediatric Population: Evaluation and Management, Lawlor [/bib_ref]. In grade I and II, the vocal folds are still defined, smaller findings may sometimes remain clinically undetected. In grade III and IV, the differentiation of the vocal folds is nearly impossible.
Due to the different degrees of stenosis, the clinical symptoms of the affected patients may vary. Patients with type I findings are often asymptomatic. Others report about dysphonia that also persists when patients cry. The symptoms may worsen in cases of infection of the upper airways. Grade III and IV may impose by highly relevant persisting dyspnea with stridor as of birth [bib_ref] Dysphonia in children due to congenital laryngeal web. Case series, Rodriguez [/bib_ref].
## Diagnostics
Newborns with dysphonia and stridor should be examined by an ENT specialist with pediatric expertise. The diagnosis is made by transnasal flexible endoscopy. Rigid endoscopy under sedation with and without spontaneous breathing is essential in order to assess the severity of the anomaly. In these cases, special attention is necessary so that no further scarring by inadequate intubation is caused.
## Therapy and prognosis
Small asymptomatic congenital glottic webs may first undergo wait-and-see strategy. If surgery becomes necessary due to worsening of the symptoms, it should be performed endoscopically in cases of small findings. Different procedures are applied such as balloon dilatation and transection of the web with the CO 2 laser. It must be observed that most webs reach into the subglottic area which can make the temporary insertion of a silicone keel necessary. Grade III and IV webs sometimes require initial tracheostomy with subsequent reconstruction that is usually performed in several steps [bib_ref] Laryngotracheal reconstruction in congenital laryngeal webs and atresias, Wyatt [/bib_ref]. In summary, tracheostomy should be avoided if possible. In cases of congenital glottic webs of grade III and IV, open approaches are preferred [bib_ref] Management of severe congenital laryngeal webs -a 12 year review, De Trey [/bib_ref].
Note: In most cases, glottis webs reach into the cricoid cartilage and therefore the simple transection does not generally suffice as therapy.
## Tracheal agenesis
In the literature, about 200 publications on the topic of tracheal agenesis are found. The reason for the development of tracheal agenesis is not clarified. Theories exist about the displacement of the esophagotracheal septum in dorsal or ventral direction during embryogenesis [bib_ref] Tracheal agenesis. A rare cause of respiratory insufficiency in neonates, Schummer [/bib_ref].
## Clinical manifestation
Tracheal agenesis (TA) was first described in 1900. Since then, only few cases have been published. The prevalence amounts to about 1 of 50 000-100 000 live births with a ratio of 2:1 regarding male and female patients. In general, the disease has a lethal outcome. In most cases, it is diagnosed before birth when a so-called CHAOS syndrome (see above) is detected. However, a CHAOS syndrome cannot develop in cases of tracheoesophageal fistula so that the diagnosis in these cases can only be made shortly after birth.
According to Floyd, three types are classified . Type I (20 %) consists of an agenesis of the upper trachea, a fistula is found between the trachea and the esophagus, the bronchi are regular.
Type II (60 %) consists of complete tracheal agenesis, the bronchi are regular and a fistula is found between the carina and the esophagus. In cases of type III, the main bronchi separately exit from the esophagus [bib_ref] Agenesis of the trachea, Floyd [/bib_ref] [bib_ref] Congenital tracheal malformations, Varela [/bib_ref] (▶Fig. 4). TA is rarely observed as isolated malformation, in up to 94 % other congenital anomalies are present, e. g. VATER association (vertebral defects, anal atresia, tracheoesophageal fistula with esophageal atresia, and radial or renal dysplasia). The leading symptom is severe stridor after birth that unfortunately has a rapid lethal outcome [bib_ref] Clinical presentation and airway management of tracheal atresia: A systematic review, Smith [/bib_ref].
## Diagnostics
In general, polyhydramnios can be identified during pregnancy by means of ultrasound. Furthermore, also the trachea can be well displayed in prenatal diagnostics [bib_ref] Doppler assessment of tracheal and nasal fluid flow during fetal breathing movements:..., Kalache [/bib_ref]. Severe dyspnea and the associated impossibility of endotracheal intubation are cardinal symptoms of TA.
## Therapy and prognosis
Patients with complete TA cannot survive. The results of surgical therapy of TA are disillusioning up to now. Only four cases are found in the literature with short survival periods. Multidisciplinary approaches are required in this context. No reports are available about patients who have reached toddler age [bib_ref] Tracheal agenesis. A rare cause of respiratory insufficiency in neonates, Schummer [/bib_ref].
Note: Tracheal agenesis is not compatible with life even if the diagnosis is already made before birth.
## Rare types of laryngitis
The unspecific acute and chronic laryngitis is a frequent disease that may have manifold origins. Among acute as well as chronic types, rare subgroups are found. Rare types of acute laryngitis comprise ulcero-membranous laryngitis, which is a group of different diseases that are characterized by fibrinous deposits on the mucosa. Classic examples are diphtheria and the ulcero-membranous Plaut-Vincenti disease. In the last 10 years, no articles have been published due to the rarity of the cases of laryngeal diphtheria so that it is only mentioned here. The rare types of chronic laryngitis include some specific inflammatory diseases, in first place laryngeal tuberculosis that will be dealt with in the following chapter. Furthermore, leprosy of the larynx will be discussed. Laryngeal syphilis is extremely rare. Primary, secondary, and tertiary affections typically manifest rather in the oral cavity and the oropharynx and only in very rare cases in the larynx. In the current literature, only one case is reported [bib_ref] Laryngeal Syphilis A Case Report, Lahav [/bib_ref].
## Laryngeal tuberculosis
In the literature, more than 700 publications are found on laryngeal tuberculosis. They appeared mainly in the 1950ies, but also in the last years, articles were regularly published on this topic.
## Clinical manifestation
Globally, tuberculosis represents a relevant disease. However, in Europe -especially in Germany -it ranks among the rare diseases. Males are twice as much affected than females. In general, young adults acquire this disease. The association with an HIV infection is found in up to 50 % [bib_ref] Laryngeal tuberculosis in HIV-infected patients: A difficult diagnosis, Singh [/bib_ref]. Endoscopy shows ulcera as well as hypertrophic granulation manifestations [bib_ref] Laryngeal tuberculosis: A review of 26 cases, Wang [/bib_ref]. All areas of the larynx can be involved, however, the vocal folds are more frequently affected than other regions. In the older literature, posterior laryngeal parts were defined as predilection sites; the reason for this is supposed to be the retention of infectious sputum from the deep airways in the postcricoid region. More recent systematic observational trials, however, could not confirm this assumption. Typically, only one side is affected, e. g. as so-called monochorditis of one vocal fold [bib_ref] Current clinical propensity of laryngeal tuberculosis: review of 60 cases, Lim [/bib_ref]. Due to the unspecific appearance, laryngeal tuberculosis may be endoscopically mis-diagnosed as laryngeal cancer. In nearly all patients, the chest X-ray is conspicuous even if only part of these patients suffer from chronic cough. ▶Fig. 3 Tracheal agenesis. type 1 A tracheal stump is found above the bifurcation into which a fistula opens from the cervical esophagus; incidence of 20 %. type 2 A tracheal stump does not exist, the bifurcation is directly connected with the esophagus; incidence of 60 %. type 3 In this case, also the bifurcation is missing, both main bronchi open separately into the esophagus; incidence of 20 %. The diagnosis is confirmed by biopsy with histological detection of caseous epitheloid-cell granulomas and acid-fast bacilli in the Ziehl-Neelsen staining or PCR as well as sputum analyses with the cultural proof of Mycobacterium tuberculosis [bib_ref] Primary Laryngeal Tuberculosis: A Series of 15 Cases, Agarwal [/bib_ref]. Nearly all patients have additional pulmonary affection so that chest X-ray is highly relevant [bib_ref] Tuberculous Laryngitis -a Series of 37 Patients, Bailey [/bib_ref]. In general, the sputum culture is positive, however, this examination takes much time due to the long duration of cultivation. The disease occurrence and tuberculosis-related deaths as well as the microbiological detection of Mycobacteria have to be reported to the health authorities.
## Therapy
After the diagnosis has been confirmed, therapy should be initiated as soon as possible. Regression of the laryngeal lesions usually starts some weeks after therapy onset. In general, tuberculostatic therapy is performed over 6 months [bib_ref] Primary Laryngeal Tuberculosis: Our Experiences at a Tertiary Care Teaching Hospital in..., Swain [/bib_ref]. However, scars develop very regularly, especially in the larynx. Typically, fixations of the cricoarytenoid joint, posterior glottic stenoses, or subglottic stenoses occur with resulting persistent dysphonia and stress-related dyspnea.
## Prognosis
With adequate therapy, the prognosis of the patients with regard to complete healing of the disease is very good. However, complex scarring is observed frequently, which is very difficult to treat and leads to functional impairment [bib_ref] Laryngeal tuberculosis: a review of 26 cases, Wang [/bib_ref]. Note: Monochorditis is a typical manifestation of laryngeal tuberculosis.
## Leprosy of the larynx
Only sporadic case reports on leprosy of the larynx are found in the literature or the publications are several decades old.
## Clinical manifestation
The disease is caused by Mycobacterium leprae and is nowadays mainly found in tropical and subtropical areas of Asia, especially in India. Long-lasting and close physical contact is a precondition for infection, so that the disease often affects people who live together in narrow conditions. Since lower temperatures promote cell division of leprosy bacteria, the nasal mucosa is particularly colonized and is thus considered as pathogen reservoir and main infection source. The development of granulation tissue (lepromas) is typical. It surrounds preferably the peripheral nerves and destroys them in the further course. In cases of leprosy manifestations of the nasal mucosa, laryngeal affection is observed in about 30 % of the cases. Lepromas are nearly exclusively found on the prominent parts of the larynx, i. e. the free edge of the epiglottis where they appear as tumor-like distended masses.
## Diagnostics
According to the recommendations of the Robert-Koch Institute (www.rki.de), the diagnostics is based on histological examination or the identification of specific DNA by PCR from skin or mucosal biopsies. Sometimes, detection of pathogens is also possible in the lymphatic fluid, which is harvested by scratching the skin at predefined sites. Disease occurrence and/or death have to be reported to the health authorities.
## Therapy and prognosis
The treatment of leprosy is oriented at the therapy of tuberculosis. Poly-drug-therapy is performed for several months. In the larynx, the lesions heal with scar formation. In addition, sensitivity disorders are described because of damage of the superior laryngeal nerve with resulting aspiration pneumonia [bib_ref] Leprosy of the larynx, Soni [/bib_ref] Note: Lepromas preferably manifest in close neighborhood to peripheral nerves and lead to sensitivity disorders. Manifestation in the larynx may lead to aspiration.
## Laryngeal leishmaniasis
Leishmaniasis is an infectious granulomatous disease that may be caused by the protozoae Leishmania (L) brasiliensis, L. amazonensis, L. panamensis, or L. guyanensis. The primary vectors are sand flies. The disease is classified into three clinical manifestations: the visceral type (Kala-Azar), the cutaneous type, and the mucocutaneous type. Mucosal leishmaniasis generally manifests after hematogenic distribution over months and years after skin infection [bib_ref] Laryngeal leishmaniasis, Grant [/bib_ref]. It is often observed in Latin America, the pathogen is L. brasiliensis [bib_ref] Mucosal Leishmaniasis Caused by Leishmania (Viannia) braziliensis and Leishmania (Viannia) guyanensis in..., Guerra [/bib_ref]. The occurrence of mucocutaneous manifestations depends on the virulence of the parasites and the immunocompetence of the host. Hence, even in endemic regions it is a rare disease. Most frequently, the nose, pharynx, larynx, and oral cavity are affected. Infection usually occurs in the context of journeys to foreign countries. The Robert-Koch Institute reports about 20 cases each year in Germany.
## Clinical manifestation
In general, laryngoscopy shows extended inflammatory lesions with erythema and edema. Purulent exudation may be observed. In advanced cases, tissue destruction occurs which can be extended and is often misinterpreted as malignant disease.
## Diagnostics
The diagnosis of leishmaniasis is based on direct cytological or histological evidence of intracellular parasites. Furthermore, immunological tests are available, e. g. the identification of antibodies in the immunofluorescence or via ELISA. PCR has the highest sensitivity and specificity that is performed with bone marrow or skin specimens or also with peripheral blood. Most decisive is biopsy and travel history taking.
## Therapy and prognosis
An adequate therapy of the cutaneous lesions can reduce the risk of mucosal disease which may appear even years after primary infection. The therapy of choice is the application of liposomal amphotericin B (AmBisome, e. g. Gilead Science GmbH, Martinsried, Germany). Miltefosin p.o. (Impavido, Paesel und Lorei GmbH & Co. KG, Rheinberg, Germany) is considered as measure of second choice. Theoretically, the disease is curable by adequate therapy, however, complete healing is not possible in many cases. Especially the cutaneous type often leads to disfiguring scar formations. In the context of diagnostics, also the immunocompetence of the patient should be verified (HIV test). The primary prophylaxis (application of repellents) is of particular importance.
Note: Laryngeal leishmaniasis may develop even years after skin manifestation. Due to its clinical appearance, it may be misdiagnosed as laryngeal carcinoma.
## Benign tumors
## Benign epithelial tumors
## Juvenile and adult papillomas
A large number of publications on recurrent respiratory papillomatosis can be found in the literature with more than 800 original articles and reviews. In the current overview, the disease will be briefly summarized. The authors want to refer to the numerous comprehensive review articles on this topic.
## Clinical manifestation
Laryngeal papillomas are the most frequently observed benign epithelial tumors of the larynx [bib_ref] Recurrent respiratory papillomatosis: a review, Derkay [/bib_ref]. The difference is made between two types: The juvenile type, which is diagnosed before the age of 12 and generally occurs in young children and the adult type, that mainly affects young, preferentially male adults [bib_ref] Epidemiology of recurrent respiratory papillomatosis, Larson [/bib_ref]. Both types have a comparable incidence of 2-4 per 100 000 people and thus count to orphan diseases. Laryngeal papillomatosis is caused by different subtypes of the human papilloma virus (HPV), mainly HPV Type 6 and 11. HPV Type 11 seems to be associated with more aggressive courses. The manifestation within transition zones between various epithelia (e. g. between squamous epithelium and respiratory epithelium) is typical. Laryngeal papillomas are found especially at the vocal folds [bib_ref] Human papillomavirus-11-associated recurrent respiratory papillomatosis is more aggressive than human papillomavirus-6-associated disease, Rabah [/bib_ref]. The infection pathway is still not fully understood. Peripartal infection is assumed to be the reason for the juvenile type. A correlation can be identified with the presence of maternal anogenital papillomas. Regarding the adult type, sexual infection pathways are additionally discussed [bib_ref] Juvenile respiratory papillomatosis: risk factors for severity, Rodier [/bib_ref]. The clinical appearance depends on the location and the size of the papillomas. Chronic hoarseness and -at later stages -obstruction of the airways are typical symptoms. In most cases, the papillomas concern the laryngeal mucosa. Tracheal or even bronchial extension is very rare and associated with a particularly poor prognosis [bib_ref] Diagnosis and management of pulmonary metastasis from recurrent respiratory papillomatosis, Silver [/bib_ref].
## Diagnostics
Indirect laryngoscopy may already allow a diagnoses in many cases. Papillomas are characterized by a cauliflower-like surface with a specific vascular pattern [fig_ref] ▶Figure 5: Papillomatosis [/fig_ref]. Histological examination provides the evidence. The detection of HPV DNA in the tissue is nowadays diagnostic standard. Special imaging procedures, e. g. narrow-band imaging or comparable techniques, may be helpful to identify also small manifestations [bib_ref] Analysis of a staging assessment system for prediction of surgical interval in..., Derkay [/bib_ref].
## Therapy
Usually, the removal of disturbing or critically extended papilloma foci is performed by transoral microsurgical procedures. Such resections are strictly limited to the epithelium because the virus does not affect the subepithelial tissue. Radical resections are obsolete, they do not avoid recurrences, but they harm the functionality of the larynx. Recurrences generally develop in adjacent mucosal areas that are also infected by the virus. Repeated session are frequently necessary to remove the papillomas by different measures, based on the individual symptoms. The occurrence of recurrences is pathognomonic for the disease. Surgical and pharmacotherapies will only briefly be mentioned in this chapter, the authors refer to the specific literature. In Europe, therapy of choice is often CO 2 laser surgery with tissue preserving evaporation of the papil-lomas. In the US, shaver procedures are primarily applied. The more recent literature describes the application of new types of lasers. There are numerous publications on drug-related therapy. Pharmacotherapy aims at prolonging the intervals between surgeries and has to be considered as adjuvant measure. Interferon-alpha represented the first treatment strategy to support the immune system in papilloma patients. However, the medication is only rarely applied today due to relevant systemic side effects. Instead, the intralesional, partly also systemic application of bevacizumab (Avastin, e. g. Haemato Pharm GmbH, Schönefeld, Germany) and cidofovir (Tillomed Pharma GmbH, Ahrensberg, Germany) are in the focus today. The VEGF antibody bevacizumab is already routinely applied for example in ophthalmology. As angiogenesis inhibitor it can impede the new development of papillomatous foci. The use of cidofovir is discussed extensively in the literature, while a final assessment of the very different trials is problematic. In any case, repeated applications are necessary which may lead to relevant toxicity. More recent data on the COX-2 inhibitor celecoxib do not show any advantage for the patients. In a phase-II study, currently the effectiveness of pembrolizumab on the papilloma growth is evaluated [bib_ref] Current and future management of recurrent respiratory papillomatosis, Ivancic [/bib_ref]. Furthermore, there are small case series where the tetravalent vaccine Gardasil (MSD Vaccins, Lyon, France) has been applied in papilloma patients. Partly, promising results are found in small case series or in case reports. The pathophysiological mechanism of the preventive approach is not known up to now.However, it is assumed that in particular an infection of non-affected areas may be inhibited [bib_ref] Laryngeal Recurrent Respiratory Papillomatosis: Current Aspects on Diagnosis and Therapy, Langer [/bib_ref].
## Prognosis
The recurrent respiratory papillomatosis is a disease that cannot be cured by current therapeutical measures. The prophylactic effects of comprehensive vaccination will be evident in the next decades. Prognostically, a rapid growth as well as an affection of the trachea are very unfavorable. Reports on malignant degeneration in adults are available, a malignant transformation rate of about 2 % is assumed [bib_ref] Current and future management of recurrent respiratory papillomatosis, Ivancic [/bib_ref]. Note: The objective of the therapy of laryngeal papillomatosis is the thorough removal of the papillomas with shaver or laser with preservation of the laryngeal function.
## Pleomorphic adenomas
Pleomorphic adenomas of the larynx are very rare, in the literature 28 cases have been described.
## Clinical manifestation
Adenomas in the area of the larynx originate from the seromucous glands and are histologically identical with the tumors of the salivary glands. The pleomorphic adenoma is the most frequently observed benign tumor of the major salivary glands. More rarely, the minor salivary glands are affected, and particularly rarely those in the larynx [bib_ref] Benign pleomorphic adenoma of the larynx: report of a case and review..., Dubey [/bib_ref]. Monomorphic adenomas with tubular, acinar, or papillary histomorphology occur extremely rarely in the larynx. More frequently, but also very rare, pleomorphic adenomas are described [bib_ref] Primary carcinoma ex-pleomorphic adenoma of anterior commissure of the larynx, Kovarik [/bib_ref]. The submucous masses are slowly growing and painless, clinical findings are unspecific. Radiologically and clinically, they cannot be differentiated from other benign tumors. Rarely, malignant transformation is observed, two cases have been described in the literature [bib_ref] Primary carcinoma ex-pleomorphic adenoma of anterior commissure of the larynx, Kovarik [/bib_ref] [bib_ref] Carcinoma arising in a pleomorphic adenoma of the epiglottis, Milford [/bib_ref]. Typical locations for pleomorphic adenomas are the supraglottis [bib_ref] A case of pleomorphic adenoma of the epiglottis. Bilateral vocal-cord paralysis after..., Ito [/bib_ref] , the subglottis [bib_ref] Subglottic pleomorphic adenoma, Pardal Refoyo [/bib_ref] and the trachea [bib_ref] Pleomorphic adenoma of the trachea, Sim [/bib_ref] with nearly equal incidences. Rarely, they manifest on the level of the glottis [bib_ref] Pleomorphic Adenoma of the Larynx: A Case Report, Motahari [/bib_ref] [bib_ref] Pleomorphic adenoma of the larynx, Argat [/bib_ref]. The symptoms depend on the location and the size. Hoarseness, dysphagia, or (sometimes life-threatening) dyspnea are observed [bib_ref] Subglottic pleomorphic adenoma, Pardal Refoyo [/bib_ref].
## Diagnostics
Radiological imaging (CT scan and MRI) mainly aims at identifying the extension. Differential diagnostic classification is not possible. For this purpose, an excision (in larger tumors biopsy) with histological and immunohistological assessment is necessary. The pathologist must identify epithelial and myoepithelial-stromal components in order to verify the diagnosis. The expression of cytokeratin, S100, glial fibrillary acidic protein (GFAP), and vimentin are typical for pleomorphic adenomas [bib_ref] Pleomorphic adenoma of the larynx, Altunpulluk [/bib_ref].
## Therapy
Surgery is the therapy of choice. The treatment concept should aim at preserving functional aspects beside complete tumor resection. Open accesses, e. g. via lateral pharyngotomy [bib_ref] Pleomorphic Adenoma of the Larynx: A Case Report, Motahari [/bib_ref] , and partly also transoral microsurgical resections are described [bib_ref] Subglottic pleomorphic adenoma, Pardal Refoyo [/bib_ref] [bib_ref] Pleomorphic adenoma of the epiglottis. Case report and review of the literature, Suttner [/bib_ref]. Recurrences have been reported as they are also found in other locations.
## Prognosis
It remains open if the recurrences described in the literature are residual findings after incomplete removal or true recurrences. In the literature, also two cases of a carcinoma ex pleomorphic adenoma of the larynx are described, otherwise the prognosis seems to be favorable.
Note:
The key for diagnosis is the immunohistological characterization, therapy optimally aims at complete tumor removal with at the same time preservation of the laryngeal function.
## Benign non-epithelial tumors
## Chondromas
In the literature, 85 publications are found. They include about 250 cases, among them three children [bib_ref] Laryngeal chondroma: a rare neck tumor in pediatric age, Ozlugedik [/bib_ref].
## Clinical manifestation
Beside chondrosarcomas, chondromas belong to the tumors of the laryngeal framework [bib_ref] Laryngeal Chondroma: An Unusual Complication Endotracheal Entubation, Gokdogan [/bib_ref]. The incidence of chondromas and chondrosarcmoas within the group of all laryngeal neoplasms amounts to clearly less than 1 %. In 70-75 % of the cases, chondromas originate from the luminal side of the cricoid plate, followed by thyroid cartilage [bib_ref] Laryngeal chondroma: a rare diagnosis in this localization, Tastekin [/bib_ref]. Single case reports also describe endolaryngeal chondromas at the arytenoid cartilage, mostly unilateral, in rare cases also bilateral [bib_ref] Bilateral Vocal Fold Chondromas, Lai [/bib_ref]. Furthermore, Yang and Lin published the case of laryngeal chondroma of the epiglottis [bib_ref] A peculiar site of chondroma: The epiglottis, Yang [/bib_ref] and Ozcan et al. about chondroma of the hyoid [bib_ref] Hyoid chondroma presenting as an external neck mass, Ozcan [/bib_ref]. Mostly adults younger than 50 years are affected while males are more frequently affected than females with a ratio of 4:1 [bib_ref] Congenital laryngeal chondroma: an unusual cause of congenital stridor, Gundlach [/bib_ref]. Clinically, the tumors impose as hard, often spherical masses growing in-and outwardly and covered by regular mucosa. The symptoms vary according to the size and location of the tumors. Especially in endolaryngeal findings, dysphonia or increasing dyspnea with inspiratory stridor dominate. At the typical locations of the cricoid, the tumors grow also in direction of the esophagus entrance and may cause swallowing disorders. In some cases, chrondromas present as a palpable masses at the outer neck or as incidental findings in the context of imaging due to other reasons. Because of the slow growth, history often comprises several years. Cartilaginous tumors also appear in the trachea. These osteochondromas are sometimes detected as incidental findings during endoscopy. They are mostly small and do not require treatment.
## Diagnostics
Comparable to all rare tumors, initial diagnoses are frequently incorrect. The most important imaging procedures are computed tomography and MRI. However, neither CT scan nor MRI can reliably differentiate between chondroma and chondrosarcoma. High-resolution computed tomography is recommended. It displays a tissue mass at the laryngeal framework originating from the cartilage. The most frequent radiological differential diagnoses are carcinomas with cartilaginous infiltration, laryngeal manifestations of chondromatosis or posttraumatic calcifications [bib_ref] Chondroma of the larynx--case report and differential diagnosis in CT, Gulden [/bib_ref]. In the context of diagnostic microlaryngoscopy, tumors of the laryngeal framework may be diagnosed with high reliability [bib_ref] Chondroma of laryngeal cartilage mimicking thyroid tumor: A case report, Tan [/bib_ref]. Also here, a differentiation between chondroma and chondrosarcoma is not possible. Finally, histological examination is essential. It reveals hyaline cartilage formations similar to normal cartilage. The characteristics of chondromas in comparison to chondrosarcomas are a low percentage of cells, regular nuclei size and morphology as well as missing mitotic . Chondromas and chondrosarcomas may also develop synchronously or metachronously [bib_ref] Chondroma and chondrosarcoma of the larynx, Baatenburg De Jong [/bib_ref].
## Therapy
Chondromas are surgically resected, laryngeal preservation is an ultimate goal. In this context, generous open surgical therapies are mentioned, partly with temporary tracheostomy and multi-step reconstructions with rib cartilage [bib_ref] Chondroma of laryngeal cartilage mimicking thyroid tumor: A case report, Tan [/bib_ref]. Small glottic findings can also be removed by transoral microsurgery [bib_ref] Bilateral Vocal Fold Chondromas, Lai [/bib_ref]. There is no evi-dence in the literature regarding the difference between initial conservative function-preserving partial resections of chondromas with follow-up over many years or a primary radical surgery [bib_ref] Laryngeal chondroma: a benign process with long-term clinical implications, Chiu [/bib_ref]. In addition, reports about laryngectomy in cases of extremely large tumors are found [bib_ref] Chondroma and chondrosarcoma of the larynx: apropos of 4 cases, Verhulst [/bib_ref].
## Prognosis
An early diagnosis is desirable in order to perform possible functionpreserving therapy. The tumors grow very slowly and patients need to undergo follow-up for a long time. Chondrosarcomas always have to be taken into consideration as differential diagnosis [bib_ref] Laryngeal chondroma: a rare diagnosis in this localization, Tastekin [/bib_ref].
Note: The relationship of the tumor to the laryngeal framework is a typical characteristic as well as the slow growth and the hard consistency. A differentiation of chondrosarcoma is only possible by means of histopathology.
## Lipomas
In the literature, currently 49 articles are found about laryngeal lipomas with a total of about 130 described cases.
## Clinical manifestation
Generally, lipomas belong to the most frequent benign tumors in humans. They make up 4-5 % of all benign tumors. However, they are only rarely found in the larynx. About 0.6 % of benign laryngeal tumors are lipomas [bib_ref] Large Laryngeal Lipoma with Extra Laryngeal Component Mimics Mixed Form Laryngocele: A..., Azandaryani [/bib_ref] [bib_ref] Large lipoma of the larynx: a case report, Khorsandi Ashtiani [/bib_ref]. Even more rarely, they are found in the trachea. They develop mostly after the 40 th year of life, males and females are equally affected. Lipomas may occur at all locations of the larynx, including the vocal folds [bib_ref] Lipoma of the Larynx and Other Differential Diagnoses, Okromelidze [/bib_ref] , the pre-epiglottic space [bib_ref] Lipoma of the pre-epiglottic space: A common pathology in an uncommon place, Costa [/bib_ref] or combined extra-and intralaryngeal [bib_ref] Large Laryngeal Lipoma with Extra Laryngeal Component Mimics Mixed Form Laryngocele: A..., Azandaryani [/bib_ref]. Mostly they present as isolated findings, but they may also occur in the context of generalized lipomatosis [bib_ref] Laryngeal lipoma associated with diffuse lipomatosis: case report and literature review, Durr [/bib_ref]. The symptoms are unspecific and depend on the location and size of the findings. Clinically, the tumors impose as soft, yellowish, partly lobular, partly polypous masses. In many cases, they reach into the laryngeal lumen or protrude the mucosa. Also life-threatening obstructions may develop. There are rare subtypes of lipomas of the larynx, e. g. spindle cell lipomas and pleomorphic lipomas. In this regard, eight cases have been described in the literature [bib_ref] Laryngeal Spindle Cell/ Pleomorphic Lipoma: A Case Report. An In-Depth Review of..., Burkes [/bib_ref] [bib_ref] Acute dyspnea caused by a giant spindle cell lipoma of the larynx, Wolf-Magele [/bib_ref] [bib_ref] Lipoma and liposarcoma of the larynx: case reports and literature review, Kodiyan [/bib_ref] [bib_ref] Spindle cell lipoma of the larynx, Mottola [/bib_ref] [bib_ref] Spindle cell lipoma within the larynx: a case report with correlated light..., Nonaka [/bib_ref].
## Diagnostics
In general, lipomas can be well displayed by ultrasound, however, the visibility is limited within the larynx due to the cartilage framework. MRI may clearly identify lipomatous tumors, however, a differentiation between lipomas and liposarcomas or lipoblastomas is nearly impossible [bib_ref] Lipoma of the Larynx and Other Differential Diagnoses, Okromelidze [/bib_ref]. Histologically, mature fatty tissue is found with delicate vessels and narrow connective tissue septa. CD34 positive, spindle-shaped cells with abundant fibrous and myxoid stroma, penetrated by abundant fatty tissue are typical for spindle cell lipomas [bib_ref] Laryngeal Spindle Cell/ Pleomorphic Lipoma: A Case Report. An In-Depth Review of..., Burkes [/bib_ref].
## Therapy
Surgical therapy is the only option for removal of the tumor. Open approaches are indicated in large or poorly accessible findings, otherwise the transoral microsurgery is preferred [bib_ref] Acute dyspnea caused by a giant spindle cell lipoma of the larynx, Wolf-Magele [/bib_ref] or a combination of both [bib_ref] Lipoma of the larynx: a case report, Vincentiis [/bib_ref]. More recent publications describe a removal via transoral robotic surgery [bib_ref] Lipoma and liposarcoma of the larynx: case reports and literature review, Kodiyan [/bib_ref]. Small incidental findings that have been discovered in the context of imaging can primarily be controlled radiologically after histological assessment (differentiation of liposarcoma) if no functional impairment is observed.
## Prognosis
In cases of incomplete removal, the residues may lead to further growth and new complaints. After complete resection, recurrences appear rarely. Follow-up over several years is recommended.
Note: The differentiation between lipoma and liposarcoma is only possible by means of histology and cannot be made by imaging. Complete extirpation should be aimed at.
## Rhabdomyomas
In the literature, 34 reports about a total of 40 cases are found.
## Clinical manifestation
Usually, rhabdomyomas are tumors of the cardiac muscles. Apart from the cardiac locations, the head and neck region is most frequently affected, especially the neck and face muscles as well as the muscles of the oropharynx and the tongue [bib_ref] Extracardiac juvenile rhabdomyoma of the larynx: a rare pathological finding, Sharma [/bib_ref]. Laryngeal manifestations have been described, although they are extremely rare [bib_ref] Adult-Type Rhabdomyoma of the Larynx: Clinicopathologic Study of an Uncommon Tumor in..., Altissimi [/bib_ref]. Male patients develop more frequently rhabdomyomas than females. The tumors usually appear as isolated findings and only rarely in the context of syndromes [bib_ref] Adult-Type Rhabdomyoma of the Larynx in Birt-Hogg-Dube Syndrome: Evidence for a Real..., Balakumar [/bib_ref] [bib_ref] Unusual presentation of extracardiac fetal rhabdomyoma of the larynx in a pediatric..., Elawabdeh [/bib_ref]. Two different types are differentiated: the adult type, which appears more frequently in the larynx [bib_ref] Rhabdomyoma of the larynx: a review of the literature with a summary..., Johansen [/bib_ref] as well as the fetal ones [bib_ref] Unusual presentation of extracardiac fetal rhabdomyoma of the larynx in a pediatric..., Elawabdeh [/bib_ref]. Hereby, the difference is made between cellular and myxoid subtypes. The symptoms are unspecific and depend on the size and location of the tumors. Usually, laryngeal rhabdomyomas originate from the M. vocalis but they may also affect the outer laryngeal muscles. This entity has not been described in the trachea up to now.
## Diagnostics
In CT scans and MRI, rhabdomyomas impose as contrast enhancing sharply defined tumors. There is no clear differential diagnostic criterion. It is necessary to perform biopsy that shows polygonal vacuolized cells with fine-granular eosinophilic cytoplasm [fig_ref] ▶Figure 6: Rhabdomyoma [/fig_ref]. The evidence is given by immunohistology with expression of desmin, myoglobin, muscle-specific actin (MSA), and negativity of cytokeratin, S100, CD68R, chromogranin A, and synaptophysin [bib_ref] Rhabdomyoma of the head and neck: morphology and differential diagnosis, Hansen [/bib_ref].
## Therapy
The therapy of choice consists of complete surgical resection. Endoscopic as well as open approaches have been described. In most cases, phonosurgical reconstructions are necessary.
## Prognosis
The prognosis is very good, only one report is available about a recurrence after complete surgical resection [bib_ref] Rhabdomyoma of the larynx recurring after 12 years: immunocytochemistry and differential diagnosis, Hamper [/bib_ref].
Note: Laryngeal rhabdomyomas mostly originate from the M. vocalis.
## Paragangliomas
In the literature, 67 publications with 77 described cases of paragangliomas of the larynx are found.
## Clinical manifestation
In most cases, paragangliomas are benign neuroendocrine tumors of the extra-adrenal chromaffin tissue of the autonomous neuro-nal system. Laryngeal paraganglioma cells are mainly found in the supraglottic (90 %) or subglottic area and are associated with the superior and inferior laryngeal nerves. They make up 0.6 % of all laryngeal tumors [bib_ref] Multiple Cutaneous Metastases on (99m)TC-HYNIC-TOC Scan in a Rare Case of Malignant..., Sankar [/bib_ref]. The functional complaints depend on the size and the location. Mostly, the lesions are monofocal, rarely multi-local development is observed [bib_ref] Laryngeal paraganglioma in a patient with multiple head and neck paragangliomas, Rubin [/bib_ref]. Paragangliomas may also appear in the trachea. Often the tumors are painful.
## Diagnostics
Generally, biopsy is performed to classify the tumor. Hereby, smaller tumors are already completely removed. More often, the findings are large so that definitive resection has to be performed. Biopsies may lead to intensive hemorrhage. Histologically, paragangliomas have an alveolar basic structure; they are composed of cell accumulations that are kept together by fiber networks and are located in a well vascularized stroma. Immunohistochemistry confirms the diagnosis by presentation of chromogranin, synaptophysin, and neuron-specific enolase, in the supporting cells S100 protein is found. After biopsy, a comprehensive staging should be performed. Beside MRI and CT scan, also methods of nuclear medicine may be applied due to the frequently expressed somatostatin receptors, e. g. DOTATATE PET which additionally allows exclusion of multilocal paragangliomas [bib_ref] Multiple Cutaneous Metastases on (99m)TC-HYNIC-TOC Scan in a Rare Case of Malignant..., Sankar [/bib_ref]. In the last years, several gene alterations could be identified that are responsible for the genesis of paragangliomas of the head and neck. The predominant signaling pathways concern the enzyme succinate-dehydrogenase (SDH) which consists of the five subunits of SDH-A, SDH-B, SDH-C, SDH-D, and SDH-AF2. Alterations in each of these five subunits lead to SDH deficient tumors, respectively defined as paraganglioma syndrome 1-5 [bib_ref] Paragangliomas of the Head and Neck: An Overview from Diagnosis to Genetics, Williams [/bib_ref]. Laryngeal manifestations have only been described for the paraganglioma type 1 (SDH-D). According changes are located on the locus 11q23.1 [bib_ref] Paragangliomas of the Head and Neck: An Overview from Diagnosis to Genetics, Williams [/bib_ref] [bib_ref] Head and neck paragangliomas: genetic spectrum and clinical variability in 79 consecutive..., Piccini [/bib_ref]. Within a European registry (Head and Neck Paraganglioma Registry) different mutation subgroups of 121 paraganglioma-patients have been collected and statistically evaluated [bib_ref] Predictors and prevalence of paraganglioma syndrome associated with mutations of the SDHC..., Schiavi [/bib_ref].
## Therapy
The therapy of choice is surgery. Frequently, open approaches are described, e. g. via lateral pharyngotomy or thyrofissure. Naik et al. report about preoperative superselective embolization of the superior thyroid artery with subsequent partial laryngeal resection via lateral pharyngotomy including tracheostomy [bib_ref] Laryngeal paraganglioma: a rare clinical entity managed by supraselective embolization and lateral..., Naik [/bib_ref]. In addi tion, transoral laser surgical procedures are performed [bib_ref] Laser surgical treatment of laryngeal paraganglioma, Sesterhenn [/bib_ref]. In the literature, successful primary radiotherapy is reported [bib_ref] Paraganglioma of the endolarynx: a rare tumor in an uncommon location, Smolarz [/bib_ref]. Pharmacotherapeutic approaches with octreotide (Sandostatin e. g. Kohlpharma GmbH, Merzig, Germany) may be applied for symptom reduction and stabilization of the course, especially in non-resectable lesions with positive initial octreotide scintigraphy [bib_ref] Laryngeal paragangliomas -controversies in diagnosis and management, Papacharalampous [/bib_ref].
## Prognosis
In about 15 % of the cases, recurrences occur. In this context, rarely appearing malignant degeneration must be assumed [bib_ref] Paragangliomas of the head and neck, Pellitteri [/bib_ref]. Long-term follow-up is essential. Note: Paragangliomas may cause pain. A nuclear-medical staging is recommended. The therapy of choice is surgery; however, high risk of tumor hemorrhage must be taken into consideration.
## Schwannomas
In the literature, 82 publications on laryngeal schwannomas are found.
## Clinical manifestation
Schwannomas, also called neurilemmomas (formerly: neurinomas), belong to the group of neurogenic tumors. They originate from Schwann cells. Schwannomas occur frequently in the head and neck region, however, the larynx is very rarely affected. In the literature, more female patients with laryngeal schwannomas are mentioned compared to males. Generally the middle age groups are concerned. Rarely, laryngeal schwannomas are found in children [bib_ref] Laryngeal schwannoma in an 8-year-old boy with inspiratory dyspnea, Rognone [/bib_ref]. Within the larynx, they develop in different locations [bib_ref] Laryngeal schwannoma: a systematic review, Wong [/bib_ref] , e. g. (most frequently) at the aryepiglottic fold and the arytenoid region (▶Fig. 7) [bib_ref] Schwannoma of the larynx, Ebmeyer [/bib_ref] , in the pre-epiglottic space [bib_ref] An unusual cause of high dysphagia: schwannoma of the preepiglottic space, Ayad [/bib_ref] , in the paraglottic space [bib_ref] Voluminous laryngeal schwannoma excision with a mini-invasive external approach: a case report, Tulli [/bib_ref] , at the false vocal cords, the vocal folds, the epiglottis, and in the subglottis. Schwannomas may also appear in the tracheal lumen. The origin are branches of the superior as well as inferior laryngeal nerve. The complaints depend on the size and location: Frequently Schwannomas are incidentally detected in the context of neck imaging.
## Diagnostics
In the preoperative diagnostics, MRI is most important with a hyperintense signal in T2 and contrast enhancement for which T1 scan is chosen due to the poor native signal. Careful phoniatric functional diagnostics are required unless the mobility can be assessed based on the tumor mass. Finally, only histological and immunohistochemical examinations can confirm the suspicion. Histologically, vertically and plait-like arranged fusiform cells with palisade structure of the nuclei are observed. Also in the larynx, two architectural types of Antoni A (compact with vertical palisades) and
Antoni B (loosely myxomatous) are differentiated. In schwannomas as well as neurofibromas, the marker constellation of CD34 (-), SMA (-), S-100 ( + ) is typical . Hereby, the immunohistological identification of S-100 protein refers to a neurogenic origin. However, this is not absolutely specific. Schwannomas are frequently detected as incidental findings in the context of neck imaging. In cases of absent functional disorders, first wait-and-scan strategy may be pursued after histological confirmation. Otherwise, therapy of choice is the surgical removal mostly via transoral microsurgerybut also via open approaches, partly with tracheostomy [bib_ref] An unusual cause of high dysphagia: schwannoma of the preepiglottic space, Ayad [/bib_ref]. In the recent literature, reports on transoral robotic removal with the DaVinci system are found [bib_ref] Lipoma and liposarcoma of the larynx: case reports and literature review, Kodiyan [/bib_ref] [bib_ref] Transoral robotic approach for laryngeal schwannoma, Millas [/bib_ref].
## Prognosis
In the literature, 5 cases with recurrences have been described. In general, postoperative sensible deficits or (depending on the origin of the schwannoma) also motor deficits are observed. Therefore, follow-up by phoniatricians is recommended. Note: Schwannomas are mostly found at the aryepiglottic folds, however, they may occur in all areas of the larynx.
## Neurofibromas
In the literature, 51 publications on neurofibromas of the larynx are found, some of them published as case series. 63 pediatric cases have been described up to now.
## Clinical manifestation
Neurofibromas are neurogenic tumors with a predominantly fibrous part that originate from the endoneurium. Histologically the difference is made between simple, cell-rich, and plexiform neurofibromas. Manifestations in the larynx mostly occur in the context of generalized neurofibromatosis in the sense of von Recklinghausen's disease (neurofibromatosis [NF] I) [bib_ref] Glottic neurofibroma in an elderly patient: a case report, Liu [/bib_ref] [bib_ref] Management of advanced laryngeal and hypopharyngeal plexiform neurofibroma in adults, Dave [/bib_ref]. Especially hereby, the disease mainly manifests in (small) children. However, also isolated laryngeal manifestations have been described. Neurofibromas make up 0.05 % of all benign laryngeal tumors [bib_ref] Massive Plexiform Neurofibroma of the Neck and Larynx, Mobashir [/bib_ref]. The clinical appearance depends on the size and the location of the findings. Comparable to schwannomas, neurofibromas are mostly found in the supraglottis and rarely in the subglottis or in the trachea [bib_ref] Solitary subglottic neurofibroma: a report of an unusual manifestation, Gstottner [/bib_ref].
## Diagnostics
MRI is the imaging procedure of choice. Neurofibromas are hyperintense in T2 with central intermediary areas, in T1 weighting they have a weak to moderate signal. A clear differentiation with schwannomas is not always possible. However, MRI may frequently differentiate between plexiform tumors without clear borders and non-plexiform well-defined subtypes [bib_ref] Growth type of plexiform neurofibromas in NF1 determined on magnetic resonance images, Friedrich [/bib_ref]. The immunohistological marker combination of neurofibromas is similar to the one of schwannomas.
## Therapy
Similar to schwannomas, also here the wait-and-scan strategy is an option that should be interdisciplinarily discussed. Modern therapeutic procedures focus on endoscopic transoral resection or partial resection [bib_ref] Pediatric laryngeal neurofibroma: case report and review of the literature, Chinn [/bib_ref]. Especially plexiform neurofibromas are particularly poorly delimited and thus difficult to resect [bib_ref] Management of advanced laryngeal and hypopharyngeal plexiform neurofibroma in adults, Dave [/bib_ref]. In the literature, there is no case where recurrence has not occurred, even after total laryngectomy. Therefore, the tendency is clearly to leave radical resections and to focus on function preserving (partial) re-sections with clinical and imaging controls. Tracheostomy should be avoided as long as possible [bib_ref] The biology and management of laryngeal neurofibroma, Rahbar [/bib_ref].
## Prognosis
Because of the frequent recurrences even after radical surgeries, conservative surgical procedures are indicated, in cases of doubt also in the sense of partial resections. Therefore, the patients have to undergo long-term follow-up, regular MRI controls are necessary. Furthermore, the possibility of malignant degeneration must be taken into account.
Note: Laryngeal neurofibromas usually appear in the context of generalized neurofibromatosis and only rarely as isolated disease.
## Granular cell tumors
In the literature, 76 publications are found on this topic with about 100 cases.
## Clinical manifestation
Granular cell tumors belong to the tumors of neuroectodermal origin, they originate from Schwann cells. The oral cavity and in particular the tongue are most frequently affected, about 10 % of the head and neck manifestations concern the larynx [bib_ref] Granular-Cell Tumors of the Larynx, Nolte [/bib_ref]. Children as well as adults develop granular cell tumors, the gender distribution is balanced [bib_ref] Complicated pediatric subglottic granular cell tumor with extensive intraluminal and extraluminal invasion, Gonik [/bib_ref]. The predilection site within the larynx is the posterior part of the vocal folds (▶Fig. 9) [bib_ref] Granular Cell Tumors of the Larynx: A Clinicopathologic Study of Five Patients, Valldeperes [/bib_ref]. The tumors impose as pale, submucously growing mass on the vocal folds, the initial symptom is dysphonia [bib_ref] Refractory Therapy Hoarseness in Granular Cell Tumor of the Vocal Cord, Reiter [/bib_ref]. In 5 % of the cases, several laryngeal manifestations are observed [bib_ref] Novel surgical management of a laryngeal granular cell tumour, Mcintyre [/bib_ref]. Granular cell tumors also appear in the trachea, hereby the female gender is more frequently affected. Typically, these tumors affect the cervical parts of the trachea.
## Diagnostics
The diagnosis can exclusively be made by histological and immunohistological examinations because the clinical picture is unspecific and the tumor rather reminds of a cyst. Submucously growing, polygonal, granular cells are typical, they react positive on S-100, CD 57, CD 68, and SOX 10 [bib_ref] Granular Cell Tumour of the Larynx, Sproat [/bib_ref]. The S-100 positivity is typical for neurogenic tumors (▶Fig. 10).
## Therapy
Usually, the tumors are excised transorally by laser surgery [bib_ref] Novel surgical management of a laryngeal granular cell tumour, Mcintyre [/bib_ref] [bib_ref] Granular cell tumor on larynx, Park [/bib_ref]. In rare subglottic affection, open surgical procedures should be preferred [bib_ref] Complicated pediatric subglottic granular cell tumor with extensive intraluminal and extraluminal invasion, Gonik [/bib_ref].
## Prognosis
The prognosis after complete excision is good with a recurrence rate of 2-3 %. For granular cell tumors, malignant degeneration rates of 1-2 % are described. However, in the literature exists only one report about laryngeal malignant transformation [bib_ref] Pediatric Laryngeal Granular-Cell Tumor, Conley [/bib_ref].
Note: Granular cell tumors look like cysts, the suspected diagnosis could be made due to the uncommon location in the posterior third of the vocal folds.
## Hemangiomas
In the literature, more than 300 publications are found on hemangiomas of the larynx.
## Clinical manifestation
Laryngeal hemangiomas appear as cavernous and capillary lesions. In adults, mostly cavernous, in newborns and children capillary heman-giomas are observed. Pediatric hemangiomas are significantly more frequent and classically located in the subglottis (▶Fig. 11). Girls are more frequently affected than boys. Hemangiomas in adults occur more rarely and are located mainly in the supraglottis [bib_ref] Subglottic infantile haemangioma: A rare but important consideration in young infants presenting..., Ting [/bib_ref]. Macro-▶Fig. 10 Granular cell tumor. a Microscopic examination shows a tumor that has grown in large accumulations with large tumor cells that are not sharply defined and have a granular cytoplasm. The nuclei are mostly centrally located with homogenous, slightly hyperchromatic and enlarged vesicular nuclei. In immunohistochemistry, the tumor cells react positively on b S100 and c S0X10 as indication for neuroectodermal origin. scopically, hemangiomas appear as convex blue-reddish swellings with broad base that are covered by mucosa, sometimes also as pedicled polyps with smooth or rough surface. In dependence of the tumor location, the patients become symptomatic in cases of subglottic findings, predominantly due to inspiratory stridor, cough, and rapid oxygen desaturation. While pediatric hemangiomas grow within the first months of life, involution is observed around the first birthday [bib_ref] Management of Infantile Hemangiomas of the Airway, Darrow [/bib_ref]. Hemangiomas may also appear in the trachea.
## Diagnostics
Beside endoscopic examination, imaging plays a major role. Typical findings in MRI are the asymmetric stenosis of the subglottic airways in the coronary imaging. MRI is the imaging procedure of choice. Hemangiomas impose as solid tumors of moderate intensity in T1w as well as hyperintense in T2 weighting. Flow-voids are typical for shunt formations [bib_ref] Vascular anomalies, Mulliken [/bib_ref]. Histological diagnostics are not always easy, the glucose carrier protein 1 may be used as immunohistological marker for infantile hemangiomas.
## Therapy
In the literature, different therapy options are discussed for subglottic hemangiomas. Most decisive is the extent of the subglottic lumen stenosis. In cases of minor stenosis, a conservative procedure (watch-and-wait) is suggested, regular follow-up examinations are necessary. In cases of moderate lumen stenosis up to 50 %, corticosteroids are recommended. If the stenosis is higher up to 70 %, laser treatment is the therapy of choice. In the literature cutting and angiolytic laser modalities are described (CO 2 , KTP, and Nd:YAG laser) [bib_ref] Management of Infantile Hemangiomas of the Airway, Darrow [/bib_ref]. Tracheostomy is recommended as emergency measure in cases of large findings or multi-local lesions. Open surgical excisions are reserved to extended subglottic findings with lumen stenosis of more than 70 %. Reconstruction (e. g. with rib cartilage) should be included in the planning. There are several pharmacotherapeutic approaches (e. g. interferon or vincristine). The strategy of radiotherapy was abandoned several decades ago, also cryotherapy is no longer recommended. As acute measure, the treatment with propranolol for several months is suitable. Application of beta blockers seems to be effective mostly in children and is often used as primary therapy [bib_ref] Efficacy and rebound rates in propranolol-treated subglottic hemangioma: A literature review, Schwartz [/bib_ref] [bib_ref] The biology and management of subglottic hemangioma: Past, present, future, Rahbar [/bib_ref].
## Note: predilection site for pediatric laryngeal hemangiomas is the subglottis. by means of mri the diagnosis can often be verified. the decision
for therapy depends on the extent of the subglottic lumen stenosis.
## Rare malignant tumors
The current incidence, prevalence, and mortality of laryngeal cancer are estimated to be 2.76 cases/year per 100,000 population, 14.33 cases per 100,000 population, and 1.66 deaths/year per 100,000 population, respectively [bib_ref] Updates on larynx cancer epidemiology, Nocini [/bib_ref]. Squamous cell carcinomas are clearly the most frequent entity, with only 5 % of cases being rare epithelial and non-epithelial laryngeal malignancies. In the following, some entities will be exemplarily highlighted in more detail.
## Carcinomas
## Laryngeal adeno-squamous carcinomas (lasc)
A literature research reveals 31 publications. Up to now, about 100 cases have been described in the upper aerodigestive tract [bib_ref] How phenotype guides management of non-conventional squamous cell carcinomas of the larynx?, Lopez [/bib_ref].
## Clinical manifestation
Adeno-squamous carcinoma is an extremely rare and highly aggressive malignancy. The larynx is the most frequent manifestation site [bib_ref] Immunohistochemical and HPV-related features of laryngeal adenosquamous carcinoma, Lovato [/bib_ref]. Risk factors include nicotine and alcohol consumption. An association with high-risk types of human papillomavirus has been discussed, however, it seems to be improbable [bib_ref] World Health Organization classification of tumours: pathology and genetics of head and..., Thompson [/bib_ref] [bib_ref] Laryngeal adenosquamous carcinoma: A population-based perspective, Dubal [/bib_ref]. The disease usually occurs in an advanced age, while men are clearly more frequently affected with a ratio of 4:1. The symptoms depend on the exact location and the extent of the carcinoma and tend to be non-specific. Locoregional lymph node metastasis occurs in 75 % of the cases; at the time of initial diagnosis, 25 % of the patients already have distant metastases, mainly in the lung [bib_ref] Adenosquamous carcinoma of the upper aerodigestive tract: a clinicopathologic study of 12..., Keelawat [/bib_ref].
## Diagnostics
In addition to clinical examination and panendoscopy with biopsy, imaging techniques such as CT and MRI are indicated. The histological verification may be difficult, because the carcinoma often grows submucously and may thus be overlooked in early stages. Characteristic histology is the biphasic morphology of the tumor with squamous and cribriform tubulo-glandular differentiation in close proximity. Intraluminal accumulation of mucin is a typical finding as well. Histomorphologically, the differentiation with mucoepidermoid carcinoma is difficult; therefore additional immunohistochemical examinations are necessary. In this context, detection of p63, carcino-embryonic antigen as well as low-and high-molecular cytokeratin (CK) can be performed. CK20 is negative [bib_ref] MUC expression in adenosquamous carcinoma of the head and neck regions of..., Kusafuka [/bib_ref]. Reliable differentiation from mucoepidermoid carcinoma and conventional squamous cell carcinoma is important because LASC has a significantly poorer prognosis. A publication by Yoshimura and colleagues revealed that biopsies from oral ASC were frequently misinterpreted [bib_ref] Clinical characteristics of oral adenosquamous carcinoma: report of a case and an..., Yoshimura [/bib_ref].
## Therapy
There is no clear consensus regarding an adequate therapy for LASC. Depending on the tumor size, it includes often radical surgery with consideration of the lymphatic drainage, followed by adjuvant radiotherapy with or without chemotherapy.
## Prognosis
The disease has a significantly worse prognosis than conventional squamous cell carcinoma, with a 5-year survival rate of only 13-25 %. Causes of death are local recurrences as well as distant metastases. Beside the aggressive growth behavior of the tumor, potential histological misinterpretations with resulting incorrect therapy concepts also contribute to the poor prognosis. An early diagnosis allows the introduction of adequate therapy and is prognostically relevant [bib_ref] Immunohistochemical and HPV-related features of laryngeal adenosquamous carcinoma, Lovato [/bib_ref].
Note: Adeno-squamous carcinomas have a poor prognosis; differentiation from mucoepidermoid carcinomas or squamous cell carcinomas is only possible by immunohistochemistry.
## Laryngeal lymphoepithelial carcinomas (llec)
With 0.2 % of all laryngeal tumors, this carcinoma is extremely rare. Up to now, in the literature, only 40 cases have been described [bib_ref] Lymphoepithelial carcinoma of the larynx, Marioni [/bib_ref] [bib_ref] Epstein-Barr virus-associated lymphoepithelial carcinoma of the larynx, Kermani [/bib_ref].
## Clinical manifestation
Lymphoepithelial carcinomas are Epstein-Barr virus (EBV)-associated tumors that are located mainly in the nasopharynx and have an endemic geographic distribution, particularly in Southeast Asia [bib_ref] Lymphoepithelioma of the larynx, Andryk [/bib_ref] [bib_ref] Lymphoepithelial carcinoma of the larynx and hypopharynx: Study of eight cases with..., Macmillan [/bib_ref]. In comparison to nasopharyngeal manifestation, LLEC is very rare. An EBV association is controversially discussed, in any case it seems to be significantly lower than in the nasopharynx. The carcinoma is frequently located in the supraglottis, rarely subglottic or in the trachea [bib_ref] Lymphoepithelial carcinoma of the larynx, Coskun [/bib_ref]. With a ratio of 3:1, men are more frequently affected than women, preferentially between the 5 th and 7 th decade of life [bib_ref] Lymphoepithelial carcinoma of the larynx and hypopharynx: Study of eight cases with..., Macmillan [/bib_ref]. In contrast to lymphoepithelial nasopharyngeal carcinomas, LLEC less commonly affect Caucasians. The clinical symptoms are non-specific [bib_ref] Lymphoepithelioma-like carcinoma of the larynx associated with an Epstein-Barr viral infection. Otolaryng, Sone [/bib_ref].
## Diagnostics
In addition to clinical examination and panendoscopy including biopsy, also CT and/or MRI are necessary for therapy planning, similar to all other laryngeal tumors. Predominantly submucosal growth makes tumor detection difficult, especially in early stages, as well as adequate specimen collection and to treatment planning. Often cervical lymph node metastases represent the initial symptom so that diagnosis is made by lymph node biopsy, initially suggesting nasopharyngeal carcinoma. An EBV association is investigated by means of in situ hybridization. In this context, the identification of EBV encoded small RNA (EBER) is confirmatory. Important further differential diagnoses are Non-Hodgkin-Lymphomas, undifferentiated carcinomas, and malignant melanomas [bib_ref] Lymphoepithelial carcinoma of the larynx, Marioni [/bib_ref] [bib_ref] The Role of Epstein-Barr-Virus in Lymphoepithelioma-Like Carcinomas, Iezzoni [/bib_ref]. Furthermore, laryngeal metastasis of a nasopharyngeal carcinoma have to be excluded. Therapy LLEC is highly radiosensitive, similar to lymphoepithelial carcinoma of the nasopharynx, making radiotherapy the treatment of choice. It ensures good local tumor control [bib_ref] Lymphoepithelial carcinoma of the larynx, Coskun [/bib_ref] [bib_ref] Lymphoepithelial carcinoma of the larynx, Marioni [/bib_ref]. Due to the frequent presence of extensive locoregional metastases, irradiation of the lymphatic drainage area is always necessary. The role of chemotherapy is not yet fully clarified, however, it is assumed that concomitant chemotherapy may reduce the rate of distant metastases [bib_ref] Lymphoepithelial carcinoma of the larynx and hypopharynx: Study of eight cases with..., Macmillan [/bib_ref] [bib_ref] Lymphoepithelial carcinoma of the larynx, Marioni [/bib_ref] [bib_ref] Lymphoepithelial carcinoma of the larynx, Ibrahimov [/bib_ref].
## Prognosis
About one-third of LLEC patients die as a result of their disease, mostly due to distant metastases [bib_ref] Lymphoepithelial carcinoma of the larynx and hypopharynx: Study of eight cases with..., Macmillan [/bib_ref] [bib_ref] Clinicopathological consultation. Lymphoepithelial carcinoma of the larynx hypopharynx, and trachea, Ferlito [/bib_ref].
Note: The lymphoepithelial laryngeal carcinoma is treated primarily by radiotherapy.
## Adenoid cystic carcinoma of the larynx
In the literature, more than 200 articles have been published on adenoid cystic carcinomas of the larynx. Most articles are reports that describe several cases.
## Clinical manifestation
Adenoid cystic carcinoma (ACC) are the most frequent tumors of the minor salivary glands. The density of minor salivary glands within the laryngeal mucosa is 23-47 glands/cm 2 , approximately 20-40 fold lower than in the oral cavity [bib_ref] Glottic-Subglottic adenoid cystic carcinoma. A case report and review of the literature, Testa [/bib_ref]. This is one reason for the rarity of laryngeal ACC with < 1 % of all laryngeal tumors.
ACC are nearly equally found in both genders with a mean age of about 50 years [bib_ref] The range and demographics of salivary gland tumours diagnosed in a UK..., Jones [/bib_ref] [bib_ref] Laryngeal adenoid cystic carcinoma: A population-based perspective, Dubal [/bib_ref]. The most frequent location is the subglottis (▶Fig. 12) with extension into the trachea [bib_ref] Nonsquamous carcinomas of the larynx, Batsakis [/bib_ref] [bib_ref] Malignant minor salivary gland tumors of the larynx, Ganly [/bib_ref]. ACC of the supraglottis and the level of the vocal folds is very rare. Depending on the location, symptoms like globus sensation, hoarseness, dyspnea, and stridor may occur. The tumors grow very slowly, therefore the symptoms also develop with a corresponding delay. . A clear risk factor for the development of ACC could not be identified up to now. Cervical lymph node metastases accur very rarely, distant metastases, particularly pulmonary metastases, may manifest after many years in analogy to ACC of other locations [bib_ref] Laryngeal adenoid cystic carcinoma: case report, Negro [/bib_ref] [bib_ref] Adenoid cystic carcinoma: a pitfall in aspiration cytology of the thyroid, Idowu [/bib_ref].
## Diagnostics
Cross-sectional imaging diagnostics are necessary for definition of the tumor size and therapy planning. However, it must be taken into consideration that the extent is often underestimated due to the diffuse submucosal tumor growth and frequent presence of perineural sheath infiltration. Confirmation is provided by histological and immunohistological examinations. ACC show numerous tubular and cribriform structures. Perineural sheath infiltration is ubiquitous. Immunohistochemical markers are KIT (CD117) in the inner epithelial cells and p63 as well as SMA in the peripheral myoepithelial cells. Furthermore, staining for MYB and MYB/NFIB is performed [bib_ref] Analysis of MYB expression and MYB-NFIB gene fusions in adenoid cystic carcinoma..., Brill Lb 2nd [/bib_ref]. Significant genetic alteration is a t(6;9) chromosome translocation or rarely a t(8;9) translocation. The resulting fusions involve the MYB or MYBL1 oncogene and the transcription factor gene NFIB (▶Fig. 13).
## Therapy and prognosis
Due to the relatively low radiosensitivity of these tumors, surgery must be considered as the therapy of choice. Because of the above-mentioned growth pattern, a radical procedure possibly in the sense of total laryngectomy is required, although even radical surgery often results in R1 situations with regard to perineural sheath infiltration . Despite ACC showing relative radioresistance, some trials could reveal an improved local and locoregional tumor control by means of adjuvant radiotherapy in R1 situations [bib_ref] Combined treatment of adenoid cystic carcinoma of the salivary glands, Avery [/bib_ref] [bib_ref] Adenoid cystic carcinoma of the head and neck: predictors of morbidity and..., Fordice [/bib_ref] [bib_ref] The influence of positive margins and nerve invasion in adenoid cystic carcinoma..., Garden [/bib_ref]. Chen and colleagues reported that local tumor control in patients with ACC in the head and neck region at 5 and 10 years with adjuvant therapy was 92 and 84 %, respectively, and without radiotherapy 80 and 61 %, respectively [bib_ref] Adenoid cystic carcinoma of the head and neck treated by surgery with..., Chen [/bib_ref]. In addition to IMRT (e. g. 50 Gy), heavy ion radiation may improve the prognosis [bib_ref] Intensity Modulated Radiotherapy (IMRT) + Carbon Ion Boost for Adenoid Cystic Carcinoma..., Lang [/bib_ref].
## Prognosis
Factors determining the prognosis are the resection status, tumor size, and the perineural sheath infiltration [bib_ref] Adenoid cystic carcinoma of the head and neck--a 20 years experience, Kokemueller [/bib_ref]. The 5-year survival rate of patients with laryngeal ACC ranges from 43 to 75 % [bib_ref] Glandular carcinoma of the larynx: the UCLA experience, Alavi [/bib_ref] [bib_ref] Malignant sialogenic tumours of the larynx, Mahlstedt [/bib_ref] [bib_ref] Cancer of the Minor Salivary-Glands of the Larynx, Cohen [/bib_ref]. The high difference can be explained by the low number of cases. In general, the local tumor control rate is good due to radical surgery with adjuvant radiation, if needed, the patients die almost exclusively due to distant metastases.
Note: Surgical approaches are the therapy of choice for adenoid cystic carcinomas of the larynx. The prognosis is determined by late distant metastasis.
## Mucoepidermoid carcinomas (mec)
There are only about 100 case reports on this tumor entity.
## Clinical manifestation
The symptoms of MEC are similar to those of other tumors in the larynx. The tumors are found supraglottically as well as subglottically [bib_ref] Mucoepidermoid carcinoma of the larynx: report of three cases, Prgomet [/bib_ref]. There is one report of a 12-year-old child with the disease, but typically adults are affected [bib_ref] Mucoepidermoid carcinoma of the larynx with transglottic involvement in a child: A..., Bhagat [/bib_ref].
## Diagnostics
In addition to clinical examination and panendoscopy including biopsy, imaging procedures such as CT and/or MRI are indicated. Diagnosis is made by histologic confirmation. The evaluation can be very challenging; there is a risk of misidentification of the tumor with e. g. squamous cell carcinoma, adenocarcinoma, adenosquamous cell carcinoma, or neuroendocrine carcinoma. MEC consist of three cell types: epidermoid cells, mucous secreting cells, and intermediary cells. Histologically, a differentiation between low-grade (lowmalignant), intermediate-grade , and high-grade (highly malignant) types is performed.
## Therapy
The therapeutic procedures are based on the treatment of MEC of the salivary glands. Predominantly, surgery is performed with frequently subsequent adjuvant radiation, if required in combination with chemotherapy.
## Prognosis
The differentiation is one of the most important prognostic factors of MEC. Low-and intermediate-grade tumors have a good prognosis after adequate surgical resection with 10-year survival rates of 90 % and 70 %, respectively. High-grade tumors, however, have a 10year survival rate of only 25 %. In cases of well-differentiated, small, and non-metastasized tumors, a close follow-up concept may be performed instead of adjuvant therapy [bib_ref] Mucoepidermoid Carcinoma of the Larynx -a Clinicopathological Study of 11 Cases with..., Ferlito [/bib_ref] [bib_ref] Recurrent High-Grade Invasive Mucoepidermoid Carcinoma of Larynx: A Case Report and Review..., King [/bib_ref].
Note: The prognosis of MEC of the larynx depends on the grade of differentiation.
## Clinical manifestation
In analogy to NECs of other locations such as the lung, 3 different subtypes are distinguished in the larynx: well-differentiated (G1) NEC-L (synonym: carcinoid), moderately differentiated (G2) NEC-L (synonym: atypical carcinoid), and poorly differentiated (G3) NEC-L, which in turn is subdivided into small-cell (sm) and large cell (lc) types [bib_ref] Update from the 4th Edition of the World Health Organization Classification of..., Gale [/bib_ref]. The G1 NEC-L occurs most rarely, followed by G3 type, whereas the G2 NEC-L appears most frequently. This distinction has therapeutic consequences, as clinical behavior and response to treatment differ significantly between the subtypes. NEC-L concerns mostly the male gender with a sex ratio of 3:1, except of well-differentiated NEC-L in which no sex preference has been observed [bib_ref] Update from the 4th Edition of the World Health Organization Classification of..., Gale [/bib_ref] [bib_ref] Clinical recommendations on the treatment of neuroendocrine carcinoma of the larynx: A..., Van Der Laan [/bib_ref]. The disease peak is in the 5 th decade of life and frequently a history of tobacco abuse is found. The most frequent location is the supraglottis. NEC-L causes non-specific symptoms such as hoarseness, dysphagia, or sore throat. In very rare cases, a paraneoplastic syndrome may appear due to tumorinduced hormone overproduction [bib_ref] Paraneoplastic syndromes in patients with laryngeal neuroendocrine carcinomas: clinical manifestations and prognostic..., Ferlito [/bib_ref]. Of particular note is the high tendency of NEC-L to metastasize dermally, which is observed in up to 20 % of the cases, according to the literature [bib_ref] A study of moderately differentiated neuroendocrine carcinomas of the larynx and an..., Chung [/bib_ref] [bib_ref] Laryngeal Neuroendocrine Tumor With Elevated Serum Calcitonin: A Diagnostic and Therapeutic Challenge, Feola [/bib_ref] [bib_ref] Minimally invasive surgery for recurrent neuroendocrine carcinoma of the supraglottic larynx, Machens [/bib_ref]. Therefore, a careful inspection of the skin is urgently required in the context of staging.
## Diagnostics
In addition to the clinical examination including panendoscopy with biopsy, different diagnostic procedures are applied. Imaging techniques include ultrasound, CT as well as PET-CT (FDG-PET-CT/DO-TATE-PET-CT). Regarding immunohistological diagnostics, different markers are assessed such as cytokeratin, EMA, and at least one positive neuroendocrine marker (e. g. chromogranin A, synaptophysin, neural cell adhesion molecule NCAM (CD56)) [bib_ref] Neuroendocrine neoplasms of the larynx: an overview, Ferlito [/bib_ref]. NEC can secrete various hormones such as serotonin, calcitonin, growth hormones, insulin, gastrin, or glucagon. To evaluate the prognosis of the patient, Ki 67 expression analysis is obligatory.
Medullary thyroid carcinoma should be taken into consideration as differential diagnosis and excluded especially in cases of increased serum calcitonin levels that appear mainly in G2 NEC-L [bib_ref] Laryngeal Neuroendocrine Tumor With Elevated Serum Calcitonin: A Diagnostic and Therapeutic Challenge, Feola [/bib_ref]. Further, NEC-L has to be delineated from laryngeal metastasis of a NEC of other origin (e. g. lung).
## Therapy
Therapeutic decisions should be made primarily interdisciplinarily and subtype-specific. In cases of G1 NEC-L, local resection in the sense of either open or transoral endoscopic laser surgical partial laryngectomy is recommended. G2 NEC-L are the largest subgroup. Therapy planning is challenging because 30 % of the patients already have distant metastases at the time of first presentation. Radical surgical resection is indicated. Elective neck dissection should be performed, often occult metastases are found. The sensitivity of G2 NEC-L to radiotherapy is questionable. It could be shown that patients who underwent primary radiotherapy had a lower diseasespecific survival compared to patients that have been treated sur-gically. Recurrences can occur even after 5 years of conventional follow-up period, so that follow-up should be extended to 10 years. Patients with G3 sm-and lc-NEC-L develop early distant metastases. In these cases, the treatment is similar to that of neuroendocrine lung carcinomas and consists of a combination of radiotherapy and chemotherapy [bib_ref] Clinical recommendations on the treatment of neuroendocrine carcinoma of the larynx: A..., Van Der Laan [/bib_ref].
## Prognosis
The course of the disease mainly depends on the differentiation grade. The recurrence rate amounts to 35 % for G1 NEC-L and 81 % for lc-NEC-L. Patients with a G2 or G3 lc-NEC-L develop more frequently distant metastases compared to patients with G1 NEC-L [bib_ref] Clinical recommendations on the treatment of neuroendocrine carcinoma of the larynx: A..., Van Der Laan [/bib_ref]. The disease-free survival (DFS) after 5 years of G1 NEC-L amounts to 80 %. In contrast, the 5-year DFS for G2, G3 sm-NEC-L and lc-NEC-L is 52 %, 19 % and 15 %, respectively [bib_ref] Clinical recommendations on the treatment of neuroendocrine carcinoma of the larynx: A..., Van Der Laan [/bib_ref].
Note: There are numerous neuroendocrine carcinomas. This confirms the necessity of interdisciplinary therapy planning. Due to the high proneness of NEC-L to dermal metastasis, a thorough inspection of the skin is essential.
## Malignant non-epithelial tumors
About 4.3 % of the sarcomas appear in the head and neck [bib_ref] Prognostic factors for recurrence and survival in head and neck soft tissue..., Kraus [/bib_ref]. This heterogeneous group of diseases is characterized by a variety of histological and clinical presentations, making therapy planning challenging. In the larynx, sarcomas of the skeleton as well as soft tissue are found. J. Le Vay, B. O'Sullivan and colleagues investigated the prognosis for different types of head and neck sarcomas. They discovered that well-differentiated sarcomas and the subtypes of liposarcoma, fibrosarcoma, and chondrosarcoma are relatively slow-growing tumors with a low incidence of metastases. On the other hand, poorly differentiated and biologically aggressive tumors such as angiosarcoma, osteogenic sarcoma or neurofibrosarcoma show an increased tendency for distant metastasis [bib_ref] An assessment of prognostic factors in soft-tissue sarcoma of the head and..., Vay [/bib_ref] [bib_ref] Management of sarcomas of the head and neck in adults, Pellitteri [/bib_ref].
Chondrosarcomas occur most frequently as sarcomas of the skeletal system, whereas soft tissue sarcomas are extremely rare. In the following chapters, chondrosarcomas and rhabdomyosarcomas will be considered more in detail. Regarding the entities fibrosarcoma, liposarcoma, synovial sarcoma, angiosarcoma, and leiomyosarcoma, the authors refer to the literature on this topic.
## Laryngeal chondrosarcoma
In the literature, more than 300 publications are found on laryngeal chondrosarcomas, an estimation of the number of known cases is difficult.
## Clinical manifestation
Laryngeal chondrosarcoma (LCS) is the most frequently appearing type of sarcoma in the larynx. After squamous cell carcinomas and adenocarcinomas, the chondrosarcoma is the third most common malignancy of the larynx. Chondrosarcomas are more frequent than chrondromas. The clinical symptoms are non-specific and do not vary from other laryngeal neoplasms. LCS metastasize locoregionally with a rate of 2 % [bib_ref] Laryngeal chondrosarcoma: a population-based analysis, Dubal [/bib_ref]. Distant metastases occur in about 8.5 % of the cases, frequently, the lung and the bones are affected [bib_ref] Laryngeal chondrosarcoma--ten years of experience, Oliveira [/bib_ref]. LCS mostly develops from hyaline cartilage of the cricoid [fig_ref] ▶Figure 4 a: Tracheal agenesis in a newborn [/fig_ref] [bib_ref] Chondrosarcoma of the larynx, Buda [/bib_ref].
The histological confirmation in the context of panendoscopy is essential for the diagnosis. Histologically, it is difficult to differentiate the most frequently occurring well-differentiated chrondrosarcomas from chondromas. The number of mitoses is decisive as well as the hypercellularity and nuclei pleomorphisms, and other anomalies of the nuclei (▶Fig. 15). The infiltration of neighboring healthy cartilage or bone excludes chondroma and argues for sarcoma. Additional immunohistochemical examinations are performed to assess the mitotic activity. CT and MRI provide important information on the tumor extension and the adjacent healthy soft tissue.
## Therapy
Surgery is the procedure of choice for treatment of LCS. The merely endoscopic resection has advantages under a functional aspect and due to the lower morbidity. However, regarding the decision in favor of endoscopic tumor resection it should be taken into account that the recurrence rate is very high at 50 % [bib_ref] Cartilaginous tumors of the larynx: endoscopic laser management using YAG/KTP, Merrot [/bib_ref] [bib_ref] Well and intermediate differentiated laryngeal chondrosarcoma: toward conservative surgery?, Damiani [/bib_ref]. In general, open surgical procedures are performed, ideally under microscopic control. Tracheostomies can frequently be avoided. One must keep in mind that the biopsy access in the context of definitive therapy has to be excised as well. This applies particularly for the accesses from outside. Because of the mostly slow growth of well differentiated subtypes, nowadays a radical surgical procedure is not performed. Moreover, organ preservation is in the focus, so that tumors are removed with a narrow safety under preservation of the mucosa and the function, followed by close MRI follow-up examinations. Recurrences can also be treated circumscribed surgical measures. Today, total laryngectomy is considered as ultima ratio.
## Prognosis
In general, the prognosis is good. Dubal and colleagues performed a retrospective analysis of the tumor registry called SEER (United States National Cancer Institute's Surveillance, Epidemiology, and End Results). They recorded 143 cases with a median age of about ▶Fig. 14 Chondrosarcoma. ( * ) low-grade (G1) chondrosarcoma at the right arytenoid and cricoid; flexible laryngoscopy (chip/tip).
▶Fig. 15 Chondrosarcoma. a Chondrosarcomas of the larynx are malignant mesenchymal tumors of the hyaline cartilage. In the overview, a tumor with lobular structure and chrondroid matrix and increased cellularity is seen. The tumor is located under the laryngeal superficial epithelium. b In higher amplification, the nuclear pleomorphism and hyperchromatism of the neoplastic chondrocytes is displayed in contrast to non-neoplastic cartilaginous tissue (right side). There are also cartilaginous lacunae with two nuclei. a b 61 years. The male sex was most commonly affected, accounting for 76 %. Analysis of disease-specific survival showed 96.5 %, 88.6 %, and 84.8 % at 1, 5, and 10 years, respectively.
Note: The treatment of chondrosarcomas of the larynx frequently allows organ preservation. The tumors grow slowly and today surgery is no longer that radical. It must be taken into account that the biopsy access has to be excised in the context of definitive therapy as well.
## Rhabdomyosarcoma
In the literature, about 100 articles are found on rhabdomyosarcomas of the larynx.
## Clinical manifestation
Rhabdomyosarcomas (RMS) are mesenchymal tumors that originate from the striated muscles and belong to the group of soft tissue sarcomas. About one-third of RMS appear in the head and neck while the orbit, nasopharynx, and nose are the most frequently observed manifestation sites. The larynx is very rarely affected. According to the WHO, RMS are classified into three histological groups. They include embryonic RMS, alveolar RMS, and pleomorphic RMS. Embryonic RMS may affect children as well as adults [bib_ref] Rhabdomyosarcoma of the head and neck in children, Hicks [/bib_ref]. Pleomorphic RMS exclusively appears in adults. The age distribution for RMS is as follows: 1 % at the age of < 1 year and 13 % for ≥ 15-year-old. Thus, the majority of the patients ( > 85 %) is between 1 and 15 years old [bib_ref] Intergroup Rhabdomyosarcoma Study: update for pathologists, Qualman [/bib_ref]. RMS affect mainly the glottis and supraglottis, and males areinvolved in more than 90 % of the cases. The symptoms are non-specific. Pleomorphic RMS is characterized by a highly malignant tumor biology.
## Diagnostics
In addition to histological confirmation in the context of panendoscopy, imaging procedures (CT and MRI scan) are considered as standard procedures. The additional immunohistological examination is crucial. The identification of desmin and myogenin (MYF 4) is typical. Furthermore, the markers of MYOD1, fast myosin, myoglobin, MSA, and SMA can be verified (▶Fig. 16). Especially the alveolar subtypes express also non-myogenic markers such as cytokeratin, EMA, CD 56, chromogranin, synaptophysin, CD 20, and CD 99. As a particularity, alveolar RMS have a PAX3-FOX01 gene fusion and their identification may be helpful in the context of differential diagnosis [bib_ref] Detection of the t(2;13) (q35;q14) and PAX3-FKHR fusion in alveolar rhabdomyosarcoma by..., Biegel [/bib_ref].
## Therapy
Patients with RMS should be treated only in specialized centers. Pediatric patients are evaluated by the Cooperative Weichteilsarkom Studiengruppe (CWS; cooperative soft tisse sarcoma study group). Hereby, a guideline-conform treatment recommendation is made. According to Hicks and colleagues, chemotherapy is favored in all embryonic RMS consisting of vincristine and actinomycin D without radiotherapy. In unfavorable histological groups, among them alveolar, undifferentiated, and anaplastic types, cytoxan as well as radiotherapy are applied in addition to vincristine and actinomycin D [bib_ref] Rhabdomyosarcoma of the head and neck in children, Hicks [/bib_ref]. In adults, surgery is the therapy of choice in cases of resectable tumors [bib_ref] Primary rhabdomyosarcoma of the larynx, Maheshwari [/bib_ref].
## Prognosis
Patients with embryonic RMS have a good prognosis. The prognosis for alveolar RMS is poorer than for embryonic RMS. According to Kissane, pleomorphic RMS have the worst prognosis [bib_ref] Rhabdomyosarcoma of the head and neck in children, Hicks [/bib_ref].
▶Fig. 16 Embryonic rhabdomyosarcoma. a The embryonic rhabdomyosarcoma is a malignant soft tissue tumor with morphological and immunophenotypical properties of embryonic skeleton muscles. Histologically predominantly differentiating rhabdomyoblasts develop with strongly eosinophilic and elongated cytoplasm that may be increasingly observed after previous therapy (such as in this case). The nuclei are clearly enlarged and hyperchromatic. b In higher amplification single tumor cells with multiple nuclei (arrow) and sporadic striation. c Immunohistochemically, the tumor cells react positively on desmin and myogenin. Note: Therapy planning of the different types of rhabdomyosarcoma is performed in an interdisciplinary approach. Only for adults, the surgical treatment is in the fore. [bib_ref] Synchronous Airway Lesions and Outcomes in Infants With Severe Laryngomalacia Requiring Supraglottoplasty, Schroeder [/bib_ref] Laryngeal and tracheal manifestations of general diseases
## Rheumatoid arthritis (ra)
RA is an inflammatory chronic autoimmune disease with a prevalence of about 1 % that may lead to severe joint and bone damage. The progressive disease manifests frequently in the larynx, up to 75 % of the patients may develop voice disorders [bib_ref] Manifestation of Rheumatic Diseases in the Larynx, Reiter [/bib_ref] [bib_ref] Prevalence and relative risk of dysphonia in rheumatoid arthritis, Speyer [/bib_ref].
## Clinical manifestation
The primary symptom is dysphonia, which is caused by an affection of the cricoarytenoid joints. Furthermore, mucosal lesions of the vocal folds (e. g. nodules) and neurogenic muscle disorders may lead to a voice changes. Rarely, affected patients also report about pains during speech, chronic cough, or dyspnea in the cases of bilateral vocal fold fixations [bib_ref] Prevalence of subjective voice impairment in rheumatoid arthritis, Fisher [/bib_ref] [bib_ref] Descriptive Epidemiology of Voice Disorders in Rheumatoid Arthritis: Prevalence, Risk Factors, and..., Roy [/bib_ref] [bib_ref] Bilateral Cricoarytenoid Arthritis: A Cause of Recurrent Upper Airway Obstruction in Rheumatoid..., Pradhan [/bib_ref] [bib_ref] Bamboo node: Primary vocal fold lesion as evidence of autoimmune disease, Murano [/bib_ref].
## Diagnostics
In general, the underlying rheumatic disease of the patients is known, the diagnosis is thus made in the overall context. Structural changes can be documented by means of video-laryngostroboscopic examination. In the initial stages of the disease, rednes and swelling in the area of the arytenoid cartilages are typical signs, swollen mucosa of the vocal folds, especially in the posterior parts, develops only in the course of the disease. With progressive arytenoid fixation, the glottic gap becomes increasingly narrow. The differentiation is important between primary symptoms due to the disease and secondary symptoms due to immunosuppressive corticoid therapy. Arthritis-related impairment of the mobility of the laryngeal joints can be verified functionally in the context of microlaryngoscopy [bib_ref] Laryngeal involvement in rheumatoid arthritis, Beirith [/bib_ref]. For differential diagnosis, it must be taken into account that beside RA also other systemic diseases may cause an affection of the cricothyroid and cricoarytenoid joint [bib_ref] Type 1 autoimmune hepatitis revealed by a dysphonia related to cricoarytenoid arthritis, Loukili [/bib_ref] [bib_ref] Cricoarytenoid arthritis in Sjogren's syndrome, Seve [/bib_ref] [bib_ref] Cricoarytenoid arthritis with rheumatoid arthritis and systemic lupus erythematosus, Nanke [/bib_ref]. In high-resolution CT, joint changes may be detected very clearly. Hence, it is an integral component of diagnostics.
## Therapy and prognosis
Mainly the treatment of the underlying disease is performed. Logopedic therapy can be offered as a supportive treatment. Steroids may be applied systemically or intralesionally [bib_ref] Intra-articular steroid injection in acute rheumatoid arthritis of the larynx, Habib [/bib_ref] [bib_ref] Accelerated nodulosis during low dose methotrexate therapy for rheumatoid arthritis. An analysis..., Kerstens [/bib_ref]. In cases of rheumatic vocal fold nodules, microsurgical treatment is recommended after failed conservative therapy.
Note: 75 % of all patients with rheumatoid arthritis develop laryngeal symptoms.
## Primary laryngotracheal amyloidosis
Primary amyloidosis is a progressive systemic disease with deposits of insoluble protein fibers (amyloid) in the tissue. In Germany, primary amyloidosis occurs with an incidence of 6-10 cases per 1 000 000 inhabitants per year, 15 % of the cases are localized, while a small percentage of these cases affecting the larynx (PLA) [bib_ref] Primary Laryngotracheal Amyloidosis With Bilateral Vocal Cord Involvement and Associated Bronchiectasis, Lanks [/bib_ref] [bib_ref] Primary localized laryngeal amyloidosis: report of 3 cases with long-term follow-up and..., Ma [/bib_ref]. In the head and neck, amyloid deposits also occur orally, pharyngeally, and in the paranasal sinuses. Only 400 cases of PLA have been described in the literature. In the context of primary amyloidosis, mostly light chain amyloid is deposited. At this point, it should be mentioned that secondary amyloidosis (on the basis of chronic infectious and non-infectious diseases, lymphatic tumors or long-term dialysis) appears significantly more frequently. Regarding the systemic type, different other amyloid types can be detected beside light chain amyloids.
## Clinical manifestation
The disease progresses slowly. Depending on the location, different symptoms may occur. These include dysphagia, hoarseness, dyspnea, and stridor. Sometimes, the disease may be misdiagnosed as asthma or chronic obstructive bronchitis [bib_ref] Primary tracheobronchial amyloidosis in China: analysis of 64 cases and a review..., Ding [/bib_ref]. Clinically, often singular or multiple yellowish polypoid, submucous lesions are present (▶Fig. 17). Mostly, the single foci do not exceed 1 cm in size. Young adults are most commonly affected. Predilection site is the supraglottis. Ulcerations may also occur focally.
## Diagnosis
Clinical examination including phoniatric assessment, imaging procedures (CT and MRI scan) as well as panendoscopy with biopsy are indicated for diagnosis. The clinical findings are typical. The Congo red staining is an established technique for confirmation of amyloid deposits (▶Fig. 18) [bib_ref] Congo Red and amyloids: history and relationship, Yakupova [/bib_ref]. A comprehensive examination is obligatory to exclude further amyloid deposits in the head and neck region, furthermore the causes of secondary amyloidosis must be investigated. For this purpose, bone marrow aspiration, blood and urine examinations are necessary to exclude a light chain disease, in particular plasmocytoma must be excluded as a cause. The University of Heidelberg runs a National Clinical Amyloidosis Registry.
## Therapy
The different clinical symptoms require patient-specific therapy planning. In most cases, local microlaryngoscopic ablation is sufficient. In some cases, tracheostomy has to be performed to secure the airways [bib_ref] Endoscopic management of laryngo-tracheobronchial amyloidosis: a series of 32 patients, Piazza [/bib_ref].
## Prognosis
Despite sufficient local therapy, the recurrence rate is approximately 50 %. Systemic therapeutics are useless in cases of localized laryngeal manifestations [bib_ref] Primary localized laryngeal amyloidosis: report of 3 cases with long-term follow-up and..., Ma [/bib_ref]. Patients who are not appropriate for surgery may undergo radiotherapy. However, data regarding this issue are sparse [bib_ref] Tracheobronchial amyloidosis: a case report of successful treatment with external beam radiation..., Monroe [/bib_ref]. Systemic secondary amyloidosis has a poor prognosis; due to the multiple organ affection (especially heart, kidneys, and brain), the disease is often fatal. In contrast to systemic amyloidosis, local amyloidosis has a significantly better prognosis. At this point, the amyloidosis support group is mentioned that may be contacted via www.amyloidose-selbshilfe.de.
Note: If amyloid deposits are found in the larynx, secondary amyloidosis with poor prognosis must be excluded
## Pemphigus vulgaris
Pemphigus vulgaris (PV) is a rare bullous autoimmune disease of the skin and the mucosa. It is caused by circulating autoantibodies against cadherins in the desmosomes. In cases of PV, antibodies against desmoglein 3 and desmoglein 1 are found. Desmoglein 3 is expressed particularly strongly in the mucosa [bib_ref] Role of intramolecular epitope spreading in pemphigus vulgaris, Salato [/bib_ref] [bib_ref] Explanations for the clinical and microscopic localization of lesions in pemphigus foliaceus..., Mahoney [/bib_ref]. Especially laryngeal involvement may cause life-threatening airway obstruction.
## Clinical manifestation
PV can manifest in different regions of the head and neck. Lesions of the oral mucosa in the sense of erosions are described in 75-80 % of the cases. The affection of the larynx may lead to symptoms such as stridor, dyspnea, hoarseness, and hemoptysis.
## Diagnosis
ENT examination of the mucosa of the upper aerodigestive tract is essential, skin or mucosal biopsy is mandatoryfor diagnosis. Histologically, a suprabasal cleft can be confirmed in the HE staining. Additionally, autoantibodies are visualized in the direct immunofluorescence. Antibodies against desmoglein 3 circulating in the blood are confirmed by means of ELISA or indirect immunofluorescence. ELISA is highly sensitive and specific [bib_ref] The clinical transition between pemphigus foliaceus and pemphigus vulgaris correlates well with..., Komai [/bib_ref].
## Therapy and prognosis
Therapy has to be performed interdisciplinarily. This includes the systemic application of glucocorticoids and other immunosuppressants. Rituximab, a monoclonal antibody directed against the B cell antigen CD 20, can also be used for the treatment of PV. Schmidt and colleagues could show that partial remission was achieved in 95 % of the patients [bib_ref] Rituximab in Severe Pemphigus, Schmidt [/bib_ref]. Furthermore, promising innovative therapy concepts are described. They include the so-called chimeric autoantibody receptor (CAAR). The underlying principle is that the disease-causing B cells are recognized by the CAAR and eliminated via T cell activation [bib_ref] Antigen-specific B-cell depletion for precision therapy of mucosal pemphigus vulgaris, Lee [/bib_ref] [bib_ref] Reengineering chimeric antigen receptor T cells for targeted therapy of autoimmune disease, Ellebrecht [/bib_ref]. If untreated, the disease leads to death. This is caused by dermal fluid loss, superinfections or other complications due to disturbed skin barrier. In cases of adequate therapy, the long-term survival rate is more than 90 %. There is a support group for patients with pemphigus diseases (www.pemphigus-pemphigoid-selbsthilfe.de).
Note: With adequate therapeutic interventions, pemphigus vulgaris of the larynx has a good prognosis.
## Granulomatosis with polyangiitis (gpa)
GPA, formerly called Wegener's granulomatosis, is an idiopathic vasculitis of the middle and small arteries characterized by necro-▶Fig. 18 Amyloidosis. a Laryngeal mucosa with subepithelial, amorphous, extracellular, eosinophilic deposits that in Congo red staining (see b) have a characteristic red reaction, corresponding to amyloid. tizing granulomatous inflammations of the respiratory system as well as a co-existing glomerulonephritis. In Europe, about 10 cases per 1 000 000 people are estimated. GPA is a severe disease; median survival of an untreated generalized form is 5 months, death occurs mainly due to kidney and lung failure. Modern immunosuppressive therapy concepts have significantly improved the patients' prognosis; currently the median survival rate is 21 years after diagnosis [bib_ref] Granulomatosis With Polyangiitis in Otolaryngologist Practice: A Review of Current Knowledge, Wojciechowska [/bib_ref].
## Clinical manifestation
Manifestation in the larynx or trachea in form of ulcers or subglottic stenosis appears irregularly. Rarely, laryngotracheal lesions are the only manifestations of GPA. Subglottic stenosis is found in 10-20 % of GPA patients and often occurs in children. The origin for subglottic stenosis is a destruction of the surrounding tissue due to vasculitis with subsequent reduced blood flow that is associated with excessive fibrosis [bib_ref] Upper airway manifestations of granulomatosis with polyangiitis, Alam [/bib_ref]. Subglottic stenosis with a reduced lumen of up to 70 % usually remain asymptomatic, while severer findings may lead to life-threatening dyspnea, stridor and rarely cough. Up to now, only one case of supraglottic stenosis has been described in the context of GPA [bib_ref] Supraglottic stenosis in localized Wegener granulomatosis, Belloso [/bib_ref]. Several pharmaceutics may cause acute episodes, e. g. propylthiouracil, methimazole, carbimazole, sulfasalazine, or minocycline [bib_ref] Drug-induced vasculitis: a clinical and pathological review, Radic [/bib_ref].
## Diagnosis
The diagnosis of GPA is based on the criteria of the American College of Rheumatology. Two of the following criteria have to be met: involvement of the paranasal sinuses, formation of nodules, mixed pulmonary infiltrates or caverns in the chest X-ray, hematuria or similar changes in the urinary status as well as the histological confirmation of granulomas in the arterial wall or in the perivascular tissue of an artery or arterioles [bib_ref] The American College of Rheumatology 1990 criteria for the classification of Wegener's..., Leavitt [/bib_ref]. For clarification, biopsy is required. Serology may differentiate several subtypes of so-called "antineutrophil cytoplasm antibodies" (ANCA), the presence of cANCAs is typical. In acute disease situations, the sensitivity and specificity of the ANCA testing are 91 % and 99 %, respectively. In the context of generalized GPA, the ANCAs are increased in 90-95 % of the cases. In cases of localized organ affection in the head and neck area, the levels are positive only in 50-70 % of the patients [bib_ref] Prospective long-term follow-up of patients with localised Wegener's granulomatosis: does it occur..., Holle [/bib_ref]. The diagnosis of laryngeal or tracheal manifestation is made primarily by means of flexible endoscopy as well as high-resolution CT of the neck.
## Therapy and prognosis
In less advanced cases, pharmacological therapy is sufficient, supported by topical inhalative glucocorticoids or optionally circumscribed surgical procedures may be applied (e. g. laser surgery). The most common pharmacotherapeutic approaches include glucocorticoids as well as cyclophosphamide, rituximab, intravenous immunoglobulins, abatacept, methotrexate, or azathioprine [bib_ref] Granulomatosis With Polyangiitis in Otolaryngologist Practice: A Review of Current Knowledge, Wojciechowska [/bib_ref]. Despite adequate pharmacotherapy, up to 80 % of the patients need surgical intervention, this includes translesional injection of corticoids, endoscopic dilatations, radial laser incisions, rarely stent applications, tracheostomy, or open resections. As it is well known, each surgical procedure may lead to an increase of the stenosis, so that initial surgical steps should be kept as minimally invasive as possible [bib_ref] Otorhinolaryngological manifestations in granulomatosis with polyangiitis (Wegener's), Trimarchi [/bib_ref]. Information for affected patients may be retrieved for example via https://gpa-info.org.
Note: Beside cartilaginous damage of the nasal skeleton, subglottic stenosis are regularly occurring manifestations of GPA in the head and neck area.
## Sarcoidosis
In Germany, sarcoidosis occurs with a prevalence of about 50 per 100 000 inhabitants and manifests mainly in the lung and the lymph nodes. An affection of the larynx or trachea is rather rare. An exclusive affection of the head and neck may occasionally be observed [bib_ref] Dysphonia as a Symptom of Sarcoidosis, Smith [/bib_ref].
## Clinical manifestation
Most patients acquire the disease in the 3 rd and 4 th decade of life. In rare cases, also children may be affected. Depending on the location, laryngeal sarcoidosis is characterized by dysphagia, dysphonia, cough, or airway obstruction with dyspnea. In cases of mediastinal lymph node sarcoidosis, also recurrent laryngeal nerve paralysis may be observed [bib_ref] Neurosarcoidosis Presenting Initially as Idiopathic Vocal Cord Paralysis, Wu [/bib_ref]. A suddenly occurring edematous swelling of the laryngeal tissue is typical, which may cause pseudotumor-like changes. Sometimes, granulomas or ulcerations are also found. The epiglottis and the aryepiglottic folds are affected most frequently with up to 80 % of the cases (▶Fig. 19), although al other laryngeal areas may be affected as well [bib_ref] Manifestation of Rheumatic Diseases in the Larynx, Reiter [/bib_ref] [bib_ref] Laryngeal sarcoidosis: Presentation and management in the pediatric population, Strychowsky [/bib_ref].
## Diagnostics
The three-pillar model for diagnosis are: the clinical and radiological presentation, the histological confirmation of non-caseating granulomas, and the exclusion of relevant differential diagnoses. For radiological clarification of sarcoidosis a CT of the thorax is required. In acute sarcoidosis, serology shows an increased blood sedimentation rate with normal CRP and frequently an increased overall IgG, serum calcium, and calcium in the urine. The most important differential diagnosis is tuberculosis [bib_ref] Sarcoidosis of the upper and lower airways, Morgenthau [/bib_ref].
## Therapy
In cases of acute airway obstruction, appropriate airway protection may be required, in extreme situations tracheostomy is necessary.
▶Fig. 19 Laryngeal sarcoidosis. Manifestation of sarcoidosis at the larynx with typical deformity of the epiglottis.
Usually, the course of sarcoidosis is favorable, in more than 50 % of the cases, spontaneous remissions occur after 1-2 years. After 5 years, most cases are spontaneously healed. Hence, therapy completely depends on the severity of the clinical findings. In cases of critical organ affection, corticosteroids are effective, they may even be applied topically. If the larynx is affected, corticosteroids should be applied early to avoid critical airway obstruction. Beside steroids, mitomycin applications are possible as well [bib_ref] Sinonasal and Laryngeal Sarcoidosis-An Uncommon Presentation and Management Challenge, Edriss [/bib_ref]. In Germany, some self-help groups exist for sarcoidosis patients, e. g. Deutsche Sarkoidose-Vereinigung e. V. and Sarkoidose-Netzwerk e. V.
Note: In most cases, sarcoidosis heals spontaneously. The treatment is performed based on the symptoms.
## Relapsing polychondritis
In the literature, about 100 publications are found on the topic of relapsing polychondritis of the larynx and/or trachea. Mostly, small patient series are presented.
## Clinical manifestation
Relapsing polychondritis (RP) is a rare autoimmune disease characterized by recurrent inflammation of hyaline cartilage. The disease has an incidence of 3-4.5:1,000,000, while females are five times more frequently affected than males [bib_ref] Relapsing polychondritis, Pearson [/bib_ref] [bib_ref] Relapsing polychondritis in the Department of Defense population and review of the..., Mathew [/bib_ref]. The incidence of RP increases with age and has its peak in the 4 th decade of life while 5 % of the cases also occur in childhood [bib_ref] Laryngotracheal stenosis requiring emergency tracheostomy as the first manifestation of childhood-relapsing polychondritis, Buscatti [/bib_ref]. The disease is characterized by recurrent inflammation and cartilage destruction, in particular of the ears, nose, and airways [bib_ref] Relapsing polychondritis: prospective study of 23 patients and a review of the..., Mcadam [/bib_ref] [bib_ref] Pulmonary manifestations of relapsing polychondritis, Rafeq [/bib_ref]. The mechanisms of initiation of RP, maintenance of the pathological immune response, and subsequent cartilage destruction remain largely unclear. Several factors seem to be involved, including genetical conspicuities (HLA-DR4), specific antibodies against cartilage structures (type II collagen, matrilin-1), and the modification of the cytokine and chemokine signatures (MCP-1, MIP-1β, and IL-8) [bib_ref] Susceptibility to Relapsing Polychondritis Is Associated with Hla-Dr4, Lang [/bib_ref] [bib_ref] Relapsing polychondritis: Clinical and immunogenetic analysis of 62 patients, Zeuner [/bib_ref] [bib_ref] Antibodies to Type-2 Collagen and Circulating Immune-Complexes in Relapsing Polychondritis, Foidart [/bib_ref] [bib_ref] The occurrence of autoantibodies to matrilin 1 reflects a tissue-specific response to..., Hansson [/bib_ref] [bib_ref] Serum cytokine profiles in relapsing polychondritis suggest monocyte/macrophage activation, Stabler [/bib_ref]. Matrilin-1 is an extracellular matrix protein that is predominantly expressed in the tracheal cartilage [bib_ref] Radioimmunoassay of the 148-Kilodalton Cartilage Protein -Distribution of the Protein among Bovine-Tissues, Paulsson [/bib_ref]. In about 30 % of the cases, additional autoimmune rheumatologic and hematologic pathologies pathologies are observed [bib_ref] Relapsing polychondritis: prospective study of 23 patients and a review of the..., Mcadam [/bib_ref] [bib_ref] Management of laryngotracheobronchial sequelae and complications of relapsing polychondritis, Spraggs [/bib_ref]. Clinically, the duration and severity of the disease may vary significantly. Several organs may be affected, in 50 % of the RP patients, the larynx and the trachea are involved, which can lead to life-threatening laryngotracheal stenosis [bib_ref] Relapsing polychondritis, Trentham [/bib_ref]. A recent study could show that RP is associated with other autoimmune diseases in more than 20 % such as Sjögren's syndrome, autoimmune thyroiditis, systemic lupus erythematosus, antiphospholipid syndrome, or rheumatoid arthritis [bib_ref] Relapsing Polychondritis Can Be Characterized by Three Different Clinical Phenotypes: Analysis of..., Dion [/bib_ref].
## Diagnostics
In 1976, McAdam and colleagues established the first criteria of RP [bib_ref] Relapsing polychondritis: prospective study of 23 patients and a review of the..., Mcadam [/bib_ref]. The diagnosis can be made if three of six of the following criteria are met: 1) bilateral auricular chondritis; 2) non-erosive seronegative inflammatory arthritis; 3) nasal chondritis; 4) ophthalmological inflammations; 5) chondritis of the airways; 6) vestibulocochlear disorders. These criteria have been modified by Damiani and Levine and by Michet and colleagues [bib_ref] Relapsing polychondritis--report of ten cases, Damiani [/bib_ref] [bib_ref] Relapsing polychondritis. Survival and predictive role of early disease manifestations, Michet [/bib_ref]. Therapy RP patients with mild symptoms are usually treated with nonsteroidal anti-inflammatory drugs (NSAR) and, if necessary, predniso-lone. In cases of severe manifestations with damage of the airways, higher doses of prednisolone may be required [bib_ref] Lack of association of HLA-A and B locus antigens with relapsing polychondritis, Luthra [/bib_ref]. If steroid resistance develops during treatment, immunosuppressive drugs such as azathioprine, cyclosporine A, and plasmapheresis may be applied [bib_ref] Cutaneous manifestations of patients with relapsing polychondritis: an association with extracutaneous complications, Shimizu [/bib_ref]. Furthermore, methotrexate is an effective and rather well tolerated drug [bib_ref] Cogan's syndrome: an autoimmune inner ear disease, Greco [/bib_ref]. Dapsone is an antirheumatic agent with antibiotic effect that may be applied as second-line therapy. However, the manifold side effects have to be taken into consideration.
There are some reports about successful therapy with infliximab [bib_ref] Bell's palsy and autoimmunity, Greco [/bib_ref] [bib_ref] Relapsing polychondritis, Kent [/bib_ref].
## Prognosis
The course of the disease is hardly predictable. Recurrent inflammatory reactions of the cartilage lead to permanent destruction of the affected tissue. If this destruction affects the trachea and larynx, serious respiratory problems may occur that can be lethal.
Note: RP is diagnosed based on a symptom score. The disease is treated primarily with medications.
## Rare diseases of the thyroid
The last chapter will discuss rare diseases of the thyroid. Hereby, the selection was made consciously.
## Riedel's thyroiditis
Riedel's thyroiditis is reported in about 100 publications in the literature. These are nearly exclusively single case reports.
## Clinical manifestation
Riedel's thyroiditis is a fibrous thyroiditis. With an incidence of 1 per 100 000 population, it is a rare disease.. The origin of the disease is unknown. There are hints for an autoimmune genesis, however, fibrosing IgG 4 associated disease is discussed as well [bib_ref] Riedel's Thyroiditis: A Clinical Review, Hennessey [/bib_ref]. A indurated and enlarged thyroid is a typical appearance. In cases of affection of the surrounding structures of the neck, frequently obstruction of the upper airways is observed. Furthermore, dysphagia, stridor and partly even recurrent laryngeal nerve paresis are found.
## Diagnostics
The examination reveals a very solid mass in the jugulum that is not displaced relative to the thyroid gland. Usually, a hypothyreodism is seen in the blood tests. Regarding differential diagnosis, a carcinoma may be suspected [bib_ref] Invasive Fibrous Thyroiditis (Riedel Thyroiditis): The Mayo Clinic Experience, Fatourechi [/bib_ref] , however, the clinical appearance is still is typical. Ultrasound shows a hypoechogenic thyroid fibrosis that is sparsely supplied with blood. CT is helpful to assess an extrathyroidal extension and in estimating the extent of fibrosis [bib_ref] Riedel's thyroiditis in multifocal fibrosclerosis: CT and MR imaging findings, Ozgen [/bib_ref]. Finally, the diagnosis can only be made by means of open biopsy. Typically, the fibrous tissue is infiltrated by eosinophils. Due to the rarity of the disease, the diagnosis is often delayed. The four diagnostic criteria include 1) extrathyroidal extension of the inflammatory process; 2) presence of occlusive phlebitis; 3) exclusion of granulomas, giant cells, lymphoid follicles, and 4) exclusion of malignancy [bib_ref] Current concepts on Riedel thyroiditis, Papi [/bib_ref].
## Therapy
The treatment comprises the application of glucocorticoids or tamoxifen. Surgical measures are challenging because the thyroid is difficult to separate from the surrounding healthy tissue. Thus, a circumscribed surgical therapy should only be performed to avoid obstructive symptoms.
## Prognosis
The disease has a good prognosis; main cause of death is severe tracheal compression. The mortality rate is 5 % [bib_ref] Riedel's Thyroiditis: A Clinical Review, Hennessey [/bib_ref].
Note: Riedel's thyroiditis is treated with glucocorticoids, surgical therapy is performed symptom-based, often in the sense of partial resections.
## Allan-herndon-dudley syndrome (mct8 deficiency)
Up to now, 120 articles have been published on the very rare MCT8 deficiency syndrome.
## Clinical manifestation
Thyroid hormones are essential for the development and function of the central nerve system. They are transported into the interior of the cells via transmembraneous cellular carrier proteins. One of them is the monocarboxylate transporter 8 (MCT8). MCT8 is specific for the transportation of thyroid hormones and is coded by the SLC16A2 gene. Gene mutation leads to a rare (1-year incidence of 0.1 per 100 000 people) X-chromosomal disease that is known as Allan-Herndon-Dudley syndrome and concerns almost allways men. It is not a thyroid disease in the proper sense,but it is often assigttned to this group. Phenotypically, the patients have severe neurological symptoms such as paraparesis with hypotonia, bradykinesia, spasticity, moderate to severe mental retardation with missing speech development as well as epilepsy [bib_ref] Expanding the phenotypic spectrum of Allan-Herndon-Dudley syndrome in patients with SLC16A2 mutations, Remerand [/bib_ref].
## Diagnostics
Typical laboratory constellations include an increased triiodothyronine (T3) plasma level (peripheral hyperthyroidism), reduced thyroxin (T4), and a normal or slightly increased TSH level [bib_ref] Expanding the phenotypic spectrum of Allan-Herndon-Dudley syndrome in patients with SLC16A2 mutations, Remerand [/bib_ref] [bib_ref] A novel syndrome combining thyroid and neurological abnormalities is associated with mutations..., Dumitrescu [/bib_ref]. Genetic examination is required for clear diagnosis. Therapy T3 analogues that do not need MCT8 to pass through the cell membrane could be used as therapeutic agents. They comprise e. g. the T3 analogue of 3.5-diiodothyreopropionic acid and 3,5,3'-triiodothyroacetic acid [bib_ref] Cardiac effects of 3,5-diiodothyropropionic acid, a thyroid hormone analog with inotropic selectivity, Pennock [/bib_ref] [bib_ref] Metabolic effects of thyroid hormone derivatives, Moreno [/bib_ref] [bib_ref] A Thyroid Hormone-Based Strategy for Correcting the Biochemical Abnormality in X-Linked Adrenoleukodystrophy, Hartley [/bib_ref] [bib_ref] In Vitro and Mouse Studies Supporting Therapeutic Utility of Triiodothyroacetic Acid in..., Kersseboom [/bib_ref]. Another therapeutic option is the application of pharmacological chaperons that allow a conformation stabilization of mutated proteins. In two studies, Braun and colleagues could show that the application of pharmacological chaperons restore MCT8 in the cell membranes and mediate the T3 transport in certain types of SLC16A2 mutations [bib_ref] Efficient Activation of Pathogenic DeltaPhe501 Mutation in Monocarboxylate Transporter 8 by Chemical..., Braun [/bib_ref] [bib_ref] The Chemical Chaperone Phenylbutyrate Rescues MCT8 Mutations Associated With Milder Phenotypes in..., Braun [/bib_ref].
Note: MCT8 deficiency manifests mainly in form of severe neurological disorders.
## Primary congenital hypothyroidism (ch)
There is a comprehensive data situation of more than 8,000 articles.
## Clinical manifestation
Congenital hyperthyroidism (CH) is classified into primary and secondary types. The origins of primary CH are developmental disorders of the thyroid or a deficient thyroid hormone synthesis. Se-condary CH is cause by disorders of the central regulation in the hypothalamus or the pituitary gland. The prevalence of the disease varies regionally, which is probably due to the different occurrences of iodine. Generally, primary CH is found in 1:4,000 newborns [bib_ref] Congenital hypothyroidism, Rastogi [/bib_ref] [bib_ref] Neonatal screening for congenital hypothyroidism and phenylketonuria in China, Zhan [/bib_ref].
## Diagnostics
The diagnosis is made in the context of the mandatory laboratory testing of newborn examinations (at the U2 examination, 3 rd to 10 th day of life). The disease is characterized by a reduction or even complete missing of the thyroid hormones T3 and T4 as well as an increase of the TSH level. If during the diagnostic procedure low TSH levels are measured in addition to T3 and T4, the rare form of central hypothyroidism may be assumed that appears with an incidence of 1:20 000 newborns.
## Therapy
The therapy of choice is the early application of the thyroid hormone of L-thyroxine.
Note: The primary congenital hypothyroidism is diagnosed in the context of obligatory examinations of newborns.
## Medullary thyroid carcinoma
A comprehensive data situation of 8,200 publications is available.
## Clinical manifestation
The medullary thyroid carcinoma (MTC) develops from the parafollicular cells of the thyroid that secrete calcitonin. It accounts for approximately 5 to 10 % of all thyroid carcinomas and has an incidence of 0.11 per 100 000 people [bib_ref] Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma, Wells [/bib_ref] [bib_ref] Thyroid cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up, Pacini [/bib_ref]. The disease occurs sporadically in 75 % of the cases and in 25 % as part of the autosomal dominantly inherited disorder MEN (multiple endocrine neoplasm) type 2 and type 3. MEN type 2 and type 3 are mainly caused by the germ line mutation in the RET proto-oncogene, which encodes a tyrosine kinase receptor [bib_ref] Single missense mutation in the tyrosine kinase catalytic domain of the RET..., Carlson [/bib_ref] [bib_ref] Germ-line mutations of the RET proto-oncogene in multiple endocrine neoplasia type 2A, Mulligan [/bib_ref]. Because of the autosomal dominant inheritance, relatives of MEN patients should undergo human genetic consultation. However, it must be taken into account that a negative family history does not exclude a genetic disposition, as generations may be skipped or new mutations may occur [bib_ref] Diagnosis and pathologic characteristics of medullary thyroid carcinoma-review of current guidelines, Thomas [/bib_ref] [bib_ref] Guidelines for the management of thyroid cancer, Perros [/bib_ref]. Furthermore, MTC appears clearly earlier in these patients than in sporadic cases [bib_ref] Medullary carcinoma, Tuttle [/bib_ref].
## Diagnostics
The cytological examination by means of fine needle aspiration (FNA) is helpful for the diagnosis of MTC, but it has a variable accuracy. This circumstance is due to the heterogenic cytological appearance of MTC, which may falsely mimic diseases such as follicular neoplasms or sarcomas [bib_ref] Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma, Wells [/bib_ref]. Confirmation of MTC is the immunohistological evidence of calcitonin. Serum calcitonin further serves as sensitive peripheral tumor marker. The calcitonin screening is a highly sensitive test for the early detection of MTC [bib_ref] Predictive value of serum calcitonin levels for preoperative diagnosis of medullary thyroid..., Costante [/bib_ref]. Calcitonin concentrations of 60-100 pg/ml can be a strong indicator for MTC [bib_ref] Guidelines for the management of thyroid cancer, Perros [/bib_ref]. In cases of slightly increased basal calcitonin, a stimulation test with pentagastrin or calcium is recommended in order to precise the indication for surgery. In case of threshold values of more than 100 pg/ml for stimulated calcitonin, there is an indication for thyroidectomy [bib_ref] Guidelines for the management of thyroid cancer, Perros [/bib_ref]. Furthermore, mutation analy-ses are obligatory for all patients with newly diagnosed MTC [bib_ref] Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma, Wells [/bib_ref]. MCT do not exhibit iodine metabolism and are therefore not amenable to radioiodine therapy. Another hallmark is the ability to rapidly metastasize to hematogenous and lymphogenous sites. Therefore, early detection is prognostically crucial. , In patients with family MTC, diagnostics of the adrenal medulla and parathyroids are obligatory. Ganglioneuromas on the tongue also allow the assumption of MEN type 2b [bib_ref] Life saving glance diagnosis in type IIb multiple endocrine neoplasia, Biesinger [/bib_ref]. Therapy Total thyroidectomy is the therapy of choice. The indication is made after histological or biochemical diagnosis confirmation of MTC. In cases of stimulated calcitonin levels greater than 100 pg/ml, the indication of surgery is made, whether or not a tumor could detected on imaging. The indication for central and lateral neck dissection is made generously in sporadic MTC. In hereditary types, the decision for prophylactic neck dissection is made depending on the basal calcitonin level. If distant metastases are found, tyrosine kinase inhibitors such as cabozantinib and vandetanib may prolong the progression-free survival with palliative intention [bib_ref] Vandetanib for the treatment of patients with locally advanced or metastatic hereditary..., Wells [/bib_ref] [bib_ref] The safety of vandetanib for the treatment of thyroid cancer, Tsang [/bib_ref] [bib_ref] Vandetanib (100 mg) in patients with locally advanced or metastatic hereditary medullary..., Robinson [/bib_ref] [bib_ref] Cabozantinib in progressive medullary thyroid cancer, Elisei [/bib_ref] [bib_ref] Efficacy of cabozantinib (Cabo) in medullary thyroid cancer (MTC) patients with RAS..., Sherman [/bib_ref]. Adjuvant radiotherapy should be taken into account in patients with an increased risk of local recurrence or airway obstruction [bib_ref] A primer on the genetics of medullary thyroid cancer, Larouche [/bib_ref]. Prophylactic thyroidectomy in children in cases of confirmed RET mutation is recommended. In high-risk patients, it should be performed already in the first months of life and in risk patients (e. g. mutations in the codon 634) before the age of 5 years [bib_ref] Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma, Wells [/bib_ref]. Calcitonin and CEA (carcino-embryonic antigen) are applied for postoperative follow-up of patients with MTC. Of course, beside MTC also accompanying diseases that are associated with MEN, e. g. pheochromocytoma, hyperparathyroidism or other manifestations, have to be treated interdisciplinarily.
Note: MTC is a complex endocrinological disease. It must always be treated interdisciplinarily.
[fig] ▶Figure 4 a: Tracheal agenesis in a newborn. The epiglottis is developed, the laryngeal lumen is closed. Ventilation of the patient via the esophagus and the esophago-bronchial fistula is performed. b Tracheal agenesis in a newborn. Now with uploaded epiglottis and visibly free esophageal lumen and completely closed laryngeal lumen. a b [/fig]
[fig] ▶Figure 5: Papillomatosis. ( * ) Laryngeal papillomatosis (here: HPV type 6) with pathognomonic vascular loops; flexible laryngoscopy (chip/tip). [/fig]
[fig] ▶Figure 6: Rhabdomyoma. a Histology reveals a tumor with large polygonal long cells with abundant granular eosinophilic cytoplasm which is partly striated and small round nuclei with distinct cell borders. Mitoses only rarely occur or are completely missing. b In higher amplification, a striation of the tumor cells is seen. c Immunohistochemistry reveals a high expression of desmin. d Myogenin reacts positively in single nuclei. [/fig]
[fig] ▶Figure 7: Schwannoma. ( * ) Plexiform laryngeal schwannoma of the left supraglottis; flexible laryngoscopy (chip/tip).▶Fig. 8 Schwannoma. The figures show a plexiform schwannoma. a Schwannomas are nerve sheath tumors that consist entirely or nearly entirely of differentiated neoplastic Schwann cells. In the overview, a characteristic nodular growth is observed that is typical for plexiform schwannomas. b/c In higher amplification, a monotone, cell-rich spindle-cell tumor without atypia, necrosis or increased mitotic activity as indication of a benign tumor is seen in contrast to classic schwannomas, that have a growth pattern changing between Antoni a and a. [/fig]
[fig] ▶Figure 9: Granular cell tumor. ( * ) Granular cell tumor of the posterior third of the right vocal fold; flexible laryngoscopy (chip/tip). [/fig]
[fig] 17: Laryngeal amyloidosis. ( * ) Laryngeal AL amyloidosis of the kappa light chain type left hemilarynx , flexible laryngoscopy (chip/ tip). [/fig]
|
Sequence-programmable covalent bonding of designed DNA assemblies
## The pdf file includes:
Fig. S1. Design diagram of the brick-like (TT motifs 1 to 3) object prepared using caDNAno. . Exemplary negative-stained TEM micrographs of the brick-like object (TT motifs 1 to 4) in different buffers/solvents. [fig_ref] Figure S3: Laser-scanned fluorescent image of a 2 [/fig_ref]. Laser-scanned fluorescent image of a 2.0% agarose gel that was run in an ice-cooled water bath. [fig_ref] Figure S4: Laser-scanned fluorescent image of a 2 [/fig_ref]. Laser-scanned fluorescent image of a 2.0% agarose gel that was run in an ambient temperature water bath. . Design diagram of the brick-like (TT motifs 1 to 4) object prepared using caDNAno. . Cryo-EM data of the brick-like object with TT-motifs (1) -(4) before crosslinking in folding buffer. . Cryo-EM data of the brick-like object with TT-motifs (1) -(4) after crosslinking in folding buffer. . Design diagram of the switch object. [fig_ref] Figure S24: Design diagram of the polymerization brick object prepared using caDNAno [/fig_ref]. Design diagram of the polymerization brick object. Legend for table S1 Legends for movies S1 to S6 Reference (70) Other Supplementary Material for this manuscript includes the following:
(available at advances.sciencemag.org/cgi/content/full/4/8/eaau1157/DC1) for the nonirradiated (uncross-linked) sample of the brick-like variant with thymines at all staple termini, with TT motifs at all crossover sites and additional 5-T loops in folding buffer and in the presence of 5 mM MgCl 2 . Movie S6 (.mov format). Corresponding movie to fig. S20 for the irradiated (cross-linked) sample of the brick-like variant with thymines at all staple termini, with TT motifs at all crossover sites and additional 5-T loops in folding buffer and in the presence of 5 mM MgCl 2 .
Fig. S1. Design diagram of the brick-like (TT motifs 1 to 3) object prepared using caDNAno. The object features 1-thymine-long overhangs at all staple termini. In addition, TT motifs were inserted at all crossover sites per strand. Interfaces were passivated with poly-thymine overhangs. Inset upper right: Cross-section of the object designed in honeycomb lattice.
## Fig. s2
. Exemplary negative-stained TEM micrographs of the brick-like object (TT motifs 1 to 4) in different buffers/solvents. Images were high-pass filtered to reduce staining gradients. Insets: corresponding average 2D particle micrographs. Scale bars: 20 nm
## Fig. s8. exemplary negative-stained tem micrographs of the pointer object in different buffers/solvents.
Images were high-pass filtered to reduce staining gradients. Insets: corresponding average 2D particle micrographs. Scale bars: 20 nm. 0% agarose gels stained with ethidium bromide. Irradiated (135 minutes at 310 nm) and non-irradiated samples were either incubated for 30 minutes at different temperatures or incubated for 3 hours at room temperature in double-distilled water containing successively lower concentrations of monovalent sodium chloride, respectively. p: pocket; u: unfolded species; f: folded species; c: crosslinked staple strands; s: uncrosslinked staple strands; L: 1kB ladder; NI and I: nonirradiated and irradiated reference samples in folding buffer with 5 mM MgCl 2 , respectively. The images of the gels were globally auto-leveled. (B) and (C) as in (A) but with the brick-like DNA origami object featuring additional thymidines at all strand termini, at all strand crossover positions, and 5T-loops and the pointer object featuring additional thymidines at all strand termini and at all strand crossover positions, respectively. Movie S1. Tomogram obtained from cryo-EM of the brick-like variant with thymines at all staple termini and with TT motifs at all crossover sites.
## Movie
[fig] Fig: S21. Cryo-EM data of the brick-like object with TT-motifs (1) -(4) before crosslinking in folding buffer. . S22. Cryo-EM data of the brick-like object with TT-motifs (1) -(4) after crosslinking in folding buffer. . S23. Design diagram of the switch object. . S24. Design diagram of the polymerization brick object. Legend for table S1 Legends for movies S1 to S6 Reference (70) [/fig]
[table] Table S1: Microsoft Excel format). Sequences of staple strands for all DNA objects used in this work. Movie S1 (.mov format). Tomogram obtained from cryo-EM of the brick-like variant with thymines at all staple termini and with TT motifs at all crossover sites. Movie S2 (.mov format). Corresponding movie to fig. S20 for the nonirradiated (uncross-linked) sample of the brick-like variant with thymines at all staple termini and with TT motifs at all crossover sites in folding buffer and in the presence of 5 mM MgCl 2 . Movie S3 (.mov format). Corresponding movie to fig. S20 for the irradiated (cross-linked) sample of the brick-like variant with thymines at all staple termini and with TT motifs at all crossover sites in folding buffer and in the presence of 5 mM MgCl 2 . Movie S4 (.mov format). Corresponding movie to fig. S20 for the irradiated (cross-linked) sample of the brick-like variant with thymines at all staple termini and with TT motifs at all crossover sites in PBS buffer and in the absence of MgCl 2 . Movie S5 (.mov format). Corresponding movie to fig. [/table]
|
Digital versus Analog Procedures for the Prosthetic Restoration of Single Implants: A Randomized Controlled Trial with 1 Year of Follow-Up
Aim. To compare the outcome of digital versus analog procedures for the restoration of single implants. Methods. Over a twoyear period (2014)(2015)(2016), all patients who had been treated in a dental center with a single implant were randomly assigned to receive either a monolithic zirconia crown, fabricated with digital workflow (test group), or a metal-ceramic crown, fabricated with analog workflow (control group). All patients were followed for 1 year after the delivery of the final crown. The outcomes were success, complications, peri-implant marginal bone loss (PIMBL), patient satisfaction, and time and cost of the treatment. Results. 50 patients (22 males, 28 females; mean age 52.6±13.4 years) were randomly assigned to one of the groups (25 per group). Both workflows showed high success (92%) and low complication rate (8%). No significant differences were found in the mean PIMBL between test (0.39±0.29mm) and control (0.54±0.32mm) groups. Patients preferred digital impressions. Taking the impression took half the time in the test group (20±5min) than in the control (50±7min) group. When calculating active working time, workflow in the test group was more time-efficient than in the control group, for provisional (70±15min versus 340±37min) and final crowns (29±9min versus 260±26min). The digital procedure presented lower costs than the analog (€277.3 versus €392.2). Conclusions. No significant clinical or radiographic differences were found between digital and analog procedures; however, the digital workflow was preferred by patients; it reduced active treatment time and costs. The present study is registered in the ISRCTN
# Introduction
The world of dentistry is now experiencing a revolution, thanks to the rapid establishment of digital technologies [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref]. New acquisition devices (intraoral scanners [bib_ref] Intraoral scanners in dentistry: a review of the current literature, Mangano [/bib_ref] , face scanners, and cone-beam computed tomography (CBCT) [bib_ref] Anterior maxilla alveolar ridge dimension and morphology measurement by cone beam computerized..., Zhang [/bib_ref] allow the capture of three-dimensional (3D) images of patients, which are then processed in computer-assisted design/computer-assisted manufacturing (CAD/CAM) software [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] -this is in order to be able to design and then produce, through subtractive technologies (milling) or additive (3D printing) methods, prosthetic restorations [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Esthetic and Clinical Performance of Implant-Supported All-Ceramic Crowns Made with Prefabricated or..., Wittneben [/bib_ref] [bib_ref] Applications of additive manufacturing in dentistry: A review, Bhargav [/bib_ref] , surgical templates [bib_ref] Clinical applications and effectiveness of guided implant surgery: A critical review based..., Colombo [/bib_ref] , orthodontic aligners [bib_ref] Efficacy of clear aligners in controlling orthodontic tooth movement: a systematic review, Rossini [/bib_ref] , and a whole series of other custom-made devices (such as implants [bib_ref] Custom-made, root-analogue direct laser metal forming implant: a case report, Mangano [/bib_ref] and custom-made bone grafts [bib_ref] Custom-Made Synthetic Scaffolds for Bone Reconstruction: A Retrospective, Luongo [/bib_ref].
In implant prosthodontics, digital technologies today allow one to capture an accurate impression of the implant with intraoral scanners and therefore with structured light or laser only, without having to use conventional impression trays and materials [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] Intraoral scanners in dentistry: a review of the current literature, Mangano [/bib_ref] [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Accuracy of four intraoral scanners in oral implantology: a comparative in vitro..., Imburgia [/bib_ref]. This new procedure is absolutely pleasing to patients, because it can reduce discomfort and stress while in the dental chair [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] Intraoral scanners in dentistry: a review of the current literature, Mangano [/bib_ref] [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Digital versus analog complete-arch impressions for singleunit premolar implant crowns: Operating time..., Schepke [/bib_ref] [bib_ref] Patient outcomes and procedure working time for digital versus conventional impressions: A..., Gallardo [/bib_ref] ; moreover, it is appreciated by the clinicians, as, besides being a powerful marketing tool with patients, it simplifies the clinical procedures and allows one to communicate in a more efficient and dynamic way with the dental laboratory [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] Intraoral scanners in dentistry: a review of the current literature, Mangano [/bib_ref] [bib_ref] Digital versus analog complete-arch impressions for singleunit premolar implant crowns: Operating time..., Schepke [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref]. The modern dental laboratory, once it has received the optical impression, can proceed to the CAD design of the implant abutment and prosthetic crown, and can subsequently produce them through CAM procedures, 2 BioMed Research International such as milling [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] Intraoral scanners in dentistry: a review of the current literature, Mangano [/bib_ref] [bib_ref] Anterior maxilla alveolar ridge dimension and morphology measurement by cone beam computerized..., Zhang [/bib_ref] [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Esthetic and Clinical Performance of Implant-Supported All-Ceramic Crowns Made with Prefabricated or..., Wittneben [/bib_ref] [bib_ref] Digital versus analog complete-arch impressions for singleunit premolar implant crowns: Operating time..., Schepke [/bib_ref] [bib_ref] Patient outcomes and procedure working time for digital versus conventional impressions: A..., Gallardo [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref]. These restorations, appropriately characterized, will be sent to the dentist for clinical application [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] Intraoral scanners in dentistry: a review of the current literature, Mangano [/bib_ref] [bib_ref] Anterior maxilla alveolar ridge dimension and morphology measurement by cone beam computerized..., Zhang [/bib_ref] [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Esthetic and Clinical Performance of Implant-Supported All-Ceramic Crowns Made with Prefabricated or..., Wittneben [/bib_ref] [bib_ref] Digital versus analog complete-arch impressions for singleunit premolar implant crowns: Operating time..., Schepke [/bib_ref] [bib_ref] Patient outcomes and procedure working time for digital versus conventional impressions: A..., Gallardo [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref].
The replacement of the single missing or failing tooth with a dental implant is today one of the most frequent indications of modern implantology, with high percentages of survival and success, as unequivocally reported by the literature [bib_ref] Systematic review of the survival rate and the incidence of biological, technical,..., Jung [/bib_ref] [bib_ref] Prospective clinical evaluation of 307 singletooth morse taper-connection implants: A multicenter study, Mangano [/bib_ref] [bib_ref] Soft and hard tissue changes following immediate placement or immediate restoration of..., Yan [/bib_ref] [bib_ref] Singletooth replacement: factors affecting different prosthetic treatment modalities, Al-Quran [/bib_ref].
Digital technologies seem to represent an ideal application in implant prosthodontics, especially in the replacement of the single element, as recently demonstrated [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Esthetic and Clinical Performance of Implant-Supported All-Ceramic Crowns Made with Prefabricated or..., Wittneben [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] [bib_ref] Monolithic implantsupported lithium disilicate (LS2) crowns in a complete digital workflow: A..., Joda [/bib_ref]. The digital procedure consists of the acquisition of an optical impression of the implant, through the positioning of a scanbody, able to transfer the spatial position of the fixture into the CAD software [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Esthetic and Clinical Performance of Implant-Supported All-Ceramic Crowns Made with Prefabricated or..., Wittneben [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] [bib_ref] Monolithic implantsupported lithium disilicate (LS2) crowns in a complete digital workflow: A..., Joda [/bib_ref]. In the CAD, the dental technician replaces the mesh of the scanbody with the file contained in the implant library and thus design the restoration [bib_ref] Digital technology in fixed implant prosthodontics, Joda [/bib_ref] [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Esthetic and Clinical Performance of Implant-Supported All-Ceramic Crowns Made with Prefabricated or..., Wittneben [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] [bib_ref] Monolithic implantsupported lithium disilicate (LS2) crowns in a complete digital workflow: A..., Joda [/bib_ref]. The restoration is directly milled from a monolithic block of lithium disilicate [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Monolithic implantsupported lithium disilicate (LS2) crowns in a complete digital workflow: A..., Joda [/bib_ref] [bib_ref] Randomized controlled within-subject evaluation of digital and conventional workflows for the fabrication..., Sailer [/bib_ref] [bib_ref] Randomized controlled withinsubject evaluation of digital and conventional workflows for the fabrication..., Zeltner [/bib_ref] [bib_ref] CAD/CAMfabricated ceramic implant-supported single crowns made from lithium disilicate: Final results of..., Spies [/bib_ref] or zirconia [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] [bib_ref] Clinical Results and Technical Complications of Posterior Implant-Supported Modified Monolithic Zirconia Single..., Cheng [/bib_ref] [bib_ref] Fracture rate of monolithic zirconia restorations up to 5 years: A dental..., Sulaiman [/bib_ref] (without the need to pass through a physical model), then characterized, and then delivered to the dentist. This restoration can be directly screwed onto a prefabricated titanium abutment with a dedicated shape, or it can be cemented on an individual hybrid abutment, the upper part of which is shaped and milled in zirconia and glued to a titanium bonding base [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Esthetic and Clinical Performance of Implant-Supported All-Ceramic Crowns Made with Prefabricated or..., Wittneben [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] [bib_ref] Clinical Results and Technical Complications of Posterior Implant-Supported Modified Monolithic Zirconia Single..., Cheng [/bib_ref] [bib_ref] Fracture rate of monolithic zirconia restorations up to 5 years: A dental..., Sulaiman [/bib_ref].
The locking-taper, pure interference-fit implants (Morse taper connection implants, i.e., implants without a connecting screw between the abutment and the fixture) represent a successful long-term solution for prosthetic rehabilitation of single-tooth gaps [bib_ref] Prospective clinical evaluation of 307 singletooth morse taper-connection implants: A multicenter study, Mangano [/bib_ref] [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] , as well as partially and totally edentulous jaws [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref] , for supporting both fixed [bib_ref] Prospective clinical evaluation of 307 singletooth morse taper-connection implants: A multicenter study, Mangano [/bib_ref] [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref] and removable prostheses [bib_ref] Morse taper connection implants supporting "planned" maxillary and mandibular bar-retained overdentures: a..., Mangano [/bib_ref]. In the case of singletooth gaps, in particular, several clinical studies have shown that the use of such implants allows one to minimize the complications of a prosthetic nature [bib_ref] Prospective clinical evaluation of 307 singletooth morse taper-connection implants: A multicenter study, Mangano [/bib_ref] [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref] , particularly in the long term [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref].
To date, there are very few randomized controlled studies in the literature that compare the procedures and results obtained in the replacement of the single teeth with digital versus analog (conventional) procedures [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] ; moreover, there are no studies that address this topic for implants with a Morse taper, pure interference-fit connection [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref].
The aim of the present randomized clinical trial (RCT) was therefore to evaluate the success and complications encountered in the prosthetic restoration of single-tooth Morse taper connection implants, with digital and analog procedures, comparing the two methods; moreover, the present clinical trial aims to analyze and compare the patients' preference, the treatment times, and the costs, relative to the two different methodologies.
# Materials and methods
## Patient selection.
Only patients who had undergone surgical treatment with the insertion of a single Morse taper connection implant (Exacone5, Leone Implants, Sesto Fiorentino, Italy) in the posterior areas (premolars and molars) of both jaws, in the period between September 2014 and September 2016, in a single dental center (Gravedona, Como, Italy), were considered for enrollment in the present RCT. A further inclusion criterion was the diameter and height of the implant received: the patients had to be installed with a fixture of a minimum diameter of 4.1 mm and a height of at least 8 mm. In order to be enrolled in the study, patients had to have dentition in the opposite jaw and therefore occlusal contacts. Finally, to be enrolled, patients had to read and sign a document of adhesion to the present study, on the nature (and possible therapeutic alternatives) of which they were informed in detail; by signing this document, they committed themselves to come to the dental clinic for the required follow-up appointments.
All patients who received a single implant with a diameter of less than 4.1 mm and a height of less than 8 mm were automatically excluded from this study, as were all patients who had undergone preimplant regenerative bone therapies or who had been treated with guided bone regeneration and membranes for the presence of peri-implant defects. Additional exclusion criteria included systemic diseases such as uncompensated diabetes, immunocompromised states, head and neck tumors, and osteoporosis treated with aminobisphosphonates (administered orally and / or parenterally). Active periodontal infections and oral mucosa pathologies also represented exclusion criteria for enrollment in the present study. On the other hand, smoking and parafunctional habits (bruxism and/or clenching) did not represent exclusion criteria; smokers and bruxists could be included in this study.
The present RCT took place in full compliance with the principles set forth in the Helsinki Declaration on human subject experimentation of 1975 (and revision of 2008) and obtained the approval of the University of Insubria Ethics Committee, with approval number 826-0034086 (title: "Multicenter Clinical Studies on Survival and Success of Morse Taper Connection Implants").
## Study design.
The present study was designed as a randomized controlled clinical trial, for the comparison of two different prosthetic treatment modalities-digital versus analog procedures-for single implants inserted in the posterior areas of the jaws. Patients were therefore randomly assigned either to receive a monolithic zirconia crown, fabricated with a full digital workflow (test group: digital procedure consisting of digital impression with intraoral scanner, and CAD/CAM procedure without any physical model) or to receive a metal-ceramic crown, fabricated with a conventional analog workflow (control group: analog procedure consisting of impression-taking with polyvinyl siloxane, plaster model pouring, and lost-wax casting technique) [fig_ref] Table 1: Digital versus analog procedures [/fig_ref].
Block randomization was used in order to achieve a balanced allocation and an equal distribution of patients between the test and the control groups [bib_ref] Blocked randomization with randomly selected block sizes, Efird [/bib_ref]. The randomization scheme consisted of a sequence of blocks, each block containing a prespecified number of treatment assignments, in random order [bib_ref] Blocked randomization with randomly selected block sizes, Efird [/bib_ref]. (1) Preparation of the provisional resin crown (i) the implant analog is assembled with the transfer (ii) plaster models are poured (iii) the transfer is removed (iv) a prefabricated titanium abutment is inserted in the implant analog (v) the abutment is manually prepared into the desired form (vi) wax-up of the provisional crown (vii) a silicon mask is prepared over the wax-up (viii) the wax is removed (ix) the resin is injected into the silicon mask and polymerized in order to obtain the provisional crown (x) the provisional crown is polished and characterized (2) Preparation of the metal coping (i) wax-up of the coping (ii) lost-wax casting for the fabrication of the metal coping (iii) the metal coping is polished Second appointment in dental clinic: (1) Healing abutment removal (2) Insertion and activation of the individual hybrid abutment (3) Delivery and cementation of the provisional crown Second appointment in dental clinic (1) Healing abutment removal (2) Insertion and activation of the titanium abutment (3) Delivery and cementation of the provisional crown In dental laboratory: (1) Computer assisted design (CAD) (i) design of the final crown (2) Computer assisted manufacturing (CAM) (i) milling of the final crown in zirconia (ii) the zirconia crown is sintered (iii) the zirconia crown is characterized Third appointment in dental clinic: (1) The provisional crown is removed (2) The metal coping is placed on the titanium abutment The main outcomes of the study were clinical (implantcrown success and complications) and radiographic (periimplant marginal bone loss), but patient satisfaction and time and costs of therapy in the two treatment groups were also compared.
In both groups of patients, the same Morse taper connection implants were used. This implant system (Exacone5, Leone Implants, Sesto Fiorentino, Italy) features a screwless, pure interference-fit connection between the fixture and the abutment, with an angle of 1.5 ∘ , combined with an internal hexagon [bib_ref] Prospective clinical evaluation of 307 singletooth morse taper-connection implants: A multicenter study, Mangano [/bib_ref] [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref] [fig_ref] Figure 1: The fixtures used in the present study had a screwless Morse taper... [/fig_ref].
The present randomized controlled trial has been registered in the ISRCTN publicly available register (http://www.isrctn.com/ISRCTN36259164) with number ISRCTN36259164.
## Treatment procedures.
In both cases (either in the test-digital procedure; or in the control-analog procedure), the process started with the removal of the transmucosal healing abutment. Digital treatment (test procedure) consisted of taking an optical impression with an intraoral scanner (CS 36005, Carestream Dental, Rochester, NY, USA) after scanbody positioning. Scanning was limited to the posterior sector of interest, including the antagonist arch. This optical impression, in the form of .STL or .PLY files, was sent to the dental laboratory, which proceeded with the design in CAD software (Exocad DentalCAD5, Exocad, Darmstadt, Germany) of a definitive individual zirconia abutment to be cemented on a titanium base (link) and a temporary crown. The individual abutment was made of zirconia (for milling) with a powerful 5-axis milling machine (Roland DWX-505, Roland Easy Shape, Ascoli Piceno, Italy), sintered in an oven (Tabeo5, Mihm-Vogt, Stutensee, Germany), and cemented extraorally on a straight (Ti-base5 or Multitech straight5, Leone, Sesto Fiorentino, Italy) or 15 ∘ -angled (Multitech 15 ∘ 5, Leone, Sesto Fiorentino, Italy) titanium base, selected according to the CAD project. The temporary crown, however, was produced by milling in polymethylmethacrylate (PMMA) with a 4axis milling machine (Roland DWX-45, Roland Easy Shape, Ascoli Piceno, Italy). Upon delivery of the temporary, the individual hybrid abutment was activated with a percussion hammer and the PMMA crown cemented on it with a zincoxide eugenol cement (TempBond5, Kerr, Orange, CA, USA). A careful occlusal check, polishing, and characterization was made. Any occlusal precontacts recorded in this phase with articulating papers (Bausch Articulating Paper5, Bausch Inc, Nashua, NH, USA) was photographed, in order to be available and guide the modeling of the final crown. The temporary crown remained in situ for a total period of two months, at the end of which it was replaced with a definitive monolithic crown in translucent zirconia (Katana5, Kuraray Noritake, Tokyo, Japan). This was obtained by modifying the design of the temporary in the previous CAD scene, taking care to adapt the cement spaces to the needs of the new material (zirconia) and to check and modify the occlusal contact points, based on the indications collected at the time BioMed Research International 5 of positioning of the PMMA crown. In the digital treatment (test), no physical models of the jaws were prepared.
The analog treatment (control procedure) consisted of taking a conventional impression of the implant with polyvinyl siloxane (Elite HDPlus5, Zhermack, Badia Polesine, Italy) on a generic tray after positioning of an analog impression transfer. Alginate impression was made for the antagonist arch. These impressions were disinfected and sent by post to the laboratory where plaster models were poured, in which the position of the implant was reproduced by means of an implant analog. On this model, the technician prepared a titanium abutment, purchased by the company from the dentist and supplied for this purpose, on which a temporary resin crown was produced. At the same time, the technician proceeded to the wax-up and obtained the metal structure (coping) of the crown, through a classical technique of lost-wax casting. After activation with percussion hammer of the prepared titanium abutment, the clinician applied the temporary acrylic resin crown, checking all the occlusal contacts. After a 2-month period of temporization, the patient was recalled for a second impression in polyvinyl siloxane, in which the metal coping was positioned on the abutment and retained in the impression. Starting from this impression, the technician could stratify the ceramic over the metal structure (coping) of the restoration, and about a week later the patient was recalled for the application of the final metal-ceramic crown. Again, all occlusal contacts were carefully checked, and the restoration was polished and cemented with temporary cement.
The sequence of the digital and analog treatment procedures is summarized in [fig_ref] Table 1: Digital versus analog procedures [/fig_ref].
All patients were included in a hygiene-maintenance protocol, with 3 professional hygiene sessions per year, 1 every 4 months. During this period, all the complications and adverse events that could affect the restorations in the two groups were scrupulously noted. Finally, 1 year after the definitive crown was delivered, all the patients were recalled for the final examination visit, in which an endoral radiograph of control was also taken, and the outcomes of the study were evaluated.
## Outcomes of the study.
The outcomes of the present randomized controlled trial were clinical and radiographic in nature, such as the implant-crown success, the biologic and prosthetic complications encountered during the observation period, and the peri-implant marginal bone loss (PIMBL); moreover, the present study investigated patient satisfaction with and the time and cost aspect of the prosthetic treatment.
## Implant-crown
Success. Implant-crown success was the main clinical outcome of the present study. An implantsupported restoration was defined as successful if it was still functioning at the end of the study without any complication, either biological or prosthetic, during the entire follow-up and at the last control appointment, 1 year after delivery. On the other hand, if only a single complication involving the implant-supported restoration occurred, the crown was included in the group of failures. The biological complications occurring during the follow-up (peri-implant mucositis and peri-implantitis), as well as the prosthetic complications affecting the restoration (complications of a mechanical nature, such as abutment loosening or abutment fracture; technical complications, such as chipping or fracturing of the restoration) were considered the reasons for the failure of the restoration. The threshold for defining peri-implantitis was set at a probing pocket depth ≥6 mm with bleeding/suppuration on probing and evidence of peri-implant bone loss >3.0 mm [bib_ref] Clinical, microbiological, and immunological aspects of healthy versus periimplantitis tissue in full..., Ata-Ali [/bib_ref].
## Peri-implant marginal bone
Loss. The radiographic outcome of the present study was peri-implant bone stability, measured as PIMBL. This outcome was measured on intraoral radiographs, comparing the peri-implant bone peaks (mesial and distal) at the time of implant placement (T1) and 1 year after delivery of the definitive crown (T2). This comparison was made as described previously in other studies [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref] [bib_ref] Morse taper connection implants supporting "planned" maxillary and mandibular bar-retained overdentures: a..., Mangano [/bib_ref]. In short, the radiographs were taken using a Rinn system for alignment with a rigid film-object X-ray source coupled to a beam-aiming device. The mesial and distal marginal bone levels of all implants were measured at T1 and T2, with the help of an ocular grid (magnification 4.5x); the reference points for these measurements were the most coronal bone-to-implant contact and the implant shoulder margin. Accordingly, the PIMBL was calculated as modification in the peri-implant bone between T1 and T2, and the mean of the mesial and distal calculations was used as the final value. In order to adjust for radiographic distortion, the radiographic length of each implant was compared with the actual (true) implant length, by means of the following equation:
[formula] ℎ : ℎ = :(1) [/formula]
All measurements were taken by an independent calibrated observer who was not part of the treating team.
## Patient satisfaction.
The degree of satisfaction and the perception of the quality of the treatment received by patients with digital and with analog procedures were investigated by means of a visual analog score (VAS) questionnaire, based on 10 specific questions. For each of these questions, patients were asked to assign a score of 0 -10 -20 -30 -40 -50 -60 -70 -80 -90 -100, based on their satisfaction with the treatment received (0 = absolutely dissatisfied with the treatment; 10-20-30-40 = strongly dissatisfied with the treatment; 50 = insufficiently satisfied with the treatment; 60 = sufficiently satisfied with the treatment; 70-80-90 = very satisfied with the treatment; and 100 = fully satisfied with the treatment). The questions were as follows:
(1) Are you satisfied with the treatment?
with 0 = very dissatisfied and 100 = very satisfied
(2) Did you experience discomfort during the impression-taking?
with 0 = high discomfort and 100 = no discomfort with 0= absolutely not and 100 = yes, of course.
## Treatment time.
The overall treatment time and the active working time (i.e., the effective working time, excluding machine time) required for the prosthetic restoration of 1 single implant, with both treatments (digital versus analog procedure) were calculated as follows. For each treatment, the time required was calculated for the following procedures: impressions; preparation of the elements (abutment and crown) required for the provisionalization; delivery of the provisional restoration; and preparation and delivery of the final restoration. The treatment time was calculated in minutes (min).
## Cost of the treatment.
In order to assess the cost of both treatments (digital versus analog procedure) for the dentist, all the expenses related to the purchase of materials and the services of the dental laboratory were examined. The prices taken into consideration here were the official prices for the implant system used in this study, in Italy; therefore, being inclusive of value-added tax (VAT) (4% in the case of intraoral components; 22% in the case of components for extraoral use), laboratory prices reflected those of a mid-level laboratory located in northern Italy and were VAT-exempted. In this evaluation, the hourly cost of the dental clinic was not examined. The cost of the treatment was calculated in euros (€).
# Statistical analysis.
We summarized the demographics and other characteristics of the patients enrolled in this study, using standard descriptive statistics, and compared the two groups (digital versus analog groups) using either Student's t-test for independent samples, or a Chi-square test for continuous and categorical variables. Similarly, the distribution of major implant characteristics (position, location, diameter, and length) between the study groups was compared using Chi-square tests. We defined an implantcrown success in terms of absence of failure or complications (biological and/or prosthetic). In each group, we estimated the prevalence of successes and tested the null hypothesis of no difference using a Chi-square test. We estimated the average PIMBL in each group with 95% confidence interval (95% CI) and tested the difference in average bone loss using Student's t-test. Patient satisfaction was assessed using a VASlike score in 10 different domains; we report on the average score in each domain and on the results of a corresponding t-test to compare the two groups. Treatment times were measured on a subsample of n = 5 patients per study arm. We used Student's t-test to assess the null hypothesis of no difference between the study groups. The costs paid by the dentist for the fabrication of an implant-supported crown, through digital and analog procedure, were directly compared by examining expenses related to the purchase of materials and the dental laboratory service. The statistical analyses were performed using SAS system software, 9.4 release.
# Results
## Patient population and implant distribution.
In total, 50 patients (22 males and 28 females, between 24 and 76 years of age, mean age 52.6 ± 13.4 years, median 54.5 years, 95% CI 48.9-56.3 years) were selected for inclusion in the present RCT. Among these patients, only 8 were smokers and 4 were bruxists.
Patients were randomly assigned to one of the two groups (25 patients per group).
In the test group (digital treatment procedure), 12 males and 13 females were allocated. The patients' ages were between 24 and 68 years (mean 51.6 ± 12.3, median 54, 95% CI 46.8-56.4). Among the patients allocated in the test group, 4 were smokers and 3 were bruxists.
In the control group (analog treatment procedure), 10 males and 15 females were allocated. The patient age was between 26 and 76 years (mean 53.6 ± 14.6, median 55, 95% CI 47.9-59.3). Among the patients allocated to the control group, 4 were smokers and 1 was a bruxist.
Study groups did not differ by age, gender, or prevalence of smokers or bruxists (all p-values >0.05).
Overall, with regard to the overall implant distribution, 15 implants were in the maxilla and 35 in the mandible; 22 were premolars and 28 molars. The most frequently represented implant length was 10 mm (32 implants) followed by 12 mm (11 implants); only 7 implants were 8 mm in length. The most frequently represented implant diameter was 4.1 mm (42 implants), with only 8 fixtures having a diameter of 4.8 mm.
In the test group (digital treatment procedure), 6 implants were in the maxilla and 19 in the mandible; with regard to the position of the fixtures, 10 were premolars and 15 were molars. The most frequently used implant diameter was 4.1 mm (25 implants), whereas only 5 implants had a diameter of 4.8 mm; the most frequent implant length was 10 mm (18 implants), followed by 12 mm (5 implants) and 8 mm (2 implants).
In the control group (analog treatment procedure), 9 implants were in the maxilla and 16 in the mandible; with regard to the position of the fixtures, 12 were premolars and 13 molars. Finally, 22 fixtures had a diameter of 4.1 mm and 3 fixtures had a diameter of 4.8 mm; the most frequently used implant length was 10 mm (14 implants) followed by 12 mm (6 implants) and 8 mm (5 implants).
There was no difference between study groups in terms of implant location, position, length, and diameter (all p-values > 0.05; [fig_ref] Table 2: Implant distribution and number of failures and complications registered in the two... [/fig_ref].
## Implant-crown success.
No patients dropped out from the study and no implants failed, for a 100% implant survival at the last follow-up control, 1 year after the delivery of the final crowns. Overall, a limited number of biologic (3/50 implants, 6%) and prosthetic (1/50 implants, 2%) complications were reported, for an implant-crown success rate of 92% (46/50 implants did not show any complication, 1 year after the delivery of the final crown).
In the test group, one 56-year-old, smoking, female patient experienced a biologic complication (peri-implant mucositis, with soft-tissue inflammation, exudation, and bleeding on probing in absence of peri-implant bone loss). This patient was successfully treated with a series of professional oral hygiene sessions and the problem was resolved. At the end of the study, the biologic complications in the test group amounted to 4% (1/25 crowns). With regard to prosthetic complications, the monolithic translucent zirconia crown of a 44-year-old male patient faced chipping, i.e., fracture of a portion of the mesio-vestibular cusp of a mandibular molar. The patient was a bruxist. The damaged crown was therefore removed and designed again from the previous intraoral scan, taking care to reduce the occlusal contacts at that critical point. The new crown was milled and delivered to the patient, and no further prosthetic complications were registered. Accordingly, at the end of the study, the prosthetic complications in the test group amounted to 4% (1/25 crowns). The implant-crown success in the test group amounted to 92% (23/25 implants without complications).
In the control group, two implants (1 in a 62-year-old smoking patient and 1 in a 72-year-old nonsmoking patient) were affected by peri-implant mucositis with soft-tissue inflammation and swelling, without evidence of peri-implant bone loss. Both these patients had a low level of compliance and insufficient oral hygiene. However, both these implants were successfully treated with professional oral hygiene sessions and no further problems were evidenced in these patients. At the end of the study, the incidence of biologic complications in the control group was 8% (2/25 implants). Conversely, no prosthetic complications were reported in this group, for an overall implant-crown success of 92% (23/25 implants without complications).
The test and the control groups did not differ in terms of complications or success probability (all p-values > 0.05; [fig_ref] Table 2: Implant distribution and number of failures and complications registered in the two... [/fig_ref]. Examples of the digital and analog workflow are provided in [fig_ref] 3: Impression taking [/fig_ref]
## Peri-implant marginal bone loss.
Overall, the evaluation of bone loss around the implants revealed a mean PIMBL of 0.47 ± 0.31 mm (median 0.4 mm; 95% CI 0.39-0.55 mm) after 1 year from the delivery of the final crown. In the test group (digital treatment procedure), the mean PIMBL amounted to 0.39 ± 0.29 mm (median 0.4 mm; 95% CI 0.28-0.5 mm). In the control group (analog treatment procedure), the mean PIMBL amounted to 0.54 ± 0.32 mm (median 0.6; 95% CI 0.42-0.66 mm). The average difference of -0.16 mm (95% CI: -0.33, 0.02 mm) in favor of the test group was not statistically significant (p = 0.08). All details on the PIMBL are summarized in [fig_ref] Table 3: Peri-implant marginal bone loss [/fig_ref].
## Patient satisfaction.
The patient satisfaction for each treatment (digital and analog procedure) is summarized in [fig_ref] Table 4: Patient satisfaction in the two different groups [/fig_ref]. During the impression-taking, the digital group reported less discomfort (average difference in VAS score: +24, p < 0.0001) and were less likely to experience gag reflex/nausea (+13.2, p = 0.007), resulting in a higher rating of overall comfort during the impression procedure (+28.4, p < 0.001). Conversely, there was no difference between the study groups in the restoration phase. In addition, the digital implant received higher scores in terms of treatment time (average difference in VAS: +16, p < 0.0001) and costs (+9.2, p = 0.01). Satisfaction over the aesthetic result did not differ, with both methods scoring above 90. The overall patient satisfaction did not differ (average difference in VAS score: +2.8, p = 0.35), but only the digital group had an average score above 90. Finally, a statistically significant difference was found in the overall mean VAS (+ 9.9, p < 0.0001).
## Treatment time.
The time required for each treatment (digital and analog procedure) is summarized in [fig_ref] Table 5: Treatment time, in minutes [/fig_ref]. On average, time for impression-taking was almost halved in the digital compared with the analog groups (20 ± 5 min versus 50 ± 7 min; p< 0 .0001). Overall, production times of both the temporary (600 ± 45 min versus 340 ± 37 min; p = 0.0001) and the final (525 ± 39 min versus 260 ± 26 min; p < 0.0001) crown were larger for the former group. However, when calculating active working time for the dental technician, the workflow in the test group was more efficient than that of the control group, for both the provisional (70 ± 15 min versus 340 ± 37 min; p < 0.0001) and the final crown (29 ± 9 min versus 260 ± 26 min; p < 0.0001). We observed no difference in delivery times.
## 3.6.
Cost of the Treatment. The costs of the digital and analog treatment procedures were as summarized in [fig_ref] Table 6: Costs for materials and dental laboratory procedures, in euro [/fig_ref]. For the test group (digital procedure), the cost of the prosthetic components amounted to €107.3 (scanbody €36.6, titanium base €70.7). The dental laboratory costs amounted to €60 for the CAD/CAM procedures for the fabrication of the provisional PMMA crown and the upper portion of the abutment (zirconia) and the assembly of the individual hybrid abutment. The CAD/CAM procedures for the fabrication of the final translucent zirconia crown cost €110. Overall, the expense for the fabrication of a single implant crown via the full digital procedure (without any model) amounted to €277.3.
For the control group (analog procedure), the cost of the prosthetic components amounted to €107.2 (transfer €25.0, abutment €70.7, and implant analog €11.5). The dental laboratory costs amounted to €75 for the procedures needed for the preparation of the provisional resin crown and €210 for the procedures required for the fabrication of the final metal-ceramic crown. The expense for the fabrication of a single implant crown via the conventional analog procedure (with a plaster model) amounted to €392.2.
Overall, the global cost for the fabrication of the 25 crowns of the test group via the digital procedure amounted to €6932.5; the global cost for the fabrication of the 25 crowns of the control group via the analogic procedure amounted to €9805.
# Discussion
Modern digital techniques today are proliferating in dentistry, especially in implant prosthodontics [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref]. The intraoral scan, in particular, gathers great consensus among dentists, <.0001 because it allows one to eliminate the conventional physical impression with impression trays and materials (polyvinyl siloxane/polyether), which can be technically difficult in implant dentistry and has always represented a moment of stress for the patient . Further advantages of the optical impression are represented by the rationalization of processes and interactions with the dental laboratory [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref]. Once the intraoral scan has been taken, the CAD/CAM procedures allow one to design and produce, by milling, individual prosthetic abutments and temporary and definitive prostheses . Such prostheses are clinically precise, as demonstrated by the literature in various applications [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] [bib_ref] Randomized controlled within-subject evaluation of digital and conventional workflows for the fabrication..., Sailer [/bib_ref] [bib_ref] Randomized controlled withinsubject evaluation of digital and conventional workflows for the fabrication..., Zeltner [/bib_ref] [bib_ref] CAD/CAMfabricated ceramic implant-supported single crowns made from lithium disilicate: Final results of..., Spies [/bib_ref] [bib_ref] Clinical Results and Technical Complications of Posterior Implant-Supported Modified Monolithic Zirconia Single..., Cheng [/bib_ref] [bib_ref] Fracture rate of monolithic zirconia restorations up to 5 years: A dental..., Sulaiman [/bib_ref] and are fabricated of highly aesthetic materials, able to integrate perfectly into the patient's oral cavity, such as lithium disilicate [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Monolithic implantsupported lithium disilicate (LS2) crowns in a complete digital workflow: A..., Joda [/bib_ref] [bib_ref] Randomized controlled within-subject evaluation of digital and conventional workflows for the fabrication..., Sailer [/bib_ref] [bib_ref] Randomized controlled withinsubject evaluation of digital and conventional workflows for the fabrication..., Zeltner [/bib_ref] [bib_ref] CAD/CAMfabricated ceramic implant-supported single crowns made from lithium disilicate: Final results of..., Spies [/bib_ref] [bib_ref] Digitally-oriented materials: focus on lithium disilicate ceramics, Zarone [/bib_ref] and zirconia [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref] [bib_ref] Clinical Results and Technical Complications of Posterior Implant-Supported Modified Monolithic Zirconia Single..., Cheng [/bib_ref] [bib_ref] Fracture rate of monolithic zirconia restorations up to 5 years: A dental..., Sulaiman [/bib_ref] [bib_ref] Effect of mouthrinses on color stability of monolithic zirconia and feldspathic ceramic:..., Derafshi [/bib_ref]. Despite the fact that digital procedures are conquering the market, to date only few clinical studies have compared the results obtained in implant prosthodontics with those obtained by modern digital techniques (intraoral digital impressions by means of an intraoral scanner) and conventional analog techniques (such as the classic physical impression with trays and materials, for the manufacture of metal-ceramic prostheses) [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref]. In a prospective clinical study, Joda et al. [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] evaluated the time efficiency of digital versus conventional workflows for the fabrication of single, implant-supported crowns in posterior sites. Twenty patients were selected and enrolled in the study, each one receiving a customized CAD/CAM individual hybrid abutment with a monolithic zirconia crown (test, digital workflow) and a standardized titanium abutment plus a porcelain-fused-to-metal crown (control, conventional workflow) [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref]. The primary outcome of the study was time efficiency; therefore all clinical and laboratory procedures were timed, in minutes [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref]. All crowns were provided within two clinical appointments. However, the mean total production time (as sum of all the clinical and laboratory steps) was significantly different (p = 0.0001): in fact, the [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref]. The authors concluded that the digital workflow seems to be more timeefficient than the conventional workflow [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref]. Similar conclusions were reported in another randomized controlled trial by the same authors [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] , who evaluated the time efficiency of the single implant restoration with monolithic lithium disilicate crowns (test) versus porcelain fused to zirconium dioxide (control), in a digital workflow. Twenty patients in need of single-tooth replacement in posterior regions were randomly assigned to be restored with monolithic CAD/CAM lithium disilicate crowns bonded to prefabricated titanium abutments (test group, 10 patients) or to receive CAD/CAM-fabricated zirconia suprastructures and hand-layered ceramic veneering (control group, 10 patients) [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref]. In the test group, no physical models were needed; in the control group, conversely, milled master models were produced [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref]. Clinical appointments were needed for both groups of patients. However, the mean total fabrication time was significantly different (p = 0.0001), with 75.3 ± 2.1 minutes for the test group and 156.6 ± 4.6 minutes for the control group [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref]. A significantly shorter (p = 0.001) mean chair time for the test group (20.8 ± 0.3 min) was found, when compared with the control group (24.1 ± 1.1 min); for the laboratory, the reduction in working time was significant too (p = 0.0001) and even more evident (54.5 ± 4.9 minutes for test group versus 132.5 ± 8.7 minutes for control group) [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref]. In addition, the digital workflow with monolithic single crowns resulted in an overall reduction of 30% of the laboratory costs [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref]. The authors concluded that the direct fabrication of lithium disilicate monolithic crowns was more time-efficient than the fabrication of CAD/CAM-fabricated zirconia copings veneered with ceramic [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref].
## Production of the temporary crown
In another prospective cohort trial, the authors evaluated the costs and time for implant-supported single-tooth reconstructions with a digital versus a conventional workflow [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref].
Twenty patients were enrolled for rehabilitation with 2 x 20 implant crowns in a crossover study; the comparison was between digital workflow (customized titanium abutment + CAD/CAM zirconia superstructures) versus conventional workflow (standardized titanium abutments plus porcelain fused metal crowns [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref]. For each treatment procedure, the costs were analyzed. At the end of the study, direct treatment costs were significantly reduced (p = 0.0004) with digital workflow (CHF 1815.35) when compared with the analogic workflow (CHF 2119.65) [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref]. The total laboratory costs were significantly (p = 0.003) reduced too with the digital procedures (digital workflow: CHF 941.95; conventional workflow: CHF 1245.65) [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref]. The clinical productivity rate was moreover increased with the digital pathway [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref].
All these studies have shown that modern digital technologies are as clinically successful as conventional (analog) techniques and can reduce both the time required for implant-prosthetic treatment and the costs associated with it [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref] [bib_ref] Time-efficiency analysis comparing digital and conventional workflows for implant crowns: a prospective..., Joda [/bib_ref] [bib_ref] Time-efficiency analysis of the treatment with monolithic implant crowns in a digital..., Joda [/bib_ref] [bib_ref] Digital vs. conventional implant prosthetic workflows: a cost/time analysis, Joda [/bib_ref] [bib_ref] Complete digital workflow for the production of implant-supported single-unit monolithic crowns, Joda [/bib_ref]. However, there is certainly a need for more scientific evidence on this topic [bib_ref] The complete digital workflow in fixed prosthodontics: a systematic review, Joda [/bib_ref].
Our present randomized controlled clinical trial aimed to compare the reliability and effectiveness of two treatment modalities for the single implant: digital versus analog procedure-this considering not only relevant clinical aspects such as complications and marginal bone resorption, but also patient satisfaction, treatment time, and costs. We therefore selected 50 patients, each treated with a single implant positioned in the posterior areas of the jaw and allocated them to two numerically identical groups: the first group (test, 25 patients) was treated with modern digital procedures, while the second group (control, the other 25) received an analog (and therefore conventional) prosthetic treatment. It is important to emphasize how, in the present study, the block randomization has determined the formation of two homogeneous groups of patients, with regard to demographics and implant distribution. In fact, study groups did not differ by age, gender, or prevalence of smokers or bruxists, nor in terms of implant location, position, length, and diameter (all p-values > 0.05).
The first finding that emerged from our study is that 1 year after the definitive crown was delivered, there were no implant failures, for a 100% survival rate in the two groups.
There were only two complications for each group (one biological and one prosthetic complication in the test group; two biological complications in the control group) for an implant-prosthetic success (i.e., survival of implant-supported restoration without any complications) of 92%. From the clinical point of view, therefore, the two groups had virtually identical survival and success rates. The only difference between the two groups (although not statistically significant) was given by the higher incidence of biological problems (8%) in the control group, with two patients suffering from peri-implant mucositis. These patients had poor oral hygiene compliance, but nevertheless both these implants were successfully treated with professional oral hygiene sessions and no further problems were evidenced.
No prosthetic complications occurred in the control group. This is not surprising; it is well known from the literature that systems with a Morse taper implant abutment connection are characterized by a very low incidence of prosthetic complications (whether mechanical or technical) in both the short and long terms [bib_ref] Prospective clinical evaluation of 307 singletooth morse taper-connection implants: A multicenter study, Mangano [/bib_ref] [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref] [bib_ref] Prosthetic Complications Affecting Single-Tooth Morse-Taper Connection Implants, Mangano [/bib_ref]. In a recent retrospective clinical study on 578 patients treated with 612 Morse taper connection implants and rehabilitated single crowns, the 15-year cumulative implant-crown success rate was 94.5% (93.1% and 94.9% for the anterior and posterior crowns, respectively), with a very low incidence of prosthetic complications (1.5%) [bib_ref] Prosthetic Complications Affecting Single-Tooth Morse-Taper Connection Implants, Mangano [/bib_ref]. This performance is guaranteed by the absolute mechanical stability of the connection between the abutment and the implant, in the absence of relative micromovements [bib_ref] Performance of conical abutment (Morse Taper) connection implants: a systematic review, Schmitt [/bib_ref] , which reduces the incidence of abutment loosening even in the long term [bib_ref] Prospective clinical evaluation of 307 singletooth morse taper-connection implants: A multicenter study, Mangano [/bib_ref] [bib_ref] Short (8-mm) locking-taper implants supporting single crowns in posterior region: a prospective..., Mangano [/bib_ref] [bib_ref] Fixed restorations supported by Morse-taper connection implants: a retrospective clinical study with..., Mangano [/bib_ref] [bib_ref] Survival and complication rates of fixed restorations supported by locking-taper implants: a..., Mangano [/bib_ref] [bib_ref] Prosthetic Complications Affecting Single-Tooth Morse-Taper Connection Implants, Mangano [/bib_ref].
Consistently, even in the test group there were no mechanical complications. However, there was a complication of a technical nature. In fact, a monolithic crown in translucent zirconia underwent chipping, with fracture of a portion of the mesiovestibular cusp of a mandibular molar. In this case, the patient was a bruxist, and the restoration was therefore replaced with a new monolithic crown, always made in CAD/CAM, but suitably modified in terms of design, in order to prevent any further chipping. In a case of technical problems of this type, the digital procedure certainly has the advantage of being able to resume and modify the original CAD design of the crown and proceed with rapid milling without having to repeat the impression; this saves a lot of time. However, the removal of a well-cemented zirconia crown can be difficult, especially if definitive cements are used. To prevent this problem, and to be able to remove the crowns if necessary, in our present study, we cemented all crowns with a temporary cement.
In our present study, there was no statistically significant difference in marginal bone resorption at 1 year from the delivery of the definitive crown, in either group; the stability of the mesial and distal bone levels at the implant was indeed optimal. In the test group, the mean PIMBL amounted to 0.39 ± 0.29 mm (median 0.4 mm; 95% CI 0.28-0.5 mm); in the control group (analog treatment procedure), the mean PIMBL amounted to 0.54 ± 0.32 mm (median 0.6; 95% CI 0.42-0.66 mm).
Since clinical and radiographic data do not show significant differences between the test and control groups, attention must be paid to patient satisfaction and the time and costs of the two different workflows.
In particular, the analysis of patient preferences showed that patients prefer the optical impression to the conventional impression, as unequivocally demonstrated in the literature [bib_ref] Intraoral scanners in dentistry: a review of the current literature, Mangano [/bib_ref] [bib_ref] Digital versus analog complete-arch impressions for singleunit premolar implant crowns: Operating time..., Schepke [/bib_ref] [bib_ref] Patient outcomes and procedure working time for digital versus conventional impressions: A..., Gallardo [/bib_ref] [bib_ref] Comparison of digital and conventional impression techniques: evaluation of patients' perception, treatment..., Yuzbasioglu [/bib_ref]. In fact, during the impression-taking, the digital group reported less discomfort (average difference in VAS score: +24, p < 0.0001) and were less likely to experience gag reflex/nausea (+13.2, p = 0.007), resulting in a higher rating of overall comfort during the impression procedure (+28.4, p < 0.001). On the other hand, there were no statistically significant differences in the patients' attitude towards the subsequent phases of treatment (i.e., the application of provisional and final restorations). However, the digital procedure received higher scores in terms of treatment time (average difference in VAS: +16, p < 0.0001) and costs (+9.2, p = 0.01); although not statistically significant, the overall satisfaction for the treatment was higher in the test group.
On average, time for impression-taking was almost halved in the digital compared with the analog group (20 ± 5 min versus 50 ± 7 min; p < 0.0001). Overall, production times of both the temporary (600 ± 45 min versus 340 ± 37 min; p = 0.0001) and the final (525 ± 39 min versus 260 ± 26 min; p < 0.0001) crown were larger for the former group. The latter results depended on the long sintering times of zirconia, in the test group. During the milling and sintering of the zirconia in the oven, however, the technician is free to do something else and to attend to other cases, without needing to guard the oven itself. Therefore, in an evaluation of the active working time, the situation was reversed, and the workflow in the test group was more efficient than that of the control group, for both the provisional (70 ± 15 min versus 340 ± 37 min; p < 0.0001) and the final crown (29 ± 9 min versus 260 ± 26 min; p< 0 .0001).
Finally, with regard to the cost of treatment, the digital procedure presented lower costs for the dentist than the conventional one (€277.3 versus €392.2, per each crown). Many of the savings were concentrated in the laboratory procedures for fabrication of the final crown (€110 for the digital crown versus €210 for the analog crown); there were no differences in the cost of impression-taking, which amounted to about €107 across the two groups, and there was a little saving in the manufacturing of the temporary restoration (€60 digital versus €75 analog). Overall, the savings for each crown using a digital procedure amounted to €114.9; considering all 25 patients, the savings guaranteed by the digital procedure amounted to €2872.5. Obviously, it must be considered that, in the face of this saving on laboratory procedures, both the dental practice and the dental laboratory must sustain substantial investments in technology to be able to work in full digital workflow (in the specific case of this study, about €30,000 for the purchase of intraoral scanner, €6000 for the purchase of CAD software, and about €35,000 for the purchase of milling machines and zirconia sintering oven), which are not necessary for working in an analog workflow. Finally, the prices for the implant-prosthetic components and materials reported in this study refer to the Italian market and are inclusive of the Italian VAT; the costs could be different in another country. Similarly, the laboratory costs reported here are those of a medium-sized dental laboratory, and they refer to a local region (that of high Lombardy, in northern Italy on the Swiss border); laboratory prices may differ, even within the same region.
The present study has limitations (such as the low number of subjects enrolled, the limited number of crowns placed, and the short follow-up time); therefore further long-term randomized controlled trials are needed to draw more specific conclusions on the reliability of full digital procedures for the fabrication of single implant crowns.
# Conclusions
In the present randomized controlled trial, both the digital and the analog workflows worked successfully in restoring single-tooth gaps with implant-supported crowns, showing high success rates (92%) and a low incidence of complications (8%), and no statistically significant differences were found in the PIMBL between the two groups (test: 0.39 ± 0.29 mm; control: 0.54 ± 0.32 mm) 1 year after delivery of the definitive crown. However, patients preferred digital impressions to the conventional ones and were globally more satisfied with the digital procedures. The digital procedures were more timeefficient. In fact, on average, time for impression-taking was almost halved in the digital compared with the analog groups; and although the provisional and definitive crown fabrication involved overall more time in the test than in the control group, most of this time was machine time (i.e., the machines were operating, without the need for a continuous control by the dental technician). Therefore, when calculating active working time for the dental technician, the workflow in the test group was more efficient than that of the control group, for both the provisional (70 ± 15 min versus 340 ± 37 min; p < 0.0001) and the final crown (29 ± 9 min versus 260 ± 26 min; p < 0.0001). Finally, the digital procedure presented lower costs for the dentist than the conventional one (€277.3 versus €392.2, per crown).
Within its inherent limitations (such as the low number of subjects enrolled, the limited number of crowns placed, and the short follow-up time), the present RCT supports the concept that the digital workflow is preferred by patients, is time-effective for the dental laboratory, and is less expensive for the dentist, when compared with the analog one. Further long-term RCTs on a larger sample of patients are required to draw more specific conclusions on the reliability and efficacy of full digital workflow for the fabrication of implantsupported single crowns.
## Data availability
The data used to support the findings of this study are available from the corresponding author upon reasonable request.
# Disclosure
The present study is self-funded.
## Conflicts of interest
No conflicts of interest are reported for this clinical research.
[fig] 3: Impression taking (i) implant arch in polyvinyl siloxane with the metal coping in position (ii) opposite arch in alginate (iii) occlusal registration Third appointment in dental clinic: (1) Removal of the provisional crown (2) Delivery and cementation of the final monolithic translucent zirconia crown In dental laboratory: (1) Preparation of the final metal-ceramic crown (i) plaster models are poured with coping in position (ii) the ceramic is stratified over the metal coping and sintered (iii) the ceramic is polished Fourth appointment in dental clinic: (1) Removal of the provisional crown (2) Delivery and cementation of the final metal-ceramic crown [/fig]
[fig] Figure 1: The fixtures used in the present study had a screwless Morse taper implant abutment connection (Exacone5, Leone Implants, Sesto Fiorentino, Italy) with a taper angle of 1.5 ∘ , combined with an internal hexagon. (a) Drawing of the fixture with the multitech abutment in position, test group; (b) section of the same assembly along the long axis, 30%; (c) section of the same assembly along the long axis, 50%. [/fig]
[fig] 5: Does your implant-supported restoration function well? with 0 = bad function and 100 = well function (6) Do you feel secure biting on your restoration? with 0 = unsecure and 100 = secure(7) Are you pleased with the final aesthetic result? with 0 = not pleased and 100 = very much pleased (8) Is the treatment time justified? with 0 = totally not justified and 100 = completely justified (9) Is the treatment cost justified? with 0 = totally not justified (too high), and 100 = completely justified (10) Would you repeat this treatment again, if necessary? [/fig]
[fig] Figure 6: Test group (digital workflow). (a) the .STL file of the final zirconia crown is ready to be milled; (b) delivery of the final monolithic zirconia crown; (c) radiograph at the time of implant placement; (d) radiograph 1 year after delivery of the definitive crown. [/fig]
[fig] Figure 7, Figure 8: Test group (digital workflow). Digital impressions of a first mandibular molar with CS 36005 (Carestream Dental, Rochester, NY, USA). (a) Intraoral view of the implant scanbody; (b) impression of the mucosa after removal of the healing abutment, .PLY; (c) impression with the scanbody in situ; .PLY (d) .STL file of the impression with the implant scanbody. Test group (digital workflow). Computer-assisted design (CAD) phases for the design of the final zirconia abutment and the provisional PMMA crown with Exocad DentalCAD5 (Exocad, Darmstadt, Germany). (a) The final zirconia abutment and the provisional PMMA crown have been designed; (b) occlusal view of the temporary crown; (c) space between the individual zirconia abutment and the opposing arch; (d) relationship between the abutment and the crown using the transparency tool. [/fig]
[table] Table 1: Digital versus analog procedures: workflow. iv) polishing and characterization of the provisional PMMA crown (3) Assembly of the individual abutment by cementing the upper zirconia portion over the selected titanium baseIn dental laboratory: [/table]
[table] Table 2: Implant distribution and number of failures and complications registered in the two groups (digital versus analog procedures). [/table]
[table] Table 4: Patient satisfaction in the two different groups (digital versus analog treatment). * = Student's t-test for independent samples, using Satterthwaite method in case of unequal group variances. [/table]
[table] Table 6: Costs for materials and dental laboratory procedures, in euro (€), in the two different groups (digital versus analog treatment). [/table]
|
Non-invasive prediction of recurrence in bladder cancer by detecting somatic TERT promoter mutations in urine
Background: Urothelial bladder cancer (UBC) is characterised by a high risk of recurrence. Patient monitoring is currently based on iterative cystoscopy and on urine cytology with low sensitivity in non-muscle-invasive bladder cancer (NMIBC). Telomerase reverse transcriptase (TERT) is frequently reactivated in UBC by promoter mutations.Methods:We studied whether detection of TERT mutation in urine could be a predictor of UBC recurrence and compared this to cytology/cystoscopy for patient follow-up. A total of 348 patients treated by transurethral bladder resection for UBC were included together with 167 control patients.Results: Overall sensitivity was 80.5% and specificity 89.8%, and was not greatly impacted by inflammation or infection. TERT remaining positive after initial surgery was associated with residual carcinoma in situ. TERT in urine was a reliable and dynamic predictor of recurrence in NMIBC (Po0.0001). In univariate analysis, TERT positive-status after initial surgery increased risk of recurrence by 5.34-fold (P ¼ 0.0004). TERT positive-status was still associated with recurrence in the subset of patients with negative cystoscopy (P ¼ 0.034).Conclusions: TERT mutations in urine might be helpful for early detection of recurrence in UBC, especially in NMIBC.
Diagnosis and treatment of patients presenting an urothelial bladder cancer (UBC) is a major challenge for clinicians. Prognosis for UBC is tightly correlated to stage and grade; the two main entities depend on infiltration of the muscle layer and are nonmuscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC). The association of urine cytology and cystoscopy is the current gold standard to detect recurrence and monitor UBC. The main advantage of urine cytology is that it is non-invasive, cheap, and easy to perform. Its high specificity (90-98%) makes cytology an interesting test to monitor high-grade tumours, with sensitivity up to 90% in pTis [bib_ref] Narrow-band imaging cystoscopy in non-muscle-invasive bladder cancer: a prospective comparison to the..., Geavlete [/bib_ref]. Unfortunately, overall sensitivity to detect tumour cells ranges from 22 to 62% [bib_ref] Diagnostic value of cytology of voided urine, Koss [/bib_ref] [bib_ref] Cost efficiency analysis of modern cytocentrifugation methods versus liquid based (Cytyc Thinprep)..., Piaton [/bib_ref] [bib_ref] Non-invasive diagnostic tests for bladder cancer: a review of the literature, Bassi [/bib_ref] making it unsuitable for low-grade lesions. Although the current gold standard is cystoscopy and cytology, it is subjective and may vary with the experience of the observers [bib_ref] The promise of novel molecular markers in bladder cancer, Miremami [/bib_ref]. This is particularly a problem in some conditions (elderly patients or those with neurogenic bladder with a chronic urinary infection, inflammatory bladder etc.). Therefore objective biological markers are required. To date, several urinary biomarkers have been described, some of them being more sensitive than urine cytology. However, they are lacking specificity, especially in the conditions cited above [bib_ref] Analysis of false-positive BTA STAT test results in patients followed up for..., Raitanen [/bib_ref] [bib_ref] The role of BTA stat Test in follow-up of patients with bladder..., Raitanen [/bib_ref] [bib_ref] Urinary protein biomarker panel for the detection of recurrent bladder cancer, Rosser [/bib_ref] [bib_ref] Urinary biomarkers for diagnosis of bladder cancer: a systematic review and meta-analysis, Chou [/bib_ref].
TERT (telomerase reverse transcriptase) is an essential safeguard of genomic integrity, responsible for telomere maintenance. In aging or damaged cells, TERT activity is physiologically shut down leading to shortened telomeres and induction of a form of cell death called senescence. Escaping senescence is a hallmark of cancer [bib_ref] Hallmarks of cancer: the next generation, Hanahan [/bib_ref] and nearly all tumours develop genetic or epigenetic strategies to avoid elimination by increasing TERT activity [bib_ref] Human telomerase: biogenesis, trafficking, recruitment, and activation, Schmidt [/bib_ref]. Two recurrent somatic mutations (C228T and C250T) have been identified in the TERT promoter in melanoma [bib_ref] TERT promoter mutations in familial and sporadic melanoma, Horn [/bib_ref] [bib_ref] Highly recurrent TERT promoter mutations in human melanoma, Huang [/bib_ref] as well as in various other tumours including bladder cancers [bib_ref] TERT promoter mutations occur frequently in gliomas and a subset of tumors..., Killela [/bib_ref] [bib_ref] Telomerase reverse transcriptase gene promoter mutations help discern the origin of urogenital..., Wu [/bib_ref]. TERT promoter mutations have previously been described at high frequencies across stages in malignant bladder tumours, but the prognostic value in urine is unclear [bib_ref] Telomerase reverse transcriptase promoter mutations in bladder cancer: high frequency across stages,..., Allory [/bib_ref] [bib_ref] Multiplex PCR and next generation sequencing for the non-invasive detection of bladder..., Ward [/bib_ref].
In this context, we evaluated, in a large prospective cohort, the use of non-invasive detection of TERT promoter mutations in urine as a predictive marker of recurrences and compared this to cytology and cystoscopy.
# Materials and methods
Patients. Eighty-three patients from the Urology department of Lyon Sud university hospital (Lyon, France) presenting UBC irrespective of histological stage, were initially enrolled in a previous study between 2000 and 2008 [bib_ref] Microarray gene expression profiling and analysis of bladder cancer supports the sub-classification..., Descotes [/bib_ref]. The cohort was then prospectively increased up to 348 patients in 2014 [fig_ref] Table 1: Characteristics of tumours and patients with TUBR Abbreviations [/fig_ref]. Informed consent was obtained from patients to use their specimens for research purposes, as required by the French Committee for the Protection of Human Subjects. This study complies with the latest version of the Declaration of Helsinki and general guidelines for good clinical practice. Treatments (intravesical BCG or mitomycin) and followup were based on European Association of Urology (EAU) guidelines. For each patient an initial urine sample was obtained at the time of tumour resection by cystoscopy (transurethral bladder resection, (TUBR)). Subsequent urine samples were collected during follow-up consultations. No extra samples of urines were collected for this study, because analyses were performed on residual materials obtained for cytology. During follow-up some patients presented a recurrence based on cystoscopy. For 50 of them, recurrence was not confirmed by pathologists on surgical material obtained by TUBR. We further referred to these cases as 'non-tumour' pathology post-TUBR.
To assess specificity, 167 patients without UBC were also included as controls [fig_ref] Table 2: Pathological data and TERT status in 167 urines without bladder cancer Abbreviation [/fig_ref] : 89 from healthy individuals or from patients consulting for low urinary tract symptoms or urinary incontinence (excluding patients reporting macroscopic and microscopic haematuria), 17 neurogenic bladder, 10 infectious urines, and 42 patients diagnosed with any other type of cancer (prostate, kidney, intestine, etc.). Of note, patients presenting with macroscopic and microscopic haematuria were excluded from this control cohort because it might be linked to an undiagnosed UBC.
Cytology. Urine samples were fixed with a Carbowax solution of 20% polyethylene glycol 1500 (Merck, Darmstadt, Germany) in 50% ethanol and treated as previously described (Collin-Chavagnac et al, 2010). Urothelial cells were considered highgrade when they displayed an increased nucleus/cytoplasm (N/C) ratio, hyperchromatism, and markedly irregular nuclear borders or prominent nucleoli, and low-grade when they formed papillary fronds, had an increased N/C ratio, and a slightly irregular nuclear shape, or showed numerous elongated cells with slight nuclear abnormalities, as previously described [bib_ref] Review of the state of the art and recommendations of the Papanicolaou..., Layfield [/bib_ref]. We categorised cytological results as positive or negative for high-grade urothelial tumour cells. Urine classified as high-grade or AUC-H were considered positive [bib_ref] Diagnostic terminology for urinary cytology reports including the new subcategories 'atypical urothelial..., Piaton [/bib_ref] , whereas normal, inflammatory, reactive, and degenerative urothelial findings were considered negative.
Histopathology. Tumour stage and histological grade were assessed according to the International Union Against Cancertumour, node, metastases system and the 2004 World Health Organization classification. Histopathology served as the gold standard for cancer diagnosis.
## Molecular testing.
A 25-ml sample of fresh urine was centrifuged at 800 g for 10 min and the cell pellet (urine sediment) was rinsed in PBS and frozen at À 80 1C until use. DNA was extracted from urine sediment using Circulating Nucleic Acid Kit QIAamp according to the manufacturer's recommendations (Qiagen, Hilden, Germany). Mutations of TERT promoter were analysed by nested PCR and Sanger sequencing. The first PCR (forward 5 0 -C ACCCGTCCTGCCCCTTCACCTT-3 0 and reverse 5 0 -GGCTT CCCACGTGCGCAGCAGGA-3 0 ) amplifies a 275-bp fragment that is used as matrix for a second PCR (forward 5 0 -CCC CTTCACCTTCCAGCTC-3 0 and reverse 5 0 -GCCGCGGAAAGG AAGG-3 0 ) amplifying a fragment of 118 bp carrying the points À 124 mutation (C228T) and À 146 (C250T). PCR products were then sequenced according to the Sanger method.
Statistical analysis. Quantitative variables were compared between groups using Mann-Whitney test and categorical variables using the w 2 -test or Fisher's exact test. Survival analyses were performed in the population of patients with positive TERT mutation at the time of TUBR. Analysis was restricted to superficial UBC stages, excluding pTis stage. Absence of recurrence was defined as no event with a minimum follow-up of 6 months. Median follow-up was 11.3 months (range: [fig_ref] Figure 1: Distribution of somatic TERT promoter mutation and positive urine cytology among tumour... [/fig_ref]. The Kaplan-Meier method was used to estimate recurrence-free survival (RFS), and curves were compared using the Log-rank test. Univariate and multivariate analysis were performed using Cox proportional hazard model with 95% confidence intervals (CI). All tests were set at the significance level of Po0.05. Statistical analyses were performed using the IBM SPSS Statistics software (release 19).
# Results
Recurrent somatic TERT promoter mutations in urine from UBC patients. Urine from 348 patients was collected at the time of initial TURB and tested for TERT promoter mutations. This cohort included 275 NMIBC (pTa or pT1), 61 MIBC (4pT1), and 12 carcinoma in situ (CIS). The overall TERT mutation rate was 80.5% (280 out of 348). TERT positivity in urine was stageindependent; sensitivity was 79.4 in pTa and 77.6% in pT1 NMIBC, increasing to 85.2 in MIBC (4pT1) and 91.7% in pTis.
There was a higher frequency of TERT mutations among highgrade tumours than among low-grade ones (P ¼ 0.0193); sensitivity in low-grade lesions was 74.3% [fig_ref] Table 1: Characteristics of tumours and patients with TUBR Abbreviations [/fig_ref]. Among TERT promoter hotspot mutations, C228T was the most prevalent (235 out of 280, 83.9%), followed by C250T (35 out of 280). There were also rarer substitutions in some patients (C228A and CC242TT), and in two patients concomitant mutations of both C228T and C250T.
Comparison of TERT mutations to cytology for the detection of UBC. TERT positivity was not significantly correlated with urine cytology classification [fig_ref] Table 1: Characteristics of tumours and patients with TUBR Abbreviations [/fig_ref]. Regardless of tumour stage, sensitivity of TERT mutations (280 out of 348, 80.5%) was significantly higher than urine cytology (117 out of 348, 33.6%) (w 2 -test Po0.0001). The overall sensitivity to detect UBC was for stratification of NMIBC in three groups (low-grade pTa, highgrade pTa/pTis and pT1) showed a sensitivity of cytology of 5.5%, 43.3%, and 50%, respectively whereas sensitivity of TERT remained high whatever the group (74.3%, 92.5%, and 77.6%, respectively; [fig_ref] Figure 1: Distribution of somatic TERT promoter mutation and positive urine cytology among tumour... [/fig_ref]. Of note, sensitivity of TERT in MIBC (4pT1) was not significantly different from cytology (P ¼ 0.0515).
## Specificity of tert mutation detection in urine.
To assess the specificity of this test we included 167 'non-UBC' patients including 125 with benign bladder lesion or healthy individuals, and 42 with other cancers [fig_ref] Table 2: Pathological data and TERT status in 167 urines without bladder cancer Abbreviation [/fig_ref]. Specificity was 92.0% (115 out of 125) in those with benign bladder lesion or healthy individuals, and 83.3% in cancer patients (35 out of 42). Among those with infectious or inflammatory urines (neurogenic bladder), only one presented a TERT mutation (specificity 96.3%, 26 out of 27). As prostate cancer could have been associated with TERT mutation, we also studied a group of patients with prostate cancer and without known UBC. In this group, specificity of TERT mutations was 87.9% (29 out of 33).
Detection of residual CIS after TUBR. Analysis of follow-up urines from the 348 patients found that in some cases TERT remained positive after TUBR whereas histopathology on the surgical specimen was negative for UBC, which could point to the presence of residual CIS. To test this hypothesis we analysed the association between TERT status and presence of pTis lesions detected during follow-up in patients with 'non-tumour' pathology post-TUBR (n ¼ 50; [fig_ref] Table 3: CorrelationAbbreviation [/fig_ref]. Among the 13 patients with recurrence of a confirmed CIS within 6 months after initial TUBR, 10 (76.9%) had follow-up urine (1 month post-surgery) that remained positive for TERT mutation, whereas among the 35 who did not have recurrence, 35.1% remained positive. TERT mutation was significantly associated with 6-month recurrence of pTis (P ¼ 0.0214).
Prediction of recurrence in NMIBC by TERT in urine. Patients presenting with NMIBC are known to be at risk of recurrence with more invasive tumours. We therefore evaluated whether TERT in urine could detect such recurrence, by analysing the association of TERT status with RFS in 100 patients with a minimum follow-up of 6 months and initially presenting a NMIBC without pTis (because it is known to increase risk of recurrence). Presence of TERT promoter mutation in urine was strongly associated with recurrence in these NMIBC patients [fig_ref] Figure 1: Distribution of somatic TERT promoter mutation and positive urine cytology among tumour... [/fig_ref].
Results of univariate RFS analysis showed that TERT mutation was associated with a risk of recurrence that increased 5.34-fold in the NMIBC subset (95% CI 2.11-13.55; P ¼ 0.0004). This association did not persist in multivariate analysis when associated with cystoscopy in the Cox regression model (HR 1.72, 95% CI 0.61-4.81 P ¼ 0.3015). However when patients were stratified on cystoscopy status, TERT mutation positivity was still significantly associated with recurrence , P ¼ 0.034) in the negative cystoscopy subset (46 out of 100 patients), despite only a few numbers of events (5 out of 6 were TERT mutated). TERT mutation did not provide additional information in terms of RFS in the positive cystoscopy subset (n ¼ 54; , P ¼ 0.9728).
Dynamic follow-up of NMIBC patients using TERT. We also observed that presence of TERT mutations in urine was a dynamic marker of recurrence. A frequent scenario is exemplified by the case presented in . This patient was initially treated by TUBR for a NMIBC classified as low-grade pTa. He presented a C250T mutation that became negative in the follow-up urines. Seven years later, TERT became positive again, associated with cystoscopic signs of recurrence. A tumour classified as high-grade pTa was resected. Since this second surgery (three years follow-up) urines have remained negative for TERT mutation and the patient has not experienced recurrence. Of note, urine cytology remained negative throughout progression of the disease, including initial tumour and recurrence.
# Discussion
Results of this large cohort study demonstrate that detecting TERT promoter mutations in urine is a non-invasive and sensitive way to detect UBC lesions, even of low-grade, where cytology is not sensitive enough. TERT may help to detect recurrence earlier and to better adapt follow-up frequency and treatment. We further showed that TERT remained positive after TUBR was significantly associated with residual CIS, which is difficult to detect by standard cystoscopy. This could explain why TERT remained a predictor of recurrence even in the negative cystoscopy group. Therefore TERT testing could help to better identify the group of patients for whom to consider hexaminolevulinate fluorescence cystoscopy. However, around 20% of patients did not show positive TERT in urine after initial TUBR and remained negative during followup. Because TERT is known to be reactivated by mechanisms other than mutations of its promoter, it may be of interest to study TERT expression and activity in these non-mutated patients [bib_ref] Comprehensive mutation analysis of the TERT promoter in bladder cancer and detection..., Hurst [/bib_ref]. Importantly we also found that detection of TERT mutation remained highly specific in inflammatory or infectious urines where previously described urinary biomarkers are known to give false-positive results [bib_ref] Analysis of false-positive BTA STAT test results in patients followed up for..., Raitanen [/bib_ref] [bib_ref] Urinary biomarkers for diagnosis of bladder cancer: a systematic review and meta-analysis, Chou [/bib_ref]. This is an important issue since this type of urine is frequent in non-UBC patients. Furthermore, TERT could also help clinicians to distinguish recurrence from inflammatory scar in case of suspicious cystoscopy.
The study does have some limitations. The main one being the single-centre design that could introduce some positive bias in analysing the performance of this marker.
A prospective study, investigating combination of TERT with FGFR3 and OTX1 as diagnostic urinary markers during follow-up of patients with primary NMIBC, recently confirmed the interest of this panel in patients with negative cystoscopy [bib_ref] ) FGFR3, TERT and OTX1 as a urinary biomarker combination for surveillance..., Beukers [/bib_ref]. It would be important to define the negative predictive value of TERT mutation as a single marker in case of suspicious cystoscopy (inflammatory lesions, scar post BCG therapy, etc.), and its positive predictive value in case of negative cystoscopy and cytology.
# Conclusions
Detection of TERT promoter mutations in urine is a reliable noninvasive prognostic marker for recurrence in UBC, especially in NMIBC where cytology does have some limitations.
## Acknowledgements
This study was supported by the French Ministry of Health . We would to thank Florence Morin for her technical support, Dr Pierre Sujobert for his constructive comments on this manuscript and Philip Robinson (Hospices Civils de Lyon) for editorial help. This work is published under the standard license to publish agreement. After 12 months the work will become freely available and the license terms will switch to a Creative Commons Attribution-NonCommercial-Share Alike 4.0 Unported License.
[fig] Figure 1: Distribution of somatic TERT promoter mutation and positive urine cytology among tumour stage in the 348 UBC cohort. Black bars, TERT mutated; HG, high-grade; LG, low-grade; White bars, positive cytology. [/fig]
[fig] Figure 2, Figure 3: Recurrence-free survival and TERT status. Kaplan-Meier curves for RFS probabilities according to somatic TERT promoter mutation alone (A) or associated with positive cystoscopy (B) and negative cystoscopy (C) in the NMIBC subset (n ¼ 100). Example of somatic TERT promoter mutation status and patient outcome. HG, high-grade; LG, low-grade; À , negative result. [/fig]
[table] Table 1: Characteristics of tumours and patients with TUBR Abbreviations: AUC-H ¼ Atypical urothelial cells of high grade; AUC-US ¼ atypical urothelial cells of undetermined significance; TERT ¼ telomerase reverse transcriptase. a P-values correspond to w 2 -test or Mann-Whitney test (age). b Atypical urothelial cells of undetermined significance (AUC-US) or cannot exclude high grade (AUC-H). [/table]
[table] Table 2: Pathological data and TERT status in 167 urines without bladder cancer Abbreviation: TERT ¼ telomerase reverse transcriptase. [/table]
[table] Table 3: CorrelationAbbreviation: TERT ¼ telomerase reverse transcriptase. a P-value corresponds to Fisher's exact test. [/table]
|
Environmental effects on the central nervous system.
The centrai nervous system (CNS) is designed to respond to the environment and is peculiarly vulnerable to many of the influences found in the environment. Utilizing an anatomical classification (cortex, cerebellum, peripheral nerves) major toxins and stresses are reviewed with selections from recent references. Selective vulnerability of certain areas to particular toxins is apparent at all levels of the CNS, although the amount of damage produced by any noxious agent depends on the age and genetic substrate of the subject. It is apparent that the effects of certain well known and long respected environmental toxins such as lead, mercury, etc., deserve continued surveillance. In addition, the overwhelming impact on the CNS of social damages such as trauma, alcohol, and tobacco cannot be ignored by environmentalists. The effect of the hospital and therapeutic environment has become apparent in view of increased awareness of iatrogenic disorders. The need for particular laboratory tests, for example, examination of CSF and nerve conduction toxicity studies, is suggested. Epidemics such as the recent solvent neuropathies suggest a need for continued animal studies that are chronic, as well as acute evaluations when predicting the potential toxic effects of industrial compounds.
Philosophers, historians, and scientists all respond to the effects of the environment, but each perceives it differently. An historical event is different to the man who reported it fresh than it appears to the historical researcher a century later. The facts of one generation become fancies or fables of the next. Each interpretation of the environment is unique for different individuals and in different cultures; and personal slant, bias, or pecadillo assures that each interpreter responds in his own fashion. Not only is the perception and reporting of responses complex and varied, but actual physical variability in responses to the environment is similarly complex. Even dizygotic twins exposed to the same toxin in the same uterus at the same time may vary in the amount of damage they display. Individuality and how it limits response, perception, or performance is not the subject of this monograph. Rather, this review reports what may happen to any vulnerable nervous system when it is exposed to a toxin or stress. The effect of environmental stress may well differ for each individual species, each personality, and each combination of genes, and this review can be concerned with only a small part of what has been reported to happen. Some effects of the environment are equivocal or *Neurological Association, Inc., 931 Chatham Lane, Columbus, Ohio 43221. uncertain, some beneficial, and a few are clearly deleterious. It is the deleterious effects that are of interest in this review.
In addition to the anatomical classifications used here, there are many potential ways to classify the effects of the environment on the CNS. Logically, the environment itself (trauma, drugs, water, radiation, etc.) could be the basis for classification. Instead, much of this review is classified by particular sections of the CNS, (cortex, cord, etc.) and uses subcategories to include lists of toxins. Other organ systems (heart, lungs, kidney) are not specifically covered, reflecting the neurologists' bias that all other organs are in fact only supportive to the CNS and indeed have purpose only if the CNS functions. It is more convenient to ignore the multiple disasters to the CNS when supportive systems do fail than to pretend to discuss these supportive systems. Since the CNS is designed to respond to the environment, responses are to be expected and these responses in turn lead to secondary and tertiary effects of each upon the other. Secondary ripples and interrelationships are not emphasized, though many probably deserve particular study.
Classification of CNS responses, in addition to the basic anatomical orientation used here, could emphasize selective vulnerability, critical periods, behavioral responses, genetic predisposition, or neuropathologic responses to injury. Selective vulnerability is discussed as an example of the potential importance of just one of these subjects.
## Selective vulnerability
Selective vulnerability suggests a special responsivity of particular areas of the CNS to certain agents and numerous examples abound from all the neurosciences. The polio viruses (which also once selectively decimated particularly defenseless population groups) selectively damage anterior horn cells. Some polio patients present with a failure of respiratory neurons, a few with rare conditions such as cerebellitis; but the major brunt of the attack is on the motor cells of the anterior horn in the spinal cord, while sensory neurons are completely spared. The clinical weakness is often asymmetrical, suggesting that even the anterior horn cells of each side of the cord are not assaulted in an identical fashion. The localization to the anterior horn cells is so prominent that the clinical features of this viral disease were identified and prevention anticipated before other viral disorders of the CNS were recognized at all.
Herpes simplex is another strangely selective agent which in vulnerable individuals does not stay confined to a "cold sore" on the lips or to a smouldering veneral disease, but rather spreads to the medial temporal lobes and produces severe and permanent necrosis. Damage to the temporal or limbic areas by this virus can permanently alter memory or emotions. Such temporal localization by herpes could be due to a biochemical affinity between the virus and the temporal or limbic areas but the exact cause for such localization is unknown. Other viruses demonstrate other tastes. For example, the cerebellar lobes in cats can be devastated by a viral panencephalitis; and the common chickenpox virus of humans can also lead to major destruction of the cerebellum. Perhaps even more unique (because it depends on a certain species, plus immaturity, and damage to a particular region), is the destruction of the ependymal cells which can follow infection with mumps virus. A remarkable series of projects by Johnson and others revealed that this virus can lead to a later hydrocephalus following destruction of the ependymal linings in the ventricles of suckling animals [bib_ref] Hydrocephalus following viral infections: pathology of aqueductal stenosis developing after experimental mumps..., Johnson [/bib_ref]. Selective vulnerability is not limited to the effect of viruses on a single level or region of the CNS.. Particular cells within a region may be responsive to certain drugs. Phenytoin can produce mild decrease in crispness of intellectual responses (3) in any patient, but occasional patients are so sedated by this generally safe drug that it cannot be utilized in therapy. The individual variability in response to phenytoin can reflect abnormal parahydroxylation of the drug in the liver, constipation which changes absorption, pregnancy, or even results from major genetic differences in metabolism. When vulnerable individuals receive excess phenytoin (for them and their metabolic pathways) there can be a transient or a permanent cerebellar ataxia. Therapeutic and moderate amounts of phenytoin may stimulate the same Purkinje cells that are damaged by larger amounts of the drug, and there has been informal suggestion that stimulation of the cerebellum by phenytoin is part of the explanation for its anticonvulsant effect. The phenytoin story is even more complex than this brief summary, however. The same drug can lead to an increased incidence of cleft palates in human offspring of epileptic mothers as well as in mouse embryos after pregnant mice are fed phenytoin. In the toxic clefting phenomenon, once again, a vulnerable creature is injured at a certain time.
There are other drugs as relatively specific as phenytoin plus many drugs that are as nonspecific in their effect as alcohol. Alcohol, most particularly when linked with dietary insufficiencies, can effect the peripheral nerves, the periventricular region, mamillary bodies, or even the cortex. It is hard to separate the effects of alcoholism from the effects of poor nutrition, trauma, and associated drugs; but many believe that excess alcohol itself can produce significant cortical deterioration. Abuse of alcohol can certainly be linked with cerebellar deterioration. The mechanism is uncertain since the afflicted persons have often suffered from dietary irregularities as well as alcohol abuse, but the general linkage of alcoholism and cerebellar symptomatology is clear. The ataxia and slurred speech of acute drunkenness is similar in pattern to the permanent deficit seen in patients with alcoholic cerebellar degeneration. A temporary metabolic derangement, drunkenness, is eventually succeeded by a permanent anatomical lesion, cerebellar degeneration. Even within groups of alcoholics the vulnerability appears widely varied. Red wine drinkers may develop deterioration of the corpus callosum, another group of imbibers develops peripheral neuropathy, and some of the alcoholics seem to manifest intermittent alcoholic blackouts or withdrawal seizures. The exact cause for the particular vulnerability of a certain individual is often obscure.
As alluded to above, selective vulnerability can also relate to a particular period in the developing CNS. Too much, too little, or too rapidly changing; all varieties of homeostatic disturbances can damage the developing organism. The damage from a Environmental Health Perspectives protein-caloric deficiency, for example, is profound when it assaults the cerebrum of an infant, although a similar effect may be hard to detect in an adult. X-ray insult to the CNS varies not only between patients (3000 rads leading to damage of the CNS in one patient, another patient almost untouched by 5000 rads), but the severity depends on the age of the subject. In the fetus an entire organ system, the cerebellum for example, may be vulnerable during early development but immune a few weeks later. The cerebral areas with the most rapid metabolism, greatest local response to the toxin, most tenuous support systems, or with preexistent injuryare most vulnerable. One general and repeated lesson from the CNS is that it is not only the character of the toxin but the overall state of the substrate that determines the response. Secondly, almost anything that produces a temporary effect can produce a similar permanent effect.
A related phenomena to selective vulnerability is the concept of critical periods. Critical periods implies staged development on a ribbon of time, and after passing a certain age marker the creature is no longer receptive or vulnerable. There may be only one precise period when development or learning can occur maximally. In birds, imprinting represents an obvious example. There is a short or finite time during which the baby duck can become attached either to its mother or, when mother is absent, to the keeper's foot. Imprinting, as and others have shown, can be changed by prior conditioning to meaningful but unnatural sounds. No matter how imprinting is studied, there is a time during which changes must occur if they are to be effective, and after a critical period the animal may never assume the prior level of performance or may be fixated on an inappropriate object. , in their now classic work on the visual cortex and lateral geniculate of kittens, demonstrated that if the animal is blindfolded early in life the lateral geniculate may never function adequately. The electrical field of the visual cortex also deteriorates to below the level present at the time light was excluded. Amblyopia exanopsia may be an example of a similar phenomena in humans. The retina may be normal, the cortex normal, but if one eye is deviated laterally long enough the image is permanently ignored and the function of that eye is irreversibly reduced.
Many psychologists, psychiatrists, and even moralists confirm that there are indeed critical periods of behavior in humans. Certain adjustments to the environment are appropriate at different ages and if not attended to properly and at the correct time, maximum ability may always be thwarted or lost. Any person has observed this in motor skills-there is an optimal time to learn to ski, or skate, ride a bike, or hit a baseball. Failure to develop and utilize a system when it matures can thwart maximum performance in adulthood.
Some of these examples of selective vulnerability and critical periods are more speculative than scientific, others are patently self-evident. There exist extensive data, not reviewed here, that suggest neurophysiologic responses of the CNS can be enhanced by usage, often studied with electroretinogram, electromyelograph, or nerve conduction. These neurophysiologic responses can be impeded by toxins and might be accelerated by early or enhanced demands.
This chapter organizes the effects of the environment on the CNS by an anatomical system and only secondarily by toxins or stresses.
## Relevant cranial nerves cn i (olfactory)
The sense of smell is a major force in human behavior, to judge from the volume of advertising by the perfume and soap industry, though the power of olfactory stimulation is repressed or unrecognized by most adults. In other creatures, odors affect sexual arousal and hormonal function, as documented in an immense zoological literature. Olfactory phenomena may also affect feeding behavior and even in humans can certainly heighten enthusiasm for drink or solid foods. Olfaction may be useful in primitive societies to avoid dangerous toxins, and it is possible that evolutionary benefit or personal survival was once supplied by a desire for cleanliness. Olfactory sensitivity diminishes with age, but even when it is total, anosmia is rarely considered a major handicap. The armed services usually allows only a ten percent disability for a total loss of sense of smell.
The olfactory nerve is probably the cranial nerve most commonly destroyed in head injuries [bib_ref] Abnormalities of taste and smell after head trauma, Schechter [/bib_ref]. Such rupture of olfactory nerve filaments at the cribriform plate is usually inconsequential for a well adjusted person. In addition to trauma and to congenital or to familial anosmia, acute viral illnesses can produce anosmia. It is possible that some metals (such as zinc) have a role in olfaction, but at present the relationship of olfaction to these agents is poorly defined. It is recognized that nasal, palatal, and other midline defects can result from teratogens, including particularly phenytoin and other anticonvulsants, but there are no data regarding an increased incidence of anosmia in children of epileptics who received anticonvulsants during pregnancy.
## Cn ii (eye)
The cornea can be scarred or damaged by infectious (viral, bacterial, or fungal) and by literally thousands of traumatic and environmental stresses. Certain traumatic agents such as pine or metallic dust can have special ophthalmologic significance. Trauma to the eye in sports and industry has been reduced by adequate helmets (football), waterproof glasses (prolonged swimming practice), and shielding glasses (industry). Metabolic insults to the eye include some of the aminoacidurias and galactosemia, neither a major public health problem but of medical importance since a few individuals with selected metabolic disorders can be helped by special diets.
It appears likely that anyone who lives long enough may develop a cataract of the lens and might also develop corneal opacities. There is no current metabolic explanation that explains the cause for the various familial types of cataracts, some of which have markedly different ages of onset, but environmental factors are not usually suggested as causal., The reduction in ability to tolerate glare (as on the highway) combined with a dislike for dim light is partly related to corneal and lens changes of aging and is so common that it deser-ves notice in lighting, highway, and vehicular designs whenever senior citizens are involved.
Cataracts related to diabetes, as well as other familial causes for cataracts, are of particular medical interest; but there is little firm evidence that modification of the environment accelerates or retards the development of such cataracts. Advice for a prudent diet, and in some families a trial of a low sugar or low fat diet, probably will not be completely preventive but is worth consideration. Food additives are reviewed in other chapters, there are no additives that are yet conclusively linked with significant ophthalmologic disease in man. Nevertheless, certain drugs, particularly phenothiazines in prolonged high dosage, can predispose to cataract (chlorpromazine) as well as to retinal changes (thioridazine) and other environmental factors may await discovery. Since the population group at greatest risk for iatrogenic (phenothiazines) cataracts is those patients in state institutions, and since medical care can be abominable in such institutions, routine ophthalmological monitoring can be considered for such populations [bib_ref] The ocular toxic findings with prolonged and high dosage chlorpromazine intake, Siddall [/bib_ref]. "'Drug vacations" have also been suggested as a prophylactic measure, although such therapus interruptus is rarely employed, even though the psychiatric effect of the drugs is long lasting.
In addition to responding to drugs the retina is also affected by prenatal infections, including some possibly preventable ones (toxoplasmosis, rubella) and others, some of which are theoretically preventable. A major epidemic of optic neuritis and myelopathy has been tentatively linked to clioquinol [bib_ref] Subacute myelo-optic neuropathy and clioquinol. An epidemiological case history for diagnosis, Meade [/bib_ref]. Other retinal toxic states such as those resulting from alcohol and tobacco might, probably very theoretically, also be preventable. Though rare, the presence of visual and neurological symptoms in tobacco-alcohol amblyopia does occur and has been related by some to cyanide toxicity from cigars. Similar visual symptoms due to cyanide toxicity may occur in African patients with exclusive diets of cassava root. Ethambutol, methyl alcohol, and similar toxins are currently rare causes of blindness, and in most of these cases the source of exposure is usually clear. There are some familial groups for whom the optic nerve may be at particular risk, for example in individuals who have inherited a tendency to develop Leber's optic atrophy. Though partially sex-linked, the inheritance pattern in Leber's atrophy is variable and blindness may result from the effects of environmental factors on a vulnerable genetic substrate. In view of this example, plus the well accepted changes secondary to phenothiazines, it is impossible to discount the role of as yet undiscovered toxins in "idiopathic" retinal deterioration.
The importance of "minor" eye problems, such as amblyopia exanopsia, has become more apparent with each decade, as reflected by the steady lowering of the age at which surgery is attempted on the eyes of afflicted children. The magnitude of failure of vision in the unused eye reflects the length of time the image from the deviant eye has been suppressed. Less certain, but recently suggested in Englandis the possibility that failure to develop steropsis and even overt dyslexia can originate from mild and undetected degrees of strabismus present very early in life.
Portions of the story of retrolental fibroplasia (RFL) can be used to illustrate the complex interaction of environment, predisposition, medical research, social, iatrogenic factors, and a resultant damage to vision. Over two decades ago it was apparent that RLF was a major cause of blindness in some areas (New York State) but almost nonexistent elsewhere (in Charity Hospital in New Orleans). RLF was clearly associated with prematurity; and attempts to expose the neonate to more or to less light, to vitamin A, and to more or to less antibiotics were all reported [bib_ref] Neurologic and psychologic assessment of institutionalized children with retrolental fibroplasia, Paulson [/bib_ref]. None of these factors were relevant. RLF represented a new medical disorder and one that could be recognized even during the period of its development. Eventually the role of new technology as the basic cause of disease became apparent as the link of RLF with Environmental Health Perspectives incubators that could supply 100 percent oxygen was discovered. The relative lack of RLF in blacks and in other groups was thus explained, they were just not treated in incubators in the same fashion. A simple preventative measure also became available-just do not give too much oxygen. Even after the cause was clear, a federal cooperative study was undertaken and there was a delay of several years before this disease of technology became fully publicized. There still remains need for investigation into the optimal level of oxygen needed to avoid RLF and yet supply enough oxygen to avoid a significant cerebral defect, since a deficit in oxygen can create cerebral palsy.
## Hearing and vestibular (cn viii)
The auditory and vestibular centers begin to develop in the fetus even before maturation of the visual system and perhaps the accelerated early development is partially due to environmental factors in utero. Sounds, including meaningful voices, are perceived early in life and are crucial for normal neural and emotional development. Clinicians often consider the psychological affect of profound deafness to be greater than that of profound blindness. Even for normals absence of sound contributes heavily to the effect of sensory deprivation. The role of excess sound in psychological development is more controversial, but there can be presumed to be an impact of background noise or music (or deliberate combinations of both) in our culture. There are suggestions that noise contamination affects behavior. Hearing, which can be permanently lost due to noisy occupations [bib_ref] Otology and industrial hearing conservation, Sataloff [/bib_ref] , may be transiently reduced by even one trip on a noisy rapid transit system [bib_ref] Acoustical study of a rapid transit system, Chang [/bib_ref]. In certain occupations which are characterized by either very noisy or very quiet working conditions, deprivation of meaningful sound might lead to features reminiscent of a sensory deprivation experiment.
Drugs can be otologically selective in that for some individuals the major neurologic side effects of these medicines is damage to CN VIII, as with streptomycin, minocycline, garamycin, etc. [bib_ref] Minocycline possible vestibular side effects, Nicol [/bib_ref]. In the fetus, thalidomide produced gross anomalies of the ear, as well as the better known phocomelia [bib_ref] Thalidomide anomalies of the ear, Takemori [/bib_ref]. Even alcohol can produce a transient decrease in hearing function, as in one recent study on pilots [bib_ref] Influence of alcoholic beverages on the vestibulo-ocular responses to coriolis stimulation, Dowd [/bib_ref]. It is possible that major insults to hearing are additive and monitoring of auditory functions may be wise for some individuals receiving ototoxic drugs.
## Cortex
Although recognized as a major problem in un-derdeveloped countries and during wartime, protein-caloric deficiencies and vitamin deficiencies are rarely a clinical problem in the U. S. [bib_ref] Brain biochemistry; effects of diet, Kolata [/bib_ref] [bib_ref] Protein-calorie malnutrition in children and its relation to psychological development and behavior, Latham [/bib_ref] [bib_ref] Neurological conditions resulting from prolonged and severe dietary restriction, Denny-Brown [/bib_ref]. Food faddism has become a major industry, however, and subgroups in our land not only avoid, but may favor unduly, certain food groups or substances-, and in rare individuals eating becomes a religion that interferes with optimal health care. Even in normals, very poor nutrition can deleteriously affect cerebral function, as in the various avitaminosis that still occur occasionally. Chronic iron deficiency is a common affliction of rich and poor and continues to exist even in today's affluent society. Excess nutrition in vitamins is of as much lay concern as the more obvious excess caloric intake of most Americans. Strongly held opinions about self-administrations of various megavitamins or iron preparations is a common basis for discussion in the physician's office, but fortunately with the exception of vitamin A or D intoxication [bib_ref] Chronic vitamin A intoxication, Teo [/bib_ref] , most fad diets and most mild deficiencies or excesses are probably harmless unless they deflect from more rational efforts to secure health. Varied dietary patterns in pregnant women do continue to deserve study, however, though conclusive linkages with toxemia, fetal wastage, etc. is unclear. Hypernatremia and similar therapeutic misadventures are preventable errors in nutritional support [bib_ref] A preventable cause of acquired brain damage?, Chambers [/bib_ref].
Infection can affect not only the cortex but also any other portion of the nervous system. Susceptibility to infections depends on prior exposures plus general resistance, and certain groups at particular risk will continue to deserve surveillance. Diseases in young military recruits (meningoencephalitis) (24) warrants continued study, and there is much to be learned from many special and potentially epidemic situations encountered in residential schools or state hospitals. Animal reservoirs of infection in the environment deserving regular surveillance are pigeons (cryptococcus) (25), other birds (sparrows or starlings and St. Louis encephalitis), or domestic animals (porcine streptococcus or influenza) [bib_ref] Porcine streptococci causing meningitis and septicemia in man, Zanen [/bib_ref]. Brucellosis [bib_ref] Brucellosis in the United States 1960-1972. An abattoirassociated disease, Buchanen [/bib_ref] , an abattoir-associated disease, is often associated with headache or visual symptoms. Public education may be needed regarding cherished pets, as in toxocara canis infestation or psiticoccus [bib_ref] Toxocara canis infestation with encephalitis. Can, Mikhael [/bib_ref]. The entire pattern of CNS infection may be changing; as is the medical awareness of certain infectious disorders. The prolonged effect of continued intracranial activity of the rubella virus is one example, as is the unusual manifestation of a measles-like virus seen in SSPE. A slow course following viral infection, or a "slow virus" disease, remains at least one conceivable potential effect of a program of innoculation with attenuated virus. The increased aware-ness of "slow viruses" in the CNS demands continued caution in the planning for the release of such attenuated viruses for vaccinations. There have been reported cases of transfer of the viral degenerative state called Jakob's disease to individuals who received corneal transplants from persons wvho suffered from this virus [bib_ref] Possible person-to-person transmission of Creutzfeldt-Jakob Disease, Duffy [/bib_ref].
Brain tumors can be produced experimentally by several organic compounds, particularly the nitrosoureas. In contrast to some types of bladder tumors, where over 50%o seem to be a response to environmental toxins, there is little fim data regarding brain tumors in humans. Radiation therapy and chemotherapy, perhaps in a summative fashion, can lead to new neurologic syndromes [bib_ref] Radiation therapy of brain tumors in children, Onoyama [/bib_ref] [bib_ref] Postradiation motor neuron syndrome, Sadowsky [/bib_ref].
According to Caveness, accidental death and disability in our time has achieved a magnitude comparable to that of the plagues of the middle ages; with the cortex and brain as major recipients of the trauma (32). There were 8,111,000 head injuries in the U. S. in 1974, and over 1,900,000 of these were serious, with 6,600,000 days of hospitalization. Caveness estimates that 140,000 of these cases will develop seizures that will be persistent (7% of serious head injuries). At times new neurologic disorders (subdural, epidural, hygroma) other than seizures also occur months after injury [bib_ref] Delayed onset of acute post-traumatic subdural effusion, Gutierrez [/bib_ref]. The incidence of post-traumatic seizures reflects constitutional and genetic factors, the extent of injury, and may be affected by prophylactic use of anticonvulsants [bib_ref] Pharmacologic prophylaxis of post-traumatic epilepsy, Rapport [/bib_ref] or antiedema agents. The deficits are by no means limited to seizures, and many of the same patients have dementia or focal sensory and motor loss. The amount of intellectual function relates largely to the extent of trauma and to the quality of the brain before injury; but environmental and educational factors may also affect recovery. Relatively simple measures, perhaps some not yet considered, can reduce or prevent the extent of injury. As early as three decades ago, crash helmets were recommended for cyclists [bib_ref] Head injuries in motorcyclists: the importance of the crash helmet, Cairns [/bib_ref] [bib_ref] Crash helmets reduce head injuries, Jamieson [/bib_ref] [bib_ref] Comprehensive health care and motor vehicle safety. New EngI, Fielding [/bib_ref] , but they are still not universal. Indeed, head injuries occur in as high as 30%o of motorcycle accidents [bib_ref] Injury patterns in motorcycle collisions, Drysdale [/bib_ref]. Reduction in speed limit, improvement in driver skill, optimal design of vehicles, and protective devices are all relevant health concerns. Less clear is the long term effect of minor trauma such as repeated brief unconsciousness in teenage football players. Recognized and unsuspected birth trauma is generally accepted as a major cause of seizure disorders, but precise figures are hard to obtain. As the major environmental danger to life in the young, the potentially most economically shattering of illness, a cause of major and distressing morbidity, and one of the least studied of CNS disorders the incidence and type of major and minor trauma of the brain and spinal cord deserves attention in any evaluation of the health of the environments available in this country.
Convulsions of other types than postraumatic include febrile seizures, which may not be as innocuous as once was supposed. There is data from recent "kindling" studies that suggests that the threshold for future seizures can be lowered by repeated subclinical seizures [bib_ref] Spontaneous recurrent seizure state induced by daily electrical amygdaloid stimulation in sengalese..., Wada [/bib_ref]. Complaints regarding the delay in release of anticonvulsants has been commented upon by both national neurological organizations (American Neurological Association and American Academy of Neurology) and intensive survey of potential benefits as well as the side effects of all anticonvulsants is highly desirable in view of their widespread need and use. The risk of death and injury in uncontrolled epileptics is significantly increased [bib_ref] Unexpected, unexplained death in epileptic patients, Terrence [/bib_ref]. Certain special phenomena, as convulsions after burns [bib_ref] Postburn convulsive disorders in children, Mcmanus [/bib_ref] , or due to certain unusual toxins (such as Borax) [bib_ref] Seizure disorders and anemia associated with chronic borax intoxication, Gordon [/bib_ref] , or from drug withdrawal may have significance for the environmentalist; both in research and in education. The complexities of "epilepsy" can be illustrated in many ways. One example is infantile spastic hemiplegia [bib_ref] Acute hemiplegia of childhood, Tibbles [/bib_ref]. This condition may follow a high fever, particularly in a child who has developed marasmus or dehydration due to illness or neglect. Focal neurologic disease such as flaccid hemiparesis gradually evolves into spasticity with seizures and these in turn are affected by the availability of medical care. Each new case is unique and special, but attention to general health care, use of adequate anticonvulsants and an optimal educational environment may make the difference between a disgruntled crippled epileptic and a useful taxpayer.
Specific toxins that affect the cortex and CNS are almost innumerable, and their effects marvellously variable. Brief, selective, and very incomplete mention is offered as a reminder of the variety, and attention should be called to the extensive recent handbook for details. Lead causes encephalopathy in children and usually a neuropathy in adults, although encephalopathy can occur [bib_ref] Lead encephalopathy in adults, Whitfield [/bib_ref]. Possible behavioral effects have been attributed to subclinical levels of lead in the environment. Mercury usually causes a neuropathy, but can produce both agitation and tremor. Mercury can be obtained through industrial exposure and as a pollutant in water. Mercury has been recently linked to disease in special population groups, including North American Indians. Arsenic is notorious for its presence in illicit alcohol and in concoctions from murderous relatives, but may also be increased in subclinical amounts in farmers who use insecticides. Thallium is a rare poison now, due to elimination from over-the-counter solid rat poisons. Manganese deficiency has a controversial role in production of tremors or seizures in alcoholism and in other disorders, and toxicity from excess manganese can lead to cortical changes following prolonged exposure to manganese in animals and in humans. Monosodium glutamate is an almost ubiquitous agent that is rarely linked with disease in humans, but has been an experimental cortical toxin [bib_ref] Brain damage in neonatal mice following monosodium glutamate administration, Lemky-Johnston [/bib_ref] [bib_ref] Effect of dietary monosodium L-glutamate on some brain and liver metabolites in..., Prosky [/bib_ref]. Glutethimide [bib_ref] Neurologic effects of glutethimide, Haas [/bib_ref] and similar soporifics can represent very difficult problems when presenting as an overdose and any emergency room can testify that these and similar agents are very readily available in the community. Jimson weed contains organic toxins with marked anticholinergic affects and the plant is not controllable in the fields, although proper medical education may prevent death from the severe anticholinergic poisoning that results from ingestion of the plant. Hexachlorophene can cause damage to neuronal systems in experimented animals such as rats (50), swine [bib_ref] Experimental hexachlorophene intoxication in young swine, Robinson [/bib_ref] , and possibly in humans [bib_ref] Neurotoxicity of hexachlorophene in the human. I. A. clinicopathological study of 248..., Shuman [/bib_ref] [bib_ref] Hexachlorophene-Not a cry of "wolf, Catalano [/bib_ref] [bib_ref] Increase in cerebral fluids in rats after treatment with hexachlorophene or triethyltin, Lock [/bib_ref]. Marijuana has been linked to an "amotivational" syndrome. Hypoglycemia and hypernatremia are often both preventable causes of cerebral damage in infancy. Hypocalcemia in the fetus, particularly when associated with low calcium levels in the mother, may also be preventable [bib_ref] Maternal factors in the aetiology of neonatal hypocalcaemia, Watney [/bib_ref]. Carbon monoxide requires additional study regarding its effect through prolonged subclinical exposure [bib_ref] Carbon monoxide poisoning, report of an outbreak involving 171 persons, Saslow [/bib_ref] [bib_ref] Fetal brain damage following maternal carbon monoxide intoxication: an exherimental study, Ginsberg [/bib_ref]. There are reasonably good laboratory techniques available for study of this common environmental hazard. Vitamin A, as is well known, can produce severe cerebral edema when taken in excess.
Alcohol may well be the most all pervasive and widely used noxious drug in America. Even before birth, effects of alcohol abuse can be manifested in the "fetal alcohol syndrome" which includes malformations, small size, retardation, and other anomalous features in children born of alcoholic mothers [bib_ref] A prospective study of the fetal alcohol syndrome, Ouellette [/bib_ref] [bib_ref] The fetal alcohol syndrome, Jones [/bib_ref] [bib_ref] Prominent lateral palatine ridge; developmental and clinical relevance, Hanson [/bib_ref]. Alcohol predisposes to trauma of the brain (as contusion, subdural, epidural) and to peripheral nerve injury by entrapment or direct injury. Alcohol is a major cause of acute seizures in adults; in one study over 40% of acute admissions to hospital were abusers of alcohol [bib_ref] Seizure admissions to a city hospital: the role of alcohol, Earnest [/bib_ref]. Combined with vitamin deficiency (Korsakoff-Wernicke Syndrome, cerebellar atrophy, etc.) or in withdrawal states (tremor, fits, delirium tremors) the secondary results of alcohol affect all regions of the CNS. Furthermore, for many drugs metabolism is changed in acute alcoholic states [bib_ref] Mechanism of the effect of acute ethanol on hexobarbital metabolism, Chung [/bib_ref]. The alcoholic is at greater risk for infection, malnutrition, deprivation, and medical misadventure; but there well may be no truly useful environmental recommendations at this time (See recommendations below).
Additional research is indicated in the management of withdrawal states from alcohol and other drugs, as well as in study of other toxins associated with alcohol and in their combined neurochemical effects. Reasons for the high incidence of alcoholism in particular groups (Indians, Irish) or low incidence in other groups (Jews), continues to be worthy of study. Continued study of the direct effects of alcohol alone, as well as in clinical settings in which there is little malnutritional and vitamin deficiency, as in alcoholic dementia, seems indicated.
The largest single group of understudied neurologic disorders is the presenile dementias. Alzheimer's disease is a major public health problem, and up to 25% of state hospital patients over 60 suffer from this disease. Possible environmental causes (slow infection or aluminum intoxication) deserve increased study in the next decade.
## Occipital lobe
Among the special groups of patients who are at particular risk from the environment are residential students in institutions for the blind (some of which house both blind and deaf). Endemic infections, intoxications, and poor medical care can occur in these settings and in any similarly closed institution. In addition the eventual intermarriage of students is very common, and such marriages can lead to complex sortings of genetic disorders. Since many inherited forms of deafness or blindness are recessive, the offspring of these matings are often normal, and the severely handicapped and frequently devoted parents create a special environment for the child. There has been little involvement of organized medicine in such institutions even for the education of medical students. The results of efforts, at times almost punitive, of teachers or house mothers to eliminate unwanted features which are common in the behavior of the blind ("blindism," stereotypes such as rocking, awkward gait) has not usually been considered a social or pediatric problem. It is possible that this particular environment offers unique opportunities for behavioral and epidemiological studies.
The cortex and the white matter, can respond with demyelination to certain anesthetics, such as halothane (in utero) [bib_ref] Pathological changes in the nervous system following in utero exposure to halothane, Chang [/bib_ref] , as well as to recognized, overlooked, or inadvertent anoxia. Speculation regarding primary environmental factors in multiple sclerosis (MS) has abounded, partly because of the well known geographical distribution of the disease.
Influenza or enteric viruses have been implicated, without certain proof. Immunologic features of the patients with MS may well be contributors to the development of the disease. Much patient interest exists in special diets (gluten-free, lecithin, unsaturated fats, vitamin E), but there is no substantial evidence that special diets produce, predispose to, or ameliorate MS.
In the first task force report, there is an interesting allusion to vibrating tools producing damage of the CNS, and more recent references exist [bib_ref] Vibration syndrome in lumbermen (working with chain saws), Laitinen [/bib_ref]. Except in entrapment neuropathies, vibration in the environment is usually considered more nuisance than danger.
Some acknowledgement is necessary regarding the alarming increase in overdose problems of all kinds. Certain drugs, such as the atropine group (tricyclics, anti-Parkinsonian) may respond to particular antidotes such as physostigmine. These drugs when used even in small doses can affect memory, although deleterious long term effects are unclear. The development of drug treatment or information centers have been useful resources for the practitioner of medicine and new clinical problems or varieties of old ones will continue to deserve publication and study by directors of such drug treatment centers. There is no current convenient and rapid monitoring system to alert the medical community of a new form of drug intoxication or overdose.
## Basal ganglia
Barbeau has suggested that Parkinsonism represents an aging phenomena that can be accelerated by toxins, infections, or trauma [bib_ref] Parkinson's Disease: etiological considerations, Barbeau [/bib_ref]. It is certainly true that the appearance in "normal" aging includes many extrapyramidal features, and it is probably not surprising that numerous toxins (particularly metallic ones) appear to cause accelerated dysfunction of the basal ganglia. Several of these toxins deserve special mention. Wilson's disease has a relationship to copper, can be diagnosed with relevant laboratory tests, and medical and dietary measures are often useful. For subjects at risk, (siblings in this recessive disorder), monitoring of copper metabolism in the preclinical state can be justified with enthusiasm. Less is known about several other rare diseases of the basal ganglia. Hallervorden-Spatz disease involves deposits of iron and can lead to a severe progressive dystonia. Fahrs' disease includes calcification in the basal ganglia. Manganese miners can develop a dystonic form of Parkinsonism.
Phenothiazines produce complex neurologic disorders, often of great interest medically and legally. An acute dystonic state can be seen, and seems to be more common in the young. Parkinsonism, or "pseudo-Parkinsonism" can develop after several weeks of therapy with phenothiazines and is reported to be more common in patients with a family history of Parkinsonism. Tardive dyskinesia is related to prolonged use of phenothiazines and is particularly likely to occur in patients who are old and in those patients who have organic brain disease [bib_ref] Drug induced extrapyramidal syndromes, Delay [/bib_ref]. Amphetamines, so commonly employed in the hyperactive child (73), can elicit not only anorexia but also choreiform movements [bib_ref] Methylphenidate-induced chorea, Weiner [/bib_ref]. Tardive dyskinesia has been seen in children [bib_ref] Tardive dyskinesia as a possible sequel of long-term therapy with phenothiazines, Paulson [/bib_ref]. This iatrogenic disorder, called tardive because of the long delay in developing symptoms after exposure, may be present in 15% of patients in state hospitals, but despite its frequency is often unknown to the practicing physician. A source of quiet and at times acrimonious controversy ten years ago, tardive dyskinesia is now well accepted and is an excellent example of some of the reasons for a delay in the discovery of any iatrogenic disease [bib_ref] Clinical psychopharmacology in its 20th year, Crane [/bib_ref]. ("It is the disease, not my medicine," or "the side effect is useful," or "the virtues of the medicine compensate for the side effect," or "don't discuss it, do you want a law suit or something?") Drug vacations and use of the minimal effective dosage may help prevent tardive dyskinesia.
Carbon monoxide can lead to a Parkinsonian state or even neuropathy [bib_ref] Carbon monoxide intoxication with peripheral neuropathy, Snyder [/bib_ref] [bib_ref] Carbon monoxideinduced parkinsonism, Ringel [/bib_ref] in rare patients who are exposed to an overdose, but more usually produces transient coma or cortical damage. The effect of continued elevated levels of carbon monoxide in the environment is a matter of controversy, but it would be difficult to prove that such levels are completely innocuous.
There are environmental and genetic aspects of infectious, and of traumatic or postinfectious states in diseases that affect the basal ganglia, in addition to Sydenham's chorea. After the encephalitis lethargica epidemic in 1918, complex neurologic disorders were seen and have been considered likely to return with some future epidemic. Environmental contributors to basal ganglia diseases are not totally clear; each severe new epidemic of influenza will need study regarding long-term effects of infection on the CNS.
## Brain stem
Though the brain stem is of great interest to the neurologist, there are few environmentally induced disorders that affect this region primarily. Thiamine deficiency has been postulated to explain a rare inherited disorder of childhood (Leighs' disease), and extreme thiamine deficiency in the alcoholic can be associated with petechial hemorrhages in the brain stem or with myelinolysis. Once a major cause of mental disease (pellagra), in this country avitaminosis is not considered a significant problem currently except for alcoholics. Multiple sclerosis can affect any portion of the CNS, including myelin in the brain stem; and a viral etiology, or an interaction of viral, immunological and environmental stresses is currently postulated as the major etiologic factor.
The persistent vitality of the brain stem and technological advances in medicine have lead to an interesting recent financial, medical (81), societal, and legal interface; the concept of brain death. Other than public education in the significance of trauma as a health hazard and of the need for renal transplants, there is no obvious responsibility for the environmentalist in the issue of brain death.
## Cerebellum
There are numerous viral agents and chemical toxins that can lead to a temporary cerebellitis or to permanent cerebellar damage [bib_ref] Toxic cerebellar degeneration, Valsamis [/bib_ref]. In most, other than the epidemiology of infection or genetic predisposition, there is little of note for the environmentalist.
Panencephalitis can lead to cerebellar damage in cats and destroy or damage an entire colony. The mumps virus, particularly in the very young, can produce changes in hamster, rat, or man that damage subcerebellar regions, damage to the aqueduct causing hydrocephalus, or damage to the cerebellum itself. Some epidemics of chickenpox are strikingly associated with signs of cerebellar dysfunction, but what produces such variation during epidemics of viral disease is unclear.
The cerebellum, particularly the Purkinje layer of the cerebellum, is very vulnerable to anoxia as well as to hyperthermia. Educational efforts to encourage fluids and limit hot clothing in summer athletic training and for unseasoned military recruits has reduced the incidence of heat stroke [bib_ref] Heat stroke. New Engl, Clowes [/bib_ref]. The familial disorder of malignant hyperthermia can produce a fatal result during anesthesia of at risk subjects [bib_ref] Maligant hyperpyrexic myopathy Quart, Harriman [/bib_ref]. Though changes occur in muscles during and even between attacks in this entity, there is no completely reliable enzyme test for this uncommon and at times fatal entity of malignant hyperthermia. Elevations in temperature of any cause can lead to cerebellar damage that is irreversible. Cerebellar degeneration resulting from toxins other than alcohol does occur (85) but has been hard to investigate, partially because of the lack of definitive laboratory tests. Repeated and severe seizures, particularly in the setting of anoxia, will probably produce a loss of Purkinje cells, and cerebellar degeneration can certainly result from therapy with phenytoin in both experimental and clinical situations. Since phenytoin and some antimetabolites such as fluorouracil can damage the cerebellum selectively, the possibility of unidentified cerebellar toxins in the environment cannot be dismissed.
## Spinal cord
The spinal cord can be damaged by any local infection, even unrecognized ones such as lues, and by other rare infectious agents such as toxoplasmosis. Specific occupational or avocational situations, such as "the bends" in aviators or scuba divers, may lead to injury of the cord [bib_ref] Central nervous system involvement following type I aviators bends complicated by complacency, Dully [/bib_ref]. Environmental features are more obvious in some of the allergic responses of the cord, such as in the myelopathy that can result from vaccinations (rabies or smallpox) [bib_ref] Epidemiology of antirabies treatment in Georgia, Currier [/bib_ref] [bib_ref] The neurology of jennerian vaccination, Spillane [/bib_ref] [bib_ref] Neurological complication of antirabies vaccination in Sao-Paulo, Brazil. Clinical and therapeutic aspects, Assia [/bib_ref]. Some venoms, including bee sting or spider venoms [bib_ref] Red back spider venom and inhibitory transmission, Einhorn [/bib_ref] [bib_ref] Inhibition of creatine phosphokinase activity and synaptic transmission by black widow spider..., Chmouliovsky [/bib_ref] , can produce severe allergic neurologic responses. Motor root, brachial plexus, peripheral nerve and even the cortex (94) may also be involved when the cord is affected by serum sickness and similar inflammatory reactions. Metabolic disorders such as pernicious anemia can selectively affect the cord, and some conditions such as amyotrophic lateral sclerosis (ALS) may have associated metabolic derangements in pancreatic or liver function in the late stages, although the basic cause of ALS still remains completely obscure. A more obvious factor is iatrogenic assault on the cord and its arachnoidal coverings, as in intrathecal injections of contrast material or medications (95), some of which (such as steroids for multiple sclerosis) are demonstrably worthless but can quite successfully produce arachnoiditis with secondary pain and motor or sensory deficits.
Odontoid fracture [bib_ref] Fracture of the odontoid process of the axis, Husby [/bib_ref] and cervical compressive fracture is all too common as a diving or trampolene [bib_ref] Injuries to the cervical spine from diving accidents, Kewalramani [/bib_ref] [bib_ref] Surfing injuries at Waikiki, Allen [/bib_ref] injury and every large community hospital is likely to see at least one each summer. Strenuous exercises of the neck and protective neck supports have probably greatly reduced the incidence of similar severe neck injury in football contact, but spinal cord injury can and does still occur in athletic events. Even minor trauma of the neck can lead to infarction of the cord in children [bib_ref] Spinal cord infarction due to minor trauma in children, Ahmann [/bib_ref] and unrecognized obstetrical trauma to the spinal cord or plexus (100) accounts for numerous cases of cerebral palsy each year.
Triothocresyl phosphate (TOCP) is one particular toxin that has been used extensively in study of pathologic responses of the cord and peripheral nerve [bib_ref] Histoenzymic studies of the central nervous system in tri-ortho-cresyl phosphate poisoning, Wender [/bib_ref] [bib_ref] Effect of TOCP poisoning on the pacinian corpuscles of slow loris, Krishnamurti [/bib_ref]. A major epidemic with this agent occurred during prohibition (104) (Jamaica Ginger and "Jake paralysis"), but exposure is now uncommon in the U. S. though still reported overseas (Bombay, Morocco) from contaminated oils or foodstuffs. The clinical picture may represent a combined neuropathy and myelopathy with a spastic gait in the setting of a neuropathy.
In the past ten years there have been suggestions of a possible impact of the potato blight fungus in producing defects such as meningomyelocele and encephalocele. There is wide variation in the incidence of these defects in different parts of the world, but toxic environmental linkages (105, 106) remain obscure. Preliminary data (107) suggest that it is now possible to diagnose the defects antenatally from aminiotic fluid and fetal research in the next decade may elucidate the cause of these disorders.
Radiation myelitis is an all-too-common complication of therapy with large doses administered to the spinal area [bib_ref] Permanent radiation myelopathy, Locksmith [/bib_ref]. There are no careful studies at this time regarding the combined effect of chemotherapy and radiation.
## Alutonomic disturbances
Certain people with autonomic dysfiunction are particularly vulnerable to environmental stresses, for example children with the Riley-Day Syndrome may not tolerate anesthesia or underwater swimming. Except for postural hypotension, there are few other examples of autonomic disturbances in clinical medicine, but many drugs, including overthe-counter soporifics, contain agents that exert major effects on the autonomic system. Some of the more available herbal hallucinogens have marked autonomic effects [bib_ref] Anticholinergic poisons, treatment with physostigmine, Rumack [/bib_ref] and effects on memory plus a ready treatment [bib_ref] An analysis of the learning deficit following hyoscine administration to man, Crow [/bib_ref].
## Peripheral nerves
Agents that produce a neuropathy may produce such an effect (1) indirectly or by damage to vascular or supportive tissue, (2) by damage to Schwann's cells and secondary segmental demyelination, or (3) by injury primarily to the neuron and then a "dying back" at the axon. The signs of a peripheral neuropathy are relatively stereotyped, with a decrease in motor power distally, decreased deep tendon reflexes, and decreased sensation. There is some suggestion of a difference in the clinical presentations after exposure to different poisonings, for example poisoning by arsenic is usually more painful than the neuropathy of porphyria, and the latter neuropathy may be more motor in pattern. Any neuropathic insult may be additive, and several toxins tend to summate in their effect. In general, however, both the nerve pathology and the neurologic examination in cases with neuropathy fail to be completely specific and for this reason this section will be organized by noxious agents, not by region or by pathologic change. Allen's review (113) is a particularly good summary, organized by categories of toxins.
## Metals
Lead: Lead as an industrial hazard (114) is less common now than in the last century, but environmental exposure has included foundries, mines, paint plants, and plastic industries which use lead stearate. Hobbies [bib_ref] Hobbyists working with lead, Feldman [/bib_ref] , foods, and homemade wines, even pet foods [bib_ref] Lead in pet foods and processed organ meats, Hankin [/bib_ref] can be contaminated. Residency near industries that use lead, old paint or house dust, colored inks, and illegal whisky are among the sources of exposure mentioned in recent literature [bib_ref] Lead encephalopathy caused by ingestion of illicitly distilled whiskey, Morris [/bib_ref] [bib_ref] Lead poisoning. Amer, Guinee [/bib_ref] [bib_ref] House and hand dust as a potential source of childhood lead exposure, Sayre [/bib_ref]. The habit of pica is important [bib_ref] Prevention of pica, the major cause of lead poisoning in children, De La Burde [/bib_ref] , but only partially explains the high lead levels found in inner city children [bib_ref] Undue absorption of lead among children: New look at an old problem...., Lin-Fu [/bib_ref]. There are new and useful diagnostic tests including measurement of aminolevulinic acid dehydratase activity in erythyrocytes. The enzyme activity approaches zero with 100 Aego of lead in the blood. Blood levels (significance is 40 ,ug% or greater) of lead can also indicate undue subclinical exposure. Other useful tests include urine measures of coproporphyrin and the presence of basophilic stippling in the peripheral blood smear; and once lead intoxication is diagnosed there are treatments available [bib_ref] Ambulatory treatment of lead poisoning, report of 1,155 cases, Sachs [/bib_ref]. The clinical picture of lead intoxication includes encephalopathy ranging from impaired alertness to delirium and coma, and in adults a motor neuropathy can also occur. There is a predisposition of the child with sickle cell disease to develop neuropathy after lead intoxication [bib_ref] Lead neuropathy and sickle cell disease, Erenberg [/bib_ref]. Alcoholics may also be more vulnerable than nonalcoholics to this toxin. Even the basic mortality rates are higher in lead workers than in the general population [bib_ref] Mortality of lead workers, Cooper [/bib_ref]. Animal work has suggested that subclinical lead poisoning can impair brain function and there is a suggestion of increased incidence of hypertension in patients with high lead levels [bib_ref] Blood lead and hypertension, Beevers [/bib_ref]. Possible correlations between subclinical lead intoxication and impaired cerebrovascular, mental, or psychological performance remains controversial [bib_ref] Blood lead levels, behavior and intelligence. Population study, Lansdown [/bib_ref].
Mercury: Mercury in the inorganic form can lead to mild neurologic disorder, but the clinical effect of organic mercury compounds can be devastatingly severe [bib_ref] The neurotoxicity of alkyl mercury compounds. Amer, Dales [/bib_ref]. The sources vary, and organic mercury has been found in some fungicides (see grain) [bib_ref] Methyl mercury poisoning in Iraq. A neurological study, Rustam [/bib_ref] , in fish, and in industrial plants (used as a catalyst for vinyl chloride) [bib_ref] Mercury as an environmental hazard, Vostal [/bib_ref]. There is even an increased hazard in dental offices [bib_ref] An environmental study of mercury contamination in dental offices, Schneider [/bib_ref]. Other countries have reported the clinical features in se-vere cases including weakness or spasticity in late stages (Minimata Disease in Japan), muscle atrophy, and even a pattern like ALS, plus sensorineural hearing loss, decreased vision, or a marked agitated delirium [bib_ref] Minamata disease: an unusual neurological disorder caused by contaminated fish, Mcalpine [/bib_ref] [bib_ref] The outbreak of a neurologic disorder in Minamata, Japan, and its relationship..., Kurland [/bib_ref] [bib_ref] Behavioral effects of mercury and methylmercury, Evans [/bib_ref] [bib_ref] The involuntary movements of chronic mercury poisoning, Snyder [/bib_ref]. Diagnostic tests for mercury include direct measurement of body products, plus electrical studies of nerve conduction [bib_ref] Electrophysiological investigations of methylmercury intoxication in humans. Evaluation of pheripheral nerve by..., Burg [/bib_ref] and when present in toxic levels, a high level of mercury can be found in samples of liver and kidney by biopsy as well as in the cerebellum and brain or by autopsy [bib_ref] Mercury in the human brain, Olszewski [/bib_ref]. Methylmercury can pass the fetal barrier, but in general young animals tolerate higher doses [bib_ref] Toxicity of methylmercury: effects on different ages of rats, Lin [/bib_ref]. The pathogenesis of mercury poisoning may include binding of sulfhydryl groups of proteins and the effect is very long lasting [bib_ref] Neurological change in cats following long-term diet of mercury contaminated tuna, Chang [/bib_ref]. Recent studies have indicated a high incidence of poisoning in those Indians in Northern Canada who eat large amounts of fish caught from contaminated waters [bib_ref] Methyl mercury intoxication in northwestern Quebec, Barbeau [/bib_ref].
Manganese: Manganese is only a sporadic cause of industrial poisoning, usually in the manganese or ferromanganese industries [bib_ref] Clinical manganism and exposure to manganese in the production and processing of..., Smyth [/bib_ref] [bib_ref] Chronic manganese intoxication, Cook [/bib_ref]. Large numbers of cases have been reported from Chile and India, but sporadic cases are reported from the U. S. [bib_ref] Elevated manganese levels associated with dementia and extrapyramidal signs, Banta [/bib_ref]. The major route of exposure for manganese is via inhalation of dust. The absolute levels in muscles, skin, blood, hair, and brain have been measured and of particular interest in the late development of a central disorder which is not just a peripheral neuropathy, but includes a fundamental disorder of the basal ganglia. The onset is 4-15 years after initial exposure and can include a severe mental disorder, a state similar to Parkinsonism, plus a severe dystonic syndrome. There is some suggestion of an increased absorption of manganese in association with anemia and nutritional deficiency. It is thought that manganese may produce some of its damage by a possible defect in neurotransmitter, since in the cases studied there is reduction in brain dopamine and homovanillic acid (HVA) as well as low levels of homovanillic acid in the cerebrospinal fluid. L-Dopa has been useful for symptomatic management.
Thallium: Thallium has become an uncommon poison now that rat poisons are more limited and since its depilatory qualities are not utilized cosmetically [bib_ref] Thallium intoxication treated with dithizone and hemodialysis, Paulson [/bib_ref] [bib_ref] Thallium poisoning, Bank [/bib_ref] , but it can still present as a puzzling case of neuropathy. Diagnosis can be suspected after the development of alopecia and confirmed by levels of the agent in urine or hair. Other metals such as tin, zinc, etc. may be crucially important in CNS functions, but are rarely indicted as cause of disease.
## Solvents
n-Hexane: n-Hexane poisoning has been found in printing plants, paint factories, furniture repair shops, and in industrial plants that use glue [bib_ref] Polyneuropathy due to n-hexane, Paulson [/bib_ref] [bib_ref] Pathology of lacquer thinner induced neuropathy, Means [/bib_ref]. Most cases have been associated with varnish or with a solvent used for rubber cements. Glue sniffers and "huffers" may also develop a neuropathy which could be due to hexane [bib_ref] Toxic polyneuropathy due to glue sniffing, Goto [/bib_ref] [bib_ref] Glue sniffer's neuropathy, Towfighi [/bib_ref] [bib_ref] Glue sniffing neuropathy, Korobkin [/bib_ref]. Measurement requires samples of the atmosphere and levels above 500 ppm may be elevated sufficiently to produce damage. Serum pseudocholinesterase may be reduced in affected persons. The clinical findings included an insidious neuropathy, largely motor and largely distal. Some of the glue sniffers have developed more profound disturbances including mental changes; but most of these patients have been exposed to multiple toxins in addition to the n-hexane. There is no specific therapy. Samples of environmental air should be checked to diagnose this neuropathy and since the condition is usually mild, and recovery complete; ventilation, avoidance of overtime, and other simple measures may be all that is required prophylactically. Methyl Butyl Ketone: Methyl n-butyl ketone (MBK) [bib_ref] Neuropathy and methyl n-butyl ketone, Mendell [/bib_ref] [bib_ref] Medical news: solvent causes motor neuropathy in workers at clothing factory, Anonymous [/bib_ref] is usually obtained by inhalation in plants using the agent (often required for glued fabrics), and MBK may be found mixed with other agents. A recent outbreak resulted from substitution of MBK for other solvents after the oil crisis produced a change in availability [bib_ref] Toxic polyneuropathy produced by methyl n-butyl ketone, Mendell [/bib_ref] [bib_ref] MBK neuropathy among spray painters, Mallov [/bib_ref] [bib_ref] Peripheral nerve changes induced by methyl n-butyl ketone and potentiation by methyl..., Saida [/bib_ref] [bib_ref] Methyl n-butyl ketone, Mcdonough [/bib_ref]. The clinical findings include an insidious neuropathy with paresthesias and distal weakness. Axoplasmic flow can be affected [bib_ref] Methyl n-butyl-ketone-induced changes in fast Axoplasmic flow, Mendell [/bib_ref]. There may be a common metabolyte for-n-hexane and MBK. The diagnosis is nonspecific but requires clinical suspicion plus a check in relevant plants for levels of MBK in the atmosphere as well as in agents used in manufacture [bib_ref] Peripheral neuropathy in a coated fabrics plant, Billmaier [/bib_ref] [bib_ref] Neuropathy and methyl n-butyl ketone, Duckett [/bib_ref]. Electromyography can be a useful screening procedure to detect subclinical cases of neuropathy.
Toluene: Toluene is usually considered nontoxic, but some changes in cerebral function have been suggested after chronic exposure to this agent [bib_ref] Permanent encephalopathy from toluene inhalation, Knox [/bib_ref]. One of the major research areas that remains is to determine possible synergisms between toluene, n-hexane, and other solvents. Toluene is such a commonly used agent that interactions with other compounds and possible changes in the chemical structure during use or in the atmosphere deserve careful evaluation.
Trichloroethylene: Trichloroethylene, in addition to damage to other organs, can produce a cranial neuritis or more diffuse CNS disorder [bib_ref] Trichlorethylene: a review, Smith [/bib_ref] [bib_ref] Polyneuritis cranialis associated with industrial trichloroethylene poisoning, Buxton [/bib_ref].
## Organic chemicals
Cogeners of phthalate with higher alcohols are Octoberl977 8S used in the leather or fabric industry. The clinical features of toxicity include a mixed sensory and motor neuropathy, with occurrence only after several years of exposure. Acrylamide is a much studied monomer [bib_ref] Six cases of acrylamide poisoning, Garland [/bib_ref] [bib_ref] The pathogenesis of dying-back polyneuropathies, Prineas [/bib_ref] [bib_ref] Modification of acrylamide neuropathy in rats by selected factors, Kaplan [/bib_ref] [bib_ref] Acrylamide neuropathy in rats, Suzuki [/bib_ref] [bib_ref] Motor end-plate fine structure in acrylamide dying-back neuropathy: a sequential morphometric study, Tsujihata [/bib_ref] [bib_ref] The assessment of peripheral neurotoxicity in dogs: Comparative studies with acrylamide and..., Thomann [/bib_ref] that is used in inks, and overdose can produce a rather mild neuropathy. There is considerable experimental data, and it has been found that a distal axonal neuropathy can be produced by acrylamide in most animals studied. The fundamental effect of the compound may be to produce a defect in protein synthesis.
Triorthocresyl phosphate (TOCP) and related compounds can produce a mixed neuropathy and myelopathy. It has been said that TOCP inhibits pseudocholinesterase.
Mipafox [bib_ref] Paralysis following poisoning by a new organic phosphorus insecticide (Mipafox), Bidstrup [/bib_ref] and similar organic phosphorus pesticides can produce an acute and profound paralysis or death [bib_ref] Neurological manifestations of organophosphorus insecticide poisoning, Wadia [/bib_ref] [bib_ref] Organophosphates -a pediatric hazard, Mayer [/bib_ref]. In some areas there seems no doubt but that the workers appear better informed about these agents than are the physicians.
Kepone can lead to severe tremors in individuals who receive toxic doses, and the variety of CNS effects is dwarfed only by the complex social and legal issues in the recent overdoses [bib_ref] Neurologic disorder induced by kepone: preliminary report, Taylor [/bib_ref].
The effect of herbicides in the CNS remains uncertain (177-179) and perhaps minimal, but since teratogenic effects are known, neuropathies may eventually be reported.
## Gases
Carbon disulfide, now used in the manufacture of viscose rayon, in the past century produced hundreds of cases of intoxication with acute and chronic encephalopathy, neuropathy, etc [bib_ref] Toxicology of carbon disulphide, Brieger [/bib_ref]. Modem ventilation systems and stringent regulations have largely eliminated this disorder.
Carbon monoxide is discussed above (cortex). Methyl chloride is a foaming agent in the plastics industry that can produce transient ataxia, unsteady gait, and confusion [bib_ref] Chronic methyl chloride intoxication in six industrial workers, Scharnweber [/bib_ref].
Methyl bromide has been reported to produce persistent myoclonus [bib_ref] Methyl bromide poisoning, Goulon [/bib_ref] but also might be expected to produce a neuropathy.
## Other causes of neuropathies
There are numerous other therapeutic substances that can produce a neuropathy, including dapsone [bib_ref] Dapsone neuropathy-A disorder of motor axons, Gutmann [/bib_ref] , chloramphenicol, metromidazole, clioquinol, hydantoins, disulfram, isoniazides, furans, thalidomide, vinca alkaloids, etc. Overdose with other agents, in addition to vitamins, can dam-age the CNS. Some of these various compounds may offer metabolic clues to other varieties of neuropathy and myopathy.
Any standard neurologic test lists dozens of metabolic causes for neuropathy, the most common one of which is diabetic neuropathy. Some have limited environmental or local interest, such as Tangiers' disease which was first reported from an island in the Chesapeake, but none are of primary environmental concern at this time.
Tick paralysis occurs in certain parts of the country but is rare indeed, as are the neuropathy and acute respiratory problems secondary to botulism. Both these forms of paralysis are probably preventableand both have educational significance for professional and laymen. The environmental importance of the viral neuropathies is harder to assess and many viral illnesses seem associated with muscle weakness and aches. Occasionally Guillain-Barre syndrome can seem to affect several persons in the same area, but probably secondary to viral illness rather than to hidden toxic exposures. Preexisting debility or other unknown factors may be additive in such cases. The recent postulated linkage between Guillain-Barre syndrome and the vaccine for swine flu may further accelerate investigation of the response of the CNS to allergic insults of all types.
The entrapment neuropathies include compressive neuropathies at any area where nerves can be pushed or pinched. "Bulb digger's" paralysis, a peroneal palsy, probably involves not only a squat occupation position but a familial predisposition. "Gunslinger's" or "postman's' palsy of the brachial plexus is also an occupational hazard for a few. A common tardy nerve palsy is that secondary to resting on an elbow (ulnar nerve) and is common in telephone operators, debilitated persons who rest on their elbows, and in habitues of saloons.
There are numerous basic elements which, when either in excess or deficient in the environment, can produce neurologic disease. Copper is an obvious example (189), but even silicon [bib_ref] Silicon as an essential element, Carlisle [/bib_ref] [bib_ref] Recent dietary trace element research, exemplified by tin, fluorine and silicon, Schwarz [/bib_ref] has been demonstrated to be necessary for the development of supportive tissue. These have been largely studied from the hematologic point of view and are not well understood as regards to significance or prevalence of neurological disease. There are, in addition to viral neuropathies, many infectious agents that produce not only systemic symptoms but the clinical picture of an acute neuropathy. Diphtheria is now largely controlled by educational and epidemiological advances, but botulism still produces severe illness and fatalities in intermittent outbreaks [bib_ref] Ocular involvement in benign botulism, Konig [/bib_ref] [bib_ref] Current trends in botulism in the U, Merson [/bib_ref] [bib_ref] Botulism: ten year experience, Cherington [/bib_ref].
## Environmental health perspectives
## Muscle disease
The entire area of causal factors in muscle disease is almost as complex as that of the CNS. Walton's recent text is a good reference source.
## Vascular disorder
It is hard to overestimate the significance of cerebral arteriosclerosis in the CNS as a cause of disability and death. The major preventable contributors to this killer includes hypertension, diabetes, smoking, hypercholesterolemia, and possibly elevation of triglycerides. Recently the quantity of metals in water has been reported to correlate with vascular disease [bib_ref] Cardiovascular mortality, municipal water and corrosion, Schroeder [/bib_ref]. The clinical features of arteriosclerosis of the cerebral vessels will not be reviewed here, but it is clear that diagnosis and treatment is inconsequential in importance when compared to the need for prevention. Nevertheless, other than control of blood pressure and a prudent diet, there is at this time unfortunately no prevention to offer most patients.
Bickerstaffand others have reviewed the role of oral contraceptives in vascular disorders of the brain. In addition to an effect caused by a vasculitis or thrombosis, there is also a tendency for the production of hypertension after use of these agents.
Aneurysms and vasculitis can result from heroin addiction and particularly when impure compounds are injected intravenously [bib_ref] Intracranial mycotic aneurysms and subacute bacterial endocarditis in heroin addiction, Gilroy [/bib_ref].
Migraine attacks can be produced in vulnerable persons by diet, coffee, chocolate, tryamines, histamines in red wines, hypoglycemia, etc., but there is no general or universal toxin in the environment (except weather) which is currently indicted in attacks of this common malady. For some, specific environmental stresses can be very relevant, and even minor trauma [bib_ref] Footballer's migraine, Matthews [/bib_ref] can be the major environmental cause of such headaches in susceptible individuals.
Patients who have had strokes are more vulnerable to thrombosis [bib_ref] Deep venous thrombosis of the legs after strokes, Warlow [/bib_ref] , but in addition, certain compounds, diazepam and phenytoin included [bib_ref] Thrombophlebitis with diazepam used intravenously, Langdon [/bib_ref] , may predispose to thrombophlebitis during IV therapy; but other than alerting the physician there is no apparent responsibility for the environmentalist in this area.
## Subject at risk and other comments
There is a great deal that can be stated, much of which can be only philosophical, about the hazard of becoming a patient. The fact that this report does not review such hazards does not indicate their nonexistence. It is instructive to be reminded however, as Crane has done, how extremely insidious iatrogenic disease can be. There are obvious as well as unconscious reasons why iatrogenic disorders are suppressed, missed, or ignored. There are many new therapeutic situations in which the beneficiary is at great risk. All the antimetabolytes might produce reduced vigor and lead to secondary infection. Radiation is similar in overall effects to the antimetabolites, and if one deals with people more vulnerable in several different ways, the risks of radiation become even greater. The variety of responses obscures unpleasant facts of medical care, since every individual has critical or vulnerable periods, as in very early or very advanced age, and with any disease the resistance to environmental stress can be lessened. In any occupation the new worker may be exposed to unexpected hazards of toxins, most of which are fortunately mild. This review has suggested that there is selective vulnerability of the CNS to certain agents and toxic agents can affect quite specific regions of the CNS. These effects are not random and may indicate unexplained biochemical affinities between the toxin and the neuronal cells. Every new medicine is a potentially new CNS toxin. In the CNS two insults are often additive and produce a greater effect than the apparent sum of both. A partially damaged, depleted, or malnourished CNS can withstand far less anoxia or toxic damage than the normal. Nerve cells do not regenerate; once gone they do not return. Tumor, radiation, chemotherapy, and extreme age could be "additive" in their effect, and so monitoring for the effect of the toxin is very difficult. This is not to say that therapeutic toxins should be as severely monitored as optional or coincidental toxins. It is more acceptable to develop a side effect from life saving therapy than to develop a mild reaction caused by an unnecessary and undesired contaminant.
Certain population groups are at particular risk, including the very young, adolescents (trauma), recruits (new infection), alcoholics (almost any physical insult is additive in presence of alcoholism), and retarded or institutionalized persons. Noxious agents such as carbon monoxide may damage the fetus even if the mother is unharmed. Halothane or other anesthetics could also possibly do the same. Therefore, in any survey of vulnerable groups the fetal population deserves particular attention.
## Behavior and related areas
Television has been postulated to cause an increase in violence, paranoia, and hyperactivity in children [bib_ref] The scary world of T.V.'s heavy viewer, Gerbner [/bib_ref] [bib_ref] Violence, television, and the health of American youth, Somers [/bib_ref]. Since there may be no educational source as all pervasive as TV, continued studies of its effect on behavior and physical health is mandatory. At present, common sense suggests that TV must be counteracted by other stimulation and that its use does not eliminate and may increase the need for other forms of cerebral activity than passive absorption of visual images.
The child with "hyperactivity" is a major pediatric problem, and this controversial and heterogeneous diagnosis includes many entities, some of which are not medical. As high as 1% of school children in some areas are on amphetamines for hyperactivity, and the long-term effects of these drugs is of major interest. For example, there are some data to suggest a syndrome akin to tardive dyskinesia in children on such medications, and a relative failure of growth has also been postulated in children who receive amphetamines for long periods of time.
## Screening tests
One major educational area for the public health service concerns study of the utilization of health services. It may become desirable, or necessary, to decrease not increase, the utilization of health professionals, and to encourage more use of medical judgment by home authorities. One sure protection from iatrogenic misadventure is to avoid the doctor when the problem is trivial. A more critical, even negative, view of screening programs of all kinds might result from intensive evaluation of the effectiveness of prior major screening efforts. There is strong feeling in some members of the medical community that routine or massive screening programs are likely to be inefficient, expensive, worthless, or all three. Alertness to toxins and other environmental stresses must include willingness to proceed with selective screening, but massive screening is not a substitute for intelligent search, an informed observer, and selective screening.
## Tests and the environment
No one test will give information on all neural processes in the human, but perhaps behavior and language come closest. In humans, language is a unique tool, both for history taking and to obtain insights into patient behavior from family members. The standard psychiatric tests of mental function are often less revealing than reports of behavior that are offered by social workers, family, or the patient himself. Language can also be a precise monitor as in new techniques such as the Porch Index of Communicative Ability. Since behavior is the feature that may be most affected when the cortex is insulted, it is logical to turn to psychological measures for quantification. Although MMPI, trail making tests, Bender-Gestalt, flicker fusion, and similar tests can be helpful, all can be influenced by psychological factors, and none is without blemish.
No test can substitute for thoughtful evaluation of the total problem. Similarly, though the neurologic exam may be the best measure to detect significant damage to the nervous system (particularly peripheral neuropathy) the neurologic examination only screens sensory, motor, cerebellar, and mental performance and is not a substitute for synthesis of the entire problem, nor is a neurological exam a replacement for a scholarly review of the clinical possibilities.
In some disorders, as in damage by drugs such as the phenothiazines which can affect either cornea or retina, a specially designed ophthalmological evaluation will be helpful to indicate the extent of tissue damage. Slit lamp or indirect ophthalmological techniques will often be required. In these techniques there is no substitute for a knowledgeable ophthalmologist. Similarly, as for example with some inhaled toxins, particular liver function tests or liver biopsy may document damage as well as does any single neurologic technique. Certain enzymes (CPK, aldolase) are of particular value in indicating muscle dysfunction, and others less commonly employed (cholinesterase) may be a clue to toxins such as TOPC or n-hexane.
Changes in cerebrospinal fluid can often suggest that the CNS has been damaged. Protein values are elevated in many, probably most, severe neuropathies, and yet toxicity studies in animals never sample CSF. In certain conditions one or another component can be studied, such as homovanillic acid levels in manganese intoxication, and when indicated these may be of particular interest. Dozens of enzymes can be detected in the CSF and yet there is little study of the effects of toxins on these CSF enzymes or of states in nature when these enzymes are deficient.
Neurophysiologic approaches, especially sophisticated electromyography, can be particularly useful in detecting peripheral nerve and muscle damage. Careful study of nerve conduction rates as well as evaluation of miniature endplate potentials (MEPP) is of value in some disorders. The routine electroencephalogram (EEG) is probably useful only in extreme toxic states, but with frequency analysis and evoked response techniques, additional insights can be expected from this readily available clinical tool. Other neurophysiologic techniques such as electronystagmography (ENG) or electroretinography (ERG) are of value when the optic nerves or retina is affected. These techniques Environmental Health Perspectives can be used in animal toxicity studies as well as in human research, but rarely are.
There is at least one incompletely tapped resource in neurologic research; natural animal models. Not only in man, but in animals (particularly mice), many genetic disorders have been identified and pure strains of these animals are available. Study of the influence of noxious substances on these creatures is not in vogue, but since difficulties in the CNS are often compound and may be more than additive, some animal models might be used to detect the presence of trace amounts of toxins.
One generality regarding tests is that no test substitutes for intelligent observation. Tests should be selected depending on suspicions, foriexample it is logical to suggest assessment of hearing in noisy environments. EMG may be a useful screening technique to detect the neuropathy caused by n-hexane, when a plant uses this compound. Since this neuropathy may occur without patient complaint, random screening in these plants could be justified. EMG will not, however, be a substitute for accurate atmospheric check. The purpose of the neurologic tests should thus be to confirm, and occasionally discover, the effect of a noxious situation. Thoughtful utilization of multiple tests, including biochemical ones, may then indicate precisely what the noxious material is, how to eliminate it, and perhaps even prevent future exposures.
## Summary, comments, and recommendations
The central nervous system (CNS) is designed for appropriate response to the environment and is peculiarly vulnerable to many influences found in the environment. Selective damage to particular areas of the CNS is noted with numerous toxins, and certain individuals are considerably more vulnerable than others.
Unfortunately many of the influences on the CNS are hard to isolate, and sometimes the toxic agents produce changes in behavior or intellectual function that are difficult to study. The CNS responds differently than other areas: for example, the blood-brain barrier tends to protect the CNS from conventional carcinogens, although an increased overall mortality and incidence of brain tumors following exposure to vinyl chloride has been reported [bib_ref] Mortality study of workers in the manufacture of vinyl chloride and its..., Tabershaw [/bib_ref] [bib_ref] Mortality experience of a cohort of vinyl chloride-polyvinyl chloride workers, Nicholson [/bib_ref]. The production of tumors by certain carcinogens, as nitrosoureas, is under active investigation in experimental laboratories, but it seems probable that the incidence of clinical malignancy in the CNS is less influenced by environmental agents than is malignancy elsewhere in the body. On the other hand, production of malfor-mations in the fetus can follow administration of many compounds, though systematic investigation of the mechanism of such effects has been limited. In the CNS the time of the administration of the toxin is as important as the nature of the toxin. For example, at certain stages of fetal development, radiation may damage the cerebellum; at another stage the same dose of radiation damages the neuronal cells in their migration from the ependyma to the cortex.
Even after the fetal period there is great variability of response, particularly between individuals. Phenytoin, and other hydantoins, may cause ataxia at one dose in one individual, while at a similar blood level not even clumsiness is evident in a second individual.
Other drugs, or toxins, or even pre-existing disease, may contribute to sensitivity, but individual and largely unknown metabolic differences are also crucial. The same infection may produce a fatal hydrocephalus in fetal animals of two species (hamster, rat) but will not always do so after the same exposure in apparently identical siblings of the same species. To detect the effect of toxins in any particular region of the CNS may require special techniques, but in many instances, as in the examples above, such definitive techniques are not available until after the final neuropathological assessment.
Peripheral neuropathy is perhaps the most obvious common result of environmental toxins in adults and can be detected clinically as well as neurophysiologically. Neuropathy in industries, or in avocational exposures, may be noted first in isolated patients and after such affected persons are identified then screening for neuropathy in population groups at risk (selective screening) should be encouraged. Screening may be required to locate subclinical and less severe cases. To this date, skilled neurologic examination and detailed electromyography remain the most useful screening tests for neuropathy.
Methyl n-butyl ketone (MBK) was discovered to be toxic after an outbreak of peripheral neuropathy was diagnosed by intensive combined efforts of scientists, union, and government agencies. This toxin had been used in a plant making coated fabrics, and the solvent had been required because of sudden changes in availability and costs of related chemicals. In such acute outbreaks, alert or suspicious workers, physicians, or management must first prove the presence of a problem and then at great expense of time and effort identify the specific culprit. MBK could have been identified as toxic in animals, before the human epidemic, if extended chronic studies had been done in animals. None had been done. At times an epidemic elsewhere, as in Minimata disease due to mercury in waters of Japan, or, closer at hand, as reported in 1976 in Indians in Northern Canada, will alert government agencies of a need to survey similar environments or groups in the U. S. Manganese intoxication, more common in Chile than in the U. S., is another example of a disorder that undoubtedly occurs in the U. S. but can be studied more thoroughly in another region.
Chronic lead poisoning has been associated not only with reduction in longevity, but with neuropathy and even psychologic disorders and hyperactivity in childhood. The toxins that can damage the nervous system number in the hundreds, as reviewed in excellent recent texts such as those by Dyck and Walton. Since so many chemical compounds can produce a peripheral neuropathy, this clinical phenomenon can be used to suggest the presence of a toxic effect in the environment or in the body. It is certain that as yet undiscovered toxic or metabolic causes of neuropathy await detection since a specific cause for at least 25% of cases with neuropathy remain unexplained even after very detailed evaluation. The very young, the old, the debilitated, and those with metabolic predisposition are particularly likely to be affected by smaller doses of noxious agents than the usual worker in the usual occupational setting.
RECOMMENDATION 1: ENVIRONMENTAL Tox-INS AND DRUGS All evaluation of drugs or other potential toxins should include chronic as well as acute studies. MBK could have been detected as a toxic agent if chronic studies had been performed, although the acute experiments revealed no abnormalities.
Check of specific toxins in specific industries or environments should be encouraged. Lead is an obvious example. Mercury levels should be evaluated in groups of fishermen and in heavy consumers of fish. Discovery of any new toxin may automatically imply the need for a survey of relevant groups by government health agencies, as well as by the most proximate investigators.
Hobbyists at risk (furniture remodeling, lead soldiers, etc.) should be warned routinely, preferably in published journals of these groups as well as on labels of the relevant agents. The national trend toward more detailed labeling of all products including drugs should be encouraged and education in schools should include instruction in how to understand labels.
Research on carcinogens in experimental animals should be combined with a survey of the incidence of tumors of the CNS in groups that appear at risk. It is important to discover why the brain seems less vulnerable to environmental carcinogens. latrogenic disease and the difficulty in detecting it is of significance in the nation's health. For example, the phenothiazines produce at least three varieties of disorder in the basal ganglia. One of these, tardive dyskinesia, was particularly difficult to accept as genuine, at least partly because the dyskinesia was slow to develop and because of the iatrogenic aspect. Since these and other chemical agents can produce unpredicted long-term and even permanent effects on the CNS, prolonged exposure to other agents such as carbon monoxide must also remain suspect, even when exposure is at a low level.
Disease related to use of drugs is a major problem in diagnosis and treatment of new hospital admissions. A high proportion of new admissions are on unknown drugs. Automated measures of enzymes, blood counts, electrolytes, etc. have revolutionized medical care in the past two decades, but there is no similar routine check for serum levels of therapeutic and nontherapeutic drugs. Even cases of lifethreatening coma due to overdose can at times be hard to identify because of a delay in laboratory reports. Intensive research into the effect of ubiquitous or commonly used drugs, especially when used for children, such as amphetamines, aspirin, soporifics, and so forth, should be encouraged.
Specific drug labels should be made mandatory on all prescription drugs unless specifically ordered otherwise.
Package inserts of all drugs should be available to patients.
At times "epidemics" appear as sudden new diseases or as alarming new or previously unrecognized manifestations of old diseases. The recent concern with swine influenza is an example. Some neurologists have anticipated a possible return of post-viral Parkinsonism following this or similar infections. There is no obvious or generally recognized and easily mobilized mechanism for the study of unique or transient short-lived health phenomena. The most readily available research source is usually the local medical school or local investigative physicians. At times, Public Health Services can be the major resource and must be involved in the education efforts during such episodes. There might be an ongoing mechanism to rapidly initiate research efforts in strange, unique, or transient new health problems, but now there is none. By no means will all these unexpected transient episodes be infectious or toxic, though acute infectious diseases and new toxins often have environmental and health significance of long lasting importance. RECOMMENDATION 3: STUDY OF NEW PHE-NOMENA Current coordinative groups in the study of outbreaks of encephalitis should include military and public health groups. Utilizing the experience of such groups improves the rapidity of response in short-lived health phenomena.
There are individuals at particular risk in the environment. For example, adolescents sustain more trauma, smokers more lung infections, and aged or institutionalized patients may be more vulnerable to various social changes or medical fads. Furthermore, even for the most hardy of persons there are environmental situations in which there is unwanted risk. One such environment that is certain to receive increased research interest is the hospital. Steadily rising costs may lead to innovative new and safer approaches, but intensive physiologic monitoring and an exposure to drugs and chemicals can be anticipated by most who enter the hospital. The anxiety and medical efforts in intensive care units had been shown to produce extreme exhaustion in the patient and loss of REM sleep in regular wards and may be almost as severe a stress. The impact of any particular environment on health, particularly those environments not intensively studied, can be assumed to be underestimated unless disorders are obviously and overtly caused by the environment. RECOMMENDATION 4: THERAPEUTIC EN-VIRONMENT There should be encouragement in research on the influence of the hospital environment (state hospitals, intensive care units, differences in ward or private rooms, etc.). Such studies should include the interference of this environment with sleep and dignity, as well as an assessment of expense and overt environment hazards (such as electricity in coronary patients).
Hospitals should forbid smoking except in a few designated areas. The role of smoking is vascular and pulmonary diseases should be included in all health education.
The most hazardous feature in the U. S. environment is trauma. Trauma in the streets due to social disorder is inconsequential compared to that on the highways due to deliberate speed. Seizures, dementia, and focal neurologic deficit can all result from traumatic insults to the brain and literally tens of thousands of cases are produced annually. Less understood is the effect of lesser degrees of trauma on later intellectual performance. There is no question that athletic injuries of the knee can affect future function of the joint, but there is no certainty regarding the equally common repeated minor cerebral traumata of athletic youth. Certain groups, the retarded for example, are at a particular risk. Future studies might properly be concerned with athletics as an environmental stress on brain and body. RECOMMENDATION 5: TRAUMA When a "sport" is suspected to be particularly hazardous (hand gliders, skateboards, competition of children on motorcycles, etc.) there should be formal (legal or health) hearings to consider limitation of such activities, as was done for fireworks decades ago. Such reports are often performed at State or National Parks, and the Park Service may be a good reference source.
Research on the effects of trauma on the brain and spinal cord should continue to be encouraged, particularly research aimed at preventive measures. The interrelationship between acute trauma, occupation, therapeutic chemicals, and previous trauma continues to be worthy of investigation.
The fledgling specialty of "sports medicine" should be encouraged via special symposia, particularly in regard to changing patterns of athletics for women and in subjects at particular risk, as the retarded child.
An expert in sport medicine or orthopedics should be included on the board of trustees and professional advisors of all athletic organizations that are licensed or incorporated, as NCAA, Little League, etc.
Since the ototoxicity of noise is potentiated by certain drugs, such as diuretics or aminoglycosidic antibiotics, audiometric studies should be made of workers exposed to noise while taking medication that is potentially ototoxic.
Swimming lessons as a part of the public school curriculum should be encouraged, although it is unlikely this would reduce the incidence of drowning or diving injury.
It may be unnecessary to mention the insidious and devastating effect of alcohol on the nervous system. Alcoholics suffer more trauma and produce more trauma than other members of society. Alcoholics are particularly vulnerable to other toxins and detection of a second toxin may be difficult in this group of patients. : ALCOHOL there should be continued study of the effects of alcohol as a major toxin as well as studies of the effect of alcohol in the vulnerable or weakened person.
The mechanism of drug withdrawal is not well understood, and research in this area might lead to therapeutic advances. Withdrawal phenomena are similar for all major drugs that effect the CNS, but we fail to understand the process.
There should be studies of the effects of laws on alcohol abuse. For example, what would happen if bartenders or hosts were held responsible for injuries to intoxicated persons served by them? What would happen to salesmen if they were held liable for injurious actions caused by minors who are sold the alcohol? There are probably more varieties of chemical, neurophysiological, and clinical tests on the nervous system than on any other organ. Numerous unusual metabolic or genetic diseases can be confirmed only by special tests, many of which have helped elucidate our understanding of the nervous system. It is impossible to state with certainty that investigation using any one test will solve a neurologic puzzle. One reason the concept of screening or monitoring is raised so often is that many toxic or environmental phenomena first came to attention through an isolated or special case and only then numerous other cases were discovered. Since there are so many potential tests, however, surveys should be aimed for results, and good judgment in where and what to study remains as crucial as the test itself. RECOMMENDATION 7: DETECTION OF NEW EN-VIRONMENTAL PROBLEMS Continued support for the training of scientific physicians and academic specialists is mandatory. The push to train practitioners of medicine should not be allowed to suppress the creative or investigative spirit.
It might be possible, as hospitals become more innovative [bib_ref] No frills hospital, Anonymous [/bib_ref] as well as more closely monitored by federal agencies, to select a few representative institutions to maintain careful records on all admissions for: trauma, overdose, drug side effects, toxic states, industrial accidents, and alcoholism. No matter how well intentioned, however, collection of such data should not be initiated without clear purposes.
Use of electromyography, spinal fluid protein, and nerve biopsy should be encouraged in toxicologic research in animals, in an effort to detect latent neuropathy.
There should be encouragement of the development of newest physiological techniques (miniature endplate potentials, etc.) for detection of subclinical effects of noxious agents. Major neurologic organizations should be encouraged to support symposia on the effect of the environment on the brain and behavior.
Several of the above recommendations have concerned education for scientists or public. Education is a a proper role for almost any tax supported agency and is also part of every physician's responsibility. What education and what style of education for the public deserves additional formal consideration by environmental groups. Planning for such education may reduce duplication and oversight.
The major preventable causes of vascular disease; namely hypertension and tobacco, deserve particularly thoughtful research and educational efforts. Actual mechanisms of injury by these agents remain incompletely understood, but some therapeutic changes in style of living in the U. S. can already be suggested. Sloth, overindulgence of food or alcohol, and tobacco are obvious old enemies. The continued search for hidden dangers must be combined with assault on known ones. This material is drawn from a Background Document prepared by the author for the NIEHS Second Task Force for Research Planning in Environmental Health Science. The Report of the Task Force is an independent and collective report which has been published by the Government Printing Office under the title, "Human Health and Environment-Some Research Needs." Copies of the original material for this Background Document, as well as others prepared for the report can be secured from the National Technical Information Service, U.S. Department of Commerce, 5285 Port Royal Road, Springfield, Virginia 22161.
[fig] RECOMMENDATION 2: DRUGS IN THE HOSPI-TAL AND NONHOSPrTALIZED POPULATION Research in development of an automated and inexpensive drug monitor to be used for all hospitalized patients should be initiated. [/fig]
|
A lectin-coupled porous silicon-based biosensor: label-free optical detection of bacteria in a real-time mode
Accuracy and speed of detection, along with technical and instrumental simplicity, are indispensable for the bacterial detection methods. porous silicon (pSi) has unique optical and chemical properties which makes it a good candidate for biosensing applications. On the other hand, lectins have specific carbohydrate-binding properties and are inexpensive compared to popular antibodies. We propose a lectin-conjugated pSi-based biosensor for label-free and real-time detection of Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus) by reflectometric interference Fourier transform spectroscopy (RIFTS). We modified meso-PSiO 2 (10-40 nm pore diameter) with three lectins of ConA (Concanavalin A), WGA (Wheat Germ Agglutinin), and UeA (Ulex europaeus agglutinin) with various carbohydrate specificities, as bioreceptor. The results showed that ConA and WGA have the highest binding affinity for E. coli and S. aureus respectively and hence can effectively detect them. This was confirmed by 6.8% and 7.8% decrease in peak amplitude of fast Fourier transform (FFT) spectra (at 10 5 cells mL −1 concentration). A limit of detection (LOD) of about 10 3 cells mL −1 and a linear response range of 10 3 to 10 5 cells mL −1 were observed for both conA-E. coli and WGA-S. aureus interaction platforms that are comparable to the other reports in the literature. Dissimilar response patterns among lectins can be attributed to the different bacterial cell wall structures. Further assessments were carried out by applying the biosensor for the detection of Klebsiella aerogenes and Bacillus subtilis bacteria. the overall obtained results reinforced the conjecture that the WGA and conA have a stronger interaction with Gram-positive and Gram-negative bacteria, respectively. Therefore, it seems that specific lectins can be suggested for bacterial Gram-typing or even serotyping. These observations were confirmed by the principal component analysis (pcA) model.According to World Health Organization (WHO) reports, dangerous water and food-borne diseases such as cholera, typhoid, dysentery, hepatitis, etc. are responsible for an estimated 2 million diarrheal deaths each year 1 . Therefore, providing inexpensive methods that can be run in the shortest possible time is a growing need for timely prevention of health crises. The conventional methods for microorganism detection (such as culture-, nucleic acid-, and immunological-based methods) mostly suffer from slowness, methodological complexity, and high cost that make them inefficient for rapid and real-time detection of pathogenic microorganisms 2 . Therefore, Over the past decades, categories of biosensors such as optical, electrochemical, and piezoelectric have attracted immense attention for diverse sensing applications 3 . Most optical biosensors have high sensitivity and high specificity. Their operation is cost-effective due to low material consumption. Furthermore, the real-time operation makes them fast and label-free monitoring makes them easy to operate. Therefore, they have gained full attention and so rapid progress 3-5 .Along with the common optical approaches of fluorescence and surface plasmon resonance (SPR) 6 , recently, biosensors based on RIFTS such as PSi-based biosensors have opened new windows of opportunities for the open
design and fabrication of more effective analytical devices [bib_ref] Porous silicon biosensor: Current status, Dhanekar [/bib_ref]. In addition to ease of fabrication, biocompatibility, and biodegradation 8 , PSi has outstanding features that make it an excellent choice as a transducer for biosensing. Its large surface area and a lot of hydride and hydroxyl surface groups allow for easy surface modification with a wide range of chemical methods. Moreover, its solid substrate makes it a potentially regenerable and reusable biosensor. Based on these facts, in recent years, many PSi-based platforms are proposed for various applications such as drug delivery, characterizing of cellular processes such as phospholipid bilayer formation and discrimination of single nucleotide changes in DNA [bib_ref] Label-free discrimination of single nucleotide changes in DNA by reflectometric interference Fourier..., Makiyan [/bib_ref] [bib_ref] Characterization of phospholipid bilayer formation on a thin film of porous SiO..., Pace [/bib_ref] [bib_ref] Porous silicon in drug delivery devices and materials, Anglin [/bib_ref]. Detection of bacteria by this nanostructure (PSi) is another area of research with a large number of publications in the literature [bib_ref] Porous silicon-based biosensors: Towards real-time optical detection of target bacteria in the..., Massad-Ivanir [/bib_ref] [bib_ref] Detection of ammonia-oxidizing bacteria (AOB) using a porous silicon optical biosensor based..., Zhang [/bib_ref].
Based on these facts, in recent years, many PSi-based platforms are proposed for drug delivery, characterizing of cellular processes such as phospholipid bilayer formation and discrimination of single nucleotide changes in DNA and bioassay [bib_ref] Label-free discrimination of single nucleotide changes in DNA by reflectometric interference Fourier..., Makiyan [/bib_ref] [bib_ref] Characterization of phospholipid bilayer formation on a thin film of porous SiO..., Pace [/bib_ref] [bib_ref] Porous silicon in drug delivery devices and materials, Anglin [/bib_ref] [bib_ref] Porous silicon-based biosensors: Towards real-time optical detection of target bacteria in the..., Massad-Ivanir [/bib_ref]. Detection of bacteria by this nanostructure (PSi) is another area of research that has been accompanied by the publication of a considerable amount of literature. Examples in this area are the following researches:
An antibody conjugated nanopore array based on PSi has been introduced for the immunosensing of E. coli by RIFTS [bib_ref] Bacteria detection based on its blockage effect on silicon nanopore array, Tang [/bib_ref]. As well as regarding the size-exclusion filtering capabilities of PSi structure, it could be used as both a filter and a sensor for the detection of target molecules from complex biological samples [bib_ref] Whole blood optical biosensor, Bonanno [/bib_ref]. Also, a porous SiO 2 / hydrogel hybrids structure immobilized with the antibody via a biotin-streptavidin system is constructed for label-free detection of bacterial by RIFTS. Recently a photonic crystal of porous silicon has been employed for bacterial detection based antibody-antigen interaction. In addition to one-dimension photonic crystal configurations of PSi, other sophisticated structures such as Bragg mirrors and microcavities have been employed for biological application [bib_ref] Detection of ammonia-oxidizing bacteria (AOB) using a porous silicon optical biosensor based..., Zhang [/bib_ref] [bib_ref] Double etched porous silicon films for improved optical sensing of bacteria, Gongalsky [/bib_ref] [bib_ref] A porous silicon optical microcavity for sensitive bacteria detection, Li [/bib_ref] [bib_ref] Identification of gram negative bacteria using nanoscale silicon microcavities, Chan [/bib_ref].
Up to now, the antibodies and nucleic acid-based probes have been the most used bioreceptors in biosensors design. But their use has been associated with the challenges. For example, antibodies, despite their unique specificity and selectivity, are unstable and expensive. In fact, glycan moieties are poor immunogen and for this reason, the production of their high specific antibodies is very hard [bib_ref] Development and applications of lectins as biological tools in biomedical research, Dan [/bib_ref]. The nucleic acid-based probes such as aptamers also suffer from long, labor-intensive, and costly production processes. Moreover, the target structure of antibodies and aptamers must be well-defined. Otherwise, the cross-reactivity probability increases and undermines the accuracy of the results [bib_ref] Glycoprofiling as a novel tool in serological assays of systemic sclerosis: A..., Bertók [/bib_ref] [bib_ref] Nanoscale controlled architecture for development of ultrasensitive lectin biosensors applicable in glycomics, Kluková [/bib_ref]. Thus in recent years, lectins have been opened up broad prospects for the application in biosensing studies. The lectins are a group of proteins with plant, microbial or animal origin. They recognize and bind to specific glycan moieties of glycoproteins or glycolipids [bib_ref] Immobilization of concanavalin A lectin on a reduced graphene oxide-thionine surface by..., Filip [/bib_ref].
In contrast to antibodies, lectins are inexpensive and more stable. They have a smaller size that can provide an appropriate surface density. Also, the specificity of lectins is based on the presence of certain carbohydrate residues in glycan moieties and is independent of a particular sequence. But their major drawback is that they have less specificity than antibodies. Although, due to their potential for polyvalent interaction (mostly having more than one binding site) with carbohydrates, this limitation is improved, and results in an affinity comparable to immunological interaction [bib_ref] Analytical application of lectins, Hendrickson [/bib_ref].
Although there is a growing interest in the design of lectin based biosensors, the main focus of this researches has been on the analytical label-free methods such as SPR, quartz crystal microbalance (QCM), and electrochemical impedance spectroscopy (EIS) [bib_ref] Immobilization of concanavalin A lectin on a reduced graphene oxide-thionine surface by..., Filip [/bib_ref] [bib_ref] Recent progress in lectin-based biosensors, Wang [/bib_ref]. To the best of our knowledge, there is only one study on the use of lectin-modified PSi to evaluate antimicrobial susceptibility by the RIFTS method without focusing on the role of lectin [bib_ref] Unraveling antimicrobial susceptibility of bacterial networks on micropillar architectures using intrinsic phase-shift..., Leonard [/bib_ref]. The aim of this research has therefore been to investigate the use of lectins instead of antibodies to modify the surface of PSi for label-free detection of bacteria by the RIFTS method. In this work, we sought to detect E. coli and S. aureus by three types of lectins (ConA, WGA, and UEA) as bioreceptors and compared our results with other investigations using antibodies as biorecognition elements.
Material and methods fabrication of pSi. Highly boron-doped (p-type) silicon wafers (0.001-0.009 Ω cm resistivity, ⟨100⟩-oriented) were purchased from Latech Scientific Supply, Singapore. The wafers were cut into 8 mm × 8 mm squares pieces by laser cutting. At first, silicon samples were washed by soap and deionized (DI) water. Then, sonicated in methanol and ethanol mixture (1:1 volume ratio) for 5 min and rinsed with DI water. The etching process carried out in a Teflon cell by using an electrolyte solution consisted of HF (39%) and absolute ethanol (3:1 volume ratio). Due to the formation of a parasitic layer (with a much smaller pore size than the desired size) on top of the surface, a two-step protocol was followed to eliminate this layer [bib_ref] Label-free discrimination of single nucleotide changes in DNA by reflectometric interference Fourier..., Makiyan [/bib_ref] [bib_ref] Surface characterization of porous silicon after pore opening processes inducing chemical modifications, Errien [/bib_ref]. First, the silicon pieces were subjected to anodic etching under a constant current density of 100 mA cm −2 for 30 s and then were rinsed with methanol and were dried under a nitrogen stream. This initial layer was destroyed by sonication in an ethanolic solution of NaOH (1 M) (4 g NaOH + 10 mL EtOH + 90 mL DI water) for 3 min. Then samples were washed three times with abundant ethanol. In the second step, the main sensing layer was prepared by the etching of previous substrates in the formerly mentioned concentration of electrolyte and at a constant current density of 130 mA cm −2 for 60 s. After etching, the PSi samples were rinsed with methanol and were dried under a nitrogen stream. and incubated at 37 °C with shaking for 24 h. After the incubation period, 10 mL of the overnight culture was centrifuged at 6,000 rpm for 10 min. The supernatant was discarded and bacteria biomass resuspended in saline solution. The bacterial suspension was used for the preparation of serial dilution (10 -1 to 10 -7 ) with saline. The 10 -3 to 10 -5 dilutions were used for the counting of viable bacteria cells by the drop-plate method and biosensing experiment simultaneously. Also, we used the fresh overnight E. coli culture medium as a complex sample. The serial dilutions (10 -3 to 10 -5 ) of overnight culture medium were prepared and straightly applied for biosensing experiment and plate counting.
## Psi surface modification.
pSi layer characterization. The structure, pore size, and surface morphology of the PSi samples were characterized by field emission scanning electron microscopy (FE-SEM; MIRA3, TESCAN, Czech Republic) working at 15 kV acceleration voltage of the electron gun. Pores diameter distribution was obtained by ImageJ software (ImageJ; National Institutes of Health, Bethesda, MD, USA). The chemical compositions of the surface were investigated with FTIR (Tensor 27, Bruker, USA) in ATR (attenuated total reflectance) mode. All the spectra were obtained from an average over 16 scans with a resolution of 4 cm −1 in the wavenumber range 400-4,000 cm −1 . To evaluate the wettability of samples, the contact angle (CA) between the surface and DI water droplet (4µL) was measured by a CA system (CA-500A, SharifSolar, Iran). CA measurements at five different points of the substrate surface were carried out and the average of them was reported. For RIFTS measurements, a bifurcated optical fiber was used to focus the white light from a tungsten-halogen light source on the sample surface and transfer the specularly reflected beam to the spectrometer (USB4000, Ocean Optics, USA) as illustrated schematically in , main part. The reflected spectra were recorded in the range of 400 to 1,000 nm at a spectral resolution of 0.5 nm with a spectral acquisition time of 10 ms. For surface modification study by RIFTS measurements, in each modification step , surface modification box, part a), the reflected spectrum , surface modification box, part b) was transformed by fast Fourier transformation (FFT) which was computed by the IGOR software (Wavemetrics Inc., Oregon, USA). This transformation yielded a peak located at the value of effective optical thickness (EOT, twice the product of the refractive index and thickness values of the porous layer 10 ; , surface modification box, part c). The porosity, refractive index, and thickness of the porous layer were determined by the method similar to the surface modification box in and using spectroscopic liquid infiltration method (SLIM) [bib_ref] Porous Silicon in Practice: Preparation, Sailor [/bib_ref]. It means that the reflectance spectra were obtained from the surface of fresh-PSi in two different media (air and methanol) and corresponding FFT values were calculated. By applying the SLIM algorithm of IGOR PRO software to these values, porosity, refractive index, and thickness of the porous layer were calculated.
For biosensing measurements, modified PSi samples were fixed on a handmade fluidic system , biosensing box, part a). This includes a fixed plexiglass well connecting to a syringe pump with an inlet tube for injection and an outlet tube for the evacuation of liquids (saline buffer and bacterial solution). In each experiment, an initial washing step was performed by passing saline buffer for 10-15 min over the sample at a flow rate of 200 µL/min. In the second step, bacterial suspension in saline (in dilutions of 10 -3 to 10 -5 cells mL −1 ) was injected with the previous flow rate (200 µL/min) and the surface was allowed to incubate with it for 25-30 min at ambient temperature. Finally, the surface was rinsed with saline again for 10-15 min at a flow rate of 50 µL/min. During all of the foregoing processes, the reflective spectrum was recorded every 2 min. Since the bacteria size is several [bib_ref] Porous silicon-based biosensors: Towards real-time optical detection of target bacteria in the..., Massad-Ivanir [/bib_ref] [bib_ref] Whole-cell detection of live Lactobacillus acidophilus on aptamer-decorated porous silicon biosensors, Urmann [/bib_ref]. After the subsequent washing step and with the removal of unbounded species, the intensity of the reflection spectrum was increased again. However, depending on the sensitivity of the sensor, this value remains lower than the base signal. Real-time monitoring of the biosensing process was carried out, by extracting the FFT peak amplitude vs. time . Finally, the FFT peak amplitude change in percentage for PSi lectinmodified biosensor was calculated by the following equation:
## Principal component analysis (pca). principal component analysis is a multivariate projection method
designed to extract and display the systematic variation in a data matrix. The starting point for PCA is a matrix of data with N rows (observations) and K columns (variables) [bib_ref] Process analytical technology (PAT) and quality by designons, Eriksson [/bib_ref]. In our case, elements of the data matrix were the FFT peak amplitude changes (responses of biosensors). The columns representing the responses of a specific lectin to different bacteria at different concentrations. However, the rows representing the responses to a distinct bacteria at a definite concentration from different lectins. Principal component analysis (PCA) was applied to this matrix and scores plot and loading plot were obtained in Sigmaplot (Systat Software Inc.).
# Results and discussion
characterization of pSi by fe-SeM. The characteristics of the PSi layer i.e. thickness and pore size were studied by FE-SEM after oxidation. [fig_ref] Figure 3: FE-SEM images of PSi [/fig_ref] illustrates the pores on the surface and the inset plot shows the pore size distribution (5 to 45 nm) which the majority of pores have the size between 20 and 25 nm. The crosssection imagining of the porous layer reveals a thickness of about 3.3 µm [fig_ref] Figure 3: FE-SEM images of PSi [/fig_ref] and the inset plot illustrates the cross-section image near the surface with higher resolution.
PSi surface modification study. Freshly prepared PSi surface is rich in hydride groups (Si x SiH y ) 9 that are highly reactive, unstable, and inherently susceptible to oxidation or hydrolysis, even in environmental conditions. It changes the physicochemical and optical properties of the PSi layer over time [bib_ref] Influence of natural aging on photoluminescence from porous silicon, Lenshin [/bib_ref]. For example, in biosensing purposes, these unwanted reactions increase the gradual dissolution of atoms from the surface and detach the biorecognition elements [bib_ref] Quantification and reduction of the residual chemical reactivity of passivated biodegradable porous..., Shabir [/bib_ref]. Additionally, it has been shown that hydride groups are strong reducing agents that make the surface incompatible to conjugate with many biological or non-biological materials [bib_ref] Suitability of porous silicon microparticles for the long-term delivery of redox-active therapeutics, Wu [/bib_ref]. Hence, surface passivation of PSi is a critical step to stable its surface for application at biological research fields [bib_ref] A mini review: Recent advances in surface modification of porous silicon, Lee [/bib_ref]. Thus, the fresh PSi samples were modified as reported in the "PSi surface modification" section. To verify the accuracy of PSi layer modification, the surface chemical changes were monitored by FTIR, CA analysis, and RIFTS after each step. During the surface modification, the chemistry of the sample surface undergoes a considerable change that can be monitored by FTIR analysis [fig_ref] Figure 4: Step by step study of PSi surface modifications by [/fig_ref]. These results are similar to those reported by other [bib_ref] Progress of porous silicon APTES-functionalization by FTIR investigations, Majoul [/bib_ref] [bib_ref] Spectroscopy studies of functionalized oxidized porous silicon surface for biosensing applications, Hiraoui [/bib_ref]. After oxidation, the appearance of typical peaks attributes to Si-O-Si bonds at 1,000 to 1200 cm −1 and under 1,000 (attribute to O n SiH x species) in FTIR spectra verify the successful oxidation process. Therefore, to achieve the best oxidation condition (among the six different oxidation processes) as mentioned in the "PSi surface modification" section, the Si-O-Si bond absorption peak was more accurately investigated. As [fig_ref] Figure 4: Step by step study of PSi surface modifications by [/fig_ref] indicates, the sample oxidized by a thermal pretreatment in an oven at 65 °C for 1 h and subsequently immersion in H 2 O 2 for 24 h presents a stronger absorption peak. Thus, this oxidation method has been selected as the final protocol for the oxidation of samples. After APTES modification [fig_ref] Figure 4: Step by step study of PSi surface modifications by [/fig_ref] , the band vibrations mods of this molecule leave two distinct absorption regions in 2,800-3,000 cm −1 and 1,550-1,650 cm −1 . The first one is assigned to the symmetric and asymmetric stretching modes of C-H from APTES backbone in 2,932 cm −1 www.nature.com/scientificreports/ and 2,883 cm −1 respectively. Another region is related to NH 2 deformation modes of APTES amine groups at 1,484 cm −1 and 1,560 cm −1 . After GA treatment, the spectrum depicts two peaks assigned to CH 2 deformation and C=O (aldehyde group) stretching mode respectively at 1,407 and 1,720 cm −1 . One another peak attributed to the N=C bond between the silane layer and GA appeared between 1,630 and 1,650 cm −1 . Finally, the lectin immobilization step was verified by the appearance of two significant absorbance peaks attributed to amide I and amide II bands at 1,653 and 1,545 cm −1 respectively. As already mentioned the surface of freshly prepared PSi due to the abundant hydride groups displays extremely hydrophobicity that limits it for the biomedical application taking place in aqueous-based medium. However, after oxidation, as shown in [fig_ref] Figure 4: Step by step study of PSi surface modifications by [/fig_ref] , a sharp decrease in CA from 96º ± 13º to 19º ± 8º validates the successful oxidation process. After APTES and GA treatment, we observed an increase in hydrophobicity (28º ± 6º and 50º ± 11º respectively) due to the carbon backbone of these molecules. But after the lectin immobilization, the presence of hydrophilic functional groups of the protein structure compensates for this decrease and CA decreased to 27º ± 4º. These results are in agreement with other reports indicating that changes in PSi surface chemistry during of modification process greatly affect the wettability behavior of it [bib_ref] Label-free discrimination of single nucleotide changes in DNA by reflectometric interference Fourier..., Makiyan [/bib_ref] [bib_ref] Aminosilane functionalizations of mesoporous oxidized silicon for oligonucleotide synthesis and detection, De Stefano [/bib_ref].
[formula] (1) FFT peak amplitude change (%) = A 1 − A 2 A 1 × 100 [/formula]
Since the refractive index (and consequently EOT parameter) of the PSi layer is greatly affected by its surface chemistry, another way to evaluate the success of the modification process is the assessment of reflective spectra. [fig_ref] Figure 4: Step by step study of PSi surface modifications by [/fig_ref] (main plot) demonstrates these reflective spectra for the fresh PSi layer and after each modification step. FFT of these spectra has been illustrated in the inset of [fig_ref] Figure 4: Step by step study of PSi surface modifications by [/fig_ref]. As it has been shown, upon oxidation a decrease in EOT was observed (8.89 ± 0.13 µm to 8.12 ± 0.14 µm). This is attributed to the formation of a thin SiO 2 layer on pore surfaces that has a lower refractive index than Si skeleton [bib_ref] Electrochemical stabilization of porous silicon multilayers for sensing various chemical compounds, Salem [/bib_ref]. In contrast, modification of the surface by APTES and GA molecules, increased EOT to 8.78 ± 0.14 µm and 8.95 ± 0.19 µm respectively, due to organo-silane layer formation that increases the effective refractive index of layer [bib_ref] Aminosilane functionalizations of mesoporous oxidized silicon for oligonucleotide synthesis and detection, De Stefano [/bib_ref]. Furthermore, through the lectin modification, due to the amide bond formation, EOT increased more (9.16 ± 0.14 µm). The porosity and refractive index of the fresh PSi layer were calculated by the SLIM method. These parameters were 82 ± 5% and 1.5 ± 0.3 respectively averaged over three samples. Also, the SLIM method yielded a thickness of 3.5 ± 0.5 µm for the PSi layer which is in agreement with the FE-SEM result (3.3 µm). evaluation of biosensor performance. Carbohydrates are one of the four major groups of macromolecules that are found on the surface of bacterial cells such as all other living organisms. In addition to the nutritional role of carbohydrates as an energy resource, many of the vital functions of living cells such as move- www.nature.com/scientificreports/ ment, structural protection, cell-cell recognition, cell junction, carry out through proteins and lipids that are glycosylated [bib_ref] Glycosylation as a main regulator of growth and death factor receptors signaling, Ferreira [/bib_ref]. Huge structural diversity at cell surface glycans in contrast to DNA and proteins (even between cells from one species) introduce them as a unique identity card for cell identification [bib_ref] Sensing glycans as biochemical messages by tissue lectins: The sugar code at..., Kaltner [/bib_ref]. The similarity of physicochemical properties, despite the unique structural complexity of carbohydrate, limits the detection of them with common methods based on the chromatography or spectroscopy for biosensing purposes [bib_ref] Analytical application of lectins, Hendrickson [/bib_ref] [bib_ref] Lectin-and saccharide-functionalized nano-chemiresistor arrays for detection and identification of pathogenic bacteria infection, Saucedo [/bib_ref]. Nevertheless, lectins, relying on their ability to detect not only the composition of carbohydrates but also the configuration of bonds between the sugar residues, are the most effective tools in this area [bib_ref] Dependence of concanavalin a binding on anomeric configuration, linkage type, and ligand..., Isabelle [/bib_ref] [bib_ref] Capillary electrophoresis separations of glycans, Lu [/bib_ref]. Therefore, in this study, lectin-carbohydrate interaction has been considered as the basis of detection. The function of this biosensor is evaluated and compared with three lectin types having different sugar-binding specificity to choose the best design with the highest affinity and performance. [fig_ref] Figure 5 •: FFT peak amplitude change percentage of PSi modified by three different types... [/fig_ref] illustrates the response [FFT peak amplitude change percentage, Eq. (1)] of PSi modified by these three types of lectin (ConA, WGA, and UEA) to the three different concentrations (3 × 10 3 , 3 × 10 4 and 3 × 10 5 cells mL −1 ) of E. coli [fig_ref] Figure 5 •: FFT peak amplitude change percentage of PSi modified by three different types... [/fig_ref] and S. aureus [fig_ref] Figure 5 •: FFT peak amplitude change percentage of PSi modified by three different types... [/fig_ref]. For comparison, the response of fresh PSi (control 1) and modified PSi (without lectin; control 2) to both types of bacteria in three different concentrations and ConA-modified PSi (control 3) to DI water were measured. The figure shows some interesting results which are summarized as follows:
- Each of the three types of modification for PSi (by ConA, WGA, and UEA) has responded to both types of bacteria (E. coli and S. aureus). These responses were considerable except modification by UEA at a bacterial concentration of 3 × 10 3 cells mL −1 . In this case, the responses were comparable to the responses of control tests. [fig_ref] Table 1 ,: compares the characteristics [/fig_ref] shows no observable contrast between the performance of lectinmodified and antibody-modified PSi biosensors based on RIFTS. However, lectins have a much lower affinity than the antibodies (dissociation constant of 10 -4 -10 −5 M for lectins vs. 10 −8 M for antibodies) [bib_ref] Antibodies and lectins in glycan analysis, Cummings [/bib_ref] , there are several possible explanations for our observations. This results could be attributed to the relatively smaller size of lectin molecules (in comparison with antibodies) which allows for the higher density of binding sites on the surface for these molecules to produce comparable sensitivity [bib_ref] Comprehensive list of lectins: origins, natures, and carbohydrate specificities, Kobayashi [/bib_ref]. Polyvalency (multiple binding sites) of lectins may be another important factor that compensates for this limitation. - As illustrated no significant change was observed in FFT peak amplitude for control tests. This observation indicates that the lectin-carbohydrate interaction is the key role to bacterial bonding in this experiment and the physical absorption of bacteria on the surface is much negligible to trigger a considerable change in the reflection spectrum. - There was a linear relationship (in semi-logarithmic scale) between the FFT peak amplitude change percentage and bacterial concentration in the range of 3 × 10 3 to 3 × 10 5 cells mL −1 for each of the three types of modification and both type of bacteria (insets in [fig_ref] Figure 5 •: FFT peak amplitude change percentage of PSi modified by three different types... [/fig_ref]. As shown the values of R-squared are between 0.9093 to 0.9950. However, when the concentration of bacteria approached 10 6 cells mL −1 for both cases, the FFT peak amplitude change percentage was no longer be linearly proportional to concentration due to the overcrowding bacteria cells adjacent to the PSi surface which increase light scattering. Also in concentrations less than 10 3 cells mL −1 , no distinct signals were observed (data not shown). Regarding the obtained results with E. coli and S. aureus, it seemed that Gram-negative and Gram-positive bacteria have different interaction patterns based on lectin type. These results encouraged us to test the biosensor's performance with two other bacteria. We used Klebsiella aerogenes and Bacillus subtilis strains for more experiments. The obtained results [fig_ref] Figure 6: FFT peak amplitude change percentage of PSi modified by three different types... [/fig_ref] showed that the affinity pattern for both recent bacteria follows the same pattern as the previous results for E. coli and S. aureus. Also, these results reinforced the conjecture that the WGA and ConA have a more interaction affinity for Gram-positive and Gram-negative bacteria, respectively. Of course, to give a definitive opinion, many more bacteria species will need to be tested in the next studies. www.nature.com/scientificreports/ Moreover, the biosensor performance analysis with E. coli culture medium as a real sample showed that although the response rate (percent of FFT peak amplitude change) has been slightly decreased compared to pure bacterial dilutions (have addressed in manuscript), it still was significantly upper than the control tests.
For example, for 10 -3 dilution of culture medium, the FFT peak amplitude decrease is 4.1% in contrast to 6.8% for 10 -3 pure bacterial suspension. Also during spectra recording relatively high ups and downs occurred can be due to the physical adsorption and separation of the culture medium compounds. This result indicated that the designed platform can be efficiently applied even for complex samples and other biosensing purposes. principal component analysis. PCA is a statistically valuable method for discovering relationships between multivariate data. That means the smaller distance between the data indicates a positive correlation and significant similarity between them, and vice versa. In this study, PCA aims to determine the relationship between bacteria strain based on their response profile to different [bib_ref] Lectin-and saccharide-functionalized nano-chemiresistor arrays for detection and identification of pathogenic bacteria infection, Saucedo [/bib_ref] [bib_ref] Microorganisms recognition and quantification by lectin adsorptive affinity impedance, Gamella [/bib_ref] [bib_ref] SERS detection of multiple antimicrobial-resistant pathogens using nanosensors, Kearns [/bib_ref]. The results of the PCA analysis are depicted in. The scores plot (main plot ofshows the projection of the data onto the 2D PCA parameters space. As shown PC1 can explain 72.80% of the total data variance and PC2 can explain 23.21% of the variance. It is obvious from this plot that gram-negative bacteria (E. coli and K. aerogenes) scatter in a positive PC2 area. However, gram-positive bacteria (S. aureus and B. subtilis) scatter in the negative PC2 area. Thus, three types of PSi lectin-modified biosensors can discriminate bacterial gram-type effectively without overlap. The loading plot (inset plot inshows which lectins are significant, and how the three types of PSi lectinmodified biosensors are correlated. In the loading plot, strongly positive correlated variables have two vectors that are very close to each other. However, strongly negative correlated variables have vectors that are out of phase by 180o. Additionally, the magnitude of the vector of each variable also has information. The larger the www.nature.com/scientificreports/ magnitude of its vector, the stronger impact that variable has on the model 61 . Therefore, the inset plot inindicates that WGA and ConA have the strongest impact on the responses of our biosensors respectively while UEA has the weakest impact on responses. Besides, ConA and UEA are strongly positive correlated variables. It means that when the response of ConA increases or decreases (by changing the bacteria type or bacteria concentration), the response of UEA tends to change in the same way.
# Conclusion
Advanced and rapid detection methods are mainly based on antibodies or nucleic acids as routine bioreceptor molecules. These two molecules despite sufficient sensitivity and specificity but face challenges due to inherent limitations such as low stability and high cost. On the other hand, the detection methods based on these molecules are mostly time-consuming and require heavy laboratory equipment, qualified personnel, and require some initial knowledge of the target molecule or structure. Hence, in recent years, lectins relying on considerable properties such as low cost, desired stability, acceptable sensitivity, and selectivity has been introduced as an attractive alternative for antibodies and nucleic acids. Therefore in this study, we proposed a lectin conjugated PSi-based biosensor for bacteria detection by the RIFTS technique. Due to the acceptable specificity of lectins, the transducer surface was modified with three different lectins (ConA, WGA, and UEA) as bioreceptors. Evaluation of biosensor performance showed a different response pattern concerning bacteria species and lectin type. So that for E. coli, ConA, and S. aureus, WGA have the highest binding affinity with a linear rang response from 3 × 10 3 to 3 × 10 5 cells mL −1 whereas UEA showed the lowest responses to both types of bacteria. Also, a relatively low LOD about 10 3 cells mL −1 was reported for WGA and ConA in their highest binding affinity profile. The further assessments with two extra bacteria species of K. aerogenes and B. subtilis revealed a similar response pattern based on the lectin type and Gram type of bacteria to the main experiment. Also, the assessment of the obtained data by the PCA test further confirms a significant pattern in the bacteria-lectins interaction based on Gram-type of bacteria. Therefore, given the efficiency and cost-effectiveness of PSi as a transducer and lectin as a bioreceptor, this report can be a promising approach for broad biosensing applications.
Received: 29 January 2020; Accepted: 27 August 2020
[fig] Figure 1: Surface modification of PSi as transducer upon each modification step. magnitude larger than the pore diameters of the porous layer, adsorption of bacteria occurred only on top of the pores of PSi. It changes the contrast between the refractive index of PSi and the refractive index of ambient. Consequently, the amount of reflected light(Fig. 2, biosensing box, part b)and the peak amplitude of its FFT values(Fig. 2, biosensing box, part c)were decreased in harmony with other research [/fig]
[fig] Figure2.: Schematic of RIFTS measurement. Surface modification box: (a) the penetration of analyte molecules in pores or change the surface chemistry during surface modification process which alters EOT, (b) reflection spectra from PSi in each step in part a and (c) the FFT intensity extracted from the spectra in part (b). Biosensing box: (a) biosensing monitoring of bacterial suspension by a handmade fluidic system, (b) reflection spectra from PSi in two steps in part (a); (c) the FFT intensity extracted from the spectra in part (b); (d) Real-Time monitoring of biosensor response extracted from part c during the initial washing, bacterial expose and final washing steps. [/fig]
[fig] Figure 3: FE-SEM images of PSi: (A) top view (main plot) and pore diameter distribution (inset plot), (B) and cross-sectional view (inset: image with higher resolution). Scientific RepoRtS | (2020) 10:16017 | https://doi.org/10.1038/s41598-020-72457-x [/fig]
[fig] Figure 4: Step by step study of PSi surface modifications by (A) FTIR analysis (inset: comparison between different oxidation conditions on Si-O-Si bonds), (B) CA measurement, and (C) RIFTS method (inset: corresponding FFT values). Scientific RepoRtS | (2020) 10:16017 | https://doi.org/10.1038/s41598-020-72457-x [/fig]
[fig] Figure 5 •: FFT peak amplitude change percentage of PSi modified by three different types of lectin (ConA, WGA, and UEA) at the presence of different concentrations of (A) E. coli and (B) S. aureus. Control tests: fresh PSi as control 1, GA-modified PSi (without lectin) as control 2 and DI water on ConA-modified PSi as control 3. (R-squared values: 0.9955 for E. coli-ConA, 0.9435 for E. coli-WGA, 0.9709 for E. coli-UEA, 0.9093 for S. aureus-ConA, 0.9950 for S. aureus-WGA, 0.9394 for S. aureus-UEA). The biosensor responses were different for two types of bacteria based on lectin types. PSi modified by ConA had the highest responses to E. coli in all three concentrations. However, modification by WGA was the best one to efficiently detect S. aureus in the same foregoing concentrations. In contrast, PSi modified by UEA showed the lowest responses to both types of bacteria in all measuring concentrations. It is suggested that these observations are caused by the different binding affinity between lectins and bacterial species. Such different affinity profiles can be evaluated from two perspectives: (1) Based on the binding affinity levels of used lectins and (2) Based on the structural difference between the two bacterial species. Based on the first perspective and as mentioned previously the affinity level of lectins greatly depends on their structure and number of the sugar-binding sites. The more the number of binding sites, the greater the binding affinity. WGA with eight binding sites which are complementary to 6 or 8 β-1,4-GlcNAc units 55 in contrast to ConA with four binding sites for 4 α-d-mannosyl or α-d-glucosyl residues of sugar motifs 56 . However, UEA has only one binding site for α (1,2) linked fucose residues of oligosaccharide motifs 57 . It seems possible that the low responses of our UEA-modified biosensors are due to this factor. Without a doubt, the presence of sufficient amounts of the target molecule in addition to having multiple binding sites is another essential condition for the occurrence of this event. Based on the second perspective, it should be noted that Gram-positive and Gram-negative bacteria have different cell wall structures. The peptidoglycan (or murein) is the main cell wall component of both groups that like a rigid envelope surrounds the cytoplasmic membrane. This polymer contains sugar (glycan) chains of β-(1,4) linked N-acetylglucosamine and N-acetylmuramic acid, which are cross-linked by peptides bridges. The Gram-positive bacteria cell wall consists mainly of a thick peptidoglycan layer and many other secondary polymers as the outmost part of bacteria. Gram-negative bacteria have complicated cell wall that in which a thin peptidoglycan layer enclosed by an outer membrane rich in a macromolecule known as lipopolysaccharide (LPS) and other lipo-or glycoproteins. LPS molecule is made up of three distinctive domains includes of lipid A, core polysaccharide, and O-antigen. O-antigen is the outmost part and consists of unique repeating oligosaccharide subunits that vary even in strain level 58 . Besides, the peptidoglycan is the main content of the cell wall structure (up to 90%). Thus N-acetyl-d-glucosamine (target of WGA) containing motifs are one of the most abundant structures in the external decoration of Gram-positive bacteria. In Gram-negative bacteria, such a parallel role can be imagined for d-glucose and d-mannose containing motifs of O-antigen and other glycosylated proteins or polymers. Hence, given the availability of two above-mentioned conditions (The plurality of binding sites and the frequency of the target molecule), the obtained results are logically interpretable and acceptable. [/fig]
[fig] Figure 6: FFT peak amplitude change percentage of PSi modified by three different types of lectin (ConA, WGA, and UEA) at the presence of different concentrations of (A) K. aerogenes and (B) B. subtilis. Control tests: fresh PSi as control 1, GA-modified PSi (without lectin) as control 2 and DI water on ConA-modified PSi as control 3. [/fig]
[fig] Figure 7: (Main plot) PCA scores plot of the two first PCs. The increase in bacteria concentration is indicated by an arrow direction, (Inset plot) PCA loading plot of the first two principal components. Scientific RepoRtS | (2020) 10:16017 | https://doi.org/10.1038/s41598-020-72457-x [/fig]
[table] Table 1 ,: compares the characteristics (LOD and linear range) of the presented biosensor with examples of other reports. It is apparent from this table that the obtained LOD values and linear ranges of our biosensors are comparable to similar quantities of other reports based on QCM, SPR, and etc. biosensors with lectin modification. Besides, the data in [/table]
[table] Table 1: Review of the characteristics of some bacterial biosensor reported in the literature. *LOD = 3.3 × standard deviation/slope. [/table]
|
Intellectual disability genomics: current state, pitfalls and future challenges
Intellectual disability (ID) can be caused by non-genetic and genetic factors, the latter being responsible for more than 1700 ID-related disorders. The broad ID phenotypic and genetic heterogeneity, as well as the difficulty in the establishment of the inheritance pattern, often result in a delay in the diagnosis. It has become apparent that massive parallel sequencing can overcome these difficulties. In this review we address: (i) ID genetic aetiology, (ii) clinical/ medical settings testing, (iii) massive parallel sequencing, (iv) variant filtering and prioritization, (v) variant classification guidelines and functional studies, and (vi) ID diagnostic yield. Furthermore, the need for a constant update of the methodologies and functional tests, is essential. Thus, international collaborations, to gather expertise, data and resources through multidisciplinary contributions, are fundamental to keep track of the fast progress in ID gene discovery.
# Background
Neurodevelopmental disorders (NDDs) are clinically defined as "a group of conditions with onset in the developmental period (…) characterized by developmental deficits that produce impairments of personal, social, academic, or occupational functioning" [Diagnostic and Statistical Manual of Mental Disorders, 5th Edition -DSM . Intellectual disability (ID), formerly known as "mental retardation", is an incomplete mental development, leading to a substantial limitation in general mental abilities, intellectual functioning, adaptive behaviour and function skills, in comparison with individuals of the same age, gender and social-cultural background. These limitations can be observed in many domains such as communication, personal care, self-governance, functional academic skills, among others.
ID can appear as an isolated feature (non-syndromic ID, NSID), or associated with facial dysmorphic features, other morphological anomalies, multisystemic disorders (syndromic ID, SID)or multiple neuropsychiatric and/or neurobehavioral problems, such as autism or epilepsy, and neuromuscular features, e.g. ataxia, spastic paraplegia, sensory or motor neuropathy, and muscular dystrophy. Previously, ID classification was based on intelligence quotient (IQ) scores: mild (IQ 50-69, 85.0% of ID cases), moderate (IQ.0% of ID cases), severe (IQ 20-34, 3.5% of ID cases) and profound (IQ < 20, 1.5% of ID cases). Nowadays, ID diagnostic criteria include (i) deficits in intellectual functioning confirmed by clinical evaluation and standard IQ testing; (ii) deficits in adaptive functioning that results in failure to meet developmental and sociocultural standards for personal independence and social responsibility; and (iii) onset of deficits during the developmental period. The severity of ID is based on the level of adaptive functioning deficits of an individual in the conceptual, social and practical domains, which determines the level Open Access *Correspondence: [email protected] 1 Centro de Genética Médica Jacinto de Magalhães (CGM), Centro Hospitalar Universitário do Porto (CHUPorto), Porto, Portugal Full list of author information is available at the end of the article of support needed. Under the age of 5 years, the term Global Developmental Delay (GDD) is used. GDD is characterized by the failure to accomplish developmental milestones expected for a given age range, in two or more of the above-mentioned domains, including gross or fine motor skills, speech and language, cognition, personal-social and activities of daily living. ID and GDD are evaluated and clinically followed by the same medical specialties, in particular in paediatric clinics, psychiatry, neurology/epilepsy, and rehabilitation medicine clinics. Of note, not all children with GDD will show ID in adulthood.
ID affects between 1 and 3% of individuals worldwide, although with some regional differences. Mild ID is believed to affect 0.7-1.3% of the general population, while severe and profound ID have an estimated prevalence of less than 0.5%. ID represents an important public health problem, affecting families and the society, being a burden to the health systems with direct costs estimated in 43.3 billion euro per year in Europe. Non-genetic or environmental factors, such as socio-cultural determinants and infections, can contribute to ID, although the majority of severe or profound ID are known to have a monogenetic origin.
Technological advances in the last decade, led to the identification of novel ID genes, bringing new insights into the ID molecular diagnosis, and the underlying biological mechanisms. Establishment of the ID genetic aetiology is mandatory for proper diagnosis, prognosis and disease management, assuming a key role in genetic counselling. Based on disease recurrence risk and the availability of a specific preimplantation or prenatal test, couples can be offered planning in future pregnancies. Currently, ID is rarely treatable but molecular diagnosis is crucial to guide patients and families in the process of disease acceptance and expectations adjustment allowing the liaison with patient organisations and associations. The ragbag of ID classifications, diagnostic methodologies and functional studies demand constant update and systematization to improve ID diagnostic and investigational strategies. Here, we propose to review seminal works in ID particularly focusing on massive parallel sequencing applications and functional validation of genetic variants, aiming at guiding ID diagnostic investigation.
## Intellectual disability is genetically and clinically extremely heterogeneous
Genetic diagnosis of ID can be dated back to 1959 with the identification of trisomy 21 in Down syndrome, still being the most frequent chromosome disorder and the most common single cause of ID. Conventional cytogenetics, namely karyotyping and fluorescence in situ hybridization (FISH), allow the identification of numeric and structural chromosome abnormalities, which are responsible for about 15% of ID. Recurrent microdeletions and microduplications have been identified by chromosomal microarray analysis (CMA), in patients affected with ID-related disorders, including Williams, DiGeorge, Prader-Willi, Angelman, Wolf-Hirschhorn or Cri du Chat syndromes. DNA copy-number variants (CNVs) containing few to hundreds of genes, have increasingly been identified as ID causes. CNVs, occur mostly de novo, and are responsible for about 10-14% of ID cases. Research studies in cohorts of patients carrying recurrent CNVs allowed the identification of new disease and dosage sensitive dominant genes.
Regarding monogenic ID cases, most are caused by single nucleotide variants (SNVs), and small insertions or deletions (indels), in genes that code for proteins that operate in key biological processes such as neurogenesis, synaptogenesis or synaptic plasticity. Development of a DNA sequencing method, the Sanger sequencing in 1975, and further automatization and commercialization in the 1980's, were key for the detection of this type of variants.
Non-Mendelian ID disorders are a challenge in diagnosis, genetic counselling and recurrence risk estimation. A special group are those caused by dynamic mutations occurring in tri, tetra and pentanucleotide repetitive regions. The first report of ID pathogenic variants caused by repeat expansions occurred in 1991. This study described the identification of a trinucleotide repetitive region, a CGG repeat tract located at the 5′ untranslated region of FMRP translational regulator 1 gene (FMR1) implicated in Fragile X syndrome (FXS). FXS is the most common cause of inherited ID, and despite being identified three decades ago, there is no effective treatment and knowledge on disease mechanisms is scarce. To date, more than 40 inherited diseases affecting the central nervous system, have been identified.
Also, DNA methylation or DNA imprinting, wellknown epigenetic disease mechanisms, do not follow a Mendelian inheritance pattern. Imprinting diseases are implicated in ID, growth impairment, development and metabolism defects, associated with disturbance of the regulation, dosage and genomic sequence of imprinting loci. The identification of consistent and significant methylation aberrations in multiple unrelated but phenotypically similar patientsis still challenging. The expression pattern of imprinted genes is monoallelic and parental origin dependent. To date there are eight well-characterized imprinting disorders: Prader-Willi, Angelman, Silver-Russell, Beckwith-Wiedemann, Templeand Kagami-Ogata syndromes, Transient Neonatal Diabetesand Pseudohypoparathyroidism type 1B.
Another group of heterogeneous non-Mendelian genetic diseases are those caused by pathogenic variants in the mitochondrial genome (mtDNA), also known to be involved in ID. Mitochondrial disorders are characterized by a deficient oxidative phosphorylation, with an estimated prevalence among adults of 2.9 cases per 100,000 individuals and 9.6 cases per 100,000 individuals, respectively caused by nuclear or mtDNA mutations. Approximately 1 in 200 healthy individuals carry a pathogenic variant in mtDNA with low levels of heteroplasmy, with implications in the offspring. Leigh syndrome caused by molecular defects in nuclear and mtDNA genes, and Mitochondrial DNA Depletion syndrome 4A (Alpers syndrome), are two examples of childhood-onset mitochondrial neurodegenerative disorders.
The large genetic heterogeneity, intrinsic to ID-related disorders, as well as the absence of a specific inheritance pattern, especially when there is only one affected family member, can hamper the selection of the gene to target. To interrogate a large number of genes in a single step, tackling the majority of ID causes, including SNV, indels, CNVs and even structural chromosome abnormalities, the development of the genome-wide sequencing approaches, such as massive parallel sequencing, was essential.
## Massive parallel sequencing -a milestone towards id-gene identification
Massive parallel sequencing commonly named next generation sequencing (NGS) is a fast, accurate, efficient and cost-beneficial screening strategy, representing a milestone in novel ID genes identification. Nontargeted NGS, a "genotyping driven" gene identification approach, unveiled the complexity of genotype-phenotype correlations, especially in heterogeneous disorders, where pathogenic variants in some ID-related genes can be implicated in "atypical" phenotypes. For instance, variants in CHD2, SETD2 and SLC6A1 genes are known to cause autism in some cases and severe ID without autistic features in others. With the use of reverse phenotyping, clinicians return to patients to validate or infirm a molecular result even in cases of rare genetic occurrences. Describing new features associated with well-known phenotypes expanded the phenotyping spectrum of a given gene/disease, impacting ultra-rare disorders with atypical phenotypes.
The first study using exome sequencing (ES) to uncover the genetic basis of Miller syndrome, a monogenic disorder, was published in 2010. In the last decade, new genes were rapidly associated with other autosomal dominant syndromesand the number of autosomal recessive ID (ARID) genes more than doubled. Concurrently, more than 2500 ID genes, were identified, including primary and candidate genes.
According to the SysID database, there are 1500 primary ID genes, causing 1797 ID related disorders, and 1248 ID candidate genes. ID related genes can be gathered based on their ontology, or biological function . The gene ontology-based analysis shows the large heterogeneity of ID, as well as the biological pathways involved. Gene cluster analysis shows 270 genes and 415 diseases associated with metabolism. Phenylketonuria and galactosemia, caused by molecular deficits in PAH and GALT genes respectively, are examples of such disorders, representing 1-5% of ID causes. A significant number of ID genes/diseases are also involved in transport (214/342), nervous system development (200/339), RNA metabolism (179/273) and transcription (152/245).
## Common features: from library preparation to sequencing reactions
Four sequencing platforms sharing common basic features, such as library preparation and template generation, were hitherto developed. Sequencing reactions are intrinsic to each methodology and the signal resulting from the amplification is obtained by fluorescence, light or ionic potential modification, depending on the underlying principle: sequencing by synthesis, pyrosequencing, sequencing by ligation and ion semiconductor sequencing.
Sequencing by synthesis is based in a step-by-step incorporation of nucleotides attached to a single florescent molecule. The error rate is low, although increasing with the read length. In pyrosequencing, a pyrophosphate molecule is released and light will be generated after a cascade of chemical reactions, following the polymerase incorporation of a nucleotide. This results in larger read lengths, but with high costs and high error rate over homopolymers of 6 or more nucleotides. In ligation, the reaction is based in fluorescent 8-mer oligonucleotide probes, resulting in very short read lengths. In ion semiconductor sequencing, the nucleotide sequence is determined by pH changes. Overall, this is the most cost-effective and time-efficient, despite the high error rate in large homopolymers.
## Targeted-ngs is effective on clinically recognizable forms of id
Targeted-NGS (TNGS) has been largely used in ID diagnostic settings, either using panels of genes involved in common pathways, or by studying an entire chromosome. Najmabadi et al.identified putative disease-causing variants in 78 out of136 consanguineous families (57%), resulting in the identification of 50 candidate ARID genes and variants in known syndromic-ARID genes in 26 families. Tzschach et al., sequenced 107 XLID genes in 50 patients with a suggestive XLID family history and in 100 sporadic ID patients, identified pathogenic variants in 13 (26%), and in five (5%) patients, respectively. Hu et al.identified seven novel XLID genes: CLCN4, CNKSR2, FRMPD4, KLHL15, LAS1L, RLIM and USP27X and a previously characterized ID pathogenic variant in 74 families (18%), after sequencing 745 genes in 405 families. The diagnostic yield is biased to the targeted regions and influenced by the clustering of genetic errors, typically occurring in regions with high homozygosity due to inbreeding, such as in Iran. In well characterized patients with dysmorphic, neurological or systemic features, TNGS low sequencing costs, high coverage, completeness and incidental findings low-rate, results in a decrease in the diagnostic turnaround time. As knowledge evolves, e.g. new disease associated genes are identified, updates are needed, which can be laborious, time-consuming and increase the TNGS costs.
## Exome-sequencing improves the diagnostic yield in syndromic ndds
Exome sequencing (ES) has been shown to be a powerful, robust, and scalable methodology in ID diagnosis. Trio-ES analysis (i.e. proband and parents) led to the identification of a significant number of de novo variants in patients with sporadic ID. De Ligt et al.performed a trio-ES study in 100 families and identified 70 de novo variants in 53 patients, with an overall diagnostic yield of 53%. Rauch et al.identified 87 de novo variants of which 16 in known ADID genes, in 45 out of 51 patients after a negative CNV screening. Considering the six loss-of-function variants, identified in six novel ADID genes and assumed to be pathogenic, a diagnostic yield of 88% is achieved. The Deciphering Developmental Disorders (DDD) study recruited families from all regional genetics services around the United Kingdom (UK) and Ireland. Around 2000 families with undiagnosed developmental disorders were included in the first year of the study, increasing to 8000 within 3 years. After genomewide microarray and trio-ES studies, focusing 1133 complete trio-families, de novo and segregating variants in known developmental disorder genes were identified, representing a diagnostic yield of 27%. In 2018, data were reanalysed in light of new molecular and clinical knowledge and a diagnosis was attained in further 454 families, representing a diagnostic yield of 40%. In 2019, a meta-analysis gathering information on 30 NDDs studies published between January 2014 and June 2018 concluded that the ES yield for overall NDDs is 36%, isolated NDDs 31%, and syndromic NDDs 53.
## Genome-sequencing: a complete approach
Genome sequencing (GS) provides homogeneous coverage, improving the detection of SNVs, CNVs, and balanced translocations, as well as the detection of mosaicism, when coverage depth is sufficient (e.g. a mean coverage of 130 ×). In the Schluth-Bolard et al.study, balanced chromosomal rearrangements with inversions and translocations were identified in three patients. Gilissen et al.identified 84 de novo CNVs and 82 SNVs in a cohort of 50 patients, previously undiagnosed after ES, reaching a conclusive diagnosis in 21 patients (42%). These authors estimate that the cumulative diagnostic yield of GS was 62%, including de novo SNVs (39%), de novo CNVs (21%) and recessive variants (2%), based on previously published data with large cohorts. In a cohort of 244 ID/developmental delay (DD) children, Bowling et al.identified 44 pathogenic and 10 likely pathogenic SNV/ indel variants, 5 pathogenic and 1 likely pathogenic CNVs, resulting in a diagnostic yield of 25%. Wang et al.tested whole genome low-coverage sequencing to detect CNVs, and medical exome sequencing (MES), i.e. exome analysis of known ID disease-causing genes, to identify SNVs, in 95 patients with a negative CNVs screening. Nineteen pathogenic CNVs in 16 patients (17%), and ten pathogenic SNVs in 8 patients (8%) were found. GS is the most comprehensive genetic test, as it interrogates all the genome, however, improvements in the bioinformatics algorithms for variant detection and interpretation are needed. Together with the decrease of the associated costs, are crucial for the routine implementation of GS in diagnostic settings.
## Variant filtering
Massive parallel sequencing raw data is standardly generated in the FASTQ format. The files contain the identification, sequence and sequence identifier, and quality values of each sample. Reads are usually mapped into the hg19/GRCh37 or GRCh38 versions of the human reference genome, and the alignment results are typically reported in binary alignment map (BAM) format. BAM files contain information on the possible location of each read in the human genome. After sequence alignment, variant calling will identify differences between the reads sequence and the reference genome. Variants are usually reported as variant call format (VCF) file. VCF files are composed of several lines, each corresponding to a genomic position. Sophisticated algorithms as used to screen the information generated after genome sequencing with inherent data storage and interpretation issues. Due to the intrinsic ID heterogeneity, the use of guidelines are important.represents a simplified workflow to guide variant filtering.
## Variant coverage
Variants occurring in 20% of the reads, with a minimum coverage of ten, should be considered to reduce the prioritization of sequencing artefacts. Nevertheless, variants occurring in less than 1% of the reads can be identified, when sufficient coverage is attained (e.g. 30-60 x for genome). Rohlin et al.study suggest a high mosaicism detection rate when compared with other molecular techniques, but dependent on coverage levels. Jamuar et al.identified mosaic pathogenic variants, the majority of which were undetected by conventional Sanger sequencing, in peripheral blood DNA from patients with brain malformations, using high-coverage sequencing target panels.
## Variant frequency
Variants causing uncommon and severe conditions usually are rare among the general population, and therefore variants with a frequency ≥ 1% (based on SNPs -Ensembl, dbSNPand gnomAD, for SNVs and small indels, Database of Genomic Variants (DGV)or DECIPHERin case of CNVs, and other in-house databases) can be excluded from further analysis. Exceptions are those involved in rare oligogenic diseases that can exceed 18%and common variants (minor allele frequency, MAF ≥ 5%) generally located in non-coding regions. Niemi et al.studied a cohort of 6987 children with severe NDDs and showed that inherited common variants were responsible for 7.7% of risk variance. Databases have emerged focusing on non-coding regions regulatory elements, such as CODE (http:// www. encod eproj ect. org)and Orion (http:// www. genom ic-orion. org).
## Variant percentage among reads
The inclusion of the putative ID Mendelian inheritance in the filtering strategy and variant prioritization may help to organize information and to reduce the number of candidate variants. For instance, homozygous variants are often associated with consanguinity, and therefore more common in inbred populations, and ID sporadic cases are frequently caused by autosomal dominant de novo pathogenic variants. Ancestry is therefore relevant information to consider before prioritization. Homozygous variants usually show a > 80% variant allele frequency (VAF), whereas compound heterozygous variants show a VAF varying from 20 to 80% among reads.
## Variant review
Candidate variants should be reviewed by manual analysis, using a suitable software such as the Integrative Genomics Viewer (IGV). Although still debatable, gold standard methodologies might be used to confirm variants, such as Sanger sequencing for SNVs, and genomic quantitative PCR (Q-PCR) for CNVs.
## Variant deleteriousness categorization
We suggest sequential steps for accessing the functional impact of variants in ID, towards variant classification in five categories: pathogenic, likely pathogenic, uncertain significance, likely benign, and benign, according to the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) recommendations.
Known pathogenic variants in well-recognized ID genes, based on the data published at ClinVar, Clin-Gen, OMIM, and SysIDdatabases, should be first prioritized. Other aspects should then be considered: (i) implication in other disorders, with central nervous system (CNS) impairment; (ii) levels of expression in CNS/brain, (iii) interaction with other proteins implicated in ID, or (iv) biochemical function similarity with other ID genes. Variants predicted to seriously disrupt the protein function (e.g. Loss of function, LoF) with a MAF of ≤1%, and its presence in > 50% of isoforms, should follow. When available, familial studies are used to confirm the segregation of each suitable candidate variant with the phenotype.
## In silico causality prediction
Particularly in missense variants, causality ascertainment is challenging, with an accuracy of about 80%, despite the improvement in the in silico pathogenicity predictions tools. In Rauch et al.work, two NAA10 variants were classified as pathogenic based on the expected protein effect and patient's phenotype, yet predicted as benign using in silico tools. Putative splicing effect can be screened using tools such as SpliceSite-Finder-like (normal score threshold ≥70 for SDS and SAS), MaxEntScan (normal score threshold ≥0 for SDS and SAS), NNSPLICE (normal score threshold ≥0.4 for SDS and SAS)or GeneSplicer (normal score threshold ≥0 for SDS and SAS)and
## Replication studies
Gather unrelated patients with a similar phenotype and carrying putative deleterious variants in the same gene, i.e. replication, is crucial to identify new ID genes. Nevertheless, assembly patients that comply to these characteristics is problematic, particularly in rare ID syndromes., and (vi) Solve-RD -solving the unsolved rare diseases (https:// solve-rd. eu), among others.
## Model organisms
In vivo and in vitro studies are particularly important to disclose the deleteriousness of ambiguous or novel variants as well as to implicate new genes in ID phenotypes. The implementation of ID functional studies, using model organisms or patient-derived tissues or cells, is however, complex in a diagnostic facility. Since the 1980s and 1990s, models have been used to understand the mechanisms of monogenic ID disorders, as orthologous genes are involved in evolutionary conserved biological processes. Simple organisms, with short life cycles, allowing genetic manipulation, can easily give insights into several biological processes. Next, several model organisms and corresponding ID seminal studies will be described.
## Yeast
Yeast has been considered a valuable ID model following the advances in "autophagy" knowledge, a mechanism compromised in neurological disorders. Saccharomyces cerevisiae, with 23% of homology with human genes, shares particular evolutionary conserved Adapted from Schuurs-Hoeijmakers et al.key elements with neurons, e.g. budding or mating in yeast to neurite outgrowth or spinogenesis in neurons. Yeast models were used to define (i) the function of septin in the differentiation and compartmentalization of neurons, (ii) the role of the MED12-complex in transcriptional regulation, and (iii) the mechanisms underlying mitochondrial disorders. Furthermore, has been used to study aging mechanisms and age-associated neurodegenerative disorders (reviewed by Ruetenik et al..
## Caenorhabditis elegans
The nematode Caenorhabditis elegans has also been largely used as a model for neurodevelopmental disorders. With approximately 41% homology with human genes, a short life cycle, easy cultivation and accessibility to the entire nervous system structure. C. elegans revealed to be a very valuable model to study crucial processes, such as cell birth and diversification, cell migration, morphogenesis and pathfinding, synapse formation, and neurite/synapse sorting maintenance and plasticity (reviewed by Rapti. The use of C. elegans brought important insights into the human system nervous illness, such as epilepsy, autism spectrum disorder (ASD) and ID (reviewed by Bessa et al..
## Drosophila melanogaster
Identification of conserved genes and pathways in Drosophila melanogaster (with 75% homology to human genes), goes back to the end of the 1970s. The genes involved in wings development and pattering contributed to the characterization of pathways and mechanisms responsible for skeletal and craniofacial abnormalities in humans. Drosophila is a reference Gene/protein functions GeneCards Genomic, proteomic, transcriptomic, genetic and functional information on all known and predicted human genes https:// www. genec ards. orgmodel in ID and ASD as the neuromuscular junction show structural, morphologic, and functional similarities to human synapses. Allowing the study of subcellular events, such as synapses and dendritic complexity, neurotransmission and circuit connectivity, neuronal activity and physiology, brain morphology, and behaviour alterations such as learning and social interaction issues, makes Drosophila a valuable and complete model to understand those disorders. Some human genes do not have a homologue in Drosophila, where vertebrate models, such as zebrafish and mice, are useful.
## Zebrafish
Zebrafish, with 70% of genomic content homology with humans, and similar CNS structures, such as the hippocampus, diencephalon, tectum and tegmentum, and cerebellum, has emerged as an important disease model but also to test potential therapeutic solutions. Zebrafish has a short reproductive cycle, transparent embryos and larvae, easy access to the central nervous system, being used to recapitulate: (i) behaviour, such as hypoactivity and hyperactivity, hyperexcitability, impulsiveness, aggressiveness, circadian disturbances, and schizophrenia; (ii) cognitive, learning and memory deficits, and structural abnormalities; or (iii) physical, such as microcephaly, macrocephaly and microphthalmia, some of the neurodevelopmental disorders clinical features. Zebrafish has been widely used as model for ASD, attention deficit hyperactivity disorder (ADHD), ID and schizophrenia-like phenotypes (reviewed by De Abreu et al..
## Mice
ID research, NDD investigation, including development of innovative therapies is anchored in mice studies, due to the similarity (90%) between both genomes. Pivotal studies include: (i) biochemical alterations, such as Mecp2-related deficit in Gamma aminobutyric acid (GABA) and glutamate synthesis pathway, and the imbalance of brain metabolites in the hippocampus of Fmr1 KO mice during the developmental period of synaptogenesis and early myelination, (ii) changes in synaptic morphology and function, such as Syngap1 associated to early maturation of the spines, and decrease of dopamine auto receptors in Mecp2 KO mice, and (iii) behavioural issues, such as social impairment, communication problems, repetitive behaviour and resistance to change in routine, cognition, memory, and learning.
## Patient-derived cellular models
The brain is an unavailable organ in live humans whereas post-mortem tissue gives information mostly on the end-stage of a disease, providing little contribution on early brain development or impairment. ID genes are differently expressed during brain development and thus the impact of variants in such genes should be accessed at the suitable stage of maturity. Cellular models to study monogenic ID disorders have emerged as an alternative to animal models, such as humaninduced pluripotent stem cells (hiPSCs).
## Human-induced pluripotent stem cells
hiPSCs differentiation allow generation of somatic cells, including human neurons at early developmental stages. Patient-derived fibroblast can be reprogramed into iPSCs using the "OSKM" factors (Oct3/4, Sox2, Klf4, and c-Myc), and then differentiated into highly pure populations of glutamatergic, GABAergic, dopaminergic, serotonergic or motor neurons, astrocytes, or oligodendrocytes, depending on the transcription factor used. The simultaneous culture of two or more cell types is possible allowing a physiological contextualization and recapitulation of the human biological systems. hiPSC models have been used in ID-related disorders, such as Rett, Fragile-X, Dravet, Phelan-McDermid, Miller Dieker, Angelman, Prader-Willi, Timothy, Williams-Beuren and Lowe syndromes, Friedreich's ataxia, Alexander and Pelizaeus-Merzbaucher diseases, primary microcephaly and X-linked adrenoleukodystrophy (reviewed by Sabitha et al.. The duration of the procedure and the expertise needed, are some of the limitations. Additionally, phenomena such as genetic instability and epigenetic alterations leading to changes in gene expression can occur during the reprogramming procedure, and hamper results interpretation. Furthermore, hiPSCs do not recapitulate behavioural phenotypes, nor the influence of environmental factors or late-onset diseases due to their incomplete maturation.
## Induced neurons
Induced neurons (iNeurons) have shown to be a promising alternative to hiPSCs, as they preserve the original somatic age-related epigenetic landscape. iNeurons resulting from differentiation of mouse embryonic fibroblasts using the transfection factors Ascl1, Brn2, and Myt 1 l (BAM pool)were first developed in 2010. To overcome the need of an invasive sample collection, such as skin biopsy, Tanabe et al.described a method to generate neurons by reprogramming blood nuclear cells (blood iNs). Nevertheless, the necessary co-culture with mouse glia convolutes the interpretation of the results, as these cells can distort neuronal morphologies.
## Genome editing using crispr platforms
Genome editing systems such as the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) are indispensable tools in biological research. The key success in the CRISPR mechanism is the association of a RNA guide (gRNA) and Cas9 protein. While the gRNA, a 20-nt targeting sequence, recognizes DNA sites by base pairing, the Cas9 cleaves DNA, through double-strand breaks (DBS) induction, activating DNA repair mechanisms such as nonhomologous end joining (NHEJ) or homology-directed repair (HDR). Several CRISPR/Cas9-based studies have been carried out in hiPSCs, showing their efficiency and potential (reviewed by Ben Jehuda et al.. The use of CRISPR and hiPSCs simultaneously allows analysis in a donor-independent manner, overcoming the heterogeneity often observed in hiPSCs, due to the specific genetic background, epigenetic factors or other inter-individual differences. One of the limitations of CRISPR/Cas-9 editing system is the off-target effects i.e. Cas-9 binds and cleaves unintentional genomic sites. The "prime editing" combining Cas9 and a reverse transcriptase, allows genome editing without the double-strand DNA breaks collateral effect. CRISPR interference (CRISPRi) and CRISPR activation (CRIS-PRa) have been developed as alternatives for previous genome editing platforms. The CRISPRi/a result from the fusion of the CRISPR technology with a dead nuclease (dCas9), allowing the repression/activation of gene expression at the transcriptional level. These tools, so far eligible for Mendelian disorders, mandate recommendations and guidelines to ensure that human genome editing is used ethically and safely.
## Concluding remarks
Diagnostic approach in a medical genetic setting begins by the observation and categorization of the clinical features. The Human Phenotype Ontology Project (HPO; https:// hpo. jax. org/ app/) terminology gathers a set of terms and codifications of signs, symptoms, and other phenotypic manifestations, contributing to an accurate phenotyping. By adopting this terminology, clinical data can be shared and integrated across the scientific and medical communities, guiding geneticists towards the definition of the ID diagnosis strategy and molecular defect identification. While at this point genotype-phenotype correlation is complex, new ID classification systems have emerged. Kochinke et al.developed a phenotype-based bipartite clinical classification system that interprets the phenotypic heterogeneity characteristic of monogenic ID. Recently, Biesecker et al.suggested a syndrome definition based on the affected gene and phenotypic description. Using the ID gene GLI3 as an example, a clear and simplistic description of several heterogeneous diseases would be GLI3related Pallister-Hall syndrome or GLI3-related Greig cephalopolysyndactyly syndrome. The literature indicates that high ID diagnostic yields are attained by applying the following sequential testing strategy using validated methods, after a detailed clinical evaluation: numeric and structural chromosomal abnormalities analysis, FXS testing, MECP2 (females) and PTEN genes investigation (in the presence of ASDs with macrocephaly), CNVs screening by CMAand exome sequencing. As illustrated in, ID diagnostic yield depends on the technology used, the presence and variability of other clinical features and inheritance pattern. De Brouwer et al.demonstrated the importance of a deep, accurate and homogeneous phenotyping, after diagnosing 42% of patients with XLID. Najmabadi et al., combined microarray analysis and massive parallel sequencing with a diagnostic yield of 57%, in consanguineous families. Nevertheless, caution is warranted when comparing data from different studies, and special attention should be drawn to the heterogeneity of the clinical descriptions and putative bias in patient ascertainment.
ID diagnosis strategy should also include systematic reanalysis of previously generated data, in light of the new knowledge, e.g. databases update, novel disease genes identification, new clinical features and molecular information. This is a clear advantage of the ES / GS over the TNGS. Reanalysis of ES data from 1133 children with severe developmental disorders and their parents, increased the diagnostic yield from 27% (2015) to 40% (2018).
To date, the ID diagnostic yield remains low, and the identification of previously undetected variants in non-coding regions by GS will clarify hitherto some molecularly undiagnosed ID cases. Moreover, the recent development of long-read sequencing (LRS), namely Single-molecule real-time (SMRT) sequencing, using PacBio sequencing (Pacific BioSciences, Menlo Park, CA, USA), and nanopore sequencing, using the MinION instrument (Oxford Nanopore Technologies, Oxford, UK), will fill the gap of massive parallel sequencing, with long reads (over 10 kb), and alignment and mapping errors reduction. LRS improves the identification of structural variants, such as large inversions and translocations, and pseudogenes, as well as precisely sequence long tandem repeat expansions and high GC-rich regions, increases variant phasing determination, allowing the simultaneous establishment of parental origin, inheritance patterns, and disease risk haplotypes.
While the current variant classification guidelines combine functional and clinical data, the stepwise ABC system proposed by Houge et al.suggests a sequential combination of the (A) functional and (B) clinical grades and optionally (C) selection of a standard comment(s) that best address the clinical question. In order to guide clinicians in attaining variant significance, the ABC system can be used separately or as an add-on the ACMG/AMP classification. This highlights the need and the importance of the crosstalk between clinicians and laboratory geneticists to guide genetic investigation, to establish (novel) genotype-phenotype correlations and ultimately to understand the mechanisms underlying the diseases.
The current challenge is the evaluation of the pathogenicity of the variants, rather than their identification. For this purpose, multidisciplinary international research collaborations/cooperation must be established. Ideally, a "rapid" functional test to study several genes in a diagnostic setting, might contribute to overcome this issue. This could represent an important step to translate these insights into future applications that will improve personalized patient support, care and treatment. . |
Background: Congenital bronchial atresia is a rare pulmonary abnormality characterized by the disrupted communication between the central and the peripheral bronchus and is typically asymptomatic. Although it can be symptomatic especially when infections occur in the involved areas, fungal infections are rare complications in patients with bronchial atresia. We report a case of congenital bronchial atresia complicated by a fungal infection.Case presentation:A 30-year-old man with no previous history of immune dysfunction was brought to a nearby hospital and diagnosed with a left lung abscess. Although antimicrobial treatment was administered, it was ineffective, and he was transferred to our hospital. Since diagnostic imaging findings and bronchoscopy suggested congenital bronchial atresia and a fungal infection, he was treated with voriconazole and surgical resection was subsequently performed. A tissue culture detected Aspergillus fumigatus and histopathological findings were compatible with bronchial atresia. After discharge, he remained well and voriconazole was discontinued 5 months after the initiation of therapy.Conclusion:Bronchial atresia is a rare disease that is seldom complicated by a fungal infection, which is also a rare complication; however, physicians should consider fungal infections in patients with bronchial atresia who present with infections resistant to antimicrobial treatment.
# Background
Congenital bronchial atresia is a rare pulmonary abnormality typically characterized by segmental or lobar emphysema with or without mucoid impaction from the obliteration of the bronchus. Bronchial atresia is usually asymptomatic and is coincidentally found in most cases. It can become symptomatic when infections occur; however, they are thought to rarely occur in patients with bronchial atresia. Here, we present a unique case of congenital bronchial atresia complicated by lung abscess due to Aspergillus fumigatus.
## Case presentation
A 30-year-old male with no medical history was admitted to the nearby hospital with a worsening cough and low-grade fever that had persisted for a month. Chest computed tomography (CT) revealed an abscess in the left lower lobe of his lung, and clinical laboratory tests showed elevated white blood cells of 13,200/μl (normal Open Access *Correspondence: [email protected] 1 Department of Infection Control Science, Graduate School of Medicine, Osaka City University, 1-4-3, Asahi-machi, Abeno-ku, Osaka-shi, Osaka 545-8585, Japan Full list of author information is available at the end of the article range, 4000-8000/μl) and C-reactive protein of 5.97 mg/ dl (normal range, < 0.40 mg/dl). Treatment with sulbactam/ampicillin was initiated. Since no improvements were confirmed by the imaging and laboratory findings, antimicrobial treatment was changed to tazobactam/ piperacillin on day 3 post admission, and was subsequently changed to meropenem (MEPM) on day 25 post admission. Treatment with voriconazole (VRCZ) was added on day 17 post admission because his β-D-glucan level reached to 123 pg/ml (normal range, < 20 pg/ml). However, it was discontinued 8 days later. Transbronchial biopsy and bronchoalveolar lavage performed to assess the mass did not detect malignancy or significant levels of bacteria, including mycobacteria and fungi. On day 27, the patient was referred to our tertiary center for further examination and treatment by specialists in respirology and infection and potential thoracic surgery. Since the definitive diagnosis could not be made, the patient was transferred to our hospital for further examination and treatment on day 27.
On admission, his vital signs were as follows: a body temperature of 36.6 °C, blood pressure of 120/66 mmHg, heart rate of 84 beats per minute, and a respiratory rate of 12 breaths per minute. The arterial oxygen saturation measured via pulse oximetry was 98% in the examination room. A physical examination including respiratory sound revealed no significant finding. Clinical laboratory tests showed that the patient had elevated C-reactive protein levels of 15.56 mg/dl (normal range, < 0.40 mg/ dl) and his Aspergillus antigen test was positive (cut-off index of 3.5). Human immunodeficiency virus antibody and interferon-gamma releasing assay were negative. In addition, the β-D-glucan level decreased to a normal range (14.7 pg/ml). Chest X-ray and contrast-enhanced CT were performed on admission. Chest X-ray showed an infiltrative shadow in the left lower lung field, and CT showed a massive lesion in the left lower lobe of the lung, emphysematous changes, ectatic bronchi, and a nodular lesion that was thought to contain a mucoid impaction within its lingular segment. The bronchus was disrupted proximal to the mass lesion of the left lower lobe . Consequently, bronchial atresia in the inferior and anterior basal segments of the left lung was suspected based on imaging findings. Since the presence of a malignant lesion could not be excluded, bronchoscopy was performed on day 2 after transfer to our hospital. An endobronchial ultrasound-guided transbronchial needle aspiration was performed on the small mass lesion near the left second carina . The disrupted bronchus could not be confirmed via bronchoscopy because the disrupted site was located too peripherally. Histopathological analyses revealed no malignant findings, but presence of a filamentous fungus was indicated in Grocott staining . Although cultures did not detect any microorganisms, based on these findings, a lung abscess due to Aspergillus spp. was suspected. Therefore, VRCZ was reintroduced on day 9 after transfer. To perform treatment and confirm the suspected diagnosis of a lung abscess due to an Aspergillus infection, the patient underwent surgical resection of the lingular and basal segments of the left lung on day 22 after transfer. The abscess formation was macroscopically confirmed , and histopathological examination revealed emphysematous changes and ectatic bronchi within the lingular and left basal segments . These findings were compatible with bronchial atresia, and no malignant lesions were confirmed. Although histopathological analyses did not reveal filamentous fungus, a tissue culture resulted in A. fumigatus growth. MEPM and intravenous VRCZ were continued for a total of 4 weeks and 3 weeks, respectively, and the patient was discharged on day 32 after transfer with continuing oral treatment with VRCZ. VRCZ was discontinued 5 months after the initiation of treatment because of poor medication compliance and the side effect of alopecia, and subsequent recurrence of the infection has not been confir med.
# Discussion and conclusions
Congenital bronchial atresia is a rare pulmonary abnormality that was first described bywhich is characterized by an atretic bronchus that does not communicate with the central airways. It has been hypothesized that congenital bronchial atresia occurs as a result of focal ischemia of the developing bronchus due to vascular injury at 15 weeks gestation. Secondary bronchial atresia has been reported to be caused by hypertrophy of bronchial mucosa, external compression of bronchi by abnormal vascular flow, and tuberculosis. Although the precise prevalence of congenital bronchial atresia is unclear, it is estimated to occur in 1.2 per 100,000 males. It is more common in male than in female patients (male:female = 16:9). The condition often occurs at either the segmental or the subsegmental levels, and rarely occurs at the lobar level. Based on CT findings, bronchial atresia was presumed to be at the subsegmental level in this case.
In patients with congenital bronchial atresia, the lung distal to the atretic bronchus becomes hyperinflated due to collateral ventilation through pores of Kohn and canals of Lambert, which results in emphysematous changes. A mucocele can also be found distal to the atretic point. Reflecting these changes, CT imaging typically reveals a mucocele, ectatic bronchi, and emphysematous changes. The identification of the blind-ending bronchus via bronchoscopy is also helpful for diagnosis. However, Wang et al. reported that only 50% of patients with congenital bronchial atresia could be diagnosed using bronchoscopy. In this case, we could not visually confirm the atretic bronchus on bronchoscopy.
In the present case, the presence of any disease capable of causing secondary bronchial atresia was not confirmed; therefore, the patient was likely affected by the congenital form of the disorder. Additionally, mucoceles were considered to have developed to the abscess. Ectatic bronchi and emphysematous changes occurred in the left lower lobe of the lung. These findings were confirmed by histology; however, they were not confirmed by CT presumably because of lung compression caused by the abscess.
Congenital bronchial atresia has been associated with congenital, non-pulmonary anomalies, such as pectus excavatum. The disorder is typically asymptomaticbecause infections are less likely to develop throughout the involved areas as a result of disrupted communication with the central bronchus. The most commonly reported clinical manifestation in symptomatic patients is recurrent pulmonary infection. Although most infections are bacterial, Mycobacterium avium and cytomegalovirus infections have been reported to be infectious complications in patients with congenital bronchial atresia. Additionally, pneumothorax can develop due to rupture of hyperinflated lung tissue, and pulmonary hypertension has been reported to result from the ventilation/perfusion mismatch within hyperinflated lobes of patients. To our knowledge, there has been only one report of a case of chronic pulmonary aspergillosis in patients with congenital bronchial atresia. Therefore, fungal infections in bronchial atresia are presumed to be rare. In this case, the patient had an underlying disease of breast cancer, which could cause immunodeficiency. Furthermore, the patient was asymptomatic, and the disease was coincidentally observed. In contrast, in our case, the patient was immunocompetent with no underlying immunocompromising diseases and had a symptom of persistent fever. Considering that cellmediated immune defects are a risk factor for fungal lung abscesses, abscess formation as a result of the fungal infection in this patient with no immune dysfunctions was unique.
No standardized guidelines for the treatment for bronchial atresia have been created. Some clinicians prefer to perform surgery on asymptomatic patients to prevent long-term consequences of the disorder, such as compression and damage to intact adjacent lung parenchyma. However, follow-up with regular chest X-rays is a reasonable strategy for monitoring asymptomatic patients. Surgical resection should be considered in cases with recurrent and severe symptoms of infection, that cannot be treated effectively pharmacologically, and to exclude the possibility of the formation of malignant lesions. The best duration of treatment suited for eliminating an Aspergillus lung abscess also remains undetermined. A minimum of 6 months of antifungal therapy is recommended for chronic pulmonary aspergillosis. In addition, surgical resection is also recommended for localized disease. We treated the patient using a b a Histopathological assessment of the resected lung showed emphysematous change and ectatic bronchus (arrows) of left basal segments (a, b) (Hematoxylin and Eosin staining, × 40) combination of antifungal and surgical treatment similar to recommendations for the treatment of chronic pulmonary aspergillosis. Subsequently, the disease followed a benign course despite the early discontinuation of antifungal treatment.
In conclusion, we experienced a case of congenital bronchial atresia complicated by a lung abscess due to an A. fumigatus infection. Supposedly, this case was unique compared with the previously reported case in that the fungal infection occurred in an immunocompetent patient and was symptomatic. Congenital bronchial atresia is a rare disease and is usually asymptomatic. However, it can cause pulmonary infections including fungal infections. Therefore, when infections occur in patients with bronchial atresia, physicians should consider fungal infections especially when antimicrobial treatment is ineffective.
Abbreviations CT: Computed tomography; MEPM: Meropenem; VRCZ: Voriconazole. |
Hsp90 Governs Dispersion and Drug Resistance of Fungal Biofilms
Fungal biofilms are a major cause of human mortality and are recalcitrant to most treatments due to intrinsic drug resistance. These complex communities of multiple cell types form on indwelling medical devices and their eradication often requires surgical removal of infected devices. Here we implicate the molecular chaperone Hsp90 as a key regulator of biofilm dispersion and drug resistance. We previously established that in the leading human fungal pathogen, Candida albicans, Hsp90 enables the emergence and maintenance of drug resistance in planktonic conditions by stabilizing the protein phosphatase calcineurin and MAPK Mkc1. Hsp90 also regulates temperature-dependent C. albicans morphogenesis through repression of cAMP-PKA signalling. Here we demonstrate that genetic depletion of Hsp90 reduced C. albicans biofilm growth and maturation in vitro and impaired dispersal of biofilm cells. Further, compromising Hsp90 function in vitro abrogated resistance of C. albicans biofilms to the most widely deployed class of antifungal drugs, the azoles. Depletion of Hsp90 led to reduction of calcineurin and Mkc1 in planktonic but not biofilm conditions, suggesting that Hsp90 regulates drug resistance through different mechanisms in these distinct cellular states. Reduction of Hsp90 levels led to a marked decrease in matrix glucan levels, providing a compelling mechanism through which Hsp90 might regulate biofilm azole resistance. Impairment of Hsp90 function genetically or pharmacologically transformed fluconazole from ineffectual to highly effective in eradicating biofilms in a rat venous catheter infection model. Finally, inhibition of Hsp90 reduced resistance of biofilms of the most lethal mould, Aspergillus fumigatus, to the newest class of antifungals to reach the clinic, the echinocandins. Thus, we establish a novel mechanism regulating biofilm drug resistance and dispersion and that targeting Hsp90 provides a much-needed strategy for improving clinical outcome in the treatment of biofilm infections.
# Introduction
In recent decades, fungal pathogens have emerged as a predominant cause of human disease, especially in immunocompromised individuals. The number of acquired fungal bloodstream infections has increased by ,207% in this timeframe [bib_ref] Stress, drugs, and evolution: the role of cellular signaling in fungal drug..., Cowen [/bib_ref] [bib_ref] Epidemiology of invasive mycoses in North America, Pfaller [/bib_ref] [bib_ref] Trends in mortality due to invasive mycotic diseases in the United States, Mcneil [/bib_ref]. Although diverse species are capable of causing infection, a few prevail as the most prevalent cause of disease. Candida and Aspergillus species together account for ,70% of all invasive fungal infections, with Candida albicans and Aspergillus fumigatus prevailing as the leading causal agents of opportunistic mycoses [bib_ref] Epidemiology of invasive mycoses in North America, Pfaller [/bib_ref]. Candida species are the fourth leading cause of hospital acquired bloodstream infections in the United States with mortality rates estimated at 40% [bib_ref] The epidemiology and attributable outcomes of candidemia in adults and children hospitalized..., Zaoutis [/bib_ref] [bib_ref] Epidemiology of invasive candidiasis: a persistent public health problem, Pfaller [/bib_ref]. The profound economic consequences of Candida infections can be demonstrated by the ,$1.7 billion spent annually on treating candidemia in the United States alone [bib_ref] The direct cost and incidence of systemic fungal infections, Wilson [/bib_ref]. Further, A. fumigatus is the most common etiological agent of invasive aspergillosis, with a 40-90% mortality rate [bib_ref] Aspergillosis case-fatality rate: systematic review of the literature, Lin [/bib_ref]. In patients with pulmonary disorders such as asthma or cystic fibrosis, A. fumigatus infection can cause allergic bronchopulmonary aspergillosis leading to severe complications. For these fungal species, there are numerous factors that contribute to the pathogenicity and recalcitrance of resulting infections to antifungal treatment, including the ability to evolve and maintain resistance to conventional antifungal therapy [bib_ref] Stress, drugs, and evolution: the role of cellular signaling in fungal drug..., Cowen [/bib_ref].
Due to the limited number of drug targets available to exploit in fungal pathogens that are absent or sufficiently divergent in the human host, the vast majority of antifungal drugs in clinical use target ergosterol or its biosynthesis. The azoles are the most widely used class of antifungal in the clinic and function by inhibiting the ergosterol biosynthetic enzyme Erg11, causing a block in the production of ergosterol and the accumulation of the toxic byproduct 14-a-methyl-3,6-diol, culminating in a severe membrane stress [bib_ref] Molecular basis of resistance to azole antifungals, Lupetti [/bib_ref] [bib_ref] An insight into the antifungal pipeline: selected new molecules and beyond, Ostrosky-Zeichner [/bib_ref]. The azoles are generally fungistatic against yeasts, including Candida species, and fungicidal against moulds, such as Aspergillus species. The fungistatic nature of the azoles towards C. albicans culminates in strong directional selection on the surviving population to evolve drug resistance [bib_ref] Evolution of antifungal-drug resistance: mechanisms and pathogen fitness, Anderson [/bib_ref] [bib_ref] The evolution of fungal drug resistance: modulating the trajectory from genotype to..., Cowen [/bib_ref]. In fact, high levels of azole resistance in C. albicans clinical isolates often accumulate through multiple mechanisms including: upregulation of drug efflux pumps, overexpression or alteration of Erg11, or modification of stress response pathways that are crucial for resistance [bib_ref] Stress, drugs, and evolution: the role of cellular signaling in fungal drug..., Cowen [/bib_ref] [bib_ref] Evolution of antifungal-drug resistance: mechanisms and pathogen fitness, Anderson [/bib_ref] [bib_ref] The evolution of fungal drug resistance: modulating the trajectory from genotype to..., Cowen [/bib_ref] [bib_ref] Prevalence of molecular mechanisms of resistance to azole antifungal agents in Candida..., Perea [/bib_ref] [bib_ref] Regulatory circuitry governing fungal development, drug resistance, and disease, Shapiro [/bib_ref]. The echinocandins are the only new class of antifungal to reach the clinic in decades. They act as noncompetitive inhibitors of b-1,3 glucan synthase, an enzyme involved in fungal cell wall synthesis [bib_ref] An insight into the antifungal pipeline: selected new molecules and beyond, Ostrosky-Zeichner [/bib_ref] , resulting in the loss of cell wall integrity and a severe cell wall stress. The impact of the echinocandins is generally opposite to that of the azoles, in that they are fungicidal against yeasts and fungistatic against moulds. Resistance of C. albicans clinical isolates to the echinocandins has been reported and is often associated with mutations in the drug target [bib_ref] Regulatory circuitry governing fungal development, drug resistance, and disease, Shapiro [/bib_ref] [bib_ref] Assessing resistance to the echinocandin antifungal drug caspofungin in Candida albicans by..., Balashov [/bib_ref] [bib_ref] Specific substitutions in the echinocandin target Fks1p account for reduced susceptibility of..., Park [/bib_ref].
An additional key factor responsible for the virulence and drug resistance of C. albicans and A. fumigatus is their tendency to form biofilms on medical devices that are highly resistant to antifungal treatment [bib_ref] Genetic control of Candida albicans biofilm development, Finkel [/bib_ref] [bib_ref] Our current understanding of fungal biofilms, Ramage [/bib_ref] [bib_ref] How to build a biofilm: a fungal perspective, Blankenship [/bib_ref] [bib_ref] Genetics and genomics of Candida albicans biofilm formation, Nobile [/bib_ref]. The use of such medical devicessuch as venous catheters, urinary catheters and artificial jointshas dramatically risen to more than 10 million recipients per year [bib_ref] Candida biofilms on implanted biomaterials: a clinically significant problem, Ramage [/bib_ref] [bib_ref] Candida infections of medical devices, Kojic [/bib_ref]. This poses a severe clinical problem as C. albicans is the third leading cause of intravascular catheter-related infections, and has the overall highest crude mortality rate of ,30% for deviceassociated infections [bib_ref] Our current understanding of fungal biofilms, Ramage [/bib_ref] [bib_ref] Candida infections of medical devices, Kojic [/bib_ref] [bib_ref] Candidemia at a tertiary-care hospital: epidemiology, treatment, clinical outcome and risk factors..., Viudes [/bib_ref]. Further, A. fumigatus infections have been reported on medical implant devices as well as on bronchial epithelial cells [bib_ref] Our current understanding of fungal biofilms, Ramage [/bib_ref] [bib_ref] Aspergillus fumigatus forms biofilms with reduced antifungal drug susceptibility on bronchial epithelial..., Seidler [/bib_ref]. The inherent drug resistance of biofilms often necessitates surgical removal of the infected medical devices in order to eradicate the fungal infection.
Extensive research has focused on mechanisms of drug resistance in C. albicans biofilms, and it is apparent that cells in a fungal biofilm represent an epigenetic modification of the cellular state compared to their planktonic counterparts, with changes in cellular morphology, cell-to-cell communication, and gene expres-sion, as well as with the production of an extra-cellular matrix [bib_ref] Genetic control of Candida albicans biofilm development, Finkel [/bib_ref] [bib_ref] How to build a biofilm: a fungal perspective, Blankenship [/bib_ref] [bib_ref] Genetics and genomics of Candida albicans biofilm formation, Nobile [/bib_ref]. Multiple factors contribute to the elevated drug resistance of C. albicans biofilms. These factors include increased cell density [bib_ref] Role for cell density in antifungal drug resistance in Candida albicans biofilms, Perumal [/bib_ref] , increased expression of drug efflux pumps [bib_ref] Investigation of multidrug efflux pumps in relation to fluconazole resistance in Candida..., Ramage [/bib_ref] [bib_ref] Mechanism of fluconazole resistance in Candida albicans biofilms: phase-specific role of efflux..., Mukherjee [/bib_ref] , decreased ergosterol content [bib_ref] Mechanism of fluconazole resistance in Candida albicans biofilms: phase-specific role of efflux..., Mukherjee [/bib_ref] , elevated b-1,3 glucan levels in the cell wall and biofilm matrix [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] [bib_ref] Genetic basis of Candida biofilm resistance due to drug-sequestering matrix glucan, Nett [/bib_ref] , as well as signalling mediated by protein kinase C (PKC) [bib_ref] A contact-activated kinase signals Candida albicans invasive growth and biofilm development, Kumamoto [/bib_ref] and the protein phosphatase calcineurin [bib_ref] Synergistic effect of calcineurin inhibitors and fluconazole against Candida albicans biofilms, Uppuluri [/bib_ref].
The molecular chaperone Hsp90 regulates complex cellular circuitry in eukaryotes by stabilizing regulators of cellular signalling [bib_ref] HSP90 at the hub of protein homeostasis: emerging mechanistic insights, Taipale [/bib_ref] [bib_ref] Structure and mechanism of the Hsp90 molecular chaperone machinery, Pearl [/bib_ref]. As a consequence, inhibiting Hsp90 disrupts a plethora of cellular processes and has broad therapeutic potential against diverse eukaryotic pathogens including the protozoan parasites Plasmodium falciparum and Trypanosoma evansi as well as numerous fungal species [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref] [bib_ref] Heat shock protein 90 as a drug target against protozoan infections: biochemical..., Pallavi [/bib_ref] [bib_ref] A repurposing strategy identifies novel synergistic inhibitors of Plasmodium falciparum heat shock..., Shahinas [/bib_ref]. In the planktonic state, Hsp90 potentiates the emergence and maintenance of resistance to azoles and echinocandins in C. albicans at least in part via calcineurin [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref] ; Hsp90 physically interacts with the catalytic subunit of calcineurin, keeping it stable and poised for activation [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref]. Recently, Hsp90 was also shown to enable azole and echinocandin resistance in C. albicans via the PKC cell wall integrity pathway [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. Hsp90 depletion results in the destabilization of the terminal mitogen-activated protein kinase (MAPK) Mkc1, providing the second Hsp90 client protein implicated in drug resistance [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. Compromising C. albicans Hsp90 function renders drug-resistant isolates susceptible in vitro and improves the therapeutic efficacy of antifungals in a Galleria mellonella model of C. albicans pathogenesis and a murine model of disseminated candidiasis [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref]. Compromising A. fumigatus Hsp90 also enhances the efficacy of echinocandins both in vitro and in the G. mellonella model of infection [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref]. Notably, Hsp90 regulates not only drug resistance in C. albicans but also the morphogenetic transition between yeast and filamentous growth, a trait important for virulence [bib_ref] Systematic screens of a Candida albicans homozygous deletion library decouple morphogenetic switching..., Noble [/bib_ref]. Compromising Hsp90 function induces filamentation by relieving Hsp90-mediated repression of cAMP-protein kinase A (PKA) signalling [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref]. The ability to transition between morphological states is also critical for biofilm formation and development [bib_ref] The filamentation pathway controlled by the Efg1 regulator protein is required for..., Ramage [/bib_ref].
Given that Hsp90 governs fungal morphogenesis and drug resistance in planktonic conditions, we sought to investigate if this molecular chaperone also regulates the development and drug resistance of biofilms. We discovered that genetically compromising Hsp90 function reduced but did not block biofilm maturation in vitro and had minimal impact on the ability of C. albicans to form robust biofilms in an in vivo rat catheter model,. Genetic depletion of C. albicans Hsp90 reduced biofilm dispersal, with the few dispersed cells being largely inviable. Moreover, compromising C. albicans Hsp90 function genetically or pharmacologically transformed the azole fluconazole from ineffectual to highly efficacious in eradicating biofilms both in vitro and in a rat catheter model of infection. In stark contrast to planktonic conditions, reduction of C. albicans Hsp90 levels genetically in biofilm conditions did not lead to depletion of the client proteins calcineurin or Mkc1, suggesting that Hsp90 regulates drug resistance through distinct mechanisms in these different cellular states. Genetic depletion of Hsp90 reduced glucan levels in the biofilm matrix, providing a compelling mechanism by which Hsp90 might regulate biofilm drug resistance. Finally, in the most lethal mould, A. fumigatus, compromising Hsp90 function enhanced the efficacy of azoles and echinocandins in an in vitro model. Our results implicate Hsp90 as a novel regulator of biofilm dispersion and drug resistance, and provide strong support for the utility of Hsp90 inhibitors as a therapeutic strategy for biofilm infections caused by diverse fungal species.
## Author summary
Candida albicans and Aspergillus fumigatus are the most common causative agents of fungal infections worldwide. Both species can form biofilms on host tissues and indwelling medical devices that are highly resistant to antifungal treatment. Here we implicate the molecular chaperone Hsp90 as a key regulator of biofilm dispersion and drug resistance. Compromising Hsp90 function reduced biofilm formation of C. albicans in vitro and impaired dispersal of biofilm cells, potentially blocking their capacity to serve as reservoirs for infection. Further, compromise of Hsp90 function abrogated resistance of C. albicans biofilms to the most widely deployed class of antifungal, the azoles, both in vitro and in a mammalian model of catheter-associated candidiasis. Key drug resistance regulators were depleted upon reduction of Hsp90 levels in planktonic but not biofilm conditions, suggesting that Hsp90 regulates drug resistance through different mechanisms in these distinct cellular states. Reduction of Hsp90 markedly reduced levels of matrix glucan, a carbohydrate important for C. albicans biofilm drug resistance. Inhibition of Hsp90 also reduced resistance of A. fumigatus biofilms to the newest class of antifungal, the echinocandins. Thus, targeting Hsp90 provides a promising strategy for the treatment of biofilm infections caused by diverse fungal species.
# Results
Hsp90 is not required for C. albicans biofilm formation in vitro or in vivo Due to the key roles of Hsp90 in both morphogenesis and drug resistance under planktonic conditions [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref] [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref] , we hypothesized that Hsp90 might also regulate C. albicans biofilm formation and drug resistance. First, we tested whether compromising Hsp90 function affected biofilm growth. To do this, C. albicans biofilms were cultured for 24 hours in static 96 well microtiter plates, washed to remove non-adherent cells, grown for an additional 24 hours with various concentrations of the Hsp90 inhibitor geldanamycin, and growth was quantified by metabolic activity using an XTT reduction assay [bib_ref] Standardized method for in vitro antifungal susceptibility testing of Candida albicans biofilms, Ramage [/bib_ref]. The geldanamycin was added at 24 hours rather than at the initial time point as is the standard for biofilm drug studies since the initial cells are planktonic and much more susceptible to drugs than their biofilm counterparts [bib_ref] Synergistic effect of calcineurin inhibitors and fluconazole against Candida albicans biofilms, Uppuluri [/bib_ref] [bib_ref] Standardized method for in vitro antifungal susceptibility testing of Candida albicans biofilms, Ramage [/bib_ref] ; consistent with this, initial attempts to include geldanamycin during inoculation led to a toxicity profile identical to that of planktonic cells (data not shown). When geldanamycin was added at 24 hours, no significant differences in metabolic activity were observed at a variety of concentrations tested up to 100 mg/mL (P.0.05, ANOVA, Bonferroni's Multiple Comparison Test, [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Thus, Hsp90 inhibitors do not compromise biofilm development.
To further explore Hsp90's role in biofilm formation, we exploited a strain of C. albicans in which Hsp90 levels could be depleted by tetracycline-mediated transcriptional repression (tetO-HSP90/hsp90D). Biofilms of the wild type and tetO-HSP90/hsp90D strain were cultured in static 96 well microtiter plates with or without 20 mg/mL of the tetracycline analog doxycycline from the time of inoculation. Doxycycline was included at this early point given the time required for transcriptional repression to manifest in depletion of Hsp90, and enabled by the absence of toxicity in planktonic cells. Doxycycline-mediated transcriptional repression of Hsp90 decreased biofilm development, but did not block formation of a mature biofilm [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref] , P,0.01). We observed comparable results when biofilms were cultured on silicon elastomer squares, and when biofilm growth was monitored by XTT reduction or by dry weight [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref] and B). To determine if depletion of Hsp90 prior to inoculation had a more profound effect on biofilm formation, we performed a comparable assay but in the presence or absence of doxycycline in the overnight culture. Depletion of Hsp90 prior to inoculation did not further reduce biofilm formation but rather led to a biofilm indistinguishable from the no doxycycline control [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Although Hsp90 is essential, this dose of doxycycline causes reduced growth rate of the tetO-HSP90/hsp90D strain in planktonic cultures but has little effect on stationary phase cell density [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref]. Western blot analysis validated that Hsp90 levels were dramatically reduced in biofilms formed by the tetO-HSP90/ hsp90D strain when cultured in the presence of doxycycline [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. We note that when biofilms were formed under shaking conditions, the tetO-HSP90/hsp90D strain had reduced biofilm growth, which was exacerbated in the presence of doxycycline [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Thus, while Hsp90's impact on biofilm development can vary, under most conditions tested compromising Hsp90 function does not block biofilm formation in vitro. Biofilms were grown in 96-well microtiter plates in RPMI at 37uC. After 24 hours wells were washed with PBS to remove non-adherent cells and fresh RPMI medium was added containing various concentrations of the Hsp90 inhibitor geldanamycin (GdA). Biofilms were grown for an additional 24 hours at 37uC. Metabolic activity was measured using an XTT reduction assay and quantified by measuring absorbance at 490 nm. Error bars represent standard deviations of five technical replicates. Biofilm growth in the presence of GdA was not significantly different from the untreated control (P.0.05, ANOVA, Bonferroni's Multiple Comparison Test). (B) Strains of C. albicans were grown in 96-well microtiter plates in RPMI at 37uC for 24 hours with or without 20 mg/mL doxycycline (DOX). Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Doxycycline-mediated transcriptional repression of HSP90 in the tetO-HSP90/hsp90D strain yielded a small reduction in biofilm growth (P,0.01). Asterisk indicates P,0.01 compared to all other conditions. Error bars represent standard deviations from five technical replicates. (C) Hsp90 levels are dramatically reduced in a C. albicans biofilm upon treatment of the tetO-HSP90/hsp90D strain with 20 mg/mL doyxcycline in RPMI at 37uC. Total protein was resolved by SDS-PAGE and blots were hybridized with a-Hsp90 and a-tubulin as a loading control. doi:10.1371/journal.ppat.1002257.g001
In order to address the role of Hsp90 in biofilm growth in vivo, biofilm formation was examined using a rat venous catheter model of biofilm-associated candidiasis that mimics central venous catheters in patients [bib_ref] Development and characterization of an in vivo central venous catheter Candida albicans..., Andes [/bib_ref]. Infection of implanted catheters with C. albicans was performed by intraluminal instillation, catheters were flushed after 6 hours, and biofilm formation was monitored with or without 20 mg/mL doxycycline after 24 hours. The tetO-HSP90/hsp90D strain was capable of establishing a biofilm in the rat venous catheter, as visualized by scanning electron microscopy [fig_ref] Figure 2: Genetic depletion of Hsp90 does not block C [/fig_ref]. Further, transcriptional repression of HSP90 with doxycycline did not block the formation of a robust biofilm [fig_ref] Figure 2: Genetic depletion of Hsp90 does not block C [/fig_ref]. These results demonstrate that compromising Hsp90 function does not impair the ability of C. albicans to form mature biofilms in vivo.
Compromising Hsp90 function produces biofilms with altered morphologies As mentioned above, Hsp90 is a key regulator of the yeast to filament transition in C. albicans [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref] , a process implicated in virulence and biofilm formation [bib_ref] The filamentation pathway controlled by the Efg1 regulator protein is required for..., Ramage [/bib_ref]. Therefore, we examined the architecture of geldanamycin treated biofilms cultured on silicon elastomer squares to enable imaging by confocal microscopy. Biofilms treated with geldanamycin had decreased thickness of the bottom yeast layer (30 mm and 45 mm versus 90 mm and 100 mm in the untreated control, P = 0.0237, t-test) without substantial change in the thickness of the upper layer of filaments . That a greater proportion of the biofilm thickness was occupied by filaments compared to yeast suggests that Hsp90 inhibition might lead to enhanced filamentation in biofilms. Moreover, biofilms treated with geldanamycin showed more polarized filaments extending away from the biofilm basal surface compared to the interconnected meshwork of filaments in an untreated control . That biofilms formed upon Hsp90 inhibition had a greater proportion of their total thickness occupied by filaments compared to yeast is consistent with Hsp90's repressive effect on filamentation in planktonic conditions. Hsp90 function is important for dispersal of C. albicans biofilms Based on our finding that C. albicans biofilms display altered morphologies upon Hsp90 inhibition, we sought to evaluate the effect of Hsp90 function on biofilm dispersion given that morphogenesis plays a critical role in this process [bib_ref] The transcriptional regulator Nrg1p controls Candida albicans biofilm formation and dispersion, Uppuluri [/bib_ref] [bib_ref] Dispersion as an important step in the Candida albicans biofilm developmental cycle, Uppuluri [/bib_ref]. We monitored dispersion of yeast cells using the only well validated model which involves culturing biofilms on silicon elastomer under conditions of flow [bib_ref] Design of a simple model of Candida albicans biofilms formed under conditions..., Uppuluri [/bib_ref] [bib_ref] An easy and economical in vitro method for the formation of Candida..., Uppuluri [/bib_ref]. When biofilms were cultured in the absence of doxycycline with the tetO-HSP90/hsp90D strain, the number of dispersed cells after 1 hour was 90,000 cells/mL and remained fairly constant over a 24 hour time period [fig_ref] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell... [/fig_ref]. In contrast, in the presence of 20 mg/mL doxycycline the number of dispersed cells was dramatically reduced to approximately 17,000 cells/mL throughout the 24 hours (P = 0.0022, t-test, [fig_ref] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell... [/fig_ref]. We confirmed that the effects of doxycycline were . Pharmacological inhibition of Hsp90 alters C. albicans biofilms architecture. C. albicans cells were grown on silicon elastomer squares in RPMI at 37uC for 24 hours. C. albicans wild-type biofilms were left untreated (A), or treated with 10 mg/mL geldanamycin (GdA) for 48 hours (B). Biofilms were stained with concanavalin A conjugate for confocal scanning laser microscopy visualization, and image reconstructions were created to provide side views (top panel). Representative images are shown. Confocal scanning laser microscopy depth views were artificially coloured (middle panel) with blue representing within 10 mm from the silicon, orange representing approximately 300 mm from the silicon, and red representing over 400 mm from the silicon. Scanning electron microscopy images are shown in bottom panel. Biofilms treated with GdA show a thinner lower layer of yeast than the untreated control. doi:10.1371/journal.ppat.1002257.g003 specifically due to transcriptional repression of HSP90, as doxycycline had no impact on biofilm dispersal of the wild-type strain lacking the tetO promoter [fig_ref] Figure 2: Genetic depletion of Hsp90 does not block C [/fig_ref]. Intriguingly, the cells that were dispersed upon reduction of Hsp90 levels had major viability defects compared to their untreated counterparts (P = 0.007, t-test) with only 55% viable at 1 hour, 5% viable at 12 hours, and less than 1% viable at 24 hours [fig_ref] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell... [/fig_ref]. The dramatic reduction in viability was specific to the dispersed cell population with doxycycline-mediated transcriptional repression of HSP90, as the viability of dispersed cells in the untreated control remained close to 50% even at 24 hours [fig_ref] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell... [/fig_ref]. Viability was unaffected when a wild-type strain lacking the tetO promoter was treated with doxycycline, confirming that the effects observed were due to transcriptional repression of HSP90 [fig_ref] Figure 2: Genetic depletion of Hsp90 does not block C [/fig_ref]. The reduced viability upon reduction of Hsp90 levels was specific to the dispersed cell population within the biofilm, as there was only a minor defect in overall metabolic activity of the tetO-HSP90/ hsp90D biofilms in the presence of doxycycline [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Further, under planktonic conditions viability remained.85% when the tetO-HSP90/hsp90D strain was grown in the presence of doxycycline for 24 hours. Taken together, Hsp90 plays a critical role in the dispersal step of the biofilm life cycle and is crucial for survival of dispersed cells.
## Hsp90 enables the resistance of c. albicans biofilms to fluconazole in vitro
Genetic or pharmacological compromise of Hsp90 function renders C. albicans susceptible to azoles and echinocandins under planktonic conditions [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref] [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref] [bib_ref] Genetic architecture of Hsp90-dependent drug resistance, Cowen [/bib_ref]. Since compromising Hsp90 function pharmacologically did not impair biofilm maturation, we investigated whether inhibition of Hsp90 would alter biofilm drug resistance using the standard 96 well microtiter plate static assay that enables testing many drug concentrations. We focused on the azoles, since biofilms are notoriously resistant to this class of drugs, compromising their therapeutic utility. As a positive control, a wild-type C. albicans biofilm was subjected to a gradient of concentrations of the calcineurin inhibitor FK506 in addition to a gradient of fluconazole, a drug combination with established synergistic activity against C. albicans biofilms [bib_ref] Synergistic effect of calcineurin inhibitors and fluconazole against Candida albicans biofilms, Uppuluri [/bib_ref]. We confirmed Strains of C. albicans were grown in 96-well microtiter plates in RPMI at 37uC. After 24 hours cells were washed with PBS to remove non-adherent cells and fresh medium was added with varying concentrations of the azole fluconazole (FL) and either the calcineurin inhibitor FK506 or the Hsp90 inhibitor geldanamycin (GdA) in a checkerboard format. Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. The FIC index was calculated as indicated in [fig_ref] Table 1: Inhibition of calcineurin or Hsp90 has synergistic activity with fluconazole against wild-type... [/fig_ref]. Bright green represents growth above the MIC 50 , dull green represents growth at the MIC 50 , and black represents growth below the MIC [bib_ref] Hsp90: a specialized but essential proteinfolding tool, Young [/bib_ref]. Data was quantitatively displayed with colour using the program Java TreeView 1.1.3 (http://jtreeview.sourceforge.net). Inhibiting calcineurin or Hsp90 function has synergistic activity with fluconazole. (B) Strains of C. albicans were grown in 96-well microtiter plates in RPMI at 37uC. When indicated, 20 mg/mL doxycycline (DOX) was added to the medium. After 24 hours cells were washed with PBS to remove non-adherent cells and fresh medium was added with varying concentrations of fluconazole. Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Genetic depletion of Hsp90 reduces the MIC 50 of fluconazole to a greater extent than deletion of its client proteins calcineurin or Mkc1. doi:10.1371/journal.ppat.1002257.g005 synergistic activity of FK506 with fluconazole by measuring metabolic activity using the XTT reduction assay [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. Biofilms were extremely susceptible to the combination of inhibitors with a calculated FIC index of 0.1093, indicating potent synergy [fig_ref] Table 1: Inhibition of calcineurin or Hsp90 has synergistic activity with fluconazole against wild-type... [/fig_ref]. To determine if Hsp90 enables biofilm azole resistance, we used an equivalent experiment but with a gradient of concentrations of the Hsp90 inhibitor geldanamycin and a gradient of fluconazole. Geldanamycin exhibited potent synergy with fluconazole, dramatically reducing azole resistance at only 3.125 mg/mL geldanamycin. Maximal effects were observed with 12.5 mg/mL geldanamycin, which reduced the MIC 50 of fluconazole from .1000 mg/mL to 31.25 mg/mL [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. Further, FIC indexes as low as 0.125 to 0.156 were calculated for the combination of fluconazole and geldanamycin confirming that inhibition of Hsp90 has a potent synergistic effect with azoles against C. albicans biofilms [fig_ref] Table 1: Inhibition of calcineurin or Hsp90 has synergistic activity with fluconazole against wild-type... [/fig_ref]. Next, we utilized the tetO-HSP90/hsp90D strain in order to validate that the synergistic activity of geldanamycin with fluconazole against C. albicans biofilms was indeed due to Hsp90 inhibition. Biofilms of a wild-type strain of C. albicans had a fluconazole MIC 50 of over 512 mg/mL [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. Deletion of one allele of HSP90 or replacing the promoter of the sole remaining HSP90 allele with the tetracycline-repressible promoter had no impact on fluconazole resistance [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. However, upon depletion of Hsp90 by doxycycline-mediated transcriptional repression in the tetO-HSP90/hsp90D strain, the fluconazole MIC 50 was dramatically reduced to only 8 mg/mL, a .60-fold increase in fluconazole sensitivity [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. Hence, both pharmacological and genetic evidence confirms that Hsp90 function is critical for azole resistance of C. albicans biofilms.
The role of downstream effectors of Hsp90 in azole resistance of C. albicans biofilms To further dissect the mechanism by which Hsp90 regulates azole resistance of C. albicans biofilms, we repeated the drug susceptibility assay with strains lacking specific Hsp90 client proteins. Under planktonic conditions both calcineurin and Mkc1 are important Hsp90 client proteins that regulate the maintenance of azole resistance [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref] [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref] [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. Moreover, these client proteins have previously been shown to be important for azole resistance of C. albicans biofilms [bib_ref] A contact-activated kinase signals Candida albicans invasive growth and biofilm development, Kumamoto [/bib_ref] [bib_ref] Synergistic effect of calcineurin inhibitors and fluconazole against Candida albicans biofilms, Uppuluri [/bib_ref]. We found that biofilms formed by strains lacking the catalytic subunit of calcineurin (cna1D/cna1D) or the terminal MAPK of the PKC cell wall integrity signalling pathway (mkc1D/mkc1D) had fluconazole MIC 50 values of 32 mg/mL and 128 mg/mL, respectively; their fluconazole resistance levels were intermediate between the robust resistance of the wild-type parental strain and the sensitivity observed upon impairment of Hsp90 function [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. The finding that compromise of calcineurin function does not confer as severe a reduction in biofilm fluconazole resistance as compromise of Hsp90 function is intriguing in light of the fact that under all planktonic conditions tested, inhibition of calcineurin phenocopies inhibition of Hsp90 in terms of azole resistance [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref] [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref] [bib_ref] Genetic architecture of Hsp90-dependent drug resistance, Cowen [/bib_ref]. These results suggest that calcineurin and Mkc1 may be able to partially compensate for the loss of the other client during times of azole-induced stress in a biofilm environment. Alternatively, these findings could be explained by the existence of a novel downstream effector of Hsp90 important for azole resistance of C. albicans biofilms.
To further investigate the mechanisms by which Hsp90 regulates azole resistance in biofilm conditions, we examined protein levels of the client proteins calcineurin and Mkc1 upon Hsp90 depletion. Strains were cultured in RPMI medium for both planktonic and biofilm growth. Biofilms were cultured on plastic under static conditions, as with our drug studies. We previously established that under planktonic conditions genetic reduction of Hsp90 levels leads to depletion of the catalytic subunit of calcineurin (Cna1) and Mkc1 [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref] [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. Here, the tetO-HSP90/hsp90D strain was grown in either planktonic or biofilm conditions in the presence or absence of 20 mg/mL doxycycline for 48 hours. Under both conditions, Hsp90 levels were dramatically reduced in the presence of doxycycline . To monitor calcineurin levels, we used a C-terminal 6xHis-FLAG epitope tag on Cna1 in the tetO-HSP90/hsp90D strain. In the tagged strains, Cna1 levels were comparable under planktonic and biofilm conditions in the absence of doxycycline . Doxycycline-mediated reduction of Hsp90 levels led to an ,90% reduction in Cna1 in planktonic conditions, however, Cna1 levels remained stable in biofilm conditions . All strains had comparable amounts of protein loaded, as confirmed with a tubulin loading control. To monitor total Mkc1 levels, we used a C-terminal 6xHis-FLAG epitope tag on Mkc1 in the tetO-HSP90/ hsp90D strain. Mkc1 levels were comparable in the tagged strains under planktonic and biofilm conditions in the absence of doxycycline . As with Cna1, doxycycline-mediated reduction of Hsp90 levels led to ,80% reduction in Mkc1 levels in planktonic conditions, however, Mkc1 levels remained stable in biofilm conditions . We next addressed whether depletion of Hsp90 affected levels of activated, dually phosphorylated Mkc1. Mkc1 was activated in all strains in the absence of doxycycline. As with total Mkc1 levels, doxycycline-mediated reduction of Hsp90 led to a reduction in levels of activated Mkc1 in planktonic conditions, however, Mkc1 remained activated in biofilm conditions. Taken together, these results suggest that Hsp90 may play different roles in client protein regulation in these distinct cellular states, and also that these client proteins may have other means of maintaining stability in a biofilm environment.
Hsp90 regulates matrix glucan levels in C. albicans biofilms Given our findings that Hsp90 client proteins remain stable in a biofilm, irrespective of Hsp90 levels, and that deletion of these client proteins does not phenocopy Hsp90 depletion in terms of biofilm azole resistance, we hypothesized that Hsp90 also regulates biofilm drug resistance through a mechanism independent of calcineurin and Mkc1 signalling. Recent studies established that glucan present in the biofilm matrix is critical for azole resistance due its capacity to sequester fluconazole, preventing it from reaching its intracellular target [bib_ref] Genetic basis of Candida biofilm resistance due to drug-sequestering matrix glucan, Nett [/bib_ref]. Consequently, we investigated whether Hsp90 affects glucan levels in the biofilm matrix. Biofilms were cultured on plastic in static conditions in the presence or absence of 20 mg/mL doxycycline for 48 hours, matrix material was harvested from biofilms with equivalent metabolic activity, and b-1,3 glucan levels were quantified. In the tetO-HSP90/hsp90D strain, the level of glucan in the biofilm matrix was ,6,000 pg/mL in the absence of doxycycline [fig_ref] Figure 7: Depletion of Hsp90 reduces biofilm matrix glucan [/fig_ref]. Transcriptional repression of HSP90 with 20 mg/mL doxycycline led to reduced glucan levels of only ,3,700 pg/mL (P,0.01, ANOVA, Bonferroni's Multiple Comparison Test, [fig_ref] Figure 7: Depletion of Hsp90 reduces biofilm matrix glucan [/fig_ref]. Doxycycline had no impact on matrix glucan levels of a wild-type strain lacking the tetO promoter, confirming that Hsp90 depletion leads to reduced glucan levels [fig_ref] Figure 7: Depletion of Hsp90 reduces biofilm matrix glucan [/fig_ref]. The ,40% reduction in matrix glucan upon Hsp90 depletion is likely to have made a major contribution to azole susceptibility, given that reduction of biofilm matrix glucan levels of ,60% in an FKS1/fks1D mutant abrogates biofilm drug resistance [bib_ref] Genetic basis of Candida biofilm resistance due to drug-sequestering matrix glucan, Nett [/bib_ref]. These results provide the . The Hsp90 client proteins Cna1 and Mkc1 exhibit reduced dependence on Hsp90 for stability under biofilm compared to planktonic conditions. (A) Genetic depletion of Hsp90 does not reduce calcineurin levels in biofilm conditions. The tetO-HSP90/hsp90D strain with one allele of CNA1 C-terminally 6xHis-FLAG tagged was grown in planktonic or biofilm conditions with or without doxycycline (DOX, 20 mg/mL) for 48 hours. Total protein was resolved by SDS-PAGE and blots were hybridized with a-Hsp90, a-FLAG to monitor calcineurin levels, and a-tubulin as a loading control (left panel). Cna1 levels from two independent Western blots were quantified using ImageJ software (http://rsb.info.nih.gov/ij/index. html). The density of bands obtained for Cna1 was normalized relative to the density of bands for the corresponding tubulin loading control. Levels were subsequently normalized to the untreated control for the planktonic or biofilm state (right panel). (B) Depletion of Hsp90 does not deplete Mkc1 in biofilm conditions. The tetO-HSP90/hsp90D strain with one allele of MKC1 C-terminally 6xHis-FLAG tagged was grown in planktonic or biofilm conditions with or without DOX for 48 hours. Total protein was resolved by SDS-PAGE and blots were hybridized with a-Hsp90, a-His 6 to monitor Mkc1 levels, a-phospho-p44/42 to monitor dually phosphorylated Mkc1, and a-tubulin as a loading control (left panel). Mkc1 levels from two independent Western blots were quantified using ImageJ software. The density of bands for Mkc1 was normalized relative to the density of bands for the tubulin loading control. Levels were subsequently normalized to the untreated control for the planktonic or biofilm state (right panel). doi:10.1371/journal.ppat.1002257.g006 first link of Hsp90 to glucan production in C. albicans and mechanistic insight as to how Hsp90 regulates biofilm drug resistance.
## Hsp90 is required for c. albicans biofilm azole resistance in vivo
Due to the robust synergy observed between Hsp90 inhibition and fluconazole in vitro, we sought to address whether synergy was also observed in vivo in the rat venous catheter model of C. albicans biofilm infection using the tetO-HSP90/hsp90D strain. Addition of fluconazole alone (250 mg/mL) after 24 hours of biofilm growth did not affect the biofilm formed by the tetO-HSP90/hsp90D strain [fig_ref] Figure 8: Compromise of Hsp90 function genetically or pharmacologically enhances the efficacy of fluconazole... [/fig_ref]. Doxycycline was delivered during both the biofilm formation and drug treatment phases, and also had no major effect on the biofilm formed by the tetO-HSP90/hsp90D strain [fig_ref] Figure 2: Genetic depletion of Hsp90 does not block C [/fig_ref]. However, the combination of fluconazole and doxycycline destroyed the biofilm as observed by scanning electron microscopy [fig_ref] Figure 8: Compromise of Hsp90 function genetically or pharmacologically enhances the efficacy of fluconazole... [/fig_ref]. Thus, Hsp90 is required for the resistance of C. albicans biofilms to fluconazole in a mammalian host.
In order to further explore the therapeutic potential of targeting Hsp90 for C. albicans biofilm infections in vivo, we explored the efficacy of combining fluconazole with an Hsp90 inhibitor structurally related to geldanamycin and in clinical development as an anti-cancer agent, 17-(allylamino)-17-demethoxygeldanamycin (17-AAG). Central venous rat catheters were infected with C. albicans and biofilm formation proceeded over a 24-hour period. At this point, fluconazole alone (250 mg/mL), 17-AAG alone (100 mg/mL), or the drug combination was instilled and allowed to dwell in the catheter for an additional 24 hours. Serial dilutions of the catheter fluid were then plated in order to assess viable colony forming units. We found that the combined drug treatment significantly reduced fungal burden compared to the individual drug treatments alone (P,0.001, ANOVA, Bonferroni's Multiple Comparison Test, [fig_ref] Figure 8: Compromise of Hsp90 function genetically or pharmacologically enhances the efficacy of fluconazole... [/fig_ref]. In fact, catheters from the animals undergoing the combination therapy were completely sterile [fig_ref] Figure 8: Compromise of Hsp90 function genetically or pharmacologically enhances the efficacy of fluconazole... [/fig_ref]. These experiments in a mammalian model provide compelling evidence that clinically relevant Hsp90 inhibitors may prove to be extremely valuable in combating C. albicans biofilm infections.
## Hsp90 is required for drug resistance of a. fumigatus biofilms
We previously established that Hsp90 inhibitors increase the efficacy of the echinocandins against A. fumigatus under standard culture conditions [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref] , motivating these studies to determine if Hsp90 inhibitors also affect drug resistance of A. fumigatus biofilms. After 24 hours of growth, A. fumigatus biofilms were subjected to a gradient of concentrations of the echinocandins caspofungin or micafungin, or the azoles voriconazole or fluconazole, in addition to a gradient of concentrations of the Hsp90 inhibitor geldanamycin in 96 well microtiter plates under static conditions. Metabolic activity was assessed using the XTT reduction assay after an additional 24 hours. The biofilms were completely resistant to all the antifungal drugs tested and geldanamycin individually, though the combination of geldanamycin with many of the antifungals was effective in reducing biofilm development. Geldanamycin displayed robust synergy with both caspofungin and micafungin , with an FIC value of 0.375 for both drugs [fig_ref] Table 2: Inhibition of Hsp90 has synergistic activity with echinocandins against wild-type A [/fig_ref]. Geldanamycin also enhanced voriconazole activity , with more potent effects observed when drugs were added to biofilms after only 8 hours of growth . Geldanamycin did not enhance the efficacy of fluconazole under any conditions tested (data not shown). These patterns of drug synergy observed with A. fumigatus biofilms are consistent with those patterns observed with Aspergillus in planktonic conditions [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref].
Next, given Hsp90's role in regulating fungal morphogenesis we explored the impact of drug treatment on morphology of A. fumigatus biofilms. Scanning electron microscopy revealed striking architectural changes of A. fumigatus biofilms upon drug treatment. The control biofilms appeared robust and healthy, however, upon Hsp90 inhibition increased hyphal and matrix production was observed . Treating biofilms with caspofungin alone resulted in minimal damage, however, the addition of both caspofungin and geldanamycin caused numerous burst and broken hyphae throughout the biofilm . Finally, voriconazole treatment resulted in a flat ribbon-like morphology, and the addition of geldanamycin induced further cell damage . The tetO-HSP90/hsp90D strain was inoculated in rat venous catheters for 24 hours with or without 20 mg/mL doxycycline (DOX) followed by intraluminal azole treatment for an additional 24 hours. Following drug exposure, catheters were removed for visualization by scanning electron microscopy. The first column represents treatment with 250 mg/mL fluconazole (FL), followed by treatment with both 20 mg/mL DOX and 250 mg/mL FL. The top row represents 50 X magnification and the bottom row represents 1,000 X magnification. The combination of FL and DOX abrogates biofilms. (B) Biofilms were cultured as in A with 250 mg/mL FL, 100 mg/mL 17-AAG, or the combination of drugs. Serial dilutions of the catheter fluid were plated for viable Taken together, these results indicate that inhibition of Hsp90 induces changes in morphology of A. fumigatus biofilms, in addition to enhancing the efficacy of azoles and echinocandins against these otherwise recalcitrant cellular structures.
# Discussion
Our results establish a novel role for Hsp90 in dispersion and drug resistance of fungal biofilms, with profound therapeutic potential. Resistance of C. albicans biofilms to many antifungal drugs including the azoles, often necessitates surgical removal of the infected catheter or substrate demanding new therapeutic strategies. Here, we demonstrate that compromising the function of C. albicans Hsp90 blocks biofilm dispersal, potentially reducing their ability to serve as reservoirs for persistent infection [fig_ref] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell... [/fig_ref]. Further, we show that compromising Hsp90 function genetically or pharmacologically in C. albicans renders biofilms exquisitely susceptible to azoles, such that fluconazole is transformed from inefficacious to highly effective in destroying biofilms both in vitro [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref] and [fig_ref] Table 1: Inhibition of calcineurin or Hsp90 has synergistic activity with fluconazole against wild-type... [/fig_ref] and in a mammalian model of infection [fig_ref] Figure 8: Compromise of Hsp90 function genetically or pharmacologically enhances the efficacy of fluconazole... [/fig_ref]. Finally, in A. fumigatus we found that compromising Hsp90 function dramatically improves the efficacy of antifungals . Thus, inhibition of Hsp90 enhances the efficacy of antifungals against biofilms formed by the two leading fungal pathogens of humans separated by ,1 billion years of evolution, suggesting that this combinatorial therapeutic strategy could have a broad spectrum of activity against diverse fungal pathogens.
Hsp90 exerts pleiotropic effects on cellular circuitry in eukaryotes by stabilizing diverse regulators of cellular signalling [bib_ref] HSP90 at the hub of protein homeostasis: emerging mechanistic insights, Taipale [/bib_ref] [bib_ref] Structure and mechanism of the Hsp90 molecular chaperone machinery, Pearl [/bib_ref] [bib_ref] Hsp90: a specialized but essential proteinfolding tool, Young [/bib_ref]. Hsp90 regulates the temperature-dependent morphogenetic transition from yeast to filamentous growth in C. albicans, such that compromise of Hsp90 function by elevated temperature relieves Hsp90-mediated repression of Ras1-PKA signalling and induces filamentous growth [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref]. While compromise of Hsp90 function could have impaired biofilm development by enhancing filamentous growth, we found negligible impact on biofilm development in vivo [fig_ref] Figure 2: Genetic depletion of Hsp90 does not block C [/fig_ref] ; in vitro, compromise of Hsp90 function did reduce biofilm maturation under static conditions with more severe effects under shaking conditions [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Biofilms formed in the presence of Hsp90 inhibitor had a greater proportion of their total thickness occupied by filaments compared to yeast , suggesting that Hsp90's role in repressing the yeast to filament transition in planktonic cells [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref] is conserved in the biofilm state. Consequently, we investigated the impact of compromising Hsp90 function on dispersion, a stage of the biofilm life cycle intimately coupled to morphogenetic transitions, with the majority of dispersed cells being in the yeast form [bib_ref] The transcriptional regulator Nrg1p controls Candida albicans biofilm formation and dispersion, Uppuluri [/bib_ref] [bib_ref] Dispersion as an important step in the Candida albicans biofilm developmental cycle, Uppuluri [/bib_ref]. We found that compromising Hsp90 function dramatically reduces the dispersed cell population [fig_ref] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell... [/fig_ref] , consistent with previous findings with hyperfilamentous C. albicans mutants [bib_ref] The transcriptional regulator Nrg1p controls Candida albicans biofilm formation and dispersion, Uppuluri [/bib_ref] [bib_ref] Dispersion as an important step in the Candida albicans biofilm developmental cycle, Uppuluri [/bib_ref]. Strikingly, the majority of cells that disperse from biofilms with reduced levels of Hsp90 are inviable [fig_ref] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell... [/fig_ref] , which likely reflects an enhanced dependence of this cell population on Hsp90. Given that the dispersed cell population is thought to be responsible for device-associated candidemia and the establishment of disseminated infection, inhibition of C. albicans Hsp90 function in individuals suffering from biofilm infections may assist in the prevention of the invasive forms of disease. In the broader sense, it is striking that depletion of Hsp90 blocks the production of yeast in C. albicans in planktonic conditions [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref] as well as throughout the biofilm lifecycle, creating a constitutively filamentous program characteristic of the strictly filamentous lifestyle of the vast majority of fungi.
Hsp90 potentiates the emergence and maintenance of C. albicans drug resistance through multiple client proteins. A key mediator of Hsp90-dependent drug resistance is the protein phosphatase calcineurin [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref] [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref] [bib_ref] Genetic architecture of Hsp90-dependent drug resistance, Cowen [/bib_ref]. In planktonic cells, Hsp90 stabilizes the catalytic subunit of calcineurin, Cna1, thereby enabling calcineurin-dependent cellular signalling required for survival of druginduced cellular stress [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref]. Hsp90 also regulates drug resistance by stabilizing the MAPK Mkc1, thereby enabling additional stress responses important for resistance [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. In planktonic conditions, inhibition of calcineurin phenocopies inhibition of Hsp90 reducing drug resistance of diverse mutants, though deletion of MKC1 has a less severe effect on resistance under specific conditions [bib_ref] Hsp90 potentiates the rapid evolution of new traits: drug resistance in diverse..., Cowen [/bib_ref] [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref] [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. In biofilms, homozygous deletion of either CNA1 or MKC1 causes an intermediate increase in sensitivity to azoles compared to reduction of HSP90 levels [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. Genetic depletion of Hsp90 reduces the fluconazole MIC 50 from .512 mg/mL to 8 mg/mL, whereas deletion of CNA1 reduces resistance to 32 mg/mL and deletion of MKC1 reduces resistance only to 128 mg/mL [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. Thus, both calcineurin and Mkc1 have reduced impact on azole resistance of biofilms compared to Hsp90, suggesting differences in the Hsp90-dependent cellular circuitry between the biofilm and planktonic cellular states.
Hsp90 regulates circuitry required for fungal drug resistance largely by stabilizing key regulators of cellular signalling. In planktonic conditions, reduction of Hsp90 levels leads to depletion of both Cna1 and Mkc1 [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref] [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. In stark contrast, Cna1 and Mkc1 remain stable in biofilms, despite reduction of Hsp90 levels . In both planktonic and biofilm conditions, Hsp90 levels were reduced by doxycycline-mediated transcriptional repression in the tetO-HSP90-hsp90D strain and levels of Hsp90 were reduced sufficiently to abrogate drug resistance in both conditions. The reduced dependence of Cna1 and Mkc1 on Hsp90 in biofilms suggests that these proteins have altered stability in this cellular state. These Hsp90 client proteins may assume an alternate conformation in biofilms that is inherently more stable, or they may interact with other proteins or chaperones that confer increased stability and reduced dependence upon Hsp90. Consistent with the possibility of altered chaperone balance in biofilm cells, the Hsp70 family member SSB1 is overexpressed sixfold in biofilms compared to their planktonic counterparts [bib_ref] Candida albicans biofilms: a developmental state associated with specific and stable gene..., Garcia-Sanchez [/bib_ref]. While it is possible that Hsp90 may still regulate Cna1 and Mkc1 function through a mechanism distinct from protein stability, we note that Mkc1 is still activated upon Hsp90 depletion in biofilms . Given Hsp90's high degree of connectivity in diverse signalling cascades, it could also affect biofilm drug resistance in a multitude of other ways, such as by regulating remodeling of the cell wall and cell membrane [bib_ref] Mechanism of fluconazole resistance in Candida albicans biofilms: phase-specific role of efflux..., Mukherjee [/bib_ref] [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] , signalling cascades important for matrix production [bib_ref] Genetic basis of Candida biofilm resistance due to drug-sequestering matrix glucan, Nett [/bib_ref] [bib_ref] Biofilm matrix regulation by Candida albicans Zap1, Nobile [/bib_ref] , or the function of contactdependent signalling molecules that initiate responses to surfaces [bib_ref] A contact-activated kinase signals Candida albicans invasive growth and biofilm development, Kumamoto [/bib_ref]. Future studies will determine on a more global scale the impact of cellular state on Hsp90 client protein stability, and the complex circuitry by which Hsp90 regulates biofilm drug resistance.
Our results suggest that Hsp90 is a novel regulator of matrix glucan levels. For C. albicans the reduction in matrix glucan levels upon Hsp90 depletion provides a mechanism by which Hsp90 might govern biofilm azole resistance. C. albicans biofilms possess fungal colony counts. Results are expressed as the mean colony forming unit (CFU) per catheter. The combination of FL and 17-AAG reduces fungal burden in the catheter compared to individual drug treatments (Asterisk indicates P,0.001, ANOVA, Bonferroni's Multiple Comparison Test). doi:10.1371/journal.ppat.1002257.g008 . Pharmacological inhibition of Hsp90 enhances the efficacy of echinocandins and azoles against A. fumigatus biofilms and affects biofilm morphology. (A) A. fumigatus was grown in 96-well dishes in RPMI at 37uC. After 24 hours cells were washed with PBS to remove non-adherent cells and fresh medium was added with varying concentrations of the echinocandin caspofungin (CF), the azole voriconazole (VL), and the Hsp90 inhibitor geldanamycin (GdA) in a checkerboard format. Drug treatment was left on for 24 hours. Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. The FIC index was calculated as indicated in [fig_ref] Table 2: Inhibition of Hsp90 has synergistic activity with echinocandins against wild-type A [/fig_ref]. Bright green represents growth above the MIC 50 , dull green represents growth at the MIC 50 , and black represents growth below the MIC 50 . (B) A. fumigatus cells were left untreated, or treated with 32 mg/mL CF or 256 mg/mL VL in the absence and presence of 50 mg/mL GdA for 24 hours. Following drug exposure, biofilms were fixed and imaged by scanning electron microscopy. Biofilms treated with antifungal show increased cellular damage in the presence of GdA. The white arrows indicate burst and broken hyphae in the biofilms treated with CF and GdA. doi:10.1371/journal.ppat.1002257.g009 elevated cell wall b-1,3 glucan content compared to their planktonic counterparts [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] , and matrix glucan sequesters fluconazole, preventing it from reaching its intracellular target [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] [bib_ref] Genetic basis of Candida biofilm resistance due to drug-sequestering matrix glucan, Nett [/bib_ref]. The ,40% reduction in matrix glucan we observed upon Hsp90 depletion [fig_ref] Figure 7: Depletion of Hsp90 reduces biofilm matrix glucan [/fig_ref] likely contributes to reduced azole resistance, given that a reduction of matrix glucan levels of ,60% in an FKS1/fks1D mutant abrogates biofilm drug resistance [bib_ref] Genetic basis of Candida biofilm resistance due to drug-sequestering matrix glucan, Nett [/bib_ref]. Hsp90 could regulate glucan levels by directly or indirectly affecting b-1,3 glucan synthase, Fks1, a protein important for the production of matrix glucan and for antifungal resistance [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] [bib_ref] Genetic basis of Candida biofilm resistance due to drug-sequestering matrix glucan, Nett [/bib_ref]. Alternatively, Hsp90 could regulate matrix production by directly or indirectly affecting Zap1, or its downstream targets Gca1 and Gca2, which play an important role in matrix production, likely through the hydrolytic release of b-glucan fragments from the environment [bib_ref] Biofilm matrix regulation by Candida albicans Zap1, Nobile [/bib_ref]. We note that in A. fumigatus, inhibition of Hsp90 appears to increase matrix production , though glucan levels remain unknown. Future studies will dissect the molecular mechanisms by which Hsp90 regulates biofilm matrix production and if there is divergent circuitry between these fungal pathogens.
This work establishes that targeting Hsp90 may provide a powerful therapeutic strategy for biofilm infections caused by the leading fungal pathogens of humans. Compromising Hsp90 function genetically or pharmacologically reduces azole resistance of C. albicans biofilms both in vitro and in the rat venous catheter model of infection [fig_ref] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C [/fig_ref]. Importantly, inhibition of Hsp90 with 17-AAG, an Hsp90 inhibitor that has advanced in clinical trials for the treatment of cancer [bib_ref] AAG for HSP90 inhibition in cancer-from bench to bedside, Usmani [/bib_ref] [bib_ref] Update on Hsp90 inhibitors in clinical trial, Kim [/bib_ref] and is synergistic with antifungals in planktonic conditions [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref] , transforms fluconazole from ineffective to highly efficacious in a mammalian model of biofilm infection [fig_ref] Figure 8: Compromise of Hsp90 function genetically or pharmacologically enhances the efficacy of fluconazole... [/fig_ref]. There may in fact be a multitude of benefits of inhibiting Hsp90 in the context of C. albicans biofilm infections given a recent report that treatment of in vitro C. albicans biofilms with voriconazole induces resistance to micafungin in an Hsp90-dependent manner [bib_ref] Anti-Candida-biofilm activity of micafungin is attenuated by voriconazole but restored by pharmacological..., Kaneko [/bib_ref]. The therapeutic potential of Hsp90 inhibitors against fungal biofilms extends beyond C. albicans to the most lethal mould, A. fumigatus. Pharmacological inhibition of Hsp90 enhances the efficacy of both azoles and echinocandins against A. fumigatus biofilms . The synergy between Hsp90 inhibitors and echinocandins is more pronounced than that with azoles, consistent with findings in the planktonic cellular state [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref]. Thus, targeting Hsp90 may provide a much-needed strategy to enhance the efficacy of antifungal drugs against biofilms formed by diverse fungal pathogens.
Our results provide a new facet to the broader therapeutic paradigm of Hsp90 inhibitors in the treatment of infectious disease caused by fungi and other pathogenic eukaryotes. In addition to the profound effects on biofilm drug resistance and dispersal, compromising Hsp90 function enhances the efficacy of azoles and echinocandins against disseminated disease caused by the leading fungal pathogens of humans in invertebrate and mammalian models of infection [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref] [bib_ref] Hsp90 governs echinocandin resistance in the pathogenic yeast Candida albicans via calcineurin, Singh [/bib_ref]. Beyond enhancing antifungal activity, Hsp90 also provides an attractive antifungal target on its own given that depletion of fungal Hsp90 results in complete clearance of a kidney fungal burden in a mouse model of disseminated candidiasis [bib_ref] Hsp90 orchestrates temperature-dependent Candida albicans morphogenesis via Ras1-PKA signaling, Shapiro [/bib_ref]. Hsp90 inhibitors also exhibit potent activity against malaria and Trypanosoma infections, thus extending their spectrum of activity to the protozoan parasites Plasmodium falciparum and Trypanosoma evansi [bib_ref] Heat shock protein 90 as a drug target against protozoan infections: biochemical..., Pallavi [/bib_ref] [bib_ref] A repurposing strategy identifies novel synergistic inhibitors of Plasmodium falciparum heat shock..., Shahinas [/bib_ref]. The development of Hsp90 as a therapeutic target for infectious disease may benefit from the plethora of structurally diverse Hsp90 inhibitors that have been developed, many of which are in advanced phase clinical trials for cancer treatment, with substantial promise due to the depletion of a myriad of oncoproteins upon inhibition of Hsp90 [bib_ref] Targeting the dynamic HSP90 complex in cancer, Trepel [/bib_ref]. Given the importance of Hsp90 in chaperoning key regulators of cellular signalling in all eukaryotes, the challenge of advancing Hsp90 as a target for infectious disease lies in avoiding host toxicity issues. Indeed, although well tolerated in the mammalian host individually or in combination therapies [bib_ref] Targeting the dynamic HSP90 complex in cancer, Trepel [/bib_ref] , Hsp90 inhibitors have toxicity in the context of an acute disseminated fungal infection [bib_ref] Harnessing Hsp90 function as a powerful, broadly effective therapeutic strategy for fungal..., Cowen [/bib_ref]. This toxicity may be due to Hsp90's role in regulating host immune and stress responses during infection. Toxicity was not observed in our studies of biofilm infections in the mammalian model, perhaps owing to both the localized infection and drug delivery, suggesting that this therapeutic strategy could rapidly translate from the laboratory bench to the patients' bedside. In the broader context, the challenge for further development of Hsp90 as a therapeutic target for infectious disease lies in developing pathogen-selective inhibitors or drugs that target pathogen-specific components of the Hsp90 circuitry governing drug resistance and virulence.
# Materials and methods
# Ethics statement
## Strains and culture conditions
Archives of C. albicans strains were maintained at 280uC in 25% glycerol. Strains were routinely maintained and grown in YPD liquid medium (1% yeast extract, 2% bactopeptone, 2% glucose) at 30uC. Strains used in this study are listed in [fig_ref] Table 1: Inhibition of calcineurin or Hsp90 has synergistic activity with fluconazole against wild-type... [/fig_ref]. Strain construction is described in the Supplemental Material.
## Biofilm growth conditions
Multiple in vitro assays were used to assess C. albicans biofilm growth and antifungal drug susceptibility. In the first model, biofilms were developed in 96-well polystyrene plates, as previously described [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] [bib_ref] Standardized method for in vitro antifungal susceptibility testing of Candida albicans biofilms, Ramage [/bib_ref]. Briefly, strains were grown overnight in YPD at 37uC. Subsequently, cultures were resuspended in RPMI medium buffered with HEPES or MOPS, in the presence or absence of doxycycline (631311, BD Biosciences) to a final concentration of 10 6 cells/mL. An aliquot of 100 ml was added to each well of a 96-well flat-bottom plate, followed by incubation at 37uC. After 24 hours, the wells were gently washed twice with phosphate-buffered saline (PBS) to remove non-adherent cells, and fresh medium was added with or without a gradient of geldanamycin (ant-gl-5, Cedarlane). After 24 hours, non-adherent cells were washed away with PBS and biofilm cell metabolic activity was measured using the XTT reduction assay as previously described [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] [bib_ref] Standardized method for in vitro antifungal susceptibility testing of Candida albicans biofilms, Ramage [/bib_ref]. Briefly, 90 ml of XTT (X4251, Sigma) at 1 mg/mL and 10 ml phenazine methosulfate (P9625, Sigma) at 320 mg/mL were added to each well, followed by incubation at 37uC for 2 hours. Absorbance of the supernatant transferred to a fresh plate was measured at 490 nm using an automated plate reader, and experiments were carried out in a minimum of 5 replicates for each strain. In the second model, biofilms were developed on silicon elastomer (SE) surfaces as has been described previously [bib_ref] Candida albicans biofilmdefective mutants, Richard [/bib_ref]. C. albicans wild-type cells were grown overnight in YPD medium at 30uC and diluted to an optical density at 600 nm of 0.5 in RPMI medium. The suspension was added to a sterile 12-well plate containing bovine serum (B-9433, Sigma)-treated SE (Cardiovascular Instrument silicon sheets; PR72034-06N) and incubated at 37uC for 90 min at 150 rpm agitation for initial adhesion. The SE were washed with PBS, transferred to fresh plates containing either fresh RPMI medium in the absence of drug, or RPMI with 10 mg/mL geldanamycin or 20 mg/mL doxycycline. Plates were incubated at 37uC for 48 hours at 150 rpm agitation to allow biofilm formation, followed by visualization by microscopy or by monitoring biofilm growth by XTT reduction or dry weight, as previously described [bib_ref] Synergistic effect of calcineurin inhibitors and fluconazole against Candida albicans biofilms, Uppuluri [/bib_ref] [bib_ref] Standardized method for in vitro antifungal susceptibility testing of Candida albicans biofilms, Ramage [/bib_ref].
## C. albicans biofilm dispersion
For obtaining cells dispersed from biofilms, C. albicans biofilms were cultured in a simple flow biofilm model, as described previously [bib_ref] Design of a simple model of Candida albicans biofilms formed under conditions..., Uppuluri [/bib_ref] [bib_ref] An easy and economical in vitro method for the formation of Candida..., Uppuluri [/bib_ref]. Briefly, this model involves a controlled flow of fresh medium via Tygon tubing (Cole-Parmer, Vernon Hills, IL) into a 15 mL polypropylene conical tube (BD, Franklin, NJ) holding a SE strip. Medium flow is controlled at 1 mL/minute, by connecting the tubing to a peristaltic pump (Masterflex L/S Easy-Load II, Cole-Parmer). The whole apparatus is placed inside an incubator to facilitate biofilm development at 37uC. SE strips (169 cm , Cardiovascular instrument Corp, Wakefield, MA), were sterilized by autoclaving and pre-treated for 24 hours with bovine serum. C. albicans was grown overnight at 30uC, washed, and diluted to an optical density at 600 nm of 0.5 in Yeast Nitrogen base (YNB) medium (BD Biosciences, San Jose, CA) with 50 mM glucose. The SE strips were incubated with the diluted C. albicans suspension at 37uC for 90 min at 100 rpm agitation for the initial adhesion of cells. Next, the strip was inserted into the conical tube and the peristaltic pump was turned on. At various time points during biofilm development, cells released from the biofilm in the flow-through were collected from the bottom of the conical tube. The dispersed cells were enumerated by a hemocytometer to obtain cell counts and there were no differences observed in the degree of clumping or morphological state of the dispersed cells, which were in the yeast form. Viability of the dispersed cells was assessed by plating and by colony counts on YPD agar.
## C. albicans in vitro biofilm drug susceptibility
Drug susceptibility assays were performed on biofilms formed in wells of 96-well plates. Fresh medium (RPMI/HEPES) and drugs were added to wells containing biofilms grown for 24 hours. Dilutions of fluconazole (Sequoia Research Products) were from 1000 mg/ml down to 0 with the following concentration steps in mg/ml: 1000, 500, 250, 125, 62. .90625, 1.953125, 0.9765625. FK506 (AG Scientific) gradients were from 75 mg/mL down to 0 with the following concentration steps in mg/ ml: 75, .6875, 2.3475, 1.171875. Geldanamycin gradients were from 100 mg/mL to 0 with the following concentration steps in mg/ml: 100, 50, 25, 12.5, 6.25, 3.125, 1.5625. Drug combinations were examined alone or in combination in a checkerboard format. After incubation at 37uC for 24 hours, biofilms were washed twice with PBS and metabolic activity was measured using the XTT assay, as described above. The drug concentration associated with 50% reduction in optical density compared to the drug-free control wells (MIC [bib_ref] Hsp90: a specialized but essential proteinfolding tool, Young [/bib_ref]
## C. albicans in vivo biofilm model
In order to evaluate biofilm formation in vivo, a rat central venous catheter infection model was employed [bib_ref] Development and characterization of an in vivo central venous catheter Candida albicans..., Andes [/bib_ref]. Specificpathogen-free Sprague-Dawley rats weighing ,400 g were used (Harlan Sprague-Dawley, Indianapolis, IN). A heparinized (100 U/mL) polyethylene catheter was surgically inserted into the jugular vein and advanced 2 cm to a site above the right atrium. After the catheter was secured to the vein, the proximal end was tunneled subcutaneously to the midscapular space and externalized through the skin. The catheters were implanted 24 hours prior to inoculation with C. albicans to allow a conditioning period for deposition of host protein on the catheter surface. Infection was performed by intraluminal instillation of 500 ml of C. albicans (10 6 cells/mL). After 6 hours, the catheters were flushed and maintained with heparinized 0.85% NaCl for 24 hours to allow for biofilm formation. While one end of the catheter is open to the venous blood, most of the fluid contents remain within the catheter unless pushed into the bloodstream with additional fluid from the external end. For drug treatment studies, fluconazole (250 mg/mL), 17-AAG (A-6880, LC Laboratories, 100 mg/mL), or saline was instilled and allowed to dwell in the catheter for an additional 24 hours [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref]. For doxycycline studies, doxycycline (20 mg/mL) was delivered during both the biofilm formation and the drug treatment phases. At the end of the observation period, the animals were sacrificed and the catheters were removed. In order to quantify fungal biofilm formation in the catheter, the contents were drained to remove blood and nonadherent organisms. The distal 2 cm of catheter was cut from the entire catheter length and the segment was placed in 1 mL of 0.85% NaCl. Following sonication for 10 minutes (FS 14 water bath sonicator and 40-kHz transducer [Fisher Scientific]) and vigorous vortexing for 30 seconds, serial dilutions of the catheter fluid were plated on Sabouraud Dextrose Agar (SDA) for viable fungal colony counts. Results are expressed as the mean colony forming unit (CFU) per milliliter.
## A. fumigatus in vitro biofilm drug susceptibility
Aspergillus fumigatus Af293 was maintained on SDA slopes at 4uC. For conidial preparation Af293 was propagated on SAB agar for 72 hours and conidia harvested in PBS containing 0.025% (v/v) Tween 20 and quantified as previously described [bib_ref] Development of a simple model for studying the effects of antifungal agents..., Mowat [/bib_ref]. Commercially available voriconazole (Pfizer Pharmaceuticals, NY, USA), micafungin (Astellas Pharma Inc, Ibaraki, Japan) and caspofungin (Merck Sharp Dohme Ltd, NJ, USA) were used throughout this study. Each antifungal drug was prepared at stock concentrations of 10 mg/mL in sterile water and used within 24 hours of reconstitution.
Af293 conidial inoculum (1610 5 conidia/mL) was dispensed into flat bottomed 96-well microtitre plates and incubated for 8 or 24 hours at 37uC as previously described [bib_ref] Development of a simple model for studying the effects of antifungal agents..., Mowat [/bib_ref]. Biofilms were gently washed twice with PBS and each antifungal agent and geldanamycin were diluted to working concentrations in RPMI, which were tested either alone or in combination in a checkerboard format. Antifungal agent dilutions were from 512 mg/ml down to 0 with the following concentration steps in mg/ml: 512, 256, 128, 64, 32, 16, 8, 4, 2, 1, 0.5. Geldanamycin dilutions were from 100 mg/ml down to 0 with the following concentration steps in mg/ml: 100, 25, 12.5, 6.25, 3.125, 1.5625. The biofilms were then treated and processed as described for C. albicans.
## Confocal microscopy
Biofilms were stained with 25 mg/mL concanavalin A-Alexa Fluor 594 conjugate (C-11253; Molecular Probes, Eugene, OR) for 1 hour in the dark at 37uC. Confocal scanning laser microscopy (CSLM) was performed with a ZeissLSM 510 upright confocal microscope using a Zeiss Achroplan 40X, 0.8-W objective. Stained biofilms were observed using a HeNe1 laser with an excitation wavelength of 543 nm. The Zeiss LSM Image Browser v4.2 software was used to assemble images into side and depth views. Artificially coloured depth view images represent cell depth using a colour gradient, where cells closest to the SE are represented in blue and the cells farthest away are represented in red.
## Scanning electron microscopy
Biofilms formed in vitro were placed overnight in a fixative (4% formaldehyde v/v, 1% glutaraldehyde v/v in PBS), rinsed in 0.1 M phosphate buffer and air dried in desiccators. Notably, harsh dehydration steps were not performed to minimize the damage to the original biofilm structure. The samples were coated with gold/palladium (40%/60%) and observed under a scanning electron microscope (Leo 435 VP) in high vacuum mode at 15 kV. The images were assembled using Photoshop software (Adobe, Mountain View, CA.).
Catheter segments were processed for scanning electron microscopy as previously described [bib_ref] Development and characterization of an in vivo central venous catheter Candida albicans..., Andes [/bib_ref]. Following overnight fixation (4% formaldehyde, 1% glutaraldehyde in PBS), catheter segments were washed with PBS and treated with osmium tetroxide (1% in PBS) for 30 minutes. Drying was accomplished using a series of alcohol washes followed by critical point drying. Catheter segments were mounted and gold coated. Images were obtained with a scanning electron microscope (JEOL JSM-6100) in the high-vacuum mode at 10 kV. The images were assembled using Adobe Photoshop 7.0.1.
# Immune blot analysis
For the protein stability assay, planktonic cultures were grown in RPMI buffered with MOPS and treated as described previously [bib_ref] PKC signaling regulates drug resistance of the fungal pathogen Candida albicans via..., Lafayette [/bib_ref]. For biofilm cultures, C. albicans was grown overnight in YPD medium at 30uC and diluted to an optical density at 600 nm of 0.5 in RPMI medium. The suspension was added to a bovine serum (16190; Gibco)-treated sterile 6-well plate and incubated at 37uC for 90 minutes for initial adhesion. The plates were washed with PBS, and fresh RPMI medium was added with or without 20 mg/ mL doxycycline. Plates were incubated at 37uC for 48 hours.
Cells were harvested by centrifugation and were washed with sterile water. Cell pellets were resuspended in lysis buffer containing 50 mM HEPES pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 1 mM PMSF, and protease inhibitor cocktail (complete, EDTA-free tablet, Roche Diagnostics). Cells suspended in lysis buffer were mechanically disrupted by adding acid-washed glass beads and bead beating for 3 minutes. Protein concentrations were determined by Bradford analysis. Protein samples were mixed with one-sixth volume of 6X sample buffer containing 0.35 M Tris-HCl, 10% (w/v) SDS, 36% glycerol, 5% b-mercaptoethanol, and 0.012% bromophenol blue for SDS-PAGE. Samples were boiled for 5 minutes and then separated by SDS-PAGE using an 8% acrylamide gel. Proteins were electrotransferred to PVDF membranes (Bio-Rad Laboratories, Inc.) and blocked with 5% skimmed milk in phosphate buffered saline (PBS) with 0.1% tween. Blots were hybridized with antibodies against CaHsp90 (1:10000), generously provided by Brian Larsen [bib_ref] Isolation and partial characterization of Hsp90 from Candida albicans, Burt [/bib_ref] , FLAG (1:10000, Sigma Aldrich Co.), His 6 (1:10, P5A11, generously provided by Elizabeth Wayner), phospho-p44/42 MAPK (Thr202/Tyr204) (1:2000, Cell Signaling), or against alpha-tubulin (1:1000; AbD Serotec, MCA78G).
## Biofilm matrix collection and matrix b-1,3 glucan measurements
Matrix b-1,3 glucan content was measured using a limulus lysate based assay, as previously described [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] [bib_ref] Beta-Dglucan as a diagnostic adjunct for invasive fungal infections: validation, cutoff development,..., Odabasi [/bib_ref]. Matrix was collected from C. albicans biofilms growing in the wells of 6-well polystyrene plates with or without 20 mg/mL doxycycline for 48 hours. The method for culturing biofilms was as described above for the immune blot analysis with the exception that all reagents were glucan-free. Biofilms were dislodged using a sterile spatula, washed with PBS, sonicated for 10 minutes, and centrifuged 3 times at 4500 x g for 20 minutes to separate cells from soluble matrix material [bib_ref] Putative role of beta-1,3 glucans in Candida albicans biofilm resistance, Nett [/bib_ref] [bib_ref] Extracellular polymer of Candida albicans: isolation, analysis and role in adhesion, Mccourtie [/bib_ref]. Samples were stored at -20uC and glucan concentrations were determined using the Glucatell [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref] The impact of Hsp90 depletion on C. albicans biofilm formation and maturation in multiple models. (A) A wild-type strain of C. albicans and the tetO-HSP90/hsp90D strain were grown on silicon elastomer squares in RPMI at 37uC for 24 hours with or without 20 mg/mL doxycycline (DOX). Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Treatment of wild-type biofilms with DOX did not alter biofilm growth, while Hsp90 depletion caused a moderate but significant reduction in biofilm growth (P,0.01, ANOVA, Bonferroni's Multiple Comparison Test). (B) Biofilms were grown as in part A, but growth was measured by dry weight. Treatment of wild-type biofilms with DOX did not alter biofilm growth, while Hsp90 depletion caused a moderate but significant reduction in biofilm growth (P,0.05). (C) The tetO-HSP90/hsp90D strain was grown with or without 20 mg/mL DOX in the overnight culture as well as during biofilm formation on plastic under static conditions. Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. Depletion of Hsp90 does not block biofilm formation. (D) Biofilms were cultured on plastic under shaking conditions with or without 20 mg/mL DOX. Treatment of wild-type biofilms with DOX did not alter biofilm growth. The tetO-HSP90/hsp90D strain showed impaired biofilm development (P,0.001), which was exacerbated in the presence of 20 mg/mL DOX. This is consistent with impaired HSP90 induction in response to many conditions when driven by the non-native tetO promoter and the further transcriptional repression of HSP90 with DOX [bib_ref] The epidemiology and attributable outcomes of candidemia in adults and children hospitalized..., Zaoutis [/bib_ref]. (TIF) [fig_ref] Figure 2: Genetic depletion of Hsp90 does not block C [/fig_ref] Treating C. albicans with doxycycline does not impair biofilm dispersal. (A) A wild-type strain of C. albicans lacking the tetO promoter was cultured in the presence or absence of 20 mg/mL doxycycline (DOX). The number of dispersed cells released from biofilms was monitored over a 24 hour period. (B) The viability of dispersed cells from a wildtype C. albicans strain was determined by plating on YPD agar. DOX has no effect on biofilm dispersal or viability in a wild-type strain. (TIF) Pharmacological inhibition of Hsp90 enhances the efficacy of echinocandins and azoles against A. fumigatus biofilms. A. fumigatus was grown in 96-well microtiter plates in RPMI at 37uC. (A) After 24 hours cells were washed with PBS to remove non-adherent cells and fresh media was added with varying concentrations of the echinocandin micafungin (MF) in combination with the Hsp90 inhibitor geldanamycin (GdA) in a checkerboard format, and incubated with the biofilm for 24 hours. Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref]. The FIC index was calculated as indicated in [fig_ref] Table 2: Inhibition of Hsp90 has synergistic activity with echinocandins against wild-type A [/fig_ref]. Bright green represents growth above the MIC 50 , dull green represents growth at the MIC 50 , and black represents growth below the MIC 50 . (B) After 8 hours cells were washed with PBS to remove non-adherent cells and fresh media was added with varying concentrations of the azole voriconazole (VL) in combination with GdA in a checkerboard format, and incubated with the biofilm for 24 hours. Metabolic activity was measured as in [fig_ref] Figure 1: Compromise of Hsp90 function does not block C [/fig_ref] and data analyzed as in . (TIF) Table S1 C. albicans strains used in this study.
## Supporting information
## (doc)
Text S1 Supporting materials and methods.
## (doc)
Author Contributions
[fig] Figure 1: Compromise of Hsp90 function does not block C. albicans biofilm development in vitro. (A) [/fig]
[fig] Figure 2: Genetic depletion of Hsp90 does not block C. albicans biofilm formation in vivo. The tetO-HSP90/hsp90D strain was inoculated in rat venous catheters in the presence or absence of 20 mg/mL doxycycline (DOX). Biofilms were examined by scanning electron microscopy imaging at 24 hours. The top row represents 50 X magnification while the bottom row represents 1,000 X magnification. Biofilm thickness and structure were similar in the presence or absence of doxycycline. doi:10.1371/journal.ppat.1002257.g002 [/fig]
[fig] Figure 4: Depletion of Hsp90 reduces biofilm dispersion and viability of the dispersed cell population. C. albicans biofilms from the tetO-HSP90/hsp90D strain were cultured in the presence or absence of 20 mg/mL doxycycline (DOX). (A) The number of dispersed cells released from biofilms was monitored over a 24 hour period. (B) The viability of dispersed cells was determined by plating on YPD agar. doi:10.1371/journal.ppat.1002257.g004 [/fig]
[fig] Figure 5: Inhibition of Hsp90 function dramatically enhances the efficacy of fluconazole against C. albicans biofilms in vitro. (A) [/fig]
[fig] Figure 7: Depletion of Hsp90 reduces biofilm matrix glucan. Strains of C. albicans were cultured in 6-well polystyrene dishes for 48 hours with or without 20 mg/mL doxycycline (DOX). Matrix samples were collected and matrix b-1,3 glucan levels were meausured using a limulus lysate based assay. Asterisk indicates P,0.01 (ANOVA, Bonferroni's Multiple Comparison Test) compared to all other conditions. doi:10.1371/journal.ppat.1002257.g007 [/fig]
[fig] Figure 8: Compromise of Hsp90 function genetically or pharmacologically enhances the efficacy of fluconazole in vivo. (A) [/fig]
[table] Table 1: Inhibition of calcineurin or Hsp90 has synergistic activity with fluconazole against wild-type C. albicans biofilms.FIC index (MIC 50 of drug A in combination)/(MIC 50 of drug A alone) + (MIC 50 of drug B in combination)/(MIC 50 of drug B alone). A FIC of ,0.5 is indicative of synergism. doi:10.1371/journal.ppat.1002257.t001 [/table]
[table] Table 2: Inhibition of Hsp90 has synergistic activity with echinocandins against wild-type A. fumigatus biofilms. [/table]
|
Fast and accurate detection of spread source in large complex networks
The algorithm proposed by Pinto, Thiran and Vetterli is a deterministic Gaussian method for a single source identification in a network of size N . It requires a full knowledge of the network structure G and of times t k , 1 k K, at which the information arrived at observer k (observed delays). The observers are nodes that report report from which neighbor and at what time it received the information. One observer (lets say o K ) serves as a reference point. The method also assumes the information spreads through the network along the shortest paths and the propagation times θ i for each edge are i.i.d Gaussian random variables, for which the mean µ and the variance σ 2 are known. The assumption of diffusion along the shortest paths is used when the algorithm is applied to an arbitrary, non-tree graph. In this case the actual propagation tree (which is not known a priori) is approximated by a breadth-first search (BFS) tree composed of the shortest paths between the node which is supposed to be the source and all observers. The algorithm calculates a score for each node s using a density function of (K − 1)-dimensional normal distribution, where K is the number of observers. The node with the highest score is pointed as the rumor source. If |P(u, v)| denotes the length of the path P(u, v)connecting nodes u and v, d is the vector of observed delays, µ s is the vector of deterministic delays and Λ s is the delay covariance matrix, the score of node s is given by:where[µ s ] k = µ(|P(s, o k )| − |P(s, o K )|),1
## S.1 pinto-thiran-vetterli algorithm
The algorithm proposed by Pinto, Thiran and Vetterli is a deterministic Gaussian method for a single source identification in a network of size N . It requires a full knowledge of the network structure G and of times t k , 1 k K, at which the information arrived at observer k (observed delays). The observers are nodes that report report from which neighbor and at what time it received the information. One observer (lets say o K ) serves as a reference point. The method also assumes the information spreads through the network along the shortest paths and the propagation times θ i for each edge are i.i.d Gaussian random variables, for which the mean µ and the variance σ 2 are known. The assumption of diffusion along the shortest paths is used when the algorithm is applied to an arbitrary, non-tree graph. In this case the actual propagation tree (which is not known a priori) is approximated by a breadth-first search (BFS) tree composed of the shortest paths between the node which is supposed to be the source and all observers. The algorithm calculates a score for each node s using a density function of (K − 1)-dimensional normal distribution, where K is the number of observers. The node with the highest score is pointed as the rumor source. If |P(u, v)| denotes the length of the path P connecting nodes u and v, d is the vector of observed delays, µ s is the vector of deterministic delays and Λ s is the delay covariance matrix, the score of node s is given by:
[formula] φ(s) = exp(− 1 2 (d − µ s ) T Λ −1 s (d − µ s )) |Λ s | 1/2 ,(1) [/formula]
where
[formula] [d] k = t k − t K ,(2) [/formula]
[µ s ] k = µ(|P(s, o k )| − |P(s, o K )|),
[formula] [Λ s ] k,i = σ 2 |P(o K , o k ) ∩ P(o K , o i )|,(3) [/formula]
for k, i = 1, . . . , K − 1. Despite the simplifications, the method works reasonably well under certain conditions (density of observers ρ = K/N > 10% and propagation ratio λ = µ/σ > √ 2).
PTVA uses the whole network for inferring the source, which makes the algorithm computationally expensive.
It calculates the score for each node in the network which requires to create BFS tree N times. Each BFS tree covers practically the entire network since it reaches to all observers, including these which are far from the root node. The complexity of creating BFS tree is O(N 2 ) per node s ∈ G. Unfortunately, the assumption of diffusion along the shortest paths works well only for the observers which are close to the suspected node (the root of BFS tree) since it makes use of the fact that typical random networks are locally tree-like. However, this may not be true for observers which are far from the supposed source. Thus making use of the distant observers may be problematic -it increases the expenses and decreases the correctness (because there can be many possible paths between the source and the remote observers). So far we have considered the situation when the number of observers is much smaller than the number of other nodes. In this case the overall complexity In general, the algorithm's complexity for arbitrary graphs as O(N (K 3 + N 2 )). Assuming K ∼ N γ , PTVA complexity ranges from O(N 3 ) when γ 2/3 to O(N 4 ) when γ = 1.
1e+03 2e+03 5e+03 1e+04 2e+04 5e+04 1e+05 1e−01 1e+00 1e+01 1e+02 1e+03 network size computation time tree creation (PTVA) matrix operations (PTVA) tree creation (GMLA) matrix operations (GMLA) : The comparison of the times needed for creation of the propagation tree and for matrix operations.
The time complexity of matrix operation for PTVA is O(N 2.68 ) per node (red line), while the time complexity of tree creation is O(N 1.81 ) per node for PTVA (green) and O(N 1.34 ) for GMLA (blue). The tests were performed on BA network with m = 3. The observers were distributed randomly over whole network with density ρ = 0.1.
The results of selected studies on the localization quality of Pinto-Thiran-Vetterli Algorithm executed on data with Limited Information (PTVA-LI) are presented in figures S2 and S3. The explanation and motivation of PTVA-LI can be found in Section Pinto-Thiran-Vetterli Algorithm in the main paper. shows how the algorithm's quality of localization depends on the density of observers. We present these results for three different values of infection rates: β ∈ {0.2, 0.5, 0.8}. As seen in , increasing the density of observers improves the quality of the results to some extent. The higher the infection rate is, the faster the accuracy curve saturates on the level which is equal to the infection rate. This means that the accuracy of PTVA-LI does not converge to unity for ρ → 1 if β < 1. On the other hand the computation time increases very fast with the density of observers, i.e. t ∼ ρ 2 . presents dependency of the localization quality on the infection rate for three different values of the density of observers ρ ∈ {0.1, 0.2, 0.3}. One can see that the infection rate is a crucial factor for the quality of results of PTVA-LI. For β < 0.5 (which corresponds to propagation ratio λ < √ 2) PTVA-LI has worse accuracy and distance error than the baseline method. For 0.5 β < 0.6 PTVA-LI has better accuracy than the baseline method but still worse distance error. For β 0.5 PTVA-LI finally overtakes the baseline method. Our studies demonstrate that PTVA-LI works best with small density of observers ρ ≈ 0.2 (due to the short calculation time) and high propagation ratio (the higher λ, the better performance).
## S.2 gradient maximum likelihood algorithm
GMLA always uses K 0 fastest informed observers (which we call "the nearest" for simplicity) and skips the others. Thus, the question arises if the observers which receive the signal later do not provide valuable additional information. [fig_ref] Figure S4: Normalized logarithm of the score vs the number of nearest observers which... [/fig_ref] shows how the score changes if the algorithm takes into account increasing number of the observers sorted in ascending order of the time of receiving the signal. It also shows that the first dozen or so nearest observers provide enough information to estimate the score of a node. Taking more of them does not substantially affect the result. The question remains whether the nearest observers are the best choice. Perhaps taking into account only the latest observers will provide equally good or better results? presents the tests performed using various groups of observers. The total number of observers is 200 and they are partitioned into 10 non-overlapping groups. The first group contains first 20 observers with the lowest delays.
The second group consists of the observers from positions 21-40. The last, tenth group contains 20 observers which were informed at the latest. These results show that only the nearest observers provide sufficiently accurate information to locate the true source.
The basis of GMLA are two fundamental ideas: smartly limiting the number of observers and a gradient-like selection of suspected nodes. We have studied the effect of each of them on performance of the algorithm. The basic premise behind the selection of suspected nodes is saving time spent on calculating the scores. The number of suspected (and checked) nodes in GMLA is proportional to the logarithm of the network size , and therefore much smaller than the number of suspected nodes in PTVA-LI. shows the quality of the localization results of the algorithm with the gradient-like selection. These results show that the gradient-like selection not only decreases the time of computation but also improves the accuracy, the rank of true source and the distance error. Table S1 explains why the gradient-like selection improves quality of the source detection. The first column shows that PTVA-LI always finds the true source if the score of the source is the global maximum (23.9% of all cases). Otherwise, PTVA-LI is always wrong (76.1% of all cases). In comparison, GMLA (but without limiting the number of observers) sometimes miss the true source even though the score of the source is the best score in the network (1.5% of all cases), but in some cases it can detect the source which has not the maximum score (5.8% of all cases). shows the quality of the localization results of the algorithm which uses the limited number of the nearest observers, but does not use the gradient-like selection of suspected nodes. The accuracy of such an algorithm increases with the network size and outperforms PTVA-LI for networks of size N = 500 or bigger. However, this is true only for scale-free networks. As we write in Discussion in the main paper, performance improvement by limiting the number of observers is due to rejecting observers which are "behind the hubs". The substantial computation time reduction by limiting the number of observers is due to a big acceleration of matrix operations (see . : The results of 1000 realizations of PTVA-LI/GMLA on BA network (N = 1000, k = 6). The numbers on the left side of the cells refer to PTVA-LI. The symbol φ(s * ) denotes the score of the true source, while max(φ) is the highest score in the network in a given realization.
[formula] φ(s * ) = max(φ) φ(s * ) < max(φ) # hits 239 / 224 0 / 58 # mishits 0 / 15 761 / 703 [/formula]
Figures S9 and S10 present the results of the performance tests of GMLA for BA network. A very substantial difference between PTVA and GMLA is that GMLA computes and assigns scores to limited number of nodes and therefore sometimes it can skip the real source. When this happens, the node which is the true source is left without the score thus its rank is equal to the size of the network (it occupies the last place on the score ranking ex aequo with the other skipped nodes). Figures S9a and S10a show how often GMLA assigns the score to the node which is the true source. The probability P (s * ∈ V s ) is a monotonically increasing function of ρ and β which is 0.85 for ρ = 0.2 and β = 0.5 (which are typical values used in our studies). The fact that P (s * / ∈ V s ) < 1 weakens the performance of GMLA but our studies show that this is a case only when ρ or β are very small. On the other hand, when the propagation ratio is very high (β close to 1) PTVA-LI performs better than GMLA. However, in real life the propagation ratio is rather low, because the spreading process is highly nondeterministic. In cases of low and medium β, GMLA performance dominates over PTVA-LI performance despite the fact that GMLA sometimes rejects the true source at the selection stage. The observers were distributed randomly over whole network with density ρ = 0.2. SI model was used for propagation with infection rates β = 0.5. The results are averaged over 100 realizations. The standard errors are smaller than symbols on plot. S.3 Optimal number of the nearest observers for GMLA GMLA requires optimization of the one crucial parameter, namely the number of selected nearest observers K 0 . This parameter controls both the precision and the speed of the algorithm. [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref] shows non-trivial dependency between K 0 and performance of GMLA for various sizes of BA network. Three types of behavior caused by controlling K 0 are visible. The first is seen in [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref] and S11b. The probability that the true source is among the nodes with the highest score [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref] decreases when K 0 increases. This is due to the fact, that the average size of the group of top scorers [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref] also decreases with K 0 . This is because the more observers we use, the better resolution in the score we have. The second type of dependency on K 0 can be seen in [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref] ,d,e,f and g. These are: the accuracy, the rank of the true source, the distance error, the number of suspected nodes, and the computation time. For each of those the optimal value of the nearest observers K * 0 exists. However, this optimal value can be different for different measures of the performance and moreover it is dependent on the size of the network (see in the main article). The interesting fact is that too small value of K 0 increases the calculation time [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref]. This can be easily understood when we notice that the number of suspected nodes [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref] also increases with small K 0 . It follows that when K 0 is too small, GMLA needs more steps before it reaches the local maximum of score and therefore it computes scores for more nodes (which takes time). The last type of behavior is a linear dependency between the size of the propagation tree and the number of the nearest observers (see [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref]. Regardless of the network size, the tree size is about 1.4 times larger than K 0 . It is also worth mentioning that the effect of K 0 on probability of including the true source by GMLA while computing scores is rather small [fig_ref] Figure S11: Performance of GMLA versus the number of the nearest observers K 0... [/fig_ref]. Another method to find the optimal number of the nearest observers is presented in . We test performance of GMLA versus the size of BA network for various types of dependency between K 0 and N . The best results in terms of accuracy, rank and distance error were obtained using function K 0 = 0.5 √ N , which is in agreement with in the main article.
We conducted the same study for Erdős-Rényi (ER) network . Figures S13c,d,e show that for ER graph K * 0 is a saturation point rather than a peak. However, the minimum of the computation time is still visible in . In this case the optimal number of the nearest observers is the minimal number K 0 with the algorithm achieves the maximal performance. presents performance of GMLA versus the size of ER network for various types of dependency between K 0 and N . Again, the best results were obtained for K 0 = 0.5 √ N , as in the case of Barabási-Albert network.
We checked also if other factors can affect the optimal number of the nearest observers. Figures S15-S16
present dependency between K 0 and performance of GMLA for BA and ER networks of various average degrees (between 4 and 10). Despite the fact that ,e and S16c,e may suggest some influence of the average degree on K * 0 , it is too small to be a significance. Moreover, ,b,d,f,g and S16a,b,d,f,g do not confirm existence of such influence. The results of the simulations with various infection rates β show that the optimal number of the nearest observers does not depend on the propagation ratio.
## S.4 performance of gmla for various networks
In addition to the tests performed on Erdős-Rényi and Barabási-Albert networks in the main paper), we present the results for Random Regular Graphs, Exponential Random Graphs (in two versions, generated from the configuration model and evolved by random node attachment) and scale-free networks with degree distribution P (k) ∼ k −3 (created using the configuration model). Both algorithms achieve best results for Random Regular Graphs , but PTVA-LI performs better than GMLA, except for the computation time. For Exponential Random Graphs the algorithms have very similar quality of the source localization with a slight advantage of PTVA-LI for the configuration model and a small advantage of GMLA for the evolving networks S21). Both algorithms have the lowest performance for scale-free networks, but in this case GMLA outperforms substantially PTVA-LI. Deterioration of the quality of detection is the result of existence of hubs in the network (see Discussion in the main paper). The observers which are "behind the hubs" are not reliable (we call them "noisy observers") and distort the estimation of the scores. GMLA uses only K 0 nearest observers, among which there is a lower fraction of the noisy observers than among all observers in the main paper). This in combination with the gradient-like selection of suspected nodes improves the quality of the algorithm's results.
[fig] Figure S2, Figure S3: Performance of PTVA-LI versus the density of observers ρ. The tests were performed on BA network with N = 500 and m = 3. SI model was used for signal propagation with infection rates β = {0.2, 0.5, 0.8} which gives values of propagation ratio λ = { √ 5/2, √ 2, √ 5}. Each point is the result of 1000 realizations. (a) Accuracy. The dotted line shows the performance of the baseline method which is equal to the density of observers. (b) Median rank of the true source. The error bars indicate a spread between 4th and 6th decile. (c) Mean distance error. Dotted lines show performance of the baseline method. Colors of the lines correspond to various infection rates. (d) Mean computation time. The solid line is a power law with the exponent α = 2. Performance of PTVA-LI versus the infection rate β. The tests were performed on BA network with N = 500 and m = 3. The observers were distributed randomly over whole network with densities ρ = {0.1, 0.2, 0.3}. Each point is the result of 1000 realizations. (a) Accuracy. The dotted lines show performance of the baseline method which is equal to density of observers. (b) Median rank of the true source. The error bars indicate a spread between 4th and 6th decile. (c) Mean distance error. The dotted lines show performance of the baseline method. The colors of the lines correspond to the density of observers. (d) Mean computation time. [/fig]
[fig] Figure S4: Normalized logarithm of the score vs the number of nearest observers which were used for the computation. The score of a node (the true source or its first neighbor or its second-order neighbor) was computed for various values of the number of the nearest observers. The results show that increasing the number of the nearest observers above a certain threshold does not change the score substantially. The tests were performed on (a) ER graph with k = 6, (b) BA network with m = 3. The size of networks is N = 1000 [/fig]
[fig] Figure S5, Figure S6, Figure S7, Figure S8, Figure S9, Figure S10: Performance of GMLA depending on the observations used. The observers were grouped according to the time of receipt of the information. The first group contains first 20 observers with the lowest delays. The tests were performed on BA and ER networks. (a) Accuracy drops rapidly at the transition from the first to second group. (b) Median rank of the true source. (c) Mean distance error. (d) Mean computation time. (a) The number of suspected nodes vs the size of BA network. The fit is a logarithmic function N 0 = 6.58 ln(N ) − 3. Parameters: k = 6, ρ = 0.2, β = 0.5, K 0 = 0.5 √ N , 1000 realizations. (b) The number of suspected nodes vs the average degree of BA network. The fit is a linear function N 0 = 3.5 k + 18.2. Parameters: N = 2000, ρ = 0.2, β = 0.5, K 0 = 20, 1000 realizations. The result of a gradient-like selection of suspected nodes in BA network (m = 3). Blue diamonds are the results of the source detection obtained by applying GMLA with K 0 = K, i.e. without the limited number of the observers. Green dots are the results of the baseline method, which always selects the first observer as the true source. (a) Accuracy of GMLA is higher by several percent points than accuracy of PTVA-LI. (b) The band inside the box shows the median of the true source rank. The bottom and top of the box show the values of the first and third quartiles. GMLA has noticeably lower rank than PTVA-LI. (c) The mean distance error of GMLA is lower than PTVA and similar to the mean distance error of the baseline method. (d) The mean computation time of a single realization of GMLA is substantially smaller that PTVA-LI. The solid lines are power laws with exponents 3.1 (red) and 2.4 (blue). Parameters: k = 6, λ = √ 2, ρ = 0.2, K 0 = K. The result of a limited number of observers in BA network (m = 3). Blue diamonds are the results obtained by selecting only K 0 = 0.5 √ N nearest observers (without a gradient-like selection of suspected nodes). Green dots are the results of the baseline method, which always selects the first observer as the true source. (a) Accuracy of GMLA increases with the network size and exceeds the accuracy of PTVA-LI in cases of network with 500 nodes or more. (b) The band inside the box shows the median of the true source rank. The bottom and top of the box show the values of the first and third quartiles. GMLA has lower rank than PTVA-LI for networks of size 1000 or bigger. (c) The mean distance error for PTVA-LI increases with the network size in contrast to GMLA, for which the mean distance error became smaller than PTVA-LI for networks of size 1000 or bigger. (d) The mean computation time of a single realization of GMLA is substantially smaller that PTVA-LI. The solid line is power law with exponents 3.1 (red). Parameters: k = 6, Performance of GMLA versus the density of observers ρ. The tests were performed on BA network with N = 1000 and m = 3. The number of the nearest observers was set to K 0 = 30. SI model was used for propagation with infection rates β = {0.2, 0.5, 0.8} which correspond to values of propagation ratio 5}. Each point is result of 1000 realizations. (a) Empirical probability that the true source is in the group of suspected nodes (and has a calculated score). (b) Accuracy. Dotted line shows performance of the baseline method which is equal to density of observers. (c) Mean size of the group of top scorers. (d) Median rank of the true source. The error bars indicate a spread between 4th and 6th decile. (e) Mean distance error. The dotted lines show performance of the baseline method. The colors of the lines correspond to the infection rate. (f) Mean computation time. Performance of GMLA versus the infection rate β. The tests were performed on BA network with N = 1000 and m = 3. The number of the nearest observers was set to K 0 = 30. The observers were distributed randomly over whole network with densities ρ = {0.1, 0.2, 0.3}. Each point is the result of 1000 realizations. (a) Empirical probability that the true source is in the group of suspected nodes (and has a calculated score). (b) Accuracy. The dotted lines show performance of the baseline method which is equal to density of observers. (c) Mean size of the group of top scorers. (d) Median rank of the true source. The error bars indicate a spread between 4th and 6th decile. (e) Mean distance error. Dotted lines show performance of the baseline method. Colors of the lines correspond to various infection rates. (f) Mean computation time. [/fig]
[fig] Figure S11: Performance of GMLA versus the number of the nearest observers K 0 for BA network with m = 3 and N ∈ {500, 1000, 5000, 10000}. The observers were distributed randomly over whole network with density ρ = 0.3. SI model was used for propagation with infection rate β = 0.5 which gives propagation ratio λ = √ 2. Each point is the result of 5000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It includes only the cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. (g) Mean number of suspected nodes N 0 = |V s | for which scores were determined. (h) Mean size of the propagation tree created for each suspected node. (i) The empirical probability that the true source is in a group of suspected nodes (and has calculated score). [/fig]
[fig] Figure S12, Figure S13: Performance of GMLA versus the network size N for different numbers of the nearest observers K 0 . The tests were performed on BA network with m = 3. The observers were distributed randomly over whole network with density ρ = 0.2. SI model was used for propagation with infection rate β = 0.5 which gives propagation ratio λ = √ 2. Each point is the result of 1000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It includes only the cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. The solid lines are power laws with exponents: 1.64 (yellow), 1.65 (blue), 1.67 (green), 1.72 (red) and 1.77 (violet). (g) Mean number of suspected nodes N 0 = |V s | for which the scores were determined. The solid lines are linear functions of the logarithm of the network size. (h) Mean size of the propagation tree created for each suspected node. The red line is an average of the red points. The yellow and green lines are logarithmic functions. The blue and violet lines correspond to a power law with exponents 0.48 and 0.97 respectively. (i) The empirical probability that the true source is in a group of suspected nodes (and has calculated score). Performance of GMLA versus the number of the nearest observers K 0 for ER graph with k = 6 and N ∈ {500, 1000, 5000, 10000}. The observers were distributed randomly over whole network with density ρ = 0.3. SI model was used for propagation with infection rate β = 0.5 which gives propagation ratio λ = √ 2. Each point is the result of 5000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It includes only the cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. (g) Mean number of suspected nodes N 0 = |V s | for which scores were determined. (h) Mean size of the propagation tree created for each suspected node. (i) The empirical probability that the true source is in a group of suspected nodes (and has calculated score). [/fig]
[fig] Figure S14, Figure S15, Figure S16, Figure S17, Figure S18: Performance of GMLA versus the network size N for different number of the nearest observers K 0 . The tests were performed on ER graph with k = 6. The observers were distributed randomly over whole network with density ρ = 0.2. SI model was used for propagation with infection rate β = 0.5 which gives propagation ratio λ = √ 2. Each point is the result of 1000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It includes only the cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. The solid lines are power laws with exponents: 1.64 (yellow), 1.65 (blue), 1.67 (green), 1.72 (red) and 1.77 (violet). (g) Mean number of suspected nodes N 0 = |V s | for which the scores were determined. The solid lines are linear functions of the logarithm of the network size. (h) Mean size of the propagation tree created for each suspected node. The red line is an average of the red points. The yellow and green lines are logarithmic functions. The blue and violet lines correspond to a power law with exponents 0.48 and 0.97 respectively. (i) The empirical probability that the true source is in a group of suspected nodes Performance of GMLA versus the number of the nearest observers K 0 for BA network with N = 1000 and m ∈ {2, 3, 4, 5}. The observers were distributed randomly over whole network with density ρ = 0.3. SI model was used for propagation with infection rate β = 0.5 which gives propagation ratio λ = √ 2. Each point is the result of 5000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It includes only the cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. (g) Mean number of suspected nodes N 0 = |V s | for which scores were determined. (h) Mean size of the propagation tree created for each suspected node. (i) The empirical probability that the true source is in a group of suspected nodes Performance of GMLA versus the number of the nearest observers K 0 for ER network with N = 1000 and k ∈ {4, 6, 8, 10}. The observers were distributed randomly over whole network with density ρ = 0.3. SI model was used for propagation with infection rate β = 0.5 which gives propagation ratio λ = √ 2. Each point is the result of 5000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It takes into account only cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. (g) Mean number of suspected nodes N 0 = |V s | for which scores were determined. (h) Mean size of the propagation tree created for each suspected node. (i) The empirical probability that the true source is in a group of suspected nodes (and has calculated score). Performance of GMLA versus the number of the nearest observers K 0 for BA network with N = 1000 and m = 3. The observers were distributed randomly over whole network with density ρ = 0.3. SI model was used for propagation with infection rate β ∈ {0.2, 0.5, 0.8} which gives propagation ratio λ = { is the result of 5000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It includes only the cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. (g) Mean number of suspected nodes N 0 = |V s | for which scores were determined. (h) Mean size of the propagation tree created for each suspected node. (i) The empirical probability that the true source is in a group of suspected nodes (and has calculated score). Performance of GMLA versus the number of the nearest observers K 0 for ER network with N = 1000 and k = 6. The observers were distributed randomly over whole network with density ρ = 0.3. SI model was used for propagation with infection rate β ∈ {0.2, 0.5, 0.8} which gives propagation ratio λ = { is the result of 5000 realizations. (a) Detection rate. The empirical probability that the true source is in a group of top scorers. (b) Mean size of the group of top scorers. (c) Accuracy. (d) Mean rank of the true source. It includes only the cases when s * ∈ V s . (e) Mean distance error. (f) Computation time. (g) Mean number of suspected nodes N 0 = |V s | for which scores were determined. (h) Mean size of the propagation tree created for each suspected node. (i) The empirical probability that the true source is in a group of suspected nodes (and has calculated score). [/fig]
[fig] Figure S19, Figure S20, Figure S21: GMLA (blue) versus PTVA-LI (red) and the baseline method (green) on Random Regular Graphs (k = 6). (a) Accuracy of both algorithms increases with the network size, but PTVA-LI performs slightly better.(b) The band inside the box shows the median of the rank of the true source. The bottom and top of the box show values of the first and third quartiles. PTVA-LI has lower rank than GMLA. (c) The mean distance error of both algorithms decreases with the network size, but PTVA-LI performs better. (d) The mean computation time of a single realization of GMLA is substantially smaller that PTVA-LI. The solid line is a power law with exponent 3.1 (red). Parameters: k = 6, λ = √ 2, ρ = 0.2, K 0 = 0GMLA (blue) versus PTVA-LI (red) and the baseline method (green) on Exponential Random Graphs (created using the configuration model). (a) Accuracy of both algorithms is similar and increases with the network size. (b) The band inside the box shows the median of the rank of the true source. The bottom and top of the box show values of the first and third quartiles. PTVA-LI has slightly lower rank than GMLA. (c) The mean distance error of both algorithms decreases with the network size, but PTVA-LI performs better. (d) The mean computation time of a single realization of GMLA is substantially smaller that PTVA-LI. The solid line is a power law with exponent 3.1 (red). Parameters: k = 6, GMLA (blue) versus PTVA-LI (red) and the baseline method (green) on Exponential Random Graphs (evolved by random node attachment). (a) Accuracy of both algorithms increases with the network size, but GMLA performs slightly better. (b) The band inside the box shows the median of the rank of the true source. The bottom and top of the box show values of the first and third quartiles. Results of both algorithms have similar rank. (c) The mean distance error of both algorithms decreases with the network size, but PTVA-LI performs better. (d) The mean computation time of a single realization of GMLA is substantially smaller that PTVA-LI. The solid line is a power law with exponent 3.3 (red). Parameters: k = 6, [/fig]
|
Advances in Mechanism Research on Polygonatum in Prevention and Treatment of Diabetes
Diabetes mellitus is a fast-growing disease with a major influence on people's quality of life. Oral hypoglycemic drugs and insulin are currently the main effective drugs in the treatment of diabetes, but chronic consumption of these drugs has certain side effects. Polysaccharides, saponins, flavonoids, and phenolics are the primary secondary metabolites isolated from the rhizomes of Polygonatum sibiricum Redouté [Asparagaceae], Polygonatum kingianum Collett & Hemsl [Asparagaceae], or Polygonatum cyrtonema Hua [Asparagaceae], which have attracted much more attention owing to their unique therapeutic role in the treatment and prevention of diabetes. However, the research on the mechanism of these three Polygonatum spp. in diabetes has not been reviewed. This review provides a summary of the research progress of three Polygonatum spp. on diabetes and its complications, reveals the potential antidiabetic mechanism of three Polygonatum spp., and discusses the effect of different processed products of three Polygonatum spp. in treating diabetes, for the sake of a thorough understanding of its effects on the prevention and treatment of diabetes and diabetes complications.
# Introduction
Diabetes mellitus (DM) is a comprehensive endocrine and metabolic disease characterized by glucose metabolism disorders, mainly resulting from insulin resistance or insufficient insulin secretion [bib_ref] Extracellular Vesicles in Type 2 Diabetes Mellitus: Key Roles in Pathogenesis, Complications,..., Xiao [/bib_ref]. According to the American Diabetes Association, it is divided into four major types: type 1 diabetes (T1DM), type 2 diabetes (T2DM), gestational diabetes (GDM), and diabetes from other causes. T2DM has the highest incidence among these diseases [bib_ref] Diabetes Trends in the USA, Skyler [/bib_ref]. The key factors that cause T2DM are pancreas β-cell failure, insulin resistance, and its complex interrelationships [bib_ref] Of Fat, β Cells, and Diabetes, Thorens [/bib_ref]. More importantly, long-term hyperglycemia may cause malfunction and long-term damage in a variety of tissues and organs, particularly the eyes, nerves, kidneys, heart, and blood vessels [bib_ref] Diabetic Neuropathy Research: from Mouse Models to Targets for Treatment, Pham [/bib_ref] [bib_ref] Review of Glucagon-like Peptide-1 Receptor Agonists for the Treatment of Type 2..., Sloan [/bib_ref].
In recent years, DM has become one of the primary diseases endangering modern people, and the number of patients has been increasing year by year. Almost all patients require oral hypoglycemic agents or injecting insulin. Diabetes is difficult to control in a maintainable long-term lifestyle [bib_ref] Management of Type 2 Diabetes: New and Future Developments in Treatment, Tahrani [/bib_ref] [bib_ref] Erratum. Impact of Lifestyle and Metformin Interventions on the Risk of Progression..., Herman [/bib_ref]. Current oral hypoglycemic agents include the earlier developed metformin and sulfonamides, as well as some novel hypoglycemic agents targeting the pancreas or liver, such as sodium-dependent glucose transporter 2 (SGLT2) inhibitors, dipeptidyl peptidase-4 (DPP-4) inhibitors, and glucagon-like peptide 1 (GLP-1) receptor agonists, which can enhance insulin activity, exert insulin-like effects, or alleviate glucose metabolism disorders [bib_ref] Pharmacology and Therapeutic Implications of Current Drugs for Type 2 Diabetes Mellitus, Tahrani [/bib_ref]. However, the clinical application of insulin often causes hypoglycemia, insulin resistance, lipoatrophy, and other side effects. Although the newly developed drugs have had a lower risk of hypoglycemia in recent years, they are more expensive than other earlydeveloped drugs such as sulfonylureas, and long-term safety has yet to be determined. Therefore, it is essential to screen the natural products with antidiabetic activity and investigate their material basis, pharmacodynamics, and mechanism of action to provide new ideas for developing high efficiency and low toxicity antidiabetic drugs.
Traditional Chinese medicines (TCMs) are gaining popularity as a result of their success in the treatment and prevention of diabetes, such as Panax ginseng C. A. Mey. , Lycium barbarum L. , Coptis chinensis Franch.
[Ranunculaceae], Abelmoschus esculentus (L.) Moench , Angelica sinensis (Oliv.) Diels , and Andrographis paniculata (Burm.f.) Nees [Acanthaceae] [bib_ref] Synergetic Effect of Andrographis Paniculata Polysaccharide on Diabetic Nephropathy with Andrographolide, Xu [/bib_ref] [bib_ref] Anti-hyperglycemic and Anti-oxidative Activities of Ginseng Polysaccharides in STZ-Induced Diabetic Mice, Sun [/bib_ref] [bib_ref] LBP-4a Improves Insulin Resistance via Translocation and Activation of GLUT4 in OLETF..., Zhao [/bib_ref] [bib_ref] Protective Effects of Angelica Sinensis Polysaccharide against Hyperglycemia and Liver Injury in..., Wang [/bib_ref] [bib_ref] The Inhibiting Effect of the Coptis Chinensis Polysaccharide on the Type II..., Cui [/bib_ref] [bib_ref] The Anti-nephritic Activity of a Polysaccharide from Okra (Abelmoschus Esculentus (L.) Moench)..., Liao [/bib_ref]. Polygonatum is a common genus from the Asparagaceae family widely distributed in China and has been used as medicine or food for more than 2,000 years because it invigorates the spleen, moisturizes the lungs, and invigorates qi. The plant part used is the dry rhizome of Polygonatum sibiricum Redouté (P.sibiricum), Polygonatum kingianum Collett & Hemsl (P. kingianum), or Polygonatum cyrtonema Hua (P. cyrtonema), introduced in the 2020 edition of the Pharmacopoeia of the People's Republic of China. The underlying pharmacological applications of Polygonatum are gaining popularity in clinical diseases, such as fatty liver disease, Alzheimer's disease, diabetes mellitus, and cancer . Such biological activities are closely related to the secondary metabolites of Polygonatum, including polysaccharides, saponins, flavonoids, phenolics, alkaloids, anthraquinones, lignans, and a variety of beneficial amino acids [bib_ref] Research Progress in Polygonati Rhizoma and Predictive Analysis on Q-Marker, Jiang [/bib_ref]. Recently, Polygonatum spp. have become widely used TCMs in improving diabetes.
In this work, we comprehensively analyzed the antidiabeticrelated research work on three Polygonatum spp. The purpose of this work is to review the secondary metabolites of Polygonatum and their antidiabetic mechanism, investigate the effects of different processed products of Polygonatum on treatment, and lay the foundation for the clinical application and product development of Polygonatum.
## Data collection
According to published reports from 2011 to 2021, "Polygonatum" or "Rhizoma polygonti" combined with "diabetes" or "anti-diabetic," "secondary metabolites," and "processed products" were used as search keywords. The data were collected by various online databases, including PUBMED, Web of Science, Science Direct, SpringerLink, Wiley Online Library, Wanfang, and China Knowledge Network. About 189 papers were found by reading abstracts to exclude repetitive and irrelevant papers. The data were further extracted from the above studies: P. sibiricum, P. kingianum, and P. cyrtonema were used according to the 2020 edition of the Pharmacopoeia of the People's Republic of China; test design with the control group and functional verification; and dose use strictly in line with the standard (rat dose = human dose g * 0.018/0.02 kg). Eventually, we found that 47 articles met the screening standard and brought into this paper by critically reviewing and analyzing the data, aiming to identify secondary metabolites and processed products of Polygonatum involved in the antidiabetic mechanism.
## Secondary metabolites of polygonatum
Currently, the secondary metabolites of Polygonatum have been reported to include polysaccharides, saponins (steroidal saponins and triterpenoids), flavonoids, phenols, alkaloids, lignans, phytosterols, and volatile oils, of which the first four ones are the major ingredients and have been studied most frequently. Additionally, polysaccharides and saponins were the highest in P. cyrtonema, and flavonoids and other phenolics were the highest in P. sibiricum [fig_ref] TABLE 1 |: Comparison of the major chemical constituents of three Polygonatum spp [/fig_ref].
## Polysaccharides
Polysaccharide is not only an active important component of Polygonatum but also an important evaluation index of its quality. It has been reported that polysaccharide is composed of many monosaccharides including fructose (Fru), glucose (Glc), mannose (Man), galactose (Gal), arabinose (Ara), and rhamnose (Rha), as well as a handful of glucuronic acid (GlcA) and xylose (Xyl). The molecular weights of polysaccharides from Polygonatum plants are estimated to be approximately 2,734~3.6 × 10 5 Da [bib_ref] The Genus Polygonatum: A Review of Ethnopharmacology, Zhao [/bib_ref]. Two new polysaccharides (PSP50-2-1 and PSP50-2-2) were isolated and purified from the rhizome of P. sibiricum, both of which were homogeneous polysaccharides by the analysis of the specific optical rotation. Meanwhile, the result of monosaccharide composition indicated that PSP50-2-1 and PSP50-2-2 were made up of Glc, Gal, and Fru [bib_ref] Structural Characterization and Osteogenic Activity In Vitro of Novel Polysaccharides from the..., Liu [/bib_ref] , with the molecular weight of 7.7 and 7.0 kDa, respectively. More importantly, found that four polysaccharides isolated from P. sibiricum (PSP1, PSP2, PSP3, and PSP4) were made up of Gal, Rha, Man, Glu, and Xyl in different proportions, and the immune activity of polysaccharides was closely related to that of Rha residues, with the molecular weight of 4.415, 2.236, 7.743, and 6.467 kDa, respectively . Zhao et al. found that polysaccharides isolated from P. sibiricum, P. kingianum, and P. cyrtonema were mainly made up of Fru and pectins, with a molecular weight of more than 4.1 × 10 5 Da . different structures of saponins in Polygonatum, saponins were divided into steroidal saponins and triterpenoid saponins. Zhao et al. summarized 162 saponins from 18 species of Polygonatum genus, among which 70 steroidal saponins and 12 triterpenoid saponins were isolated from P. sibiricum, P. kingianum, and P. cyrtonema [bib_ref] The Genus Polygonatum: A Review of Ethnopharmacology, Zhao [/bib_ref]. Subsequently, some studies provided novel findings of five novel steroidal saponins isolated from P. sibiricum, 3-O-β-D-glucopyranosyl(1→2)-β-
[formula] D-glucopyranosyl(1→4)-β-D-fucopyranosyl-(25R)-spirost-5-en- 3β,17α-diol, 3-O-β-D-glucopyranosyl(1→4)-β-D-fucopyranosyl- (25R/S)-spirost-5-en-3β,12β-diol, 3-O-β- D-glucopyranosyl(1→4)-β-D-fucopyranosyl-(25R)-spirost-5-en- 3β,17α-diol, 3-O-β-D-glucopyranosyl(1→4)-β-D-fucopyranosyl-( [/formula]
25S)-spirost-5-en-3β,17α-diol, and kingianoside Z [bib_ref] Steroidal Saponins from the Rhizome of Polygonatum Sibiricum, Tang [/bib_ref]. Two new steroidal saponins were isolated from P. kingianum, named polygokingiaside A and polygokingiaside B, respectively [bib_ref] Polygokingiasides A and B: Two New Spirostane Glycosides from the Roots of..., Ha [/bib_ref]. A novel steroidal saponin was isolated from P. cyrtonema, named Huangjingsterol B . On the other hand, no new triterpenoid saponins were found in Polygonatum plants because triterpenoid saponins are found principally in the Magnoliopsida class, and steroidal saponins are distributed widely in the Liliopsida class [bib_ref] Saponins and Their Role in Biological Processes in Plants, Faizal [/bib_ref].
## Phenolics
Phenolics include flavonoids, phenolics, and lignins. Flavonoids are ubiquitous in natural plants and have a broad spectrum of biological activities. Until now, 34 flavonoids have been isolated from P. sibiricum, P. kingianum, and P. cyrtonema, which can be divided into six types in accordance with the structure of the parent nucleus: homoisoflavones, isoflavones, flavones, chalcones, dihydroflavones, and rosandalanes [fig_ref] TABLE 2 |: Flavonoids isolated from three Polygonatum spp [/fig_ref]. Among them, homoisoflavones are the most abundant in Polygonatum, such as 4′,5,7-trihydroxy-6-methyl-8-methoxyhomoisoflavanon, disporopsin, and polygonatone H. Phenolics in plants are secondary metabolites synthesized during the normal development of plants. Relatively rare studies have been conducted on the structural properties of the phenolics and lignans from Polygonatum. Wang et al. identified two known compounds (narcissoside and nicotiflorin) from P. sibiricum by 1D/2D NMR and MS data [bib_ref] Constituents from Polygonatum Sibiricum and Their Inhibitions on the Formation of Advanced..., Wang [/bib_ref]. Zhai and Wang isolated syringaresinol-di-O-β-D-glucoside from P. sibiricum [bib_ref] Syringaresinol-di-O-β-D-glucoside, a P-henolic C-ompound from Polygonatum sibiricum, Zhai [/bib_ref]. Chen et al. isolated a benzofuran-type lignan (polygonneolignanoside A) from P. sibiricum [bib_ref] A New Benzofuran Lignan from Rhizomes of Polygonatum Sibiricum, Chen [/bib_ref].
## Other secondary metabolites
The contents of alkaloids, phytosterols, and volatile compounds in Polygonatum were extremely low, and their structures were less studied. Polygonatine A and Polygonatine B isolated from P. sibiricum were identified as alkaloids [bib_ref] Two New Alkaloids from the Rhizome of Polygonatum Sibiricum, Sun [/bib_ref]. Four phytosterol compounds have already been identified in P. sibiricum and P. kingianum, including β-sitosterol, carotenoside, palmitate-3β sitosterol, ester and (22S)-cholest-5ene-1β,3β,16β,22-tetrol 1-O-α-L-rhamnopyranosyl 16-O-β-D-glucopyranoside [bib_ref] Diffusiondriven Instability and Hopf Bifurcation in Brusselator System, Li [/bib_ref] [bib_ref] A Bisdesmosidic Cholestane Glycoside from the Rhizomes of Polygonatum Sibiricum, Ahn [/bib_ref]. Volatile compounds were found in the rhizomes of P. cyrtonema, which accounted for 95.97% of the total volatile oils [bib_ref] Study on Constituents and Biological Activity of Volatile Oil from Tubers of..., Yu [/bib_ref].
## Potential antidiabetic mechanism of polygonatum on diabetes
Studies have shown that certain active ingredients of traditional Chinese herbal medicines have apparent effects of lowering blood sugar and blood lipids, such as polysaccharides, saponins, flavonoids, phenols, and alkaloids [bib_ref] Hypoglycemic and Hypolipidemic Effects of Total Saponins from Stauntonia Chinensis in Diabetic..., Xu [/bib_ref] [bib_ref] Characterization, Hypolipidemic and Antioxidant Activities of Degraded Polysaccharides from Ganoderma Lucidum, Xu [/bib_ref] [bib_ref] Phenolic Profiles, Antioxidant, Antiproliferative, and Hypoglycemic Activities of Ehretia Macrophyla Wall. (EMW)..., Deng [/bib_ref] [bib_ref] Protective Effect and Possible Mechanisms of Ligustrazine Isolated from Ligusticum Wallichii on..., Zhuang [/bib_ref]. Polygonatum is rich in these substances and hence is a Chinese herbal medicine with great medicinal value. The number of research papers on secondary metabolites and biological activities of Polygonatum is increasing in recent decades.
To date, there are three models to study the antidiabetic mechanism of secondary metabolites from Polygonatum: cells, diabetic animal models, and humans [fig_ref] FIGURE 1 |: Experimental workflow in the study of Polygonatum spp [/fig_ref]. For example, in in vitro studies, in which the IR-3T3-L1 adipocytes and IR-HepG2 cells were cultured, it was found that Polygonatum could increase glucose intake by alleviating oxidative stress and inflammation [bib_ref] Hypoglycemic Effects and Modulation of Gut Microbiota of Diabetic Mice by Saponin..., Luo [/bib_ref]. In animal models of diabetes, for the sake of identifying the metabolic impact of the Polygonatum rhizome extract, high-fat diet (HFD)-, streptozotocin (STZ)-, or alloxan-induced rats were administered Polygonatum orally at a certain dose for a period. It is suggested that Polygonatum could decrease high blood glucose by analyzing various factors related to metabolic syndrome [bib_ref] Extraction Optimization of Saponin from Polygonatum Sibiricum by Design-Expert and Preliminary Study..., Pang [/bib_ref] [bib_ref] Polysaccharides from Polygonatum Kingianum Improve Glucose and Lipid Metabolism in Rats Fed..., Gu [/bib_ref] [bib_ref] Structural characterization, hypoglycemic effects and antidiabetic mechanism of a novel polysaccharides from..., Li [/bib_ref]. In addition, Polygonatum also improves homeostasis model assessment of insulin sensitivity (HOMA-IS) and homeostasis model assessment of insulin resistance (HOMA-IR) of patients with diabetes in clinical studies [bib_ref] Effect of Adjuvant Therapy of Jiangtang Tongmai Capsule on HOMA-IS and HOMA-IR..., Ping [/bib_ref].
## In vitro models
The nuclear factor erythroid 2-related factor 2/heme oxygenase-1 (Nrf2/HO-1) signaling pathway is closely related to pancreatic βcell injury, obesity, glucose metabolism disorders, and insulin resistance. Polysaccharides of P. sibiricum (PSP) (50, 100, and 250 μg/ml) can alleviate IR and proliferation of IR-3T3-L1 adipocytes by activating Nrf2/HO-1 signaling pathway in IR-3T3-L1 adipocytes, they promoted the expression of Nrf2 and HO-1 and lessened the expression levels of inflammatory cytokines [interleukin-1β (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α)], and subsequently enhanced glucose intake by stimulating the expression of transporter subtype-4 (GLUT4). When the Nrf2 gene was silenced, the expressions of inflammatory cytokines, HO-1, GLUT4, and glucose intake were elevated, thereby reversing the therapeutic effect of PSP in IR-3T3-L1 adipocytes. In IR-HepG2 cells, polysaccharides from P. kingianum (PKP) enhanced the levels of glucose utilization efficiency at doses of 6.25, 12.5, and 25 mg/L ; saponins from P. sibiricum (PSS) could significantly inhibit insulin resistance in a dose-dependent manner in HepG2 cells, but it was noteworthy that when the concentration of PSS was above 500 μg/ml, it affected cell viability. Moreover, PSS also could markedly attenuated the activities of α-glucosidase and α-amylase in vitro [bib_ref] Hypoglycemic Effects and Modulation of Gut Microbiota of Diabetic Mice by Saponin..., Luo [/bib_ref].
## Animal models
Polysaccharides of P. cyrtonema (PCP) (450-900 mg/kg) significantly improved the survival rate of STZ-induced T1DM female rats by inhibiting weight loss, suppressing inflammatory cytokine expression in the liver, and increasing insulin receptor substrate (IRS) expression, thereby improving the hepatic immune response. More importantly, both low (120 mg/kg) and high (480 mg/kg) doses of PKP improved diabetic symptoms by increasing short-chain fatty acid (SCFA) levels, modulating gut microbiota composition, and reducing inflammation in HFD rats [bib_ref] Polysaccharides from Polygonatum Kingianum Improve Glucose and Lipid Metabolism in Rats Fed..., Gu [/bib_ref].
Saponins from P. kingianum (TSPK) also have antidiabetic effects. STZ-induced diabetic rats were given TSPK for 8 weeks at 0.025 and 0.1 g/kg, TSPK could alleviate hyperlipidemia and hyperglycemia in diabetic rats, and the genome-wide expression indicated that expression of GLUT4 was significantly upregulated. In contrast, the expression of G6P was downregulated in the insulin signal pathway [bib_ref] Antidiabetic Effect of Total Saponins from Polygonatum Kingianum in Streptozotocin-Induced Daibetic Rats, Lu [/bib_ref]. The structure and number of gut microbiota of rats treated with TSPK were significantly changed, so TSPK may prevent T2DM by regulating gut microbiota and the secretion of SCFAs [bib_ref] Intake of Total Saponins and Polysaccharides from Polygonatum Kingianum Affects the Gut..., Yan [/bib_ref]. Furthermore, PSS can activate hexokinase and then converts glucose to glucose-6phosphatase (G6P), which promotes glycogen synthesis and ultimately reduces insulin resistance. Interestingly, the number of bacteria changed in the dung of the T2DM rats treated with PSS (1, 1.5, and 2 g/kg), with the result that the number of probiotics increased and the number of harmful bacteria decreased [bib_ref] Hypoglycemic Effects and Modulation of Gut Microbiota of Diabetic Mice by Saponin..., Luo [/bib_ref]. found that total flavonoids of P. sibiricum (TFP) have significant hypoglycemic effects on both T1DM and T2DM. Compared to those of the control group, the hypoglycemic effects of 100 and 200 mg/kg of TFP were similar to those of 20 mg/kg of acarbose in STZ-induced T1DM rats. In HFD-and alloxan-induced T2DM rats, 200 mg/kg of TFP had a similar hypoglycemic effect to 15 mg/kg of gliclazide. After 9 days of treatment with 100 and 200 mg/kg of TFP, the fasting blood glucose (FBG) of rats decreased in a dose-dependent manner. Besides, TFP significantly inhibited α-amylase activity in a dosedependent manner in vitro [bib_ref] Anti-diabetic Effects of Total Flavonoids from Polygonatum Sibiricum Red in Induced Diabetic..., Shu [/bib_ref]. Overall, TFP may have multiple beneficial effects on lessening hyperglycemia induced by alloxan, STZ, and HFD in diabetic rats, respectively. However, there is another class of phenolic compound (syringaresinol-di-O-β-D-glucoside (SOG)) isolated from P. sibiricum that exerts an antidiabetic effect. Treatment with SOG (25, 50, and 75 mg/kg) facilitated insulin secretion and reduced the levels of lipid metabolism and oxidative stress in the STZ-induced diabetic rats, as well as downregulated the expression of nitrotyrosine (NT) and TGF-β1 in kidneys [bib_ref] Syringaresinol-di-O-β-D-glucoside, a P-henolic C-ompound from Polygonatum sibiricum, Zhai [/bib_ref]. Thus, SOG showed a significant antidiabetic effect by suppressing oxidative stress.
In summary, polysaccharides, saponins, flavonoids, and other phenolics of Polygonatum have a prominent role in lowering blood sugar and blood lipids in DM [fig_ref] TABLE 3 |: Antidiabetic properties of three Polygonatum spp [/fig_ref]. The minimum dose of Polygonatum secondary metabolites is 25 mg/kg, and the maximum dose is 2 g/kg.
## Clinical application
To date, clinical studies verified that a few Chinese patent medicines containing Polygonatum have a beneficial effect on diabetes, such as Tangwei capsules, Jinlida granules, Tangmaikang granules, Jiangtangjia tablets, and Qizhi Jiangtang capsules (Tables 4, 5). Jinlida granules could significantly decrease the level of hemoglobin A1c (HbA1c) and fasting plasma glucose (FPG) in the 2-h postprandial blood glucose (2hPG) in the individuals who received Jinlida granules (9 g) compared to the control groups. Jiangtang Tongmai capsules (1.05 g) combined with glibenclamide can reduce the blood glucose level, improve HOMA-IS and HOMA-IR of patients with T2DM, and reduce the severity of clinical symptoms of T2DM [bib_ref] Effect of Adjuvant Therapy of Jiangtang Tongmai Capsule on HOMA-IS and HOMA-IR..., Ping [/bib_ref]. HbA1c and HOMA-IR were significantly decreased after treatment with Jiangtangshu tablets (1.5 g) combined with repaglinide, while GLP-1 and fasting serum insulin (FINS) levels were significantly increased [bib_ref] Clinical Study on Jiangtangshu Tablets Combined with Repaglinide in Treatment of Type..., Li [/bib_ref]. After treatment with Qizhi Jiangtang capsules (2.5 g), NO serum content was increased, and endothelin-1 (ET-1), thromboxane B2 (TXB2), blood urea nitrogen (BUN), serum creatinine (SCr) contents were lower than those in the control group, which eventually improved renal microcirculation and dysfunction [bib_ref] Effect of Qizhi Jiangtang Capsule on Microcirculation and Renal Function in Diabetic..., Si [/bib_ref]. In conclusion, these Chinese patent medicines could effectively control blood glucose and inhibit insulin resistance without significant adverse effects and could be used as an adjuvant drug for the treatment of T2DM and its complications.
## Potential antidiabetic mechanism of polygonatum on diabetes complications
Diabetes can also lead to complications of other diseases, such as acute kidney injury (AKI), diabetic retinopathy (DR), and diabetic nephropathy (DN). The p38 MAPK is the most critical and common signaling pathway in protecting against inflammatory kidney injury (Ahmed and Mohamed, 2018). Gentamicin (GM) can stimulate the secretion of inflammatory cytokines via activation of p38 mitogen-activated protein kinase (MAPK)/activation transcription factor 2 (p38 MAPK/ATF2) pathway, triggering a set of inflammatory cascade reactions that result in kidney injury. However, PSP could markedly Frontiers in Pharmacology | www.frontiersin.org February 2022 | Volume 13 | Article 758501 decrease the expression levels of neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury molecule-1 (KIM-1), preventing the p38 MAPK/ATF2 pathway to suppress the secretion of inflammatory cytokines in the kidney [bib_ref] Protective Effect of Polygonatum Sibiricum Polysaccharides on Gentamicin-Induced Acute Kidney Injury in..., Han [/bib_ref]. As a result, PSP has a potential pharmacotherapy on GMinduced AKI rats. VEGF is a crucial angiogenic growth factor that facilitates the migration, proliferation, and angiogenesis of vascular endothelial cells [bib_ref] The MAPK Signaling Pathway Mediates the GPR91-dependent Release of VEGF from RGC-5..., Hu [/bib_ref]. Moreover, some growth factors can promote retinal cell proliferation, such as transforming growth factor-β (TGF-β), which can contribute to cell proliferation and differentiation and suppress DNA synthesis of vascular endothelial cells [bib_ref] Elevated Serum Levels of Soluble TNF Receptors and Adhesion Molecules Are Associated..., Sharma [/bib_ref]. Epidermal growth factor (EGF) works on the proliferation of retinal capillary endothelial [bib_ref] Inhibition of EGF Signaling Protects the Diabetic Retina from Insulin-Induced Vascular Leakage, Sugimoto [/bib_ref]. However, the treatment of PSP notably reduced the expression of VEGF, TGF-β, and EGF in the DR retina. In STZ-induced DR rats, the expression of apoptotic protein B-cell lymphoma-2 factor (Bcl-2) was enhanced, while the expression of Bcl2-associated X protein (Bax) and p38 was reduced in PSP-treated rats. p38 MAPK is pivotal in the regulation of apoptosis. In addition, PSP can also reduce the activity of the superoxide dismutase (SOD) enzyme and increase the content of malondialdehyde (MDA), thus reducing oxidative stress of DM rats .
Wnt/β-catenin pathway (Wnt) signaling is involved in pancreas development and islet function [bib_ref] Glucagon-like Peptide-1 Activation of TCF7L2-dependent Wnt Signaling Enhances Pancreatic Beta Cell Proliferation, Liu [/bib_ref] [bib_ref] Association of Canonical Wnt/β-Catenin Pathway and Type 2 Diabetes: Genetic Epidemiological Study..., Wang [/bib_ref] [bib_ref] Cross-talk between Insulin and Wnt Signaling in Preadipocytes: Role of Wnt Co-receptor..., Palsgaard [/bib_ref] and plays a vital role in modulating GLP-1 through regulating the transcription of the proglucagon gene in . Zou et al. proved that Shen'an granules could regulate urinary protein, renal function, and dyslipidemia in DN rats, and such effects are achieved by suppressing the activation of the Wnt/β-catenin signaling pathway [bib_ref] Effects of Shen'an Granules on Wnt Signaling Pathway in Mouse Models of..., Zou [/bib_ref]. Furthermore, the hypoglycemic effect of PSS on T2DM was also related to the Wnt/β-catenin signaling pathway. There is evidence that the expression of Wnt4 and β-catenin in the DN model group has been notably enhanced compared with that of the control group. In contrast, the expression of Wnt4 and β-catenin in the highdose and low-dose PSS groups notably decreased [bib_ref] Protective Effects of Polygonatum Sibiricum Saponin on Kidney and Wnt/β-Catenin Signaling Pathway..., Jing [/bib_ref]. Therefore, PSS can suppress the process of tubulointerstitial fibrosis by blocking the activation of the Wnt/β-catenin signaling pathway and finally plays a vital role in kidney protection.
In brief, the studies of molecular mechanisms suggest that Polygonatum influences the development of diabetic complications by regulating MAPK, adenosine monophosphate-activated protein kinase (AMPK), and Wnt/βcatenin signaling pathway .
## Effects of different processed products of polygonatum on diabetes
TCMs need to be processed to have a better therapeutic effect, unlike Western medicine. Processed TCMs have an apparent therapeutic effect, low toxicity, and convenience for storability. Moreover, different processing methods of the same drug show different efficacy.
An effort was made to compare the effect of hypoglycemic and hypolipidemic among ninefold-processed P. kingianum and four products of P. kingianum processed with different auxiliary materials (wine, black beans, Rehmannia glutinosa (Gaertn.) DC [Orobanchaceae] and L. barbarum L.
[Solanaceae]). All five processed P. kingianum products were administered in high-glucose rats and high-fat rats at doses of 1.95 and 1.35 mg/g, respectively. The results confirmed that the hypoglycemic effect of ninefoldprocessed P. kingianum and P. kingianum processed with R. glutinosa (Gaertn.) DC. had markedly enhanced effects, while L. barbarum L. processed P. kingianum shows no significant effect. Additionally, ninefold-processed P. kingianum has the best hypolipidemic effect, and L. barbarum L. processed P. kingianum has the lowest effect through detecting the content of four factors related to lipid metabolism [total cholesterol (TC), triglyceride (TG), highdensity lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C)] [bib_ref] A Preliminary Study on the Effect Difference of Processed Products of Rhizoma..., Zhang [/bib_ref]. Other studies revealed that water extracts from each processed product of P. sibiricum (10 g/kg) were given by intragastric administration for 6 weeks; fourfold processing of P. sibiricum can better improve Qi and Yin deficiency syndrome by increasing the body weight and tail diameter of rats and regulating the glucose and lipid metabolism, compared to ninefold-processed P. sibiricum [bib_ref] Pharmacodynamic Effects of Different Processed Products of Rhizoma Polygonati on Rat Model..., Ma [/bib_ref].
Li et al. found that fermented P. sibiricum (FPS) could lower insulin, FBG, and lipid metabolism than P. sibiricum. FPS showed greater efficacy than P. sibiricum in decreasing insulin resistance by increasing the p-AKT/AKT ratio, and FPS had a hypolipidemic effect on liver and fat in STZ-induced diabetic rats by improving lipolysis and inhibiting adipogenesis .
## Future prospects
The beneficial effects of antidiabetes may be related to the metabolites of natural products in the human body. For instance, conjugated (glucuronidated and sulfated) metabolites of hydroxytyrosol and oleuropein are detected in plasma and urine following oleuropein consumption at a single dose of 76.6 mg per person. The concentration of oleuropein metabolites was significantly increased compared with oleuropein (149 vs. 3.55 ng/ml) in plasma [bib_ref] Human Absorption and Metabolism of Oleuropein and Hydroxytyrosol Ingested as Olive, De Bock [/bib_ref]. However, there are no studies on the beneficial effects of chemical components of Polygonatum against diabetes, and this may be related to metabolites in humans, and the specific mechanism needs to be further studied. Beyond that, it also is worth further exploring whether different active components of Polygonatum work alone or in a particular proportion with better curative effect against diabetes.
# Conclusion
Diabetes mellitus, known as thirst dissipation in ancient China, was characterized by polydipsia, polyuria, polyphagia, emaciation, fatigue, and frequent urination. Now, it is common knowledge that DM is a group of metabolic diseases characterized by hyperglycemia, which is a chronic disease that cannot be cured by pharmaceutical means, but treatments can alleviate the development and symptoms of diabetes. With the increasing number of diabetic patients, natural products of Polygonatum (polysaccharides, saponins, flavonoids, and phenols) have attracted wide attention on account of their efficacy in lowering blood sugar and blood lipids. However, there are more studies on the hypoglycemic effect of polysaccharides and saponins than that of flavonoids and phenols. However, flavonoids and other phenolics are worthy of being studied. In addition, this review also summarizes the three insulin signaling pathways-p38MAPK, AMPK, and Wnt/β-catenin signaling pathways-that might be involved in the treatment of diabetes with Polygonatum, whereas these signaling pathways could result in a variety of biological activities to change, such as glucose uptake and glycogen synthesis, cell survival, oxidative stress, inflammation, and lipid metabolism. Consequently, the mechanism of action and targets of Polygonatum have been studied from the perspective of its unique chemical components, which is crucial to lay the foundation for clinical research.
Preclinical and clinical studies have shown that Polygonum has a positive therapeutic effect on diabetes. However, there is still a lack of research on Polygonum intake in humans. It is worth noting that the minimum effective dose of Polygonum must be determined in clinical studies due to individual differences.
Overall, the antidiabetic efficacy of Polygonum is well-known, and the antidiabetic benefits of bioactive components, especially polysaccharides and saponins, have widely been reported. Meanwhile, the combination use of Polygonatum and other clinical hypoglycemic drugs could enhance the therapeutic effect of hypoglycemic drugs, giving Polygonatum a broader FIGURE 2 | Insulin signaling pathways of three Polygonatum spp. Note. mTOR, mammalian target of rapamycin; PDK1/2, 3-phosphoinositide-dependent protein kinase-1/2; GSK3, glycogen synthase kinase 3β.
Frontiers in Pharmacology | www.frontiersin.org February 2022 | Volume 13 | Article 758501 9 application prospect in the treatment of diabetes and its complications.
# Author contributions
SL and Q-JJ: conceptualization, writing-original draft. Y-QP, TF, and SH: collected the literatures. JD and Z-SL: writing-review and editing.
# Funding
The present study was supported by the Key R&D program project from Zhejiang Province (2017C02034), Basic Public Welfare Research Plan of Zhejiang Province (LGN21C020008), Characteristic Industrial Chain Project of Shaanxi Provincial Science and Technology Department (2019TSLSF02-02), and Key Program of Shandong Province, China (S190002030001).
[fig] FIGURE 1 |: Experimental workflow in the study of Polygonatum spp. in the treatment of diabetes. Frontiers in Pharmacology | www.frontiersin.org February 2022 | Volume 13 | Article 758501 [/fig]
[table] TABLE 1 |: Comparison of the major chemical constituents of three Polygonatum spp. [/table]
[table] TABLE 2 |: Flavonoids isolated from three Polygonatum spp. [/table]
[table] TABLE 3 |: Antidiabetic properties of three Polygonatum spp. in cells and animal models. [/table]
[table] TABLE 4 |: Chinese patent medicines containing Polygonatum with hypoglycemic effect in human studies. [/table]
[table] TABLE 5 |: Chinese patent medicine prescription containing Polygonatum with hypoglycemic effect (data from db.yaozh.com). [/table]
|
MicroRNA regulation of CYP 1A2, CYP3A4 and CYP2E1 expression in acetaminophen toxicity
MicroRNAs (miRNAs) that regulate the cytochrome P-450 isoforms involved in acetaminophen (APAP) toxicity were examined in HepaRG cells treated with APAP (20 mM).In-vitro studies found that APAP protein adducts were increased at 1 h, followed by ALT increases at 12 and 24 h. CYP1A2, CYP3A4 and CYP2E1 mRNA levels were decreased, while miRNAs were increased for miR-122-5p, miR-378a-5p, miR-27b-3p at 6 h and miR-125b-5p at 12 h and miR-27b-3p at 24 h. Putative miRNA binding sites on the 3′UTRs of the CYPs were identified in-silico. Overexpression of miR-122-5p and miR-378a-5p in cells suppressed protein expression of CYP1A2, CYP3A4 and CYP2E1. Luciferase reporter assays confirmed the interaction between miR-122 and the 3′UTR of the CYP1A2 and CYP3A4. Thus, the in-vitro experiments showed that miR-122-5p and miR-378a-5p upregulation were associated with translational repression of CYPs. Serum samples of children with APAP overdose had significant elevation of miR-122-5p, miR-378a-5p, miR-125b-5p and miR-27b-3p, compared to healthy controls and receiver operator curves of the miRNAs had AUCs of 91 to 100%. Collectively, the data suggest that miRNA elevations in APAP toxicity represent a regulatory response to modify CYP1A2, CYP3A4 and CYP2E1 translation due to cellular stress and injury.MicroRNAs (miRNAs, miRs) are 18-25-nucleotides short non-coding RNAs which regulate gene expression 1-3 through mRNA degradation or the inhibition of protein translation 4,5 . Previous reports have described a role for microRNAs (miRNAs) in the regulation of drug metabolizing enzymes (DMEs) 6-8 . Characterizing miRNA profiles in relationship to cytochrome P450 (CYP) expression for DMEs relevant to APAP toxicity is important to further understanding mechanisms of toxicity and may have relevance for understanding individual variability in susceptibility to drug toxicity, as metabolism is an early, initiating event in the development of APAP hepatotoxicity. A number of laboratories using murine models of APAP toxicity have reported the upregulation of liver enriched miR-122 in association with hepatotoxicity 9-12 and several researchers have reported elevation of miR-122 in clinical samples of subjects with acetaminophen toxicity[13][14][15][16][17][18]. Despite this literature, the relationship of miRNAs to DMEs relevant to APAP toxicity (i.e., CYP1A2, CYP3A4 and CYP2E1) 19-23 is not well understood. In the present study, the relationship of CYP-regulating miRNAs to CYP1A2, CYP3A4 and CYP2E1 gene expression in APAP exposed HepaRG cells was evaluated and compared to indicators of APAP toxicity and metabolism 24,25 . HepaRG cells have high expression and activity levels for DMEs 26-28 and have been previously used in studies of APAP toxicity 29-31 . In addition, levels of miRNAs were examined in clinical samples obtained from APAP overdose subjects and compared to common clinical indicators of liver injury and APAP toxicity.ResultsTime course and dose response; ALT, APAP protein adducts, and CYP-binding miRNAs in HepaRG cells. Dose and response studies were conducted in cells to evaluate the temporal relationships among toxicity (ALT elevation), APAP protein adducts (an indicator of oxidative metabolism) and miRNA profiles. As shown inFig. 1A, APAP protein adducts were increased at 12 and 24 h in media from cells exposed to APAP 5 mM (12.
IU/L + 0.44). Consistent with previous data [bib_ref] HepaRG cells: a human model to study mechanisms of APAP hepatotoxicity, Mcgill [/bib_ref] , a dose response pattern was observed for APAP protein adducts in the APAP 5 and 20 mM exposed cells [fig_ref] Figure 1: Figure 1 [/fig_ref]. [fig_ref] Figure 1: Figure 1 [/fig_ref] ,B demonstrates that ALT levels increased over time in both dose groups at 24 h (*p < 0.05).
Cells treated with 5 or 20 mM APAP released miR-122-5p, miR-378a-5p, miR-27b-3p, and miR-125b-5p into media as a function of both APAP dose and time. Consistent with the ALT and adduct profiles described above, the most marked changes in miRNA profiles were observed in the APAP 20 mM cells. [fig_ref] Figure 1: Figure 1 [/fig_ref] shows down-regulation of all miRNAs at 1 h in the APAP 5 mM cells, followed by elevation of miR-122-5p at 6 h, miR-378a-5p at 6 h, and miR-27b-3p at 24 h. In contrast, the APAP 20 mM cells [fig_ref] Figure 1: Figure 1 [/fig_ref] had a significant elevation of miR-122-5p at 6 h, miR-378a-5p at 6 h, miR-125b-5p at 12 h, and miR-27b-3p at 24 h. Thus, the data demonstrate Time and dose-dependent changes in HepaRG cells with APAP treatment for ALT, APAP protein adducts, miRNA (miR-122-5p, miR-378a-5p, miR-27b-3p, and miR-125b-5p and mRNA levels (CYP2E1, CYP3A4, and CYP1A2) levels. Effect of 5 and 20 mM APAP treatment of HepaRG cells at 1, 6, 12 and 24 h time points. (A and B) Show levels of ALT and APAP protein adducts in cell medium; *p < 0.05 compared to controls for APAP 5 mM and 10 mM. Parenthesis around the asterisk denotes significance of ALT for 5 mM APAP treatment at 12 and 24 h. (C and D) Depict miRNA expression in cell medium. Data were normalized for HepaRG cell culture medium with Let-7d and spiked-in C. elegans miR-39; *p < 0.05 compared to controls. (E and F) Describe CYP1E2, CYP1A2 and CYP3A4 mRNA levels determined by qRT-PCR. TaqMan assays were duplexed with GAPDH to normalize the mRNA expression. Parenthesis around the asterisk differentiate 24 h CYP3A4 from CYP1A2; *p < 0.05 compared to controls. "h" Denotes hour. Error bars represent Standard error of the mean.
that HepaRG cells generate dose-response and temporal data for toxicity and oxidative drug metabolism endpoints known to be important in models of APAP toxicity. Furthermore, miRNA profiles were distinct between the 5 ("low dose") and 20 mM ("high dose") exposed cells.
Effect of APAP treatment on CYP2E1, CYP1A2 and CYP3A4 mRNA expression. To understand temporal relationships between miRNA profiles and gene expression for DME important in APAP toxicity, gene expression of CYP2E1, CYP1A2 and CYP3A4 was examined in cells exposed to APAP 5 mM or 20 mM. As shown in [fig_ref] Figure 1: Figure 1 [/fig_ref] , CYP2E1 mRNA expression in cells treated with 5 mM APAP was down regulated by 2.2 fold (p < 0.05) at 24 h. Transient down-regulation of CYP1A2 mRNA expression was observed with 5 mM APAP at 1 h (p < 0.05; [fig_ref] Figure 1: Figure 1 [/fig_ref] , while increased expression was apparent at 24 h. CYP3A4 mRNA expression was increased at all-time points for the APAP 5 mM [fig_ref] Figure 1: Figure 1 [/fig_ref] ,while the APAP 20 mM cells had down regulation of CYP2E1 at 6, 12 and 24 h (*p < 0.05, [fig_ref] Figure 1: Figure 1 [/fig_ref]. In contrast, the 20 mM APAP dose produced marked down regulation of CYP1A2 at all-time points, with a 26 fold decrease by 12 h (*p < 0.05, [fig_ref] Figure 1: Figure 1 [/fig_ref]. Moreover, cells exposed to 20 mM APAP had decreased CYP3A4 mRNA expression at 12 and 24 h (*p < 0.05). miRNA target sites located on promoters, 5′UTR, CDS and 3′UTR of CYP genes. The miR-Walk2.0 in-silico meta-analysis showed that hsa-miR-122-5p, hsa-miR-378a-5p, hsa-miR-27b-3p and hsa-miR-125b-5p have target sites within the 3′UTR region of CYP1A1, CYP1A2, CYP3A4 and CYP2E1 genes. Moreover, a total of 47 interactions were identified between the CYP1A1, CYP1A2, CYP3A4 and CYP2E1 genes and the 4 miRNAs . Among these interactions, 15, 5, 16 and 11 were detected on the promoter, 5′ UTR, CDS and 3′ UTR regions, respectively. The mRNA-miRNA interactions suggest preferential control at transcriptional, as well as translational levels. The mRNA 5′ and 3′UTR regions were predicted to harbor several miRNA binding sites. Collectively, the in-silico analysis suggests cooperativity and multiplicity of miRNA binding sites on gene regulation during APAP toxicity. miR-122 targets 3′ UTR of CYP1A2 and CYP3A4. To further examine the predicted relationships between miR-122 and CYP1A2 and CYP3A4 , HEK293T cells were transfected with 3′UTR luciferase plasmids with miR-122 mimic and inhibitor. Luciferase activity in HEK293T cells expressing the CYP1A2-Luc-3′UTR was inhibited by overexpressing the miR-122 mimic [fig_ref] Figure 2: Luciferase assay confirming direct suppression of CYP1A2 and CYP3A4 transcription through 3′UTR... [/fig_ref] and reporter activity returned to control level (determined by transfection of pEZX-MT06) following transfection with the miR-122 inhibitor. However, overexpression of the miR-122 mimic marginally decreased CYP3A4 3′UTR luciferase activity, while addition of the miRNA-122 inhibitor partially reversed luciferase activity to the level of control plasmid [fig_ref] Figure 2: Luciferase assay confirming direct suppression of CYP1A2 and CYP3A4 transcription through 3′UTR... [/fig_ref]. 3′UTR reporter assays demonstrated that the miR-122 mimic reduced luciferase activity, confirming the presence of miR-122-binding sites within the 3′UTR of CYP1A2 and CYP3A4 genes. Mohri and colleagues 32 previously confirmed the human CYP2E1 3′UTR target for miR-378a.
Gain-of-function and loss-of function studies: impact of miRNAs on protein expression and toxicity. In further studies, the effect of the miR-122-5p and miR-378 mimics and inhibitors on protein expression was examined. CYP1A2 protein expression was reduced by 90% at 12 h in APAP 20 mM cells, compared to controls [fig_ref] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect... [/fig_ref]. Transfection with the miR-122-5p mimic attenuated the effect of APAP on CYP1A2 protein expression; CYP1A2 protein expression was reduced by 40% compared to controls. In contrast, transfection of the miR-122-5p inhibitor restored CYP1A2 protein expression to that of controls [fig_ref] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect... [/fig_ref].
APAP reduced CYP3A4 protein levels by 83% [fig_ref] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect... [/fig_ref] , whereas transfection of cells with the miR-122 mimic reduced CYP3A4 protein levels by only 28%. In addition, transfection of cells with the miR-122 inhibitor restored CYP3A4 expression to that of controls [fig_ref] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect... [/fig_ref]. [fig_ref] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect... [/fig_ref] demonstrates that APAP reduced CYP2E1 protein expression by only 26%, whereas cells transfected with the miR-378 mimic had reduced CYP2E1 protein expression by 33%. Transfection with the miR-378 inhibitor restored CYP2E1 protein to that of controls.
Non-APAP exposed cells transfected with mimics and inhibitors of miR-122 and miR-378a had no changes in protein expression for CYP1A2, CYP3A4 and CYP2E1 [Supplementary Figure S1A-F and [fig_ref] Figure 2: Luciferase assay confirming direct suppression of CYP1A2 and CYP3A4 transcription through 3′UTR... [/fig_ref] ]. Thus, collectively the data show that gain of function of miR-122 repressed CYP1A2 and CYP3A4, while gain of function of miR-378a repressed CYP2E1 protein levels. [fig_ref] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect... [/fig_ref] illustrates the effect of mimic and inhibitor transfections on trends in ALT levels. APAP treated cells had increased ALT levels compared to controls. Both the APAP/miR-122 mimic treated cells and the APAP/miR-378 mimic treated cells had higher ALT levels than the APAP only cells. Cells treated with the miR-122 inhibitor or the miR-378 inhibitor had ALT levels comparable to the APAP only cells.
Elevated levels of APAP protein adducts, ALT, and CYP-binding miRNA in APAP overdose subjects. To examine the clinical relevance of the in-vitro data, serum samples from children with APAP overdose were analyzed for APAP protein adducts, ALT, and miRNA. Demographic, toxicity, and APAP protein adduct data are provided in . Consistent with previous data [bib_ref] Pharmacokinetics of APAP-protein adducts in adults with APAP overdose and acute liver..., James [/bib_ref] [bib_ref] Targeted liquid chromatography-mass spectrometry analysis of serum acylcarnitines in APAP toxicity in..., Bhattacharyya [/bib_ref] [bib_ref] Comparison of Bile Acids and Acetaminophen Protein Adducts in Children and Adolescents..., James [/bib_ref] , ALT levels and APAP protein adducts were markedly elevated in the serum samples of APAP overdose subjects compared to healthy controls.
Box plots shown in [fig_ref] Figure 4: Box plots and ROC curves for miR-122-5p, miR-378a-5p, miR-125b-5p and miR-27b-3p [/fig_ref] demonstrate elevations of miR-122-5p, miR-378a-5p, miR-125b-5p and miR-27b-3p in APAP overdose subjects compared to the control group (*p < 0.05). Serum samples from the APAP . Demographic and clinical parameters of APAP overdose and control subjects (*p < 0.001).
group showed a 140 fold elevation of miR-122-5p compared to the control group, whereas elevations of miR-378a-5p, miR-125b-5p, and miR-27b-3p were 12.6, 8.4, and 4.4 fold higher than the control group. Receiver Operating Curve (ROC) analysis demonstrated the performance for the miRNAs to discriminate between the APAP overdose and control groups [fig_ref] Figure 4: Box plots and ROC curves for miR-122-5p, miR-378a-5p, miR-125b-5p and miR-27b-3p [/fig_ref]. The ROC analysis had an Area Under the Curve (AUC) of 1 for miR-122-5p, a predicted regulator of CYP2E1, CYP1A2 and CYP3A4, while AUCs for miR-378a-5p, miR-125b-5p, and miR-27b-3p were 96.7% (91.0-100%), 91.7% (80.2-100%), and 91.1% (79.7-100%), respectively. [fig_ref] Figure 5: Correlation heatmap showing correlations between ALT, APAP protein adducts and the miRNAs [/fig_ref] provides summary data for correlation coefficients (blue circles) for individual miRNA's, ALT, APAP protein adducts, and other miRNAs. R values > 0.75 were observed for ALT and miR-122-5p, miR-378a-5p and miR-125b-5p. R values of 0.58-0.68 were observed for APAP protein adducts and miR-122-5p, miR-27b-3p, miR-125b-5p and miR-378a-5p (p < 0.05). Consistent with previous data of Yang and colleagues 17 , a high correlation (R = 0.67) was observed for miR-122-5p and APAP protein adducts 17 .
# Discussion
The present study used HepaRG cells and clinical samples to examine the expression of miRNAs identified to regulate CYPs known to be involved in the metabolism and subsequent development of APAP toxicity [bib_ref] Role of CYP2E1 in the hepatotoxicity of APAP, Lee [/bib_ref] [bib_ref] Contribution of CYP2E1 and CYP3A to APAP reactive metabolite formation, Manyike [/bib_ref] [bib_ref] The effect of omeprazole pretreatment on APAP metabolism in rapid and slow..., Sarich [/bib_ref]. Initial in-silico analysis determined that miR-122-5p, miR-378a-5p, miR-27b-3p and miR-125b-5p target CYP1A2, CYP3A4 and CYP2E1 genes. Direct suppression of CYP1A2 and CYP3A4 through 3′UTR binding of miR-122 was confirmed by luciferase assays. Gain-of-function and loss-of-function studies demonstrated that miR-122 and miR-378a downregulated CYP 1A2, CYP3A4, and CYP2E1 expression in cells at the translational level. Treatment of cells with two concentrations of APAP ("low" and "high" dose) generated time course toxicity and metabolism data consistent with previous data [bib_ref] HepaRG cells: a human model to study mechanisms of APAP hepatotoxicity, Mcgill [/bib_ref] [bib_ref] Time course of acetaminophen-protein adducts and acetaminophen metabolites in circulation of overdose..., Xie [/bib_ref] , and indicated differential down regulation of CYP2E1 and CYP1A2 mRNA as a function of APAP dose. In contrast, upregulation of CYP3A4 expression occurred with APAP 5 mM, replicating findings from a recent report [bib_ref] Toxicogenomics-based prediction of acetaminophen-induced liver injury using human hepatic cell systems, Rodrigues [/bib_ref] , while APAP 20 mM led to down regulation of CYP3A4 expression. APAP treatment resulted in decreased expression of miRNA (miR-122-5p, miR-125b-5p, miR-378a-5p and miR-27b-3p) in cells and increased expression of miRNA's in media [fig_ref] Figure 1: Figure 1 [/fig_ref]. These data are consistent with a report by Wang and colleagues 9 who noted the down-regulation of hepatic miRNAs and subsequent elevation in miRNAs in the plasma of mice treated with a toxic dose of APAP [bib_ref] Circulating microRNAs, potential biomarkers for drug-induced liver injury, Wang [/bib_ref].
Elevations of CYP-regulating miRNAs correlated with elevations of APAP protein adducts in APAP treated cells and in patients with APAP toxicity. Initial studies of miRNA expression in APAP toxicity reported elevations of the liver-specific miR-122-5p and its correlation with ALT elevation 9,10 ; miR-122 may also represent a predictive biomarker of APAP liver injury [bib_ref] Evaluation of miR-122 and other biomarkers in distinct acute liver injury in..., Starckx [/bib_ref] and resolving liver injury [bib_ref] Circulating microRNAs as potential markers of human drug-induced liver injury, Starkey Lewis [/bib_ref] [bib_ref] Mechanistic biomarkers provide early and sensitive detection of APAP-induced acute liver injury..., Antoine [/bib_ref]. Yamaura et al. used a rat model of APAP toxicity and described the elevation of miR-122, as well as miR-192, miR-685, miR-193 and miR-29c 12 . High throughput omic analysis using human APAP overdose samples suggested that miR-122-5p 15-18 , miR-27b-3p [bib_ref] Circulating microRNA profiles in human patients with APAP hepatotoxicity or ischemic hepatitis, Ward [/bib_ref] [bib_ref] Application of High-Throughput Sequencing to Circulating microRNAs Reveals Novel Biomarkers for Druginduced..., Krauskopf [/bib_ref] and miR-125b-5p 15,17 were sensitive and noninvasive biomarkers for APAP toxicity. In the present study, we observed a strong association between miR-122-5p, miR-378a-5p, miR-125b-5p, and miR-27b-3p and APAP liver injury (defined by ALT elevation; [fig_ref] Figure 4: Box plots and ROC curves for miR-122-5p, miR-378a-5p, miR-125b-5p and miR-27b-3p [/fig_ref].
Prediction algorithms have identified a number of miRNAs that may target CYPs 3′UTR regions [bib_ref] Novel advances in cytochrome P450 research, Singh [/bib_ref] [bib_ref] Genetic variation in the 3′-UTR of CYP1A2, CYP2B6, CYP2D6, CYP3A4, NR1I2, and..., Swart [/bib_ref] [bib_ref] In silico identification of microRNAs predicted to regulate the drug metabolizing cytochrome..., Ramamoorthy [/bib_ref] in drug metabolism. To our knowledge, the present study is the first to use this approach in APAP toxicity. Of the potential miRNA binding sites identified within the gene regions of the CYPs (CYP2E1, CYP3A4 and CYP1A2) (Table S1) the greatest number of interactions were found for the promoter, CDS and 3′UTR. Taken together in APAP exposed cells down regulation of CYP1A2, CYP3A4 and CYP2E1 [fig_ref] Figure 1: Figure 1 [/fig_ref] are mediated by interactions between miRNA binding sites and the CYPs gene regions (promoter, CDS and 3′ UTR). Our findings are in accordance with the previous studies reporting cooperative miRNA-target interactions for 3′UTR and the CDS or 5′UTR region, resulting in strong gene downregulation [bib_ref] The Impact of miRNA Target Sites in Coding Sequences and in 3′UTRs, Fang [/bib_ref] [bib_ref] New class of microRNA targets containing simultaneous 5′-UTR and 3′-UTR interaction sites, Lee [/bib_ref] [bib_ref] Analysis of CDS-located miRNA target sites suggests that they can effectively inhibit..., Hausser [/bib_ref] [bib_ref] The code within the code: microRNAs target coding regions, Forman [/bib_ref]. The analysis also identified region-specific miRNA binding sites (promoter and CDS) for CYP members, consistent with a previous report that suggested that some miRNAs preferentially base-pair to the CDS region, suggesting a key role in rapid inhibition of translation 44 . These interactions, or co-targets, have previously been demonstrated in a number of studies 42-45 , but not in APAP toxicity. As each miRNA can target roughly 200 transcripts [bib_ref] Combinatorial microRNA target predictions, Krek [/bib_ref] , the number of putative interactions may be very large to have an effect on gene regulation in APAP toxicity.
Target prediction and in-vitro functional studies using 3′UTR reporter assays showed that CYP1A2 and CYP3A4 were a direct target of miR-122. The present study also corroborates the work of Mohri and colleagues [bib_ref] Human CYP2E1 is regulated by miR-378, Mohri [/bib_ref] , who showed that 3′ UTR of CYP2E1 was regulated by miR-378 in HEK293 cells, mainly via translational repression [bib_ref] Human CYP2E1 is regulated by miR-378, Mohri [/bib_ref]. The data presented herein are consistent with the hypothesis that miR-122-5p and miR-378a-5p upregulation in APAP exposed cells may represent a protective mechanism to lower CYPs expression [fig_ref] Figure 6: Postulated role of miRNA in APAP toxicity [/fig_ref] and the formation of the reactive metabolite, N-acetyl-para-quinone imine (NAPQI), and resulting generation of APAP protein adducts. The data also suggest that the predominant effect of miR-122 and miR-378a on CYP1A2, CYP3A4 and CYP2E1 is exerted at the translational level [fig_ref] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect... [/fig_ref] and is consistent with recent data from primary cells and tissues from miRNA mutant miceshowing that 48% of target genes are regulated by translational repression. Thus, the data provide new insights into the function of miRNA in the regulation of CYP expression in APAP toxicity and suggests that miRNAs may be a potential target for novel therapeutic approaches in APAP toxicity.
One limitation of the present study its reliance on HepaRG cells which do not fully capture the inherent biologic variability and microenvironment of the human liver. It is generally regarded that APAP toxicity is a direct necrotic event, although signaling cascades invoked by cytokine release from inflammatory cells and other pathways may modulate the toxicity. Since the liver comprises a variety of cell types, further well characterized in-vivo studies are necessary to fully understand the miRNA regulated signaling pathways in APAP toxicity.
To summarize, serum miR-122-5p was highly increased in children with APAP overdose, similar to data published in adults. Three other miRNAs (miR-378a-5p, miR-125b-5p, and miR-27b-3p) were also increased, but the relative increase was lower than that of miR-122-5p. Strong correlations were noted for ALT and miR-122-5p, miR-378a-5p and miR-125b-5p (R = 0.76-89, p < 0.05), while moderate correlations were noted for APAP protein adducts and the miRNAs (R = 0.58-0.67, p < 0.05). These elevations may represent cell-to-cell communication in an attempt to minimize toxic events in cells or stimulate hepatocyte repair responses. A better understanding of the function of miRNAs in APAP-induced hepatotoxicity may have relevance for the development of new treatments for APAP toxicity and/or understanding individual susceptibility to drug toxicity. The results provide new insights into the protective role of miR-122 and miR-378a in suppressing the expression of DMEs.
# Methods
Chemicals and materials. APAP, nuclease free water and ethanol (Molecular grade) were purchased from Sigma-Aldrich (St. Louis, MO). miRNA and RNA isolation, cDNA synthesis and pre-amplification reagents were purchased from Qiagen (Valencia, CA). Collagen coated 24 well plates and phosphate buffered saline were obtained from ThermoFisher Scientific (Carlsbad, CA). The miR-122 mimic, miR-122 inhibitor, miR-378a mimic, miR-378a inhibitor, AllStars Negative control siRNA and Hs Cell Death Control siRNA were purchased from Qiagen (Valencia, CA). HepaRG cells. HepaRG cells (ThermoFisher Scientific,Carlsbad, CA) were seeded in base medium per the manufacturer's protocol. Cell viability was determined as per Trypan blue method and 330,000 cells were seeded per well in collagen-coated 24 well plates. As per protocol, from day 2-9 base medium was supplemented with Tox Medium Supplement (ThermoFisher Scientific, Carlsbad, CA). APAP treatment, mRNA/miRNA isolation, cDNA synthesis and quantitative real-time PCR (qRT-PCR). Cells were incubated with 5 or 20 mM APAP 26,29-31 on day 7 for 1, 6, 12, or 24 h time points.
Isolation of RNA from cells was performed using the miRNeasy Mini Kit (Qiagen, Valencia, CA). Total RNA (500 ng) was used for cDNA synthesis using RT2 First Strand Kit (Qiagen, Valencia, CA). For gene expression analysis, quantitative real-time polymerase chain reactions (qRT-PCRs) were performed using TaqMan Fast Master Mix (ThermoFisher Scientific, Carlsbad, CA) for Cyp1A2 (Hs00167927_m1), Cyp2E1 (Hs00559368_m1) and Cyp3A4 (Hs00604506_m1) and were duplexed with GAPDH (Hs02758991_g1). RNeasy MinElute Cleanup Kit was used for miRNA isolation from cells (Qiagen, Valencia, CA) and cDNA was synthesized with the miScript II RT kit (Qiagen, Valencia, CA). For media samples, the total RNA, including miRNA was isolated from 100 µl media spiked in with C. elegans miR-39 (Ce_miR-39_1) using miRNeasy Serum/Plasma kit (Qiagen, Valencia, CA). cDNA synthesis was performed using the miScript II RT kit (Qiagen, Valencia, CA) with miScript HiSpec buffer. cDNA from cells and media was pre-amplified with miScript PreAMP PCR Kit (Qiagen, Valencia, CA). For miRNA expression analysis, cDNAs from cells and media were used in qRT-PCR reaction for miScript primer assays: hsa-miR-122-5p (MS00003416), hsa-miR-378a-5p (MS00009646), hsa-miR-27b-3p (MS00031668), and hsa-miR-125b-5p (MS00006629), Hs_let_7d_1 (MS00003136), Ce_miR-39_1 (MS00019789) and Hs_SNORD95 (MS00033726) (Qiagen, Valencia, CA). Expression levels of miRNAs and CYP mRNA were measured in triplicate on a QuantStudio 6 Flex Real-Time PCR System (ThermoFisher Scientific, Carlsbad, CA). Toxicity assays. Hepatic necrosis was assessed by measurement of alanine aminotransferase (ALT) in media using an Ace Alera Serum Chemistry Analyzer (Alfa Wassermann).
Human samples. Peripheral blood samples were collected as part of a multicenter investigation of APAP toxicity in children of ages 2-18 years, approved by University of Arkansas for Medical Sciences (UAMS) institutional review board (IRB) [bib_ref] Targeted liquid chromatography-mass spectrometry analysis of serum acylcarnitines in APAP toxicity in..., Bhattacharyya [/bib_ref]. Subjects were categorized as controls (n = 15 healthy children with no use of APAP in the preceding 2 weeks) or APAP overdose (n = 12; children that required hospitalization for treatment of APAP overdose) and informed consent was obtained from all participants per UAMS IRB policy guidelines and all experiments were performed in accordance with the relevant guidelines and regulations. A single blood sample was collected from control subjects, whereas daily samples were collected from the APAP overdose subjects. Blood samples were centrifuged within 30 minutes of collection and the serum was stored at −80 °C. Demographic information available for study subjects included subject age, gender, weight, height, body surface area, past medical history, reason for hospitalization, relevant history concerning recent APAP dosing and concomitant medications. Clinical laboratory results included APAP concentration and ALT determinations.
Quantitation of APAP protein adducts. APAP protein adducts were analyzed using a method previously published by our laboratory [bib_ref] Determination of APAP-protein adducts in mouse liver and serum and human serum..., Muldrew [/bib_ref] [bib_ref] Unrecognized APAP toxicity as a cause of indeterminate acute liver failure, Khandelwal [/bib_ref] [bib_ref] An Immunoassay to Rapidly Measure Acetaminophen Protein Adducts Accurately Identifies Patients With..., Roberts [/bib_ref]. Briefly, 100 µl of serum or cell supernatants was gel filtered, hydrolyzed with protease, precipitated, and injected onto an HPLC system. APAP cysteine was resolved on a 150 mm C18 column and detected using an ESA CoulArray electrochemical detector. Concentrations of adducts were determined relative to a standard curve of authentic APAP cysteine and reported as nmoles a-cys/mg protein (in-vitro data) and nmol/L a-cys (human serum samples).
Serum miRNA isolation and cDNA synthesis. From 100 µl serum sample total RNA including miRNA was isolated, reverse transcribed and pre-amplified as mentioned above in section describing isolation from media. miRNA expression analysis. The pre-amplified cDNA from cells, media and human sera were used to run qRT-PCR using miScript SYBR Green PCR Kit and miScript Primer Assays for miRNAs (hsa-miR-122-5p, hsa-miR-378a-5p, hsa-miR-27b-3p, and hsa-miR-125b-5p) (Qiagen, Valencia, CA). All samples were run in triplicate on QuantStudio 6 Flex Real-Time PCR System (ThermoFisher Scientific, Carlsbad, CA). For each primer assay, negative controls were run with water and cDNA samples. For no template controls (NTC) a master mix with no cDNA was used. Data normalization was performed where appropriate with exogenous control (Ce-miR-39) and endogenous controls (let-7d, RNU6 and/or SNORD95). Relative quantitation was calculated using the 2 −ΔΔCt method. microRNA target predictions. miRWalk2.0 49,50 is a comprehensive database (http://zmf.umm.
uni-heidelberg.de/apps/zmf/mirwalk2/generetsys-self.html) that provides in-silico meta-analysis of miRNA-target interactions by combining data sets from 13 different prediction algorithms. The complete sequence (promoter, CDS, 5′ and 3′UTR regions) of CYP1A1, CYP1A2, CYP3A4 and CYP2E1 genes was screened to identify binding sites of 4 miRNAs (hsa-miR-122-5p, hsa-miR-125b-5p, hsa-miR-378a-5p, and hsa-miR-27b-3p), previously reported in APAP toxicity studies 3′UTR luciferase reporter assays. To confirm the miRNA-mRNA interactions, human CYP1A2 (HmiT003774) and CYP3A4 (HmiT055335) 3′ UTR target clones in pEZX-MT06 were purchased (GeneCopoeia, Rockville, MD). HEK293T cells were seeded in 96-well plates (3 × 10 4 cells/well) for 24 h. HEK293T cells were co-transfected in OPTI MEM media (ThermoFisher Scientific, Carlsbad, CA) with 200 ng of 3′ UTR luciferase reporter plasmids and miR-122 mimic, miR-122 inhibitor or negative control plasmid Scientific RepoRts | 7: 12331 | DOI:10.1038/s41598-017-11811-y pEZX-MT06 (GeneCopoeia, Rockville, MD) with Attractene transfection reagent (Qiagen, Valencia, CA). Firefly and renilla luciferase activities were measured 24 h after transfection using the Luc-Pair Duo-Luciferase Assay Kit 2.0 (GeneCopoeia, Rockville, MD). Data represent the ratio of firefly luciferase activity to that of Renilla.
## Transfection of microrna mimics and inhibitors. mimics and inhibitors of mir-122 and mir-378
were used to overexpress or knockdown miRNA in HepaRG cells. Cells were grown in 24 well plate (see cell culture) and transfected with 5 nM mimic, 50 nM or 100 nM inhibitor using RNAiMAX Transfection Reagent (ThermoFisher Scientific, Carlsbad, CA). At 24 h, the transfection medium was replaced with HepaRG cells Tox Working Medium. At 72 h, cells were cultured in glutathione free media. APAP (20 mM) was added for 12 h and cells were harvested at 96 h post transfection for protein analysis. Transfection efficiency of 65-89% was observed with 5-50 nM concentration of AllStars Hs Cell Death Control siRNA. Preliminary studies found no interference in protein expression with negative control AllStars siRNA.
Protein isolation and Western blot analysis. After transfections, cells were lysed with ice-cold RIPA lysis buffer (ThermoFisher Scientific, Carlsbad, CA) containing protease inhibitors (Pierce Protease Inhibitor Mini Tablets, Rockford, IL and Mammalian Protease Arrest, GBiosciences, St. Louis, MO) and centrifuged at 12,500 rpm for 10 min at 4 °C. Cell supernatants were used for immunoblots. Proteins were quantified using Pierce BCA Protein Assay Kit (ThermoFisher Scientific, Carlsbad, CA) and 20 ug of protein was loaded on to precast polyacrylamide Bis-tris 10% gels (GenScript, Piscataway, NJ). The membranes were blocked with 5% skim milk and incubated with primary antibodies against CYP2E1, CYP1A2 and CYP3A4 (Abcam, Cambridge, MA) and β-actin (Sigma, St. Louis, MO). Next membranes were incubated with HRP-conjugated secondary antibodies to goat anti-rabbit IgG H&L (Abcam, Cambridge, MA). The protein bands were visualized using the SuperSignal West Pico Chemiluminescent Substrate (ThermoFisher Scientific, Carlsbad, CA) and densitometry was performed with the ImageJ software program (NIH, Bethesda, MD).
# Statistical analysis.
Statistical analyses of cellular data used one-way ANOVA with post hoc Dunnett's test to determine differences between treatment groups and the control. The unpaired Student's t-test was used for comparison between two cell groups. The non-parametric Mann-Whitney U test was used for pairwise comparisons between clinical groups for toxicity (ALT), metabolism (APAP protein adducts) and miRNA data. Spearman rank correlation analysis was used to assess the pairwise relationships among miRNAs, ALT and APAP protein adducts; data are reported as medians, minimums, and maximums. A p value < 0.05 was considered significant for all analyses. Receiver operating curve (ROC) analysis was used to assess the performance of each miRNA as a binary classifier of toxicity outcomes.
# Disclosure.
Dr. James has a patent application pending for the measurement of APAP protein adducts in human blood samples.
[fig] Figure 1: Figure 1. Time and dose-dependent changes in HepaRG cells with APAP treatment for ALT, APAP protein adducts, miRNA (miR-122-5p, miR-378a-5p, miR-27b-3p, and miR-125b-5p and mRNA levels (CYP2E1, CYP3A4, and CYP1A2) levels. Effect of 5 and 20 mM APAP treatment of HepaRG cells at 1, 6, 12 and 24 h time points. (A and B) Show levels of ALT and APAP protein adducts in cell medium; *p < 0.05 compared to controls for APAP 5 mM and 10 mM. Parenthesis around the asterisk denotes significance of ALT for 5 mM APAP treatment at 12 and 24 h. (C and D) Depict miRNA expression in cell medium. Data were normalized for HepaRG cell culture medium with Let-7d and spiked-in C. elegans miR-39; *p < 0.05 compared to controls. (E and F) Describe CYP1E2, CYP1A2 and CYP3A4 mRNA levels determined by qRT-PCR. TaqMan assays were duplexed with GAPDH to normalize the mRNA expression. Parenthesis around the asterisk differentiate 24 h CYP3A4 from CYP1A2; *p < 0.05 compared to controls. "h" Denotes hour. Error bars represent Standard error of the mean. [/fig]
[fig] Figure 2: Luciferase assay confirming direct suppression of CYP1A2 and CYP3A4 transcription through 3′UTR binding of miR-122. Effects of miR-122 mimic or inhibitor on luciferase activity in HEK293 cells cotransfected with a firefly luciferase reporter vector containing the CYP1A2 3′UTR-Luc (Fig. 2A)and CYP3A4 3′ UTR-Luc(Fig. 2B). Control plasmid pEZX-MT06 with no 3′UTR, miR-122 mimic (1 nM and 5 nM), miR-122 inhibitor (100 nM). A one-way analysis of variance (ANOVA) with post hoc Dunnett's test, *p < 0.0196 and **p < 0.006. Values are mean of 2 independent experiments. RLU = relative luciferase units. [/fig]
[fig] Figure 3: Transfection of miRNA mimic and inhibitor into HepaRG cells and its effect on the protein levels in APAP treated cells (N = 3). Western blot analysis of miR-122 mimic and inhibitor transfected cells treated with APAP for 12 h. (A) CYP1A2 and (B) CYP3A4: Ctrl = Control untreated; APAP = 20 mM; miR-122 Mimic 5 nM; miR-122 Inhibitor 100 nM. Uncropped image of blots for CYP1A2 and CYP3A4 are shown in Figure S3. Western blot analysis of miR-378 mimic and inhibitor transfected cells (C) CYP2E1: Ctrl = Control untreated; APAP = 20 mM; miR-378a Mimic 5 nM; miR-378a Inhibitor 50 nM. Uncropped image of blot for CYP2E1 shown inFigure S4. Data are normalized to β-Actin (*p < 0.05, **p < 0.01) and error bars represent standard error of the mean (SEM). (D) ALT levels in miR-122 and miR-378a mimic and inhibitor transfected cells treated with 20 mM APAP for 12 h. "h" denotes hour. [/fig]
[fig] Figure 4: Box plots and ROC curves for miR-122-5p, miR-378a-5p, miR-125b-5p and miR-27b-3p. (A) Depicts box plots of circulating miRNAs in serum from healthy and APAP overdose samples. Data was normalized for miRNA expression with let-7d and spiked C. elegans miR-39. Data were presented as a box plot of the median and range of log transformed relative expression level. The top and bottom of the box represent the 75 th and 25 th percentile. (B) Shows receiver operating characteristics (ROC) curve analysis, indicating diagnostic value of each miRNA in discriminating APAP toxicity. [/fig]
[fig] Figure 5: Correlation heatmap showing correlations between ALT, APAP protein adducts and the miRNAs. The correlation coefficients are presented in the circles. On the right is shown a color key for the strength of correlations. Significance level p < 0.05. [/fig]
[fig] Figure 6: Postulated role of miRNA in APAP toxicity. Schematic illustration of miRNA regulating CYP 450 in APAP toxicity. miRNAs (miR-122, miR-378a, miR-27b and miR-125b) were detected in cell culture medium and serum from APAP over dose subjects. CYP1A2, CYP3A4 and CYP2E1 were detected by qPCR in HepaRG cells and mRNA expression levels match the color coated keys inFig. 1E,F. Dashed lines show miRNA predicted to target 3′ UTR of CYPs. The bold lines show validation with gain of function experiments for miR-122 and miR-378a mimics in a cellular model. Scientific RepoRts | 7: 12331 | DOI:10.1038/s41598-017-11811-y [/fig]
|
N‐glycan signatures identified in tumor interstitial fluid and serum of breast cancer patients: association with tumor biology and clinical outcome
Particular N-glycan structures are known to be associated with breast malignancies by coordinating various regulatory events within the tumor and corresponding microenvironment, thus implying that N-glycan patterns may be used for cancer stratification and as predictive or prognostic biomarkers. However, the association between N-glycans secreted by breast tumor and corresponding clinical relevance remain to be elucidated. We profiled N-glycans by HILIC UPLC across a discovery dataset composed of tumor interstitial fluids (TIF, n = 85), paired normal interstitial fluids (NIF, n = 54) and serum samples (n = 28) followed by independent evaluation, with the ultimate goal of identifying tumor-related N-glycan patterns in blood of patients with breast cancer. The segregation of N-linked oligosaccharides revealed 33 compositions, which exhibited differential abundances between TIF and NIF. TIFs were depleted of bisecting N-glycans, which are known to play essential roles in tumor suppression. An increased level of simple high mannose N-glycans in TIF strongly correlated with the presence of tumor infiltrating lymphocytes within tumor. At the same time, a low level of highly complex N-glycans in TIF inversely correlated with the presence of infiltrating lymphocytes within tumor. Survival analysis showed that patients exhibiting increased TIF abundance of GP24 had better outcomes, whereas low levels of GP10, GP23, GP38, and coreF were associated with poor prognosis. Levels of GP1, GP8, GP9, GP14, GP23, GP28, GP37, GP38, and coreF were significantly correlated between TIF and paired serum samples. Cross-validation analysis using an independent serum dataset supported the observed correlation between TIF and serum, for five of nine N-glycan groups: GP8, GP9, GP14, GP23, and coreF. Collectively, our results imply that profiling of N-glycans from proximal breast tumor fluids is a promising strategy for determining tumorderived glyco-signature(s) in the blood. N-glycans structures validated in our study may serve as novel biomarkers to improve the diagnostic and prognostic stratification of patients with breast cancer.
Particular N-glycan structures are known to be associated with breast malignancies by coordinating various regulatory events within the tumor and corresponding microenvironment, thus implying that N-glycan patterns may be used for cancer stratification and as predictive or prognostic biomarkers. However, the association between N-glycans secreted by breast tumor and corresponding clinical relevance remain to be elucidated. We profiled N-glycans by HILIC UPLC across a discovery dataset composed of tumor interstitial fluids (TIF, n = 85), paired normal interstitial fluids (NIF, n = 54) and serum samples (n = 28) followed by independent evaluation, with the ultimate goal of identifying tumor-related N-glycan patterns in blood of patients with breast cancer. The segregation of N-linked oligosaccharides revealed 33 compositions, which exhibited differential abundances between TIF and NIF. TIFs were depleted of bisecting N-glycans, which are known to play essential roles in tumor suppression. An increased level of simple high mannose N-glycans in TIF strongly correlated with the presence of tumor infiltrating lymphocytes within tumor. At the same time, a low level of highly complex N-glycans in TIF inversely correlated with the presence of infiltrating lymphocytes within tumor. Survival analysis showed that patients exhibiting increased TIF abundance of GP24 had better outcomes, whereas low levels of GP10, GP23, GP38, and coreF were associated with poor prognosis. Levels of GP1, GP8, GP9, GP14, GP23, GP28, GP37, GP38, and coreF were significantly correlated between TIF and paired serum samples. Cross-validation analysis using an independent serum dataset supported the observed correlation between TIF and serum, for five of nine N-glycan groups: GP8, GP9, GP14, GP23, and coreF. Collectively, our results imply that profiling of N-glycans from proximal breast tumor fluids is a promising strategy for determining tumorderived glyco-signature(s) in the blood. N-glycans structures validated in our study may serve as novel biomarkers to improve the diagnostic and prognostic stratification of patients with breast cancer.
# Introduction
Breast cancer (BC) is the most common cancer worldwide among women, with more than 1 300 000 new cases diagnosed every year. BC is the leading cause of cancer-related deaths among women [bib_ref] Global cancer incidence and mortality rates and trends-an update, Torre [/bib_ref]. Numerous studies have established that stepwise accumulation of multiple genetic and epigenetic alterations in epithelial cancer cells , as well as changes in stromal composition, drive and direct the progression of breast cancer [bib_ref] Systematic analysis of breast cancer morphology uncovers stromal features associated with survival, Beck [/bib_ref]. These studies highlight the heterogeneity and complexity of breast malignancies and point to a major challenge in the development of targeted therapeutics.
A growing body of evidence points to a crucial role of the multidirectional network communications between malignant epithelial cells and the tumor microenvironment in tumor evolution and progression. Multidirectional signaling events within the tumor stroma are implemented through the tumor interstitial fluid , which forms at the interface between circulating bodily (lymph and blood) and intracellular fluids. TIF facilitates the exchange of ions, proteins, cytokines, and miRNA within the interstitial space [bib_ref] Cytokine profiling of tumor interstitial fluid of the breast and its relationship..., Espinoza [/bib_ref] [bib_ref] Tumor interstitial fluid -a treasure trove of cancer biomarkers, Gromov [/bib_ref] [bib_ref] Gaining insights into cancer biology through exploration of the cancer secretome using..., Papaleo [/bib_ref]. Various biomolecules are released by tumor and stromal cells into the interstitium [bib_ref] Emerging roles of the tumor-associated stroma in promoting tumor metastasis, Horimoto [/bib_ref] [bib_ref] In-depth proteomic analysis of tissue interstitial fluid for hepatocellular carcinoma serum biomarker..., Zhang [/bib_ref] and subsequently drain through the lymphatic system into the bloodstream, where they can be detected and quantified [bib_ref] On the development of plasma protein biomarkers, Surinova [/bib_ref]. Given the high concentration of potential cancer-specific biomolecules within the local tumor milieu [bib_ref] Body fluid proteomics: Prospects for biomarker discovery, Ahn [/bib_ref] , interstitial fluid is considered to be a valuable resource for BC biomarker discovery [bib_ref] Tumor interstitial fluid formation, characterization, and clinical implications, Wagner [/bib_ref].
Glycosylation is a template-free enzymatic process that produces glycosidic linkages of monosaccharides to macromolecules such as carbohydrates, lipids, and proteins through the sequential attachment of glycan moieties in a function-specific context. This posttranslational modification is a well-known hallmark of cancer [bib_ref] Glycosylation in cancer: mechanisms and clinical implications, Pinho [/bib_ref] and is implicated in almost all molecular and metabolic events in normal and malignant cells. These events include protein folding and stability, cell-cell interaction, angiogenesis, immune modulation, cell signaling, and gene expression [bib_ref] Vertebrate protein glycosylation: diversity, synthesis and function, Moremen [/bib_ref]. Two major types of glycosylation (N-linked and O-linked) coexist in mammalian cells and often occur simultaneously on the same target macromolecules [bib_ref] Glycosylation in cancer: mechanisms and clinical implications, Pinho [/bib_ref].
The involvement of N-glycosylation in the development and progression of BC has been documented by in vitro and in vivo studies [bib_ref] ST6GalNAc I expression in MDA-MB-231 breast cancer cells greatly modifies their O-glycosylation..., Julien [/bib_ref]. N-glycan branching, particularly the increased expression of complex b-1,6 branched N-linked glycans, is often associated with more aggressive tumor behavior, such as enhanced migration, invasion, and metastatic potential [bib_ref] Inhibition of N-linked glycosylation disrupts receptor tyrosine kinase signaling in tumor cells, Contessa [/bib_ref]. In contrast, the expression of bisecting glycans strengthens cell adhesion and is associated with cancer suppression [bib_ref] Glycans and cancer: role of N-glycans in cancer biomarker, progression and metastasis,..., Taniguchi [/bib_ref]. Several N-glycan patterns with altered circulating glycan structures originating from either a primary tumor or from other organs in response to a neoplastic process have recently been described [bib_ref] Breast cancer diagnosis and prognosis through quantitative measurements of serum glycan profiles, Kyselova [/bib_ref]. Levels of biantennary N-glycan chains as well as a-2,3linked sialic acid-modified N-glycans are often decreased in sera of patients with BC, compared to healthy controls. The same tendency is observed in the sera of lung cancer patients, and it was suggested that an aberrant N-glycan signature based on serum glycan analysis could be used to distinguish cancer types [bib_ref] Serum glycoprotein-derived N-and O-linked glycans as cancer biomarkers, Lan [/bib_ref]. However, no robust blood glycan markers for BC have been identified to date, mainly because of the high degree of complexity and dynamic range of biomolecules (>10 orders of magnitude) circulating in the bloodstream. Furthermore, similar types of molecules are externalized from other body tissues and organs under physiological conditions.
To identify tumor-derived N-glycans patterns, we investigated the secreted glycome by profiling N-glycans released from matched tumors (TIF), normal mammary tissues (NIF), and serum samples using hydrophilic interaction liquid chromatography (HILIC) ultra-performance liquid chromatography (UPLC) [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref]. The aims of our study were (a) to compare N-glycans secreted directly from tumor and stromal cells to correlate the N-glycan profiles and the corresponding abundances in paired TIF and serum samples; (b) to explore whether the appearance of particular glycoforms in TIF is correlated with the presence of tumor infiltrating lymphocytes (TILs) in corresponding tumors; (c) to examine a potential association between N-glycan levels and clinical outcome; and (d) to evaluate our data and results of analysis using an independent cohort of the normal, benign, and BC blood samples. [bib_ref] Proteomic analysis of tissue samples in translational breast cancer research, Gromov [/bib_ref] at Copenhagen University Hospital. The criteria for high-risk cancers, applied by the DBCG, are age below 35 years, and/or a tumor diameter of more than 20 mm, and/or a histological malignancy 2 or 3, and/or negative estrogen (ER) and progesterone (PgR) receptor statuses, and/or a positive axillary status. Mastectomy enables pathologist to dissect a tissue sample from a nonmalignant area located relatively distant to the tumor, that is, 5 cm. at least. We used such criteria for dissection of normal breast lesions to avoid any impact of cancer field cancerization, which has been observed in histologically normal breast biopsies located 1 cm from the tumor margins, but not in lesions resected 5 cm from the tumor or obtained from reduction mammoplasty [bib_ref] Telomere DNA content and allelic imbalance demonstrate field cancerization in histologically normal..., Heaphy [/bib_ref] [bib_ref] Breast field cancerization: isolation and comparison of telomeraseexpressing cells in tumor and..., Trujillo [/bib_ref]. All normal tissue specimen dissected from the breast after mastectomy were morphologically and histologically evaluated [bib_ref] Histological evaluation of the normal breast, Russo [/bib_ref] to ensure normal epithelial acini and ducts structures.
# Materials and methods
All the patients presented a unifocal tumor, and none of the patients had a history of breast surgery or had received preoperative treatment (naive samples). Patients were followed after surgery, and cancer-specific survival was measured from the date of primary surgery until the date of death from BC. Death records were complete up to October 08, 2014 and served as the censoring date. Registered clinicopathological data for the patients were available from the Department of Pathology, Rigshospitalet, Copenhagen University Hospital, Denmark. This study was conducted in compliance with the Helsinki II Declaration, and written informed consent was obtained from all participants and approved by the Copenhagen and Frederiksberg regional division of the Danish National Committee on Biomedical Research Ethics (KF 01-069/03).
At the time of collection, each tissue specimen was divided into two pieces. One piece was stored at À80°C and was subsequently prepared as a formalinfixed paraffin-embedded (FFPE) sample that was sectioned, mounted on glass slides, and stained for histological characterization, tumor subtyping, TIL scoring, and immunohistochemistry (IHC) analysis. The second biopsy piece was placed in PBS at 4°C within 30-45 min of surgical excision and then was subjected to interstitial fluid recovery (see below).
Matched sera were obtained from women enrolled in the Danish Center for Translational Breast Cancer Research program who underwent surgery between 2001 and 2006. Blood samples were collected preoperatively following a standardized protocol [bib_ref] Plasma and serum levels of tissue inhibitor of metalloproteinases-1 are associated with..., Wurtz [/bib_ref]. Briefly, serum was collected in serum-separating tubes and was left on the bench for 30 min before centrifuging for 10 min at 2000 G.
The separate serum cohort, Mammographic Density and Genetis (MDG), consisted of serum N-glycan profiles from 107 BC patients and 62 healthy women [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref] and was used to validate the results of the TIF analysis.
2.2. Immunohistochemistry of tissue biopsies: histological assessment and tumor subtyping Immunohistochemistry analysis was performed as described elsewhere [bib_ref] Proteomic characterization of the interstitial fluid perfusing the breast tumor microenvironment: a..., Celis [/bib_ref]. First, small FFPE blocks were prepared from two to three various parts of the tissue piece and the sections were stained with a CK19 (KRT19) antibody. Tissue morphology, tumor cell content and visual assessment of tumor stroma percentages were evaluated as previously described [bib_ref] Cytokine profiling of tumor interstitial fluid of the breast and its relationship..., Espinoza [/bib_ref]. All slides were blindly reviewed by two independent investigators (IIG, PSG). Subtype scoring of the tumor tissues as luminal A (LumA), luminal B (LumB), luminal B HER2-enriched (LumB HER2-enriched), HER2, and triple negative breast cancer (TNBC) was performed based on the estrogen receptor (ER), progesterone receptor (PgR), epidermal growth factor receptor-2 (HER2), and Ki67 status determined for each tissue sample mainly in accordance with the St. Gallen International Breast Cancer Guidelines. The criteria used for each subtype classification are summarized in [fig_ref] Table 1: (Continued). [/fig_ref]. The monoclonal mouse antibody raised against CK19 (clone 4E8) was obtained from ThermoFischer Scientific. The monoclonal mouse antibody raised against Ki67 (clone MIB-1) was purchased from DAKO. The monoclonal antibody raised against ER (clone 1D5) was obtained from DAKO. The monoclonal antibody raised against synthetic peptide directed toward the N-terminal end of PgR was purchased from DAKO. The polyclonal rabbit antibody raised against Her2 (Hercep Test) was obtained from DAKO. For all staining, positive control slides were included in parallel in accordance with the manufactory instructions. For the negative controls, the slides were incubated with PBS instead of primary antibodies. All information about patients and samples analyzed in the study is presented in .
## Estimation of tumor infiltrating lymphocytes and their subpopulations
Immunohistochemistry analyses were performed to examine the most prominent components of the immune microenvironment in the corresponding tumor biopsies used for interstitial fluid recovery and molecular characterization. Scoring of total leukocytes, T lymphocytes, T helper lymphocytes, cytotoxic T lymphocytes, and macrophages were determined based on staining performed with antibodies raised against CD45+, CD3+, CD4+, CD8+, and CD68+, respectively. The monoclonal antibodies raised against CD45 (clone 2B11 + PD7/26), CD4 (clone IS 649), CD8 (clone 144B), and CD68 (clone PG-M1) were purchased from DAKO. The polyclonal antibody raised against synthetic peptide from the intracellular part of the e-chain of human CD3 was obtained from DAKO. The proportion of TILs in tissue sections was evaluated in accordance with the recommendations of the International TILs Working Group 2014 [bib_ref] The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an..., Salgado [/bib_ref]. An assessment of overall inflammatory reactions and the number of lymphoid cells present within biopsies were determined by hematoxylin and eosin (HE) staining as described elsewhere [bib_ref] Tumor-associated lymphocytes as an independent predictor of response to neoadjuvant chemotherapy in..., Denkert [/bib_ref] : 1+ (>10%), 2+ (10-50%), 3+ (>50%). These scores were independently and blindly assigned two independent investigators (IIG and PSG). The macrophage marker, CD68, was also evaluated with the same criteria. For each immune cell population that was analyzed, the expression results were dichotomized as low (<10%) and high (>10%). (columns T-X) contains the detailed information regarding stratification of the samples based on the TILs presence.
## Interstitial fluid recovery
Tumor interstitial fluids and NIF samples were extracted from fresh breast tumor and normal tissue specimen, as previously described [bib_ref] Proteomic characterization of the interstitial fluid perfusing the breast tumor microenvironment: a..., Celis [/bib_ref]. Briefly, 0.1-0.3 g of clean tissue was cut into small pieces (~1 mm 3 each), washed twice in cold PBS to remove blood and cell debris, and then incubated in PBS for 1 h at 37°C in a humidified CO 2 incubator. The samples were then centrifuged at 200 g and 4000 g for 2 min and 20 min, respectively, at 4°C. The supernatants were carefully aspirated, and total protein concentration for each sample was determined with the Bradford assay [bib_ref] A rapid and sensitive method for the quantitation of microgram quantities of..., Bradford [/bib_ref].
## N-glycan profiling by uplc
## Sample processing for uplc
About 50-100 lL of TIF, NIF, depending on the original protein concentration, was lyophilized followed by resuspension in 10 lL of distilled water. N-glycans were released using an updated version of the high-throughput automated method described by Stockmann and coauthors using a liquid-handling robot. The samples were then denaturated with dithiothreitol, alkylated with iodoacetamide, and N-glycans were released from the protein backbone enzymatically via PNGase F (Prozyme Glyco N-Glycanase, code GKE-5006D, 10 lL per well, 0.5 mU in 1 M ammonium bicarbonate, pH 8.0). The released glycans were captured on solid supports, and excess reagents, salts, and other impurities were removed by vacuum or centrifuge filtration, and the glycans were then released and labeled with the fluorophore 2-aminobenzamide (2-AB). Next, glycans were purified in a 96-well chemically inert filter plate (Millipore Solvinert, hydrophobic polytetrafluoroethylene membrane, 0.45 lm pore size) using Hyper-SepDiol SPE cartridges (ThermoScientific, Waltham, MA, USA) , with each well containing all glycans released from individual sample. The samples were then lyophilized and dissolved in 10 lL of an acetonitrile-water mixture (70 : 30).
# Uplc analysis
Purified N-glycans were automatically injected into the UPLS system in a mixture of 70% acetonitrile in water (see above). For UPLC analysis, a 2.1 9 150 mm HILIC column (Waters, Milford, MA, USA) was coupled with an Acquity UPLC system (Waters) equipped with a Waters temperature control module and a Waters Acquity fluorescence detector. The column temperature was set to 40°C, and two buffer solutions, A (50 mM formic acid adjusted to pH 4.4 with ammonia solution) and B (pure acetonitrile), were used to run the following 30 min linear gradient: 0.56 mLÁmin À1 flow rate for 23 min with 30-47% of buffer A followed by 47-70% of A and finally reverting back to 30% of A to complete the run. The elution of N-glycan was measured by fluorescence detection at 420 nm with excitation at 330 nm. The system was calibrated using an external standard of hydrolyzed and 2AB-labeled glucose oligomers to create a dextran ladder, as described previously . The use of an external standard enabled reproducible relative quantitation of glycans between the runs. The GU (glycose unit) assigned to every peak in the chromatogram, based on the standard and each peak (collection of glycans at the same GU), is proportional of the entire glycome calculated as 100% of fluorescence intensity.
A total of 165 N-glycans assigned to 46 glycan peaks (GP1 to GP46) were detected in tissue interstitial fluids. Each glycan peak (GP) contains several predominant structures. The total composition of all structures and predominant glycan features are summarized in .
# Feature analysis
N-glycan peaks (GP) were pooled based on similar structural or compositional features of the peak glycan members. Features relating to a peak were determined based on the major glycan members of that peak described at [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref].
## Bioinformatic and statistical analysis
## Curation of the dataset for analysis
The analysis of glycan abundances was performed using two datasets in parallel; (d1) all available samples corresponding to 85 TIF samples and 54 NIF samples and (d2) paired tumor and normal samples including a total of 54 individual TIF-NIF pairs (108 samples). The peaks of the glycan UPLC output represent the relative area for each glycan peak in the spectrum. The glycan abundances were log2-transformed to reduce the impact of outliers and to deal with the skewness of the glycan distribution. The log2 transformation resulted in the majority of glycan abundances approaching normal a distribution. After log2 transformation, the data were corrected for batch effects using the ComBat function of the SVA R package [bib_ref] Adjusting batch effects in microarray expression data using empirical Bayes methods, Johnson [/bib_ref]. ComBat batch-corrected data were only used for plotting purposes.
All the initial data, scripts for analyses, and outputs are released as free materials at https://github.com/ ELELAB/N-glycan-TIF so that our findings could be reproduced.
## Multidimensional scaling
Classical multidimensional scaling (R version 3.3.1) was used to reduce the number of the dimensions within the data. Specifically, 108 samples (54 TIF-NIF pairs) with each 63 measurements of glycan/glycan feature abundances were reduced to two dimensions (M1 and M2). Multidimensional scaling (MDS) was performed with the function cmdscale() using Euclidean distance as the distance metric. The plotting was done with R package ggplot2 2.2.1.
# Differential abundance analysis
The differential abundance analysis (DAA) was performed using the statistical software LIMMA (Linear Models for Microarray Data) implemented in R [bib_ref] ) limma powers differential expression analyses for RNA-sequencing and microarray studies, Ritchie [/bib_ref]. LIMMA has few underlying statistical assumptions and is known to be powerful for small sample sizes as a result of shrinkage of featurespecific variances [bib_ref] ) limma powers differential expression analyses for RNA-sequencing and microarray studies, Ritchie [/bib_ref]. Although LIMMA was originally developed for analysis of microarray, a number studies had shown the versatility of this software for the analysis of other -omics data [bib_ref] Insights into RNA biology from an atlas of mammalian mRNA-binding proteins, Castello [/bib_ref].
For the analysis of paired data (d2), the information on patient ID was incorporated into the design matrix to account for patient-specific effects. For the analysis with unpaired data (d1), information on batch was added to the model. We carried out DAA between NIF samples and TIF samples using a corrected Pvalue (FDR: false discovery rate) of 0.05 as the cutoff for significance.
To determine whether any clinical variables could be related to the abundance of specific N-glycans groups or N-glycan features, DAA was performed for tumor grade (Gr), receptor status (HER2, ER, PgR), and tumor infiltrating lymphocyte status (TILs, CD3, CD4, CD8, CD45, CD68) in the sample (see .
All analyses were performed using two datasets in parallel; (d1) all available samples corresponding to 85 TIFs (stratifying into luminal A = 43, luminal B = 12, luminal B HER2-enriched = 8, HER2 = 9, and TNBC = 13) and 54 NIF samples, and (d2) paired tumor and normal samples including a total of 54 individual TIF-NIF pairs (108 samples): luminal A = 22, luminal B = 10, luminal B HER2-enriched = 6, HER2 = 4, and TNBC = 12 (see .
# Hierarchical clustering analysis
Hierarchical clustering was performed to visually inspect the results of the DAA. Agglomerative hierarchical clustering was implemented in R with hclust() (R-stats), using the Ward's method (ward.2D), the statistical premise of which is to minimize the total within-cluster variance [bib_ref] Ward's Hierarchical Agglomerative Clustering Method: Which Algorithms Implement Ward's Criterion?, Murtagh [/bib_ref].
## Correlation analysis-glycan abundances in tif and serum
The log2-transformed abundances of TIF N-glycans structures were correlated with corresponding serum profiles of 28 serum samples available. Classic Pearson's product-moment correlation was performed in R. The significance of correlation scores was tested and obtained P-values corrected using FDR. Correlations with an FDR < 0.05 were considered significant and kept for further analysis.
# Survival analysis
Survival analysis was performed using a Cox proportional hazard model. Dataset d1 (e.g., all available TIF, n = 85) was used for analysis. Survival was modeled using one N-glycan feature at a time, for example, not accounting for potential inter-glycan abundance effects. Clinical parameters where tested for confounding effects on N-glycan levels and/or clinical outcome. As expected, age at diagnosis was found to have a significant effect on overall survival. Of the remaining parameters, TIL status (CD4 + and CD45 + ) was found to be a confounder. Before regression analysis, the covariates were tested for violation of the proportional hazard assumption. Also, the log-linearity of continuous variables (N-glycans and age) was evaluated. In the final models, age was modeled with splines (df = 2). Four confounder packages were tested accounting for an age at surgery and/or tumor infiltrating lymphocytes status (total TILs, CD4 + , and CD45 + )-GPX represents a glycan peak:
[formula] cpk1 ¼ GPX þ Age cpk2 ¼ GPX þ Age þ TILs cpk1 ¼ GPX þ Age þ CD4 cpk2 ¼ GPX þ Age þ CD45 [/formula]
In addition to the cox regression model with overall patient survival as outcome (censored = 0 and event = 1), survival analysis was performed using, as events, only those deaths for which information on primary cause of death was available and denoted as 'malignant neoplasm of breast' (censored = 0 and malignant neoplasm of breast = 1).
Results of the cox models were reported as hazard ratios, confidence intervals, and FDR values. Survival curves were generated using the corrected regression models. Survival curves were made assuming an age of 66 at surgery (median age at entry for the cohort). For each N-glycan composition, high abundance was defined as the upper 25th percentile, while low abundance was defined as the lower 25th percentile. Survival analysis was performed using R-packages survcomp and survminer [bib_ref] A comparative study of survival models for breast cancer prognostication based on..., Haibe-Kains [/bib_ref].
The experimental workflow including number of samples used in each analysis is presented in [fig_ref] Figure 1: The experimental workflow including final number of samples used in each analysis [/fig_ref].
# Results
## Comparative analysis of n-glycan structures in matched tif and nif: distribution across five bc subtypes and correlation with clinicopathological parameters
To obtain a general overview of N-glycan profiles across TIF-and NIF-matched counterparts, we plotted all paired samples using multidimensional scaling (MDS). Forty-six glycan groups (GP1-GP46) and 17 N-glycan features were quantified . The MDS plotting revealed considerable segregation of TIF and NIF samples [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref].
To evaluate a possible segregation of glycan patterns across five main BC subtypes, we stratified tumor samples in accordance with the St. Galen criteria: luminal A, luminal B, luminal B HER2-enriched, HER2, and TNBC. As seen in [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref] , no clear clustering between subtypes was identified [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref] , even after merging samples into three major groups: luminal, HER2, and TNBC (data not shown). The absence of a significant difference in N-glycan abundance across subtypes may be partly explained by a large difference in the numbers of samples in each subtype group . We speculate that the partitioning of BC samples into subtypes based on immunohistochemistry is not directly transferrable to N-glycan abundance and/or that N-glycan levels may reflect an alternative glycan-based tumor stratification.
We also did not find any significant correlation between the abundance of TIF N-glycan structures and clinical tumor variables including grade, type, and/or hormone or growth receptor status (data not shown). These results indicate that N-glycans externalized into breast tumor interstitial fluid may not be directly associated with these clinicopathological characteristics of the tumor.
In order to determine which N-glycans were differentially represented in fluids originating from tumor [fig_ref] Table 1: (Continued). [/fig_ref]. N-glycans with differential abundance in TIF and NIF. Statistics for each differentially abundant N-glycan group and feature (log-fold change, P-value, and FDR), as well as directionality in TIF (up or down). Bisected N-glycans are highlighted in bold. compared to normal breast tissue, we performed differential abundance analysis (DAA) using paired TIF-NIF samples. In accordance with the clustering observed in the MDS plot [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref] , DAA yielded 33 N-glycan groups with significantly differential abundance in TIF vs. NIF: 13 groups with significantly elevated levels in TIF samples and 20 groups with significantly decreased levels in TIF as compared to NIF counterparts [fig_ref] Figure 1: The experimental workflow including final number of samples used in each analysis [/fig_ref]. Our results showed that TIF samples were enriched for particular type of sialylated (S3-S4), highly galactosylated (G3-G4) N-glycans with a high number of antennae (A3-A4), as well as for simpler N-glycans (such as monoantennary glycan, A1, no galactosylation, G0). A significant decrease in core fucosylated and bisected Nglycans, represented by GP4, GP7, GP10, GP15, GP20, GP23, GP26, GP28, and GP35, was observed in TIF. DAA with all available samples (d1 set-see Material and methods 6.1) yielded a set of N-glycans almost identical to the one obtained with paired samples only (data not shown). . N-glycans with differential abundance in TIL-enriched and TIL-depleted samples. (A) Bar plot shows N-glycan groups with differential abundance in tumor samples with low (0/+1) and high (+2/ +3) overall TIL status, as determined by CD45 positivity (see for details). (B) Bar plot shows N-glycan groups with differential abundance in samples with low vs. high TIL status, as determined by CD4 positivity. Height and directionality of bars indicate log-fold change. Shade depicts inverse FDR: Darker shade indicates lower FDR. All N-glycans depicted in the plot had FDR ≤ 0.05.
[formula] 1.0 A1 G0 GP6 S0 G1 highM GP21 GP41 GP38 GP45 S4 GP37 GP43 G4 A4 outerarmF GP25 S3 GP32 S2 [/formula]
## Association of n-glycan pattern with tils
Tumor infiltrating lymphocytes have been shown to play an essential role in BC progression, influencing cross talk between tumor and stromal cells and providing prognostic and potentially predictive values [bib_ref] Disentangling the relationship between tumor genetic programs and immune responsiveness, Bedognetti [/bib_ref] [bib_ref] Tumor-Infiltrating Lymphocytes: A Predictive and Prognostic Biomarker in Neoadjuvant-Treated HER2-Positive Breast Cancer, Heppner [/bib_ref]. To disclose a possible relationship between N-glycan structures, which displayed differential abundance between TIF and NIF [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref] with the composition of TILs within tumor microenvironment for details), we performed DAA, considering the extent of lymphocyte infiltration within tumor biopsies. A detailed evaluation of TIL subtypes, often described in breast cancer literature, was performed by IHC using the antibodies specific for particular lymphocyte antigen [fig_ref] Figure 1: The experimental workflow including final number of samples used in each analysis [/fig_ref].
The association between N-glycans and overall TILs determined by CD45 + revealed ten N-glycan groups, which were differentially abundant in TIF released from biopsies enriched for TILs compared to samples with low TIL status . In particular, it turned out that samples with high levels of lymphocyte infiltration (see for details) showed significantly high levels of G1 (glycans with 1 galactose) and S0 (glycans without sialic acid), whereas GP38 (mostly tetraantennary tetragalactosylated trisialylated glycans, A4G4S3), GP41 (mostly tetraantennary tetragalactosylated tetrasialylated glycans), GP45 (mostly outerarm fucosylated tetraantennary tetragalactosylated tetrasialylated glycans), S4 (tetrasialylated glycans), G4 (tetragalactosylated glycans), A4 (tetraantennary glycans), GP25 (mainly biantennary digalactosylated disialylated glycans without fucosylation, A2G2S2), and outer-armF (outerarm fucosylated glycans) were predominant in samples with low level of lymphocyte infiltration and .
To determine which subpopulation of TILs contributed most to the differential abundance of N-glycoforms in tumor proximal fluid, glycome profiles were analyzed relative to four TIL subgroups: CD3 + , CD4 + , CD8 + , and CD68 + . The results showed that total (CD45 + ) and T helper cells (CD4 + )-enriched biopsies displayed relatively similar N-glycan patterns , B). No significant differential abundance was observed for CD3+, CD8+, and CD68+ TIL subpopulations (data not shown). A high abundance of A1 (monoantennary glycans), G0 (agalactosylated glycans), GP6 (mainly high-mannose glycans, M5), highM (total high-mannose glycans), and GP21 (mainly biantennary monosialylated glycans [A2G2S1] and high mannosylated hybrid glycans [M5A1G1S as well as S0 and G1 was associated with CD4 + -enriched samples. Inversely, samples with high CD4 + had lower levels of multibranched N-glycans and highly sialylated N-glycans, many of which contain outerarm fucose (e.g., GP25, GP32, GP37, GP38, GP41, GP43, GP45) peaks as well as S4, G4, A4, outerarmF, S3, and S2 features.
## Prognostic potential of tif n-glycan abundances
To determine whether N-glycan abundance predicted outcomes for patients with BC, overall survival analysis was performed across all patients for which survival information was available with a Cox proportional hazard model. Two types of Cox models were compared: one corrected only for age at diagnosis and one corrected for age at diagnosis + TIL status (overall, estimated by CD45 + and CD4 + ).
The Cox model, corrected only for age at diagnosis, yielded six N-glycan groups, which were significantly associated with overall outcome . One of these, GP24 (biantennary bigalactosylated bi-sialic-acid . Differential abundance of N-glycan groups segregating TIF, NIF, matched serum, and MDG cancer and normal serum. Column 3 = N-glycan rank in TIF based on log-fold change, in total: 13 0 N-glycan groups increased in TIF; 20 N-glycan groups decreased in TIF (see [fig_ref] Table 1: (Continued). [/fig_ref]. Column 4 = N-glycan rank in MDG cancer serum based on log-fold change, in total: 26 N-glycan groups increased and 18 N-glycan groups decreased. Column 5 indicates correlations between abundance of N-glycan composition in TIF and paired serum samples. DA, differential abundance. N-glycan groups for which abundance in MDG correlated with abundance in TIF-paired serum datasets are highlighted in bold. glycan, A2G2S2), had a hazard ratio below 1, for example, a high level of GP24 was predictive of superior prognosis. The remaining five groups had hazard ratios greater than 1, implying that a high abundance of these was associated with poor outcomes. These included GP5 (core fucosylated biantennary glycan, FA2), GP10 (core fucosylated bisected biantennary monogalactosylated glycan, FA2[6]BG1), GP23 (core fucosylated bisected biantennary bigalactosylated monosialylated glycans, FA2BG2S1), GP38 (mostly tetraantennary tetragalactosylated trisialylated glycans, A4G4S3), and coreF (core fucosylated glycans) . All glycans, except GP5, were among those that segregated TIF and NIF. The results of the survival analysis, in which a death was only classified as an events if the cause of death was known to be 'malignant neoplasm of breast', yielded a similar pattern as the one observed for the cox model with overall survival, with N-glycans GP5, GP8, GP10, GP23, GP38, and coreF displaying high hazard ratios (HR:~2.0 -7.0) [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref]. However, despite the high HRs and the fact that the 95% confidence intervals of these N-glycans did not overlap 1, P-values were no longer significant after correction for multiple testing (FDR). We attribute this lack of significance to the lower power associated with this model, for example, if only outcomes with known cause of death are classified as events, the ratio of events/censures is notably reduced-this will have a large impact on a small(er) dataset.
Correction for TILs estimated by CD45 + or CD4 + did not alter the overall results of survival analysis. shows the survival curves using the age-corrected regression model for each structure associated with overall survival. The association of low GP10 and GP23 levels with poorer outcome is in agreement with the functional role of bisecting N-glycans in tumor development.
## Correlation of n-glycan abundance in paired tif and serum
To identify N-glycans with potential as noninvasive blood-based biomarkers for breast malignancy, we determined which of the 33 N-glycans displayed a significant correlation of abundances within 28 paired serum samples, after segregating NIF and TIF samples [fig_ref] Figure 1: The experimental workflow including final number of samples used in each analysis [/fig_ref].
Classic Pearson's product-moment analysis was performed to correlate N-glycan structures in TIF and serum. Of 33 N-glycan groups, nine were correlated with N-glycan levels in serum .
To determine whether any of the N-glycans present in serum reflect tumor immune status, we performed correlation analysis for the overall proportion of TILs (CD45+) and T helper cells (CD4+) within corresponding tumors. Two N-glycan groups, GP37 (mostly triantennary outerarm fucosylated trigalactosylated trisialylated glycans, A3F1G3S3) and GP38 (mostly tetraantennary tetragalactosylated trisialylated glycans, A4G4S3), were less abundant in serum from patients exhibiting high overall levels of TIL (CD45 + ) or T helper cells, as determined by CD4 + staining in corresponding tumors . N-glycan GP23, GP38, and coreF structures were identified as predictive for overall survival , demonstrating a significant correlation of abundance between TIF and paired serum (see for details). . Overview of the main results obtained in the study. N-glycan groups (1-46) and N-glycan features. Rows = traits and columns = Nglycans ID. Black dots denote which N-glycan group and features were significantly associated with a given trait based on analysis described in the corresponding result sections. The prominent N-glycan structures within a given N-glycan peak are specified below the GP's ID.
## Validation of n-glycan structures in an independent serum cohort
To validate the results obtained from glycan profiling of matched TIF/serum samples, 33 DA N-glycans were analyzed across an independent serum MDG dataset. The MDG cohort contains samples obtained from healthy controls and BC patients [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref] and profiled with the same UPLC-based technology, thus minimizing technical variability between experimental platforms. DAA was applied to the MDG dataset on log2-transformed data, in agreement with the protocol applied to our TIF-NIF BC cohort.
Twelve of 33 glycan groups were found to be abundant in both TIF and MDG BC serum samples as compared to NIF and MDG normal serum . Specifically, we found N-glycan peaks for GP3 (monoantennary monogalactosylated glycan, A1[6] G1), GP30 (tetraantennary tetragalactosylated monosialylated glycan, A4G4S[3]1), GP39 (tetraantennary tetragalactosylated trisialylated glycans, A4G4S3), and GP40 (mostly tetraantennary outerarm fucosylated trigalactosylated trisialylated glycans, A4F1G3S3) to be more abundant in TIF and BC serum samples as compared to NIF and normal serum samples. Inversely, bisected biantennary core fucosylated N-glycans (GP8, GP9, GP10, GP14, GP15, G23, as well as A2 and fucosylated glycans [coreF]) were less abundant in TIF and BC serum compared to normal samples.
Five of 12 N-glycan groups (highlighted in bold in displayed a significant correlation of abundance in MDG-and TIF-paired serum datasets. Segregation analysis based on these groups (GP8, GP9, GP14, GP23, and coreF) revealed a significant separation between normal and cancer MDG serum samples and showed a significant correlation in matched serum .
# Discussion
To the best of our knowledge, this study is the first analysis of the N-glycome in the tumor interstitium of patients with BC. Experiments were designed to identify aberrant glycosylation associated with tumor growth and progression. The study is part of a comprehensive project focused on characterization of the entire molecular complement of breast tumor interstitial fluid, aiming to identify integrated signatures associated with events underlying breast tumor metabolism, as detectable in blood [bib_ref] Cytokine profiling of tumor interstitial fluid of the breast and its relationship..., Espinoza [/bib_ref] [bib_ref] Profiling of microRNAs in tumor interstitial fluid of breast tumors -a novel..., Halvorsen [/bib_ref]. The analysis included a detailed morphological characterization of tumor lesions and evaluation of the spatial heterogeneity of TILs in tumor specimens, to elucidate the influence of tumor immune composition on secreted N-glycome complement. Data were subjected to bioinformatics analysis to characterize the N-glycome in the breast tissue interstitium and to reveal potentially valuable correlations of aberrant glycan patterns with breast tumor biology, including clinical outcome and presence in the blood. Finally, data were computationally validated with the independent serum MDG dataset [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref] , which contains BC carcinoma and nonmalignant serum profiled by analogous technology. summarizes the main results of our analyses.
## N-glycan patterns in tif
Multidimensional scaling and DAA of N-glycan profiles revealed distinct segregation between TIF and NIF. We detected 33 N-glycan groups and features with differential level between groups. N-glycan structures displaying significantly higher abundance in TIF as compared to NIF belong mainly to the monoantennary type [GP1, GP3, and A1 (sum of all monoantennary glycans)]. Our results are consistent with those of a recent study reporting a clear segregation of N-glycans circulating in the blood of patients with BC and normal individuals [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref] , particularly, for the core fucosylated N-glycans [bib_ref] Serum N-Glycome Characterization in Patients with Resectable Periampullary Adenocarcinoma, Hamfjord [/bib_ref] [bib_ref] Enzymes for N-Glycan branching and their genetic and nongenetic regulation in cancer, Kizuka [/bib_ref] that are represented by a set of GP4, GP7, GP10, GP15, GP20, GP23, GP26, GP28, and GP35 structures in our TIF samples.
It has previously been shown that specific glycan structures have different impacts on cell adhesion, which is one of the major molecular events during malignant transformation that affects cancer cell fate [bib_ref] Vertebrate protein glycosylation: diversity, synthesis and function, Moremen [/bib_ref]. Interestingly, we found nine bisecting structures (GP4, GP7, GP10, GP15, GP20, GP23, GP26, GP28, and GP35) to be significantly more abundant in NIF. The observed depletion of bisecting N-glycans in TIF is in agreement with current consensus regarding the functionality and role of N-bisecting glycans in cancer progression. Recent research has shown that extension of GlcNAc bisecting has a significant effect on cell survival and tumor aggressiveness [bib_ref] Enzymes for N-Glycan branching and their genetic and nongenetic regulation in cancer, Kizuka [/bib_ref]. This phenomenon has also been reported for cadherins, proteins that have a substantial impact on cell adhesion. When modified by bisecting glycans, cadherins reinforce cell adhesion and are consequently associated with cancer suppression. In contrast, cadherins bearing branched complex N-glycans are less involved in the control of cell adhesion and are associated with cancer progression [bib_ref] Cadherins Glycans in Cancer: Sweet Players in a Bitter Process, Carvalho [/bib_ref]. It has been proposed, mainly by using in vitro model systems, that the presence of this unique bisecting structural feature has important implications for the entire cellular glycan complement. Thus, enzymes responsible for producing N-glycan groups other than those with bisecting branches (e.g., GnT-IV, GnT-V) are almost completely inhibited by the presence of a bisecting GlcNAc residue in the N-glycan molecule. The results of our study support this notion, clearly demonstrating a significant abundance of particular set of bisecting glycan species in the interstitium of nonmalignant lesions as compared to their neoplastic counterparts.
## Correlation between tils and n-glycan composition
Tumor infiltrating lymphocytes are frequently found within tumors, suggesting that tumors trigger an immune response in the host. The presence of TILs within the tumor microenvironment has been reported as an important biomarker linked to clinical outcome [bib_ref] Tumor-Infiltrating Lymphocytes: A Predictive and Prognostic Biomarker in Neoadjuvant-Treated HER2-Positive Breast Cancer, Heppner [/bib_ref]. In this study, we immuneprofiled particular subsets of TILs, for example, those most often described in connection with BC in current literature [bib_ref] Tumor-associated lymphocytes as an independent predictor of response to neoadjuvant chemotherapy in..., Denkert [/bib_ref] [bib_ref] The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an..., Salgado [/bib_ref].
We identified a number of glycoconjugates in TIF that were significantly associated with the proportion of TILs , as determined by immunohistochemistry. In samples with high levels of total TILs, we observed an increase in simple high mannose N-glycan features (G0, G1, S0, and highM) and, inversely, a decrease in the abundance of highly complex N-glycan groups (A4, G4, S4, GP25, GP37, GP38, GP41, GP45, etc.). To our knowledge, our data are the first evidence highlighting a direct impact of the tumor immune complement on the secreted N-glycan profile in breast tumor.
## Relationship between n-glycan patterns and clinical outcome
Our survival analysis with overall patient survival as outcome, showed a significant association between Nglycome profiles and the overall survival of patients with BC. Cox proportional hazard regression revealed five N-glycan peaks to be significantly associated with poor survival (GP5, GP10, GP23, GP38, and coreF) and one glycan peak (GP24) as predictive of positive clinical outcome . All glycan peaks, except GP5, were among those that segregated TIF and NIF. We speculate that the absence of GP5 (core fucosylated biantennary glycan, FA2) among the 33 differentially abundant glycans segregating TIF and NIF may be related to the fact that only a subset of patients exhibit high abundance of this N-glycan, which is prognostic for overall survival according to our analyses. Indeed, the remaining patients display GP5 levels similar to those observed in NIF. Thus, the differential abundance of GP5 will not be detected when stratifying NIF and TIF by DAA.
The cox model in which deaths were classified as events only if cause of death was known and annotated as 'malignant neoplasm of breast' did not yield significant results after FDR correction [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref]. Although it might be that the N-glycan groups identified as prognostic from the cox model with overall patient survival have inflated P-values and may be considered as partly false positives, we hypothesize that the observed lack of significance merely reflects the decrease in power (larger confidence intervals) of this model, for an already small dataset. This observation is supported by the fact that the two cox models show similar results in terms of Nglycans groups identified as having the highest hazard ratios (GP5, GP10, GP23, GP38, and coreF), with significant P-values before FDR correction. We cannot say for sure that patients, for which we do not have cause of death, did not die due cancer presence even if the primary cause of death is not BC itself, but a 'side effect' of disease. The later notion might still be of interest for prediction of patient prognosis by using identified N-glycan signature.
Although correction for TIL status (CD45 + or CD4 + ) did not affect the overall results of survival analysis, the corrected P-value for GP38 did decrease when TILs were added to the model, highlighting the relationship between GP38 levels and TIL status seen in DAA. Reduced levels of GP38 correlated with a high proportion of overall and CD4 + TILs, which contribute to tumor suppression [bib_ref] Tapping CD4 T cells for cancer immunotherapy: the choice of personalized genomics, Zanetti [/bib_ref] , thus supporting an association between tumor immune status and overall survival in patients with BC. Identification of GP10 and GP23 in relation to clinical outcome supports the suggested 'protective' status of bisecting glycans [bib_ref] Enzymes for N-Glycan branching and their genetic and nongenetic regulation in cancer, Kizuka [/bib_ref] , whereas decreased levels of coreF have recently been reported to contribute to the malignancy of gastric cancer [bib_ref] Decreased corefucosylation contributes to malignancy in gastric cancer, Zhao [/bib_ref].
## Correlation of n-glycan abundance in tif and serum samples
Changes in N-linked glycan structure in serum or plasma of patients diagnosed with breast, prostate, ovarian, pancreatic, liver, or lung cancer have recently been reported . Alterations in the N-glycome profile may be the result of a primary response or a general systemic reaction of the body to the progression and metabolism of a tumor. Additionally, the high degree of complexity and dynamic range of biomolecules externalized physiologically to the blood from other tissues can mask molecules released from the primary tumor. Comparative TIF serum analysis helps to discriminate biomolecules released directly from primary tumor into the tumor interstitium from the systemic body response. In this study, of 33 N-glycans that segregated TIF and NIF samples, levels of nine glycan groups (GP1, GP8, GP9, GP14, GP23, GP28, GP37, GP38, and coreF) were significantly correlated with N-glycan levels in serum . One of these N-glycan groups, GP1 (monoantennary glycan, A1), displayed an inverse significant correlation with corresponding levels in serum, that is, high levels in TIF corresponded to low levels in serum. This observation may indicate that the molecules carrying this particular glycan feature accumulate within the tumor interstitium as a primary tumor response; however, this process is not associated with subsequent transport into the bloodstream. A more detailed look at the N-glycan profiles in TIF and NIF showed that two other N-glycan groups, GP3 and A1, with high TIF abundance [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref] also exhibited a negative association with corresponding serum levels, although these trends were not significant (P = À0.19 and À0.44, respectively). The majority of N-glycans detected in these peaks are all core fucosylated biantennary except for GP38. Our findings support previous reports describing the decreased levels of some core fucosylated glycans in the sera of patients with BC [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref]. The low levels of these types of N-glycans in the sera of patients with BC may indicate that biomolecules in the tumor interstitium bearing these N-glycan structures do not reach the bloodstream, but, rather, are involved in intercellular cross-communication within the local tumor space. This assumption may be supported by the functional features reported for core fucosylated glycans [bib_ref] Biological function of fucosylation in cancer biology, Miyoshi [/bib_ref]. Alternatively, the inverse association between particular glycans in serum and TIF may be the result of a high dilution factor as well as the expected presence of other, more abundant, glycan species originated from no tumor sites, thus masking the presence of tumor-derived biomolecules in the blood.
Computational validation of the paired TIF-NIF serum data was achieved through comparison with the independent N-glycan MDG serum dataset [bib_ref] Association of Nglycosylation with breast carcinoma and systemic features using high-resolution quantitative..., Saldova [/bib_ref]. Among the nine N-glycan groups, levels of which in TIF were significantly correlated with those in matched serum samples, five (GP8, GP9, GP14, GP23, and coreF) were validated within the MDG serum dataset. The fact that we did not detect more overlaps in this validation experiment may be explained by the fact that (in contrast to the MDG serum dataset) our TIF-matched serum dataset did not include blood samples from healthy individuals, which are important when establishing the correct baseline for normality.
Levels of most biantennary glycans, such as a2,3 sialic acid-modified N-glycan chains, decreased in the sera of patients with BC in this study, that is in agreement with previously reported data. The opposite trend was observed in the sera of lung cancer patients, which is characterized by a high level of biantennary N-glycan chains containing Sialyl Lewis structure (SLex) [bib_ref] Serum glycoprotein-derived N-and O-linked glycans as cancer biomarkers, Lan [/bib_ref]. Serum levels of biantennary N-glycan chains carrying core fucose or both core fucose and sialic acid, as well as the level of complex triantennary N-glycan containing only one sialic acid or both fucose and sialic acid, were decreased in tumor samples as compared to normal controls.
# Conclusions
The results of our study showed (a) clear segregation of patterns of N-glycan release from tumor vs. normal mammary tissue; (b) elevated levels of particular bisecting glycans (GP4, GP7, GP10, GP15, GP20, GP23, GP26, GP37, and GP28), which contribute to tumor suppression in normal breast tissue interstitium; (c) association of several N-glycans (A1, G0, GP6, M5, highM, GP21, GP41, GP38, GP45, GP37, GP43, GP26, GP32, and S2) in breast tumor interstitium with the proportion and composition of infiltrating lymphocyte populations; and (d) correlation of N-glycan pattern in TIF and corresponding serum with clinical outcome. Levels of five differentially abundant N-glycans correlated with levels in TIF and matched serum (GP8, GP9, GP14, GP23, and coreF). Importantly, the prognostic potential of GP23 and coreF was validated in an independent serum cohort. These N-glycans most likely reflect the signaling events underlying tumor biology and progression and may have potential for use as biomarkers to improve the diagnostic and prognostic stratification of BC. In the current study, we were not able to estimate whether particular adjuvant therapies would have had any impact on the abundance patterns of released N-glycan in association with clinical outcome due to diversity of the treatment applied to the patients included in the discovery set. Further evaluation of the presented data using large independent dataset of serum from patients with breast cancer should be performed in a future.
## Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article: [fig_ref] Figure 1: The experimental workflow including final number of samples used in each analysis [/fig_ref]. The representative images of TILs distribution within a single tumor biopsy based on the IHC analysis. [fig_ref] Figure 2: N-glycans with differential abundance between TIF and NIF [/fig_ref]. Cox proportional-hazard regression with known cause of death. . The segregation of MDG BC cancer and normal serum based on the level of five N-glycans groups exhibited differential abundance across TIF, NIF and matched serum. [fig_ref] Table 1: (Continued). [/fig_ref]. The biopsies with ≥ 1% of the invasive cancer cells positively stained for ER-and PgR were classified as positive. . Complete characteristics of 85 breast cancer patients enrolled in the study. . Glycan peaks and corresponding N-glycan features. . N-glycans identified as differentially abundant between samples with high and low tumor TILs. N-glycans are reported with associated log fold changes and adjusted P-values. . N-glycans with significantly correlated abundances between paired TIF and serum samples. N-glycans are reported with associated pearsons correlation score and adjusted P-value.
[fig] Figure 1: The experimental workflow including final number of samples used in each analysis. [/fig]
[fig] Figure 2: N-glycans with differential abundance between TIF and NIF. (A) Multidimensional scaling plot. M1 and M2 are the scaling components that best captured the distance (squared euclidian) relationship between samples. Gray points denote NIF samples; colored points are TIFs, stratifying into BC subtypes luminal A (LumA), luminal B (LumB), luminal B HER2-enriched (LumB HER2-enriched), HER2, and TNBC. Ellipses capture the majority of samples within a single group. (B) Bar plot shows 33 N-glycan groups with differential abundance between TIF and NIF. Height and directionality of bars indicate the log-fold change; color depicts the scaled inverse FDR: Darker shade indicates a lower FDR. All glycans depicted in the plot had FDR ≤ 0.05. Stars mark peaks containing bisecting N-glycans. [/fig]
[fig] Figure 4, Figure 5: Prognostic association of TIF N-glycans with overall survival. (A) Cox proportional hazards regression. Dot plot shows hazard ratios and confidence intervals for each N-glycan group. Dot size indicates the inverse FDR, that is, small FDRs are depicted as large dots. Dot color indicates whether the abundance of a given N-glycan was significantly associated with overall survival (FDR ≤ 0.05). (B) Survival curves based on low and high abundance of five N-glycan groups associated with poorer overall survival and one N-glycan group (GP24) found to be protective. Curves are fitted (simulated) using the age at diagnosis of 66 years (median of observed data). Curves show the probability of survival in years after diagnosis (at the age of 66) as a function of a high (orange) or low (blue) level of a given N-glycan(minimum observed value-lower 25th percentile, and maximum observed value-upper 25th percentile for the N-glycan in question). Correlation of N-glycan abundance in TIF and matched serum samples. (A) Dot plot shows Pearson's correlation coefficient between N-glycan abundances in serum and TIF. Shade of dot represents inverse FDR. (B) Scatter plots of nine N-glycan groups with abundances significantly correlated in serum and TIFs. Line color indicates directionality (red = negative correlation, blue = positive correlation). Gray shading shows the confidence of the regression line in a given area of the plot. [/fig]
|
New Insights into Plant Extracellular DNA. A Study in Soybean Root Extracellular Trap
Citation: Chambard, M.; Plasson, C.; Derambure, C.; Coutant, S.; Tournier, I.; Lefranc, B.; Leprince, J.; Kiefer-Meyer, M.-C.; Driouich, A.; Follet-Gueye, M.-L.; et al. New Insights into Plant Extracellular DNA. A Study in Soybean Root Extracellular Trap. Cells 2021, 10, 69.
# Introduction
Extracellular DNA (exDNA) is known to be a part of different extracellular traps (ETs) in mammals and other animals like crocodiles [bib_ref] Extracellular histones identified in crocodile blood inhibit in-vitro HIV-1 infection, Kozlowski [/bib_ref] , fish [bib_ref] Fish cast NETs: Neutrophil extracellular traps are released from fish neutrophils, Palić [/bib_ref] , or chickens [bib_ref] In vitro and in vivo generation of heterophil extracellular traps after Salmonella..., Pieper [/bib_ref]. ETs are released by immune cells like neutrophils, monocytes, mastocytes, macrophages, or eosinophils [bib_ref] Phagocytosisindependent antimicrobial activity of mast cells by means of extracellular trap formation, Von Köckritz-Blickwede [/bib_ref] [bib_ref] Distinct Cell Death Programs in Monocytes Regulate Innate Responses Following Challenge with..., Webster [/bib_ref] [bib_ref] Eosinophil and neutrophil extracellular DNA traps in human allergic asthmatic airways, Dworski [/bib_ref] [bib_ref] Eosinophil extracellular DNA traps in skin diseases, Simon [/bib_ref] [bib_ref] Mannheimia haemolytica and Its Leukotoxin Cause Macrophage Extracellular Trap Formation by Bovine..., Aulik [/bib_ref]. exDNA from ETs are released upon pathogen detection [bib_ref] Infection-induced NETosis is a dynamic process involving neutrophil multitasking in vivo, Yipp [/bib_ref] or during inflammation, allergy, coagulation, or autoimmune diseases [bib_ref] Extracellular DNA traps in allergic, infectious, and autoimmune diseases, Simon [/bib_ref] [bib_ref] Fafutis-Morris, M. Neutrophil Extracellular Traps and Its Implications in Inflammation: An Overview, Delgado-Rizo [/bib_ref] [bib_ref] Stimulus-dependent chromatin dynamics, citrullination, calcium signalling and ROS production during NET formation, De Bont [/bib_ref]. In neutrophils ET (NET), exDNA release has been well described and involves two processes: suicide lytic NETosis or vital NETosis produced by living cells. In vital NETosis, exDNA may be released by mitochondria [bib_ref] Viable neutrophils release mitochondrial DNA to form neutrophil extracellular traps, Yousefi [/bib_ref]. exDNA is also known to be a part of biofilms from bacteria, archaea, or fungi, allowing biofilms' structuration and stability, and improving their resistance to antibiotics or antifungal compounds [bib_ref] A characterization of DNA release in Pseudomonas aeruginosa cultures and biofilms, Allesen-Holm [/bib_ref] [bib_ref] Presence of Extracellular DNA in the Candida albicans Biofilm Matrix and its..., Martins [/bib_ref] [bib_ref] Effect of Extracellular DNA Destruction by DNase I on Characteristics of Forming..., Tetz [/bib_ref] [bib_ref] Mechanisms and Regulation of Extracellular DNA Release and Its Biological Roles in..., Ibáñez De Aldecoa [/bib_ref]. In soils or aquatic environments, exDNA has also been found [bib_ref] Mechanisms and Regulation of Extracellular DNA Release and Its Biological Roles in..., Ibáñez De Aldecoa [/bib_ref] [bib_ref] Preliminary evidences of the presence of extracellular DNA single stranded forms in..., Pathan [/bib_ref] and used as a nutritional source for microorganisms as marine bacteria or Archaea [bib_ref] Preservation and Significance of Extracellular DNA in Ferruginous Sediments from Lake Towuti, Vuillemin [/bib_ref]. exDNA is also a part of plant root ET (RET), a defense network located at the root apex made up of mucilaginous secretions (including glycomolecules, proteins, and specialized metabolites) and root-associated cap-derived cells, or AC-DC [bib_ref] Root border cells and secretions as critical elements in plant host defense, Driouich [/bib_ref] [bib_ref] Root extracellular traps versus neutrophil extracellular traps in host defence, a case..., Driouich [/bib_ref]. Indeed, exDNA has been highlighted in the RET of two Fabaceae: pea [bib_ref] Extracellular DNA Is Required for Root Tip Resistance to Fungal Infection, Wen [/bib_ref] and soybean [bib_ref] Root Border Cells and Mucilage Secretions of Soybean, Glycine Max (Merr) L.:..., Ropitaux [/bib_ref]. This DNA might play a stuctural role in the root trap and enables to lyse bacterial cells [bib_ref] Extracellular DNases of Ralstonia solanacearum modulate biofilms and facilitate bacterial wilt virulence:..., Tran [/bib_ref]. Like NET exDNA, a study shows that plant exDNA within the RET may be synthesized and exported by living cells [bib_ref] Visualization of extracellular DNA released during border cell separation from the root..., Wen [/bib_ref].
Although exDNA seems to be a key player in root defense [bib_ref] Root extracellular traps versus neutrophil extracellular traps in host defence, a case..., Driouich [/bib_ref] [bib_ref] Root Border Cells and Their Role in Plant Defense, Hawes [/bib_ref] , there are still many issues about its origin and modulation in the RET during plant immunity processes. A first sequence analysis was made with pea exDNA random clones in 2009 [bib_ref] Extracellular DNA Is Required for Root Tip Resistance to Fungal Infection, Wen [/bib_ref] , but no high-throughput genomic sequencing has ever been done and, till now, no analysis has been conducted on soybean.
Herein, in order to explore the hypothesis of exDNA occurring from living cells or cell lysis as for NET [bib_ref] Root extracellular traps versus neutrophil extracellular traps in host defence, a case..., Driouich [/bib_ref] , we provide the first DNA-seq of plant exDNA. Moreover, we investigated here if soybean root elicitation by PEP-13, a peptide from oomycete genus Phytophthora [bib_ref] Pep-13, a plant defense-inducing pathogen-associated pattern from Phytophthora transglutaminases, Brunner [/bib_ref] known to induce soybean root rot, does alter RET exDNA sequence.
# Materials and methods
## Plant material, pep-13 elicitation, and exdna separation
Glycine max (L.) Merr. seeds from variety Castetis (La Dauphinoise, Vienne, France) were sterilized overnight twice in chlorine gas and grown on liquid [bib_ref] Extracellular histones identified in crocodile blood inhibit in-vitro HIV-1 infection, Kozlowski [/bib_ref] 2 MS medium, supplemented with 2 µg/mL of peptide PEP-13 for elicited plants, at 26 - C (80-90% relative humidity (RH), 16 h to 8 h day and night cycle), in a growth chamber for five days. The peptide PEP-13 (VWNQPVRGFKVYE) used for this experiment was synthesized on the facility PRIMACEN (Université de Rouen Normandie, France).
Soybean RETs were collected aseptically by manual agitation of the roots in 400 µL of sterile DNAse-free water (Water, Sterile Nuclease-free, TFS Fisher BioReagents). exDNA was separated from the root AC-DC (root-associated cap-derived cells; [bib_ref] Root extracellular traps versus neutrophil extracellular traps in host defence, a case..., Driouich [/bib_ref] by centrifugation (10 min at 3000 g). Supernatant was collected and incubated for 10 min at 100 - C and then frozen. The absence of microbial contamination was surveyed on LB medium incubated at 24 and 37 - C for 48 h. After 48 h, only sterile samples were used and treated for 30 min at 60 - C with Proteinase k (Qiagen).
## Dna purification and sequencing
Each sample (300 plants/sample) was concentrated to a minimal volume (~500 µL) with a speedvac (Thermo Scientific, Waltham, MA, USA). Purification with DNeasy ® PowerClean ® Pro Cleanup kit (Qiagen, Hilden, Germany) was done on 100 µL of each sample. DNA samples were sonicated for 300 s (S220-Series Covaris) to obtain 200 pb fragments. Samples were visualized with a TapeStation 2200 (Agilent, Santa Clara, CA, USA) and High Sensitivity D1000 ScreenTape System kit (Agilent). Libraries were constructed using the NEBNext ® Ultra™ II kit (NEB) and 1 ng of DNA and sequenced on a MISeq System (Illumina) using 2 × 150 bases paired-end sequencing (V2 Micro 300 cycles kit, Illumina) on the Service Commun de Génomique IRIB-Inserm U1245 (Rouen, France).
## Bioinformatic analyses
Sequenced reads were aligned to the Glycine max genome Wiliams 82 Assembly 4 Annotation 1 (Wm82.a4.v1) (https://phytozome.jgi.doe.gov/) using BWA Aligner (V0.7.17) and GATK (V4.0.6.0) [bib_ref] The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing..., Mckenna [/bib_ref] to obtain binary alignment files and indexes. These files were visualized with Integrative Genomic Viewer (V2.6.3) [bib_ref] A Bioconductor package for differential expression analysis of digital gene expression data, Robinson [/bib_ref]. The coverage for total, organelles, and nuclear DNA was calculated with the following formula: X = (n × l)/L; where n = number of reads, l = reads length (Mb), and L = genome length (Mb).
Genomic sequence was divided in 1000 pb sequences and the number of reads per 1000 pb sequence was calculated with Bedtools (V2.22) [bib_ref] BEDTools: A flexible suite of utilities for comparing genomic features, Quinlan [/bib_ref] using the coverage command. A differential analysis was done with DESeq2 (V2.11.40.6) [bib_ref] Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2, Love [/bib_ref] on these 1000 pb parts. Volcano plots and read alignment analysis were made with Excel. Sequences identification was done with NCBI (National Center for Biotechnology Information) Nucleotide databank and nucleotide BLAST tool, against the "plants" organisms' database of NCBI.
# Results
## Differences between nuclear and organelle dna coverage
In order to find out the origin of RET soybean exDNA, we sequenced this DNA using a next-generation sequencing approach (MiSeq, Illumina). In total, we generated 1.7 Gb reads from six samples of 300 plant RET; a total of 6.786 million reads were mapped to the soybean whole genome including nuclear (20 chromosomes) and organelles (plastid and mitochondria) sequences. For each sample, we had an average of 1,224,885 reads mapped on the whole genome. Among them, an average of 47,543 reads were mapped on organelles (mitochondria and plastid) DNA and an average of 1,177,342 reads were mapped on chromosomes. Using IGV (integrative genomic viewer), we examined the alignment of the reads [fig_ref] Figure 1: Examples of integrative genomic viewer [/fig_ref].
# Results
## Differences between nuclear and organelle dna coverage
In order to find out the origin of RET soybean exDNA, we sequenced this DNA using a next-generation sequencing approach (MiSeq, Illumina). In total, we generated 1.7 Gb reads from six samples of 300 plant RET; a total of 6.786 million reads were mapped to the soybean whole genome including nuclear (20 chromosomes) and organelles (plastid and mitochondria) sequences. For each sample, we had an average of 1,224,885 reads mapped on the whole genome. Among them, an average of 47,543 reads were mapped on organelles (mitochondria and plastid) DNA and an average of 1,177,342 reads were mapped on chromosomes. Using IGV (integrative genomic viewer), we examined the alignment of the reads [fig_ref] Figure 1: Examples of integrative genomic viewer [/fig_ref]. Mitochondrial and plastid sequences seemed to be well and uniformly covered by exDNA reads. Furthermore, replicates did not show any differences in alignment. By contrast, nuclear sequence showed very few alignments of reads, which seemed to be randomly aligned between the different replicates.
Thus, this analysis showed a high difference in coverage of exDNA between nuclear DNA and organelles DNA. Indeed, the number of reads per mega base is surprisingly higher for plastidial and mitochondrial DNA compared with nuclear DNA (nDNA). To validate this observation, we calculated the coverage of nDNA and organelles DNA by the exDNA [fig_ref] Table 1: Soybean variety Castetis extracellular DNA counts and coverage for nuclear DNA, mitochondrial... [/fig_ref]. This calculation supported our previous observations [fig_ref] Figure 1: Examples of integrative genomic viewer [/fig_ref] , indicating that the coverage is different between organelles and nuclear DNA. Actually, organelles' coverage is more than seventy times higher for mitochondria (17.88 X) and sixty-two times higher for plastids 70 X) than nDNA (0.25 X). Mitochondrial and plastid sequences seemed to be well and uniformly covered by exDNA reads. Furthermore, replicates did not show any differences in alignment. By contrast, nuclear sequence showed very few alignments of reads, which seemed to be randomly aligned between the different replicates.
Thus, this analysis showed a high difference in coverage of exDNA between nuclear DNA and organelles DNA. Indeed, the number of reads per mega base is surprisingly higher for plastidial and mitochondrial DNA compared with nuclear DNA (nDNA). To validate this observation, we calculated the coverage of nDNA and organelles DNA by the exDNA [fig_ref] Table 1: Soybean variety Castetis extracellular DNA counts and coverage for nuclear DNA, mitochondrial... [/fig_ref]. This calculation supported our previous observations [fig_ref] Figure 1: Examples of integrative genomic viewer [/fig_ref] , indicating that the coverage is different between organelles and nuclear DNA. Actually, organelles' coverage is more than seventy times higher for mitochondria (17.88 X) and sixty-two times higher for plastids 70 X) than nDNA (0.25 X).
On the other hand, it has been shown that exDNA is implicated in plant root defense [bib_ref] Root extracellular traps versus neutrophil extracellular traps in host defence, a case..., Driouich [/bib_ref] [bib_ref] Extracellular DNA Is Required for Root Tip Resistance to Fungal Infection, Wen [/bib_ref] [bib_ref] Extracellular DNA: The tip of root defenses?, Hawes [/bib_ref] , suggesting that it could be modified in response to pathogens or elicitors. Thus, we investigated how exDNA respond to an elicitation with PEP-13.
## Impact of pep-13 elicitation on exdna sequence
In order to investigate the potential changes of exDNA during a defense response, we also sequenced exDNA from elicited soybean seedlings with PEP-13, a peptide elicitor from the oomycete Phytophthora megasperma sp. glycinea [bib_ref] Different cell-waU components from Phytophthora megasperma f. sp. glycinea elicit phytoalexin production..., Parker [/bib_ref] , now known as P. sojae. Soybean whole genome (nuclear and organelles) was divided into 1000 pb parts and exDNA sequences were aligned on these parts. Then, a differential analysis was done with DESeq2 to obtain a fold change and a p-value for each 1000 pb part. These results are shown in two volcano plots, one for nDNA [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref] and the other for organelles DNA [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref]. On the other hand, it has been shown that exDNA is implicated in plant root defense [bib_ref] Root extracellular traps versus neutrophil extracellular traps in host defence, a case..., Driouich [/bib_ref] [bib_ref] Extracellular DNA Is Required for Root Tip Resistance to Fungal Infection, Wen [/bib_ref] [bib_ref] Extracellular DNA: The tip of root defenses?, Hawes [/bib_ref] , suggesting that it could be modified in response to pathogens or elicitors. Thus, we investigated how exDNA respond to an elicitation with PEP-13.
## Impact of pep-13 elicitation on exdna sequence
In order to investigate the potential changes of exDNA during a defense response, we also sequenced exDNA from elicited soybean seedlings with PEP-13, a peptide elicitor from the oomycete Phytophthora megasperma sp glycinea [bib_ref] Different cell-waU components from Phytophthora megasperma f. sp. glycinea elicit phytoalexin production..., Parker [/bib_ref] , now known as P. sojae. Soybean whole genome (nuclear and organelles) was divided into 1000 pb parts and exDNA sequences were aligned on these parts. Then, a differential analysis was done with DESeq2 to obtain a fold change and a p-value for each 1000 pb part. These results are shown in two volcano plots, one for nDNA [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref] and the other for organelles DNA [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref]. mitochondrial and plastidial DNA merged as "organelles DNA" owing to their similarity. A,B: 1000 pb parts of the soybean genomic sequence are represented by a point. Each point is placed according to its Log2 of fold change and the absolute value of the Log10 of its p-value. A color is assigned to every point depending on its fold change and p-value. Grey points correspond to sequences with no significant differences between PEP-13 elicited and control conditions, blue points correspond to sequences that are under-represented in control condition (fold change < 0.5 (log2 fold change < −1), p-value < 0.05), and red points correspond to sequences that are over-represented in PEP- This analysis shows no significant differences in the alignment of exDNA reads on organelles DNA. By contrast, alignment on nDNA revealed some significant differences between exDNA of elicited and control seedlings [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref]. Indeed, we obtained 1723 parts of 1000 pb nDNA that are over-represented (FC > 2; p-value < 0.05), and 2393 parts of 1000 pb nDNA that are under-represented (FC < 0.5; p-value < 0.05). Despite the low coverage of nDNA, it appeared that this part of exDNA could be affected by the elicitation with PEP-13. In order to better understand this potential response to PEP-13, we further investigated these differentially represented sequences [fig_ref] Table 2: Identification of differentially represented 1000 pb parts of nuclear DNA in exDNA... [/fig_ref]. Grey points correspond to sequences with no significant differences between PEP-13 elicited and control conditions, blue points correspond to sequences that are under-represented in control condition (fold change < 0.5 (log2 fold change < −1), p-value < 0.05), and red points correspond to sequences that are over-represented in PEP-13.
This analysis shows no significant differences in the alignment of exDNA reads on organelles DNA. By contrast, alignment on nDNA revealed some significant differences between exDNA of elicited and control seedlings [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref]. Indeed, we obtained 1723 parts of 1000 pb nDNA that are over-represented (FC > 2; p-value < 0.05), and 2393 parts of 1000 pb nDNA that are under-represented (FC < 0.5; p-value < 0.05). Despite the low coverage of nDNA, it appeared that this part of exDNA could be affected by the elicitation with PEP-13. In order to better understand this potential response to PEP-13, we further investigated these differentially represented sequences [fig_ref] Table 2: Identification of differentially represented 1000 pb parts of nuclear DNA in exDNA... [/fig_ref]. Here, we show twenty 1000 pb parts with the ten most important FC (over-represented, FC > 2, p-value < 0.05) and the ten lowest FC (under-represented, FC < 0.5, p-value < 0.05) [fig_ref] Table 2: Identification of differentially represented 1000 pb parts of nuclear DNA in exDNA... [/fig_ref]. Among these twenty 1000 pb parts, a mix of coding and non-coding sequences was found.
However, these sequences are only counted 7 to 9 times in their highest-detected condition, and 0 times in their lowest-detected condition. These counts seemed to be really low, and despite their associated p-value and fold change, we assume that PEP-13 elicitation has a little or no effect in terms of the number of exDNA reads alignment.
In order to examine alignment profiles, we realized read alignment graphics for each chromosome and organelle. Here, we show the example of chromosome 19 [fig_ref] Figure 3: Reads' alignment along chromosome 19 [/fig_ref] and mitochondria alignment [fig_ref] Figure 3: Reads' alignment along chromosome 19 [/fig_ref] , which are representative of the other chromosomes and organelles, in control (dark blue) and PEP-13 elicited (yellow) conditions. For the nuclear sequence, we highlighted a baseline of nearly zero counts all along the chromosome sequence, which is similar between the two conditions. Secondly, we can see some 1000 pb parts showing higher alignment counts (almost 400 for the 24,013,001-24,014,000 pb part in control condition), which seemed slightly lower in the elicited condition than in the control condition. This 1000 pb part was not detected previously as a differentially represented part (FC > 2; p-value < 0.05) because there was no significant change between PEP-13 elicited and control conditions. For organelles' alignments, we highlighted a high baseline (around 200 counts) all along the sequence, which also seemed slightly lower in the elicited condition compared with the control condition. According to the previous differential analysis [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref] , there was no significant change between the two conditions. and mitochondria alignment [fig_ref] Figure 3: Reads' alignment along chromosome 19 [/fig_ref] , which are representative of the other chromosomes and organelles, in control (dark blue) and PEP-13 elicited (yellow) conditions. For the nuclear sequence, we highlighted a baseline of nearly zero counts all along the chromosome sequence, which is similar between the two conditions. Secondly, we can see some 1000 pb parts showing higher alignment counts (almost 400 for the 24,013,001-24,014,000 pb part in control condition), which seemed slightly lower in the elicited condition than in the control condition. This 1000 pb part was not detected previously as a differentially represented part (FC > 2; p-value < 0.05) because there was no significant change between PEP-13 elicited and control conditions. For organelles' alignments, we highlighted a high baseline (around 200 counts) all along the sequence, which also seemed slightly lower in the elicited condition compared with the control condition. According to the previous differential analysis [fig_ref] Figure 2: Volcano plots of exDNA from soybean var [/fig_ref] , there was no significant change between the two conditions. A total of eleven 1000 pb parts from nuclear DNA with a count higher than 100 in every samples (three PEP-13 elicited and three control conditions) were found in the soybean genome. These parts with a higher alignment of reads are detailed in . . Identification of 1000 pb parts of nuclear DNA in exDNA from soybean var. Castetis with a higher read alignment (count > 100). N/A = non-coding sequence. A total of eleven 1000 pb parts from nuclear DNA with a count higher than 100 in every samples (three PEP-13 elicited and three control conditions) were found in the soybean genome. These parts with a higher alignment of reads are detailed in . . Identification of 1000 pb parts of nuclear DNA in exDNA from soybean var. Castetis with a higher read alignment (count > 100). N/A = non-coding sequence.
## Chromosome
Position ( In , we show that the eleven 1000 pb parts of nuclear DNA with a high number of reads aligned are a mix of coding and non-coding sequences. Furthermore, these sequences are mainly repeated sequences and/or identified only in soybean or Fabaceae.
# Discussion
This study is a pioneer high-sequencing approach of plant exDNA performed in soybean RET. Sequence analysis was successively done in control condition and after PEP-13 elicitation. For the control condition, reads aligned on soybean genomic DNA (nuclear and organelles DNA) suggested that soybean RET exDNA is constitutively a mix of coding and non-coding sequences from nDNA and organelles DNA. A very similar result was observed [bib_ref] Extracellular DNA Is Required for Root Tip Resistance to Fungal Infection, Wen [/bib_ref] with some identified sequences of pea exDNA.
In our study, we registered a higher number of reads aligned on nDNA than on mitochondrial/plastidial DNA. Nevertheless, considering the size of each genomic DNA (942.21 Mb for nuclear DNA and 0.55 Mb for organelles DNA), we show in control condition that the coverage is seventy times higher for mitochondria and plastids (17.29 X) compared with nDNA (0.25 X). This difference might be explained by the number of organelles in plant cells, and especially in root AC-DC, which appears to contains a high number of organelles [bib_ref] Root Border-Like Cells of Arabidopsis. Microscopical Characterization and Role in the Interaction..., Vicré [/bib_ref] [bib_ref] Association between border cell responses and localized root infection by pathogenic Aphanomyces..., Cannesan [/bib_ref]. A sequencing of roots' AC-DC cellular DNA (including both nuclear and organelles DNA) might indicate if the number of organelles in these cells could itself explain this coverage difference.
Moreover, two other assumptions could also explain this high difference; that is, organelles DNA could be better preserved in the RET than nDNA or the major origin of RET extracellular DNA could be from organelles.
In favor of the first hypothesis, it is known that DNA degradation in the soil is dependent on extrinsic and intrinsic properties [bib_ref] Factors governing extracellular DNA degradation dynamics in soil, Sirois [/bib_ref]. In our experimental design, as nuclear and organelles DNA sequences are in the same extracellular matrix (the RET), we assumed that extrinsic properties are not responsible for the preservation differences of exDNA. Among intrinsic properties that could affect DNA persistence in the RET, percentage of GC, DNA molecular weight and conformation, or methylation could occur [bib_ref] Factors governing extracellular DNA degradation dynamics in soil, Sirois [/bib_ref] [bib_ref] Physical and Topological Properties of Circular DNA, Vinograd [/bib_ref] [bib_ref] Base-stacking and base-pairing contributions into thermal stability of the DNA double helix, Yakovchuk [/bib_ref].
When comparing the GC% of nDNA (35% GC; [bib_ref] Genome sequence of the palaeopolyploid soybean, Schmutz [/bib_ref] , mitochondrial DNA (45% GC; [bib_ref] The Mitochondrial Genome of Soybean Reveals Complex Genome Structures and Gene Evolution..., Chang [/bib_ref] , and plastidial DNA (34% GC; [bib_ref] Complete Chloroplast Genome Sequence of Glycine max and Comparative Analyses with other..., Saski [/bib_ref] sequences, some differences appear between nDNA and mitochondrial DNA GC%, but not between nDNA and plastidial DNA. Thus, the GC% intrinsic property could explain a better persistence of mitochondrial DNA, but not of plastidial DNA, in comparison with nDNA.
DNA molecular weight, including sequence length, may allow a better stability of exDNA. Molecular weight is directly correlated to soil adsorption of DNA [bib_ref] DNA adsorption to soils and sediments, Ogram [/bib_ref]. In the RET and in our in vitro conditions, exDNA could be adsorbed by a positively charged surfaceor a positively charged component of the RET like histones [bib_ref] Comparative analysis of HD2 type histone deacetylases in higher plants, Dangl [/bib_ref] , increasing its conservation. Currently, there is no study comparing exDNA sequence sizes depending on their origin (nuclear or organelles DNA), which could explain the differences of nDNA and organelles DNA conservation.
Another explanation of these conservation differences could be the fact that plastids and mitochondrial DNA are both circular DNA, in contrast with nDNA, which is linear. Indeed, circular DNA is known to be more stable than linear DNA [bib_ref] Physical and Topological Properties of Circular DNA, Vinograd [/bib_ref]. Regarding this property, organelles DNA degradation in the RET might be longer than nDNA degradation, resulting in these conservation differences.
ExDNA could be fragmented during cellular lysis of root tip cells as a result of DNAse activity. It is known that DNAse activity could be both enhanced or inhibited by DNA methylation in a misunderstood way. Thus, a difference in DNA methylation between chromosomes and organelles could explain conservation differences. In plants, DNA methylation is species-and organelles-specific [bib_ref] Distribution of 5-methylcytosine residues in 5S rRNA genes in Arabidopsis thaliana and..., Fulneček [/bib_ref]. Two major methylation types are found in plants DNA: m5C (5-methylcytosine) and m6A (N6-methyladenine). In mitochondrial DNA, only m6A is found, unlike plastids and nuclear DNA, which contain both of these types of methylation. Moreover, recently, m6A methylation density has been investigated in soybean, showing 1.4% for plastid DNA, 1.05% for mitochondrial DNA, and only 0.05% for nDNA. Given this hypothesis, an exDNA methylation analysis might be done in order to understand if methylation processes are involved in the conservation of exDNA sequences.
Our second hypothesis is that exDNA could mainly originate from mitochondrial and plastids. Indeed, it has been shown in NET that exDNA could originate from mitochondria [bib_ref] Viable neutrophils release mitochondrial DNA to form neutrophil extracellular traps, Yousefi [/bib_ref] [bib_ref] Neutrophil extracellular traps enriched in oxidized mitochondrial DNA are interferogenic and contribute..., Lood [/bib_ref]. In this case, NET formation does not induce neutrophil death (vital NETosis). It is thus possible to imagine that RET formation could consist of the programmed DNA release from organelles. In this assumption, mitochondrial and plastidial DNA would be released in the extracellular matrix without being degraded during cell death. This could explain the high conservation and coverage of organelles DNA. Nuclear DNA in the RET, for its part, could be explained by the release of fragmented DNA during the death of the older root cells. Furthermore, this hypothesis is consistent with the results obtained in the literature [bib_ref] Extracellular DNA Is Required for Root Tip Resistance to Fungal Infection, Wen [/bib_ref]. Indeed, they observed that exDNA release in pea RET seemed to be newly synthetized DNA. Moreover, plastids division is not dependent on the cell cycleand could occur regularly (from 1 to 4 h in wheat)thanks to specific regulation mechanisms [bib_ref] The PLASTID DIVISION1 and 2 Components of the Chloroplast Division Machinery Determine..., Okazaki [/bib_ref] [bib_ref] PDV2 has a dosage effect on chloroplast division in Arabidopsis, Chang [/bib_ref].
Considering nDNA, we recorded a particular profile of read alignment . While read alignment showed a low coverage on chromosomes, eleven 1000 pb parts had a significantly higher number of counts (>100 counts for each replicate and experimental condition). Thus, it seemed that some parts of the nDNA are better conserved than others. Among these 1000 pb parts, we found sequences that are specific to the soybean or Fabaceae genomes. Among these, three are also involved in plant defense response, which strengthened the importance of exDNA in root defense. We also found several repeated sequences with a higher read depth. This could reflect a variation in the copy number of the soybean variety used in this study compared with the reference genome used [bib_ref] Genome-wide detection of structural variants and indels by local assembly, Wala [/bib_ref]. Indeed, the soybean variety used for this study (Castetis) was not the same as the genome reference variety (Williams 82). Thus, it could be possible that Castetis has more repeated sequences in its genome as a result of varietal selection. Indeed, pathogen or lodging resistances are significant criteria for varietal selection because of the sub-functionalization reducing the gene pleiotropy and responsible for duplication of genes [bib_ref] Gene Duplication: The Genomic Trade in Spare Parts, Hurles [/bib_ref].
Another explanation of the high number of these reads is that these sequences could be recognized as DAMPs (damage-associated molecular patterns). Indeed, plant genomic DNA is a DAMP and plants seem to be able to recognize its own DNA among the DNA of another species [bib_ref] Extracellular self-DNA as a damage-associated molecular pattern (DAMP) that triggers self-specific immunity..., Duran-Flores [/bib_ref]. In our study, we identified some sequences specific of the Glycine max genome . These sequences could be recognized by the root as DAMPs in the RET and initiate a plant defense response, improving plant resistance to pathogens, making exDNA an endogenous elicitor. In order to confirm this hypothesis, it might be interesting to test if these exDNA sequences are able to induce plant defense response.
The second part of our study was to observe exDNA during a defense response to PEP-13 elicitation. Even if our differential analysis seemed to show some differences for the nDNA sequences, we assume that it was not a significant variation, considering the really low counts of these sequences. Two hypotheses could explain these results: exDNA alignment might not be influenced by PEP-13 elicitation, or exDNA might not be affected by any elicitation. Regarding these results and the hypothesis, elicitation tests should be done with other elicitors (β-glucan elicitor, others peptides, lipopolysaccharides, and so on) in order to find out if exDNA alignment could be influenced by elicitation.
In summary, our findings reveal that exDNA is constitutively a mix of coding and non-coding sequences from mainly organelles DNA and, in the minority, from nuclear DNA, which are not modified by PEP-13 elicitation. Further investigations are now needed to know if methylation processes are involved in the conservation of exDNA sequences, if specific nuclear sequences of exDNA could be recognized by the root as a DAMP, or if other MAMPs could significantly alter RET exDNA sequences.
# Data availability statement:
The data presented in this study are openly available in the European Nucleotide Archive (ENA) at http://www.ebi.ac.uk/ena/data/view/PRJEB42337.
[fig] Figure 1: Examples of integrative genomic viewer (IGV) views of mitochondrial, plastidial, and nuclear DNA coverages with exDNA from soybean var. Castetis. For each organelle and chromosome, a representative 50 Kb part of the total sequence is shown. For nuclear DNA, the view corresponds to the position 4,527,274 pb to 4,577,274 pb on the first chromosome (representative example of all chromosomes). For plastid DNA, the view corresponds to the position 55 pb to 50,055 pb. For the mitochondrial DNA, the view corresponds to the position 36,375 pb to 86,375 pb. R = replicate. [/fig]
[fig] Figure 2: Volcano plots of exDNA from soybean var. Castetis differentially aligned in 1000 pb parts. A: Nuclear DNA, B: [/fig]
[fig] Cells 2021 ,: 10, x FOR PEER REVIEW 6 of 12 [/fig]
[fig] Figure 3: Reads' alignment along chromosome 19 (A) and mitochondria (B) in PEP-13 elicited and control conditions. For each 1000 pb parts of the chromosome 19 and mitochondria (representative of chromosomes and organelles, respectively), the number of reads is represented on this graphic. PEP-13 elicited (yellow) and control condition (dark blue) have been overlapped in order to show their similarity. Reads were obtained from soybean var. Castetis exDNA. [/fig]
[fig] Author: Contributions: Writing-Original Draft Preparation, M.C., I.B. and M.-L.F.-G.; Writing-Review & Editing, M.C., I.B., M.-L.F.-G., C.P., I.T., M.-C.K.-M., A.D., J.L. and B.L.; Investigation, M.C., C.P., I.B., I.T., S.C., C.D. and M.-L.F.-G.; Funding Acquisition, I.B., A.D., M.-L.F.-G.; and J.L., B.L. provided the PEP-13 elicitor. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by La Région Normandie, RIN EPURE 2017/19 project. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. [/fig]
[table] Table 1: Soybean variety Castetis extracellular DNA counts and coverage for nuclear DNA, mitochondrial DNA, plastidial DNA, and total DNA. Coverage has been calculated according to the formula explained in Material and Methods section. [/table]
[table] Table 2: Identification of differentially represented 1000 pb parts of nuclear DNA in exDNA from soybean var. Castetis. E = PEP-13 elicited condition; NE = control condition; N/A = non-coding sequence. [/table]
|
Effects of Continuous Hepatitis with Persistent Hepatitis C Viremia on Outcome after Resection of Hepatocellular Carcinoma
The effect of persistent hepatitis C viremia on the outcome after resection of hepatocellular carcinoma (HCC) was investigated in 59 consecutive patients with a single small HCC (≤ ≤ ≤ ≤3.0 cm in diameter). The presence of serum hepatitis C virus (HCV) RNA was evaluated using a reverse transcription polymerase chain reaction method as well as a branched DNA probe method. Clinicopathologic findings were compared between patients with and without viremia and the risk factors for poor outcome were evaluated. Hepatitis C virus (HCV) RNA was not detected in the sera from 7 patients (group 1), but was detected in the sera from the other 52 patients (group 2). Alanine aminotransferase (ALT) activity was significantly higher in group 2 than in group 1. The proportion of patients with active hepatitis was significantly higher in group 2. In group 2, new HCC often developed after the operation and four patients died of liver dysfunction. HCV viremia, high ALT activity, high concentration of total bilirubin, and liver cirrhosis were related to recurrence after the operation. Multivariate analysis indicated that HCV viremia and high ALT activity were independent risk factors for recurrence of HCC. Continuous hepatitis with persistent HCV viremia worsened the outcome after the resection of HCC by causing new development of HCC and deterioration of liver function. In patients with HCV-related HCC, but without HCV viremia, satisfactory results can be expected after liver resection.
Despite improvements in medical imaging, surgical techniques, and perioperative management, outcomes after liver resection for hepatocellular carcinoma (HCC) still are unsatisfactory because of a high recurrence rate.There have been many studies concerning factors related to recurrence and outcome after resection of HCC. [bib_ref] Factors linked to early recurrence of small hepatocellular carcinoma after hepatectomy: univariate..., Shirabe [/bib_ref] [bib_ref] Intrahepatic recurrence after resection of hepatocellular carcinoma complicating cirrhosis, Belghiti [/bib_ref] [bib_ref] Risk factors linked to tumor recurrence of human hepatocellular carcinoma after hepatic..., Jwo [/bib_ref] [bib_ref] Predictive factors for postoperative recurrence of hepatocellular carcinoma, Okada [/bib_ref] [bib_ref] Patterns of recurrence after initial treatment in patients with small hepatocellular carcinoma, Kumada [/bib_ref] [bib_ref] Risk factors for tumor recurrence and prognosis after curative resection of hepatocellular..., Ikeda [/bib_ref] [bib_ref] Long-term results after resection of hepatocellular carcinoma: experience of 480 cases, Kosuge [/bib_ref] [bib_ref] Prognostic factors of hepatocellular carcinoma in patients undergoing hepatic resection, Izumi [/bib_ref] [bib_ref] Surgical treatment and prognostic variables of hepatocellular carcinoma in 122 cirrhotics, Carlo [/bib_ref] [bib_ref] Prognostic significance of pathologic features of hepatocellular carcinoma: a multivariate analysis of..., Ng [/bib_ref] [bib_ref] Hepatocellular carcinoma and cirrhosis: results of surgical treatment in a European series, Fuster [/bib_ref] In some of these reports, hepatitis viral status was not evaluated. In others, hepatitis C virus (HCV) infection was determined by whether the patients had anti-HCV antibodies in their sera before the operation. The subjects in these studies thus might have included patients without HCV viremia at the time of or after the operation.
In patients infected with HCV, cirrhosis with chronic inflammation, liver cell necrosis and regeneration, and extensive fibrosis are important in the development of HCC. [bib_ref] Detection of replicative hepatitis C virus sequences in hepatocellular carcinoma, Gerber [/bib_ref] [bib_ref] Molecular biology of the hepatitis C viruses: implications for diagnosis, development and..., Houghton [/bib_ref] [bib_ref] Current pathogenetic and molecular concepts in viral liver carcinogenesis, Schirmacher [/bib_ref] [bib_ref] Transcriptional regulation of cellular and viral promoters by the hepatitis C virus..., Ray [/bib_ref] HCC tends to develop or to become detectable when DNA synthesis increases in patients with HCVassociated cirrhosis. Increased DNA synthesis in hepatocytes also appears to accelerate recurrence after resection for HCC in patients with HCV-associated cirrhosis. [bib_ref] Role of increased DNA synthesis activity of hepatocytes in multicentric hepatocarcinogenesis in..., Tarao [/bib_ref] We have found that a high spermidine/spermine ratio, which is closely related to cell proliferation, in the noncancerous hepatic tissue is a risk factor for recurrence after liver resection for HCC. [bib_ref] Effect of alcohol abuse on polyamine metabolism in hepatocellular carcinoma and noncancerous..., Kubo [/bib_ref] It has also been reported that the sustained elevation of the serum alanine aminotransferase (ALT) activity in HCV infection is closely associated with progression from cirrhosis to HCC, as well as with recurrence after the resection of HCC. [bib_ref] Relationship between the recurrence of hepatocellular carcinoma (HCC) and serum alanine aminotransferase..., Tarao [/bib_ref] These findings indicate that continuous hepatitis caused by persistent HCV viremia strongly affects the outcome after the operation. Recently we found that HCV RNA was not detectable in the sera of some patients with HCC and anti-HCV. [bib_ref] Clinicopathological criteria for multicentricity of hepatocellular carcinoma and risk factors for such..., Kubo [/bib_ref] Benvegnù et al. [bib_ref] Lack of correlation between hepatitis C virus genotypes and clinical course of..., Benvegnù [/bib_ref] have reported that HCV RNA was not detected in some patients with chronic liver disease and anti-HCV, and that persistently high or fluctuating levels of ALT during follow-up were rarely seen in such patients. The natural disappearance rate of serum HCV RNA in patients infected with HCV has been reported to be 2.8% per year in Japan. [bib_ref] Natural disappearance of serum HCV RNA: prospective study in a hyperendemic area, Fukuizumi [/bib_ref] In addition, interferon therapy results in the disappearance of HCV RNA from not only the serum, but also the hepatic tissue. [bib_ref] Effect of interferon administration on serum hepatitis C virus RNA in patients..., Chayama [/bib_ref] [bib_ref] Detection of hepatitis C virus RNA in serum of patients with chronic..., Hagiwara [/bib_ref] [bib_ref] Longterm follow-up of patients with chronic hepatitis C treated with α-interferon, Shindo [/bib_ref] However, persistent HCV viremia has not been studied as a prognostic factor.
The aim of the present study was to investigate the effect of continuous hepatitis with persistent HCV viremia on the clinicopathologic findings in patients with HCC and on the outcome after resection.
# Patients and methods
Patients From April 1990 to June 1997, 204 patients with anti-HCV antibodies underwent liver resection for HCC. Fifty-nine of the 204 patients had a single small HCC (≤3.0 cm in diameter) without portal or hepatic vein invasion. Patients whose resected liver specimens showed portal or hepatic vein invasion on pathologic examination were excluded from the study to avoid instances of intrahepatic metastasis that might be present at the time of the operation. Four of the 59 patients had received interferon therapy before the detection of HCC. Patients were examined preoperatively by ultrasonography, computed tomography (plain and enhanced CT), Lipiodol CT (Ultra-Fluide, Laboratorie Guerbet, Villepointe, France), magnetic resonance imaging (in most patients) and angiography. Direct ultrasonography was performed during the operation. Since May 1993, CT during arteriography and CT during arterioportography have been done if possible. Patients with hepatitis B surface antigen (HBsAg) in their sera were excluded from the study. The patients comprised 44 men and 15 women with ages ranging from 46 to 79 years. The subjects were divided into two groups. Group 1 consisted of patients in whom HCV RNA was not detected in their sera taken before and after operation, and group 2 consisted of patients in whom HCV RNA was detected in their sera. None of the patients received adjuvant therapy or interferon therapy after the operation.
The study was conducted in accordance with the Helsinki Declaration and the guidelines of the ethics committee of our institution. Written informed consent was obtained from each patient. Viral markers Serum samples obtained before and after operation from all patients were assayed for hepatitis B virus (HBV) and HCV. Serum was examined for HBsAg with an enzyme immunoassay (International Reagents Corp., Kobe). Samples were examined for anti-HCV by second-or third-generation ELISA (Ortho Diagnostic Systems, Tokyo). Serum HCV RNA was detected using a polymerase chain reaction with reverse transcription and primers derived from a conserved 5′-untranslated region of the viral genome [bib_ref] Randomised trial of effects of interferonα on incidence of hepatocellular carcinoma in..., Nishiguchi [/bib_ref] as well as a branched DNA probe method (Quantiplex HCV-RNA, Chiron Corp., Emeryville, CA). When HCV RNA was not detected in the sera by these two methods, the results were further confirmed by a single-tube assay kit (Amplicor HCV test, Roche Diagnostic Systems/Nippon Roche Co., Branchburg, NJ). HCV RNA in HCC tissue and noncancerous hepatic tissue, obtained at operation, was also examined using a reported method. [bib_ref] Detection and analysis of replicating hepatitis C virus RNA in hepatocellular carcinoma..., Kurosaki [/bib_ref] Results of laboratory tests Laboratory tests included measurements of serum concentrations of α-fetoprotein (AFP), total bilirubin, and albumin, activities of aspartate aminotransferase (AST) and ALT, platelet count, and 15-min indocyanine green retention test (ICGR. Changes in ALT activity after the operation also were evaluated, using only results measured in our hospital. Values obtained in other hospitals were not used because the methods (kits) used in other hospitals sometimes were different from ours. Detection of recurrence When tumor recurrence was suspected on the basis of tumor marker assays, ultrasonography, CT, or some combination of these, angiography and/or a biopsy under ultrasonographic guidance were employed to make a definitive diagnosis. Pathologic examination The histologic grade of tumor differentiation was assigned using a modificationof the classification by Edmondson and Steiner [bib_ref] Primary carcinoma of the liver: a study of 100 cases among 48,900..., Edmondson [/bib_ref] and Well-differentiated HCC had a high cell density, a high nucleus/cytoplasm ratio, strong cytoplasmic eosinophilia, and an irregular pattern with thin trabeculae; in some tumors, the pattern was pseudoglandular. Portal tracts were found within the cancerous tissue. There was replacement growth at the boundary between the tumor and healthy tissue. Noncancerous tissues were also examined pathologically. A histology activity index (HAI) score [bib_ref] Classification of chronic hepatitis: diagnosis, grading and staging, Desmet [/bib_ref] was used to evaluate the severity of hepatitis and the degree of fibrosis. At least two pathologists inspected each specimen. Risk factors for poor outcome after operation We evaluated various risk factors and for each, we calculated the relative risk of poor outcome. The variables were selected based on their potential relationship to outcome according to previous studies or our clinical experience. The variables chosen were HCV RNA in the serum (presence or absence), age (<65 or ≥65 years), sex, history of heavy drinking (intake of at least 86 g of ethanol daily for at least 10 years), [bib_ref] High malignancy of hepatocellular carcinoma in alcoholic patients with hepatitis C virus, Kubo [/bib_ref] history of blood transfusion, Child-Pugh classification (A or B), [bib_ref] Transection of the esophagus for bleeding oesophageal varices, Pugh [/bib_ref] serum ALT activity (≤45 IU/liter or >45 IU/liter), serum concentration of total bilirubin (<1.0 mg/dl or ≥1.0 mg/dl), differentiation of the main tumor (well-differentiated or other), the severity of hepatitis [grade ≤1 (no or minimal hepatitis) or grade >1 (mild to severe hepatitis)], the degree of fibrosis [stage ≤3 (no fibrosis to severe fibrosis) or stage 4 (cirrhosis)], and type of resection (anatomic or nonanatomic). Statistics We used Student's t test to analyze differences in age and the Mann-Whitney test to analyze differences in serum concentrations of albumin and total bilirubin, activities of AST and ALT, platelet count, ICGR, and tumor size. The χ 2 or Fisher's exact test was used to compare categorical data between groups. When the changes in ALT activity after the operation were evaluated, the mean ALT activity obtained for each patient after the operation was calculated and the mean of the values obtained for each test in one group was compared with that in the other group. The Wilcoxon rank sum test was then used to assess differences in these means. Tumor-free and cumulative survival rates were calculated using the Kaplan-Meier method and the significance of differences in survival rates between groups was assessed using the log-rank test. Cox's proportional hazards model was used for multivariate analysis. A difference with a P value <0.05 was considered significant.
# Results
Clinicopathologic findings in patients HCV RNA was not detected in the sera obtained from 7 patients before operation or at any follow-up appointment (group 1); therefore these 7 patients did not have persistent HCV viremia. HCV RNA was detected in the sera from the remaining 52 patients before and during the whole obser-vation period (group 2); these 52 patients had persistent HCV viremia. In group 1, 3 of 7 patients had received interferon therapy before the detection of HCC and HCV RNA was not detected in their sera just before the operation. Tests for HCV RNA in the HCC tissue and the noncancerous tissue, which were performed in 4 patients in group 1, were negative. Clinical features and the results of laboratory tests are shown in [fig_ref] Table I: Clinicopathologic Findings and the Results of Laboratory Tests in Patients with and... [/fig_ref]. Ages ranged from 46 to 79 years in group 1 and from 51 to 64 years in group 2. There was no significant difference between the two groups in the mean age, sex distribution, history of blood transfusion, or history of alcohol abuse. No significant differences were noted between the two groups in Child-Pugh score, proportion of patients with elevated AFP concentration (greater than 20 ng/ml), or tumor size. Results of laboratory tests are given as medians and 10th and 90th percentiles (in parentheses). AST, aspartate aminotransferase; ALT, alanine aminotransferase; ICGR 15 , 15-min indocyanine green retention test. Activity of hepatitis and degree of fibrosis were evaluated according to histologic activity index. [bib_ref] Classification of chronic hepatitis: diagnosis, grading and staging, Desmet [/bib_ref] The activities of AST and ALT and serum concentration of total bilirubin were significantly higher in group 2 than in group 1 (P=0.0244, P=0.0197, and P=0.0012, respectively). The serum concentration of albumin was significantly lower in group 2 than in group 1 (P=0.0034). Platelet count and ICGRwere not significantly different between the two groups. Although differentiation of the main tumor and the degree of fibrosis (staging) were not significantly different between the two groups, the proportion of patients with no or minimal hepatitis (grade 0 or 1) was significantly higher in group 1 than in group 2 (P=0.0033). Although the mean serum ALT activity after the operation in group 1 was within the reference range (≤45 IU/ liter), the mean in group 2 was higher than the reference range (>45 IU/liter, [fig_ref] Figure 1: Changes in serum alanine aminotransferase activity after operation [/fig_ref]. The mean serum ALT activity after the operation was significantly higher in group 2 than in group 1 (P=0.0051). Outcome after operation and factors related to recurrence and outcome In group 1, recurrence was seen in the remaining liver of one patient, who died 3 years and 3 months after the operation. Recurrence in the remaining liver was noted in 36 patients in group 2. In 8 of these 36 patients, pathologic examination of the recurrent tumors obtained by a biopsy or second resection was possible. The recurrent tumors in 7 of the 8 patients included a component of well-differentiated HCC, suggesting that the recurrent tumors were newly developed HCC. [bib_ref] Patterns of recurrence after initial treatment in patients with small hepatocellular carcinoma, Kumada [/bib_ref] [bib_ref] Different pattern of chromosomal allele loss in multiple hepatocellular carcinomas as evidence..., Tsuda [/bib_ref] [bib_ref] Clinicopathologic comparison between resected hepatocellular carcinomas (HCC) and recurrent tumors: a special..., Okuda [/bib_ref] [bib_ref] Patterns of and risk factors for recurrence after live resection for well-differentiated..., Kubo [/bib_ref] Tumor-free survival rates are shown in [fig_ref] Table I: Clinicopathologic Findings and the Results of Laboratory Tests in Patients with and... [/fig_ref]. A history of blood transfusion (P=0.0082), presence of HCV viremia (group 2, P=0.0027, [fig_ref] Figure 2: Tumor-free survival rates after liver resection in patients with and without serum... [/fig_ref] , high activity of ALT (P=0.0303), high concentration of total bilirubin (P= 0.0192), and liver cirrhosis (stage 4 fibrosis, P=0.0472) were significantly related to recurrence after operation. Child-Pugh score B (P=0.0621) and mild to severe hepatitis (grade 2 to 4, P=0.0641) tended to be related to recurrence, but without statistical significance. In the patient with recurrence in group 1, transcatheter arterial embolization was performed for the recurrent tumors. In group 2, second resection was performed in 7 patients, transarterial treatment such as transcatheter arterial embolization and hepatic arterial infusion chemotherapy in 19 patients, percutaneous ethanol injection therapy in 5 patients, and microwave coagulonecrotic therapy in 5 patients. Twenty patients in group 2 died from recurrence, and 4 patients died of liver dysfunction without recurrence. Cumulative survival rates for the two groups are shown in [fig_ref] Table I: Clinicopathologic Findings and the Results of Laboratory Tests in Patients with and... [/fig_ref]. Child-Pugh score B (P=0.0155), high concentration of total bilirubin (P=0.0341), liver cirrhosis (stage 4 fibrosis, P=0.0409), and nonanatomic resection (P= 0.0303) were significantly related to shorter survival time. A history of alcohol abuse (P=0.0881), presence of HCV viremia (group 2, P=0.0983, [fig_ref] Figure 3: Cumulative survival rates after liver resection in patients with and without serum... [/fig_ref] , and mild to severe hepatitis (grade 2 to 4, P=0.0591) tended to be related to shorter survival time, but without statistical significance.
Multivariate analysis indicated that history of blood transfusion, HCV viremia, and high activity of ALT were independent risk factors for recurrence, and Child-Pugh score B was an independent risk factor for shorter survival time [fig_ref] Table I: Clinicopathologic Findings and the Results of Laboratory Tests in Patients with and... [/fig_ref]. Odds ratios of HCV viremia (group 2), compared with absence of HCV viremia (group 1), were 15.87 for recurrence and 4.05 for shorter survival time.
# Discussion
Chronic liver disease, such as active hepatitis or cirrhosis, is generally considered a premalignant condition. Since the advent of the test for anti-HCV antibodies, HCV has been thought to be a major cause of chronic liver disease and HCC in areas where the prevalence of HCV is high, such as Japan and Europe. [bib_ref] Prevalence of antibodies to hepatitis C virus in Spanish patients with hepatocellular..., Bruix [/bib_ref] [bib_ref] Prevalence of antibodies to hepatitis C virus in Italian patients with hepatocellular..., Colombo [/bib_ref] [bib_ref] A high prevalence of antibody to the hepatitis C virus in patients..., Nishioka [/bib_ref] [bib_ref] High prevalence of hepatitis B and C viruses in patients with hepatocellular..., Kubo [/bib_ref] In this study, laboratory test results and the severity of hepatitis were better in group 1 than in group 2, although the degree of fibrosis in the HAI score was not different between the two groups. These findings indicate that, in group 1, liver function and activity of hepatitis had improved after the remission of HCV viremia, although fibrosis had already developed before remission. Recurrence of HCC after the operation may manifest itself as either intrahepatic metastasis from the original tumor or newly developed HCC. [bib_ref] Factors linked to early recurrence of small hepatocellular carcinoma after hepatectomy: univariate..., Shirabe [/bib_ref] [bib_ref] Intrahepatic recurrence after resection of hepatocellular carcinoma complicating cirrhosis, Belghiti [/bib_ref] [bib_ref] Risk factors linked to tumor recurrence of human hepatocellular carcinoma after hepatic..., Jwo [/bib_ref] [bib_ref] Predictive factors for postoperative recurrence of hepatocellular carcinoma, Okada [/bib_ref] [bib_ref] Patterns of recurrence after initial treatment in patients with small hepatocellular carcinoma, Kumada [/bib_ref] [bib_ref] Different pattern of chromosomal allele loss in multiple hepatocellular carcinomas as evidence..., Tsuda [/bib_ref] [bib_ref] Clinicopathologic comparison between resected hepatocellular carcinomas (HCC) and recurrent tumors: a special..., Okuda [/bib_ref] [bib_ref] Patterns of and risk factors for recurrence after live resection for well-differentiated..., Kubo [/bib_ref] Tarao et al. [bib_ref] Relationship between the recurrence of hepatocellular carcinoma (HCC) and serum alanine aminotransferase..., Tarao [/bib_ref] have reported that the recurrence rate after liver resection for HCV-related HCC was higher in patients with high ALT activity (≥80 IU/liter) than in patients with low ALT activity (<80 IU/liter). They suggested that a necroinflammatory process in the hepatocytes is important for new carcinogenesis after the operation. Chiu et al. [bib_ref] Can determination of the proliferative capacity of the nontumor portion predict the..., Chiu [/bib_ref] have reported that a marked increase in the proliferative capacity of the noncancerous hepatic tissue is a significant risk factor for tumor recurrence. In this study, we found that HCV viremia, high activity of ALT (>45 IU/liter), high concentration of total bilirubin (≥1.0 mg/dl), and liver cirrhosis (stage 4 fibrosis) were significant risk factors for recurrence, using univariate analysis. HCV viremia and high activity of ALT were independent risk factors in multivariate analysis. Mild to severe hepatitis (grade 2 to 4) also tended to be related to recurrence. The mean ALT activity was significantly higher in group 2 than in group 1. In addition, some recurrent tumors in group 2 were strongly suspected to be newly developed HCC. These findings suggest that persistent HCV viremia is related to active hepatitis in the adjacent hepatic tissue, resulting in a high risk of hepatocarcinogenesis even after resection of the primary HCC.
Child-Pugh score B, high serum concentration of total bilirubin, and liver cirrhosis (stage 4 fibrosis) were significant risk factors for poor outcome. The three factors are related to advanced liver disease that had already developed before operation. Although nonanatomical resection was not a significant factor for recurrence, it was a significant factor for poor outcome. These findings indicate that new HCC often developed even after anatomical resection and that the patients who underwent anatomical resection might have better remaining liver function. Alcohol abuse, HCV viremia, and mild to severe hepatitis (grade 2 to 4) tended to be related to shorter survival time. We have reported that the outcome after liver resection was significantly poorer in patients with alcohol abuse than in patients without alcohol abuse. [bib_ref] High malignancy of hepatocellular carcinoma in alcoholic patients with hepatitis C virus, Kubo [/bib_ref] Progression of liver dysfunction due to persistent HCV infection seems to be an additional risk factor; in 4 patients in group 2, worsening liver function was the cause of death.
Recent randomized controlled trials have shown that interferon therapy leads to a rapid decrease in ALT activity and to a disappearance of serum HCV RNA in about one-third of patients with chronic hepatitis C. [bib_ref] Randomised trial of effects of interferonα on incidence of hepatocellular carcinoma in..., Nishiguchi [/bib_ref] [bib_ref] Alpha interferon treatment may prevent hepatocellular carcinoma in HCV-related liver cirrhosis, Mazzella [/bib_ref] [bib_ref] Risk factors and the effect of interferon therapy in the development of..., Kuwana [/bib_ref] Patients who respond to interferon therapy with long-term remission of disease and sustained loss of HCV RNA generally are regarded as being unlikely to develop cirrhosis of the liver or HCC. Thus, interferon therapy may suppress postoperative carcinogenesis. Another possible mechanism is that interferon may have a suppressive effect on HCC progression, because Lai et al. [bib_ref] Recombinant interferon-α in inoperable hepatocellular carcinoma: a randomized controlled trial, Lai [/bib_ref] have shown that patients with advanced HCC treated with interferon have a significantly higher survival rate than patients not given interferon.
In conclusion, continuous hepatitis with persistent HCV viremia adversely affects outcome after liver resection for
[fig] Figure 1: Changes in serum alanine aminotransferase activity after operation. Vertical bars indicate SD. Numbers of patients at each time point are shown at the bottom.alanine aminotransferase activity in patients with serum HCV RNA, alanine aminotransferase activity in patients without serum HCV RNA. [/fig]
[fig] Figure 2: Tumor-free survival rates after liver resection in patients with and without serum HCV RNA.patients without HCV viremia (n=7), patients with HCV viremia (n=52). [/fig]
[fig] Figure 3: Cumulative survival rates after liver resection in patients with and without serum HCV RNA.patients without HCV viremia (n=7), patients with HCV viremia (n=52). [/fig]
[table] Table I: Clinicopathologic Findings and the Results of Laboratory Tests in Patients with and without Hepatitis C Viremia [/table]
|
Downregulation of Nuclear-Encoded Genes of Oxidative Metabolism in Dialyzed Chronic Kidney Disease Patients
Background: Mitochondria, essential eukaryotic cells organelles defined as the ''powerhouse of the cell'' because of their ability to produce the vast majority of energy necessary for cellular metabolism, may have a primary role in the oxidative stress-related intracellular machinery associated to chronic kidney disease (CKD).Methods: To better assess this research assumption, we decided to study the key factors regulating mitochondrial oxidative metabolism in CKD patients in peritoneal dialysis (PD, n = 15) using several bio-molecular methodologies.Results: RT-PCR experiments demonstrate that the expression level of peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1a) and nuclear respiratory factor-1 (NRF-1), two genes primarily involved in mitochondrial biogenesis and functions, were significantly hypo-expressed in peripheral blood mononuclear cells of PD patients compared to healthy subjects (HS, n = 15). Additionally, mRNA levels of several PGC1-a downstream target genes (TFAM, COX6C,COX7C, UQCRH and MCAD) were profoundly down-regulated in PD cells. TFAM protein analysis confirmed gene-expression results. High plasmatic concentration of Malondialdehyde found in PD patients, confirmed the contribution of the oxidative stress to these biological effects. Finally, Nuclear factor erythroid-derived 2-like 2 (NRF2 or NFE2L2), a transcription factor for numerous antioxidant/detoxifying enzymes and one of its target genes, superoxide dismutase-2 mitochondrial (SOD2) were up-regulated in PD compared to HS.Conclusions: Our results revealed, for the first time, that CKD-PD patients' PBMC, through a complex intracellular biochemical machinery, are able to modulate their mitochondrial functions probably in the attempt to reduce oxidative metabolic damage and to turn on a valuable defense cellular strategy against oxidative stress.
# Introduction
Chronic kidney disease (CKD) is a progressive and irreversible deterioration of kidney function classified by the last international guidelines into five stages according to glomerular filtration rate. In the last stage (end stage renal disease) the kidney impairment is advanced and cellular/metabolic functions are significantly altered and enable to guarantee normal body homeostasis. Therefore, at this clinical phase, renal replacement therapies (RRTs, peritoneal-or hemo-dialysis) or renal transplantation are needed to ensure patient's survival.
Although hemodialysis (HD) still represents the leading RRT, in the last years, the number of patients undergoing peritoneal dialysis (PD) procedure is increasing worldwide being a preferable choice for young patients with high life expectancy and an elevated probability to undergo renal transplantation. This is mainly related to a lower activation of microinflammation, compared to HD [bib_ref] Cardiovascular disease in haemodialysis and peritoneal dialysis: arguments pro peritoneal dialysis, Van Biesen [/bib_ref] , better preservation of residual renal function [bib_ref] Peritoneal dialysis and preservation of residual renal function, Tam [/bib_ref] and higher quality of life [bib_ref] Quality of life in patients undergoing hemodialysis or peritoneal dialysis treatment, Theofilou [/bib_ref].
However, peritoneal catheter and dialysis solutions (characterized by high concentration of glucose, glucose degradation products, low pH and high osmolality) used to remove waste products generated from normal metabolic processes, uremic toxins and to normalize body fluid and electrolytesmay still determine the systemic activation of a complex intracellular machinery leading to inflammation and oxidative stress [bib_ref] Relative contribution of nutrition and inflammation to clinical outcome in dialysis patients, Kalantar-Zadeh [/bib_ref] [bib_ref] Oxidative stress during peritoneal dialysis: implications in functional and structural changes in..., Noh [/bib_ref] [bib_ref] Free radicals in peritoneal dialysis: agents of damage?, Breborowicz [/bib_ref].
In this context, the recruitment, rolling and activation of peripheral blood mononuclear cells (PBMCs) may have a pivotal etiopathogenic role.
As recently reported, transcriptomic analysis of PBMCs of PD patients has shown high expression of important key regulators of inflammation and oxidative stress (e.g. RELA, GSS) in this population [bib_ref] Dialysisrelated systemic microinflammation is associated with specific genomic patterns, Zaza [/bib_ref] with mitochondria having a possible pivotal role in the onset and development of these processes [bib_ref] Mitochondria in innate immunity, Arnoult [/bib_ref] [bib_ref] Generation of reactive oxygen species by the mitochondrial electron transport chain, Liu [/bib_ref].
Mitochondria are essential eukaryotic cells organelles involved in several metabolic pathways, including calcium signaling [bib_ref] Mitochondrial calcium signalling and cell death: approaches for assessing the role of..., Hajnóczky [/bib_ref] , heme [bib_ref] Biosynthesis of heme in mammals, Ajioka [/bib_ref] and steroid synthesis [bib_ref] T channels and steroid biosynthesis: in search of a link with mitochondria, Rossier [/bib_ref] , apoptosis [bib_ref] Apoptotic pathways: the roads to ruin, Green [/bib_ref]. This intracellular elements are defined as the ''powerhouse of the cell'' because of their ability to produce the vast majority of energy necessary for cellular metabolism through the oxidative phosphorylation system (OXPHOS).
Structurally, they present an outer and inner membrane, the latter of which would be impermeable to all molecules in the absence of specific carriers and contains the OXPHOS complexes. The respiratory flux is due to the donation of electrons from NADor FAD-dependent substrates, via respiratory chain, to molecular oxygen which is finally reduced to water. Simultaneously, the energy conserving complexes I, III and IV build up a transmembrane electrochemical gradient by coupling the electron transfer activity to proton translocation from the matrix to the outer side of the inner mitochondrial membrane. Complex V utilizes backward the electrochemical gradient for ATP synthesis. During this process a small percentage of electrons may ''leak'' from the respiratory chain and partially reduce oxygen, forming superoxide anion (O 2 2 ) [bib_ref] The role of mitochondria in reactive oxygen species metabolism and signaling, Starkov [/bib_ref]. Consequently mitochondria are the major source of ROS in the cell.
Recently increasing evidences showed mitochondrial dysfunction in a broad spectrum of renal disorders [bib_ref] The not so 'mighty chondrion': emergence of renal diseases due to mitochondrial..., Hall [/bib_ref] and we previously reported, for the first time, that CKD/HD patients exhibited an impaired mitochondrial respiratory system. Therefore, in order to improve our knowledge about the mitochondria-related cellular changes occurring in the heterogeneous CKD population, to study possible intracellular defense mechanisms against inflammation/oxidative stress and to identify new diagnostic/prognostic biomarkers or therapeutic targets, we decided to analyze in our in vivo study some of the key biological cellular regulators of oxidative metabolism in uremic patients in dialysis treatment.
We decided to focus on PD patients since the increasing interest of the nephrological research on this large dialysis patients' population and because, at present, no study has analyzed the relationship between oxidative stress and mitochondrial deregulation in this population.
# Methods
## Patients and controls
A total of 30 subjects [n: 15 CKD in peritoneal dialysis (PD) and n:15 healthy subjects (HS)], after signing informed consent, were enrolled in our study. The main demographic and clinical features are summarized in [fig_ref] Table 1: Demographics and clinical characteristics of study group [/fig_ref].
Thirteen PD patients have been treated with automated peritoneal dialysis (APD) using a bicarbonate buffer (Physioneal, Baxter) and 2 with a continuous ambulatory peritoneal dialysis (CAPD) using a bicarbonate buffer (Physioneal, Baxter, Chicago, IL, USA).
To avoid confounding factors, all patients suffering from systemic autoimmune disorders, infectious diseases, diabetes, chronic lung diseases, neoplasm, or inflammatory diseases and patients receiving antibiotics, corticosteroids, or non-steroidal antiinflammatory agents were excluded. No patients had symptomatic coronary artery diseases or a family history of premature cardiovascular diseases.
Serum C-reactive protein (CRP) levels were measured in all patients included in the study using high-sensitivity immunonephelometric (Dade Behring, Marbung, Germany) according to the manufacturer's protocol.
The study was carried out according to the Declaration of Helsinki and approved by the Institutional Ethic Review Board of the University Hospital ''Policlinico di Bari'', Bari, Italy (number of registration: 599/CE).
## Peripheral blood mononuclear cells (pbmcs) isolation
Fifteen ml whole blood were collected from all subjects included in the study. For PD patients the biological material was obtained during the outpatient clinical evaluations. PBMCs were isolated by density separation over a Ficoll-Paque TM (GE healthcare, Sweden) gradient (460 g for 30 min) and washed three times with PBS pH 7.4/1 mM EDTA (Sigma, Milan, Italy). Cells were counted, and viability was assessed by trypan blue exclusion method (.90% PBMCs were viable).
## Rna extraction and real-time pcr
Total RNA was isolated from PBMCs by RNeasy mini kit Qiagen (QIAGEN AG, Basel, Switzerland). RNA was quantified by Quant-iT RNA Assay kit with Qubit Fluorometer (Invitrogen).
Reverse transcription of RNA was performed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems), following the manufacturer's instructions. One mg of RNA was reverse transcribed using random primer and MultiScribe Reverse Transcriptase. Real-time PCR amplification reactions were performed in duplicate in 20 ml of final volume via SYBR Green chemistry on ABI-Prism 7700 (Applied Biosystem). PCR protocol was performed using QuantiTect Primer Assays (Qiagen, Basel, Switzerland) for PGC-1a, NRF-1, TFAM, MCAD, NRF2, SOD2 and b-actin: 50uC for 2 min, 95uC for 2 min and 40 two-step cycles: 95uC for 15 sec and 60uC for 30 sec. For COX6C, COX7C (subunits of complex IV or Cytochrome c oxidase) and UQCRH (subunit of Complex III or ubiquinol-cytochrome c reductase) the primers were designed by the aid of the Primer3 software (frodo.wi.mit.edu). The forward and reverse primer sequences were: COX6C forward 59-CTTTGTA- at 10,000 g at 4uC for 10 min. The supernatants were collected and stored at 280uC until used. Aliquots containing 30 mg of proteins from each lysate were subjected to SDS-PAGE on a Criterion TM TGX Any kD Stain-Free TM (Biorad) and then electrotransferred onto PVDF membrane (Biorad) by Trans-Blot Turbo transfer system (Biorad). The filter was blocked with 5% milk powder in TBS containing 0.1% Tween-20 (TBS-T). Membranes were probed with primary antibody against TFAM (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and incubated with secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Horseradish peroxidase was detected with the ECLenhanced chemiluminescence system (Amersham, Buckinghamshire, UK). The same membranes were stripped and proteins were rehybridized with anti-b-actin antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Images were acquired using a scanner EPSON Perfection 2580 Photo (EPSON, Long Beach, CA, USA) and quantified by Image J 1.34 Software (http://rsb. info.nih.gov/ij/). The intensity of bands, corresponding to the TFAM protein, was normalized to the b-actin signal.
## Western blot analysis for nuclear protein expression of nrf2
Nuclear extracts of freshly isolated PBMCs from 5 randomly selected HS and 5 PD patients were obtained according to the published protocol by Almeida et al [bib_ref] Dysregulation of CREB activation and histone acetylation in 3-nitropropionic acid-treated cortical neurons:..., Almeida [/bib_ref]. PBMCs were washed with ice-cold PBS and resuspended in buffer A (10 mM HEPES, 10 mM NaCl, 3 mM MgCl 2 , 1 mM EGTA, 0.1% Triton X-100, pH 7.5), supplemented with 50 mM NaF, 1 mM Na 3 VO 4 , 1 mM DTT, 1 mM PMSF, and complete protease inhibitor cocktail (Roche) on ice for 40 min, then centrifuged at 2,400 g for 10 min. The pellets containing nuclei were resuspended in buffer B (25 mM HEPES, 300 mM NaCl, 5 mM MgCl 2 , 1 mM EGTA, 20% glycerol, pH 7.4), supplemented with 50 mM NaF, 1 mM Na 3 VO 4 , 1 mM DTT, 1 mM PMSF, and complete protease inhibitor cocktail on ice for 60 min. The lysates were centrifuged at 12,000 g for 20 min, the supernatants were collected, and the amount of nuclear protein was measured using the Bio-Rad protein assay reagent.
Subsequently, nuclear extracts (20 mg) were resolved in 9% SDS-PAGE and electrotransferred onto nitrocellulose membranes. The filter was blocked with 5% milk powder in TBS-T. Membranes were probed with anti-NRF2 (Genetex GTX103300; 1:500) and anti-Lamin B (sc-2616, Santa Cruz Biotechnolgy 1:1000) antibodies. After three washes in TBST, the membranes were incubated with the secondary peroxidase conjugated antibodies. The signal was detected by ECL-enhanced chemiluminescence system (Amersham, Buckinghamshire, UK) according to the manufacturer's instructions. Images were acquired using a scanner EPSON Perfection 2580 Photo (EPSON, Long Beach, CA, USA) and quantified by Image J 1.34 Software (http://rsb.info.nih.gov/ij/). The intensity of bands, corresponding to the NRF2 protein, was normalized to the Lamin B signal.
## Thiobarbituric acid reactive substances (tbars) assay
Plasmatic measurement of Malondialdehyde (MDA) as TBARS was performed for all subjects included in the study using a commercially available kit (AbNOVA Abnova, Walnut, CA) according to the manufacturer's instructions.
## Resting metabolic rate
Resting metabolic rate (RMR) was assessed by indirect calorimetry with the use of an MMC Horizon System 6 (Beckman Sensormedics, Milan, Italy) that measures resting oxygen uptake and resting carbon dioxide production. The subjects were familiarized with the canopy of the calorimeter so that they did not feel suffocated during the measurement period. They were instructed to avoid hyperventilating, fidgeting, and falling asleep. Gas was measured in the morning in the supine position after the subjects had fasted 12 h. Values were considered reliable after a 20-min nonstop period when differences in consecutive values were ,5%; at this point, gas measurements were continued for another 20 min, and mean values for resting oxygen uptake and resting carbon dioxide production were calculated. The CV for duplicate measurements in 12 subjects was 5%.
# Statistical analysis
Results were expressed as mean 6 SD. Student's t-test and Chisquared test were used to assess differences in clinical, demographic and experimental features. A value of p,0.05 was considered to be statistically significant. R 2.0.1 statistical software was used to perform the above analyses (www.r-project.org).
# Results
Peroxisome Proliferator-activated Receptor Gamma coactivator 1 Alpha (PGC1-a) and Nuclear Respiratory Factor 1 (NRF-1) Gene Expression
To assess whether PBMCs isolated from CKD patients undergoing PD treatment showed an alteration of the cellular machinery associated to the oxidative energy metabolism, we measured by RT-PCR the expression level of peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC1-a) and Nuclear respiratory factor 1 (NRF-1), two genes encoding, respectively, for a transcriptional coactivator and a transcription factor that together stimulate the expression of a broad set of nuclear genes involved in mitochondrial biogenesis and functions.
As reported in [fig_ref] Figure 1: Peroxisome proliferator-activated receptor gamma coactivator 1 alpha [/fig_ref] , both genes resulted significantly downregulated in PD compared to HS.
## Mitochondrial transcription factor a (tfam) gene and protein levels
Since TFAM is the major gene regulated by NRF-1 and it regulates mitochondrial transcription and replication, we decided to measure its gene expression in all subjects included in the study.
As expected, PD patients showed a significant TFAM mRNA down-regulation compared to HS [fig_ref] Figure 2: Transcription factor A, mitochondrial [/fig_ref]. Protein analysis by western blot confirmed the results obtained by gene expression [fig_ref] Figure 2: Transcription factor A, mitochondrial [/fig_ref].
## Medium chain acyl coa dehydrogenase (mcad) gene expression
To confirm the reduced PGC1-a activity we evaluated the mRNA level of MCAD, an oxidoreductase enzyme tightly regulated by this coactivator, that catalyzes the first step of mitochondrial fatty acid beta-oxidation, in our PD patients and HS. MCAD gene expression was significantly lower in PD compared to HS [fig_ref] Figure 3: Medium chain acyl CoA dehydrogenase [/fig_ref].
## Oxidative phosphorylation system (oxphos) subunits
Mitochondrial oxidative phosphorylation system (OXPHOS) plays a central role for energy homeostasis in mammals. This machinery is finely regulated by several factors including PGC-1a and NRF-1. Therefore, we decided to analyze whether the down regulation of these factors in our CKD-PD patients induced downregulation of expression of genes encoding for mitochondrial OXPHOS subunits. Our results revealed that the expression level of genes encoding for UQCRH (complex III subunit) and COX6C/COX7C (complex IV subunits) were significantly hypo-expressed in our PD patients compared to HS [fig_ref] Figure 4: COX6C, COCX7C and UQCRH gene expression by Real-Time PCR in PBMCs from... [/fig_ref]. These results suggest a possible reduction of OXPHOS activity.
## Oxidative stress evaluation by thiobarbituric acid reactive substances (tbars)
To assess the relationship between PCG1-a, NRF-1 deregulation and oxidative stress, we measured the TBARS levels, products of oxidative damage to lipid, in all subjects included in our study. In particular, we analyzed the plasmatic levels of malondialdehyde (MDA), one of several low-molecular-weight end products formed via the decomposition of certain primary and secondary lipid peroxidation products. MDA levels were significantly higher in our PD patients compared to HS [fig_ref] Figure 5: Malondialdehyde [/fig_ref] demonstrating an high oxidative stress in our dialysis patients. To exclude a possible influence of the nutritional status on the previous results, we performed an indirect calorimetry test to measure the resting metabolic rate (RMR) on all the study population. Concordantly with other reports [bib_ref] Resting energy expenditure in peritoneal dialysis patients, Bazanelli [/bib_ref] , we did not find any differences between PD and HS [fig_ref] Figure 6: Resting metabolic rate [/fig_ref].
## Evaluation of the antioxidant cellular response by nrf2 and sod2 gene expression analysis
Finally, to better define the antioxidant response against the cytotoxic effects of oxidative stress, we evaluated the expression level of Nuclear factor-erythroid-2-related factor 2 (NRF2 or NFE2L2) and Superoxide dismutase 2, mitochondrial (SOD2) in our CKD-PD patients and HS.
As predictable, both genes were up-regulated in PD compared to HS demonstrating an enhanced activation of the cellular antioxidant machinery in this study group [fig_ref] Figure 7: NRF-2 and SOD2 gene expression by Real-Time PCR in PBMCs from healthy... [/fig_ref].
## Nrf2 protein levels
Because of NRF2 mediates the transcriptional response of cells to oxidative stress by the translocation into nucleus, we evaluated its protein level in nuclear extracts from PBMCs of HS and CKD-PD patients. Concordantly to gene expression analysis, NRF2 protein level resulted significantly higher in CKD-PD patients compared to HS [fig_ref] Figure 8: Nuclear factor erythroid-derived 2-like 2 [/fig_ref].
# Discussion
Oxidative stress and mitochondrial activity are key elements of many pathological conditions, including neurodegenerative disorders, diabetes, cardiovascular disease and cancer [bib_ref] Principles and therapeutic relevance for targeting mitochondria in aging and neurodegenerative diseases, Serviddio [/bib_ref] [bib_ref] Oxidative stress, insulin signaling, and diabetes, Rains [/bib_ref] [bib_ref] Mitochondrial subversion in cancer, Chatterjee [/bib_ref] [bib_ref] Regulation of endothelial function by mitochondrial reactive oxygen species, Widlansky [/bib_ref].
Moreover, our research group has recently demonstrated, for the first time, a close link between mitochondrial deregulation and oxidative stress in chronic kidney disease (CKD) patients in conservative and hemodialysis (HD) treatment [bib_ref] Mitochondrial dysregulation and oxidative stress in patients with chronic kidney disease, Granata [/bib_ref]. In particular, using an innovative high-throughput technology, we discovered that several biological elements involved in the oxidative phosphorylation system and two key constituents of the mitochondrial complex IV (COXI and COXIV) were deregulated in CKD/HD patients compared to healthy controls. In addition, complex IV activity, the terminal enzyme of the mitochondrial respiratory chain catalyzing the electron transfer from reduced cytochrome c to oxygen [bib_ref] Cytochrome c oxidase: catalytic cycle and mechanisms of proton pumping-a discussion, Michel [/bib_ref] , resulted significantly lower in CKD/ HD patients compared to healthy subjects demonstrating a reduced activity of oxidative phosphorylation system in this population. However, at the state of art, the complete cellular mechanism involved in this intricate biological system is still completely unknown.
Therefore, to better comprehend the mitochondria-related biochemical/metabolic cellular alterations in CKD, we decided to use a well standardized biomolecular methodologies (e.g., RT-PCR, western-blotting) to measure the expression level of some key biological cellular regulators of the oxidative metabolism in patients with elevated kidney damage undergoing peritoneal dialysis (PD) treatment. PD is a renal replacement therapy that, instead of hemodialysis, is associated with better preservation of residual renal function, initial survival advantage, reduced erythropoiesis stimulatory agent requirements and preservation of vascular access sites [bib_ref] Relationship between dialysis modality and mortality, Mcdonald [/bib_ref] [bib_ref] Integrated end-stage renal disease care: the role of peritoneal dialysis, Blake [/bib_ref]. However, the use of ''unphysiologic'' conventional PD fluids (characterized by acidic pH, high lactate concentrations, high osmolality, high glucose concentrations, and contamination by glucose degradation products) may contribute to the onset/ development of several adverse outcomes [bib_ref] 4-Dideoxyglucosone-3-ene promotes leukocyte apoptosis, Catalan [/bib_ref] [bib_ref] In vitro testing of a potentially biocompatible continuous ambulatory peritoneal dialysis fluid, Topley [/bib_ref] [bib_ref] Effect of lactate-buffered peritoneal dialysis fluids on human peritoneal mesothelial cell interleukin-6..., Witowski [/bib_ref] [bib_ref] Effects of conventional and new peritoneal dialysis fluids on leukocyte recruitment in..., Mortier [/bib_ref] and to the activation of oxidative stress [bib_ref] Oxidative stress during peritoneal dialysis: implications in functional and structural changes in..., Noh [/bib_ref]. Although well described, the molecular mechanisms associated to latter condition are still not completely known.
To better address this point, we focused on the Peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1a)-related intracellular machinery. PGC-1a is a well known master regulator of mitochondrial oxidative metabolism [bib_ref] Metabolic control through the PGC-1 family of transcription coactivators, Lin [/bib_ref]. Its expression seems finely tuned to reflect cellular energy needs, with conditions of increased energy demands inducing its expression [bib_ref] PGC-1 coactivators: inducible regulators of energy metabolism in health and disease, Finck [/bib_ref] [bib_ref] PGC-1alpha: a key regulator of energy metabolism, Liang [/bib_ref]. PGC-1a performs this task by coactivating a large number of transcription factors, including, among others, nuclear respiratory transcription factor 1 (NRF-1) and in this way, it regulates the activity of numerous nuclear genes encoding mitochondrial proteins [bib_ref] Transcriptional activators and coactivators in the nuclear control of mitochondrial function in..., Scarpulla [/bib_ref].
Interestingly, our RT-PCR experiments demonstrated that the expression levels of PGC-1a, NRF-1 and the other analyzed downstream target genes were significantly down-regulated in PD patients compared to HS. Therefore PBMCs of CKD patients showed a specific down-regulation of several nuclear-encoded genes involved in the mitochondrial biogenesis and functions (TFAM, COX6C, COX7C, UQCRH and MCAD).
TFAM has a key biological role because, after migration into mitochondria, it regulates mitochondrial DNA transcription and replication [bib_ref] Transcriptional regulatory circuits controlling mitochondrial biogenesis and function, Kelly [/bib_ref]. COX6C and COX7C encode for two subunits of the cytochrome c oxidase (COX or Complex IV), UQCRH is a component of the ubiquinol-cytochrome c reductase complex (complex III), which catalyzes electron transfer from succinate and nicotinamide adenine dinucleotide-linked dehydrogenases to cytochrome c [bib_ref] Possible molecular mechanisms of the protonmotive function of cytochrome systems, Mitchell [/bib_ref]. MCAD, then, is an oxidoreductase enzyme regulated by PGC1-a, that catalyzes the first step of mitochondrial fatty acid beta-oxidation.
Based on previous literature evidences reporting that reactive oxygen species may induce PGC-1a down-regulation [bib_ref] Reactive oxygen species-dependent transcriptional regulation of peroxisome proliferator-activated receptor c coactivator 1a..., Kim [/bib_ref] , we assumed that this biological/biochemical complex, through a mitigation of the mitochondrial OXPHOS activity, could represent a protective adaptive response against chronic cellular perturbation associated to the kidney disease-related oxidative injury.
Our CKD patients, in fact, showed higher plasma concentration of Malondialdehyde (MDA), a thiobarbituric acid reactive substance (TBARS) commonly known as a marker of oxidative stress [bib_ref] Simple assay for the level of total lipid peroxides in serum or..., Yagi [/bib_ref] , compared to healthy controls.
Additionally, our results confirmed previous literature evidences reporting that advanced CKD ''per se'' mainly through the accumulation of several circulating uremic toxins (e.g., indoxyl sulfate, p-cresyl sulfate) [bib_ref] The uremic solute indoxyl sulfate induces oxidative stress in endothelial cells, Dou [/bib_ref] [bib_ref] Pcresylsulphate, the main in vivo metabolite of p-cresol, activates leucocyte free radical..., Schepers [/bib_ref] and the interaction of PBMCs with bio-incompatible dialysis devices can cause their activation with imbalance between pro-and anti-oxidant activities resulting in high oxidative stress [bib_ref] Free radicals in peritoneal dialysis: agents of damage?, Breborowicz [/bib_ref] [bib_ref] Increased oxidative damage to peripheral blood leukocyte DNA in chronic peritoneal dialysis..., Tarng [/bib_ref].
This hypothesis was also in part confirmed by our finding of an additional activated cellular anti-oxidant machinery in PBMCs of PD patients. Specifically, Nuclear factor erythroid-derived 2-like 2 (NRF2 or NFE2L2), a transcription factor regulating the expression of numerous antioxidant/detoxifying enzymes, and one of its down-stream target genes superoxide dismutase-2 mitochondrial (SOD2) [bib_ref] The Nrf2-antioxidant response element signaling pathway and its activation by oxidative stress, Nguyen [/bib_ref] resulted significantly up-regulated in our CKD-PD population. SOD2 binds to the superoxide byproducts of oxidative phosphorylation and converts them to hydrogen peroxide and diatomic oxygen [bib_ref] The role of manganese superoxide dismutase in health and disease, Robinson [/bib_ref]. Therefore, all together, our results, although generated on a small but well selected patients' population, revealed, for the first time, a fine regulated intracellular biochemical system associated to oxidative stress response in CKD patients. It is plausible that this redox-dependent mechanism could have a pivotal role in antioxidant defense cellular strategy occurring in cells of these patients [fig_ref] Figure 9: Supposed mechanism of PGC1-a/NRF1-NRF2 pathway anti-oxidative stress cellular defense in chronic kidney... [/fig_ref]. However, further studies are necessary to better delineate all the biological/biochemical mechanisms involved.
The main limitation of our study is the lack of the analysis of all potential clinical variables able to influence the ''mitochondrial'' transcriptomic profile primarily due to the time and cost consuming of a global analytic/research strategy.
Moreover, our in vitro model failed to corroborate our in vivo findings (See File S1 and [fig_ref] Figure 1: Peroxisome proliferator-activated receptor gamma coactivator 1 alpha [/fig_ref]. In fact, PBMCs stimulated with high glucose PD dialysis solutions showed the upregulation of all the previous analyzed genes. These contradictory results clearly demonstrated the unquestionable complexity of this machinery in which all together uremia, microinflammation, subclinical/clinical peritoneal infections, acidosis, electrolytic unbalance contribute to the onset and development of this biological/clinical condition. Therefore, we strong encourage a collaborative international research program to address these points.
Finally, we can not exclude that, in future, the modulation of this machinery could turn on as a valuable point of therapeutic intervention to reduce oxidative stress-related clinical complications in CKD patients in both conservative and dialysis treatment.
## Supporting information
[fig] Figure 1: Peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC1-a) and nuclear respiratory factor-1 (NRF-1) gene expression by Real-Time PCR in PBMCs from healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). The histograms represent the mean 6 SD of PGC1-a (A) and NRF-1 (B) expression level, determined by Real-Time PCR in PBMCs from 15 HS and 15 PD patients. For both genes the expression level was significantly lower in PD compared to HS (**p,0.001). doi:10.1371/journal.pone.0077847.g001 [/fig]
[fig] Figure 2: Transcription factor A, mitochondrial (TFAM) gene and protein expression in PBMCs from healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). (A) The histogram represents the mean 6 SD of TFAM expression level, determined by Real-Time PCR in PBMCs from 15 HS and 15 PD patients. (B) Dot-plot represents the normalized TFAM protein level in total cell lysates of 8 HS and 11 PD patients assessed by Western blotting and (C) the representative western blotting experiment for TFAM. The line represents the mean value. Both TFAM mRNA and protein level was significantly lower in PD patients compared to HS. doi:10.1371/journal.pone.0077847.g002 [/fig]
[fig] Figure 3: Medium chain acyl CoA dehydrogenase (MCAD) gene expression in PBMCs from healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). The histogram represents the mean 6 SD of MCAD expression level, determined by Real-Time PCR in PBMCs from 15 HS and 15 PD patients. The expression level was significantly lower in PD compared to HS (**p,0.001). doi:10.1371/journal.pone.0077847.g003 [/fig]
[fig] Figure 4: COX6C, COCX7C and UQCRH gene expression by Real-Time PCR in PBMCs from healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). The histograms represent the mean 6 SD of COX6C (A) COX7C (B) and UQCRH (C) expression level, determined by Real-Time PCR in PBMCs from 15 HS and 15 PD patients. For all these genes the expression level was significantly lower in PD compared to HS (**p,0.001). doi:10.1371/journal.pone.0077847.g004 [/fig]
[fig] Figure 5: Malondialdehyde (MDA) levels in plasma of healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). Dot-plot represents plasma MDA levels in 15 HS and 15 PD patients. Malondialdehyde (MDA) was measured as TBARS. The line represents the mean value. MDA level was higher in PD compared to HS. doi:10.1371/journal.pone.0077847.g005 [/fig]
[fig] Figure 6: Resting metabolic rate (RMR) assessment in healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). Dot-plot represents the RMR assessed by indirect calorimetry in 15 HS and 15 PD patients. RMR levels were similar in the two study groups. doi:10.1371/journal.pone.0077847.g006 [/fig]
[fig] Figure 7: NRF-2 and SOD2 gene expression by Real-Time PCR in PBMCs from healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). The histograms represent the mean 6 SD of NRF-2 (A) and SOD2 (B) expression level, determined by Real-Time PCR in PBMCs from 15 HS, 15 PD patients. For both genes the expression level was significantly higher in PD compared to HS (*p,0.01). doi:10.1371/journal.pone.0077847.g007 [/fig]
[fig] Figure 8: Nuclear factor erythroid-derived 2-like 2 (NRF2) protein expression in nuclear extracts of PBMCs from healthy subjects (HS) and chronic kidney disease patients in peritoneal dialysis (PD). (A) Dot-plot represents the normalized NRF2 protein level in nuclear extracts of 5 HS and 5 PD patients assessed by Western blotting and (B) the representative western blotting experiment for NRF2. The line represents the mean value. NRF2 protein level was significantly higher in PD patients compared to HS. doi:10.1371/journal.pone.0077847.g008 Nutritional Status Evaluation Using the Resting Metabolic Rate (RMR) [/fig]
[fig] Figure 9: Supposed mechanism of PGC1-a/NRF1-NRF2 pathway anti-oxidative stress cellular defense in chronic kidney disease patients in peritoneal dialysis treatment. Oxidative stress alters the interaction of Kelch-like ECH-associated protein 1 (Keap1) and Nuclear factor erythroid-derived 2-like 2 (Nrf2), thereby liberating Nrf2 activity from repression by Keap1. NRF2 migrates into the nucleus were it activates the transcription of Superoxide dismutase 2, mitochondrial (SOD2). At the same time, oxidative stress causes the down-regulation of Peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC1-a) and Nuclear respiratory factor-1 (NRF-1) with the consequent down-regulation of PGC-1a downstream target genes (TFAM, COX6C, COX7C, UQCRH and MCAD). The reduced TFAM expression causes a decrease in mitochondrial transcription and replication. The down-regulation of all these factors suggests the decrease in mitochondrial OXPHOS activity in order to reduce ROS accumulation and creating an antioxidant feedback. doi:10.1371/journal.pone.0077847.g009 [/fig]
[fig] Figure S1, Figure: Peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC1-a), nuclear respiratory factor-1 (NRF-1) and Transcription factor A, mitochondrial (TFAM) gene expression in PBMC from 3 healthy subjects stimulated with PD fluid. The histograms represent the mean 6 SD of PGC1-a (A), NRF-1 (B) and TFAM (C) level of expression, determined by Real-Time PCR in PBMC from 3 HS stimulated with PD fluid. All the three genes resulted significantly up-regulated after 6 and 24 h of incubation with PD fluid (#p,0.05; ##p,0.01 versus CTR). (TIF) Figure S2 Medium chain acyl CoA dehydrogenase (MCAD) gene expression in PBMC from 3 healthy subjects stimulated with PD fluid. The histogram represents the mean 6 SD of MCAD level of expression, determined by Real-Time PCR in PBMC from 3 HS stimulated with PD fluid. The expression level was significantly up-regulated after 6 and 24 h of incubation with PD fluid (#p,0.05 versus CTR). (TIF) Figure S3 COX6C, COCX7C and UQCRH gene expression in PBMC from 3 healthy subjects stimulated with PD fluid. The histograms represent the mean 6 SD of COX6C (A) COX7C (B) and UQCRH (C) level of expression, determined by Real-Time PCR in PBMC from 3 HS stimulated with PD fluid. All the three genes resulted significantly up-regulated after 6 and 24 h of incubation with PD fluid (#p,0.05; ##p,0.01 versus CTR). (TIF) [/fig]
[table] Table 1: Demographics and clinical characteristics of study group. [/table]
|
Laparoscopic Morcellation of Didelphic Uterus With Cervical and Renal Aplasia
This is a case report (and review of the literature) of a 12-year and 10-month-old girl with a rare congenital anomaly of uterus didelphys, unilateral cervix aplasia, and ipsilateral renal aplasia. She had severe dysmenorrhea from the first menses. In an effort to preserve fertility, a cervical fistula was made that closed over. A laparoscopic hemi-hysterectomy was done successfully and rapidly with laparoscopic morcellation. Because no ureter was present, it was not necessary to trace it. For this congenital anomaly, laparoscopic morcellation of the obstructed hemiuterus is the preferred treatment either as a primary procedure or as a secondary procedure following failure of a surgical cervical fistula for the young patient.
# Introduction
A congenital anomaly syndrome was reported in 1922 of uterus didelphys, unilateral hematocolpos with distal vaginal aplasia, and ipsilateral renal agenesis. [bib_ref] A case of unilateral haematocolpos, haematometra and haematosalpinx, Purslow [/bib_ref] Since then, over 180 such cases have been reported. [bib_ref] Term pregnancy at the site of atresia following vaginal canalization in a..., Chen [/bib_ref] It has no eponym, and we refer to it as "no eponym" syndrome. [bib_ref] Congenital gynecologic anomalies I -Rokitansky Syndrome (MRKH Syndrome), Altchek [/bib_ref] The patients present with unilateral upper hematocolpos and painful menses that start many months after menarche, because initially the upper vagina can slowly distend without pain [fig_ref] Figure 1: The usual "no eponym" anomaly syndrome before menstruation begins [/fig_ref]. The diagnosis is suggested by a paravaginal mass (hematocolpos) readily detected by vaginal or rectal examination. The appropriate surgical management for the usual no eponym syndrome is conservative vaginal marsupialization of the upper ipsilateral hematocolpos and leaving both hemiuteri in place.
Our patient had an unusual variation. Instead of obstruction by the unilateral distal vaginal aplasia, the obstruction was higher up due to cervix aplasia [fig_ref] Figure 3: Our case report [/fig_ref]. In view of the young age and the desire to preserve fertility, we created a uterine fistula, which closed over. Rather than repeat the fistula with the possibility of infection, the hemiuterus was removed by laparoscopic hemi-hysterectomy morcellation.
## Case report and results
The patient was first seen at age 12 years 10 months because of severe left lower abdominal pain with the first menstrual period, which began 1year and 9 months earlier. For the previous 6 months, the pain had become continuous.
The enlarged left hemiuterus was palpated arising from the pelvis to 6cm above the symphysis pubis. On rectal examination, the lower pole was tense, tender, cystic, and high (6cm from the pelvic floor). Abdominal sonogram indicated left renal agenesis. Pelvic sonogram showed a 13-cm x 10-cm hematometra. MRI depicted a uterus didelphys. The left hemiuterus wall was dilated by old blood and elevated the left adnexa [fig_ref] Figure 3: Our case report [/fig_ref] aplasia. The combination of uterus didelphys, cervical aplasia, and elevated adnexa has not been described in the literature to the best of our knowledge.
Examination with the patient under general anesthesia revealed one normal vagina and one right normal cervix. There was no left cervix. The bulging left vaginal fornix was punctured and dilated with the release of a large volume of old, pasty menstrual blood from the left hematometra. A wide number 24 Foley bladder catheter with a large 10-mL balloon was left indwelling for drainage. The catheter passed 2 weeks later. After surgery, the first menses were painless. Eight weeks later, a 1-finger vaginal examination palpated a distended 7-cm left hemiuterus fundus without any fistula opening. An MRI showed marked distention of the left hemiuterus with old and new blood and a new left hydro-hemato-salpinx. Two months later, the patient had painful menses. A 1-finger examination palpated a 10-cm left posterior cystic mass. A repeat MRI showed no left cervix.
At the second surgery, there was no left hemiuterus fistula visible or palpable in the left vaginal fornix. Laparoscopy was done for the first time. There were extensive adhesions in the left pelvis. The anterior cul-de-sac had superficial endometriosis. A left salpingectomy was done because of the severe hematosalpinx. After dissection and hemostasis by bipolar and harmonic wave ultrasound instruments, the left hemiuterus was completely removed by a relatively simple and rapid laparoscopic morcellation [fig_ref] Figure 4: Laparoscopic morcellation of enlarged left uterus [/fig_ref] down to the left vaginal vault but not penetrating it. There was no unusual blood loss or need of suturing. Because there was no left kidney and therefore no left ureter, there was no need to identify it. There was no sign of the fistula. Six months later, an MRI showed an absent left hemiuterus and a normal right hemiuterus with a patent cervix, and the patient's menses were painless.
# Discussion
Our case was similar to no eponym syndrome because of the uterus didelphys and unilateral renal agenesis. Our case differs from no eponym syndrome because of the unilateral cervical agenesis and the lack of hematocolpos. The surgical approach is different for each, vaginal for the no eponym syndrome [fig_ref] Figure 1: The usual "no eponym" anomaly syndrome before menstruation begins [/fig_ref] versus abdominal for cervix aplasia (our case) [fig_ref] Figure 3: Our case report [/fig_ref].
Morcellation of a congenital hemi-uterus is an ideal surgical option especially for the young patient in whom there is a volume of solid benign tissue to be removed. In our patient and in the more common no eponym anomaly, no kidney and no ureter are present on the side being operated upon.
No unusual blood loss occurred. The morcellation itself only took a few minutes. The main lateral uterine blood vessels and the smaller vessels in the base and medially were readily controlled.
Whereas with the usual no eponym syndrome, dysmenorrhea begins many months or even a year after menarche; in our case, there was severe pain with the first menses because of cervix aplasia, hematometra, retrograde menstruation, hematosalpinx, and eventually pelvic endometriosis.
Although MRI imaging is more precise than pelvic ultrasound, the first MRI missed the left cervix aplasia.
Prior to the first surgery, it was assumed that the patient had the usual no eponym syndrome with a unilateral distal vaginal agenesis, which is treated by vaginal medial marsupialization of the upper obstructed vagina. Thus, the surgeon should be prepared for variations of the basic anomaly.
In retrospect, other clues were the lack of left hematocol-pos, the left adnexa being elevated by the left hematometra on imaging, and a palpable lower abdominal pelvic mass.
The fistula rather than hemihysterectomy was created because of the desire to preserve all possible fertility and the young age of the patient. The fistula failed and rather than make another fistula to evacuate the hematometra with the danger of infection, the left hemiuterus was removed by laparoscopic morcellation.
Vaginal surgery in a teenager may be difficult because of a deep, narrow vagina and difficult exposure access. There are inherent problems in surgical management. The precise details of the anomaly may not be discovered until the surgery is in progress. The surgeon wishes to preserve possible fertility, and the informed consent usually comes from the parents based on the expected findings.
## Literature review
In 1999, [bib_ref] Laparoscopic hemi-hysterectomy in treatment of a didelphic uterus with a hypoplastic cervix..., Lee [/bib_ref] there was a report of a 17-year-old with a uterus didelphys with a right hematometra, hypoplastic cervix, a right obstructed upper hemivagina, and an ipsilateral renal agenesis. The authors indicated that the ". . . majority of clinicians view hysterectomy as the optimal primary surgical management." The right hemiuterus was removed by laparoscopy rather than attempting to make a fistula for fear of infection. They wrote, "laparoscopic hemihysterectomy is a rather sophisticated technique. However, the automatic endoscopic stapler and suturing technique make these procedures possible."
Our experience with laparoscopic morcellation was the opposite. We found it relatively easy, fast, and without the need for suturing. In addition, there was no need to search for a ureter.
In 2001, 5 the same group wrote, "although treatment of didelphic uterus with a hypoplastic cervix is controversial, we believe that creation of the neocervix with prophylactic endometrial ablation may be a conservative effective modality." Their article favored the latter. They reported 2 similar patients age 18 and 16 with a hypoplastic cervix, didelphic uterus, and ipsilateral renal agenesis. In the first, laparoscopic salpingectomy for hematosalpinx was done. Then they opened the hematometra, suctioned it out, and stitched it closed. Leaving the uterus in place, they did a uterovaginal canalization and endometrial ablation. Their second case was similar with a right cervix aplasia. A laparoscopic salpingectomy was done followed by a vaginal creation of a neocervix and an endometrial ablation. They recommended a laparoscopic salpingectomy to prevent an ectopic pregnancy, leaving the obstructed hemiuterus, followed by a vaginal creation of a neocervix and endometrial ablation. Apparently they developed their procedure because of the technical difficulty of removing the obstructed hemiuterus by laparoscopy using traditional surgery. Their procedure is lengthy and leaves the problems of endometrial ablation in the young patient. There have not been any other reports of their procedure in the United States literature.
A 14-year-old 6 had severe menstrual backache and urinary retention due to a large right hematometra of a didelphic uterus in the Pouch of Douglas with cervical aplasia, and ipsilateral renal aplasia. An abdominal right hemihysterectomy was done.
A 48-year-old had bilateral cervical agenesis of uterus didelphys that was treated by an abdominal bilateral hemi-hysterectomy. [bib_ref] Uterus didelphys with cervical agenesis associated with adenomyosis, a leiomyoma and ovarian..., Yang [/bib_ref] Despite the title of a paper in 1979, the operative note favors the no eponym syndrome. When the upper hematocolpos septum was incised without any other surgery, the hematocolpos and hematometra drained old menstrual blood. 8
# Conclusion
Our patient had a rare congenital anomaly that was a variation of an unusual syndrome (no eponym syndrome) anomaly. The latter is characterized by uterus didelphys, unilateral aplasia of one distal vagina, and almost 100% ipsilateral renal aplasia.
The initial MRI did not discover the single cervix aplasia, and the preoperative diagnosis was assumed to be the no eponym syndrome, which is treated by the vaginal route of a large marsupialization of the obstructed hematocolpos. Thus, the surgeon has to have experience with anomalies and be prepared for unexpected situations.
In retrospect, the clues to our case were immediate severe dysmenorrhea starting with the first menses; the image showed elevated adnexa on the affected side, and a tense tender cystic hematometra whose lower pole was relatively high.
In general, the usual approach to complete cervix agenesis is abdominal hysterectomy, because attempts to make a fistula usually are unsuccessful, and there is a danger from severe infection.
Nevertheless, after informed consent because of the young age and the desire to preserve all possible fertility, an attempt was made to create a fistula cervix. After it failed, we did a laparoscopic morcellation of the obstruction hemiuterus.
We found laparoscopic morcellation to be relatively fast, simple, and with essentially no blood loss. An additional factor was the ipsilateral absent kidney and the ureter.
Two recent papers 9,10 report a long-term rare complication of small pieces of tissue left after morcellation of the uterus forming benign painful tumors.
Donnez et al 9 reported on a woman age 48 who had an uneventful laparoscopic morcellation subtotal hysterectomy because of menorrhagia and myomas. The histology was myomas and adenomyosis. She was given estrogen hormone replacement therapy. Five years later, she had pain and a 4-cm mass in the pararectal fossa that was an adenomyoma due to a piece of tissue left after morcellation.
Paul and Koshy 10 reported multiple parasitic myomas after laparoscopic myomectomy with morcellation.
The surgeon should endeavor to remove any small pieces of morcellated tissue in the abdomen, perhaps by extensive washing. Patients should have long-term follow-up.
The usual morcellator is for single patient use. It is contraindicated for use on vascularized tissue or as a dissecting tool. It is not used for possible malignancy.
[fig] Figure 1: The usual "no eponym" anomaly syndrome before menstruation begins. [/fig]
[fig] Figure 2: The usual "no eponym" anomaly syndrome after the start of menstruation. The hematocolpos is retroperitoneal and deep in the pelvis, requiring a vaginal surgical approach. [/fig]
[fig] Figure 3: Our case report. The obstruction was left cervix aplasia. The hematometra is intraperitoneal and surgery for it requires a laparotomy or laparoscopy. The hematometra has elevated the adnexa. [/fig]
[fig] Figure 4: Laparoscopic morcellation of enlarged left uterus. Normal right uterus with ovary and tube. Right sacro-uterine ligament and cul-de-sac visible. Diluted blood in deep pelvis. [/fig]
|
Exploring knowledge of parents and caregivers on cancer symptoms in children: an observational study regarding the need for educational tools and health promotion in low- and middle-income countries
Background: Although most cases of childhood cancer are unlikely to be prevented, by today's standards, most children with cancer can now be cured. However, disparities about survival exist among countries; in Mexico, the overall survival is 49.6%, with 70% of childhood cancers diagnosed at advanced stages. Therefore, parents and caregivers must have optimal knowledge of the early signs and symptoms of childhood malignancies as they are largely nonspecific. This study was designed to explore the current knowledge of childhood cancer among parents and caregivers in Mexico and identify the need for education and health promotion in low-and middle-income countries.Methods: An online survey of 112 parents and caregivers was performed to assess their knowledge of childhood cancer, focusing on the signs and symptoms and early diagnostic strategies.Results: Sixty-nine (61.6%) mothers, 23 (20.5%) fathers, 17 (15.2%) familiar caregivers, and 3 (2.7%) non-familiar caregivers responded. Forty-six (41.1%) respondents said that they knew a child diagnosed with cancer, 92.9% mentioned leukemia as the most common type of cancer among children, the most highly ranked option when asked which sign/symptom they considered as a warning for suspicion was growth/lump in any part of the body, 97.3% considered that an early diagnosis is related to a higher cure rate, and 92.9% expressed the desire to receive reliable information about childhood cancer.Conclusions:Although parents and caregivers have some knowledge of childhood cancer, there are concepts that should be reinforced to improve their understanding of this group of diseases, as they are the frontline for children to seek medical attention. In the future, the use of tools that help educate more caregivers will strengthen knowledge and contribution regarding this issue and promote the generation of public policies that support the early diagnosis of childhood cancer.
# Background
Cancer is a global health problem in both the pediatric and adult populations. Recently, the diagnosis of this group of diseases has increased in the pediatric population because of several factors. Nevertheless, making an early diagnosis is of great importance, which is crucial for treatment and the improvement of patient outcomes, a critical factor in low-and middle-income countries (LMICs).
There are specific challenges unique to LMICs that result in pediatric cancer diagnostic delays. However, only a few articles have been published in developing countries about early diagnosis [bib_ref] Factors Associated with delayed cancer diagnosis in Egyptian children, Abdelkhalek [/bib_ref] [bib_ref] Delay and causes of delay in the diagnosis of childhood cancer in..., Stefan [/bib_ref] [bib_ref] Factors influencing time to diagnosis of childhood cancer in Ibadan, Bo [/bib_ref] [bib_ref] Clinical and social factors that affect the time to diagnosis of mexican..., Fajardo-Gutiérrez [/bib_ref]. These specific challenges include cultural and socioeconomic differences. Moreover, according to other authors, other factors related to diagnostic delay were patient-related (e.g. age, sex, family size, parental education, residence, socioeconomic level, type of cancer, and presenting symptoms), physician-related, or health system-related. A study conducted in 2022 in Mexico reported that children whose parents had the lowest educational level had longer diagnostic delays than those children with parents with the highest educational level [bib_ref] Clinical and social factors that affect the time to diagnosis of mexican..., Fajardo-Gutiérrez [/bib_ref]. Of these factors, parental knowledge and education are important modifiable factors that we can address, and studies have reported that parental knowledge and education may or may not be related to cancer diagnostic delays in children [bib_ref] Factors Associated with Delayed Cancer Diagnosis in Egyptian Children, Abdelkhalek [/bib_ref] [bib_ref] Determinants of delayed diagnosis among pediatric cancer patients from Ayder comprehensive specialized..., Berhane [/bib_ref] [bib_ref] Factors associated with the latency to diagnosis of childhood cancer in Peru, Vasquez [/bib_ref] [bib_ref] Diagnosis delays in childhood cancer: a review, Dang-Tan [/bib_ref].
Although most cases of childhood cancer are unlikely to be prevented, by today's standards, most children with cancer can now be cured [bib_ref] Sustaining Innovation and Improvement in the Treatment of Childhood Cancer: Lessons from..., Pritchard-Jones [/bib_ref]. Unfortunately, substantial disparities in survival rates exist among countries, as 94% of all cancer deaths worldwide among children occur in LMICs [bib_ref] Sustaining Innovation and Improvement in the Treatment of Childhood Cancer: Lessons from..., Pritchard-Jones [/bib_ref].
In Mexico, the overall survival rate for childhood cancer is 49.6%, with variability ranging from 6.8% to 64.1% among states. Moreover, approximately 75% of cases are diagnosed at advanced stages, an important risk factor that has a profound impact on patient survival.
Because the signs/symptoms of cancer in children are largely non-specific, it is essential that parents and caregivers have optimal knowledge regarding the early signs/symptoms of childhood malignancies [bib_ref] Diagnosis delays in childhood cancer: a review, Dang-Tan [/bib_ref]. Therefore, this study was designed to characterize the current knowledge of childhood cancer among a population of parents and caregivers in Mexico.
# Methods
## Setting and study population
We conducted a one-time survey involving Mexican parents and caregivers of children and adolescents of the ages 0-18 years, who attended an education and child development forum named "ExpoLearning Kids in January 2020. The forum was attended approximately by 600 individuals, of whom 112 parents and caregivers, who were personally invited, agreed to participate in this study. None of the 112 parents and caregivers declined to participate. This forum focused on providing information based on lectures and debates with parents and caregivers about child and youth education, health, nutrition, mindfulness, and physical and artistic activities. This forum was held in Monterrey, Nuevo Leon, Mexico, which includes its metropolitan area (composed of 13 cities), and is the second largest metropolitan area in Mexico, with an approximate population of 5 million inhabitants. The Mexican health system comprises two sectors: public and private. In Monterrey, public healthcare is delivered by several institutions: [bib_ref] Sistema de Salud de México, Dantés [/bib_ref] (1) the Mexican Social Security Institute provides compulsory health insurance to workers in the formal labor market and their families (68 million individuals in Mexico and 4.4 million individuals in Nuevo Leon); (2) the Secretary of Health, the National Institutes of Health, and the secretaries of health at the state level provide care to individuals affiliated to the Social Protection System in Health through its operating arm Seguro Popular (54 million individuals in Mexico and 0.7 million individuals in Nuevo Leon); (3) and state, oil, army, and navy workers (14 million individuals in Mexico and 0.3 million individuals in Nuevo Leon) have their own social security institutions and healthcare delivery mechanisms. In contrast, the private healthcare sector consists of individuals who own private insurance (approximately 3% of the total population) and small businesses that provide insurance to their employees (approximately 2% of the total population). Regarding socioeconomic status, approximately 60% of the residents in Monterrey are in the middle class, with an average monthly salary of $1,155.
## Survey design
The survey was designed to evaluate the following: (1) the number of children cared for, healthcare, and type of education; (2) whether the respondents knew a child diagnosed with cancer and their relationship with him/ her; (3) knowledge regarding several childhood cancer facts; (4) knowledge about the signs/symptoms of childhood cancer; (5) early diagnostic strategies; and (6) exposure to reliable information about cancer in children and adolescents. The performance of this survey was evaluated in a pilot study involving 30 parents and caregivers in a waiting room of a private children's outpatient clinic located in Monterrey, which is a reference clinic in the metropolitan area, having a high level of pediatric clinical expertise. Parents and caregivers were chosen and invited while waiting for their child's consultation. The questions' adequacy was evaluated using the Delphi technique [bib_ref] Descripción y Usos Del Método Delphi En Investigaciones Del Área de La..., Varela-Ruiz [/bib_ref] with a panel of nine experts in Pediatric Oncology.
## Survey tool
The questionnaire included 29 items. Eight questions were designed to obtain the demographic characteristics of the respondents. Four questions were designed to obtain the following information: (1) the number of children cared for, (2) physician responsible for evaluating children, (3) physician's practice type, and (4) type of school attended by children. Furthermore, two questions regarding the experience of knowing a child diagnosed with cancer and their relationship with him/her were included. Six questions were designed to evaluate knowledge regarding the following aspects: (1) definition of cancer, (2) first thought when hearing the word "cancer, " (3) the most common type of childhood cancer, (4) how frequent they think childhood cancer is, (5) the causes of cancer in children, and (6) the most common pediatric cancer age group. Of these six questions, all but the third and fourth ones were answered using a 5-point Likert scale that measures the level of agreement/disagreement. Three questions about children's and adolescent's cancer symptoms were included: (1) warning signs/symptoms (using a 5-point Likert scale measuring the level of agreement/disagreement), (2) whether they consider that any symptom lasting ≥ 2-3 weeks in children should be suspicious of cancer, and (3) whether they think that Down syndrome is a risk factor for childhood cancer. Two questions were designed to evaluate knowledge about early diagnostic strategies: (1) preventive measures to assure the early diagnosis of childhood cancer (using a 5-point Likert scale measuring the level of agreement/disagreement) and (2) whether they consider that early diagnosis is related to a higher cure rate. The parents' and caregivers' exposure to reliable information about pediatric cancer was evaluated by asking the following: (1) whether they had ever received information about childhood cancer, (2) the source of information, (3) knowledge about vaccines that can help prevent certain types of childhood cancer, and (4) whether they wanted to receive information about the warning signs/symptoms, early diagnosis, and treatment of cancer in children.
## Study design and participants
The survey was written in Spanish in Google Forms and distributed to 112 parents and caregivers; each interview was completed in approximately 15 min. Restrictions were made for each question so that responses could not be sent to the final database if there were missing answers. Only complete records were processed; no surveys were excluded. Participation in the study was voluntary, responses were anonymous, and no incentives or compensations were offered. Before the initiation of the study, permission was obtained from the forum concerned. The questionnaire and methodology for this study were approved by the Institutional Human Research and Ethics Review Boards of the Escuela de Medicina y Ciencias de la Salud, Tecnológico de Monterrey (number P000112-LLA2020-CEIC-CR003), and all methods were performed according to relevant guidelines and regulations. Before obtaining their consent, all parents and caregivers were provided with information about the study, concerning its objective, plan, and benefits.
# Statistical analysis
Quantitative variables are reported as medians and interquartile ranges (25 th to 75 th percentiles) and were analyzed using the Mann-Whitney U test. Categorical variables were analyzed using Fisher's exact test or the chi-square test. In addition to descriptive analysis, we arranged the survey records from the respondents with experience of knowing a child diagnosed with cancer and compared them with those without this experience. P values of less than 0.05 were used to denote statistical significance, and the alpha reliability of all questions was adequate (α = 0.899). The survey was analyzed using Statistical Package for the Social Sciences, version 27.0.
# Results
Of the parents and caregivers in this study, 61.6% were mothers, 29.5% were between the ages of 30 and 34 years, 59.8% had a professional degree, and 77.7% stated that the physician responsible for evaluating their children at least once a year was a pediatrician [fig_ref] Table 1: Demographic characteristics of Parents/Caregivers, health care and educational characteristics of children taking... [/fig_ref].
Among the 112 parents and caregivers, 41.1% knew a child diagnosed with cancer. Regarding the definition of cancer, 83.9% strongly agreed that it is a worrisome disease, and 71.4% strongly agreed that it is a painful illness. The question regarding the first thought parents and caregivers had when hearing the word "cancer" showed that 64.3% strongly agreed is "suffering, " and 62.5% strongly agreed is "chemotherapy. " More than 90% mentioned leukemia as the most common type of cancer in children, and 50.9% stated that childhood cancer is frequent. Moreover, 20.5% strongly agreed the most common pediatric cancer age group is the 6-12 years year group. Among the provided options for the causes of cancer in children, the most highly ranked option was radiation exposure [fig_ref] Figure 1: The parents' and caregivers' perceptions of the causes of cancer in children [/fig_ref].
A question was asked to determine which signs/symptoms the parents and caregivers considered a warning sign or symptom of childhood cancer [fig_ref] Figure 2: Distribution of the parents' and caregivers' statements about the warning signs and... [/fig_ref]. The three most highly ranked options were growth/lump in any part of the body, unexplained bruises, and weight loss. Among the respondents, 58.9% answered that any symptom lasting ≥ 2-3 weeks in children should be suspicious for cancer, and 88.4% did not consider Down syndrome as a risk factor for cancer.
Almost all respondents (97.3%) considered that the early diagnosis of cancer in children is related to a higher cure rate [fig_ref] Table 3: Parents'/caregivers' statements about childhood cancer symptoms, early diagnosis strategies, and exposure to... [/fig_ref] , and the three most highly ranked options regarding the preventive measures they considered important to ensure the early diagnosis of childhood cancer were as follows: improving education about childhood cancer for parents and caregivers, attending wellchild visits, and healthy eating [fig_ref] Figure 3: Perceptions of the parents and caregivers on the preventive measures to ensure... [/fig_ref].
An analysis of the parents' and caregivers' exposure to information about childhood cancer showed that 66.1% had never received information. More than 70% stated that they did not know that there are vaccines that can help prevent certain types of childhood cancer; and 92.9% expressed the desire to receive information about the signs/symptoms, early diagnosis, and treatment of cancer in children [fig_ref] Table 3: Parents'/caregivers' statements about childhood cancer symptoms, early diagnosis strategies, and exposure to... [/fig_ref].
The following relationships were found to be statistically significant, considering the two groups regarding knowing/not knowing a child with cancer (data not shown): defining cancer as a fatal disease (W = 1073.50, p = 0.006); defining cancer as a disease that always causes death (W = 1060.00, p = 0.005), with the group that do not know a child with cancer more likely to refer to both definitions; vitamin supplementation for children (W = 1053.00, p = 0.005) and getting routine specific imaging examinations done (i.e., ultrasonography, computed tomography, and magnetic resonance imaging) (W = 1052.50, p = 0.004) as preventive measures to assure the early diagnosis of childhood cancer; exposure to any information about childhood cancer (X 2 = 8.994; df = 1; p = 0.003), pediatricians as the source of information about cancer in children (X 2 = 3.865; df = 1; p = 0.049), and knowledge about vaccines that can help prevent certain types of childhood cancer (X 2 = 7.129; df = 1; p = 0.008).
# Discussion
Continued medical advancements have demonstrated that childhood cancer cannot be prevented. Therefore, the cornerstone for improving outcomes using more effective and less toxic treatments relies on several factors, of which early diagnosis is of great importance.
However, the pathway toward better outcomes for children with cancer among LMICs is often impeded by numerous obstacles, with a timely diagnosis relying on multiple factors: cancer biology, patient characteristics, physician experience/education, access to healthcare services, and children's family environment [bib_ref] Clinical and social factors that affect the time to diagnosis of mexican..., Fajardo-Gutiérrez [/bib_ref] [bib_ref] En La Infancia y Adolescencia, Vega-Vega [/bib_ref]. However, there is a need to focus on modifiable factors, such as knowledge of childhood cancer in parents and caregivers, to help diagnose this group of diseases earlier. For this reason, this study was designed to evaluate the parents' and caregivers' knowledge about this topic to have a better understanding of the influence of this factor in the delay of pediatric cancer diagnosis; thus we can focus on an intervention to implement effective programs and public policies to ensure the timely diagnosis of cancer in children.
To the best of our knowledge, few studies in Latin America have assessed what parents and caregivers know about childhood cancer [bib_ref] Clinical and social factors that affect the time to diagnosis of mexican..., Fajardo-Gutiérrez [/bib_ref] [bib_ref] Factors associated with the latency to diagnosis of childhood cancer in Peru, Vasquez [/bib_ref] , and two studies of this kind were found, which were conducted in Turkey [bib_ref] Knowledge of Turkish Mothers with Children in the 0-13 Age Group about..., Demirbaǧ [/bib_ref] and Greece, respectivey [bib_ref] Public Knowledge, Beliefs and Practices in Greece about Cancer Etiology and Prevention, Charalabopoulos [/bib_ref]. Evaluating parents' and caregivers' current knowledge of cancer is important because parent-related factors, such as lower educational level, younger age, and lower socioeconomic status, can be associated with a greater diagnostic delay of cancer in pediatric patients [bib_ref] Clinical and social factors that affect the time to diagnosis of mexican..., Fajardo-Gutiérrez [/bib_ref] [bib_ref] Factors Associated with Delayed Cancer Diagnosis in Egyptian Children, Abdelkhalek [/bib_ref]. In this study, more than half of the respondents had a professional degree, and 29.5% were between the ages of 30 and 34 years. However, 66.1% stated that they had never received any information about childhood cancer, indicating that even though more than half of them had a high educational level, there is a need for better education regarding pediatric cancer.
More than 75% of the parents and caregivers stated that the physician responsible for evaluating their children at least once a year was a pediatrician, indicating the importance of regular well-child visits to establish continuous care and surveillance of all children, which can help identify subtle symptoms that may not be noted initially as important by the children's families [bib_ref] Utilization of Healthcare Services among Children Members of Medical Insurance for a..., Pérez-Cuevas [/bib_ref].
Several parents and caregivers reported that they knew a child with cancer, whereas the top definitions of cancer that the respondents strongly agreed to were that it is a worrisome and painful disease; and when the respondents heard the word "cancer, " most of them thought of "suffering. " These findings highlight how parents and caregivers perceive this disease, and although cultural background must always be considered [bib_ref] Cultural Influences in Pediatric Cancer: From Diagnosis to Cure/End of Life, Gray [/bib_ref] , cancer in children is mostly identified as the most frightening illness. An analysis of contemporary representations of childhood cancer in Romanian media [bib_ref] 3390/ child ren61 10126. • fast, convenient online submission • thorough peer..., Teodorescu [/bib_ref] reported that a diagnosis that can lead to premature death, such as cancer, is perceived as a transgression that often inspires negative reactions, sentiments, and stigma. This could support this study's findings where the respondents perceived that childhood cancer is worrisome and painful and the overall association of cancer with words, such as "suffering. " The incidence of childhood cancer varies among different countries; particularly, in Mexico, it increased from 133.5/million children in 2007 to 150.1/million children in 2015 [bib_ref] Descriptive epidemiology in Mexican children with cancer under an open national public..., Rivera-Luna [/bib_ref]. Moreover, although the most prevalent cancer may also vary, in Mexico, acute lymphoblastic leukemia (ALL) takes the lead [bib_ref] Descriptive epidemiology in Mexican children with cancer under an open national public..., Rivera-Luna [/bib_ref]. In this study, the most common type of childhood cancer that the respondents reported was leukemia; 20.5% of the respondents strongly agreed that the most common age group affected by cancer is the 6-12 years year group. These findings suggest that although they are aware of the most common type of childhood cancer, the respondents lack knowledge regarding the most common age group they must keep a close eye on because the age at which the incidence of pediatric cancer is the highest is between 2 and 6 years [bib_ref] Descriptive epidemiology in Mexican children with cancer under an open national public..., Rivera-Luna [/bib_ref].
When the respondents were asked to determine which options they considered as the causes of childhood cancer; the top three responses with the highest ranks of respondents who strongly agreed were radiation exposure (52.0%), smoking during pregnancy (27.0%), and family history of cancer along with bad eating habits (both with 20.0%). Since ALL is the most prevalent type of childhood cancer in Mexico, it is imperative that parents and caregivers know which factors can be related to its etiology. Several factors might contribute to the etiology of ALL or might have a protective effect against this disease, including paternal smoking [bib_ref] Parental Smoking and Childhood Leukemia, Chang [/bib_ref] [bib_ref] Tobacco Smoke Exposure and the Risk of Childhood Acute Lymphoblastic Leukemia and..., Chunxia [/bib_ref] and breastfeeding [bib_ref] Breastfeeding and childhood leukemia incidence: a meta-analysis and systematic review, Amitay [/bib_ref] [bib_ref] The impact of breastfeeding on the childhood acute leukemia risk among children..., Massoud [/bib_ref] , respectively. The parents and caregivers in this study showed a lack of knowledge regarding the protective effect breastfeeding might have against childhood cancer, particularly ALL, as only a few strongly agreed that lack of breastfeeding might be a cause of cancer in children, in almost an even proportion to witchcraft and divine retribution.
A Mexican multicenter cohort study showed that the average time from the onset of symptoms to the diagnosis of childhood cancer was 43.5 ± 22.5 days [bib_ref] Estímulo Iatrotrópico y Tiempo Al Diagnóstico En Pacientes Pediátricos Con Leucemia Linfoblástica..., Lora [/bib_ref]. Heightened recognition of the early signs/symptoms among the general population and primary care providers is essentially the first factor in a chain of events that ultimately lead to a child's prompt diagnosis. In this study, the top three signs and symptoms that the respondents strongly agreed to be considered a warning sign/symptom of childhood cancer were growth/lump in any part of the body, unexplained bruises, and weight loss. The literature highlights that the most common symptoms of cancer in [bib_ref] Signs and symptoms of childhood cancer: a guide for early recognition, Fragkandrea [/bib_ref]. Although all signs/symptoms included in this question should be considered a warning and even though results showed that respondents have some idea of which symptoms are more common, there is still a low level of knowledge considering that they did not even mention fever as one of the top three signs/symptoms of childhood cancer, and that the biggest challenge in diagnosing this group of diseases is that they often present with non-specific symptoms that less frequently have serious outcomes.
Among the respondents, 58.9% answered that any symptom lasting ≥ 2-3 weeks should be suspicious of cancer. This is remarkable as it has been reported that the average time from symptom onset and the first medical appointment received was 9.2 ± 16.6 days, followed by 34.3 ± 16.3 days from the initial appointment to the diagnostic confirmation of cancer, after an average of 2.3 medical consults [bib_ref] Estímulo Iatrotrópico y Tiempo Al Diagnóstico En Pacientes Pediátricos Con Leucemia Linfoblástica..., Lora [/bib_ref] , highlighting the need for searching medical care as soon as parents and caregivers detect persistent symptoms. A qualitative study of parents with children newly diagnosed with cancer showed that most of them initially attributed their child's symptoms to minor illnesses, particularly when the latter was coupled with normal behavior, interpreted as "incompatible" with cancer [bib_ref] Shouting from the Roof Tops": A Qualitative Study of How Children with..., Clarke [/bib_ref].
An important contributing factor to mortality in children with cancer in Mexico is diagnosis at advanced disease stages, estimated to happen in almost 70% of cases, in contrast to high-income countries, where approximately 70% of cases are diagnosed with local or regional disease [bib_ref] Stage at diagnosis for childhood solid cancers in Australia: a population-based study, Youlden [/bib_ref]. Among the respondents, 97.3% considered that the early diagnosis of cancer in children is related to a higher cure rate and that the three top preventive measures were as follows: improving education about childhood cancer in parents and caregivers, attending well-child visits, and healthy eating. Furthermore, 92.9% of the respondents expressed the desire to receive information about childhood cancer. Although most respondents were aware of the important role of the early diagnosis of pediatric cancer in achieving better outcomes, the results also highlighted that the respondents have the desire to receive trustable information, emphasizing the deficiency of knowledge of childhood cancer.
This study has limitations, including a limited number of respondents. Moreover, those willing to respond may not be representative of the entire population of Mexican parents and caregivers. It could also be possible that the characteristics of the parents and caregivers who responded may be different from those of non-respondents, which might have affected the study results and interpretations. Besides, this study does not address the perceptions of parents and caregivers on the communication about childhood cancer from different sources of information or other cultural factors that can influence the knowledge they have about this disease. Regarding the next steps following this study, it would be interesting to replicate the distribution of the survey and increase the sample size to have a more balanced and representative group of respondents of the entire population of Mexican parents and caregivers, including those of different geographic areas in Mexico and educational level and those receiving not only private but also public healthcare. This would help generalize the results to the general Mexican population.
# Conclusions
Although a long pathway of factors is involved toward achieving the timely and optimal diagnosis of cancer in children, strengthening the use of preventive childhood healthcare services and public campaigns to enhance parents' and caregivers' knowledge of childhood cancer is mandatory to decrease the delay interval between symptom presentation and the diagnosis of cancer in children, to achieve better outcomes with fewer complications. In the future, the use of tools such as applications that help educate more parents and caregivers, will strengthen the knowledge and contribution regarding this issue because educating the population is of vital importance for the prompt diagnosis of the disease. Finally, the development of public policies that support the early diagnosis of cancer, particularly in children, who is a doubly vulnerable population, must be promoted.
[fig] Figure 1: The parents' and caregivers' perceptions of the causes of cancer in children. The respondents were instructed to rank several causes of cancer in children. The line in the middle of the box represents the median; the edges of the box represent the 25 th and 75 th percentile interquartile ranges; the whiskers represent the minimum and maximum observations children are as follows: (1) pallor, fatigue, and malaise; (2) fever; and (3) recurrent or treatment-resistant infections [/fig]
[fig] Figure 2: Distribution of the parents' and caregivers' statements about the warning signs and symptoms of cancer in children. The respondents were instructed to rank several signs and symptoms of cancer in children. The line in the middle of the box represents the median; the edges of the box represent the 25 th and 75 th percentile interquartile ranges; the whiskers represent the minimum and maximum observations [/fig]
[fig] Figure 3: Perceptions of the parents and caregivers on the preventive measures to ensure the early diagnosis of childhood cancer. The respondents were instructed to rank several preventive measures for the early diagnosis of cancer in children. The line in the middle of the box represents the median; the edges of the box represent the 25 th and 75 th percentile interquartile ranges; the whiskers represent the minimum and maximum observations [/fig]
[table] Table 1: Demographic characteristics of Parents/Caregivers, health care and educational characteristics of children taking care of [/table]
[table] Table 3: Parents'/caregivers' statements about childhood cancer symptoms, early diagnosis strategies, and exposure to information about cancer in children Opinion about any symptom lasting more than 2 to 3 weeks (including fever) in children should be suspicious for cancer An early diagnosis of cancer in children is related to a higher cure rate Other healthcare professional (nurse, general physician, family physician) Knowledge about vaccines that can help prevent certain types of childhood cancer Willingness to receive reliable information about warning signs and symptoms, early diagnosis and treatment of cancer in children [/table]
|
Paracrine control of α-cell glucagon exocytosis is compromised in human type-2 diabetes
Glucagon is released from pancreatic α-cells to activate pathways that raise blood glucose. Its secretion is regulated by α-cell-intrinsic glucose sensing and paracrine control through insulin and somatostatin. To understand the inadequately high glucagon levels that contribute to hyperglycemia in type-2 diabetes (T2D), we analyzed granule behavior, exocytosis and membrane excitability in α-cells of 68 non-diabetic and 21 T2D human donors. We report that exocytosis is moderately reduced in α-cells of T2D donors, without changes in voltagedependent ion currents or granule trafficking. Dispersed α-cells have a non-physiological V-shaped dose response to glucose, with maximal exocytosis at hyperglycemia. Within intact islets, hyperglycemia instead inhibits α-cell exocytosis, but not in T2D or when paracrine inhibition by insulin or somatostatin is blocked. Surface expression of somatostatin-receptor-2 is reduced in T2D, suggesting a mechanism for the observed somatostatin resistance. Thus, elevated glucagon in human T2D may reflect α-cell insensitivity to paracrine inhibition at hyperglycemia.
G lucagon is released from pancreatic α-cells and counteracts the glucose-lowering actions of insulin by stimulating gluconeogenesis and hepatic glucose output. Initially thought of only as part of the body's defense against hypoglycemia 1 , it is now clear that inadequate glucagon levels also contribute to diabetic hyperglycemia and present a challenge for diabetes management. Glucagon secretion is triggered by low blood glucose and suppressed at physiological glucose levels, and both α-cell intrinsic and paracrine mechanisms have been cited to explain these effects. In the intrinsic models, glucose metabolism and the generation of ATP play a central role, either through subtle, K ATP channel-dependent depolarization of the resting membrane potential and subsequent inactivation of Na + -channels 7-10 , or as consequence of glucose-induced activation of the sarco/endoplasmic reticulum Ca 2+ -ATPase that leads to closure of store-operated channels and hyperpolarization. In addition, intrinsic glucose-dependent cAMP signaling may play a role. However, none of these models fully explain the glucose concentration dependence of glucagon secretion, in particular in the hyperglycemic range.
Glucagon secretion is also under paracrine control from neighboring βand δ-cells, and the inhibitory effects of somatostatin, insulin, and GABAhave long been recognized. Paracrine inhibition is likely to play a role at elevated glucose levels, when βand δ-cells are active. Indeed, glucagon is secreted in pulses that are anti-synchronous to pulses of insulin and somatostatin. This relationship is important for the postprandial suppression of glucagon secretion and is lost in type-2 diabetes and pre-diabetes. α-cells express the somatostatin receptor SSTR2, which leads to hyperpolarization via activation of GIRK-channels. Somatostatin also inhibits the exocytosis machinery via calcineurin, and inhibits α-cell exocytosis by effectively decreasing cytosolic cAMP. Insulin receptor signaling is required for the suppression of glucagon secretion in vivo, but the precise mechanisms behind this are still debated. Decreased sensitivity to insulin (or somatostatin) may therefore underlie the inadequate glucagon secretion in type-2 diabetes.
Glucagon is stored in~7000 granules (diameter~270 nm) and secreted by Ca 2+ -and SNARE protein-dependent exocytosis. At any time, only~1% of these granules are in a releasable state that can undergo exocytosis upon Ca 2+ -influx. Paracrine signaling and glucose regulate glucagon secretion at least in part by affecting the size of this releasable pool of granules. In many endocrine cells, secretory granules become release ready by sequential docking at the plasma membrane and assembly of the secretory machinery (priming). Although disturbances in these steps have been documented in β-cells of type-2 diabetic donors have not yet been studied in α-cells. An obstacle for understanding the regulation of α-cells has been the difficulty to isolate intrinsic and paracrine factors of α-cell regulation, as well as species differences between humans and rodent models. Glucagon secretion in vivo and in intact islets is affected by the presence of neighboring cell types, while single-cell electrophysiological measurements are invasive and may not reflect the in vivo situation.
In the current work, we took an optical approach to study glucagon granule exocytosis in α-cells of non-diabetic (ND) and type-2 diabetic (T2D) human subjects. We report that α-cells within intact islets respond with physiological inhibition of exocytosis by elevated glucose, whereas dispersed α-cells have a V-shaped response to glucose due to the lack of paracrine inhibition by insulin and somatostatin from neighboring βand δ-cells. Importantly, α-cells of T2D are resistant to inhibition by insulin and somatostatin, which might underlie the hyperglucagonemia in type-2 diabetes.
# Results
Exocytosis of glucagon granules in human α-cells. Docking and exocytosis of glucagon granules at the plasma membrane was studied in dispersed islet preparations from 68 non-diabetic (ND) donors that all had glycated hemoglobin HbA1c values <6% (average 5.57 ± 0.29%,. To identify α-cells, we transduced with Pppg-EGFP and the secretory granule marker NPY-mCherry, or with Pppg-NPY-EGFPand S1a, b) to drive expression of fluorescent proteins from the pre-proglucagon promoter (see methods). After culture for 26-48 h, α-cells were imaged by total internal reflection (TIRF) microscopy, which selectively images fluorescence near the plasma membrane (exponential decay constant τ~0.1 µm). The granule marker had a punctate staining pattern and excellent overlap with antiglucagon immunostaining. Local application of elevated K + (75 mM, replacing Na + ) to depolarize the cells resulted in exocytosis, seen as rapid disappearance of individual fluorescently labeled granules (gr, seeand examples in S1B,C). Exposure to elevated K + for 40 s released 0.078 ± 0.004 granules µm −12 (169 α-cells/29 ND donors,, black). Exocytosis proceeded initially with a burst (5.2 × 10 −3 gr µm −2 s −1 during the first 10 s) and decreased later to <0.6 × 10 −3 gr µm −2 s −1 ; these rates are about one-third of those observed in human β-cells. Fitting the cumulative exocytosis (n = 1530 granules) with a double exponential function revealed two components with time constants of τ = 3.6 ± 0.2 s and 19.9 ± 0.9 s. The faster component made up 39 ± 3% of the total response, and likely corresponds to the RRP. Exocytosis occurred in granules that from the start of the experiment had been docked at the plasma membrane. We therefore quantified changes in docked granules during the experiment, by measuring the density of granules in the TIRF field. Stimulated exocytosis partially depleted docked granules, indicating that replacement by docking of new granules is relatively slow (a notion confirmed by a double stimulation protocol,. On average, 13 ± 0.7% of the docked granules were released during the stimulation. Thus, depolarization of α-cells results in exocytosis with biphasic kinetics similar to those in other endocrine cells, indicating the existence of granule pools with differing release probabilities.
Reduced granule docking and exocytosis glucagon in T2D α-cells. During the course of this study, we also received islets from 21 donors that had been clinically diagnosed with type-2 diabetes (T2D), or whose glycated hemoglobin (HbA1c) values were above 6% (average 6.6 ± 0.7%,. In dispersed T2D α-cells, K + -stimulated exocytosis was reduced to 66 ± 8% of that in ND α-cells (p = 0.004, 0.052 ± 0.005 gr µm −2 ; 75 cells/12 donors;, mostly due to a reduced amplitude of the fast component (τ = 2.1 ± 0.1 s, 25 ± 3%, n = 441 granules, p = 0.0008 vs ND). Docked granules were slightly fewer in T2D α-cells (0.57 ± 0.02 gr µm −2 , 106 cells/17 donors) compared with ND (0.61 ± 0.008 gr µm −2 ; p = 0.01, 399 cells/50 donors,. Exocytosis and granule density correlated on a per-donor basis (Pearson r = 0.42, p = 0.006, 41 donors;, and both exocytosis (r = 0.49, p = 0.002;and docked granules (r = 0.37, p = 0.003;anti-correlated with the donor's HbA1c, as is the case in human β-cells. The relationships are surprising given that reduced glucagon secretion should lead to reduced blood glucose and HbA1c values, and suggest that the diabetic state is causal for the reduced release capacity of T2D α-cells.
Voltage-dependent currents are normal in T2D α-cells. Since exocytosis in α-cells depends on Ca 2+ -influx, we characterized voltage-dependent ion currents using patch-clamp electrophysiology. Dispersed ND or T2D α-cells were voltage clamped in whole-cell mode, and subjected to step depolarizations up to +70 mV from a holding potential of −70 mV. Analysis of the resulting inward currents revealed peak Ca 2+and Na + -currentsthat were of similar amplitude in ND and T2D cells. Half-maximal Ca 2+ -current activation was reached at −23 ± 0.4 (ND) and −25 ± 1.4 mV (T2D, n.s.), and half-maximal Na + -current activation was at −25 ± 0.6 (ND) and −24 ± 0.8 mV (T2D, n.s.). We also determined depolarization evoked membrane capacitance increases, a measure of exocytosis. A train of 14 depolarizations to 0 mV lasting 200 ms each resulted in a total capacitance increase of 112 ± 19 fF in ND cells and 78 ± 19 fF in T2D cells. This corresponds to a reduction of exocytosis by 25 ± 10% in T2D (p = 0.1), which is similar to the reduction observed by imaging granule release. Cell size, as assessed by cell capacitance was not different in the two groups. Thus, reduced exocytosis in T2D α-cells cannot be explained by changes in Ca 2+ -channel behavior.
Glucose regulation of glucagon secretion. Next, we determined the physiological glucose dependence of dispersed α-cells by measuring spontaneous exocytosis in a range of ambient glucose concentrations (1, 3, 7, 10, or 20 mM; at least 20 min pre-incubation), without imposing any depolarization. In movies lasting 3 min, we quantified granule exocytosis, docked granules, and the rate of docking of new granules at the plasma membrane. Spontaneous exocytosis was observed in all glucose concentrations, with a bimodal (V-shaped) dose response to glucose. Inhibition was about half at the nadir of 7 mM glucose, and ND and T2D cells behaved essentially identically. Docked granules, and to a lesser degree the rate of docking (in the same cells) likewise had a bimodal response to glucose, with a nadir at 7 mM, c, d). In separate experiments, we noticed that the response to changes in the glucose concentration was slow, and only minor during 2 min observation. Thus, the glucose dependence of dispersed α-cells does not reflect physiological glucagon secretion, and exocytosis is accelerated in the hyperglycemic range instead of being inhibited. The data also suggest that the availability of docked granules is part of the regulation of exocytosis in α-cells.
To experimentally isolate direct glucose effects on the exocytosis machinery, we applied K + -stimulations to α-cells bathed in 1, 7, or 10 mM glucose. This approach primarily elicits exocytosis of granules that are "primed" for exocytosis, the readily releasable pool (RRP). At all glucose concentrations, elevated K + stimulated biphasic exocytosis that by far exceeded spontaneous exocytosis. However, we observed again a bimodal glucose dependence with a nadir at 7 mM (38 cells/6 donors), where K + -induced exocytosis was reduced by about half compared with 1 mM glucose (−45 ± 9%, p = 8 × 10 −4 , 30 cells/6 donors), or 10 mM glucose (−47 ± 6%, p = 5 × 10 −6 , 71 cells/14 donors). This reduction was most prominent during the initial burst phase, which may correspond to the immediately releasable pool of granules in β-cells. At all glucose concentrations, K + -stimulated exocytosis was significantly slower in T2D α-cells than in ND, but the bimodal glucose dependence of docking and exocytosis was preserved. We conclude that the α-cell exocytosis machinery is regulated by the ambient glucose concentration.
Exocytosis in α-cells within intact islets. The unexpected Vshaped glucose response of dispersed α-cells indicates that intrinsic regulation cannot explain the physiological inhibition of glucagon secretion in hyperglycemia. An alternative are paracrine effects from neighboring βand δ-cells, which prompted us to quantify glucose dependent exocytosis in α-cells within intact islets. As expected, exocytosis of α-cells in intact islets varied with the glucose concentration, with inhibition in 7 mM glucose by 67 ± 15% (p = 0.016; 10 islets/3 donors), compared with 1 mM glucose (11 islets/3 donors). Exocytosis was also inhibited in 10 mM (by 56 ± 16% vs 1 mM, p = 0.024, 16 cells/4 donors), in contrast to dispersed α-cells. However, this inhibition was prevented by the SSTR2-specific somatostatin receptor antagonist CYN154806 (200 nM in the bath solution; p = 0.003 vs 10 mM glucose, 19 cells/3 donors). Similarly, block of insulin action with the insulin receptor antagonist S961 (1 μM) prevented inhibition of exocytosis at 10 mM glucose (p = 0.003 vs 10 mM glucose, 11 cells/2 donors), indicating that paracrine signaling is required for proper glucagon control in hyperglycemic conditions. In intact islets of T2D donors, α-cell exocytosis had a bimodal dose response to glucose that lacked inhibition hyperglycemia. At 7 mM glucose (16 islets/4 donors), exocytosis was inhibited to about half compared with 1 mM (21 islets/5 donors), whereas 10 mM glucose has no effect (29 islets/5 donors). Thus, α-cells within intact ND islets have a physiological response to glucose. Disruption of paracrine signaling by antagonists or islet dispersion leads to a V-shaped glucose response, similar to that observed in intact T2D islets.
We confirmed 48 expression of SSTR2 in human α-cells by coimmunostaining the receptor and glucagon in pancreatic sections of 10 human donors (5 ND, 5 T2D;. In ND islets, SSTR2 distribution was mostly confined to the cell membrane of αand other islet cells, whereas in T2D islets the SSTR2 staining was both weaker and largely vesicular. Quantitative analysis confirmed this conclusion and estimated that SSTR2 surface expression is decreased by 44 ± 7%, in T2D (824 cells/5 T2D donors vs 828 cells/5 ND donors. Glucagon levels and distribution were similar in both groups. Insensitivity to somatostatin is therefore the result of excessive receptor internalization, as has recently been shown for pituitary cells.
Paracrine regulation of exocytosis in dispersed α-cells. Glucagon secretion is regulated by a network of paracrine mechanisms, some of which act directly on α-cells. We therefore quantified K + -stimulated exocytosis dispersed α-cells in presence of a panel of islet paracrine effectors (somatostatin (SST, 400 nM), insulin (INS, 100 nM), forskolin (FSK, 2 µM), ϒ-aminobutyric acic GABA (400 nM), adrenaline (ADR, 5 µM), or glutamate (Glut, 1 mM), all present in the bath) at 1 or 10 mM glucose. In 10 mM glucose, the δ-cell hormone somatostatin inhibited K + -stimulated exocytosis by 65 ± 4% (p = 2 × 10 −6 , 53 cells/9 donors, vs control 71 cells/14 donors). β-cell factors likewise inhibited exocytosis, with insulin reducing it by 53 ± 5% (p = 8 × 10 −5 , 53 cells/8 donors) and GABA reducing it by 24 ± 13% (n.s., 14 cells/3 donors). In contrast, adrenaline doubled (p = 4 × 10 −9 , 30 cells/5 donors) and glutamate tripled α-cell exocytosis (p = 14 × 10 −20 , 16 cells/3 donors). Elevated cAMP, after exposure to forskolin, had no effect on exocytosis (n.s., 30 cells/5 donors), in contrast to previous reports. In α-cells of T2D donors, adrenaline accelerated exocytosis about three fold (p = 6 × 10 −9 ; 19 cells/3 donors). In contrast, the inhibition by somatostatin or insulin was lost in T2D. None of the tested compounds affected the density of docked granules, b lower), suggesting that paracrine factors modulate αcell exocytosis by affecting granule priming, rather than docking. In hypoglycemic conditions (1 mM glucose,, neither somatostatin (27 cells/6 donors, blue) nor insulin (24 cells/5 donors, green) affected K + -stimulated exocytosis of dispersed αcells, while adrenaline (p = 2 × 10 −8 , 15 cells/3 donors, pink) and glutamate (p = 9 × 10 −10 , 9 cells/2 donors, orange) accelerated exocytosis 2-3-fold compared with control (1 mM glucose, 30 cells/6 donors). T2D α-cellsbehaved identical to ND α-cells with regard to somatostatin (17 cells/3 donors, blue), insulin (23 cells/3 donors, green) and adrenaline (p = 2 × 10 −12 , 10 cells/2 donors, pink), except for a moderate reduction of exocytosis at 1 mM glucose. No differences in the density of docked granules were observed in presence of any of the effectors, or comparing T2D with ND cells.
## Rapid paracrine inhibition by insulin and somatostatin.
Glucagon secretion oscillates with a frequency of minutes, which is inconsistent with the relatively slow glucose dependent regulation. We therefore determined the time course of paracrine inhibition of spontaneous exocytosis by rapidly applying somatostatin, b, blue shading) or insulinto dispersed α-cells. The effect frequently wore off towards the end of the 1 min challenge, and was rapidly reversed by removing the hormone. Similarly, rapid effects on exocytosis were observed with insulin (21 cells, 4 donors,or adrenaline. Thus, paracrine regulation of glucagon exocytosis occurs on a timescale that is consistent with the observed glucagon pulsatility. Due to the small effect size, the time course of somatostatin or insulin inhibition of exocytosis could not be determined for T2D α-cells. Finally, we tested the effect of somatostatin and insulin on α-cell electrical activity by perforated patch-clamp electrophysiology. In 10 mM glucose, electrical activity consisted action potential trains as described previously. In ND α-cells, pulses of somatostatin (400 nM) caused rapid cessation of electrical activity. In the example intop, action potentials reappeared 20 s after removal of the somatostatin. However, in many cells electrical activity reappeared already during the somatostatin pulse, which is apparent the averaged membrane potential. Similar recovery during the somatostatin pulse was also seen for exocytosis, and may reflect somatostatin receptor inactivation. Insulin also dampened electrical activity, but these effects were both slower and weaker than those of somatostatin. In α-cells from T2D donors, the somatostatin or insulin pulses had little effect on the time course of exocytosisor electrical activity. In summary, insulin and somatostatin inhibit both exocytosisand electrical activity. Both effects are lost in T2D, which is consistent with the notion that these cells are resistant to paracrine inhibition.
# Discussion
Glucose controls glucagon secretion by intrinsic and paracrine mechanisms, but their relative significance is still debated, and secretory defects in type-2 diabetes are not well understood. The current work is first in using high-resolution microscopy to study glucagon secretion both in intact islets and in single dispersed α-cells of healthy and type-2 diabetic donors, thus isolating intrinsic from paracrine mechanisms while having full control over paracrine signaling. We show that in the absence of paracrine influence, isolated α-cells respond appropriately to hypoglycemia with an increase in glucagon granule exocytosis. This is consistent with the glucose dependence of glucagon secretion from intact islets (but not FACS sorted α-cells), and indicates that glucagon secretion in the lower glucose-concentration range is mostly under intrinsic control. With only 2-fold difference in the exocytosis rate between minimal secretion at 7 mM and maximal secretion at 1 mM glucose, the dynamic range is small compared with β-cells. Surprisingly, exocytosis of dispersed α-cells is stimulated in the hyperglycemic range, leading to an unphysiological V-shaped response with maximal exocytosis above 10 mM glucose. This is in contrast to intact islets, in which glucagon secretion is depressed between 3 and 20 mM glucose. We confirm this here by exocytosis measurements in intact islets, and provide evidence that this depression depends on the inhibitory effects on insulin and somatostatin that are released by neighboring βand δ-cells. Evidently, α-cell intrinsic mechanisms are sufficient for the regulation of glucagon secretion hypo-and normoglycemic range (0-7 mM glucose), while paracrine inhibition is responsible for the physiological response in the hyperglycemic range. Consequently, appropriate glucagon secretion in the hyperglycemic range is lost when α-cells are removed from their context within the islet. Elevated glucagon is a hallmark of type-2 diabetes. Despite this, both glucose-dependent and depolarization (K + )-induced exocytosis was reduced in α-cells from donors that had been diagnosed with T2D. Both exocytosis and docked granules were moderately anti-correlated with donor HbA1c values. This indicates that the exocytosis machinery in α-cells from type-2 diabetics is slightly impaired, while electrophysiological parameters (that determine electrical activity, depolarization and Ca 2+ -influx) were normal. The reason for this is unknown, but may reflect reduced expression of certain exocytosis-related proteins, as is the case in β-cells. The reduced exocytotic capacity in T2D α-cells is unrelated to changes in electrical activity, because it could be observed in K + -stimulation experiments in which the membrane potential is clamped. Since exocytosis in single α-cells is impaired rather than increased in T2D, the hyperglucagonemia in diabetic humans must be due to mechanisms that are lost in isolated cells, such as paracrine or neuronal regulation. In addition, gut derived glucagon may contribute to hyperglucagonemia following oral glucose intake.
Strikingly, the inhibitory effects of insulin and somatostatin on glucagon exocytosis were strongly reduced in cells from T2D donors, in parallel with internalization and reduced surface expression of SSTR2, the major somatostatin receptor in human α-cells. This points to α-cell resistance to insulin and somatostatin as the main cause for inadequate glucagon secretion in type-2 diabetes, which in turn exacerbates hyperglycemia. Insulin resistance is a hallmark of T2DM, and has previously been proposed as mechanism for hyperglucagonemia. For example, insulin resistance is associated with fasting glucagon levels, and this inverse relationship is lost in type-2 diabetes 28 . Interestingly, SSTR2 surface expression was also reduced in β-cells within T2D islets, suggesting that reduced somatostatin sensitivity may contribute also to increased insulin secretion, as observed early during the development of T2D. While there is reason to believe that this is a consequence of altered δ-cells activity 60 , its role may be to adapt islets to periods of greater food availability.
Exposure to insulin, somatostatin, and GABA reduced α-cell exocytosis, while adrenalin and glutamate stimulated it. This is consistent with the known effects of these signaling molecules on islets, as well as systemically. We show here that these effects are a b d c very rapid (seconds), which is consistent with the frequency of pulsatile glucagon release in vivo and from intact islets. In the absence of glucose-dependent control, paracrine inhibition by insulin and somatostatin is therefore the most likely mechanism for glucagon regulation in the hyperglycemic range. It can be speculated that the differential glucose dependence of insulin and somatostatin secretion is reflected in different target glucose ranges for their action on α-cells. All tested paracrine modulators affected exocytosis machinery at the priming step, rather than by increasing granule docking. This is consistent with previous findings that somatostatin inhibits exocytosis in rat α-cells through G i -dependent depriming, and reports that antagonists of SSTR2 12 or the associated G-protein cascade 12 increase glucagon secretion without altering the glucose-dependent inhibition of glucagon secretion. We did not observe any bursts of exocytosis, as might be expected given the pulsatile glucagon secretion from intact islets. This is in line with the absence of membrane potential oscillations in single cells(that we confirm here), and indicates that the islet context is required not just for intra-islet synchronization, but for oscillatory α-cell behavior as such.
Capacitance measurements indicate that glucagon granules exist in at least two states with different release probabilities, which are often referred to as the readily releasable pool of granules (RRP) and a larger reserve pool (RP). We show here that glucagon granules were present at the plasma membrane for extended periods before undergoing exocytosis. We interpret this as the relatively slow conversion from RP to RRP that reflects the molecular assembly of the secretory machinery at the release site, in analogy with the situation in β-cells. Throughout the glucose range, the rate of exocytosis was nearly identical to that of granule docking, suggesting that docking is rate limiting for secretion. This may indeed be the case during strong (non-physiological) stimulation, as illustrated by the finding that the glucosedependence of depolarization-induced exocytosis followed that of granule docking. However, in physiological conditions elevated K + accelerated exocytosis~50-fold (during the first second), which indicates a large excess in exocytotic capacity that is not triggered by normal α-cell electrical activity. A possible explanation could be that only a limited number of granules is positionally primed, i.e. located near voltage-gated Ca 2+ -granuleswork is required to understand the combination of factors affecting granule exocytosis, and the granule conversion rates provided here may be useful in this regard.
# Methods
Tissue. Pancreatic islets and pancreas sections were obtained from human cadaveric donors by the Nordic Network for Clinical Islet Transplantation Uppsala 64 (ethical approval by Uppsala Regional Ethics Board) or the ADI Isletcore at the University of Alberta 65 (ethical approval by Alberta Human Research Ethics Board, Pro00001754), with written donor and family consent for use in research. Work with human tissue complied with all relevant ethical regulations for use of human tissue in research and the study was approved by the Uppsala Regional Ethics Board (2006/348). Isolated islets were cultured free-floating in sterile dishes in CMRL 1066 culture medium containing 5.5 mM glucose, 10% fetal calf serum (FCS), 2 mM L-glutamine, streptomycin (100 U/ml), and penicillin (100 U/ml) at 37°C in an atmosphere of 5% CO 2 up to 2 weeks. Islets were dispersed into single cells by gentle pipetting in cell dissociation buffer (Thermo Fisher Scientific) supplemented with trypsin (0.005%, Life Technologies). Cells were then washed and plated in serum-containing medium onto 22-mm polylysine-coated coverslips, allowed to settle overnight, and then transduced using adenovirus. In, c, the dispersion step was omitted and intact islets were transduced with Pppg-NPY-EGFP adenovirus and allowed to settle onto 22-mm polylysine-coated coverslips.
Labeling of human pancreatic α-cells and glucagon granules. To identify αcells, we transduced cells with adenovirus coding for enhanced green fluorescent protein (EGFP) under the control of the pre-proglucagon promoter. The system takes advantage of Tet-On conditional expression in the presence of 4 μM doxycycline, to drive expression of EGFP. The cells were simultaneously transduced with adNPY-mCherry, a well-established secretory granule marker. Alternatively, adenovirus coding for EGFP-tagged neuropeptide Y under control of the preproglucagon promoter (Pppg-NPY-EGFP) was used, thus combining cell type identification and secretory granule label. For both approaches, immunostaining with an anti-glucagon antibody confirmed that over 90% of the fluorescently labeled cells were α-cellsand. Approximately onethird of the glucagon positive cells were labeled with Pppg-NPY-EGFP (Supplementary. NPY-EGFP labeled granules had excellent overlap with punctate glucagon staining, with 94 ± 1% of glucagon positive granules being labeled with NPY-EGFP (37 cells, 5 donors). We verified that exocytosis in the identified cells was stimulated by adrenaline, which increases intracellular Ca 2+ in αbut not β-cells.
TIRF microscopy. Cells were imaged using a custom-built lens-type total internal reflection (TIRF) microscope based on an AxioObserver Z1 with a ×100/1.45 objective (Carl Zeiss). Excitation was from two DPSS lasers at 491 and 561 nm (Cobolt) passed through a cleanup filter (zet405/488/561/640x, Chroma) and controlled with an acousto-optical tunable filter (AA-Opto). Excitation and emission light were separated using a beamsplitter (ZT405/488/561/640rpc, Chroma). The emission light was chromatically separated onto separate areas of an EMCCD camera (Roper QuantEM 512SC) using an image splitter (Optical Insights) with a cutoff at 565 nm (565dcxr, Chroma) and emission filters (ET525/ 50m and 600/50m, Chroma). Scaling was 160 nm per pixel. Confocal microscopy was done with a Zeiss LSM780 using a 63/1.40 objective (Zeiss) with sequential scanning of the red (excitation 561 nm, emission 578-696 nm) and green channel (excitation 488 nm, emission 493-574 nm). Pinhole size was 0.61 mm, corresponding to 1 Airy unit.
Cells were imaged in a standard solution containing (in mM) 138 NaCl, 5.6 KCl, 1.2 MgCl 2 , 2.6 CaCl 2 , 10 D-glucose, 5 HEPES (pH 7.4 with NaOH). In Figs. 3-5 the D-glucose concentration was varied as indicated and cells equilibrated for at least 20 min before recording commenced. Where stated, exocytosis was evoked by elevating K + to rapidly depolarize the cells (75 mM KCl equimolarly replacing NaCl in the standard solution). In these experiments, K + was elevated from 10 s after the onset of the recording until the end of the experiment at 50 s. Timed applications of elevated K +, f, 4, 5d black, and 6), adrenalin (Supplementary, insulin or somatostatin (in, and glucose changeswere by computer-controlled local pressure ejection from a pulled glass pipette (similar to those used for patch clamp). Spontaneous glucose-dependent exocytosis-c, and 6) was recorded for 3 min per after equilibration in the stated conditions. In, insulin or somatostatin were applied during the indicated times.
Image analysis. Exocytosis events were identified based on the characteristic rapid loss of the granule marker fluorescence (1-2 frames). Granule docking events were rare and defined as granules that approached the TIRF field and becoming laterally confined once they reached their maximum brightness.
Docked granules were counted using the 'find maxima' function in ImageJ (http://rsbweb.nih.gov/ij). Values were normalized to each cells' contact area with the coverslip (footprint). The ΔF parameter estimates the fluorescence that is specifically localized to a granule, but subtracting a local background value (average of a 5 pixel wide annulus) from the average fluorescence value in a 3 pixel wide circle, both centered at the granule position.
Immunostaining of pancreatic sections. For analysis of SSTR2 expression, deparaffinized human pancreatic tissue sections (biobank samples obtained from the EXODIAB consortium, Uppsala) were heated in a buffer containing 10 mM Tri-sodium citrate and 0.05% Tween 20 (pH 6) for 15 min, allowed to cool, and rinsed with Dako wash buffer 1x. After a 30-min blocking step (Background Sniper, Biocare Medical), sections were rinsed with wash buffer 1x (Dako) and incubated with anti-SSTR2 (Abcam ab134152, diluted 1:500 in wash buffer), and anti-glucagon antibodies (Dako A0565, diluted 1:1500 in wash buffer) overnight at 4°C. The slides were then washed in wash buffer and incubated with fluorophorelabeled secondary antibodies (diluted in Dako wash buffer 1x) for 30 min at room temperature. Fluorescence was visualized using a Zeiss LSM 780 confocal microscope. For analysis, 3-pixel wide linescans of fluorescent intensity were calculated as illustrated in, f, top. Background subtracted and estimated as the minimum value to the left of the alignment point (corresponding to the nucleus location), after 3 × 3 median filtering.
Electrophysiology. Standard whole-cell voltage clamp and capacitance recordings were performed using an EPC-9 patch amplifier (HEKA Electronics, Lambrecht/ Pfalz, Germany) and PatchMaster software. Voltage-dependent currents were investigated using an IV-protocol, in which the membrane was depolarized from −70 mV to +80 mV (10 mV steps) during 50 ms each. Currents were compensated for capacitive transients and linear leak using a P/4 protocol. Exocytosis was detected as changes in cell capacitance using the lock-in module of Patchmaster (30 mV peak-to-peak with a frequency of 1 kHz).
Patch electrodes were made from borosilicate glass capillaries coated with Sylgard close to the tips and fire-polished. The pipette resistance ranged between 2 and 4 MΩ when filled with the intracellular solution containing (in mM) 125 Csglutamate, 10 CsCl, 10 NaCl, 1 MgCl 2 , 0.05 EGTA, 3 Mg-ATP, 0.1 cAMP, and 5 HEPES, pH 7.2 adjusted using CsOH.
During the experiments, the cells were continuously superfused with an extracellular solution containing (in mM) 138 NaCl, 5.6 KCl, 1.2 MgCl 2 , 2.6 CaCl 2 , 10 D-glucose, and 5 Hepes, pH 7.4 adjusted with NaOH at a rate of 0.4 ml/min. All electrophysiological measurements were performed at 32C. In the analysis of the measured voltage-dependent current consists of both Na + and Ca 2+ current components, were the rapid peak current (0-3 ms) represent the Na + current and the sustained current during the latter part of the depolarization reflects the Ca 2+ current. In, for membrane potential, solution was containing (in mM) 76 K 2 SO 4 , 10 KCl, 1 MgCl 2 , and 5 HEPES, pH 7.3 adjusted with KOH.
Statistics. Data are presented as mean ± SEM unless otherwise stated. Statistical significance was tested using t-test for comparing ND and T2D groups, or ANOVA for multiple comparisons, as stated (in Origin 2018). Correlation was quantified as Pearson coefficient r using Excel.
Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.
## Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. A source data file containing all numeric analysis for Figs. 1c-g, 2b-e, 3b-f, 4b-f, 5a-d, and 6a-h, and, b, and 4a is provided with the paper. |
Avoiding contact allergens: from basic research to the in vitro identification of contact allergens
## Forschergruppe allergologie, hautklinik, universitätsklinikum freiburg
## Avoiding contact allergens: from basic research to the in vitro identification of contact allergens
Allergic contact dermatitis (ACD) is a chemical-induced inflammatory skin disease. Contact allergens are low-molecular-weight chemicals that must react with proteins in order to become immunogenic. This interaction leads to the activation of innate immune and stress responses and to the formation of antigenic epitopes for T cells which are the effector cells of ACD. Due to the multitude of chemicals that surround us in our daily life and their potential sensitizing capacity, it is crucial to identify contact sensitizers before these chemicals are used in consumer products. Appropriate in vitro assays for hazard identification are urgently needed to replace animal-based assays. The EU-wide ban on sensitization testing of cosmetic ingredients in animals is in effect since March 2009 and the necessity to test more than 30,000 already marketed chemicals for their sensitizing potential under the EU regulation REACh has intensified the worldwide efforts to replace animal testing. We summarize here the current strategies to develop a battery of assays which allows the identification of contact allergens by in vitro alternatives to animal testing. Our main focus lies on the test systems recently developed within the EU project Sens-it-iv in which we participate.
## Characteristics of contact allergens and strategies to avoid them
Allergic contact dermatitis (ACD) is among the most frequent skin diseases with increasing prevalence. It is very often workrelated and clinically problematic due to the risk of chronification and the lack of causative treatments. Due to often long treatment times using immunosuppressive drugs such as corticosteroids which may have significant side effects the socioeconomic impact and the impact on the quality of life of the patients is enormous. Therefore, avoiding contact allergens is the optimal prevention and mandatory for patients. However, this is often not possible without a change of the profession when ACD is caused by contact allergens in the workplace. A reliable system for the identification of contact allergens by the chemical and pharmaceutical industryso far done using animal-based assays -is essential to assure consumer safety [bib_ref] CTLA-8, cloned from an activated T cell, bearing AU-rich messenger RNA instability..., Rouvier [/bib_ref] [bib_ref] IL-TIF/IL-22: genomic organization and mapping of the human and mouse genes, Dumoutier [/bib_ref] [bib_ref] Th17: an effector CD4 T cell lineage with regulatory T cell ties, Weaver [/bib_ref] [bib_ref] TH17 cells in the big picture of immunology, Schmidt-Weber [/bib_ref]. Mice and guinea pigs are used for sensitization testing. The only validated and accepted assays so far are the Guinea Pig Maximisation Test (GPMT) and the Buehler Test (OECD Test Guideline 406) as well as the mouse Local Lymph Node assay (LLNA). The LLNA and its modifications represent the current gold standard (OECD Test Guidelines 429, 442A, 442B) [fig_ref] Figure 1: Figure 1 [/fig_ref]. In the original protocol test substances are applied daily for three consecutive days on mouse ear skin. Then, the proliferation of draining lymph node cells is assessed by in vivo incorporation of radioactive [ 3 H]-thymidine. The so-called EC3 values quantify the stimulation over the control level and can be used to additionally assess the relative potency of the identified contact allergens. The latter is an important parameter due to the possibility to use substances with weak sensitizing potency at low concentrations in consumer products. Many fragrances used in perfumes, soaps and other products fall within this category. However, the EU legislation has banned sensitization testing in animals for the cosmetics industry since March 2009 with the exception of repeated dose testing until March 2013. In addition, the EU regulation REACh requires sensitization testing of more than 30.000, more likely 50 -60.000 already marketed chemicals produced in amounts over 1 ton/year [bib_ref] Tumor-infiltrating regulatory dendritic cells inhibit CD8+ T cell function via L-arginine metabolism, Norian [/bib_ref]. Therefore, and for financial and ethical reasons, in vitro alternatives to animal testing are being developed internationally. The EU has funded such a program, the project Sens-it-iv (Novel Testing Strategies for in vitro assessment of allergens, see also www.sens-it-iv.eu) from 2005 to 2011. The development of in vitro alternatives is a yearlong process including basic research. Establishing an in vitro assay with a standardized operation procedure (SOP) requires technology transfer and ring trials by independent laboratories with blinded chemicals for assay validation as well as regulatory acceptance for example by the European Centre for the Validation of Alternative Methods (ECVAM) and the responsible authorities in the EU and other countries. Examples for criteria that must be followed are given by the OECD Guidelines for the Testing of Chemicals (e.g. Guideline 34).
More than 4.000 low molecular weight (< 500 Dalton) organic and inorganic chemicals are identified as contact allergens. They react with proteins either by complex formation, for example with histidines in the case of metal ions such as nickel and cobalt, or they covalently bind to amino acid side chains of e.g. cysteine and lysine as shown for many organic chemicals. In contrast to ir- Steps leading to sensitization to contact allergens and in vitro assay development. In vitro assays for contact allergen identification must reproduce key steps of the sensitization phase of ACD. Skin penetration, protein modification and activation of innate immune and stress responses in the skin are followed by DC emigration and homing to skin draining lymph nodes. Contact allergens are presented on DC in the context of MHC molecules to T cells. These are primed and differentiate to effector T cells which exert their function after recruitment to the skin in the elicitation phase. The validated animal tests (GPMT, Buehler Test and LLNA) must be replaced by in vitro alternatives. ritant chemicals which have toxic effects on the skin and induce irritant contact dermatitis (ICD) this chemical protein reactivity is a hallmark of contact allergens and a prerequisite for their immunogenicity and antigenicity. Therefore, they are called haptens or halfantigens. Protein reactivity is acquired by some chemicals, the so-called pre-haptens by (auto-) oxidation or for others, the so-called pro-haptens only after metabolic conversion [fig_ref] Figure 2: Currently used terminology for categories of chemicals that cause contact dermatitis [/fig_ref]. These pro-and pre-haptens are very problematic with respect to clinical diagnosis and in vivo assays for contact allergen identification. Usually, the parent compound that is used in clinical patch testing or in the in vivo assays lacks sensitizing potential and the reactive adducts are most of the time unknown.
Parameters that are not fully addressed by the current in vitro assay strategies are the ability and efficiency of the chemical to penetrate into the skin and the frequency of application or skin contact that in some cases allows accumulation over a critical threshold for sensitization that is not reached by a single contact. Also the metabolic conversion and oxidation of pro-and pre-haptens is not yet well enough understood to be implicated in assays, but with the current progress in basic research such efforts are underway [bib_ref] Production of interleukin 22 but not interleukin 17 by a subset of..., Duhen [/bib_ref] [bib_ref] Identification of a human helper T cell population that has abundant production..., Trifari [/bib_ref]. These problems must be solved in the future.
## The sensitization phase of allergic contact dermatitis
The ultimate goal of all efforts to replace animal testing for contact allergen identification by in vitro assays is the implementation of key steps in the immune response to contact allergens in assay development [fig_ref] Figure 1: Figure 1 [/fig_ref]. The unavoidable reductionism that is required for all in vitro assays must be able to reproduce key processes that happen in a given cell type in its tissue microenvironment in vivo during sensitization or elicitation of contact dermatitis. This is a great challenge and requires a detailed mechanistic understanding of the immunologic pathomechanisms of the disease. Especially in the field of innate immune responses to contact allergens significant progress has been made during recent years and has improved our understanding of the sensitization process [bib_ref] Th22 cells represent a distinct human T cell subset involved in epidermal..., Eyerich [/bib_ref] [bib_ref] Th17 cytokines interleukin (IL)-17 and IL-22 modulate distinct inflammatory and keratinocyte-response pathways, Nograles [/bib_ref] [bib_ref] IL-17 is produced by nickel-specific T lymphocytes and regulates ICAM-1 expression and..., Albanesi [/bib_ref]. One of the striking features of contact allergens besides their ability to form antigenic determinants for T cells is their adjuvant effect. This allows them to activate innate immune and stress responses which lead to skin inflammation. Most likely, some of the chemical protein modifications introduced by contact allergens act like conventional post-translational modifications by altering protein function, localization and by inducing signalling and, eventually, inflammation [bib_ref] Th17 cytokines interleukin (IL)-17 and IL-22 modulate distinct inflammatory and keratinocyte-response pathways, Nograles [/bib_ref] [bib_ref] IL-17 is produced by nickel-specific T lymphocytes and regulates ICAM-1 expression and..., Albanesi [/bib_ref] [bib_ref] IL-17 in atopic eczema: linking allergen-specific adaptive and microbialtriggered innate immune response, Eyerich [/bib_ref]. The inflammatory response is essential for the sensitization. Most likely it is also required for the elicitation of ACD for example to provide the cytokines and chemokines that are needed to recruit the contact allergen-specific effector T cells from the bloodstream to the inflamed skin. Studies in the mouse contact hypersensitivity (CHS) model have now shown that contact allergens trigger innate immune and stress responses that are also used in immune responses to infection [bib_ref] Th17 cytokines interleukin (IL)-17 and IL-22 modulate distinct inflammatory and keratinocyte-response pathways, Nograles [/bib_ref] [bib_ref] IL-17 is produced by nickel-specific T lymphocytes and regulates ICAM-1 expression and..., Albanesi [/bib_ref]. Thus, TLR2 and TLR4 play a role for CHS to TNCB, oxazolone and FITC [bib_ref] A novel subset of CD4(+) T(H)2 memory/effector cells that produce inflammatory IL-17..., Wang [/bib_ref]. These innate immune receptors recognize the bacterial cell wall components such as lipopeptides and lipopolysaccharide (LPS) from gram-positive and gram-negative bacteria, respectively. We have demonstrated that TNCB induces the degradation of the extracellular matrix component hyaluronic acid (HA). Fragments of HA then trigger TLR2 and TLR4 to promote skin inflammation (Esser, PR. In preparation, Abstract P095, Allergologie 20:177, 2011). In contrast, nickel ions interact directly with conserved histidines in the human TLR4 and induce signalling [bib_ref] IL-23 and IL-17 in the establishment of protective pulmonary CD4+ T cell..., Khader [/bib_ref]. Interestingly, these histidines are missing in the mouse TLR4 which explains why mice could not be used as an animal model for ACD to nickel. In order to induce CHS in mice, nickel must be co-applied with adjuvants such as complete Freund's adjuvant or LPS [bib_ref] Requirement of interleukin-17A for systemic anti-Candida albicans host defense in mice, Huang [/bib_ref] [bib_ref] Uric acid is a danger signal activating NALP3 inflammasome in lung injury..., Gasse [/bib_ref]. These results highlight the crucial role of the adjuvant effect of contact allergens and the essential role of innate immunity in ACD.
Contact allergens also induce the production of reactive oxygen species (ROS) which participate in the HA degradation and activate Keap1/Nrf2 mediated anti-oxidant responses. They can also deplete glutathione thereby causing a disturbed redox balance of the cells [bib_ref] The differentiation of human T(H)-17 cells requires transforming growth factor-beta and induction..., Manel [/bib_ref] [bib_ref] Staphylococcal exotoxins are strong inducers of IL-22: A potential role in atopic..., Niebuhr [/bib_ref] [bib_ref] The aryl hydrocarbon receptor links TH17-cell-mediated autoimmunity to environmental toxins, Veldhoen [/bib_ref] [bib_ref] The aryl hydrocarbon receptor in immunity, Esser [/bib_ref] [bib_ref] IL-17 and IL-22: siblings, not twins, Eyerich [/bib_ref] [bib_ref] IL-22 increases the innate immunity of tissues, Wolk [/bib_ref].
Another action of contact allergens is the triggering of ATP release from stressed or damaged cells in the skin [bib_ref] Interleukin (IL)-22 and IL-17 are coexpressed by Th17 cells and cooperatively enhance..., Liang [/bib_ref]. Extracellular ATP serves as a danger signal and triggers an efflux of K + ions from the cells via stimulation of the purinergic ATP receptor P2X 7 . This, in turn leads to the activation of the NLRP3 inflammasome, a cytosolic protein complex that activates caspase-1. This enzyme then cleaves the immature pro-forms of IL-1b and IL-18 which are produced in response to TLR stimulation to the mature and secreted forms. These cytokines are important mediators of inflammation, and IL-1R signalling is essential for sensitization to contact allergens. The cooperative action of these innate immune and stress pathways is crucial in DC [bib_ref] Th17 cytokines interleukin (IL)-17 and IL-22 modulate distinct inflammatory and keratinocyte-response pathways, Nograles [/bib_ref] [bib_ref] IL-17 is produced by nickel-specific T lymphocytes and regulates ICAM-1 expression and..., Albanesi [/bib_ref]. Failure of a single one (TLR2/4 activation, ROS production, P2X 7 triggering, NLRP3 inflammasome activation) is sufficient to abrogate the sensitizing potential of DC [bib_ref] A novel subset of CD4(+) T(H)2 memory/effector cells that produce inflammatory IL-17..., Wang [/bib_ref] [bib_ref] IL-23 and IL-17 in the establishment of protective pulmonary CD4+ T cell..., Khader [/bib_ref] [bib_ref] Requirement of interleukin-17A for systemic anti-Candida albicans host defense in mice, Huang [/bib_ref] [bib_ref] Uric acid is a danger signal activating NALP3 inflammasome in lung injury..., Gasse [/bib_ref] [bib_ref] The differentiation of human T(H)-17 cells requires transforming growth factor-beta and induction..., Manel [/bib_ref] [bib_ref] Staphylococcal exotoxins are strong inducers of IL-22: A potential role in atopic..., Niebuhr [/bib_ref] [bib_ref] The aryl hydrocarbon receptor links TH17-cell-mediated autoimmunity to environmental toxins, Veldhoen [/bib_ref] [bib_ref] The aryl hydrocarbon receptor in immunity, Esser [/bib_ref] [bib_ref] IL-17 and IL-22: siblings, not twins, Eyerich [/bib_ref] [bib_ref] IL-22 increases the innate immunity of tissues, Wolk [/bib_ref] [bib_ref] Interleukin (IL)-22 and IL-17 are coexpressed by Th17 cells and cooperatively enhance..., Liang [/bib_ref]. These findings form a scientific and conceptual framework for the development of new in vitro assays or the improvement of existing assays. Further progress in assay development will be made by implementing the results from genomic and proteomic studies. These global techniques aim at identifying profiles of genes or proteins that are regulated in their expression by contact allergens but not by irritants. In addition, such contact allergen-specific signatures help to confirm known and to identify new signalling pathways involved in the sensitization process. This will also be useful for the identification of new drug targets for causative treatment of ACD.
## In vitro assay development for contact allergen identification
In vitro assays for contact allergen identification must cover the key steps of the sensitization process [fig_ref] Figure 1: Figure 1 [/fig_ref]. Several of the in vitro assays that have been and are being developed address the interaction of chemicals with proteins or cells such as monocytes, DC [bib_ref] Impaired T(H)17 cell differentiation in subjects with autosomal dominant hyper-IgE syndrome, Milner [/bib_ref] and keratinocytes [bib_ref] Patients with chronic mucocutaneous candidiasis exhibit reduced production of Th17-associated cytokines IL-17..., Eyerich [/bib_ref] [bib_ref] IL-17 and Th17 cells in human inflammatory diseases, Miossec [/bib_ref]. One of the first steps of the sensitization phase is the protein binding of contact allergens based on their essential chemical reactivity. Understanding the chemistry of contact allergens is one of the goals that may allow in silico prediction based on structural alerts [bib_ref] C-C chemokine receptor 6-regulated entry of TH-17 cells into the CNS through..., Reboldi [/bib_ref] [bib_ref] Divergent pro-and antiinflammatory roles for IL-23 and IL-12 in joint autoimmune inflammation, Murphy [/bib_ref]. Currently, covalent binding of chemicals to amino acid side chains is addressed by the Direct Peptide Reactivity Assay (DPRA) [bib_ref] IL-17A and IL-17F do not contribute vitally to autoimmune neuro-inflammation in mice, Haak [/bib_ref] [bib_ref] Hydrodynamic-based delivery of an interleukin-22-Ig fusion gene ameliorates experimental autoimmune myocarditis in..., Chang [/bib_ref]. The depletion of chemically modified model peptides with defined target amino acids for contact allergen modification is measured. This assay may also allow potency assessment since the efficiency of depletion that reflects the chemical reactivity correlates well with the EC3 values from the LLNA [bib_ref] Interleukin 22 (IL-22) plays a protective role in T cell-mediated murine hepatitis:..., Radaeva [/bib_ref]. The next step is the activation of skin cells. The human Cell Line Activation Test (hCLAT) uses the human monocytic leukemia cell line THP-1 which is exposed for 24 h to the test chemicals. Readout parameters are the induction of the co-stimulatory molecules CD54 and CD86 on the surface of these cells [bib_ref] Hydrodynamic gene delivery of interleukin-22 protects the mouse liver from concanavalin A-,..., Pan [/bib_ref] [bib_ref] IL-22 mediates mucosal host defense against Gram-negative bacterial pneumonia, Aujla [/bib_ref] [bib_ref] The serine protease marapsin is expressed in stratified squamous epithelia and is..., Li [/bib_ref] [bib_ref] IL-17 amplifies human contact hypersensitivity by licensing hapten nonspecific Th1 cells to..., Pennino [/bib_ref]. Similarly, the Myeloid U937 Skin Sensitisation Test (MUSST) detects the induction of IL-8 and CD86 [bib_ref] Th22 cells represent a distinct human T cell subset involved in epidermal..., Eyerich [/bib_ref] [bib_ref] IL-22 and IL-20 are key mediators of the epidermal alterations in psoriasis..., Wolk [/bib_ref]. Similar readouts as well as p38 MAP kinase activation have been used for THP-1 cells [bib_ref] IL-23 signaling enhances Th2 polarization and regulates allergic airway inflammation, Peng [/bib_ref].
Epidermal keratinocytes and dermal fibroblasts also respond to contact allergens and irritants and significantly contribute to ACD and ICD. They produce inflammatory mediators such as cytokines and chemokines that attract immune cells into the skin and regulate the emigration of DC to the skin draining lymph nodes. Thus, human keratinocytes are triggered via TLR to produce pro-IL-18 and process it via the NLRP3 inflammasome to the bioactive mature cytokine. The NCTC assay uses the human keratinocyte cell line NCTC 2544 to assess intracellular production of IL-18 as triggered in response to contact allergens [bib_ref] Patients with chronic mucocutaneous candidiasis exhibit reduced production of Th17-associated cytokines IL-17..., Eyerich [/bib_ref] [bib_ref] IL-17 and Th17 cells in human inflammatory diseases, Miossec [/bib_ref].
DC migration from the epidermis to the dermis and then to the draining lymph nodes is the next step in sensitization that can be addressed by in vitro assays. It has recently been shown that both irritants and contact allergens induce the migration of Langerhans cells (LC) from the epidermis to the dermis. This step precedes the CCR7 and CCL19/ CCL21 dependent migration via the efferent lymphatics to the draining lymph node where the priming of contact allergen-specific T cells occurs. It is dependent on a set of chemokines produced by dermal fibroblasts in response to both classes of chemicals [bib_ref] IL-17 is increased in asthmatic airways and induces human bronchial fibroblasts to..., Molet [/bib_ref] [bib_ref] TH17 cells mediate steroidresistant airway inflammation and airway hyperresponsiveness in mice, Mckinley [/bib_ref] [bib_ref] Interleukin-22 is a negative regulator of the allergic response, Schnyder [/bib_ref]. The DC migration assay is based on the finding that irritants and contact allergens induce different chemokine receptors on Langerhans cells of the epidermis [bib_ref] Surface phenotype and antigenic specificity of human interleukin 17-producing T helper memory..., Acosta-Rodriguez [/bib_ref]. Thus, irritants induce CCR2 and CCR5 and, hence, LC migration to CCL2 and CCL5, while contact allergens induce CXCR4 and migration to CXCL12. The DC migration assay uses the human acute myeloid leukemia cell line MUTZ-3 which is differentiated to a LC-like phenotype [bib_ref] Surface phenotype and antigenic specificity of human interleukin 17-producing T helper memory..., Acosta-Rodriguez [/bib_ref]. The MUTZ-LC are then treated with the test chemical for 24 h. Subsequently they are used in a transwell migration assay using CCL5 or CXCL12 as chemoattractants. The migration index calculated as the CXCL12:CCL5 ratio serves as readout. The DC migration assay is up to now the only functional single cell test for contact allergen identification.
A two-tiered strategy was proposed to assess the potency of contact allergens. The NCTC 2544 assay is used for contact allergen identification in the first tier [bib_ref] Patients with chronic mucocutaneous candidiasis exhibit reduced production of Th17-associated cytokines IL-17..., Eyerich [/bib_ref] [bib_ref] IL-17 and Th17 cells in human inflammatory diseases, Miossec [/bib_ref] and the EE potency assay for potency assessment in the second tier [bib_ref] Human interleukin 17-producing cells originate from a CD161+CD4+ T cell precursor, Cosmi [/bib_ref]. The latter assay addresses chemical toxicity and IL-1a production in epidermal equivalents as readout. The ef-fective chemical concentration required to reduce cell viability by 50% (EE-EC50) and the chemical concentration which increases IL-1α secretion by 10-fold (EE-IL-1α[10×]) are used to determine sensitizer potency.
A promising strategy to identify at least cysteine-reactive contact allergens uses a keratinocyte-based reporter cell line that expresses luciferase under the control of tandem anti-oxidant response elements (ARE). Contact allergens can react with cysteines in the cytosolic protein Keap1, a sensor for oxidative and electrophilic stress. This protein is associated with the transcription factor Nrf2. Nrf2 is usually fed to the proteasome for degradation but Keap1 oxidation or chemical modification at critical cysteines results in translocation of Nrf2 to the nucleus where it binds to ARE in genes of the antioxidant phase 2 response such as glutathione reductase, heme oxygenase and NAD(P) H:quinone oxidoreductase 1 [bib_ref] Circulating and gut-resident human Th17 cells express CD161 and promote intestinal inflammation, Kleinschek [/bib_ref] [bib_ref] Cutting edge: NKT cells constitutively express IL-23 receptor and RORgammat and rapidly..., Rachitskaya [/bib_ref] [bib_ref] Human fetal lymphoid tissue-inducer cells are interleukin 17-producing precursors to RORC+ CD127+..., Cupedo [/bib_ref]. The HaCaT human keratinocyte reporter cell line is now used to identify cysteine-reactive contact allergens based on their ability to trigger ARE-dependent luciferase activity.
Genomic studies have been performed with MUTZ-3 progenitor cells and monocytes that were treated with test chemicals. Characteristic gene profiles have been identified that can be used as a contact allergenspecific signature in the Genomic Analysis Rapid Detection (GARD) assay (www. sens-it-iv.eu and [bib_ref] Expression of IL-17 in human memory CD45RO+ T lymphocytes and its regulation..., Shin [/bib_ref]. Similar approaches are ongoing in proteomic studies to identify characteristic protein signatures in MUTZ-3 and keratinocytes. Further refinement of these promising genomic and proteomic assays will clarify whether they could be stand-alone assays that reliably identify contact allergens without integration in a tiered strategy and whether or not potency assessment will be possible using these approaches.
## T cell responses to contact allergens and assay development
The most specific response of the immune system to contact allergens is the T cell response that concludes the sensitization phase. Contact allergens form T cell epitopes by chemical modification of peptides presented in the binding groove of MHC molecules on DC and other cells. Alternatively they can modify extracellular or cellular proteins which are then processed by DC to yield hapten-modified peptides for presentation to T cells after DC migration from the skin to the draining lymph nodes. Naïve T cells can recognize these haptenated peptides or complexes of metal ions with MHC molecules, peptides and TCR and become activated [bib_ref] Interleukin 17-producing gamma delta T cells increased in patients with active pulmonary..., Peng [/bib_ref]. This T cell priming step leads to the proliferation of contact allergen-specific T cells and their differentiation to effector cells which are recruited to the inflamed skin during the contact allergen-induced inflammatory response in the elicitation phase of ACD. Based on these processes in vitro T cell priming assays (TCPA) have been developed [bib_ref] Interleukin 17-producing gamma delta T cells increased in patients with active pulmonary..., Peng [/bib_ref]. The latest improved protocols for these functional tests use monocyte-derived DC as antigen-presenting cells (APC) and naïve human T cells. The sorting of naïve T cells or the depletion of CD25+ and CD56+ T cells may increase the sensitivity of the assay due to the depletion of immunoregulatory cells such as CD4+CD25+FOXP3+ regulatory T cells and NKT cells [bib_ref] CD27 is a thymic determinant of the balance between interferon-gamma-and interleukin 17-producing..., Ribot [/bib_ref] [bib_ref] Identification of CD25+ gamma delta T cells as fetal thymus-derived naturally occurring..., Shibata [/bib_ref]. The test chemical can be directly added to the cocultures, DC can be pulsed before co-culture with T cells or haptenated proteins such as contact allergen-coupled human serum albumin can be used as antigens [bib_ref] Interleukin 17-producing gamma delta T cells increased in patients with active pulmonary..., Peng [/bib_ref] [bib_ref] Identification of CD25+ gamma delta T cells as fetal thymus-derived naturally occurring..., Shibata [/bib_ref]. The readout of the TCPA is T cell proliferation or, more efficient and specific, multiparametric flow cytometry to detect intracellular cytokines such as IFN-g on defined populations of T cells [bib_ref] Interleukin 17-producing gamma delta T cells increased in patients with active pulmonary..., Peng [/bib_ref]. The latter technique may allow potency assessment since the frequency of antigen-specific T cells is determined as well. We are currently testing whether the frequency of contact allergen-specific T cells and the diversity of their TCR repertoire correlates with the potency of the respective contact allergen. If this was the case such studies would allow assessment of contact allergen potency in the TCPA (Kimber, I. et al., submitted for publication) [bib_ref] Th22 cells represent a distinct human T cell subset involved in epidermal..., Eyerich [/bib_ref].
## Summary
The current worldwide efforts to replace animal testing in the sensitization testing of chemicals is fuelled by the progress in basic research. The improved understanding of cellular and molecular processes in the sensitization and elicitation of ACD can now be implemented in assay development. The assays in development as well as future assays can be based on this knowledge. A tiered strategy is the likely future of in vitro assay development for the identification of contact allergens. Ideally a battery of assays will be used that cover different steps of the sensitization process. A scoring system will then allow the identification of contact allergens and potency assessment. The fruitful international cooperations should also help to reduce animal testing not only for cosmetic ingredients but also in other areas of (immuno-) toxicology and research, to refine existing assays, and to eventually replace it [bib_ref] CTLA-8, cloned from an activated T cell, bearing AU-rich messenger RNA instability..., Rouvier [/bib_ref].
[fig] Figure 1: Figure 1. Steps leading to sensitization to contact allergens and in vitro assay development. In vitro assays for contact allergen identification must reproduce key steps of the sensitization phase of ACD. Skin penetration, protein modification and activation of innate immune and stress responses in the skin are followed by DC emigration and homing to skin draining lymph nodes. Contact allergens are presented on DC in the context of MHC molecules to T cells. These are primed and differentiate to effector T cells which exert their function after recruitment to the skin in the elicitation phase. The validated animal tests (GPMT, Buehler Test and LLNA) must be replaced by in vitro alternatives. [/fig]
[fig] Figure 2: Currently used terminology for categories of chemicals that cause contact dermatitis. Irritants have toxic effects on the skin and induce Irritant Contact Dermatitis (ICD) without triggering a T cell response. In contrast, contact allergens are protein-reactive haptens or pre-and pro-haptens that are converted to reactive haptens by (auto-) oxidation or metabolic activation. They activate the innate immune system due to their adjuvant effect and form antigenic T cell epitopes leading to Allergic Contact Dermatitis (ACD). [/fig]
|
Clinical Efficacy of Custom-built Software for the Early Detection of Glaucoma: A Comparison of Axial-length and Major Retinal Artery Location Data
[bib_ref] Improvement of diagnostic performance regarding retinal nerve fiber layer defect using shifting..., Rho [/bib_ref]
# Materials and methods
This study was approved by the institutional review board of CHA Bundang Medical Center (2017-05-045-001) and conducted at CHA Bundang Medical Center. This study conformed to the Declaration of Helsinki and written informed consent was obtained. One hundred and fifty-one healthy eyes of 151 Korean adults (normative group, prospectively enrolled and approved by the institutional review board) and 137 additional eyes (validation group; 79 early-glaucoma eyes of 79 Korean adults and 58 age-matched additional healthy eyes of 58 Korean adults retrospectively enrolled and approved by the institutional review board) were consecutively enrolled between 2015 and 2017. All subjects underwent a full ophthalmic examination including slit-lamp biomicroscopy, best-corrected visual acuity (BCVA), spherical equivalent measurements, Goldmann applanation tonometry by a glaucoma specialist (SR), central corneal thickness (CCT) measurements using an ultrasonic pachymeter (UP-1000; Nidek, Gamagori, Japan), axial-length (AXL) determinations using an Echoscan (US-4000, Nidek), gonioscopy, visual-field tests using static automated white-on-white threshold perimetry (Swedish Interactive Threshold Algorithm standard 24-2, Humphrey Field Analyzer II; Carl Zeiss Meditec, Dublin, CA, USA), fundus examinations after pupil dilation, fundus photography/red-free photography using a fundus camera (VX-10i; Kowa, Nagoya, Japan), and Spectralis OCT scans (Heidelberg Engineering GmbH, Heidelberg, Germany).
The inclusion criteria for all subjects were visual acuity of >20 / 30 and an age between 18 and 80 years. The exclusion criteria were clinical evidence of neuro-ophthalmic disease, retinal disease, history of retinal and refractive surgery within 3 months, or closed irido-corneal angle and refractive error of >± 8.0 diopters (D) and ±3.0 D of cylinder. In the normal group, subjects were additionally excluded if intraocular pressure (IOP) was > 21 mmHg, a glaucomatous optic disc was present, or abnormal visualfield test results existed. Subjects with mean deviation (MD) values >-6 dB (normal or early field damage) were consecutively enrolled in the validation group.
The retinal nerve fiber layer (RNFL)-defect-detecting ability of the two software programs were assessed based on the standard criterion named "early-glaucoma state." The early-glaucoma state was defined as subjects showing glaucomatous optic disc contours or RNFL defects in redfree photography plus corresponding abnormal visual-field results with early damage (MD better than -6 dB). A reliable visual field was defined by fixation loss <30% and a false-positive or false-negative error <20%. An abnormal visual field was defined by one or more of the following criteria: glaucoma hemifield test outside normal limits, pattern standard deviation decreased to p < 0.05, and three or more non-edged points in a cluster decreased to p < 0.05 with one of these decreased to p < 0.01.
The subjects' temporal-superior-nasal-inferior-temporal (TSNIT) graphs in the validation group were assessed using the built-in software and our custom-built software. If a subject's graph line contacted the reference graph of each software (lower 1% limit of normative data), it was defined as abnormal. The custom-built software displayed four different normative graphs according to the major retinal artery position [bib_ref] Improvement of diagnostic performance regarding retinal nerve fiber layer defect using shifting..., Rho [/bib_ref]. Using the overlaying technique in the software, the observer could detect abnormal regions easily [fig_ref] Figure 1: A demonstration of an overlaying function using our custom-built software [/fig_ref]. The validation group was set to compare the RNFL-defect-detecting ability of three different criteria for the lower 1% limit. Criterion 1 was from the manufacturer's data; criteria 2 and 3 were from the established normative data in 151 healthy Korean eyes driven by their axial length and vessel position, respectively. For criteria 2 and 3, the volunteers' data were separated into four subgroups, which were combinations of two groups according to the median values of axial-length and vessel-position data.
## Rnfl-thickness measurement and angle assessment
All OCT scans were performed by a highly experienced operator, and image analysis was performed according to the same protocol as described elsewhere [bib_ref] Improvement of diagnostic performance regarding retinal nerve fiber layer defect using shifting..., Rho [/bib_ref]. Briefly, the peripapillary RNFL thicknesses were measured using the instrument's built-in software ver. 5.4.7.0. To avoid the Fovea to Disc (FoDi) technology effect, the mode for internal fixation was deactivated because the FoDi technology provides a predicted horizontal fovea to disc axis even in a severely torsioned eye at a baseline scan. This means that the FoDi effect would interfere the measurement of the RNFL peak location. All measurements were automatically segmented and selected images (with quality scores >22) were reconfirmed one by one.
Each subject's RNFL-thickness data at intervals of 5° (a total of 72 values) were extracted using the Cool Ruler software ver. 1.5 (freeware from Fabsoft, http://www.fabsoft. com) from the TSNIT graphs. The cross points in a 3.5-mm peripapillary circle and the major retinal arteries of the superotemporal and inferotemporal halves were designated obtaining the angle degrees between a horizontal line (from the center of the disc) and the cross points. The angle degrees were measured with the PicPick software ver. 3.2.4 (freeware from NTeWORKS, http://www.picpick.org) and ImageJ (National Institutes of Health, Bethesda, MD, USA).
# Statistical analysis
Statistical analyses were performed using the IBM SPSS Statistics ver. 21.0 (IBM Corp., Armonk, NY, USA). Chisquare tests and independent sample t-tests were used to compare demographic features. A p-value <0.05 was defined as statistically significant. For comparison between the accuracy of the built-in software and our custom-built software, kappa values were utilized to assess the degree of agreement between the two software programs to identify the early-glaucoma state. The kappa statistic was calculated from a 2-by-2 table in which the results from the two software programs corresponded to that for early-glaucoma diagnoses. Kappa values between 0.0 and 0.2 indicated slight agreement; those between 0.21 and 0.40 indicated fair agreement; those between 0.41 and 0.60 indicated moderate agreement; those between 0.61 and 0.80 indicated good agreement; and those between 0.81 and 1.00 indicated almost perfect agreement. With estimated kappa values of 0.4 and 0.9 for each software's decision from a pilot study, a minimal number of 50 subjects was required to detect a 0.5 difference in kappa values above positive ratings of 0.5 with 90% statistical power [bib_ref] The kappa statistic in reliability studies: use, interpretation, and sample size requirements, Sim [/bib_ref].
The characteristics of the subjects were statistically analyzed according to whether the software programs have an identical decision from all three criteria (concordance group) or not (discordance) to assess the relationship between different criteria and the early detection ability. [fig_ref] Table 1: Patient demographics of normal group and early glaucoma group [/fig_ref] shows the percentage of males in the early-glaucoma group that was relatively higher than that of the normal group with marginal statistical significance (p = 0.04).
# Results
The mean values of CCT, MD, and visual-field index (VFI) in the normal group were higher than those in the early-glaucoma group. BCVA, IOP, AXL, major artery location (STa and ITa) and inter-arterial angle were not statistically different between the nor mal group and the early-glaucoma group. [fig_ref] Table 2: Kappa values, sensitivity and specificity of each software for detecting early glaucoma [/fig_ref] shows the agreement of the three criteria with the standard diagnosis of early-glaucoma eyes. Kappa values for criterion 3 (κ = 0.940, p < 0.001) were highest among the three criteria. However, criterion 2 (κ = 0.852, p < 0.001) also showed higher kappa values than criterion 1 (κ = 0.475, p < 0.001). The sensitivity and specificity for each criterion are listed in [fig_ref] Table 2: Kappa values, sensitivity and specificity of each software for detecting early glaucoma [/fig_ref]. Criterion 3 showed the best discriminatory ability, although criterion 2 was also better than criterion 1. [fig_ref] Figure 1: A demonstration of an overlaying function using our custom-built software [/fig_ref] describes how to utilize the custom-built software. Two representative cases showing a remarkable ability to detect glaucoma early are displayed in [fig_ref] Figure 2: A representative case of early glaucoma detection using our custom-built software [/fig_ref]. And the ability to detect early glaucoma can be improved by using the vessel position data [fig_ref] Figure 3: A representative discordance case shows the detection ability of criterion 3 in... [/fig_ref]. [fig_ref] Table 3: Demographics between positive OCT group and false negative OCT group in early... [/fig_ref] shows the demographic characteristics of the positive-OCT group and the false-negative-OCT group. Age, sex, BCVA, IOP, CCT, and AXL were similar in the two groups. However, the MD and VFI of the false-negative eyes in the OCT group were higher than those of the positive eyes, suggesting that more false-positive results were observed in early-glaucoma cases using criterion 1 (manufacturer's database). The ITa was related to the temporalization of vessels in the false-negative group with statistical significance ( p = 0.003). However, the STa was not related to the temporalization. This might be also related to the vulnerability difference between the inferior and superior RNFL defects [bib_ref] Quantifying positional variation of retinal blood vessels in glaucoma, Wang [/bib_ref].
We also compared the concordance of criteria 2 and 3 with criterion 1 in the early-glaucoma group [fig_ref] Table 4: Demographics between concordance group and discordance group [/fig_ref] , 5). [fig_ref] Table 4: Demographics between concordance group and discordance group [/fig_ref] shows that there was no factor related to the discordance between criteria 1 and 2. However, discordance was found between criteria 1 and 3, especially for MD, VFI, ITa, and inter-arterial angle. The difference was statistically significant [fig_ref] Table 5: Demographics [/fig_ref] meaning that the STa-adjusted norms were more sensitive for the detection of early RNFL changes than the AXL-adjusted norms.
# Discussion
In our study, we have constructed a new reference line for detecting retinal nerve fiber layer defects corrected by retinal vessel location or axial length using our custom-built software. Individual variation of the RNFL thickness obtained using spectral domain OCT is a prob- Values are presented as mean ± standard deviation (range) unless otherwise indicated. BCVA = best-corrected visual acuity; IOP = intraocular pressure; AXL = axial length; CCT = central corneal thickness; MD = mean deviation; VFI = visual field index; STa = angle degree of superotemporal major retinal artery; ITa =angle degree of inferotemporal major retinal artery; IAA = inter-arterial angle. * Independent samples t-test except gender variable (chi-square test). Values are presented as mean ± standard deviation (range) unless otherwise indicated. BCVA = best-corrected visual acuity; IOP = intraocular pressure; AXL = axial length; CCT = central corneal thickness; MD = mean deviation; VFI = visual field index; STa = angle degree of superotemporal major retinal artery; ITa = angle degree of inferotemporal major retinal artery; IAA = inter-arterial angle. * Independent samples t-test except gender variable (chi-square test).
lem not only for glaucoma specialists but also for patients in terms of their quality of life [bib_ref] The times they are a-changin': time to change glaucoma management, Heijl [/bib_ref]. In one study, 15% of glaucoma patients were blind during their whole life [bib_ref] Lifetime visual disability in open-angle glaucoma and ocular hypertension, Forsman [/bib_ref]. Late diagnosis is one of the most important reasons for increasing the risk for blindness [bib_ref] A comparison of glaucoma patients identified through mass screening and in routine..., Grodum [/bib_ref]. Therefore, if we can detect RNFL defects earlier, we may improve the patient's vision. From our previous study [bib_ref] Improvement of diagnostic performance regarding retinal nerve fiber layer defect using shifting..., Rho [/bib_ref] , we have confirmed that consideration of the major retinal artery position is helpful for early detection of glaucomatous RNFL defects. However, in clinics, different types and versions of OCT devices or software are used due to a financial or a technical reason. Therefore, clinicians cannot always apply the normative database directly to each patient. Our study was originally designed to overcome this issue using a simple overlaying technique with our developed software. This software can store the normative data of RNFL thicknesses, constructing a large database. With minimal modification, we can then utilize the data from the different types of OCT instruments or software, which means customization can be done per institution or per ethnic population.
Besides the location data of the major retinal vessel, myopia parameters such as axial length and spherical equivalent were significantly correlated to the peak positions of RNFL thickness according to our previous study [bib_ref] Improvement of diagnostic performance regarding retinal nerve fiber layer defect using shifting..., Rho [/bib_ref]. Therefore, using the database-constructing function in our software, we could configure variables after accumulating a large database to achieve a better diagnosis. This database program also allowed us to investigate the influence of ethnic differences on RNFL thickness or TSNIT graph contour [bib_ref] Korean normative database for time domain optical coherence tomography to detect localized..., Kang [/bib_ref].
Moreover, our software has a very simple but unique function that enables us to get two different graphs superimposed on one screen or monitor. This function shines especially in the real-world practice that is very heterogeneous in terms of the types of monitors and the computer settings. This simple technique enables us not only to detect the RNFL defect in a different set of normative databases but also to use the software in different types of OCT devices (e.g., Cirrus or RTView). With a single placement of a cursor in the software window, we could make the reference graph transparent as necessary [fig_ref] Figure 1: A demonstration of an overlaying function using our custom-built software [/fig_ref]. At first inspection, confusion due to the background color occurred; however, this issue was easily overcome by moving the cursor to optimize the transparency of the reference graph. The inter-user variability and the intra-user variability were checked by two examiners (a glaucoma specialist and a resident). The kappa values were all 1.0, implying that the problem was negligible.
Although the major temporal vessel positions were known to be correlated with the peak of the RNFL [bib_ref] Blood vessel contributions to retinal nerve fiber layer thickness profiles measured with..., Hood [/bib_ref] , we found a slight discrepancy between superior and inferior artery locations. Wang et al. [bib_ref] Quantifying positional variation of retinal blood vessels in glaucoma, Wang [/bib_ref] reported that peripapillary superior artery positions were significantly nasalized according to glaucoma severity. This is a possible reason for the ITa's statistically significant relationship to temporalization of vessels in the false-negative group. In our study, STa was not significant.
In our previous study, the correlation of RNFL peak location and other factors, such as major retinal artery or vein position, was observed. AXL was evaluated, and we found that the location of the major retinal artery was the factor most correlated with the RNFL peak location. Hood [bib_ref] Improving our understanding, and detection, of glaucomatous damage: an approach based upon..., Hood [/bib_ref] also reported difficulty in interpreting the metrics that each OCT device provided although numerous studies have documented their reasonably good sensitivities (e.g., 80%). If the study included only severe glaucoma cases, then 80% sensitivity would be a serious issue. To prevent this issue, we have included eyes with early glaucomatous changes. Because the major retinal artery location and AXL were the two most relevant factors to the RNFL-thickness data, we have assessed them in this study.
A major limitation of our study was the fundamental feature of indirectly obtained RNFL-thickness data because of the manufacturer's policy. However, as we discussed in another report [bib_ref] Improvement of diagnostic performance regarding retinal nerve fiber layer defect using shifting..., Rho [/bib_ref] , qualitative and quantitative comparisons have shown the reproducibility error was smaller than the previously reported variation of OCT scans [bib_ref] Retinal nerve fiber layer imaging with spectral-domain optical coherence tomography a study..., Leung [/bib_ref]. There were two false-positive cases in criterion 3 with an RNFL defect in the nasal region that generally correlated less to the glaucoma type. We believe that this might be one good reason to have a larger database to allow the utilization of more variables. Although the RN-FL-thickness data had to be manually extracted, we uploaded new normal data and visualized this data easily as a graph using our software.
In this study, we utilized only one variable at a time for categorizing the reference graph, which was determined by the median value of the artery positions or AXL values and which produced four graphs. The reason for this was efficiency. We compared the accuracy of this method to that of another method that had four subgroups determined by the bottom quarter, the median, and the top quarter values.
There was no significant difference between the two methods (kappa value [95% confidence interval] of methods 1 and 2, 0.980 [0.941 to 1.000] vs. 0.960 [0.905 to 1.000], respectively). Although this study included early-glaucoma eyes for the validation set, in future studies data with pre-perimetric-glaucoma eyes or ocular-hypertensive eyes would be informative. After unification of ethnicity and graph shifting according to retinal artery location, the most remarkable change was the narrowing of the 99% interval. Because the increase of sensitivity can lead to the early detection of glaucoma, more subtle changes can theoretically be detected.
In conclusion, even in early-glaucoma states, the accuracy of RNFL defect-detection can be dramatically improved by considering the uniform ethnic population and the location of the major retinal artery. To achieve that goal, a customized database per institution with a graph-overlaying technique could provide a simple and feasible way for clinical practice.
## Conflict of interest
No potential conflict of interest relevant to this article was reported.
[fig] Figure 1: A demonstration of an overlaying function using our custom-built software. (A) The main window of the custom-built software shows the principal features including database management, artery information input window, and the graph visualization function key.(B,C) The background was seen using a manual adjustment cursor. [/fig]
[fig] Figure 2: A representative case of early glaucoma detection using our custom-built software. (A) A red-free photograph shows a localized retinal nerve fiber layer (RNFL) defect (arrows) in the inferotemporal region. (B) Early damage of the superior visual field was noted in the subject's right eye that corresponded to the inferotemporal RNFL defect. (C) The inferotemporal RNFL defect was not detected by the built-in software; (D) however, the defect was detected by our custom-built software. GHT = glaucoma hemifield test; VFI = visual-field index; MD = mean deviation; PSD = pattern standard deviation. [/fig]
[fig] Figure 3: A representative discordance case shows the detection ability of criterion 3 in the early diagnosis of glaucoma. (A) Red-free photography shows a wedge-shaped retinal nerve fiber layer (RNFL) defect in the inferotemporal region. The short red arrows indicate the margin of the defect. (B) Pattern deviation and global indices of visual-field tests demonstrate outside normal limit defects in the superonasal quadrant. (C) Criterion 1 (top) and criterion 2 (middle) cannot detect any RNFL defect based on the lower 1% limit of RNFL thickness. Only criterion 3 (bottom) detects the glaucomatous RNFL defect (long red arrow). GHT = glaucoma hemifield test; VFI = visual-field index; MD = mean deviation; PSD = pattern standard deviation. [/fig]
[table] Table 1: Patient demographics of normal group and early glaucoma group [/table]
[table] Table 2: Kappa values, sensitivity and specificity of each software for detecting early glaucoma [/table]
[table] Table 3: Demographics between positive OCT group and false negative OCT group in early glaucoma eye (n = 79)Values are presented as mean ± standard deviation (range) unless otherwise indicated. OCT = optical coherence tomography; BCVA = best-corrected visual acuity; IOP = intraocular pressure; AXL = axial length; CCT = central corneal thickness; MD = mean deviation; VFI = visual field index; STa = angle degree of superotemporal major retinal artery; ITa =angle degree of inferotemporal major retinal artery; IAA = inter-arterial angle. [/table]
[table] Table 4: Demographics between concordance group and discordance group (criterion 1 vs. criterion 2) in early glaucoma eye (n = 79) [/table]
[table] Table 5: Demographics [/table]
|
Protometabolism as out-of-equilibrium chemistry
## Protometabolism
What makes chemistry metabolic is the ability to sustain a living cell, which means that chemistry prior to the emergence of life cannot be described as metabolic. Nevertheless, extant metabolic chemistry must have had progenitors, and it is this set of chemical reactions that existed before the transition to biology that is frequently referred to as protometabolism. However, the term has been used loosely, and we would argue in a way that clouds our understanding of the emergence of life. The chemistry that sustains life is complex with many pathways and components integrated together within a highly organized compartment that defies thermodynamic equilibrium. Narrow investigations of individual synthetic pathways in isolation and in absence of a compartment will likely be less informative than, as Ruiz-Mirazo et al. [bib_ref] Prebiotic systems chemistry: new perspectives for the origins of life, Ruiz-Mirazo [/bib_ref] argue, systems level approaches that search for commonalities and synergies between chemistries and phases that could help to explain why life is as it is today.
Extant metabolism consists of both catabolic and anabolic reactions, i.e. the breakdown and synthesis of molecules, respectively. The exergonic chemistry of catabolism drives the endergonic reactions of anabolism that are needed to sustain the cell. But, this is not a one-way process where anabolism relies on catabolism alone. The products of anabolism are needed to mediate and regulate catabolism. The current search for separate prebiotic analogues of extant anabolic and catabolic chemistries largely does not address this critical feature of biology, which is how prebiotic systems chemistry gave rise to a mixture of reactions that tie together in a way that defies equilibrium. Such primordial reaction networks that join endergonic with exergonic pathways, ideally through common currencies, as discussed below, would in our estimation be better described as protometabolic in comparison to narrower investigations of nonenzymatic mimics of extant catabolism or anabolism in isolation.
## Prebiotic analogues of anabolism
Much effort has centred on prebiotic mimics of anabolism whereby plausible routes towards the synthesis of the building blocks of life are deciphered. What has been more contentious is whether these prebiotic reaction pathways resembled modern-day metabolism or were often distinct, reflecting the different constraints imposed by an ancient, prebiotic environment [bib_ref] Illuminating life's origins: UV photochemistry in abiotic synthesis of biomolecules, Green [/bib_ref] [bib_ref] Life's biological chemistry: a destiny or destination starting from prebiotic chemistry?, Krishnamurthy [/bib_ref] [bib_ref] Nonenzymatic metabolic reactions and life's origins, Muchowska [/bib_ref]. Both arguments invoke the ability of enzymes to facilitate but not invent chemistry, and both tend to rely on laboratory approximations of reactions under thermodynamic, as opposed to kinetic, control. Proponents of prebiotic chemistry unconstrained by extant biology find it unsurprising that biological reactions can be run nonenzymatically, particularly when subjected to extremes in temperature or pH. Activation energy barriers can be traversed without enzymes, so it would be more surprising if the reaction could not proceed at all. Therefore, observing an uncontrolled nonenzymatic analogue of extant metabolism may tell us little about the role of such chemistry in the origins of life, particularly if metabolism requires regulatory processes that are not easily obtainable without enzymes. Conversely, if a reaction can run nonenzymatically, why would it not? A functioning reaction cycle from which the building blocks of life can be derived is attractive as the emergence of modern-day metabolism could be envisaged to proceed by the stepwise addition of enzymes [bib_ref] On the origin of metabolic pathways, Lazcano [/bib_ref] [bib_ref] The RNA world and the origin of metabolic enzymes, Ralser [/bib_ref]. As with most arguments, reality likely does not fit cleanly into a single category. Just as the RNA world hypothesis gives the most compelling explanation for the rise of Darwinian evolution, it is naive to think that life began with RNA alone. Similarly, while the abundance of biological molecules that can be synthesized from cyanide prebiotically [bib_ref] Provisioning the origin and early evolution of life, Wu [/bib_ref] cannot be ignored, it stands to reason that at least some prebiotic reactions were similar to contemporary biological analogues [bib_ref] Out of fuzzy chemistry: from prebiotic chemistry to metabolic networks, Peretó [/bib_ref].
Although differences in opinion are healthy, particularly when subjected to experimental evaluation, opinions that lead to the imposition of poorly supported rigid constraints are unhelpful. Since we do not know the conditions that gave rise to life, we are left with the realities of chemical reactivity and geological plausibility. A path between nonenzymatic prebiotic chemistry that resembles extant biological pathways and the last universal common ancestor royalsocietypublishing.org/journal/rsta Phil. Trans. R. Soc. A 380:
(LUCA) is easy to imagine, which leads many to confidently extrapolate back in time from our best guesses as to what LUCA looked like [bib_ref] The last universal common ancestor between ancient Earth chemistry and the onset..., Weiss [/bib_ref]. However, if such an approach is taken at the exclusion of a broader investigation of what prebiotically plausible molecules give rise to, then we are likely to be led astray. As others have argued before, the early steps leading to life may have been erased by time and thus cannot be easily perceived from extant artefacts [bib_ref] Life's biological chemistry: a destiny or destination starting from prebiotic chemistry?, Krishnamurthy [/bib_ref] [bib_ref] Provisioning the origin and early evolution of life, Wu [/bib_ref]. Phylogeny can tell us much about the early evolution of life, but it is not logical to allow the genetics of a highly complex organism with 300-400 genes [bib_ref] Determination of the core of a minimal bacterial gene set, Gil [/bib_ref] [bib_ref] The physiology and habitat of the last universal common ancestor, Weiss [/bib_ref] dictate how chemistry prior to genetics must have proceeded. Ultimately, a thorough testing of different chemical scenarios congruent with geology is needed so that experimental data and hypotheses can be compared and evaluated .
Regardless of the vantage point, what works on prebiotic analogues of anabolism have in common is an attempt to understand how larger biological molecules could have been synthesized prebiotically from smaller, abundant chemical resources. Although there were early, important studies on the prebiotic synthesis of biological molecules [bib_ref] Organic compound synthesis on the primitive earth: several questions about the origin..., Miller [/bib_ref] [bib_ref] Ueber künstliche Bildung des Harnstoffs, Wöhler [/bib_ref] , modern work applies knowledge of synthetic organic chemistry to reactions in water or at the air-water interface [bib_ref] Water-air interfaces as environments to address the water paradox in prebiotic chemistry:..., Deal [/bib_ref] , under prebiotically plausible conditions. As CO 2 was more abundant on the prebiotic Earth, and as the Wood-Ljungdahl pathway (or the reductive acetyl-coenzyme A pathway) has been hypothesized to have been exploited by LUCA, CO 2 often features as a carbon source for the synthesis of more complex molecules. Metallic iron or precipitates of metal ions under a variety of conditions have been found to reduce CO 2 to formate, methanol, methane, acetate and pyruvate [bib_ref] Primordial carbonylated iron-sulfur compounds and the synthesis of pyruvate, Cody [/bib_ref] [bib_ref] Abiogenic methane formation and isotopic fractionation under hydrothermal conditions, Horita [/bib_ref] [bib_ref] 2020 CO 2 reduction driven by a pH gradient, Hudson [/bib_ref] [bib_ref] A hydrogen-dependent geochemical analogue of primordial carbon and energy metabolism, Preiner [/bib_ref]. Metal ions are further invoked for the catalysis of several steps of an abiotic reductive citric acid cycle, which can be coupled to the synthesis of alanine [bib_ref] Metals promote sequences of the reverse Krebs cycle, Muchowska [/bib_ref]. Alternatively, glycolate can be synthesized from CO 2 and in the presence of UV light and sulfite without the participation of metals. Further irradiation of the glycolate with sulfite produces several components of the citric acid cycle, including citrate, malate and succinate. Metal-free reactions between pyruvate and glyoxylate can also lead to a prebiotic analogue of the reductive citric acid cycle and produce amino acids via transamination with glycine [bib_ref] 2020 A plausible metal-free ancestral analogue of the Krebs cycle composed entirely..., Stubbs [/bib_ref]. Conversely, if one considers the rich chemistry of the planetary abundant cyanide [bib_ref] High energy radical chemistry formation of HCN-rich atmospheres on early earth, Ferus [/bib_ref] [bib_ref] The cold origin of life: A. implications based on the hydrolytic stabilities..., Miyakawa [/bib_ref] [bib_ref] Organometallic compounds as carriers of extraterrestrial cyanide in primitive meteorites, Smith [/bib_ref] , particularly in the presence of UV light, metal ions and reductants such as sulfite, then over half of the 20 amino acids [bib_ref] Prebiotic synthesis of cysteine peptides that catalyze peptide ligation in neutral water, Foden [/bib_ref] [bib_ref] Common origins of RNA, protein and lipid precursors in a cyanosulfidic protometabolism, Patel [/bib_ref] , pyrimidine and purine ribonucleotides [bib_ref] A high-yielding, strictly regioselective prebiotic purine nucleoside formation pathway, Becker [/bib_ref] [bib_ref] Divergent prebiotic synthesis of pyrimidine and 8-oxo-purine ribonucleotides, Stairs [/bib_ref] [bib_ref] Selective prebiotic formation of RNA pyrimidine and DNA purine nucleosides, Xu [/bib_ref] [bib_ref] Synthesis of activated pyrimidine ribonucleotides in prebiotically plausible conditions, Powner [/bib_ref] and precursors to lipids [bib_ref] Common origins of RNA, protein and lipid precursors in a cyanosulfidic protometabolism, Patel [/bib_ref] can all be built by photoreductive homologation pathways orthogonal to that of extant biology. Although we favour the robust nonbiological-like routes to the synthesis of the building blocks of life, and it is important to directly compare published data from different perspectives [bib_ref] Illuminating life's origins: UV photochemistry in abiotic synthesis of biomolecules, Green [/bib_ref] , the intent here is not to promote or diminish one perspective but to highlight that several paths have been published for the synthesis of the building blocks and that these pathways often give rise to and consume pyruvate and other α-ketoacids.
## Prebiotic analogues of catabolism
Surprisingly, less effort has been expended in investigating prebiotic mimics of catabolism. Metabolic intermediates, such as glucose-6-phosphate, have been shown to break down to pyruvate in the presence of Fe 2+ at elevated temperatures in a manner reminiscent of glycolysis [bib_ref] Non-enzymatic glycolysis and pentose phosphate pathway-like reactions in a plausible Archean ocean, Keller [/bib_ref]. Similar conditions lead to both the synthesis and degradation of the intermediates of the citric acid cycle when starting from pyruvate and glyoxylate . The reaction pathway can be exploited to synthesize amino acids if hydroxylamine and metallic iron are added . Non-metal-dependent catabolic-like pathways have also been described. An analogue of glycolysis starting from the simple sugar glyceraldehyde gives rise to a series of reactions that generate phosphoenolpyruvate and pyruvate when fed with cyanide, the phosphorylating agent diamidophosphate [bib_ref] Regioselective α-phosphorylation of aldoses in aqueous solution, Krishnamurthy [/bib_ref] and glycoaldehyde [bib_ref] Prebiotic synthesis of phosphoenol pyruvate by αphosphorylation-controlled triose glycolysis, Coggins [/bib_ref]. Additionally, two interconnecting reaction networks, referred to as the 4-hydroxy-2-ketoglutarate and malonate cycles, that function as a type of prebiotic analogue of the citric acid cycle are sustained by feeding with pyruvate, glyoxylate and hydrogen peroxide [bib_ref] Linked cycles of oxidative decarboxylation of glyoxylate as protometabolic analogs of the..., Springsteen [/bib_ref]. Here, reaction intermediates can be syphoned off to synthesize aspartate in the presence of ammonia [bib_ref] Linked cycles of oxidative decarboxylation of glyoxylate as protometabolic analogs of the..., Springsteen [/bib_ref]. As with the prebiotic analogues of anabolic However, the ingredients of life alone are not enough to make a living cell. If that were so, then we would be able to take apart and put back together a living cell. Further, metabolism is more dissimilar to laboratory-based synthesis than is often appreciated. Scientists are skilled at manipulating thermodynamics and reactivity to push reactions forward, but metabolism does not solely rely on such direct manipulations of chemistry. Instead, metabolism elegantly ties the two branches (i.e. anabolic and catabolic) together through an intermediary process that deposits, stores and spends energy to sustain a vast array of endergonic chemistry (figure 1). That is, life-as-we-know-it plugs into a fuel source with a wire of molecules that consecutively transfers electrons [bib_ref] The origin of life: a case is made for the descent of..., Trefil [/bib_ref]. The molecules that engage in electron transfer most commonly include metallocofactors, ribodinucleotides and quinones. The fuel can be reduced organic molecules, such as sugars and lipids, or inorganic molecules, such as molecular hydrogen, hydrogen sulfide and ferrous ions. The thermodynamically favourable flux of electrons through the cell is coupled to processes that store the released energy into common currencies. This is a key invention of biology [bib_ref] The origin of life: a case is made for the descent of..., Trefil [/bib_ref] and is distinct from the more direct coupling that is typically used in prebiotic chemical studies. In biology, the common currencies consist of H + gradients, Na + gradients and adenosine triphosphate, which are used by the cell to pay for the costly work of manufacturing and repairing cellular infrastructure [bib_ref] The origin of life: a case is made for the descent of..., Trefil [/bib_ref]. What is unclear is whether such regulated fluxes of energy that are used to keep cellular systems out-of-equilibrium emerged early or were a later invention.
The concept of life being a by-product of dissipating energy gradients is not new. Others have eloquently discussed the potential chemiosmotic [bib_ref] Coupling of phosphorylation to electron and hydrogen transfer by a chemiosmotic type..., Mitchell [/bib_ref] origins of life, and how geological gradients could have given rise to the Earth's first cells [bib_ref] Energy at life's origin, Martin [/bib_ref]. Broadly speaking, such hypotheses tend to place the origins of life at hydrothermal vents, which is in contrast to more genetic centric work, such as the RNA world hypothesis, which has favoured surface conditions. Although both perspectives may focus on syntheses starting from different carbon sources (carbon dioxide versus hydrogen cyanide) and with a bigger or smaller role of metal ions, in a narrow sense, there is commonality in that both frequently seek to delineate how the building blocks of life were made. The deeper difference between the two perspectives concerns the placement of metabolism within a timeline encompassing the emergence of the Earth's first cells (figure 2). In other words, were the building blocks of life synthesized by a protometabolic system prior to the emergence of a protocell (figure 2a) or did the building blocks accumulate by distinct chemistry before the advent of protometabolism (figure 2b)? Thus far, much more experimental work has been carried out on chemistry congruent with surface conditions where the building blocks could have accumulated [bib_ref] Illuminating life's origins: UV photochemistry in abiotic synthesis of biomolecules, Green [/bib_ref] [bib_ref] UV photostability of three 2-aminoazoles with key roles in prebiotic chemistry on..., Todd [/bib_ref] [bib_ref] Ultravioletdriven deamination of cytidine ribonucleotides under planetary conditions, Todd [/bib_ref] [bib_ref] Photochemically driven redox chemistry induces protocell membrane pearling and division, Zhu [/bib_ref] ; however, studies have recently begun to investigate the plausibility of hydrothermal vent scenarios guided by the presumption of an earlier appearance of protometabolism [bib_ref] 2020 CO 2 reduction driven by a pH gradient, Hudson [/bib_ref] [bib_ref] Effects of amino acids on phosphate adsorption onto iron (oxy)hydroxide minerals under..., Flores [/bib_ref] [bib_ref] Promotion of protocell self-assembly from mixed amphiphiles at the origin of life, Jordan [/bib_ref]. While we appreciate some aspects of the hydrothermal vent theories, such as the role of metal ions, including iron-sulfur clusters [bib_ref] The origin of the RNA world: co-evolution of genes and metabolism, Copley [/bib_ref] [bib_ref] How did LUCA make a living? Chemiosmosis in the origin of life, Lane [/bib_ref] , and the emphasis on energetics, hydrothermal vents are not the only places capable of providing an energy source.
## Non-hydrothermal vent, out-of-equilibrium processes
Living cells are out-of-equilibrium chemical systems that harness the dissipation of fuel sources to maintain their highly ordered state (figure 3). The persistence of these metastable, out-ofequilibrium arrangements is also known as dynamic kinetic stability [bib_ref] Towards an evolutionary theory of the origin of life based on kinetics..., Pascal [/bib_ref] [bib_ref] Extending the concept of kinetic stability: toward a paradigm for life, Pross [/bib_ref] [bib_ref] How and why kinetics, thermodynamics, and chemistry induce the logic of biological..., Pross [/bib_ref] , which is at times additionally referred to as dynamic self-assembly when describing chemical systems without biological components [bib_ref] Dynamic self-assembly of magnetized, millimetre-sized objects rotating at a liquid-air interface, Grzybowski [/bib_ref]. Attempts to construct chemical systems that imitate features of life tend to focus on mimicking three different facets of life-as-we-know-it as opposed to building an integrated protocell, including (i) error-prone replication capable of Darwinian evolution, (ii) metabolic-like chemistry that sustains an overall dissipative system with a fuel source and (iii) compartmentalization to separate the chemical system from the environment. Although the assimilation of all three within a protocell is presumably needed to generate a highly adaptable system capable of surviving environmental fluctuations, more narrow investigations have already begun to reveal how mechanisms, such as autocatalysis, can give rise to characteristics that imitate some features of life. Importantly, these studies directly link thermodynamics, kinetics, competition and selection in a way that is absent in discussions of the energetic fluxes present at hydrothermal vents.
What these studies typically have in common is either the presence of autocatalysis or dependence on reactions that proceed along different paths in the forward and reverse directions. Autocatalysis is simply when the product of a reaction catalyses its own formation [bib_ref] Mechanisms of autocatalysis, Bissette [/bib_ref] [bib_ref] Autocatalytic models for the origin of biological homochirality, Blackmond [/bib_ref] and represents an efficient means of funnelling reactions down a restricted path towards a common product. A well-known example is the ability of the Soai reaction to resolve a nearly racemic mixture to enantiopurity, perhaps revealing the types of mechanisms that led to life being largely dependent on homochirality [bib_ref] Autocatalytic models for the origin of biological homochirality, Blackmond [/bib_ref]. However, the examples typically encountered in origins of life research are more complex, rely on a phase transition of some type and focus on replication. For example, Luisi and co-workers [bib_ref] Autocatalytic self-replicating micelles as models for prebiotic structures, Bachmann [/bib_ref] demonstrated how the hydrolysis of ethyl caprylate produces ethanol and the fatty acid caprylate. When sufficient fatty acid is produced, the caprylate assembles into micelles that then catalyse the formation of more fatty acid [bib_ref] Autocatalytic self-replicating micelles as models for prebiotic structures, Bachmann [/bib_ref] , although later computational work suggests that the mechanism is more complex [bib_ref] Mechanisms of autocatalysis, Bissette [/bib_ref] [bib_ref] Phase-transfer model for the dynamics of 'micellar autocatalysis, Buhse [/bib_ref] [bib_ref] Origin of autocatalysis in the biphasic alkaline hydrolysis of C-4 to C-8..., Buhse [/bib_ref]. Here, it is not solely the formation of a molecule, the fatty acid, that is critical, but that this molecule is able to self-assemble into a higher order aggregate structure that then pulls the system away from equilibrium in addition to acting as a catalyst for the synthesis of more fatty acid. This type of A + B C → D, where D is an aggregate, or new phase, composed of C is common.
More recent examples [bib_ref] 2020 From self-replication to replicator systems en route to de novo life, Adamski [/bib_ref] with modified peptides instead of fatty acids exploit the ability of some of the intermediate products to aggregate into fibres (figure 4a). In one example, peptides are modified with a benzenedithiol group that upon oxidation forms an array of disulfide macrocycles in which one specific type, the hexameric form, assembles into fibres [bib_ref] Exponential self-replication enabled through a fibre elongation/breakage mechanism, Colomb-Delsuc [/bib_ref]. In this case, the fibres represent the newly formed phase similar to the micelle example above. If none of the disulfide macrocycles formed were capable of phase separation, then the final pool of molecules would reflect the free energy of the products and their placement within the free energy landscape [bib_ref] Selection and amplification of hosts from dynamic combinatorial libraries of macrocyclic disulfides, Otto [/bib_ref]. Conversely, when one molecule is capable of self-assembly, the gained free energy favours the accumulation of the self-assembled state at the expense of the non-assembling competitors [bib_ref] Self-duplicating amplification in a dynamic combinatorial library, Xu [/bib_ref]. The dynamics of similar systems has been studied, including systems with aggregates composed of prebiotic precursors of RNA [bib_ref] Efficient self-assembly in water of long noncovalent polymers by nucleobase analogues, Cafferty [/bib_ref] [bib_ref] Spontaneous prebiotic formation of a β-ribofuranoside that self-assembles with a complementary heterocycle, Chen [/bib_ref]. However, the described peptide version can also be mechanically broken, leading to a form of replication whereby the fibre fragments seed the growth of more fibres in a way similar to that seen with crushed crystals [bib_ref] Exponential self-replication enabled through a fibre elongation/breakage mechanism, Colomb-Delsuc [/bib_ref].
The key to ensuring a continuous out-of-equilibrium state is the coupling of the system to its energy source. Therefore, the building blocks cannot be completely consumed and the final selfassembled state itself should not inhibit exponential self-replication [bib_ref] Minimal replicator theory I: parabolic versus exponential growth, Von Kiedrowski [/bib_ref]. Serial transfer, whereby a portion of the self-replicating set is introduced to a new environment with replenished fuel sources, can be used to overcome equilibrium [bib_ref] Serial transfer can aid the evolution of autocatalytic sets, Hordijk [/bib_ref]. Perhaps more elegantly, the mechanically induced breakage of fibres described above additionally helps to avoid reaching equilibrium, (1), e.g. thiol containing peptides, react with a fuel molecule or oxidant (2) to form a dynamic combinatorial library (3). One molecular member of the library can self-assemble (4), thus pulling the system out-of-equilibrium. The self-assembled state (4) itself, or after mechanical breakage, can catalyse the formation of more of the self-assembled state. Further, new properties of the self-assembled state, either intrinsically or by uptake of cofactors from the environment, may emerge which facilitate the productive consumption of building blocks (5). (b) Differential formation and degradation pathways. A mixture of soluble building blocks (1), e.g. fatty acids, react with fuel molecules (2), such as a carbodiimide, to generate products of decreased solubility (3) that can be degraded back to the building block through hydrolysis. The self-assembly of a subset of products (4) inhibits the back reaction, thereby pulling the system away from equilibrium and sustaining an assembly of unstable molecules over time. (Online version in colour.) because the broken fibres increase the number of available self-replicating template sites [bib_ref] Exponential self-replication enabled through a fibre elongation/breakage mechanism, Colomb-Delsuc [/bib_ref]. That is, the change in copy number of the self-replicator further pulls the system towards dynamic kinetic stability instead of thermodynamic equilibrium [bib_ref] Towards an evolutionary theory of the origin of life based on kinetics..., Pascal [/bib_ref] [bib_ref] Extending the concept of kinetic stability: toward a paradigm for life, Pross [/bib_ref] [bib_ref] How and why kinetics, thermodynamics, and chemistry induce the logic of biological..., Pross [/bib_ref]. One objective of those working on such systems is to allow for open-ended evolution in a way similar but not limited to Red Queen coevolution [bib_ref] Grand views of evolution, De Vladar [/bib_ref] [bib_ref] Molecular evolution as predicted by natural selection, Van Valen [/bib_ref] whereby a self-replicating entity partially, or wholly, displaces the free energy landscape of a system. As a consequence, other self-replicating entities present within the ecosystem must engage in reciprocal adaptive interactions. The resulting ecosystem has the potential to overcome Eigen's paradox, which relates the complexity of the system to copying fidelity [bib_ref] Selforganization of matter and the evolution of biological macromolecules, Eigen [/bib_ref] , because interconnected catalytic hypercycles can function as a dynamic community ensuring accuracy in self-replication, in a way similar to Eigen's quasi-species model. The goal in research on self-replicators is often to identify replicators that acquire metabolism by catalysing chemical reactions within the environment in a way that promotes their own replication. In impressive work by Otto and co-workers [bib_ref] Emergence of light-driven protometabolism on recruitment of a photocatalytic cofactor by a..., Santiago [/bib_ref] , peptide fibres were found to recruit chromophore cofactors from the environment, which led to the photooxidation of the building blocks. As oxidation of the building blocks is necessary to form the disulfide linked macrocycles necessary for subsequent self-assembly into fibres, a positive feedback loop emerged from the chemical system that allowed the self-replicator to better harness the surrounding environment. Significantly, such feedback loops can also emerge from the intrinsic properties of the fibres themselves through interactions with the environment without the need of cofactors [bib_ref] 2020 Chance emergence of catalytic activity and promiscuity in a self-replicator, Ottelé [/bib_ref].
## Non-autocatalytic out-of-equilibria
It is additionally possible to maintain the existence of a thermodynamically unfavourable state without autocatalysis if the product of the pathway is formed and degraded by different paths [bib_ref] Transient assembly of active materials fueled by a chemical reaction, Boekhoven [/bib_ref]. In this case, the depletion of the fuel leads to the disassembly of the nonkinetically trapped aggregate. Such examples typically exploit phase separation, with the formation of actin filaments and microtubules representing biological examples [bib_ref] Transient assembly of active materials fueled by a chemical reaction, Boekhoven [/bib_ref]. An instructive case is that of the carbodiimide mediated formation of fatty acid anhydrides that then phase separate into oil droplets that are more resistant to hydrolysis back to the fatty acid building block [bib_ref] Self-selection of dissipative assemblies driven by primitive chemical reaction networks, Tena-Solsona [/bib_ref] (figure 4b). If the starting pool of building blocks consists of fatty acids of different lengths, then the fatty anhydrides of the longer chain fatty acids preferentially form oil droplets, meaning that the major products consist of the longer chain fatty anhydrides. As expected, the preferential formation of one subset of fatty anhydrides is accentuated by repeated refuelling, particularly when performed in batch additions as opposed to continuous feeding. Therefore, even without autocatalysis, competition can emerge between different chemical systems, with one, out-of-equilibrium subset persisting and outcompeting competitors over time [bib_ref] Self-selection of dissipative assemblies driven by primitive chemical reaction networks, Tena-Solsona [/bib_ref]. It is interesting to note that older examples of chemical systems persisting in the presence of separate chemical reactions that feed and degrade fatty acid vesicles were reported in support of the importance of autopoiesis in the origins of life [bib_ref] A chemical model of homeostasis, Zepik [/bib_ref].
# Conclusion
We have not attempted to go through all the examples but rather to point out that dissipative systems that mimic some critical features of life can emerge in scenarios that do not depend on hydrothermal vent conditions [bib_ref] Design of multi-phase dynamic chemical networks, Chen [/bib_ref] [bib_ref] 2020 Selection from a pool of self-assembling lipid replicators, Colomer [/bib_ref] [bib_ref] Triarylamine-based supramolecular polymers: structures, dynamics, and functions, Moulin [/bib_ref] [bib_ref] Self-replicating amphiphilic β-sheet peptides, Rubinov [/bib_ref]. Energy is not a unique property of hydrothermal vents. The Sun is a clear example of a prebiotic energy source that would have been inaccessible in the deep sea [bib_ref] Illuminating life's origins: UV photochemistry in abiotic synthesis of biomolecules, Green [/bib_ref] [bib_ref] Water-air interfaces as environments to address the water paradox in prebiotic chemistry:..., Deal [/bib_ref] [bib_ref] Ultravioletdriven deamination of cytidine ribonucleotides under planetary conditions, Todd [/bib_ref] , and chemical energy sources, e.g. isonitriles, likely drove prebiotic chemistry forward [bib_ref] Harnessing chemical energy for the activation and joining of prebiotic building blocks, Liu [/bib_ref]. Work on non-hydrothermal, out-of-equilibrium systems thus far has made impressive progress, but none of these previous studies attempted to work within prebiotically plausible conditions and rarely made use of molecules that were clearly on a path from prebiotic chemistry to life-as-we-know-it, with a possible exception being the formose reaction [bib_ref] Borate minerals stabilize ribose, Ricardo [/bib_ref] , although not universally accepted [bib_ref] Mimicking the surface and prebiotic chemistry of early Earth using flow chemistry, Ritson [/bib_ref]. If prebiotically plausible dissipative systems were experimentally demonstrated that clearly tied into a protocellular architecture in a way that helped explain extant biology, then the case for one prebiotic scenario over another would be strengthened. Such a system may look quite different from the directly coupled examples developed thus far, because the way in which biology exploits the dissipation of a fuel source to drive cellular organization and function is considerably more complex. It is, therefore, fair to ask if extant metabolic-like chemistry emerged early or was a later invention.
To help gain insight into this problem, we believe that more work is needed that integrates multiple components and pathways together into protocellular compartments. That is, we should take what has been learned from out-of-equilibrium, dissipative chemical systems and apply these principles to the construction of fuel-driven, prebiotically plausible protocells. The lack of use of vesicles in studies on protometabolism is surprising, as the simple presence of a membrane alters chemistry, facilitating peptide synthesis [bib_ref] Fatty acids' double role in the prebiotic formation of a hydrophobic dipeptide, Murillo-Sánchez [/bib_ref] [bib_ref] Competition between model protocells driven by an encapsulated catalyst, Adamala [/bib_ref] , competition [bib_ref] Competition between model protocells driven by an encapsulated catalyst, Adamala [/bib_ref] [bib_ref] The emergence of competition between model protocells, Chen [/bib_ref] and provides for ways to tie chemistry to the survivability of the protocell [bib_ref] Permeability-driven selection in a semi-empirical protocell model: the roots of prebiotic systems..., Piedrafita [/bib_ref]. There are already some clues regarding how catabolic and anabolic chemistry could have been coupled. The prebiotic synthesis of ironsulfur clusters has been demonstrated [bib_ref] UV-light-driven prebiotic synthesis of iron-sulfur clusters, Bonfio [/bib_ref] , and iron-sulfur peptides can engage in electron transfer reactions that lead to the generation of a proton gradient across diacyl phospholipid membranes [bib_ref] Prebiotic iron-sulfur peptide catalysts generate a pH gradient across model membranes of..., Bonfio [/bib_ref]. Further, prebiotic fuel sources, such as α-ketoacids, can initiate electron transfer reactions in a way that resembles what is found in biology [bib_ref] Protometabolic reduction of NAD+ with α-Keto acids, Basak [/bib_ref]. It seems that several pieces of the puzzle are there, ready to be put in place, with the striking exception of nucleic acids. Although the challenge is great, we suspect that many critical insights will come from a more concerted effort to investigate how prebiotically plausible chemistry could have fuelled early protocells. Regardless of the outcome, the data will likely help settle some of the ongoing debates regarding the role of (proto)metabolism in the origins of life.
Data accessibility. This article has no additional data. Authors' contributions. All authors contributed to conceptualization, writing-original draft and writing-review and editing. All authors gave final approval for publication and agreed to be held accountable for the work performed therein.
Competing interests. We declare we have no competing interests. Funding. We acknowledge the support of the Simons Foundation (grant no. 290358FY19) and the Natural Sciences and Engineering Research Council of Canada (NSERC) (grant no. RGPIN-2020) to S.S.M.
[fig] 2022: The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/ by/4.0/, which permits unrestricted use, provided the original author and source are credited. royalsocietypublishing.org/journal/rsta Phil. Trans. R. Soc. A 380: 20200423 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . [/fig]
[fig] Figure 1: royalsocietypublishing.org/journal/rsta Phil. Trans. R. Soc. A 380: 20200423 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extant metabolism works, in part, by funnelling the energy released from the oxidative degradation of varied fuel sources into common currencies, which are then used to drive the energetically costly reductive synthetic processes needed to sustain the cell. reactions, the published prebiotic analogues of catabolic networks frequently feature pyruvate and other α-ketoacids. The aforementioned work demonstrates what types of reactions and molecules were potentially chemically accessible on the prebiotic Earth. Such research is critically important, because knowing what was present addresses what types of protocells could have formed. [/fig]
[fig] Figure 2: Comparing timelines of early life innovations. (a) A generic representation of hydrothermal vent scenarios where a constant influx of energy and matter through porous rock sustains prebiotic analogues of metabolism. Over time, the system evolves, generating biopolymers and other molecular machinery before an escape event that leads to encapsulation within a lipid vesicle. (b) A generic representation of the surface of the early Earth where building blocks assemble preferentially within lipid vesicles. This spatial isolation imposes a selective pressure that gives rise to protocells supported by an internal chemistry. (Online version in colour.) [/fig]
[fig] Figure 3: royalsocietypublishing.org/journal/rsta Phil. Trans. R. Soc. A 380: 20200423 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Living systems use metabolism to exploit the free energy released from the degradation of fuel to maintain their out-of-equilibrium state. As life is a chemical unit capable of catalysing the degradation of fuel and is capable of proliferation, a feedback loop is established between the fuel containing environment and the living organism. (Online version in colour.) [/fig]
[fig] Figure 4: Self-assembly pulls systems out-of-equilibrium. (a) Autocatalysis. Building blocks [/fig]
|
Prejudice and Feeling of Threat towards Syrian Refugees: The Moderating Effects of Precarious Employment and Perceived Low Outgroup Morality
Refugees frequently experience traumatic situations that result in the deterioration of their psychological well-being. In addition, perceived prejudice and discrimination against them by the host society can worsen their mental health. In this research study, using a Spanish sample, prejudice towards Syrian refugees is analyzed taking into account feeling of threat (realistic or symbolic), precarious employment, and perceived outgroup morality. Using a total of 365 participants, our results reveal that individuals feel more prejudice towards refugees when the former scored higher in realistic threat and symbolic threat, were in a highly precarious situation of employment and perceived refugees as being more immoral. Furthermore, it was found that persons who scored high in realistic threat and at the same time were in a situation of precarious employment, were those who displayed greater prejudice. The results likewise pointed to individuals who scored high in symbolic threat and in outgroup morality as being those who felt greater rejection towards the refugees. Accordingly, our results confirm the importance of feeling of threat in relation to prejudice, and highlight two important moderating factors: precarious employment and perceived outgroup morality.
# Introduction
According to the United Nations High Commissioner for Refugees (UNHCR), in recent years we have been immersed in a humanitarian crisis which has led to the displacement of millions of people due to wars or violence in their countries of origin. Specifically, it is estimated that there are 70.8 million forcibly displaced people in all of the world, with 25.9 million of them being refugees. This phenomenon called "the refugee crisis", is characterized by the growing number of forced displacements toward Europe and other Western countries [bib_ref] Introduction to the thematic issue "Refugee crisis: The borders of human mobility, Duarte [/bib_ref]. The main countries of origin are Syria, Afghanistan, South Sudan and Myanmar. In fact, the case of Syria represents the largest exodus that has taken place since World War II.
The circumstances causing the refugees to flee their countries make this displacement different from other types of voluntary migration, as it is frequently characterized by forced departure, fear and traumatic experiences. This feature often leads to refugees experiencing, once they are settled, mental health problems and a significant deterioration in their mental health [bib_ref] Morbidity of asylum seekers in a medium-sized German city, Führer [/bib_ref] and wellbeing [bib_ref] The Relationship between Trauma, Post-Migration Problems and the Psychological Well-Being of Refugees..., Carswell [/bib_ref]. In particular, refugees have a high prevalence of mental disorders, specifically post-traumatic stress disorder and depression [bib_ref] Prevalence of post-traumatic stress disorder, depression and somatisation in recently arrived refugees..., Nesterko [/bib_ref] [bib_ref] Depression and anxiety in labor migrants and refugees-A systematic review and meta-analysis, Lindert [/bib_ref]. Research carried out by Chung et al. [bib_ref] The impact of trauma exposure characteristics on post-traumatic stress disorder and psychiatric..., Chung [/bib_ref] on Syrian refugees shows that 43% of them have post-traumatic stress disorder and that they have higher levels of psychological stress when they have witnessed situations of horror, death, etc. In addition, the study highlights differences in the mental health of refugees depending on the host country. On the other hand, the degree of acceptance and integration of the host country is also a key element for the mental health of this sector of the population, since there is a negative relationship between perceived prejudice and the psychological wellbeing of potentially stigmatized minority groups [bib_ref] La satisfacción vital: Su relación con el prejuicio, la identidad nacional, la..., Murillo [/bib_ref]. In this sense, the meta-analysis undertaken by Pascoe and Smart Richman [bib_ref] Perceived discrimination and health: A meta-analytic review, Pascoe [/bib_ref] , which takes into account the results of 134 studies carried out in diverse countries, revealed that stress levels increase when there is perceived discrimination. Until now there have been few studies that analyze which factors contribute to prejudice and discrimination of the host country towards refugees.
At the same time, the impact of perceived threat of ethnic prejudice is well documented [bib_ref] Intergroup threat and outgroup attitudes: A meta-analytic review, Riek [/bib_ref]. Nevertheless, until now there have been no studies that have taken an analysis of outgroup threat in relation to prejudice towards refugees into account in the Spanish context, with this being an essential variable for the understanding of intergroup relations, formation of negative stereotypes towards outgroups [bib_ref] Percepción de amenaza exogrupal, contacto intergrupal y prejuicio afectivo hacia colectivos migrantes..., Marcos [/bib_ref] [bib_ref] The effects of realistic threat and group identification on social dominance orientation, Morrison [/bib_ref] [bib_ref] The effects of feeling threatened on attitudes toward immigrants, Stephan [/bib_ref] and as such, for mental health and psychological wellbeing.
## The importance of outgroup threat in understanding prejudice
Cottrell and Neuberg [bib_ref] Different emotional reactions to different groups: A socio-functional threat-based approach to "prejudice, Cottrell [/bib_ref] hold that the specific emotions that we feel toward other groups arise from the perception that such groups threaten certain aspects that are important to us, such as, for example, economic resources or values. Accordingly, when individuals perceive that members of other groups put into question or threaten elements they value, they experience outgroup hostility. In recent years, the Intergroup Threat Theory (ITT) [bib_ref] Intergroup threat theory, Stephan [/bib_ref] has become an important theoretical framework for understanding the key role of threatening elements in the genesis of ethnic prejudice, taking into account one of the most relevant approaches in this field [bib_ref] Exploring the Differential Effects of Perceived Threat on Attitudes Toward Ethnic Minority..., Jedinger [/bib_ref]. The ITT distinguishes between two basic sources for threat: realistic threat and symbolic threat. Realistic threat implies the perception of competition between ingroup and outgroup for scarce resources, such as employment, social services, education and healthcare [bib_ref] Amenaza percibida como predictor de prejuicios y estereotipos: Reacciones de los americanos..., Ybarra [/bib_ref]. At the same time, symbolic threat refers to the perception of differences in values and beliefs that members of outgroups are thought to have. In this sense, symbolic threat is focused on worldview and moral values, along with fear of losing ingroup customs, language and traditions as a consequence of interaction with members of the outgroup [bib_ref] Balances teóricos de la amenaza como variable para la investigación social, Rodríguez [/bib_ref].
The ITT, and its previous versions [bib_ref] The role of threat in intergroup relations, Stephan [/bib_ref] , have demonstrated that perceived threat, both symbolic and realistic, plays a central role in intergroup attitudes, and it is a predictive variable for prejudice in various social contexts [bib_ref] Intergroup threat theory, Stephan [/bib_ref]. This relation between intergroup threat and greater resentment and hostility towards immigrants has been reflected in different studies, for recently arrived immigrants [bib_ref] Integrated Threat Theory and Acceptance of Immigrant Assimilation: An Analysis of Muslim..., Croucher [/bib_ref] [bib_ref] Prejudice toward immigrants, Stephan [/bib_ref] as well as for those who are settled residents [bib_ref] Prejudice towards Muslims in the Netherlands: Testing integrated threat theory, González [/bib_ref]. However, different research studies have brought to light that the degree to which each type of threat is related to prejudice depends to a large extent on the nature of the relation between the groups being considered [bib_ref] Prejudice toward immigrants to Spain and Israel: An integrated Threat Theory analysis, Stephan [/bib_ref]. Stated in another way, different minority groups can provoke different perceptions of threat [bib_ref] Intergroup threat theory, Stephan [/bib_ref] [bib_ref] Different groups, different threats: Public attitudes towards immigrants, Hellwig [/bib_ref] [bib_ref] The differential impact of threats on ethnic prejudice toward three minority groups..., Rooij [/bib_ref]. Due to the fact that the type and importance of the threat induced depend on the specific characteristics of the intergroup context, different outgroups can evoke differential attitudinal reactions among members of the majority group [bib_ref] Walloons as general or specific others? A comparison of anti-Walloon and anti-immigrant..., Meuleman [/bib_ref]. Along these lines, one of the most recent studies is that of Jedinger and Eisentraut [bib_ref] Exploring the Differential Effects of Perceived Threat on Attitudes Toward Ethnic Minority..., Jedinger [/bib_ref]. These authors hypothesize an effect of different types of threat on ethnic prejudices based on perceived characteristics in minority groups. The results of this study, undertaken within the German context, show that negative attitudes towards Muslims and third-generation Turks arise mainly from a perception of economic threat and cultural threat, while the refugees and the Romani are also linked to threat involving public safety and criminality. In the Spanish context, the relation between hostility towards immigrants and perceived realistic and symbolic threat have been dealt with in different studies [bib_ref] Escala de amenaza grupal internalizada por inmigrantes latinoamericanos en España: Evidencias iniciales..., Berríos-Riquelme [/bib_ref] [bib_ref] Los inmigrantes vistos por los españoles, Checa [/bib_ref] [bib_ref] Escala de Percepción de Amenaza Exogrupal (EPAE): Fiabilidad y evidencias de validez, Navas [/bib_ref] [bib_ref] The perception of realistic and symbolic threat and its influence on prejudice,..., Wlodarczyk [/bib_ref] ; nevertheless, until now, this relationship has not been studied in the refugee population.
## Stereotype content and the dimension of outgroup morality
Although in colloquial terms concepts such as prejudice, discrimination, racism, and stereotypes are used interchangeably, we should bear in mind that they are not synonymous concepts. Traditionally, prejudice has been conceived as a negative attitude towards a certain social group, but nowadays the role of emotions has become especially relevant. From the theory of the inter-group emotions [bib_ref] Social identity and social emotions: Towards new conceptualizations of prejudice, Smith [/bib_ref] , prejudice is considered as an emotion that depends on the social, political, and cultural context in which intergroup relations take place [bib_ref] Was it the problem? Prejudice as an actitud-in-context, Eagly [/bib_ref]. On the other hand, social stereotypes are beliefs about characteristics that a group of people have stemming from simply belonging to a social group [bib_ref] Conceptual approaches to stereotypes and stereotyìng, Ashmore [/bib_ref]. The question as to why we associate different attributes to certain social groups has been a recurrent one in the study of stereotypes. In general, it has been confirmed that the formation of stereotypes is a complex phenomenon that can be explained by cognitive, affective, socio-motivational, and cultural processes [bib_ref] Social psychological foundations of stereotype formation, Mackie [/bib_ref].
According to the Stereotype Content Model [bib_ref] A model of (often mixed) stereotype content: Competence and warmth respectively follow..., Fiske [/bib_ref] [bib_ref] Dis)respecting versus (dis)liking: Status and interdependence predict ambivalent stereotypes of competence and..., Fiske [/bib_ref] , judgement about others derives from structural relationships that the groups maintain with one another. These relationships are determined by whether these groups compete for resources or if they are high or low status. For the model's first approaches [bib_ref] A model of (often mixed) stereotype content: Competence and warmth respectively follow..., Fiske [/bib_ref] [bib_ref] Dis)respecting versus (dis)liking: Status and interdependence predict ambivalent stereotypes of competence and..., Fiske [/bib_ref] , the social perception towards the outgroup is developed on the basis of two dimensions: warmth and competence. Later, Leach, Ellemers and Barreto [bib_ref] Group virtue: The importance of morality (vs. Competence and sociability) in the..., Leach [/bib_ref] established that the dimension of warmth in fact encompasses two different dimensions: sociability and morality. Sociability is associated with the desire to interact with others (for example, being friendly or nice). At the same time, morality refers to the degree to which the behavior of the person or group evaluated is considered proper (for example, a trustworthy or sincere person). The existence of these two differential dimensions and their impact on prejudice has been confirmed in diverse studies [bib_ref] Looking for honesty: The primary role of morality (vs. Sociability and competence)..., Brambilla [/bib_ref] [bib_ref] Aplicación extendida del Modelo de Contenido de los Estereotipos (MCE) hacia tres..., López-Ródriguez [/bib_ref] [bib_ref] Initial impressions determine behaviours: Morality predicts the willingness to help newcomers, Pagliaro [/bib_ref] [bib_ref] La perspectiva de la minoría: Estereotipos y emociones entre grupos inmigrantes, Cuadrado [/bib_ref]. Furthermore, it seems that the dimension of morality is the one with most weight when it comes to determining our evaluation of certain groups [bib_ref] When is bad to be friendly and Smart: The desirability of sociability..., Landy [/bib_ref] [bib_ref] Groups and morality, Leach [/bib_ref].
One of the most recent studies on refugees in the Spanish context corresponds to that by Ordóñez-Carrasco, Blanc, Navas and Rojas-Tejada [bib_ref] Acculturation preferences of Spaniards towards Middle Eastern and Syrian refugees and their..., Ordóñez-Carrasco [/bib_ref]. In this research, the dimension of outgroup morality is linked [bib_ref] Looking for honesty: The primary role of morality (vs. Sociability and competence)..., Brambilla [/bib_ref] to the acculturated preferences towards Syrian refugees. According to these authors, when the host population perceives a high degree of outgroup morality, and as such, the outgroup is no longer a threat, the former are seen as more flexible and permissive with regard to the outgroup maintaining its culture of origin in the private sphere. Our study, thus, points out the importance of perceived morality in prejudice and in the strategies of acculturation, not only with respect to immigrants, but also individuals who are refugees.
## The present study
In this study, we seek to determine the role of outgroup threat in its two dimensions, realistic and symbolic, in the ethnic prejudice of the Spanish population towards Syrian refugees. In Europe, two types of opposing discourse converge with respect to refugees [bib_ref] Threat and suffering: The liminal space of 'The Jungle, Howarth [/bib_ref]. On one hand, there is the discourse where human rights take precedence and there is a call for countries to offer protection to people coming from other ones that are in conflict; on the other hand, there is other discourse that highlights the risk and threat represented by mass acceptance of people who have been forced to emigrate [bib_ref] Acculturation preferences of Spaniards towards Middle Eastern and Syrian refugees and their..., Ordóñez-Carrasco [/bib_ref]. In recent years, in Spain, the number of people requesting asylum has grown exponentially, going from 5947 in 2014 to 118,264 in 2019. Specifically, according to Spanish Commission for Refugees (CEAR), the number of Syrian refugees in Spain is estimated to be between 12,000 and 15,000 people. The latest data on Syrian asylum seekers in Spain, corresponding to 2019, indicated a number of 2775. This increase within such a short time is important in the host population's perception of refugees, because among the main variables involved in the activation of threat and negative attitudes towards outgroups are the size of the foreign population and competition for resources brought about by the country's economic context [bib_ref] Los inmigrantes vistos por los españoles, Checa [/bib_ref] [bib_ref] Immigration as a threat: Explaining the changing pattern of xenophobia in Spain, Cea D'ancona [/bib_ref]. In line with the ITT [bib_ref] Intergroup threat theory, Stephan [/bib_ref] , in our study we expected the participants to show greater prejudice when they felt a higher degree of realistic threat (H1a) and symbolic threat (H1b) toward Syrian refugees.
At the same time, the refugee crisis is framed within the context of the economic and financial crisis in Europe, which, together with the austerity measures, have brought about a lesser degree of openness from the European population towards persons coming from other countries [bib_ref] La crisis de valores: La propia Unión Europea en tela de juicio, Onghena [/bib_ref]. During periods of crises, the perception that minority groups pose a threat for scarce resources and for the national cultural homogeneity becomes exacerbated [bib_ref] Immigration as a threat: Explaining the changing pattern of xenophobia in Spain, Cea D'ancona [/bib_ref]. Specifically in Spain, the economic crisis of 2008 was severe and intense and intergroup relations were sharply impacted by this social reality. This socioeconomic context diminished the quality and level of life of Spain's citizens (Spaniards and immigrants), who have had to compete for jobs and social resources in a society in crisis for a decade [bib_ref] Escala de amenaza grupal internalizada por inmigrantes latinoamericanos en España: Evidencias iniciales..., Berríos-Riquelme [/bib_ref]. In this sense, we expect to obtain that those persons that were in a situation of precarious employment, defined in this study as individuals who are unemployed or employed in a job that they viewed as having very negative conditions, would show greater levels of prejudice towards refugees (H2). In addition, as realistic threat implies perceived competition for resources, among which is employment, we expected that the situation of precarious employment would be a moderating factor for the same. Accordingly, we hypothesized that individuals would feel greater hostility towards Syrian individuals when they scored high in realistic threat and were in situation of precarious employment (H3).
Finally, the dimension outgroup morality was taken into account in the study, which, as stated earlier, is considered the most important for the stereotype content and one of the most relevant factors in outgroup hostility [bib_ref] Looking for honesty: The primary role of morality (vs. Sociability and competence)..., Brambilla [/bib_ref] [bib_ref] On the importance of being moral: The distinctive role of morality in..., Brambilla [/bib_ref]. In line with results from other studies on ethnic prejudice towards minority groups in the Spanish context [bib_ref] Aplicación extendida del Modelo de Contenido de los Estereotipos (MCE) hacia tres..., López-Ródriguez [/bib_ref] [bib_ref] Acculturation preferences of Spaniards towards Middle Eastern and Syrian refugees and their..., Ordóñez-Carrasco [/bib_ref] [bib_ref] The role of values in attitudes towards violence: Discrimination against Morrocans and..., Álvaro [/bib_ref] , it was expected that when individuals perceived low morality in the Syrian refugees, they would show greater prejudice towards them (H4). Furthermore, we hypothesized that perceived morality would be a moderating effect for symbolic threat. Thus, it was expected that when individuals perceived that their main traditions and values were endangered [bib_ref] Balances teóricos de la amenaza como variable para la investigación social, Rodríguez [/bib_ref] by the arrival of Syrian refugees, that is-they felt threatened by them in the symbolic dimension, and at the same time, felt that they were immoral [bib_ref] Looking for honesty: The primary role of morality (vs. Sociability and competence)..., Brambilla [/bib_ref] , harmful and dishonest persons (H5)-they would show greater outgroup hostility.
# Materials and methods
## Participants
A total of 472 individuals participated in this study, from which those who were not in a position to be employed (such as students, retirees, people with disabilities, etc.) were discarded, resulting in a final total of 365. Among the participants, 51.5% were women and 48.5% men. The average age was 40.35 (SD = 11.61), in ages ranging between 19 and 67. As for education level, the distribution was the following: 21.9% basic education, 28.1% secondary education 28.8% university studies, and 21.3% vocational/professional training. Regarding employment status, 17.6% were unemployed and 28.5% were employed and viewed their job in a negative light, while 53.9% had a job which they evaluated positively and as being in line with their expectations.
## Instruments
Sociodemographic questionnaire. The participants were asked about sociodemographic characteristics such as their sex, age, educational level, employment status, and political viewpoint.
Outgroup Threat Perception Scale. To measure outgroup threat, the Escala de Percepcion de Amenaza Exogrupal-EPAE (Outgroup Threat Perception Scale) [bib_ref] Escala de Percepción de Amenaza Exogrupal (EPAE): Fiabilidad y evidencias de validez, Navas [/bib_ref] was used. It is composed of a total of 13 items which range from response 1 ("Not at all") to 5 ("Very much so") that assess two factors: realistic threat (9 items) and symbolic threat (4 items). The items to measure symbolic threat refer to the degree to which the individuals feel that refuges endanger educational and family values, religious beliefs and cultural traditions. At the same time, the items that assess realistic threat indicate the extent to which individuals feel that the refugees put at risk access to jobs, healthcare, education, welfare benefits, the economic stability of the country, health, public order, and personal and national security. The Cronbach's alpha obtained was 0.87 for symbolic threat and 0.89 for realistic threat.
Stereotyped dimension of outgroup morality. A scale was designed based on the research of Leach, Ellemers and Barreto [bib_ref] Group virtue: The importance of morality (vs. Competence and sociability) in the..., Leach [/bib_ref] in the Spanish version [bib_ref] Aplicación extendida del Modelo de Contenido de los Estereotipos (MCE) hacia tres..., López-Ródriguez [/bib_ref]. Participants were asked to what degree they considered the refugees honest, trustworthy, sincere, respectful, fair and well-intentioned for a range of responses that went from 1 ("Not at all") a 5 ("Very much so"). The reliability coefficient was 89.
Prejudice. A scale of emotions was used, composed of 10 items with responses ranging from 1 ("Not at all") to 5 ("Very much so"), which measured the affective component of the prejudicial attitude towards refugees through negative emotions (fear, unease, anger, disgust, hate, deception, disdain, frustration, resentment, and agitation). The Cronbach's alpha obtained was 0.92.
## Procedure
The data were collected in Malaga (Spain). The questionnaire was administered by doctoral students from the Faculty of Psychology and Speech Therapy at the University of Malaga, who had been previously trained in social and community research. The sample was random using the snow-ball sampling method, asking for voluntary collaboration from participants. The questionnaire was anonymous and data confidentiality was assured as was the fact that results would only be used for academic purposes. In addition, participants read a debriefing explaining the goals of the study and they also were able to request an additional oral debriefing. The ethical guidelines of this research were approved by the Research Ethics Committee of the University of Málaga (85-2019-M). The data were gathered between November 2019 and February 2020.
# Statistical analysis
Once the data screening was performed, the SPSS v23.0 statistical software (SPSS Inc., Chicago, IL, USA) was used to calculate the reliability, descriptive statistics, and correlations of the measured variables. First, the reliability was calculated for each of the scales used for the variables realistic threat, symbolic threat, outgroup morality and prejudice. With respect to employment status, participants were classified into two groups: (1) highly precarious employment: this includes unemployed individuals or individuals who perceive their job as having very negative work conditions and (2) low degree of precariousness: composed of individuals who have a job that they view as being in accordance with their expectations and with good working conditions. Later, the descriptive statistics were measured (mean and standard deviation) and the correlation among variables. Lastly, a hierarchical regression analysis was carried out through the PROCESS macro for SPSS.
# Results
First, the descriptive statistics and the correlations between variables were calculated. The results show that the participants obtained a medium-low score in realistic threat (M = 2.22; SD = 1.10) and symbolic threat (M = 2.11; SD = 1.07), medium-high in outgroup morality (M = 3.17; SD = 0.89) and low in prejudice towards the refugees (M = 1.49; SD = 0.91). As for the relationship between the variables, it was observed that for higher levels of realistic threat, there were higher levels of symbolic threat (r = 0.597; p < 0.01), higher levels of prejudice towards refugees (r = 0.381; p < 0.01), and lower levels of perceived morality (r = −0.450; p < 0.01). At the same time, at higher levels of symbolic threat there were greater level of prejudice towards refugees (r = 0.401; p < 0.01) and lower levels of perceived morality (r = −0.499; p < 0.01). It was also found that with higher levels of outgroup morality there was lower prejudice (r = −0.422; p < 0.01).
Then, a hierarchical regression analysis was carried out [fig_ref] Table 1: Multiple regression analysis on prejudice [/fig_ref] , in which the criterion variable prejudice towards refugees (R 2 c = 0.399; F (4, 365) = 21.938, p = 0.001). In the first step, age and gender included as control variables, not being statistically significant (gender: β = 0.017, t = 0.068, p = 0.946; age: β = 0.190, t = 0.792, p = 0.440). In the second step, four predictor variables were included: realistic threat, symbolic threat, outgroup morality and precarious employment. Finally, in the third step, the interactions between realistic threat and precarious employment and symbolic threat and outgroup morality were included. All of these variables had a significant effect on greater prejudice towards refugees. Individuals showed more prejudice when they felt threatened by the refugees in the realistic dimension (β = 0.447, t = 2.07, p = 0.020) and the symbolic one (β = 0.759, t = 3.53, p = 0.001), perceived low morality in the refugees (β = −0.276, t = −3.70, p = 0.000) and were in highly precarious employment (β = 0.146, t = 2.33, p = 0.021). It was likewise found that the interaction between realistic threat and precarious employment was significant (β = 0.561, t = 2.65, p = 0.009). To understand the interaction between variables we follow the process suggested by Aiken and West. The test of simple slopes revealed that the realistic threat of participants predicted prejudice when they were in highly precarious employment ((+1 SD) (β = 0.439, t = 5.55, p = 0.000) but not when precarious employment was low (−1 SD) (β = 0.137, t = 2.09, p = 0.057). That is, individuals showed more prejudice when they perceived a higher threat regarding resources with respect to refugees and were in a situation of precarious employment. [fig_ref] Figure 1: Interaction between Realistic Threat and Precariousness on Prejudice [/fig_ref]. At the same time, interaction between symbolic threat and outgroup morality was significant (β = −0.300, t = −4.40, p = 0.000). The test of simple slopes revealed that symbolic threat in the participants predicted prejudice when they showed low levels of outgroup morality (−1 SD) (β = 0.260, t = 3.91, p = 0.000), but not with high levels (+1 SD) (β = 0.012, t = 1.40, p = 0.888). That is, when individuals perceived a risk to their values and customs from the Syrian refugees and thought that the latter were not honest or trustworthy, more prejudice was felt [fig_ref] Figure 2: Interaction between Symbolic Threat and Morality on Prejudice [/fig_ref] Individuals showed more prejudice when they felt threatened by the refugees in the realistic dimension (β = 0.447, t = 2.07, p = 0.020) and the symbolic one (β = 0.759, t = 3.53, p = 0.001), perceived low morality in the refugees (β = −0.276, t = −3.70, p = 0.000) and were in highly precarious employment (β = 0.146, t = 2.33, p = 0.021). It was likewise found that the interaction between realistic threat and precarious employment was significant (β = −0.561, t = −2.65, p = 0.009). To understand the interaction between variables we follow the process suggested by Aiken and West. The test of simple slopes revealed that the realistic threat of participants predicted prejudice when they were in highly precarious employment ((+1 SD) (β = 0.439, t = 5.55, p = 0.000) but not when precarious employment was low (−1 SD) (β = 0.137, t = 2.09, p = 0.057). That is, individuals showed more prejudice when they perceived a higher threat regarding resources with respect to refugees and were in a situation of precarious employment. [fig_ref] Figure 1: Interaction between Realistic Threat and Precariousness on Prejudice [/fig_ref]. At the same time, interaction between symbolic threat and outgroup morality was significant (β = −0.300, t = −4.40, p = 0.000). The test of simple slopes revealed that symbolic threat in the participants predicted prejudice when they showed low levels of outgroup morality (−1 SD) (β = 260, t = 3.91, p = 0.000), but not with high levels (+1 SD) (β = 0.012, t = 1.40, p = 0.888). That is, when individuals perceived a risk to their values and customs from the Syrian refugees and thought that the latter were not honest or trustworthy, more prejudice was felt [fig_ref] Figure 2: Interaction between Symbolic Threat and Morality on Prejudice [/fig_ref].
# Discussion
In this study we have sought to study the relationship of the feeling of threat, level of employment precariousness, and perception of outgroup morality with prejudice towards Syrian refugees in the Spanish context. The majority of refugees experience diverse emotional problems and a deterioration in their psychological wellbeing as a consequence of the traumatic experiences they live through [bib_ref] Bouncing forward of young refugees: A perspective on resilience research directions, Sleijpen [/bib_ref]. The response of the host society during their arrival and the perception of being accepted can be key to their integration and recovery [bib_ref] Crisis de refugiados: Un análisis de las concepciones presentes en los estudios..., Quercetti [/bib_ref] and, in contrast, perceived discrimination can generate disorders such as depression and anxiety [bib_ref] Perceived discrimination, social support net and psychological well-being among three immigrant groups, Jasinskaja-Lahti [/bib_ref] [bib_ref] The meaning and consequences of perceived discrimination in disadvantaged and privileged social..., Schmitt [/bib_ref].
In keeping with the ITT [bib_ref] Intergroup threat theory, Stephan [/bib_ref] , outgroup threat plays an important role in understanding outgroup hostility towards the collective studied. In this sense, our results are in line with what was posited in H1: the host population shows more prejudice when it feels threatened by the Syrian individuals, due to competition for resources (H1a) and because the refugees could put their values, traditions and worldview at risk. (H1b). Different studies along these lines have corroborated the relationship between hostility towards minority groups and realistic threat [bib_ref] Intergroup threat and outgroup attitudes: A meta-analytic review, Riek [/bib_ref] [bib_ref] Percepción de amenaza exogrupal, contacto intergrupal y prejuicio afectivo hacia colectivos migrantes..., Marcos [/bib_ref] [bib_ref] Intergroup threat theory, Stephan [/bib_ref] [bib_ref] From prejudice to discrimination: The legitimizing role of perceived threat in discrimination..., Pereira [/bib_ref] as well as symbolic threat [bib_ref] Prejudice towards Muslims in the Netherlands: Testing integrated threat theory, González [/bib_ref] [bib_ref] Predictive factor of immigration attitudes, Ramos De Oliveira [/bib_ref].
On the other hand, our H2 postulated that individuals who were in a situation of highly precarious employment would experience more negative emotions towards Syrian refugees. Indeed, our results are in line with what we expected, with individuals in highly insecure employment being those who showed greater outgroup hostility. From the classic theory of realistic threat for analyzing prejudice [bib_ref] Group Conflict and Cooperation: Their Social Psychology, Sherif [/bib_ref] , it was established that when groups compete for resources, more intense intergroup rejection behavior and emotions arise, leading to increased prejudice and discrimination. In addition, according to our data, the situation of precarious employment is a moderating factor in the feeling of realistic threat, as was formulated in H3. When individuals scored high in this dimension of threat and are in a situation of precarious employment more prejudice towards Syrian refugees was displayed. This relationship is worthy of note, given the current healthcare crisis, which is also provoking a social and economic crisis, considerably increasing levels of unemployment. Specifically, in Spain following the outbreak of the COVID-19 pandemic, the current unemployment rate (May 2020) is at 14.4%. In situations of crisis, the perception of threat increases [bib_ref] Immigration as a threat: Explaining the changing pattern of xenophobia in Spain, Cea D'ancona [/bib_ref] and minority groups (for example, immigrants and refugees) can become scapegoats for taking out frustrationand consequently prejudice is increased. Therefore, fostering policies of employability and improving
# Discussion
In this study we have sought to study the relationship of the feeling of threat, level of employment precariousness, and perception of outgroup morality with prejudice towards Syrian refugees in the Spanish context. The majority of refugees experience diverse emotional problems and a deterioration in their psychological wellbeing as a consequence of the traumatic experiences they live through [bib_ref] Bouncing forward of young refugees: A perspective on resilience research directions, Sleijpen [/bib_ref]. The response of the host society during their arrival and the perception of being accepted can be key to their integration and recovery [bib_ref] Crisis de refugiados: Un análisis de las concepciones presentes en los estudios..., Quercetti [/bib_ref] and, in contrast, perceived discrimination can generate disorders such as depression and anxiety [bib_ref] Perceived discrimination, social support net and psychological well-being among three immigrant groups, Jasinskaja-Lahti [/bib_ref] [bib_ref] The meaning and consequences of perceived discrimination in disadvantaged and privileged social..., Schmitt [/bib_ref].
In keeping with the ITT [bib_ref] Intergroup threat theory, Stephan [/bib_ref] , outgroup threat plays an important role in understanding outgroup hostility towards the collective studied. In this sense, our results are in line with what was posited in H1: the host population shows more prejudice when it feels threatened by the Syrian individuals, due to competition for resources (H1a) and because the refugees could put their values, traditions and worldview at risk. (H1b). Different studies along these lines have corroborated the relationship between hostility towards minority groups and realistic threat [bib_ref] Intergroup threat and outgroup attitudes: A meta-analytic review, Riek [/bib_ref] [bib_ref] Percepción de amenaza exogrupal, contacto intergrupal y prejuicio afectivo hacia colectivos migrantes..., Marcos [/bib_ref] [bib_ref] Intergroup threat theory, Stephan [/bib_ref] [bib_ref] From prejudice to discrimination: The legitimizing role of perceived threat in discrimination..., Pereira [/bib_ref] as well as symbolic threat [bib_ref] Prejudice towards Muslims in the Netherlands: Testing integrated threat theory, González [/bib_ref] [bib_ref] Predictive factor of immigration attitudes, Ramos De Oliveira [/bib_ref].
On the other hand, our H2 postulated that individuals who were in a situation of highly precarious employment would experience more negative emotions towards Syrian refugees. Indeed, our results are in line with what we expected, with individuals in highly insecure employment being those who showed greater outgroup hostility. From the classic theory of realistic threat for analyzing prejudice [bib_ref] Group Conflict and Cooperation: Their Social Psychology, Sherif [/bib_ref] , it was established that when groups compete for resources, more intense intergroup rejection behavior and emotions arise, leading to increased prejudice and discrimination. In addition, according to our data, the situation of precarious employment is a moderating factor in the feeling of realistic threat, as was formulated in H3. When individuals scored high in this dimension of threat and are in a situation of precarious employment more prejudice towards Syrian refugees was displayed. This relationship is worthy of note, given the current healthcare crisis, which is also provoking a social and economic crisis, considerably increasing levels of unemployment. Specifically, in Spain following the outbreak of the COVID-19 pandemic, the current unemployment rate (May 2020) is at 14.4%. In situations of crisis, the perception of threat increases [bib_ref] Immigration as a threat: Explaining the changing pattern of xenophobia in Spain, Cea D'ancona [/bib_ref] and minority groups (for example, immigrants and refugees) can become scapegoats for taking out frustrationand consequently prejudice is increased. Therefore, fostering policies of employability and improving work conditions not only foment a more socially cohesive society, but also one where potential intergroup conflicts can be prevented.
Lastly, in this study we have taken into account the stereotyped dimension of morality, so that when individuals perceive the Syrian refugees as not being honest or trustworthy, they showed more prejudice, in accordance with H4. Different studies [bib_ref] When is bad to be friendly and Smart: The desirability of sociability..., Landy [/bib_ref] [bib_ref] Groups and morality, Leach [/bib_ref] have determined that the dimension of morality is the one carrying most weight when determining valuations of certain groups and is a determining factor in regards to prejudice [bib_ref] Looking for honesty: The primary role of morality (vs. Sociability and competence)..., Brambilla [/bib_ref] [bib_ref] On the importance of being moral: The distinctive role of morality in..., Brambilla [/bib_ref]. Furthermore, in our study we have been able to verify that morality is a moderating factor for symbolic threat, as was posited in H5. When individuals feel that Syrian refugees constitute a threat to their worldview, traditions and ways of understanding life and presuppose that these have unethical or socially harmful values, they display a higher degree of outgroup hostility.
## Limitations and future research
As a main limitation of our study, we point out that in our research the consideration of individuals with a highly precarious employment has been made up of two elements: an objective one, being unemployed, and subjective, one's own assessment of their work conditions as being negative, understanding that the feeling of threat from outgroups can be activated in the lowest socio-economic groups and among those with fewer resources. In future studies, it would be worthwhile verifying if there are differences with respect to the level of prejudice and the perception of threat by individuals who are unemployed contrasted with those who, although are working, consider their employment to be of low quality. At the same time, it should be analyzed, if after the COVID-19 pandemic, with the resulting higher unemployment rate (which is currently about 20%), the negative conception and rejection of refugees has increased. It would also be important to know if there are differences in the level of prejudice according to the level of education and the perceived social class.
Along these lines, it would be of importance to check for possible differences in the stereotypes towards different groups of refugees, for example, towards Venezuelans, who represent the largest group of refugees in Spain. We also consider it relevant to study the possible differences that may exist in psychological well-being between the local population and refugees. Likewise, it would be convenient to analyze the perception of rejection felt by the refugee population since the threat of stereotyping in minority groups can increase anxiety and deteriorate mental health [bib_ref] An integrated process model of stereotype threat effects on performance, Schmader [/bib_ref].
Lastly, we should add that, in future research, it would be of value to consider ideological viewpoints in relation to prejudice. In this vein, the study carried out in different European countries by Wike, Stokes and Simmonspoints out that individuals with a right-wing political ideology are more concerned about the arrival of refugees and show more negative attitudes towards minorities (especially Muslims). At the same time, the ideological variables of right-wing authoritarianismand social dominancehave been widely related to prejudice [bib_ref] Ideology and prejudice: The role of value conflicts, Chambers [/bib_ref] [bib_ref] A dual process motivational model of ideology, politics, and prejudice, Duckitt [/bib_ref] [bib_ref] The march of modern fascism: A comparison of social dominance orientation and..., Duriez [/bib_ref] and outgroup threat [bib_ref] Cornelis, I. Threat and right-wing attitudes: A cross-national approach, Onraet [/bib_ref].
# Conclusions
According to the Geneva Convention, states must share responsibility for taking in those individuals who have to flee their countries of origin because they are in danger. However, European countries, until now, have not developed a true policy of integration for this collective. This inevitably leads to the host population feeling threatened and to the existence of different elements that can increase prejudice. In the shared beliefs held regarding outgroups, cultural factors, socialization processes and political, social and economic circumstances all play a central role [bib_ref] Self-stereotyping in the face of threats to group status and distinctiveness: The..., Spears [/bib_ref]. The degree of understanding and framework in which intergroup relations are established have an effect at the collective level, according to the group they are immersed in, as well at the individual level. In the case of the individuals making up the minority groups, the perception of societal rejection and discrimination could lead to a loss of psychological wellbeing, and as such, deterioration in mental health [bib_ref] Perceived discrimination, social support net and psychological well-being among three immigrant groups, Jasinskaja-Lahti [/bib_ref]. The fact that at this moment, Syria is one of the countries with the highest number of displaced persons as a consequence of the war, and that a significant volume of refugees have arrived in Spain within a short span of time, makes it necessary to attend to this social reality for the sake of constructing a more cohesive society, fomenting equality among fellow citizens and improving the mental health of the refugee collective.
In this study, we have been able to confirm that the feeling of threat towards refugees is moderated by the perception of employment precariousness and outgroup morality-elements that must be taken into consideration for improving the inclusion and integration of this collective and for developing positive intergroup relations. The results of this research show that for future intervention programs aimed at reducing racism and xenophobia, it is important to deconstruct the perception of refugees as a threat, to seek empathy with this group and to highlight their possible contribution to the host society.
## Conflicts of interest:
The authors declare no conflict of interest.
[fig] Figure 1: Interaction between Realistic Threat and Precariousness on Prejudice. [/fig]
[fig] Figure 2: Interaction between Symbolic Threat and Morality on Prejudice. [/fig]
[fig] Author: Contributions: Conceptualization, M.V.-M.; methodology, M.V.-M. and J.M.C.; writing-original draft preparation, M.V.-M.; writing-review and editing, J.M.C.; supervision, J.E.S.M.G. and F.P.N.; funding acquisition, J.M.C. All authors have read and agreed to the published version of the manuscript. Funding: This research has been funded by the University of Malaga research project "The role of stereotypes in intergroup relationships in the multicultural context of the province of Malaga" (18_B3_02). [/fig]
[table] Table 1: Multiple regression analysis on prejudice. [/table]
|
A Review of Testosterone Pellets in the Treatment of Hypogonadism
Currently, the most popular form of testosterone replacement is the topical gels that require daily applications and incur a risk of transfer of testosterone to partners and family. One of the problems with testosterone replacement is the short half-life of testosterone. A long-acting formulation is appealing to patients and physicians. In 1972, fused crystalline testosterone pellets were approved in the USA by the FDA but they were not marketed until 2008. Pharmacokinetics studies were available on a different formulation from which much can be learned and applied to the current formulation, Testopel®. The decay kinetics, pituitary suppression, and effect on other sex steroids are reviewed as well as the short-term complication rates. This review should provide the testosterone pellet implanter a better understanding of the physiology of testosterone pellet supplementation for hypogonadism.
# Introduction
The benefits of testosterone replacement for the treatment of hypogonadism are well documented. As men age, there is a steady decline in testosterone due to an aging pituitarygonadal axis. Current treatment modalities require repeated testosterone injections or topical application of gels. Long-term topical and injection therapy are fraught with poor long-term compliance due to the inconvenience of the application and vacillating serum levels. An ideal therapy would be one that is easy to administer, provides reliable levels, and is affordable.
Long-lasting testosterone (T) pellets were FDA-approved in 1972. At that time, the only other options available were inexpensive generic intramuscular T injections. Testopel® which was approved by the FDA is crystalline T, formulated in 75-mg pellets (3×8 mm) with a surface area of 98 mm 2 . The pellets are surgically placed in the subcutaneous space and gradually dissolve. Their long-lasting effect is presumed to be due to the gradual dissolution of the pellets in the relatively hypovascular subdermal space [bib_ref] Absorption of hormone implants in man, Bishop [/bib_ref]. The formulation lay "dormant" until 2008 when the patent was purchased by a newly formed company, Slate. The T pellets were named Testopel®. T replacement therapy evolved between 1972 and 2008. With increased therapeutic formulations, incessant direct-to-consumer advertising, and the increasing number of males of the baby boom generation seeking "eternal youth," T replacement therapy had evolved into a multibillion-dollar pharmaceutical industry.
## Gels versus pellets
Gel-based T replacements have since become the preferred method of therapy since their introduction. The noninvasive nature of gels made it an appealing alternative to T injections. Yet, the need for daily applications, erratic absorption, low long-term compliance rates, the risk of T transfer to family members, and the expense of the monthly prescriptions opened the window for a "new" form of T therapy. Testopel® was thus a welcome new addition to the armamentarium as it required a simple procedure three to four times a year and delivered eugonadal T levels for periods from 3 to 6 months. Day-to-day compliance was not an issue and as such, the therapeutic efficacy was more easily assessed. Unfortunately, the prescribing information, a reflection of the regulatory standards of 1972, was lacking in specific information as to the mechanism of dissolution, the recommended dosing, the insertion technique, the pharmacokinetics, the dosing recommendations, and the data on the dosing frequency. Bioavailability studies were deferred with the original FDA application. Information in the package insert was based on observations and extrapolations from data on injectable T propionate, a short-acting T ester that is no longer used. It is unclear whether any pharmacokinetic studies were ever done on the actual Testopel® pellets. There are certainly no published studies prior to 2009 on the currently FDA-approved Testopel® pellets.
## Pellet pharmacokinetics
Much of the data on T pellet pharmacokinetics comes from work from Handelsman in Australia. The only peer-reviewed publications on T pellets are based on a pellet formulation that has never been approved in the USA produced by Organon. Nonetheless, much can be learned about the pharmacokinetics of T pellets by a critical review of those studies. In 1988, Handelsman conducted a randomized crossover comparator study in 15 previously treated hypogonadal men (9 primary and 6 secondary), using the Organon product of T pellets (6× 100 mg). Resulting hormonal levels were the primary endpoints. This formulation is more in line with the currently used Testopel® pellets in terms of pellet surface area and T dose. The comparator arms were injectable T esters (250 mg every 2 weeks) and oral Tundecanoate (not available in the USA). The hormone levels after injectable T were assessed weekly for 1 month whereas in the T pellets levels were measured weekly for 1 month then monthly thereafter until the levels returned to baseline. Levels on the pellets peaked at 781 ng/dl at 3 weeks and returned to baseline (247 ng/dl) by week 20. Extrapolating from the data presented in the paper, a level of 300 ng/dl was reached at around 13 weeks. Serum chemistries and hematologic parameters were unchanged throughout the study. Both the pellet and injection group reported consistent subjective improvement in libido, potency, muscular strength, and general well-being. With equal numbers (six) of pellet and injection subjects, patients opted to stay on their respective therapies. Reflecting the early experience with the pellets, 6 of 15 men (40 %) extruded at least one pellet though no wound infections occurred. Six 100-mg pellets were found to maintain T levels for up to 4 months [bib_ref] Randomized clinical trial of testosterone replacement therapy in hypogonadal men, Conway [/bib_ref].
In 1990, Handelsmen published an open-label crossover pharmacokinetic study in 43 men (22 hypergonadotropic and 21 hypogonadotropic) on the three regimens (6×100 mg, 6× 200 mg, and 3×200 mg), measuring T levels at monthly intervals. The treatments were crossed over at 6 months. The surface areas were 200 and 100 mm 2 for the 200-and 100-mg pellet, respectively. Pellets were inserted into the sub dermal fat in the lower abdominal wall at the level of the umbilicus. Each pellet was inserted into its own tract 5-10 cm from the insertion site. He found that T levels correlated highly with the dose inserted. The levels in men with 1200 mg were nearly twice that of the men with a 600-mg insertion. Rates of absorption of T from the pellets were constant with a "near linear zero-order release of T over months that was not influenced by the size or number of pellets. The estimated halftime of absorption was approximately 2.5 months and the rate of T release was 1.3 mg/day for the 200 mg pellet and 0.65 mg/day for the 100 mg pellet." Despite a constant absorption of T from the pellets, serum T levels peaked at 1 month and decreased monthly thereafter, supportive of first-order decay kinetics. This would suggest that as the pellets dissolve, the surface area decreases and the amount of T delivered decreases. Ninety percent of T is metabolized by hepatic CYP3A4, with approximately 10 and 1 % metabolized by 5 alpha reductase and aromatase inhibitor, respectively. As the K m of the CYP3A4 metabolism is 50,000 nanomoles, the metabolism of T is physiologically impossible to saturate [bib_ref] Human cytochrome P450 3A4-catalyzed testosterone 6 beta-hydroxylation and erythromycin N-demethylation. Competition during..., Wang [/bib_ref]. The kinetics of T decay must therefore be all attributable to the dissolution of the pellets in the subcutaneous space. Based on some intact extruded pellets and assuming that the extruded pellets had comparable dissolution properties to intact in situ pellets, he extrapolated a dissolution rate of 1.5 mg/day and maintenance of eugonadal ranges for 4-5 months. He calculated a therapeutic half-life of 2.5 months. As production rate of endogenous T is 3-9 mg/day [bib_ref] Further study of factors affecting the metabolic clearance rate of testosterone in..., Southren [/bib_ref] , 400-1200 mg of pellets could produce eugonadal ranges for 4-6 months. LH levels, measured only in the primary hypogonadal men only, were markedly and uniformly suppressed for 1-4 months. Their levels inversely mirrored declining T levels. The reproducible effect of the pellets on gonadotropins was the basis for him recommending using them to augment clinical monitoring. SHBG levels were not affected. He reported a 10 % (10/111) extrusion rate. The frequency of pellet extrusion fell dramatically with increased experience from an incidence of 40 % after the first 15 procedures to 5 % in the later 96 implants. Some palpable fibrosis at the insertion site in some men was seen long after the pellets were dissolved. Seventy percent (30/43) expressed a desire to continue with the pellets versus injections [bib_ref] Pharmacokinetics and pharmacodynamics of testosterone pellets in man, Handelsman [/bib_ref].
In 1996, Jockenhovel published a comprehensive pharmacokinetic study in 14 profoundly hypogonadal men (baseline T of 34 ng/dl) using the 6×200 mg formulation. Only three of the men had idiopathic secondary hypogonadism. Measuring T levels nine times in the first 48 h, a burst release of T was seen with a C max of 144 ng/dl. T levels stabilized for 63 days following which first order decay occurred. Estradiol levels peaked at 42 days, at 38 pgm/ml. At 180 days, the mean serum T was 300. Both LH and FSH exponentially decreased after insertion, inversely mirroring Tand estradiol levels. An interesting observation was the volume of distribution increased while T 1/2 decreased with increasing BMI. Their mean BMI was less than 25.kg/m 2 . This observation is not surprising as T in the periphery equilibrates quickly between most organs and the blood [bib_ref] Salivary testosterone in men: further evidence of a direct correlation with free..., Wang [/bib_ref]. Men with larger BMIs have, in general, a larger volume of distribution. When considering pellets in obese patients, they might need significantly more pellets. Despite a 5 % extrusion rate, all but one of the men expressed the desire to continue with the pellets [bib_ref] Pharmacokinetics and pharmacodynamics of subcutaneous testosterone implants in hypogonadal men, Jockenhovel [/bib_ref].
## Complications, insertion technique, and dosing
In a retrospective survey, Handelsman determined that extrusions were increased by early post implantation increased physical activity [bib_ref] An analysis of testosterone implants for androgen replacement therapy, Handelsman [/bib_ref]. Experiencing an incidence of extrusions (5-12 %) and infections (1.4-6.8 %), he set about to determine if complication rates were affected by the technique, site of insertion, washing of the pellets, impregnation of the pellets with antibiotics, and experience of the implanter. Using the anterior abdominal wall insertion technique, he found that none of the aforementioned variables influenced extrusion or infection rates. Whereas extrusions associated with infections usually occur at a median of 4 weeks, extrusions without apparent infections occurred at 9 weeks [bib_ref] A randomised controlled clinical trial of antibiotic impregnation of testosterone pellet implants..., Kelleher [/bib_ref] [bib_ref] Influence of implantation site and track geometry on the extrusion rate and..., Kelleher [/bib_ref].
In 2004, Handelsman reviewed his experience in 136 men with a standard dose of 800 mg (4×200 mg pellets). Pellets were placed under the skin of the lateral abdominal wall or the lateral aspects of the buttocks along the pants line. Reimplantations were based on symptoms alone without any reminders to the patients. His extrusion rate was approximately 10 %. T levels were drawn at the implantation visit (median 299 vs baseline of 144). If extrusions occurred, the men were instructed to keep the pellets and make a note of the day of the extrusion. The pellets were desiccated upon return to the clinic and weighed to determine the amount that was left. No quantitative measurement was made of the remaining T in the pellet. A linear regression curve was generated by the desiccated weight and the day of extrusion, and a pellet dissolution rate of 1.31 mg/day was calculated. Only intact pellets were used and the pellets maintained their cylindrical shape for a median of 98 days. As nadir levels were unpredictably lowered by increased number of extrusions, one might question the strength of his assumptions. Is an extruded pellet biologically equivalent to one in situ? Handelsman suggested that a dose four pellets of 200 mg should last 5.8 months. The questions remain as to the relevance of these observations to the current Testopel® preparation. Are the formulations truly bioequivalent? Does pellet fragmentation occur during insertion? Are the decay curves the same? How does BMI affect the peak serum levels and decay curve? Should symptoms or T levels be used as replacement criteria?
## Studies and clinical experience
The literature on Testopel® implantation was nonexistent until the 2008 introduction of the product in the US market. In 2009, Cavender et al. published a single-site retrospective review of his experience with the 75-mg Testopel® pellets in 80 men (272 insertions) treated for clinical hypogonadism (T<350 ng/dl). The series was uncontrolled and retrospective and with variable follow-up and treatments. It was not intended as a pharmacokinetic study but a report of a clinical experience and represented the first published reported clinical experience with Testopel®. Dosing of the pellets was arbitrarily based on severity of symptoms, body weight, age, and lifetstyle. The insertion technique was a modification of the Handelsmen lateral jackknife position and utilizeda proprietary trocar. Reported infection rates and extrusion rates were considerably lower than Handelsman at 0.3 and 0.3 % respectively, reflecting the different formulation, technique, experience of implanter, patient selection, or a combination of all the factors. Although he demonstrated normalization of T levels, because of the variability of the patients, treatments, and inclusion criteria, little could be concluded from the paper other than Testopel® could be expected to produce results at least similar tothe Organon product and that it appeared to be at least as well tolerated .
Kaminetsky published an industry-supported, FDA-approved, pharmacokinetic study in 30 men with Testopel® . Dosing was based on BMI and baseline T levels and the insertion technique that was published by Cavender. Twenty-eight men received treatment (8 pellets, n=3; 10 pellets, n=14; 12 pellets, n=12). None met the criteria for six pellets (baseline T of <315 ng/dl and a BMI of <18). Peak T levels at 1 month were dose dependent with 100, 100, 86, 75, and 14 % above 315 ng/dl at weeks 1, 4, 12, 20, and 24, respectively. The continuation phase of the study in (22 of 28 men) revealed that 100 and 31.8 had levels >315 ng/dl at week 4 and 16 after treatment. This would suggest that biologic variability of treatment effect might be expected from one treatment to the next for unexplained reasons. Previous studies have not investigated the reproducibility of the levels from one insertion to the next. During phase 1 of the study, erectile function scores increased clinically significantly in the first 12 weeks of treatment though the score returned to baseline at the end of the study. Though clinically significant, there was no placebo arm. As in other studies, pituitary gonadotropins were suppressed as T and estradiol levels increased. Unlike the Handelsman experience, no extrusions or infections were reported.
The findings in the Kaminetsky study though reassuring about the safety and efficacy of the Testopel® was in stark contrast to the FDA-approved 1972 package insert provided with the product, recommending 150-450 mg (two to six pellets) every 3-6 months. None of the studies thus reported results with such low dosing. In an attempt to provide some clarity McCullough et al. published an independent multiinstitutional study on 380 patients with 702 insertions at 6 institutions. This represented the early experience at each institution. Investigators pooled their data on pre-insertion and post-insertion T levels along with the number of implanted pellets. The number of pellets to be inserted and the techniques used were based on the clinical experience of each investigator though all but one started with six pellets. Unlike the Kaminetsky study, pre-treatment level of T or BMI were not used as a criteria for the number of pellets required. With multiple investigators using their own criteria for treatment and follow-up, this represented a more generalizable guide for the new implanter. Though the T levels at 4 weeks were comparable for all pellet levels, patient and investigator disenchantment with subsequent levels lead investigators to insert more than the minimum number of pellets. Six or seven pellets were utilized in only 10 % of the insertions with 10 or more pellets being inserted in 63 %. The insertion of 10 or more pellets resulted in eugonadal levels for a longer period of time. Regardless of the number inserted, all men were hypogonadal at 6 months with most men requiring re-implantation after 4 months, much like the 1988 Handelsman study with the Organon 100-mg pellets. T levels appeared to decay exponentially behaving like first-order decay kinetics. Unlike in the Handelsman study with 200-mg pellets, inserting more pellets did not increase the period of time between insertions consistent with an exponential decay .
In 2012, Pastuszak et al. reported their follow-up of 273 men, many of whom were in the original multi-institutional study. The indications for treatment were clinical hypogonadism. Mean initial T level was 328 (208)ng/dl and BMI 30 (5). Sixty-eight men had been diagnosed with prostate cancer. Based on observed levels determined at various time points, decay curves were calculated. Unlike the Jockenhovel pharmacokinetic study, many of the conclusions drawn were from extrapolated data and not from actual measured levels. This retrospective observational study examined the effect of initial T level, BMI, multiple insertions, and the number of pellets inserted. Like the multi-institutional study, adequate numbers in the 6-7-pellet group were lacking. Ninety-five percent of the men were treated with 10 or more pellets. The authors found that early post-insertion T levels were impacted by the number of pellets inserted. Higher levels were achieved with more pellets. As endogenous T production is virtually shut down with exogenous replacement, the initial T level did not impact subsequent T levels. There is therefore no need to titrate the number of pellets based on the initial T level as was done in the Kaminetsky pharmacokinetic trial. Men with BMIs less than 25 achieved higher levels than those with a BMI greater than 25, and the decay is faster. This supports the concept that volume of distribution affects subsequent hormone levels, as demonstrated by Jockhovel. Regardless of the number of insertions, number of pellets implanted, initial T level, or BMI, most men needed re-implantation between 3 and 4 months after the insertion. No men experienced a significant increase in the hematocrit or hemoglobin and no men with documented prostate cancer experienced progression for their disease. No men developed prostate cancer during the period of observation. Extrusion and infection rate were 1.1 and 0.4 % respectively, suggesting inherent differences in the nature of the pellets .
## Patient experience and satisfaction
There have been several single-center papers addressing treatment satisfaction and compliance with Testopel®. Khera et al. investigated a small series of young men (4) with Klinefelters who would be requiring lifetime T replacement. Compliance with the Testopel® formulation was better than gels or injections [bib_ref] Subcutaneous implantable testosterone pellets overcome noncompliance in adolescents with Klinefelter syndrome, Moskovic [/bib_ref]. Liphultz looked at patient satisfaction rates at Baylor via a survey based on patient recall. The choice of therapy was heavily influenced by physician recommendation with 53, 31, and 17 % choosing injections, gels, and Testopel®, respectively. Overall, approximately 70 % of patients were satisfied with their respective treatments regardless of modality used. Though the pellets were favored because of the ease and convenience of use, injections were favored because of their decreased cost [17- ]. Regional differences in insurance reimbursement can clearly impact patient preferences. Placebo-controlled studies on patient reported outcomes are understandably lacking with Testopel®.
As with type 2 diabetes, barring major lifestyle changes, hypogonadism is a lifetime problem for which men need treatment. Clearly, T pellets offer some advantages with respect to the maintenance of consistent eugonadal levels of T. There are unfortunately no long-term studies of the use of T pellets. As with any surgical procedure, with repeated insertions, the cumulative risk of an extrusion, hematoma, or infection increases. Anecdotally, many implanters have found that with repeated insertions, subcutaneous fibrosis occurs, making the insertions more difficult. In an effort to increase the interval between insertions, McCullough et al. combined anastrozole, an aromatase inhibitor with Testopel®. The theory was that if the metabolism of Testopel® could be slowed down, T levels might remain eugonadal for a longer period of time and the interval between insertions increased. Thirty-eight men with 65 insertions were analyzed. T levels at up at 120 days were comparable between the groups. With the addition of anastrozole, T levels were maintained at eugonadal levels for over 120 days. Mean re-insertion time was increased from 124 (22) to 194 (62)days. Twenty-five percent did not require a reinsertion. The mechanism though had nothing to do with a decrease in T metabolism but was secondary to the inhibition of the suppressive effect of T replacement on pituitary gonadotropins and endogenous production. Men on Testopel® and anastrozole did not demonstrate the secondary increase in their estradiol levels or suppression of their LH and FSH levels. The addition of a generic aromatase inhibitor significantly decreased the number of annual Testopel® insertions .
# Conclusions
The use of T implants as a form of T replacement has been reported since 1938 [bib_ref] Further experiments on the administration of hormones by the subcutaneous implantation of..., Deanesley [/bib_ref]. The formulation approved by the FDA in 1972 was Testopel®. Despite a lack of information provided by the package insert, there is a wealth of information available through peer-reviewed studies on similar preparations and the product itself. Pellets provide sustained eugonadal T levels for 3-6 months. Contemporary studies suggest that the FDA recommended 3-6 pellets are inadequate for most men and that 10 pellets (750 mg) produce the most reliable levels. Post implantation levels are affected by the volume of distribution, i.e., thin men require fewer, whereas morbidly obese men might require more. Extrusion and implantation infection rates at high-volume centers with Testopel® are less than 1 %, and patient acceptance of the procedure is very high. As with all forms of replacement therapy, close monitoring of therapeutic efficacy is important. More long-term studies on clinical efficacy and safety are needed.
## Compliance with ethics guidelines
Conflict of Interest Andrew McCullough reports the receipt of personal fees from Auxillium, Endo, and Repros as well as grants from Lilly and grants and personal fees from Antares.
## Human and animal rights and informed consent
This article does not contain any studies with human or animal subjects performed by any of the authors.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.
Curr Sex Health Rep (2014) 6:265-269 |
A stochastic mixed effects model to assess treatment effects and fluctuations in home‐measured peak expiratory flow and the association with exacerbation risk in asthma
This is an open access article under the terms of the Creat ive Commo ns Attri bution-NonCo mmercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.AbstractHome-based measures of lung function, inflammation, symptoms, and medication use are frequently collected in respiratory clinical trials. However, new statistical approaches are needed to make better use of the information contained in these data-rich variables. In this work, we use data from two phase III asthma clinical trials demonstrating the benefit of benralizumab treatment to develop a novel longitudinal mixed effects model of peak expiratory flow (PEF), a lung function measure easily captured at home using a hand-held device. The model is based on an extension of the mixed effects modeling framework to incorporate stochastic differential equations and allows for quantification of several statistical properties of a patient's PEF data: the longitudinal trend, long-term fluctuations, and day-to-day variability. These properties are compared between treatment groups and related to a patient's exacerbation risk using a repeated time-to-event model. The mixed effects model adequately described the observed data from the two clinical trials, and model parameters were accurately estimated. Benralizumab treatment was shown to improve a patient's average PEF level and reduce longterm fluctuations. Both of these effects were shown to be associated with a lower exacerbation risk. The day-to-day variability was neither significantly affected by treatment nor associated with exacerbation risk. Our work shows the potential of a stochastic model-based analysis of home-based lung function measures to support better estimation and understanding of treatment effects and disease stability. The proposed analysis can serve as a complement to descriptive statistics of home-based measures in the reporting of respiratory clinical trials.Study HighlightsWHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?Home-based measurements of lung function are frequently collected in respiratory clinical trials. Novel statistical methods for analyzing these data-rich measurements are emerging, but more work is needed to better characterize the data and what it can tell us about treatment effects and disease stability.| 213A STOCHASTIC MIXED EFFECTS MODEL OF HOME-MEASURED PEFSUPPORTING INFORMATIONAdditional supporting information may be found in the online version of the article at the publisher's website.How to cite this article: Leander J, Jirstrand M, Eriksson UG, Palmér R. A stochastic mixed effects model to assess treatment effects and fluctuations in home-measured peak expiratory flow and the association with exacerbation risk in asthma. CPT
# Introduction
Respiratory clinical trials commonly include the collection of home-based measurements, allowing for the frequent monitoring of patients. Home-based measurements may include, for example, peak expiratory flow (PEF), forced expiratory volume in 1 s, fractional exhaled nitric oxide (FeNO), and patient reported outcomes, such as symptoms and reliever medication use. Currently, the use of these variables is often limited to the triggering of alerts, for example, that the patient should contact the treating physician, or as exploratory efficacy end points reported with descriptive statistics. However, in the search for new end points to characterize respiratory disease severity and treatment efficacy, new ways of analyzing these data-rich variables are emerging.
One concept currently being evaluated for use in respiratory clinical trials is disease fluctuations, 1 that is, seemingly random deviations in frequently measured disease variables. CompEx 2 is one example of a novel end point where changes in daily diary variables-specifically PEF, symptoms, and reliever use-are used to identify episodes of asthma deteriorations based on prespecified thresholds. Combined with severe exacerbations-a frequently used primary end point in late-phase asthma clinical trials defined as disease worsening requiring systemic corticosteroids or hospital admission-these deteriorations create a composite event end point with a higher event rate compared to exacerbations alone, thus providing an opportunity to detect treatment effects in shorter and/or smaller trials. Similar diary-based end points developed for chronic obstructive pulmonary disease (COPD) are COPDCompEx 3 and EXACT-PRO.In addition to these event end points-which essentially reduce the information in frequently sampled longitudinal data to one or a few timepoints-other approaches to analyze disease fluctuations have been suggested. One such approach is detrended fluctuation analysis (DFA), which assesses the self-similarity (long-range correlation) of time series data. [bib_ref] Long-range anticorrelations and non-Gaussian behavior of the heartbeat, Peng [/bib_ref] [bib_ref] Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time series, Peng [/bib_ref] The self-similarity is quantified by a positive parameter , where = 0.5 indicates a noncorrelated random time series and > 0.5 indicates the presence of long-range correlations. DFA has mainly been applied to daily sampled PEF data, 1,7-10 where a patient's PEF is shown to be significantly associated and predictive of exacerbation risk. PEF also appears to change with type of treatment, [bib_ref] Risk of severe asthma episodes predicted from fluctuation analysis of airway function, Frey [/bib_ref] and results from Donaldson et al. [bib_ref] Detrended fluctuation analysis of peak expiratory flow and exacerbation frequency in COPD, Donaldson [/bib_ref] indicate that fewer patients may be needed to detect a treatment difference in PEF compared with exacerbation frequency.
Another stochastic property of daily sampled PEF data often assessed together with PEF is the coefficient of variation (CV PEF ), giving an overall measure of a patient's lung function variability over a certain time period. High CV PEF has been shown to be associated with loss of asthma control, [bib_ref] Risk of severe asthma episodes predicted from fluctuation analysis of airway function, Frey [/bib_ref] [bib_ref] Variability of lung function predicts loss of asthma control following withdrawal of..., Thamrin [/bib_ref] further highlighting the value of assessing lung function fluctuations in respiratory diseases.
Although PEF and CV PEF seem to hold potential for use in disease management and as clinical trial end points, most analyses evaluating these variables have included a limited number of patients. In addition, the standard DFA comes with some limitations, including the assumed loglog linearity of the measured signal's power spectrum, its sensitivity to missing values, and the need for rather long time series to avoid biased estimates of . [bib_ref] Revisiting detrended fluctuation analysis, Bryce [/bib_ref] In large phase III clinical trials, diary compliance is not always perfect, and many patients can have periods of missing values. Patients may also end the trial early, with limited data available for estimating both PEF and CV PEF .
In this work, we present an alternative way to model and analyze the statistical properties of PEF time series in a clinical trial setting. Specifically, we identified stochastic WHAT QUESTION DID THIS STUDY ADDRESS? Can a stochastic model-based analysis of home-based measurements of peak expiratory flow (PEF) be used to improve the understanding of lung function dynamics, treatment response, and exacerbation risk in asthma?
## What does this study add to our knowledge?
A stochastic model of home-measured PEF was developed, providing a robust way of characterizing statistical properties of the data. Results show that both trends and fluctuations in PEF can be affected by treatment and are associated with exacerbation risk.
## How might this change drug discovery, development, and/or therapeutics?
A stochastic model-based analysis of PEF can support better estimation and understanding of treatment effects, disease stability, and exacerbation risk. Furthermore, it can serve as a complement to descriptive statistics of home-based PEF in the reporting of clinical trials. differential equations mixed effects (SDEME) models as a suitable statistical framework to model PEF data. [bib_ref] Mixed effects in stochastic differential equation models, Ditlevsen [/bib_ref] [bib_ref] Non-linear mixed effects models with stochastic differential equations: Implementation of an estimation..., Overgaard [/bib_ref] [bib_ref] Practical estimation of high dimensional stochastic differential mixed effects models, Picchini [/bib_ref] [bib_ref] A review on estimation of stochastic differential equations for pharmacokinetic/pharmacodynamic models, Donnet [/bib_ref] [bib_ref] Mixed effects modeling using stochastic differential equations: illustrated by pharmacokinetic data of..., Leander [/bib_ref] The framework provides several advantages: (1) stochastic differential equations offer great flexibility when modeling stochastic processes and can be designed to have arbitrary complexity, (2) the use of mixed effects allows sharing information on parameter values across individuals, (3) treatment effects can be easily incorporated and estimated, and (4) missing values are typically not an issue.
We develop our SDEME model based on data from two pivotal phase III clinical trials demonstrating the effects of benralizumab (Fasenra®, AstraZeneca) on top of standard of care in patients with moderate to severe asthma. [bib_ref] Benralizumab, an anti-interleukin-5 receptor α monoclonal antibody, as add-on treatment for patients..., Fitzgerald [/bib_ref] [bib_ref] Efficacy and safety of benralizumab for patients with severe asthma uncontrolled with..., Bleecker [/bib_ref] The estimated model parameters-including a patient's long-term and short-term PEF variability as well as average PEF response-are compared with estimates of PEF and CV PEF and related to asthma exacerbation risk using a repeated time-to-event (RTTE) model. We also investigate treatment differences in these model parameters.
# Methods
## Clinical trial data
Patient-level data from two phase III, randomized, double-blind, parallel-group, placebo-controlled, multicenter clinical trials including patients with moderate to severe asthma served as the basis for model development and analysis. Both trials investigated the efficacy of benralizumab on top of standard of care and were selected based on (1) a long study period and a large sample size; (2) the availability of daily, diary-based PEF measurements; and (3) the presence of a significant treatment effect on lung function and exacerbations. The details of the trials have been presented elsewhere, 17,18 but a summary follows.
Trial A (NCT01914757) enrolled patients aged 12-75 years with severe uncontrolled asthma and at least two exacerbations while on medium-to-high-dosage inhaled corticosteroids (ICS) and long-acting β2-agonists (LABA) in the previous year. In total, 1306 patients were randomly assigned to either placebo, benralizumab 30 mg every 4 weeks, or benralizumab 30 mg every 8 weeks (first three doses every 4 weeks) as an add on to their standard treatment. The treatment period was 56 weeks.
Trial B (NCT01928771) enrolled patients aged 12-75 years with a diagnosis of asthma for at least 1 year and at least two exacerbations while on high-dosage ICS and LABA in the previous year. In total, 1205 were randomly assigned to either placebo, benralizumab 30 mg every 4 weeks, or benralizumab 30 mg every 8 weeks (first three doses every 4 weeks) as an add on to their standard treatment. The treatment period was 48 weeks.
Both trials were conducted in accordance with the Declaration of Helsinki, the International Conference on Harmonization Guidelines for Good Clinical Practice, and applicable regulatory requirements. Before the analysis, all informed consent forms were reviewed for data reuse in accordance with AstraZeneca data-sharing rules. Patients from countries where ethics committees do not approve data reuse as well as patients who had withdrawn consent were excluded. Consequently, the data used in this work are a subset of the original trial data and any direct comparison with the original trial results should be made with care (seefor a comparison of the number of patients included in the original trials and in this analysis). We refer to the trial data used in this analysis as data sets A and B.
## Pef and exacerbations
Home-measured morning PEF (unit: L min −1 ) measurements, collected daily in an electronic diary system, were considered for development of the SDEME model. Observations from 2 weeks before the first dose (baseline measurements) up to 4 weeks after the last dose were included based on study designs and the pharmacokinetic properties of benralizumab. [bib_ref] Population pharmacokinetics and pharmacodynamics of benralizumab in healthy volunteers and patients with..., Wang [/bib_ref] Before model development, obvious outliers in the PEF time series were removed. This was done by calculating the interquartile range (IQR) for each patient and removing observations outside the interval (Q1 -2 × IQR, Q3 + 2 × IQR), where Q1 and Q3 denote the first and third quartiles of the observations. In total, 7077 of 740,894 (0.96%) PEF observations were removed.
For the analysis of association between the properties of a patient's PEF response and asthma exacerbation risk, exacerbations taking place after the first dose and up to 4 weeks after the last dose were considered. Asthma exacerbations were defined as in the original study protocols, that is, as worsening of asthma that led to (1) use of systemic corticosteroids for 3 days or more or a temporary increase in a stable, background dosage of oral corticosteroids; (2) an emergency department or urgent care visit (<24 h) attributed to asthma that required systemic corticosteroids; and/or (3) an inpatient admission to hospital (≥24 h) attributed to asthma. Worsening of asthma was defined as any new or increased symptoms or signs that were concerning to the patient or related to an Asthma Daily Diary alert.
## Statistical modeling and analysis
The statistical modeling and analysis of the PEF and exacerbation data consists of the following three parts: (1) development of an SDEME model of PEF, (2) the analysis of the association between a patient's SDEME model parameter estimates and asthma exacerbation risk, and (3) the comparison of the SDEME model results to a standard DFA and coefficient of variation (CV) analysis of PEF data.
A stochastic differential equation mixed effects model of PEF SDEME models are extensions of the frequently used nonlinear mixed effects models with differential equations, where the underlying dynamical system includes stochasticity (e.g., biological and environmental stochastic effects). Thus, SDEME models provide a means to distinguish the following three sources of variability: interindividual variability (IIV), stochasticity in the dynamics (also known as system noise), and measurement noise. [bib_ref] Mixed effects modeling using stochastic differential equations: illustrated by pharmacokinetic data of..., Leander [/bib_ref] [bib_ref] Beyond deterministic models in drug discovery and development, Irurzun-Arana [/bib_ref] As this is-to the authors' best knowledge-the first application of an SDEME model to analyze daily lung function measurements, the model structure was developed with the following criteria in mind: (1) the model should be conceptually simple with interpretable parameters, (2) it should be able to describe the most important stochastic characteristics and trends of the PEF data, and (3) it should be computationally feasible and robust enough to be applied to large sets of clinical trial data.
The PEF model is illustrated in . The model has the following three components: one component that describes a deterministic response (i.e., the trend induced by treatment), one component that describes long-term fluctuations around the deterministic response, and one additional component to capture dayto-day variability. The three components are detailed in the next paragraphs. The deterministic response was described using a turnover model given by the ordinary differential equation
In this equation, PEF base (unit: L min −1 ) and eff (unitless) represent the baseline PEF and asymptotic treatment effect (set to zero before treatment initiation), respectively. The two arms with benralizumab treatment in each trial were pooled into a single active treatment group. The treatment effect was modeled as relative to baseline, which was supported by exploratory data analyses and model diagnostics. The parameter k tr (unit: day −1 ) is the rate constant of the turnover process describing the onset of treatment effect (different values estimated for active treatment and placebo, respectively). Exploratory analyses revealed a strong influence of age (centered around 50 years of age) and sex (female as the reference group) on the PEF baseline, and this together with a random effect,
[formula] dx (t) dt = k tr PEF base (1 + eff) − x (t) , x (0) = PEF base . [/formula]
base , were incorporated to model the PEF baseline on an individual level:
where the covariate sex is 0 for female and 1 for male. In addition, eff was allowed to take different values depending on treatment group (active or placebo) and modelled using an additive random effect eff
The stochastic dynamics (here referred to as long-term fluctuations) is modeled by an Ornstein-Uhlenbeck (OU) process. The OU process is a stationary stochastic process, which tends to drift toward its mean (set to zero in this case). The process is described by the following stochastic differential equation:
where W (t) is a standard Wiener process, k v (unit: day −1 ) is the so-called mean reversion speed, and g (unit: L min −1 day −1/2 ) is a scaling factor for the stochastic component, referred to as the magnitude of the long-term fluctuations. Note that in the case g = 0, this corresponds to a standard turnover model with the constant solution v (t) = 0, whereas for g > 0, the process will fluctuate around 0. The larger the value of g, the more the fluctuations will deviate from the mean. The frequency distribution and correlation over time of the fluctuations is determined by k v (increasing correlation over time with decreasing mean reversion speed). The quantity 1∕k v is sometimes referred to as the characteristic correlation time of the process.
The actual PEF observations are finally modeled by the sum of the deterministic response, the long-term fluctuations, and the day-to-day variability, denoted e(t k ) :
where e(t k ) ∼ (0, 1), k = 1, ⋯ , n are independent standard normal random variables, and (unit: L min −1 ) is the magnitude (standard deviation) of the day-to-day variability. The day-to-day variability is assumed to include, for example, uncorrelated physiological changes on short timescales and device measurement errors.
To account for IIV in the day-to-day variability, as well as the long-term fluctuations, random effects on and g are introduced according to the following relationships: PEF base = PEF base, female + sex ΔPEF base, male + PEF base, age (age − 50) exp base ,
[formula] eff = eff k + eff , k ∈ {plac, active} dv (t) = −k v v (t) dt + g dW (t) , v(0) = 0, PEF(t k ) = x(t k ) + v(t k ) + e(t k ) F I G U R E 1 [/formula]
Illustration of the longitudinal PEF model. The observed PEF time series is modeled by the following three components: trend, long-term fluctuations, and day-to-day variability. PEF, peak expiratory flow where treatment denotes the treatment covariate, which is set to 0 for placebo and 1 for active. The parameters plac and g plac denote the population-typical values for the magnitude of the day-to-day variability and the magnitude of the longterm fluctuations for the placebo group, respectively, and p , active and p g, active denote multiplicate (additive on logscale) treatment covariate effects. Note that the inclusion of IIV on the day-to-day variability implies that the magnitude of the day-to-day variability varies between patients, but not within. The inclusion of IIV on the residual error has previously been adopted by others in the context of pharmacokinetic-pharmacodynamic modeling. [bib_ref] Assessment of the relative in vivo potency of the hydroxylated metabolite of..., Kerbusch [/bib_ref] In total, the model has four random effects, and the random-effects vector = base , eff , , g is assumed to be multivariate normally distributed with a mean of zero and covariance matrix .
Separate PEF models were estimated for data sets A and B.
## Estimation of model parameters and model selection
The model parameters were estimated using the firstorder conditional estimation with interaction method, as implemented in the open-source Wolfram Mathematica package NLMEModeling.Because of the stochastic dynamics of the underlying system, the extended Kalman filter (EKF) was used to estimate the state of the system. [bib_ref] Mixed effects modeling using stochastic differential equations: illustrated by pharmacokinetic data of..., Leander [/bib_ref] [bib_ref] Exact gradients improve parameter estimation in nonlinear mixed effects models with stochastic..., Ólafsdóttir [/bib_ref] Similar to other implementations, [bib_ref] Stochastic differential equations in NONMEM ® : implementation, application, and comparison with..., Tornøe [/bib_ref] NLMEModeling automatically generates the necessary equations for the EKF.
Model selection was based on a combination of measures, including log-likelihood value, goodness-of-fit assessments, visual predictive checks, inspection of empirical Bayes estimates (EBEs), and parameter precision. Inclusion of offdiagonal elements in the random-effects covariance matrix was guided by empirical assessment of the EBEs. The shrinkage for EBEs was calculated on the standard deviation scale.
The standard errors of the parameter estimates were calculated from the variance-covariance matrix of the parameter estimates.
## Association with asthma exacerbation risk
The association between a patient's PEF model parameters (using the EBEs) and the risk of asthma exacerbations was assessed in an RTTE analysis based on pooled data from data sets A and B. The RTTE analysis was done based on a Cox proportional hazards model with a γdistributed frailty parameter, resulting in an RTTE model often referred to as a shared frailty model. [bib_ref] Improving the evaluation of COPD exacerbation treatment effects by accounting for early..., Król [/bib_ref] [bib_ref] A tutorial on frailty models, Balan [/bib_ref] The frailty describes the (unexplained) between-patient variability in exacerbation risk and acts proportionally on the baseline hazard. Hence, the exacerbation risk for individual i is modeled according to the relationship where h 0 (t) denotes the baseline hazard, u i is the γdistributed frailty, X i denotes the individual covariates, and are the corresponding regression coefficients. Covariates considered in the RTTE analysis were the factors study, treatment group, age, and sex as well as the individual model parameters PEF base , eff , and g.
The R package frailtyEM was used to perform the RTTE analysis using the Anderson-Gill method for counting processes to handle repeated events. [bib_ref] frailtyEM: an R package for estimating semiparametric shared frailty models, Balan [/bib_ref] [bib_ref] Cox's regression model for counting processes: a large sample study, Andersen [/bib_ref] Detrended fluctuation analysis DFA is a method for studying long-range correlations in a signal. The output of the analysis is a positive scaling exponent α, where α = 0.5 indicates the signal is uncorrelated and increasing values of α indicate increasingly stronger long-range correlations. 5,6 Several published articles have used DFA to quantify long-range correlations in frequently measured PEF data and assessed its association with exacerbation risk. [bib_ref] Risk of severe asthma episodes predicted from fluctuation analysis of airway function, Frey [/bib_ref] [bib_ref] Complexity of chronic asthma and chronic obstructive pulmonary disease: implications for risk..., Frey [/bib_ref] [bib_ref] Detrended fluctuation analysis of peak expiratory flow and exacerbation frequency in COPD, Donaldson [/bib_ref] The DFA implementation in the Python package fathon was used to calculate in this analysis. [bib_ref] A Python package for a fast computation of detrendend fluctuation analysis and..., Bianchi [/bib_ref] Because of the sensitivity of the method to missing data and short time series, a DFA was only performed on patients with more than 6 consecutive months of PEF observations including less than 3% missing values. [bib_ref] Risk of severe asthma episodes predicted from fluctuation analysis of airway function, Frey [/bib_ref] Linear interpolation was used to get a regularly sampled time series in case of missing values. The steady-state CV (CV ss ), based on data 100+ days after treatment initiation, was also calculated for each of these patients, and between-group (active vs. placebo) comparisons of α and CV ss were performed on pooled data (data sets A and B) using a two-sample t-test. The values were also correlated with the EBEs of g and for each patient using the Spearman rank correlation.
# Results
## Clinical data characteristics
The clinical trial data used in the analysis are summarized in
[formula] h i (t) = h 0 (t) u i exp X T i , [/formula]
morning PEF were available for analysis in data sets A and B, with an average number of 339 and 282 observations per patient, respectively. The total number of moderate to severe exacerbations in the two analysis data sets were 989 in data set A and 1012 in data set B, with 516 and 474 patients having at least one exacerbation, respectively. The observed PEF base was slightly higher in data set A compared with data set B (mean 248 L/min and 236 L/ min, respectively), whereas the age distribution in both trials were similar (median 49 years), and the majority of patients were female (61% and 66%, respectively).
## Mixed effects pef model
Parameter estimates for the two PEF models are listed in and visual predictive checks are shown in .
Additional model diagnostics are provided in Figures S1-S4. Parameter estimates were estimated with adequate precision (relative standard error percentage, <30 for all parameters). The PEF base were estimated slightly higher in data set A, and there was a significant effect of both sex and age in both trials. The estimated asymptotic treatment effects eff active were similar (9.1% and 9.6%, respectively), whereas the placebo effect eff plac was slightly higher in data set B (3.1% and 4.0%, respectively).
During model development, correlations between the EBEs were apparent, and a full random-effects covariance matrix was assessed. However, the correlation between the baseline and treatment random effect was small and hence fixed to zero to improve numerical stability. The shrinkage of the random-effect parameters was small (largest 16.0% for the random effect on treatment effect in data set B).
The rate constant k tr for the treatment response were estimated higher in the active than in the placebo group, with the time to reach 90% of steady state being 80-140 days for the active groups and 220 days for placebo.
The magnitude of the long-term fluctuations g was significantly lower in the active group compared with placebo: −4.7% (95% confidence interval [CI], −8.8% to −0.36%) in data set A and −11.6% (95% CI, −15% to −7.2%) in data set B. No significant difference was seen in the magnitude of the day-to-day variability , but there was a trend toward lower values in the active group: −0.4% (95% CI, −3.5% to +2.9%) in data set A and −2.5% (95% CI, −5.9% to +0.98%) in data set B. Similar parameter estimates were obtained using data up to the first exacerbation only, as listed in.
In , the observed PEF time series (black dots) for six randomly selected patients are depicted together with the model prediction for PEF (yellow) and the deterministic treatment response (black, solid line).
## Association with exacerbation risk
The results of the RTTE analysis are presented in . Both sex and age were found to be nonsignificant (p values of 0.63 and 0.41, respectively) and were not F I G U R E 2 Visual predictive checks for the two analysis data sets. The black lines show the 10th, 50th, and 90th percentiles of the observations. The shaded areas are the 95% confidence intervals of the median (light blue) and the 10th and 90th (light gold) percentiles predicted by the model. PEF, peak expiratory flow
[formula] (a) (b) (c) (d) [/formula]
included in the final time-to-event model. Study, treatment group, PEF base , eff , and g were found to be significantly associated with exacerbation risk. Overall, the risk of exacerbation was slightly higher in data set B, with a hazard ratio (HR) of 1.20 (95% CI, 1.05-1.36) compared with data set A. The estimated HR for active treatment versus placebo was 0.70 (95% CI, 0.61-0.79), with an additional benefit of active treatment being described via some of the PEF parameters. Specifically, both a larger PEF treatment effect asymptote and a reduced magnitude of long-term fluctuations were associated with a lower exacerbation risk (HR < 1). Finally, a high PEF base was also associated with lower risk, whereas had no significant association with exacerbation risk (p = 0.57). The RTTE results for the sensitivity analysis, listed in , identified the same significant covariates.
# Dfa analysis
About one third of all patients (data set A, n = 464; data set B, n = 355) fulfilled the data quality criteria (≥6 months of data, <3% missing values) to be included in the DFA F I G U R E 3 Observed and model-predicted PEF response for six randomly selected patients from data set B. The black solid lines show the deterministic treatment response, whereas the yellow fluctuating curves show the long-term fluctuations around the deterministic response. The light-blue shaded areas reflect the uncertainty in the prediction (± standard deviation). PEF, peak expiratory flow analysis. The distribution of α and CV ss values in placebo and active patients (pooled from the two data sets) are shown in [fig_ref] E 4: Detrended fluctuation analysis results and comparison to individual stochastic differential equations mixed... [/fig_ref] ,b. On average, both α and CV ss were lower in patients on active treatment, but only the difference in CV ss was statistically significant (p < 0.05).
# Discussion
As the use of home-based assessments in respiratory trials increases, the amount of clinical data available to characterize patients and treatment effects is rapidly growing. To make the most out of the data, new statistical approaches are required.
By developing an SDEME model, we have been able to characterize trends and fluctuations in frequently sampled, home-measured PEF data from a large number of patients with asthma enrolled in two phase III clinical trials. With the model, we show that benralizumab, an established effective treatment of asthma, both improves a patient's average PEF level and reduces the magnitude of long-term fluctuations around that level. Furthermore, we show that these effects can be associated with a reduced asthma exacerbation risk.
Our developed PEF model adequately described the observed data in both trial data sets, with consistent estimates of the model parameters overall. The populationtypical eff of benralizumab (defined as percent change from baseline) was estimated to approximately 9.5% compared with 3.5% for placebo, whereas g was 5%-12% smaller in the benralizumab group. The RTTE analysis indicates that these population-typical improvements in the average PEF level and long-term fluctuations are associated with an exacerbation risk reduction of 11% (p < 0.001) and 3%-7% (p < 0.001), respectively. It should be noted that these effects constituted only a part of the overall risk reduction, as benralizumab also showed an additional 30% risk reduction (p < 0.001) not explained by changes in PEF parameters. Interestingly, the magnitude of the day-to-day variability was neither affected by treatment nor associated with exacerbation risk (p = 0.573), suggesting that it is particularly longer term PEF fluctuations that are important.
As a drop in PEF is often seen during an asthma exacerbation, it may be no surprise that PEF fluctuations are associated with exacerbation risk. However, in our sensitivity analysis-where we only use data up to before the first exacerbation-we still saw a significant association between a patient's estimated g and exacerbation risk.
Comparing the results the SDEME model to the DFA method, an expected positive correlation between α and g and negative correlation between g and were observednoting that a direct comparison between the approaches is not straightforward. [bib_ref] Consistency of detrended fluctuation analysis, Løvsletten [/bib_ref] In the subset of patients selected for DFA, there was a nonsignificant trend of lower α values in patients treated with benralizumab as well as a significant difference in CV values (lower in the active group). This is consistent with the significant treatment difference seen in g because a smaller g corresponds to both a smaller α and CV.
An important strength of the SDEME analysis, in contrast to the DFA method, is that information on g is shared between patients and that the weighting of patients with different amounts of data (including missing data) is automatically taken care of in the estimation process. Similar to the DFA, however, our approach also involves making assumptions about the underlying data-generating process. Importantly, we describe the trends in the PEF data using a turnover model, the long-term fluctuations using the OU process, and the day-to-day variability using a white noise process. The sampled version of the OU process is in fact known to be equivalent to an autoregressive process of order 1. Hence, the PEF model includes a correlation between adjacent timepoints, similar to the approach described in Karlsson et al., [bib_ref] Three new residual error models for population PK/PD analyses, Karlsson [/bib_ref] but is generalized to also allow for an independent and uncorrelated (between timepoints) part of the residual error.
Our work is limited by the fact that we did not explore alternative stochastic processes to describe the PEF data. Thus, we cannot conclude how well the OU process compares with other possible stochastic models. The SDEME framework, however, makes it convenient to extend our model to possibly capture more complex stochastic characteristics of the data. The OU process is a linear dynamical system, meaning that the Kalman filter provides an optimal estimator.To handle more complex and possibly nonlinear models, one could consider nonlinear estimators, for example, particle filters. [bib_ref] Using PMCMC in EM algorithm for stochastic mixed models: theoretical and practical..., Donnet [/bib_ref] There are most likely also other types of variables and information that could be added to our model to explain some of the fluctuations in PEF. Examples of such variables could be known external stimuli, for example, seasonality, rescue medication use, and change in treatment. Another natural extension of the current model would be to incorporate PEF measured in the evening. However, there is evidence on a difference in the PEF level between morning and evening, [bib_ref] Diurnal variation in peak expiratory flow and forced expiratory volume, Goel [/bib_ref] which an extended model would have to account for.
For future perspectives, one could potentially develop stochastic models that simultaneously integrate multiple types of home-based measurements, for example, FeNO, rescue medication use, and patient-reported symptoms. Such models would have the potential advantage to provide a more robust estimate of disease fluctuations compared with studying each variable in isolation.
In conclusion, we have developed a novel, extendable, model-based approach for analyzing home-based measurements of PEF collected in respiratory clinical trials. Our model enables characterization of multiple statistical properties of PEF time series data to support better estimation and understanding of treatment effects, disease stability, and exacerbation risk. We recommend including this type of model-based analysis as a complement to descriptive statistics of home-measured PEF data in the reporting of respiratory clinical trials.
## Conflicts of interest
J.L. and U.G.E. are employees of AstraZeneca and may own stock or stock options. R.P. was an employee of AstraZeneca at the time when the work was conducted. M.J. declared no competing interests for this work.
# Author contributions
[fig] Figure 4c ,: d illustrate the correlations between α and g and α and . The scaling exponent α exhibited a positive correlation with g (Spearman ρ = 0.55; p < 0.001) and a negative correlation with (Spearman ρ = −0.16; p < 0.001). [/fig]
[fig] E 4: Detrended fluctuation analysis results and comparison to individual stochastic differential equations mixed effects modelderived parameters (pooled data). Box plots for (a) CV ss and (b) α for the two treatment groups. Correlation between α and (c) the magnitude of the long-term fluctuations g and (d) the magnitude of the day-to-day variability ; solid lines show a local regression weighted smoother. CV ss , steady-state coefficient of variation [/fig]
|
Perceval sutureless aortic valve replacement after ascending aortic replacement
This is an open access article under the terms of the Creat ive Commo ns Attri bution-NonCo mmercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.AbstractIn patients with a narrow sinotubular junction, small sinus of Valsalva, or extensibility loss in the aortic root, aortic valve replacement (AVR) with a standard valve is challenging due to limited surgical field. Detailed preoperative measurements of the aortic root render performing AVR using the Perceval valve easy.K E Y W O R D Saortic valve replacement, ascending aortic replacement, Perceval, sutureless How to cite this article: Yajima S, Satoh A, Sekiya N, et al. Perceval sutureless aortic valve replacement after ascending aortic replacement.
# | introduction
The Perceval bioprosthesis (LivaNova PLC) is a sutureless valve built on a self-expandable nitinol stent that has a dual role of supporting the valve and fixing it in place with three bovine pericardial leaflets.The implantation method is simple, with three guiding sutures placed on each nadir during implantation.Subsequently, the sutures are removed from the annulus, which means that ligation is not required. This sutureless system is favorable in cases with a narrow sinotubular junction (STJ), small sinus of Valsalva, or loss of extensibility of the aortic root because, in such cases, it is often difficult to obtain an ideal surgical field for suturing or tying knots in aortic valve replacement (AVR) with standard valves. In patients undergoing ascending aortic replacement using a proximal internal felt strip, mobilization of the STJ or aortic root is restricted; therefore, we often encounter troublesome cases that require removal of the implanted graft and felt strips to perform AVR or aortic root replacement.Herein, we reported two successful AVR cases using the Perceval valve in patients with a history of ascending aortic replacement, without any manipulation of the aortic root.
## | case report
## | case 1
A 73-year-old man who had bicuspid aortic valve stenosis accompanied with ascending aortic dilatation underwent AVR using a 21-mm Trifecta aortic valve (St. Jude Medical, Inc.) and ascending aortic replacement using a 26-mm Jgraft (Japan Lifeline) in 2014. He suddenly presented with dyspnea on exertion and leg edema in 2020. Chest radiography showed cardiomegaly with a cardiothoracic ratio of 61%, pulmonary congestion, and bilateral pleural effusion. On echocardiography, the non-coronary cusp of the prosthetic valve was disrupted, resulting in severe aortic insufficiency. The prosthetic valve also revealed mild stenosis, with a mean gradient of 28 mmHg. Mitral insufficiency was moderate, and tricuspid insufficiency was mild, with a peak pressure gradient of 46 mmHg. The diameters of the STJ and aortic valve annulus were 30.3 mm and 23.5 mm (STJ/annulus ratio, 1.29), respectively. The distance between the aortic annulus and the proximal felt strip was 29.6 mm . In the operative findings, the aortic root severely adhered to the surrounding tissue, and the proximal felt strip reduced the extensibility of the STJ. Aortotomy was performed on the vascular graft approximately 35 mm above the aortic annulus (approximately 5 mm above the proximal felt strip). After removal of the prosthetic valve and pannus beneath the annulus, a Perceval M bioprosthesis was implanted with three guiding sutures, which were placed on each nadir . The aortotomy line was sufficiently separated from the outflow ring to be closed. The operative, cardiopulmonary bypass, and cardiac cross-clamping times were 380, 150, and 89 min, respectively. No atrioventricular block was observed. Postoperative echocardiography revealed opened left ventricular outflow, with a mean prosthetic valve pressure gradient of 10 mmHg, an effective orifice area of 1.8 cm 2 , and no paravalvular leakage. The mitral and tricuspid insufficiencies were both improved to mild, with a peak tricuspid valve pressure gradient of 21 mmHg. A Perceval bioprosthesis was implanted in the optimal position below the proximal felt strips . The patient was discharged on day 18 postoperatively, without any complications.
## | case 2
A 79-year-old woman, who underwent ascending aortic replacement using a 26-mm J-graft (Japan Lifeline) for acute aortic dissection 6 months prior, developed acute heart failure due to severe aortic insufficiency. Echocardiography revealed a vegetation-like mass on the aortic valve, which caused severe aortic insufficiency. In addition, a blood culture revealed the presence of Enterococcus faecalis. On contrast-enhanced computed tomography, the diameters of the STJ and aortic valve annulus were found to be 27.3 and 24.0 mm (STJ/annulus ratio, 1.14), respectively, and the distance between the aortic annulus and the bottom of the proximal felt strip was 30.6 mm . After 4 weeks of antibiotic treatment followed by a negative blood culture, AVR was performed. Aortotomy was performed with procedures similar to those used in case 1, and a Perceval L bioprosthesis was implanted in the same manner as in case 1 . The aortotomy line was sufficiently separated from the outflow ring to be closed. The operative, cardiopulmonary bypass, and cardiac cross-clamping times were 351, 168, and 87 min, respectively. In this case, a postoperative advanced atrioventricular block developed; therefore, a permanent pacemaker was implanted 35 days postoperatively. Postoperative echocardiography demonstrated a mean prosthetic valve pressure gradient of 11 mmHg, and the effective orifice area was 1.4 cm 2 , with trivial paravalvular leakage. A Perceval bioprosthesis was implanted below the proximal felt strips . The patient was discharged 50 days postoperatively.Considering the outflow stent height of the XL size valve, if the distance between the aortic annulus and the bottom of the proximal felt strip is over 30 mm, a Perceval valve can be implanted without any concerns of deformation. In case 1, we assumed that the implanted valve size was M or L, as a result of the preoperative measurement of the annulus diameter; therefore, we could predict that the valve was implanted below the proximal intimal felt strip, even though the distance between the aortic annulus and the bottom of the proximal felt strip was 29.6 mm. Note that the proximal felt strip restricts the extensibility of the STJ, and valve sizers cannot easily pass the STJ. As a result, underestimation of the size can arise, particularly in cases using a standard valve. Therefore, preoperative measurement of the annulus is crucial for the selection of valve size in such cases. The valve-collapsing delivery system of the Perceval valve would also be helpful in passing a narrow and inflexible STJ. We cannot explain the reason for the development of delayed advanced atrioventricular block in case 2 clearly, but that of the right size was implanted in the right position as shown in . The nature of Perceval self-expandable characteristic would lead to gradual oppression on the atrioventricular node. To determine the optimal valve size by measurement requires more time because of the narrow STJ. With the current experience, we concluded that the optimal valve size can be selected preoperatively if detailed aortic annulus measurement was made, and AVR using the Perceval valve can be performed with aortotomy on the graft and without any manipulation of the aortic root. More experience could make the surgery-related time shorter.
## Acknowledgement
None.
## Conflict of interest
The authors declare that there are no conflicts of interest.
# Author contributions
Shin Yajima was responsible for the acquisition of the data and writing of the original manuscript. Taichi Sakaguchi assisted with the critical revision of the manuscript. The other authors provided final approval of the version to be published.
# Ethical approval
The patients provided informed consent for publication of their case details.
## Consent
Written informed consent was obtained from the patient to publish this report in accordance with the journal's patient consent policy.
# Data availability statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
## Orcid
Shin Yajima https://orcid.org/0000-0001-5058-0267 Taichi Sakaguchi https://orcid. org/0000-0002-8733-0533 |
Changes in Health Behaviors, Mental and Physical Health among Older Adults under Severe Lockdown Restrictions during the COVID-19 Pandemic in Spain
Citation: García-Esquinas, E.; Ortolá, R.; Gine-Vázquez, I.; Carnicero, J.A.; Mañas, A.; Lara, E.; Alvarez-Bustos, A.; Vicente-Rodriguez, G.; Sotos-Prieto, M.; Olaya, B.; et al.
# Introduction
The COVID-19 pandemic has rocked our society to its core. Until safe and effective vaccines are globally distributed, social distancing interventions to reduce the transmission of SARS-CoV-2 are essential [bib_ref] Association of Public Health Interventions with the Epidemiology of the COVID-19 Outbreak..., Pan [/bib_ref] , but the interventions themselves are not exempt from health and economic challenges [bib_ref] Mental health and health behaviours before and during the initial phase of..., Niedzwiedz [/bib_ref] [bib_ref] Effects of the lockdown on the mental health of the general population..., Fiorillo [/bib_ref] [bib_ref] The Coconel Group Covid-19 health crisis and lockdown associated with high level..., Beck [/bib_ref].
The State of Alarm imposed by the Spanish Government to curb the spread of the SARS-CoV-2 infection entailed a national lockdown starting on 15 March and the imposition of distancing measures such as the cancelation of mass events or the closure of non-essential customer-facing business and educational institutions.Two weeks later, on 29 March, the Government strengthened these measures to avoid the saturation of intensive care units. A period of five weeks started in which citizens were only allowed to leave their homes for essential work, buying of staple products, or emergencies. On 4 May, citizens were first authorized to leave their homes to exercise or walk, for a maximum of 1 h a day, following timetables according to age and only being accompanied by people they had been living with during the lockdown. From 10 May to 21 June, a progressive de-escalation of confinement measures led to the so-called "New Normality", in which Spaniards were allowed to attend their jobs, gather in small groups, and move between provinces as long as they complied with safe distancing and face covering requirements. All around the world, countries dealt with the first wave of the pandemic with different degrees of movement restrictions, but the Spanish lockdown, especially during those five weeks between 30 March and 4 May, was one of the most restrictive in Europe.
A few studies have investigated the health consequences of the lockdown during the first wave of the pandemic in the Spanish adult population. Using convenience sample surveys conducted between March and June 2020, in which participants reported their perceived changes in lifestyle and health-related factors since the start of the pandemic, these studies have shown a general worsening in psychological and mental health [bib_ref] How did Different Generations Cope with the COVID-19 Pandemic? Early Stages of..., Justo-Alonso [/bib_ref] [bib_ref] Mental health consequences during the initial stage of the 2020 Coronavirus pandemic..., González-Sanguino [/bib_ref] [bib_ref] Levels and variables associated with psychological distress during confinement due to the..., Pérez [/bib_ref] , as well as in sleep quality [bib_ref] The assessment of lifestyle changes during the COVID-19 pandemic using a multidimensional..., Balanzá-Martínez [/bib_ref] [bib_ref] The mediation role of sleep quality in the association between the incidence..., Werneck [/bib_ref] [bib_ref] Profiles of sleep changes during the COVID-19 pandemic: Demographic, behavioural and psychological..., Robillard [/bib_ref] , mild or no changes in tobacco and alcohol consumption [bib_ref] The assessment of lifestyle changes during the COVID-19 pandemic using a multidimensional..., Balanzá-Martínez [/bib_ref] [bib_ref] Physical and Psychological Effects Related to Food Habits and Lifestyle Changes Derived..., López-Moreno [/bib_ref] , slight or no changes in weight [bib_ref] Physical and Psychological Effects Related to Food Habits and Lifestyle Changes Derived..., López-Moreno [/bib_ref] [bib_ref] Food habits in the Galician population during confinement for COVID-19, López [/bib_ref] [bib_ref] Weight changes during the COVID-19 home confinement. Effects on psychosocial variables, Fernandez-Rio [/bib_ref] , conflicting results for diet quality [bib_ref] Changes in Dietary Behaviours during the COVID-19 Outbreak Confinement in the Spanish..., Rodríguez-Pérez [/bib_ref] [bib_ref] Food choice motives and the nutritional quality of diet during the COVID-19..., Marty [/bib_ref] , important reductions in physical activity, and increases in sedentary time [bib_ref] The assessment of lifestyle changes during the COVID-19 pandemic using a multidimensional..., Balanzá-Martínez [/bib_ref] [bib_ref] Physical Activity Change during COVID-19 Confinement, Castañeda-Babarro [/bib_ref] [bib_ref] The Impact of the COVID-19 Confinement on the Habits of PA Practice..., García-Tascón [/bib_ref] [bib_ref] Depression, Anxiety and stress during COVID-19: Associations with changes in physical activity,..., Stanton [/bib_ref]. Interestingly, one survey found that the percentage of interviewees meeting the guidelines regarding screen time became lower between 22 March and 5 April, while unhealthy alcohol consumption and insufficient physical activity decreased during this three-week period [bib_ref] COVID-19 Confinement and Health Risk Behaviors in Spain, López-Bueno [/bib_ref].
Unfortunately, none of these studies described the determinants of the observed changes in health behaviors, most were at high risk of information bias (i.e., people not accurately reporting their past exposures or symptoms, especially at a time of emotional distress), and, although a few provided stratified information by age groups, most presented aggregated information for young, middle-aged, and older adults. Moreover, most of these studies were based on convenience samples, which increase the risk of selection bias. As the COVID-19 epidemic evolves and new social distancing measures may be needed, and being aware of the risk of future infectious pandemics, it is critical that we identify population subgroups that are at increased risk of health deterioration with social isolation and confinement. This is especially important for older adults because they represent around 10% of the worldwide population (up to 20% in Spain), frequently live alone, and are most vulnerable to the development and progression of diseases including COVID- In the general older adult population in Spain, the prevalence of tobacco smoking before the pandemic was very low, and about one-third of individuals reported consuming alcohol beverages (mostly wine) on a daily basis; adherence to the Mediterranean diet was much higher than in young and middle-age adults, but physical activity was much lower. Lastly, most older adults reported having good or only regular health, and in general reported better mental health than their young and middle-age counterparts. In this context, this study describes the main changes in health behaviors and in mental and physical health between a pre-pandemic period and weeks 7 to 15 after the beginning of the COVID-19 lockdown among participants in four cohorts of community-dwelling older adults in Spain.
# Materials and methods
## Study cohorts
Seniors-ENRICA-2 (ENRICA): Longitudinal study of older adults selected between 2015 and 2017 through sex-and district-stratified random sampling of all communitydwelling individuals aged ≥65 years holding a national healthcare card and living in the Madrid Region. In 2019 (baseline wave for the present study), participants were reinterviewed face-to-face at their homes using Computer-Assisted Personal Interviews (CAPI), and they underwent a physical exam and blood tests. Participants gave informed written consent, and the Clinical Research Ethics Committee of the La Paz University Hospital in Madrid approved the study and its follow-ups (Protocol #HULP-PI 1793).
Edad con Salud (ES): Longitudinal population-based household survey representative of the non-institutionalized older adult population aged ≥60 years recruited in 206 municipalities in Spain. The first wave took place between 2011 and 2012 and was part of the Collaborative Research on Ageing in Europe (COURAGE) study. Participants were re-evaluated twice, between 2014 and 2015, and in 2018 [bib_ref] Fruit and Vegetable Consumption and Potential Moderators Associated with All-Cause Mortality in..., Olaya [/bib_ref]. In 2019-2020 (baseline wave for the present study), just before the start of the COVID-19 outbreak, a refreshment sample with participants from 9 municipalities in Barcelona and 7 municipalities in Madrid was added. Interviews were conducted face-to-face by trained lay interviewers at respondents' homes using CAPI. The research protocol and its follow-ups were approved by the Clinical Research Ethics Review Committees of both Parc Sanitari Sant Joan de Déu (Barcelona, Spain) and Hospital Universitario La Princesa (Madrid, Spain). All participants provided written informed consent for their participation and the treatment of their personal data.
Toledo Study for Healthy Aging (TSHA): Prospective study of older adults aged ≥65 years recruited in 15 municipalities in the province of Toledo. The cohort was established between 2006 and 2009, and it had follow-up visits in 2011-2013 and 2016-2017 (baseline wave for the present study), when participants were interviewed at their homes using CAPI, underwent a physical exam, and provided blood and urine samples [bib_ref] The prevalence of frailty syndrome in an older population from Spain. The..., Garcia-Garcia [/bib_ref]. Participants gave informed written consent, and the Clinical Research Ethics Committee of Toledo Hospital Complex approved the study and subsequent follow-ups (Protocol # 2203/30/2005).
The elderly-Exernet multi-center study (EXERNET): Multi-center study of noninstitutionalized individuals aged ≥65 years recruited in Aragón, Castilla-La Mancha, and Madrid. At baseline (2008-2009) and follow-up visits (2011-2012 and 2016-2017 (baseline wave for the present study)) participants responded to a face-to-face questionnaire and underwent a physical exam [bib_ref] The effects of Age, Organized Physical Activity and Sedentarism on Fitness in..., Gomez-Bruton [/bib_ref]. Participants gave informed written consent, and the Clinical Research Ethics Committee of Aragón approved the study and its follow-ups (#18/2008).
## Study participants
A total of 4936 participants from the cohorts with at least one pre-pandemic follow-up visit were eligible to participate in the study. From these, 753 individuals could not be contacted, and 717 did not agree to participate. After excluding 120 individuals infected with COVID-19, 97 living with a COVID-19 infected patient, as well as 208 participants with no information on important confounders, a total of 1323 individuals from ENRICA, 464 from ES, 829 from TSHA, and 425 from EXERNET comprised the analytical sample (n = 3041).
## Study variables
At baseline and at follow-up, we collected data on health behaviors, mental and physical health, and their potential determinants, including demographic and social variables, housing conditions, and aging experiences. [fig_ref] Figure 1: Baseline information collected in the study cohorts [/fig_ref] shows the main variables collected at baseline in each cohort. In all cohorts, baseline information was collected on sociodemographic factors (sex, age, education, civil status); social isolation (cohabitation, frequency of contacts with family and friends); tobacco consumption; diet quality (14-point Mediterranean diet adherence screener questionnaire-MEDAS) [bib_ref] A Short Screener Is Valid for Assessing Mediterranean Diet Adherence among Older..., Schröder [/bib_ref] ; physical activity, as per the EPIC-cohort questionnaire (EPAQ) [bib_ref] Estimation of reproducibility and relative validity of the questions included in the..., Pols [/bib_ref] in ENRICA and the Elderly EXERNET Physical Activity Questionnaire (EEPAQ) [bib_ref] Validation of the self-report EXERNET questionnaire for measuring physical activity and sedentary..., López-Rodríguez [/bib_ref] in EXERNET, with results expressed as Metabolic Equivalent Tasks (METs), the Global Physical Activity Questionnaire in ES [bib_ref] International Physical Activity Questionnaire: 12-Country Reliability and Validity, Craig [/bib_ref] , and the Physical Activity Scale for the Elderly (PASE) in the TSHA); measured weight and height, with the body mass index (BMI) being calculated as the weight in kg divided by squared height in m; hours of nighttime sleep (short sleep defined as ≤6 h and long sleep as ≥9 h); quality of life (the 12-item Short Form (SF) in ENRICA, the WHO Disability Assessment Schedule 2.0 (WHODAS) in ES, and the EuroQol-5D [bib_ref] Validity of the EQ-5D-5L and reference norms for the Spanish population, Hernandez [/bib_ref] in TSHA and EXERNET); and history of physiciandiagnosed chronic conditions (i.e., hypertension, diabetes, osteoarthritis, and depression). Definition of hypertension was based on a self-reported physician diagnosis, current use of anti-hypertensive medication, or a clinical blood pressure reading ≥140/90 mmHg. Type 2 diabetes mellitus was defined as a self-reported physician diagnosis, current use of anti-diabetic medication, or fasting glucose ≥126 mg/d.
## Baseline information
## Follow-up information
Phone interviews were performed in the four cohorts between 27 April (week 7 of the pandemic) and 22 June (week 15), with information being collected on: presence and symptoms of COVID-19 cases within the household; number of outpatient or hospital health care visits during confinement; problems receiving health care or accessing medications during the pandemic; social isolation, degree of social support, and responsibilities towards dependents; sub-optimal housing conditions (including accessibility problems, noise annoyance, lack of internet access, absence of outdoor views or lack of a terrace/balcony or a private garden/yard to go out during confinement); tobacco consumption; passive tobacco smoke; alcohol drinking; diet quality; grocery modality during confinement (on-line, in-store) and person in charge; eating at a regular time (all cohorts but ES); eating more ultra-processed food than usual; eating snacks or food to calm anxiety (ENRICA); time spent in physical activities such as walking (i.e., to the supermarket, the pharmacy, or to walk the dog), exercising at home (i.e., with a stationary bike, a treadmill, fitness devices, etc.), engaging in do-it-yourself activities, or doing housework (ENRICA and EXERNET); the PASE questionnaire (TSHA and EXERNET); reasons for not doing any physical activity confinement; weight; time on sedentary activities (ENRICA) or overall sedentary time (EXERNET); hours of night-time sleep; hours of day-time sleep (EN-RICA and EXERNET); poor sleep quality (all cohorts but the TSHA); quality of life (SF-12 in ENRICA and WHODAS in ES); frequency, intensity, and locations of pain (ENRICA). Additionally, in some cohorts, information was collected on feelings of loneliness (the 3-item UCLA loneliness scale in ES, and a 0 to 5 scale in ENRICA) [bib_ref] A Short Scale for Measuring Loneliness in Large Surveys, Hughes [/bib_ref] ; exposure to secondhand smoke (SHS) (ENRICA) [bib_ref] Exposure to secondhand tobacco smoke is associated with reduced muscle strength in..., Rios [/bib_ref] ; alcohol intake (all cohorts but EXERNET); frequency of snack consumption and use of food to calm anxiety (ENRICA); time on sedentary activities (the Nurses' Health Study questionnaire in ENRICA) [bib_ref] Validation of the Spanish version of the physical activity questionnaire used in..., Martínez-González [/bib_ref] or overall sedentary time (ES and EXERNET); hours of day-time sleep (ENRICA and EXERNET); overall sleep quality (in all cohorts but TSHA); other indicators of poor night-time sleep quality (i.e., "difficulty falling asleep", "awakening during night-time", "early awakening with difficulty getting back to sleep" and "use of sleeping pills") and drowsiness (i.e., "being so sleepy at day-time as to need a nap", "not feeling rested in the morning", and the Epworth Sleepiness Scale) in ENRICA [bib_ref] Social Network and Risk of Poor Sleep Outcomes in Older Adults: Results..., Leon-Gonzalez [/bib_ref] ; prevalence and characteristics of pain (ENRICA) [bib_ref] Gender Differences in Pain Risk in Old Age: Magnitude and Contributors, García-Esquinas [/bib_ref] ; history of physiciandiagnosed coronary heart disease, congestive heart failure, angina, or cancer at any site (all cohorts but EXERNET); mobility limitations, in ENRICA [bib_ref] Television viewing time as a risk factor for frailty and functional limitations..., García-Esquinas [/bib_ref] ; the Mini-Mental State Examination Score (MMSE) [bib_ref] Evalidation and standardization of the cognition mini-exam (first Spanish version of the..., Lobo [/bib_ref] (ENRICA and ES); the Cantril's Ladder of Life Scale [bib_ref] Cantril Self-Anchoring Striving Scale, Glatzer [/bib_ref] (ENRICA and ES); and a negative aging experience scale that ranged from 1 (very positive experience) to 5 (very negative experience) (ENRICA).
Information on overall sleep quality was harmonized into a variable with the following categories: "very good", "good", "fair", "poor/very poor". Moreover, in ENRICA a 0 to 7 score with the number of poor sleep quality indicators was constructed as follows: those who answered "sometimes" or "almost always" to the items "difficulty falling asleep", "awakening during night-time", "early awakening with difficulty getting back to sleep", "use of sleeping pills", "being so sleepy at day-time as to need a nap", or "not feeling rested in the morning", as well as those with an Epworth Sleepiness Scale score > 10, received 1 point; their counterparts received 0 points. Additionally, in ENRICA, a pain scale ranging from 0 (no pain) to 7 (worse pain) was created based on pain frequency, intensity (assessed according to its impact on habitual activities), and number of pain sites. Sporadic (<2 times/week) and frequent (≥2 times/week) pain were assigned scores of 1 and 2, respectively; light, moderate, and high intensity pain, scores of 1, 2, and 3, respectively; and 1-2 and ≥3 pain sites scores of 1 and 2, respectively [bib_ref] Gender Differences in Pain Risk in Old Age: Magnitude and Contributors, García-Esquinas [/bib_ref].
## Follow-up information
Phone interviews were performed in the four cohorts between 27 April (week 7 of the pandemic) and 22 June (week 15), with information being collected on: presence and symptoms of COVID-19 cases within the household; number of outpatient or hospital health care visits during confinement; problems receiving health care or accessing medications during the pandemic; social isolation, degree of social support, and responsibilities towards dependents; sub-optimal housing conditions (including accessibility problems, noise annoyance, lack of internet access, absence of outdoor views or lack of a terrace/balcony or a private garden/yard to go out during confinement); tobacco consumption; passive tobacco smoke; alcohol drinking; diet quality; grocery modality during confinement (online, in-store) and person in charge; eating at a regular time (all cohorts but ES); eating more ultra-processed food than usual; eating snacks or food to calm anxiety (ENRICA); time spent in physical activities such as walking (i.e., to the supermarket, the pharmacy, or to walk the dog), exercising at home (i.e., with a stationary bike, a treadmill, fitness devices, etc.), engaging in do-it-yourself activities, or doing housework (ENRICA and EXERNET); the PASE questionnaire (TSHA and EXERNET); reasons for not doing any physical activity confinement; weight; time on sedentary activities (ENRICA) or overall sedentary time (EXERNET); hours of night-time sleep; hours of day-time sleep (ENRICA and EXERNET); poor sleep quality (all cohorts but the TSHA); quality of life (SF-12 in ENRICA and WHODAS in ES); frequency, intensity, and locations of pain (ENRICA).
# Statistical analysis
We first evaluated the longitudinal changes in the frequency of smoking, SHS, alcohol intake, diet, physical activity, measured weight, sedentary time, night-time and day-time sleep, sleep quality, overall health, and pain between the pre-confinement and confinement periods, by using mixed models with robust standard errors accounting for within-participant correlations induced by repeated measures. These mixed models were fitted with adjustment for sociodemographic, lifestyle, and health-related risk factors, including fixed effects for sex, educational level (primary or less, secondary, or university), baseline marital status (married, widowed, single, divorced), and study cohort (model 1); as well as fixed effects for baseline BMI and chronic morbidities, and changes over time in cohabitation status (living alone or not), smoking status (never, former or current), adherence to the Mediterranean diet score (0 to 14 points), physical activity (cohort-specific quartiles), night-time sleep, and overall health (cohort-specific quartiles) (model 2). To reduce the influence of increasing age during follow-up, we also conducted sensitivity analyses by excluding participants whose baseline measures were obtained earlier than 1.5 years before confinement.
In order to identify potential determinants of lifestyle changes or health decline during confinement, we then evaluated the relationship between baseline variables and changes in lifestyles and health outcomes. With this purpose, we used either linear or multinomial regression models that adjusted for baseline socio-demographic factors, housing conditions, smoking status, adherence to the Mediterranean diet, BMI, recreational activity and household activity, night-time sleep, chronic morbidities, time since last follow-up visit, cohort of study, and week of confinement. Although results from these models are given for each predictor and study outcome in tables, we tried to simplify the presentation of results by conducting additional backward stepwise linear and multinomial regression models with the studied baseline predictors (inclusion set at p < 0.1). These models were forced to adjust for age and sex, and their results are reported in figures, with red or blue colors identifying subjects with unhealthier and healthier changes, respectively. Categories "a" and "b" from multinomial regression models (defined in the paragraph below) represent these unhealthier (red) and healthier (blue) categories.
Linear (both fully adjusted and stepwise) regression was used to model the changes in the MEDAS score (all cohorts), weight (all cohorts), recreational and household activities (ENRICA and EXERNET), the PASE score (TSHA), total sedentary time (ENRICA and EXERNET), specific sedentary activities (ENRICA), hours of night-time sleep (all cohorts), number of total poor sleep quality indicators (ENRICA), physical component score (PCS) and mental component score (MCS) of the SF-12 (ENRICA), the WHODAS-12 (ES), and the pain scale (ENRICA). In addition, multinomial (both fully adjusted and stepwise) regression was used for the following a priori defined outcomes:
## 1.
Changes in alcohol consumption. Participants in ENRICA, ES, and TSHA cohorts were classified into the following categories: (a) drinkers who increased their frequency of consumption (i.e., changed from not daily to daily drinkers); (b) drinkers who decreased their frequency of intake (i.e., changed from daily to not daily or to no consumption at all); (c) drinkers who did not change their frequency of consumption (reference); and (d) non-drinkers who maintained their non-drinking status during confinement.
## 2.
Changes in diet quality. A "worsening" (a) and an "improvement" (b) categories were created including individuals who had either decreased or increased, respectively, their MEDAS score ≥ 1 point over follow-up; while a "reference" (c) category classified those who experienced no changes or slight increases/decreases (below 1 point) on this score.
## 3.
Changes in weight. Individuals who gained or lost >1 kg were classified into an "increased weight" (a) or "decreased weight" (b) category, respectively, while those who experienced no or small (<1 kg) change in weight were classified into a "maintained weight" or "reference" (c) category.
## 4.
Changes in physical activity and sedentary time. Participants in ENRICA, TSHA, and EXERNET cohorts were classified as experiencing: (a) "Unhealthier changes", if they decreased physical activity or increased sedentary time more than the observed 75th percentile in each cohort; (b) "Healthier (than average or reference) changes", if they decreased activity or increased sedentary time less than the 25th percentile or if they had slightly increased activity or decreased sedentary time with confinement; and (c) "Average" or "reference" changes if they changed between the 25th and 75th percentile.
## 5.
Changes in night-time sleep. The following categories were defined: (a) "Worsening", those who at baseline had normal sleep and developed short or long sleep with confinement; (b) "Improvement", those who at baseline had short or long sleep and developed normal sleep during confinement; and (c) "Reference", those who stayed in the same normal/non-normal sleep category. 6.
Changes in sleep quality. Finally, older adults in ENRICA, ES, and EXERNET cohorts were classified into: (a) "Worsening", if they had suffered an increase in the "poor sleep quality" score during confinement; (b) "Improvement", if they decreased in that score; and (c) "Reference", if they had no changes.
Statistical analyses were performed with Stata version 14.1 (StataCorp LP, College Station, TX, USA).
# Results
Supplementary Table S1 shows the prevalence and mean (SD) estimates for the main sociodemographic, lifestyle, and health characteristics of participants in the four cohorts at the pre-confinement period (baseline) and the confinement follow-up period. At baseline, the mean age was 74.5 years and the proportion of men 42.3%, although this latter figure varied widely across cohorts (from 21.2% in EXERNET to 50.5% in ENRICA). The prevalence of university studies also differed across cohorts and ranged from >15% in the ENRICA and ES to 8% in TSHAA and EXERNET. Similarly, participants in ENRICA and ES showed a higher prevalence of Internet access. Overall, 65% of participants were married, 4% to 10% lived in houses with no outdoor views, and 24% to 37% lacked an outdoor terrace or balcony to go outside during confinement. Regarding social factors, the proportion of participants who lived alone was similar at baseline and during confinement (around 20-25% in all cohorts except for EXERNET, where the frequency was 35%). However, the percentage of participants without daily contact with family or friends halved during the pandemic. Moreover, in ENRICA, the single cohort where this information was available, no increases in lonely feelings during the study period were observed.
3.1. Changes in Tobacco Exposure, Alcohol Intake, Diet Quality, and Weight [fig_ref] Table 1: Longitudinal changes in the prevalence of smoking, second-hand smoke, alcohol intake, diet... [/fig_ref] shows results for changes in the prevalence of smoking, SHS, alcohol intake, and diet quality between the pre-confinement and confinement periods. While the prevalence of active smoking remained stable, the likelihood of quitting smoking increased during follow-up. About half of the active smokers at confinement reported no changes in the frequency of tobacco consumption, while 22% reported an increase and 27% a decrease in the frequency of smoking since the start of the pandemic. The risk of increased consumption was higher among smokers with bad sleep quality at baseline and was lower among those lacking a terrace or balcony to smoke (Supplementary .
In ENRICA, the prevalence of SHS increased during follow-up, although this association was lost when excluding participants who had been followed-up for more than 1.5 years. The prevalence of current alcohol drinking was very similar in both periods, whereas a decreased likelihood of daily intake and a non-significant increased probability of daily intake were observed. Despite this change in the frequency of consumption, data from ENRICA showed that the prevalence of binge drinking (defined as having >5 units of alcohol in a session in men, or >4 in women) was lower during confinement (1.8%) than before (2.7%) (data not shown in tables). Among drinkers, only a few baseline factors were associated with the odds of changing the frequency of alcohol consumption; notably participants who lacked a terrace or balcony and former smokers showed a lower odds of increasing their frequency of alcohol intake during follow-up . The upper graph shows the odds ratios (95% confidence interval) of worsening (red color) or improving (blue color) alcohol intake according to the presence of the variables represented on the horizontal axis. Worsening was defined by increasing alcohol frequency from not daily to daily and improving as decreasing alcohol frequency from daily to not daily. For instance, the graph shows that participants who lacked a terrace or balcony and former smokers presented a lower odds of increasing their frequency of alcohol intake (worsening) during follow-up. The lower graph shows the same type of results for adherence to the Mediterranean diet. In this case, worsening and improving were defined by changes in the MEDAS score in one or more points during follow-up. Variables on the horizontal axis were selected with a backward stepwise regression from the list of candidate variables shown in .
Most study participants reported eating at a regular time during confinement, and a very low proportion declared eating more ultra-processed food than before (Supplementary [fig_ref] Table 1: Longitudinal changes in the prevalence of smoking, second-hand smoke, alcohol intake, diet... [/fig_ref]. During confinement, 51.5% of study participants shopped face-to-face, 41.7% asked someone to shop for them, and 6.8% shopped online (data not shown in tables). In ENRICA, the odds of snacks consumption and that of use of food to calm anxiety was lower during the State of Alarm than in the pre-confinement period ((odds ratio (OR): 0.72; 95%CI: 0.60; 0.85) and (OR: 0.56; 95%CI: 0.41; 0.76)), respectively [fig_ref] Table 1: Longitudinal changes in the prevalence of smoking, second-hand smoke, alcohol intake, diet... [/fig_ref] , although the association with snacks consumption was lost when excluding participants whose followup was longer than 1.5 years. Data from all cohorts combined showed a mean reduction of 0.5 points (95%CI: −0.60; −0.45) in the MEDAS score during follow-up [fig_ref] Table 1: Longitudinal changes in the prevalence of smoking, second-hand smoke, alcohol intake, diet... [/fig_ref]. The items that were most negatively affected by the State of Alarm were fruits and vegetables, fish, wine, olive oil, and sofrito (a sauced made of olive oil and garlic); while the preference for olive oil and red meat, consumption of legumes, and consuming fewer sweets and cakes improved. Predictors of increased odds of worsening diet quality with confinement included having better MEDAS scores at baseline, being older, lacking outdoor views, and sleeping > 9 h/night [fig_ref] Figure 2: Prospective association between baseline participant characteristics and changes in alcohol consumption or... [/fig_ref] and . A lower odds of worsening diet quality was observed for daily moderate drinkers and for individuals who performed more physical activity, while those with better general health and non-hypertensives were more likely to experience improvements in the MEDAS. No differences in changes in the MEDAS score were observed by grocery modality (data not shown).
Although 29% and 19% of participants reported "feeling they were getting fatter or thinner" with confinement, respectively, no changes in weight between study periods were observed in those with objective information (mean change: −0.57 kgs; 95%CI: −1.29; 0.15). Still, self-reported and objective information showed some correlation: on average, those who self-reported "feeling they were getting fatter" had gained 1.3 kg since the pre-confinement measure was taken, while those who self-reported "feeling they were losing weight" had lost, on average, −2.3 kg. Women, those with university studies, short night-time sleepers, and those who took longer naps at baseline experienced greater weight loss during follow-up than their counterparts .
## Changes in physical activity and sedentary time
Physical activity and sedentary time were the lifestyle behaviors that changed the most with confinement [fig_ref] Table 1: Longitudinal changes in the prevalence of smoking, second-hand smoke, alcohol intake, diet... [/fig_ref]. Compared with the preconfinement period, participants in ENRICA and EXERNET showed a mean reduction in recreational and household physical activity of 17. Alleged reasons for not doing any physical activity during confinement among those who used to engage in some type of recreational activity before confinement included not being able to go outside (52.4%), not knowing how to perform physical activity at home (11.4%), not having the resources to do physical activity at home (13.3%), or not having enough time (3.4%). Only 3% of study participants increased their physical activity during confinement, and in general this was a slight increase (<4 METs h/week).
Regarding sedentary time, the ENRICA and EXERNET cohorts showed an average increase of 91.68 min/day (95%CI: 81.21; 102.11), which, according to the ENRICA, was more marked for active (i.e., time on the computer or tablet, time reading) than passive (i.e., television viewing) sedentary activities.
Interestingly, changes in physical activity and sedentary time seemed to reverse with time. Compared with those interviewed in the first weeks of total confinement, those interviewed when confinement measures started to be lifted showed a progressive increase in activity and a decrease in The upper graph shows the odds ratios (95% confidence interval) of decreasing physical activity more than the observed 75th percentile change (red color) or decreasing physical activity less than the observed 25th percentile change (blue color) according to the presence of the variables represented on the horizontal axis. For instance, the graph shows that females and those with greater adherence to the Mediterranean Diet (higher MEDAS scores) were more likely to decrease physical activity less than the observed 25th percentile change, while those with no daily contact with family or friends or with a lack of outdoor views were more likely to decrease physical activity more than the observed 75th percentile. Similarly, the lower graph shows the odds ratios (95% confidence interval) of increasing sedentary time more than the observed 75th percentile change (red color) or increasing sedentary time less than observed 25th percentile change (blue color) according to these variables. Variables on the horizontal axis were selected with a backward stepwise regression from the list of candidate variables shown in Compared with men, women showed higher odds of presenting healthier changes in activity (OR: 1.34; 95%CI: 1.03; 1.74) and sedentary time (OR: 1.54; 95%CI: 1.08; 2.19), as they seemed to compensate the inability to perform some recreational activities during confinement with increases in household activity, as well as higher reductions in TV viewing time (−0.31 min/day; 95%CI: −0.60; −0.02) and higher increases in reading time (0.19 min/day; 95%CI: 0.03; 0.45). In general, sub-optimal housing conditions such as not having outdoor views or lacking a garden/yard were associated with unhealthier changes in physical activity, as well as with higher increases in TV time with confinement, while not having internet access was associated with lower increases in sedentary time, particularly in screen time. Participants with higher baseline MEDAS scores were less likely to experience unhealthy changes in physical activity or sedentary time, while those who performed more activity were less likely to increase sitting time, and those who spent more time seated were less likely to reduce physical activity. Poor night-time sleep duration and day-time sleepiness at baseline were also associated with unhealthier changes in activity and sedentary time.
## Changes in sleep duration and sleep quality
In multivariate models adjusted for changes in other potential confounders, data from all cohorts combined showed no changes in night-time sleep duration with confinement (−0.00 h/night; 95%CI: 0.06; 0.05). However, slight decreases were observed in day-time sleep duration (−12.08 min/day; 95%CI −13.73, −10.42) as well as in the "poor sleep quality" scale (0.12 points; 95%CI: 0.09; 0.15) [fig_ref] Table 1: Longitudinal changes in the prevalence of smoking, second-hand smoke, alcohol intake, diet... [/fig_ref]. Baseline predictors of unhealthier changes in sleep quality included being a married woman, living alone, feeling lonelier, doing more physical activity, a lower adherence to the Mediterranean diet, having worse overall health, and greater pain and [fig_ref] Figure 4: Prospective association between baseline participant characteristics and changes in poor sleep quality... [/fig_ref]. The upper graph shows the odds ratios (95% confidence interval) of worsening (red color) or improving (blue color) sleep quality. Variables on the horizontal axis were selected with a backward stepwise regression from the list of candidate variables shown in Supplementary Tables S5 and S6.
## Changes in physical and mental health and pain
During confinement, 23% and 4.9% of study participants attended a primary health center or an emergency room for reasons not related to a COVID-19 infection. Most of them (87.4% and 91.2%, respectively) declared they were satisfied with the care received. Most participants took their usual medication, although a few (n = 4) declared not being able to get medications because of the imposed mobility restrictions (data not shown in tables).
In Other predictors of unhealthier changes in the PCS were lacking a garden/yard, poorer sleep quality, hypertension, cardiovascular disease, cancer, depression, mobility limitations, and lower general life satisfaction. Predictors of unhealthier changes in MCS were living alone, feeling lonelier, having too much noise at home, doing less physical activity, taking very long naps, having mobility limitations, a negative aging experience, and low MMSE scores. Finally, unhealthier changes in pain were observed in participants who lived in homes with more noise, as well as in those who suffered from baseline short sleep, worse overall health, osteo-muscular disease, depression, mobility limitations, or lower general life satisfaction. . Prospective association between participant characteristics and changes in the mental component (MCS) and physical component (PCS) scores of the SF-12 (n = 1243 from Seniors-ENRICA-2), the WHODAS-12 (n = 460 from Edad con Salud), and the pain score (n = 983 from Seniors-ENRICA-2) during the COVID-19 confinement. Increases in MCS or PCS components reflect improvements in health, while increases in the WHODAS-12 or pain scores reflect worsening in health or pain, respectively. † The pain scale ranged 0 (no pain) to 7 (worst pain) and was created based on pain frequency, pain intensity (assessed according to its impact on habitual activities), and number of pain sites. Sporadic (<2 times/week) and frequent (≥2 times/week) pain were assigned a score of 1 and 2, respectively; light, moderate, and high intensity, a score of 1, 2, and 3, respectively; and 1-2 and ≥3 sites a score of 1 and 2, respectively. ‡ Information only available in the Seniors ENRICA-2 cohort. * p-value for trend <0.05. All models were adjusted for baseline age, sex (men or women), educational level (primary or less, secondary, or university), civil status (married, widowed, single, divorced), housing conditions (lack of outdoor views, terrace/balcony, private garden/yard, lack of internet, and noise), smoking status (never, former, or current), adherence to the Mediterranean diet (MEDAS score), body mass index (normoweight, overweight, or obese), physical activity (cohort-specific quartiles), hours of night-time sleep (normal, short sleep, long sleep), chronic comorbidities (hypertension, diabetes, osteo-muscular disease, and depression), time since last follow-up visit, cohort of study, and week of the State of Alarm. Models for MCS and PCS components of the SF-12 adjusted for baseline values of the PCS and the MCS, respectively.
## Changes in physical component
## Sensitivity analyses
When analyses were restricted to the 857 participants whose baseline measurements had been taken as far apart as 1.5 years before the State of Alarm, consistent findings were observed for most variables except for SHS, snacks consumption, and household physical activity, which no longer showed significant changes. The relationship with changes in some MEDAS items (i.e., vegetables, red meat, wine, sweets and cakes) also disappeared, and a slight decrease in night-time sleep was now observed (mean change −0.16 h/day; 95%CI: −0.27; −0.06).
# Discussion
This study, conducted with older adults from four Spanish population-based cohorts, showed that, on average, strict confinement was not associated with a deterioration in lifestyle risk factors, except for physical activity and sedentary time, whose effects seemed to reverse with the end of the State of Alarm. Despite this, results from the ENRICA suggested moderate declines in mental health that did not seem to reverse after restrictions were lifted. On the other hand, the apparent improvements in physical health in participants from this last cohort did reverse with the end of the State of Alarm and may have been related to a better self-perceived performance at a less physically demanding time.
This study identified subgroups of individuals at increased risk of developing unhealthier lifestyles with confinement, including males (for physical activity and sedentariness), those without daily contact with family and friends (for diet and physical activity), married participants (for sleep quality), those with feelings of loneliness (for diet and sleep quality), those with sub-optimal housing conditions (for diet, physical activity, TV viewing time), those with unhealthy sleep duration, as well as those with worse overall health or chronic morbidities (i.e., subjects with obesity or diabetics suffered greater reductions in physical activity and greater increases in screen time). On the other hand, having a greater adherence to the Mediterranean diet and doing more physical activity protected older adults from developing unhealthier lifestyles with confinement.
What do other studies say about changes in smoking, alcohol intake, diet quality, and weight during the COVID-19 pandemic?
Different on-line surveys conducted during the first months of the pandemic in adults from China [bib_ref] Report: Increased Addictive Internet and Substance Use Behavior During the COVID-19 Pandemic..., Sun [/bib_ref] , France, the Netherlands [bib_ref] The double-edged relationship between COVID-19 stress and smoking: Implications for smoking cessation, Bommelé [/bib_ref] , or the United Kingdom [bib_ref] The psychosocial impact of the COVID-19 pandemic on changes in smoking behavior:..., Chen [/bib_ref] have shown that current smokers were similarly likely (around 20-25%) to increase or decrease their tobacco consumption with the pandemic. In the Dutch survey, authors found that this apparently contradictory relationship was driven both by a negative effect of stress, boredom, and restriction on smoking and by a positive effect of motivation to quit smoking from fear of contracting COVID-19 and becoming severely ill. Consistently, in a Chinese survey, participants with higher epidemiologic concern were more often willing to improve their diets [bib_ref] The association between subjective impact and the willingness to adopt healthy dietary..., Xu [/bib_ref]. In the UK study, on the other hand, symptoms of mental distress (i.e., anxiety, low mood, sleeping problems) during quarantine were more frequently associated with increasing than decreasing tobacco smoke. Although in secondary analysis we did not find a relationship between symptoms of psychological distress and depression (measured at follow-up with the General Health Questionnaire-12 [bib_ref] The 12-Item General Health Questionnaire (GHQ-12): Reliability, external validity and factor structure..., Sánchez-López [/bib_ref] and the Geriatric Depression Scale [bib_ref] Development and validation of a geriatric depression screening scale: A preliminary report, Yesavage [/bib_ref] and changes in smoking frequency, we did observe that smoking older adults who spent more time listening or reading news related to the COVID-19 outbreak were more likely to increase their tobacco consumption (data not shown in tables). Our study provides evidence that smokers with poor sleep quality may be at higher risk of increasing smoking frequency in stressful situations than those who have good sleep quality. It also shows that while "lacking a terrace or balcony" was positively associated with unhealthier changes in physical activity and sedentary time with confinement, it reduced the risk of smokers and drinkers to increase their frequency of tobacco and alcohol intake, respectively.
In our study, the proportion of alcohol drinkers who increased and that of those who decreased alcohol intake with lockdown was very similar, around 15%. Compared with other surveys, the frequency of increased alcohol consumption was much lower in our population than in adults in Germany (34.7%) [bib_ref] Did the General Population in Germany Drink More Alcohol during the COVID-19..., Koopmann [/bib_ref] , Australia (20.0% [bib_ref] Alcohol use and mental health status during the first months of COVID-19..., Tran [/bib_ref] to 30.8% [bib_ref] Alcohol use in Australia during the early days of the COVID-19 pandemic:..., Neill [/bib_ref] , or the USA (29%) [bib_ref] Alcohol Consumption in Response to the COVID-19 Pandemic in the United States, Barbosa [/bib_ref] and was more similar to that reported in other Spanish [bib_ref] Physical and Psychological Effects Related to Food Habits and Lifestyle Changes Derived..., López-Moreno [/bib_ref] [bib_ref] COVID-19 Confinement and Health Risk Behaviors in Spain, López-Bueno [/bib_ref] , French [bib_ref] Behavioral Changes during COVID-19 Confinement in France: A Web-Based Study, Rossinot [/bib_ref] , and Italian surveys. In fact, in these studies, declining alcohol consumption during quarantine was more common than increasing it. Although some of the observed differences across countries may be accounted for by discrepancies in study designs (pre-post comparisons vs online surveys) and age of participants, they may also be influenced by the more social nature of drinking in Southern European countries than in other cultures. In any case, to our knowledge, ours are the first results that are entirely based on an elderly population.
Results for changes in diet quality are difficult to compare because of differences in diet assessment across studies, with some showing the proportion of those who increased or decreased the consumption of certain products [bib_ref] The assessment of lifestyle changes during the COVID-19 pandemic using a multidimensional..., Balanzá-Martínez [/bib_ref] [bib_ref] Food habits in the Galician population during confinement for COVID-19, López [/bib_ref] [bib_ref] Eating Habits and Lifestyle during COVID-19 Lockdown in the United Arab Emirates:..., Ismail [/bib_ref] , others asking participants to report their self-perceived changes in quality [bib_ref] Behavioral Changes during COVID-19 Confinement in France: A Web-Based Study, Rossinot [/bib_ref] [bib_ref] Determinants of the Lifestyle Changes during COVID-19 Pandemic in the Residents of..., Cancello [/bib_ref] , and only a few studies providing validated diet quality scores [bib_ref] Changes in Dietary Behaviours during the COVID-19 Outbreak Confinement in the Spanish..., Rodríguez-Pérez [/bib_ref] [bib_ref] Food choice motives and the nutritional quality of diet during the COVID-19..., Marty [/bib_ref] [bib_ref] Diet and physical activity during the COVID-19 lockdown period, Deschasaux-Tanguy [/bib_ref]. Even for the latter, comparisons are limited due to the use of different scales; for instance, in a Spanish study, confinement was associated with a mean increase of 0.8 points in the MEDAS score [bib_ref] Changes in Dietary Behaviours during the COVID-19 Outbreak Confinement in the Spanish..., Rodríguez-Pérez [/bib_ref] , in one French study it was associated with a mean reduction of 0.32 points in the simplified PNNS-GS2 index [bib_ref] Food choice motives and the nutritional quality of diet during the COVID-19..., Marty [/bib_ref] , while in another study it was associated with equivalent increases or decreases in the Alternative Healthy Eating Index [bib_ref] Diet and physical activity during the COVID-19 lockdown period, Deschasaux-Tanguy [/bib_ref]. Other factors compounding this comparison are the retrospective collection of pre-confinement information and the lack of focus on older adults.
Our results were somewhat expected because, during the State of Alarm, both older adults (the segment of the population that is more used to daily shopping for foodstuffs in Spain [bib_ref] Present Food Shopping Habits in the Spanish Adult Population: A Cross-Sectional Study, Achón [/bib_ref] and the whole the population inevitably had to decrease the frequency of shopping for foods that are typically consumed fresh (i.e., fruits, vegetables, or fish). In line with our results, two Spanish surveys indicated a reduction in the consumption of these fresh foods in younger adults during confinement [bib_ref] Changes in Dietary Behaviours during the COVID-19 Outbreak Confinement in the Spanish..., Rodríguez-Pérez [/bib_ref] [bib_ref] COVID-19 Confinement and Health Risk Behaviors in Spain, López-Bueno [/bib_ref]. Despite this, we observed some compensations in diet quality in older adults thanks to a decrease in the intake of foods not typical in the Mediterranean diet (i.e., red meat, sweets and cakes), or an increase in the consumption of legumes, which are typically prepared at home.
Previous surveys described mental health during confinement as an important predictor of reductions in diet quality, a fact confirmed in our study, where those with higher follow-up GDS-10 scores showed a higher risk of diet quality decline (relative risk ratio (RRR) per 1 point increase in the GDS-10: 1.25; 95%CI: 1.13-1.39). We also identified a number of baseline predictors of diet decline with confinement: social isolation (which may reflect not having someone to buy fresh foods for them), a lack of outdoor views (which is associated with lower income), the insufficiency of physical activity, or suffering with poor general health or hypertension.
Changes in weight occurring in short periods of time may be important if, as the literature on obesity suggests, they remain over time [bib_ref] COVID-19-Related Home Confinement in Adults: Weight Gain Risks and Opportunities, Bhutani [/bib_ref]. Fortunately, though, most surveys either showed no or small changes in weight with confinement [bib_ref] Physical and Psychological Effects Related to Food Habits and Lifestyle Changes Derived..., López-Moreno [/bib_ref] [bib_ref] Food habits in the Galician population during confinement for COVID-19, López [/bib_ref] [bib_ref] Weight changes during the COVID-19 home confinement. Effects on psychosocial variables, Fernandez-Rio [/bib_ref]. In our study, despite mean declines in diet quality and physical activity, only specific subgroups of older adults (mainly hypertensives and those with a negative aging experience) presented significant weight gains. Moreover, we observed that women, those with university studies, and those with short sleep and very long naps presented significant weight loss. The fact that weight loss was not associated with either baseline or follow-up (secondary analysis) overall health, psychological distress, or depression symptoms reduces the probability that it was involuntary, and thus indicative of an increased risk of frailty [bib_ref] Frailty in Older Adults: Evidence for a Phenotype, Fried [/bib_ref]. In any case, underlying changes in body composition (i.e., reduction in muscle mass and increase in fat mass) associated with the COVID-19 lockdown deserve further study in the older population.
What do other studies say about changes in physical activity and sedentary time during the COVID-19 pandemic?
Surveys on non-institutionalized adults have shown declines in physical activity and increases in sedentary time with confinement, which varied according to varying degrees and lengths of lockdown [bib_ref] The assessment of lifestyle changes during the COVID-19 pandemic using a multidimensional..., Balanzá-Martínez [/bib_ref] [bib_ref] Physical Activity Change during COVID-19 Confinement, Castañeda-Babarro [/bib_ref] [bib_ref] The Impact of the COVID-19 Confinement on the Habits of PA Practice..., García-Tascón [/bib_ref] [bib_ref] Changes in Physical Activity and Sedentary Behavior in Response to COVID-19 and..., Meyer [/bib_ref] [bib_ref] Socio-Cognitive Factors Associated With Lifestyle Changes in Response to the COVID-19 Epidemic..., Constant [/bib_ref] [bib_ref] Physical activity, dietary habits and sleep quality before and during COVID-19 lockdown:..., Martínez-De-Quel [/bib_ref] , with unhealthier changes generally observed among those with higher activity and lower sedentariness at baseline [bib_ref] Physical Activity Change during COVID-19 Confinement, Castañeda-Babarro [/bib_ref] [bib_ref] Changes in Physical Activity and Sedentary Behavior in Response to COVID-19 and..., Meyer [/bib_ref] [bib_ref] Physical activity, dietary habits and sleep quality before and during COVID-19 lockdown:..., Martínez-De-Quel [/bib_ref] and among those who had follow-up symptoms of loneliness and worse mental health [bib_ref] The mediation role of sleep quality in the association between the incidence..., Werneck [/bib_ref] [bib_ref] Physical Activity Change during COVID-19 Confinement, Castañeda-Babarro [/bib_ref] [bib_ref] Changes in Physical Activity and Sedentary Behavior in Response to COVID-19 and..., Meyer [/bib_ref] [bib_ref] Physical activity, dietary habits and sleep quality before and during COVID-19 lockdown:..., Martínez-De-Quel [/bib_ref]. In our study, we corroborated the association between previous levels of activity and sedentariness and changes in these behaviors with confinement, probably reflecting the regression to the mean phenomenon. Additionally, we first identified subpopulations of older adults who may be at increased risk of decreasing PA or increasing ST during quarantine. This information can help prioritize subgroups for health promotion or for being allowed to go outside, should future home confinements take place (i.e., the oldest old, those with worse housing conditions, or individuals with diabetes).
What do other studies say about changes in sleep duration and sleep quality during the COVID-19 pandemic?
Worse sleep quality with confinement has been reported among the adult general population in different countries [bib_ref] The assessment of lifestyle changes during the COVID-19 pandemic using a multidimensional..., Balanzá-Martínez [/bib_ref] [bib_ref] The mediation role of sleep quality in the association between the incidence..., Werneck [/bib_ref] [bib_ref] Profiles of sleep changes during the COVID-19 pandemic: Demographic, behavioural and psychological..., Robillard [/bib_ref] [bib_ref] Physical activity, dietary habits and sleep quality before and during COVID-19 lockdown:..., Martínez-De-Quel [/bib_ref] , and it is partly attributed to inactivity [bib_ref] The mediation role of sleep quality in the association between the incidence..., Werneck [/bib_ref] and higher anxiety and mental distress during isolation [bib_ref] Depression, Anxiety and stress during COVID-19: Associations with changes in physical activity,..., Stanton [/bib_ref]. In our study, the association with mental health during quarantine was confirmed, with an increased risk of reducing sleep quality in individuals with higher GDS-10 scores (RRR: 1.25; 95%CI: 1.13-1.39; data not shown in tables). Moreover, our study identified for the first time predictors of these changes, showing that older women who live alone and feel lonely, those with worse diet quality, as well as those with worse overall health and higher pain are at increased risk of developing sleep problems with confinement.
What do other studies say about changes in overall health and pain during the COVID-19 pandemic?
A quantity of evidence supports a link between confinement and decline in mental health, both in middle-aged and older adults, and provides evidence that female gender [bib_ref] Mental health consequences during the initial stage of the 2020 Coronavirus pandemic..., González-Sanguino [/bib_ref] and being younger [bib_ref] Older Adults and the Mental Health Effects of COVID-19, Vahia [/bib_ref] are risk factors of these declines. However, we have only found one study evaluating the effect of confinement on chronic pain, which showed a mild improvement of migraine features during quarantine. Now we have identified subgroups of older adults who are more vulnerable to mental health decline with confinement, including those who feel lonelier, perform less physical activity, have mobility limitations, low MMSE scores, or a negative aging experience. In addition, we found that individuals who lacked a garden/yard or who had poor sleep quality, hypertension, cardiovascular disease, cancer, depression, or mobility limitations were at increased risk of deteriorating physical function during confinement. Finally, we observed that women with short sleep, worse overall health, and lower overall life satisfaction, osteo-muscular disease, and depression, as well as those who suffered from mobility limitations were at increased risk of pain worsening.
## Study limitations and strengths
Among the limitations of the present study are that not all cohorts had information in all of the studied variables and that there was a lack of data on the longer-term (July-December) effects of isolation measures. Moreover, the study cohort did not include institutionalized older adults, so results may not be generalizable to them. However, ours is one of the very few studies that have evaluated the effects of the pandemic in non-COVID-19-infected community dwelling older adults. Moreover, unlike most of the reviewed web-based surveys, it relied on telephone interviews and used pre-pandemic inhome collected information on participants' health behaviors and health status, reducing the risk of information bias (i.e., people not accurately reporting their past exposures or symptoms).
# Conclusions
The lockdown during the first wave of COVID-19 in Spain, which was one of the most restrictive in Europe, was not associated with a deterioration in lifestyle risk factors (smok-ing, alcohol intake, diet, or weight), except for decreased physical activity and increased sedentary time, which reversed with the end of confinement. However, mental health was affected, particularly in those living alone, feeling lonelier, with mobility limitations, and with lower cognitive function. If another lockdown is imposed in this or in future pandemics, public health programs should specially address the needs of older individuals of male sex, with greater social isolation, sub-optimal housing conditions, and chronic morbidities because of their greater vulnerability to the enacted movement restrictions. CIBERFES (CB16/10/00464) and DIABFRAIL-LATAM (contract number 825546, Horizon 2020). The Elderly EXERNET multicenter study was supported by the Ministerio de Economía, Industria y Competitividad (DEP2016-78309-R (GVR)), the Ministerio de Educación y Ciencia (Red EXERNET DEP2005-00046 JAC)), the High Council of Sports (Consejo Superior de Deportes) of the Ministerio de Cultura y Deportes (09/UPB/19 (JAC) and 45/UPB/20 (JAC)), CIBERFES, the 4IE+ project (0499_4IE_PLUS_4_E (NG)) funded by the Interreg V-A España-Portugal (POCTEP) 2014-2020 program, and FEDER funds from the European Union (CB16/10/00477 (IA)). The funding agencies had no role in study design, data collection and analysis, interpretation of results, manuscript preparation or the decision to submit this manuscript for publication.
## Institutional review board
[fig] Figure 1: Baseline information collected in the study cohorts. [/fig]
[fig] Figure 2: Prospective association between baseline participant characteristics and changes in alcohol consumption or in adherence to the Mediterranean diet adherence screener (MEDAS) score during confinement. [/fig]
[fig] 48: METs h/week (95%CI: −18.61; −16.35) and 11.32 METs h/week (95%CI: −12.24; −10.41), respectively; participants in TSHA showed a mean reduction in the PASE score of 16.66 points (95%CI: −19.59; −13.73). [/fig]
[fig] Figure 3: Prospective association between baseline participant characteristics and changes in physical activity and sedentary time during confinement. [/fig]
[fig] Figure 4: Prospective association between baseline participant characteristics and changes in poor sleep quality score during confinement. [/fig]
[table] Table 1: Longitudinal changes in the prevalence of smoking, second-hand smoke, alcohol intake, diet quality, physical activity, weight, sedentary time, sleep characteristics, overall health, and pain between the COVID-19 pre-confinement (t0) and the confinement (t1) periods.n/Total Participants for Each Variable and Study CohortMain Analyses poor sleep quality indicators between the pre-confinement and confinement periods were obtained from repeated measures linear regression models with robust standard errors accounting for within-participant correlations induced by repeated measures. § Higher scores in the mental component and physical component of the SF-12 and lower scores in the WHODAS-12 are indicative of better health. [/table]
|
Impact of drains positioning on pulmonary function after coronary artery bypass grafting: an observational study
Introduction: Coronary artery bypass grafting (CABG) is a procedure associated with a decline in pulmonary function. Among the main causes is the presence of the drain that is usually positioned in the intercostal or subxiphoid region. Objective: To measure the interference of drains positioning on pulmonary function in patients undergoing CABG. Methods: Observational study that assessed preoperative pulmonary function through vital capacity (VC), maximal inspiratory pressure (MIP), maximal expiratory pressure (MEP), and peak expiratory flow (PEF). These variables were evaluated in three different moments: in the presence of two drains, when removing one, and after removing all drains. Results: We evaluated 45 patients with a mean age of 62 ± 7 years with male prevalence of 29 (64%) individuals. The insertion of drains caused a decline in pulmonary function after surgery by reducing MIP by 48%, MEP by 11%, VC by 39%, and PEF by 6%.Conclusion: This study has demonstrated that drains positioning after CABG surgery may produce weakness of the respiratory muscles, change ventilatory mechanics, and impair normal pulmonary function postoperatively.
# Introduction
Cardiovascular diseases are considered the main causes of hospitalizations and, consequently, have been increasing the mortality rate in the world. Coronary Artery Bypass Grafting (CABG) is among the most performed procedures, currently representing about 80% of surgeries performed in Brazil. [bib_ref] Pacientes Submetidos à Cirurgia Cardíaca: Caracterização Sociodemográfica, Perfil Clínico-Epidemiológico e Complicações, Dordetto [/bib_ref] Procedures that are part of the operation and can impair pulmonary function are sternotomy, cardiopulmonary bypass (CPB), grafts, mechanical ventilation (MV), and the placement of drains. [bib_ref] Qualidade de vida de pacientes submetidos à cirurgia de revascularização do miocárdio, Araújo [/bib_ref] The drains aim to minimize fluid accumulation in the pleural cavity, monitor bleeding, and prevent possible complications such as pericardial effusion, hemothorax, and tamponade. [bib_ref] Does more than a single chest tube for mediastinal drainage affect outcomes..., Le [/bib_ref] The drains are inserted in the subxiphoid and/or intercostal region, the first inserted into the costophrenic sinus directed to the base and another inserted into the intercostal space directed to the apex of the lung. [bib_ref] Intercostal versus subxiphoid approach for pleural drainage post coronary artery bypass grafting, Elnasr [/bib_ref] Although necessary, the presence of the drains can cause a change in pulmonary function, limiting the functioning of the respiratory muscles, modifying ventilatory mechanics, and generating an intense pain and causing postoperative discomfort. [bib_ref] Intercostal versus subxiphoid approach for pleural drainage post coronary artery bypass grafting, Elnasr [/bib_ref] [bib_ref] Desfechos clínicos de pacientes submetidos à cirurgia cardíaca em um hospital do..., Silveira [/bib_ref] The deleterious effects on pulmonary function are characterized by reduced vital capacity (VC), maximal inspiratory pressure (MIP), maximal expiratory pressure (MEP), peak expiratory flow (PEF), increasing ventilatory work and pulmonary compliance. Reduction of these variables may lead to complications such as hypoxemia, atelectasis, pneumonia, pleural effusion, diaphragmatic dysfunction, and prolonged hospitalization. [bib_ref] Atuação da Fisioterapia Respiratória em Complicações Pulmonares Pós-Operatórias, Vieira [/bib_ref] [bib_ref] Retirada precoce do leito no pós-operatório de cirurgia cardíaca: repercussões cardiorrespiratórias e..., Silva [/bib_ref] Previous studies addressing the effects of pleural and/or subxiphoid drainage on lung function have been performed in other clinical scenarios, or when performed on individuals undergoing cardiac surgery, these studies were somewhat limited or displayed a small sample size. Therefore, the present study aims to evaluate the influence of the drains positioning on pulmonary function in patients undergoing coronary artery bypass grafting.
# Methods
This is a prospective observational study, and it has been approved by the Research Ethics Committee of Faculdade Nobre de Feira de Santana (#2,518,734). All participants signed an informed consent form.
## Eligibility criteria
Inclusion criteria were individuals of both genders, aged 18 years and over, and submitted to coronary artery bypass grafting via median sternotomy and cardiopulmonary bypass. Exclusion criteria were patients with hemodynamic instability, chronic or acute pulmonary disease, emergency surgery, who stay more than five days in the Intensive Care Unit, difficulty in understanding or not collaborating, cardiac arrhythmia, and bronchopleural fistula.
## Study protocol
Preoperatively, patients had their pulmonary function assessed by vital capacity, inspiratory and expiratory muscle strength, and peak expiratory flow.
After these assessments, patients were referred to the operating room and later to the Intensive Care Unit (ICU). After discharge from the ICU in the postoperative period, lung function was again assessed in three moments: 1 st , with the presence of subxiphoid and intercostal drains; 2 nd , after the removal of one of the drains; 3 rd , after the removal of all drains.
Therefore, the evaluations were carried out with the patients preoperatively and postoperatively with both drains, one drain, and after the removal of both drains.
The researchers had no influence on the conduct performed by the physiotherapy team during the study period. Patients were treated according to the hospital's routine and the physical therapy approach included deep breathing exercises, cycle ergometry, kinesiotherapy, and walking. All evaluations were performed by a blinded examinee.
The subxiphoid drain was placed in the substernal region for all patients due to the risk of pericardial effusion, whereas the intercostal drain was inserted in the sixth left intercostal space in the midaxillary line. The intercostal drain was inserted whenever the internal thoracic artery was used as a graft.
## Evaluation instruments
Preoperative assessment of inspiratory muscle strength, MIP, was performed using an Indumed ® (São Paulo, Brazil) analogue manovacuometer. During the evaluation, a maximal expiration until the residual volume was requested, and then a maximal and slow inspiration to the total lung capacity was required; this test was done using the unidirectional valve method, being possible a flow through a hole of one millimeter, aiming to exclude the action of the buccinator, and repeated for three times, being used the highest value reached, as long as this value was not the last. MEP was evaluated using the same apparatus and the patient was instructed to perform a maximal inspiration until he reached his total pulmonary capacity, the mask was placed, and after that a maximum expiration was requested until the residual capacity was reached. The test was repeated three times and it was considered the highest value result, as long as this value was not the last. [bib_ref] References values for lung function tests. II. Maximal respiratory pressures and voluntary..., Neder [/bib_ref] Both tests were performed with the patient seated, lower limbs resting on the ground.
To assess VC, it was used the analogue ventilometer Ferraris Mark 8 Wright Respirometer (Louisville, Colorado, Unite States of America). The ventilometer was unlocked, cleared, and soon after the facial mask was placed on the face of the individual. The patient underwent deep inspiration until he/she reached his/her total pulmonary capacity, and soon after a slow and gradual expiration until reaching the residual volume. After this, the ventilometer was locked and the result observed and noted. The test was repeated three times, being considered the highest value result.Peak expiratory flow was evaluated using the peak flow of the Mini Wright ® brand. During the evaluation, the patient was seated, with his head in a neutral position and a nasal
# Data analysis
For data analysis, the Statiscal Package for Social Sciences 20.0 software was used. Normality was tested by the Shapiro-Wilk Test. Data were expressed as mean and standard deviation or absolute value and percentage. In order to compare pulmonary function at different times of the research, the paired Student's t-test was used. A p < 0.05 was considered for statistically significant differences.
# Results
During the study period, 55 patients were hospitalized for CABG, but 10 were excluded: two underwent emergency surgery, five were not able to understand the techniques and three displayed hemodynamic instability. Therefore, 45 patients were evaluated, being 29 (64%) male and with a mean age of 62 ± 7 years. Other clinical and surgical data are shown in [fig_ref] Table 1: Clinical data of patients undergoing coronary artery bypass grafting [/fig_ref].
We observed that all pulmonary function variables decreased after the surgical procedure. We found that the greatest decline in pulmonary function was when we compared the preoperative period with the presence of the two drains. Even after the removal of all drains, pulmonary function was not reestablished. MIP showed a 48% decrease from pre to no drain, MEP reduction 11%, VC decline 39%, and PEF 6%. These values are expressed in [fig_ref] Table 2: Evolution of pulmonary function in the presence of drains after coronary artery... [/fig_ref].
# Discussion
According to the data analyzed, we noticed that there was a decline in ventilatory muscle strength and pulmonary function in the presence of pleural and mediastinal drains after surgery, with partial recovery after drains removal. Pulmonary function was impaired by the reduction in the variables of muscle strength and vital capacity, but PEF achieved a marked fall that was directly associated with weakness of the internal and abdominal intercostal muscles, where the intercostal and subxiphoid drainage are located. [bib_ref] Intercostal versus subxiphoid approach for pleural drainage post coronary artery bypass grafting, Elnasr [/bib_ref] Decreased muscle strength at the site of insertion may be related to tissue injury and decreased blood support where the degree of muscle contraction is lower, as well as discomfort from pain. [bib_ref] Subxyphoid pleural drain confers lesser impairment in respiratory muscle strength, oxygenation and..., Cancio [/bib_ref] The drain friction with the pleural and intercostal nerve, along with the stress of the parietal pleura during breathing, justifies the intense pain of the patients contributing to the reduction of the variables and increasing the risk of respiratory complications. [bib_ref] Subxyphoid pleural drain confers lesser impairment in respiratory muscle strength, oxygenation and..., Cancio [/bib_ref] In addition, the graft using the left internal thoracic artery may result in a phrenic nerve ischemia, which anatomically innervates the diaphragm, which is one of the main breathing muscles, compromising the natural functioning of the respiratory system. [bib_ref] Intercostal versus subxiphoid approach for pleural drainage post coronary artery bypass grafting, Elnasr [/bib_ref] [bib_ref] Retirada precoce do leito no pós-operatório de cirurgia cardíaca: repercussões cardiorrespiratórias e..., Silva [/bib_ref] [bib_ref] Intervenção educativa de enfermagem ao cliente submetido à cirurgia cardíaca, Rosseto [/bib_ref] Diaphragmatic dysfunction may explain the reduction of 48% and 11% in MIP and MEP values, respectively. Just as there is resistance from the inspiratory and expiratory muscles amid the presence of drains, compromised diaphragm innervation increases the individual's effort in generating negative pressure during inspiration.
These factors tend to limit chest expansion, altering ventilatory mechanics, generating a shallow breathing pattern and, consequently, may cause deleterious effects such as tachypnea and dyspnea sensation. [bib_ref] Retirada precoce do leito no pós-operatório de cirurgia cardíaca: repercussões cardiorrespiratórias e..., Silva [/bib_ref] In the postoperative period, patients with fear of rupture of the surgical incision, expand the rib cage little, contributing to the decline in pulmonary function. [bib_ref] Estudo de revisão: A eficácia dos protocolos de fisioterapia na prevenção das..., Marques [/bib_ref] Similarly to the present findings, previous reports have demonstrated that pulmonary dysfunction is associated with the presence of intercostal and subxiphoid drains and median sternotomy. These events promote a decline in thoracic compliance, resulting in greater elastic retraction, impairing the generation of pulmonary volumes and capacities, and may present complications such as atelectasis, pneumonia, and diaphragmatic dysfunction. [bib_ref] Estudo de revisão: A eficácia dos protocolos de fisioterapia na prevenção das..., Marques [/bib_ref] [bib_ref] Avaliação da função pulmonar, força muscular respiratória e qualidade de vida no..., Medeiros [/bib_ref] Importantly, our study has shown a significant reduction in the vital capacity, which declined by 39% after surgery. This mechanism is mainly due to the deficiency of ventilatory mechanics influenced by the weakness of the respiratory muscles, sternotomy, and drains that limit the chest expansion. [bib_ref] Retirada precoce do leito no pós-operatório de cirurgia cardíaca: repercussões cardiorrespiratórias e..., Silva [/bib_ref] A recent study concluded that the positioning of the drains can reduce the forced vital capacity (FVC) and the expiratory volume in the first second (FEV1) also in patients undergoing CABG. [bib_ref] Influência da Posição do Dreno Pleural na Função Pulmonar de Pacientes Submetidos..., Vieira [/bib_ref] Although PEF was not part of the research by Vieira et al., They believed that this variable would also suffer negative changes based on the decrease in FVC. In addition to the presence of drains, the authors associated pulmonary dysfunction with chest wall edema and changes of the surfactant. [bib_ref] Influência da Posição do Dreno Pleural na Função Pulmonar de Pacientes Submetidos..., Vieira [/bib_ref] Other publications have also shown a reduction in FVC and FEV1, with an improvement in pulmonary function from the 5th postoperative day. However, authors evaluated the positioning of the intercostal and subxiphoid drain separately. [bib_ref] Sub-xyphoid pleural drain as a determinant of functional capacity and clinical results..., Guizilini [/bib_ref] [bib_ref] Pleural subxyphoid drain confers better pulmonary function and clinical outcomes in chronic..., Guizilini [/bib_ref] This procedure differs from our study, since we evaluated the positioning of the two drains on pulmonary function. The presence of drains, general anesthesia, MIP, maximum inspiratory pressure; MEP, maximum expiratory pressure; VC, vital capacity; PEF, peak expiratory flow. a p < 0.001 when comparing that moment with the preoperative period. b p < 0,01 comparing no drain with two and one drain. c p < 0,01 comparing a drain with two drains. and the use of internal thoracic artery were the factors cited to justify postoperative pulmonary dysfunction. [bib_ref] Sub-xyphoid pleural drain as a determinant of functional capacity and clinical results..., Guizilini [/bib_ref] [bib_ref] Pleural subxyphoid drain confers better pulmonary function and clinical outcomes in chronic..., Guizilini [/bib_ref] Some authors point out that the presence of drains can compromise the oxygenation and ventilation index in MV patients after CABG. [bib_ref] The effects of pleural fluid drainage on respiratory function in mechanically ventilated..., Brims [/bib_ref] [bib_ref] Influência da atuação fisioterapêutica no processo de ventilação mecânica de pacientes admitidos..., Borges [/bib_ref] Our assessments took place after discharge from the ICU but Brims et al. [bib_ref] The effects of pleural fluid drainage on respiratory function in mechanically ventilated..., Brims [/bib_ref] and Borges et al. [bib_ref] Influência da atuação fisioterapêutica no processo de ventilação mecânica de pacientes admitidos..., Borges [/bib_ref] have demonstrated drains can influence lung function even in the first moment after cardiac surgery. These rates were worse in patients with pleural effusion, impairing ventilation-perfusion, and increasing pulmonary shunt. [bib_ref] The effects of pleural fluid drainage on respiratory function in mechanically ventilated..., Brims [/bib_ref] [bib_ref] Influência da atuação fisioterapêutica no processo de ventilação mecânica de pacientes admitidos..., Borges [/bib_ref] Other complications frequently observed in the postoperative period are the reduction of oxygen pressure, carbon dioxide and oxygen saturation, these reports were observed in patients who presented a ventilatory disorder due to malfunction of respiratory muscles in the presence of drains. [bib_ref] Subxiphoid versus Intercostal Chest Tubes Comparison of Postoperative Pain and Pulmonary Morbidities..., Guden [/bib_ref] In addition, ventilatory impairment was associated with CPB, which may prolong weaning from MV due to physiological disturbance caused by the systemic inflammatory response, where through contact of blood on a non-endothelial surface occurs the activation of immunological components and the release of mediators biochemicals. [bib_ref] Effectiveness of physiotherapy in reversal of complications on myocardial revascularization, Morais [/bib_ref] [bib_ref] Fatores associados à disfunção pulmonar em pacientes revascularizados e com uso de..., Fusatto [/bib_ref] We found that as the drains were removed, pulmonary function gradually recovered, but without returning to preoperative values. We can associate this improvement with the reduction of the impact of intercostal and subxiphoid drains, to the healing of the surgical incision, and to the breathing exercises provided by physiotherapy during the entire in-hospital process.
A variable not measured in our study was the intensity of pain, especially at the place where the drains were inserted, as well as at the sternotomy, although all patients were under full analgesia effect. Studies indicate that the intensity of pain increases during respiratory movements due to the friction of the drains with the nerves and the rib cage. [bib_ref] Intercostal versus subxiphoid approach for pleural drainage post coronary artery bypass grafting, Elnasr [/bib_ref] [bib_ref] Intervenção educativa de enfermagem ao cliente submetido à cirurgia cardíaca, Rosseto [/bib_ref] [bib_ref] Subxiphoid versus Intercostal Chest Tubes Comparison of Postoperative Pain and Pulmonary Morbidities..., Guden [/bib_ref] After removal of the drains, there is a reduction in pain that helps improve pulmonary function and consequently helps patients recover. As it is not part of our research, we consider pain assessment to be a limitation of our study.
Other limitations of our study are the small sample size, increasing the risk of type 1 statistical error, secondary outcomes not assessed as postoperative complications and length of stay, observational design, and the short follow-up of these patients.
# Conclusion
The insertion of drains in the intercostal and subxiphoid space in patients undergoing coronary artery bypass grafting influenced the decline in pulmonary function, changing the ventilatory mechanics due to weakness of the respiratory muscles, which contributed to the generation of low volumes and capacities.
[table] Table 1: Clinical data of patients undergoing coronary artery bypass grafting. [/table]
[table] Table 2: Evolution of pulmonary function in the presence of drains after coronary artery bypass grafting. [/table]
|
Urinary Exosomes Diagnosis of Urological Tumors: A Systematic Review and Meta-Analysis
Purpose: Exosomes could be released directly into the urine by the urological tumoral cells, so testing urinary exosomes has great potential for non-invasive diagnosis and monitor of urological tumors. The objective of this study is to systematically review and meta-analysis of urinary exosome for urological tumors diagnosis.Materials and Methods:A systematic review of the recent English-language literature was conducted according to the PRISMA statement recommendations (CRD42021250613) using PubMed, Embase, Cochrane Library, Web of Science, and Scopus databases up to April 30, 2021. Risk-of-bias assessment was performed according to the QUADAS 2 tool. The true diagnostic value of urinary exosomes by calculating the number of true positive, false positive, true negative, and false negative, diagnoses by extracting specificity and sensitivity data from the selected literature.Results: Sixteen eligible studies enrolling 3224 patients were identified. The pooled sensitivity and specificity of urinary exosomes as a diagnostic tool in urological tumors were 83% and 88%, respectively. The area under the summary receiver operating characteristic curve was 0.92 (95% CI: 0.89-0.94). Further subgroup analyses showed that our results were stable irrespective of the urinary exosome content type and tumor type.Conclusion: Urinary exosomes may serve as novel non-invasive biomarkers for urological cancer detection. Future clinical trial designs must validate and explore their utility in treatment decision-making.Systematic Review Registration: [https://www.crd.york.ac.uk/prospero/], identifier [CRD42021250613].
# Introduction
Tissue biopsy is the current standard method for pathological diagnosis of urological cancer. However, based on one single needle biopsy is limited in reflecting the complete genomic landscape of cancer accurately and is inappropriate for early tumor screening [bib_ref] Intratumor Heterogeneity and Branched Evolution Revealed by Multiregion Sequencing, Gerlinger [/bib_ref]. To detect cell-free biomarkers (such as circulating nucleic acids, circulating tumor cells and circulating exosomes) in the body fluid, also called "Liquid biopsy", has recently show its value in clinical application [bib_ref] Liquid Biopsies, What We Do Not Know (Yet), Bardelli [/bib_ref]. Collecting the circulating tumor related gene has the potential to provide molecular characterization of primary or metastatic tumor, and these cell-free biomarkers may be used to manage the posttreatment process of tumor [bib_ref] Scalable Whole-Exome Sequencing of Cell-Free DNA Reveals High Concordance With Metastatic Tumors, Adalsteinsson [/bib_ref].
One of the main types of liquid biopsies, circulating exosome, is extracellular vesicles enclosed by a lipid bilayer membrane range from 40 to 150 nm. Exosomes contain a complex cargo of contents derived from the original cell, including nucleic acids, lipids, and proteins [bib_ref] Urinary Exosomes: The Potential for Biomarker Utility, Intercellular Signaling and Therapeutics in..., Franzen [/bib_ref]. The exosome released by tumor cells has been shown to play an important role in microenvironment, immune regulation, and other malignant processes [bib_ref] Exosome-Mediated Metastasis: Communication From a Distance, Wortzel [/bib_ref]. Compared with other tumors, urological tumors can direct release exosomes into the urine, so urinary exosomes may be more sensitive and specific to reflect the status of urological tumors [bib_ref] Isolation and Characterization of Urinary Extracellular Vesicles: Implications for Biomarker Discovery, Merchant [/bib_ref]. Since then, several studies assessing the diagnostic value of urinary exosome in urological tumor have been published [bib_ref] Urinary Extracellular Vesicle Biomarkers in Urological Cancers: From Discovery Towards Clinical Implementation, Dhondt [/bib_ref]. But the diagnostic performance of this novel biomarker has not been evaluated systematically. Therefore, the purpose of this study was to assess the diagnostic performance of urinary exosome for the detection of urological cancer including renal cancer (RCa), bladder cancer (BCa), and prostate cancer (PCa).
# Materials and methods
The protocol has been registered in the International Prospective Register of Systematic Reviews database (registration number: CRD42021250613).
## Search strategy
This systematic review and meta-analysis were performed according to the Preferred Reporting Items for Systematic Reviews and Metaanalyses (PRISMA) statement [bib_ref] The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That..., Liberati [/bib_ref]. A comprehensive literature search was followed the PRISMA 2009 checklist, and the PubMed, Embase, Cochrane Library, Web of Science, and Scopus databases were searched systematically in April 30, 2021.
The search strategy included the following terms: ("exosomes" or "extracellular vesicle") AND "urine" AND ("diagnosis" OR "biomarker") AND ("urological cancer" OR "urologic neoplasms" OR "urogenital neoplasms") AND ("kidney neoplasms" OR "kidney cancer" OR "renal cancer") AND ("prostate neoplasms" OR "prostate cancer") AND ("bladder cancer" OR "bladder neoplasms"). Two researchers (Yipeng Xu and Jianmin Lou) independently assessed the eligibility of each potentially relevant study by screening the titles and abstracts. Disagreements between the two researchers were resolved by discussion with two additional researchers (An Zhao and Zongping Wang). Other publications were identified by searching the list of references of the selected papers.
## Inclusion and exclusion criteria
Inclusion criteria for primary studies were as follows: (1) The research article was a diagnostic study using urinary exosomes;
(2) Subjects included cancer patients and healthy controls; [bib_ref] Scalable Whole-Exome Sequencing of Cell-Free DNA Reveals High Concordance With Metastatic Tumors, Adalsteinsson [/bib_ref] The data was sufficient to generate a two-by-two table consisting of true negative (TN), and false negative (FN), true positive (TP), and false positive (FP).
The exclusion criteria were as follows: (1) repeated or overlapped publications which included the same study population and genes; (2) experiments based exclusively on cell lines or tumor tissue rather than clinical samples; and (3) studies with a poor sample size (≤10).
## Data extraction and quality assessment
We extracted the following data from the selected studies: the first author's last name, year of publication, country of study, cancer type, sample sizes, exosome extraction method, type of exosome content/detection method, target molecular detection, diagnostic results (numbers of FP, FN, TP, and TN), and diagnostic performance (sensitivity and specificity).
Deek's funnel plot and Quality Assessment of Diagnostic Accuracy Studies (QUADAS) 2 tool were adopted to analyze qualitative publication bias, and a P-value of <0.05 was considered statistically significant. Risk-of-bias assessment was performed independently by two authors (YJ, YX) according to the QUADAS 2 tool. Disagreement was solved by a third party (AZ). This tool provides a measure of the risk of bias and applicability over four domains (index test, reference standard, flow, and timing) of interest [bib_ref] QUADAS-2: A Revised Tool for the Quality Assessment of Diagnostic Accuracy Studies, Whiting [/bib_ref].
## Data synthesis and analysis
All statistical analyses were performed using STATA software (version 12.0, STATA Corp, MIDAS module). Quality assessment was managed with Review Manager 5.3 (Cochrane Collaboration, Copenhagen, Denmark). The number of diagnoses (TP, TN, FP, and FN) from each study was extracted to calculate diagnostic sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR) with 95% confidence interval (CIs). PLR is calculated as sensitivity/(1-specifcity), and NLR is calculated as (1-sensitivity)/specificity. The DOR value is used as a measure of the effectiveness of a diagnostic test and is calculated as PLR/ NLR. Summary ROC curves (SROC) and AUCs of the SROC were measured. All P values were two sided, and a P value < 0.05 was considered as statistically significant.
# Results
## Literature search
Four hundred and thirty studies were confirmed through systematic search and manual review for initial screening, and 354 studies were remained after duplicates removed. After titles and abstracts were checked, 104 articles of the non-duplicate records were subjected to further full-text review, of which 88 were excluded according to the exclusion criteria. Finally, 22 studies from 16 articles were included in the present metaanalysis [bib_ref] Expression Signatures of Exosomal Long Non-Coding RNAs in Urine Serve as Novel..., Zhan [/bib_ref] [bib_ref] Urinary Exosomal Expression of Long non-Coding RNAs as Diagnostic Marker in Bladder..., Yazarlou [/bib_ref] [bib_ref] Urinary Micro-RNA Expressions and Protein Concentrations May Differentiate Bladder Cancer Patients From..., Gullu Amuran [/bib_ref] [bib_ref] Urine and Serum Exosomes as Novel Biomarkers in Detection of Bladder Cancer, Elsharkawi [/bib_ref] [bib_ref] MiR-21-5p in Urinary Extracellular Vesicles Is a Novel Biomarker of Urothelial Carcinoma, Matsuzaki [/bib_ref] [bib_ref] Urinary Exosome miR-30c-5p as a Biomarker of Clear Cell Renal Cell Carcinoma..., Song [/bib_ref] [bib_ref] Exosomal MicroRNAs Are Diagnostic Biomarkers and Can Mediate Cell-Cell Communication in Renal..., Butz [/bib_ref] [bib_ref] Expression of Small Noncoding RNAs in Urinary Exosomes Classifies Prostate Cancer Into..., Wang [/bib_ref] [bib_ref] Identification of Non-Invasive miRNAs Biomarkers for Prostate Cancer by Deep Sequencing Analysis..., Rodriguez [/bib_ref] [bib_ref] Comparative Study of Extracellular Vesicles From the Urine of Healthy Individuals and..., Bryzgunova [/bib_ref] [bib_ref] A Molecular Signature of PCA3 and ERG Exosomal RNA From Non-DRE Urine..., Donovan [/bib_ref] [bib_ref] Affinity Captured Urinary Extracellular Vesicles Provide mRNA and miRNA Biomarkers for Improved..., Davey [/bib_ref] [bib_ref] Noninvasive Prostate Cancer Detection by Measuring miRNA Variants (isomiRs) in Urine Extracellular..., Koppers-Lalic [/bib_ref] [bib_ref] Release of Urinary Extracellular Vesicles in Prostate Cancer Is Associated With Altered..., Vermassen [/bib_ref] [bib_ref] Molecular Lipid Species in Urinary Exosomes as Potential Prostate Cancer Biomarkers, Skotland [/bib_ref] [bib_ref] Exosomal Proteins as Prostate Cancer Biomarkers in Urine: From Mass Spectrometry Discovery..., Wang [/bib_ref]. No additional studies were identified via screening the bibliographies of these 16 articles. The process of literature inclusion and selection is presented in .
## Characteristics of included studies
Among them, 5 eligible studies featured a total of 408 patients with bladder cancer, 9 eligible studies featured a total of 1277 patients with prostate cancer, and 2 eligible studies featured a total of 179 patients with renal cell carcinoma. The main extraction methods of urinary exosome are ultracentrifugation or commercial exosome extraction kit. The technique for molecular examination depends on the type of exosome contents, nucleic acid exosome cargo was detected using methods such as qRT-PCR or sequencing, and non-nucleic acid exosomal cargo (proteins or lipids) was detected using methods such as enzyme-linked immunosorbent assay (ELISA) or mass spectrometry (MS). In total, all main characteristics of the eligible studies were summarized [fig_ref] TABLE 1 |: Characteristics of studies evaluating the urinary exosomes of patients with urological tumor [/fig_ref].
## Risk of bias within studies
The quality of the selected studies was evaluated in accordance with the QUADAS-2 criteria; the results of these evaluations are shown in [fig_ref] FIGURE 2 |: Grouped bar charts show risk of bias and concerns for applicability of... [/fig_ref]. Five studies were considered to be low-risk with regards to bias and applicability, and the other 11 studies were estimated as suboptimal for unclear risk in several areas, including patient selection, reference standards, and index testing. Deek's funnel plot was also used to evaluate the publication bias of included studied, and no publication bias was found (P = 0.81) (Supplementary .
In addition, meta-regression analyses were performed to analyze the heterogeneity with the potential variables, and the type of exosome content (nucleic acid/non-nucleic acid), the type of urological cancer (BCa/PCa/RCa), and proportion of patients with urological cancer (>50%/≤50%) were not significant factors affecting the heterogeneity (P > 0.05, Supplementary [fig_ref] TABLE 1 |: Characteristics of studies evaluating the urinary exosomes of patients with urological tumor [/fig_ref].
## Meta analysis of diagnostic value
All 22 eligible studies were used to evaluate the diagnostic accuracy between urinary exosome expression and urological tumors. As shown in [fig_ref] FIGURE 3 |: Coupled forest plots of pooled sensitivity and specificity [/fig_ref] , the overall diagnostic sensitivity and specificity were 0.83 (95% CI, 0.78-0.88) and 0.88 (95% CI, 0.81-0.92), respectively. Urinary exosome was significantly correlated with sensitivity (P < 0.01, I 2 = 87.89%) and specificity (P < 0.01, I 2 = 92.10%) [fig_ref] FIGURE 3 |: Coupled forest plots of pooled sensitivity and specificity [/fig_ref]. The area under the SROC curve was 0.92 (95% CI: 0.89-0.94) [fig_ref] FIGURE 4 |: Hierarchical summary receiver operating characteristic curve of the diagnostic performance of urinary... [/fig_ref]. The pooled PLR was 6.94 (95% CI: 4.29-11.22), and the pooled NLR was 0.19 (95% CI: 0.14-0.26) through random effect model (Supplementary [fig_ref] FIGURE 2 |: Grouped bar charts show risk of bias and concerns for applicability of... [/fig_ref].
# Subgroup analysis
When the studies were separately assessed according to the type of exosome content, nucleic acid analysis group of 12 studies
# Discussion
RCa, BCa, and PCa are the main types of urological tumors; their morbidity and mortality rates have continued to rise in recent years [bib_ref] Cancer Statistics 2020, Siegel [/bib_ref]. Although prostate-specific antigen (PSA) testing has been used as biomarker in prostate cancer diagnosis, prostate biopsies are still essential to make a definite diagnosis since PSA level is low, and it also leads to overdiagnosis and overtreatment [bib_ref] A Panel of Kallikrein Markers Can Reduce Unnecessary Biopsy for Prostate Cancer:..., Vickers [/bib_ref] [bib_ref] Biomarkers in Localized Prostate Cancer, Ferro [/bib_ref]. Most RCas are still found during other abdominal tests [bib_ref] Renal Cancer Biomarkers: The Promise of Personalized Care, Vasudev [/bib_ref]. Although the targeted therapy and immunotherapy have become the main treatment for advanced RCa, the complete responses is still low, and the biomarker-based strategies are still missing [bib_ref] Adjuvant Therapy in Renal Cell Carcinoma: Current Knowledges and Future Perspectives, Larroquette [/bib_ref]. Urological tumors still lack the key targeted markers such as epidermal growth factor receptor (EGFR) for lung cancer and human epidermal growth factor receptor 2 (HER2) for breast cancer. Urinary cytology was one kind of the main non-invasive diagnostic methods for urothelial cancers (including bladder cancer, renal pelvis cancer, ureteral cancer, and urethral cancer), but its sensitivity was proved deficient (7-17%), and its diagnostic accuracy for low-grade urothelial cancer was relatively low [bib_ref] Current Bladder Tumor Tests: Does Their Projected Utility Fulfill Clinical Necessity?, Lokeshwar [/bib_ref]. Compared to shedded tumor cells which are harder to capture in urine, exosomes are continually released into the urine from tumor cells. Exosomes can carry antigens from tumor-derived cells, so tumor-related exosomes can be purified by tumor antigen-bound magnetic beads to improve diagnostic specificity. Moreover, the nucleic acid cargo in exosomes may directly reflect the molecular characteristics of urological tumors. In addition, the concentration of exosome-related proteins in the first-morning urination and the second-morning urination were quite similar, and the exosomes remain intact during long-term storage or at -80°C (32), suggesting that urinary exosomes were stable enough to be examined their nucleic acid or non-nucleic acid cargo.
Urine is easy to obtain and has the advantages of convenience, non-invasive, and repeatability. To systematically evaluate the potential of urinary exosomes as non-invasive markers for urological tumors, we established a meta-analysis including 22 studies from 16 articles with 3224 patients and 1360 healthy controls; the results showed an advanced diagnostic accuracy of urinary exosomes with an AUC of 0.92, a sensitivity of 83%, and a specificity of 88%. The overall PLR value of urological exosome was 6.94, suggesting that the probability of having tumor in a people with a positive test was approximately 7-fold higher than negative controls. Several laboratories including ours have reported some over-expressed proteins in tumor tissues, which are valuable in predicting the prognosis of the urological cancer [bib_ref] Prognostic and Therapeutic Role of HER2 Expression in Micropapillary Carcinoma of the..., Sanguedolce [/bib_ref] [bib_ref] PTX3 Modulates the Immunoflogosis in Tumor Microenvironment and Is a Prognostic Factor..., Netti [/bib_ref] [bib_ref] Prognostic Value of Serum CA9 in Patients With Metastatic Clear Cell Renal..., Gigante [/bib_ref]. Whether these biomarkers can be detected in urinary exosomes and the use of urinary exosomes for monitoring tumor recurrence are worthy of further investigation.
This meta-analysis study suggests the urinary exosomes may serve as non-invasive biomarkers for urological cancer diagnosis. Several limitations of this study need to be discussed. We also reviewed the study of urinary exosomes in other urological tumors (such as ureteral cancer, renal pelvis cancer, epididymal tumor, and testis pellet cancer), but no relevant results were found. Thus, there is still a lack of relevant studies for some urological tumors with low incidences. Because of the large number of included studies reporting positive results, it is impossible to rule out the possibility of selection bias. The potential variables, including the type of exosome content, the type of urological cancer, and proportion of patients with urological cancer were not significant factors affecting the heterogeneity, but whether other factors (such as primers, kits, and quantitative methods) can contribute to bias remains to be evaluated with the enough data.
# Conclusion
Urinary exosomes has great application potential in the noninvasive diagnosis and monitoring of urological tumors. Future evolutions will be necessary to validate whether urinary exosomes may serve as a potential non-invasive marker for early diagnosis and treatment response.
# Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.
# Author contributions
Two researchers YX and JL independently assessed the eligibility of each potential study by screening the titles and abstracts. Any disagreements between the two researchers were resolved by discussion with two additional researchers HW and ZW. The manuscript was written by YX, YJ, and HX, and AZ, YX, and GL revised. JL, MY, YG, and YH were responsible for the statistical analysis. All authors contributed to the article and approved the submitted version.
[fig] FIGURE 2 |: Grouped bar charts show risk of bias and concerns for applicability of 22 included studies using QUADAS-2. QUADAS-2, Quality Assessment of Diagnostic Accuracy Studies-2. [/fig]
[fig] FIGURE 3 |: Coupled forest plots of pooled sensitivity and specificity. Numbers are pooled estimates with 95% CI in parentheses. Corresponding heterogeneity statistics are provided at bottom right corners. Horizontal lines indicate 95% CIs. CI, confidence intervals. [/fig]
[fig] FIGURE 4 |: Hierarchical summary receiver operating characteristic curve of the diagnostic performance of urinary exosomes for detecting urological tumor. [/fig]
[fig] FUNDINGFIGURE 5 |: This work was supported by the National Natural Science Foundation of China (Reference number: 81402117), Natural Science Foundation of Zhejiang Province (Reference number: LY17H160043), and Medical and Health Project of Zhejiang Province (Reference number: 2016KYA038).SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fonc.2021. 734587/full#supplementary-material A B Coupled forest plots of pooled sensitivity and specificity in the subgroup. Numbers are pooled estimates with 95% CI in parentheses. Corresponding heterogeneity statistics are provided at bottom right corners. Horizontal lines indicate 95% CIs. CI, confidence intervals. [/fig]
[table] TABLE 1 |: Characteristics of studies evaluating the urinary exosomes of patients with urological tumor. Urine Exosome RNA Isolation Kit (Norgen Biotek, Thorold, Canada) [/table]
|
Sentinel health events of environmental contamination: a consensus statement.
## Data sources for monitoring shes
The panel identified a number of data sources already in operation that could be used for monitoring health conditions potentially related to environmental exposures. The extent that type I SHEs would be represented among such health condition is simply a matter of defining the type I SHEs to be considered. These sources of data for general surveillance, as well as SHE monitoring (of either type), include death certificates, cancer registries, hospital/outpatient records, laboratory data (10), birth defects registries [bib_ref] Risk of congenital malformations associated with proximity to hazardous waste sites, Geshwind [/bib_ref] , poisoning centers, and animal necropsy data bases. These sources were considered to have the best potential use in identifying sentinel health events associated with environmental exposure. The strengths, limitations, and applications of these sources were discussed by the panel in a context with the manner by which health events could come to the attention of an alert health care provider [bib_ref] The occupational hazards of formulating contraceptives-a survey of plant employees, Harrington [/bib_ref] [bib_ref] Hypervitaminosis D associated with drinking milk, Jacobus [/bib_ref]. The United States has sometimes been criticized for having a medical care delivery system that impedes public health surveillance and detection of SHEs. This is because no unique universal identifiers exist, thus records about health events are fragmented across the health care delivery system, and record linkage studies are difficult to undertake. The alert provider is usually the person who detects type I SHEs. In infectious disease outbreaks (e.g., pneumocystis pneumonia), the role of the alert clinician is well .-IUet demonstrated. The alert practitioner also often is the primary identifier of type II problems; for example, in Puerto Rico enlarged breasts in offspring of factory workers exposed to hormones. Such reports as these come from practitioners and from the general public. Public health agencies must begin to be more responsive with investigating type I SHEs and should deliberately search for type II SHEs by systematically examining the available databases.
## Dissemination of she technology
Modifying data systems to look for SHEs will require resources, as well as a reorientation in thinking by many public health workers [bib_ref] Sentinel event strategies in environmental health, Aldrich [/bib_ref]. In public health the emphasis is on common, preventable diseases, and most epidemiologists and public health workers are not taught to look for rare, improbable events. Public health funding is not currently available for examining rare events or the associated occurrence patterns. Technology is making processing and examining large quantities of data faster and more routine. However, even with easy to use, sophisticated, and efficient computer systems, systematic examination for -Oregon, rare and unusual events will still take material resources. It is expected that the primary health care provider, whether nurse, physician, dentist, veterinarian, or other, will be the primary identifier of SHEs. SHEs should be brought to the attention of county health departments, state health departments, or the federal government through the state health departments. Clinical practitioners could be apprised of these candidate sentinel events and explanations of what represents an unusual health event pattern in the same ways that the original occupational modes of SHEs were publicized (e.g., publication in medical journals, continuing medical education courses).
Particular committees of the American Academy of Pediatrics and of the American Medical Association concern themselves with the environment and could provide leadership for the use of SHE reasoning. National Institute of Health-funded centers for environmental health train medical residents and offer courses through which information regarding SHE surveillance may be distributed.
# Discussion
The alert health care provider will frequently be the first and most critical element in detecting SHEs, especially type I. This is particularly true because obtaining morbidity data with high coverage in the United States through existing data sources is limited. These providers need to be educated about their role in detecting SHEs.
It is difficult to find examples of SHEs detected through analysis of a database. However, most investigations rely on a partnership between the practitioner and a public health agency. The practitioner may identify the first indication of a problem, but the agency is needed to evaluate, confirm, and determine the scope of the problem. For example, with the problem of vitamin D in Massachusetts, three cases were found, but the public health agency identified more cases and did investigative work [bib_ref] Hypervitaminosis D associated with drinking milk, Jacobus [/bib_ref]. Public health agencies should play a greater role in the study of type I SHEs and should search for type II SHEs by examining available databases. When local public health agencies receive reports about type I SHE occurrences, they should be evaluated by familiar methods of confirming diagnosis, comparing rates, etc. [bib_ref] Case investigation in epidemiology, Armenian [/bib_ref].
Type II SHEs may be detected through existing surveillance systems. New technologies have made traditionally time-consuming and laborious activities easier (e.g., mapping disease cases). ATSDR has recently produced a software package expressly to facilitate these types of statistical analyses [bib_ref] CLUSTER: users manual for software to assist with investigations of rare health..., Aldrich [/bib_ref]. There are many innovative and exciting uses of geographic information systems for health applications that may facilitate identification of type II SHEs. More innovations are likely to become available in the future. Research attention is needed for ways of setting priorities and handling reports of type II SHEs. |
ACCEPTANCE: protocol for a feasibility study of a multicomponent physical activity intervention following treatment for cervical cancer
[bib_ref] Substantial Improvement in UK Cervical Cancer Survival with Chemoradiotherapy: Results of a..., Vale [/bib_ref]
# Introduction
Advancements in the clinical management of cervical cancer have resulted in improved survival, leading to a change in focus to treatment-related sequelae and survivorship issues, to reduce iatrogenic morbidity and to improve patients' long-term health. [bib_ref] Iatrogenic menopause after treatment for cervical cancer, Moss [/bib_ref] Although many patients diagnosed with a cervical cancer have early stage disease (stage IA), which can be successfully treated with local excision, 64% of women have at least stage IB disease at diagnosis.As a result, many patients will need to undergo radical surgery and/or chemoradiotherapy, that consequentially leads to the loss of fertility and an early menopause. [bib_ref] Iatrogenic menopause after treatment for cervical cancer, Moss [/bib_ref] The long-term effects of radiotherapy on the vagina, 5 bladder 6 and bowel can occur in between 30% and 40% of patients and can result in distressing symptoms, urinary/faecal such as incontinence 7 and tenesmus. [bib_ref] Self-Assessment of morbidity following radical hysterectomy for cervical cancer, Butler-Manuel [/bib_ref] It can also lead to permanent physical changes, vaginal shortening/stenosis and reduced clitoral sensitivity, thereby physically limiting the ability to have and enjoy intercourse. [bib_ref] Longitudinal study of sexual function and vaginal changes after radiotherapy for cervical..., Jensen [/bib_ref] Chronic fatigue and pelvic insufficiency fractures are also frequently reported following radiotherapy treatment for cervical cancer, the symptoms of which may not diminish with time, [bib_ref] Treatment complications among long-term survivors of cervical cancer: treated by surgery or..., Elghamrawi [/bib_ref] and can have a negative impact on daily functions and social quality of life. [bib_ref] Sexual dysfunction and treatment for early stage cervical cancer, Schover [/bib_ref] Strengths and limitations of this study ► The integration of behaviour change theory and insight from the target population to inform the development of practical strategies ensures an appropriate and user-friendly intervention. ► A mixed methods process evaluation will provide a great depth of understanding relating to the experiences and acceptability of the intervention and will enhance subsequent refinement where necessary. ► As this is a feasibility study, the analysis will not report statistically powered differences in outcomes (e.g., device-assessed PA).
## Open access
There is also substantial evidence of the devastating psychological consequences of a cervical cancer diagnosis. A high proportion of patients experience longterm psychological distress following their diagnosis, [bib_ref] Quality of life in longterm cervical cancer survivors: a population-based study, Borgne [/bib_ref] which can impact on their social, work and functional well-being. The psychological impact is more pronounced in younger patients and those who receive chemoradiotherapy, 13 with reported lower levels of self-confidence and a more negative body image compared with those who undergo surgery alone. 14 Hence, a greater focus on cervical cancer survivorship is warranted to support patients who are living alongside the physical, psychological and social impacts of cervical cancer treatments.
Identifying and implementing ways to manage treatment side-effects can be challenging, given the individualistic and complex nature of the survivorship experience. [bib_ref] Life beyond cancer: women's experiences 5 years after treatment for gynaecological cancer, Sekse [/bib_ref] Moreover, concerns have been raised regarding the knowledge provision of treatment-related morbidity and clinical support services. [bib_ref] Unmet needs for information and psychosocial support in relation to quality of..., Faller [/bib_ref] Physical activity (PA), defined as 'any bodily movement produced by skeletal muscles that results in energy expenditure' ( 17 p126) has been shown to be a safe avenue to improve quality of life following cancer treatment, [bib_ref] Effects and moderators of exercise on quality of life and physical function..., Buffart [/bib_ref] with qualitative data indicating that PA can specifically improve physical, psychological, social and spiritual domains. [bib_ref] Physical activity and quality of life in cancer survivors: a meta-synthesis of..., Burke [/bib_ref] However, PA levels are reportedly low and remain low up to 3 years following treatment for a gynaecological malignancy. [bib_ref] Physical activity trajectories following gynecological cancer: results from a prospective, longitudinal cohort..., Fleming [/bib_ref] In particular, cervical cancer survivors meet recommended PA guidelines less frequently than patients following other cancer types. [bib_ref] Accelerometer-derived physical activity and sedentary time by cancer type in the United..., Thraen-Borowski [/bib_ref] Exercise recommendation for cancer survivors have been suggested [bib_ref] Exercise guidelines for cancer survivors: consensus statement from international multidisciplinary roundtable, Campbell [/bib_ref] ; however, these do not target the specific PA barriers faced after cervical cancer, thus supporting the need for tailored PA promotion in this population.
To date, there has been little research focus on promoting PA among those treated for cervical cancer. This is highlighted by cervical cancer cases being underrepresented in meta-analysis studies of gynaecological cancer. [bib_ref] Effects and moderators of exercise on quality of life and physical function..., Buffart [/bib_ref] Literature based in the broader context of gynaecological cancer and PA does provide some degree of insight regarding PA preferences and feasibility of different types of PA. Such studies suggest that PA which: incorporates home-based walking, [bib_ref] A randomised controlled trial testing the feasibility and efficacy of a physical..., Donnelly [/bib_ref] [bib_ref] A focus group study exploring gynecological cancer survivors' experiences and perceptions of..., Donnelly [/bib_ref] [bib_ref] Physical activity and lower limb lymphedema among uterine cancer survivors, Brown [/bib_ref] and gym activities [bib_ref] Physical activity and lower limb lymphedema among uterine cancer survivors, Brown [/bib_ref] ; is graded and flexible 24 ; is combined with a form of support or counselling [bib_ref] A focus group study exploring gynecological cancer survivors' experiences and perceptions of..., Donnelly [/bib_ref] ; includes social interaction with peers, 26 tends to be favourable among survivors of gynaecological cancers. There is growing support for technology-based PA promotion among cancer survivors, [bib_ref] Acceptability and feasibility of a Fitbit physical activity monitor for endometrial cancer..., Rossi [/bib_ref] which can place less burden on participants, and can be easily tailored to the individual. However, the PA barriers faced after cervical cancer may differ to other gynaecological cancers due to variations in patient characteristics (e.g., age) and treatment side effects. Therefore, it is important to consider the specific needs of cervical cancer survivors during intervention design and implementation.
To the best of our knowledge, an intervention which promotes PA participation following treatment for cervical cancer has not been evaluated. Therefore, we aim to test the feasibility and acceptability of conducting and evaluating a multicomponent intervention programme, informed by relevant theories of behaviour change, to increase PA levels following treatment for cervical cancer. This intervention has been named: Acceptability in Cervical Cancer of an Exercise-based Programme delivered Through An oNline Community Environment (ACCEPTANCE). This intervention was developed using the intervention mapping protocol.
## Aim
The overall aim of this study is to assess the feasibility and acceptability of a PA intervention following treatment for cervical cancer.
## Methods and analysis study design
This is a single-arm pre study and post study to assess the feasibility and acceptability of conducting and evaluating a PA intervention following treatment for cervical cancer.
A process evaluation will also be conducted throughout the study.
## Study setting
This study will be conducted through the University Hospitals of Leicester (UHL) National Health Service Trust, in collaboration with Loughborough University and the University of Leicester.
## Patient and public involvement (ppi)
PPI consisted of two groups of women treated for cervical cancer within the previous 10 years. PPI input has informed the development of the study since its inception. Insights and feedback gathered from one to one and group discussions with PPI members enabled development of the research question, assessment of the potential burden on participants and informed the intervention content to ensure that intervention materials were user friendly [fig_ref] Table 1: Patient and public involvement [/fig_ref]. [bib_ref] Increasing physical activity levels following treatment for cervical cancer: an intervention mapping..., Millet [/bib_ref] Continuation of PPI involvement will inform the intervention delivery and dissemination of results to the wider community as appropriate.
## Participants and recruitment
Participants between the ages of 18 and 60 years who have undergone treatment for cervical cancer at least 6 months previously with curative intent (either surgery, chemoradiotherapy or both modalities), who do not meet the national PA guidelines (150-300 min of moderate to vigorous intensity PA per week),who can gain permission from their gynaecologist/oncologist to take part and are proficient in speaking English will be recruited. The age range of 18-60 years was chosen to reflect the Chief Medical Officer's PA adult guidelines and to optimise engagement and adherence to a technology-based PA programme. Exclusion criteria are benign or premalignant (cervical interepithelial neoplasia) disease of the cervix; clinical/radiological evidence of disseminated malignancy; pregnancy or breast feeding; performance status ≥3; comorbidity that in the opinion of the patient's supervising gynaecologist/oncologist would preclude the patient from meeting the study requirements. Initial database screening will take place by the clinical gynaecological oncology team to identify candidates who are at an appropriate age and stage in their treatment journey. Potentially eligible participants will be approached through their clinical team; recruitment posters will be placed in the gynaecological and oncology outpatient departments at UHL; and study details (overview of research, inclusion criteria and relevant contact details) will be posted on social media platforms. Local cervical cancer support groups will also be contacted and given information on the study to raise awareness of the research. Screening will take place via a telephone call with a member of the research team. The Scottish Physical Activity Questionnaire 31 will be used to determine PA levels. Participants will be asked how many minutes per week they spend being active at a moderate intensity in the context of the last 7 days. Consent procedures will take place either virtually via video/telephone call or in person at UHL.
## Sample size
This study will aim to recruit approximately 30 participants. This sample size was chosen as it is a realistic target when considering cervical cancer incidence both in the UK and at UHL where recruitment will take place over the given time period. In line with recommendations from the National Institute for Health Research, a sample size of 30 is appropriate to answer the questions posed by a feasibility trial. [bib_ref] Design and analysis of pilot studies: recommendations for good practice, Lancaster [/bib_ref] The feasibility of lifestyle and PA interventions in endometrial and breast cancer survivors have successfully employed similar sample sizes. Intervention The aim of the multicomponent intervention is to increase PA levels of the target population, specifically, individual and group-based walking over 12 consecutive weeks (figure 1). [fig_ref] Table 2: Description of intervention components [/fig_ref] provides a description of each intervention component while [fig_ref] Figure 2: Logic model for the ACCEPTANCE trial [/fig_ref] outlines the Logic Model. Walking has been identified as a common PA preference among gynaecological cancer survivors, [bib_ref] Physical activity correlates, barriers, and preferences for women with gynecological cancer, Farrokhzadi [/bib_ref] particularly for those who do not meet the recommended PA guidelines. [bib_ref] A population-based study of the determinants of physical activity in ovarian cancer..., Stevinson [/bib_ref] The intervention components are underpinned by social cognitive theory,and are informed by the health belief model,and theories of self-regulation.Selfmonitoring behaviour is strongly linked with successful PA behaviour change, particularly when combined with at least one other self-regulatory component. [bib_ref] Effective techniques in healthy eating and physical activity interventions: a meta-regression, Michie [/bib_ref] The intervention includes education provision, problem-solving and goal setting in relation to increasing PA levels after treatment for cervical cancer. Self-monitoring of PA behaviour will be facilitated by providing participants with an intervention diary to complete daily and weekly, fortnightly online health coaching via video or telephone call administered by the study coordinator and a consumer PA monitor (fitbit inspire 2) to receive real-time feedback on their activity levels and prompts to be more active and to review goals. Throughout, peer support and group walking among participants will be encouraged via a messaging platform, which will allow participants to maintain contact and organise group walking sessions. The study coordinator will suggest appropriate walk locations convenient to each group.
## Measures
All PA and health measures will be assessed at baseline, week 12 and 3 months after completion (week 24). Previous literature suggests that psychological measures may take longer to change than physical measures (e.g., menopausal symptoms or PA) and therefore to reduce participant burden, only these will be taken at week 6 [fig_ref] Table 3: PA and health outcomes, outcome measures and measurement time points [/fig_ref].
## Pa and health measures
An accelerometer (GENEActiv, Activinsights) and sleep log will be administered at all evaluation time points to measure PA. Participants will be asked to wear the accelerometer on their non-dominant wrist 24/7 for 8 days at each time point. Valid data will be at least 3 days >16 hours of accelerometer data collected. [bib_ref] Promotion of physical activity through structured education with differing levels of ongoing..., Yates [/bib_ref] Device-assessed PA will be the primary research outcome measure of interest in the subsequent pilot study and definitive main trial. Participants' belief in their ability to successfully complete incremental 5-50 min periods of walking at a moderate/brisk pace will be assessed via the self-efficacy for walking scale. [bib_ref] Exercise environment, selfefficacy, and affective responses to acute exercise in older adults, Mcauley [/bib_ref] Motivation for PA will be assessed using the Behavioural Regulation in Exercise Questionnaire. [bib_ref] A modification to the behavioural regulation in exercise questionnaire to include an..., Markland [/bib_ref] This is a 23-item scale which assesses motivation on a self-determination continuum. Enjoyment of PA will also be assessed via the original 18-item PACES scale. [bib_ref] Physical activity enjoyment scale: two validation studies, Kendzierski [/bib_ref] In relation to well-being, menopausal symptoms, such as hot flushes and sweating will be investigated using the Menopausal Rating Scale.Quality of life will be assessed using the EORTC QLQ-30, which is commonly used in gynaecological cancer populations and includes five functional scales, three symptom scales, a global health scale and a quality of life scale. The Hospital Anxiety and Depression scale (HADS), [bib_ref] The hospital anxiety and depression scale, Zigmond [/bib_ref] will be administered to measure symptoms of anxiety and depression. Fatigue will be assessed using the Multi-dimensional Fatigue Symptom Inventory (MFSI-SF), [bib_ref] Further validation of the multidimensional fatigue symptom inventory-short form, Stein [/bib_ref] a measure which has been previously used to assess fatigue in a gynaecological cancer population. [bib_ref] A randomised controlled trial testing the feasibility and efficacy of a physical..., Donnelly [/bib_ref] Evaluation-related feasibility measures Participant recruitment rates will be monitored by recording the number of identified potential participants, the number of those who express an interest in taking part, the number of those who are eligible and not eligible along with the reasons for ineligibility and the number who consent to participate. Retention and attrition rates will be monitored by recording the number of participants who withdraw from the study and the number who do not attend follow-up. Where possible, reasons for withdrawal will also be reported. The number of participants who complete the PA and health measures at each time point will be summarised.
## Process evaluation
The process evaluation will be conducted to explore participant experiences of the study and intervention as well as compliance with the intervention components, to identify facilitators and barriers to increasing PA and any suggestions for improvement and refinement. Compliance with intervention components will be reviewed throughout the intervention. Data will be presented on launch attendance rates, the number of diary entries completed, number of daily step counts recorded to reflect compliance of wearing the PA monitor, number of health coaching sessions attended, the number of participants who post on the online forum and engage in group Open access messaging, the number of participants who take part in group walks and the frequency of these.
Questionnaires will be administered to understand participants' experience of attending the intervention launch, of using the PA monitor and the group messaging. A semi-structured interview will also be conducted by the study coordinator to further explore perceptions of the intervention and the evaluation measures, perceived benefits of participation, maintenance of the intervention components and PA levels. Interview data will be analysed using Template Analysis, [bib_ref] Doing Template Analysis, King [/bib_ref] employing a deductive coding framework to inform study refinement. To ensure a balanced insight and that subsequent meaningful refinement can be implemented, some transcripts will be coded by an independent researcher who did not contribute to the intervention design.
A researcher log will be completed by the study coordinator where a record of interactions with participants (e.g., email conversations/telephone conversations) will be kept throughout the duration of the study. Feedback will be broadly divided into four categories: evaluation of intervention components; experience of evaluation measures; alterations to future intervention implementation; barriers to PA and reasons for discontinuation (where relevant).
## Data collection
The accelerometer and the quantitative measures will be posted to participants. Process evaluation questionnaires will be completed online throughout the intervention. The evaluation interview will be offered after the week-24 time point to all participants (including those who have withdrawn) either in person or via video call. Goal Setting ► Introduction to SMART goals ► Using PA monitor to set goals related to steps and intensity *Each section divided by guided 'stand up and stretch' prompts Self-monitoring of PA ► The PA monitor will provide participants with feedback on their activity levels via: number of steps taken by participants and the number of minutes that participants spend in a moderate intensity activity each day Prompts ► Prompts offered by PA monitor to encourage PA ► Prompts to update participant on their progress in reaching their daily goal
## Intervention diary (monitoring of well-being)
Daily entries ► Input daily step count as shown on PA monitor ► Rate mood out of 10 Weekly entries ► Rate the following symptoms on a traffic light system: energy levels, anxiety levels, body confidence, physical pain, bladder issues
Health coaching ► Revision of previous goal(s) and the outcome (successful/ unsuccessful); attribute reasons for this outcome ► Identification of challenges and facilitators regarding previous goal ► Revision of steps, intensity and well-being over the previous 2 weeks ► Facilitated goal setting and identification of potential challenges for the following 2 weeks Group messaging ► To be used as a platform to schedule group walks (organising a time and a place) with the aim of one group walk per week. Standardised communication prompts to be offered by the study coordinator Online community ► Participants will be encouraged to use the study-specific community on the PA monitor application, accessed via smart phone where personal PA insights and statistics can be shared with other participants PA, physical activity.
## Open access
# Statistical analysis
As this is a feasibility study, main analyses will use descriptive statistics. Data will be analysed using IBM SPSS statistics V. [bib_ref] Physical activity and lower limb lymphedema among uterine cancer survivors, Brown [/bib_ref].
## The feasibility of inclusion criteria
We have suggested that if an eligibility criterion accounts for 40% of ineligibility, it will be reported.
The feasibility of study recruitment and retention A priori, we have defined a recruitment success rate as recruitment of 30 participants in the first 3 months or that 75% of those identified as eligible are recruited, and a successful retention rate at week-24 as 70%.
The feasibility of and compliance with intervention components Attendance rates at the intervention launch, engagement with the online forum, frequency of wearing the PA monitor, frequency of group walking sessions, compliance with diary completion and attendance rates at health coaching sessions will all be summarised. The threshold of feasibility for these rates will be 70% of participants complying, respectively.
The feasibility of the evaluation measures Compliance with evaluation measures will be deemed successful if no more than 20% of participants fail to provide questionnaire and accelerometer data at both baseline and week-24 follow-up. Open access PA data and questionnaire outcomes Data from the accelerometer will be cleaned and processed in GGIR using R programming and analysed using IBM SPSS statistics V. [bib_ref] Physical activity and lower limb lymphedema among uterine cancer survivors, Brown [/bib_ref]. Means and standard deviations (SDs) will be used to describe the PA from the accelerometer at each evaluation time point. Constructs measured via questionnaire will be analysed by standard scoring procedures. Questionnaire outcomes of interest will be reported on means (or medians where necessary) and SDs.
## Ethics and dissemination
Ethical approval has been granted by the West of Scotland Research Ethics Committee 1 for this study. Results will primarily be used to refine intervention components and to inform a larger full-scale trial, if appropriate. Results from the study will be shared in peer-reviewed journals, presented at international conferences and will be shared with the UHL and Leicester Hospital's charity.
# Discussion
This article describes the protocol for a study testing the feasibility and acceptability of conducting and evaluating a multicomponent PA intervention following treatment for cervical cancer. Regular PA has been shown to have physical and psychological benefits and to be safe after cancer treatment. [bib_ref] Interventions for promoting habitual exercise in people living with and beyond cancer, Turner [/bib_ref] PA can also alleviate a range of symptoms experienced as a result of cancer treatment . [bib_ref] A randomised controlled trial testing the feasibility and efficacy of a physical..., Donnelly [/bib_ref] PA levels appear to be low following cervical cancer [bib_ref] Accelerometer-derived physical activity and sedentary time by cancer type in the United..., Thraen-Borowski [/bib_ref] and this may be partly due to the lack of a specific intervention that takes into consideration the particular side-effects that can occur, in particular pelvic pain and incontinence, that may not be such an issue in other cancer types.
The study protocol details a 12-week intervention programme to promote walking and PA, with evaluation measures taken at baseline, 6 weeks, 12 weeks and 24 weeks. The process evaluation is a mix of questionnaires and a qualitative interview. The intervention was designed using the intervention mapping protocol.Development followed a systematic six-step framework, which prioritises input from the target population as well as integrating behaviour change theories (BCTs) and constructs to underpin change within the intervention, informed by relevant literature. [bib_ref] Predicting physical activity among cancer survivors: meta-analytic path modeling of longitudinal studies, Hirschey [/bib_ref] Interventions underpinned by relevant BCTs have been found to be more successful in implementing change and in identifying more specific evaluation criteria. [bib_ref] Interventions for promoting habitual exercise in people living with and beyond cancer, Turner [/bib_ref] The current PA guidelines offered by the Government Chief Medical Officers do not specify recommendations for cervical cancer survivors.However, using a development framework that prioritises insights and experiences from the target population is a great strength of this intervention, as it allowed direct and effective tailoring of the intervention components despite specific recommendations not being available. Hence, the aerobic PA recommendations were deemed an appropriate target for women treated for cervical cancer. [bib_ref] Increasing physical activity levels following treatment for cervical cancer: an intervention mapping..., Millet [/bib_ref] Additionally, the criterion to include women who are at least 6 months post treatment was also a decision informed by the need's assessment, with the time between 3 and 9 months post treatment being identified as critical for intervention and for positive change. [bib_ref] Prospective examination of objectively assessed physical activity and sedentary time after breast..., Sabiston [/bib_ref] A limitation is that the protocol does not incorporate strength training, and therefore does not reflect the current PA guidelines for after cancer. However, as PA levels are generally low following cervical cancer treatment, it seemed appropriate to promote one PA modality which was realistically achievable (i.e., walking) to enhance adherence to the intervention.
This study will add to the current debate regarding the role of technology in promoting PA after cancer. The intervention uses a PA monitor and its associated online community and a messaging platform. Previous research supports the use of such e-health technologies in combination with health coaching over a shorter duration (4-week programme). [bib_ref] Efficacy, feasibility, and acceptability of a novel technology-based intervention to support physical..., Gell [/bib_ref] However, it has been suggested that further investigation is needed to test the acceptability of such technologies in more diverse populations, [bib_ref] The use of eHealth to promote physical activity in cancer survivors: a..., Haberlin [/bib_ref] to explore optimal intervention length and to question the maintenance of PA behaviour change. [bib_ref] Efficacy, feasibility, and acceptability of a novel technology-based intervention to support physical..., Gell [/bib_ref] Although a limitation of this protocol is that statistically relevant changes in PA between time points will not be reported, trends in PA levels will provide data inferring the feasibility of promoting and maintaining PA behaviours. The process evaluation will also provide valuable insight regarding the acceptability of the components.
# Conclusion
We described the protocol of a study which tests the feasibility and acceptability of delivering and evaluating a multicomponent PA intervention for women following treatment for cervical cancer. A process evaluation will provide insight into whether the intervention components and evaluation measures are accepted by participants. Findings from the study will inform intervention refinement in preparation for a full-scale pilot trial.
Twitter Nessa Millet @Nessamillet1
[fig] Objectives 1: To determine participant recruitment, retention and attrition rates. 2. To review descriptive data to determine the feasibility of the inclusion criteria. 3. To determine compliance and completeness of evaluation measures at all time points. 4. To report the feasibility of delivering the intervention and evaluation measures. 5. To explore participant compliance with the intervention. 6. To obtain views on the acceptability of the intervention and evaluation measures. 7. To describe device-assessed PA levels and questionnaire data outcomes of interest. [/fig]
[fig] Figure 1: Intervention programme flow chart. [/fig]
[fig] Figure 2: Logic model for the ACCEPTANCE trial. PA, physical activity; SCT, social cognitive theory. [/fig]
[table] Table 1: Patient and public involvement (PPI) involvement and engagement in research activities [/table]
[table] Table 2: Description of intervention components [/table]
[table] Table 3: PA and health outcomes, outcome measures and measurement time points [/table]
|
Role of B7-H4 siRNA in Proliferation, Migration, and Invasion of LOVO Colorectal Carcinoma Cell Line
Objectives. Colorectal cancer is one of the most common malignancies. Recent studies investigated that B7-H4 is highly expressed in various cancers. We aimed at exploring the effect of B7-H4 siRNA on proliferation, invasion, and migration of LOVO cells which expressed B7-H4 notably. Design and Methods. Colon adenocarcinoma dataset was downloaded from The Cancer Genome Atlas. 35 colorectal cancer patients admitted to Shanghai Tongren Hospital were enrolled in this study. Cell proliferation and cell cycle distribution were identified by CCK8 and flow cytometry, respectively. Transwell assay was performed to detect the invasion and migration of LOVO cells. CXCL12/CXCR4 expression and JAK2/STAT3 phosphorylation were determined by real-time PCR and western blot. Results. B7-H4 expressed is elevated in colorectal cancer tissues than in the adjacent normal tissues. B7-H4 siRNA effectively inhibited the proliferation at 24 h and 48 h, arrested cell cycle at G0/G1, and suppressed cell invasion and migration. Gene set enrichment analysis showed that CXCL12/CXCR4 and JAK/STAT were correlative with the B7-H4 expression. Additionally, CXCL12/CXCR4 expression and JAK2/STAT3 phosphorylation were reduced. Conclusions. B7-H4 siRNA can effectively inhibit proliferation, invasion, and migration of LOVO cells by targeting CXCL12/CXCR4 and JAK2/STAT3 signaling, which can serve as a new target for colorectal carcinoma treatment.
# Introduction
Colorectal cancer is one of the most common malignancies encountered in the world and is the second most common cause of cancer-related mortality [bib_ref] Colorectal cancer incidence and mortality in china, Zheng [/bib_ref] [bib_ref] The elderly patients with colorectal cancer need careful multidisciplinary evaluation and optimizing..., Wang [/bib_ref]. In the past several decades, the incidence of colorectal carcinoma has also risen in the whole world. With the development of diagnostic technology in recent years, increasing number of patients are diagnosed with early-stage colon cancer, and proper treatment was taken in time. However, there are still patients with early surgery suffering from metastases especially colorectal liver metastases, which can finally result in death [bib_ref] Colorectal liver metastases guidelines, tumor heterogeneity and clonal evolution: can this be..., Lianos [/bib_ref]. Therefore, focusing on the molecular mechanism of colorectal liver metastases is the pressing target in colorectal carcinoma therapy.
B7 family molecules, including B7-H1 (PD-L1), B7-DC (PD-L2), and B7-H4 (B7S1), combine with the corresponding receptors of activated T lymphocytes, which negatively regulate T cell immunity [bib_ref] Inhibitory B7-family molecules in the tumour microenvironment, Zou [/bib_ref]. This plays an important role in downregulating antitumor immunity in cancer microenvironments. B7-H4 is a newly discovered transmembrane protein of the B7-CD28 family. It is expressed in human tissues in the cancers of the colon, lung, pancreas, brain, stomach, breast, ovary, esophagus, prostate, liver, and kidney [bib_ref] B7-H1 and B7-H4 expression in colorectal carcinoma: correlation with tumor FOXP3 +..., Zhao [/bib_ref] [bib_ref] Induced expression of B7-H4 on the surface of lung cancer cell by..., Chen [/bib_ref] [bib_ref] B7-H4 enhances oncogenicity and inhibits apoptosis in pancreatic cancer cells, Qian [/bib_ref] , which is classified as coinhibitors of cell-mediated immunity [bib_ref] B7-H1 and B7-H4 expression in colorectal carcinoma: correlation with tumor FOXP3 +..., Zhao [/bib_ref]. It combines with surface receptors of the activated T cell, thus negatively regulating T cells' growth, cell cycle, and immune function. Recently, B7-H4 was found to be highly expressed in colorectal carcinoma [bib_ref] T-cell infiltration and expressions of T lymphocyte co-inhibitory B7-H1 and B7-H4 molecules..., Liang [/bib_ref]. As a consequence, we attempt to explore the effect of the B7-H4 siRNA silencing on proliferation, cell circle distribution, migration, and invasion in colorectal cancer LOVO cells. To gain further insight into the biological pathways involved in colorectal cancer pathogenesis through B7-H4 pathway, a gene set enrichment analysis (GSEA) was performed.
Presently, we demonstrated the effect of B7-H4 siRNA on proliferation, cell circle distribution, invasion, and migration. GSEA showed that CXCL12/CXCR4 and JAK/STAT were correlative with the B7-H4 expression. The expression of CXCL12, CXCR4, p-JAK2, JAK2, p-STAT3, and STAT3 was also studied by means of RT-PCR and western blot analysis.
# Materials and methods
# Bioinformatics analysis.
The gene expression data were obtained at The Cancer Genome Atlas website (TCGA, https://tcga-data.nci.nih.gov/tcga/) for the colon adenocarcinoma (COAD) projects. To gain further insight into the biological pathways involved in colorectal cancer pathogenesis through B7-H4 pathway, a gene set enrichment analysis (GSEA) was performed. GSEA is a method of analyzing and interpreting microarray and such data using biological knowledge [bib_ref] GSEA-P: a desktop application for Gene Set Enrichment Analysis, Subramanian [/bib_ref]. The data in question is analyzed in terms of their differential enrichment in a predefined biological set of genes (representing pathways). These predefined biological sets can be published information about biochemical pathway or coexpression in a previous experiment. GSEA was performed using GSEA version 2.0 from the Broad Institute at MIT. The TCGA data were analyzed by GSEA. In this study, GSEA firstly generated an ordered list of all genes according to their correlation with B7-H4 expression and then a predefined gene set (signature of gene expression upon perturbation of certain cancer-related gene) receives an enrichment score (ES), which is a measure of statistical evidence rejecting the null hypothesis that its members are randomly distributed in the ordered list. The expression level of B7-H4 was used as phenotype label, and "metric for ranking genes" was set to Pearson's correlation. All other basic and advanced fields were set to default. The KEGG gene sets, biological process database (c2.KEGG.v4.0) from the Molecular Signatures Database-Msig DB (http://www .broad.mit.edu/gsea/msigdb/index.jsp), were used for enrichment analysis.
## Patients and tissue
Samples. 35 colorectal cancer patients admitted to Shanghai Tongren Hospital were enrolled in this study. All patients have complete clinical and pathological follow-up data. Adjacent normal colorectal tissues were also collected as negative controls. These normal colorectal tissues were resected within at least 5 cm of the tumor margin when the patients underwent definitive surgery. Ethical approval for the study was provided by the independent ethics committee, Shanghai Tongren Hospital. Informed and written consent was obtained from all patients or their advisers according to the ethics committee guidelines.
## Cell
Culture. LOVO, SW480, HCT116, SW620, and HT29 cells are human colorectal carcinoma cells. All cells were obtained from the Shanghai Cell Bank, Chinese Academy of Sciences (Shanghai, China). Cells were cultured in DMEM supplemented with 10% fetal bovine serum, 100 units/mL penicillin, and 100 g/mL streptomycin and incubated in a humidified atmosphere at 37 ∘ C with 5% CO 2 .
2.4. miRNA Transfection. LOVO cells were seeded in antibiotic-free medium the day before transfection. The cells were transfected with 50 nmol/L of B7-H4 siRNA or negative control by using lipofectamine 2000 (Invitrogen, Shanghai, China) according to the instructions provided by the manufacturer. After 48 hours, the transfected cells were collected and processed for quantitative real-time PCR (qPCR), western blot, proliferation, cell cycle, migration, and invasion assay.
## Cell proliferation assay. cell viability was assessed by
Cell Counting Kit-(CCK-) 8 (Tongren, Shanghai, China). Briefly, 4 × 10 3 cells were seeded in each 96-well plate and further incubated for 24 and 48 hours, respectively. CCK-8 reagent was added to each well at 1 hour before the endpoint of incubation. The optical density (OD) 450 nm values in each well were determined by a microplate reader. Experiments were repeated at least three times each time in triplicate.
## Cell cycle assay.
After 48 hours of transfection, cells were harvested and cell cycle distribution was analyzed using flow cytometer (FACSCalibur, BD Biosciences, Franklin Lakes, NJ, USA). DNA histograms were analyzed using flow cytometer (FACSCalibur, BD Biosciences).
## Cell invasion assay.
Cell invasion assay was performed by a 24-well Transwell chamber with a pore size of 8 m (Sigma, San Francisco, USA). The inserts were coated with 50 L Matrigel (dilution at 1 : 2; BD Bioscience). Cells were trypsinized after transfection for 24 hours and transferred to the upper Matrigel chamber in 100 L of serum-free medium supplementing 1 × 10 5 cells and incubated for 24 hours. The lower chamber was filled with medium containing 10% FBS as chemoattractants. After incubation, the cells that passed through the filter were fixed and stained by 0.1% crystal violet. The numbers of invaded cells were counted in five randomly selected high power fields under a microscope (OLYMPUS, Shenzhen, China).
## Migration
Assay. Cells in logarithmic phase were digested by 0.25% trypsin (Gibco, Shanghai, China) and then suspended in RPMI-1640 (HyClone) medium containing 10% fetal calf serum (Gibco). Cells were seeded in a 12-plate microplate at a density of 1 × 10 5 cells/mL and then incubated for 1 h. The supernatant was discarded and cells were washed 2 times by PBS (Gibco). 4% paraformaldehyde (JRDUN Biotech, Shanghai, China) was supplemented for 15 min and cells were stained by Giemsa (JRDUN) for 30 min. Then, cells were washed several times and the optical density (OD) values were read at 570 nm by a microplate reader (Thermo, Waltham, USA). Adhesion rate (%) = (OD 1 /OD 0 ) × 100%, with OD 1 being HB treated groups and OD 0 being control group.
## Reverse transcription and real-time pcr.
The mRNA expression of B7-H4 in LOVO cells after B7-H4 siRNA transfection was quantified by RT-PCR. Total RNA was isolated from cells using Trizol Reagent (Invitrogen) and quantified. cDNA was synthesized from 5 mg of RNA using AMV reverse transcriptase (Fermentas, USA) according to the manufacturer's instructions. B7-H4 was amplified from the cDNA by RT-PCR. The PCR conditions consisted of 5 min at 95 ∘ C one cycle, 30 seconds at 95 ∘ C, 30 seconds at 55 ∘ C, 30 seconds at 72 ∘ C, and 7 min at 72 ∘ C 35 cycles. Primer pairs for human genes were designed using the Primer Express Software (Applied Biosystems, Shanghai, China) and are listed in .
## Western blot.
Cultured or transfected cells were harvested and washed twice with PBS and lysed in ice-cold radio immunoprecipitation assay buffer (RIPA, Beyotime, Shanghai, China) with freshly added 0.01% protease inhibitor cocktail (Sigma, Shanghai, China) and incubated on ice for 30 min. Cell lysis was centrifuged at 13,000 rcf for 10 min at 4 ∘ C and the supernatant (20-30 g of protein) was run on 10% SDS-PAGE gel and transferred electrophoretically to a polyvinylidene fluoride membrane (Millipore, Shanghai, China). The blots were blocked with 5% skim milk, followed by incubation with antibodies against B7-H4 (Abcam), GAPDH (Fermentas), CXCL12 (Abcam), CXCR4 (Abcam), p-JAK2 and JAK2 (CST), and p-STAT3 and STAT3 (CST). Blots were then incubated with goat anti-mouse or anti-rabbit secondary antibody (Beyotime, Shanghai, China) and visualized using enhanced chemiluminescence (ECL, Thermo Scientific, Shanghai, China).
## Growth of cells in athymic nude mouse and tumor size
Determination. LOVO cells transfected with B7-H4 siRNA and cells with nontreatment were trypsinized and were washed and resuspended in DMEM without FBS. Cell concentration and viability were determined using trypan blue. Eight male athymic nude mice were randomly divided into 2 groups (4 mice/group) and were subcutaneously injected by 2 × 10 6 cells transfected with control cells or B7-H4 siRNA cells, respectively. The tumor size was determined every 4 days after tumor formed (around 1-2 weeks) as previously described [bib_ref] Determination of subcutaneous tumor size in athymic (nude) mice, Tomayko [/bib_ref]. 48 days later, the mice were sacrificed and photographed, and the tumors were weighted on a digital balance.
# Statistical analysis.
All results are presented as the mean ± SD and the data were analyzed by a SPSS 13.0 statistical package (SPSS Inc., Chicago, IL). Data for multiple comparisons were subjected to one-way ANOVA followed by Dunnett's test and a value of < 0.05 was considered statistically significant.
# Results
## High expression of b7-h4 in tumor
Tissue. The gene expression data were obtained at The Cancer Genome Atlas website (TCGA, https://tcga-data.nci.nih.gov/tcga/) for the colon adenocarcinoma (COAD) projects. As shown in [fig_ref] Figure 1: B7-H4 expression in colon cancer tissues and their adjacent normal tissues [/fig_ref] , a notable increase was observed in B7-H4 mRNA expression of colorectal carcinoma tissue compared with normal tissue.
To further verify the biological role of B7-H4 in colorectal carcinoma, we used real-time PCR to detect the expression levels of B7-H4 in colorectal cancer patients' tissues. Because the expression of B7-H4 is considered quite low in gastrointestinal cancer, B7-H4 levels in ovarian cancer and normal tissues ( = 10) were also detected as positive control (see [fig_ref] Figure 1: B7-H4 expression in colon cancer tissues and their adjacent normal tissues [/fig_ref] in Supplementary Material available online at http://dx.doi.org/10.1155/2015/326981). As [fig_ref] Figure 1: B7-H4 expression in colon cancer tissues and their adjacent normal tissues [/fig_ref] shows, B7-H4 expression level was higher in colorectal tumor tissues than that in adjacent normal tissue control ( = 35, < 0.01).
## B7-h4 expression in colorectal carcinoma cell
Lines. An obvious difference was presented in colorectal cancer tissue and normal tissue. We then detected the mRNA expression and protein level of B7-H4 in various colorectal carcinoma cell lines including LOVO, SW480, HTC116, SW620, and HT29 by RT-PCR and western blot, respectively. As shown in [fig_ref] Figure 2: B7-H4 expression in colorectal carcinoma LOVO, SW480, HCT116, SW620, and HT29 cell... [/fig_ref] , B7-H4 mRNA expression in LOVO cell line was significantly higher than any other cell line. In addition, western blot displayed that protein level of B7-H4 was the highest among all cell lines [fig_ref] Figure 2: B7-H4 expression in colorectal carcinoma LOVO, SW480, HCT116, SW620, and HT29 cell... [/fig_ref] and 2(c)). As a result, LOVO cell line was determined to carry out further investigations.
## Effect of b7-h4 sirna on cell viability of lovo cell line.
B7-H4 mRNA was interfered in LOVO cell line as previously described. RT-PCR and western blot were employed to identify the interference efficient. [fig_ref] Figure 3: Effect of B7-H4 siRNA on cell viability of LOVO cells [/fig_ref] showed that B7-H4 mRNA expression in B7-H4 siRNA group was decreased dramatically compared with the control group and mock group. Western blot showed that protein level was declined notably in B7-H4 siRNA group in comparison with the control group and mock group [fig_ref] Figure 3: Effect of B7-H4 siRNA on cell viability of LOVO cells [/fig_ref].
B7-H4 siRNA on cell viability of LOVO cell line was measured by CCK-8 assay. In [fig_ref] Figure 3: Effect of B7-H4 siRNA on cell viability of LOVO cells [/fig_ref] , cell viability was weakened in B7-H4 siRNA group markedly after transfection, 24 h and 48 h, in comparison with that of control and mock cells.
## Effect of b7-h4 sirna on cell cycle of lovo cell line.
After 48 h of transfection, cell cycle distribution was analyzed using flow cytometer. As shows, compared with the control and mock groups, cells of B7-H4 group were arrested in G0/G1 phase.
## B7-h4 sirna inhibited invasion and migration of lovo
Cell Line. For pervasion to distant organs, the invasion of tumor cells occurs via cell-secreted proteolytic degradation of the cellular basement membrane, which is the leading cause of cancer death [bib_ref] Inhibition of invasion and metastasis of gastric cancer cells through snail targeting..., Guo [/bib_ref]. The Transwell Invasion test results are presented in [fig_ref] Figure 5: Effect of B7-H4 siRNA on invasion and migration of LOVO cells [/fig_ref]. Treatment with B7-H4 siRNA for 24 h significantly suppressed the invasion ability of cells when compared with the control group of untreated cells and mock-vehicle group (36.24 ± 5.37% versus 100 ± 8.36% in the control group; 36.24 ± 5.37% versus 93.31 ± 11.24% in the mock group; all < 0.01). The effects of B7-H4 siRNA on chemotactic motility in cells were evaluated as previously described. As displayed in Figures 5(c) and 5(d), migrated cells decreased dramatically after B7-H4 siRNA transfection treatment for 24 h (41±9.32% versus 100 ± 12.37% in the control group; 41 ± 9.32% versus 102±11.24% in the mock group; all < 0.01). Typical images of B7-H4 siRNA inhibiting cell migration are presented in [fig_ref] Figure 5: Effect of B7-H4 siRNA on invasion and migration of LOVO cells [/fig_ref]. (GSEA). To probe the B7-H4-associated pathways on an unbiased basis, we performed GSEA using data from the TCGA. GSEA is designed to detect coordinated differences in expression of predefined sets of functionally related genes. Among all the 188 predefined "KEGG pathways" gene sets, the CXCL12/CXCR4 pathway and JAK/STAT pathway were identified with the significant association with B7-H4 expression in the TCGA data [fig_ref] Figure 6: Effect of B7-H4 siRNA on the protein expressions of CXCL12, CXCR4, p-JAK,... [/fig_ref].
## Identification of genes and signaling associated biological pathways and processes by gene set enrichment analysis
## B7-h4 sirna regulated the mrna expression of cxcl12
and CXCR4 in LOVO Cells. The effects of B7-H4 siRNA on the mRNA expression of CXCL12 and CXCR4, which are closely related to metastasis of tumor cells, were also investigated and the results are presented in. As a result, B7-H4 siRNA treatment significantly decreased the expression of CXCL12 and CXCR4 in comparison with the control group and mock group. After LOVO cells were treated with B7-H4 siRNA transfection for 6 h; the RNA expression of CXCL12 and CXCR4 was detected by RT-PCR. * * p < 0.01 compared with the control group; ## p < 0.01 compared with the mock group; data are expressed as the mean ± SD, = 6.
After 6 h of B7-H4 siRNA treatment, the phosphorylation protein levels of JAK2 and STAT3 in cells were analyzed by western blot. As shown in Figures 6(e) and 6(f), B7-H4 siRNA significantly decreased the relative expression levels of p-JAK 2/JAK2 (0.2234 ± 0.0197 versus 1.1005 ± 0.0956 in the control group; 0.2234±0.0197 versus 1.04±0.0846 in the mock group; all < 0.01) and p-STAT3/STAT3 (0.2144 ± 0.0214 versus 0.7423 ± 0.0623 in the control group; 0.2144 ± 0.0214 versus 0.6316 ± 0.0594 in mock group; all < 0.01).
## B7-h4 sirna suppressed growth of colorectal carcinoma
In Vivo. Next, we determined whether silence of B7-H4 in colorectal cancer cells could reduce tumor growth in vivo. LOVO cells transfected with control and B7-H4 siRNA were subcutaneously injected in athymic nude mice, respectively. Tumor volumes were measured for 48 days. As shown in [fig_ref] Figure 7: Effect of B7-H4 siRNA on tumor growth in vivo [/fig_ref] , B7-H4 silenced tumors grew slower in mice, whereas control tumors grow faster in mice. After 48 days, the volume and weight of B7-H4 siRNA treated tumor were smaller and lighter than those of control [fig_ref] Figure 7: Effect of B7-H4 siRNA on tumor growth in vivo [/fig_ref] , < 0.01). These data suggested that inhibition of B7-H4 siRNA in colon cancer cells suppressed tumor growth in nude mice.
# Discussion
B7-H4 is associated with the generation and progression of tumor; it may be a new target in the diagnosis and treatment of tumor. B7-H4 expression has been found in various kinds of tumors including colon, lung, pancreas, brain, stomach, breast, ovary, esophagus, prostate, liver, and kidney [bib_ref] B7-H1 and B7-H4 expression in colorectal carcinoma: correlation with tumor FOXP3 +..., Zhao [/bib_ref] [bib_ref] Induced expression of B7-H4 on the surface of lung cancer cell by..., Chen [/bib_ref] [bib_ref] B7-H4 enhances oncogenicity and inhibits apoptosis in pancreatic cancer cells, Qian [/bib_ref] [bib_ref] T-cell infiltration and expressions of T lymphocyte co-inhibitory B7-H1 and B7-H4 molecules..., Liang [/bib_ref] [bib_ref] Enhanced T cell immunity by B7-H4 down regulation in non small-cell lung..., Sun [/bib_ref]. Qian et al. demonstrated that B7-H4 siRNA inhibited cell proliferation, colony formation, and migration of pancreatic cancer cells [bib_ref] B7-H4 enhances oncogenicity and inhibits apoptosis in pancreatic cancer cells, Qian [/bib_ref]. Shi et al. found that serum B7-H4 expression is a significant prognostic indicator for patients with gastric cancer. Cheng et al. demonstrated that B7-H4 directly promoted malignant transformation of ovarian cancer cell line [bib_ref] B7-H4 expression promotes tumorigenesis in ovarian cancer, Cheng [/bib_ref]. In the present study, we found that B7-H4 is highly expressed in colorectal tumor tissues.
Among several cell lines, B7-H4 was found to be notably expressed in LOVO cells. Therefore, LOVO cell line was determined for further investigations. B7-H4 siRNA effectively inhibited proliferation, invasion, and migration of LOVO cells. In order to elucidate the possible mechanism involved, we performed GSEA to identify associated biological processes and signaling pathways using the mRNA signature based risk score for classification. The risk score was accompanied with the downregulation of several cancerrelated networks, namely, CXCL12/CXCR4 pathway and JAK/STAT. These indicated that CXCL12/CXCR signaling and JAK/STAT signaling play crucial role in the process of antiproliferation, anti-invasion, and antimigration triggered by B7-H4. Further in vivo tumor formation study in nude mice indicated that the inhibition of B7-H4 siRNA in colorectal cancer cells delayed the progress of tumor formation. These results suggested that B7-H4 might be a potential therapy target for colorectal cancer.
The chemokine CXCL12 is an extracellular chemokine which binds to its cell surface receptor CXCR4. Various CXCR4/CXCL12 axis-related intracellular signal transduction cascades and effectors, important regulators for survival, proliferation, and death have been determined, such as MAPK [bib_ref] Anticancer properties of the novel nitric oxide-donating compound (S,R)-3-phenyl-4,5-dihydro-5-isoxazole acetic acid-nitric oxide..., Maksimovic-Ivanic [/bib_ref] , JAK/STAT, PI-3K/Akt, and Wnt [bib_ref] A review on CXCR4/CXCL12 axis in oncology: no place to hide, Domanska [/bib_ref] [bib_ref] The CXCL12/ CXCR4 chemokine ligand/receptor axis in cardiovascular disease, Doring [/bib_ref] [bib_ref] The CXCR4/CXCL12 axis in cutaneous malignancies with an emphasis on melanoma, Mitchell [/bib_ref]. The axis of CXCL12/CXCR4 has been considered to play an important role for cancer cell migration [bib_ref] A review on CXCR4/CXCL12 axis in oncology: no place to hide, Domanska [/bib_ref]. reported that CXCL12/CXCR4 axis played an important role in the progression and organ-specific metastasis of pancreatic adenocarcinoma [bib_ref] CXCL12/CXCR4 axis plays pivotal roles in the organ-specific metastasis of pancreatic adenocarcinoma:..., Zhong [/bib_ref]. Wu et al. demonstrated that tumor cells from pancreatic cancer express high level of CXCR4 [bib_ref] Role of CXCL12/CXCR4 signaling axis in pancreatic cancer, Wu [/bib_ref]. CXCL12 is the ligand for CXCR4, which is extensively secreted by neighboring stromal cells and other distant organs. CXCL12 primarily binds to CXCR4 and induces intracellular signaling through several divergent pathways, which are involved in progression and metastasis of cancer. Xie et al. investigated that the specific downregulation of CXCR4 inhibited cell growth, invasiveness, and migration of NSCLC. The accumulating reports have shown that CXCL12 and CXCR4 are overexpressed both in mRNA and protein levels during initiation and progression of cancers. In our study, B7-H4 siRNA significantly suppressed the mRNA and protein expressions of CXCL12 and CXCR4, which could interpret the effects of antiproliferation, anti-invasion, and antimigration of B7-H4 in LOVO cells.
In addition, phosphorylation of JAK/STAT signaling was found negatively adjusted by B7-H4 siRNA. STAT3, an important transcription factor of JAK/STAT signal transduction pathway, can promote the rapid induction of genes by directly transducing signals from the receptor into the nucleus and play a pivotal role in mediating the biological response for these ligands. Activation of the signal transcription-3 (STAT3) signal transduction pathway largely contributes to inflammation and carcinogenesis. Increasing evidence displayed that constitutive activation of STAT3 gave rise to neoplasm invasion and metastasis [bib_ref] Astaxanthin inhibits JAK/STAT-3 signaling to abrogate cell proliferation, invasion and angiogenesis in..., Kowshik [/bib_ref] [bib_ref] Concomitant activation of the JAK/STAT, PI3K/AKT, and ERK signaling is involved in..., Saxena [/bib_ref]. In the present study, the phosphorylation of JAK2 and STAT3 was dramatically inhibited, which was also the mechanism of antiproliferation, anti-invasion, and antimigration of B7-H4 siRNA.
# Conclusions
In conclusion, siRNA targeting of B7-H4 effectively suppressed the proliferation, invasion, and migration of LOVO cells by inhibiting the effect of CXCL12/CXCR4 and the phosphorylation of JAK2/STAT3 on increasing the metastatic potential of colorectal carcinoma. This work can supply important evidence for the prevention and treatment of colorectal cancer.
[fig] Figure 1: B7-H4 expression in colon cancer tissues and their adjacent normal tissues. 35 pairs of colon cancer tissues and their adjacent normal tissues were collected and mRNA expression of B7-H4 mRNA expression was identified by RT-PCR. [/fig]
[fig] Figure 2: B7-H4 expression in colorectal carcinoma LOVO, SW480, HCT116, SW620, and HT29 cell lines. (a) B7-H4 mRNA expression was determined by RT-PCR in LOVO, SW480, HCT116, SW620, and HT29 cell lines. (b and c) Protein levels of B7-H4 were determined by western blot in LOVO, SW480, HCT116, SW620, and HT29 cell lines. * * < 0.01 compared with the LOVO cells; data are expressed as the mean ± SD, = 6. [/fig]
[fig] Figure 3: Effect of B7-H4 siRNA on cell viability of LOVO cells. (a) mRNA expression of B7-H4 in LOVO cells after B7-H4 siRNA transfection for 48 h was quantified by RT-PCR. (b and c) Protein expression of B7-H4 in LOVO cells after B7-H4 siRNA transfection for 48 h was quantified by western blot analysis. (d) After B7-H4 siRNA transfection for 0, 12, 24, and 48 h, cell viability of LOVO cells was identified by flow cytometry. * * < 0.01 compared with the control cells; ## < 0.01 compared with the mock cells; data are expressed as the mean ± SD, = 6. [/fig]
[fig] 3. 8, Figure 4: B7-H4 siRNA Adjusted the Protein Expression of CXCL12, CXCR4, p-JAK2, and p-STAT3. After 48 h of B7-H4 siRNA treatment, the protein expressions of CXCL12 and CXCR4 in cells were analyzed by western blot. As revealed in Figures 6(b) and 6(c), B7-H4 siRNA significantly decreased the expression levels of CXCL12 (0.1984 ± 0.0146 versus 1.1492 ± 0.0965 in the control group; 0.1984 ± 0.0146 versus 1.103 ± 0.0569 in the mock group; all < 0.01) and CXCR4 (0.5032± 0.0459 versus 1.2789 ± 0.1102 in the control group; 0.5032 ± 0.0459 versus 1.4143 ± 0.1236 in mock group; all < 0.01). Effect of B7-H4 siRNA on cell circle distribution of LOVO cells. (a and b) After B7-H4 siRNA transfection for 48 h, cell circle distribution of LOVO cells was identified by flow cytometry. * * < 0.01 compared with the control cells; # < 0.05 and ## < 0.01 compared with the mock cells; data are expressed as the mean ± SD, = 6. [/fig]
[fig] Figure 5: Effect of B7-H4 siRNA on invasion and migration of LOVO cells. (a and b) After B7-H4 siRNA transfection for 48 h, invasive ability of LOVO cells was identified by Transwell assay. (c and d) After B7-H4 siRNA transfection for 48 h, cell migration was identified as previously described. * * < 0.01 compared with the control cells; ## < 0.01 compared with the mock cells; data are expressed as the mean ± SD, = 6. [/fig]
[fig] Figure 6: Effect of B7-H4 siRNA on the protein expressions of CXCL12, CXCR4, p-JAK, JAK, p-STAT3, and STAT3 in LOVO cells. (a and d) Identification of CXCL12/CXCR4 and JAK/STAT signalling as regulatory targets of B7-H4. Gene set enrichment analysis (GSEA) identified significant association between B7-H4 and CXCL12/CXCR4 and JAK/STAT signaling pathway in both the TCGA colorectal cancer dataset and the multitumor dataset. (b and c) After 48 h of B7-H4 siRNA treatment, the protein expressions of CXCL12 and CXCR4 in cells were analyzed by western blot. GAPDH was also detected as the control of sample loading. (e and f) Western blot was performed to identify the protein levels of p-JAK2, JAK, p-STAT3, and STAT3, and GAPDH was also detected as the control of sample loading. * * < 0.01 compared with the control cells; ## < 0.01 compared with the mock cells; data are expressed as the mean ± SD, = 6. [/fig]
[fig] Figure 7: Effect of B7-H4 siRNA on tumor growth in vivo. LOVO cells transfected with B7-H4 siRNA were subcutaneously injected in athymic nude mice. (a) Tumor diameter was evaluated for 48 days. (b) At day 48, mice were sacrificed and tumors were weighted. Tumor growth was significantly reduced in B7-H4 siRNA tumors. * * < 0.01 compared with the control cells; data are expressed as the mean ± SD, = 4. [/fig]
|
Secretome profiles of immortalized dental follicle cells using iTRAQ-based proteomic analysis
Secretomes produced by mesenchymal stromal cells (MSCs) were considered to be therapeutic potential. However, harvesting enough primary MSCs from tissue was time-consuming and costly, which impeded the application of MSCs secretomes. This study was to immortalize MSCs and compare the secretomes profile of immortalized and original MSCs. Human dental follicle cells (DFCs) were isolated and immortalized using pMPH86. The secretome profile of immortalized DFCs (iDFCs) was investigated and compared using iTRAQ labeling combined with mass spectrometry (MS) quantitative proteomics. The MS data was analyzed using ProteinPilotTM software, and then bioinformatic analysis of identified proteins was done. A total of 2092 secreted proteins were detected in conditioned media of iDFCs. Compared with primary DFCs, 253 differently expressed proteins were found in iDFCs secretome (142 up-regulated and 111 down-regulated). Intensive bioinformatic analysis revealed that the majority of secreted proteins were involved in cellular process, metabolic process, biological regulation, cellular component organization or biogenesis, immune system process, developmental process, response to stimulus and signaling. Proteomic profile of cell secretome wasn't largely affected after immortalization converted by this piggyBac immortalization system. The secretome of iDFCs may be a good candidate of primary DFCs for regenerative medicine.
A variety of factors secreted by MSCs (mesenchymal stromal cells) including growth factors, cytokines and chemokines is broadly defined as the MSC secretomes [bib_ref] Stem cells: insights into the secretome, Makridakis [/bib_ref]. MSCs secretomes were considered to be therapeutic potential for regenerative medicine because of their angiogenic, trophic and immunomodulatory properties 2 . The effectiveness of MSCs secretomes on a series of diseases had been verified by many in-vitro or preclinical study [bib_ref] Human fetal mesenchymal stem cell secretome enhances bone consolidation in distraction osteogenesis, Xu [/bib_ref] [bib_ref] Conditioned medium derived from umbilical cord mesenchymal stem cells regenerates atrophied muscles, Kim [/bib_ref] [bib_ref] The human amniotic fluid stem cell secretome effectively counteracts doxorubicin-induced cardiotoxicity, Lazzarini [/bib_ref]. Several types of MSCs could be isolated from dental tissue, including dental pulp cells (DPCs), stem cell from apical papilla (SCAP), dental follicle cells (DFCs), periodontal ligament cells (PLCs) and so on ref. [bib_ref] Stem Cells of Dental Origin: Current Research Trends and Key Milestones towards..., Bakopoulou [/bib_ref]. The secretomes of these dental tissue derived cells were acquired, and also showed good performance on tissue repair and neurodegenerative diseases in the previous studies [bib_ref] Therapeutic Potential of Dental Pulp Stem Cell Secretome for Alzheimer's Disease Treatment:..., Nel-M [/bib_ref] [bib_ref] Angiogenic Potential and Secretome of Human Apical Papilla Mesenchymal Stem Cells in..., Bakopoulou [/bib_ref] [bib_ref] Dental pulp stem cells' secretome enhances pulp repair processes and compensates TEGDMA-induced..., Paschalidis [/bib_ref].
As a representative of dental tissue derived MSCs, dental follicle cells (DFCs) were isolated from dental follicle and had been proposed to have the capacity to differentiate into periodontium consisting of cementum, alveolar bone, and periodontal ligament [bib_ref] Dental follicle stem cells and tissue engineering, Honda [/bib_ref]. Under certain conditions, DFCs can be induced to differentiate into chondrogenic, osteogenic, adipogenic and neurogenic cells [bib_ref] Dental follicle stem cells and tissue engineering, Honda [/bib_ref]. DFCs secretomes was considered to be therapeutic potential. Conditioned medium (CM) from DFCs culture could induce osteogenic and cementogenic differentiation in vitro [bib_ref] Dental follicle cell-conditioned medium enhances the formation of osteoclast-like multinucleated cells, Kawakami [/bib_ref] [bib_ref] Adipose tissue-deprived stem cells acquire cementoblast features treated with dental follicle cell..., Wen [/bib_ref] and the formation of dental pulp-dentin complex in vivo [bib_ref] Differentiation of dental pulp stem cells into regular-shaped dentin-pulp complex induced by..., Yu [/bib_ref].
However, because the lifespan of primary MSCs and cell secretomes produced by MSCs is limited, a large number of MSCs need to be isolated from the tissues and expanded in vitro. Besides, this process was time-consuming and costly, which heavily impeded future clinical application of MSCs secretomes. So a population of immortalized MSCs which can secrete similar factors with original cells should be developped to satisfy the need of secretomes-based strategy.
In our previous study [bib_ref] TrAmplification of Human Dental Follicle Cells by piggyBac Transposon -Mediated Reversible Immortalization..., Wu [/bib_ref] , we immortalized human DFCs (iDFCs) using a piggyBac immortalization system. After immortalization, iDFCs maintained immunophenotype and differentiation capacity of DFCs, as well as high telomerase activity. However, the profile of iDFCs secretomes was unclear. Supposed that iDFCs secretomes were similar to (or not largely different from) original DFCs, they could be the excellent candidate for secretomes The phenotype and ability of multiple lineage differentiation. iDFCs displayed the phenotypes of CD73, CD90, and CD146 positivity and CD34, CD45 negativity. The percentage of STRO-1 positive cells in iDFCs are about 0.99% [fig_ref] Figure 1: Morphology, phenotype and telomerase activity of immortalized DFCs [/fig_ref]. The ability of multiple lineage differentiation was also confirmed by examining the potential for multilineage differentiation to adipocyte, chondrocyte, and osteocyte lineages.
Telomerase activity enhanced after immortalization. Telomerase activity of iDFCs significantly increased compared with that of DFCs, and maintained at a high level till passage 50 [fig_ref] Figure 1: Morphology, phenotype and telomerase activity of immortalized DFCs [/fig_ref].
A large number of proteins were detected in iDFCs secretome. A total of 2092 secreted proteins were detected in conditioned media of iDFCs. These detected proteins involved in cellular process, metabolic process, biological regulation, regulation of biological process, cellular component organization or biogenesis, immune system process, developmental process, response to stimulus, signaling, localization, multicellular organismal process, growth and so on. Gene Ontology (GO) term (molecular function, cellular component, biological process) of detected proteins in CM of iDFCs was shown in [fig_ref] Figure 2: Gene Ontology term of identified protein in iDFC secretome [/fig_ref]. COG (Clusters of Orthologous Groups of Proteins System classification) of all detected proteins in the iDFCs secretomes was shown in [fig_ref] Figure 3: COG classification of all detected proteins in the CM of iDFCs [/fig_ref].
Secretome profile was partly affected by immortalization. Only 12.1% of detected secretory proteins (253 proteins) were significantly affected by immortalization. Compared with primary DFCs, 142 protien up-regulated and 111 down-regulated in secretome of iDFCs (Supplementary file). The GO term of these differently expressed proteins was classified between the secretome of DFCs and iDFCs [fig_ref] Figure 4: Gene Ontology term of up-regulated or down-regulated proteins in CM of iDFCs... [/fig_ref]. The pathway analysis of all up-regulated or down-regulated proteins after immortalization was shown in Supplementary file.
## Dfcs or idfcs secretome contained a series of bioactive factors.
In the CM of DFCs or iDFCs, we detected collagen proteins (Type I, II, III, IV, V, VI, XI, XII), Nestin, MMPs (matrix metalloproteinase-2), TIMPs (metallopeptidase inhibitors), HSPs (heat shock proteins), PDGF (Platelet-derived growth factor), IGF-1,2(insulin-like growth factor-1 and -2), IGFBPs (insulin-like growth factor binding proteins), VEGF (vascular endothelial growth factor), bFGF (basic fibroblast growth factor) TGF-β1,2 (transforming growth factor beta 1, 2), HGF (hepatocyte growth factor) and SCF (stem cell factor), etc, which are closely related with proangiogensis, ECM remodeling, tissue repair and regeneration.
Validation of selected differently secreted proteins. Concentrations of selected secreted proteins were tested using Enzyme linked immunosorbent assay (ELISA). Compared with DFCs CM, the level of TGF-β1
# Discussion
DFCs are a population of cells isolated from dental follicle, which develops into periodontal tissues, such as periodontal ligament fibers, cementum and alveolar bone. DFCs have potential capacity of multi-differentiation, and are good choice for dental tissue engineering. However, like other types of MSCs, acquiring enough DFCs is also a major challenge before their application. Isolation of primary DFCs is costly and time-consuming, and limited lifespan of DFCs lead to the loss of their original character after several passages. To resolve this problem, this current study immortalized the primary DFCs using a piggyBac transposon-mediated system with overexpression of SV40 T-Ag. The results of our previous study 14 showed multi-differentiation potential and immunophenotype of original DFCs could be retained based on this immortalization method.
MSCs secretomes were considered to be trophic, pro-angiogenic, immunomodulative, anti-inflammatory and anti-apoptosis, which largely contribute to the therapeutic potential of MSCs in regenerative medicine. The cell-free strategy based on MSCs secretomes attracted more attention and have demonstrated to be effective in many previous studies [bib_ref] Human fetal mesenchymal stem cell secretome enhances bone consolidation in distraction osteogenesis, Xu [/bib_ref] [bib_ref] Conditioned medium derived from umbilical cord mesenchymal stem cells regenerates atrophied muscles, Kim [/bib_ref] [bib_ref] The human amniotic fluid stem cell secretome effectively counteracts doxorubicin-induced cardiotoxicity, Lazzarini [/bib_ref] [bib_ref] Therapeutic Potential of Dental Pulp Stem Cell Secretome for Alzheimer's Disease Treatment:..., Nel-M [/bib_ref] [bib_ref] Angiogenic Potential and Secretome of Human Apical Papilla Mesenchymal Stem Cells in..., Bakopoulou [/bib_ref] [bib_ref] Dental pulp stem cells' secretome enhances pulp repair processes and compensates TEGDMA-induced..., Paschalidis [/bib_ref]. Compared MSCs-based strategy, cell-free strategy is obviously an easier and safer way, which is less immunogenic and devoid of the risk of neoplasia. However, MSCs only produce a small amount of secretomes before they lose original character, so collecting enough MSCs secretomes for basic research or clinical therapy is also difficulty. To solve this problem, a large number of MSCs need to be isolated repeatedly from the tissues and amplified in vitro, which heavily impedes future clinical application of MSCs secretomes. Prolonging the lifespan and maintaining the character of original cells by immortalization could be a good choice. It was still unclear whether immortalization affect largly cell secretomes, which is closely associated with its further application. This present study for the first time comprehensively analyzed the secretome profile of a immortalized MSCs.
The results of this current study showed that immortalization partly changed the proteomic profile of DFCs secretome. Some secreted proteins were up-regulated, while some were down-regulated after immortalization using piggyBac system. Although the minority of secreted proteins (only 12.1%) were found to be differently expressed, the majority of the secreted proteins were not significantly affected by cell immortalization. Thus, the secretory function of DFCs was almost retained in the iDFCs. The secretomes of iDFCs may be a candidate for that of primary DFCs, which may be applied in regenerative medicine in future.
The iDFCs secretome included several key factors involved in tissue repair and regeneration such as collagens, MMPs, TGF-β, bFGF, SCF, and pro-angiogenic factors. Compared with primary DFCs, iDFCs secreted similar amount of angiogenic factors (VEGF, HGF, IGF-1, etc). These pro-angiogenic factors promote the proliferation and migration of endothelial cells, and maturation of newly-formed blood vessel [bib_ref] Cooperative signaling for angiogenesis and neovascularization by VEGF and HGF following islet..., Golocheikine [/bib_ref]. Collagens and MMPs play a important role in ECM remodeling during tissue repair or regeneration [bib_ref] Matrix remodeling by MMPs during wound repair, Rohani [/bib_ref]. SCF could induce homing and navigation of stem cells, which guarantee the participation of endogenous stem cells in the process of regeneration or repair [bib_ref] SCF promotes dental pulp progenitor migration, neovascularization, and collagen remodeling -potential applications..., Pan [/bib_ref]. bFGF can induce the proliferation of stem/progenitor cells and dentine formation [bib_ref] Basic fibroblast growth factor enhances stemness of human stem cells from the..., Wu [/bib_ref] [bib_ref] Effects of basic fibroblast growth factor on proliferation, the expression of osteonectin..., Shiba [/bib_ref] [bib_ref] Effects of recombinant basic fibroblast growth factor, insulinlike growth factor-II and transforming..., Tziafas [/bib_ref]. TGF-β1 was closely associated with migration, proliferation and odontogenic/osteogenic differentiation of progenitor cells/ stem cells, and stimulating matrix secretion [bib_ref] Effects of recombinant basic fibroblast growth factor, insulinlike growth factor-II and transforming..., Tziafas [/bib_ref] [bib_ref] TGF-β in dentin matrix extract induces osteoclastogenesis in vitro, Sriarj [/bib_ref] [bib_ref] Photoactivation of endogenous latent transforming growth factor-β1 directs dental stem cell differentiation..., Arany [/bib_ref] [bib_ref] Prospective potency of TGF-β1 on maintenance and regeneration of periodontal tissue, Maeda [/bib_ref]. Higher level of secreted TGF-β1 was found in immortalized cells, which implies immortalization may partly enhance the therapeutic potential of cell secretomes.
In this study, ElISA was done to verify the results of Mass spectrometry. Three proteins detected in MS analysis were selected. One of the three proteins (TGF-β1) up-regulated, one (IL-6) down-regulated and another (VEGF) didn't change significantly after immortalization. The result of ELISA support the finding of iTRAQ-based MS analysis.
The piggyBac transposon system is adopted as the tool for immortalizing DFCs. This system can effectively catalyze integration and excision of transgenes in human cells between vectors and host genome through a direct "cut and paste" mechanism called transposition 24 . It is considered to be superior to other transposon system in different types of mammalian cells [bib_ref] Genome-wide mapping of PiggyBac transposon integrations in primary human T cells, Galvan [/bib_ref]. It has been reported that various MSCs derived from dental associated tissue were successfully immortalized using piggyBac transposon system [bib_ref] Bone morphogenetic protein-9 effectively induces osteo/odontoblastic differentiation of the reversibly immortalized stem..., Wang [/bib_ref] [bib_ref] PiggyBac transposon-mediated gene delivery efficiently generates stable transfectants derived from cultured primary..., Inada [/bib_ref].
This current study is subsequent to our previous study [bib_ref] TrAmplification of Human Dental Follicle Cells by piggyBac Transposon -Mediated Reversible Immortalization..., Wu [/bib_ref]. That previous study had evaluated the immunophenotype, multi-differentiation potential, telomerase activity, proliferation ability of iDFCs. For deep investigating iDFCs and evaluating the feasibility of therapy mode relying on iDFCs secretomes, this present study compared the profile of secretory factors before and after immortalization. As an important supplement, the results of this present study could provide information for further application of iDFCs.
Overall, our study provides a systemic secretome analysis of immortalized DFC revealing a number of secreted proteins which participate in various biological process. Immortalization using this piggyBac immortalization system didn't largely change proteomic profile of cell secretome. The secretome of iDFCs could be a good candidate of original DFCs for regenerative medicine.
# Methods
Isolation and culture of human dental follicle cells. All experimental protocols were approved by the Ethics Committee of Chongqing Medical University, and all methods were carried out in accordance with relevant guidelines and regulations. Isolation of human DFCs was done according to the method described in the previous study [bib_ref] Bone morphogenetic protein-9 effectively induces osteo/odontoblastic differentiation of the reversibly immortalized stem..., Wang [/bib_ref] [bib_ref] Isolation and characterization of dental follicle cells from adult human dental follicle..., Zhang [/bib_ref]. Freshly extracted impact teeth for therapeutic pupose were collected in Chongqing stomatological hospitals in accordance with institutional ethical committee of this hospital. A written informed consent was acquired from each donner' s parents. Dental follicle tissue was isolated from teeth, washed repeatedly with phosphate-buffered saline (PBS; Hyclone, U.S.A). Subsequently, the tissue was cut into 1 mm 3 pieces, dissociated enzymatically for 40 min at 37 °C with 3 mg/ml Type I collagenase (Gibco, USA) and 5 mg/mL Dispase (Roche, Switzerland) in PBS supplemented with 15% fetal bovine serum (FBS, Gibco, USA). The suspension was filtered through a 70 μm filter. Then was pelleted by centrifugation at 1000 rpm/min for 5 min. The pellet was resuspended with culture medium. Then all cells were cultured in DMEM/F12 medium (Hyclone, U.S.A) with the addition of 10% FBS, penicillin (100 U/ml) and streptomycin (100 μg/ml) in an atmosphere of 37 °C, 20% O2 and 5% CO 2 . When passaging cells, a mixture of 0.25% trypsin and 0.01% ethylenediaminetetraacetic acid (EDTA) was used.
PiggyBac Mediated immortalized human DFCs. For setting up immortalized DFCs using the method described by Wu et al. [bib_ref] TrAmplification of Human Dental Follicle Cells by piggyBac Transposon -Mediated Reversible Immortalization..., Wu [/bib_ref] Multiple lineage differentiation. Osteogenic differentiation. iDFCs at were seeded onto 12-well plates, grown to 70% confluence, and incubated in the differentiation medium containing 10 nM dexamethasone, 10 mM b-glycerophosphate, 50 mg/mL ascorbate phosphate, 10 nM 1, 25-dihydroxyvitamin D3, and 10% FBS for 5 weeks. Cultures were fixed in 60% isopropanol, and mineralization of extracellular matrix stained with 1% Alizarin Red S.
Adipogenic differentiation. iDFCs were seeded onto 12-well plates, grown to subconfluence, and incubated in the adipogenic medium containing 1 mM dexamethasone, 1 mg/mL insulin, 0.5 mM 3-isobutyl-1-methylxanthine, and 10% FBS for 6 weeks. Cells were fixed in 10% formalin for 60 min, washed with 70% ethanol, and lipid droplets were stained with 2% (w/v) Oil Red O reagent for 5 min and washed with water.
Chondrogenic differentiation. iDFCs were seeded onto 12-well plates, grown to 70% confluence, and incubated in the differentiation medium containing cultured with TGF-β3 (10 ng/mL) for 2 weeks in 5% CO 2 at 37 °C, with a fresh medium change every 3-4 days. Chondrogenesis was perfomed by staining with safranin-O.
Telomerase activity analysis. The telomerase activities of DFCs and iDFCs were examined using TeloTAGGG ™ Telomerase PCR ELISAPLUS (Roche, USA). Briefly, telomeric repeats (TTAGGG) were added to the 3'-end of the P1-TS-primer, and the resulting products were amplified by PCR with the Internal standard. The obtained PCR products were divided into two vessels, followed by denaturation and hybridization with digoxigenin-labeled probes. Then the products were checked with an antibody against digoxigenin and the peroxidase substrate TMB. Cell extract treated under 85 °C for 10 min were used as negative controls.
Preparation of cell secretomes. DFCs (Passage 4-6) and iDFCs (Passage 5) were seeded onto T75 culture flasks respectively at a concentration of 1 × 10 6 cells/mL. When reaching 70% confluence, the cells were washed thoroughly 5 times with PBS to remove any serum residues and were re-fed with 12 ml serum free DMEM/F12 medium. After culturing the cells for 24 h, the CM was collected, centrifuged, filtered using 0.22 um syringe filters (Milipore, Germany). The CM was concentrated at 3000 × g with a 3KD Amicon ® Ultra centrifugal filters (Milipore, Germany), and then vacuum frozen dried. The powder sample was stored at −80 °C until further experiments.
Protein extraction. The powder was resuspended in 200 μl of tetraethylammonium bromide (TEAB) buffer, disrupted using ultrasonication, and centrifugated at 12000 rpm/min. The supernatant was collected, and 4 times volume of cold acetone with 10 mM DTT was used to precipitate proteins at −20 °C for about 2 h. After centrifugation at 12000 rpm/min for 20 min at 4 °C, the precipitates were collected and washed with 800 μl of cold acetone for two times. After centrifugation at 12000 rpm/min for 20 min at 4 °C, the supernatants were removed and the precipitates were dried and stored at −80 °C for later use. After resuspended in 200 μl of TEAB buffer, the protein concentration was measured using the Branford assay kit (Solarbio, Beijing, China). iTRAQ labeling and Peptide Fractionation. The protein samples were dissolved and reduced with tris-(2-carboxyethyl) phosphine, alkylated with methyl methanethiosulfonate, trypsin digested and labeled with iTRAQ Reagent-8 plex Multiplex Kit (AB Sciex U.K.) according to the manufacturer's instructions. All of the labeled samples were mixed with equal amount (CM of DFCs: 113,114; CM of iDFCs: 115,116). Next, the labeled samples were fractionated using a high-performance liquid chromatography system (Thermo Dinoex Ultimate 3000 BioRS) equipped with a Durashell C18 (5 mm, 100 Å, 4.6 × 250 mm) column.
## Mass spectrometry analysis. lc-ms/ms analysis was performed on an ab sciex nanolc-ms/ms
(Triple TOF 5600 plus) system. Samples were chromatographed using a 120-min gradient from 2 to 35% (mobile phase A: 0.1% (v/v) FA, 2% (v/v) ACN; mobile phase B: 0.1% (v/v) FA, 90% (v/v) ACN after direct injection onto a 20 cm PicoFrit emitter (New Objective) packed to 20 cm with Magic C18 AQ 3-mm 200 Å stationary phase. MS1 spectra were collected in the range 360e1460 m/z for 250 ms. The 20 most intense precursors with charge state 2e5 were selected for fragmentation, and MS2 spectra were collected in the range 50e2000 m/z for 100 ms; precursor ions were excluded from reselection for 15 s. Bioinformatic analysis. The mass spectrometry data was analyzed using ProteinPilot TM v4.5 (Applied Biosystems); peptide identifications were made using the Paragon algorithm searching against the UniProt human protein database. Only unique peptides whose confidence was more than 95% were contained in iTRAQ labeling quantification, and protein with the unused value more than 1.3 were considered for further analysis.
To determine the biological and functional properties of the identified proteins secreted by DFCs and iDFCs, the identified protein sequences were mapped with Gene Ontology Terms (http://geneontology.org/). For this, homology search was first performed for all the identified sequences with a localized NCBI blastp program against NCBI nr database. The e-value threshold was set to less than 1e-5, and the best hit for each query sequence was taken account for GO term matching. The GO term matching was performed with blast2go v4.5 pipeline [bib_ref] Blast2GO: A comprehensive suite for functional analysis in plant genomics, Conesa [/bib_ref]. COG System (http://www.ncbi.nlm.nih.gov/COG/) were employed for the functional annotation of genes from new genomes and for research into genome evolution. Pathway analysis specifying the relationships between the interacting molecules was made according to the KEGG database (http://www.kegg.jp/).
## Elisa.
To confirm these differentially-expressed proteins, ELISA was performed for further verification. In total, 4 × 10 6 DFCs or iDFCs were grown on 150-mm culture plates and incubated overnight. Cells were washed with PBS and cultured with serum-free DMEM/F12 media. After 24 h treatment, supernatants were collected and centrifuged to remove the debris. Levels of IL-6, TGF-β1 and VEGF were determined using Human IL-6 ELISA kit (Solarbio, China), Human TGF-beta1 Quantikine ELISA Kit and Human VEGF Quantikine ELISA Kit (R&D Systems).
# Statistical analysis.
Proteins with an average fold change larger than 1.5 (DFCs: iDFCs ratio >1.5 or <0.67) were considered to be significantly differentially expressed. All quantitative experiments were done in three independent experiment and the results were determined by three independent experiment. Data were showed as mean ± SD. Statistical significance was determined by student's t test and a value of p < 0.05 was considered statistically significant.
[fig] Figure 1: Morphology, phenotype and telomerase activity of immortalized DFCs. (A) The morphology of DFCs at passage 5 and iDFCs at passage 5, 50. Bar: 10 μm; (B) The phenotype of iDFCs; (C) The telomerase activity of DFCs and iDFCs at various passage. The telomerase activity of cells significantly increased after immortalization. [/fig]
[fig] Figure 2: Gene Ontology term of identified protein in iDFC secretome. [/fig]
[fig] Figure 3: COG classification of all detected proteins in the CM of iDFCs.SCiEntifiC REPORts | 7: 7300 | DOI:10.1038/s41598-017-07467-3 up-regulated and the level of IL-6 (interleukin-6) down-regulated in CM of iDFCs. No significant difference in protein level (VEGF) was found between DFCs and iDFCs CM(Fig. 5). The ELISA results were consistent with the results of iTRAQ-labeling MS analysis. [/fig]
[fig] Figure 4: Gene Ontology term of up-regulated or down-regulated proteins in CM of iDFCs compared with CM of DFCs. [/fig]
[fig] Figure 5: Concentration of cytokines in the CM of DFCs and iDFCs. Secretion of TGF-β1 increased in iDFCs compared to DFCs; Secretion of IL-1 decreased in iDFCs compared to DFCs; Secretion of VEGF didn't significantly differ between DFCs and iDFCs. [/fig]
|
Smartphone Smoking Cessation Application (SSC App) trial: a multicountry double-blind automated randomised controlled trial of a smoking cessation decision-aid ‘app’
## Reviewer
Megan Lim Burnet Institute, Australia REVIEW RETURNED 05-May-2017
## General comments
This paper and the trial it reports on are excellent and I have nothing but positive things to say about the work. The paper as a whole is concise and well-written, covering all key points. The use of a real-world setting in a randomised controlled trial design is highly commendablethis is the sort of work that needs to be done more in the mHealth space. They also achieved quite a good response rate considering the hands-off approach used. 4. P10 the paragraph is confusing. It starts with "in all countries, abstinence rates at one month were higher" although it is not significant for Singapore. And the second sentence mix the comparison between countries and the fact that the increase in abstinence rate is not significant.
## Reviewer
## 5.
A part from the wide coverage of the population with smartphone, is there some justifications for choosing USA, UK, Australia, and Singapore for the trial?
6. In addition, do those countries have similar smoking ban policies? Could that explain the absence of difference in the abstinence rate between the intervention and control group in Singapore?
## Minor comments
One limit of the smartphone smoking cessation tool is that it targets a specific population who own an expensive device and, as highlighted by the other were likely to be more motivated than the general population of smokers. It might be important to highlight that SSC won't replace the need for more smoking cessation programs, especially in low income countries.
An interesting perspective could be to propose a smartwatch application (like for physical activity) and assess whether it improve the effects of the intervention app. Comment: This study of a smartphone application with decision aids for smokers to help them quit has a number of strengths that makes it potentially an important contribution to the literature. Chief among the strengths is the size and scope (smokers from four countries) of the sample, high rate of follow-up, intent-to-treat analyses with imputation, and size of the improvement in continuous abstinence. However, the paper has several weaknesses that must be addressed before it is suitable for publication in BMJ Open.
## Version 1 -author response
Authors' Response: Thank you! Comment: A large shortcoming is the lack of sufficient detail in the description of the intervention mobile app. It is alternatively described as providing a decision aid and support for smoking cessation. It seems that the former is more accurate.
The features of the mobile app do not conform to common protocols for brief smoking cessation counseling and it appears that smokers were free to select any method to help them quit once they had set a quit date. This make the mobile app tested in this study different from other smoking cessation services provided by mobile phones and telephone quit lines. This should be clearly noted in the paper.
Authors' Response: We have added additional information about the intervention app on p8. Details of the app function were published in the study protocol.
Comment: In addition, the authors fail to describe any theoretical model that guided the design of the mobile app in the intervention arm. The authors do imply that making an informed decision to quit and reducing decisional conflict are precursors to quitting successfully and these precursors were targeted by the mobile app. However, there is no explanation given for these key conceptual mediators of quitting nor any description of how the mobile app components were designed to influence them. Moreover, it would be useful for the authors to conduct a mediational analysis on these two variables, not use of the mobile app features.
Authors' Response: The decision aid app design was based on the Ottawa Decision Support Framework (ODSF). This framework is the most widely used for the development of patient decision aids and it uses concepts and theories from general psychology , values (Fischoff, Slovic and Lichtenstein), social support , and self efficacy . We have added a statement to the manuscript on page 8 to clarify this. We do not feel that a meditational analysis of informed choice and decisional conflict is required in our RCT design and analysis. Our decision aid design use of the ODSF and outcome measure of the Multi-dimensional Measure of Informed Choice (MMIC) are consistent with numerous studies in this field as seen in the recently updated Cochrane review of 105 RCTs of patient decision aids. One of our investigators (Professor Trevena) is a co-author on that review.
Comment: The description of the study as "double-blind" seems misleading. The two mobile apps were clearly different in nature, with the intervention mobile app having far more components than the control mobile app. It would seem that participants could clearly determine which arm they were in if one was described as informational and the other as motivational and supportive in the consent form. However, it is not clear from the write up whether individual smokers were provided any sort of informed consent that described the study arms prior to completing the baseline survey and entering the study.
Authors' Response: As stated earlier, we have included additional details about the consent process that clarify the double-blinding aspect of our study design.
Comment: There are a number of aspects of the sample that suggest the recruitment through the Apple App Store may have created a selection bias that reduced generalizability. Notably, the sample had more females, higher education, middle income, and non-married status. The sample is also limited to smokers in four highly-developed, Westernized countries. This probably does reflect the population of smartphone users who use mobile apps but it may not generalize to the entire population of adult smokers. This should be noted in the Discussion.
Authors' Response: We have included an additional sentence to the Discussion section (p14) about generalizability beyond these four country settings.
Comment: Most importantly, the smokers were predominately young (average age of 28) and half were light smokers (i.e., had low nicotine dependence). Younger smokers often do not use smoking cessation services, which may explain why a sizable number chose to quit unassisted by anything beyond the mobile app. Perhaps a decision aid such as the one used in this trial is one way to get younger smokers to try some help to quit. It would be useful to analyze whether the success of the mobile app was moderated by nicotine dependency in this trial. |
Secular changes and predictors of adult height for 86 105 male and female members of the Thai Cohort Study born between 1940 and 1990
Background Height trends can be useful indicators of population health but, despite Thailand's rapid socioeconomic development since the 1950s, few studies have examined accompanying secular changes in adult height or the effects of the transition on the heights of rural versus urban populations. This study therefore sought to document average heights in different age groups of rural and urban Thais and to investigate factors associated with attained height. Methods Data from 86 105 Thai Cohort Study participants was used to estimate mean heights for men and women in different birth year groups. Simple regression was used to calculate the change in height per decade of birth year among those based in rural or urban locations as children. Multiple linear regression was used to investigate effects of other childhood factors on height. Results Overall, average heights were found to have increased by approximately 1 cm per decade in those born between 1940 and 1990. However, the rate of increase was 0.4e0.5 cm per decade greater among urban-based Thais compared with those from the countryside. Parental education levels, household assets, birth size, sibling number, birth rank and region of residence were also significantly associated with adult height. Conclusions These data suggest a marked secular increase in Thai heights in the second half of the 20th century probably reflecting improved childhood health and nutrition over this time. Rural-born Thais, who benefited to a lesser extent from the changes, may face future health challenges with greater risks of, among other things, obesity and its health consequences.
Thailand's relatively rapid transition from traditional subsistence to modern consumer economy has already produced a demographic transition in death rates [bib_ref] Cohort profile: the Thai Cohort of 87,134 Open University students, Sleigh [/bib_ref] and, in the 20th century, much of the dramatic decline in these rates was due to falling infant and child mortality.As attained adult height reflects, among other things, childhood health and nutrition, alongside falling infant and childhood mortality, it would be expected that marked secular increases in height have occurred among Thais over the past 50 years.
As well as reflecting childhood conditions, attained adult height is associated with mortality in later life. Overall, short stature is associated with a higher total mortality, 3e6 although associations with cause-specific mortality appear to vary such that most studies have found an inverse association between attained adult height and coronary heart disease and stroke, but a positive association with cancers such as those of the breast, prostate and colon. [bib_ref] Height, leg length, and cancer risk: a systematic review, Gunnell [/bib_ref] [bib_ref] Height and site-specific cancer risk: a cohort study of a Korean adult..., Sung [/bib_ref] The documentation of trends in height thus can not only help chart improvements in population health, but may also be a useful indicator of future health burdens.
Despite being well along its health transition, little information on secular changes in Thai height has been published. Furthermore, despite distinct income inequalities that have existed between rural and urban Thais, [bib_ref] Income Inequality and Kuznets' Hypothesis in Thailand, Ikemoto [/bib_ref] height changes in the two groups have not been compared. For these reasons the aims of the current study are to: document the average attained heights in different age groups of rural and urban-based participants in a large cohort study of male and female open university students in Thailand (the Thai Cohort Study), and to investigate factors associated with these secular changes.
# Methods
The overall aim of the Thai Cohort Study is to examine the health consequences of Thailand's rapid socioeconomic development and environmental change. [bib_ref] Cohort profile: the Thai Cohort of 87,134 Open University students, Sleigh [/bib_ref] To this end participants were recruited from the student body of the Sukhothai Thammathirat Open University (STOU). STOU students reside all over Thailand and have a modest socioeconomic status. Many are rural dwellers and most have full-time jobs. In 2005, all STOU students who had completed at least their first semester of study (approximately 200 000) were mailed a questionnaire, information sheet and consent form. A total of 87 134 students (47 314 women and 39 820 men) returned completed questionnaires (44%). Participants were asked to record their height in centimetres (cm). We also requested information on factors related to socioeconomic status during childhood, including the area (rural or urban) in which they resided as a child (aged 10e12 years). The sensitive period for socioeconomic influences on height may be earlier than age 10e12 years, but this age was specified for greater accuracy of recall.
## Statistical analyses
Analyses were conducted separately for men and women. As we were interested in comparing heights across different age groups, participants were initially grouped by birth year. Participants' ages ranged from 15 to 87 years, so we excluded those aged over 65 years (n¼76) to avoid a long tail in the first birth year category. Mean heights (cm) with 95% CI were calculated for each group. A similar comparison was made stratifying by area of residence (rural/urban) at age 10e12 years. The data were further assessed using simple linear regression with height as the outcome and mean-centred birth year (in decades) as the predictor. Stratification of these models by childhood area of residence suggested effect modification, so an interaction term for birth year and area of residence was added to the original model. Other childhood factors associated with attained adult height were assessed using multiple linear regression. All variables related to childhood circumstance were entered into the model, including: attained maternal and paternal level of education; sibling number; birth rank; birth size; whether breast fed in infancy; and the numbers of material resources in the home such as electricity, radio, television or a refrigerator at age 10e12 years. Variables relating to current circumstances (such as current income) were excluded to minimise the possibility of reverse causation.
We had two possible measures of birth size: self-reported birth weight (g) and relative birth size recalled by relatives (small, normal, large). When both questions were answered, the two measures were highly correlated, but the data for both were missing for a large proportion of participants. More (75% vs 40%) responded to the question on birth size recalled by relatives than on birth weight, so we used the former to estimate the effect on height change. Because a large amount of birth size data was missing we have reported the estimates for other childhood factors unadjusted for birth size. Of note, the only estimate materially changed by the addition of birth size to the model was that for breast feeding.
Birth rank was initially analysed in strata of number of siblings and age groups. The results from stratified analyses were similar to unstratified analyses and suggested that those later born were more likely to be taller than those first born, thus only a term for first born versus later born was included in the final model.
A variable indicative of childhood material resources was created by summing the positive responses from the following question: 'Which of the following did your home have when you were 10e12 years old? Electricity (generator or outside line), microwave, refrigerator, electric fan, air conditioning, television, video/tape/CD player, computer, telephone, mobile telephone, water heater, or washing machine.'
Ethics approval was obtained from Sukhothai Thammathirat Open University Research and Development Institute (protocol 0522/10) and the Australian National University Human Research Ethics Committee (protocol 2004344). Informed written consent was obtained from all participants.
# Results
Overall, 86 105 participants (99% of all cohort members) aged under 66 years reported their height. [fig_ref] Table 1: Frequencies of demographic and childhood characteristics for female and male Thai Cohort... [/fig_ref] shows the frequencies of various demographic factors as well as the indicators of childhood circumstances.
We found among both female and male participants that height increased with later birth year [fig_ref] Table 2: Mean height in centimetres for women and men in the Thai Cohort... [/fig_ref].
For women, those born between 1986 and 1990 were almost 3 cm taller than women born between 1940 and 1956. Similarly, men born between 1986 and 1990 were on average 4 cm taller than men born between 1940 and 1956. This difference was even greater when considered jointly with area of residence at age 10e12 years [fig_ref] Table 3: Mean height in cms [/fig_ref] : the youngest group of urban-based men was on average 6.4 cm taller than the oldest group of rural-based men and the equivalent difference for women was 4.3 cm.
Mean heights with 95% CI for men (figure 1A) and women (figure 1B) by rural/urban location at age 10e12 years are shown in figure 1. Linear regression suggested an overall increase in height of 1.50 cm (95% CI 1.36 to 1.63) per decade for male urban dwellers compared with 1.01 cm (95% CI 0.93 to 1.09) per decade for male rural dwellers. For women the estimates were 1.32 cm (95% CI 1.20 to 1.43) for urban dwellers and 0.91 cm (95% CI 0.82 to 0.99) for rural dwellers. The interaction terms for birth year and area of residence (rural/urban) aged 10e12 years were highly significant for both men and women (p<0.0001). [fig_ref] Table 4: Adjusted* estimates of height change [/fig_ref] shows the estimates of effect on height change of various measures of childhood circumstances. Participants' birth year remained a significant predictor of height after adjustment for the measured childhood factors, with approximately half a centimetre increase in height per decade for both women and men.
Several other childhood factors were significantly associated with attained adult height. Educational attainment of both parents was related to height of both men and women. Maternal education level was more important for women, such that those with mothers with tertiary education were on average 1.1 cm (95% CI 0.7 to 1.5 cm) taller than those whose mothers had no formal education; for men, paternal level of education was associated with the larger effect, such that men whose fathers had tertiary education were on average 1.4 cm (95% CI 1.0 to 1.8 cm) taller than men whose fathers had no formal education. Of the factors considered, birth size (recalled by relatives) was associated with the largest effects on height for both sexes. Women who were considered large or overweight at birth were on average approximately 2.8 cm taller than those considered small or underweight at birth, and the corresponding estimate for men was 3.3 cm. Those who reported having been breast fed were on average approximately 0.3 cm taller than those who reported they were not breast fed, although this effect disappeared when birth size was added to the model (0.05 cm; 95% CI À0.18 to 0.27 for women and À0.01 cm; 95% CI À0.33 to 0.31 for men). Having larger numbers of siblings was associated with shorter stature, with a more pronounced effect in men than women. Birth order was also important with both later born men and women tending to be taller than those first born.
In both sexes, those with a larger number of assets in the home at age 10e12 years were on average taller than those who had few assets, although the effect was not as large in women. As some of the technologies included in the assets score did not exist when the older participants were children, in addition to adjusting for birth year, we also stratified by birth year group. The same pattern was evident for all groups (results not shown) so we have presented only the birth year adjusted estimates.
# Discussion
Using data from 86 105 Thai open university students we have found evidence of a substantial increase in average attained adult height among both men and women born between 1940 and 1990, amounting overall to approximately 1 cm increase in height per decade. A rural/urban difference was strikingly evident such that the rate of increase in height has been approximately 0.5 cm per decade greater among men reared in urban areas compared with those from rural areas and 0.4 cm per decade greater for urban compared with rural women. Multivariable analyses showed that other indicators of childhood circumstances, such as parental education levels, household assets, birth size, sibling number and birth rank, were also significantly associated with attained adult height but did not completely account for the effect of birth year.
The strengths of our study include the large numbers of both male and female participants. The study group represents well the overall Thai population in terms of sex distribution, median income, region of residence and religion. [bib_ref] Cohort profile: the Thai Cohort of 87,134 Open University students, Sleigh [/bib_ref] However, individuals in their 20s and 30s were overrepresented in the study, and it is also likely that the study population includes fewer individuals of very high or very low socioeconomic status. Study participants are also slightly more educated than the average Thai. Our results may thus be most applicable to the group of upwardly mobile Thais who have been most affected by the socioeconomic transition of Thailand over the past 50 years.
One limitation of the study was that height was self-reported and overestimation is the norm in this situation including among STOU students,. 14 Another limitation was the crosssectional nature of the data collection, such that height was not recorded at the same age for all participants. Some younger participants would not have reached full adult height at the time of the survey, so we may have underestimated the mean adult height for those born between 1986 and 1990. Furthermore, if height is associated with survival in this population (as it appears to be in others) then it may be that the older participants here are, on average, taller than their Thai birth cohort as a whole. It is possible, therefore, that the height differences between older and younger birth cohorts are actually greater than we have found here. Conversely, age-related shrinkage might mean that the recorded heights of our older cohorts are lower than their attained adult heights. However, we feel this is less likely to be a problem because height was self-reported. Studies (from western populations) suggest that overreporting of height significantly increases with age, probably reflecting a tendency among older people to report attained rather than current height.
Some of the factors we investigated related to childhood circumstances present many years earlier, so there may have been recall errors. However, it is unlikely that taller participants would recall events in a systematically different way to shorter participants and thus bias our results. We also did not have any information on parental or sibling heights so were unable to consider the role of genetic factors in attained adult height. To our knowledge only one study has previously published data on secular changes in attained adult height in Thailand. [bib_ref] Change in mean height of Thai military recruits from 1972 through, Seubsman [/bib_ref] That study used the measured heights of male, Bangkok-based military recruits born between 1951 and 1985, and found that average height increased by approximately 1.3 cm per decade. Urban/rural differences could not be considered in that study and the population may have been of slightly lower socioeconomic status, but the findings overall were in close accord with what we have found here for men. Studies of urban-based populations from China 17 and Hong Kong 18 covering similar time periods suggested similar secular increases of 1.3/1.1 cm per decade and 1.2/0.9 cm per decade, respectively, for men and women. For Japan [bib_ref] Changes in body mass index by birth cohort in Japanese adults: results..., Funatogawa [/bib_ref] and Korea 20 data from longer time periods have suggested slightly greater increases of 1.5 cm (Japan national) and 2 cm (Korea urban) per decade for women and 2 cm per decade for men (Japan).
Attained adult height reflects, among other things, childhood health and nutrition, and thus the greater average heights that we have observed in the younger compared to the older participants suggests a marked improvement in childhood conditions in Thailand over the second half of the 20th century. This is in keeping with the large increase in gross domestic product per capita that has occurred over this period (gross domestic product per capita increased from US$990 in 1950 to US$4267 in 1990; international purchasing power parity dollarsdconstant year 2000 prices).However, the benefit seems to have been unequally distributed as the rate of increase in height appears to have been substantially greater for urban compared with rural dwellers. Ruraleurban differences in height have been reported in European and Asian populations but, at least in the European studies, height differences by locale appear to be decreasing rather than increasing. However, in Thailand increasing income inequality between rural and urban populations has been apparent from the 1960s through to at least the early 1990s, [bib_ref] Income Inequality and Kuznets' Hypothesis in Thailand, Ikemoto [/bib_ref] spanning the years in which the majority of our cohort members were children. The increasing height differences we have observed between rural and urban dwellers therefore probably reflects widening economic disparities between urban and rural communities and concomitant effects on childhood health and nutrition.
Our other findings in relation to childhood circumstances reflect what has been observed in a variety of countries (both developed and developing), including the positive relation between height and birth size, 27e29 parental levels of education, [bib_ref] Secular trends in male adult height 1904e1996 in relation to place of..., Padez [/bib_ref] and number of assets in the family home [bib_ref] Childhood socioeconomic circumstances and adult height and leg length in central and..., Webb [/bib_ref] ; and the inverse relation between height and family size. [bib_ref] Early life influences on adult leg and trunk length in the 1958..., Li [/bib_ref] The suggestion that later born children are on average taller than those first born has also been reported elsewhere, although others have found the inverse to be true. It may be that issues such as birth spacing and cultural influences on care patterns in families account for these differences, but we did not have the relevant data to explore this issue further. Overall, our findings in relation to childhood circumstances most likely reflect the effect on growth of access to resources in early childhood and ultimately the adequacy of nutrition (both quality and quantity) and the ability to control childhood infection. Our findings may also have implications for future health trends in Thailand, particularly in relation to chronic diseases. In general, greater height is associated with lower overall mortality. In particular, greater height is associated with a decreased risk of death from stroke, 4e6 coronary artery disease, respiratory disease 4e6 and external causes. [bib_ref] Adult height and the risk of mortality in South Korean women, Song [/bib_ref] This suggests that the secular increase in height should have beneficial effects on the profile of some chronic diseases in Thailand. A caveat to this is that the prevalence of other risk factors for diseases of the circulatory system is likely to increase as western diets and Mean heights and 95% CI among (a) Thai men and (b) Thai women who resided in rural or urban locations at age 10e12 years.
lifestyles are increasingly adopted. Our work suggests that this could be of particular concern in rural areas where the relative height deficit combined with an increasingly energy-dense diet and sedentary lifestyle could result in substantial levels of obesity. Recent work based on the Thai National Health Examination Surveys suggest indeed that rates of obesity are increasing more rapidly in rural than urban areas. [bib_ref] Prevalence of obesity in Thailand, Aekplakorn [/bib_ref] Height is also positively associated with a number of cancers, particularly those of the breast, prostate and colon. [bib_ref] Height, leg length, and cancer risk: a systematic review, Gunnell [/bib_ref] It is likely, therefore, that rates of these types of cancers (currently relatively rare in Thailand compared with western populations)will increase with the increase in average heights.
In conclusion, we have found evidence of a marked secular increase in the height of Thais in the second half of the 20th century reflecting the rapid socioeconomic development of Thailand over this period. While these changes suggest overall yEstimates calculated using data from those who reported birth size (w75% of cohort), adjusted for all listed factors. zWhen birth size is added to the model the effect of breast feeding is negligible (0.05 cm, 95% CI À0.18 to 0.27 for women and À0.01 cm, 95% CI À0.33 to 0.31 for men).
## What is already known on this subject
Thailand has undergone rapid socioeconomic development over the past 50 years with associate changes in population health. Secular trends in adult height reveal population health transitions, but such changes in Thailand, particularly urban/rural contrasts, are little studied.
## What this study adds
Average heights increased approximately 1 cm per decade in those born between 1940 and 1990, with urban Thais increasing approximately 0.5 cm more per decade than rural dwellers. Parental education, household assets, birth size, family size and birth rank were also significantly associated with height. This marked secular increase in Thai heights reflects improved child health and nutrition, but rural-born Thais have benefited less. They still tend to be of modest stature as adults, and now face a nutrition transition with a food surplus, obesity and its consequences.
positive health outcomes, they may also reveal opportunities for health interventions, particularly for rural-born Thais who appear to have, at least until the latter part of the 20th century, benefited to a lesser extent from the secular trend in height.
[table] Table 1: Frequencies of demographic and childhood characteristics for female and male Thai Cohort Study participants with recorded height *May not add to total due to missing data. [/table]
[table] Table 2: Mean height in centimetres for women and men in the Thai Cohort Study by birth year group [/table]
[table] Table 3: Mean height in cms (SD) by age group for men and women according to whether they lived in an urban/rural area aged 10e12 years [/table]
[table] Table 4: Adjusted* estimates of height change (cm) for a unit change in factors associated with childhood circumstances among Thais *Each factor adjusted for all others listed in the table apart from birth size recalled by relatives. [/table]
|
Subsets and Splits