Cardiology Research and Practice

Cell Signalling Pathways Leading to Novel Therapeutic Strategies in Cardiovascular Disease

Guest Editors: Sidney G. Shaw, David J. Abraham, Daryll M. Baker, and Janice Tsui Cell Signalling Pathways Leading to Novel Therapeutic Strategies in Cardiovascular Disease Cardiology Research and Practice

Cell Signalling Pathways Leading to Novel Therapeutic Strategies in Cardiovascular Disease

Guest Editors: Sidney G. Shaw, David J. Abraham, Daryll M. Baker, and Janice Tsui Copyright © 2012 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in “Cardiology Research and Practice.” All articles are open access articles distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Editorial Board

Atul Aggarwal, USA Hosen Kiat, Australia Fausto J. Pinto, Portugal Peter Backx, Canada Anne A. Knowlton, USA Bertram Pitt, USA J. Brugada, Spain GavinW.Lambert,Australia Robert Edmund Roberts, Canada Ramon Brugada, Canada Chim Choy Lang, UK Terrence D. Ruddy, Canada Hans R. Brunner, Switzerland F. H. H. Leenen, Canada Frank T. Ruschitzka, Switzerland Vicky A. Cameron, New Zealand Seppo Lehto, Finland Christian Seiler, Switzerland David J. Chambers, UK John C. Longhurst, USA Sidney G. Shaw, Switzerland Mariantonietta Cicoira, Italy Lars S. Maier, Germany Pawan K. Singal, Canada Antonio Colombo, Italy Olivia Manfrini, Italy Felix C. Tanner, Switzerland Omar H. Dabbous, USA Gerald Maurer, Austria Hendrik T. Tevaearai, Switzerland N. S. Dhalla, Canada G. A. Mensah, USA G. Thiene, Italy Firat Duru, Switzerland Robert M. Mentzer, USA H. O. Ventura, USA Vladim´ır Dzav˘ ´ık, Canada Piera Angelica Merlini, Italy Stephan von Haehling, Germany Gerasimos Filippatos, Greece Marco Metra, Italy James T. Willerson, USA Mihai Gheorghiade, USA Veselin Mitrovic, Germany Michael S. Wolin, USA Enrique P. Gurfinkel, Argentina Joseph Brent Muhlestein, USA Michael Wolzt, Austria P. Holvoet, Belgium Debabrata P. Mukherjee, USA Syed Wamique Yusuf, USA H. A. Katus, Germany J. D. Parker, Canada Contents

Cell Signalling Pathways Leading to Novel Therapeutic Strategies in Cardiovascular Disease, Sidney G. Shaw, David J. Abraham, Daryll M. Baker, and Janice Tsui Volume 2012, Article ID 475094, 3 pages

The Continuing Challenges of Translational Research: Clinician-Scientists Perspective, Shervanthi Homer-Vanniasinkam and Janice Tsui Volume 2012, Article ID 246710, 5 pages

Targeted In Situ Gene Correction of Dysfunctional APOE Alleles to Produce Atheroprotective Plasma ApoE3 Protein, Ioannis Papaioannou, J. Paul Simons, and James S. Owen Volume 2012, Article ID 148796, 16 pages Cell Therapies for Heart Function Recovery: Focus on Myocardial Tissue Engineering and Nanotechnologies, Marie-Noelle¨ Giraud, Anne Graldine Guex, and Hendrik T. Tevaearai Volume 2012, Article ID 971614, 10 pages

The Emerging Role of TLR and Innate Immunity in Cardiovascular Disease, Rolf Spirig, Janice Tsui, and Sidney Shaw Volume 2012, Article ID 181394, 12 pages

Nitric Oxide Manipulation: A Therapeutic Target for Peripheral Arterial Disease?, Gareth Williams, Xu Shi-Wen, David Abraham, Sadasivam Selvakumar, Daryll M. Baker, and Janice C. S. Tsui Volume 2012, Article ID 656247, 7 pages

MicroRNAs in Postischemic Vascular Repair, Andrea Caporali and Costanza Emanueli Volume 2012, Article ID 486702, 7 pages

Connexins and Diabetes, Josephine A. Wright, Toby Richards, and David L. Becker Volume 2012, Article ID 496904, 8 pages

Potential of Novel EPO Derivatives in Limb Ischemia, Dhiraj Joshi, Janice Tsui, Rebekah Yu, Xu Shiwen, Sadasivam Selvakumar, David J. Abraham, and Daryll M. Baker Volume 2012, Article ID 213785, 5 pages Stromal-Cell-Derived Factor-1 (SDF-1)/CXCL12 as Potential Target of Therapeutic Angiogenesis in Critical Leg Ischaemia,TeikK.Ho,X.Shiwen,D.Abraham,J.Tsui,andD.Baker Volume 2012, Article ID 143209, 7 pages

The Diabetic Heart: Too Sweet for Its Own Good?, Hannah J. Whittington, Girish G. Babu, Mihaela M. Mocanu, Derek M. Yellon, and Derek J. Hausenloy Volume 2012, Article ID 845698, 15 pages

The LOX-1 Scavenger Receptor and Its Implications in the Treatment of Vascular Disease,M.WTwigg, K. Freestone, S. Homer-Vanniasinkam, and S. Ponnambalam Volume 2012, Article ID 632408, 6 pages

Clinical Applications of Remote Ischemic Preconditioning, Kristin Veighey and Raymond J. MacAllister Volume 2012, Article ID 620681, 9 pages

Toll-Like Receptors in Ischaemia and Its Potential Role in the Pathophysiology of Muscle Damage in Critical Limb Ischaemia, Hemanshu Patel, Sidney G. Shaw, Xu Shi-Wen, David Abraham, Daryll M. Baker, and Janice C. S. Tsui Volume 2012, Article ID 121237, 13 pages Genomic Research to Identify Novel Pathways in the Development of Abdominal Aortic Aneurysm, Seamus C. Harrison, Anastasia Z. Kalea, Michael V. Holmes, Obi Agu, and Steve E. Humphries Volume 2012, Article ID 852829, 8 pages

Modulation of Fibrosis in Systemic Sclerosis by Nitric Oxide and Antioxidants,AudreyDooley, K. Richard Bruckdorfer, and David J. Abraham Volume 2012, Article ID 521958, 9 pages Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 475094, 3 pages doi:10.1155/2012/475094

Editorial Cell Signalling Pathways Leading to Novel Therapeutic Strategies in Cardiovascular Disease

Sidney G. Shaw,1 David J. Abraham,2 Daryll M. Baker,3 and Janice Tsui4

1 Vasoactive Peptide Group, Department of Clinical Research, University of Bern, 3004 Bern, Switzerland 2 Centre for Rheumatology and Connective Tissues Diseases, Royal Free Hospital, Pond Street, London NW3 2QG, UK 3 Vascular Unit, Academic Department of Surgery, Royal Free Hospital, Pond Street, London NW3 2QG, UK 4 Division of Surgery & Interventional Science, University College London, Royal Free Campus Pond Street, London NW3 2QG, UK

Correspondence should be addressed to Sidney G. Shaw, [email protected]

Received 21 November 2012; Accepted 21 November 2012 Copyright © 2012 Sidney G. Shaw et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. Overview and animal models suggests that such intervention inhibits atherosclerosis. Anti-atherogenic strategies that target LOX- Cardiovascular disease is a leading cause of morbidity and 1 function using gene therapy or small molecule inhibitors mortality with considerable financial impact worldwide. would be new ways to address the increasing incidence of This special issue explores recent developments and novel vascular disease. strategies to target emerging risk factors central to the The role of special inflammatory molecules in the complex molecular mechanisms regulating the cardiovas- initiation and promotion of vascular inflammation is also cular disease process. The concept was prompted by dis- under investigation. In particular the link between sterile cussions arising from the UCL-Royal Free International inflammation, ischemia, and the innate immune system is Cardiovascular Diseases Workshops held each year at the providing important new information concerning the role RoyalFreeHospitalcampusofUniversityCollegeLondonin of specific Toll-like receptors and endogenous molecules collaboration with the University of Bern, Switzerland. These released in response to tissue injury. As sentinels of innate annual international workshops provide an integrated forum immunity TLRs are molecular pattern recognition receptors for clinicians and scientists involved in multidisciplinary that recognize exogenous as well as tissue-derived molec- studies in cardiovascular medicine to discuss recent advances ular danger signals that promote chronic inflammation aimed at integrating the latest findings in basic research more [1, 2]. Interestingly, these endogenously derived “danger” ffi quickly and e ciently into medical practice. molecules appear to activate and induce differential TLR The causes of cardiovascular disease are numerous but dimerisation compared to exogenously derived bacterial emerging areas of therapeutic interest are equally diverse and or viral components. Consequently modulation of TLR include novel molecular mechanisms by which cardiovas- signaling by specific TLR agonists or antagonists, alone cular risk factors and mediators, including reactive oxygen or in combination, may, therefore, be a novel therapeu- species and scavenger receptors, interact to regulate vascular tic approach to treat various cardiovascular inflammatory cell proliferation and extra cellular matrix remodeling that conditions such as atherosclerosis, secondary microvascular play important roles during angiogenesis and cell response complications of diabetes, and peripheral arterial disease to injury and ischemia. (PAD) without compromising the normal immune response. The scavenger receptor LOX-1, for example, mediates Additional signaling pathways and novel strategies OxLDL endocytosis via a clathrin-independent internaliza- that may prevent or slow the development of trans- tion pathway. Administration of LOX-1 antibodies in cellular plant arteriopathy and post ischemic reperfusion injury 2 Cardiology Research and Practice are likewise now beginning to be defined. Mechanisms EPO derivatives, however, are tissue protective with- by which haemodynamic forces such as pressure, stretch, out the haematopoietic side effects and preclinical stud- and fluid shear stress are sensed by cells within the ies have demonstrated their potential effectiveness and cardiovascular system are helping in the understanding safety. of the atheroprotective benefits of steady laminar blood Cytokine targeting may also play a future role in the ther- flow. apy of peripheral vascular disease, in particular chemokine Here, there is also evidence that nitric oxide (NO) stromal-cell-derived factor-1 (SDF-1 aka CXCL12). Biologi- and its endogenous inhibitor asymmetric dimethylarginine cal effects of SDF-1 are mediated by the chemokine receptor (ADMA) play significant roles. Recent experimental work CXCR4, a 352-amino-acid rhodopsin-like transmembrane- implicates the ADMA-NO pathway and highlights the specific G protein-coupled receptor (GPCR). There is evi- potential of manipulating this as a novel adjunct therapy. dence that the administration of SDF-1 increases blood flow Related, mechanisms may be at play in relation to the and perfusion via recruitment of endothelial progenitor cells damaging effects of oxidative stress on the development (EPCs). of enhanced fibroblast activity resulting in fibrosis of the As more becomes known, molecular genomic approach- skin, heart, and lungs, vascular dysfunction and ultimately es are increasingly being used. MicroRNAs (miRNAs) are internal organ failure, and death in scleroderma. Research endogenous, small, noncoding RNAs that negatively control suggests that the free radical nitric oxide (NO), a key gene expression of target mRNAs. This phenomenon may mediator of oxidative stress, can profoundly influence the be of benefit in regulating post-ischemic angiogenesis and early microvasculopathy, and possibly the ensuing fibrogenic vascular repair and analysis of circulating miRNAs may response. Recently, animal models and human studies have potentially be useful as potential biomarkers in ischemic also identified dietary antioxidants, such as epigallocatechin- diseases. Targeted in situ gene correction is currently being 3-gallate (EGCG), which function as a protective system explored as a means to modulate Apolipoprotein E (ApoE) against oxidative stress and fibrosis. Hence, targeting EGCG gene activity. APOE, a 34-kDa circulating glycoprotein, may prove a possible candidate for therapeutic treatment has pleiotropic anti-atherogenic functions and hence is a ff aimed at reducing both oxidant stress and the fibrotic e ects candidate to treat hypercholesterolaemia and atheroscle- associated with scleroderma. rosis. In particular, we have the possibility and potential Manipulation of cyclic GMP and cyclic AMP is a further benefit of combining two technological advances to repair emerging target for intervention in cardiovascular diseases aberrant APOE genes: (i) an engineered endonuclease to such as hypertension, atherosclerosis, and heart failure. introduce a double-strand break in exon 4, which con- Drugs that enable the manipulation of these systems repre- tains the common, but dysfunctional, ε2andε4alleles; sent an exciting new area of pharmaceutical development. (ii) an efficient and selectable template for homologous Basic research has also identified a number of novel inter- recombination (HR) repair, namely, an adeno-associated ventions that stimulate innate resistance of tissues to viral (AAV) vector, which harbors the wild-type APOE ischemia-reperfusion injury that may have important impli- sequence. This technology is currently applicable ex vivo,for cations in the management of stroke and myocardial example, to target haematopoietic or induced pluripotent infarction. In particular, clinical trial data underpin one of stem cells, and also for in vivo organ gene targeting. these “conditioning” strategies, the phenomenon of remote It is to be hoped that such emerging technology will ischemic preconditioning. This may in particular provide a eventually translate into patient therapy to reduce CVD novel cardioprotective strategy for the diabetic heart which risk. evidence suggests may be more susceptible to ischemic Other genome-wide association studies (GWAS) have reperfusion injury. revolutionized research into genetic variants that under- Additionally, the study of connexins (Cxs) and cell- pin the development of many complex diseases including to-cell interactions via gap junctional communication is abdominal aortic aneurysm and allowed the exploration of becoming an active area of investigation. Changes in Cxs mechanisms by which identified loci may contribute to its found in diabetes are associated with both direct effects development. Studies have also highlighted the potential of within the vasculature and indirect effects, by impairment of post-GWAS analytical strategies to improve our understand- homeostasis in vital organs such as liver and kidney. Recent ing of the disease further. data suggests that Cxs targeting may alleviate some of the Finally, progress in the identification of key mechanisms symptoms of microvascular complications, as demonstrated regulating tissue repair, and the potential of stem cell therapy in recent work using topical Cx43 AsODN (antisense and tissue engineering, offer promising new possibilities in oligodeoxynucleotide) gel treatment. Cxs may also be used as bypass grafting and recovery from cardiac injury. future predictors of both diabetes progression and severity. We hope that readers will find that this special issue Similarly, erythropoietin (EPO) has tissue-protective addresses the important challenges, opportunities, and properties, but increases the risk of thromboembolism recent developments that are currently being exploited in by raising haemoglobin concentration. New generation cardiovascular research with the potential to promote the Cardiology Research and Practice 3 implementation of novel translational strategies for optimiz- ing the care and treatment of patients with cardiovascular disease.

Acknowledgments These studies were supported by the European Association for the study of Diabetes, Medical Research Council UK, Royal Society UK, Swiss National Science Foundation, Wellcome Trust UK, Circulation Foundation UK, and the British Heart Foundation. Sidney G. Shaw David J. Abraham Daryll M. Baker Janice Tsui

References

[1] P. Matzinger, “The evolution of the danger theory,” Expert Review of Clinical Immunology, vol. 8, no. 4, pp. 311–317, 2012. [2] W. Land, H. Schneeberger, S. Schleibner et al., “The beneficial effect of human recombinant superoxide dismutase on acute and chronic rejection events in recipients of cadaveric renal transplants,” Transplantation, vol. 57, no. 2, pp. 211–217, 1994. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 246710, 5 pages doi:10.1155/2012/246710

Review Article The Continuing Challenges of Translational Research: Clinician-Scientists’ Perspective

Shervanthi Homer-Vanniasinkam1, 2 and Janice Tsui2

1 Leeds Vascular Institute, Leeds General Infirmary, Leeds LS1 3EX, UK 2 Division of Surgery and Interventional Science, University College London, Royal Free Campus, London NW3 2QG, UK

Correspondence should be addressed to Shervanthi Homer-Vanniasinkam, [email protected]

Received 5 July 2012; Accepted 25 July 2012

Academic Editor: Sidney G. Shaw

Copyright © 2012 S. Homer-Vanniasinkam and J. Tsui. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Over the last twenty years, revolutionary advances in biomedicine including gene therapy, stem cell research, proteomics, genomics and nanotechnology have highlighted the progressive need to restructure traditional approaches to basic and clinical research in order to facilitate the rapid, efficient integration and translation of these new technologies into novel effective therapeutics. Over the past ten years, funding bodies in the USA and UK such as the National Institutes of Health (NIH) and the Medical Research Council (MRC) have been driving translational research by defining and tackling the hurdles but more still remains to be achieved. This article discusses the ongoing challenges translational researchers face and outlines recent initiatives to tackle these including the new changes to translational funding schemes proposed by the NIH and the MRC and the launch of the “European Advanced Translational Research InfraStructure in Medicine” (EATRIS). It is anticipated that initiatives such as these will not only strengthen translational biomedical research programmes already initiated but should lead to rapid benefits to patients and society.

1. Introduction More recently, this has been redefined to include 3 phases in TR, with the second phase being subdivided. Thus, in this Translational research (TR) is a relatively new area of inves- new model T1 describes basic science to clinical science, T2 tigation that ideally involves the integrated application of clinical science to clinical practice, and T3 is used to denote innovative technologies that encompass multiple disciplines the translation of clinical practice to more widespread health including physiology, pathophysiology, natural history of improvements [2]. disease, genetics, and proof-of-concept studies of drugs and Whilst the importance and benefits of TR are now devices [1]. clear to most, significant challenges faced by investigators TR describes a continuum of research in which basic have also been recognised. This article focuses on some of science discoveries are utilized to prevent or treat human those barriers and discusses whether initiatives over the past disease. It is an iterative process wherein scientific discoveries decade to drive TR forward as broken them down. are integrated into clinical applications and, conversely, clinical observations are used to generate research foci for basic science: the “bench to bedside and back to bench” 2. Barriers to Translational Research approach. Initially, 2 phases in TR were described: The essential role of TR in ensuring rapid progression of T1. basic science discoveries used to develop new treat- basic scientific knowledge to patient benefit has been empha- ments for disease (“bench to bedside”), sised for more than 3 decades [3]. During this period, the T2. research aimed at improving utilization of proven barriers to TR have also been identified. The main issues that therapies in clinical practice and community settings (“bed- have been repeatedly debated include cultural differences side to community”). between basic scientists and clinicians, lack of resources in 2 Cardiology Research and Practice terms of workforce and infrastructure, and the complex clinical environment under the guidance of experienced clin- regulatory environment [4]. Cultural differences between the icians. Both groups should also receive training in research two groups of investigators largely stem from the lack of methodology including clinical trial design and conduct, communication, differences in education and training, and and medical statistics. Multidisciplinary academic medical different goals and reward mechanisms. Regulatory issues centres are well placed to offer such training programmes are intimidating to both basic scientists and clinicians and and indeed, the USA Clinical and Translational Sciences encompass ethics involved in human research, tissue bank- Awards (CTSA) scheme has as its central strategy five goals ing and material transfer regulations, intellectual property within which training and career development are firmly rights, toxicology and manufacturing regulations, Food and embedded [8]. Drug Administration (FDA) and Medicines and Healthcare Goal 1. Build national clinical and translational research Regulatory Agency (MHRA) approvals, study sponsorship capability. and insurance, as well as trial and data monitoring. These Goal 2. Provide training and improving the career are becoming even more complex with expanding work in development of clinical and translational scientists. the fields of cell and gene therapies and tissue engineering. Goal 3. Enhance consortium-wide collaborations. Resource problems include lack of trained interdisciplinary Goal 4. Improve the health of our communities and the staff to support investigations throughout the TR cycle, nation. protected time for research particularly for clinicians, as Goal 5. Advance T1 translational research. well as access to shared resources. Together, these issues Various M.D./M.B. Ph.D. programmes and master pro- contribute to various checkpoints between the phases of TR grammes focussing on TR are now also available and further including the “valley of death” that exists between preclinical training supported by dedicated fellowships are providing a research and clinical trials [5]. clearer career pathway for both clinician-scientists and basic scientists. Despite these encouraging schemes to entice and train 3. Streamlining Regulatory Issues young researchers in translational projects, there are still a number of challenges that need to be addressed if we are In addition to cultural differences, efforts have been made to to ensure that the brightest minds will continue to enter streamline health research approvals and reduce regulatory this field of investigation. In clinical medicine, there are still burdens. The NIHR developed the Integrated Research some myths and misconceptions about the role of research Application System (IRAS) in 2008, an integrated online in training programmes such as: system which captures information required for approval application from several review bodies including the NHS (i) research has no place in a clinician’s training pro- ffi R&D o ces, research ethics committees, MHRA, Adminis- gramme; tration of Radioactive Substances Advisory Committee, Gene Therapy Advisory Committee, and the National Information (ii) there is no place for basic research in the clinical Governance Board. The Research Passport system was also training programme; set up to streamline the issuing of honorary research (iii) every surgical trainee should do some basic research; contracts for researchers not employed by the NHS. In December 2011, the NHS launched the Health Research (iv) clinicians should only engage in clinical research. Authority which will play a key role in research governance These need to be tackled at faculty, institutional, and and further streamline the approval process [6]. In the USA, national levels. the National Center for Advancing Translational Sciences Another challenge is that of career progression and (NCATS; http://www.ncats.nih.gov/) was established this promotion, and obtaining a tenured post in academic year (2012) with its mission to “catalyze the generation of medicine or in a university science department. Current innovative methods and technologies that will enhance the criteria for promotion still rely heavily on individual research development, testing, and implementation of diagnostics output such as high impact publications, grants, and invited and therapeutics across a wide range of human diseases and lectures. Investigators involved in TR may not be able to conditions.” This would include dealing with bottlenecks produce the required evidence of their contribution such as such as regulatory issues in the TR process to make it “more an adequate publication record, since translational projects efficient, less expensive and less risky” [7]. generally take longer to complete. They are also likely to be part of a fairly large interdisciplinary team and their 4. Training and Mentoring role in publications and on grants may not be apparent. Indeed the latter may be difficult to assess since they may Whilst regulatory issues are a major factor determining the be working outside their recognised disciplines. To improve success of translational projects, it is also imperative that this situation, institutions need to ensure that their tenure we train and develop a pipeline of clinician-scientists and and promotions systems are able to evaluate and recognise applied scientists who are comfortable in dealing with the the contributions investigators conducting TR make. Many continuing challenges of TR. During their training, clinician- institutions are working towards this, for example, by using scientists would benefit from experience in laboratory reviewers with diverse expertise and understanding of TR research whilst basic scientists should have exposure to the and moving away from “traditional” assessment criteria. Cardiology Research and Practice 3

Thelackofrolemodelsandmentorsinthisareahas As discoveries move from “bench to bedside”, invaluable been identified as a significant barrier by researchers trying information may be gained by moving back from the to engage in TR. Mentorship is increasingly recognised as “bedside to the bench”,for example, to understand important invaluable in the development of early career researchers underlying mechanisms and unexpected findings and to into independent investigators and most institutions have make improvements. Similar barriers work against this arm in place mentoring programmes. In the UK, the Academy of TR, such as poor communication of clinical challenges to of Medical Sciences has an effective national mentoring basic scientists and lack of funding beyond the clinical trial. and outreach scheme supported by the Department of Unless researchers, be they clinical or scientific, can see “the Health and the NIHR which has been in place since 2002. light at the end of the (TR) tunnel,” we are likely to lose This scheme specifically offers one-to-one mentoring for out on capturing, and captivating, high calibre translational early career clinician-scientists by Academy Fellows who are researchers. Some of the TR initiatives are still in their early leading scientists in their fields [9]. Further, the importance phases, so it is likely that further improvements will be seen of training of effective mentors is also now recognised [10]. in the near future.

5. Continuing Problems 6. Current Funding and Critical Issues: How Far Despite increased efforts over the past decade, the effective- Have We Come? ness of these initiatives is still unclear. Outcomes of TR, particularly in the short term, have been difficult to measure, To overcome these barriers, increased sustained funding mainly due to the long timelines between discovery and clini- dedicated to TR is clearly essential to enable innovative cal implementation. Traditional measures of research output approaches towards research to be developed. These include such as contribution to high-impact publications may also be the building of research units that incorporate multidisci- difficult to assess since TR publications usually involve mul- plinary groups which may involve bioinformaticians, statis- tiple, interdisciplinary authors with varying roles and contri- ticians, engineers, basic scientists, and clinicians; increas- butions. From the investigators’ perspective, whilst there are ing expert support in regulatory issues and clinical trial examples of successful partnerships [11], recent discussions design and conduct; as well as the initiation of forums for and publications continue to cite similar issues as barriers to interdisciplinary discussion. The latter encourages discussion TR across different specialties, both by basic scientists and by and networking between basic scientists and clinicians, clinician-scientists. For example, in 2009, the NIH Task Force initiation of interdisciplinary collaborations and the building on Research in Emergency Medicine held a series of round- of appropriate research teams. table discussions on emergency care research, with the aims In the USA and the UK, significant resources have already of identifying key research questions in emergency care and been directed towards TR via key initiatives. In the USA, barriers to research in emergency care [12–15]. The main the National Institutes of Health (NIH) launched the NIH barriers identified across the areas of emergency care were Roadmap in 2003 aimed at supporting TR [18], followed shortage of trained investigators, lack of role models and by the Clinical and Translational Science Awards (CTSA) training opportunities, inadequate protected research time, scheme in 2006 [8]. The CTSA scheme aimed to develop poorly defined research-based career paths, the culture of a consortium of institutes that would transform clinical valuing clinical care over research, poor infrastructure, lack and translational research; 60 academic medical centres of interdisciplinary research collaborations, lack of relevant are now funded, receiving USA$500 million annually. In funding streams, and ethical and regulatory issues. Of the the UK, the National Institute of Health Research (NIHR) recommendations made to address these issues, it was felt spent £45 million to fund the first Biomedical Research that formation and development of emergency care clinical Centres (BRC) and Biomedical Research Units (BRU) in research networks will be particularly beneficial to provide 2007 and 2008 within NHS and university partnerships to sharedinfrastructureandprojectsupport,aswellasto drive translational research. A second round was launched in increase patient accruals and trial size and improve efficiency 2011, spending £775 million on 11 BRCs and 20 BRUs. BRCs [15]. The Federation of American Societies for Experimental supportTRacrossawiderangeofdiseaseareaswhileBRUs Biology (FASEB) held a symposium entitled Engaging Basic are smaller groups working in priority areas including car- Scientists in Translational Research: Identifying Opportunities, diovascular disease, dementia, hearing problems, nutrition Overcoming Obstacles in March 2011 to identify and address and lifestyle, gastrointestinal, musculoskeletal, and respira- issues important to basic scientists conducting TR [16]. tory disease [19]. Similarly, the MRC in a further expansion The main challenges identified again included differences plan at the end of 2011 announced its commitment to a in culture and mindset between basic and clinical scien- new £354m investment over the next 4 years. Specific calls tists, insufficient or nonsupportive infrastructure (including were broadened to include stratified medicine research. This regulatory issues), difficulties developing and sustaining aims at understanding why groups of patients with the same collaborations, inadequate training, insufficient funding, and diagnosis often differ in response to the same treatment lack of incentives and rewards. Recommendations were made and sets out to develop UK-wide research consortia each as to how to tackle these hurdles, emphasizing the roles focussed on specific disease areas. The goal is to stratify of institutions, professional societies, funding organisations, disease processes and develop a clearer understanding of the and individual scientists [17]. underlying mechanisms that will eventually lead to future 4 Cardiology Research and Practice tailoring of treatment to individual patients and improve cost The MRC is investing £10 M in unique open innovation col- effectiveness. Additionally, the associated “Challenge Grants” laboration with AstraZeneca which has also formed partner- will support ambitious, challenge-led studies of disease ships with the Karolinska Institute to screen tissue banks for mechanisms in humans. These studies will produce major novel proteomic biomarkers that may signal underlying dis- new mechanistic insights into human disease, with potential ease or susceptibility to risk. Similarly, Gentris Corporation applicability to new therapeutic approaches and opportuni- has linked with academia in China to develop and validate ties for “reverse translation” to more basic research. Other new genomic markers while Affymetrix and Leica Microsys- major funding bodies, including the Wellcome Trust, the tems are developing complementary companion diagnostic Medical Research Council, the British Heart Foundation, tests for personalized medicine by significantly increasing and Heart Research UK, have also focussed support on TR. high through-put, in situ hybridization analyses. These part- For example, Heart Research UK funded almost £500 K in nerships provide academic researchers with unprecedented 2010 on four projects under its Translational Research Grants access to high-quality clinical and pre-clinical compounds, programme [20] whilst the Wellcome Trust spent £59 million the building blocks of new drugs, in order to help better on TR in 2010 [21]. understand a spectrum of diseases with a view to exploring Nevertheless, with the growing realisation that the new treatments. Such collaboration have the potential to research infrastructure needed to cope with the broad spec- be transformational in stimulating relationships between trum of different diseases and treatment requirements can- academia and industry. The findings of the research will help not be comprehensively covered by any individual country deliver growth to the pharmaceutical and biotech industry. alone, the beginning of 2012 also saw major commitments to TR widening and expanding with the initiation and launch of a new European research infrastructure initiative, the 7. Articles in the Current Issue European Advanced Translational Research InfraStructure in Medicine (EATRIS). Despite concerns, there does seem to be a burgeoning of pub- Within this pan-European network of eleven major insti- lications pertaining to TR. Thus, it is heartening to read in tutes across Europe, several governments and research bodies this special issue of Cardiology Research and Practice articles aim to ensure that the best research facilities for TR will devoted to translational science and research. These papers be shared by the whole research community. Central broad address a range of cardiovascular diseases including periph- areas for study include imaging and tracer development, eral arterial disease, abdominal aortic aneurysms, systemic biomarkers, vaccines, advanced therapy medicinal products, sclerosis, and pulmonary artery hypertension, discussing and infectious diseases. recent findings on disease initiation and development as well In London, the new £73 million Imperial Centre for as novel therapies that are currently being explored. They Translational and Experimental Medicine (ICTEM) will play illustrate how improved knowledge of basic pathological a major role in the programme by combining laboratory processes and their regulatory mechanisms are contributing space for 450 scientists including chemists, biologists, and to our understanding of specific diseases. For example, Spirig engineers with dedicated facilities for assessing and devel- et al. [22] and Patel et al. [23] discuss the increasingly oping novel medical therapies through clinical trials. Initial recognised link between innate immune signalling and priority areas will include cancer, infection, cardiovascular, cardiovascular diseases whilst Williams et al. [24]and metabolic, and neurological diseases. The overall concept Dooley et al. [25] demonstrate the potential of modulating of the EATRIS programme is to bring together the best the ubiquitous nitric oxide pathway in the treatment of biomedical and clinical scientists and allow them access to peripheral arterial disease and systemic sclerosis, respectively. state-of-the-art research facilities, techniques, and expertise The impact of advanced technologies on TR is clearly seen in covering the entire developmental chain from validation, the field of genetics and in this issue, Harrison and colleagues compound libraries, good manufacturing practice, and so describe the use of genomics to study pathways involved in forth to research hospitals. Education is also a key element of abdominal aortic aneurysm development [26] whist Capo- the programme that will train scientists as well as technicians rali & Emanueli discuss the role of microRNAs in vascular and nurses to think outside their immediate disciplines and repair [27]. In terms of novel therapies, contributions on provide information about clinical needs and regulatory remote ischaemia preconditioning [28], stem cell therapies requirements. [29],andtissueprotectivestrategies[30] illustrate their In the light of these initiatives, industry has also taken up potential in the cardiovascular field. This certainly indicates the challenge of TR and funded projects independently, or in that TR is being pursued intensively by these investigators partnership with various funding bodies or with major aca- and departments, and in subjects that address areas of demic health centres. A fine example of an industry-medical either unmet, or indeed, unrealized, clinical need. It is also charity partnership is the MSD Wellcome Trust Hilleman encouraging to see that most of the papers in this issue come Laboratories which were established in India in 2009, as a from multidisciplinary groups of clinicians and scientists at £90 million joint venture between MSD Laboratories India various stages of their careers working in close partnership. LLC (a wholly owned subsidiary of Merck & Co. Inc. USA) Collaborations such are these are the driving forces of TR in and the Wellcome Trust. The aim of this collaboration is to terms of bringing together expertise, integrating basic science develop high-impact affordable vaccines for people in devel- and clinical applications, and training future generations of oping countries. Other companies have since followed suit. clinician-scientists. Cardiology Research and Practice 5

8. Concluding Remarks [16] C. M. Weston, E. B. Bass, D. E. Ford, and J. B. Segal, “Faculty involvement in translational research and interdisciplinary TR is not for the faint hearted. The constant challenges of collaboration at a US academic medical center,” Journal of teaching, researching, publishing, and competing for limited Investigative Medicine, vol. 58, no. 6, pp. 770–776, 2010. sources of funding, coupled with pursuing career aims and [17] http://www.faseb.org/Portals/0/PDFs/opa/TranslationalReport- ambitions, can seem daunting. However, it can also be a FINAL.pdf. deeply satisfying and exhilarating endeavour, especially when [18] E. Zerhouni, “The NIH roadmap,” Science, vol. 302, no. 5642, the fruits of the experimental laboratory are translated into pp. 63–72, 2003. improved healthcare delivery to our patients. [19] http://www.nihr.ac.uk/infrastructure/Pages/default.aspx. Finally, we feel that TR has a central and pivotal role in [20] http://www.heartresearch.org.uk/grants/translational. harnessing significant discoveries in biomedical science for [21] http://www.wellcome.ac.uk/News/2011/News/WTVM051890 the benefit of our patients. To the sceptics who ask: “Where is .htm. the evidence that TR matters?” we would answer: “As with Sir [22] R. Spirig, J. Tsui, and S. Shaw, “The emerging role of TLR and innate immunity in cardiovascular disease,” Cardiology Christopher Wren’s monuments, the evidence is all around Research and Practice, vol. 2012, Article ID 181394, 12 pages, us.” 2012. [23] H. Patel, S. Shaw, X. Shi-Wen et al., “Toll-like recep- tors in ischaemia and its potential role in the pathophysiol- References ogy of muscle damage in critical limb ischaemia,” Cardiology Research and Practice, vol. 2012, Article ID 121237, 13 pages, [1] E. A. Zerhouni, “Translational and clinical science—time for a 2012. new vision,” New England Journal of Medicine, vol. 353, no. 15, [24] G. Williams, X. Shi-Wen, D. Abraham et al., “Nitric oxide pp. 1621–1623, 2005. manipulation: a therapeutic target for peripheral arterial [2] J. M. Westfall, J. Mold, and L. Fagnan, “Practice-based disease,” Cardiology Research and Practice, vol. 2012, Article ID research—“Blue highways” on the NIH roadmap,” Journal of 656247, 7 pages, 2012. the American Medical Association, vol. 297, no. 4, pp. 403–406, [25] A. Dooley, R. Bruckdofer, and D. Abraham, “Modulation of 2007. fibrosis in systemic sclerosis by nitric oxide and antioxidants,” [3] S. Wolf, “The real gap between bench and bedside,” New Cardiology Research and Practice, vol. 2012, Article ID 521958, England Journal of Medicine, vol. 290, no. 14, pp. 802–803, 9 pages, 2012. 1974. [26] S. C. Harrison, A. Z. Kalea, M. V. Holmes et al., “Genomic [4] W. N. Hait, “Translating research into clinical practice: delib- research to identify novel pathways in the development of erations from the American Association for Cancer Research,” abdominal aortic aneurysm,” Cardiology Research and Practice, Clinical Cancer Research, vol. 11, no. 12, pp. 4275–4277, 2005. vol. 2012, Article ID 852829, 8 pages, 2012. [5] D. Butler, “Translational research: crossing the valley of death,” [27] A. Caporali and C. Emanueli, “microRNAs in post-ischemic Nature, vol. 453, no. 7197, pp. 840–842, 2008. vascular repair,” Cardiology Research and Practice, vol. 2012, [6] http://www.nihr.ac.uk/systems/Pages/default.aspx. Article ID 486702, 7 pages, 2012. [7] http://www.ncats.nih.gov/. [28] K. V. Veighey and R. J. MacAllister, “Clinical applications of [8] http://www.ctsacentral.org/. remote ischaemic preconditioning,” Cardiology Research and [9] http://www.acmedsci.ac.uk/index.php?pid=55. Practice, vol. 2012, Article ID 620681, 9 pages, 2012. [10]M.O.Johnson,L.L.Subak,J.S.Brown,K.A.Lee,andM. [29] M. N. Giraud, G. Guex, and S. G. Shaw, “Cell therapies for D. Feldman, “An innovative program to train health sciences heart function recovery,” Cardiology Research and Practice, vol. researchers to be effective clinical and translational research 2012, Article ID 971614, 10 pages, 2012. mentors,” Academic Medicine, vol. 85, no. 3, pp. 484–489, [30] D. Joshi, J. Tsui, R. Yu et al., “Potential of novel EPO derivatives 2010. in limb ischemia,” Cardiology Research and Practice, vol. 2012, [11] H. H. Kong and J. A. Segre, “Bridging the translational Article ID 213785, 5 pages, 2012. research gap: a successful partnership involving a physician and a basic scientist,” Journal of Investigative Dermatology, vol. 130, no. 6, pp. 1478–1480, 2010. [12] W. J. Koroshetz, R. Conwit, H. Auchincloss et al., “NIH and research in the emergency setting: progress, promise, and process,” Annals of Emergency Medicine, vol. 56, no. 5, pp. 565– 567, 2010. [13] G. D’Onofrio, E. Jauch, and A. Jagoda, “NIH roundtable on opportunities to advance research on neurologic and psychiatric emergencies,” Annals of Emergency Medicine, vol. 56, pp. 551–564, 2010. [14] C. B. Cairns, R. V. Maier, and O. Adeoye, “NIH roundtable on emergency trauma research,” Annals of Emergency Medicine, vol. 56, pp. 538–550, 2010. [15] A. H. Kaji, R. J. Lewis, T. Beavers-May et al., “Summary of NIH Medical-Surgical Emergency Research Roundtable held on April 30 to May 1, 2009,” Annals of Emergency Medicine, vol. 56, pp. 522–537. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 148796, 16 pages doi:10.1155/2012/148796

Review Article Targeted In Situ Gene Correction of Dysfunctional APOE Alleles to Produce Atheroprotective Plasma ApoE3 Protein

Ioannis Papaioannou, J. Paul Simons, and James S. Owen

Division of Medicine, UCL Medical School, Royal Free Campus, Rowland Hill Street, London NW3 2PF, UK

Correspondence should be addressed to James S. Owen, [email protected]

Received 3 August 2011; Accepted 30 January 2012

Academic Editor: Sidney G. Shaw

Copyright © 2012 Ioannis Papaioannou et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Cardiovascular disease is the leading worldwide cause of death. Apolipoprotein E (ApoE) is a 34-kDa circulating glycoprotein, secreted by the liver and macrophages with pleiotropic antiatherogenic functions and hence a candidate to treat hypercholestero- laemia and atherosclerosis. Here, we describe atheroprotective properties of ApoE, though also potential proatherogenic actions, and the prevalence of dysfunctional isoforms, outline conventional gene transfer strategies, and then focus on gene correction therapeutics that can repair defective APOE alleles. In particular, we discuss the possibility and potential benefit of applying in combination two technical advances to repair aberrant APOE genes: (i) an engineered endonuclease to introduce a double-strand break (DSB) in exon 4, which contains the common, but dysfunctional, ε2andε4 alleles; (ii) an efficient and selectable template for homologous recombination (HR) repair, namely, an adeno-associated viral (AAV) vector, which harbours wild-type APOE sequence. This technology is applicable ex vivo, for example to target haematopoietic or induced pluripotent stem cells, and also for in vivo hepatic gene targeting. It is to be hoped that such emerging technology will eventually translate to patient therapy to reduce CVD risk.

1. Introduction the inflammation eventually leads to smooth muscle cells infiltration, which proliferate within the intima to foster Deaths from cardiovascular disease (CVD), which encom- development of established lesions. passes ischaemic heart disease, stroke, and peripheral vas- A significant risk factor for CVD is increased plasma cular disease, total about 17 million per year worldwide, LDL. This is susceptible to oxidation and is the main source almost one-third of the total [1]. Notably, over 80% of of the cholesterol which deposits in arteries during endothe- CVD deaths are in low- and middle-income countries; it is lium injury. The statin class of drugs helps prevent early mor- not a disease confined to developed countries. Atheroscle- bidity or death by lowering plasma LDL. However, statins rosis is a progressive inflammatory response to complex do not rectify low levels of atheroprotective high-density interactions between cell types endogenous to the arterial lipoproteins (HDLs), which is an important independent wall, monocytes, lymphocytes and platelets from blood, and risk factor because of its role in removing excess cholesterol circulating lipoproteins [2]. Early atherosclerotic lesions are from arterial walls [3, 4]. Although aggressive high-dose lipid streaks, characterized by cholesterol-engorged foam statins are reported to successfully regress atherosclerotic cells within the vascular endothelium. Foam cells originate plaque with the goal of reducing morbidity and mortality, from blood monocyte macrophages, which are recruited the incidence of side effects (liver damage, rhabdomyolysis, into the arterial intima by upregulated adhesion molecules and cancer) increases and the therapeutic value remains on activated endothelium. There via unregulated scav- contentious. Similarly, the value of statin treatment in com- enger receptors they relentlessly ingest oxidized low-density bination with ezetimibe, a cholesterol-absorption inhibitor lipoprotein (LDL) or triglyceride-depleted (but cholesterol- which further decreases LDL cholesterol by 15–20%, has containing) remnant lipoprotein particles. Failure to resolve been questioned as no significant reductions in intima-media 2 Cardiology Research and Practice thickness were observed [5]. Alternative strategies to combat [19, 20], and the allele is also tentatively linked to intrac- occlusive CVD are still urgently needed for many patients erebral haemorrhage [21]. Nevertheless, ε2 carriers have [3, 4, 6]. reduced levels of total and LDL cholesterol (see Section 3), These include drugs to raise HDL levels, such as niacin, and their risk of coronary heart disease is 20% lower than inhibitors of cholesteryl ester transfer protein (CETP), and people with the common ε3/3 genotype [22], ApoE4 asso- RVX-208, a small quinazoline-family member which upreg- ciates with incidence of ischaemic stroke and subarachnoid ulates ApoAI, the main HDL protein constituent [4–9]. haemorrhage [21] and produces an adverse lipoprotein Gene-based therapies are also receiving considerable atten- profile, with increased LDL and a slight reduction in HDL tion, including antisense oligonucleotides (ASO) to target [22]. In 1996, a meta-analysis of 14 studies found that LDL’s structural protein ApoB100 [10]. Gene silencing of carriers of the ε4 allele had an increased risk of coronary proprotein convertase subtilisin/kexin type 9 (PCSK9) also heart disease [23], and this was confirmed in a subsequent lowers LDL indirectly. Secreted PCSK9 binds to the LDL meta-analysis of 48 studies involving 15,942 disease cases, receptor (LDLR) and accelerates its degradation, and hence which concluded the risk was 42% higher compared to prolonging LDLR activity by PCSK9 inhibition lowers ε3/3 carriers [24]. Moreover, the LDL cholesterol level is plasma LDL levels [11, 12]. HDL supplementation agents 30% lower in people with the ε2/2 genotype than with are also in clinical use, including infusion of synthetic or ε4/4, a reduction comparable to that achieved with statins. plasma-purified ApoAI and ApoAI mimetic peptides [13]. Nevertheless, this supposition of risk for ε4 carriers is Gene addition has also been used to boost plasma ApoAI contentious; an updated meta-analysis, which focused on and HDL in preclinical studies using adenovirus and adeno- studies recruiting large numbers of participants to reduce associated virus (AAV) vectors for delivery [14, 15]. publication bias, reported only a modest increase of risk [22]. On the other hand, APOE: environment interactions on 2. The ApoE Gene and Protein CVD risk are attracting increased attention [25]; the ε4allele potentiates the risk of CVD from physical inactivity and also Apolipoprotein E (ApoE) is a 34-kDa polymorphic glycopro- from smoking, though mainly in women [26]. In addition, tein largely secreted by liver (∼90%; 40–60 μg/mL plasma), ε4 carriers have a lower life expectancy, a finding reflecting although other tissues particularly macrophages contribute their predisposition to neurodegenerative disease as well as [16]. The human gene is located on chromosome 19, at CVD [27, 28]. the 5 end of a 50 kb gene cluster comprising ApoCI, an As detailed in Sections 3 and 4, and independently of ApoCI pseudogene, ApoCII, and ApoCIV [3]. Like many genotype, ApoE has a plethora of actions to inhibit athero- of the soluble apolipoproteins, the ApoE gene has four genesis. Most information has come from studies in vitro or exons separated by three introns and is synthesized and in mice, and, although many mechanisms are ill-understood, released via the canonical pathway (Figure 1). The primary the weight of evidence strongly suggests that ApoE is translation product is 317 residues containing an N-terminal atheroprotective. But there is a final twist to the tale: plasma 18 amino acid (a.a.) signal peptide to direct the growing ApoE levels positively correlate with CVD mortality [29], ApoE polypeptide to the endoplasmic reticulum. Prior to leaving several unanswered questions concerning human secretion, ApoE undergoes O-linked glycosylation in the ApoE biology in health and disease [30]. Golgi, principally at Thr194 although carbohydrate chains containing sialic acid are also present on Ser290. 3. ApoE and Lipid-Related Atheroprotection ApoE protein comprises an N-terminal domain (1–191 residues) of four amphipathic α-helices linked to the C- ApoE plays an essential role in the metabolism of dietary terminus by a protease-sensitive loop [17]. An arginine- and lipids, which enter the circulation as large triglyceride- lysine-rich segment in helix-4 (residues 134–150; Figure 1) rich chylomicron particles. Following lipolysis and uptake contains the recognition site for the LDLR and LDL receptor- of energy-rich monoglycerides and fatty acids, cholesterol- related protein (LRP). ApoE contains two heparin-binding containing and potentially atherogenic remnant particles are sites, one within the receptor-binding domain and the other left, which depend on ApoE for rapid hepatic clearance via within the C-terminus (residues 243–272). There are three pathways involving LRP, the LDLR and heparan sulphated common isoforms of ApoE, termed E2, E3, and E4, which proteoglycans (HSPG) [31]. ApoE4 has marginally greater are the products of three alleles (ε2, ε3, and ε4) at a single receptor-binding capability than ApoE3, but ApoE2 is defec- gene locus [16]. The ApoE3 (Cys112,Arg158) is considered tive with only 2% and 40% binding activity to the LDLR the parent form, while ApoE4 (Arg112,Arg158)andApoE2 and LRP, respectively. Although the ApoE2 isoform has an (Cys112,Cys158) arise from genetic point mutations and are amino acid substitution (Arg158Cys) outside of the 134–150 variants (Figure 1). There is considerable allelic variation in receptor-binding domain (Figure 1), this has a major disrup- different populations, but, in Europe, the relative frequencies tive influence. In ApoE3, a salt bridge is formed between are approximately 0.77, 0.15, and 0.08 for ε3, ε4, and ε2[18], Arg158 and Asp154. However, this interaction is lost in giving in order of occurrence E3/3, E4/3, E3/2, E4/4, E4/2, ApoE2, and so Asp154 forms a bridge with Arg150, reducing and E2/2 phenotypes. the positive potential of the binding site to markedly impair The rarest variant ApoE2 is the cause of Type III hyper- its binding affinity [32]. lipoproteinaemia in a small proportion (∼5%) of ε2/2 indi- When ordered E2/2, E2/3, E2/4, E3/3, E3/4, and E4/4, viduals, which confers a markedly increased risk of CVD there is a stepwise increase in plasma LDL cholesterol [22, Cardiology Research and Practice 3

ApoE4 ApoE2 GTC CGC GGC AAG TGC CTG

GTG TGC GGC AAG CGC CTG

C112R R158C

Exon1 Exon2 Exon3 Exon4

Receptor-binding domain of ApoE Arg134-Val-Arg-Leu-Ala-Ser-His140-Leu-Arg-Lys-Leu-Arg-Lys- Arg-Leu-Leu-Arg150

5 or 3 untranslated regions Translated regions

Figure 1: Structure of the human APOE gene. The APOE gene has 4 exons, comprising 44, 66, 193, and 860 nucleotides, with over 80% of the protein coded for by exon 4. Exon 1 contains 5 untranslated sequence, while the translated sequence in exon 2 codes for most of the 18 residue signal peptide. As indicated, exon 4 contains the two common disease-associated SNPs: a T → C point mutation produces ApoE4 (Cys112Arg), while C → T mutation gives ApoE2 (Arg158Cys). The wild-type sequence is ApoE3 (Cys112, Arg158). Also shown in the lower part of the figure is the positively charged arginine- and lysine-rich α-helical segment (residues 134–150) which recognizes the LDL receptor.

33]. One explanation is that this reflects up-or downregu- with ApoE2/2 human monocyte macrophages secreting lation of the hepatic LDLR due to changes in cholesterol substantially lower amounts of ApoE compared to E3/3 or delivery by ApoE2- or ApoE4-containing remnant particles, E4/4 cells [35], apparently because in macrophages, though respectively. Thus, a diminished cholesterol supply by ApoE2 not in hepatocytes, cysteine-rich ApoE2 forms dimers and would increase hepatic LDLR numbers to lower plasma multimers which are bound to LRP and retained in the LDL; additionally, reduced competition between ApoE2- secretory pathway [36]. remnants and LDL for binding by the LDLR would result in These atheroprotective biological functions of ApoE are accelerated LDL clearance [19]. The higher binding affinity corroborated by the hyperlipidaemia and atheroma seen of ApoE4-containing lipoproteins for the LDLR would have in ApoE-deficient (ApoE−/−) mice. Normal mice resist opposite effects and so raise LDL levels. Nevertheless, the atherosclerosis and even when fed a proatherogenic diet mechanism is more complex since conversion of VLDL to develop only immature fatty streak lesions [37]. However, LDL is impaired by ApoE2, which reduces the amount of ApoE−/− mice are grossly hypercholesterolaemic on normal LDL formed [19]. chow and spontaneously develop widespread fibroprolifer- ApoE also contributes to “reverse cholesterol transport,” ative lesions, which evolve into advanced complex plaques the antiatherogenic HDL-dependent pathway by which with smooth muscle cell caps and necrotic cores [38]. excess cholesterol in peripheral tissues, including arteries, On the other hand, transgenic animals expressing the is brought to the liver for biliary excretion. Such regu- common human ApoE isoforms have revealed more subtle lation of cellular cholesterol homeostasis, particularly in aspects of ApoE atheroprotection [37]. For example, early macrophages, is vital in preventing foam cell formation and studies in both mice and rabbits showed that overexpression atherogenesis. Efficient cellular cholesterol efflux depends of ApoE3 was detrimental causing hypertriglyceridaemia. on ATP-binding cassette transporters (ABCA1 and ABCG1), Excessive production of hepatic ApoE stimulates VLDL but lipid-poor ApoAI- (preβ-1 HDL) and ApoE-containing triglyceride synthesis and also inhibits lipoprotein-lipase- particles (γ-LpE) are avid initial cholesterol acceptors. Sig- (LPL-) mediated lipolysis, largely by displacing ApoCII an nificantly, the plasma γ-LpE fraction from E3/3 subjects essential cofactor for LPL from the VLDL surface [19]. sequestered substantially more cellular cholesterol than the However, phenotypic interpretations and comparison of data fractionsinE2/2andE4/4plasmas[34]. Additionally, from different groups were sometimes confounded because ApoE can activate plasma lecithin-cholesterol acyltransferase the human APOE transgene was inserted into different (LCAT), CETP, and hepatic lipase (HL), which are all genomic locations and at varying copy numbers and because involved in HDL maturation [16]. As macrophages secrete endogenous mouse ApoE protein was also present. ApoE at lesion sites, it is the dominant acceptor in clearing These difficulties were mitigated by introduction of gene excess arterial cholesterol. This effect is isoform dependent replacement strategies to generate mice in which the mouse 4 Cardiology Research and Practice

Apoe gene was replaced by a human allele. Nevertheless, such mouse peritoneal macrophages have also been shown to have mice also have limitations: unlike wild-type mice, ApoE3 anti-inflammatory properties with ApoE3-expressing cells knock-in animals were susceptible to dietary-induced hyper- secreting lower levels of the proinflammatory cytokines IL- cholesterolaemia and atherosclerosis, apparently because 6andTNF-α than ApoE2 and ApoE4 cells [55]. human ApoE3 has low receptor-binding affinity and is Studies in vivo using ApoE−/− mice have also given less efficient at clearing remnant particles [39]. Moreover, important clues to possible atheroprotective actions of ApoE. type III hyperlipoproteinaemia in humans homozygous for For example, lipoproteins from ApoE-deficient mice are the ε2 allele is a recessive condition, whereas a dominant more oxidized and prone to oxidation than those from con- inheritance is seen in ApoE2 knock-in mice [40]. While trol mice [56], while clearance of apoptotic cell remnants is this may also reflect binding differences between human reduced in the absence of ApoE [57]. An anti-inflammatory and mouse receptors, it should be noted that ∼80% of role for ApoE in dampening inflammation induced by liver-derived mouse VLDL is ApoB48-containing particles, lipopolysaccharide (LPS) or bacteria is postulated based and, hence, mice depend much more on ApoE for remnant on the increased susceptibility of ApoE−/− mice compared clearance than do humans. Interestingly, many of the rare to control animals when challenged with such pathogens ApoE variants with amino acid substitutions within the [58–60]. By contrast, a specific proinflammatory role for receptor-binding region do associate with dominant type ApoE was suggested by van den Elzen and colleagues who III hyperlipoproteinaemia even though most have adequate showed that ApoE delivers lipid antigens to CD1 endosomal receptor-binding activity compared to ApoE2 [19]. Stud- ies In vitro studies [41], and also in ApoE2 and ApoE compartments of antigen-presenting dendritic cells, most (Arg142Cys) transgenic mice [19] suggest that recessive likely via LDLR endocytosis, to stimulate natural killer T- ApoE2 retains HSPG binding, allowing remnant clearance cells [61]. This activation was lost from ApoE-depleted via this alternative pathway [19], whereas the dominant human serum and also drastically reduced in ApoE-deficient ApoE variants have negligible binding which translates to an mice challenged with an exogenous lipid antigen. The overall clearance rate lower than that of ApoE2 [31]. authors also speculated that ApoE might use this pathway to deliver self-lipid antigens and exacerbate atherosclerosis, 4. ApoE and although such removal of antigenic lipids might also be Lipid- Independent Atheroprotection atheroprotective. A full description of the cross-talk between ApoE and cytokines, including modulation of inflammatory It is now recognized that ApoE fulfils several biological and immune responses and isoform dependency, is beyond functions unrelated to lipid transport and that these make the scope of this paper but was recently reviewed by Zhang significant contributions to its antiatherogenic activity [16, and colleagues [44]. Moreover, the (patho)physiological 42–44](Figure 2). Most are anti-inflammatory in nature importance of such lipid-independent actions of ApoE is and include the early observation that ApoE restricts T- highlighted by studies in which low-level expression contin- /− cell activation and proliferation [45]. Later, our own studies ues to provide atheroprotection in ApoE− animals, despite showed that cell-derived ApoE inhibited platelet aggregation no change in their hyperlipidaemia [42, 62]. Significantly, and also the expression of vascular cell adhesion molecule Raffai et al. [63] provided compelling evidence that low 1 (VCAM-1) on endothelial cells, actions mediated by the levels of ApoE protein have the capacity to regress preexisting common mechanism of ApoE interaction with the cell- atherosclerotic lesions, independently of lowering plasma surface receptor, LRP8 (ApoER2) to activate nitric oxide cholesterol. synthase (NOSIII) and release NO [46, 47]. Additional data suggest that ApoE stimulates tyrosine phosphorylation of 5. ApoE Gene-Augmentation Therapeutics LRP8 to initiate PI3 kinase signalling and activation of NOSIII [48]. Induction of smooth muscle cell migration 5.1. Early Insights from Protein Therapy and Transgenic Mice. and proliferation by oxidized LDL or platelet-derived growth Twenty years ago plasma-purified or recombinant ApoE factor are also suppressed by ApoE [49], while subendothelial protein were infused into rabbits with genetic or diet- retention of LDL, an early proatherogenic event [50], is induced hypercholesterolaemia. Plasma cholesterol levels impeded by the presence of ApoE, which additionally were markedly reduced, while a long-term study halted regulates the availability of cytokines and growth factors atherosclerotic plaque development [64]. This preclinical within the pericellular proteoglycan matrix [51]. There is evidence endorsed the APOE gene as a candidate for ther- clear evidence that ApoE protects cells and lipoproteins apeutic manipulation. Similarly, synthetic peptide mimics against lipid oxidation and other oxidative stresses in an of the ApoE binding region (based on a dimeric repeat isoform-dependent manner, although findings can vary of a.a. 141–155) cleared cholesterol-rich lipoproteins in with the experimental system used. In cell-free systems, ApoE−/− mice [65], a strategy which evolved into covalently metal-induced oxidation is inhibited by ApoE in an allele- coupling ApoE 141–150 residues to an 18 a.a. amphipathic specific manner (E2>E3>E4), largely because it sequesters helical peptide capable of associating with atherogenic ApoB- metal ions [52–54]. However, when LDL was incubated containing lipoproteins [66]. Thrice weekly intravenous with transfected macrophages secreting equal amounts of injection of this dual domain peptide into ApoE−/− mice ApoE, the ApoE3 and ApoE4 isoforms provided greater for 6 weeks reduced plasma cholesterol and atherosclerotic protection against oxidation than ApoE2 [54]. Transfected lesions in the aortic sinus [67]. Cardiology Research and Practice 5

Hepatic clearance of Stimulation of hepatic atherogenic lipoproteins VLDL production

Cofactor for enzymes involved in Sequesters excess HDL maturation cholesterol at lesion sites (e.g. LCAT, HL, CETP)

ApoE and lipid-related atheroprotection

ApoE and lipid-independent atheroprotection

Antioxidant activity: Reduces smooth muscle inhibits LDL oxidation cell migration and proliferation

Suppresses apoptosis Inhibits platelet aggregation

Prevents LDL retention Delivers lipid antigens to in the artery wall Anti-inflammatory activity: CD1 in dendritic cells for Endothelial VCAM-1 expression presentation to T cells T-Lymphocyte activation and proliferation

Figure 2: Proposed atheroprotective functions of ApoE3. The antiatherogenic properties of ApoE3 are divided into those related to its role in lipid metabolism (shaded area) and those independent (unshaded area).

Use of in vivo ApoE-secreting miniorgans substanti- human ApoE3 induced regression of preexisting atheroscle- ates these antiatherogenic actions of ApoE peptides and rotic lesions without altering plasma lipoprotein levels purified protein. Recombinant ApoE-expressing endothelial [77], implying lipid-independent atheroprotective actions cells embedded in Matrigel were injected intradermally (Figure 2). The potential to regress advanced atheroma in into ApoE−/− mice which, 3 months later, had 50% less older animals was also noted [78], while the low toxicity plasma cholesterol and reduced atherosclerotic plaque [68]. and low immunogenicity of helper-dependent rAd vectors Similarly, implantation of alginate-encapsulated engineered allowed high stable expression of ApoE and lifelong athero- cells into the peritoneum of ApoE−/− mice secreted sufficient protection in ApoE−/− mice [79]. ApoE to lower plasma cholesterol and increase atheropro- An alternative viral vector to rAd is adeno-associated tective HDL [69]. Transgenic mice overexpressing ApoE virus (AAV), which contains a linear single-stranded DNA provide additional evidence for ApoE atheroprotection, as genome and is now at the forefront of clinical gene therapy these animals are protected from diet-induced or diabetic trials. Early studies used vectors derived from the common hyperlipidemia [70, 71]. Moreover, macrophage-restricted serotype 2, which transduced cells very efficiently in vitro, −/− expression of ApoE in transgenic mice [72], or in ApoE but unfortunately were largely ineffective in vivo. However, mice following transplantation of wild-type bone marrow the field was boosted by the isolation of over 10 new [73, 74], inhibits atherogenesis. serotypes, which led to development of new pseudotyped rAAV vectors that is, capsids of alternative AAV serotypes 5.2. ApoE Gene Transfer Studies. Gene transfer of ApoE was harbouring the recombinant AAV serotype 2 genome. These first reported 16 years ago using recombinant adenovirus had strikingly improved performances in vivo and different (rAd) vectors, which deliver foreign DNA into mammalian tissue tropisms to AAV2/2; for example, rAAV2/8 transduced liver with near-100% efficiency. High levels of plasma ApoE liver with near-100% efficiency [80–82], while rAAV2/1 [83], were obtained following intravenous injection of rAd.ApoE3 rAAV2/6 [84] or rAAV2/7 [85]wereeffective for skeletal into ApoE−/− mice, which reduced plasma cholesterol muscle transduction. and slowed aortic atherogenesis [75]. Unfortunately, the This increased serotype efficiency was shown by intra- therapeutic effect was transient, as these 1st generation venous (liver-directed) injection of a human ApoE3 rAAV2/8 vectors triggered a strong T-cell immunological response. By vector into ApoE−/− mice; normal human levels (50- contrast, 2nd generation vectors gave sustained human ApoE 80 μg/mL) of plasma ApoE were produced [58]. It is pro- expression and largely normalized the lipoprotein profile in posed that, unlike serotype 2, these different AAV serotypes ApoE−/− mice throughout the 6 week study [76]. Moreover, uncoat rapidly to facilitate annealing of the single-stranded in hyperlipidaemic LDLR null mice, hepatic expression of plus and minus rAAV genomes into stable, transcriptionally 6 Cardiology Research and Practice active double-stranded DNA molecules [86]. Similarly, the DNA segment and its target sequence in a mammalian recent introduction of self-complementary rAAV (scAAV) genome can mediate genomic modification or repair [95, vectors, which circumvent the need for annealing rAAV 96]; it also draws on the application of short synthetic vector genomes, allows a rapid and higher level of transgene oligonucleotides to repair defective alleles in yeast [97]. expression [87, 88]. For example, expression of ApoE3 from The use of hybrid RNA-DNA oligonucleotides (RDOs or scAAV2/8 using the hepatocyte-specific promoter (LP1) chimeraplasts) to introduce site-specific changes of 1–3 −/− normalized cholesterol levels in male ApoE mice and nucleotides into mammalian genomes was reported 15 years retarded development of aortic atherosclerosis by 58% [89]. ago by the Kmiec group [98]. They constructed RDOs Skeletal muscle, an accessible, stable, and well-vascu- to form a double-stranded, hairpin-capped oligonucleotide larized tissue, has also been used for gene transfer of incorporating a short region of correcting DNA bounded ApoE. Although muscle does not normally secrete ApoE,  by long stretches of protected 2 -O-methyl RNA. The strong nonhepatic, nonmacrophage-derived ApoE is known to be RNA-DNA base pairing was considered to promote strand atheroprotective [62, 90]. Indeed, intramuscular injection invasion and annealing to the target (e.g., a genetic point of a plasmid DNA vector expressing human ApoE3 into ffi ApoE−/− mice gave modest but sustained lowering of plasma mutation) genomic locus. This high-a nity hybridisation cholesterol [91], while another study reported reduced involves both strands of the gene and creates the mismatch xanthoma and atherosclerotic plaque formation [92]. The at the point mutation, leading to recognition and correction high purity and low immunogenicity of plasmids, which by one or more of the mammalian gene repair pathways. The are expressed episomally making insertional mutagenesis early studies of Kmiec and colleagues in repairing the sickle improbable, make them attractive delivery vehicles. Expres- cell mutation in a lymphoblastoid cell line [99] and in target- sion can be further increased by electropulsing the injection ing primary CD34+ cells [100] generated much excitement, sites(s) [93]. Additionally, we have injected AAV2/7, AAV2/8, but also notable controversy as the reported conversion and AAV2/9 vectors expressing human ApoE3 into the efficiencies of up to 50% were considered implausible [101]. tibialis anterior muscles of ApoE−/− mice. The first two However, these criticisms were deflected by corroboration vectors were the most effective, producing up to 2 μg from others, including targeting human hepatocyte cells ApoE3/mL plasma, and, at 13 weeks, the mice had 50% less [102, 103] and, in landmark studies, in vivo targeting of the plaque lipid in brachiocephalic arteries than AAV2/9-treated rat factor IX gene [104] and correction of the UGT1A1 gene animals [94]. mutation in the hyperbilirubinaemic Gunn rat [105]. Based on these successes, our laboratory pioneered the technique in cardiovascular disease, including conversion 6. Oligonucleotide-Mediated Gene Editing of dysfunctional ApoE2 [106]andApoE4[107]toApoE3 of the Human APOE Gene in Cultured Cells using a standard 68-mer RDO for targeting. However, the practicality of this emerging methodology was questioned 6.1. Background. Gene editing, as described herein, uses when several groups, our own included, began to report poor short synthetic oligonucleotides to manipulate genomic reproducibility and unstable conversions [108, 109]. One DNA and introduce small, site-specific changes, typically adverse factor was variable quality of the long and complex 1–3 nucleotides, into a selected gene of a living cell. It RDO molecules, which meant higher doses of reagents and has potential, therefore, for introducing specific mutations, delivery vehicles to effect repair; in turn, these would amplify gain or loss of function, into cell lines or mouse strains, cytotoxic and proapoptotic actions or induce cell cycle arrest and for studying single-nucleotide polymorphisms (SNPs). [107, 109]. The problematic nature of these 1st generation Below and in Section 7, we describe some of the strategies reagents was mitigated, however, by findings from the Kmiec used to achieve genotypic modifications. However, it is group that the all-DNA strand of the RDO initializes the important to note that, unlike gene transfer technologies, genomic repair [110] and that a single-stranded all-DNA the change is permanent; moreover, existing gene enhancers oligonucleotide (ssODN) could outperform the RDO if and promoters, and cell-specific control and context, are protected from nuclease degradation by chemical modifica- retained. These advantages underpin its ultimate goal: the tion of bases [111]. As described in Section 6.3, these data cure of hereditary diseases caused by genetic point mutations prompted us to switch to short (27-mer) ssODNs, which are or small deletions/insertions. First, we explain why synthetic purer and give increased reproducibility. oligonucleotides, which are relatively inexpensive and offer flexible design and chemistry, remain of value for generating new isogenic cell lines, for example, to investigate differential 6.3. Gene Editing: Use of Single-Stranded All-DNA Oligonu- effects of ApoE isoforms on cellular metabolism. Second in cleotides (ssODNs). The targeting ssODNs are homologous Section 7. we describe the rapid progress being made towards to the genomic sequence except for the desired change— therapeutic applications, for example, the ex vivo or in situ a mismatch to introduce or correct a point mutation, or repair of genetic point mutations. short insertion/deletion. As single-stranded DNA is rapidly degraded within cells, ssODNs are generally protected with 6.2. Gene Editing: Early Use of RNA-DNA Oligonucleotides. modified nuclease-resistant bases, most commonly three Conceptually, this technology is founded on the independent phosphorothioate (PTO) bonds at their 5 and 3 ends. observations of Smithies and Capecchi in the 1980s that Molecular details of targeted gene alteration remain poorly homologous recombination (HR) between a large exogenous delineated but include elements of nucleotide and base exci- Cardiology Research and Practice 7 sion repair systems and also some degree of HR involvement 7. Therapeutic In Situ Correction of [112, 113], while at least part of the correcting DNA oligonu- Defective APOE Genes cleotide is physically incorporated into the genomic target site [114]. Given the difficulties and controversies associated 7.1. Rationale for ApoE Gene Repair. Most of the in vivo with RDO-directed repairs, we decided to optimize our studies described in Sections 2–4 have used ApoE null mice. methodology in recombinant mammalian cells expressing These gave important insights into ApoE functions and a sensitive reporter gene, green fluorescent protein (GFP) allowed ready evaluation of protein and gene therapeutics. which was rendered nonfluorescent by introduction of a However, as genetic deficiency of ApoE and absent plasma point mutation. This allowed analysis at the single cell level ApoE is extremely rare in humans, the value of such a preclinical model is questionable. It is noteworthy, therefore, and the harvesting of corrected (green) cells for further study. that augmentation of plasma ApoE via transgenic expression We provided stringent evidence to validate the technology or gene transfer has confirmed its atheroprotective function and to establish unequivocally that ssODN-mediated gene in LDLR deficient mice and in diabetic or fat-fed mice, all alteration is a real and reproducible phenomenon [115]. of which have normal levels of plasma ApoE. Nevertheless, Although 5-3-PTO protection gave 10-fold greater cor- ffi animal studies also indicate caution in augmenting ApoE3 rection e ciencies than unmodified ssODN (about 2% expression using gene transfer technologies; overexpression versus 0.2% of targeted cells were green), we confirmed in mice and rabbits causes hypertriglyceridaemia [19]. the observation of others [111] that only a low percentage An alternative and attractive approach, albeit a distant of these green cells was actively replicating [116]. Cell therapeutic goal, is to alleviate dyslipoproteinaemia and cycle analysis 16 h posttransfection revealed that the ssODN reduce CVD risk by in situ correction of a deleterious with PTO end protection resulted in DNA damage with APOE gene using targeted gene editing. Is this feasible? accumulation of cells in the G2 phase, whereas treatment One possibility, founded on the multiple atheroprotective with unmodified ssODN was markedly less toxic. However, actions of macrophage-secreted ApoE3, is to cure type III by varying the type or position of protecting groups, we hyperlipoproteinaemia by transplantation of haematopoietic discovered that internal protection of the ssODN with stem cells (HSCs), following ex vivo repair of the aberrant four PTO residues at the targeting site maintained efficient APOE gene. As the condition is recessive, conversion of gene correction and, compared to 5-3-PTO protection, one ε2alleletoε3 should provide effective treatment. substantially reduced cell cycle arrest to enable cell growth Moreover, the current impetus in bringing stem cell tech- and proliferation (Figure 3)[116]. nologies to the clinic, particularly induced pluripotent stem We are currently evaluating this novel ssODN design cells (iPSCs), suggests additional possibilities for ex vivo in targeting the endogenous APOE3 gene in human THP1 manipulations and cell-based therapies [118]. These include monocyte macrophages and human HepG2 hepatoblastoma efficient differentiation to functional hepatocyte-like cells cells (both ε3/ε3) to generate cells with the ApoE4 phe- [119–121], which would allow a repaired APOE gene to notype. We have devised an affinity-capture matrix to dis- be expressed from cells engrafted into liver. Finally, there tinguish successfully targeted ApoE4-secreting cells (<1%) is also the exciting prospect of direct in vivo hepatic gene from noncorrected ApoE3 cells. This uses low permeability targeting for in situ correction of genetic disease [122]. These carboxymethyl-cellulose media preloaded with a commercial sophisticated therapies to repair defective genes in the future ApoE4-specific monoclonal antibody for detection and gives will rely on more advanced technologies than simple nuclear sensitive and reproducible findings. Secreted ApoE was delivery of ssODNs and these are discussed in the remaining sections of this paper. readily trapped on the surface of individual cells, enhancing staining efficiency and decreasing assay time. Enriched 7.2. DNA Repair Templates. The ssODNs we have used for corrected populations of ApoE4-secreting cells have been targeting of the APOE gene have several advantages: high detected, although isolation and expansion of corrected purity and simple production, including variable chemistry, clones is proving difficult. Should a successful clone be easy delivery, and high fidelity with no off-target effects. identified and characterized, then it will undergo a 2nd Although ssODNs and other constructs such as triple targeting to generate the ApoE4/4 phenotype. helix forming oligonucleotides and small homologous DNA These experiments are notable for two reasons. First, fragments [113] can act as DNA repair templates, the most they will provide conclusive proof that ssODN-directed gene effective is adeno-associated virus (AAV), which carries a editing can generate new cell lines, here, with E3/E4 and linear single-stranded DNA genome. Though technically E4/4 phenotypes, although a similar strategy can generate challenging to produce, these viral vectors can genetically the other common phenotypes (E2/E3, E2/2, and E2/4) if an target embryonic stem cells (ESCs), HSCs, or iPSCs achiev- ApoE2-specific antibody is used [117]. Second, because these ing correction efficiencies of 0.07–1% [113, 123–125]. The isogenic cells can be used to investigate isoform differences in mechanism, as illustrated in Figure 5, for targeting and gene macrophage (and hepatocyte) ApoE trafficking and secretion repair of defective APOE2 is via homologous recombination [35, 36], to monitor differential cellular responses to oxidant (HR), as silencing essential components of the potentially stress or to cholesterol loading and to investigate variation competitive pathway, nonhomologous end joining (NHEJ) in structure and composition of secreted ApoE lipoprotein has no effect on conversion frequencies [126]. Note too that particles [48]. the targeting vector can harbour a selectable marker, such 8 Cardiology Research and Practice

1000

800 20 G2 G1 GFP + ve cells 600 15

400 10 DNA content DNA

200 Cell Number 5

0 0 100 101 102 103 104 0 200 400 600 800 1000 Single division only GFP DNA content (a) End-protected 53 PTO ssODN: good efficiency and minimal proliferation 1000

800 GFP + ve cells 10 G2 600 G1 8 400 6 DNA content DNA 4

200 Cell Number 2 GFP+ve cells 0 0 100 101 102 103 104 0 200 400 600 800 1000 Dividing colonies GFP PI (2) DNA content (b) Internal-protected PTO ssODN: medium efficiency and good proliferation

Figure 3: Effects of nuclease-protected ssODNs on targeted gene repair. Recombinant cells expressing mutated (nonfluorescent) green fluorescent protein (GFP) were targeted with two ssODNs, one with three phosphorothioate (PTO) residues at each end (a) and the other with four internal PTOs at the central correcting segment (b). Posttargeting, the green cells were counted by flow cytometry (left panels), stained for cell cycle analysis (central panels) and cultured to observe cell growth (right panels). The PTO end-protected ssODN gave a much higher rate of gene correction, as measured by the number of green cells (boxes in left panels). However, this resulted in DNA damage and accrual of cells in the G2 phase, whereas the GFP+ve cells obtained after internally protected ssODN targeting had a markedly greater proportion in G1 (central panel). Significantly, only the occasional divided pair of green cells was noted in cultures treated with end-protected ssODN, whereas actively replicating green cells were evident when the internally protected ssODN was used (right panels).

APOE C158 Chromosomal DNA 2allele

Homologous R158 recombination AAV Repair template APOE3allele R158

Figure 4: Conversion of the mutant APOE2 allele to wild-type APOE3 using an AAV-based DNA repair sequence. The single-stranded adeno-associated virus genome, which contains Exon 4 of APOE3 and its flanking regions within its ITRs (inverted terminal repeats), is an efficient recombination template. As indicated, the mechanism is via a homologous recombination (HR) pathway. Cardiology Research and Practice 9

Chromosomal target DNA

DSB formation Donor DNA

Homology arm Repair Homology arm (left) sequence (right)

DSB stimulates HR HR

Mutation at target locus is Marked increase in gene repair replaced with central repair efficiency sequence from donor DNA

Figure 5: Double-strand breaks (DSBs) stimulate homologous recombination (HR) repair. HR facilitates exchange of DNA sequence between donor and acceptor molecules, provided they share a certain amount of sequence similarity. Donor DNA molecules are designed to contain a central “repair” sequence, flanked by “homology” arms, whose sequence is identical to that of the acceptor (e.g., genomic) DNA. Through HR, the repair sequence replaces the entire sequence lying between the crossover points (marked with an X) on the acceptor. Gene targeting exploits HR in this way to make genetic changes to a cell’s DNA. Importantly, introduction of a DSB into the acceptor sequence, which can be achieved by adding an engineered nuclease with locus-specific cleavage (see Figure 6), markedly increases gene repair efficiency. as puromycin N-acetyl-transferase, between the homology principle, customized ZFNs which function in pairs as arms, which if flanked by two LoxP sites can be excised FokI cleavage requires dimerization can be created with from the genome of repaired cells by Cre-recombinase. More predetermined sequence specificity to introduce a DSB at a recently, helper-dependent adenoviral vectors (HDAdVs) precise genomic locus (Figure 6). Within the last few years, have also been used for efficient and accurate gene targeting an impressive number of ZFN gene targeting successes has of human ESCs and iPSCs [127, 128]. These vectors are been reported, including in vitro corrections of 18–30%, capable of correcting large genomic regions (∼35 kb) and the generation of gene-knockout rats, and manipulation are unaffected if the target locus is transcriptionally inactive of human embryonic or induced pluripotent stem cells [129]. [113, 133–136]. One concern of ZFNs is genotoxicity due to off-target cleavages that may disrupt normal genes [137], 7.3. Double-Strand Breaks (DSBs) Markedly Stimulate HR although improved designs and screening procedures are Gene Repair. Although HR-directed gene repair is accurate helping to reduce this possibility [138, 139]. As second and versatile, as illustrated in Figures 4 and 5,itoccurs problem is that ZFNs can recleave a repaired site, although with a very low frequency in mammalian cells (∼10−6) using this can be minimized by introducing a few silent mutations conventional linear targeting plasmids as templates [130]. into the donor DNA template as this will impair subsequent However, HR is one way in which cells repair double-strand ZFN binding [140]. ff DNA breaks, for example, by using the sister chromatid Transcription activator-like e ector nucleases (TALENs) as a template, and this phenomenon has been exploited are another group of chimeric nucleases, which have a to increase HR gene-targeting events over 1000-fold. Initial different class of DNA-binding domains coupled to FokI. studies inserted the rare cutting site for the meganuclease, TAL effector proteins contain highly modular DNA-binding I-SceI, within a reporter target gene and then cotransfected domains, with a very simple code between their amino acids cells with the endonuclease and a repair DNA template and target DNA bases [141, 142]. This allows the ready [131, 132]. Nevertheless, such a system has little practical assembly of a series of individual modules to bind a unique value; it requires prior engineering of the target gene and DNA sequence and to activate FokI dimerization and cleav- cannot be used for endogenous genes. age. Although reports of TALEN-mediated gene targeting of Recent work in engineering artificial endonucleases now mammalian cells are still limited, this simple and effective circumvents this bottleneck by creating chimeric nucleases technology may soon supersede ZFNs; indeed, in the first with the potential to cleave DNA at virtually any desired direct comparison, TALENs had greater specificity and less sequence. The most studied are the zinc finger nucleases cytotoxicity than ZFNs [143]. (ZFNs), which comprise a DNA-binding domain (assembled Finally, an alternative to engineered chimeric nucleases as 3–5 finger modules, each recognizing 3 consecutive bases) are the homing endonucleases, which induce DSBs with joined to the nonspecific DNA cleavage enzyme, FokI.In exceptional specificity. These are natural enzymes with large 10 Cardiology Research and Practice

Zinc finger domains Fok1 Fok1

Zinc finger domains Flexible linker

Specific DNA binding by zinc finger domains

Dimerization of Fok1 domains to cleave DNA and create DSB

Figure 6: Basic structure and design of a zinc finger nuclease (ZFN). ZFNs are created by joining a DNA-binding region to the catalytic domain of the nonspecific Fok1 endonuclease. Zinc fingers are a protein motif capable of DNA binding, whose sequence specificity can be predetermined. Each zinc finger, illustrated by an individual circle, recognizes 3-4 nucleotides, and, by assembling three or four suitable zinc finger motifs, a sequence-specific DNA-binding domain can be created. Fok1 nuclease activity requires dimerization, and so the customized ZFNs function in pairs. As shown, the zinc finger-binding domain brings two Fok1 units together in the right orientation over the target sequence; this induces Fok1 dimerization and target sequence cleavage. asymmetric recognition sequences (12–40 bp) that form edit human iPS cells [150]. Given that ssODNs are easier five families based on sequence and structure homology and cheaper to produce in bulk, have high purity and safety [144, 145]. “Mix and match” engineering, aided by com- profiles, and relatively easy to deliver, it seems likely that their putation, allows production and validation of dozens of use will continue to expand. customized meganucleases, most based on the laglidadg family sequence motif. In turn, prospective DSB sites can be predicted across the entire human genome. This suggests 8. Concluding Remarks > that it will be possible to identify a unique 14-bp sequence The technology to repair inborn genetic mutations in cells, within exon 4 of the APOE gene that can be cleaved by a including haematopoietic stem cells and induced pluripotent homing meganuclease. stem cells, has developed rapidly during the last few years. In situ correction of the dysfunctional ε2andε4alleles 7.4. ZFN-ssODN Combinations. To date, the most common to alleviate hyperlipidaemia and counteract the progression template for DSB repair-dependent gene targeting has been a of atherosclerosis is now a preclinical reality. It is also circular or linear plasmid, although advantages of using AAV translatable to patients. The combination of a selectable as the exogenous DNA donor substrate are now emerging AAV-based DNA template for efficient HR repair, which is [146–148]. However, exciting new work establishes that further stimulated by the safe and precise introduction of the ZFN-ssODN combination offers flexible and efficient a DSB at the target genomic locus, provides the necessary genome editing. Chen and colleagues [149] used ZFNs that technological tools [146, 147, 151]. Indeed, genome editing cut within the ssODN homology area and reported editing using engineered endonucleases was recently announced as frequencies between 1 and 30% depending on the cell type Nature’s Method of the Year for 2011 [152]. and whether the targets were single nucleotide substitutions Repairing the defective APOE2 gene in bone mar- or small to large deletions. Importantly, the ssODN appeared row (lineage-negative) stem cells from patients with amoreefficient and faithful recombination partner than type III hyperlipoproteinaemia, followed by nonmyeloab- traditional double-stranded DNA constructs, most likely lative (reduced-intensity) transplantation, will provide because ssODN does not participate in NHEJ and hence macrophage-secreted antiatherogenic ApoE3 at lesions sites will not insert into nonspecific cleavage sites. Indeed, the and halt atherosclerotic plaque progression. This therapeutic duo of ZFN-ssODN has already been used to successfully approach can be critically evaluated in a preclinical model, Cardiology Research and Practice 11 the human ApoE2 knock-in mouse. Moreover, the prospect scAAV: Self-complementary adeno-associated viral of direct in vivo hepatic gene targeting for in situ correction vector of dysfunctional ε2 alleles is no longer a distant dream. ssODN: Single-stranded oligodeoxyribonucleotide Codelivery via hepatotropic AAV8 vectors of a chimeric TALEN: Transcription activator-like nuclease nuclease and donor DNA template to the livers of mice with VLDL: Very-low-density lipoprotein blood factor IX deficiency restored haemostasis [122]. ZFN: Zinc finger nuclease. Is there a case for also converting the ε4alleletoε3 to reduce CVD risk when, as outlined in Section 2, the Acknowledgments increase may only be marginal even for the ε4/ε4 carriers? Undoubtedly, the case can be argued: within this group, there The authors’ work was supported by project grants from will certainly be individuals who would benefit, perhaps the British Heart Foundation (PG/06/015/20305 and PG/ because their LDL is in the top decile of ε4/ε4 carriers or 09/070/27912) and by an award from the Peter Samuel Fund, because they have other risk factors. The future prospect of Royal Free Hospital to I. Papaioannou (no 933). editing the gene in macrophages and/or liver will also allow a more flexible therapeutic approach. Given that the ε2allele References is associated with low LDL and reduced CVD, we can also [1] P. Poole-Wilson, “The prevention of cardiovascular disease speculate that a better therapeutic option is to convert cells ε ε ε ε ε ε worldwide: whose task and WHO’s task?” Clinical Medicine, from 4/ 4 to the 4/ 2 genotype, rather than 4/ 3. What is vol. 5, no. 4, pp. 379–384, 2005. clear, however, is that ApoE, which in 1973 was identified in [2] G. K. Hansson, “Mechanisms of disease: inflammation, ath- human VLDL [153], has multiple roles in human cell biology erosclerosis, and coronary artery disease,” The New England and disease and that, despite impressive understanding of Journal of Medicine, vol. 352, no. 16, pp. 1685–1626, 2005. these functions over the last two decades, we still have much [3] D. J. Hausenloy and D. M. Yellon, “Enhancing cardiovascular more to discover. In time, it is to be hoped that such insights, disease risk reduction: raising high-density lipoprotein lev- coupled with the emerging technology of ex vivo or in situ els,” Current Opinion in Cardiology, vol. 24, no. 5, pp. 473– editing of the human APOE gene, will eventually translate to 482, 2009. patient therapy to reduce CVD risk. [4] P. Linsel-Nitschke and A. R. Tall, “HDL as a target in the treatment of atherosclerotic cardiovascular disease,” Nature Reviews Drug Discovery, vol. 4, no. 3, pp. 193–205, 2005. [5] M. Rizzo and G. B. Rini, “Ezetimibe, cardiovascular risk and atherogenic dyslipidaemia,” Archives of Medical Science, vol. 7, no. 1, pp. 5–7, 2011. Abbreviations [6] V. Charlton-Menys and P. N. Durrington, “Human choles- AAV: Adenoassociated virus terol metabolism and therapeutic molecules,” Experimental ABCA1/G1: ATP-binding cassette transporter A1/G1 Physiology, vol. 93, no. 1, pp. 27–42, 2008. ff Apo: Apolipoprotein [7] D. Du y and D. J. Rader, “Update on strategies to increase HDL quantity and function,” Nature Reviews, vol. 6, no. 7, CVD: Cardiovascular disease pp. 455–463, 2009. CETP: Cholesteryl ester transfer protein [8] M. H. Davidson, “Apolipoprotein A-I therapy: promise, DSB: Double-strand break challenges, and disappointment,” Journal of the American ESC: Embryonic stem cell College of Cardiology, vol. 57, no. 9, pp. 1120–1121, 2011. GFP: Green fluorescent protein [9]E.M.deGomaandD.J.Rader,“NovelHDL-directedphar- HDAdV: Helper-dependent adenovirus macotherapeutic strategies,” Nature Reviews Cardiology, vol. HSPG: Heparan sulphate proteoglycan 8, no. 5, pp. 266–277, 2011. HL: Hepatic lipase [10] D. A. Bell, A. J. Hooper, and J. R. Burnett, “Mipomersen, an HDL: High-density lipoprotein antisense apolipoprotein B synthesis inhibitor,” Expert Opin- HR: Homologous recombination iononInvestigationalDrugs, vol. 20, no. 2, pp. 265–272, 2011. HSC: Haematopoietic stem cell [11] C. J. Duff andN.M.Hooper,“PCSK9:anemergingtarget iPSC: Induced pluripotent stem cell for treatment of hypercholesterolemia,” Expert Opinion on LCAT: Lecithin-cholesterol acyltransferase Therapeutic Targets, vol. 15, no. 2, pp. 157–168, 2011. LDL: Low-density lipoprotein [12] A. K. Soutar, “Unexpected roles for PCSK9 in lipid LDLR: LDL receptor metabolism,” Current Opinion in Lipidology, vol. 22, no. 3, pp. 192–196, 2011. LRP: LDL receptor-related protein [13] J. D. Smith, “Apolipoprotein A-I and its mimetics for the LPS: Lipopolysaccharide treatment of atherosclerosis,” Current Opinion in Investiga- NHEJ: Nonhomologous end joining tional Drugs, vol. 11, no. 9, pp. 989–996, 2010. PCSK9: Proprotein convertase subtilisin/kexin type 9 [14] K. Oka, L. M. Belalcazar, C. Dieker et al., “Sustained phe- PTO: Phosphorothioate notypic correction in a mouse model of hypoalphalipopro- rAdV: Recombinant adenovirus teinemia with a helper-dependent adenovirus vector,” Gene RDO: RNA-DNA oligonucleotide Therapy, vol. 14, no. 3, pp. 191–202, 2007. 12 Cardiology Research and Practice

[15] S. F. C. Vaessen, R. J. Veldman, E. M. Cornijn et al., “AAV [30] M. Reilly and D. J. Rader, “Apolipoprotein E and coronary gene therapy as a means to increase apolipoprotein (Apo) A- disease: a puzzling paradox,” PLoS Medicine,vol.3,no.6,pp. I and high-density lipoprotein-cholesterol levels: correction 0736–0738, 2006. of murine ApoA-I deficiency,” Journal of Gene Medicine, vol. [31] R. W. Mahley and Y. Huang, “Atherogenic remnant lipopro- 11, no. 8, pp. 697–707, 2009. teins: role for proteoglycans in trapping, transferring, and [16] K. Greenow, N. J. Pearce, and D. P. Ramji, “The key role internalizing,” Journal of Clinical Investigation, vol. 117, no. of apolipoprotein E in atherosclerosis,” Journal of Molecular 1, pp. 94–98, 2007. Medicine, vol. 83, no. 5, pp. 329–342, 2005. [32] R. W. Mahley, K. H. Weisgraber, and Y. Huang, “Apolipopro- [17] D. M. Hatters, C. A. Peters-Libeu, and K. H. Weisgraber, tein E: structure determines function, from atherosclerosis to “Apolipoprotein E structure: insights into function,” Trends Alzheimer’s disease to AIDS,” Journal of lipid research, vol. 50, in Biochemical Sciences, vol. 31, no. 8, pp. 445–454, 2006. pp. S183–188, 2009. [18] J. E. Eichner, S. T. Dunn, G. Perveen, D. M. Thompson, K. [33] S. J. Murdoch, A. P. Boright, A. D. Paterson et al., “LDL E.Stewart,andB.C.Stroehla,“ApolipoproteinEpolymor- composition in E2/2 subjects and LDL distribution by Apo phism and cardiovascular disease: a HuGE review,” American E genotype in type 1 diabetes,” Atherosclerosis, vol. 192, no. 1, Journal of Epidemiology, vol. 155, no. 6, pp. 487–495, 2002. pp. 138–147, 2007. [34] Y. Huang, A. von Eckardstein, S. Wu, and G. Assmann, [19] R. W. Mahley, Y. Huang, and S. C. Rall, “Pathogenesis “Effects of the apolipoprotein E polymorphism on uptake of type III hyperlipoproteinemia (dysbetalipoproteinemia): and transfer of cell-derived cholesterol in plasma,” Journal of questions, quandaries, and paradoxes,” Journal of Lipid Clinical Investigation, vol. 96, no. 6, pp. 2693–2701, 1995. Research, vol. 40, no. 11, pp. 1933–1949, 1999. [35] P. Cullen, A. Cignarella, B. Brennhausen, S. Mohr, G. Ass- [20] P. N. Hopkins, L. L. Wu, S. C. Hunt, and E. A. Brinton, mann, and A. von Eckardstein, “Phenotype-dependent dif- “Plasma triglycerides and type III hyperlipidemia are inde- ferences in apolipoprotein E metabolism and in cholesterol pendently associated with premature familial coronary artery homeostasis in human monocyte-derived macrophages,” disease,” Journal of the American College of Cardiology, vol. 45, Journal of Clinical Investigation, vol. 101, no. 8, pp. 1670– no. 7, pp. 1003–1012, 2005. 1677, 1998. [21]C.Sudlow,N.A.Mart´ınez Gonzalez,´ J. Kim, and C. Clark, [36] D. Fan, S. Qiu, C. D. Overton et al., “Impaired secretion “Does Apolipoprotein E genotype influence the risk of of apolipoprotein E2 from macrophages,” The Journal of ischemic stroke, intracerebral hemorrhage, or subarachnoid Biological Chemistry, vol. 282, no. 18, pp. 13746–13753, 2007. hemorrhage? Systematic review and meta-analyses of 31 [37] K. Wouters, R. Shiri-Sverdlov, P. J. van Gorp, M. van Bilsen, studies among 5961 cases and 17 965 controls,” Stroke, vol. and M. H. Hofker, “Understanding hyperlipidemia and 37, no. 2, pp. 364–370, 2006. atherosclerosis: lessons from genetically modified apoe and [22] A. M. Bennet, E. Di Angelantonio, Z. Ye et al., “Association ldlr mice,” Clinical Chemistry and Laboratory Medicine, vol. of apolipoprotein E genotypes with lipid levels and coronary 43, no. 5, pp. 470–479, 2005. risk,” Journal of the American Medical Association, vol. 298, [38] A. S. Plump, J. D. Smith, T. Hayek et al., “Severe hypercholes- no. 11, pp. 1300–1311, 2007. terolemia and atherosclerosis in apolipoprotein E-deficient [23]Y.Song,M.J.Stampfer,andS.Liu,“Meta-analysis: mice created by homologous recombination in ES cells,” Cell, apolipoprotein E genotypes and risk for coronary heart vol. 71, no. 2, pp. 343–353, 1992. disease,” Annals of Internal Medicine, vol. 141, no. 2, pp. 137– [39] P. M. Sullivan, H. Mezdour, Y. Aratani et al., “Targeted 147, 2004. replacement of the mouse apolipoprotein E gene with [24]P.W.F.Wilson,E.J.Schaefer,M.G.Larson,andJ.M. the common human APOE3 allele enhances diet-induced Ordovas, “Apolipoprotein E alleles and risk of coronary hypercholesterolemia and atherosclerosis,” The Journal of disease: a meta-analysis,” Arteriosclerosis, Thrombosis, and Biological Chemistry, vol. 272, no. 29, pp. 17972–17980, 1997. Vascular Biology, vol. 16, no. 10, pp. 1250–1255, 1996. [40]P.M.Sullivan,H.Mezdour,S.H.Quarfordt,andN. [25] P. J. Talmud, “Commentary on the paper by Gustavsson et Maeda, “Type III hyperlipoproteinemia and spontaneous atherosclerosis in mice resulting from gene replacement al. entitled “Interaction of apolipoprotein E genotype with ∗ smoking and physical inactivity on coronary heart disease of mouse Apoe with human APOE 2,” Journal of Clinical risk in men and women”,” Atherosclerosis, vol. 220, no. 2, pp. Investigation, vol. 102, no. 1, pp. 130–135, 1998. 323–324, 2012. [41] Z. S. Ji, S. Fazio, and R. W. Mahley, “Variable heparan sulfate proteoglycan binding of apolipoprotein E variants may [26] J. Gustavsson, K. Mehlig, K. Leander et al., “Interaction modulate the expression of type III hyperlipoproteinemia,” of apolipoprotein E genotype with smoking and physical The Journal of Biological Chemistry, vol. 269, no. 18, pp. inactivity on coronary heart disease risk in men and women,” 13421–13428, 1994. Atherosclerosis, vol. 220, no. 2, pp. 486–492, 2012. [42] L. K. Curtiss, “ApoE in atherosclerosis: a protein with [27] G. J. McKay, G. Silvestri, U. Chakravarthy et al., “Variations multiple hats,” Arteriosclerosis, Thrombosis, and Vascular in apolipoprotein E frequency with age in a pooled analysis Biology, vol. 20, no. 8, pp. 1852–1853, 2000. of a large group of older people,” American Journal of [43] J. Davignon, “Apolipoprotein E and atherosclerosis beyond Epidemiology, vol. 173, no. 12, pp. 1357–1364, 2011. lipid effect,” Arteriosclerosis, Thrombosis, and Vascular Biol- [28]J.Deelen,M.Beekman,H.W.Uhetal.,“Genome-wide ogy, vol. 25, no. 2, pp. 267–269, 2005. association study identifies a single major locus contributing [44] H. Zhang, L. M. Wu, and J. Wu, “Cross-talk between to survival into old age; the APOE locus revisited,” Aging Cell, apolipoprotein E and cytokines,” Mediators of Inflammation, vol. 10, no. 4, pp. 686–698, 2011. vol. 2011, Article ID 949072, 10 pages, 2011. [29] S. P. Mooijaart, J. F. P. Berbee,D.VanHeemstetal.,“ApoE´ [45] M. E. Kelly, M. A. Clay, M. J. Mistry, H. M. Hsieh-Li, and J. A. plasma levels and risk of cardiovascular mortality in old age,” K. Harmony, “Apolipoprotein E inhibition of proliferation of PLoS Medicine, vol. 3, no. 6, article 176, 2006. mitogen-activated T lymphocytes: production of interleukin Cardiology Research and Practice 13

2 with reduced biological activity,” Cellular Immunology, vol. [62] F. E. Thorngate, L. L. Rudel, R. L. Walzem, and D. L. Williams, 159, no. 2, pp. 124–139, 1994. “Low levels of extrahepatic nonmacrophage apoE inhibit [46] D. R. Riddell, A. Graham, and J. S. Owen, “Apolipoprotein atherosclerosis without correcting hypercholesterolemia in E inhibits platelet aggregation through the L- arginine:nitric apoE-deficient mice,” Arteriosclerosis, Thrombosis, and Vascu- oxide pathway: implications for vascular disease,” The Journal lar Biology, vol. 20, no. 8, pp. 1939–1945, 2000. of Biological Chemistry, vol. 272, no. 1, pp. 89–95, 1997. [63] R. L. Raffai,S.M.Loeb,andK.H.Weisgraber,“Apolipopro- [47] A. K. Stannard, D. R. Riddell, S. M. Sacre et al., “Cell- tein E promotes the regression of atherosclerosis indepen- derived apolipoprotein E (ApoE) particles inhibit vascular dently of lowering plasma cholesterol levels,” Arteriosclerosis, cell adhesion molecule-1 (VCAM-1) expression in human Thrombosis, and Vascular Biology, vol. 25, no. 2, pp. 436–441, endothelial cells,” The Journal of Biological Chemistry, vol. 2005. 276, no. 49, pp. 46011–46016, 2001. [64] N. Yamada, I. Inoue, M. Kawamura et al., “Apolipoprotein [48] S. M. Sacre, A. K. Stannard, and J. S. Owen, “Apolipoprotein E prevents the progression of atherosclerosis in Watanabe E(apoE)isoformsdifferentially induce nitric oxide produc- heritable hyperlipidemic rabbits,” Journal of Clinical Investi- tion in endothelial cells,” FEBS Letters, vol. 540, no. 1–3, pp. gation, vol. 89, no. 2, pp. 706–711, 1992. 181–187, 2003. [65] I. R. Nikoulin and L. K. Curtiss, “An apolipoprotein E syn- [49] D. Y. Hui and J. E. Basford, “Distinct signaling mechanisms thetic peptide targets to lipoproteins in plasma and mediates for apoE inhibition of cell migration and proliferation,” both cellular lipoprotein interactions in vitro and acute Neurobiology of Aging, vol. 26, no. 3, pp. 317–323, 2005. clearance of cholesterol-rich lipoproteins in vivo,” Journal of [50] K. Skal˚ en,´ M. Gustafsson, E. Knutsen Rydberg et al., “Suben- Clinical Investigation, vol. 101, no. 1, pp. 223–234, 1998. dothelial retention of atherogenic lipoproteins in early [66] O. F. Sharifov, G. Nayyar, D. W. Garber et al., “Apolipoprotein atherosclerosis,” Nature, vol. 417, no. 6890, pp. 750–754, E mimetics and cholesterol-lowering properties,” American 2002. Journal of Cardiovascular Drugs, vol. 11, no. 6, pp. 371–381, [51] T. Vogel, N. Guo, R. Guy et al., “Apolipoprotein E: a potent 2011. inhibitor of endothelial and tumor cell proliferation,” Journal [67] G. Nayyar, S. P. Handattu, C. E. Monroe et al., “Two adjacent of Cellular Biochemistry, vol. 54, no. 3, pp. 299–308, 1994. domains (141-150 and 151-160) of apoE covalently linked [52] M. Miyata and J. D. Smith, “Apolipoprotein E allele-specific to a class A amphipathic helical peptide exhibit opposite antioxidant activity and effects on cytotoxicity by oxidative ff β atherogenic e ects,” Atherosclerosis, vol. 213, no. 2, pp. 449– insults and -amyloid peptides,” Nature Genetics, vol. 14, no. 457, 2010. 1, pp. 55–61, 1996. [68] L. Cioffi, F. G. Sturtz, S. Wittmer et al., “A novel endothelial [53] T. Pham, A. Kodvawala, and D. Y. Hui, “The receptor binding cell-based gene therapy platform for the in vivo delivery of domain of apolipoprotein E is responsible for its antioxidant apolipoprotein E,” Gene Therapy, vol. 6, no. 6, pp. 1153–1159, activity,” Biochemistry, vol. 44, no. 20, pp. 7577–7582, 2005. 1999. [54] L. Mabile, C. Lefebvre, J. Lavigne et al., “Secreted apolipopro- [69] A. D. Tagalakis, I. A. Diakonov, I. R. Graham et al., tein E reduces macrophage-mediated LDL oxidation in an “Apolipoprotein E delivery by peritoneal implantation of isoform-dependent way,” Journal of Cellular Biochemistry, encapsulated recombinant cells improves the hyperlipi- vol. 90, no. 4, pp. 766–776, 2003. daemic profile in apoE-deficient mice,” Biochimica et Bio- [55] L. M. Tsoi, K. Y. Wong, Y. M. Liu, and Y. Y. Ho, “Apopro- physica Acta, vol. 1686, no. 3, pp. 190–199, 2005. tein E isoform-dependent expression and secretion of pro- inflammatory cytokines TNF-α and IL-6 in macrophages,” [70] H. Shimano, N. Yamada, M. Katsuki et al., “Overexpression Archives of Biochemistry and Biophysics, vol. 460, no. 1, pp. of apolipoprotein E in transgenic mice: marked reduction 33–40, 2007. in plasma lipoproteins except high density lipoprotein and [56] T. Hayek, J. Oiknine, J. G. Brook, and M. Aviram, “Increased resistance against diet-induced hypercholesterolemia,” Pro- plasma and lipoprotein lipid peroxidation in apo E-deficient ceedings of the National Academy of Sciences of the United mice,” Biochemical and Biophysical Research Communica- States of America, vol. 89, no. 5, pp. 1750–1754, 1992. tions, vol. 201, no. 3, pp. 1567–1574, 1994. [71] K. Yamamoto, H. Shimano, M. Shimada et al., “Overexpres- [57] D. J. Grainger, J. Reckless, and E. McKilligin, “Apolipoprotein sion of apolipoprotein E prevents development of diabetic E modulates clearance of apoptotic bodies in vitro and hyperlipidemia in transgenic mice,” Diabetes,vol.44,no.5, in vivo, resulting in a systemic proinflammatory state in pp. 580–585, 1995. apolipoprotein E-deficient mice,” Journal of Immunology, vol. [72] S. Bellosta, R. W. Mahley, D. A. Sanan et al., “Macrophage- 173, no. 10, pp. 6366–6375, 2004. specific expression of human apolipoprotein E reduces [58] K. Ali, M. Middleton, E. Pure,´ and D. J. Rader, “Apolipopro- atherosclerosis in hypercholesterolemic apolipoprotein E- tein E suppresses the type I inflammatory response in vivo,” null mice,” Journal of Clinical Investigation,vol.96,no.5,pp. Circulation Research, vol. 97, no. 9, pp. 922–927, 2005. 2170–2179, 1995. [59] S. E. Roselaar and A. Daugherty, “Apolipoprotein E-deficient [73] M. F. Linton, J. B. Atkinson, and S. Fazio, “Prevention of mice have impaired innate immune responses to Listeria atherosclerosis in apolipoprotein E-deficient mice by bone monocytogenes in vivo,” Journal of Lipid Research, vol. 39, no. marrow transplantation,” Science, vol. 267, no. 5200, pp. 9, pp. 1740–1743, 1998. 1034–1037, 1995. [60] De Bont, Netea, Demacker, Kullberg, Van Der Meer, and Stal- [74] W. Shi, X. Wang, N. J. Wang, W. H. McBride, and A. enhoef, “Apolipoprotein E-deficient mice have an impaired J. Lusis, “Effect of macrophage-derived apolipoprotein E immune response to Klebsiella pneumoniae,” European Jour- on established atherosclerosis in apolipoprotein E-deficient nal of Clinical Investigation, vol. 30, no. 9, pp. 818–822, 2000. mice,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. [61] P. Van Den Elzen, S. Garg, L. Leon´ et al., “Apolipoprotein- 20, no. 10, pp. 2261–2266, 2000. mediated pathways of lipid antigen presentation,” Nature, [75] V. S. Kashyap, S. Santamarina-Fojo, D. R. Brown et al., vol. 437, no. 7060, pp. 906–910, 2005. “Apolipoprotein E deficiency in mice: gene replacement 14 Cardiology Research and Practice

and prevention of atherosclerosis using adenovirus vectors,” highly efficient transduction of murine and nonhuman Journal of Clinical Investigation, vol. 96, no. 3, pp. 1612–1620, primate liver,” Blood, vol. 107, no. 7, pp. 2653–2661, 2006. 1995. [89] E. Osman, V. Evans, I. R. Graham et al., “Preliminary evalua- [76] K. Tsukamoto, P. Smith, J. M. Glick, and D. J. Rader, “Liver- tion of a self-complementary AAV2/8 vector for hepatic gene directed gene transfer and prolonged expression of three transfer of human apoE3 to inhibit atherosclerotic lesion major human apoE isoforms in apoE-deficient mice,” Journal development in apoE-deficient mice,” Atherosclerosis, vol. of Clinical Investigation, vol. 100, no. 1, pp. 107–114, 1997. 204, no. 1, pp. 121–126, 2009. [77] R. K. Tangirala, D. Pratico,´ G. A. FitzGerald et al., “Reduction [90]H.Wientgen,F.E.Thorngate,S.Omerhodzicetal.,“Sub- of isoprostanes and regression of advanced atherosclerosis by physiologic apolipoprotein E (ApoE) plasma levels inhibit apolipoprotein E,” The Journal of Biological Chemistry, vol. neointimal formation after arterial injury in ApoE-deficient 276, no. 1, pp. 261–266, 2001. mice,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. [78] J. D. Harris, I. R. Graham, S. Schepelmann et al., “Acute 24, no. 8, pp. 1460–1465, 2004. regression of advanced and retardation of early aortic [91] M. Rinaldi, A. L. Catapano, P. Parrella et al., “Treatment atheroma in immunocompetent apolipoprotein-E (apoE) of severe hypercholesterolemia in apolipoprotein E-deficient deficient mice by administration of a second generation mice by intramuscular injection of plasmid DNA,” Gene [E1−,E31−, polymerase1−] adenovirus vector expressing Therapy, vol. 7, no. 21, pp. 1795–1801, 2000. human apoE,” Human Molecular Genetics, vol. 11, no. 1, pp. [92] T. Athanasopoulos, J. S. Owen, D. Hassall et al., “Intra- 43–58, 2002. muscular injection of a plasmid vector expressing human [79] I. H. Kim, A. Jozkowicz,´ P. A. Piedra, K. Oka, and L. Chan, apolipoprotein E limits progression of xanthoma and aortic “Lifetime correction of genetic deficiency in mice with atheroma in apoE-deficient mice,” Human Molecular Genet- a single injection of helper-dependent adenoviral vector,” ics, vol. 9, no. 17, pp. 2545–2551, 2000. Proceedings of the National Academy of Sciences of the United [93] V. Evans, H. Foster, I. R. Graham et al., “Human apolipopro- States of America, vol. 98, no. 23, pp. 13282–13287, 2001. tein E expression from mouse skeletal muscle by elec- [80] G. P. Gao, M. R. Alvira, L. Wang, R. Calcedo, J. Johnston, and trotransfer of nonviral DNA (plasmid) and pseudotyped J. M. Wilson, “Novel adeno-associated viruses from rhesus recombinant adeno-associated virus (AAV2/7),” Human monkeys as vectors for human gene therapy,” Proceedings Gene Therapy, vol. 19, no. 6, pp. 569–578, 2008. of the National Academy of Sciences of the United States of [94] V. C. Evans, I. R. Graham, T. Athanasopoulos et al., “Adeno- America, vol. 99, no. 18, pp. 11854–11859, 2002. associated virus serotypes 7 and 8 outperform serotype 9 [81]C.Lebherz,G.Gao,J.P.Louboutin,J.Millar,D.Rader,and in expressing atheroprotective human apoE3 from mouse J. M. Wilson, “Gene therapy with novel adeno-associated skeletal muscle,” Metabolism, vol. 60, no. 4, pp. 491–498, virus vectors substantially diminishes atherosclerosis in a 2011. murine model of familial hypercholesterolemia,” Journal of [95] O. Smithies, R. G. Gregg, S. S. Boggs, M. A. Koralewski, Gene Medicine, vol. 6, no. 6, pp. 663–672, 2004. and R. S. Kucherlapati, “Insertion of DNA sequences into [82] H. Nakai, S. Fuess, T. A. Storm, S. I. Muramatsu, Y. Nara, the human chromosomal β-globin locus by homologous and M. A. Kay, “Unrestricted hepatocyte transduction with recombination,” Nature, vol. 317, no. 6034, pp. 230–234, adeno-associated virus serotype 8 vectors in mice,” Journal of 1985. Virology, vol. 79, no. 1, pp. 214–224, 2005. [96] K. R. Thomas, K. R. Folger, and M. R. Capecchi, “High fre- [83]H.Chao,P.E.Monahan,Y.Liu,R.J.Samulski,andC. quency targeting of genes to specific sites in the mammalian E. Walsh, “Sustained and complete phenotype correction genome,” Cell, vol. 44, no. 3, pp. 419–428, 1986. of hemophilia B mice following intramuscular injection of [97] R. P. Moerschell, S. Tsunasawa, and F. Sherman, “Transfor- AAV1 serotype vectors,” Molecular Therapy,vol.4,no.3,pp. mation of yeast with synthetic oligonucleotides,” Proceedings 217–222, 2001. of the National Academy of Sciences of the United States of [84] M. J. Blankinship, P. Gregorevic, J. M. Allen et al., “Efficient America, vol. 85, no. 2, pp. 524–528, 1988. transduction of skeletal muscle using vectors based on [98] K. Yoon, A. Cole-Strauss, and E. B. Kmiec, “Targeted gene adeno-associated virus serotype 6,” Molecular Therapy, vol. correction of episomal DNA in mammalian cells mediated 10, no. 4, pp. 671–678, 2004. by a chimeric RNA·DNA oligonucleotide,” Proceedings of the [85] J. P. Louboutin, L. Wang, and J. M. Wilson, “Gene transfer National Academy of Sciences of the United States of America, into skeletal muscle using novel AAV serotypes,” Journal of vol. 93, no. 5, pp. 2071–2076, 1996. Gene Medicine, vol. 7, no. 4, pp. 442–451, 2005. [99] A. Cole-Strauss, K. Yoon, Y. Xiang et al., “Correction of the [86]C.E.Thomas,T.A.Storm,Z.Huang,andM.A.Kay, mutation responsible for sickle cell anemia by an RNA- DNA “Rapid uncoating of vector genomes is the key to efficient oligonucleotide,” Science, vol. 273, no. 5280, pp. 1386–1389, liver transduction with pseudotyped adeno-associated virus 1996. vectors,” Journal of Virology, vol. 78, no. 6, pp. 3110–3122, [100] Y. Xiang, A. Cole-Strauss, K. Yoon, J. Gryn, and E. B. Kmiec, 2004. “Targeted gene conversion in a mammalian CD34+-enriched [87]D.M.McCarty,H.Fu,P.E.Monahan,C.E.Toulson,P. cell population using a chimeric RNA/DNA oligonucleotide,” Naik, and R. J. Samulski, “Adeno-associated virus terminal JournalofMolecularMedicine, vol. 75, no. 11-12, pp. 829– repeat (TR) mutant generates self-complementary vectors 835, 1997. to overcome the rate-limiting step to transduction in vivo,” [101] K. R. Thomas and M. R. Capecchi, “Recombinant DNA Gene Therapy, vol. 10, no. 26, pp. 2112–2118, 2003. technique and sickle cell anemia research,” Science, vol. 275, [88]A.C.Nathwani,J.T.Gray,C.Y.C.Ngetal.,“Self-comple- no. 5305, pp. 1404–1405, 1997. mentary adeno-associated virus vectors containing a novel [102] B. T. Kren, A. Cole-Strauss, E. B. Kmiec, and C. J. Steer, “Tar- liver-specific human factor IX expression cassette enable geted nucleotide exchange in the alkaline phosphatase gene Cardiology Research and Practice 15

of HuH-7 cells mediated by a chimeric RNA/DNA oligonu- the replication of corrected cells following gene editing,” cleotide,” Hepatology, vol. 25, no. 6, pp. 1462–1468, 1997. Journal of Gene Medicine, vol. 11, no. 3, pp. 267–274, 2009. [103] P. Bandyopadhyay, Y. Ma, C. Linehan-Stieers, B. T. Kren, and [117] R. L. Raffai,R.McPherson,K.H.Weisgraberetal.,“Antibody C. J. Steer, “Nucleotide exchange in genomic DNA of rat hep- phenotyping test for the human apolipoprotein E2 isoform,” atocytes using RNA/DNA oligonucleotides: targeted delivery Clinical Chemistry, vol. 449, no. 9, pp. 1524–1526, 2003. of liposomes and polyethyleneimine to the asialoglycoprotein [118] L. Naldini, “Ex vivo gene transfer and correction for cell- receptor,” The Journal of Biological Chemistry, vol. 274, no. 15, based therapies,” Nature Reviews Genetics,vol.12,no.5,pp. pp. 10163–10172, 1999. 301–315, 2011. [104] B. T. Kren, P. Banoyopaohyay, and C. J. Steer, “In vivo [119] G. J. Sullivan, D. C. Hay, I. H. Park et al., “Generation of site-directed mutagenesis of the factor IX gene by chimetic functional human hepatic endoderm from human induced RND/DNA oligonucleotides,” Nature Medicine, vol. 4, no. 3, pluripotent stem cells,” Hepatology, vol. 51, no. 1, pp. 329– pp. 285–290, 1998. 335, 2010. [105] B. T. Kren, B. Parashar, P. Bandyopadhyay, N. R. Chowdhury, [120] A. Ghodsizadeh, A. Taei, M. Totonchi et al., “Generation J. R. Chowdhury, and C. J. Steer, “Correction of the UDP- of liver disease-specific induced pluripotent stem cells along glucuronosyltransferase gene defect in the Gunn rat model with efficient differentiation to functional hepatocyte-like of Crigler-Najjar syndrome type I with a chimeric oligonu- cells,” Stem Cell Reviews and Reports, vol. 6, no. 4, pp. 622– cleotide,” Proceedings of the National Academy of Sciences of 632, 2010. the United States of America, vol. 96, no. 18, pp. 10349–10354, [121] S. M. Choi, Y. Kim, H. Liu, P.Chaudhari, Z. Ye, and Y. Y. Jang, 1999. “Liver engraftment potential of hepatic cells derived from [106] A. D. Tagalakis, I. R. Graham, D. R. Riddell, J. G. Dickson, patient-specific induced pluripotent stem cells,” Cell Cycle, and J. S. Owen, “Gene correction of the apolipoprotein (apo) vol. 10, no. 15, pp. 2423–2427, 2011. E2 phenotype to wild-type apoE3 by in situ chimeraplasty,” [122] H. Li, V. Haurigot, Y. Doyon et al., “In vivo genome editing The Journal of Biological Chemistry, vol. 276, no. 16, pp. restores haemostasis in a mouse model of haemophilia,” 13226–13230, 2001. Nature, vol. 475, no. 7355, pp. 217–221, 2011. [107] A. D. Tagalakis, J. G. Dickson, J. S. Owen, and J. P. Simons, [123] K. Paiboonsukwong, F. Ohbayashi, H. Shiiba, E. Aizawa, T. “Correction of the neuropathogenic human apolipopro- Yamashita, and K. Mitani, “Correction of mutant Fanconi tein E4 (APOE4) gene to APOE3 in vitro using syn- anemia gene by homologous recombination in human thetic RNA/DNA oligonucleotides (chimeraplasts),” Journal hematopoietic cells using adeno-associated virus vector,” of Molecular Neuroscience, vol. 25, no. 1, pp. 95–103, 2005. Journal of Gene Medicine, vol. 11, no. 11, pp. 1012–1019, 2009. [108] J. Albuquerque-Silva, G. Vassart, J. Lavinha, and M. J. Abramowicz, “Chimeraplasty validation,” Nature Biotechnol- [124] K. Mitsui, K. Suzuki, E. Aizawa et al., “Gene targeting in ogy, vol. 19, no. 11, article 1011, 2001. human pluripotent stem cells with adeno-associated virus vectors,” Biochemical and Biophysical Research Communica- [109] A. Manzano, Z. Mohri, G. Sperber et al., “Failure to generate tions, vol. 388, no. 4, pp. 711–717, 2009. atheroprotective apolipoprotein AI phenotypes using syn- [125] I. F. Khan, R. K. Hirata, P. R. Wang et al., “Engineering of thetic RNA/DNA oligonucleotides (chimeraplasts),” Journal human pluripotent stem cells by AAV-mediated gene target- of Gene Medicine, vol. 5, no. 9, pp. 795–802, 2003. ing,” Molecular Therapy, vol. 18, no. 6, pp. 1192–1199, 2010. [110] H. B. Gamper, A. Cole-Strauss, R. Metz, H. Parekh, R. Kumar, [126] A. Vasileva, R. M. Linden, and R. Jessberger, “Homologous and E. B. Kmiec, “A plausible mechanism for gene correction recombination is required for AAV-mediated gene targeting,” by chimeric oligonucleotides,” Biochemistry, vol. 39, no. 19, Nucleic acids research., vol. 34, no. 11, pp. 3345–3360, 2006. pp. 5808–5816, 2000. [127] G. H. Liu, K. Suzuki, J. Qu et al., “Targeted gene correction of [111] H. B. Gamper, H. Parekh, M. C. Rice, M. Bruner, H. Youkey, laminopathy-associated LMNA mutations in patient-specific and E. B. Kmiec, “The DNA strand of chimeric RNA/DNA iPSCs,” Cell Stem Cell, vol. 8, no. 6, pp. 688–694, 2011. oligonucleotides can direct gene repair/conversion activity [128] E. Aizawa, Y. Hirabayashi, Y. Iwanaga et al., “Efficient and in mammalian and plant cell-free extracts,” Nucleic Acids accurate homologous recombination in hESCs and hiPSCs Research, vol. 28, no. 21, pp. 4332–4339, 2000. using helper-dependent adenoviral vectors,” Molecular Ther- [112] H. Parekh-Olmedo, L. Ferrara, E. Brachman, and E. B. apy, vol. 20, no. 2, pp. 424–431, 2012. Kmiec,“Genetherapyprogressandprospects:targetedgene [129] H. Pan, W. Zhang, W. Zhang, and G. H. Liu, “Find and repair,” Gene Therapy, vol. 12, no. 8, pp. 639–646, 2005. replace: editing human genome in pluripotent stem cells,” [113] N. M. Jensen, T. Dalsgaard, M. Jakobsen et al., “An update Protein and Cell, vol. 2, no. 12, pp. 950–956, 2011. on targeted gene repair in mammalian cells: methods and [130] K. R. Thomas, C. Deng, and M. R. Capecchi, “High-fidelity mechanisms,” Journal of Biomedical Science, vol. 18, no. 1, gene targeting in embryonic stem cells by using sequence article 10, 2011. replacement vectors,” Molecular and Cellular Biology, vol. 12, [114] S. Radecke, F. Radecke, I. Peter, and K. Schwarz, “Physical no. 7, pp. 2919–2923, 1992. incorporation of a single-stranded oligodeoxynucleotide [131] M. Jasin, “Genetic manipulation of genomes with rare- during targeted repair of a human chromosomal locus,” cutting endonucleases,” Trends in Genetics, vol. 12, no. 6, pp. Journal of Gene Medicine, vol. 8, no. 2, pp. 217–228, 2006. 224–228, 1996. [115] P. Disterer, J. P. Simons, and J. S. Owen, “Validation of [132] M. H. Porteus and D. Baltimore, “Chimeric nucleases stimu- oligonucleotide-mediated gene editing,” Gene Therapy, vol. late gene targeting in human cells,” Science, vol. 300, no. 5620, 16, no. 6, pp. 824–826, 2009. p. 763, 2003. [116] I. Papaioannou, P. Disterer, and J. S. Owen, “Use of internally [133] M. H. Porteus, “Mammalian gene targeting with designed nuclease-protected single-strand DNA oligonucleotides and zinc finger nucleases,” Molecular Therapy,vol.13,no.2,pp. silencing of the mismatch repair protein, MSH2, enhances 438–446, 2006. 16 Cardiology Research and Practice

[134] M. Porteus, “Using homologous recombination to manipu- early onset parkinson point mutations,” Cell, vol. 146, no. 2, latethegenomeofhumansomaticcells,”Biotechnology and pp. 318–331, 2011. Genetic Engineering Reviews, vol. 24, pp. 195–212, 2007. [151] T. Li, S. Huang, X. Zhao et al., “Modularly assembled [135] D. Carroll, “Progress and prospects: zinc-finger nucleases as designer TAL effector nucleases for targeted gene knockout gene therapy agents,” Gene Therapy, vol. 15, no. 22, pp. 1463– and gene replacement in eukaryotes,” Nucleic Acids Research, 1468, 2008. vol. 39, no. 14, pp. 6315–6325, 2011. [136] S. H. Rahman, M. L. Maeder, J. K. Joung, and T. Cathomen, [152] M. Baker, “Gene-editing nucleases,” Nature Methods, vol. 9, “Zinc-finger nucleases for somatic gene therapy: the next no. 1, pp. 23–26, 2012. frontier,” Human Gene Therapy, vol. 22, no. 8, pp. 925–933, [153] V. G. Shore and B. Shore, “Heterogeneity of human plasma 2011. very low density lipoproteins. Separation of species differing [137] C. Mussolino and T. Cathomen, “On target? Tracing zinc- in protein components,” Biochemistry, vol. 12, no. 3, pp. 502– finger-nuclease specificity,” Nature Methods,vol.8,no.9,pp. 507, 1973. 725–726, 2011. [138] S. M. Pruett-Miller, D. W. Reading, S. N. Porter, and M. H. Porteus, “Attenuation of zinc finger nuclease toxicity by small-molecule regulation of protein levels,” PLoS Genetics, vol. 5, no. 2, Article ID e1000376, 2009. [139] J. D. Sander, E. J. Dahlborg, M. J. Goodwin et al., “Selection- free zinc-finger-nuclease engineering by context-dependent assembly (CoDA),” Nature Methods, vol. 8, no. 1, pp. 67–69, 2011. [140] M. Isalan, “Zinc-finger nucleases: how to play two good hands,” Nature Methods, vol. 9, no. 1, pp. 32–34, 2012. [141] M. Christian, T. Cermak, E. L. Doyle et al., “Targeting DNA double-strand breaks with TAL effector nucleases,” Genetics, vol. 186, no. 2, pp. 756–761, 2010. [142] T. Li, S. Huang, X. Zhao et al., “Modularly assembled designer TAL effector nucleases for targeted gene knockout and gene replacement in eukaryotes,” Nucleic Acids Research, vol. 39, no. 14, pp. 6315–6325, 2011. [143] C. Mussolino, R. Morbitzer, F. Lutge,¨ N. Dannemann, T. Lahaye, and T. Cathomen, “A novel TALE nuclease scaffold enables high genome editing activity in combination with low toxicity,” Nucleic Acids Research, vol. 39, no. 21, pp. 9283– 9293, 2011. [144] W. A. Kues and H. Niemann, “Advances in farm animal transgenesis,” Preventive Veterinary Medicine, vol. 102, no. 2, pp. 146–156, 2011. [145] G. Silva, L. Poirot, R. Galetto et al., “Meganucleases and other tools for targeted genome engineering: perspectives and challenges for gene therapy,” Current Gene Therapy, vol. 11, no. 1, pp. 11–27, 2011. [146] E. M. Handel,,¨ K. Gellhaus, K. Khan et al., “Versatile and efficient genome editing in human cells by combining zinc- finger nucleases with adeno-associated viral vectors,” Human Gene Therapy, vol. 23, no. 3, pp. 321–329, 2012. [147] P. Asuri, M. A. Bartel, T. Vazin, J. H. Jang, T. B. Wong, and D. V. Schaffer, “Directed evolution of adeno-associated virus for enhanced gene delivery and gene targeting in human pluripotent stem cells,” Molecular Therapy,vol.20,no.2,pp. 329–338, 2012. [148] B. L. Ellis, M. L. Hirsch, S. N. Porter, R. J. Samulski, and M. H. Porteus, “Zinc-finger nuclease-mediated gene correction using single AAV vector transduction and enhancement by Food and Drug Administration-approved drugs,” Gene Therapy. In press. [149] F. Chen, S. M. Pruett-Miller, Y. Huang et al., “High-frequency genome editing using ssDNA oligonucleotides with zinc- finger nucleases,” Nature Methods, vol. 8, no. 9, pp. 753–757, 2011. [150] F. Soldner, J. Laganiere,` A. W. Cheng et al., “Generation of isogenic pluripotent stem cells differing exclusively at two Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 971614, 10 pages doi:10.1155/2012/971614

Review Article Cell Therapies for Heart Function Recovery: Focus on Myocardial Tissue Engineering and Nanotechnologies

Marie-Noelle¨ Giraud,1 Anne Geraldine´ Guex,2, 3 and Hendrik T. Tevaearai2

1 Cardiology, Department of Medicine, Faculty of Science, University of Fribourg, Chemin du Mus´ee 5, 1700 Fribourg, Switzerland 2 Clinic for Cardiovascular Surgery, Inselspital Berne, Berne University Hospital and University of Berne, Switzerland 3 Empa, Swiss Federal Laboratories for Material Science and Technology, 9014 St. Gallen, Switzerland

Correspondence should be addressed to Marie-Noelle¨ Giraud, [email protected]

Received 29 July 2011; Accepted 6 February 2012

Academic Editor: Daryll M. Baker

Copyright © 2012 Marie-Noelle¨ Giraud et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Cell therapies have gained increasing interest and developed in several approaches related to the treatment of damaged myo- cardium. The results of multiple clinical trials have already been reported, almost exclusively involving the direct injection of stem cells. It has, however, been postulated that the efficiency of injected cells could possibly be hindered by the mechanical trauma due to the injection and their low survival in the hostile environment. It has indeed been demonstrated that cell mortality due to the injection approaches 90%. Major issues still need to be resolved and bed-to-bench followup is paramount to foster clinical implementations. The tissue engineering approach thus constitutes an attractive alternative since it provides the opportunity to deliver a large number of cells that are already organized in an extracellular matrix. Recent laboratory reports confirmed the inter- est of this approach and already encouraged a few groups to investigate it in clinical studies. We discuss current knowledge regard- ing engineered tissue for myocardial repair or replacement and in particular the recent implementation of nanotechnological approaches.

1. Introduction continues to encourage the scientific and clinical communi- ties to multiply the laboratory investigations and consider the It was long believed that the adult heart does not regenerate. value of stem cell therapy in clinical protocols. Depending on The recent discovery of cardiac stem cells (CSCs), however, the clinical need and the rationale, transplantation of isolated challenged this dogma [1]. Since then, several populations cells or implantation of an engineered muscle graft is under of CSCs have been identified and distinguished by means consideration as presented in Figure 1. As illustrated, the of their surface markers. In addition, using a fascinating ap- concept for cell-based therapy is thus quite straightforward; proach based on the comparison of C14 incorporation before however, its implementation faces numerous challenges. and after the explosion of the atomic bomb, Bergmann et al. In this paper we present the important questions that recently demonstrated that human cardio-myocytes in fact remain to be investigated to ascertain a successful translation regenerate at a rate of approximately one percent per year at of current experimental knowledge regarding cell therapy the age of 25 and 0.45% at the age of 75 [2]. for myocardial repair/replacement. In particular, we empha- Beside their possible implication in this regenerative pro- size the critical importance of favoring a multidisciplinary cess, the exact physiological function of CSCs has not yet approach including biotechnologies, material science, and been fully clarified. Their role in pathological situations is nanotechnologies to engineer myocardial tissue. also unclear since, in case of myocardial injury such as after 2. Clinical Trials a myocardial infarction, their potential regenerative capacity is clearly overwhelmed. Nevertheless, the rapid progress Compelling evidence of the beneficial effect of isolated cell in understanding myocardial regenerative mechanisms transplantation to the heart including improvement in cardiac 2 Cardiology Research and Practice

Cell expansion and differentiation

Isolation A B C D In vitro

Cell-secreted Biopsy Isolated cells Engineered Functionalized factors tissue matrix

Intracoronary injection In vivo

Intramyocardial Epicardial Myocardial injection implantation infarction Figure 1: Cell therapy approaches for myocardial infarction: cells are isolated from biopsies, expanded, and eventually differentiated in vitro following specific culture conditions. Conditioned medium containing secreted or lyophilized factors (A) or isolated cells (B) are injected directly within the myocardium or within the coronaries. Further in vitro process from cultured cells enables the development of structured engineered muscle tissue with or without contracting properties that can be directly sutured or glued at the surface of the infarct (C). Functionalized matrix combining biologically active factors and engrafted cells (D) represents a more sophisticated alternative.

contractile function, decrease in left ventricular remodeling, interest for myocardial tissue engineering. Recent stud- reduction of the infarct size, and increase in vascular density ies provide convincing experimental short-term outcomes was provided by early experimental studies [3]. Conse- showing recovery of heart function; our group contributed to quently, rapid clinical trials testing the safety and efficiency this proof of concept with several types of engineered tissues of cell therapy have been undertaken and are ongoing [4– investigated for functional recovery including long-term fol- 6]. However, in a general manner, beneficial effects on lowup [10–12]. To date, the first two clinical trials have heart function and regeneration observed after cell therapy been initiated [13, 14]. The first twenty patients with post- in animal models were not always followed by convincing infarction myocardial scar received autologous bone marrow clinical outcomes [7, 8]. The modest or absent improvement stem cells either directly injected in and around the infarct of heart function has been confirmed in the Cochrane report, or seeded on a collagen matrix, which was then placed and presenting a recent meta-analysis focusing on bone marrow sutured on the infarcted area. This pioneer study not only stem cells transplantation [9]. It has been hypothesized that confirmed the feasibility and safety of the procedure but also the modest or indeed lack of functional improvement may already suggested a benefit in favor of the combination of be the result of poor cell specificity and quality as well cells and matrix. Results of the recently launched second clin- as technical pitfalls during injection. The report concluded ical trial, describing the implantation of an engineered con- with the following major issue to be investigated: define struct composed of stacks of myoblast sheets, are pending the optimal type and the dose of stem cells, the route and [14]. timing of delivery after myocardial infarction, and long-term outcomes. In addition, mechanisms of action and in parti- 3. Major Challenges: What Research Is Needed cular the role of injected stem cells in the management of to Make Cell-Based Treatments a Reality? acute myocardial infarction are of particular relevance to im- prove treatment efficiency. Furthermore, the possibility that 3.1. In Vivo Investigations. The importance of experimental the injured microenvironment has low ability to permit settings and in particular large animal models to provide cell survival and differentiation has been raised. Indeed, predictor features for cell therapy applied in human clinical the rather hostile, hypoxic, stressed and remodeled cardiac trials has been emphasized by van der Spoel et al. [15]. environment as well as the immunologic and inflammatory The authors performed a meta-analysis for cell therapy on milieu related to the patient’s disease is certainly unfavorable large animal models of acute and chronic cardiac ischemia. conditions for cell growth and differentiation. They determined a short-term 7.5% global ejection function The search for new strategies to overcome drawbacks improvement due to an increased end systolic volume. They from direct cell implantation has resulted in an increased reported a prevalence of mesenchymal stem cells (MSCs), Cardiology Research and Practice 3 a high number of cells, and better outcomes for chronic Growing numbers of studies provide evidence that the ischemia. However, no effect of distinct delivery routes was beneficial effect of delivered cells is mediated via a paracrine examined. effect. Cell secretions of cardioprotective, angiogenic, or Experimental and preclinical investigations have mostly stem-cell-recruiting factors are expected to trigger heart re- been performed in small (rodent) or large (pig) animal mod- generation. A large panel of secreted cytokines and chemoat- els of heart failure following a myocardial infarction induced tractants [1] are believed to bring their beneficial effect to by ligation of the left anterior descending coronary artery the failing heart and suggest a multifactorial effect on angio- (LAD ligation). Various treatments have been tested and genesis [24], inhibition of cardiomyocyte apoptosis [25], compared (Figure 1):cellswereappliedinacuteorchronic antifibrotic effects [26], and mobilization of endogenous phases, cells were injected into the scar or at its periphery, stem cells [1] as well modulation of the inflammatory pro- and tissue constructs were glued or sutured at the surface cesses [27]. of the infarcted area. Typically, morphological and function- al changes of the treated hearts were followed by repeated 3.3. Cell echocardiography or MRI and generally over a 4-week follow-up period [10, 16, 17]. At the end of this observation 3.3.1. Source and Cell Type. Cell types and their potential period and before sacrifice, additional invasive investigations for new medical treatment are presented in Tables 1 and 2. were performed using, for example, a pressure or a con- Stem cells represent a promising cell source due to their high ductance microtip catheter to record and analyze comple- potential for differentiation and expansion capacity. mentary contractile parameters. These studies allowed assessment of functional out- comes. Furthermore, increased interest in cell tracking, cell 3.3.2. How Many Cells Are Required? Cell survival and integration, and survival permitted the development and engraftment in a hostile environment with inflammation, optimization of state-of-the-art technologies as described by fibrosis, and hypoxia are a major concern. It has been de- Terrovitis et al. [18]. In addition, extensive efforts focusing monstrated that more than 90% of injected cells are lost on proteomics and high-throughput screening will enable within the first minutes following injection. Optimization of the discovery of major mechanisms of action and important cell retention after injection, engraftment, and survival are of factors for myocardial recovery and repair [19]. paramount importance to further define the optimal quanti- tyofcellstobeimplanted.Sofar,toovercomethiseffect, ff 3.2. Mechanisms of Action. Experimental cell therapy inves- large numbers of cells have been injected. Dose e ect has tigations showed beneficial outcomes including significant largely been reported [15]. Therefore, the injection of a high number of cells will require a high expansion capacity of improvement in ventricular function, increased wall thick- autologous cells and a massive capacity of expansion if hete- ness, and decreased end diastolic and systolic volumes as rologous cells are used. Alternatively, strategies to improve well as neovascularisation of the scar area. However, decrease cell engraftment and survival have been developed and of the infarct size suggesting myogenesis is still a matter of include preconditioning of the cells prior to transplantation debate and seems to be dependent on the type of cells that (heat shock, hypoxia), increased expression of survival were implanted [20–22]. factors, exposition to prosurvival factors, and the implant- Several potential hypotheses have been raised to explain ation of engineered tissue. the effects and remain to be further investigated. First, a ff girding e ect attenuating the adverse remodeling has been 3.3.3. Further Aspects to Be Considered. Some cell-specific proposed, suggesting prevention of dilatation, modification drawbacks are listed in Table 2. Clinical availability is one of the scar elasticity, and an increase in wall thickness due to a important feature to take into account in the choice of the cell ff cluster of cells or implanted tissues. This e ect may have a source and may limit their relevance in a translational per- minor impact as the large number of cells washed out after spective. Neonatal cardiomyocytes, for example, have been injection results in very small clusters of cells that may not widely used in preclinical studies; however, they were not ex- be sufficient to produce the adequate mechanical strength to ploitable for clinical studies due to low accessibility and prevent remodeling. Furthermore, implantation of an acel- ethical concerns. The same concerns applied for embryonic lular scaffold has little or no effect on cardiac function com- stem cells and increased research investigations to assess pared to the implantation of engineered tissues [12]. Second, their teratogenicity must be undertaken before safe clinical the replacement of lost cardiomyocytes by transplanted cells use. Induced pluripotent stem cells (iPSCs) overcome some is a major issue for tissue repair. Only investigations using major shortcomings such as accessibility, expansion, and neonatal cardiomyocytes and embryonic stem cells could capacity; however, significant improvements to generate report the presence of new cardiomyocytes in the periphery clinical-grade iPSCs are required for clinical translation. of implanted cells [23]. Although the delivery of embryonic Purification is also an important feature. The selection of stem cells or cardiac progenitor cells as committed cells to population clones or heterogenous population of adult stem cardiac lineage could reinforce muscle contractility after dif- cells has been investigated; however, it is not yet clear what ferentiation and may contribute to systolic force, myogenesis type of cell population has the best regenerative capacity. is unlikely to explain the positive outcomes observed after cell Furthermore, immunogenicity of the heterologous cell may therapy using other cell types. limit cell survival. Several lines of research must be carried 4 Cardiology Research and Practice

Table 1: Potential cell source. Definition Source Drawbacks Donor/recipient Autologous Same individual Not always available (genetic diseases, age) Allogenic Same species Immunological issues Xenogenic Different species Ethical issues and rejection Syngenic or isogenic Genetically identical individuals (clones, inbred) Most appropriate for research with animal model Origin/differentiation Primary Tissue or organ/specialized Large expansion needed Secondary Cell bank Cryopreservation/immunological issues Embryonic stem cells (iPSCs) Undifferentiated Ethical issues/purification/teratoma Adult stem cells Commited Selection of type/source

Table 2: Potential cells for new therapeutic treatment.

Change in cardiac function Candidates Concerns Side effects Mechanism of action Clinical trials (% EF versus ctrl.)∗ Human embryonic stem cells Ethics purification Teratoma Differentiation/myogenesis FDA approval Fetal/neonatal cardiac muscle Ethics accessibility Differentiation/myogenesis × cells Induced pluripotent stem cells Teratoma Differentiation/myogenesis × Cardiac stem cells Differentiation/myogenesis  (2009) −0.2; +6.0 Skeletal muscle myoblasts Poor electrocoupling Arrhythmia Paracrine effect  +3; +14 Purification/loss of Bone marrow stem cells Arrhythmia? Paracrine effect  −3.0; +12 function with age Survival and Progenitors controlled Paracrine effect  +2.8, +6.3 differentiation ∗EF: ejection fraction of the treated heart compared to control groups (adapted from Segers and Lee, 2008 [28]).

out to identify the most efficient therapeutic candidate for 3.5. Where? Feasibility, safety, and cell retention are the com- patients with cardiovascular diseases. mon features that may drive the choice of the cell delivery. Different ways to inject the cells have been investigated 3.4. When? The development of cardiac infarct following in clinical trials, such as intracoronary (IC), intramyocar- ischemic injury is rapidly and sequentially associated with dial (IM), transendocardial, interstitial retrograde coronary cell death, release of paracrine factors, inflammation with venous (IVR), epicardial, or systemic injection. Hou et al. leucocytes infiltration, the formation of granulation tissue [30] quantified the retention rates of peripheral blood mono- composed of myofibroblast, macrophage, and collagen, nuclear cells within the swine ischemic heart: IM resulted in spreading of the initial injury to adjacent tissue, and finally amostefficient delivery mode with 11% of cell retention 1 fibrosis. The reorganization of the extracellular matrix allows hour after cell delivery but with a large variability compared for compensation of the loss of cardiomyocytes. This remod- to other techniques. The retention efficiency was confirmed eling will progressively lead to a reduction of cardiac wall in a rodent model after injection of cardiosphere-derived thickness, ventricle dilatation, and more severe heart failure. cells [31]. However, the authors also reported that IM injec- The therapeutic target will define the cell therapy strategy. Therapeutic angiogenesis and/or myogenesis using cell or tion can result in cell loss through the needle track and coro- tissue transplantation might be promising therapeutic strate- nary venous vessels. In addition, IM is also known to induce gies in patients with severe ischemic heart disease or patients myocardium injuries at the site of needle insertion. To date, with end-stage heart failure. Alternatively, stimulation of the the optimal delivery route has not been identified. Their regenerative process may preferentially be beneficial for acute respective advantages and disadvantages are reviewed by Dib ischemia. Using a rat myocardial infarction model, Hu et al. et al. [32]. [29] provided evidence of better outcomes when MSC were implanted 1 week after infarction. The authors suggested that 4. Cell Delivery through Engineered Tissue reduced inflammation as well as early time point in tissue remodeling towards scar formation favors cell engraftment The controllable scaffolds and culture conditions made and angiogenesis. possible by tissue engineering approaches allow the design of Cardiology Research and Practice 5 an adequate microenvironment that not only permits pre- order to create a vascularized tissue of up to 1 mm thickness conditioning the cells in vitro and provides a differentiation [44]. As opposed to the first approaches described here, the direction of the tissue before it is implanted for the prepa- cell sheet approach does not involve the transplantation of ration of cells prior to their implantation but also would an artificial extracellular matrix. Nevertheless, the potential overcome major drawbacks of isolated cells’ transplantation of matrices may be extended since these structures can be related to their survival in inadequate ischemic tissue. In “functionalized” through the addition of chemical com- fact, the matrix of the engineered tissue can be compared pounds or proteins, which may provide specific signals that to the ECM in a corresponding natural tissue. Its function will trigger the behavior of the seeded as well as the host cells. during the tissue engineering process must, however, be distinguished between the in vitro period (preimplantation 5. How Nanotechnology Will Help? or maturation period), the surgery period (implantation period), and the in vivo period (postimplantation period). In the future, the potential of engineered tissue and in part- Regarding the in vitro period, the matrix could be assimilated icular the scaffold type as well as better understanding of to a biocompatible and nontoxic support material for the biomaterial-cell interactions are of paramount importance cells that are planned to be transplanted. The ideal matri- toward successful cardiac regeneration. Initiated with only ces should thus consist of a two- or a three-dimensional a few mandatory factors such as being biocompatible, non- structure that should favor not only cells’ attachment and cytotoxic, and providing a three-dimensional framework for growth but also their further organization and possibly dif- cells to attach and develop, scaffolds for tissue engineering ferentiation toward a highly ordered formation including in- have evolved into more and more highly sophisticated and tercellular contacts. custom tailored constructs. Incorporating peptides, proteins, Considering the implantation phase, the matrix offers a or growth factors renders an inert synthetic scaffold biologi- significant practical advantage over the direct injection of cally active [45, 46] and facilitates regulation of cell expres- cells. For instance, engineered myocardial biografts may be sion via material properties and at the site of interest. Sur- considered as a bandage that can be easily and rapidly ap- face immobilized growth factors bypass the problems of plied at the surface of the infarcted zone. Conversely, the rapid diffusion, short blood plasma half-life, and potential traditional application of cell therapy requires multiple injec- health risk as seen for soluble factors injected into the blood tions to cover a similar zone. The role of the matrix during stream [47, 48]. Focusing on functionalization of cardiac the in vivo phase is variable. On one hand, it may represent a constructs, four main approaches have been investigated so structurally resistant element that can withstand the high and far: (I) smart materials, (II) surface modified materials via permanent mechanical stresses observed during the cardiac adsorption, (III) surface modified materials via covalent im- contraction/relaxation cycles. A second major role during mobilization, and (IV) blended materials. Concepts of sub- this period is its integration within the host tissue and strate functionalization are presented in Figure 2. eventual replacement by a host ECM. For example, recent data have shown increasing evidence that the matrices may 5.1. Smart Materials. Smart materials are materials that alter provide specific signals that will trigger the behavior of the their shape, color, or size in response to an external stimulus. seeded as well as the host cells. Such an external stimulus can be a change in temperature, Various approaches have already demonstrated the possi- pH, electrical or magnetic field, light, or naturally occurring bility of designing myocardial-like structures and are detail- enzymes. In the field of tissue engineering, materials showing ed in recent reviews [33–36]. Briefly, one typical method smart behavior are mainly, if not exclusively, restricted to consists of engineering a construct in vitro by combining a hydrogels showing thermo- or enzyme-responsive behavior. polymeric or a biological scaffold organized in a 3-dimen- As early as the sixties, Wichterle and Lim [49]characterized sional matrix onto which cells will be seeded [20, 37, 38]. a hydrophilic gel for biological use. Although consisting up An updated list of biomaterials used for the treatment of to 99% of water, hydrogels sustained their position over the myocardial infarction was recently assembled by Rane and years and gained increasing interest in tissue engineering and Christman [39]. A second alternative takes advantage of the numerous reviews summarize the concept of bioresponsive self-aggregation of cells when cultured in high density to- hydrogels for tissue engineering or drug delivery [50–55]. gether with collagen, fibrin, laminin, or fibronectin [40]. In The concepts of stimuli-dependent conformational changes another recently described technique, the controlled recellu- of polymers have only recently bridged from chemical labo- larization of a previously decellularized natural matrix was ratories and theoretical application to in vitro and in vivo proposed and showed spectacular results using an entire studies. rat heart [41]. Bioprinting is also an interesting technology which essentially uses the inkjet printing principle to pre- 5.1.1. Thermosensitive Hydrogels. Thermosensitive hydrogels cisely distribute biological material within a culture’s semi- are mostly based on poly(N-isopropylacrylamide) (pNI- solid substrate [42]. Finally, an interesting approach that PAAm). Upon cooling from 37◦Cto32◦C, the polymer takes advantage of the temperature-dependent hydrophobic/ switches from a hydrophobic to a hydrophilic state. The pre- hydrophilic properties of the culture dishes consists of creat- vious state allows for cell attachment, whereas the latter one ing monolayered cell sheets to be implanted directly at the causes cell sheet release from the substrate. For a detailed des- surface of the heart [43]. Amazingly, this can be repeated so cription of the mechanism, see Graziano as well as Baysal that several sheets may be stacked on top of each other in and Karasz [56, 57]. Shimizu et al. [58]culturedneonatalrat 6 Cardiology Research and Practice

(I) (generally an ECM protein) and a component that controls ΔT, changes in (non)covalent interactions (a synthetic poly- mer) that then cause macroscopic transitions. Excellent articles by Lutolf et al. [55]andUlijn[50] describe the under- lying principles and mechanisms. Most MMP-sensitive hy- (II) drogels provide sites for disease-specific enzymes that de- grade the hydrogel, allowing either cell invasion or drug release and represent the most prominent candidates for ap- plication in tissue engineering. Indicating the importance of (III) degradable hydrogels, Shapira et al. [61]conductedastudy EDC/NHS of neonatal rat cardiomyocytes on a MMP-sensitive PEGy- lated fibrinogen hydrogel. A different cell morphology was induced in MMP-2- and MMP-9-deficient cell cultures com- (IVa) pared to control cultures with MMP. Other experimental studies focused on the functionalization of hydrogels with cell adhesion rather than MMP-sensitive motives. Yu and co- (IVb) workers[62] designed an RGD-modified alginate hydrogel and could show increased proliferation of human umbilical vein cord endothelial cells (HUVECs) and increased angio- genesis in a rat infarct model. Combining cell adhesion mo- Matrix metalloproteinase (MMP) tives, enzyme-sensitive scaffolds and drug release in one con- Biologically active groups, exposed upon struct, Phelps et al. [63] engineered PEG-based bioartifi- conformational change cial hydrogel matrices presenting MMP-degradable sites as Drug growth factor release system, RGD peptide as cell adhesion ECM proteins/growth factors motifs, and VEGF to induce the growth of vasculature in vivo. They reported that implantation of their construct in- EDC/NHS coupling agents, 1-ethyl-3-(3-dimethylaminopropyl) duced the growth of new vessels into the matrix in vivo and carbodiimide and N-hydroxysuccinimide resulted in significantly increased rate of reperfusion in a rat Figure 2: Schematic illustration of different functionalization prin- limb ischemic model. ciples. (I) Smart materials, changing conformation, and exposing different chemical groups upon temperature change or a change 5.2. Surface-Modified Materials via Protein Adsorbance. in enzyme concentration. (II) Surface functionalized materials. A Surface-modified materials are among the most frequently ff synthetic sca old is immersed in a ECM protein solution, allowing functionalized materials. Synthetic, biologically inert poly- for protein adsortion on the surface. (III) Covalently functionaliz- mer scaffolds are rendered bioactive by a coating of naturally ed scaffolds. Functional proteins are coupled to the surface via ff occurring ECM proteins. The manifold techniques and ap- EDC/NHS chemistry. (IV) Blend materials, (a) hybrid sca olds of ff various polymers or (b) hybrid scaffolds of polymers and drugs for proaches of hybrid sca olds of naturally occurring and syn- controlled release. thetic polymers are summarized in reviews by Furth et al. and Rosso et al. and particularly focused on cardiac tissue engineering, in Chan et al. [64–66]. Interestingly, in a com- parative study of fibronectin-, collagen-, or laminin- cardiomyocytes on temperature-sensitive pNI-PAAm-coated coated elastomer poly(1,8-octanediol-co-citric acid) (POC), dishes. Cell sheets were detached and overlaid to construct a Hidalgo-Bastida et al. [67] could demonstrate most pro- four-layered cardiac graft. Studies of subcutaneous implan- moted cell adhesion of HL-1 mouse cardiac muscle cells on tation in rats showed constant beating and vascularization fibronectin-coated substrates. In a completely different study, of the construct. Following the same principle, Kubo et al. C2C12 mouse myoblasts were shown to align and differen- [59] cultured neonatal rat cardiomyocytes on pNI-PPAm to tiate into myotubes on collagen-coated electrospun scaffolds design myocardial tubes of wrapped cell sheets. In a com- of DegraPol [68]. Following the concept of contact guidance, bined study of thermo-sensitive and micropatterned pNI- McDevitt et al. [69] constructed micropatterned laminin PAAm-ECM films, the creation of multilayered oriented con- lanes on poly(dimethylsiloxane) (PDMS) substrates to pro- structs of cardiomyocytes and C2C12 myoblasts was shown mote cell alignment of cardiomyocytes. Several years later, [60]. These studies demonstrated promising results for the Cimetta et al. [70]producedcontractilecardiacmyograftson in vitro preconditioning of cell cultures and, subsequently, laminin-coated (microprinted) poly(acrylamide) hydrogels. scaffold-free implantation. This concept is one of the first to Easy setup and high reproducibility made the setup a poten- reach clinical trials. tial candidate for application in high-throughput biological and physiological studies. The straightforward method of 5.1.2. Enzyme-Sensitive Hydrogels. Enzyme or more specifi- protein adsorbance, however, carries several drawbacks cally, matrix metalloproteinase- (MMP-) sensitive hydrogels such as uncontrolled release and unknown conformation of always consist of two parts: a MMP-sensitive component the protein on the surface. Furthermore, adsorbed protein Cardiology Research and Practice 7 concentration can only be inaccurately controlled. Alterna- we, however, aim for immobilized factors, stimulating the tively, the concept of covalently linked proteins arose. regeneration of ischemic tissue over a longer period of time. In 2003, a clinical study on the safety and efficacy of 5.3. Surface-Modified Materials via Covalent Immobilization. intracoronary and intravenous infusion of rhVEGF was con- Growth factors play an essential role in tissue engineering, ducted [79]. In the so-called VIVA trial (vascular endothelial and ideally, they would not be supplied in a soluble, quickly growth factor in ischemia for vascular angiogenesis), 178 degradable form but be active at the site of interest, that is, patients with stable exertional angina were randomized to at the scaffold surface and in the targeted host tissue. New- receive placebo, low-dose (17 ng kg−1 min−1), and high-dose ly developed materials influence the neovascularization pro- (50 ng) rhVEGF by intracoronary infusion, followed by in- cesses in ischemic tissue since they allow the delivery of vas- travenous infusion on days 3, 6, and 9. VEGF was safe and cular endothelial growth factor (VEGF) and basic fibroblast well tolerated; high-dose VEGF resulted in significant im- growth factor (FGF), two main factors that control neo- provement in angina after 120 days. Despite promising re- angiogenesis. Immobilized VEGF has been confirmed [47, sults, this was only a small trial with a short-term followup. 48, 71] to induce extended signaling and enhanced biologi- cal activity compared to soluble VEGF, as shown by in- 5.4. Blend Materials and Drug Release. Astraightforward creased proliferation of endothelial cells (ECs). Shen et al. technique that potentially involves all aforementioned con- [72] furthermore demonstrated increased cell infiltration of cepts of functionalization constitutes the production of endothelial cells into a VEGF-functionalized scaffold, com- material blends. Following the same rationale of optimizing pared to cell cultures where VEGF is supplied in soluble the bioactivity of scaffolds, synthetic materials can be blend- form in the medium. In an advanced study, Chiu et al. [73] ed with naturally occurring extracellular matrix (ECM) pro- covalently immobilized VEGF and angiopoietin-1 (Ang-1) teins, growth factors, or simply other synthetic materials to on a porous collagen scaffold, resulting in enhanced vascu- alter mechanical properties. In a straightforward setup, Choi larization in a CAM assay and improved tube formation by et al. [80] produced an electrospun substrate of aligned poly- endothelial cells. Furthermore, gelatin has been grafted to caprolactone/collagen fibres and could induce skeletal mus- air plasma-activated PCL scaffolds via coupling agents such cle differentiation and myotube formation thereon. Using as 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide and N- a similar scaffold, Tillman et al. [81] confirmed the poten- Hydroxysuccinimid (EDC/NHS coupling chemistry). The tial of polycaprolactone/collagen scaffold for in vitro cell cul- modified substrate enhanced spreading and proliferation of ture of endothelial progenitor cells; furthermore, the con- endothelial cells. Additionally, endothelial cells followed the struct was shown to retain its structural integrity over one fibre orientation of gelatin-coated scaffolds as compared to month in a rabbit aortoiliac bypass model. The endothe- pure PCL scaffolds where a random cell orientation was lialized substrates resisted blood platelet adherence in the found [74]. Although used in many approaches, immobiliz- animal model. ing proteins or growth factors, via EDC/NHS chemistry, Synthetic or natural polymers can, however, not only be promote several issues. EDC/NHS coupling generates highly blended among each other but also drugs serve as essen- heterogeneous structures, regarding function and orienta- tial supplements in material design for biomedical engineer- tion of the immobilized proteins. Backer et al. [47]develop- ing. Kraehenbuehl et al. [82] developed a matrix metallo- ed a new, site-specific covalent immobilization approach. proteinase- (MMP-) degradable polyethylene glycol (PEG) A genetically induced N-terminal Cys-tag of VEGF cou- construct with incorporated thymosin β4. Entrapped Tβ4 pled the growth factor to fibronectin. Controlled orienta- promoted enhanced endothelial cell survival, cadherine, tion was achieved. VEGF receptors were stimulated by im- and angiopoietin-2 expression and increased MMP-2 and mobilized VEGF and are fully capable of signal transduction MMP-9 production. Metalloproteinase production directly pathways. Rather simple chemistry confirmed biological stimulated the hydrogel degradation and Tβ4 release, pro- activity, increased endothelial cell proliferation, and re- viding a controlled drug release system. ported that vascularization in a CAM model render VEGF- Interestingly, Thakur et al. [83] combined two different functionalized scaffolds a promising substrate for in vivo drugs in a poly(L-lactic acid) (PLLA) electrospun scaffold. vascularization of ischemic myocardium. However, in vivo lidocaine and mupirocin showed different release kinetics studies of functionalized substrates are still in their infancy. when incorporated into the fibrous mesh. The hybrid scaf- Experimental studies by Banfi and coworkers [75–77]em- fold can be employed for wound dressing, where a fast release phasized the critical role of microenvironmental VEGF con- of Lidocaine promotes immediate pain relief, whereas a centration. Timing of the expression, concentration gradient, sustained release of Mupirocin provides a constant antibiotic and interactions with cells are all critical issues that need function until wound healing. The dual-release profile con- to be taken into account when designing VEGF scaffolds. cept can find wide application in tissue engineering, enabling A critical threshold defines both normal and aberrant angi- scientists to develop spatiotemporal controlled drug release. ogenesis. In summary, the spatiotemporal distribution of VEGF to stimulate the formation of stable new vessels in 5.5. Nanoparticles and Drug Release. Furthermore, nan- ascaffold must be addressed and investigated. Studies on otechnologies offer the possibility to design nanoparticles for VEGF release from alginate/chitosan hydrogel were perform- drug delivery with large possibilities for customization of the ed by De laRi Va et al. [78], indicating a first burst effect, fol- particles. In particular, functionalized surfaces allow target- lowed by constant release over 5 weeks. For cardiac implants, ing of the particle, and tailored solubility, size, and shape 8 Cardiology Research and Practice present advantages for drug encapsulation and optimized [4]J.M.CollinsandB.Russell,“Stemcelltherapyforcardiac biodistribution [84]. The nanosized particles are especially repair,” Journal of Cardiovascular Nursing,vol.24,no.2,pp. designed for the delivery of drugs with intracellular targets. 93–97, 2009. For instance, Dvir et al. recently targeted cardiac cells within [5] B. Trachtenberg, D. L. Velazquez, A. R. Williams et al., “Ra- the infarcted heart [85]. The authors designed nanoscaled tionale and design of the transendocardial injection of auto- liposomes, functionalised with an angiotensin II type I (AT1) logous human cells (bone marrow or mesenchymal) in chro- nic ischemic left ventricular dysfunction and heart failure sec- receptor-specific peptide sequence. Twenty-four hours after ondary to myocardial infarction (TAC-HFT) trial: a random- injection, the particles were found mainly accumulated in the ized, double-blind, placebo-controlled study of safety and left ventricle of infarcted mice hearts. efficacy,” American Heart Journal, vol. 161, no. 3, pp. 487–493, A controlled drug release at the site of interest may be 2011. controlled by enzyme-, pH or temperature-sensitive hydrogel [6] J. T. Willerson, E. C. Perin, S. G. Ellis et al., “Intramyocardial systems. Wang et al. [86] performed an intramyocardial injection of autologous bone marrow mononuclear cells for administration of bFGF-loaded temperature-sensitive chi- patients with chronic ischemic heart disease and left ventri- tosan hydrogel and reported on an attenuated remodeling, cular dysfunction (First Mononuclear Cells injected in the US reduced infarct size, and increased arteriole numbers. [FOCUS]): rationale and design,” American Heart Journal, vol. Finally, using a multiapproached experimental design, 160, no. 2, pp. 215–223, 2010. [7] A. Rosenzweig, “Cardiac cell therapy—mixed results from Ye et al. [87] injected skeletal myoblasts, transfected with a mixed cells,” The New England Journal of Medicine, vol. 355, hypoxia-regulated VEGF plasmid that was encapsulated in no. 12, pp. 1274–1277, 2006. polyethylenimine nanoparticles, into normal and infarcted [8] K. C. Wollert and H. Drexler, “Cell therapy for the treatment hearts. The transfected myoblasts showed an improved cell of coronary heart disease: a critical appraisal,” Nature Reviews survival and induced improved global LV function in a rabbit Cardiology, vol. 7, no. 4, pp. 204–215, 2010. model of myocardial infarction. [9]E.Martin-Rendon,S.Brunskill,C.Doree´ et al., “Stem cell treatment for acute myocardial infarction,” Cochrane Database 6. Conclusion of Systematic Reviews, no. 4, Article ID CD006536, 2008. [10] M. N. Giraud, E. Ayuni, S. Cook, M. Siepe, T. P. Carrel, and Various strategies for cardiac diseases, including tissue engi- H. T. Tevaearai, “Hydrogel-based engineered skeletal muscle neering and stem-cell-based therapy, have been investigated grafts normalize heart function early after myocardial infarc- tion,” Artificial Organs, vol. 32, no. 9, pp. 692–700, 2008. in the past decade. Solving challenges that have arisen is a [11] M. N. Giraud, R. Flueckiger, S. Cook et al., “Long-term pressing objective for cardiac reparative medicine. Neverthe- evaluation of myoblast seeded patches implanted on infarcted less, we can realistically predict that future treatments will rat hearts,” Artificial Organs, vol. 34, no. 6, pp. E184–E192, include cell-based therapies. However, so far only short-to 2010. mid-term results have been provided. The long-term evalu- [12] M. Siepe, M. N. Giraud, M. Pavlovic et al., “Myoblast-seed- ation of possible heart recoveries remains to be confirmed, ed biodegradable scaffolds to prevent post-myocardial infarc- in particular for engineered tissue using rapidly degradable tion evolution toward heart failure,” Journal of Thoracic and scaffolds. Controls of cell matrix interaction, dose, and time Cardiovascular Surgery, vol. 132, no. 1, pp. 124–131, 2006. delivery will represent major breakthroughs for refining [13] J. C. Chachques, J. C. Trainini, N. Lago, M. Cortes-Morichetti, treatment and will govern successful clinical applications. O. Schussler, and A. Carpentier, “Myocardial assistance by grafting a new bioartificial upgraded myocardium (MAG- NUM trial): clinical feasibility study,” The Annals of Thoracic Conflict of interests Surgery, vol. 85, no. 3, pp. 901–908, 2008. [14] T. Shimizu, H. Sekine, M. Yamato, and T. Okano, “Cell sheet- None of the authors have conflict of interest to disclose. based myocardial tissue engineering: new hope for damaged heart rescue,” Current Pharmaceutical Design, vol. 15, no. 24, Acknowledgments pp. 2807–2814, 2009. [15] T. I. G. van der Spoel, S. J. Jansen of Lorkeers, P.Agostoni et al., The authors’ work was financially supported by the Swiss “Human relevance of pre-clinical studies in stem cell therapy: National Science Foundation (SNF Grant no. 122334) and systematic review and meta-analysis of large animal models of the Swiss Heart Foundation. Theye are grateful to Laura ischaemic heart disease,” Cardiovascular Research. In press. [16] W. H. Zimmermann, I. Melnychenko, G. Wasmeier et al., Seidel for proof reading and administrative assistance. “Engineered heart tissue grafts improve systolic and diastolic function in infarcted rat hearts,” Nature Medicine, vol. 12, no. References 4, pp. 452–458, 2006. [17] M. Jaconi and P. Zammaretti-Schaer, “3D-cardiac tissue engi- [1] S. Bollini, N. Smart, and P. R. Riley, “Resident cardiac pro- neering for the cell therapy of heart failure,” Tech. Rep. US genitor cells: at the heart of regeneration,” Journal of Molecular 2008/0226726 A1, 2008. and Cellular Cardiology, vol. 50, no. 2, pp. 296–303, 2011. [18] J. V. Terrovitis, R. R. Smith, and E. Marban,´ “Assessment and [2] O. Bergmann, R. D. Bhardwaj, S. Bernard et al., “Evidence for optimization of cell engraftment after transplantation into the cardiomyocyte renewal in humans,” Science, vol. 324, no. 5923, heart,” Circulation Research, vol. 106, no. 3, pp. 479–494, 2010. pp. 98–102, 2009. [19] H. Skalnikova, J. Motlik, S. J. Gadher, and H. Kovarova, “Map- [3] J. M. Collins and B. Russell, “Stem cell therapy for cardiac ping of the secretome of primary isolates of mammalian cells, repair,” Journal of Cardiovascular Nursing,vol.24,no.2,pp. stem cells and derived cell lines,” Proteomics,vol.11,no.4,pp. 93–97, 2009. 691–708, 2011. Cardiology Research and Practice 9

[20] M. N. Giraud, C. Armbruster, T. Carrel, and H. T. Tevaearai, [37] M. Siepe, M. N. Giraud, E. Liljensten et al., “Construction of “Current state of the art in myocardial tissue engineering,” Tis- skeletal myoblast-based polyurethane scaffolds for myocardial sue Engineering, vol. 13, no. 8, pp. 1825–1836, 2007. repair,” Artificial Organs, vol. 31, no. 6, pp. 425–433, 2007. [21] W. H. Zimmermann, M. Didie,´ S. Doker¨ et al., “Heart muscle [38] A. G. Guex, F. M. Kocher, G. Fortunato et al., “Fine-tuning of engineering: an update on cardiac muscle replacement the- substrate architecture and surface chemistry promotes muscle rapy,” Cardiovascular Research, vol. 71, no. 3, pp. 419–429, tissue development,” Acta Biomaterialia,vol.8,no.4,pp. 2006. 1481–1489, 2012. [22] J. Leor, Y. Amsalem, and S. Cohen, “Cells, scaffolds, and mole- [39] A. A. Rane and K. L. Christman, “Biomaterials for the treat- cules for myocardial tissue engineering,” Pharmacology and ment of myocardial infarction a 5-year update,” Journal of the Therapeutics, vol. 105, no. 2, pp. 151–163, 2005. American College of Cardiology, vol. 58, no. 25, pp. 2615–2629, [23] M. A. Laflamme and C. E. Murry, “Heart regeneration,” Na- 2011. [40] J. M. Kelm, V. Djonov, L. M. Ittner et al., “Design of custom- ture, vol. 473, no. 7347, pp. 326–335, 2011. shaped vascularized tissues using microtissue spheroids as [24]Y.Zhou,S.Wang,Z.Yu,R.F.HoytJr.,X.Qu,andK.A. ff minimal building units,” Tissue Engineering,vol.12,no.8,pp. Horvath, “Marrow stromal cells di erentiate into vasculature 2151–2160, 2006. after allogeneic transplantation into ischemic myocardium,” [41] H. C. Ott, T. S. Matthiesen, S. K. Goh et al., “Perfusion-de- The Annals of Thoracic Surgery, vol. 91, no. 4, pp. 1206–1212, cellularized matrix: using nature’s platform to engineer a bio- 2011. artificial heart,” Nature Medicine, vol. 14, no. 2, pp. 213–221, [25] N. Kawaguchi, A. J. Smith, C. D. Waring et al., “c-kitpos 2008. GATA-4 high rat cardiac stem cells foster adult cardiomyocyte [42] V. Mironov, “Organ printing,” Tissue Engineering A, vol. 17, p. survival through IGF-1 paracrine signalling,” PLoS One, vol. 5, 537, 2011. no. 12, Article ID e14297, 2010. [43] T. Hoashi, G. Matsumiya, S. Miyagawa et al., “Skeletal myo- [26] L. Li, S. Zhang, Y. Zhang, B. Yu, Y. Xu, and Z. Guan, “Paracrine blast sheet transplantation improves the diastolic function of action mediate the antifibrotic effect of transplanted mes- a pressure-overloaded right heart,” Journal of Thoracic and enchymal stem cells in a rat model of global heart failure,” Cardiovascular Surgery, vol. 138, no. 2, pp. 460–467, 2009. Molecular Biology Reports, vol. 36, no. 4, pp. 725–731, 2009. [44] N. Asakawa, T. Shimizu, Y. Tsuda et al., “Pre-vascularization [27] G. U. Premaratne, L.-P Ma, M. Fujita, X. Lin, E. Bollano, and of in vitro three-dimensional tissues created by cell sheet engi- M. Fu, “Stromal vascular fraction transplantation as an alter- neering,” Biomaterials, vol. 31, no. 14, pp. 3903–3909, 2010. native therapy for ischemic heart failure: anti-inflammatory [45] Y. Ito, “Covalently immobilized biosignal molecule materials role,” Journal of Cardiothoracic Surgery, vol. 6, p. 43, 2011. for tissue engineering,” Soft Matter, vol. 4, no. 1, pp. 46–56, [28] V. F. M. Segers and R. T. Lee, “Stem-cell therapy for cardiac 2008. [46] I. T. Hwang, C. H. Jung, D. K. Kim, Y. C. Nho, and J. H. Choi, disease,” Nature, vol. 451, no. 7181, pp. 937–942, 2008. “Patterning of biomolecules on a poly(epsilon-caprolactone) [29] X. Hu, J. Wang, J. Chen et al., “Optimal temporal delivery of film surface functionalized by ion implantation,” Colloids and bone marrow mesenchymal stem cells in rats with myocardial Surfaces B, vol. 74, no. 1, pp. 375–379, 2009. infarction,” European Journal of Cardio-thoracic Surgery, vol. [47] M. V. Backer, V. Patel, B. T. Jehning, K. P. Claffey, and J. M. 31, no. 3, pp. 438–443, 2007. Backer, “Surface immobilization of active vascular endothelial [30] D. Hou, E. A. S. Youssef, T. J. Brinton et al., “Radiolabeled cell growth factor via a cysteine-containing tag,” Biomaterials, vol. distribution after intramyocardial, intracoronary, and inter- 27, no. 31, pp. 5452–5458, 2006. stitial retrograde coronary venous delivery: Implications for [48] J. L. Sharon and D. A. Puleo, “Immobilization of glycopro- current clinical trials,” Circulation, vol. 112, no. 9, pp. I150– teins, such as VEGF, on biodegradable substrates,” Acta Bio- I156, 2005. materialia, vol. 4, no. 4, pp. 1016–1023, 2008. [31] M. Bonios, J. Terrovitis, C. Y. Chang et al., “Myocardial sub- [49] O. Wichterle and D. L´ım, “Hydrophilic gels for biological use,” strate and route of administration determine acute cardiac re- Nature, vol. 185, no. 4706, pp. 117–118, 1960. tention and lung bio-distribution of cardiosphere-derived [50] R. V.Ulijn, “Enzyme-responsive materials: a new class of smart cells,” Journal of Nuclear Cardiology, vol. 18, no. 3, pp. 443– biomaterials,” Journal of Materials Chemistry, vol. 16, no. 23, 450, 2011. pp. 2217–2225, 2006. ff [32] N. Dib, H. Khawaja, S. Varner, M. McCarthy, and A. Campbell, [51] L. Moroni and J. H. Elissee , “Biomaterials engineered for “Cell therapy for cardiovascular disease: a comparison of integration,” Materials Today, vol. 11, no. 5, pp. 44–51, 2008. methods of delivery,” Journal of Cardiovascular Translational [52] M. Caldorera-Moore and N. A. Peppas, “Micro- and nan- Research, vol. 4, no. 2, pp. 177–181, 2010. otechnologies for intelligent and responsive biomaterial-based medical systems,” Advanced Drug Delivery Reviews, vol. 61, no. [33] W. H. Zimmermann and R. Cesnjevar, “Cardiac tissue engi- 15, pp. 1391–1401, 2009. neering: implications for pediatric heart surgery,” Pediatric [53] R. M. P. da Silva, J. F. Mano, and R. L. Reis, “Smart thermores- Cardiology, vol. 30, no. 5, pp. 716–723, 2009. ponsive coatings and surfaces for tissue engineering: switching [34] P. Akhyari, H. Kamiya, A. Haverich, M. Karck, and A. cell-material boundaries,” Trends in Biotechnology, vol. 25, no. Lichtenberg, “Myocardial tissue engineering: the extracellular 12, pp. 577–583, 2007. matrix,” European Journal of Cardio-thoracic Surgery, vol. 34, [54] J. F. Mano, “Stimuli-responsive polymeric systems for biomed- no. 2, pp. 229–241, 2008. ical applications,” Advanced Engineering Materials, vol. 10, no. [35] M. Siepe, P. Akhyari, A. Lichtenberg, C. Schlensak, and F. 6, pp. 515–527, 2008. Beyersdorf, “Stem cells used for cardiovascular tissue engi- [55] M. P. Lutolf, J. L. Lauer-Fields, H. G. Schmoekel et al., “Syn- neering,” European Journal of Cardio-thoracic Surgery, vol. 34, thetic matrix metalloproteinase-sensitive hydrogels for the no. 2, pp. 242–247, 2008. conduction of tissue regeneration: engineering cell-invasion [36] L. Hecker and R. K. Birla, “Engineering the heart piece by characteristics,” Proceedings of the National Academy of piece: state of the art in cardiac tissue engineering,” Regen- Sciences of the United States of America, vol. 100, no. 9, pp. erative Medicine, vol. 2, no. 2, pp. 125–144, 2007. 5413–5418, 2003. 10 Cardiology Research and Practice

[56] G. Graziano, “On the temperature-induced coil to globule motes penetration and proliferation of endothelial cells,” Acta transition of poly-N-isopropylacrylamide in dilute aqueous Biomaterialia, vol. 4, no. 3, pp. 477–489, 2008. solutions,” International Journal of Biological Macromolecules, [73] L. L. Y. Chiu and M. Radisic, “Scaffolds with covalently im- vol. 27, no. 1, pp. 89–97, 2000. mobilized VEGF and angiopoietin-1 for vascularization of [57] B. M. Baysal and F. E. Karasz, “Coil-globule collapse in flexible engineered tissues,” Biomaterials, vol. 31, no. 2, pp. 226–241, macromolecules,” Macromolecular Theory and Simulations, 2010. vol. 12, no. 9, pp. 627–646, 2003. [74] Z. Ma, W. He, T. Yong, and S. Ramakrishna, “Grafting of [58] T. Shimizu, M. Yamato, Y. Isoi et al., “Fabrication of pulsatile gelatin on electrospun poly(caprolactone) nanofibers to im- cardiac tissue grafts using a novel 3-dimensional cell sheet prove endothelial cell spreading and proliferation and to con- manipulation technique and temperature-responsive cell cul- trol cell orientation,” Tissue Engineering,vol.11,no.7-8,pp. ture surfaces,” Circulation research, vol. 90, no. 3, pp. E40–E48, 1149–1158, 2005. 2002. [75] C. R. Ozawa, A. Banfi, N. L. Glazer et al., “Microenvironmen- [59] H. Kubo, T. Shimizu, M. Yamato, T. Fujimoto, and T. Okano, tal VEGF concentration, not total dose, determines a threshold “Creation of myocardial tubes using cardiomyocyte sheets and between normal and aberrant angiogenesis,” Journal of Clinical an in vitro cell sheet-wrapping device,” Biomaterials, vol. 28, Investigation, vol. 113, no. 4, pp. 516–527, 2004. no. 24, pp. 3508–3516, 2007. [76] A. Banfi, G. von Degenfeld, and H. M. Blau, “Critical [60] H. Fujita, K. Shimizu, and E. Nagamori, “Application of a cell role of microenvironmental factors in angiogenesis,” Current sheet-polymer film complex with temperature sensitivity for Atherosclerosis Reports, vol. 7, no. 3, pp. 227–234, 2005. increased mechanical strength and cell alignment capability,” [77] G. von Degenfeld, A. Banfi, M. L. Springer et al., “Microen- Biotechnology and Bioengineering, vol. 103, no. 2, pp. 370–377, vironmental VEGF distribution is critical for stable and func- 2009. tional vessel growth in ischemia.,” The FASEB journal, vol. 20, [61] K. Shapira, D. Dikovsky, M. Habib, L. Gepstein, and D. no. 14, pp. 2657–2659, 2006. Seliktar, “Hydrogels for cardiac tissue regeneration,” Bio-Med- [78] B. De la Riva, C. Nowak, E. Sanchez´ et al., “VEGF-controlled ical Materials and Engineering, vol. 18, no. 4-5, pp. 309–314, release within a bone defect from alginate/chitosan/PLA-H 2008. scaffolds,” European Journal of Pharmaceutics and Biopharma- [62] J. Yu, K. T. Du, Q. Fang et al., “The use of human mesen- ceutics, vol. 73, no. 1, pp. 50–58, 2009. chymal stem cells encapsulated in RGD modified alginate [79] T. D. Henry, B. H. Annex, G. R. McKendall et al., “The microspheres in the repair of myocardial infarction in the rat,” VIVA trial: vascular endothelial growth factor in ischemia for Biomaterials, vol. 31, no. 27, pp. 7012–7020, 2010. vascular angiogenesis,” Circulation, vol. 107, no. 10, pp. 1359– [63]E.A.Phelps,N.Landazuri,P.M.Thule,W.R.Taylor,andA.J. 1365, 2003. Garcia, “Bioartificial matrices for therapeutic vascularization,” [80]J.S.Choi,S.J.Lee,G.J.Christ,A.Atala,andJ.J.Yoo, Proceedings of the National Academy of Sciences of the United “The influence of electrospun aligned poly(epsilon-caprolac- States of America, vol. 107, no. 8, pp. 3323–3328, 2010. tone)/collagen nanofiber meshes on the formation of self- [64] M. E. Furth, A. Atala, and M. E. van Dyke, “Smart biomaterials aligned skeletal muscle myotubes,” Biomaterials, vol. 29, no. design for tissue engineering and regenerative medicine,” 19, pp. 2899–2906, 2008. Biomaterials, vol. 28, no. 34, pp. 5068–5073, 2007. [81] B. W. Tillman, S. K. Yazdani, S. J. Lee, R. L. Geary, A. Atala, [65] F. Rosso, G. Marino, A. Giordano, M. Barbarisi, D. Parmeg- andJ.J.Yoo,“Thein vivo stability of electrospun polycapro- giani, and A. Barbarisi, “Smart materials as scaffolds for tissue lactone-collagen scaffolds in vascular reconstruction,” Bioma- engineering,” Journal of Cellular Physiology, vol. 203, no. 3, pp. terials, vol. 30, no. 4, pp. 583–588, 2009. 465–470, 2005. [82] T. P.Kraehenbuehl, L. S. Ferreira, P.Zammaretti, J. A. Hubbell, [66] Q. Z. Chen, S. E. Harding, N. N. Ali, A. R. Lyon, and A. R. and R. Langer, “Cell-responsive hydrogel for encapsulation of Boccaccini, “Biomaterials in cardiac tissue engineering: ten vascular cells,” Biomaterials, vol. 30, no. 26, pp. 4318–4324, years of research survey,” Materials Science and Engineering R, 2009. vol. 59, no. 1–6, pp. 1–37, 2008. [83] R. A. Thakur, C. A. Florek, J. Kohn, and B. B. Michniak, “Elec- [67] L. A. Hidalgo-Bastida, J. J. A. Barry, N. M. Everitt et al., “Cell trospun nanofibrous polymeric scaffold with targeted drug re- adhesion and mechanical properties of a flexible scaffold for lease profiles for potential application as wound dressing,” In- cardiac tissue engineering,” Acta Biomaterialia, vol. 3, no. 4, ternational Journal of Pharmaceutics, vol. 364, no. 1, pp. 87–93, pp. 457–462, 2007. 2008. [68] S. A. Riboldi, N. Sadr, L. Pigini et al., “Skeletal myogenesis on [84] R. Singh and J. W. Lillard, “Nanoparticle-based targeted drug highly orientated microfibrous polyesterurethane scaffolds,” delivery,” Experimental and Molecular Pathology, vol. 86, no. 3, Journal of Biomedical Materials Research A,vol.84,no.4,pp. pp. 215–223, 2009. 1094–1101, 2008. [85] T. Dvir, M. Bauer, A. Schroeder et al., “Nanoparticles targeting [69]T.C.McDevitt,J.C.Angello,M.L.Whitneyetal.,“Invitro the infarcted heart,” Nano Letters, vol. 11, no. 10, pp. 4411– generation of differentiated cardiac myofibers on micropat- 4414, 2011. terned laminin surfaces,” Journal of Biomedical Materials Re- [86] H. Wang, X. Zhang, Y. Li et al., “Improved myocardial per- search, vol. 60, no. 3, pp. 472–479, 2002. formance in infarcted rat heart by co-injection of basic fibro- [70] E. Cimetta, S. Pizzato, S. Bollini, E. Serena, P.De Coppi, and N. blast growth factor with temperature-responsive chitosan hy- Elvassore, “Production of arrays of cardiac and skeletal muscle drogel,” The Journal of Heart and Lung Transplantation, vol. 29, myofibers by micropatterning techniques on a soft substrate,” no. 8, pp. 881–887, 2010. Biomedical Microdevices, vol. 11, no. 2, pp. 389–400, 2009. [87] L. Ye, W. Zhang, L. P. Su et al., “Nanoparticle based delivery [71]M.Ehrbar,S.C.Rizzi,R.Hlushchuketal.,“Enzymaticforma- of hypoxia-regulated VEGF transgene system combined with tion of modular cell-instructive fibrin analogs for tissue engi- myoblast engraftment for myocardial repair,” Biomaterials, neering,” Biomaterials, vol. 28, no. 26, pp. 3856–3866, 2007. vol. 32, no. 9, pp. 2424–2431, 2011. [72] Y. H. Shen, M. S. Shoichet, and M. Radisic, “Vascular endo- thelial growth factor immobilized in collagen scaffold pro- Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 181394, 11 pages doi:10.1155/2012/181394

Review Article The Emerging Role of TLR and Innate Immunity in Cardiovascular Disease

Rolf Spirig,1 Janice Tsui,2 and Sidney Shaw3

1 CSL Behring AG, 3000 Bern, Switzerland 2 Division of Surgery & Interventional Science, University College London, Royal Free Campus, London NW3 2QG, UK 3 Department of Clinical Experimental Research, University of Bern, 3010 Bern, Switzerland

Correspondence should be addressed to Sidney Shaw, [email protected]

Received 15 July 2011; Accepted 29 November 2011

Academic Editor: Daryll M. Baker

Copyright © 2012 Rolf Spirig et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Cardiovascular disease is a complex disorder involving multiple pathophysiological processes, several of which involve activation of toll-like receptors (TLRs) of the innate immune system. As sentinels of innate immunity TLRs are nonclonally germline- encoded molecular pattern recognition receptors that recognize exogenous as well as tissue-derived molecular dangers signals promoting inflammation. In addition to their expression in immune cells, TLRs are found in other tissues and cell types including cardiomyocytes, endothelial and vascular smooth muscle cells. TLRs are differentially regulated in various cell types by several cardiovascular risk factors such as hypercholesterolemia, hyperlipidemia, and hyperglycemia and may represent a key mechanism linking chronic inflammation, cardiovascular disease progression, and activation of the immune system. Modulation of TLR signaling by specific TLR agonists or antagonists, alone or in combination, may be a useful therapeutic approach to treat various cardiovascular inflammatory conditions such as atherosclerosis, peripheral arterial disease, secondary microvascular complications of diabetes, autoimmune disease, and ischemia reperfusion injury. In this paper we discuss recent developments and current evidence for the role of TLR in cardiovascular disease as well as the therapeutic potential of various compounds on inhibition of TLR-mediated inflammatory responses.

1. Introduction: Innate Immunity and Mice do not express TLR10 but do express TLR11, TLR12, Toll-Like Receptors (TLRs) and TLR13 [4]. TLR1, TLR2, TLR4, TLR5, TLR6, and TLR11 are displayed on the cell surface while TLR3, TLR7, TLR8, Historically the immune system has been divided into and TLR9 are localized intracellularly. TLRs are distributed the innate and the adaptive immune system. Neutrophils, and differentially expressed in several cell types and tissues. eosinophils, basophils, mast cells, monocytes, macrophages, They are present on polymorphonuclear cells, macrophages, dendritic cells (DCs), NK cells, NK-T cells, γδ T cells, and mast cells, DC, NK cells, T cells, and B cells. Interestingly, B-1 cells are considered to be cellular members of the innate TLR expression has also been detected on cardiac, epithelial, immune system which can be activated by signaling through endothelial, and vascular smooth muscle cells. Moreover, TLR. In addition, endothelial cells may form part of this mesenchymal and parenchymal cells of different organs and system since they also possess antigen-presenting capabilities tissue such as kidney, heart, lung, liver, skin, brain and and therefore immune regulation properties apart from their intestine express TLR, but their functional role and relevance function as a barrier between tissue and blood [1]. is not yet fully understood [5]. A year after the discovery of the role of drosophila Toll The molecular weight of TLR ranges between 90 and protein in the host defense against fungal infection [2], a 115 kDa. The extracellular region of Toll contains leucine- mammalian homologue was identified, referred to as TLR4 rich repeat (LRR) motifs whereas the cytoplasmic domain [3]. Since then, 13 members of the TLR family have been has similarities with that of the mammalian Interleukin- identified in mammals, ten in humans, and twelve in mice. 1 receptor (IL-1R) family and is designated as Toll/IL-1R 2 Cardiology Research and Practice

(TIR) homology domain, containing around 200 amino foam cell formation is a hallmark of atherosclerosis. TLR4 acids. Within this domain, the regions of homology comprise has been shown to contribute to early-stage intimal foam cell three conserved boxes, which are crucial for signaling. accumulation at lesion-prone aortic sites in ApoE KO mice, After ligand binding, TLRs dimerize and undergo the as does TLR2 to a lesser extent [16]. Intimal smooth muscle conformational change required for recruitment of down- cells surround and penetrate early lesions, where TLR4 sig- stream signaling molecules. In general, these include the naling, enhanced by hypercholesterolemia, promotes lesion adaptor molecule, myeloid differentiation primary-response progression by stimulation of acyl-coenzyme A: cholesterol protein 88 (MyD88), TIR-domain-containing adaptor pro- acyltransferase-1 mRNA expression, cytoplasmic cholesterol tein (TIRAP; also known as MyD88-adaptor-like protein ester accumulation, and monocyte chemoattractant protein- or Mal), IL-1R-associated kinases (IRAKs), transforming 1 (MCP-1) mRNA and protein expression in a TLR4- growth factor-β-(TGF-β-) activated kinase (TAK1), TAK1- dependent manner. Other TLRs also appear to be involved. binding protein 1 (TAB1), TAB2, and tumor-necrosis factor Lipid accumulation in macrophages is closely linked to the (TNF) receptor-associated factor 6 (TRAF6) [6, 7]. The TLR PAT family of proteins (named after perilipin, adipophilin, family signaling pathway is highly homologous to that of the and TIP47 (tail-interacting protein of 47 kDa). TLR9- IL-1R family and represents the core pathway of all TLR, mediated signaling stimulates perilipin 3 expression and except for TLR3. Studies in 2001 revealed the existence of a macrophage accumulation of lipids, especially triglycerides. MyD88-independent pathway since stimulation of MyD88- Oligodeoxynucleotide (ODN) 1826, an agonist ligand of deficient DC with LPS still induced their maturation [8]. TLR9, significantly enhanced perilipin 3 expression in Therefore, exposure to LPS induces TLR4-signaling via a RAW264.7 cells via upregulation of IL-1α and IFNβ, whilst MyD88-dependent as well as MyD88-independent pathway, chloroquine, a TLR9 inhibitor, virtually completely abol- which subsequently activates IRF3 [9]. Up to now, TLR3- ished ODN1826-induced perilipin 3 expression. Inhibitors signaling is considered to be MyD88 independent [7]. of c-jun NH2-terminal kinase and PI3-kinase suppressed the Importantly, there seems to be a difference between TLR level of perilipin 3 mRNA induced by ODN1826 [17]. signaling induced by endogenous versus exogenous ligands Not all effects of TLR activation, however, may be which may in part be due to differential TLR signaling detrimental. Soluble forms of human TLR2 (sTLR2) have complex formation [10]. been shown to be released by monocytes, and depletion of sTLR2 resulted in an exaggerated inflammatory response [18]. Patients with postmyocardial infarction and heart 2. Toll-Like Receptors in Autoimmune and failure have also been shown to have markedly decreased Cardiovascular Disease sTLR2comparedtocontrols[19]. Recent studies indicate that neointima formation in a The concept that innate immune signaling can be triggered perivascular collar-induced injury model is reduced by sys- not only by external pathogens but also by endogenous temic administration of the dsRNA analog poly(I:C) (a TLR3 molecules released in response to tissue injury was first agonist) in a TLR3-dependent manner. Furthermore, genetic proposed almost two decades ago by Matzinger [11]. Subse- deletion of TLR3 markedly enhanced the development of quently, it was shown that antigen-presenting cells, namely, elastic lamina damage after collar-induced injury and accel- DC, can be activated by a variety of endogenous stimuli erated the onset of atherosclerosis in hypercholesterolemic [12]. Since then further advances in this area have been ApoE knockout mice [20]. Collectively, these data suggest surprisingly slow and only recently their emerging potential a protective role for TLR3 signaling in the vessel wall. role in cardiovascular disease has become to be recognized as Taken together, current data indicates that the contribution summarized in Figure 1 [13]. of TLR signaling to the progression of the atherosclerotic In this context, several endogenously derived molecules process may depend, at least in part, on a balance between released from necrotic cells, so-called damage-associated detrimental and protective TLR-mediated mechanisms. This molecular patterns (DAMPs) or alarmins, have been iden- in turn may depend not only on changes in the relative tified, which lead to “sterile inflammation” via activation expression of appropriate receptors on relevant cell types but of TLR. In addition, other molecules released from dying also on the relative availability of endogenous ligands. cells, for example, proteinases also lead to the generation of extracellular DAMP by degradation of components of the 2.2. Diabetes, Insulin Resistance, and Other Cardiovascular extracellular matrix or glycocalyx. DAMPs now encompass Risk Factor Interactions with TLR. Other cardiovascular risk a wide range of molecules including heat shock proteins, factors, notably diabetes, obesity, and insulin resistance, high-mobility group box 1 (HMGB1) protein, a chromatin- are also associated with a low-grade inflammatory state binding nuclear protein, ATP,uric acid, heparan sulfate (HS), that reflects activation of innate immunity associated with hyaluronan, and others [14, 15]. These molecules have been metabolic, environmental, and genetic factors. Evidence shown to bind to different TLRs or other molecular pattern suggests, for example, that resistin, originally described recognition receptors (PRRs) expressed on various cell types as an adipose tissue-specific hormone, is involved in and trigger the release of proinflammatory mediators. pathologic processes leading to CVD including inflam- mation, endothelial dysfunction, thrombosis, angiogenesis, 2.1. Atherosclerosis: Interaction of Risk Factors and TLR. and smooth muscle cell dysfunction. Recent data indicates Excessive accumulation of lipids in macrophages resulting in that a key mechanism underlying its detrimental effects Cardiology Research and Practice 3

Valvular disease Aortic interstitial cells express Thrombosis TLR2 and 4. Platelets express TLR1, 2, 4, and 6. Stimulation by TLR2 or 4 Stimulation by TLR4 induces stronger binding of agonists induces secretion of platelets to fibrinogen and seems to be correlated TNF, IL6, IL8, and MCP1 as with LPS-induced thrombocytopenia and neutrophil- well as increase of mRNA expression of dependent pulmonary sequestration. Activation via BMP2, Runx2, FGF, PDGFA MMP1 TLR2 induces upregulation of P-selectin, integrin αIIbβ3, ROS, and formation of neutrophil-platelets aggregates.

Aorta Left coronary artery Circumflex artery Right coronary Left artery anterior Endothelium descending artery Plaque Blood clot Artery wall

Diseased artery Congestive heart failure and remodeling Cardiomyocytes express TLR2, 3, 4, 5, 7, and 9 on the Atherosclerosis/CAD mRNA level. Stimulation by TLR2, 4, and 5 induces Upregulation of TLR2 expression. secretion of CXCL1, MIP2, and upregulation of ICAM-1. oxLDL acts as a TLR4 agonists on macrophages Stimulation by TLR4 on VSMCs results in increased In addition, stimulation by these receptors correlates with LDL and ECM deposition. decreased contractility. Stimulation by TLR9 induces IFNα enhances TLR4 signaling and thereby the secretion of TNFα , IL1β , IL6, and induction of iNOS. production of TNFα, IL12, and MMP9. Furthermore, TLR activation results in sarcomere shortening.

Figure 1: Summary of current data implicating TLR signaling in various cardiovascular disease processes (figure modified from [13]). BMP: bone morphogenetic protein; Runx2: Runt related transcription factor 2; FGF: fibroblast growth factor; PDGFA: platelet derived growth factor A; MMP: matrix metalloproteinase; ROS: reactive oxygen species; VSMC: vascular smooth muscle cells; ECM: extracellular matrix; LDL: low density lipoprotein; CXCL1: chemokine (C-X-C motif) ligand 1; MIP: macrophage-inflammatory protein; ICAM: intracellular adhesion molecule; iNOS: inducible nitric oxide synthase.

in these processes is that TLR4 serves as a receptor for [24]. This may contribute to the accentuated proinflam- the proinflammatory effects of resistin in human cells [21, matory state and complications of T1DM. Underlying 22]. This may partly explain the multifunctional role of molecular mechanisms linking these observations appear to resistin in chronic inflammation, atherosclerosis, and insulin involve complex-formation between advanced glycation end resistance. Similarly, nutritional fatty acids, whose circulating product-modified oxidized low-density lipoprotein (AGE- levels are often increased in obesity, activate TLR4 signaling LDL), the receptor for advanced glycation end products in adipocytes and macrophages and the capacity of dietary (RAGE), and the scavenger receptor CD36 [25]. Subsequent fatty acids to induce inflammatory signaling in adipose cells activation of downstream signaling pathways induced by or tissue and macrophages is blunted in the absence of binding of this complex to TLR4 results in activation of TLR4. p38-α, JNK, and ERK1 kinases and AP1, Elk1, and NFκB Other studies suggest an association between the transcription factors leading to increased production of Asp299Gly polymorphism of the TLR4 gene and early TNFα and proinflammatory cytokines. These mechanisms onset of diabetic retinopathy in type 2 diabetic patients may partly underlie the increased risk of atherosclerosis [23]. TLR2 and TLR4 expressions as well as signaling have observed in diabetics. Two common polymorphisms in also been shown to be enhanced in monocytes of patients TLR4, D299G and T399I, were shown in vitro to reduce with Type 1 diabetes with microvascular complications the response of TLR4 to LPS but had no effect on 4 Cardiology Research and Practice

4) Activation of innate immune cells via toll-like receptors e.g., Therapeutic intervention Granulocytes e.g. IVIg, C1-INH, LMW-DXS etc. Monocytes Dendritic cells NK cells etc.

I/R injury 3) Release of danger signals 1) Healthy endothelium 2) Complement Heparan sulfate Glyocalyx (endothelial surface) and coagulation activation LMW hyaluronan Extracellular matrix (subendothelial) HMGB1 HSP etc.

Figure 2: Cardiac I/R injury activates multiple inflammatory pathways such as complement, coagulation, and/or innate immune cells by binding of danger signals via expressed TLR. LMW-DXS: low molecular weight dextran sulfate, HMGB1: high-mobility group protein box 1, HSP: heat shock proteins. the AGE-LDL-complex signaling. This supports data from mice deficient in TLR4 signaling. Furthermore, these mice other studies suggesting that TLR activation by DAMP may had reduced intragraft mRNA levels of TNF-α,IL-1β,IL- activate alternative downstream proinflammatory pathways 6, EGR-1, ICAM-1, and iNOS [26]. In a mouse model of to those induced by pathogen-associated ligands. myocardial infarction, TLR4 deficiency resulted in less tissue damage and cardioprotection [28]. A cardioprotective effect has also been observed for mice deficient for TLR2, for 3. Toll-Like Receptors in Cardiac I/R Injury example, by a reduced infiltration of neutrophils into the tissue [27]. The exact role of TLR2 however is somewhat 3.1. Toll-Like Receptors as Sentinels of Innate Immunity controversial since, in other studies, TLR2 agonist ligands in Cardiac I/R Injury. There is an increasing number of were reported to induce cardioprotection, mediated via a studies demonstrating a major role of TLR in several animal TLR2/PI3K/Akt-dependent mechanism [29]. The reason for models of ischemia reperfusion (I/R) injury. Cardiac I/R these conflicting reports is currently unclear but may reflect injury has a significant clinical relevance as, for example, differences between chronic and acute models of I/R injury. in heart transplantation (HTx), myocardial infarction (MI), or coronary artery bypass graft surgery. Tissue damage and inflammation occurs after coronary artery occlusion 3.2. Interplay of TLR with Other Members of Innate Immunity (ischemia) when reperfusion occurs (restoration of blood in Cardiac I/R Injury. I/R injury leads to the activation of flow). A hallmark of I/R injury is a strong activation of multiple inflammatory pathways. Furthermore, there is an the innate immune system, that is, activation of com- active interplay between pathways such as TLR and com- plement and coagulation, recruitment of innate immune plement. The release of the nonmuscle myosin from dying cells, cytokine release, formation of reactive oxygen species cells is recognized by naturally occurring IgM antibodies, (ROS), mitochrondrial dysfunction, as well as apoptosis and resulting in complement activation and tissue damage [30]. cell necrosis (Figure 2). Studies with TLR deficient mice Interestingly, the anaphylatoxin and complement cleavage have demonstrated a crucial role of TLR2 and TLR4 in product C5a has been shown to negatively regulate produc- I/R injury-mediated inflammatory responses in the heart tion of IL-12 family members such as IL-12, IL-23, and IL- [26, 27]. Kaczorowski et al. showed in a murine cardiac 27 in inflammatory macrophages [31]. Furthermore, mice transplantation model that serum levels of TNF-α,IL-1β, deficient in the membrane complement regulator CD55 IL-6, troponin I, and MCP-1 were dramatically reduced in (DAF) have elevated levels of TNF-α,IL-1β, and IL-6 in Cardiology Research and Practice 5

Table 1: Non selective-TLR antagonists in cardiovascular disease.

In vitro (TLR inhibition) Compound In vivo (cardiac I/R injury, HTx) TLR/cells TLR agonists MI/pig [48] TLR2 and TLR4/human MoDC [33] TLR2: LTA, Pam3CSK4 LMW-DXS HTx/rat [49] TLR2/human NK cells [50] TLR4: LPS, HS Xeno HTx/hamster-to-rat [92] TLR4/human MoDC [64] TLR4: LPS IVIg HTx/human [63] TLR9/human B cells [66] TLR9: CpG Oligos MI/pig [68] TLR4/murine macrophage cell line C1-INH TLR4: LPS MI/cat [69] RAW264.7 [70] HTx/human [63] ATIII TLR4/human monocytic cell line THP1 [75] TLR4: LPS HTx/mice [76] α1AT TLR4/human monocytes [79] TLR4: LPS MI/mice [80] rHDL TLR4/human monocytes [87] TLR4: LPS MI/rat [86] TLR4/human monocytes [89] MI/human [93] Statins TLR4: LPS TLR4/human MoDC [90] HTx/human [88] HS: heparan sulfate; HTx: heart transplantation; MI: myocardial infarction; MoDC: monocyte-derived dendritic cells: LTA: lipoteichoic acid; LPS: lipopolysaccharide. response to TLR agonists, whereas the levels of IL-12p40 are sufficient. Hence, combination therapies may show greater slightly decreased [32]. efficacy. Interestingly, the endogenous TLR4 ligand heparan To date several compounds have been described, many sulfate (HS) has been shown to induce the production of them already routinely used in the clinics for other of complement proteins C1q or C3 by human monocyte- indications, which have been shown to inhibit TLR signaling derived DC (MoDC) [33] whereas exogenous LPS reduced in vitro and to possess therapeutic potential in animal models secretion of C1q [34]. In addition, exogenous C1q seems to of MI or HTx. Therefore, we will discuss in the following potentiate the LPS-induced secretion of IL-12p70 by human paragraphs compounds which have shown to have anti- MoDC in vitro [35]. Another study has reported decreased inflammatory activity on various pathways of cardiac I/R levels of IL-12p70 secretion by DC after LPS-challenge in injury such as the coagulation or complement cascade, patients deficient for C1q [36]. TLR signaling, leukocyte recruitment, NK cell activation, Another important characteristic of I/R injury is the maturation of DC, and others (summarized in Table 1). very low oxygen level during ischemia. Hypoxia induces the expression of the transcription factor hypoxia-inducible factor-1α (HIF-1α) which has been described as a key 3.3.1. Low Molecular Weight Dextran Sulfate (LMW-DXS). regulator of a broad range of cellular and systemic responses Low molecular weight dextran sulfate (LMW-DXS, MW: to hypoxic conditions. Recent studies suggest a cross-talk 5000 Dalton) has various anti-inflammatory properties. between HIF-1α and TLR. It has been shown that stimulation Inhibition of all three activation pathways of complement of human MoDC with LPS under hypoxic conditions is mediated by binding to factor H [44] and enhance- resulted in a significantly higher expression of the costimula- ment of the activity of C1-INH [45]. Furthermore, LMW- tory molecules CD80 and CD86 [37].Thesamesynergistic DXS inhibits coagulation by enhancing the anticoagulatory effect was observed for murine bone-marrow-derived DC activity of antithrombin III (ATIII) and C1-INH against [38]. In addition, HIF-1α regulates the expression of VEGF activated factor XI [46]. Interestingly, LMW-DXS interferes and endothelin-1 (ET-1), a potent vasoactive peptide known with platelet adhesion [47] and has beneficial effects in to be involved in cardiac I/R injury [39]. Stimulation of different animal models of cardiac I/R injury. In a pig macrophagesaswellasMoDCwithTLRagonistsinducesthe model of acute myocardial I/R injury, LMW-DXS signif- secretion of VEGF and ET-1 [40, 41]. icantly reduced infarct size [48] and facilitated anti-CD4 mAb-induced long-term cardiac allograft survival in rats despite prolonged cold graft ischemia [49]. LMW-DXS also 3.3. Strategies to Inhibit TLR Activation in Cardiac I/R inhibited TLR2- and TLR4-mediated maturation of human Injury. Specific inhibition of TLR2 with a monoclonal MoDC and prevented the upregulation of costimulatory antibody has been demonstrated to attenuate myocardial molecules including CD40, CD80, and CD86 on MoDC in a I/R injury in mice [42]. Similarly, eritoran, a synthetic dose-dependent manner. Secretion of the proinflammatory lipid-A analogue which inhibits TLR4 signaling, was ben- cytokines TNF-α, IL-6, and IL-1β was also significantly eficial in a mouse model of myocardial infarction [43]. reduced. As a functional consequence of these effects, It should be noted however that I/R injury is a mul- antigen-presentation to T cells was prevented. TLR-induced tifactorial injury involving many different inflammatory signal transduction, phosphorylation of IκB-α,aswellas pathways and to target one single pathway might not be activation of the downstream proinflammatory transcription 6 Cardiology Research and Practice factor NFκB were also inhibited by LMW-DXS [33]. An asso- anti-inflammatory cytokine IL-10 was significantly upreg- ciated in vitro study showed similar effects of LMW-DXS on ulated. No inhibition was observed for the secretion of TLR2-mediated activation of human NK cells. Phenotypic the proinflammatory cytokine TNF-α. Importantly, DC- activation was significantly inhibited by LMW-DXS as shown mediated T cell proliferation analyzed in mixed leukocyte by a reduced upregulation of CD25, CD56, CD69, as well reaction (MLR) with allogeneic T cells was reduced by as NKp44. Furthermore, release of IFNγ was significantly IVIg [64]. A recent study has further investigated the effect reduced and degranulation of NK cells was attenuated as of IVIg on the differentiation of CD14 positive human evaluated by the upregulation of CD107a [50]. LMW-DXS monocytes into MoDC. Ballow and Allen demonstrated also interferes with activation of MoDC by the endogenous that IVIg primarily modulates the differentiation of “TLR- TLR4 ligand, heparan sulfate proteoglycan (HSPG). This is primed” monocytes before subsequent differentiation into rapidly released from the vascular endothelial surface under MoDC by IL-4 and GM-CSF. Interestingly, priming with poly conditions of inflammation and tissue damage [51–53]by I/C (TLR3 agonist) in combination with IVIg did not result proteolytic cleavage of the protein core or by endoglycolytic in an increase of CD83 expression, as observed for the other cleavage of the HS chains [53, 54]. In the plasma of vascular used TLR agonists [65]. Controversial results were observed surgery patients, elevated levels of syndecan-1 and HS were regarding DC-induced proliferation of allogeneic T-cell. This found as early as 15 minutes after reperfusion [55]. Free study demonstrated an increase of T cell proliferation by HS has been considered to act as DAMP, since it induces IVIg-differentiated DC [65] whereas Bayry et al. have shown maturation of macrophages and DC via TLR4 [56–58]. a decrease of T cell proliferation [64]. Overall, LMW-DXS is well tolerated in humans and The effect of TLR9 stimulation by CpG oligonucleotides induces an increase in the anti-inflammatory hepatocyte on B cells of SLE patients as well as healthy donors has been growth factor (HGF) in plasma [59]. In addition it not investigated. Coincubation of B cells with IVIg (10 mg/ml) only acts as a complement and coagulation inhibitor but and CpG oligos resulted in a decreased secretion of IL-6 as also modulates innate immune cells as, for example, shown well as IL-10. IVIg had a similar inhibitory effect on the by inhibition of NK cell activation. Moreover, the crosstalk activation of human B cells isolated from SLE patients and between innate and adaptive immunity is prevented by healthy controls [66]. inhibition of TLR-mediated maturation of DC. Hence, It would be of interest if IVIg also interferes with cell LMW-DXS and similarly acting, nontoxic inhibitors of activation induced by DAMP. As shown by Bayry et al., IVIg complement, coagulation, and TLR activation may be preparations contain specific LPS antibodies [64]. Therefore, of therapeutic interest for the prevention of cardiac I/R some of the observed inhibitory effects might be explained injury. by binding and neutralization of the exogenous agonist used, namely LPS.

3.3.2. Intravenous Immunoglobulins (IVIgs). IVIg products are derived from pooled human plasma of thousands of 3.3.3. C1-Esterase Inhibitor (C1-INH). C1-INH has been donorsandhavebeenusedfordecadesasreplacement reported to have various anti-inflammatory properties. Ini- therapy for patients with hypogammaglobulinemia, such tially C1-INH has been characterized as a potent inhibitor as X-linked agammaglobulinemias or common variable of the classical pathway of complement activation. Follow- immunodeficiencies. In addition, anti-inflammatory therapy up studies provided evidence for inhibition of the mannan- with high-doses of IVIg is used clinically in a variety of acute binding lectin as well as the alternative pathway of comple- or chronic autoimmune disease as, for example, Idiopathic ment activation. Protease inhibition has been reported for Thrombocytopenic Purpura (ITP), Kawasaki Disease, or kallikrein, factor XI, factor XII, plasmin, tissue plasminogen Guillain Barre´ Syndrome. Several anti-inflammatory prop- activator (tPA), and thrombin. Furthermore, binding to erties of IVIg have been described for IVIg preparations neutrophils, macrophages, and endothelial cells (ECs) has including complement inhibition, anticytokine antibodies, also been shown [67]. In the clinic, C1-INH is commonly inhibition of leukocyte rolling, induction of T regulatory used to treat patients with C1-INH deficiencies who suffer cells, and others as published and reviewed by others [60– from Hereditary Angioedemia (HAE). 62]. Importantly, IVIg has successfully been used in the clinic In vivo, treatment with C1-INH has been demonstrated as a combination therapy for a patient undergoing cardiac to be beneficial in different models of cardiac I/R injury transplantation [63].Afewreportshavedemonstratedan [68, 69]. In addition, C1-INH has successfully been used inhibitory effect of IVIg on TLR-mediated activation of as a combined therapy for a patient undergoing cardiac immune cells in vitro. One study investigated the effect of transplantation [63]. Only one study has investigated the IVIg on differentiation of monocytes into MoDC as well as effect of C1-INH on TLR4-mediated activation of a murine the effect of IVIg on TLR4-mediated maturation of MoDC. macrophage cell line. Inhibition of LPS-mediated activation Pretreatment of MoDC with IVIg followed by stimulation of murine macrophages seems to be mediated by binding with LPS resulted in reduced upregulation of the costim- and neutralization of LPS by C1-INH [70]. N-glycosylation ulatory molecules CD40, CD80, and CD86. Furthermore, of the protein seems to be important for binding of LPS upregulation of MHC class II, the key molecule for antigen- [71]. Interestingly, C1-INH also binds to graft EC, while presentation, was prevented by IVIg. Interestingly, LPS- still maintaining its function as complement inhibitor as induced secretion of IL-12p70 was inhibited whereas the shown in a model of ex vivo perfused porcine livers [72]and Cardiology Research and Practice 7 could therefore potentially be used as an additive to organ endotoxemia. rHDL has been reported to inhibit upreg- preservation solutions in order to protect the graft from I/R ulation of inflammatory adhesion molecules like ICAM-1 injury. (CD54), VCAM-1 (CD106) and E-selectin (CD62E), on EC [82] as well as reduce thrombin-induced tissue-factor (TF) 3.3.4. Antithrombin III (ATIII). ATIII is a major inhibitor of expression [83]. Furthermore, a recent study in humans α the coagulation system and acts as a potent inactivator of has shown that rHDL reduces plasma levels of TNF- and thrombin and factor Xa. Clinically, ATIII concentrates are expression of CD11b on monocytes [84]. Protection against used to treat inherited ATIII deficiencies [73]. Apart from cardiac I/R injury has been demonstrated in ex vivo perfused coagulation inhibition, ATIII has been shown to interfere rat hearts in association with a reduced cardiac content of α with leukocyte adhesion in vitro and in vivo [74]. ATIII also TNF- and enhanced secretion of prostaglandin [85]. In vivo, possesses effective anti-inflammatory properties. Mansell rHDL has been shown to prevent left ventricle remodeling et al. have demonstrated an inhibitory effect of ATIII on and improve heart function after MI. Nitric oxide levels TLR4-mediated nuclear translocation of the proinflamma- in plasma were significantly reduced in the rHDL-treated tory transcription factor NFκB in the human monocytic group, whereas no decrease in IL-6 levels was observed ff cell line THP1 [75]. There are no other reports so far [86]. An e ect of rHDL on TLR4-mediated inflammation showing an inhibitory effect of ATIII on other TLR agonist- has been demonstrated in gram-negative sepsis via binding induced activation of immune cells. Interestingly, in a mouse and neutralizing LPS and reduction of CD14 expression on model of cardiac transplantation, high doses of ATIII have monocytes [87]. Overall, rHDL attenuates the proinflamma- ff been shown to induce long-term graft survival. Induction tory e ects of many mediators of innate immunity. of regulatory cells by ATIII has been suggested but not demonstrated by the authors [76]. 3.3.7. Statins. HMG-CoA reductase inhibitors, namely, statins, have been reported to have broad anti-inflammatory 3.3.5. Alpha-1 Antitrypsin (α1AT). α1AT is routinely used in biological properties. In a single center study of HTx patients, the clinic to treat patients with α1AT deficiency and lung statins are even thought to have immunosuppressive prop- emphysema. α1AT is a potent inhibitor of the neutrophil erties. Lipid lowering therapy was strongly associated with enzyme elastase. An imbalance between α1AT and elastase an improvement in one-year survival of the patients whereas increases the risk of emphysema [77]. In a mouse model of no difference was reported regarding graft rejection [88]. silica-induced inflammation, influx of granulocytes in the In addition, statins have been demonstrated to inhibit the lungs was significant inhibited by α1AT. Interestingly α1AT upregulationofCD86aswellasMHCclassIIonMoDCand inhibited activation of the proinflammatory transcription as consequence, allogeneic T cell activation and proliferation factor NFκBinlungs[78]. An in vitro study investigated the was prevented. effect of α1AT on the TLR4-mediated activation of human Interestingly, pretreatment of primary human mono- primary monocytes. Only long-term (18 hours) cell exposure cytes with atorvastatin and simvastatin significantly inhibited to a1AT inhibited LPS-induced upregulation of the proin- the secretion of IL-6, IL-12, and TNF-α. In addition, LPS- flammatory mediators TNF-α,IL-1β, and IL-8. In contrast, induced upregulation of the costimulatory molecule CD80 short-term exposure to α1AT/LPS increased the expression was significantly inhibited by both statins. Furthermore, of TNF-α,IL-1β, and IL-8. Furthermore, long-term exposure atorvastatin as well as simvastatin induced a decrease in of monocytes to α1AT/LPS decreased the surface expression TLR4 expression at the protein as well as the mRNA level of CD14 and TLR4 [79]. Recently, a study performed by in human monocytes [89]. In addition, statins have been Toldo et al. demonstrated a beneficial effect of α1AT in demonstrated to inhibit the upregulation of CD86 as well as a mouse model of myocardial infarction. Infarct size was MHC class II on MoDC and as a consequence, allogeneic T significant reduced in the group treated with α1AT compared cell activation and proliferation were prevented [90]. with the control treated with albumin. In addition, α1AT treatment significantly reduced cytokine/chemokine tissue levels of IL-6, IL-10, IL-17, TNF-α,andMCP-1[80]. Further 4. Discussion investigations are necessary to determine the molecular Accumulating evidence indicates that TLR may play impor- α mechanism of 1AT mediated inhibition of TLR-induced tant roles in the pathogenesis of atherosclerosis, microvascu- activation of cells. Interestingly, it has been speculated by lar complication of diabetes, viral myocarditis, dilated car- Johnson et al. that elastase released by activated neutrophils diomyopathy, cardiac allograft rejection, and sepsis-induced might be a major in vivo mechanism by which HS is cleaved left ventricular dysfunction. Moreover, heart failure of from the surface of vascular EC and subsequently induces diverse etiology is also now recognized to have an important inflammation via TLR4 [81]. immune component, with TLR signaling influencing the process of cardiac remodeling and prognosis. In cardiac I/R 3.3.6. Reconstituted High-Density Lipoprotein (rHDL). A injury, studies with TLR2 and TLR4 deficient mice have beneficial effect of treatment with reconstituted High- suggested a crucial involvement of these receptors in the early Density Lipoprotein (rHDL), containing apolipoprotein A- process of inflammation. In addition, blockade with TLR2 I and phosphatidylcholine (PC), has been described in or TLR4 antagonists has been shown to have a beneficial multiple diseases including arteriosclerosis, MI, stroke, and effect in MI [42, 43]. Several danger molecules or DAMP 8 Cardiology Research and Practice

Table 2: Selective-TLR antagonists in cardiovascular disease.

Compound Target Indication Drug class Clinical phase Company Inflammation, OPN-305 TLR2 Antibody Preclinical Opsona Therapeutics autoimmunity, I/R injury Sepsis Synthetic Eritoran TLR4 Phase III Eisai Pharmaceuticals I/R injury (preclinical) lipopolysacharide TAK-242 TLR4 Sepsis, inflammation Small-molecule inhibitor Discontinued in phase III Takeda Inflammation, autoimmunity, NI-0101 TLR4 Antibody Preclinical Novimmune CVD, I/R injury, and so forth TLR7 and IMO-3100 SLE, RA, MS, atherosclerosis DNA-based compound Preclinical Idera Pharmaceuticals TLR9 CVD: cardiovascular disease; MS: multiple sclerosis; RA: rheumatoid arthritis; SLE: systemic lupus erythematosus; I/R: ischaemia reperfusion. identified in vitro are suggested to be the ligands of TLR and receptors and simultaneous agonist treatment of beneficial therefore the initiators of sterile inflammation, but overall, TLR receptor responses. Another major drawback may be there is still no direct evidence in vivo that these molecules the limitations of currently used animal models to mimic are the drivers of the inflammatory process. Cardiac I/R the clinical entity being targeted. In most cases, CVD occurs is a multifactorial injury resulting in the activation of against a background of one or more other pathological several proinflammatory pathways such as the complement processes such as hyperlipidemia, atherosclerosis, and hyper- or coagulation system (Figure 2). Inhibition of complement glycemia each of which may differentially modify the TLR has been demonstrated to be cardioprotective in I/R injury receptor and signaling profile in any given tissue or cell but had no effect on patients treated with fibrinolysis for type. Rarely are experimental therapeutic profiles performed acuteMI[91]. There are several compounds described, in atherosclerotic prone diabetic mice when evaluating, for most of them already routinely used in the clinic to treat example, peripheral arterial disease. Until the profiles of TLR patients, for example, C1-INH or IVIg, which inhibit various changes are fully understood, it is unlikely that monotherapy proinflammatory pathways including TLR. Antagonism of targeting a single TLR entity will prove to be effective in these pattern recognition receptors might contribute to the treating CVD. observed therapeutic effect of these substances shown in Thus, more experiments are warranted to study the various animal models of cardiac I/R injury. Most of the detailed profile of TLR changes in CVD and the differential “nonselective” TLR antagonists summarized in Table 1 seem or additive effects of one or more risk factors that are known to mainly target TLR4, but in almost all studies, only LPS to influence disease progression or outcome. was used as activating agent. Whether these compounds might also inhibit TLR signaling induced by other TLR ligands, exogenous as well as endogenous, remains to be Abbreviations investigated. Despite the large amount of evidence, based CVD : Cardiovascular Disease largely on animal studies, that TLRs are involved in a DC : Dendritic cell wide range of pathological cardiovascular processes, clinical DAMP : damage-associated-molecular-pattern translation of therapeutic targeting of TLR in CVD in HTx : Heart Transplantation man is still in the early stages. Two TLR4 antagonists, I/R injury : Ischemia/Reperfusion injury E5564 (Eritoran) by Eisai, Inc., and TAK-242 by Takeda IVIg : Intravenous Immunoglobulins Pharmaceutical Company, have been used in phase III LMW-DXS : Low molecular weight dextran sulfate clinical trials for the management of severe sepsis where MI : Myocardial infarction the drugs ware largely well tolerated. For cardiovascular MoDC : monocyte-derived dendritic cells diseases, Eritoran also showed some preclinical benefits TLR : Toll-like receptors. and attenuated myocardial I/R injury by inhibiting TLR4. Currently, there are no additionally published clinical trial data looking at TLR antagonists as a therapeutic for CVD References (Table 2). [1] S. Danese, E. Dejana, and C. Fiocchi, “Immune regula- This slow progress may have several causes. In most tion by microvascular endothelial cells: directing innate and cases, it is unclear as to how and to what extent various adaptive immunity, coagulation, and inflammation,” Journal TLR receptors and signaling pathways are altered in different of Immunology, vol. 178, no. 10, pp. 6017–6022, 2007. CVDs. In virtually all cases, how the complete family of [2] B. Lemaitre, E. Nicolas, L. Michaut, J. M. Reichhart, and TLR is altered in different disease states is unknown. This J. A. Hoffmann, “The dorsoventral regulatory gene cassette may have various consequences particularly if activation spatzle/Toll/Cactus controls the potent antifungal response in of specific TLR has protective effects. Effective therapeutic Drosophila adults,” Cell, vol. 86, no. 6, pp. 973–983, 1996. intervention may require antagonist blockade of detrimental [3] R. Medzhitov, P. Preston-Hurlburt, and C. A. Janeway Jr., “A human homologue of the Drosophila toll protein signals Cardiology Research and Practice 9

activation of adaptive immunity,” Nature, vol. 388, no. 6640, [21] H. Shi, M. V. Kokoeva, K. Inouye, I. Tzameli, H. Yin, and J. pp. 394–397, 1997. S. Flier, “TLR4 links innate immunity and fatty acid-induced [4] B. Beutler, “Inferences, questions and possibilities in Toll-like insulin resistance,” Journal of Clinical Investigation, vol. 116, receptor signalling,” Nature, vol. 430, no. 6996, pp. 257–263, no. 11, pp. 3015–3025, 2006. 2004. [22] A. Tarkowski, J. Bjersing, A. Shestakov, and M. I. Bokarewa, [5] M. L. Alegre, J. Leemans, A. Le Moine et al., “The multiple “Resistin competes with lipopolysaccharide for binding to facets of toll-like receptors in transplantation biology,” Trans- Toll-like receptor 4,” Journal of Cellular and Molecular plantation, vol. 86, no. 1, pp. 1–9, 2008. Medicine, vol. 14, no. 6, pp. 1419–1431, 2010. [6] S. Akira and K. Takeda, “Toll-like receptor signalling,” Nature [23] M. Buraczynska, I. Baranowicz-Gaszczyk, J. Tarach, and A. Reviews Immunology, vol. 4, no. 7, pp. 499–511, 2004. Ksiazek, “Toll-like receptor 4 gene polymorphism and early [7] L. A. J. O’Neill, “How Toll-like receptors signal: what we know onset of diabetic retinopathy in patients with type 2 diabetes,” and what we don’t know,” Current Opinion in Immunology, Human Immunology, vol. 70, no. 2, pp. 121–124, 2009. vol. 18, no. 1, pp. 3–9, 2006. [24] S. Devaraj, I. Jialal, J. M. Yun, and A. Bremer, “Demonstration [8] T. Kaisho, O. Takeuchi, T. Kawai, K. Hoshino, and S. of increased toll-like receptor 2 and toll-like receptor 4 Akira, “Endotoxin-induced maturation of MyD88-deficient expression in monocytes of type 1 diabetes mellitus patients dendritic cells,” Journal of Immunology, vol. 166, no. 9, pp. with microvascular complications,” Metabolism, vol. 60, no. 2, 5688–5694, 2001. pp. 256–259, 2011. [9] T. Kawai, O. Takeuchi, T. Fujita et al., “Lipopolysaccharide [25] C. P. Hodgkinson, R. C. Laxton, K. Patel, and S. Ye, “Advanced stimulates the MyaD88-independent pathway and results in glycation end-product of low density lipoprotein activates activation of IFN-regulatory factor 3 and the expression of the toll-like 4 receptor pathway implications for diabetic a subset of lipopolysaccharide-inducible genes,” Journal of atherosclerosis,” Arteriosclerosis, Thrombosis, and Vascular Immunology, vol. 167, no. 10, pp. 5887–5894, 2001. Biology, vol. 28, no. 12, pp. 2275–2281, 2008. [10] K. R. Taylor, K. Yamasaki, K. A. Radek et al., “Recognition of [26] D. J. Kaczorowski, A. Nakao, K. P. Mollen et al., “Toll-like hyaluronan released in sterile injury involves a unique receptor receptor 4 mediates the early inflammatory response after cold complex dependent on toll-like receptor 4, CD44, and MD-2,” ischemia/reperfusion,” Transplantation, vol. 84, no. 10, pp. Journal of Biological Chemistry, vol. 282, no. 25, pp. 18265– 1279–1287, 2007. 18275, 2007. [27] J. Favre, P. Musette, V. Douin-Echinard et al., “Toll-like receptors 2-deficient mice are protected against postischemic [11] P. Matzinger, “Tolerance, danger, and the extended family,” coronary endothelial dysfunction,” Arteriosclerosis, Thrombo- Annual Review of Immunology, vol. 12, pp. 991–1045, 1994. sis, and Vascular Biology, vol. 27, no. 5, pp. 1064–1071, 2007. [12] S. Gallucci, M. Lolkema, and P. Matzinger, “Natural adjuvants: [28] P. Zhao, J. Wang, L. He et al., “Deficiency in TLR4 signal endogenous activators of dendritic cells,” Nature Medicine, vol. transduction ameliorates cardiac injury and cardiomyocyte 5, no. 11, pp. 1249–1255, 1999. contractile dysfunction during ischemia,” Journal of Cellular [13] E. Lin, J. E. Freedman, and L. M. Beaulieu, “Innate immunity and Molecular Medicine, vol. 13, no. 8, pp. 1513–1525, 2009. and toll-like receptor antagonists: a potential role in the treat- [29] T. Ha, Y. Hu, L. Liu et al., “TLR2 ligands induce cardio- ment of cardiovascular diseases,” Cardiovascular Therapeutics, protection against ischaemia/reperfusion injury through a vol. 27, no. 2, pp. 117–123, 2009. PI3K/Akt-dependent mechanism,” Cardiovascular Research, [14] H. Kono and K. L. Rock, “How dying cells alert the immune vol. 87, no. 4, pp. 694–703, 2010. system to danger,” Nature Reviews Immunology, vol. 8, no. 4, [30] M. Zhang, E. M. Alicot, I. Chiu et al., “Identification of pp. 279–289, 2008. the target self-antigens in reperfusion injury,” Journal of [15] G. Y. Chen and G. Nunez,˜ “Sterile inflammation: sensing and Experimental Medicine, vol. 203, no. 1, pp. 141–152, 2006. reacting to damage,” Nature Reviews Immunology, vol. 10, no. [31] H. Hawlisch, Y. Belkaid, R. Baelder, D. Hildeman, C. Gerard, 12, pp. 826–837, 2010. and J. Kohl,¨ “C5a negatively regulates toll-like receptor 4- [16] M. Higashimori, J. B. Tatro, K. J. Moore, M. E. Mendelsohn, induced immune responses,” Immunity,vol.22,no.4,pp. J. B. Galper, and D. Beasley, “Role of toll-like receptor 4 in 415–426, 2005. intimal foam cell accumulation in apolipoprotein E-deficient [32] X. Zhang, Y. Kimura, C. Fang et al., “Regulation of Toll-like mice,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. receptor-mediated inflammatory response by complement in 31, no. 1, pp. 50–57, 2011. vivo,” Blood, vol. 110, no. 1, pp. 228–236, 2007. [17] J.-Q. Gu, D.-F. Wang, X.-G. Yan et al., “A Toll-like receptor 9- [33]R.Spirig,C.VanKooten,C.Obregon,L.Nicod,M.Daha, mediated pathway stimulates perilipin 3 (TIP47) expression and R. Rieben, “The complement inhibitor low molecular and induces lipid accumulation in macrophages,” American weight dextran sulfate prevents TLR4-induced phenotypic Journal of Physiology, vol. 299, no. 4, pp. E593–E600, 2010. and functional maturation of human dendritic cells,” Journal [18] E. LeBouder, J. E. Rey-Nores, N. K. Rushmere et al., “Soluble of Immunology, vol. 181, no. 2, pp. 878–890, 2008. forms of Toll-like receptor (TLR)2 capable of modulating [34] G. Castellano, A. M. Woltman, A. J. Nauta et al., “Maturation TLR2 signaling are present in human plasma and breast milk,” of dendritic cells abrogates C1q production in vivo and in Journal of Immunology, vol. 171, no. 12, pp. 6680–6689, 2003. vitro,” Blood, vol. 103, no. 10, pp. 3813–3820, 2004. [19] T. Ueland, T. Espevik, J. Kjekshus et al., “Mannose binding [35] P. Baruah, I. E. Dumitriu, G. Peri et al., “The tissue pen- lectin and soluble Toll-like receptor 2 in heart failure following traxin PTX3 limits C1q-mediated complement activation and acute myocardial infarction,” Journal of Cardiac Failure, vol. phagocytosis of apoptotic cells by dendritic cells,” Journal of 12, no. 8, pp. 659–663, 2006. Leukocyte Biology, vol. 80, no. 1, pp. 87–95, 2006. [20] J. E. Cole, T. J. Navin, A. J. Cross et al., “Unexpected protective [36] P. Baruah, I. E. Dumitriu, T. H. Malik et al., “C1q enhances role for Toll-like receptor 3 in the arterial wall,” Proceedings IFN-7 production by antigen-specific T cells via the CD40 of the National Academy of Sciences of the United States of costimulatory pathway on dendritic cells,” Blood, vol. 113, no. America, vol. 108, no. 6, pp. 2372–2377, 2011. 15, pp. 3485–3493, 2009. 10 Cardiology Research and Practice

[37] R. Spirig, S. Djafarzadeh, T. Regueira et al., “Effects of TLR rejection,” Journal of Experimental Medicine, vol. 171, no. 4, agonists on the hypoxia-regulated transcription factor HIF- pp. 1363–1368, 1990. 1α and dendritic cell maturation under normoxic conditions,” [52] J. L. Platt, A. P. Dalmasso, B. J. Lindman, N. S. Ihrcke, and PLoS One, vol. 5, no. 6, Article ID e10983, 2010. F. H. Bach, “The role of C5a and antibody in the release of [38] J. Jantsch, D. Chakravortty, N. Turza et al., “Hypoxia heparan sulfate from endothelial cells,” European Journal of and hypoxia-inducible factor-1 alpha modulate Immunology, vol. 21, no. 11, pp. 2887–2890, 1991. lipopolysaccharide-induced dendritic cell activation and [53] N. S. Ihrcke and J. L. Platt, “Shedding of heparan sulfate function,” Journal of Immunology, vol. 180, no. 7, pp. proteoglycan by stimulated endothelial cells: evidence for 4697–4705, 2008. proteolysis of cell-surface molecules,” Journal of Cellular [39] F. Brunner, E. F. Du Toit, and L. H. Opie, “Endothelin release Physiology, vol. 168, no. 3, pp. 625–637, 1996. during ischaemia and reperfusion of isolated perfused rat [54] N. S. Ihrcke, W. Parker, K. J. Reissner, and J. L. Platt, hearts,” Journal of Molecular and Cellular Cardiology, vol. 24, “Regulation of platelet heparanase during inflammation: role no. 11, pp. 1291–1305, 1992. of pH and proteinases,” Journal of Cellular Physiology, vol. 175, [40] H. Ehrenreich, R. W. Anderson, C. H. Fox et al., “Endothelins, no. 3, pp. 255–267, 1998. peptides with potent vasoactive properties, are produced by [55] M. Rehm, D. Bruegger, F. Christ et al., “Shedding of the human macrophages,” Journal of Experimental Medicine, vol. endothelial glycocalyx in patients undergoing major vascular 172, no. 6, pp. 1741–1748, 1990. surgery with global and regional ischemia,” Circulation, vol. [41] R. Spirig, I. Potapova, J. Shaw-Boden, J. Tsui, R. Rieben, and 116, no. 17, pp. 1896–1906, 2007. S. G. Shaw, “TLR2 and TLR4 agonists induce production of [56] G. B. Johnson, G. J. Brunn, Y. Kodaira, and J. L. Platt, the vasoactive peptide endothelin-1 by human dendritic cells,” “Receptor-mediated monitoring of tissue well-being via detec- Molecular Immunology, vol. 46, no. 15, pp. 3178–3182, 2009. tion of soluble heparan sulfate by toll-like receptor 4,” Journal [42] F. Arslan, M. B. Smeets, L. A. J. O’Neill et al., “Myocardial of Immunology, vol. 168, no. 10, pp. 5233–5239, 2002. ischemia/reperfusion injury is mediated by leukocytic toll-like [57] L. E. Wrenshall, R. B. Stevens, F. B. Cerra, and J. L. receptor-2 and reduced by systemic administration of a novel Platt, “Modulation of macrophage and B cell function by anti-toll-like receptor-2 antibody,” Circulation, vol. 121, no. 1, glycosaminoglycans,” Journal of Leukocyte Biology, vol. 66, no. pp. 80–90, 2010. 3, pp. 391–400, 1999. [43] A. Shimamoto, A. J. Chong, M. Yada et al., “Inhibition of toll- [58] L. E. Wrenshall, F. B. Cerra, R. K. Singh, and J. L. Platt, “Hep- like receptor 4 with eritoran attenuates myocardial ischemia- aran sulfate initiates signals in murine macrophages leading to reperfusion injury,” Circulation, vol. 114, no. 1, pp. I270–I274, divergent biologic outcomes,” Journal of Immunology, vol. 154, 2006. no. 2, pp. 871–880, 1995. [44] M. K. Pangburn, M. A. L. Atkinson, and S. Meri, “Localization [59] P. Schmidt, C. Magnusson, T. Lundgren, O. Korsgren, and of the heparin-binding site on complement factor H,” Journal B. Nilsson, “Low molecular weight dextran sulfate is well of Biological Chemistry, vol. 266, no. 25, pp. 16847–16853, tolerated in humans and increases endogenous expression of 1991. islet protective hepatocyte growth factor,” Transplantation, vol. [45] W. A. Wuillemin, H. Te Velthuis, Y. T. P.Lubbers, C. P.De Ruig, 86, no. 11, pp. 1523–1530, 2008. E. Eldering, and C. E. Hack, “Potentiation of C1 inhibitor [60] F. Nimmerjahn and J. V. Ravetch, “Anti-inflammatory actions by glycosaminoglycans: dextran sulfate species are effective of intravenous immunoglobulin,” Annual Review of Immunol- inhibitors of in vitro complement activation in plasma,” ogy, vol. 26, pp. 513–533, 2008. Journal of Immunology, vol. 159, no. 4, pp. 1953–1960, 1997. [61]A.Durandy,S.V.Kaveri,T.W.Kuijpersetal., [46] W. A. Wuillemin, E. Eldering, F. Citarella, C. P. De Ruig, H. “Intravenous immunoglobulins-understanding properties T. Cate, and C. Erik Hack, “Modulation of contact system and mechanisms,” Clinical and Experimental Immunology, proteases by glycosaminoglycans: selective enhancement of the vol. 158, supplement 1, pp. 2–13, 2009. inhibition of factor XIa,” Journal of Biological Chemistry, vol. 271, no. 22, pp. 12913–12918, 1996. [62]T.Tha-In,H.J.Metselaar,A.R.Bushell,J.Kwekkeboom, [47] S. Zeerleder, T. Mauron, B. Lammle,¨ and W. A. Wuillemin, and K. J. Wood, “Intravenous immunoglobulins promote “Effect of low-molecular weight dextran sulfate on coagula- skin allograft acceptance by triggering functional activation tion and platelet function tests,” Thrombosis Research, vol. 105, of CD4+Foxp3+ T cells,” Transplantation, vol. 89, no. 12, pp. no. 5, pp. 441–446, 2002. 1446–1455, 2010. [48] Y. Banz, O. M. Hess, S. C. Robson et al., “Locally targeted [63] P. Mohacsi, R. Rieben, G. Sigurdsson et al., “Successful cytoprotection with dextran sulfate attenuates experimental management of a B-type cardiac allograft into an O-type man porcine myocardial ischaemia/reperfusion injury,” European with 3 1/2-year clinical follow-up,” Transplantation, vol. 72, Heart Journal, vol. 26, no. 21, pp. 2334–2343, 2005. no. 7, pp. 1328–1330, 2001. [49] T. Gajanayake, B. Sawitzki, K. Matozan et al., “Dextran sulfate [64] J. Bayry, S. Lacroix-Desmazes, C. Carbonneil et al., “Inhibition facilitates anti-CD4 mAb-induced long-term rat cardiac allo- of maturation and function of dendritic cells by intravenous graft survival after prolonged cold ischemia,” American Journal immunoglobulin,” Blood, vol. 101, no. 2, pp. 758–765, 2003. of Transplantation, vol. 8, no. 6, pp. 1151–1162, 2008. [65] M. Ballow and C. Allen, “Intravenous immunoglobulin [50] A. L. Millard, R. Spirig, N. J. Mueller, J. D. Seebach, and modulates the maturation of TLR 4-primed peripheral blood R. Rieben, “Inhibition of direct and indirect TLR-mediated monocytes,” Clinical Immunology, vol. 139, pp. 208–214, 2011. activation of human NK cells by low molecular weight dextran [66] A. Kessel, R. Peri, T. Haj et al., “IVIg attenuates TLR-9 sulfate,” Molecular Immunology, vol. 47, no. 14, pp. 2349–2358, activation in B cells from SLE patients,” Journal of Clinical 2010. Immunology, vol. 32, pp. 30–38, 2011. [51] J. L. Platt, G. M. Vercellotti, B. J. Lindman, T. R. Oegema Jr., F. [67] A. E. Davis, P. Mejia, and F. Lu, “Biological activities of C1 H. Bach, and A. P. Dalmasso, “Release of heparan sulfate from inhibitor,” Molecular Immunology, vol. 45, no. 16, pp. 4057– endothelial cells. Implications for pathogenesis of hyperacute 4063, 2008. Cardiology Research and Practice 11

[68] G. Horstick, O. Berg, A. Heimann et al., “Application [84] J. A. Shaw, A. Bobik, A. Murphy et al., “Infusion of recon- of C1-esterase inhibitor during reperfusion of ischemic stituted high-density lipoprotein leads to acute changes in myocardium: dose-related beneficial versus detrimental human atherosclerotic plaque,” Circulation Research, vol. 103, effects,” Circulation, vol. 104, no. 25, pp. 3125–3131, 2001. no. 10, pp. 1084–1091, 2008. [69]M.Buerke,T.Murohara,andA.M.Lefer,“Cardioprotective [85] L. Calabresi, G. Rossoni, M. Gomaraschi, F. Sisto, F. Berti, and effects of a C1 esterase inhibitor in myocardial ischemia and G. Franceschini, “High-density lipoproteins protect isolated reperfusion,” Circulation, vol. 91, no. 2, pp. 393–402, 1995. rat hearts from ischemia-reperfusion injury by reducing [70] D. Liu, S. Cai, X. Gu, J. Scafidi, X. Wu, and A. E. Davis III, cardiac tumor necrosis factor-α content and enhancing “C1 inhibitor prevents endotoxin shock via a direct interaction prostaglandin release,” Circulation Research,vol.92,no.3,pp. with lipopolysaccharide,” Journal of Immunology, vol. 171, no. 330–337, 2003. 5, pp. 2594–2601, 2003. [86] Y. Kiya, S. I. Miura, S. Imaizumi et al., “Reconstituted high- [71] D. Liu, X. Gu, J. Scafidi, and A. E. Davis III, “N-linked density lipoprotein attenuates postinfarction left ventricular glycosylation is required for C1 inhibitor-mediated protection remodeling in rats,” Atherosclerosis, vol. 203, no. 1, pp. 137– from endotoxin shock in mice,” Infection and Immunity, vol. 144, 2009. 72, no. 4, pp. 1946–1955, 2004. [87] D. Pajkrt, J. E. Doran, F. Koster et al., “Antiinflamma- [72] L. Bergamaschini, G. Gobbo, S. Gatti et al., “Endothelial tory effects of reconstituted high-density lipoprotein during targeting with C1-inhibitor reduces complement activation in human endotoxemia,” Journal of Experimental Medicine, vol. vitro and during ex vivo reperfusion of pig liver,” Clinical and 184, no. 5, pp. 1601–1608, 1996. Experimental Immunology, vol. 126, no. 3, pp. 412–420, 2001. [88] J. A. Kobashigawa, “Statins in solid organ transplantation: [73] M. M. Patnaik and S. Moll, “Inherited antithrombin defi- is there an immunosuppressive effect?” American Journal of ciency: a review,” Haemophilia, vol. 14, no. 6, pp. 1229–1239, Transplantation, vol. 4, no. 7, pp. 1013–1018, 2004. 2008. [89] H. Methe, J.-O. Kim, S. Kofler, M. Nabauer, and M. Weis, [74] L. Ostrovsky, R. C. Woodman, D. Payne, D. Teoh, and P.Kubes, “Statins decrease Toll-like receptor 4 expression and down- “Antithrombin III prevents and rapidly reverses leukocyte stream signaling in human CD14+ monocytes,” Arteriosclero- recruitment in ischemia/reperfusion,” Circulation, vol. 96, no. sis, Thrombosis, and Vascular Biology, vol. 25, no. 7, pp. 1439– 7, pp. 2302–2310, 1997. 1445, 2005. [75] A. Mansell, A. Reinicke, D. M. Worrall, and L. A. J. O’Neill, [90] A. Yilmaz, C. Reiss, A. Weng et al., “Differential effects of “The serine protease inhibitor antithrombin III inhibits LPS- statins on relevant functions of human monocyte-derived mediated NF-κB activation by TLR-4,” FEBS Letters, vol. 508, dendritic cells,” Journal of Leukocyte Biology, vol. 79, no. 3, pp. no. 3, pp. 313–317, 2001. 529–538, 2006. [76] O. Aramaki, T. Takayama, T. Yokoyama et al., “High dose of [91] K. W. Mahaffey,C.B.Granger,J.C.Nicolauetal.,“Effect of antithrombin III induces indefinite survival of fully allogeneic pexelizumab, an anti-C5 complement antibody, as adjunctive cardiac grafts and generates regulatory cells,” Transplantation, therapy to fibrinolysis in acute myocardial infarction: the vol. 75, no. 2, pp. 217–220, 2003. COMPlement inhibition in myocardial infarction treated with [77] E. K. Silverman and R. A. Sandhaus, “Alpha1-antitrypsin thromboLYtics (COMPLY) trial,” Circulation, vol. 108, no. 10, deficiency,” New England Journal of Medicine, vol. 360, no. 26, pp. 1176–1183, 2003. pp. 2749–2757, 2009. [92] T. Laumonier, P.J. Mohacsi, K. M. Matozan et al., “Endothelial [78] A. Churg, J. Dai, K. Zay et al., “Alpha-1-antitrypsin and a cell protection by dextran sulfate: a novel strategy to prevent broad spectrum metalloprotease inhibitor, RS113456, have acute vascular rejection in xenotransplantation,” American similar acute anti-inflammatory effects,” Laboratory Investiga- Journal of Transplantation, vol. 4, no. 2, pp. 181–187, 2004. tion, vol. 81, no. 8, pp. 1119–1131, 2001. [93] U. Stenestrand and L. Wallentin, “Early statin treatment [79] I. M. Nita, D. Serapinas, and S. M. Janciauskiene, “α1- following acute myocardial infarction and 1-year survival,” Antitrypsin regulates CD14 expression and soluble CD14 Journal of the American Medical Association, vol. 285, no. 4, levels in human monocytes in vitro,” International Journal of pp. 430–436, 2001. Biochemistry and Cell Biology, vol. 39, no. 6, pp. 1165–1176, 2007. [80] S. Toldo, I. M. Seropian, E. Mezzaroma et al., “Alpha- 1 antitrypsin inhibits caspase-1 and protects from acute myocardial ischemia-reperfusion injury,” Journal of Molecular and Cellular Cardiology, vol. 51, no. 2, pp. 244–251, 2011. [81] G. B. Johnson, G. J. Brunn, and J. L. Platt, “Cutting edge: an endogenous pathway to systemic inflammatory response syndrome (SIRS)-like reactions through Toll-like receptor 4,” Journal of Immunology, vol. 172, no. 1, pp. 20–24, 2004. [82] C. Wadham, N. Albanese, J. Roberts et al., “High-density lipoproteins neutralize C-reactive protein proinflammatory activity,” Circulation, vol. 109, no. 17, pp. 2116–2122, 2004. [83] H. Viswambharan, X. F. Ming, S. Zhu et al., “Reconstituted high-density lipoprotein inhibits thrombin-induced endothe- lial tissue factor expression through inhibition of RhoA and stimulation of phosphatidylinositol 3-kinase but not Akt/endothelial nitric oxide synthase,” Circulation Research, vol. 94, no. 7, pp. 918–925, 2004. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 656247, 7 pages doi:10.1155/2012/656247

Review Article Nitric Oxide Manipulation: A Therapeutic Target for Peripheral Arterial Disease?

Gareth Williams,1 Xu Shi-Wen,2 David Abraham,2 Sadasivam Selvakumar,3 Daryll M. Baker,1 and Janice C. S. Tsui1

1 Division of Surgery & Interventional Science, University College London, Royal Free Campus, London NW3 2QG, UK 2 Centre for Rheumatology & Connective Tissue Disease, University College London, Royal Free Campus, London NW3 2QG, UK 3 Department Of Surgery, Lister Hospital, Coreys Mill Lane, Stevenage, Hertfordshire SG1 4AB, UK

Correspondence should be addressed to Janice C. S. Tsui, [email protected]

Received 15 July 2011; Accepted 16 December 2011

Academic Editor: Sidney G. Shaw

Copyright © 2012 Gareth Williams et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Peripheral Arterial Disease (PAD) is a cause of significant morbidity and mortality in the Western world. Risk factor modification and endovascular and surgical revascularisation are the main treatment options at present. However, a significant number of patients still require major amputation. There is evidence that nitric oxide (NO) and its endogenous inhibitor asymmetric dimethylarginine (ADMA) play significant roles in the pathophysiology of PAD. This paper reviews experimental work implicating the ADMA-DDAH-NO pathway in PAD, focussing on both the vascular dysfunction and effects within the ischaemic muscle, and examines the potential of manipulating this pathway as a novel adjunct therapy in PAD.

1. Introduction There have been significant efforts to elucidate the pa- thophysiological role of the NO pathway in cardiovascular Peripheral arterial disease (PAD) is a disorder characterised diseases, with recent discoveries of endogenous NO inhib- by the progressive occlusion of the large- and medium-sized itors adding further complexity [5],butthishasyettobe arteries outside of the cardiopulmonary and cerebral vascu- translated into clinically significant therapies. lar systems. It causes considerable morbidity and mortality, This paper aims to outline the current knowledge on NO affecting over 2.7 million individuals in the Western world pathway dysfunction in PAD, including recent developments [1] and most commonly manifests itself in the reduction of in understanding the role of asymmetric dimethylarginine blood flow to the lower limbs, which in severe cases lead to (ADMA) and dimethylarginine dimethylaminohydrolase chronic pain, gangrene, and eventually limb loss. Risk factors (DDAH); it will also explore potential therapeutic strategies include tobacco smoke, diabetes mellitus, hypertension, and based on the manipulation of this pathway. hypercholesterolaemia. At a molecular level, dysfunction of the vascular endothe- 2. The Nitric Oxide Pathway lium, in particular, disruption of the nitric oxide (NO) pathway [2], plays a significant role. NO is a vasodilator NO is synthesised from L-arginine by NOS, of which there synthesized by the enzyme nitric oxide synthase (NOS) [3] are three isoforms [6]: NOS I or neuronal NOS (nNOS) was and has a number of important endocrine and paracrine originally isolated from rat and porcine cerebellum; NOS II effects, including reduction of vascular smooth muscle tone; or inducible NOS (iNOS) from activated macrophages; NOS inhibition of platelet adhesion and aggregation; suppression III or endothelial NOS (eNOS) from endothelial cells. of inflammatory mediators; inhibition of smooth muscle All 3 NOS isoforms are competitively inhibited by proliferation and migration; and promotion of endothelial the methylarginines ADMA and monomethyl-arginine (L- survival and repair [4]. NMMA) which are by-products of protein degradation [7]. 2 Cardiology Research and Practice

L-arginine

ADMA

Renal excretion (20%) NOS III

DDAH DDAH1 - liver, kidney DDAH2 - vascular endothelium, heart, placenta, kidney

NO

L-citrulline Dimethylamine • Vasorelaxation • Decreased arterial stiffness • Anti-inflammatory • Anti-thrombotic • Anti-atherosclerosis

Figure 1: The nitric oxide pathway.

Both ADMA and L-NMMA are predominantly broken down due to reduced DDAH expression [18]. Upregulation of by the enzyme DDAH into L-citrulline and dimethylamine DDAH I and DDAH II in human umbilical vein endothelial [8], of which there are two isoforms [9]: DDAH I, found cells (HUVECs) by adenoviral vectors leads to a fall in predominantly in tissues expressing NOS I, liver, kidney, and ADMA and an increase in NO [7]. Similarly, upregulation of lung; and DDAH II, found in tissues expressing NOS III and DDAH I and II in DDAH I+/− mice was found to attenuate NOS II, vascular endothelium, heart, placenta, and kidney. the cardiovascular stress response, enhance acetylcholine- Both isoforms have been found to be expressed in vascular mediated relaxation, and counteract the response to excess endothelium [9](Figure 1). ADMA; DDAH upregulation helps to reverse the cardiovas- ADMA’s role in vascular endothelial dysfunction was cular dysfunction inherented in these animals [7]. first described by Vallance et al. in patients with end-stage However, DDAH I appears to be the main isoform re- chronic renal failure, whose serum ADMA is raised due, in sponsible for ADMA metabolism: DDAH I knockout mice part, to decreased renal clearance [10]. Further studies in demonstrated significantly raised ADMA levels and systolic similar patients have shown a positive correlation between blood pressure compared to wild types, despite DDAH II ADMA levels and cardiovascular morbidity and mortality expression remaining normal [19]. [11]; whilst reduction of ADMA levels by renal dialysis helps These studies suggest that DDAH manipulation may be a to restore endothelial function [12]. useful tool in the treatment of cardiovascular disease. Other studies have demonstrated a correlation between ADMA accumulation and cardiovascular risk as well as other disease states [5]. Lu et al. showed that serum ADMA levels 3. Peripheral Arterial Disease (PAD) have a positive correlation with the severity and extent of coronary artery atherosclerosis [13]; Worthmann et al. In PAD, lower limb ischaemia results in intermittent clau- demonstrated a similar correlation between ADMA levels dication: this describes pain in the affected muscle groups and adverse prognosis in ischaemic stroke [14]; and ADMA brought on by exercise and relieved by resting. In more has been found to have significant roles in renal disease [15], severe PAD, critical limb ischaemia (CLI) occurs: the patient pulmonary artery hypertension [16], and erectile dysfunc- experiences chronic pain at rest, with blood flow to the limb tion [17]. compromised to the extent that the affected limb is at risk, The accumulation of ADMA and consequent endothelial with development of ulcers and gangrene. dysfunction, increased systemic vascular resistance, and in- The Fontaine classification [20] stratifies these symptoms creased systemic blood pressure have been shown to be to correlate with severity of the disease (Table 1). Cardiology Research and Practice 3

Table 1: Fontaine Classification: stages III and IV are also known as 4. NO in Skeletal Muscle CLI. Skeletal muscle has been shown to express both NOS I Stage I Asymptomatic and NOS III: NOS I is localized to the sarcolemma, and Stage II Intermittent claudication, no rest pain NOS III associated with mitochondria and microvessels IIa When walking a distance of greater than 200 m within the tissue [27]. Within skeletal muscle, NO acts as an endogenous modulator of vascular tone [28, 29], IIb When walking a distance of less than 200 m neuromuscular transmission [30], muscle contraction [31, Stage III Nocturnal pain and/or pain at rest 32], muscle metabolism [27] and myogenesis [33]. Its role Stage IV Tissue loss: ischaemic ulcers and/or gangrene in the regulation of muscle contraction is complex and is influenced by the muscle fibre type and contraction pattern and frequency involved [31]. NO has a number of effects on skeletal muscle metabolism [27]. It affects glucose metabolism, upregulating glucose uptake and inhibiting Current therapies for CLI centre on restoring blood glycolysis [34]; it has a direct inhibitory effect on mitochon- ff flow to the a ected limb by surgical and/or endovascular drial respiration [35]; and it antagonises creatinine kinase procedures. Unfortunately, of all patients presenting with activity, decreasing phosphocreatine breakdown, limiting CLI, only 50% will have disease amenable to revascularisa- muscle function [36]. tion with 25% requiring a primary amputation. The overall NO plays a role in myogenesis as well as repair and prognosis of these patients is also poor: within a year of regeneration following injury. It has been shown to mod- presentation, a quarter of the patients would have died, 30% ulate the inflammatory response to injured skeletal muscle, would have had a major amputation, and 20% continue to controlling blood flow to the injured tissue, and affecting ff su er from CLI [1]. subsequent remodelling and repair [37]. In experimental There is evidence that the NO-ADMA-DDAH pathway models of muscle crush injury, upregulation of local NO has plays an important role in PAD. Evidence of systemically been demonstrated [38], inhibiting adhesion and inducing impaired NO synthesis has been described by Boger¨ et al. apoptosis in inflammatory cells [39]. NOS is activated in who showed urinary nitrate and cGMP excretion to diminish response to stretch in skeletal muscle, mediating the release in PAD patients as the disease progressed [2]. This group of hepatocyte growth factor and activating mononuclear pro- also demonstrated that intravenous L-arginine administra- genitor cells called satellite cells (SC), which are responsible tion induced NO-mediated vasodilation, increasing femoral for post-injury remodelling [40]. blood flow in patients with CLI, resulting in improved Tidball et al. showed that the addition of exogenous walking distances and symptom scores [21]. Furthermore, NO upregulated the expression of structural proteins within raised plasma ADMA levels have also been shown to correlate mouse myotubes and found that NOS antagonists inhibited with disease severity in PAD patients [5], implicating the this, suggesting a regulatory role for NO in myogenesis [33]. involvement of ADMA. In addition to local effects, NO appears to play a part in NO also promotes angiogenesis, activating vascular en- the systemic inflammatory response: experimental rats with dothelial growth factor and fibroblast growth factor [22]. peritonitis or septic shock have been shown to have increased Administration of NO antagonists to canine coronary expression of NOS in their skeletal muscle [41]. ischaemia models has demonstrated that NO is an important However, NO has also been shown to have deleterious mediator of new vessel growth [23]. This mechanism of effects, reacting with superoxide anions to form harmful stimulation of new blood vessel development to supply nitrogenous intermediates which are directly toxic to skeletal ischaemic tissue has been proposed as a possible treatment muscle and inhibit mitochondrial function [41, 42]. for PAD, but clinical results have so far been disappointing. The formation and persistence of new blood vessels is likely 5. NO in Ischaemia to be multifactorial, involving a complex interaction between chemokines and mechanical forces [24]. In addition, con- Impaired NO-mediated vasodilation has been found in tinued exposure to cardiovascular risk factors, such as postischaemic tissue such as myocardium [43], brain [44], hypercholesterolaemia and hyperglycaemia, has been shown and skeletal muscle [45]. There is a complex interaction to have a deleterious effect on the NO pathway [25, 26]. between ischaemia and the NOS isoforms, with both up- and However, the NO-ADMA-DDAH pathway role in PAD downregulation reported in ischaemic models; for instance, is not confined to vascular dysfunction: NO-ADMA-DDAH rat brain oxygen-glucose-deprivation models have shown interaction is known to play significant parts in the function, NOS II to be upregulated in response to ischaemia, which homeostasis, and repair of skeletal muscle. Weight for weight, in turn causes a downregulation of NOS I [46]; however, this is the most abundant tissue in the lower limb and ischaemic preconditioning of rat liver has been shown to the majority of PAD symptoms are attributable to skeletal upregulate NOS III [47]. muscle ischaemia; it can almost be regarded as a “target In skeletal muscle from patients with CLI, we have organ” of PAD. Hence, the NO-ADMA-DDAH pathway previously found raised NOS III expression associated with within ischaemic skeletal muscle represents a significant facet both microvessels within ischaemic muscle sections as well of potential CLI therapy. as the muscle fibres themselves [48]. However, this was not 4 Cardiology Research and Practice

250

200

150

DDAH2 100 Protein (nmol/g) Protein

50 Tubulin

0 1234567 8 Normal CLI (a) (b)

0.3

∗ 0.2 mol/L) µ (

0.1

CHCC H H

DDAH2 0 C2C12 C2C12 HMEC

Control Tubulin Hypoxia

∗P<0.05 (c) (d)

Figure 2: ADMA/DDAH pathway in ischaemic muscle—preliminary data. (a) Western blot showing reduced DDAH2 expression in muscle biopsies from patients with CLI. Lanes 1–4: control, 5–8: CLI muscle. (b) ELISA showing significantly higher ADMA levels in muscle from patients with CLI (P = 0.03, Mann Whitney U test). (c) Western blot showing reduced DDAH2 expression in hypoxic myotubes (C: control, H: hypoxia). (d) Conditioned medium from hypoxic C2C12 myotubes contained elevated levels of ADMA compared to medium from myotubes cultured in normoxia. (P<0.05, Students t-test). Medium from HMEC-1 cells was used as positive control.

associated with an increase in NOS activity. The involvement Whilst the majority of studies have focussed on ADMA of endogenous inhibitors may explain this and recent associated with the vasculature/endothelial cells, our findings preliminary data suggesting that ADMA levels in muscle from our in vitro studies suggest that ischaemic myotubes homogenates from patients with CLI are raised (Figures may be a source of ADMA which can then act in a paracrine 2(a) and 2(b)). In addition, DDAH2 protein levels are also fashion on surrounding endothelial cells or in an autocrine upregulated in these ischaemic muscle biopsies. manner on myotubes themselves with potentially detrimen- To study the potential role of the ADMA/DDAH pathway tal effects. in skeletal muscle further, we evaluated their levels in my- Fiedler et al. found that ADMA impairs vascular endo- otubes cultured in ischaemic conditions in vitro.Wefound thelial growth factor (VEGF) mediated angiogenesis, and reduced DDAH2 expression in ischaemic myotubes, whilst that this was prevented by upregulation of DDAH [49]. ADMA levels were elevated in ischaemic myotube condi- Increased levels of ADMA produced by ischaemic muscle tioned media (Figures 2(c) and 2(d)). may prevent angiogenesis in PAD patients and may be a Cardiology Research and Practice 5 factor in the failure of angiogenesis induction as a clinical ADMA in particular, through its actions on NOS III, has therapeutic approach. been shown to be an important modulator of NO. Early work suggests that local or systemic manipulation of ADMA, by 6. Manipulation of the ADMA-DDAH Pathway upregulating DDAH, may prove to be a novel treatment for PAD. Control of NO levels by manipulating the ADMA-DDAH- NO pathway represents a significant avenue for potential References early PAD treatment. As L-arginine is the main precursor for NO production [1]L.Norgren,W.R.Hiatt,J.A.Dormandy,M.R.Nehler,K.A. (Figure 1), an increase in the availability of substrate ought Harris, and F. G. R. Fowkes, “Inter-society consensus for the to result in an increase in NO. However, there is conflicting management of peripheral arterial disease (TASC II),” Journal evidence as to whether L-arginine supplementation has a of Vascular Surgery, vol. 45, no. 1, pp. S5–S67, 2007. clinically significant benefit in cardiovascular disease: Sydow [2]R.H.Boger,¨ S. M. Bode-Boger,W.Thiele,W.Junker,K.¨ and Munzel¨ [50] showed that L-arginine improved radial Alexander, and J. C. Frolich,¨ “Biochemical evidence for artery dilatation, in contrast to Walker et al’s findings impaired nitric oxide synthesis in patients with peripheral that vasodilatation was independent of oral L-arginine arterial occlusive disease,” Circulation, vol. 95, no. 8, pp. 2068– supplementation [51]. In PAD, the Nitric Oxide in Peripheral 2074, 1997. Arterial Insufficiency (NO-PAIN) study randomized 133 [3] R. M. J. Palmer, D. S. Ashton, and S. Moncada, “Vascular endothelial cells synthesize nitric oxide from L-arginine,” patients with intermittent claudication to oral L-arginine Nature, vol. 333, no. 6174, pp. 664–666, 1988. versus placebo for 6 months. Whilst L-arginine supple- [4] J. P. Cooke, “DDAH: a target for vascular therapy?” Vascular mentation increased plasma L-arginine levels, there was no Medicine, vol. 15, no. 3, pp. 235–238, 2010. improvement in NO availability or vascular function. This [5] J. P. Cooke, “Asymmetrical dimethylarginine: the Uber was in contrast to earlier, smaller studies of shorter duration marker?” Circulation, vol. 109, no. 15, pp. 1813–1819, 2004. ff demonstrating a beneficial e ect of L-arginine in PAD [6]U.Forstermann,M.Nakane,W.R.Tracey,andJ.S.Pollock, patients [21]. A mechanism of “arginine-tolerance” was pro- “Isoforms of nitric oxide synthase: functions in the cardiovas- posed, in a similar manner to “nitrate-tolerance”,observed in cular system,” European Heart Journal, vol. 14, no. I, pp. 10–15, prolonged nitroglycerine use. More disappointingly, whilst 1993. both groups in the NO-PAIN study showed improvement [7] B. Torondel, M. Nandi, P. Kelly, B. Wojciak-Stothard, I. Flem- in absolute walking distance, the improvement in the L- ing, and J. Leiper, “Adenoviral-mediated overexpression of arginine group was significantly less than the placebo group DDAH improves vascular tone regulation,” Vascular Medicine, [52]. vol. 15, no. 3, pp. 205–213, 2010. An alternative strategy to increase NO levels may be [8] T. Teerlink, “ADMA metabolism and clearance,” Vascular to reduce ADMA levels by enhancing DDAH activity or Medicine, vol. 10, no. 1, pp. S73–S81, 2005. expression. Nebivolol [53], pioglitazone [54], pravastatin [9] J. M. Leiper, J. Santa Maria, A. Chubb et al., “Identification of [55], and fenofibrate [56] have all been shown to decrease two human dimethylarginine dimethylaminohydrolases with ADMA by upregulating DDAH in vitro. Clinically, Sen et al. distinct tissue distributions and homology with microbial arginine deiminases,” Biochemical Journal, vol. 343, no. 1, pp. demonstrated an increase in circulating NO and a decrease in 209–214, 1999. ADMA, coupled with improvement in symptoms in patients [10] P. Vallance, A. Leone, A. Calver, J. Collier, and S. Moncada, with ischaemic heart disease treated with the selective beta1- “Accumulation of an endogenous inhibitor of nitric oxide adrenergic receptor antagonist nebivolol [47]. Nishiyama et synthesis in chronic renal failure,” Lancet, vol. 339, no. 8793, al. found that the statin, pravastatin reduced serum ADMA pp. 572–575, 1992. in patients with cerebrovascular disease [57]. [11] C. Zoccali, S. M. Bode-Boger,¨ F. Mallamaci et al., “Plasma con- In PAD, upregulation of DDAH at a local level to reduce centration of asymmetrical dimethylarginine and mortality ADMA levels within the ischaemic muscle may be of benefit in patients with end-stage renal disease: a prospective study,” in a number of ways: it may improve microvascular function Lancet, vol. 358, no. 9299, pp. 2113–2117, 2001. and local tissue perfusion; it may improve the results of [12] M. F. Hand, W. G. Haynes, and D. J. Webb, “Hemodialysis therapeutic angiogenesis by reducing the antiangiogenic and L-arginine, but not D-arginine, correct renal failure- environment; or it may improve muscle contractile and associated endothelial dysfunction,” Kidney International, vol. metabolic function and perhaps reduce tissue damage. In 53, no. 4, pp. 1068–1077, 1998. patients with CLI, such strategies may be used as an adjunct [13] T. M. Lu, Y. A. Ding, M. J. Charng, and S. J. Lin, “Asymmetrical to improve the results of revascularisation techniques as well dimethylarginine: a novel risk factor for coronary artery disease,” Clinical Cardiology, vol. 26, no. 10, pp. 458–464, 2003. as reduce tissue damage in patients with non-reconstructable ff disease. They may also be of benefit to patients with earlier [14] H. Worthmann, S. Chen, J. Martens-Lobenho er et al., “High plasma dimethylarginine levels are associated with adverse disease to prevent disease progression. clinical outcome after stroke,” Journal of Atherosclerosis and Thrombosis, vol. 18, no. 9, pp. 753–761, 2011. 7. Conclusion [15] J. T. Kielstein, J. C. Frolich,¨ H. Haller, and D. Fliser, “ADMA (asymmetric dimethylarginine): an atherosclerotic disease There is considerable evidence that disruption of the NO mediating agent in patients with renal disease?” Nephrology pathway plays a major role in the pathogenesis of PAD. Dialysis Transplantation, vol. 16, no. 9, pp. 1742–1745, 2001. 6 Cardiology Research and Practice

[16] J. T. Kielstein, S. M. Bode-Boger,¨ G. Hesse et al., “Asymmetrical [33] J. G. Tidball, M. J. Spencer, M. Wehling, and E. Lavergne, dimethylarginine in idiopathic pulmonary arterial hyperten- “Nitric-oxide synthase is a mechanical signal transducer that sion,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. modulates talin and vinculin expression,” Journal of Biological 25, no. 7, pp. 1414–1418, 2005. Chemistry, vol. 274, no. 46, pp. 33155–33160, 1999. [17]R.Maas,E.Schwedhelm,J.Albsmeier,andR.H.Boger,¨ “The [34] T. W. Balon and J. L. Nadler, “Evidence that nitric oxide pathophysiology of erectile dysfunction related to endothelial increases glucose transport in skeletal muscle,” Journal of dysfunction and mediators of vascular function,” Vascular Applied Physiology, vol. 82, no. 1, pp. 359–363, 1997. Medicine, vol. 7, no. 3, pp. 213–225, 2002. [35] L. Kobzik, B. Stringer, J. L. Balligand, M. B. Reid, and J. [18] J. Leiper, M. Nandi, B. Torondel et al., “Disruption of meth- S. Stamler, “Endothelial type nitric oxide synthase skeletal ylarginine metabolism impairs vascular homeostasis,” Nature muscle fibers: mitochondrial relationships,” Biochemical and Medicine, vol. 13, no. 2, pp. 198–203, 2007. Biophysical Research Communications, vol. 211, no. 2, pp. 375– [19] X. Hu, D. Atzler, X. Xu et al., “Dimethylarginine dimeth- 381, 1995. ylaminohydrolase-1 is the critical enzyme for degrading the [36] W. L. Gross, M. I. Bak, J. S. Ingwall et al., “Nitric oxide inhibits cardiovascular risk factor asymmetrical dimethylarginine,” creatine kinase and regulates rat heart contractile reserve,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 31, no. Proceedings of the National Academy of Sciences of the United 7, pp. 1540–1546, 2011. States of America, vol. 93, no. 11, pp. 5604–5609, 1996. [20] R. Fontaine, M. Kim, and R. Kieny, “Surgical treatment of [37] L. I. Filippin, A. J. Moreira, N. P. Marroni, and R. M. Xavier, peripheral circulation disorders,” Helvetica chirurgica acta, vol. “Nitric oxide and repair of skeletal muscle injury,” Nitric 21, no. 5-6, pp. 499–533, 1954. Oxide, vol. 21, no. 3-4, pp. 157–163, 2009. [21] S. M. Bode-Boger,R.H.B¨ oger,¨ H. Alfke et al., “L-arginine [38] I. Rubinstein, Z. Abassi, R. Coleman, F. Milman, J. Winaver, induces nitric oxide-dependent vasodilation in patients with and O. S. Better, “Involvement of nitric oxide system in exper- critical limb ischemia: a randomized, controlled study,” Circu- imental muscle crush injury,” Journal of Clinical Investigation, lation, vol. 93, no. 1, pp. 85–90, 1996. vol. 101, no. 6, pp. 1325–1333, 1998. [22] M. Ziche and L. Morbidelli, “Nitric oxide and angiogenesis,” [39] J. G. Tidball, “Inflammatory processes in muscle injury and Journal of Neuro-Oncology, vol. 50, no. 1-2, pp. 139–148, 2000. repair,” American Journal of Physiology, vol. 288, no. 2, pp. [23] T. Matsunaga, D. C. Warltier, D. W. Weihrauch, M. Moniz, R345–R353, 2005. J. Tessmer, and W. M. Chilian, “Ischemia-induced coronary [40] R. Tatsumi, A. Hattori, Y. Ikeuchi, J. E. Anderson, and R. E. collateral growth is dependent on vascular endothelial growth Allen, “Release of hepatocyte growth factor from mechanically factor and nitric oxide,” Circulation, vol. 102, no. 25, pp. 3098– stretched skeletal muscle satellite cells and role of pH and 3103, 2000. nitric oxide,” Molecular Biology of the Cell,vol.13,no.8,pp. [24] T. T. Rissanen, I. Vajanto, and S. Yla-Herttuala,¨ “Gene therapy 2909–2918, 2002. for therapeutic angiogenesis in critically ischaemic lower limb [41] H. J. Kaminski and F. H. Andrade, “Nitric oxide: biologic - On the way to the clinic,” European Journal of Clinical effects on muscle and role in muscle diseases,” Neuromuscular Investigation, vol. 31, no. 8, pp. 651–666, 2001. Disorders, vol. 11, no. 6-7, pp. 517–524, 2001. [25] A. Rivard, J. E. Fabre, M. Silver et al., “Age-dependent impair- [42]T.M.Dawson,V.L.Dawson,andS.H.Snyder,“Anovel ment of angiogenesis,” Circulation, vol. 99, no. 1, pp. 111–120, neuronal messenger molecule in brain: the free radical, nitric 1999. oxide,” Annals of Neurology, vol. 32, no. 3, pp. 297–311, 1992. [26] A. Ito, P. S. Tsao, S. Adimoolam, M. Kimoto, T. Ogawa, and [43] P. S. Tsao, N. Aoki, D. J. Lefer, G. Johnson, and A. M. J. P. Cooke, “Novel mechanism for endothelial dysfunction: Lefer, “Time course of endothelial dysfunction and myocardial dysregulation of dimethylarginine dimethylaminohydrolase,” injury during myocardial ischemia and reperfusion in the cat,” Circulation, vol. 99, no. 24, pp. 3092–3095, 1999. Circulation, vol. 82, no. 4, pp. 1402–1412, 1990. [27] M. B. Reid, “Role of nitric oxide in skeletal muscle: synthesis, [44] W. G. Mayhan, S. M. Amundsen, F. M. Faraci, and D. D. distribution and functional importance,” Acta Physiologica Heistad, “Responses of cerebral arteries after ischemia and Scandinavica, vol. 162, no. 3, pp. 401–409, 1998. reperfusion in cats,” American Journal of Physiology, vol. 255, [28] R.C.Hickner,J.S.Fisher,A.A.Ehsani,andW.M.Kohrt,“Role no. 4, pp. H879–H883, 1988. of nitric oxide in skeletal muscle blood flow at rest and during [45] W. C. Sternbergh, R. G. Makhoul, and B. Adelman, “Nitric dynamic exercise in humans,” American Journal of Physiology, oxide-mediated, endothelium-dependent vasodilation is se- vol. 42, no. 1, pp. H405–H410, 1997. lectively attenuated in the postischemic extremity,” Surgery, [29] S. N. A. Hussain, “Activity of nitric oxide synthase in the vol. 114, no. 5, pp. 960–967, 1993. ventilatory muscle vasculature,” Comparative Biochemistry and [46] J. De Alba, A. Cardenas,´ M. A. Moro et al., “Down-regulation Physiology, vol. 119, no. 1, pp. 191–201, 1998. of neuronal nitric oxide synthase by nitric oxide after oxygen- [30] J. Ribera, J. Marsal, A. Casanovas, M. Hukkanen, O. Tarabal, glucose deprivation in rat forebrain slices,” Journal of Neuro- and J. E. Esquerda, “Nitric oxide synthase in rat neuromus- chemistry, vol. 72, no. 1, pp. 248–254, 1999. cular junctions and in nerve terminals of Torpedo electric [47] N. S¸en, Y. Tavil, H. Erdamar et al., “Nebivolol therapy organ: its role as regulator of acetylcholine release,” Journal of improves endothelial function and increases exercise tolerance Neuroscience Research, vol. 51, no. 1, pp. 90–102, 1998. in patients with cardiac syndrome X,” Anadolu Kardiyoloji [31] C. L. Murrant, J. C. Frisbee, and J. K. Barclay, “The effect of Dergisi, vol. 9, no. 5, pp. 371–379, 2009. nitric oxide and endothelin on skeletal muscle contractility [48] J. C. S. Tsui, D. M. Baker, S. G. Shaw, and M. R. Dashwood, changes when stimulation is altered,” Canadian Journal of “Alterations in nitric oxide synthase isoforms in acute lower Physiology and Pharmacology, vol. 75, no. 5, pp. 414–422, 1997. limb ischemia and reperfusion,” Angiology, vol. 58, no. 5, pp. [32] G. Marechal´ and P. Gailly, “Effects of nitric oxide on the 586–592, 2007. contraction of skeletal muscle,” Cellular and Molecular Life [49] L. R. Fiedler, T. Bachetti, J. Leiper et al., “The ADMA/ Sciences, vol. 55, no. 8-9, pp. 1088–1102, 1999. DDAH pathway regulates VEGF-mediated angiogenesis,” Cardiology Research and Practice 7

Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 29, no. 12, pp. 2117–2124, 2009. [50] K. Sydow and T. Munzel,¨ “ADMA and oxidative stress,” Athe- rosclerosis Supplements, vol. 4, no. 4, pp. 41–51, 2003. [51] H. A. Walker, E. McGing, I. Fisher et al., “Endothelium- dependent vasodilation is independent of the plasma L- arginine/ADMA ratio in men with stable angina: lack of effect of oral L-arginine on endothelial function, oxidative stress and exercise performance,” Journal of the American College of Cardiology, vol. 38, no. 2, pp. 499–505, 2001. [52] A. M. Wilson, R. Harada, N. Nair, N. Balasubramanian, and J. P. Cooke, “L-arginine supplementation in peripheral arterial disease: no benefit and possible harm,” Circulation, vol. 116, no. 2, pp. 188–195, 2007. [53] U. Garbin, A. F. Pasini, C. Stranieri, S. Manfro, V. Boccioletti, and L. Cominacini, “Nebivolol reduces asymmetric dimethy- larginine in endothelial cells by increasing dimethylarginine dimethylaminohydrolase 2 (DDAH2) expression and activity,” Pharmacological Research, vol. 56, no. 6, pp. 515–521, 2007. [54] S. Wakino, K. Hayashi, S. Tatematsu et al., “Pioglitazone lowers systemic asymmetric dimethylarginine by inducing dimethylarginine dimethylaminohydrolase in rats,” Hyperten- sion Research, vol. 28, no. 3, pp. 255–262, 2005. [55] Q. F. Yin and Y. Xiong, “Pravastatin restores DDAH activity and endothelium-dependent relaxation of rat aorta after exposure to glycated protein,” Journal of Cardiovascular Phar- macology, vol. 45, no. 6, pp. 525–532, 2005. [56] T. L. Yang, M. F. Chen, B. L. Luo, Q. Y. Xie, J. L. Jiang, and Y. J. Li, “Fenofibrate decreases asymmetric dimethylarginine level in cultured endothelial cells by inhibiting NF-κB activity,” Naunyn-Schmiedeberg’s Archives of Pharmacology, vol. 371, no. 5, pp. 401–407, 2005. [57] Y. Nishiyama, M. Ueda, T. Otsuka et al., “Statin treatment decreased serum asymmetric dimethylarginine (ADMA) levels in Ischemic stroke patients,” Journal of Atherosclerosis and Thrombosis, vol. 18, no. 2, pp. 131–137, 2011. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 486702, 7 pages doi:10.1155/2012/486702

Review Article MicroRNAs in Postischemic Vascular Repair

Andrea Caporali and Costanza Emanueli

Laboratory of Vascular Pathology and Regeneration, Regenerative Medicine Section, School of Clinical Sciences, Bristol Heart Institute, University of Bristol, Bristol, BS2 8HW, UK

Correspondence should be addressed to Costanza Emanueli, [email protected]

Received 14 July 2011; Accepted 6 January 2012

Academic Editor: Daryll M. Baker

Copyright © 2012 A. Caporali and C. Emanueli. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The term angiogenesis describes the growth of endothelial sprouts from preexisting postcapillary venules. More recently, this term has been used to generally indicate the growth and remodeling process of the primitive vascular network into a complex network during development. In adulthood, angiogenesis is activated as a reparative process during wound healing and following ischemia, and it plays a key role in tumor growth and metastasis as well as in inflammatory diseases and diabetic retinopathy. MicroRNAs (miRNAs) are endogenous, small, noncoding RNAs that negatively control gene expression of target mRNAs. In this paper, we aim at describing the role of miRNAs in postischemic angiogenesis. First, we will describe the regulation and the expression of miRNAs in endothelial cells. Then, we will analyze the role of miRNAs in postischemic vascular repair. Finally, we will discuss the role of circulating miRNAs as potential biomarkers in ischemic diseases.

1. Introduction with a fully developed tunica media, which may provide alternative blood perfusion to ischemic areas [6]. microRNAs Postnatal neovascularization occurs in pathological diseases, (miRNAs) are inhibitory regulators of gene expression which including diabetic retinopathy, arthritis, and tumor growth act by binding to complementary messenger RNA (mRNA) and metastasis, as well as during wound healing and transcripts. Following initial studies in developmental biol- postischemic repair [1]. Following occlusion of a major ogy and cancer, miRNAs have recently come into focus ff artery, two di erent types of vascular regrowth responses of cardiovascular diagnostics and therapeutics [7, 8]. Since are activated to contrast the ischemic condition and salvage miRNAs repress many target mRNAs, deregulation of one of injured ischaemic tissue: sprouting of capillaries (angio- single miRNA can result in a cascade of transcriptional genesis) and growth of collateral arteries from preexisting and posttranscriptional changes relevant to disease states arterioles (arteriogenesis) [2]. Angiogenesis is the process of [9]. miRNAs have been detected in serum and plasma, and growth of endothelial sprouts from preexisting postcapillary circulating miRNA profiles have now been associated with venules. It is initiated by the vasodilation of venules leading cancer [10], diabetes [11], and heart disease [12]asan to increased permeability, followed by the proliferation emerging class of biomarkers. Here, we discuss the role of and migrationof endothelial cells (ECs). Then, venules are miRNAs in postischemic angiogenesis. divided by periendothelial cells (intussusception) or are separated by transendothelial cell bridges (bridging) to form capillaries [3, 4]. 2. MicroRNAs Biogenesis Arteriogenesis was originally defined as the process of collateral arteries formation which follows the occlusion Transcriptions of miRNA genes and protein-coding genes of a major artery [5]. Arteriogenesis is partly driven by share common regulatory mechanisms [13]. miRNA genes shear stress and involves the proliferation and migration of can be embedded in the introns of protein coding genes vascular endothelial cells (ECs) and vascular smooth muscle or can derive from their own transcript units in intergenic cells (VSMCs). It leads to the production of mature arteries regions of the genome. When miRNA genes are located 2 Cardiology Research and Practice within introns of protein-coding genes, primary miRNA bio- Table 1: microRNAs and target genes involved in neovascularisa- genesis is controlled by the same transcriptional mechanisms tion. as the parent gene. In contrast, an independent miRNA Proangiogenic Target genes References gene will have its own transcriptional controls. Interestingly, miRNAs multiple miRNAs can be produced within a single long primary nuclear miRNA transcript (pri-miRNA transcript), miR-126 Spred1/PIK3R2 [33] each of which can act independently [14, 15]. The long miR-221/222 c-kit [34] primary nuclear miRNA transcript (“pri-miRNA”) under- miR-378 Fus-1/Sufu [35] goes maturation by the RNase-III Drosha/Dgcr8 enzyme miR-424 CUL2 [36] complex, generating a precursor miRNA (“pre-miRNA”)that is exported from the nucleus by exportin-5 [16]. Cytoplasmic miR-132 p120RasGAP [37] pre-miRNAs are cleaved by a Dicer-containing complex miR-210 EFNA3 [27] [17] to generate a double-strand miRNA-duplex containing miR-130a Spred1/PIK3R2 [38] the mature miRNA on one strand and a complementary miRNA∗ on the other strand. Recently, Dicer-independent miR-296 c-kit [28] miRNA biogenesis pathway has also been reported [18]. miR-17-5p Fus-1/Sufu [39] The specificity of miRNA targeting is defined by Watson- miR-23-27 CUL2 [40] Crick complementarities between positions 2 to 8 from the Antiangiogenic  Target genes References 5 miRNA (also known as the seed sequence), with the miRNAs 3 untranslated region (UTR) of their target mRNAs [15]. Besides the binding through 3 UTR interactions, some miR-200c ZEB1 [41] miRNAs have been shown to associate to the open reading miR-217 SIRT1 [42]  frame or to the 5 UTR of the target genes [19]. miRNAs miR-34 SIRT1 [43] act through formation of stable complexes with proteins miR-503 CCNE1/cdc25A [31] of the Argonaute (Ago) family (in particular with Ago2), the core of the RNA-induced silencing complex (RISC). miR-20a VEGF-A [44] The RISC blocks the translation initiation by competition miR-92a ITGB5 [45] with the cytoplasmic cap-binding protein eIF4E (eukaryotic miRNAs involved in translation initiation factor 4E) [20] or the antiassociation postischemic factor eIF6, thus preventing the assembly of the ribosome angiogenesis [21]. Besides these mechanisms, RISC-miRNA complexes Ischemic stimuli miRNAs References can move the mRNAs they bind to the “P-bodies,” specialized cytoplasmic compartments enriched in mRNA-catabolizing Myocardial infarct miR-126 [33] Myocardial infarct enzymes, where translational repression or exonucleolytic miR-92a [45] mRNA degradation or mRNA destabilization may occur Peripheral ischemia [22, 23]. Myocardial infarct miR-210 [46]. Peripheral ischemia miR-503 [31] 3. Expression and Regulation of Peripheral ischemia miR-92a [45] MicroRNAs in Endothelial Cells Myocardial infarct miR-’100 [47] miRNA microarray analysis is, at the moment, the most used and complete technique to analyse global miRNAs expres- 3.1. Proangiogenic MicroRNAs. miR-126 is recognized as the sion profile. Using this approach, 28 miRNAs (miRNA-15b, most important miRNA for maintaining vascular integrity -16, -20, -21, -23a, -23b, -24, -27a, -29a, -30a, -30c, -31, - during ongoing angiogenesis, as it targets SPRED1 and 100, -103, -106, 125a and -b, -126, -181a, -191, -199a, -221, PIK3R2, two negative regulators of VEGFs signalling [30, -222, -320, let-7, let-7b, let-7c, and let-7d) have been found 33]. Angiogenic sprouting of vessels requires that the zinc to be commonly expressed in ECs cultured under normal finger transcription factor klf2a induces expression of an conditions [24–30]. The miRNA expression profile in ECs endothelial-specific miR-126 [48]. Growth factors increase could change if cells are studied under stressed conditions, the expression of the pro-angiogenic miR-130a and miR- as we recently proved for miRNA-503 (miR-503), which is 296 in ECs [28, 38]. miR-130a stimulates angiogenesis by induced in ECs in response to diabetes and ischemia [31]. A inhibiting GAX and HOXA5; while, miR-296 acts through highly scalable approach using next generation sequencing the inhibition of hepatocyte growth factor (HGF)-regulated (small RNA-seq) would facilitate the analysis of miRNA tyrosine kinase [28]. Moreover, EC stimulation with VEGF- expression patterns in ECs. The group of Martelli was the A and bFGF lead to a rapid transcription of miR-132. first to use this approach to analyze the miRNAs expressed Overexpression of miR-132 increases EC proliferation and by ECs exposed to hypoxia [32]. Based on experimental in vitro networking by targeting p120RasGAP, a GTPase- evidence, it is possible to divide the endothelial miRNAs into activating protein [49]. miR-210 overexpression consider- pro- and antiangiogenic miRNAs (Table 1). ably increased HUVEC migration and the formation of Cardiology Research and Practice 3 capillary-like structures by targeting ephrin A3 [27]. miR- miRNA has multiple targets, modulating the expression 424 promotes angiogenesis by inhibiting cullin 2 (CUL2) and of one miRNA could impact on the complex molecular thereby increasing HIF-1α levels [36]. The proangiogenesis pathways governing ischemic complications by potentially activity of the miR-17–92 cluster is due to miR-17-5p, miR- increasing the expression and/or activity of multiple growth 18a, and miR-19a [26]. In fact, the latter two miRNAs have factors. Hence manipulation of miRNAs could represent a been shown to target proteins containing thrombospondin novel technology potentially able to interfere with expres- type 1 repeats (TSR) [26, 50]; while miR-17-5p appears to sional regulation of multiple genes, thus opening new modulate EC migration and proliferation by targeting tissue avenues for molecular therapeutics (Figure 1). inhibitor of metalloproteinase 1 (TIMP1) [39]. Recently, the miR-23-27-24 cluster has also been reported to have 4.1. Limb Ischemia. Peripheral artery disease (PAD) induces a prominent role in angiogenesis [40]. In particular, both limb muscle hypoperfusion and favours the formation and miR-23 and miR-27 enhance angiogenesis by targeting aggravation of skin ulcers. When the wound healing capacity the anti-angiogenic proteins Sprouty2 and Sema6A [40]. is lost, PAD patients progress to a critical condition, identi- Finally, miR-378 promotes angiogenesis by targeting tumor fied as critical limb ischaemia (CLI), which usually requires suppressor candidate 2 (Fus-1) and suppressor of fused minor or major amputations of the affected limb [56]. (Sufu), thus inducing indirect up-regulation of VEGF and Despite improvements in surgical treatments, a significant Angiopoietin-1/2 [35]. portion of patients with CLI are not eligible for revasculariza- tion, and no medical therapy has been shown to be capable of 3.2. Antiangiogenic MicroRNAs. The antiangiogenic miR- reducing the need for amputation [57]. Diabetes aggravates 221/222 was discovered by Poliseno et al. during the first PAD and CLI. Furthermore, postischemic recovery is delayed screening of miRNAs in HUVECs [34]. Overexpression of in diabetic patients due to diabetes-induced impairment of miR-221/222 inhibits the angiogenesis responses to stem reparative angiogenesis. A laboratory-based study showed cell factor (SCF) by targeting the SCF receptor c-kit [34]. that conditional inactivation of Dicer in ECs reduced the Belonging to miR-17-92 cluster, miR-20a, and miR-92a dis- spontaneous angiogenic responses to limb ischemia, demon- play antiangiogenenic activity, respectively, targeting VEGF- strating that endothelial miRNAs are required for reparative A and integrin β5(ITGB5)transcripts[44, 45]. Oxidative neovascularisation [26]. Recently, miRNA expression in stress upregulates miR-200c in ECs, which translates in murine models of limb ischemia has been profiled, showing the downmodulation of the miR-200c target ZEB1 [41]. upregulation of miR-92a and -100 [47, 58]. miR-92 represses miR-217 and miR-34 regulate the expression of silent the ITGB5 target gene, thus impairing angiogenesis both information regulator 1 (SirT1). Inhibition of either miR- in vitro and in vivo [45]. Conversely, miR-100 functions 217 or miR-34 in ECs promotes angiogenesis via an increase as an endogenous repressor of the serine/threonine protein in SirT1 activity [42, 43]. Finally, miR-503 expression kinase mammalian target of rapamycin (mTOR), increasing in ECs is upregulated under anti-angiogenic conditions EC proliferation, sprouting activity and network formation mimicking diabetes mellitus and ischemia [31]. miR-503 on Matrigel [47]. Moreover, we found miR-503 levels to be targets cdc25A and cyclinE1 (CCNE1) and its overexpression remarkably higher in ischemic muscles of diabetic mice and inhibitsEC proliferation, migration, and network formation of diabetic patients with CLI in comparison to nondiabetic on Matrigel. Conversely, blocking miR-503 activity restores and/or nonischemic controls [31]. In summary, it is evident the angiogenic process in diabetic and ischemic limb mus- that a number of miRNAs are modulated by ischemia and cles. [31]. participate in postischemic angiogenesis.

4. MicroRNAs in Postischemic Angiogenesis 4.2. Acute Myocardial Infarction. Myocardial ischemic injury results from severe impairment of the coronary blood supply Ischaemic disease is highly prevalent and is associated usually produced by thrombosis or other acute alterations with elevated morbidity and mortality in developed and of coronary atherosclerotic plaques. Rapid formation of developing countries [51]. Functional recovery of ischemic collateral vessels that bypass the obstructed coronary artery tissues and organs is dependent on establishing collateral is necessary for the survival of the myocardial region that networks that supply oxygenated blood [52, 53]. Therapeutic surrounds the necrotic infarct tissue and for the subsequent induction of angiogenesis may attenuate ischemic sufferance. myocardial repair process. Notwithstanding, despite the remarkable results in animal The most convincing data in the cardiac response to models of limb ischemia and myocardial infarct, the results ischemia are for miR-126, miR-210, and miR-92a. Neovascu- of 15 years of clinical trials with monotherapeutic proan- larization at 3 weeks postmyocardial infarct (MI) is inhibited giogenic agents have been unsatisfactory [54]. One of the in miR-126 knockout mice in comparison with wild type reasons accounting for this failure is that monotherapy may [33]. Moreover, miR-126 is downregulated in the MI area, be not sufficient to promote and maintain a functional vascu- but upregulated in the MI border zone [33]. In contrast, lar network under severely compromised medical conditions. miR-210 is upregulated in the ischemic myocardium [46]. Thus, one alternative therapeutic approach could consist of miR-210 expression is induced by HIF1α, and it has been drug combinations in order to promote angiogenesis and associated with increased survival of cardiomyocytes and vascular maturation and stabilization [55]. Because each increased angiogenesis in a murine model of myocardial 4 Cardiology Research and Practice

miR-503 inhibition restores post ischemic angiogenesis in ischemic and Anti-miR-503 diabetic limb muscle

miR-92a inhibition enhances myocardial Anti-miR-92a neovascularization after limb ischemia Critical limb ischemia miR-100 inhibition stimulates (CLI) angiogenesis and improve perfusion Anti-miR-100 after femoral artery occlusion

miR-92a inhibition enhances myocardial Anti-miR-92a neovascularization after MI

miR-210 overexpression increases miR-mimics 210 myocardial neovascularization after MI Myocardial infarction (MI) Figure 1: miRNAs-based therapeutic strategies in vascular repairs, There are two major approaches to developing miRNA-based therapeutics: miRNA antagonists and miRNA mimics. miRNA antagonists are generated to inhibit endogenous miRNAs that show a pathogenic gain-of-function in diseased tissues. miRNA mimics are used to upregulate expression of beneficial miRNAs. Inhibition of miR- 92a, miR-100 and miR-503 or increase of miR-210 can promote angiogenesis, thus improving postischemic blood flow recovery in limb ischemia or myocardial infarction. infarction [46]. Finally, miR-92a expression increases after delivery of antagomiR was shown to be sufficient to induce experimental MI and inhibits neovascularisation [45]. efficient and long-lasting miRNA inhibition in vivo and to modulate neovascularization [45, 47, 63]. In particular, in vivo miR-100 inhibition by specific antagomiRs stimulated 4.3. miRNAs-Based Therapeutic Strategies in Vascular Repair. angiogenesis and resulted in functional improvement of There are two major approaches to developing miRNA- perfusion after femoral artery occlusion in mice [47]. Sim- based therapeutics: miRNA antagonists and miRNA mimics. ilarly, antagomiR-induced inhibition of miR-92a enhanced miRNA mimics are small, chemically modified, double- blood vessel repair in murine models of limb ischemia stranded RNA molecules that mimic endogenous mature and MI. [45, 47, 63]. Finally, miRNAs expression could be miRNA molecules. Mimics of proangiogenic miRNAs can inhibited using “miRNA sponge” or “decoy.” Typical sponge be used to elevate angiogenesis in the pathological setting of constructs contain four to ten miRNA binding sites separated ffi insu cient angiogenesis, such as cardiac and limb ischemia. by a few nucleotides each to avoid Ago2-mediated endonu- Caution should be used when overexpressing miRNAs, as cleolytic cleavage [64, 65]. By this technique, we inhibited an above physiological abundance of a miRNA could result miRNA-503 in the ischemic hindlimb of diabetic mice. ff ff in “o -target e ects” through the miRNA binding to seed This anti-miRNA-based intervention restored postischemic regions of genes not normally targeted [8]. miRNAs can be angiogenesis in ischemic and diabetic limb muscles [31]. overexpressed by using naked oligodeoxynucleotides or viral vectors. Recently, gene therapy using a minicircle DNA vector encoding the miR-210 precursor was shown to enhance 5. Circulating MicroRNAs myocardial neovascularization in mice with MI [59]. Anti-miRNAs (antimiRs) are modified antisense oli- The discovery of miRNAs in biological fluids is opening godeoxynucleotides harboring the full or partial comple- new frontiers in the miRNA arena. Mitchell et al. first mentary reverse sequence of a mature miRNA. These oli- demonstrated that circulating miRNAs could be detected godeoxynucleotides are able to reduce the endogenous levels reliably in serum and showed that miRNA profiles correlate of the miRNA, thus increasing expression of its mRNA with specific diseases. Moreover, miRNAs appear protected targets [60]. Most common chemical modifications to from RNAse and remain stable in the blood [66]. Therefore, increase antimiR stability are LNA- (locked nucleic acid-) their stability makes miRNAs concentrations well suited for modified nucleotides, in which the 2-O-oxygen is locked being tested in patient samples as potential biomarkers of into a C3-endo (RNA) sugar conformation [61]ora different pathological conditions. Recent studies identified complete 2-O-methylation of sugar in a phosphorothioate miRNAs in different types of microvescicles (MVs) secreted backbone, and a cholesterol molecule at the 3 end. These from cultured cells, thus supporting the idea that MVs may are the so-called “antagomiRs” [62]. LNA-antimiR against serveasphysiologicalcarriersofmiRNA[11, 67, 68]. In line miRNAs of the cluster miR-23/24 showed regulation of with this hypothesis, MVs released from cells were proved angiogenesis and choroidal neovascularisation [40]. Systemic able to alter gene expression in neighbouring and distant cells Cardiology Research and Practice 5 in vitro and in vivo by transferring miRNAs [69, 70]. Two to translate the findings into clinical biomarker. Then, in distinct processes of MVs release from the cells have been order to obtain reliable and reproducible results, there is described. MVs may derive from the endosomal membranes a need to determine suitable normalization methods for that are extruded from the cell surface of activated cells blood-based miRNA investigations. Finally, the mechanisms as exosomes [71]. Furthermore, high-density lipoprotein by which miRNAs are released into the circulation and (HDL) transports endogenous miRNAs and delivers them whether circulating miRNA molecules have any functional to recipient cells with functional targeting capabilities [72]. capabilities need further characterization. Interestingly, a recent study demonstrated that the majority of circulating miRNAs are complexed with protein rather References than with vesicles. In particular, Ago2 was present in human plasma and eluted with plasma miRNAs. This finding raises [1] M. Boodhwani and F. W. Munir, “Therapeutic angiogenesis the possibility that cells release a functional RISC complex in diabetes and hypercholesterolemia: influence of oxidative into the circulation [73]. stress,” Antioxidants and Redox Signaling,vol.11,no.8,pp. Recent reports suggest that serum levels of cardiac 1945–1959, 2009. miRNA correlate with the severity of heart injury and may [2] C. Emanueli and P. Madeddu, “Therapeutic angiogenesis: represent a good biomarker of MI. Corsten et al. described translating experimental concepts to medically relevant goals,” strikingly increased plasma levels of miR-208b and -499 after Vascular Pharmacology, vol. 45, no. 5, pp. 334–339, 2006. MI. However, studies on circulating miRNAs and MI have led [3] R. H. Adams and K. Alitalo, “Molecular regulation of angio- genesis and lymphangiogenesis,” Nature Reviews Molecular to contrasting conclusions on miR-208. Wang et al. suggested Cell Biology, vol. 8, no. 6, pp. 464–478, 2007. miR-208 as a candidate biomarker for MI in humans [74], [4] P. H. Burri, R. Hlushchuk, and V. Djonov, “Intussusceptive while Adachi et al. found very low levels of miR-208a angiogenesis: its emergence, its characteristics, and its signifi- and miR-208b in human hearts [75]. The latter result was cance,” Developmental Dynamics, vol. 231, no. 3, pp. 474–488, confirmed by D’Alessandra et al., who reported that miR- 2004. 208 was detectable only in 30% of MI patients and always [5] D. Scholz, W. J. Cai, and W. Schaper, “Arteriogenesis, a new at very low levels [76]. In contrast, miR-499 was detectable concept of vascular adaptation in occlusive disease,” Angiogen- in all control subjects and increased in all patients with MI esis, vol. 4, no. 4, pp. 247–257, 2001. [76]. Finally, we found that the plasma levels of miR-503 [6]N.Smart,K.N.Dube,´ and P. R. Riley, “Coronary vessel were dramatically increased in diabetic patients with CLI in development and insight towards neovascular therapy,” Inter- comparison with controls. Due to the low number of patients national Journal of Experimental Pathology, vol. 90, no. 3, pp. in the study, further analysis is required to investigate miR- 262–283, 2009. 503 as a potential circulating biomarker of ongoing ischemia [7]E.M.SmallandE.N.Olson,“PervasiverolesofmicroRNAs in diabetic subjects [31]. in cardiovascular biology,” Nature, vol. 469, no. 7330, pp. 336– 342, 2011. [8] E. Van Rooij, “The art of MicroRNA research,” Circulation Research, vol. 108, no. 2, pp. 219–234, 2011. 6. Future Research [9] M. Inui, G. Martello, and S. Piccolo, “MicroRNA control of signal transduction,” Nature Reviews Molecular Cell Biology, Great interest focuses on exploiting miRNAs for therapeutic vol. 11, no. 4, pp. 252–263, 2010. purposes and as clinical biomarkers. As therapeutic targets [10] G. Reid, M. B. Kirschner, and N. van Zandwijk, “Circulating for vascular diseases, miRNAs could represent a major break- microRNAs: association with disease and potential use as through because a single miRNA can regulate several target biomarkers,” Critical Reviews in Oncology/Hematology, vol. 80, genes, thus influencing multiple molecular pathways. In no. 2, pp. 193–208, 2011. consideration of the above, miRNAs antagonists or mimics [11] A. Zampetaki, S. Kiechl, I. Drozdov et al., “Plasma MicroRNA could become an important new class of drugs to regulate profiling reveals loss of endothelial MiR-126 and other angiogenesis. Moreover, manipulation of miRNAs structures MicroRNAs in type 2 diabetes,” Circulation Research, vol. 107, might increase their delivery efficiency. In particular, the no. 6, pp. 810–817, 2010. promising results obtained with modified microparticles or [12] S. K. Gupta, C. Bang, and T. Thum, “Circulating MicroRNAs nanoparticles for drug delivery in cancer [37] should be as biomarkers and potential paracrine mediators of cardiovas- extended to investigate their capacity for promoting vascular cular disease,” Circulation, vol. 3, no. 5, pp. 484–488, 2010. regeneration in ischemic tissues. A potential drawback in [13] B. N. Davis and A. Hata, “Regulation of microRNA biogenesis: the therapeutic use of miRNAs is their possible “off-target a miRiad of mechanisms,” Cell Communication and Signaling, vol. 7, p. 18, 2009. effects.” Since not all targets of a particular miRNA are [14] Y. Zeng and B. R. Cullen, “Recognition and cleavage of pri- disease-related, any therapeutic perturbation of miRNA mary microRNA transcripts,” Methods in Molecular Biology, expression will likely have side effects. ff vol. 342, pp. 49–56, 2006. As biomarkers, circulating miRNAs o er a great potential [15] D. P. Bartel, “MicroRNAs: genomics, biogenesis, mechanism, ff for the diagnosis and prognosis of di erent diseases. How- and function,” Cell, vol. 116, no. 2, pp. 281–297, 2004. ever, there are still limitations in the technology and various [16] R. Yi, Y. Qin, I. G. Macara, and B. R. Cullen, “Exportin- issues regarding miRNA measurement and quantification 5 mediates the nuclear export of pre-microRNAs and short in the circulation. First, circulating miRNA profiling has hairpin RNAs,” Genes and Development, vol. 17, no. 24, pp. to be verified in larger and independent studies in order 3011–3016, 2003. 6 Cardiology Research and Practice

[17] G. Hutvagner,´ J. McLachlan, A. E. Pasquinelli, E. Balint,´ T. [33] S. Wang, A. B. Aurora, B. A. Johnson et al., “The endothelial- Tuschl, and P. D. Zamore, “A cellular function for the RNA- specific microRNA miR-126 governs vascular integrity and interference enzyme dicer in the maturation of the let-7 small angiogenesis,” Developmental Cell, vol. 15, no. 2, pp. 261–271, temporal RNA,” Science, vol. 293, no. 5531, pp. 834–838, 2001. 2008. [18] D. Cifuentes, H. Xue, D. W. Taylor et al., “A novel miRNA [34] L. Poliseno, A. Tuccoli, L. Mariani et al., “MicroRNAs mod- processing pathway independent of dicer requires argonaute2 ulate the angiogenic properties of HUVECs,” Blood, vol. 108, catalytic activity,” Science, vol. 328, no. 5986, pp. 1694–1698, no. 9, pp. 3068–3071, 2006. 2010. [35] D. Y. Lee, Z. Deng, C. H. Wang, and B. B. Yang, “MicroRNA- [19] U. A. Ørom, F. C. Nielsen, and A. H. Lund, “MicroRNA-10a 378 promotes cell survival, tumor growth, and angiogenesis binds the 5UTR of ribosomal protein mRNAs and enhances by targeting SuFu and Fus-1 expression,” Proceedings of the their translation,” Molecular Cell, vol. 30, no. 4, pp. 460–471, National Academy of Sciences of the United States of America, 2008. vol. 104, no. 51, pp. 20350–20355, 2007. [20] M.Kiriakidou,G.S.Tan,S.Lamprinaki,M.DePlanell-Saguer, [36] G. Ghosh, I. V. Subramanian, N. Adhikari et al., “Hypoxia- P.T. Nelson, and Z. Mourelatos, “An mRNA m7G cap binding- induced microRNA-424 expression in human endothelial cells like motif within human Ago2 represses translation,” Cell, vol. regulates HIF-α isoforms and promotes angiogenesis,” Journal 129, no. 6, pp. 1141–1151, 2007. of Clinical Investigation, vol. 120, no. 11, pp. 4141–4154, 2010. [21] T. P.Chendrimada, K. J. Finn, X. Ji et al., “MicroRNA silencing [37] S. Anand, B. K. Majeti, L. M. Acevedo et al., “MicroRNA-132- through RISC recruitment of eIF6,” Nature, vol. 447, no. 7146, mediated loss of p120RasGAP activates the endothelium to pp. 823–828, 2007. facilitate pathological angiogenesis,” Nature Medicine, vol. 16, [22] A. Eulalio, I. Behm-Ansmant, and E. Izaurralde, “P bodies: no. 8, pp. 909–914, 2010. at the crossroads of post-transcriptional pathways,” Nature [38] Y. Chen and D. H. Gorski, “Regulation of angiogenesis Reviews Molecular Cell Biology, vol. 8, no. 1, pp. 9–22, 2007. through a microRNA (miR-130a) that down-regulates antian- [23] R. Parker and U. Sheth, “P bodies and the control of mRNA giogenic homeobox genes GAX and HOXA5,” Blood, vol. 111, translation and degradation,” Molecular Cell,vol.25,no.5,pp. no. 3, pp. 1217–1226, 2008. 635–646, 2007. [39] M. Otsuka, M. Zheng, M. Hayashi et al., “Impaired microRNA [24] A. Kuehbacher, C. Urbich, A. M. Zeiher, and S. Dimmeler, processing causes corpus luteum insufficiency and infertility “Role of Dicer and Drosha for endothelial microRNA expres- in mice,” Journal of Clinical Investigation, vol. 118, no. 5, pp. sion and angiogenesis,” Circulation Research, vol. 101, no. 1, 1944–1954, 2008. pp. 59–68, 2007. [40] Q. Zhou, R. Gallagher, R. Ufret-Vincenty, X. Li, E. N. Olson, [25] Y. Suarez,´ C. Fernandez-Hernando,´ J. S. Pober, and W. C. and S. Wang, “Regulation of angiogenesis and choroidal neo- Sessa, “Dicer dependent microRNAs regulate gene expres- vascularization by members of microRNA-23∼27∼24 clus- sion and functions in human endothelial cells,” Circulation ters,” Proceedings of the National Academy of Sciences of the Research, vol. 100, no. 8, pp. 1164–1173, 2007. United States of America, vol. 108, no. 20, pp. 8287–8292, 2011. [26] Y. Suarez,´ C. Fernandez-Hernando,´ J. Yu et al., “Dicer- [41] A. Magenta, C. Cencioni, P. Fasanaro et al., “MiR-200c is dependent endothelial microRNAs are necessary for postnatal upregulated by oxidative stress and induces endothelial cell angiogenesis,” Proceedings of the National Academy of Sciences apoptosis and senescence via ZEB1 inhibition,” Cell Death and of the United States of America, vol. 105, no. 37, pp. 14082– Differentiation, vol. 18, no. 10, pp. 1628–1639, 2011. 14087, 2008. [42] R. Menghini, V. Casagrande, M. Cardellini et al., “MicroRNA [27] P. Fasanaro, Y. D’Alessandra, V. Di Stefano et al., “MicroRNA- 217 modulates endothelial cell senescence via silent informa- 210 modulates endothelial cell response to hypoxia and tion regulator 1,” Circulation, vol. 120, no. 15, pp. 1524–1532, inhibits the receptor tyrosine kinase ligand ephrin-A3,” Jour- 2009. nal of Biological Chemistry, vol. 283, no. 23, pp. 15878–15883, [43] T. Zhao, J. Li, and A. F. Chen, “MicroRNA-34a induces 2008. endothelial progenitor cell senescence and impedes its angio- [28] T. Wurdinger,¨ B. A. Tannous, O. Saydam et al., “miR-296 reg- genesis via suppressing silent information regulator 1,” Ameri- ulates growth factor receptor overexpression in angiogenic can Journal of Physiology, vol. 299, no. 1, pp. E110–E116, 2010. endothelial cells,” Cancer Cell, vol. 14, no. 5, pp. 382–393, [44] Z. Hua, Q. Lv, W. Ye et al., “Mirna-directed regulation of 2008. VEGF and other angiogenic under hypoxia,” PLoS ONE, vol. [29] C. Chen, H. Chai, X. Wang et al., “Soluble CD40 ligand 1, no. 1, article e116, 2006. induces endothelial dysfunction in human and porcine coro- [45] A. Bonauer, G. Carmona, M. Iwasaki et al., “MicroRNA-92a nary artery endothelial cells,” Blood, vol. 112, no. 8, pp. 3205– controls angiogenesis and functional recovery of ischemic 3216, 2008. tissues in Mice,” Science, vol. 324, no. 5935, pp. 1710–1713, [30]J.E.Fish,M.M.Santoro,S.U.Mortonetal.,“miR-126 2009. regulates angiogenic signaling and vascular integrity,” Devel- [46] S. Hu, M. Huang, Z. Li et al., “MicroRNA-210 as a novel ther- opmental Cell, vol. 15, no. 2, pp. 272–284, 2008. apy for treatment of ischemic heart disease,” Circulation, vol. [31]A.Caporali,M.Meloni,C.Vollenkle¨ et al., “Deregulation 122, no. 11, pp. S124–S131, 2010. of microRNA-503 contributes to diabetes mellitus-induced [47] S. Grundmann, F. P. Hans, S. Kinniry et al., “MicroRNA-100 impairment of endothelial function and reparative angiogen- regulates neovascularization by suppression of mammalian esis after Limb Ischemia,” Circulation, vol. 123, no. 3, pp. 282– target of rapamycin in endothelial and vascular smooth mus- 291, 2011. cle cells,” Circulation, vol. 123, no. 9, pp. 999–1009, 2011. [32] C. Voellenkle, J. van Rooij, A. Guffanti et al., “Deep- [48] S. Nicoli, C. Standley, P. Walker, A. Hurlstone, K. E. Fogarty, sequencing of endothelial cells exposed to hypoxia reveals the and N. D. Lawson, “MicroRNA-mediated integration of hae- complexity of known and novel microRNAs,” RNA, vol. 18, modynamics and Vegf signalling during angiogenesis,” Nature, no. 3, pp. 472–484, 2012. vol. 464, no. 7292, pp. 1196–1200, 2010. Cardiology Research and Practice 7

[49] S. Anand, B. K. Majeti, L. M. Acevedo et al., “MicroRNA-132- United States of America, vol. 105, no. 30, pp. 10513–10518, mediated loss of p120RasGAP activates the endothelium to 2008. facilitate pathological angiogenesis,” Nature Medicine, vol. 16, [67] M. P. Hunter, N. Ismail, X. Zhang et al., “Detection of no. 8, pp. 909–914, 2010. microRNA expression in human peripheral blood microvesi- [50] M. Dews, A. Homayouni, D. Yu et al., “Augmentation of tumor cles,” PLoS ONE, vol. 3, no. 11, Article ID e3694, 2008. angiogenesis by a Myc-activated microRNA cluster,” Nature [68] F. Collino, M. C. Deregibus, S. Bruno et al., “Microvesicles Genetics, vol. 38, no. 9, pp. 1060–1065, 2006. derived from adult human bone marrow and tissue specific [51] E. Selvin and T. P. Erlinger, “Prevalence of and risk factors for mesenchymal stem cells shuttle selected pattern of miRNAs,” peripheral arterial disease in the United States: results from PLoS ONE, vol. 5, no. 7, Article ID e11803, 2010. the National Health and Nutrition Examination Survey, 1999- [69] Y. Zhang, D. Liu, X. Chen et al., “Secreted monocytic miR-150 2000,” Circulation, vol. 110, no. 6, pp. 738–743, 2004. enhances targeted endothelial cell migration,” Molecular Cell, [52] S. Takeshita, L. Q. Pu, L. A. Stein et al., “Intramuscular admin- vol. 39, no. 1, pp. 133–144, 2010. istration of vascular endothelial growth factor induces dose- [70] A. Zernecke, K. Bidzhekov, H. Noels et al., “Delivery dependent collateral artery augmentation in a rabbit model of of microRNA-126 by apoptotic bodies induces CXCL12- chronic limb ischemia,” Circulation, vol. 90, no. 5, pp. II228– dependent vascular protection,” Science Signaling, vol. 2, no. II234, 1994. 100, article ra81, 2009. [53] C. Emanueli, M. B. Salis, A. Pinna, G. Graiani, L. Manni, and [71] C. Thery,´ L. Zitvogel, and S. Amigorena, “Exosomes: compo- P. Madeddu, “Nerve growth factor promotes angiogenesis and sition, biogenesis and function,” Nature Reviews Immunology, arteriogenesis in ischemic hindlimbs,” Circulation, vol. 106, vol. 2, no. 8, pp. 569–579, 2002. no. 17, pp. 2257–2262, 2002. [72] K. C. Vickers, B. T. Palmisano, B. M. Shoucri, R. D. Shambu- [54] T. D. Henry, B. H. Annex, G. R. McKendall et al., “The VIVA rek, and A. T. Remaley, “MicroRNAs are transported in plasma trial: vascular endothelial growth factor in ischemia for vascu- and delivered to recipient cells by high-density lipoproteins,” lar angiogenesis,” Circulation, vol. 107, no. 10, pp. 1359–1365, Nature Cell Biology, vol. 13, no. 4, pp. 423–435, 2011. 2003. [73]J.D.Arroyo,J.R.Chevillet,E.M.Krohetal.,“Argonaute2 [55] C. Emanueli and P. Madeddu, “Changing the logic of thera- complexes carry a population of circulating microRNAs peutic angiogenesis for ischemic disease,” Trends in Molecular independent of vesicles in human plasma,” Proceedings of the Medicine, vol. 11, no. 5, pp. 207–216, 2005. National Academy of Sciences of the United States of America, [56] D. P. Slovut and T. M. Sullivan, “Critical limb ischemia: vol. 108, no. 12, pp. 5003–5008, 2011. medical and surgical management,” Vascular Medicine, vol. 13, [74] G. K. Wang, J. Q. Zhu, J. T. Zhang et al., “Circulating no. 3, pp. 281–291, 2008. microRNA: a novel potential biomarker for early diagnosis [57] L. Norgren, W. R. Hiatt, J. A. Dormandy et al., “Inter-Society of acute myocardial infarction in humans,” European Heart Consensus for the management of peripheral arterial disease Journal, vol. 31, no. 6, pp. 659–666, 2010. (TASC II),” International Angiology, vol. 26, no. 2, pp. 82–157, [75] T. Adachi, M. Nakanishi, Y. Otsuka et al., “Plasma microRNA 2007. 499 as a biomarker of acute myocardial infarction,” Clinical [58] S. Greco, M. De Simone, C. Colussi et al., “Common micro- Chemistry, vol. 56, no. 7, pp. 1183–1185, 2010. RNA signature in skeletal muscle damage and regenera- [76] Y. D’Alessandra, P. Devanna, F. Limana et al., “Circulating tion induced by Duchenne muscular dystrophy and acute microRNAs are new and sensitive biomarkers of myocardial ischemia,” FASEB Journal, vol. 23, no. 10, pp. 3335–3346, 2009. infarction,” European Heart Journal, vol. 31, no. 22, pp. 2765– [59] S. Hu, M. Huang, Z. Li et al., “MicroRNA-210 as a novel 2773, 2010. therapy for treatment of ischemic heart disease,” Circulation, vol. 122, no. 11, pp. S124–S131, 2010. [60] J. Stenvang and S. Kauppinen, “MicroRNAs as targets for antisense-based therapeutics,” Expert Opinion on Biological Therapy, vol. 8, no. 1, pp. 59–81, 2008. [61] T. P. Prakash and B. Bhat, “2-modified oligonucleotides for antisense therapeutics,” Current Topics in Medicinal Chemistry, vol. 7, no. 7, pp. 641–649, 2007. [62] J. Krutzfeldt,¨ N. Rajewsky, R. Braich et al., “Silencing of microRNAs in vivo with “antagomirs”,” Nature, vol. 438, no. 7068, pp. 685–689, 2005. [63] C. Van Solingen, L. Seghers, R. Bijkerk et al., “Antagomir- mediated silencing of endothelial cell specific microRNA-126 impairs ischemia-induced angiogenesis,” Journal of Cellular and Molecular Medicine, vol. 13, no. 8 A, pp. 1577–1585, 2009. [64] B. Gentner, G. Schira, A. Giustacchini et al., “Stable knock- down of microRNA in vivo by lentiviral vectors,” Nature Methods, vol. 6, no. 1, pp. 63–66, 2009. [65] M. S. Ebert, J. R. Neilson, and P. A. Sharp, “MicroRNA sponges: competitive inhibitors of small RNAs in mammalian cells,” Nature Methods, vol. 4, no. 9, pp. 721–726, 2007. [66] P. S. Mitchell, R. K. Parkin, E. M. Kroh et al., “Circulating microRNAs as stable blood-based markers for cancer detec- tion,” Proceedings of the National Academy of Sciences of the Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 496904, 8 pages doi:10.1155/2012/496904

Review Article Connexins and Diabetes

JosephineA.Wright,TobyRichards,andDavidL.Becker

Department of Vascular Surgery and Department of Cell and Developmental Biology, University College London Hospital, 74 Huntley Street, London NW1 2BU, UK

Correspondence should be addressed to Josephine A. Wright, [email protected]

Received 30 July 2011; Accepted 29 November 2011

Academic Editor: Janice Tsui

Copyright © 2012 Josephine A. Wright et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Cell-to-cell interactions via gap junctional communication and connexon hemichannels are involved in the pathogenesis of diabetes. Gap junctions are highly specialized transmembrane structures that are formed by connexon hemichannels, which are further assembled from proteins called “connexins.” In this paper, we discuss current knowledge about connexins in diabetes. We also discuss mechanisms of connexin influence and the role of individual connexins in various tissues and how these are affected in diabetes. Connexins may be a future target by both genetic and pharmacological approaches to develop treatments for the treatment of diabetes and its complications.

1. Introduction described in terms of molecular mass (Cx43 represents the Cx protein of 43 kDa). Six Cx units oligomerize in the en- Morbidity and mortality caused by diabetes is of serious doplasmic reticulum to form an annular hemichannel or concern worldwide. Global prevalence of diabetes is expected “connexon.” Cxs can interact structurally in various ways, to increase from 7.7% to 439 million by 2030 [1]. Diabetes and connexons are regarded as homeric if the Cxs are of the is a disorder characterised by multiple defects at both the same type or heteromeric if they are of different types. Sim- cellular and molecular levels in a wide variety of tissues. ilarly, the gap junctional channels may be formed by homeric Resulting tissue damage in diabetes is associated with inter- or heteromeric connexons (see Figure 1). When the extracel- actions and “crosstalk” between inflammatory and metabolic lular portion of a connexon docks with connexon from an pathological processes such as endothelial dysfunction (via opposing cell, a functional gap junction channel forms. VCAM-1, ICAM-1, E-selection, vWF), procoagulation (PAI- Tissue homeostasis is dependent on a complex system α 1, fibronectin, and P-selectin), and inflammation (TNF ,IL- of cell communication over long distances by endocrine β 6, IL-1 ,IL-18,andCRP)[2]. Cell-to-cell interactions, via signalling, short distances via membrane signalling path- gap junctional communication and connexon hemichannels, ways (ion channels, G-protein coupled or tyrosine-kinase are now thought to play an important role in the molecular receptors), and by cell-to-cell contact. Gap junctions allow biology of diabetes. In this paper, we discuss current knowl- a unique form of communication; they mediate intracel- edge about connexins (Cxs) and their role in various aspects lular signals by directly connecting adjacent cells. Their of diabetes. purpose is to closely regulate the intracellular passage of ions, nutrients, metabolites, second messengers, and small 2. Gap Junctions, Connexons, and Connexins biological molecules (up to 1000 Da). The multiplicity of isoforms and different structural combinations determine Gap junctions are highly specialized structures, made up of the biophysical permeation characteristics of a gap junction channels spanning adjacent cell membranes, leaving a 2– channel. Aside from forming gap junctions, the connexon 4 nm extracellular “gap,” hence their name [3, 4]. In hemichannels also provide a pathway for the release from vertebrates, gap junction channels are assembled of trans- cellsofATP,glutamate,NAD+, and prostaglandin E2, which membrane proteins called “connexins.” Cxs are commonly act as paracrine messengers [5]. A dynamic cell-to-cell 2 Cardiology Research and Practice

Connexin In the sections below, we will outline the various Cxs Connexon associated with diabetes and its complications and discuss hemichannel relevant studies (both animal and human) in the current Adjacent cell literature. Gap junction membranes channel 3.1. Diabetes. Connexins are associated with the pathogen- esis of both type I and type II diabetes (see Table 1)[12]. Gap junction Cx36 forms the gap junctions between pancreatic beta cells Heteromeric Homomeric which are required for normal basal insulin secretion and hemichannel hemichannel glucose-induced insulin release. Animal studies have shown that loss of Cx36 is associated with the loss of pulsatile Figure 1: Multiple levels of gap junction channel structure. Individ- insulin release, increase in basal insulin output, and reduced ual transmembrane connexins (coloured blue and orange) assemble glucose-induced insulin release (these secretory defects are into hexamers called “connexon hemichannels.” Connexons from two adjacent cells join to make a channel across two plasma similar to those observed in type II diabetes) [13, 14]. The membranes (green), forming a “gap junction channel.” exactroleofCx36intypeIIdiabetesinhumansisyet to be elicited. Cx36 is coded for by GJD2 gene, which is located on the 14q region of chromosome 15, a susceptibility locus for type II diabetes, and the diabetic syndrome [15]. communication is advantageous as it avoids the transport Cx36 is a native protein of human pancreatic islets that mechanisms, ligand-receptor interactions, and dilutional mediates the coupling of beta cells through preferential effects associated with extracellular communication. A key exchange of cationic molecules and contributes to the control function of gap junctions is sharing of metabolic demands of beta-cell function by modulating gene expression [16]. across cells and tissues [6]. They are also of critical impor- Endogenous glucose production is a major determinant tanceincelldifferentiation, adhesion, and apoptosis, and of fasting hyperglycaemia in type II diabetes [17]. Cx36 they have regulatory roles in embryonic development, synap- upregulation is associated with an increase in endogenous tic transmission, immune response, and carcinogenesis. glucose production, predominantly of hepatic origin. Both Cxs are now known to be multifaceted proteins that not Cx32 and Cx26 can form connexon hemichannels connect- only link to structural gap junctional complexes, but also to ing liver hepatocytes. In animal studies, these connexon adhesion molecules, tight junctions, cytoskeletal elements, hemichannels are responsible for the regulation of blood and cellular machinery, which facilitates their transport, glucose and hepatocyte glycogenolysis [18, 19]. Following assembly, function, and internalization. Initially, these gap electrical stimulation of sympathetic liver nerves, Cx32- junction protein complexes were termed the “Nexus” [7], deficient mice had inefficient mobilisation of glucose from which was the original structural name for the intracellular hepatic glycogen stores [20]. The mechanism underlying links. This term has come back into use, alongside the term this dysregulation in hepatocytes may involve interaction of “gap junction proteome” [8]. Once the gap junction channel Cxs with second messengers as Cx32-transfected HeLA cells assembly has nucleated and further binding partners are had increased transmission of inositol 1,4,5-trisphosphate- engaged, additional gap junction channels cluster and form mediated calcium waves to neighbouring cells [21]. “gap junction plaques.” Cxs are a homogenous family of proteins ubiquitously expressed in humans, and to date, 21 Clinically, the complications of diabetes are vast. Due to Cx genes have been reported in humans, and 20 in rodents. the almost ubiquitous tissue distribution of Cxs, there has been interest in Cx changes in both diabetic macrovascular and microvascular disease. 3. Clinical Connexins Since the first description of a mutation in the GJB1 gene 3.2. Macrovascular Complications of Diabetes. Cardiovascu- coding for Cx32 in patients with the X-linked Charcot Marie lar complications are the leading cause of morbidity and tooth syndrome in 1993, disease-causing mutations have mortality in diabetes. Type II diabetes is an independent risk been detected in almost every Cx gene [9, 10]. Patients factor for major large vessel atherosclerosis. In this section, with the X-linked Charcot Marie tooth have an inherited we will discuss the association of Cxs with atherosclerosis, progressive peripheral neuropathy, characterised by distal dyslipidemia, peripheral vascular disease, coronary artery symmetrical neuropathy, muscle weakness, and sensory loss. disease, hypertension, and arrhythmias. Within myelinating Schwann cells, Cx32 has been localized Changes in the expression pattern of vascular Cxs (Cx37, to the paranodal regions and Schmidt-Lantermann incisures, Cx40, Cx43, and Cx45) have been observed in animal model which are found between the inner and outer layers of the studies during the formation of atherosclerotic plaques [22]. coiled up cell. Connexon hemichannels composed of Cx32 In atherosclerosis, Cxs may contribute to the inflammatory therefore enable the rapid transport of small molecules and response through adhesion and transmigration of inflamma- nutrients in a reflexive manner within individual Schwann tory cells. Cx37 is thought to function in monocyte adhe- cells. Although Cx32 is also found outside the peripheral ner- sion and transmigration across the endothelium, whereas vous system in major organs like the liver, Cx32 mutations endothelial Cx40 is implicated in transmitting antiadhesive are associated only with X-linked Charcot Marie tooth [11]. signals within the vascular endothelium [23]. Cardiology Research and Practice 3

Table 1: Connexin changes associated with diabetic complications (upward and downward arrows indicate increase or decrease in Cx levels, VECs: vascular endothelial cells, SMCs: smooth muscle cells).

Diabetic complication Connexin changes associated with diabetes Type of study References Macrovascular Human and animal Atherosclerosis Early atheroma ↓Cx37 VECs and circulating monocytes [22] diabetic studies Advanced atheroma ↑Cx37 SMCs, ↓Cx40 VECs Early atheroma ↑then↓Cx43 SMCs Advanced atheroma ↑Cx37, ↑Cx43 macrophage foam cells Human and animal Dyslipidemia ↓Cx37, ↓Cx40 VECs [27, 28] diabetic STZ rat studies ↑Cx43 VECs and SMCs Animal studies (e.g. Hypertension ↑Cx40 VECs [35–37] Cx40−/− mice) ↑Cx37, 43, 45 SMCs Animal diabetic STZ rat Arrhythmias Ventricular myocytes ↑Cx43 (both total and phosphorylated forms) [39–41, 43] studies Atrial myocytes (↑Cx43 but ↓phosphorylated form) Sinoatrial node ↑Cx45>↑ Cx43 and Cx45 Microvascular Animal diabetic STZ and Nephropathy ↓Cx37 renin secreting cells [44–49] ZDF rat studies ↑Cx40 SMCs preglomerular arterioles ↓Cx43 SMC postglomerular arterioles and mesangial cells ↑Cx43 collecting ducts, podocytes Human diabetic studies [50, 51] Retinopathy ↓Cx43 retinal pericytes, microvascular endothelial cells Human diabetic cell studies [54–58] Erectile dysfunction ↓Cx43 penile corpora Animal STZ diabetic rat [59–61] Detrusor overactivity ↑Cx43bladderdetrusormuscle studies Animal diabetic STZ rat Peripheral neuropathy ↑Cx26, ↑Cx32 perineurium [62] studies Animal diabetic STZ rat Wound healing ↓Cx43, ↓Cx26 epidermis [70, 71] studies ↑Cx43 dermis, wound edge keratinocytes first 24 hours after injury ↑Cx43 wound edge keratinocytes Human studies [72]

The in vivo effect of diabetes on the expression pattern and Cx40 and human observational data. An association of vascular Cxs within major large vessels is not fully under- between Cx37 polymorphism and peripheral arterial disease stood. The effects of statins on connexin expression in the in patients with type II diabetes has been observed [27]. Ani- arterial wall have been observed by Sheu et al. [24], who mal studies have also shown that treatment of dyslipidemia observed reduced expression of Cx43 in diabetic rat aortic reduces Cx levels. Dlugosova´ et al. [28] identified Cx43 in the walls and an increase in Cx43 levels following simvastatin aorta (in both the tunica media and to a lesser degree in the treatment. In contrast, Hou et al. [25] found that in apoE- endothelium) of adult hereditary hypertriglyceridemia rats, deficient mice, STZ-induced diabetes was associated with an and that treatment with omega-3 polyunsaturated fatty acids increase in the burden of atheroma and downregulation of and atorvastatin markedly lowered Cx43 expression. endothelial Cx37 and Cx40 outside the plaque areas. The Atherosclerosis causes coronary arteries to progressively downregulation of these endothelial Cxs was exacerbated by stenose, and despite the development of drug-eluting stents, short-term treatment with simvastatin. This study also raises restenosis is a serious clinical problem associated with signif- the issue of how Cx expression is altered in dyslipidemia. icant morbidity and mortality. Diabetes still remains one of In dyslipidemia, there is a reduction in Cx37 and Cx40 in the most important risk factors for restenosis following stent vascular endothelial cells and increase in Cx43 expression in implantation. The levels of Cx43 are increased in the smooth both vascular endothelial cells and smooth muscle cells. This muscle cells of atherosclerotic arteries of both humans and is supported by both hyperlipidemic mice studies by Yeh et mice [29–31]. Recent study of connexins following balloon al. [26] who found reduced expression of endothelial Cx37 angioplasty has shown that this process is accompanied by 4 Cardiology Research and Practice elevated expression of Cx43 in endothelial cells and SMC nephrons supplied by a single interlobular artery. In rat of the media and later the neointima [32, 33]. Neointima preglomerular renal microvasculature, there is extensive formation following balloon catheter injury is significantly expression of Cx37, 40, and 43 mRNA in endothelial cells reduced in heterozygous Cx43 knockout mice (Cx43+/−), andCx37mRNAinsmoothmusclecells[47]. In the suggesting a correlation between neointima formation and postglomerular renal microvasculature, only Cx43 is found high levels of Cx43 during the inflammatory response to within the endothelial cells [44]. injury [34]. In diabetic nephropathy, early functional changes occur In humans, Type II diabetes is associated with hyperten- in the nephron at the level of the glomerulus including sion, and animal studies have demonstrated that multiple Cx glomerular hyperfiltration and hyperperfusion, before the changes are associated with the remodelling of the arterial onset of measurable clinical change. Over time, further struc- wall that occurs in response to hypertension. These Cx tural changes like thickening of the glomerular basement changes are complex, and various expression patterns are membrane, glomerular hypertrophy, and mesangial expan- describedindifferent models of hypertension [35]. Renin sion take place [48]. In animal models of diabetes, changes in secretion is regulated by coordinated signalling between the Cxs are associated with these key pathophysiological features. various cells of the juxtaglomerular apparatus and is de- Glomerular hyperfiltration in early diabetic nephropathy pendent on Cx40. Cx40 knockout (Cx40−/−) mice are ini- results from deceased resistance in both afferent and efferent tially found to be hypertensive [36]. Further studies [37] arterioles of the glomerulus. With the development of early using Cx40−/− mice have shown that chronic hypertension diabetic nephropathy in STZ rats, Hillis [44] found that upregulates Cx40 in endothelial cells, but Cx37, Cx43, and levels of Cx40 were increased in smooth muscle cells of the Cx45 in smooth muscle cells. It is hypothesized that for Cx37, preglomerular arterioles, while levels of Cx43 were decreased Cx40, and Cx45, control of expression does not depend on in endothelial cells of the postglomerular arterioles. This circulating plasma levels of renin and angiotensin II, but suggests that changes in these particular Cxs are associated these hormones are required for selective expression of Cx43 with impaired autoregulation and subsequent glomerular which may occur through angiotensin II activation of the hyperfiltration. As discussed in Section 3.2 above, upregula- extracellular signal-regulated kinase and NF-κBpathways. tion of Cx40 in preglomerular vessels is of pathophysiological As gap junctions mediate the cell-to-cell propagation of significance in animal models of hypertension, and normal current flow that govern orderly contraction of the healthy blood-pressure-controlled release of renin is dependent on heart, there has been considerable investigation into the role Cx40. The association between changes in Cxs and glomeru- of Cxs (Cx40, Cx43, and Cx45) in arrhythmic diabetic heart lar hyperfiltration has also been studied in Zucker diabetic disease [38]. In rat studies of STZ-induced diabetes, there is fatty (ZDF) rats. Takenaka et al. [49] found that compared to associated structural remodelling of Cxs (Cx40, Cx43, and control rats, ZDF rats had glomerular hyperfiltration. These Cx45) in the sinoatrial node and ventricular myocytes, which features were associated with reduced Cx37 in renin secreting may partially account for sinus arrhythmias, ventricular cells and enhanced phosphorylation of Cx43 (an event arrhythmias, and prolonged QT/QRS conditions found in that interfered with gap junction functioning). Aside from human diabetes [39, 40]. Inoguchi et al. [41] investigated the the renal microvascular cells, glomerular mesangial cells, underlying mechanisms by which Cx43 is involved in gen- which are normally found around the renal blood vessels, eration of arrhythmias in diabetic rat ventricular myocytes. are involved in the pathogenesis of diabetic nephropathy. Increased levels of phosphorylated Cx43 were normalized by The studies in diabetic rats by Hillis [44] also showed use of a protein kinase C β-specific inhibitor. The mechanism increased Cx37, Cx40 and absence of Cx43 in both extra- and proposed was that hyperglycaemia causes PKC activation, intramesangial cells. Later studies by this group [50] showed which in turn causes phosphorylation of Cx43 and impaired that high glucose concentrations downregulated Cx43 and ventricular contraction. Associations between Cx43 and promoted senescence of the glomerular mesangial cells. dilated cardiomyopathy have also been identified. Mutations Changes in Cxs in human kidney cells have been studied in Lamin A/C gene (LMNA) causing dilated cardiomyopathy in culture. Not all of the cell types have been studied, and are related to a decrease in Cx43 levels [42]. In these con- potential clinical applications require more consideration. ditions, it is thought that the altered distribution of Cxs (such An interesting possibility is that Cx43 has a protective effect as Cx43) and consequent interruption of gap junctional preventing renal damage. Hills et al. [51] studied human communication within the myocardium lead to decreased collecting duct cell lines and observed that as glucose con- velocity of electrical conduction, generation of reentry centrations increased, there was a time-dependent increase in arrhythmias, and arrhythmias causing sudden death [43]. the levels of Cx43. The increase in Cx43 and gap junctional communication in response to high glucose was correlated 3.3. Microvascular Complications of Diabetes with functional acceleration of calcium transients between cells. It was concluded that Cx43 may protect the collecting 3.3.1. Nephropathy. Diabetic nephropathy is the commonest duct from damage associated with established diabetic cause of end-stage renal disease. In total, nine Cxs have nephropathy. Furthermore, levels of Cx43 have also been been found in the kidney (Cx26, Cx30.3, Cx31, Cx32, Cx37, studied in human diabetic nephropathy as a “predictive Cx40, Cx43, Cx45, and Cx46) [44–46]. In the preglomerular marker” of disease progression and severity. Sawai et al. [52] vessels, cell-to-cell communication has an important phys- found that downregulation of Cx43 within podocytes was iological role in the functional coupling between adjacent closely associated with disease progression in established Cardiology Research and Practice 5 diabetic nephropathy and correlated with the degree of Cx32 gene mutation in humans is strongly associated with future decline in renal function. peripheral neuropathy in X-linked Charcot Marie tooth disease. It is also known that gap junctional communication 3.3.2. Retinopathy. Throughout the human retina, numerous is important in the maintenance of the perineurium blood- Cxs have been studied [53, 54]. Abnormalities in Cx nerve barrier. Reduced levels of Cx32 and 26 have been have been linked to the development of human diabetic observed in the perineurium in STZ rat model of diabetes- retinopathy. In human pericyte cell culture studies by Li et al. related peripheral neuropathy and are likely to contribute to [55], high glucose concentrations directly induced downreg- this condition [63]. ulation of Cx43 expression and inhibition of gap junctional intracellular communication in retinal pericytes. Accelerated 3.3.5. Chronic Ulceration. Diabetic patients are more prone apoptosis of retinal vascular cells is a key feature of diabetic to develop foot ulcers, for several reasons including neuropa- microangiopathy and is caused by inadequate transportation thy, vascular disease, and foot deformities [64]. Diabetic foot of necessary cell survival and signaling molecules. Studies woundsareveryslowtoheal[65], and numerous dysfunc- of retinal microvascular endothelial cells have found that tional events occur in all phases of diabetic wound healing decreasing Cx43 levels leads to a significant increase in the (acute inflammation, proliferation, and remodelling). There occurrence of apoptosis [56]. It has also been shown that is impairment of the inflammatory response, macrophage high-glucose-induced oxidative stress hyperphosphorylates function, cytokine, chemokine and growth factor produc- Cx43, leading to the disassembly of gap junctions and degra- tion, extracellular matrix production, angiogenesis, kerati- dation of Cx43 through the proteasome pathway [57, 58]. nocyte, and fibroblast migration and reepithelialisation [66]. At least ten of the Cxs are found within the skin (Cx Diabetic retinopathy also is associated with accelerated apop- 26, 30, 30.3, 31.1, 32, 37, 40, 43, and 45) [67]. However, tosis in retinal mitochondria. In the retinal mitochondria, it the spatial and differential expression pattern of Cxs in has been shown that matrix metalloproteinase-2 activation the skin does vary between animals and humans [68]. In modulates the abundance of heat shock protein-60. This humans, Cx43 is the most abundant Cx of the epidermis, is thought to then lead to opening of the mitochondrial expressed throughout the spinous, granular, and basal cell transition pores through disruption of Cx43 [59]. layers, sebaceous glands, hair follicles, and blood vessels [69]. Cx26 is less abundant in the epidermis and is found in a 3.3.3. Erectile and Bladder Dysfunction. Patients with dia- patchy distribution in the basal keratinocytes, but is highly betes have a higher incidence of erectile dysfunction and expressed in hair follicles and sweat glands [70]. In animal bladder dysfunction, and both these conditions are char- models, when diabetes is induced, there are significantly acterised by early onset and increased severity. The role of altered Cx levels in various layers of the skin in response Cxs in the development of erectile dysfunction and higher to injury. Studies by Wang et al. [71] demonstrated that bladder detrusor muscle activity in diabetes is emerging in the skin of rats with STZ-induced diabetes within two through recent animal studies. Suadicani et al. [60]found weeks of the onset of diabetes, Cx43 and Cx26 proteins markedly decreased expression of Cx43 in the penile corpora were significantly downregulated in the epidermis, whereas and increased expression of Cx43 in the urinary bladder Cx43 was upregulated in the dermis. These findings indicate following induction of diabetes in STZ rats. There is now that diabetes does not uniformly repress connexin expression evidence that nitric-oxide-dependent mechanisms alone are throughout the layers of the skin, but can also enhance it. ffi not su cient for maintaining normal erectile function, Furthermore, after wounding in diabetic rats, the wound suggesting that other processes such as gap junctional com- edge keratinocytes upregulate Cx43 in the first 24 hours munication facilitate erectile function by enabling passage of after injury. The keratinocytes then form a thickened bulb ions and small molecules necessary for the relaxation of cor- of nonmigrating cells at the wound margin and do not start poral smooth muscle cells [61].Inbladderdetrusormuscle migrating until Cx43 downregulates, around 48 hours after overactivity, in addition to hemichannel paracrine signalling wounding. This increased level of Cx43 then delays reep- mechanisms via ATP activation of purinergic receptors, other ithelialisation and the wound-healing process. Preventing the potential mechanisms that could be responsible for bladder abnormal expression of Cx43 in wound edge keratinocytes dysfunction include transcriptional regulation of Cx43 by rescued wound healing to normal rates or even better [70]. basic fibroblast growth factor (bFGF), and signaling via the If Cx43 expression in human diabetic wounds is altered in ERK-AP-1 pathway [62]. a similar way to the STZ diabetic rat model, targeting Cx43 may provide a new clinical application for the treatment of 3.3.4. Peripheral Neuropathy. Diabetic peripheral neuropa- diabetic ulceration. Qiu et al. [72] initially studied the topical thy involves both somatic and autonomic nerves, causing application of a Cx43 antisense olidodeoxynucleotide (Cx43 sensory deficits, palsies, and cardiovascular, gastrointestinal, asODN) pluronic gel to murine wounds. This treatment had and genitourinary dysfunction. The initiation of diabetic a rapid effect, noticeable as early as six hours after injury. peripheral neuropathy is not well understood. Various mech- The treated wounds appeared less erythematous and oede- anisms are involved in the pathogenesis of this condition, matous, with reduced exudate and less gape than control including terminal nerve destruction, decreased neuronal wounds. In diabetic STZ rats, Kamibayashi et al. [70] also conductance, demyelination, microvascular insufficiency, found that the application of Cx43 asODN pluronic inflammation, and oxidative damage. As discussed above, gel inhibited the upregulation of Cx43. This treatment 6 Cardiology Research and Practice significantly improved wound healing, compared to control [10] J. Bergoffen,S.S.Scherer,S.Wangetal.,“Connexinmutations wounds, doubling the rate of reepithelialisation. in X-linked Charcot-Marie-Tooth disease,” Science, vol. 262, In human diabetic ulcer biopsies, Brandner et al. [73]has no. 5142, pp. 2039–2042, 1993. reported persistence of Cx43 expression in wound-edge ker- [11] Y. Kim, K. G. Choi, K. Park, K. Lee, K. Chung, and B. O. Choi, atinocytes. Other studies of human diabetic fibroblasts have “X-linked dominant Charcot-Marie-Tooth disease with con- also shown increased gap junctional communication, which nexin 32 (Cx32) mutations in Koreans,” Clinical Genetics, vol. 81, no. 2, pp. 142–149, 2012. correlated with reduced cell proliferation [74]. Overall, Cxs [12] R. Hamelin, F. Allagnat, J. A. Haefliger, and P. Meda, “Con- are associated with tissue response to injury directly, via gap nexins, diabetes and the metabolic syndrome,” Current Protein junctional communication, and may also contribute to the and Peptide Science, vol. 10, no. 1, pp. 18–29, 2009. wound repair process indirectly [75]. The downregulation [13] R. N. Nlend, L. Michon, S. Bavamian et al., “Connexin36 and of Cx43 may have direct effect on the migratory ability of pancreatic β-cell functions,” Archives of Physiology and Bio- the cells involved in wound healing, through its central role chemistry, vol. 112, no. 2, pp. 74–81, 2006. in the protein “nexus.” Cx43 downregulation may also exert [14] S. Speier, A. Gjinovci, A. Charollais, P. Meda, and M. Rupnik, an influence at the transcriptional level on numerous other “Cx36-mediated coupling reduces β-cell heterogeneity, con- genes and proteins [76]. fines the stimulating glucose concentration range, and affects insulin release kinetics,” Diabetes, vol. 56, no. 4, pp. 1078– 1086, 2007. 4. Conclusion [15] N. Belluardo, A. Trovato-Salinaro, G. Mudo,` Y. L. Hurd, and D. F. Condorelli, “Structure, chromosomal localization, and The study of Cxs in diabetes is of interest. The changes in brain expression of human Cx36 gene,” Journal of Neuroscience Cxs found in diabetes are associated with both direct effects Research, vol. 57, no. 5, pp. 740–752, 1999. within the vasculature and indirect effects, by impairment [16] V. Serre-Beinier, D. Bosco, L. Zulianello et al., “Cx36 makes of homeostasis in vital organs such as the liver and kidney. channels coupling human pancreatic β-cells, and correlates It is unlikely that targeting individual connexins will cure with insulin expression,” Human Molecular Genetics, vol. 18, diabetes. However, it may be possible to alleviate some of the no. 3, pp. 428–439, 2009. symptoms of microvascular complications, as demonstrated [17] C. J. Nolan, P. Damm, and M. Prentki, “Type 2 diabetes across in recent work using topical Cx43 asODN gel treatment. As generations: from pathophysiology to prevention and man- considered in nephropathy studies, Cxs may also be used as agement,” The Lancet, vol. 378, no. 9786, pp. 169–181, 2011. [18]M.Vinken,T.Henkens,E.DeRop,J.Fraczek,T.Vanhaecke, future predictors of both diabetes progression and severity. and V. Rogiers, “Biology and pathobiology of gap junctional channels in hepatocytes,” Hepatology, vol. 47, no. 3, pp. 1077– References 1088, 2008. [19] J. Yang, A. Ichikawa, and T. Tsuchiya, “A novel function of [1] J. E. Shaw, R. A. Sicree, and P. Z. Zimmet, “Global estimates connexin 32: marked enhancement of liver function in a hep- of the prevalence of diabetes for 2010 and 2030,” Diabetes atoma cell line,” Biochemical and Biophysical Research Commu- Research and Clinical Practice, vol. 87, no. 1, pp. 4–14, 2010. nications, vol. 307, no. 1, pp. 80–85, 2003. [2] R. B. Goldberg, “Cytokine and cytokine-like inflammation [20] E. Nelles, C. Butzler,¨ D. Jung et al., “Defective propagation of markers, endothelial dysfunction, and imbalanced coagula- signals generated by sympathetic nerve stimulation in the tion in development of diabetes and its complications,” Journal liver of connexin32-deficient mice,” Proceedings of the National of Clinical Endocrinology and Metabolism,vol.94,no.9,pp. Academy of Sciences of the United States of America, vol. 93, no. 3171–3182, 2009. 18, pp. 9565–9570, 1996. [3] G. Mes¸e, G. Richard, and T. W. White, “Gap junctions: basic [21] H. Niessen and K. Willecke, “Strongly decreased gap junc- structure and function,” Journal of Investigative Dermatology, tional permeability to inositol 1,4,5-trisphosphate in con- vol. 127, no. 11, pp. 2516–2524, 2007. nexin32 deficient hepatocytes,” FEBS Letters, vol. 466, no. 1, [4] D. A. Goodenough and D. L. Paul, “Gap junctions,” Cold pp. 112–114, 2000. Spring Harbor Perspectives in Biology, vol. 1, no. 1, article [22] S. Morel, L. Burnier, and B. R. Kwak, “Connexins participate a002576, 2009. in the initiation and progression of atherosclerosis,” Seminars [5] W. H. Evans, E. De Vuyst, and L. Leybaert, “The gap junction in Immunopathology, vol. 31, no. 1, pp. 49–61, 2009. cellular internet: connexin hemichannels enter the signalling [23] K. E. L. Scheckenbach, S. Crespin, B. R. Kwak, and M. limelight,” Biochemical Journal, vol. 397, no. 1, pp. 1–14, 2006. Chanson, “Connexin channel-dependent signaling pathways [6] S. Maeda and T. Tsukihara, “Structure of the gap junction in inflammation,” Journal of Vascular Research, vol. 48, no. 2, channel and its implications for its biological functions,” pp. 91–103, 2010. Cellular and Molecular Life Sciences, vol. 68, no. 7, pp. 1115– [24] J. J. Sheu, T. H. Tsai, L. T. Chang et al., “Diabetes mellitus 1129, 2010. downregulates expression of connexin43 in rat aortic medial [7] H. S. Duffy, M. Delmar, and D. C. Spray, “Formation of the smooth muscle cells and can be reversed by simvastatin and gap junction nexus: binding partners for connexins,” Journal losartan therapy,” Chang Gung Medical Journal,vol.31,no.2, of Physiology Paris, vol. 96, no. 3-4, pp. 243–249, 2002. pp. 136–144, 2008. [8] D. W. Laird, “The gap junction proteome and its relationship [25] C. J. Y. Hou, C. H. Tsai, C. H. Su et al., “Diabetes reduces aortic to disease,” Trends in Cell Biology, vol. 20, no. 2, pp. 92–101, endothelial gap junctions in ApoE-deficient mice: simvastatin 2010. exacerbates the reduction,” Journal of Histochemistry and [9] A. Pfenniger, A. Wohlwend, and B. R. Kwak, “Mutations in Cytochemistry, vol. 56, no. 8, pp. 745–752, 2008. connexin genes and disease,” European Journal of Clinical [26] H. I. Yeh, C. S. Lu, Y. J. Wu et al., “Reduced expression of Investigation, vol. 41, no. 1, pp. 103–116, 2011. endothelial connexin37 and connexin40 in hyperlipidemic Cardiology Research and Practice 7

mice: recovery of connexin37 after 7-day simvastatin treat- [42] L. P. Sun, L. Wang, H. Wang, Y. H. Zhang, and J. L. Pu, ment,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. “Connexin 43 remodeling induced by LMNA gene mutation 23, no. 8, pp. 1391–1397, 2003. Glu82Lys in familial dilated cardiomyopathy with atrial ven- [27] N. Katakami, K. Sakamoto, H. Kaneto et al., “Association be- tricular block,” Chinese Medical Journal, vol. 123, no. 8, pp. tween the connexin37 polymorphism and peripheral arterial 1058–1062, 2010. disease in subjects with type 2 diabetes,” Diabetes Care, vol. [43]N.J.Severs,A.F.Bruce,E.Dupont,andS.Rothery,“Remod- 32, no. 5, pp. e53–e54, 2009. elling of gap junctions and connexin expression in diseased [28] K. Dlugosova,´ P. Weismann, I. Bernatov´ a,´ R. Sotn´ıkova,´ J. myocardium,” Cardiovascular Research, vol. 80, no. 1, pp. 9– Slezak,´ and L. Okruhlicova,´ “Omega-3 fatty acids and atorvas- 19, 2008. tatin affect connexin 43 expression in the aorta of hereditary [44] G. S. Hillis, “The expression of connexin 43 in human kidney hypertriglyceridemic rats,” Canadian Journal of Physiology and and cultured renal cells,” Nephron, vol. 75, no. 4, pp. 458–463, Pharmacology, vol. 87, no. 12, pp. 1074–1082, 2009. 1997. [29] J. P. Blackburn, N. S. Peters, H. I. Yeh, S. Rothery, C. R. Green, [45] R. Guo, L. Liu, and L. Barajas, “RT-PCR study of the distri- and N. J. Severs, “Upregulation of connexin43 gap junctions bution of connexin 43 mRNA in the glomerulus and renal during early stages of human coronary atherosclerosis,” Arte- tubular segments,” American Journal of Physiology - Regulatory riosclerosis, Thrombosis, and Vascular Biology, vol. 15, no. 8, pp. Integrative and Comparative Physiology, vol. 275, no. 2, pp. 1219–1228, 1995. R439–R447, 1998. [30] B. R. Kwak, F. Mulhaupt, N. Veillard, D. B. Gros, and F. Mach, [46]E.C.Beyer,J.Gerriel,K.H.Seul,D.M.Larson,K.Banach,and “Altered pattern of vascular connexin expression in atheroscle- P. R. Brink, “Modulation of intercellular communication by ff rotic plaques,” Arteriosclerosis, Thrombosis, and Vascular Biol- di erential regulation and heteromeric mixing of co-expressed ogy, vol. 22, no. 2, pp. 225–230, 2002. connexins,” Brazilian Journal of Medical and Biological Re- [31] T. Inoue and K. Node, “Molecular basis of restenosis and novel search, vol. 33, no. 4, pp. 391–397, 2000. issues of drug-eluting stents,” Circulation Journal, vol. 73, no. [47] B. Arensbak, H. B. Mikkelsen, F. Gustafsson, T. Christensen, 4, pp. 615–621, 2009. and N. H. Holstein-Rathlou, “Expression of connexin 37, 40, [32]D.Polacek,F.Bech,J.F.McKinsey,andP.F.Davies,“Con- and 43 mRNA and protein in renal preglomerular arterioles,” nexin43 gene expression in the rabbit arterial wall: effects of Histochemistry and Cell Biology, vol. 115, no. 6, pp. 479–487, hypercholesterolemia, balloon injury and their combination,” 2001. Journal of Vascular Research, vol. 34, no. 1, pp. 19–30, 1997. [48] S. Dronavalli, I. Duka, and G. L. Bakris, “The pathogenesis of diabetic nephropathy,” Nature Clinical Practice Endocrinology [33] H. I. Yeh, F. Lupu, E. Dupont, and N. J. Severs, “Upregulation and Metabolism, vol. 4, no. 8, pp. 444–452, 2008. of connexin43 gap junctions between smooth muscle cells [49] T. Takenaka, T. Inoue, H. Okada et al., “Altered gap junctional after balloon catheter injury in the rat carotid artery,” communication and renal haemodynamics in Zucker fatty rat Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 17, no. model of type 2 diabetes,” Diabetologia,vol.54,no.8,pp. 11, pp. 3174–3184, 1997. 2192–2201, 2011. [34] C. E. Chadjichristos, C. M. Matter, I. Roth et al., “Reduced [50] X. Zhang, X. Chen, D. Wu et al., “Downregulation of con- connexin43 expression limits neointima formation after bal- nexin 43 expression by high glucose induces senescence in loon distension injury in hypercholesterolemic mice,” Circula- glomerular mesangial cells,” Journal of the American Society of tion, vol. 113, no. 24, pp. 2835–2843, 2006. Nephrology, vol. 17, no. 6, pp. 1532–1542, 2006. [35] A. C. Brisset, B. E. Isakson, and B. R. Kwak, “Connexins in [51] C. E. Hills, R. Bland, D. C. Wheelans, J. Bennett, P. M. Ronco, vascular physiology and pathology,” Antioxidants and Redox and P. E. Squires, “Glucose-evoked alterations in connexin43- Signaling, vol. 11, no. 2, pp. 267–282, 2009. mediated cell-to-cell communication in human collecting [36] N. Krattinger, A. Capponi, L. Mazzolai et al., “Connexin40 duct: a possible role in diabetic nephropathy,” American regulates renin production and blood pressure,” Kidney Inter- Journal of Physiology - Renal Physiology, vol. 291, no. 5, pp. national, vol. 72, no. 7, pp. 814–822, 2007. F1045–F1051, 2006. [37] F. Alonso, N. Krattinger, L. Mazzolai et al., “An angiotensin II- [52] K. Sawai, M. Mukoyama, K. Mori et al., “Redistribution of κ and NF- B-dependent mechanism increases connexin 43 in connexin43 expression in glomerular podocytes predicts poor murine arteries targeted by renin-dependent hypertension,” renal prognosis in patients with type 2 diabetes and overt Cardiovascular Research, vol. 87, no. 1, pp. 166–176, 2010. nephropathy,” Nephrology Dialysis Transplantation, vol. 21, no. [38]N.J.Severs,S.R.Coppen,E.Dupont,H.I.Yeh,Y.S.Ko,and 9, pp. 2472–2477, 2006. T. Matsushita, “Gap junction alterations in human cardiac [53] G. Sohl,¨ A. Joussen, N. Kociok, and K. Willecke, “Expression of disease,” Cardiovascular Research, vol. 62, no. 2, pp. 368–377, connexin genes in the human retina,” BMC Ophthalmology, 2004. vol. 10, no. 1, article 27, 2010. [39] F. C. Howarth, N. Nowotny, E. Zilahi, M. A. El Haj, and M. [54] J. E. Cook and D. L. Becker, “Gap-junction proteins in retinal Lei, “Altered expression of gap junction connexin proteins development: new roles for the “nexus”,” Physiology, vol. 24, may partly underlie heart rhythm disturbances in the strep- no. 4, pp. 219–230, 2009. tozotocin-induced diabetic rat heart,” Molecular and Cellular [55] A. F. Li, T. Sato, R. Haimovici, T. Okamoto, and S. Roy, “High Biochemistry, vol. 305, no. 1-2, pp. 145–151, 2007. glucose alters connexin 43 expression and gap junction inter- [40] F. C. Howarth, N. J. Chandler, S. Kharche et al., “Effects of cellular communication activity in retinal pericytes,” Inves- streptozotocin-induced diabetes on connexin43 mRNA and tigative Ophthalmology and Visual Science, vol. 44, no. 12, pp. protein expression in ventricular muscle,” Molecular and 5376–5382, 2003. Cellular Biochemistry, vol. 319, no. 1-2, pp. 105–114, 2008. [56]S.Roy,K.Trudeau,S.Roy,Y.Behl,S.Dhar,andA. [41] T. Inoguchi, H. Y. Yu, M. Imamura et al., “Altered gap junction Chronopoulos, “New insights into hyperglycemia-induced activity in cardiovascular tissues of diabetes,” Medical Electron molecular changes in microvascular cells,” JournalofDental Microscopy, vol. 34, no. 2, pp. 86–91, 2001. Research, vol. 89, no. 2, pp. 116–127, 2010. 8 Cardiology Research and Practice

[57] R. A. Kowluru and P. S. Chan, “Oxidative stress and diabetic [75] G. S. Ghatnekar, M. P. O’Quinn, L. J. Jourdan, A. A. retinopathy,” Experimental Diabesity Research, vol. 2007, Arti- Gurjarpadhye, R. L. Draugh, and R. G. Gourdie, “Con- cle ID 43603, 2007. nexin43 carboxyl-terminal peptides reduce scar progenitor [58] R. Fernandes, H. Girao,˜ and P. Pereira, “High glucose down- and promote regenerative healing following skin wounding,” regulates intercellular communication in retinal endothe- Regenerative Medicine, vol. 4, no. 2, pp. 205–223, 2009. lial cells by enhancing degradation of connexin 43 by a [76] E. Kardami, X. Dang, D. A. Iacobas et al., “The role of connex- proteasome-dependent mechanism,” The Journal of Biological ins in controlling cell growth and gene expression,” Progress in Chemistry, vol. 279, no. 26, pp. 27219–27224, 2004. Biophysics and Molecular Biology, vol. 94, no. 1-2, pp. 245–264, [59] G. Mohammad and R. A. Kowluru, “Novel role of mito- 2007. chondrial matrix metalloproteinase-2 in the development of diabetic retinopathy,” Investigative Ophthalmology & Visual Science, vol. 52, no. 6, pp. 3832–3841, 2011. [60] S. O. Suadicani, M. Urban-Maldonado, M. T. Tar, A. Melman, andD.C.Spray,“Effects of ageing and streptozotocin-induced diabetes on connexin43 and P2 purinoceptor expression in the rat corpora cavernosa and urinary bladder,” BJU International, vol. 103, no. 12, pp. 1686–1693, 2009. [61] G. Pointis, “Connexin43: emerging role in erectile function,” International Journal of Biochemistry and Cell Biology, vol. 38, no. 10, pp. 1642–1646, 2006. [62] H. Negoro, A. Kanematsu, M. Imamura et al., “Regulation of connexin 43 by basic fibroblast growth factor in the bladder: transcriptional and behavioral implications,” Journal of Urol- ogy, vol. 185, no. 6, pp. 2398–2404, 2011. [63] D. A. Pitre, J. L. Seifert, and J. A. Bauer, “Perineurium inflam- mation and altered connexin isoform expression in a rat model of diabetes related peripheral neuropathy,” Neuroscience Letters, vol. 303, no. 1, pp. 67–71, 2001. [64] W. J. Jeffcoate and K. G. Harding, “Diabetic foot ulcers,” The Lancet, vol. 361, no. 9368, pp. 1545–1551, 2003. [65] T. Mustoe, “Understanding chronic wounds: a unifying hy- pothesis on their pathogenesis and implications for therapy,” American Journal of Surgery, vol. 187, no. 5, pp. 65S–70S, 2004. [66] H. Rafehi, A. El-Osta, and T. C. Karagiannis, “Genetic and ep- igenetic events in diabetic wound healing,” International Wound Journal, vol. 8, no. 1, pp. 12–21, 2011. [67] G. Richard, “Connexins: a connection with the skin,” Experi- mental Dermatology, vol. 9, no. 2, pp. 77–96, 2000. [68] G. Richard, “Connexin disorders of the skin,” Clinics in Der- matology, vol. 23, no. 1, pp. 23–32, 2005. [69] D. Salomon, E. Masgrau, S. Vischer et al., “Topography in mammalian connexins in human skin,” Journal of Investigative Dermatology, vol. 103, no. 2, pp. 240–247, 1994. [70] Y. Kamibayashi, M. Oyamada, Y. Oyamada, and M. Mori, “Expression of gap junction proteins connexin 26 and 43 is modulated during differentiation of keratinocytes in newborn mouse epidermis,” JournalofInvestigativeDermatology, vol. 101, no. 6, pp. 773–778, 1993. [71] C. M. Wang, J. Lincoln, J. E. Cook, and D. L. Becker, “Abnor- mal connexin expression underlies delayed wound healing in diabetic skin,” Diabetes, vol. 56, no. 11, pp. 2809–2817, 2007. [72] C. Qiu, P. Coutinho, S. Frank et al., “Targeting connexin43 expression accelerates the rate of wound repair,” Current Biology, vol. 13, no. 19, pp. 1697–1703, 2003. [73] J. M. Brandner, P. Houdek, B. Husing,C.Kaiser,andI.Moll,¨ “Connexins 26, 30, and 43: differences among spontaneous, chronic, and accelerated human wound healing,” Journal of Investigative Dermatology, vol. 122, no. 5, pp. 1310–1320, 2004. [74] K. M. Abdullah, G. Luthra, J. J. Bilski et al., “Cell-to-cell com- munication and expression of gap junctional proteins in human diabetic and nondiabetic skin fibroblasts: effects of basic fibroblast growth factor,” Endocrine,vol.10,no.1,pp. 35–41, 1999. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 213785, 5 pages doi:10.1155/2012/213785

Review Article Potential of Novel EPO Derivatives in Limb Ischemia

Dhiraj Joshi,1 Janice Tsui,1 Rebekah Yu,1 Xu Shiwen,2 Sadasivam Selvakumar,1 David J. Abraham,2 and Daryll M. Baker1

1 Vascular Unit, Division of Surgery and Interventional Science, Royal Free Hospital, University College London (Royal Free Campus), Pond Street, London NW3 2QG, UK 2 Centre for Rheumatology, University College London (Royal Free Campus), Pond Street, London NW3 2QG, UK

Correspondence should be addressed to Janice Tsui, [email protected]

Received 15 July 2011; Accepted 12 December 2011

Academic Editor: Sidney G. Shaw

Copyright © 2012 Dhiraj Joshi et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Erythropoietin (EPO) has tissue-protective properties, but it increases the risk of thromboembolism by raising the haemoglobin concentration. New generation of EPO derivatives is tissue protective without the haematopoietic side effects. Preclinical studies have demonstrated their effectiveness and safety. This paper summarizes the development in EPO derivatives with emphasis on their potential use in critical limb ischaemia.

1. Introduction gene therapy, clinical studies have been carried out to inject bone-marrow mononuclear cells intramuscularly in CLI. Critical limb ischaemia (CLI) causes considerable morbidity The innate ability of bone marrow cells to supply endothelial and mortality. Surgical and endovascular revascularization progenitor cells and release endothelial growth factors and are the main treatment options. One-fourth of patients pre- cytokines also helped induce angiogenesis [8]. However, de- senting with CLI undergo a major amputation, primarily spite a demonstration of increased angiogenesis, improve- because they are not suitable for revascularization [1]. Phar- ment in clinical endpoints was less clear. This is likely to be macotherapy has negligible role in the management of CLI. due to the complex pathology involved within the ischaemic Prostacyclin is the only disease-modifying drug with proven tissue [9]. efficacy in reducing rest pain and/or ulcer size in CLI via its Tissue protection may be an alternative therapeutic strat- vasodilatory effects on the distal circulation [2]. The Trans- egy for CLI where reducing tissue damage and maintaining Atlantic Inter Society Consensus document on peripheral tissue viability may improve outcome of current treatments vascular disease (TASC) has recommended the use of iloprost and contribute to limb salvage. It has been shown previously in CLI to avoid or delay major amputation [3]. There is lim- that limb ischaemia causes apoptosis of skeletal muscle [19]. ited evidence supporting the use of other vasoactive drugs Therefore,theuseoftissueprotectiveagentstodecrease such as pentoxifylline, cilostozol, and L-Arginine (precursor apoptosis and inflammation, either alone or in combination of nitric oxide) in the treatment of CLI [4]. with other therapeutic strategies, may be beneficial. More recently, the possibility of using recombinant for- Erythropoietin (EPO) has tissue protective properties mulations of growth factors to augment the development of that have been described extensively [20, 21]. We have recent- collaterals and new capillaries from pre-existing blood vessels ly shown that there is expression and upregulation of EPO has been explored. This is otherwise known as therapeu- receptors (EPORs) in human skeletal muscle [18]. However, tic angiogenesis [5]. Prominent among these techniques EPO triggers profound haematopoiesis that increases the is the intra-arterial or intramuscular administration of risk of thromboembolism [22]. This may prevent its use in vascular endothelial growth factor (VEGF) plasmid DNA CLI. Lately, nonhematopoietic EPO derivatives have been (phVEGF165)[6] where initial studies in patients with CLI developed, which avoid its hematopoietic side effects [23]. demonstrated an increase in angiogenesis [7]. Other than The aim of this paper is to summarize the developments that 2 Cardiology Research and Practice have taken place so far in the use of EPO derivatives in pro- tection of ischaemic tissue and discuss its potential use in CLI.

2. EPO and Its Receptors

Erythropoietin (EPO) belongs to the class-I cytokine family. It is a glycoprotein hormone with a molecular mass of 30.4 kDa [24]. The protein core contains 165 amino acids and the peripheral carbohydrate chains are made up of 4 N- linked oligosaccharides (Figure 1)[25]. Human EPO receptor (EPOR) is a cell surface protein containing 508 amino acids and has a molecular weight of 55 KDa [26]. It belongs to the cytokine receptor superfamily and contains extracellular, transmembraneous, and intracellular domains [27]. The common β-chain of IL-3/IL-5/GM-CSF receptor (also called CD131 or β-common receptor) belongs to the same super-family. The molecular weight of β-com- mon receptor (βcR) is approximately 130 kDa [23]. The EPO Carbohydrate molecule binds to two EPOR subunits (EPOR dimers or Amino acid (EPOR)2) to bring about the haemopoietic action. In con- trast, the EPO molecule binding to two βcR subunits sand- Figure 1: A schematic representation of the molecule of EPO. wiching two EPOR subunits mediates the tissue protective Adapted from Elliott et al. [25] function [23]. The four subunits constitute the heterorecep- tor complex, which binds to the EPO molecule and activates cell signalling (Figure 2). The intracellular or cytoplasmic causes hypertension by its action on vascular smooth muscle domain of all cytokine receptors is intangibly related to a cells [37–39]. Janus family tyrosine protein kinase 2 (JAK2). The binding Due to the potential side effects, most research on the of EPO to its receptor (EPOR homodimer or βcR and EPOR) tissue protective effects of EPO in CLI and other tissues has induces a conformational change in the receptor, activating a been limited to preclinical models. However, the latest non- downstream signalling pathway that leads to the transcrip- haemopoietic EPO derivatives may overcome the unneces- tion of new proteins that determine the fate of the cell [28]. sary side effects and permit the translation of preclinical re- Endogenous EPO is essential for the maturation and sults into clinical trials [11]. survival of erythrocytes and thus maintaining the normal haemoglobin concentration. In chronic renal failure, the kid- neys cease to make EPO, and commercially available prepa- 3. Nonhaematopoietic Derivatives of EPO rations of recombinant human EPO (rHuEPO) are mainly 3.1. Carbamylated Erythropoietin (CEPO). CEPO is formed used to treat anaemia secondary to chronic renal failure [30]. by the chemical modification of the amino acid lysine to In addition to its role in prolonging the survival of erythro- homocitrulline. It changes the receptor-binding capability of cytes, EPO has been shown to exert anti-apoptotic and anti- the EPO molecule [11]. inflammatory effects in neural and cardiac tissue both in vitro and in vivo. Clinical trials have also confirmed its beneficial effects in stroke and myocardial infarction [21, 31]. 3.2. Asialo EPO. Asialo EPO is formed by decreasing the However, the ability of rHuEPO to bond with EPOR in sialic acid content of rHuEPO by a technique known as the extrahaematopoietic tissues including vascular endothelium total enzymatic desialylation. Remarkably, the molecule re- is about 50 times less than that in the erythroid cells [32]. tains its tissue protective property but loses the haemopoietic Hence, the dose of EPO required to attain a tissue protective ability. This also results in the reduction of the half-life of the effect in heart and brain ischaemia is much higher than that Asialo EPO to minutes [10]. used for the management of anaemia in renal failure. In CLI, the volume of ischaemic tissue is even greater than that in 3.3. Helix-B Surface Peptide (HBSP) [40]. The EPO mol- heart and brain ischaemia, therefore higher doses still may ecule, a glycoprotein that contains 165 amino acids is folded be required to achieve a therapeutic effect. This will cause an in such a way that the hydrophobic areas-helixes A, C, and D unfavorable increase in the haematocrit and blood viscosity, face away from the surface and are embedded in the extra- which in turn will raise the risk of thromboembolism and cellular domains of (EPOR)2.However,helixBthatishydro- occlusion of arteries with pre-existing disease [33, 34]. An- philic points away from the receptor and has wide-ranging other concern is a potential effect on tumour progression due tissue-protective properties. Isolating the helix B which is to its angiogenic properties [35]. However, recent evidence devoid of the primary sequence resulted in a low molecu- suggests that this may not be a significant risk [36]. EPO also lar weight peptide. The spontaneous cyclization of Helix B Cardiology Research and Practice 3

JAK-2 JAK-2 JAK-2 JAK-2

JAK-2 JAK-2

Figure 2: Schematic representation of the structure of erythropoietin receptor (EPOR) and beta-common receptor (CD131) on the left and the right hand side, respectively. ∗Based on Brines et al. [29]. JAK-2 stands for Janus family Tyrosine Protein Kinase 2.

Table 1: Recent studies investigating the role of nonhaematopoietic derivatives of EPO.

Nonhaematopoietic Author Model Inference EPO derivative In vivo model of cerebral and spinal Erbayraktar et al. 2003 [10] Asialo EPO Neuroprotection-recovery of paraplegia cord ischaemia plus sciatic nerve injury Carbamylated In vitro and in vivo models of neural Demonstration of tissue protective Leis et al. 2004 [11] EPO (CEPO) injury properties of CEPO in neural tissue In vivo model of rat myocardial Reduction in the size of Fiordaliso et al. 2005 [12] CEPO infarction myocardial infarct CEPO reduced metabolic stress caused In vivo model of metabolic stress in Fantacci et al. 2006 [13] CEPO byCHandimprovedcellularsurvival mice induced by chronic hypoxia (CH) independent of haematopoiesis. In vivo model of radiosurgically induced Erbayraktar et al. 2006 [14] CEPO Decreased necrosis of the brain tissue brain injury in rats In vivo models of decubitus ulcers Improved healing and less intraperitoneal Erbayraktar et al. 2009 [15] CEPO and ARA 290 and peritonitis adhesion formation Reduction in the size of infarct and Ahmet et al. 2010 [16] ARA 290 In vivo model of myocardial ischemia improvement of ejection fraction Decreased apoptosis of myotubes but Joshi et al. 2010 [17] ARA 290 In vitro no effect on angiogenesis Decreased ischaemia-induced Joshi et al. 2011 [18] ARA 290 In vitro inflammation and apoptosis of myotubes surface peptide (HBSP) gives rise to pyroglutamate HBSP elicited strong tissue protection [40]. The studies of non- (pHBSP), which is also known as ARA 290. The effective erythropoietic EPO derivatives have been summarised in dose of ARA 290 peptide that showed evidence of tissue Table 1. protection was comparable on a molar basis to those studied for EPO and was higher than that required for EPO-mediated 4. Mechanism of Action erythropoiesis. For example, in the renal ischaemia model of mice, 0.08 nmol/kg of bw (equivalent to 300 units/kg of body Receptor-initiated cell signaling by EPO and its derivatives weight of EPO) was ineffective, whereas a 10-fold higher dose occurs via multiple molecular cascades, which differ in 4 Cardiology Research and Practice importance depending on the specific tissue or cell type, as [2] S. Marchesi, L. Pasqualini, R. Lombardini et al., “Prostaglan- well as the type of injury [41]. As in erythropoiesis, the din E1 improves endothelial function in critical limb ische- majority of tissue-protective responses begin by the phos- mia,” Journal of Cardiovascular Pharmacology, vol. 41, no. 2, phorylation of JAK-2 or JAK-1 [42, 43]. Following JAK-2 pp. 249–253, 2003. phosphorylation, 3 main pathways are activated that result in [3] J. A. Dormandy and R. B. Rutherford, “Management of pe- ripheral arterial disease (PAD). TASC Working Group. Trans- attenuation of apoptosis: STAT-5/Bcl-xL (Signal Transducers Atlantic Inter-Society Consensus (TASC),” Journal of Vascular and Activator of Transcription; type 5) [44]; PI3K/Akt (phos- Surgery, vol. 31, no. 1, part 2, pp. S1–S296, 2000. phatidyl inositol-3 kinase/Protein Kinase B) [45]; in some [4] Management of Peripheral Arterial Disease (PAD), “TransAt- tissues including the heart, the MAPK (mitogen-activated lantic Inter-Society Consensus (TASC),” European Journal of protein kinase) pathway is involved [46]. MAPK and P13K/ Vascular and Endovascular Surgery, vol. 19, supplemt A, pp. Akt inhibit caspase activation, thus directly attenuating S1–S250, 2000. apoptosis. We have shown that ARA 290 decreases cleaved [5] S. Takeshita, D. Gal, G. Leclerc et al., “Increased gene expres- caspase-3 in the myotubes subjected to ischaemia [18]. This sion after liposome-mediated arterial gene transfer associated may indicate that EPO and its derivatives may attenuate with intimal smooth muscle cell proliferation. In vitro and in apoptosis through the MAPK or P13K/Akt pathway. How- vivo findings in a rabbit model of vascular injury,” Journal of Clinical Investigation, vol. 93, no. 2, pp. 652–661, 1994. ever, further study is needed to understand completely the [6] J. M. Isner, A. Pieczek, R. Schainfeld et al., “Clinical evidence interactions of these signaling pathways that ultimately of angiogenesis after arterial gene transfer of phVEGF165 in transduce tissue protection in vivo. patient with ischaemic limb,” The Lancet, vol. 348, no. 9024, pp. 370–374, 1996. [7] I. Baumgartner, A. Pieczek, O. Manor et al., “Constitutive ex- 5. Discussion pression of phVEGF165 after intramuscular gene transfer pro- motes collateral vessel development in patients with critical It is now known that EPO receptors are present in human limb ischemia,” Circulation, vol. 97, no. 12, pp. 1114–1123, skeletal muscle and their expression is elevated in CLI. It 1998. has also been shown that a nonhaematopoietic derivative of [8] E. Tateishi-Yuyama, H. Matsubara, T. Murohara et al., “Ther- EPO (ARA 290) decreases inflammation and apoptosis of apeutic angiogenesis for patients with limb ischaemia by ischaemic myotubes and may thus be used to selectively en- autologous transplantation of bone-marrow cells: a pilot study hance the tissue-protective activity of EPO whilst avoiding and a randomised controlled trial,” The Lancet, vol. 360, no. ff 9331, pp. 427–435, 2002. the haematopoietic side e ects in CLI. [9] K. A. Webster, “Therapeutic angiogenesis: a complex problem The timing of administering EPO derivatives depends on requiring a sophisticated approach,” Cardiovascular Toxicol- their half-life. CEPO has a half-life of 4–6 hours while Asialo- ogy, vol. 3, no. 3, pp. 283–297, 2003. EPO and ARA 290 have a half-life of about 2 minutes. Most [10] S. Erbayraktar, G. Grasso, A. Sfacteria et al., “Asialoerythro- in vivo studies have treated animals immediately following poietin is a nonerythropoietic cytokine with broad neuropro- ischaemic injury [15, 23]. In the context of CLI, EPO deriva- tective activity in vivo,” Proceedings of the National Academy of tives may be used at the onset of CLI, either alone or as an Sciences of the United States of America, vol. 100, no. 11, pp. adjunct to surgical or endovascular revascularization via sys- 6741–6746, 2003. [11] M. Leis, P. Gliezzi, G. Grasso et al., “Derivatives of erythropoi- temic administration through intra-arterial, intravenous, or tein that are tissue protective but not erythropoietic,” Science, intramuscular routes. Due to the relatively short half-life of vol. 305, no. 5681, pp. 239–242, 2004. currently available compounds, repeated injections may be [12] F. Fiordaliso, S. Chimenti, L. Staszewsky et al., “A nonerythro- required. Clinical studies are needed to assess the efficacy of poietic derivative of erythropoietin protects the myocardium EPO derivatives in patients with CLI and to identify optimal from ischemia-reperfusion injury,” Proceedings of the National delivery routes and dosing schedules. Academy of Sciences of the United States of America, vol. 102, no. 6, pp. 2046–2051, 2005. [13] M. Fantacci, P. Bianciardi, A. Caretti et al., “Carbamylated 6. Conclusion erythropoitein ameliorates the metabolic stress induced in vivo by severe chronic hypoxia,” Proceedings of the National EPO receptors are present in human skeletal muscle and their Academy of Sciences of the United States of America, vol. 103, expression is elevated in CLI. Nonhaematopoietic derivatives no. 46, pp. 17531–17536, 2006. [14] S. Erbayraktar, N. de Lanerolle, A. de Lotbiniere` et al., “Car- of EPO decrease inflammation and apoptosis of ischaemic ff bamylated erythropoietin reduces radiosurgically-induced myotubes whilst avoiding haematopoietic side e ects. They brain injury,” Molecular Medicine, vol. 12, no. 4–6, pp. 74–80, may thus reduce tissue damage and provide a unique ther-a- 2006. peutic alternative for CLI. [15] Z. Erbayraktar, S. Erbayraktar, O. Yilmaz, A. Cerami, T. Coleman, and M. Brines, “Nonerythropoietic tissue protective compounds are highly effective facilitators of wound healing,” References Molecular Medicine, vol. 15, no. 7-8, pp. 235–241, 2009. [16] I. Ahmet, H. J. Tae, M. Juhaszova et al., “A small nonery- [1]A.J.Henry,N.D.Hevelone,M.Belkin,andL.L.Nguyen, thropoietic helix B surface peptide based upon erythropoietin “Socioeconomic and hospital-related predictors of amputa- structure is cardioprotective against ischemic myocardial tion for critical limb ischemia,” Journal of Vascular Surgery, vol. damage,” Molecular Medicine, vol. 17, no. 3-4, pp. 194–200, 53, no. 2, pp. 330–339, 2011. 2011. Cardiology Research and Practice 5

[17] D. Joshi, J. Tsui, X. Shiwen et al., “Helix-B peptide of eryth- [34] P. Zaoui, F. Bayle, J. Maurizi, M. Foret, S. DalSoglio, and ropoietin could be used as pharmacotherapy in critical limb P. Vialtel, “Early thrombosis in kidney grafted into patient ischaemia,” British Journal of Surgery, vol. 98, no. 2, 2011. treated with erythropoietin,” The Lancet, vol. 2, no. 8617, p. [18] D. Joshi, J. Tsui, H. Patel et al., “Role of erythropoietin recep- 956, 1988. tors and ligands in attenuating inflammation and apoptosis in [35] M. Dicato and L. Plawny, “Erythropoietin in cancer patients: critical limb ischemia,” Journal of Vascular Surgery, vol. 53, no. pros and cons,” Current Opinion in Oncology,vol.22,no.4,pp. 6, supplement, p. 80S, 2011. 307–311, 2010. [19] R. G. Mitchell, B. D. Duscha, J. L. Robbins et al., “Increased [36]M.E.Hardee,J.P.Kirkpatrick,S.Shanetal.,“Humanre- levels of apoptosis in gastrocnemius skeletal muscle in patients combinant erythropoietin (rEpo) has no effect on tumour with peripheral arterial disease,” Vascular Medicine, vol. 12, no. growth or angiogenesis,” British Journal of Cancer, vol. 93, no. 4, pp. 285–290, 2007. 12, pp. 1350–1355, 2005. [20] H. Ehrenreich, K. Weissenborn, H. Prange et al., “Recombi- [37] H. Schiffl and S. M. Lang, “Hypertension induced by recom- nant human erythropoietin in the treatment of acute ischemic binant human erythropoietin (rHU-EPO) can be prevented stroke,” Stroke, vol. 40, no. 12, pp. e647–e656, 2009. by indomethacin: pathogenetic role of cytosolic calcium,” [21] N. P. Riksen, D. J. Hausenloy, and D. M. Yellon, “Erythro- European Journal of Medical Research, vol. 2, no. 3, pp. 97–100, poietin: ready for prime-time cardioprotection,” Trends in 1997. Pharmacological Sciences, vol. 29, no. 5, pp. 258–267, 2008. [38] S. Kuriyama, G. Tokudome, H. Tomonari et al., “Effect of hu- [22] C. L. Bennett, S. M. Silver, B. Djulbegovic et al., “Venous man recombinant erythropoietin (Epo) on cellular Ca2+,Na+, thromboembolism and mortality associated with recombi- and K+ regulation of vascular smooth muscle cells grown in nant erythropoietin and darbepoetin administration for the vitro—a new insight into the pathogenesis of Epo induced hy- treatment of cancer-associated anemia,” JAMA, vol. 299, no. 8, pertension,” Japanese Journal of Nephrology,vol.35,no.6,pp. pp. 914–924, 2008. 671–678, 1993. [23] M. Brines, G. Grasso, F. Fiordaliso et al., “Erythropoietin med- [39] Z. Ni, X. Q. Wang, and N. D. Vaziri, “Nitric oxide metabolism iates tissue protection through an erythropoietin and com- in erythropoietin-induced hypertension: effect of calcium mon β-subunit heteroreceptor,” Proceedings of the National channel blockade,” Hypertension, vol. 32, no. 4, pp. 724–729, Academy of Sciences of the United States of America, vol. 101, 1998. no. 41, pp. 14907–14912, 2004. [40] M. Brines, N. S. A. Patel, P. Villa et al., “Nonerythropoietic, [24] J. M. Davis, T. Arakawa, T. W. Strickland, and D. A. Yphantis, tissue-protective peptides derived from the tertiary structure “Characterization of recombinant human erythropoietin pro- of erythropoietin,” Proceedings of the National Academy of duced in Chinese hamster ovary cells,” Biochemistry, vol. 26, Sciences of the United States of America, vol. 105, no. 31, pp. no. 9, pp. 2633–2638, 1987. 10925–10930, 2008. [25] S. Elliott, T. Lorenzini, D. Chang, J. Barzilay, and E. Delorme, [41] M. Brines and A. Cerami, “Erythropoietin-mediated tissue “Mapping of the active site of recombinant human erythro- protection: reducing collateral damage from the primary poietin,” Blood, vol. 89, no. 2, pp. 493–502, 1997. injury response,” Journal of Internal Medicine, vol. 264, no. 5, [26] M. Budarf, K. Huebner, B. Emanuel et al., “Assignment of the pp. 405–432, 2008. erythropoietin receptor (EPOR) gene to mouse chromosome [42] P. Rafiee, Y. Shi, J. Su, K. A. Pritchard, J. S. Tweddell, and 9 and human chromosome 19,” Genomics,vol.8,no.3,pp. J. E. Baker, “Erythropoietin protects the infant heart against 575–578, 1990. ischemia-reperfusion injury by triggering multiple signaling [27] J. F. Bazan, “A novel family of growth factor receptors: a pathways,” Basic Research in Cardiology, vol. 100, no. 3, pp. common binding domain in the growth hormone, prolactin, 187–197, 2005. erythropoietin and IL-6 receptors, and the p75 IL-2 receptor [43] M. Digicaylioglu, G. Garden, S. Timberlake, L. Fletcher, and β-chain,” Biochemical and Biophysical Research Communica- S. A. Lipton, “Acute neuroprotective synergy of erythropoietin tions, vol. 164, no. 2, pp. 788–795, 1989. and insulin-like growth factor I,” Proceedings of the National [28] D. M. Wojchowski and T. C. He, “Signal transduction in the Academy of Sciences of the United States of America, vol. 101, erythropoietin receptor system,” Stem Cells,vol.11,no.5,pp. no. 26, pp. 9855–9860, 2004. 381–392, 1993. [44] T. C. Wen, Y. Sadamoto, J. Tanaka et al., “Erythropoietin pro- [29] M. Brines and A. Cerami, “Emerging biological roles for tects neurons against chemical hypoxia and cerebral ischemic erythropoietin in the nervous system,” Nature Reviews Neu- injury by up-regulating Bcl-x,” Journal of Neuroscience Re- roscience, vol. 6, no. 6, pp. 484–494, 2005. search, vol. 67, no. 6, pp. 795–803, 2002. [30] C. G. Winearls, “Treatment of the anaemia of chornic renal [45] C. J. Parsa, A. Matsumoto, J. Kim et al., “A novel protective failure with recombinant human erythropoietin,” Drugs, vol. effect of erythropoietin in the infarcted heart,” Journal of Clini- 38, no. 3, pp. 342–345, 1989. cal Investigation, vol. 112, no. 7, pp. 999–1007, 2003. [31] H. Ehrenreich, M. Hasselblatt, C. Dembowski et al., “Erythro- [46]J.E.Baker,D.Kozik,A.K.Hsu,X.Fu,J.S.Tweddell,andG.J. poietin therapy for acute stroke is both safe and beneficial,” Gross, “Darbepoetin alfa protects the rat heart against infarc- Molecular Medicine, vol. 8, no. 8, pp. 495–505, 2002. tion: dose-response, phase of action, and mechanisms,” Jour- [32] R. Yamaji, T. Okada, M. Moriya et al., “Brain capillary nal of Cardiovascular Pharmacology, vol. 49, no. 6, pp. 337– endothelial cells express two forms of erythropoietin receptor 345, 2007. mRNA,” European Journal of Biochemistry, vol. 239, no. 2, pp. 494–500, 1996. [33] A. Mingoli, P. Sapienza, A. Puggioni, C. Modini, and A. Cav- allaro, “A possible side-effect of human erythropoietin ther- apy: thrombosis of peripheral arterial reconstruction,” Euro- pean Journal of Vascular and Endovascular Surgery, vol. 18, no. 3, pp. 273–274, 1999. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 143209, 7 pages doi:10.1155/2012/143209

Review Article Stromal-Cell-Derived Factor-1 (SDF-1)/CXCL12 as Potential Target of Therapeutic Angiogenesis in Critical Leg Ischaemia

Teik K. Ho, 1 X. Shiwen,2 D. Abraham,2 J. Tsui,1 and D. Baker1

1 Division of Surgery and Interventional Science, University College London (Royal Free Campus), The Royal Free Hospital, Pond Street, London NW3 2QG, UK 2 Centre for Rheumatology and Connective Tissue Disease, The Royal Free Hospital, Pond Street, London NW3 2QG, UK

Correspondence should be addressed to Teik K. Ho, [email protected]

Received 19 July 2011; Accepted 16 December 2011

Academic Editor: Sidney G. Shaw

Copyright © 2012 Teik K. Ho et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In the Western world, peripheral vascular disease (PVD) has a high prevalence with high morbidity and mortality. In a large percentage of these patients, lower limb amputation is still required. Studies of ischaemic skeletal muscle disclosed evidence of endogenous angiogenesis and adaptive skeletal muscle metabolic changes in response to hypoxia. Chemokines are potent chemoattractant cytokines that regulate leukocyte trafficking in homeostatic and inflammatory processes. More than 50 different chemokines and 20 different chemokine receptors have been cloned. The chemokine stromal-cell-derived factor-1 (SDF-1 aka CXCL12) is a constitutively expressed and inducible chemokine that regulates multiple physiological processes, including embryonic development and organ homeostasis. The biologic effects of SDF-1 are mediated by chemokine receptor CXCR4, a 352 amino acid rhodopsin-like transmembrane-specific G protein-coupled receptor (GPCR). There is evidence that the administration of SDF-1 increases blood flow and perfusion via recruitment of endothelial progenitor cells (EPCs). This review will focus on the role of the SDF-1/CXCR4 system in the pathophysiology of PVD and discuss their potential as therapeutic targets for PVD.

1. Introduction as a promising strategy to treat a variety of cardiovascu- lar diseases. Experimental animal model studies [6]have Atherosclerotic peripheral vascular disease (PVD) is a major progressed to several clinical trials which evaluated the cause of morbidity and mortality in the Western world [1]. angiogenic potentials of growth factors such as hepatocyte It has been reported to have 19.1% prevalence in those over growth factor (HGF), vascular endothelial growth factor 55 years of age [2], and 15% of male patients die within (VEGF), and fibroblast growth factor (FGF) in inducing 5 years of diagnosis, with the deaths mostly due to other new blood vessel growth in CLI patients. Unfortunately, the associated atherosclerotic disease such as stroke, coronary results of these clinical trials have been inconsistent and heart disease, and abdominal vascular disease [3]. With inconclusive. For example, in the regional angiogenesis with an aging population and improved medical care that has VEGF (RAVE) trial [7], intramuscular adenoviral delivery of increased life expectancy, more patients are presenting with VEGF in patients with unilateral exercise-limiting claudica- critical leg ischemia (CLI), the end stage of PVD. In 10 to tion had no effects on primary efficacy end points but was 40% of these patients [4], lower limb amputation may be associated with dose-dependent peripheral oedema. VEGF required because the anatomic extent and the distribution induces formation of hyperpermeable and proinflammatory of arterial occlusive disease make the patients unsuitable for vessels; the concerted actions of additional mediators (such revascularization, despite current advances in surgery and as the angiopoietins and platelet-derived growth factor) may endovascular revascularization techniques. be required for their stabilization and maturation. This view In the past 20 years, rapid development in molecular has been supported by the use of a combination of two biology and understanding of mechanisms of angiogenesis angiogenic factors, platelet-derived growth factor BB and [5] have led to the development of therapeutic angiogenesis FGF-2, which synergistically induce vascular networks that 2 Cardiology Research and Practice remain stable for more than a year even after depletion of with critical leg ischaemia [23]. Studies at the molecular level angiogenic factors [8]. show that the SDF-1 promoter contains hypoxia inducible Chemokines are potent chemoattractant cytokines that factor-1α (HIF-1α) binding sites, as revealed by chromatin regulate leukocyte trafficking in homeostatic and inflam- immunoprecipitation analysis, and SDF-1 expression is matory processes. More than 50 different chemokines and upregulated in endothelial cells by HIF-1α [24]. Thus, it 20 different chemokine receptors have been cloned [9]. is possible that elevated levels of HIF-1α in critical leg Chemokines usually bind to multiple receptors, and the same ischaemia [25] may have led to increased SDF-1 in the receptor may bind to more than one chemokine. However, ischaemic skeletal muscle fibres. there is one exception to this rule: SDF-1, which binds exclusively to CXCR4, and has CXCR4 as its only receptor 3. CXCR4 [9–16]. Recently, a new putative receptor called CXCR7 (aka RDC1 and GPR159) for SDF-1 had been described [17]; The biologic effects of SDF-1 are mediated by the chemokine however, its potential role in regulating cell trafficking awaits receptor CXCR4. CXCR4 (aka fusin and CD184), is a confirmation by other laboratories. 352 amino acid rhodopsin-like transmembrane specific G This review will focus on the role of stromal-cell- protein-coupled receptor (GPCR). CXCR4 is expressed by derived factor- (SDF-1), otherwise known as CXCL12, and various cell lines, including muscle cell lines, endothelial its receptor CXC chemokine receptor 4 (CXCR4; i.e., the cells, leucocytes, and progenitor cells [26–28] and the SDF-1/CXCR4 system) in the pathophysiology of PVD and expression of CXCR4 receptor on endothelial cells has discuss their potential as therapeutic targets for PVD. been reported to be regulated by hypoxia [29, 30]at transcriptional level. The regulation of CXCR4 can also 2. Stromal-Cell-Derived Factor- (SDF-1) occur during protein expression when it is subjected to a number of cotranslational modifications. These processes The chemokine stromal-cell-derived factor-1 (SDF-1, also affect the expression and function of the receptor [31]. known as CXCL12) is a constitutively expressed and After expression, CXCR4 is exposed to proteolytic degra- inducible chemokine that regulates multiple physiological dation by proteases that are present in the haematopoietic processes, including embryonic development and organ microenvironment and serum (e.g., cathepsin G, elastase, homeostasis [18]. SDF-1 is the only member of the α- dipeptidyl-peptidase) [32, 33]. CXCR4 after interaction with chemokines which does not possess the conserved Glu-Leu- SDF-1 is internalized from the surface in a mechanism Arg motif, also called the ELR motif, preceding the first involving G-protein-coupled receptor kinases followed by cysteine residue but has angiogenic activity. Two main splice the binding of β-arrestin [34], and subsequently, is recircu- forms of SDF-1 have been identified, SDF-1α and SDF- lated from the endosomal compartment at different rates: 1β, which have identical amino acid sequences except for this achieves another level of regulation. The function of the presence of 4 additional amino acids at the carboxy the CXCR4 receptor depends on its incorporation into terminus of SDF-1β [19]. Another splice variant form, SDF- membrane lipid domains called lipid rafts and several signals 1γ, has been subsequently identified in the nervous system from other membrane receptors (e.g., G-protein-coupled [20]. SDF-1δ, SDF-1ε,andSDF-1ϕ splice variants have been C3a anaphylatoxin receptor—C3aR) or integrins that may described only recently in human tissues and are abundant increase the incorporation of CXCR4 into membrane lipid in the pancreas. Additionally, SDF-1ϕ and SDF-1ε are found rafts, enhancing its signalling [35, 36]. Finally, CXCR4 in the heart and liver, as well as in fetal and adult kidneys, but is the subject of negative regulation by regulators of G- the expression of SDF-1ϕ is more pronounced than SDF-1ε. protein signalling (RGS) proteins which may regulate CXCR4 SDF-1δ, on the other hand, is also detected in the spleen, fetal signalling differently in different tissues [37, 38]. liver, and lungs [21]. SDF-1α is the predominant isoform found in all organs 4. SDF-1 CXCR4 Signalling Axis but undergoes rapid proteolysis in blood. SDF-1β is more resistant to blood-dependent degradation, stimulates angio- Upon activation of CXCR4, a number of signalling pathways genesis, and is present in highly vascularized organs such as are activated leading to a variety of biological responses. As the liver, spleen, and kidneys. In contrast, SDF-1γ is located CXCR4 couples to the Gi family of proteins, pertussis toxin in very metabolically active organs susceptible to infarction (PTX), which ADP-ribosylates Gαi and inhibits GPCR/Gi such as the heart and the brain. The significance of the coupling, has been used to delineate pathways that are G- existence of the splice forms of SDF-1 has largely remained protein-dependent and -independent. Strong evidence exists unclear. The understanding of the functional diversity of the that the protein-tyrosine phosphatases SHIP1 and SHIP2, different splicing variants will help in developing therapeutic as well as membrane-expressed hematopoietic phosphatase strategies. Structure-function analysis of SDF-1α (1–67) has CD45, are involved in the modulation of CXCR4 signalling identified the NH2-terminal amino acids (residues 1–8) as [39]. CXCR4 signalling is negatively regulated by the RGS critical to receptor binding and activation. The expression of proteins [37, 38] and lipid phosphatase activity of tumor SDF-1 in several organs including liver, brain, kidney, and suppressor phosphatase and tensin homolog (PTEN) [40]. heart increased in the presence of ischaemia [22]. Recently, it has been shown that SDF-1 expression in skeletal muscle 4.1. G Protein Signalling. To date, the majority of signalling is also elevated in ischaemic skeletal muscle fibres of patients pathways and biological outcomes of CXCR4 activation Cardiology Research and Practice 3 are PTX sensitive and, therefore, dependent on activation and complements the resident endothelial cells in sprouting of Gi proteins. Activated Gi is able to inhibit adenylyl new vessels. Also, ischaemia and various cytokines, includ- cyclase as well as activate the Src family of tyrosine ing VEGF and granulocyte-macrophage colony-stimulating kinases while liberated Gβγ activates phospholipase C-β factor (GM-CSF), are reported to mobilize EPC into sites of (PLC-β)[31], phosphoinositide-3 kinase (PI3K) [41], and neovascularization [54]. EPCs express CXCR4 which allows mitogen activated protein kinase (MAPK) [42] signalling homing to sites of neovascularization in ischaemic tissues pathways. As described earlier, the CXCR4/SDF-1 complex which release the homing signal, that is, SDF-1, which then is incorporated into lipid rafts and may associate with act as ligand for CXCR4. Other CXCR4+ proangiogenic cells members of the Src family of kinases [43]. Activated CXCR4 are composed of immature and mature hematopoietic cells, increases intracellular calcium mobilization and induces and smooth muscle cell (SMC) progenitors, which all have phosphorylation of focal adhesion components such as FAK direct or indirect proangiogenic properties. and Pyk2 [44–46]. The PI3K/AKT and MAPK signal trans- SDF-1 is upregulated in ischaemic tissues as hypoxic duction pathways contribute to chemotaxis, cell migration and/or apoptotic conditions are triggers to the induction [42, 47], and secretion of various matrix metalloproteinases of cytokine and chemokine expression. Ceradini et al. have (MMP’s) including MMP-2 and MMP-9 [46, 48]. MMP-2 demonstrated that SDF-1 expression in ischaemic sites is and MMP-9 are involved in the migration of cells through the directly correlated with the amplitude of hypoxia [24]. Sig- basement membrane [49]. They also induce secretion of the nalling activated by hypoxia leading to SDF-1 upregulation angiopoietic factor, VEGF [50]. Furthermore, blocking either involves the recruitment of integrin-linked kinase and HIF- the PI3K or MAPK pathway inhibits CXCR4-activated cell 1[24, 55].Acuteandgradualarterialocclusioncanboth migration in a pre-B cell line [42]. Ultimately, through these lead to increased SDF-1 expression in ischaemic limb [56]. pathways, SDF1-bound CXCR4 can induce cell proliferation, However, the expression of SDF-1 can be affected by age, with chemotaxis, migration, the secretion of angiopoietic factors, decreased expression in older age [57]. We investigated the all important components of the angiogenic process [51]. expression of two main splice variant, SDF-1α and SDF-1β in ischaemic human skeletal muscle of patients with critical leg 4.2. G Protein Independent Signalling. Activation of the ischaemia [23]. The study confirms the elevated expression JAK/STAT pathway by CXCR4 has been proposed to be of SDF-1α but there is lack of SDF-1β expression in ischaemic G protein independent [52]. SDF-1 induces the transient human skeletal muscle. It is postulated that the lack of SDF- association of JAK2 and JAK3 with CXCR4, leading to 1β expression in critical leg ischaemia may be explained by its the activation and nuclear translocation of a number of transient expression [58] because all the muscles examined STAT proteins. While JAK/STAT activation is G protein- have been chronically ischaemic. Another human study [59] independent, pretreatment with PTX leads to a prolonged shows that SDF-1 expression increases in acute on chronic association of JAK with CXCR4 suggesting that G protein ischaemia with perivascular retention of CXCR4+ cells. coupling is involved in JAK/STAT-receptor complex recycling The increased expression of SDF-1 in ischaemic muscles [52]. acts as a chemoattractant and homing signal for CXCR4+ EPCs. Locally administered SDF-1 has been shown to 5. SDF-1, Vasculogenesis, and Angiogenesis augment the accumulation of EPCs to the site of ischaemia, resulting in enhancing the efficacy of neovascularization Until recently, new blood vessel growth in the adult was after systemic EPC transplantation [60]. The study pro- thought to occur exclusively through angiogenesis, defined as vided experimental proof of principle for the feasibility the sprouting of vessels from existing vascular structures. In and therapeutic effectiveness of local administration of contrast to angiogenesis, endothelial progenitor cells (EPCs) SDF-1 in augmenting the SDF-1/CXCR4+ EPCs enhanced are mobilised from the bone marrow and recruited to foci of neovascularization in ischaemic muscles. neovascularization where they form new blood vessels in situ through a process called vasculogenesis. Once thought to be 5.2. Angiogenic Properties of SDF-1. Many critical cell func- limited to embryonic development, vasculogenesis appears tions such as migration, proliferation, and apoptosis inhi- to be preserved in the adult and contributes to postischemic bition are regulated by SDF-1. There are limited studies in vascular regeneration. the literature investigating human microvascular endothelial cells which are better representatives of the microvasculature 5.1. The Effect of SDF-1 and Endothelial Progenitor Cell (EPC) involved in angiogenesis. In a study comparing the angio- Trafficking on Vasculogenesis. In 1997, Asahara et al. reported genic properties of both SDF-1α and SDF-1β splice variants the isolation of putative EPC from human peripheral [23], both splice variants attenuate human microvascular blood, on the basis of cell-surface expression of CD34 and endothelial cells apoptosis and stimulate cell proliferation other endothelial markers [53]. These cells are reported to and capillary tube formation in a Matrigel assay. The differentiate in vitro into endothelial cells and seem to be Matrigel assay mimics various steps in angiogenesis such as incorporated at sites of active angiogenesis in various animal proliferation, migration, and differentiation in the process models of ischaemia. These findings suggest that incorpo- of capillary tube formation. Compared with SDF-1α, SDF- ration of bone-marrow-derived endothelial precursor cells 1β has a greater effect on apoptosis and cell proliferation. into the new vessels lumen contributes to the growing vessels Treatment with both variants results in time-dependent 4 Cardiology Research and Practice activation of PI3K/Akt and MAPK p44/42 but not p38 MAP a major characteristic of VEGF-stimulated angiogenesis, kinase [23]. These properties are likely to contribute to the may not be observed after injection of SDF-1. SDF-1 complex mechanism of angiogenesis. contributes to the stabilization of neovessel formation by recruiting CXCR4+PDGFR+ckit+ smooth muscle progeni- tor cells during recovery from vascular injury [65]. However, 6. Potential Use of SDF-1 as the angiogenic pathways involving VEGF and SDF-1 are Therapeutic Angiogenic Factor not independent. Extensive evidence suggests that SDF-1 upregulates VEGF synthesis in several different cell types, The ideal mode of delivery of angiogenic growth factors whereas VEGF and basic FGF induce SDF-1 and its receptor has yet to be determined. They are usually delivered by CXCR4 in endothelial cells [66]. In a transgenic system of gene transfer, which commonly uses adenoviruses (Ads) and VEGF-mediated neovascularization, SDF-1 is a key mediator plasmids as vectors or delivered as recombinant protein. of retention of recruited bone-marrow-derived cells in close Each method has its own disadvantages [6] and the delivery proximity to angiogenic vessels [67]. These interactions of SDF-1 is no exception. Recently, two promising studies provide a link between angiogenic growth factors and attempt to provide successful modes of SDF-1 delivery in chemokine-induced angiogenesis. ischaemic limbs to improve revascularisation using vascular Although promising, the use SDF-1 in a clinical setting gene transfer of SDF-1 by ultrasound-mediated destruction is uncertain and yet to be tested. The animals used for of plasmid bearing microbubbles and mutated recombinant experimentation are typically young and healthy and have SDF-1 proteins [61, 62]. the same genetic background. In contrast, patients with PVD The success of EPCs therapy relies on the ability of cells are a highly heterogeneous group comprising older individ- to repair damaged tissue, and it is critically dependent on uals with enormous genetic diversity and various comorbid the homing, migration, and retention to sites of ischaemia, conditions, such as diabetes mellitus, hypertension, and regardless of mode of delivery. In an animal model of chronic hyperlipidemia. ischaemic limb, it has been demonstrated that a noninvasive gene- and cell-based therapeutic approach, using vascular gene transfer of SDF-1 by ultrasound-mediated destruction 7. Future Work of plasmid bearing microbubbles to augment homing and engraftment of exogenously administered EPCs, leads to a Better understanding of the mechanisms of action of the greater angiogenic response as compared to SDF-1 gene chemokine in the ischaemic limb is necessary for optimal therapy or intravenous EPCs alone [61]. However, long-term exploitation of the SDF-1/CXCR4 axis in PVD. As discussed, ff data on EPC engraftment, beyond 14 days after delivery, are SDF-1 not only exerts its e ect on vasculogenesis by the not available. recruitment of CXCR4+ progenitor cells, but it also has The use of SDF-1 protein therapy to enhance angio- direct angiogenic properties on endothelial cells. The mecha- genesis is hampered by concerns regarding rapid inac- nism of actions causing neovascularisation in ischaemic limb tivation of the chemokine in the protease-rich environ- is uncertain: one or both of the processes may be involved. In ment of the ischaemic limb. SDF-1 processing by matrix addition, there is uncertainty on the role and interaction of metalloproteinase-2 results in generation of a neurotoxic VEGF and other angiogenic growth factors with SDF-1 in the fragment that does not bind to its main receptor, CXCR4, microenvironment of ischaemic limbs. The answers to these and lacks chemotactic activity for EPCs [63, 64]. To ensure issues are paramount to understanding the interplay between sustained SDF-1 activity in the hostile environment of the various mediators of angiogenesis and to design therapeutic ischaemic limb, Segers et al. [62] designed recombinant SDF- strategies combining protein and cell therapy approaches. 1 proteins carrying mutations that provide resistance to pro- Moreover, the effects of SDF-1 on the inflammatory tease cleavage. One of these SDF-1 variants, SSDF-1(S4V), response in the ischaemic limb need to be understood. The is resistant to processing by matrix metalloproteinase-2 and roles of elevated expression of SDF-1 and its interaction dipeptidyl peptidase IV but retains chemotactic activity in with other angiogenic growth factors in ischaemic muscles vitro and induced angiogenesis in vivo.Deliveryofprotease- need to be determined. There has been no study on how resistant SDF-1 with the use of self-assembling nanofibres the increased expression of SDF-1 by ischaemic muscles to achieve sustained local concentrations increases arteriolar affect serum levels of SDF-1, which can then affect the density and enhances blood flow in the ischaemic mouse mobilisation of progenitor cells from bone marrow. Evidence hindlimb. from in vivo studies in models of inflammatory injury is Similar to treatment with the angiogenic growth factors conflicting, suggesting that SDF-1 exerts context-dependent basic fibroblast growth factor (FGF) and VEGF, admin- proinflammatory and antiinflammatory actions [68]. istration of protease-resistant SDF-1 augments perfusion, The role of SDF-1 in establishing long-term stable and increasing vascular density in the ischaemic limb. Consider- mature blood vessels remains unknown. A major limiting ing the established effects of VEGF and basic FGF admin- factor in angiogenic therapies is the formation of unstable istration in enhancing experimental ischaemic angiogenesis, vessels that may rapidly regress after cessation of treatment what additional role could SDF-1 treatment play? Compared [69]. Whether SDF-1 mediates coating of the neovessels with with the effects of other angiogenic growth factors, SDF-1 has pericytes and vascular maturation is not known. Formation unique properties. Generation of hyperpermeable vessels, of viable and stable neovessels in vivo may require the Cardiology Research and Practice 5 concerted effort of several distinct mediators. Finally, much the CXC chemokine PBSF/SDF-1,” Nature, vol. 382, no. 6592, work is needed to investigate the role of the newly discovered pp. 635–638, 1996. SDF-1 receptor CXCR7 on SDF-1 angiogenic properties. [15] Q. Ma, D. Jones, and T. A. Springer, “The chemokine receptor To apply SDF-1 treatment in patients with PVD, certain CXCR4 is required for the retention of B lineage and granulo- issues will need to be considered, such as the effect of cytic precursors within the bone marrow microenvironment,” SDF-1 on atherosclerosis. Additional experiments using Immunity, vol. 10, no. 4, pp. 463–471, 1999. [16] W. Lu, J. A. Gersting, A. Maheshwari, R. D. Christensen, and D. atherosclerotic animal models may shed light on this con- A. Calhoun, “Developmental expression of chemokine recep- cern. Nevertheless, we believe that the concept of augmenting tor genes in the human fetus,” Early Human Development, vol. local accumulation of transplanted EPCs opens perspectives 81, no. 6, pp. 489–496, 2005. for the clinical strategy of EPC therapies. [17] K. Balabanian, B. Lagane, S. Infantino et al., “The chemokine SDF-1/CXCL12 binds to and signals through the orphan re- References ceptor RDC1 in T lymphocytes,” Journal of Biological Chem- istry, vol. 280, no. 42, pp. 35760–35766, 2005. [1]B.A.Golomb,T.T.Dang,andM.H.Criqui,“Peripheral [18] M. Z. Ratajczak, E. Zuba-Surma, M. Kucia, R. Reca, W. arterial disease: morbidity and mortality implications,” Circu- Wojakowski, and J. Ratajczak, “The pleiotropic effects of the lation, vol. 114, no. 7, pp. 688–699, 2006. SDF-1-CXCR4 axis in organogenesis, regeneration and tu- [2]W.T.Meijer,A.W.Hoes,D.Rutgers,M.L.Bots,A.Hofman, morigenesis,” Leukemia, vol. 20, no. 11, pp. 1915–1924, 2006. and D. E. Grobbee, “Peripheral arterial disease in the elderly: [19] M. Shirozu, T. Nakano, J. Inazawa et al., “Structure and chro- the rotterdam study,” Arteriosclerosis, Thrombosis, and Vascular mosomal localization of the human stromal cell-derived Biology, vol. 18, no. 2, pp. 185–192, 1998. factor 1 (SDF1) gene,” Genomics, vol. 28, no. 3, pp. 495–500, [3] J. Golledge, “Lower-limb arterial disease,” Lancet, vol. 350, no. 1995. 9089, pp. 1459–1464, 1997. [20] M. Gleichmann, C. Gillen, M. Czardybon et al., “Cloning γ [4] S. Novo, G. Coppola, and G. Milio, “Critical limb ischemia: and characterization of SDF-1 , a novel SDF-1 chemokine definition and natural history,” Current Drug Targets, Cardio- transcript with developmentally regulated expression in the vascular and Haematological Disorders, vol. 4, no. 3, pp. 219– nervous system,” European Journal of Neuroscience, vol. 12, no. 225, 2004. 6, pp. 1857–1866, 2000. [21] M. Janowski, “Functional diversity of SDF-1 splicing variants,” [5] W. Risau, “Mechanisms of angiogenesis,” Nature, vol. 386, no. Cell Adhesion and Migration, vol. 3, no. 3, pp. 243–249, 2009. 6626, pp. 671–674, 1997. [22] I. Petit, D. Jin, and S. Rafii, “The SDF-1-CXCR4 signaling [6]T.K.Ho,V.Rajkumar,D.C.Black,D.Abraham,andD. pathway: a molecular hub modulating neo-angiogenesis,” Baker, “Critical limb ischemia classification and therapeutic Trends in Immunology, vol. 28, no. 7, pp. 299–307, 2007. angiogenesis,” International Journal of Angiology, vol. 14, no. [23] T. K. Ho, J. Tsui, S. Xu, P. Leoni, D. J. Abraham, and D. 2, pp. 49–59, 2005. M. Baker, “Angiogenic effects of stromal cell-derived factor-1 [7] S. Rajagopalan, E. R. Mohler III, R. J. Lederman et al., (SDF-1/CXCL12) variants in vitro and the in vivo expressions “Regional angiogenesis with vascular endothelial growth of CXCL12 variants and CXCR4 in human critical leg factor in peripheral arterial disease: a phase II randomized, ischemia,” Journal of Vascular Surgery, vol. 51, no. 3, pp. 689– double-blind, controlled study of adenoviral delivery of vas- 699, 2010. cular endothelial growth factor 121 in patients with disabling [24] D. J. Ceradini, A. R. Kulkarni, M. J. Callaghan et al., intermittent claudication,” Circulation, vol. 108, no. 16, pp. “Progenitor cell trafficking is regulated by hypoxic gradients 1933–1938, 2003. through HIF-1 induction of SDF-1,” Nature Medicine, vol. 10, [8] R. Cao, E. Brokenhielm, R. Pawliuk et al., “Angiogenic syn- no. 8, pp. 858–864, 2004. ergism, vascular stability and improvement of hind-limb [25] T. K. Ho, V. Rajkumar, M. Ponticos et al., “Increased endoge- ischemia by a combination of PDGF-BB and FGF-2,” Nature nous angiogenic response and hypoxia-inducible factor-1α in Medicine, vol. 9, no. 5, pp. 604–613, 2003. human critical limb ischemia,” Journal of Vascular Surgery, vol. [9] R. Horuk, “Chemokine receptors,” Cytokine and Growth Factor 43, no. 1, pp. 125–133, 2006. Reviews, vol. 12, no. 4, pp. 313–335, 2001. [26] C. Murdoch, P. N. Monk, and A. Finn, “CXC chemokine [10] A. Bagri, T. Gurney, X. He et al., “The chemokine SDF1 receptor expression on human endothelial cells,” Cytokine, vol. regulates migration of dentate granule cells,” Development, vol. 11, no. 9, pp. 704–712, 1999. 129, no. 18, pp. 4249–4260, 2002. [27] M. Kucia, J. Ratajczak, R. Reca, A. Janowska-Wieczorek, and [11] R. M. Strieter, M. D. Burdick, J. Mestas, B. Gomperts, M. P. M. Z. Ratajczak, “Tissue-specific muscle, neural and liver Keane, and J. A. Belperio, “Cancer CXC chemokine networks stem/progenitor cells reside in the bone marrow, respond to and tumour angiogenesis,” European Journal of Cancer, vol. 42, an SDF-1 gradient and are mobilized into peripheral blood no. 6, pp. 768–778, 2006. during stress and tissue injury,” Blood Cells, Molecules, and [12] A. F. Schier, “Chemokine signaling: rules of attraction,” Cur- Diseases, vol. 32, no. 1, pp. 52–57, 2004. rent Biology, vol. 13, no. 5, pp. R192–R194, 2003. [28] M. Z. Ratajczak, M. Majka, M. Kucia et al., “Expression of [13] H. E. Broxmeyer, S. Cooper, L. Kohli et al., “Transgenic expres- functional CXCR4 by muscle satellite cells and secretion of sion of stromal cell-derived factor-1/CXC chemokine ligand SDF-1 by muscle-derived fibroblasts is associated with the 12 enhances myeloid progenitor cell survival/antiapoptosis in presence of both muscle progenitors in bone marrow and vitro in response to growth factor withdrawal and enhances hematopoietic stem/progenitor cells in muscles,” Stem Cells, myelopoiesis in vivo,” Journal of Immunology, vol. 170, no. 1, vol. 21, no. 3, pp. 363–371, 2003. pp. 421–429, 2003. [29] T. Schioppa, B. Uranchimeg, A. Saccani et al., “Regulation [14] T. Nagasawa, S. Hirota, K. Tachibana et al., “Defects of B-cell of the chemokine receptor CXCR4 by hypoxia,” Journal of lymphopoiesis and bone-marrow myelopoiesis in mice lacking Experimental Medicine, vol. 198, no. 9, pp. 1391–1402, 2003. 6 Cardiology Research and Practice

[30] P. Staller, J. Sulitkova, J. Lisztwan, H. Moch, E. J. Oakeley, and non-small cell lung cancer metastases,” American Journal of W. Krek, “Chemokine receptor CXCR4 downregulated by von Respiratory and Critical Care Medicine, vol. 167, no. 12, pp. Hippel-Lindau tumour suppressor pVHL,” Nature, vol. 425, 1676–1686, 2003. no. 6955, pp. 307–311, 2003. [45] K. I. Oonakahara, W. Matsuyama, I. Higashimoto, M. Kawa- [31] J. M. Busillo and J. L. Benovic, “Regulation of CXCR4 bata, K. Arimura, and M. Osame, “Stromal-derived factor- signaling,” Biochimica et Biophysica Acta, Biomembranes, vol. 1α/CXCL12-CXCR 4 axis is involved in the dissemination 1768, no. 4, pp. 952–963, 2007. of NSCLC cells into pleural space,” American Journal of [32] T. Lapidot, A. Dar, and O. Kollet, “How do stem cells find their Respiratory Cell and Molecular Biology, vol. 30, no. 5, pp. 671– way home?” Blood, vol. 106, no. 6, pp. 1901–1910, 2005. 677, 2004. [33] T. Papayannopoulou, “Current mechanistic scenarios in [46] A. Z. Fernandis, A. Prasad, H. Band, R. Klosel,¨ and R. hematopoietic stem/progenitor cell mobilization,” Blood, vol. K. Ganju, “Regulation of CXCR4-mediated chemotaxis and 103, no. 5, pp. 1580–1585, 2004. chemoinvasion of breast cancer cells,” Oncogene, vol. 23, no. [34] M. J. Orsini, J. L. Parent, S. J. Mundell, A. Marchese, and J. 1, pp. 157–167, 2004. L. Benovic, “Trafficking of the HIV coreceptor CXCR4: role [47] J. Libura, J. Drukala, M. Majka et al., “CXCR4-SDF-1 signaling of arrestins and identification of residues in the C-terminal is active in rhabdomyosarcoma cells and regulates locomotion, tail that mediate receptor internalization,” Journal of Biological chemotaxis, and adhesion,” Blood, vol. 100, no. 7, pp. 2597– Chemistry, vol. 274, no. 33, pp. 31076–31086, 1999. 2606, 2002. [35] M. Z. Ratajczak, R. Reca, M. Wysoczynski et al., “Transplanta- [48]C.H.Tang,T.W.Tan,W.M.Fu,andR.S.Yang,“Involvement tion studies in C3-deficient animals reveal a novel role of the of matrix metalloproteinase-9 in stromal cell-derived factor- third complement component (C3) in engraftment of bone 1/CXCR4 pathway of lung cancer metastasis,” Carcinogenesis, marrow cells,” Leukemia, vol. 18, no. 9, pp. 1482–1490, 2004. vol. 29, no. 1, pp. 35–43, 2008. [36] M. Wysoczynski, R. Reca, J. Ratajczak et al., “Incorporation [49] A. Janowska-Wieczorek, L. A. Marquez, A. Dobrowsky, M. Z. of CXCR4 into membrane lipid rafts primes homing-related Ratajczak, and M. L. Cabuhat, “Differential MMP and TIMP responses of hematopoietic stem/progenitor cells to an SDF-1 production by human marrow and peripheral blood CD34+ gradient,” Blood, vol. 105, no. 1, pp. 40–48, 2005. cells in response to chemokines,” Experimental Hematology, [37] M. Berthebaud, C. Riviere, P.Jarrier et al., “RGS16 is a negative vol. 28, no. 11, pp. 1274–1285, 2000. regulator of SDF-1-CXCR4 signaling in megakaryocytes,” [50] Z. Liang, J. Brooks, M. Willard et al., “CXCR4/CXCL12 Blood, vol. 106, no. 9, pp. 2962–2968, 2005. axis promotes VEGF-mediated tumor angiogenesis through [38] Y. Le, M. Honczarenko, A. M. Glodek, D. K. Ho, and Akt signaling pathway,” Biochemical and Biophysical Research L. E. Silberstein, “CXC chemokine ligand 12-induced focal Communications, vol. 359, no. 3, pp. 716–722, 2007. adhesion kinase activation and segregation into membrane [51] P. Carmeliet, “Mechanisms of angiogenesis and arteriogene- domains is modulated by regulator of G protein signaling 1 in sis,” Nature Medicine, vol. 6, no. 4, pp. 389–395, 2000. pro-B cells,” Journal of Immunology, vol. 174, no. 5, pp. 2582– [52] A. J. Vila-Coro, J. M. Rodr´ıguez-Frade, A. M. De Ana, M. 2590, 2005. C. Moreno-Ort´ız, C. Mart´ınez-A,andM.Mellado,“The [39] A. Z. Fernandis, R. P. Cherla, and R. K. Ganju, “Differen- chemokine SDF-1α triggers CXCR4 receptor dimerization and tial regulation of CXCR4-mediated T-cell chemotaxis and activates the JAK/STAT pathway,” FASEB Journal, vol. 13, no. mitogen-activated protein kinase activation by the membrane 13, pp. 1699–1710, 1999. tyrosine phosphatase, CD45,” Journal of Biological Chemistry, [53] T. Asahara, T. Murohara, A. Sullivan et al., “Isolation of puta- vol. 278, no. 11, pp. 9536–9543, 2003. tive progenitor endothelial cells for angiogenesis,” Science, vol. [40] R. J. Phillips, J. Mestas, M. Gharaee-Kermani et al., 275, no. 5302, pp. 964–967, 1997. “Epidermal growth factor and hypoxia-induced expression [54] T. Takahashi, C. Kalka, H. Masuda et al., “Ischemia- of CXC chemokine receptor 4 on non-small cell lung and cytokine-induced mobilization of bone marrow-derived cancer cells is regulated by the phosphatidylinositol 3- endothelial progenitor cells for neovascularization,” Nature kinase/PTEN/AKT/mammalian target of rapamycin signaling Medicine, vol. 5, no. 4, pp. 434–438, 1999. pathway and activation of hypoxia inducible factor-1α,” [55] S. P. Lee, S. W. Youn, H. J. Cho et al., “Integrin-linked Journal of Biological Chemistry, vol. 280, no. 23, pp. 22473– kinase, a hypoxia-responsive molecule, controls postnatal 22481, 2005. vasculogenesis by recruitment of endothelial progenitor cells [41] M. Vicente-Manzanares, M. Rey, D. R. Jones et al., “Involve- to ischemic tissue,” Circulation, vol. 114, no. 2, pp. 150–159, ment of phosphatidylinositol 3-kinase in stromal cell-derived 2006. factor-1α-induced lymphocyte polarization and chemotaxis,” [56] Y. Yang, G. Tang, J. Yan et al., “Cellular and molecular Journal of Immunology, vol. 163, no. 7, pp. 4001–4012, 1999. mechanism regulating blood flow recovery in acute versus [42] R. K. Ganju, S. A. Brubaker, J. Meyer et al., “The α-chemokine, gradual femoral artery occlusion are distinct in the mouse,” stromal cell-derived factor-1α, binds to the transmembrane Journal of Vascular Surgery, vol. 48, no. 6, pp. 1546–1558, 2008. G-protein-coupled CXCR-4 receptor and activates multiple [57]J.S.Wang,X.Liu,Z.Y.Xueetal.,“Effects of aging on signal transduction pathways,” Journal of Biological Chemistry, time course of neovascularization-related gene expression vol. 273, no. 36, pp. 23169–23175, 1998. following acute hindlimb ischemia in mice,” Chinese Medical [43] S. N. Zaman, M. E. Resek, and S. M. Robbins, “Dual acylation Journal, vol. 124, no. 7, pp. 1075–1081, 2011. and lipid raft association of Src-family protein tyrosine kinases [58] P. Cipriani, A. F. Milia, V. Liakouli et al., “Differential expres- are required for SDF-1/CXCL12-mediated chemotaxis in the sion of stromal cell-derived factor 1 and its receptor CXCR4 Jurkat human T cell lymphoma cell line,” Journal of Leukocyte in the skin and endothelial cells of systemic sclerosis patients: Biology, vol. 84, no. 4, pp. 1082–1091, 2008. pathogenetic implications,” Arthritis and Rheumatism, vol. 54, [44] R. J. Phillips, M. D. Burdick, M. Lutz, J. A. Belperio, M. no. 9, pp. 3022–3033, 2006. P. Keane, and R. M. Strieter, “The stromal derived factor- [59]V.vanWeel,L.Seghers,M.R.DeVriesetal.,“Expression 1/CXCL12-CXC chemokine receptor 4 biological axis in of vascular endothelial growth factor, stromal cell-derived Cardiology Research and Practice 7

factor-1, and CXCR4 in human limb muscle with acute and chronic ischemia,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 27, no. 6, pp. 1426–1432, 2007. [60] J. I. Yamaguchi, K. F. Kusano, O. Masuo et al., “Stromal cell-derived factor-1 effects on ex vivo expanded endothelial progenitor cell recruitment for ischemic neovascularization,” Circulation, vol. 107, no. 9, pp. 1322–1328, 2003. [61] M. A. Kuliszewski, J. Kobulnik, J. R. Lindner, D. J. Stewart, and H. Leong-Poi, “Vascular gene transfer of SDF-1 promotes endothelial progenitor cell engraftment and enhances angio- genesis in ischemic muscle,” Molecular Therapy,vol.19,no.5, pp. 895–902, 2011. [62] V. F. M. Segers, V. Revin, W. Wu et al., “Protease-resistant stromal cell-derived factor-1 for the treatment of experimental peripheral artery disease,” Circulation, vol. 123, no. 12, pp. 1306–1315, 2011. [63] G. A. McQuibban, G. S. Butler, J. H. Gong et al., “Matrix metalloproteinase activity inactivates the CXC chemokine stromal cell-derived factor-1,” Journal of Biological Chemistry, vol. 276, no. 47, pp. 43503–43508, 2001. [64] K. Zhang, G. A. McQuibban, C. Silva et al., “HIV-induced metalloproteinase processing of the chemokine stromal cell derived factor-1 causes neurodegeneration,” Nature Neuro- science, vol. 6, no. 10, pp. 1064–1071, 2003. [65]A.Zernecke,A.Schober,I.Botetal.,“SDF-1α/CXCR4 axis is instrumental in neointimal hyperplasia and recruitment of smooth muscle progenitor cells,” Circulation Research, vol. 96, no. 7, pp. 784–791, 2005. [66] R. Salcedo and J. J. Oppenheim, “Role of chemokines in angiogenesis: CXCL12/SDF-1 and CXCR4 interaction, a key regulator of endothelial cell responses,” Microcirculation, vol. 10, no. 3-4, pp. 359–370, 2003. [67] M. Grunewald, I. Avraham, Y. Dor et al., “VEGF-induced adult neovascularization: recruitment, retention, and role of accessory cells,” Cell, vol. 124, no. 1, pp. 175–189, 2006. [68] N. Karin, “The multiple faces of CXCL12 (SDF-1α)inthe regulation of immunity during health and disease,” Journal of Leukocyte Biology, vol. 88, no. 3, pp. 463–473, 2010. [69] M. J. Gounis, M. G. Spiga, R. M. Graham et al., “Angiogenesis is confined to the transient period of VEGF expression that follows adenoviral gene delivery to ischemic muscle,” Gene Therapy, vol. 12, no. 9, pp. 762–771, 2005. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 845698, 15 pages doi:10.1155/2012/845698

Review Article The Diabetic Heart: Too Sweet for Its Own Good?

Hannah J. Whittington, Girish G. Babu, Mihaela M. Mocanu, Derek M. Yellon, and Derek J. Hausenloy

The Hatter Cardiovascular Institute, University College London, 67 Chenies Mews, London, WC1E 6HX, UK

Correspondence should be addressed to Derek J. Hausenloy, [email protected]

Received 1 August 2011; Accepted 14 November 2011

Academic Editor: Janice Tsui

Copyright © 2012 Hannah J. Whittington et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Diabetes mellitus is a major risk factor for ischemic heart disease (IHD). Patients with diabetes and IHD experience worse clinical outcomes, suggesting that the diabetic heart may be more susceptible to ischemia-reperfusion injury (IRI). In contrast, the animal data suggests that the diabetic heart may be either more, equally, or even less susceptible to IRI. The conflicting animal data may be due to the choice of diabetic and/or IRI animal model. Ischemic conditioning, a phenomenon in which the heart is protected against IRI by one or more brief nonlethal periods of ischemia and reperfusion, may provide a novel cardioprotective strategy for the diabetic heart. Whether the diabetic heart is amenable to ischemic conditioning remains to be determined using relevant animal models of IRI and/or diabetes. In this paper, we review the limitations of the current experimental models used to investigate IRI and cardioprotection in the diabetic heart.

1. Introduction and young adults is also increasing especially in westernized societies [4]. Ischemic heart disease (IHD) is the leading cause of death There are many features of the diabetic heart that can and disability in the UK. Diabetes mellitus is a major risk contribute to an increased susceptibility to IHD, these factor for IHD patients with diabetes who are 2-3 times include diabetic cardiomyopathy and angiopathy [5]. Dia- more likely to develop IHD [1]. Diabetes mellitus (DM), in betic cardiomyopathy is a disease that affects the myo- general, is a condition where the body cannot adequately cardium in diabetic patients, causing structural changes control its level of glucose. There are two types of diabetes, which lead to abnormal functionality of the heart. These type I diabetes and type II diabetes. Type I diabetes is changes can eventually result in left ventricular hypertrophy characterized by the inability of the body to produce insulin; (LVH) and diastolic/systolic dysfunction [6]. Changes at this is caused by cellular-mediated programming of the the molecular level occur in the diabetic heart leading to autoimmune system and subsequent destruction of pancre- this phenotype, including endothelial dysfunction, metabolic atic beta-cells, the cells responsible for insulin production. perturbations, and differences in cellular signaling, which The prevalence of this form of diabetes is relatively low and are well described in an extensive review by Hayat et al., usually starts early in life [2]. Whereas type II diabetes is 2004 [6]. Diabetic angiopathy causes vascular complications known as the insulin resistant form, insulin is produced in chronic diabetic patients and can be classified into two however the target tissues become insensitive or resistant types: microangiopathy and macroangiopathy [7]. Diabetic to the action of insulin thereby not utilizing blood glucose microangiopathy is the term coined to describe the damage correctly and as a consequence increasing circulating blood caused to small blood vessels and capillaries within the body glucoselevels[3]. This form accounts for a high percentage as a consequence of chronic hyperglycemia. This damage can of all diabetes and is historically more common in older age lead to a decreased supply of oxygen and vital substrates to groups; however the prevalence of type II diabetes in children the tissues, which can lead to adverse clinical outcomes such 2 Cardiology Research and Practice as retinopathy, nephropathy, neuropathy, and diabetic foot whether the diabetic heart is amenable to cardioprotection [8]. Macroangiopathy as a consequence of diabetes mainly elicited by ischemic conditioning. To achieve this requires involves an accelerated form of atherosclerosis [9], another the use of appropriate animal models of IRI and diabetes risk factor for CV disease. It affects the larger blood vessels which also take into account other comorbidities such as age, within the body, where development and progression of dyslipidemia, and hypertension, factors which also impact on atherosclerotic plaques lead to stenosis or occlusions, impair- the ability to ischemic condition the heart [25]. ing blood flow [10]. Diabetic complications, atherosclerosis, The aim of this paper will be to highlight the limitations hypertension, and many other CV risk factors such as obesity of currently used animal models of IRI and diabetes as a can all interact and render the heart more susceptible to IHD potential explanation for the disparity that exists between [11]. clinical and experimental data, with regard to the sus- The incidence of diabetes is increasing at an alarming rate ceptibility of the diabetic heart to IRI and endogenous throughout the world. Globally, the estimated prevalence cardioprotection. of diabetes for 2010 was 285 million and is expected to affect 438 million people by 2030 [1]. In the UK, there are 2. Animal Models of Diabetes 2.6 million people who have been diagnosed with diabetes (2009) which equates to a 4.1% average prevalence and it is Animal models used to investigate diabetes have been estimated that 4 million people will be diabetic by 2025 [12]. created using a variety of different methods such as the The major consequence of IHD results from the detri- administration of drugs toxic to the pancreas, modified diets, mental effect of acute myocardial ischemia-reperfusion inbreeding of spontaneous mutations, or genetic engineering injury (IRI). IRI is a paradoxical event that occurs in [34]. This has resulted in the availability of many different the myocardium. Briefly, when blood flow is reduced due diabetic animal models [35]. In addition, the diabetic status to myocardial ischemia, the best treatment strategy is to induced by these methods needs to be assessed with caution reestablish blood flow to the damaged area, however this as none of the models accurately and completely reflect the return in blood flow causes damage [13]. Patients with human pathology of diabetes. diabetes experience worse clinical outcomes in a number For example, type I diabetes is characterized by insulin of clinical settings of acute IRI including acute myocardial deficiency as a consequence of autoimmune destruction of infarction (MI) [14–16], coronary angioplasty [16], and pancreatic β cells in the islets of Langerhans [36]. This cardiac bypass surgery [17–19]. This clinical data suggests disease is mimicked by administration of streptozotocin that the diabetic heart may be more susceptible to acute (STZ). However, this is an alkylating agent based nitrosourea IRI. In contrast, the animal data is inconclusive with derivative [37], which interferes with numerous cellular experimental studies suggesting that the diabetic heart may processes such as glucose transport, glucokinase function be more, equally or even less susceptible to acute IRI [20]. and can also induce DNA strand breaks [38]. This toxic However, one major reason for the disparity between the compound can be given either as a single high dose to induce clinical and animal data may be due to the choice of IRI diabetes but this is associated with high mortality. Therefore, and/or diabetic animal model used in the animal studies it is more common to use a series of low doses to induce [20]. diabetes in rodents [39]. Alloxan is a toxic glucose analogue, Given the worse clinical outcomes in diabetic patients which preferentially accumulates in beta cells of the pancreas, with IHD, novel therapeutic strategies for protecting the causing excessive production of hydroxyl free radicals and diabetic heart against the detrimental effects of acute IRI are destruction of the beta-cells hence mimicking type I diabetes required to improve clinical outcomes in this patient group. [40]. In essence, the primary etiopathology caused by these Extensive research has investigated protecting the heart from experimental approaches has little autoimmune component IRI using conditioning strategies, either by mechanically and does not truly reflect the pathogenesis leading to type activating cell survival pathways by ischemic preconditioning I diabetes in the human [39]. Therefore, it could be argued (IPC) [21] and remote ischemic conditioning (RIC) [22] that the translation of findings from this particular animal or pharmacologically using cardioprotective agents [23]. model of type I diabetes to the clinical setting may be Ischemic conditioning is an endogenous phenomenon in problematic. which one or more brief cycles of nonlethal ischemia and Type II diabetes in humans is a metabolic disorder and reperfusion applied directly to the heart protects itself from normally arises during adulthood [41]. A good animal model a sustained lethal episode of acute IRI [24]. Furthermore of type II diabetes is the Goto-Kakizaki (GK) rat, which exciting research in the area of RIC, whereby an alternative originates from the inbreeding of Wistar rats that exhibited organ to the heart, for example, the arm, can be made hyperglycemia. The GK rat spontaneously becomes diabetic ischemic by one or more brief cycles of nonlethal ischemia early in life, showing glucose insensitivity in their pancreatic and reperfusion using a blood pressure cuff, can limit the beta cells [42] with the diabetic status increasing with age damage caused by a sustained ischemic insult [22]and [43]. The Otsuka Long Evans Tokushima Fatty (OLETF) rat this may provide a novel cardioprotective approach for the originates from the inbreeding of Long-Evans rats which diabetic heart. These strategies are further described in exhibit glucose intolerance [44]; these rats are mildly obese Section 4. However, in order to translate ischemic condi- and the diabetic phenotype is more dramatic in males. Both tioning into the clinical arena for the benefit of diabetic the Zucker diabetic fatty (ZDF) rat [45] and the db/db mice patients, it is important to first determine in animal studies [46] are models that express other comorbidities such as Cardiology Research and Practice 3 obesity and dyslipidemia as well as glucose intolerance [47]. Ischaemia For a thorough review into animal models in diabetes please refer to [34]. Myocardial reperfusion ROS↑ 3. The Susceptibility of the Diabetic Heart to Ca2+↑ Acute IRI ATP↓ ↑Inflammatory pH restoration Despite the clinical data suggesting that the diabetic heart mediators is more susceptible to acute IRI, the animal data has been + conflicting with experimental studies showing more, equal, Na ↑ or less sensitivity to acute IRI. The reasons for this disparity between the animal and clinical data were the subject of a review in 1997 by Paulson, who concluded that the ↑mPTP Mito Ca2+ sensitivity of the diabetic heart to acute IRI was dependent on the animal models and conditions used. The diabetic Apoptosis/cell ↑Cyto c heart was shown to be less sensitive to acute IRI in studies death which (i) used a short duration of diabetes (<6weeks); (ii) used glucose as the only substrate; (iii) used a no- Figure 1: Endogenous factors contributing to ischemia-reperfusion flow IRI protocol, whereby global ischemia is initiated by injury. Following ischemia, blood flow is reestablished in the total interruption of the perfusate to the heart. Whereas, if myocardium. The myocardium is subject to a number of abrupt diabetes was more prolonged and severe, fatty acids were changes during the transition from ischemia to reperfusion. present in the perfusate and a low-flow IRI protocol was Both biochemical and metabolic alterations occur including the used that is, experimental protocols which reflect the clinical generation of reactive oxygen species (ROS), decrease in ATP levels, scenario better, the diabetic heart was found to be more an increase in inflammatory mediators, the rapid restoration of sensitive to IRI [20]. We have reviewed the literature since physiological pH, which in turn increases intracellular sodium 1997, and the same pattern emerges (see Tables 1, 2,and3). and overload of intracellular calcium and mitochondrial calcium. However, to this we need to add another complicating factor, These factors interact with each other to mediate reperfusion injury the choice of IRI model and the lack of other comorbidities through the opening of the mitochondrial permeability transition such as age, dyslipidemia, and hypertension, factors which pore (mPTP) and initiation of cell death pathways [13]. are critical when investigating cardioprotection. Numerous endogenous factors contribute to myocardial IRI. These include production of reactive oxygen species Ma et al. [56] found increased phosphorylation of the (ROS), changes in the intracellular calcium and pH, trig- prosurvival kinase Akt and decreased levels of caspase-3, gering of inflammatory mechanisms; all of which interact vascular endothelial growth factor (VEGF), and nitric oxide with each other to mediate opening of the mitochondrial (NO) following 2 weeks STZ induction but the opposite permeability transition pore (PTP), leading to eventual car- finding following 6 weeks treatment, supporting the loss diomyocyte death (Figure 1)[13]. Interestingly, the patho- of the cardioprotective state after 6 weeks [56]. Nawata et logical changes already occurring in diabetic cardiomyopathy al. (2002) [51] also found that 4 weeks of STZ-induced include increased release of ROS [59], abnormal handling diabetes reduced sensitivity to acute IRI in the Langendorff- of calcium, and increased release of inflammatory mediators perfused isolated heart model [51]. This phenomenon of [60]; this experimental information alongside the evidence protection in the acute diabetic setting was also present in that diabetic patients have a worse prognosis following alloxan-induced diabetes in Yucatan pigs [57]. Following 1 myocardial infarction [61], strongly suggest that the diabetic hour of regional coronary artery occlusion and reperfusion heart should be more vulnerable to damage. However, a the in vivo myocardial infarct size was smaller compared to plethora of experimental and clinical studies demonstrated control and this was accompanied by an increased expression a variety of outcomes, which could be highly dependent on of cell survival proteins. Of note, global left ventricular the animal model, experimental protocol, and severity of function was worse in diabetes; however function within the diabetes. In Figure 2, we attempt to summarize the possible area at risk was better [57]. In a cardiomyocyte model of mechanisms that make the diabetic heart more or less diabetes, cells were incubated for 3 days with either 5 mM or susceptible to infarction following ischemia reperfusion. 25 mM glucose in the medium. High-glucose treatment was protective against simulated IRI. The high-glucose treatment 3.1. Animal Models of Type I Diabetes. Avastamountof caused a reduction in necrosis, apoptosis, and calcium research has been performed in order to assess the sensitivity content, whereas the antiapoptotic protein bcl-2 increased to myocardial ischemic damage in models of type I diabetes. and proapoptotic bad was shifted to its inactive state [49]in In vivo [53, 56]andex vivo [58] investigations using a the presence of 25 mM glucose. STZ-rat model demonstrated that acute diabetes (1–4 weeks In contrast to this data, showing that “acute” or short- of STZ treatment) resulted in less susceptibility to acute term induced type I diabetes is cardioprotective compared IRI. However, if STZ treatment was increased to longer to the longer-term induction, Hadour et al. (1998) [48] than 6 weeks, the sensitivity to acute IRI was increased. induced diabetes in rabbits for 8 weeks using alloxan, then 4 Cardiology Research and Practice

Table 1: Studies indicating the diabetic heart is more sensitive to ischemic injury compared to normoglycemic controls.

Duration/onset Model of Study Model Ischemic protocol Substrates End points of diabetes diabetes In vivo non recovery, Jones et al. 30 min regional Type II db/db mouse In-bred strain In vivo substrates Infarction (1999) [26] ischemia/2 h diabetes reperfusion In vivo non recovery, Kersten et al. Dog, Alloxan (40 mg/kg) 60 min regional Type I 3weeks In vivo substrates Infarction (2000) [27] and STZ (25 mg/kg) ischemia/3 h diabetes reperfusion In vivo non recovery, Dog, Dextrose 15% to Kersten et al. 60 min regional Type I cause acute 70 mins In vivo substrates Infarction (2000) [27] ischemia/3 h diabetes hyperglycaemia reperfusion In vivo non recovery, Lefer et al. 30 min regional Type II db/db mouse In-bred strain In vivo substrates Infarction (2001) [28] ischemia/2 h diabetes reperfusion 1, 2, and 4 days Fiordaliso et of 25 mmol/L Type I al. (2001) Rat cardiomyocytes — — Cell death incubation in diabetes [29] medium In vivo non recovery, Marfella et al. Sprague-Dawley Rat, STZ 25 min regional Type I Infarction and 9days In vivo substrates (2002) [30] (70 mg/kg i.v) ischemia/2 h diabetes protein expression reperfusion Langendorff isolated Marfella et al. Sprague-Dawley Rat, heart, 25 min regional 33.3 mmol/L Type I Infarction and — (2002) [30] isolated heart ischemia/2 hr glucose diabetes protein expression reperfusion Rabbit- 50% Dextrose In vivo non recovery, hyperglycemia infused 30 min prior to Ebel et al. 30 min regional of 600 mgd1-1 Type I ischemia until reperfusion In vivo substrates Infarction (2003) [31] ischemia/2 h throughout diabetes normoglycaemic reperfusion ischemia rat-under intravenous infusion at a rate of In vivo non recovery, 4mL·kg–1·h–1: of Infarction, apoptosis Su et al. 30 min regional Type I glucose 500 g/l during — In vivo substrates and kinase (2007) [32] ischemia/6 h diabetes ischemia, saline during expression reperfusion reperfusion 1.2 mM Langendorff isolated palmitate, 3% heart, 32 min low albumin, 11 mM Desrois et al. Aging Goto Kakizaki Rat, Type II flow global In-bred strain glucose, 3 U/l Myocardial function (2010) [33] male diabetes ischemia/32 min insulin, 0.8 mM reperfusion lactate, and 0.2 mM pyruvate. subjected the hearts to in vivo 30 minute myocardial ischemia model of alloxan-induced diabetes in rabbits that 6-weeks and 3 hours reperfusion and saw a reduction in MI size duration of diabetes had no influence on the vulnerability compared to control nondiabetic rabbits. Conversely, in to IRI compared to controls, further supporting the theory nondiabetic rabbits infused with high glucose throughout of a chronic (greater than 6 weeks) protected state in this myocardial ischemia and reperfusion to mimic acute diabetes model [31]. Chronic administration of STZ for 12 weeks no difference in infarction was seen compared to controls. in rats also showed a decreased susceptibility to MI [52]. They suggested that the presence of type I diabetes in the Diabetic hearts had greater left ventricular function and rabbit induced a chronic and metabolic cardioprotective the incidence of ventricular fibrillation and creatine kinase state in the heart [48]. Ebel et al. [31] showed in the same (CK) release were decreased. This was accompanied by Cardiology Research and Practice 5

Table 2: Studies indicating the diabetic heart is less sensitive to ischemic injury compared to normoglycaemic controls.

Duration/onset of Model of Study Model Ischemic Protocol Substrates End points diabetes diabetes In vivo nonrecovery, Hadour et al. Rabbit, alloxan 30 min regional In vivo Type I 2months Infarction (1998) [48] (100 mg/kg) ischemia/3 hr substrates diabetes reperfusion 10 mM deoxyglucose and and 3 mM 3-day incubation ff amobarbital medium Scha er et al. Rat, neonatal with 25 mM Type I for1hr,ORhypoxic — Infarction (2000) [49] cardiomyocytes glucose in diabetes chamber: 2.3% O2–5% medium CO2-balance N2 for 1hr Goto Kakizaki Rat, Oliveria et al. male, isolated Type II Cell death and — In-bred strain — (2001) [50] cardiomyocyte diabetes mPTP mitochondria Langendorff isolated Nawata et al. Rat, STZ heart, 30 min low flow 11 mmol/L Type I Myocardial 4weeks (2002) [51] (65 mg/kg) global ischemia/30 min glucose diabetes function reperfusion Langendorff isolated heart: Low-flow global Myocardial Ooie et al. Rat, STZ ischemia for 5 min, 11 mmol/L Type I 12 weeks function, creatine (2003) [52] (65 mg/kg) followed by no-flow glucose diabetes kinase release ischemia for 25 min. 30 min reperfusion In vivo non recovery, Ravingerovaet´ Rat, STZ 30 min regional In vivo Type I 1week Infarction al. (2003) [53] (45 mg/kg) ischemia/4 hr substrates diabetes reperfusion Langendorff isolated Kristiansen et al. Goto Kakizaki Rat, heart, 50 min regional 11 mmol/L Type II In-bred strain Infarction (2004) [54] male ischemia/2 hr glucose diabetes reperfusion Langendorff isolated Obese Zucker Kristiansen et al. heart, 50 min regional 11 mmol/L Type II Diabetic Fatty Rat, In-bred strain Infarction (2004) [54] ischemia/2 hr glucose diabetes male reperfusion Langendorff isolated Tsang et al. Goto Kakizaki Rat, heart, 30 min regional 11 mmol/L Type II Infarction, kinase In-bred strain (2005) [55] male ischemia/2 hr glucose diabetes expression reperfusion In vivo non recovery, Ma et al. (2006) Rat, STZ 30 min regional In vivo Type I 2weeks Infarction [56] (50 mg/kg) ischemia/2 hr substrates diabetes reperfusion In vivo non recovery, Yucatan pigs, Chu et al. 1 hr regional In vivo Type I Infarction and alloxan 5weeks (2010) [57] ischemia/2 hr substrates diabetes protein expression (200 mg/kg) reperfusion Langendorff isolated Infarction and Shi-Ting et al. Rat, STZ heart, 30 min regional 11 mmol/L Type I 4weeks creatine kinase (2010) [58] (60 mg/kg) ischemia/40 min glucose diabetes release reperfusion 6 Cardiology Research and Practice

Table 3: Studies indicating no difference in the sensitivity of the diabetic heart to ischemic injury compared to normoglycemic controls.

Duration/onset Model of Study Model Ischemic Protocol Substrates End points of diabetes diabetes Blood glucose Rabbit, 10% In vivo nonrecovery, maintained at Hadour et al. glucose infusion to 30 min regional In vivo Type I 300 mg/dL Infarction (1998) [48] 300 mg/dL blood ischemia/3 hr substrates diabetes throughout glucose reperfusion procedure In vivo nonrecovery, Dog, alloxan Tanaka et al. 60 min regional In vivo Type I (40 mg/kg) and 3weeks Infarction (2002) [62] ischemia/3 hr substrates diabetes STZ (25 mg/kg) reperfusion Ravingerova´ In vivo nonrecovery, Rat, STZ 30 min regional In vivo Type I et al. (2003) 8weeks Infarction [53] (45 mg/kg) ischemia/4 hr substrates diabetes reperfusion In vivo nonrecovery, Ebel et al. Rabbit- alloxan 30 min regional In vivo Type I 6weeks Infarction (2003) [31] (100 mg/kg) ischemia/2 hr substrates diabetes reperfusion Langendorff isolated heart, 32 min low Desrois et al. Aged Goto 11 mmol/L Type II Myocardial flow global In bred strain (2004) [63] Kakisaki Rat, male glucose diabetes function ischemia/32 min reperfusion In vivo nonrecovery, Ma et al. Rat, STZ 30 min regional In vivo Type I 6weeks Infarction (2006) [56] (50 mg/kg) ischemia/2 hr substrates diabetes reperfusion In vivo nonrecovery, Bulhak et al. Goto Kakizaki Rat, 35 min regional In vivo Type II In bred strain Infarction (2009) [64] male ischemia/2 hr substrates diabetes reperfusion Matsumoto In vivo nonrecovery, Goto Kakizaki Rat, 30 min regional In vivo Type II et al. (2009) In bred strain Infarction [65] male ischemia/2 hr substrates diabetes reperfusion Langendorff isolated Shi-Ting et Infarction and Rat, STZ heart, 30 min regional 11 mmol/L Type I al. (2011) 8weeks creatine kinase (60 mg/kg) ischemia/40 min glucose diabetes [58] release reperfusion a persistent translocation of protein kinase C-ε (PKC-ε), However, in other studies, type I diabetes has been a modulator of the mitochondrial permeability transition shown to render the heart more susceptible to IRI. It has pore, during ischemia but only in the diabetic hearts. Follow- been suggested that hyperglycemia has a negative impact on ing these results, the authors suggested that in STZ-induced endogenous cardioprotective signaling. Kersten et al. [27] diabetes PKC-ε plays a crucial role in the susceptibility to showed that hyperglycemia in dogs, induced either acutely by MI [52]. However, this investigation was performed using 15% dextrose prior to IRI in vivo, or in chemically induced aLangendorff-perfused model of no flow ischemia and diabetes (3 weeks STZ), increased MI size. Interestingly, contained no added substrates. It would be interesting to the studies mentioned above suggested that “acute” diabetes assess translocation properties of PKC-epsilon in the in vivo would render the heart less susceptible to diabetes. However, setting. those studies were in rodent models of diabetes and had Other investigators have also shown that acute initiation a shorter duration of IRI. Ebel et al. [31]recordeda of type I diabetes does not influence the susceptibility to greater susceptibility to MI in their rabbit model following MI. Both alloxan and STZ-treated dogs, although at a lower an infused solution of 50% dextrose 30 minutes prior to dose to previously mentioned investigations, reported no ischemia in vivo to elicit a hyperglycemic state of 600 mg/dL difference in MI size following IRI in vivo [62]. [31]. Su et al. [32] used an infused glucose model to elicit Cardiology Research and Practice 7

(A) More (1) ↑Glucose (B) Less NCX Na/H pump Na/H

(4) ↓ + (3) Na ↑Glucose ↑ROS (2) ↓Ca2+

(1) (2)

(5) ↑p53 ↓HIF-1α P

AKT ε ↑ PKC ↓Caspase-3 (7) ↓VEGF (6) mPTP ↑Apoptosis ↑bcl-2 Mito KATP ↑PKCε ↓Vascularization

↓Ca2+ ↓Cell death ↑Cell death

Figure 2: Possible mechanisms that make the diabetic heart more or less susceptible to infarction following ischemia reperfusion. (A) Diabetes can render the heart more susceptible to infarction. (A1) A diabetes-associated increase in the activity of p53, leading to the initiation of cell death pathway [29]. (A2) High-glucose causes a decrease in the activity of transcription factor HIF-1α, a subsequent downregulation of VEGF and less revascularization following ischemia [66]. This results in cell death and larger infarct volume. (B) Diabetes can protect the heart against infarction. (B1) Hyperglycaemia is cardioprotective due to the increased availability of glucose which is the hearts preferred substrate in times of stress. (B2/3) The Na+/Ca2+ and Na+/H+ exchangers in the diabetic heart reportedly have decreased activity; therefore the diabetic heart accumulates less of these ions preventing overload and the associated detrimental effects [20]. (B4) Diabetes is associated with an increased release of reactive oxygen species (ROS); a possible subsequent release of free radical scavenging enzymes increase the level of antioxidants within the myocardium protecting the heart from the consequence of IRI [20]. (B5) An increased basal level of prosurvival kinases in diabetes [57]. (B6) PKC-ε increases in diabetes, activating the mitochondrial KATP channel causing subsequent reduction in calcium accumulation and increasing ATP synthesis. PKC-ε also persistently translocate during ischemia but only in diabetic hearts [52]. (B7) High glucose caused reduction in cell death proteins and increased anti apoptotic bcl-2 [49].

hyperglycemia in rats, this also rendered the heart more revascularization will occur resulting is potential worse susceptible to MI of note the reperfusion time following outcomes for the diabetic patient [66]. ischemia in this study was notably longer compared to other rodent models [32]. With regard to the possible mechanism, 3.2. Animal Models of Type II Diabetes. In 2004, Kristiansen hyperglycemia in rat cardiomyocytes was shown to promote et al. [54] were the first to investigate the susceptibility to p53-dependent activation of apoptosis [29]. acute IRI and ischemic conditioning (IPC) in a model of Marfella et al. [30] also examined the effect of hyper- type II diabetes. Two distinct models: the GK rat and ZDF glycemia and STZ-induced diabetes on MI size in rats, rat showed in Langendorff-perfused isolated hearts (with both in vivo and ex vivo. Both hyperglycemic conditions no added substrates), that MI size was smaller than their caused an increase in MI size, together with a decreased respective nondiabetic controls. Interestingly, even though transactivation of the hypoxic inducible factor HIF-1α [30]. these hearts appeared to be more resistant to ischemic The diabetic induction in this model was 9 days prior to damage, they were not amenable to protection by IPC [54]. the ischemic episode, although an increased dose of STZ Desrois et al. performed two separate studies also using the (70 mg/kg) was used. HIF-1α controls the upregulation of ex vivo Langendorff preparation in the GK rat in 2004 [63] vascular endothelial growth factor (VEGF) in response to and 2010 [33]. In 2004, they compared gender difference hypoxia, initiating neovascularization following ischemia. If within diabetes and the effect this has on the susceptibility this axis is impaired, as seems the case in diabetes, inadequate to MI. The preparation involved perfusing the hearts with 8 Cardiology Research and Practice

Krebs-Hensleit buffer with no added substrates. They found IPC reduces lethal cell injury in the ischemic myocar- that female hearts had larger MI size than the male GKs and dium [24]; how this phenomenon works has been extensively that no significant difference was noted between male hearts studied and is summarized in Figure 3. Briefly, IPC causes of the diabetic and nondiabetic controls [63]. Interestingly, the release of G-protein-coupled receptor (GPCR) agonists their study in 2010 again utilized the Langendorff technique, which bind to the receptor and activate numerous signaling and showed that the male GK heart was more susceptible pathways. Phosphatidylinositol-3-kinase (PI3K) activation to MI than the control heart [33]. The latter investigation can lead to activation of a number of downstream molecules included additional fatty acid substrates in the perfusate, such as Akt, protein kinase C (PKC), extracellular regulated similar to those likely to be found in the in vivo scenario. This kinase (ERK), nitric oxide synthase (NOS), and inactivation may suggest that in the setting of type II diabetes, substrates of glycogen synthase kinase-3β (GSK-3β).Theseconvergeto found in the blood could play an important role in cell activate the mitochondrial ATP-dependent potassium chan- damage. Studies in db/db mice in vivo support this idea; Lefer nel (KATP ), closing the mitochondrial permeable transition et al. (2001) [28] and Jones et al. (1999) [26]bothsawa pore (mPTP) resulting in protection from IRI [67]. Before diabetes-associated increase in MI size. IPC can be applied in the clinical setting it is important to Other studies show a similar result to Kristiansen et al’s determine whether diabetic heart is amenable to this endoge- study in 2004 [54]. Tsang et al. (2005) using the isolated heart nous cardioprotective strategy. The animal data suggests model, with no added substrates, however with a shorter that the diabetic heart is resistant to ischemic conditioning ischemic time, showed the heart of the GK rat to be less such that the IPC stimulus needs to be increased to induce susceptible to infarction than the Wistar rat control [55]. cardioprotection [70]. Supporting these findings, cardiomyocytes isolated from GK Bouchard et al. (1998) [70] showed that in a STZ rat rats were less susceptible to mPTP opening in response to model of diabetes, an IPC stimulus consisting of 1 cycle of calcium, achieved by adding soluble Ca2+ to a phosphate- 5-minute ischemia followed by 10-minute reperfusion (IPC- containing medium. This was accompanied by a larger 1) was not sufficient to preserve the vasodilatory response to calcium accumulation, leading to decreased opening of the 5-HT suggesting an impairment of endothelial cells in the mitochondrial pore and reduced cardiomyocyte death [50]. diabetic heart and consequently the inability to generate and Controversially, two recent investigations that have been release enough NO. However, when the IPC stimulus was performed in vivo [64, 65] in the GK rat showed no difference increased to 3 cycles of 5-minute ischemia followed by 5- in the susceptibility to infarct compared to control rats minute reperfusion (IPC-3) the vasodilatory response to 5- following 30 or 35 minutes ischemia and 2-hour reperfusion. HT was present. Furthermore, in the same investigation, 30- Some blood components, such as platelets and neu- minute pretreatment with adenosine, an important mediator trophils, have been suggested to play a role as mediators of in IPC [71], mimicked the cardioprotective outcome of cell damage in ischemia and reperfusion [20]. This is another IPC-3, whereas 15-minute pretreatment was ineffective [70]. confounding factor when comparing data from cell, in vitro These findings suggested that signaling pathways activated and in vivo experiments. The sensitivity of the type II diabetic by adenosine either endogenously or exogenously may be heart in in vivo settings has had limited study compared to hindered in diabetic coronary vessels, but have the potential type I diabetes, therefore more work is required to clarify to be recovered by an increased stimuli. The KATP channel these initial findings. is involved in the mechanism of protection by IPC [72]and adenosine [73]; studies in diabetic animals have reported 4. Cardioprotection in the Diabetic Heart reduced sensitivity to KATP channel activators [74]which could correlate with the need for an increased stimulus to 4.1. Animal Models of IRI. Theheart,asinanyliving achieve KATP channel activation, subsequent mPTP inhibi- tissue, has endogenous protective mechanisms which, when tion, and reduction in apoptosis initiation hence reducing activated, render it resistant to IRI, in other words the heart myocardialcelldeath[75]. Another study using alloxan- can be “conditioned” [21]. Initial studies by Murry et al. /STZ-induced diabetes in dogs also supported the role of (1986) [24], showed that short bursts of ischemia followed dysfunctional KATP channel in diabetes [76]. by reperfusion prior to a prolonged insult of ischemia and Other important endogenous mechanisms involved in reperfusion, were associated with a reduction in MI size protection by IPC is the reperfusion injury salvage kinase by 75%. This was termed “ischemic preconditioning” (IPC) pathway (RISK) [77], with its main constituent the PI3K- [24]. The phenomenon of IPC is highly reproducible in Akt axis. Tsang et al. (2005) [55] found that to protect all species [21]. Over the years, the phenomenon of IPC the myocardium in the Goto-Kakizaki rat model of type has evolved to include pharmacological conditioning [68], II diabetes, an increased IPC stimulus was also required. which involves targeting cellular mechanisms involved in IRI Similar to the previous study, 1 cycle of 5-minute ischemia or promoting those involved in IPC. Also, short bursts of followed by 10-minute reperfusion was insufficient to protect ischemia and reperfusion following an ischemic insult can the heart; however 3 cycles of this protocol reduced infarct reduce MI size and this term was coined ischemic post- volume considerably. They investigated the level of phospho- conditioning (IPost) [69]. More recently, remote ischemic rylated Akt (Akt-P) within the myocardium and found that conditioning has been discovered, whereby an organ or tissue in the diabetic heart, an increased IPC stimulus is required remote from the heart is conditioned and reduces infarct size to achieve an essential level of Akt phosphorylation neces- [22]. sary to mediate cardioprotection [55], thereby highlighting Cardiology Research and Practice 9

IPC IPost Ischaemia Reperfusion Pharmacological Agonists agents

GPCR

Extracellular

G-protein complex

Intracellular Ras PI3K

Raf Bad Bad Mek1/2 Akt ERK1/2 P Bcl-2 p70s6k R.I.S.K Bax mPTP GSK-3β

GSK-3β Mito KATP PKCε eNOS mPTP NO Mitochondria Mito KATP PKG Cardioprotection

Figure 3: The cellular mechanisms involved in Ischemic Preconditioning. IPC, IPost, or pharmacological agents initiates the release of G-protein-coupled receptor (GPCR) agonists which bind to the receptor and activate numerous signaling pathways. Phosphatidylinositol- 3-kinase (PI3K) and Ras activation can lead to activation of a number of downstream molecules such as Akt, protein kinase C (PKC), extracellular regulated kinase (ERK), nitric oxide synthase (NOS), and inactivation of glycogen synthase kinase-3β (GSK-3β). These converge to activate the mitochondrial ATP-dependent potassium channel (KATP ), closing the mitochondrial permeable transition pore (mPTP) resulting in protection from IRI [67]. a diabetes-associated impairment in PI3K-Akt signaling. In contrast, Ghaboura et al. (2011), examined EPO- Increasing the stimulus of IPC by increasing cycle number induced IPost cardioprotection in two rat models of appears important in diabetes; however, the duration of type I diabetes; STZ-induced and high-fat-diet (HFD-) the pre-ischemia/reperfusion protocol may also play a vital induced insulin resistance syndrome. They found that EPO role. Four cycles of 2-minute ischemia followed by 3-minute administered at the onset of reperfusion did not increase reperfusion did not elicit protection in two separate models phosphorylation of Akt, ERK, or GSK-3β and hence did of diabetes [54], compared to 3 cycles of 5-minute ischemia not elicit a cardioprotective effect in the hearts isolated and 10-minute reperfusion used by Tsang et al. [55] in the from STZ rats; interestingly, HFD rat hearts were protected same animal and experimental setting. following EPO. Administration of a GSK-3β antagonist, a A recent study by Hotta et al. (2010) [78] using the kinase which is downstream of the PI3K-Akt axis, given OLETF rat model of type II diabetes showed that pre- prior to ischemia and continued throughout reperfusion treatment with either opioid agonist [79] or erythropoietin had an infarct limiting effect [83]. In support of the (EPO) [80], known to mediate IPC signaling through Janus idea that inhibiting GSK-3β in diabetic hearts as a direct kinase (Jak-2), failed to elicit cardioprotection in vivo due cardioprotective strategy, Gross et al. (2007) [84] showed that to insufficient phosphorylation of Jak-2 and Akt, and like diabetes had a detrimental effect on morphine-induced pro- others, also found an elevated level of calcineurin activity tection. STZ-induced diabetes caused alterations in pathways in diabetic hearts [81]. Interestingly, following 2-week treat- upstream of GSK-3β, such as PI3K, Jak/STAT, and MAPK ment with either valsartan or losartan, Angiotensin II recep- pathways, limiting the cardioprotective effects of morphine tor type 1 (AT1) receptor blockers, Jak-PI3K-Akt signaling, administered at reperfusion. Similarly, downstream antag- and hence cardioprotection subsequent to EPO administra- onism of GSK-3β was cardioprotective [84]. The direct tion were restored [78]. Huisamen et al. (2011) supported inhibition of GSK-3β was also proved cardioprotective in a the finding that AT1 antagonism can lead to cardioprotection rat model of type II diabetes [85]. In this study, another in their ex vivo rat model of insulin resistance, diet-induced diabetic-associated myocardial phenotype was reported; they obesity (DIO) rat [82]. found that increased endoplasmic reticulum stress, caused 10 Cardiology Research and Practice alterations in PI3K-Akt, ERK, and GSK-3β signaling motifs, see if the human muscle mirrored the finding from Tsang and and hence dysregulation of the mPTP leading to the absence colleagues [55]. Hassouna et al. (2006) also suggested that the of cardioprotection by EPO [85]. dysfunctional mitochondrial KATP channel exists in diabetics, The idea that the diabetic heart is still amenable to causing impaired depolarization and superoxide production, protection but has an increased threshold for the necessary resulting in an inability to respond to IPC [95]. Increasing activation of pro survival kinases is an interesting prospect. the IPC stimulus to 3 cycles in this experimental model did Investigations by Tsang et al. [55] and Bouchard et al. not restore the protection [95], conflicting with the finding [70] clearly demonstrated this; however, in the study by of Tsang et al. [55]. However, Sivaraman et al. (2010) [88] Ghaboura et al. [83] increasing the dose of EPO did not showed that increasing the duration of ischemic period of restore cardioprotective effects. Of course, these studies are IPC from 4 minutes to 7 minutes followed by reoxygenation performed in a variety of models of diabetes, with variable in human diabetic atrial tissue restored cardioprotection and levels of diabetic severity. Tsang et al. [55] used a model similarly was related to a downregulation of PI3K-Akt axis in with moderate hyperglycemia, however both Bouchard et diabetic tissue [88]. al. [70] and Ghaboura et al. [83] induced diabetes by In a recent study performed by Linares-Palomino et injections of STZ at 55 mg/kg and 65 mg/kg, respectively. al. specific inhibitors of Akt were utilized to delineate its Therefore, speculation that the ability to protect the diabetic specific role in ischemic pre conditioning. They showed that heart can be restored by an increased stimulus in moderate blockade of Akt caused a significant reduction in cell death, hyperglycemia but not severe hyperglycemia could be flawed similar to the degree of protection elicited by either IPC or and needs further investigation. PI3K inhibition and this was evident in both rat and human All of the mentioned studies have furthered the under- tissue. Interestingly, they showed Akt to be downstream of standing of the mechanisms that could render the diabetic mitochondrial KATP channel but upstream of p38 MAPK heart ineffective to cardioprotective strategies, and how we [96]. could potentially overcome these hurdles. Though a variety This data reinforces the fact that cardioprotection in the of animal models were used, including in vivo and ex vivo diabetic heart is a complex and delicate phenomenon; the experimental protocols, direct targeting of GSK-3β in the knowledge gained from the extensive research throughout diabetic heart may provide an alternative solution to prevent many models has highlighted some potential reasons why it ischemia reperfusion injury [86]. However, like any therapy, is more difficult to protect the diabetic heart, reasons which a targeted approach would have to be considered to limit any are summarized in Figure 4. potential negative effects [87]. Proof-of-concept clinical studies are now required to determine whether the human diabetic heart is amenable to 4.2. Experimental Studies Using Human Heart Tissue. As part cardioprotection elicited by ischemic conditioning. Prelimi- of the translational process, our laboratory have established nary unpublished data from our research group suggests that since 1995 an isolated human atrial muscle preparation of the diabetic patient may not be amenable to cardioprotection simulated IRI [90]. This preparation has become a popular elicited by remote ischemic conditioning (RIC). There was tool in investigating the response of human tissue in cardio- no difference in perioperative myocardial IRI (72 hour area vascular diseases. In brief, samples of right atrial appendage under the curve serum Troponin T) in diabetic patients are obtained from the right atrial cannula insertion site in randomized to receive a standard RIC stimulus (three 5- patients undergoing cardiac bypass during coronary artery minute inflations and deflations of a blood pressure cuff bypass graft surgery [91]. From the right atrial appendage, on the upper arm) prior to CABG surgery (Babu et al. trabeculae are isolated and subjected to simulated IRI and unpublished), a RIC protocol which has been previously the recovery of basal function assessed following IRI. reported to protect nondiabetic patients [97]. Numerous studies have been performed investigating the mechanisms of preconditioning in the normoglycemic human myocardium [92]; however a limited number of 5. Limitation of Translation and Future studies also investigated this phenomena in the diabetic Directions human heart. Using a variation of this model in which the right atrial appendage is sliced and subjected to simulated The obvious limitation of translating research from bench to IRI, Ghosh et al. (2001) [93], showed the failure of IPC bedside is that many of the animal models we use do not to protect the diabetic human myocardium with 1 cycle adequately reproduce the clinical setting or fully represent of 5-minute ischemia and 5-minute reperfusion reflecting the disease pathology and manifestation in patients. In what was already demonstrated in animal investigations, addition, the vast majority of studies are conducted on that is, an increased threshold for cardioprotection in inbred animals with very little genetic variability; they are the hyperglycemic tissue. As previously mentioned KATP fed ad libitum with scientific diets and are kept in optimal channels play a role in IPC. Diazoxide, a KATP opener, did living conditions of light and temperature. This is far from not mimic IPC in diabetic tissues suggesting a dysfunctional the complexity of patients who possess a large genetic pool, mitochondrial KATP channel [93]. This result was supported undergo the stresses of day to day working life, some eating by the finding that the ATP-sensitive potassium channels are high-fat diets and undertaking limited exercise. In addition, altered in ventricular myocytes from diabetic rats [94]. Of the majority of people in westernized societies take a plethora interest, this group did not try increasing the IPC stimulus to of medications. Cardiology Research and Practice 11

Increased AT-1 ER stress receptor activity

AT-1 ↓PI3K Akt Alterations in kinase pathways

PI3K-Akt ERK GSK-3β

Jak Akt

Dysregulation PP of mPTP Loss of IPC in diabetic heart Decrease in survival kinase activity

Dysfunctional mito KATP channel KATP Diabetic heart is Calcineurin already in a activity conditioned state

Figure 4: Why is the diabetic heart harder to protect with conditioning strategies? The diabetic heart has been suggested to have a raised threshold for cardioprotection [55], this is caused by the downregulation of prosurvival kinase pathways [55, 88], resulting in dysregulation of mitochondrial permeability transition pore (mPTP), increased receptor activities for pharmacological agents [78], increased calcineurin activity [81] and evidence suggests a dysfunctional KATP channel in the mitochondria [76]. In diabetes, endoplasmic reticulum (ER) stress also causes alterations in kinase pathways leading to dysregulation of the mPTP [85]. Interestingly, some evidence suggests that the diabetic heart is in a paradoxical protective state therefore conditioning potential is lower [89].

There is an increasing amount of research now being comorbidities such as age, dyslipidaemia, and hypertension. conducted in animals with comorbidities to promote suc- Therefore, by carefully selecting and optimizing clinically cessful translation of the research field of cardioprotection. relevant animal models of IRI and diabetes, we may be able The design of the preclinical model, either in human tissue to better translate findings made at the “bench” to patients’ or animal models, is critical to harness the huge potential “bedside”. of IPC as a strategy to reduce ischemic damage. Preliminary unpublished data from our research group, utilizing aging, Conflict of Interests diabetic rats suggests an age-associated increase to the threshold of IPC and susceptibility to infarction in models There is no conflict of interests to declare. of type II diabetes (Whittington et al. unpublished). These results could highlight a possible discrepancy from original IPC studies performed in young healthy animals and could Acknowledgments indicate why the findings were not transferable to the clinic. The authors thank the British Heart Foundation (Program Grant RG/03/007 and FS/10/039/28270) for ongoing funding 6. Summary and support. This work was undertaken at University College London Hospital/University College London (UCLH/UCL) It is well established that diabetic patients with IHD who received a proportion of funding from the Department experience worse clinical outcomes yet the animal data is of Health’s National Institute of Health Research (NIHR) conflicting. The choice of the diabetic and IRI animal model Biomedical Research Centres funding scheme. may in part explain this discrepancy. The development of novel cardioprotective strategies for protecting the diabetic References heart and improving clinical outcomes in diabetic patients with IHD will be dependent on using relevant animal models [1] International Diabetes Federation, Diabetes Atlas, 4th edition, of IRI and diabetes which also take into account other 2009. 12 Cardiology Research and Practice

[2] A. L. Notkins and A. Lernmark, “Autoimmune type 1 artery bypass surgery in relation to type of diabetes,” American diabetes: resolved and unresolved issues,” Journal of Clinical Heart Journal, vol. 152, no. 3, pp. 599–605, 2006. Investigation, vol. 108, no. 9, pp. 1247–1252, 2001. [18]A.M.Calafiore,M.DiMauro,G.DiGiammarcoetal., [3] B. J. Goldstein, “Insulin resistance as the core defect in type 2 “Effect of diabetes on early and late survival after isolated diabetes mellitus,” American Journal of Cardiology, vol. 90, no. first coronary bypass surgery in multivessel disease,” Journal of 5, 2002. Thoracic and Cardiovascular Surgery, vol. 125, no. 1, pp. 144– [4] A. Fagot-Campagna, K. M. Venkat Narayan, and G. Imper- 154, 2003. atore, “Type 2 diabetes in children: exemplifies the growing [19] V. H. Thourani, W. S. Weintraub, B. Stein et al., “Influence problem of chronic diseases,” British Medical Journal, vol. 322, of diabetes mellitus on early and late outcome after coronary no. 7283, pp. 377–378, 2001. artery bypass grafting,” Annals of Thoracic Surgery, vol. 67, no. [5] R. Butler, T. M. Macdonald, A. D. Struthers, and A. D. Morris, 4, pp. 1045–1052, 1999. “The clinical implications of diabetic heart disease,” European [20] D. J. Paulson, “The diabetic heart is more sensitive to ischemic Heart Journal, vol. 19, no. 11, pp. 1617–1627, 1998. injury,” Cardiovascular Research, vol. 34, no. 1, pp. 104–112, [6]S.A.Hayat,B.Patel,R.S.Khattar,andR.A.Malik,“Diabetic 1997. cardiomyopathy: mechanisms, diagnosis and treatment,” Clin- [21] D. M. Yellon and J. M. Downey, “Preconditioning the ical Science, vol. 107, no. 6, pp. 539–557, 2004. myocardium: from cellular physiology to clinical cardiology,” Physiological Reviews, vol. 83, no. 4, pp. 1113–1151, 2003. [7] C. G. Schalkwijk and C. D. A. Stehouwer, “Vascular complica- tions in diabetes mellitus: the role of endothelial dysfunction,” [22] D. J. Hausenloy and D. M. Yellon, “Remote ischaemic precon- Clinical Science, vol. 109, no. 2, pp. 143–159, 2005. ditioning: underlying mechanisms and clinical application,” Cardiovascular Research, vol. 79, no. 3, pp. 377–386, 2008. [8] P. Dandona, A. Aljada, A. Chaudhuri, and A. Bandyopadhyay, [23] I. Andreadou, E. K. Iliodromitis, M. Koufaki, and D. T. “The potential influence of inflammation and insulin resis- Kremastinos, “Pharmacological pre- and post- condition- tance on the pathogenesis and treatment of atherosclerosis- ing agents: reperfusion-injury of the heart revisited,” Mini- related complications in type 2 diabetes,” Journal of Clinical Reviews in Medicinal Chemistry, vol. 8, no. 9, pp. 952–959, Endocrinology and Metabolism, vol. 88, no. 6, pp. 2422–2429, 2008. 2003. [24] C. E. Murry, R. B. Jennings, and K. A. Reimer, “Precondi- [9]J.A.Colwell,M.Lopes-Virella,andP.V.Halushka,“Pathogen- tioning with ischemia: a delay of lethal cell injury in ischemic esis of atherosclerosis in diabetes mellitus,” Diabetes Care, vol. myocardium,” Circulation, vol. 74, no. 5, pp. 1124–1136, 1986. 4, no. 1, pp. 121–133, 1981. [25] P. Ferdinandy, R. Schulz, and G. F. Baxter, “Interaction of car- [10] R. Ross, “Atherosclerosis—an inflammatory disease,” The New diovascular risk factors with myocardial ischemia/reperfusion England Journal of Medicine, vol. 340, no. 2, pp. 115–126, 1999. injury, preconditioning, and postconditioning,” Pharmacolog- [11] “ Diabetes mellitus: a major risk factor for cardiovascular ical Reviews, vol. 59, no. 4, pp. 418–458, 2007. disease. A joint editorial statement by the American Diabetes [26] S. P. Jones, W. G. Girod, D. N. Granger, A. J. Palazzo, and Association; The National Heart, Lung, and Blood Institute; D. J. Lefer, “Reperfusion injury is not affected by blockade of The Juvenile Diabetes Foundation International; The National P-selectin in the diabetic mouse heart,” American Journal of Institute of Diabetes and Digestive and Kidney Diseases; and Physiology, vol. 277, no. 2, pp. H763–H769, 1999. The American Heart Association,” Circulation, vol. 100, pp. [27] J. R. Kersten, W. G. Toller, E. R. Gross, P. S. Pagel, and 1132–1133, 1999. D. C. Warltier, “Diabetes abolishes ischemic preconditioning: [12] Diabetes UK, Diabetes in the UK 2010, Key statistics on role of glucose, insulin, and osmolality,” American Journal of diabetes, 2010. Physiology, vol. 278, no. 4, pp. H1218–H1224, 2000. [13] D. M. Yellon and D. J. Hausenloy, “Myocardial reperfusion [28] D. J. Lefer, R. Scalia, S. P. Jones et al., “HMG-CoA reductase injury,” The New England Journal of Medicine, vol. 357, no. 11, inhibition protects the diabetic myocardium from ischemia- pp. 1021–1135, 2007. reperfusion injury,” The FASEB Journal,vol.15,no.8,pp. [14] S. M. Haffner, S. Lehto, T. Ronnemaa,¨ K. Pyor¨ al¨ a,¨ and M. 1454–1456, 2001. Laakso, “Mortality from coronary heart disease in subjects [29] F. Fiordaliso, A. Leri, D. Cesselli et al., “Hyperglycemia with type 2 diabetes and in nondiabetic subjects with and activates p53 and p53-regulated genes leading to myocyte cell without prior myocardial infarction,” The New England Jour- death,” Diabetes, vol. 50, no. 10, pp. 2363–2375, 2001. nal of Medicine, vol. 339, no. 4, pp. 229–234, 1998. [30] R. Marfella, M. D’Amico, C. Di Filippo et al., “Myocardial [15]K.Malmberg,S.Yusuf,H.C.Gersteinetal.,“Impactof infarction in diabetic rats: role of hyperglycaemia on infarct diabetes on long-term prognosis in patients with unstable size and early expression of hypoxia-inducible factor 1,” angina and non-Q-wave myocardial infarction: results of Diabetologia, vol. 45, no. 8, pp. 1172–1181, 2002. the OASIS (Organization to Assess Strategies for Ischemic [31] D. Ebel, J. Mullenheim,¨ J. Fraßdorf¨ et al., “Effect of acute Syndromes) registry,” Circulation, vol. 102, no. 9, pp. 1014– hyperglycaemia and diabetes mellitus with and without 1019, 2000. short-term insulin treatment on myocardial ischaemic late [16]V.Mathew,B.J.Gersh,B.A.Williamsetal.,“Outcomes preconditioning in the rabbit heart in vivo,” Pflugers Archiv, in patients with diabetes mellitus undergoing percutaneous vol. 446, no. 2, pp. 175–182, 2003. coronary intervention in the current era: a report from the [32] H. Su, X. Sun, H. Ma et al., “Acute hyperglycemia exacerbates prevention of REStenosis with tranilast and its outcomes myocardial ischemia/reperfusion injury and blunts cardiopro- (PRESTO) trial,” Circulation, vol. 109, no. 4, pp. 476–480, tective effect of GIK,” American Journal of Physiology, vol. 293, 2004. no. 3, pp. E629–E635, 2007. [17]T.Alserius,N.Hammar,T.Nordqvist,andT.Ivert,“Riskof [33] M. Desrois, K. Clarke, C. Lan et al., “Upregulation of eNOS death or acute myocardial infarction 10 years after coronary and unchanged energy metabolism in increased susceptibility Cardiology Research and Practice 13

of the aging type 2 diabetic GK rat heart to ischemic injury,” induced apoptosis and necrosis,” American Journal of Physi- American Journal of Physiology, vol. 299, no. 5, pp. H1679– ology, vol. 278, no. 6, pp. H1948–H1954, 2000. H1686, 2010. [50]P.J.Oliveira,A.P.Rolo,R.Seiac˜¸a,C.M.Palmeira,M.S. [34] D. A. Rees and J. C. Alcolado, “Animal models of diabetes Santos, and A. J. M. Moreno, “Decreased susceptibility of heart mellitus,” Diabetic Medicine, vol. 22, no. 4, pp. 359–370, 2005. mitochondria from diabetic GK rats to mitochondrial perme- [35] A. Chatzigeorgiou, A. Halapas, K. Kalafatakis, and E. F. ability transition induced by calcium phosphate,” Bioscience Kamper, “The use of animal models in the study of diabetes Reports, vol. 21, no. 1, pp. 45–53, 2001. mellitus,” In Vivo, vol. 23, no. 2, pp. 245–258, 2009. [51] T. Nawata, N. Takahashi, T. Ooie, K. Kaneda, T. Saikawa, [36] J. W. Yoon and H. S. Jun, “Cellular and molecular pathogenic and T. Sakata, “Cardioprotection by streptozotocin-induced mechanisms of insulin-dependent diabetes mellitus,” Annals of diabetes and insulin against ischemia/reperfusion injury in the New York Academy of Sciences, vol. 928, pp. 200–211, 2001. rats,” Journal of Cardiovascular Pharmacology, vol. 40, no. 4, pp. 491–500, 2002. [37] A. D. Bolzan´ and M. S. Bianchi, “Genotoxicity of streptozo- tocin,” Mutation Research, vol. 512, no. 2-3, pp. 121–134, 2002. [52] T. Ooie, N. Takahashi, T. Nawata et al., “Ischemia-induced translocation of protein kinase C-ε mediates cardioprotection [38] A. A. Like and A. A. Rossini, “Streptozotocin induced in the streptozotocin-induced diabetic rat,” Circulation Jour- pancreatic insulitis: new model of diabetes mellitus,” Science, nal, vol. 67, no. 11, pp. 955–961, 2003. vol. 193, no. 4251, pp. 415–417, 1976. [53] T. Ravingerova,´ J. Necka´r,ˇ and F. Kola´r,ˇ “Ischemic tolerance [39] G. L. Wilson and E. H. Leiter, “Streptozotocin interactions β of rat hearts in acute and chronic phases of experimental with pancreatic cells and the induction of insulin-dependent diabetes,” Molecular and Cellular Biochemistry, vol. 249, no. 1- diabetes,” Current Topics in Microbiology and Immunology, vol. 2, pp. 167–174, 2003. 156, pp. 27–54, 1990. [54] S. B. Kristiansen, B. Løfgren, N. B. Støttrup et al., “Ischaemic [40] S. Lenzen, “The mechanisms of alloxan- and streptozotocin- preconditioning does not protect the heart in obese and lean induced diabetes,” Diabetologia, vol. 51, no. 2, pp. 216–226, animal models of type 2 diabetes,” Diabetologia, vol. 47, no. 2008. 10, pp. 1716–1721, 2004. [41] W. C. Knowler, E. Barrett-Connor, S. E. Fowler et al., [55] A. Tsang, D. J. Hausenloy, M. M. Mocanu, R. D. Carr, and D. “Reduction in the incidence of type 2 diabetes with lifestyle M. Yellon, “Preconditioning the diabetic heart: the importance intervention or metformin,” The New England Journal of of Akt phosphorylation,” Diabetes, vol. 54, no. 8, pp. 2360– Medicine, vol. 346, no. 6, pp. 393–403, 2002. 2364, 2005. [42] B. Portha, P. Serradas, D. Bailbe, K. I. Suzuki, Y. Goto, and [56] G. Ma, M. Al-Shabrawey, J. A. Johnson et al., “Protection β M. H. Giroix, “ -Cell insensitivity to glucose in the GK rat, A against myocardial ischemia/reperfusion injury by short- spontaneous nonobese model for type II diabetes,” Diabetes, term diabetes: enhancement of VEGF formation, capillary vol. 40, no. 4, pp. 486–491, 1991. density, and activation of cell survival signaling,” Naunyn- [43] F. C. Howarth, M. Shafiullah, and M. A. Qureshi, “Chronic Schmiedeberg’s Archives of Pharmacology, vol. 373, no. 6, pp. effects of type 2 diabetes mellitus on cardiac muscle contrac- 415–427, 2006. tion in the Goto-Kakizaki rat,” Experimental Physiology, vol. [57] L. M. Chu, R. M. Osipov, M. P.Robich et al., “Is hyperglycemia 92, no. 6, pp. 1029–1036, 2007. bad for the heart during acute ischemia?” Journal of Thoracic [44] K. Kawano, T. Hirashima, S. Mori, Y. Saitoh, M. Kurosumi, and Cardiovascular Surgery, vol. 140, no. 6, pp. 1345–1352, and T. Natori, “Spontaneous long-term hyperglycemic rat 2010. with diabetic complications: Otsuka Long-Evans Tokushima [58] W. Shi-ting, X. Mang-hua, C. Wen-ting, G. Feng-hou, and G. Fatty (OLETF) strain,” Diabetes, vol. 41, no. 11, pp. 1422– Zhu-ying, “Study on tolerance to ischemia-reperfusion injury 1428, 1992. and protection of ischemic preconditioning of type 1 diabetes [45] B. L. Leonard, R. N. Watson, K. M. Loomes, A. R. J. Phillips, rat heart,” Biomedicine and Pharmacotherapy. In press. and G. J. Cooper, “Insulin resistance in the Zucker diabetic [59] J. L. Mehta, N. Rasouli, A. K. Sinha, and B. Molavi, “Oxidative fatty rat: a metabolic characterisation of obese and lean stress in diabetes: a mechanistic overview of its effects phenotypes,” Acta Diabetologica, vol. 42, no. 4, pp. 162–170, on atherogenesis and myocardial dysfunction,” International 2005. Journal of Biochemistry and Cell Biology, vol. 38, no. 5-6, pp. [46] K. Kobayashi, T. M. Forte, S. Taniguchi, B. Y. Ishida, K. Oka, 794–803, 2006. and L. Chan, “The db/db mouse, a model for diabetic dyslipi- [60] S. Boudina and E. D. Abel, “Diabetic cardiomyopathy revis- demia: molecular characterization and effects of western diet ited,” Circulation, vol. 115, no. 25, pp. 3213–3223, 2007. feeding,” Metabolism: Clinical and Experimental,vol.49,no.1, [61] M. Jonas, H. Reicher-Reiss, V. Boyko, S. Behar, and E. Gross- pp. 22–31, 2000. man, “Hospital and 1-year outcome after acute myocardial [47] B. Zhang, M. P. Graziano, T. W. Doebber et al., “Down- infarction in patients with diabetes mellitus and hyperten- regulation of the expression of the obese gene by an antidia- sion,” Journal of Human Hypertension, vol. 17, no. 10, pp. 665– betic thiazolidinedione in Zucker diabetic fatty rats and db/db 670, 2003. mice,” Journal of Biological Chemistry, vol. 271, no. 16, pp. [62] K. Tanaka, F. Kehl, W. Gu et al., “Isoflurane-induced pre- 9455–9459, 1996. conditioning is attenuated by diabetes,” American Journal of [48] G. Hadour, R. Ferrera, L. Sebbag, R. Forrat, J. Delaye, and Physiology, vol. 282, no. 6, pp. H2018–H2023, 2002. M. De Lorgeril, “Improved myocardial tolerance to ischaemia [63] M. Desrois, R. J. Sidell, D. Gauguier, C. L. Davey, G. K. Radda, in the diabetic rabbit,” Journal of Molecular and Cellular and K. Clarke, “Gender differences in hypertrophy, insulin Cardiology, vol. 30, no. 9, pp. 1869–1875, 1998. resistance and ischemic injury in the aging type 2 diabetic rat [49] S. W. Schaffer, C. B. Croft, and V. Solodushko, “Cardiopro- heart,” Journal of Molecular and Cellular Cardiology, vol. 37, tective effect of chronic hyperglycemia: effect on hypoxia- no. 2, pp. 547–555, 2004. 14 Cardiology Research and Practice

[64] A. A. Bulhak, C. Jung, C. G. Ostenson,¨ J. O. Lundberg, [79] E. R. Gross, A. K. Hsu, and G. J. Gross, “The JAK/STAT P. O. Sjoquist, and J. Pernow, “PPAR-α activation protects pathway is essential for opioid-induced cardioprotection: the type 2 diabetic myocardium against ischemia-reperfusion JAK2 as a mediator of STAT3, Akt, and GSK-3β,” American injury: involvement of the PI3-kinase/Akt and NO pathway,” Journal of Physiology, vol. 291, no. 2, pp. H827–H834, 2006. American Journal of Physiology, vol. 296, no. 3, pp. H719– [80] J. E. Baker, “Erythropoietin mimics ischemic precondition- H727, 2009. ing,” Vascular Pharmacology, vol. 42, no. 5-6, pp. 233–241, [65] S. Matsumoto, S. Cho, S. Tosaka et al., “Pharmacological 2005. preconditioning in type 2 diabetic rat hearts: the roles [81] J. L. Gooch, J. L. Barnes, S. Garcia, and H. E. Abboud, of mitochondrial atp-sensitive potassium channels and the “Calcineurin is activated in diabetes and is required for phosphatidylinositol 3-kinase-akt pathway,” Cardiovascular glomerular hypertrophy and ECM accumulation,” American Drugs and Therapy, vol. 23, no. 4, pp. 263–270, 2009. Journal of Physiology, vol. 284, no. 1, pp. F144–F154, 2003. [66] H. Thangarajah, D. Yao, E. I. Chang et al., “The molecular [82]B.Huisamen,S.J.C.Perel,S.O.Friedrich,R.Salie,H.ˆ basis for impaired hypoxia-induced VEGF expression in Strijdom, and A. Lochner, “ANG II type i receptor antagonism diabetic tissues,” Proceedings of the National Academy of improved nitric oxide production and enhanced eNOS and Sciences of the United States of America, vol. 106, no. 32, pp. PKB/Akt expression in hearts from a rat model of insulin 13505–13510, 2009. resistance,” Molecular and Cellular Biochemistry, vol. 349, no. [67] E. Murphy and C. Steenbergen, “Preconditioning: the mito- 1-2, pp. 21–31, 2011. chondrial connection,” Annual Review of Physiology, vol. 69, [83] N. Ghaboura, S. Tamareille, P. H. Ducluzeau et al., “Diabetes pp. 51–67, 2007. mellitus abrogates erythropoietin-induced cardioprotection [68] D. J. Hausenloy and D. M. Yellon, “The therapeutic potential against ischemic-reperfusion injury by alteration of the β of ischemic conditioning: an update,” Nature Reviews Cardiol- RISK/GSK-3 signaling,” Basic Research in Cardiology, vol. ogy, vol. 8, no. 11, pp. 619–629, 2011. 106, no. 1, pp. 147–162, 2011. [69] M. Ovize, G. F. Baxter, F. Di Lisa et al., “Postconditioning and [84] E. R. Gross, A. K. Hsu, and G. J. Gross, “Diabetes abolishes protection from reperfusion injury: where do we stand: posi- morphine-induced cardioprotection via multiple pathways β tion Paper from the Working Group of Cellular Biology of the upstream of glycogen synthase kinase-3 ,” Diabetes, vol. 56, Heart of the European Society of Cardiology,” Cardiovascular no. 1, pp. 127–136, 2007. Research, vol. 87, no. 3, pp. 406–423, 2010. [85]T.Miki,T.Miura,H.Hottaetal.,“Endoplasmicreticulum [70] J. F. Bouchard and D. Lamontagne, “Protection afforded stress in diabetic hearts abolishes erythropoietin-induced myocardial protection by impairment of phospho-glycogen by preconditioning to the diabetic heart against ischaemic β injury,” Cardiovascular Research, vol. 37, no. 1, pp. 82–90, synthase kinase-3 -mediated suppression of mitochondrial 1998. permeability transition,” Diabetes, vol. 58, no. 12, pp. 2863– 2872, 2009. [71] J. W. De Jong, R. De Jonge, E. Keijzer, and S. Bradamante, [86] T. Miura, M. Nishihara, and T. Miki, “Drug development “The role of adenosine in preconditioning,” Pharmacology and targeting the glycogen synthase kinase-3β (GSK-3β)-mediated Therapeutics, vol. 87, no. 2-3, pp. 141–149, 2000. signal transduction pathway: role of GSK-3β in myocardial [72] B. O’Rourke, “Evidence for mitochondrial K+ channels and protection against ischemia /reperfusion injury,” Journal of their role in cardioprotection,” Circulation Research, vol. 94, Pharmacological Sciences, vol. 109, no. 2, pp. 162–167, 2009. no. 4, pp. 420–432, 2004. [87] H. Cheng, J. Woodgett, M. Maamari, and T. Force, “Targeting [73] V. J. McIntosh and R. D. Lasley, “Adenosine receptor-mediated GSK-3 family members in the heart: a very sharp double- cardioprotection: are all 4 subtypes required or redundant?” edged sword,” Journal of Molecular and Cellular Cardiology, Journal of Cardiovascular Pharmacology and Therapeutics, vol. vol. 51, no. 4, pp. 607–613, 2011. 17, no. 1, pp. 21–33, 2012. [88] V. Sivaraman, D. J. Hausenloy, A. M. Wynne, and D. M. Yellon, [74] K. Kamata, N. Miyata, and Y. Kasuya, “Functional changes in “Preconditioning the diabetic human myocardium,” Journal of potassium channels in aortas from rats with streptozotocin- Cellular and Molecular Medicine, vol. 14, no. 6, pp. 1740–1746, induced diabetes,” EuropeanJournalofPharmacology, vol. 166, 2010. no. 2, pp. 319–323, 1989. [89] T. Ravingerova, A. Adameova, J. Matejikova et al., “Subcellular [75] D. J. Hausenloy, H. L. Maddock, G. F. Baxter, and D. M. mechanisms of adaptation in the diabetic myocardium: rele- Yellon, “Inhibiting mitochondrial permeability transition pore vance to ischemic preconditioning in the nondiseased heart,” opening: a new paradigm for myocardial preconditioning?” Experimental and Clinical Cardiology, vol. 15, pp. 68–76, 2010. Cardiovascular Research, vol. 55, no. 3, pp. 534–543, 2002. [90]D.M.Walker,J.M.Walker,W.B.Pugsley,C.W.Pattison,and [76] J. R. Kersten, M. W. Montgomery, T. Ghassemi et al., “Diabetes D. M. Yellon, “Preconditioning in isolated superfused human and hyperglycemia impair activation of mitochondrial KATP muscle,” Journal of Molecular and Cellular Cardiology, vol. 27, channels,” American Journal of Physiology, vol. 280, no. 4, pp. no. 6, pp. 1349–1357, 1995. H1744–H1750, 2001. [91] M. E. Speechly-Dick, G. J. Graver, and D. M. Yellon, “Does [77] D. J. Hausenloy, A. Tsang, and D. M. Yellon, “The reperfusion ischemic preconditioning in the human involve protein kinase injury salvage kinase pathway: a common target for both C and the ATP-dependent K+ channel? Studies of contractile ischemic preconditioning and postconditioning,” Trends in function after simulated ischemia in an atrial in vitro model,” Cardiovascular Medicine, vol. 15, no. 2, pp. 69–75, 2005. Circulation Research, vol. 77, no. 5, pp. 1030–1035, 1995. [78] H. Hotta, T. Miura, T. Miki et al., “Short communication: [92] R. A. Kloner and S. H. Rezkalla, “Preconditioning, post- angiotensin II type 1 receptor-mediated upregulation of conditioning and their application to clinical cardiology,” calcineurin activity underlies impairment of cardioprotective Cardiovascular Research, vol. 70, no. 2, pp. 297–307, 2006. signaling in diabetic hearts,” Circulation Research, vol. 106, no. [93] S. Ghosh, N. B. Standen, and M. Galinanes,˜ “Failure to 1, pp. 129–132, 2010. precondition pathological human myocardium,” Journal of the Cardiology Research and Practice 15

American College of Cardiology, vol. 37, no. 3, pp. 711–718, 2001. [94] J. M. Smith and G. M. Wahler, “ATP-sensitive potassium channels are altered in ventricular myocytes from diabetic rats,” Molecular and Cellular Biochemistry, vol. 158, no. 1, pp. 43–51, 1996. [95]A.Hassouna,M.Loubani,B.M.Matata,A.Fowler,N.B. Standen, and M. Galinanes,˜ “Mitochondrial dysfunction as the cause of the failure to precondition the diabetic human myocardium,” Cardiovascular Research, vol. 69, no. 2, pp. 450– 458, 2006. [96] J. Linares-Palomino, M. A. Husainy, V. K. Lai, J. M. Dickenson, and M. Galinanes,˜ “Selective blockade of protein kinase B protects the rat and human myocardium against ischaemic injury,” Journal of Physiology, vol. 588, no. 12, pp. 2173–2191, 2010. [97] D. J. Hausenloy, P. K. Mwamure, V. Venugopal et al., “Effect of remote ischaemic preconditioning on myocardial injury in patients undergoing coronary artery bypass graft surgery: a randomised controlled trial,” The Lancet, vol. 370, no. 9587, pp. 575–579, 2007. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 632408, 6 pages doi:10.1155/2012/632408

Review Article The LOX-1 Scavenger Receptor and Its Implications in the Treatment of Vascular Disease

M. W Twigg,1, 2 K. Freestone,1 S. Homer-Vanniasinkam,1, 2 and S. Ponnambalam1

1 Endothelial Cell Biology Unit, Institute of Molecular & Cellular Biology, LIGHT Laboratories, Clarendon Way, Leeds LS2 9JT, UK 2 Leeds Vascular Institute, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK

Correspondence should be addressed to M. W Twigg, [email protected]

Received 22 July 2011; Accepted 14 November 2011

Academic Editor: Daryll M. Baker

Copyright © 2012 M. W Twigg et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Cardiovascular disease is the leading cause of death. The disease is due to atherosclerosis which is characterized by lipid and fat accumulation in arterial blood vessel walls. A key causative event is the accumulation of oxidised low density lipoprotein particles within vascular cells, and this is mediated by scavenger receptors. One such molecule is the LOX-1 scavenger receptor that is expressed on endothelial, vascular smooth muscle, and lymphoid cells including macrophages. LOX-1 interaction with OxLDL particles stimulates atherosclerosis. LOX-1 mediates OxLDL endocytosis via a clathrin-independent internalization pathway. Transgenic animal model studies show that LOX-1 plays a significant role in atherosclerotic plaque initiation and progression. Administration of LOX-1 antibodies in cellular and animal models suggest that such intervention inhibits atherosclerosis. Antiatherogenic strategies that target LOX-1 function using gene therapy or small molecule inhibitors would be new ways to address the increasing incidence of vascular disease in many countries.

1. Background functional LDL receptors, and, therefore, their circulating cholesterol levels are up to 5 times higher than in normal Cholesterol is a key cellular molecule that is vital to growth individuals. However, this did not explain the observation and repair and is also a major component of the mammalian that patients with FH accumulated cholesterol in their cells plasma membrane bilayer. It is also required for the synthesis despite the lack of functional LDL receptors. Furthermore, of steroid hormones and bile salts. However, an elevated lipid-laden foam cells derived from patient tissues did not level of serum cholesterol and in particular, the low-density develop in vitro even in the presence of high concentrations lipoprotein (LDL) fraction are well-established risk factors of native LDL in contrast to controls. Their work led to for the development of atherosclerosis. Atherosclerosis is a the now widely accepted view that the internalisation of a systemic disease, commonly affecting multiple vascular beds modified form of LDL rather than native LDL is responsible [1] and is the leading cause of death in Europe [2]. Uptake for the development of atherosclerotic plaque, and that this of native LDL into cells was first proposed to be a crucial occurs via membrane-bound receptors distinct to that of initial step in the pathogenesis of atherosclerosis with the the native LDL receptor. Basu et al. first demonstrated this discovery of the LDL receptor in the 1970’s by Goldstein and in an experiment in which an iodine-labelled acetylated Brown [3]. Their studies led to the elucidation of a pathway form of LDL could induce massively increased cholesterol by which the LDL-receptor complex undergoes receptor- accumulation in macrophages in vitro [4]. Further work mediated endocytosis to enable it to be taken into the cell. in vitro has shown that the oxidative modification of LDL They observed that individuals with homozygous familial increases the cellular content of cholesterol. Today, several hypercholesterolaemia (FH), who suffer with manifestations receptors have been identified that bind and internalise of atherosclerosis in their teens, are completely lacking in oxidised LDL (OxLDL). This review focuses on one of those 2 Cardiology Research and Practice receptors, the lectin-like oxidised low-density lipoprotein (CTLD). To function, LOX-1 requires oligomerisation of a receptor-1 (LOX-1) and explores its potential as a therapeutic homodimer; this basal complex contains 2 LOX-1 polypep- target. tides linked via an intermolecular disulfide bond in the neck domain via residue C140 [16]. Evidence increasingly suggests that LOX-1 activity promotes vascular dysfunction and 2. Modification of LDL atherosclerosis. LOX-1 ablation in transgenic murine models The oxidative modification of LDL results in a particle reduces atherosclerotic plaque development [17]whereas with greatly increased proatherogenic and proinflammatory overexpression of LOX-1 in apolipoprotein E-null mice properties than that of native LDL. It also becomes a ligand increases atheroma-like lesions 10-fold [18]. Interestingly, for scavenger receptors but is not recognised by the native raised serum levels of a soluble LOX-1-derived proteolytic LDL receptor, thought to be due to the greater net negative fragment correlate with elevated acute coronary syndromes charge [5]. The “average” LDL particle has been calculated (ACS) [19] and type II diabetes [20]. To date, seven single to contain 600 molecules of free cholesterol, 1600 molecules nucleotide polymorphisms (SNP) within the LOX-1 gene of cholesteryl ester, 700 molecules of phospholipid, 180 have been found (six noncoding, one coding) [21]. One molecules of triglyceride, and 1 molecule of apolipoprotein SNP appears to confer protection against OxLDL-induced B-100 [6]. All of these molecules can undergo oxidative macrophage apoptosis [22], whereas the K167N polymor- damage, and, therefore, understandably there is a spectrum phism, resulting in a lysine to asparagine substitution, of oxidation level to the OxLDL created experimentally. increases the risk of myocardial infarction in a specific It is believed that the amount of OxLDL in the human patient cohort [23]. Numerous signal transduction pathways circulation is negligible, due to the presence of antioxidants. are associated with LOX-1 activation by OxLDL binding OxLDL has, however, been extracted from human and rabbit including RhoA/Rac1, p38MAPK, protein kinase B and C, and ERK1/2, and blocking LOX-1 function in primary atherosclerotic plaque, in a form recognisable by scavenger κ receptors [7]. The oxidation is likely to take place in pockets endothelial cells inhibits proinflammatory signalling, NF- B within the subendothelial layer, where the cells produce both activation, and apoptosis [24, 25]. free radical and nonradical oxidants, and the relative con- centration of antioxidants is lower. The resulting OxLDL has an immediate proatherogenic effect; it causes the migration 3.1. Plasma Membrane Endocytosis of LOX-1. OxLDL is of monocytes to the area via stimulation of the release of rapidly internalised into cells upon binding LOX-1 and can monocyte chemoattractant protein-1 [8] and promotes the be observed in punctate perinuclear structures less than an differentiation of monocytes to macrophages by stimulating hour after exposure to a LOX-1 expressing cell (Figure 1). the release of macrophage-colony stimulating factor from This internalisation is blocked by the LOX-1-blocking- endothelial cells [9]. OxLDL is cytotoxic to endothelial cells antibody JTX92 [27], which prevents binding of OxLDL. in vitro and inhibits the vasodilation normally induced by LOX-1 mediates early steps in internalisation of OxLDL nitric oxide [10]. but within an hour of internalisation is mostly uncoupled from OxLDL, and both molecules are located in separate subcellular compartments within the cytosol [26]. This 3. LOX-1 internalisation pathway is not dependant on LOX-1 binding to OxLDL as the receptor is constitutively endocytosed from Modified LDL is internalized or endocytosed by membrane- the plasma membrane. bound scavenger receptors, and as it stands today, numerous The main form of receptor-mediated endocytosis is scavenger receptors have been identified that recognize clathrin-mediated endocytosis, but it has become apparent OxLDL. Scavenger receptors are membrane-bound proteins that clathrin-independent pathways may represent up to that are capable of binding a wide variety of ligands. The 50% of cellular uptake [28]. Reverse genetic experiments scavenger receptor family is subdivided into 8 subclasses. The have shown that both the clathrin heavy chain and the AP-2 class A scavenger receptors all bind modified LDL and are adaptor complex are not required for endocytosis of LOX- primarily expressed on macrophages. Class B includes the 1[26]. LOX-1 was also shown not to colocalise with the CD36 receptor, which has been shown to bind and internalize caveolae marker caveolin-1. However, there is evidence that modified LDL [11]. The affinity of the other classes to bind caveolae may play a part in the endocytosis of OxLDL in modified LDL is less well described, except for the class E an endothelial model [29]. The dynamin-2 GTPase has been receptor, lectin-like OxLDL receptor-1 (LOX-1). shown to be essential for LOX-1 endocytosis as expression LOX-1 (Genbank designation OLR1) was first cloned of a dominant-negative protein defective in GTPase activity as a major receptor for OxLDL in 1997 by Sawamura and blocked OxLDL uptake via the LOX-1 scavenger receptor colleagues [12].LOX-1isexpressedonavarietyofcell (Figure 2). The full mechanism by which LOX-1 endocytosis types, including endothelial cells, platelets, macrophages, occurs has yet to be established. and smooth muscle cells [13–15]. The human ortholog is In eukaryote cells, receptor-mediated endocytosis is a 50 kDa type II transmembrane glycoprotein comprising regulated by the recognition of cytoplasmic motifs by cellular 273 amino acids [12].Themammalianproteiniscomprised machinery which promotes the selection of “cargo” for of a short N-terminus cytoplasmic domain, transmembrane transport within transport intermediates such as membrane- domain, neck domain, and a C-type lectin-like domain bound vesicles [30]. The cytoplasmic domain of LOX-1 does Cardiology Research and Practice 3

(a) (b)

Figure 1: Internalisation of OxLDL. Epithelial HeLa cells transiently expressing a LOX-1-FLAG protein (green) were incubated with 10 μg/mL DiI-OxLDL (red) for 5 mins at 37◦C (pulse) to allow binding (a), and chased for a further 55 mins to allow internalization (b) before fixation. Nuclei were stained with DAPI (blue). Specimens were visualized using a deconvolution microscope using previously described procedures [26].

5 µm 5 µm

(a) (b)

Figure 2: Dynamin-2 regulates OxLDL endocytosis. Epithelial HeLa cells expressing both LOX-1-FLAG and either (a) wild-type dynamin 2 (green) or (b) a dynamin-2 K44A mutant (green) were incubated and chased with labelled Dil-OxLDL (red). Cells were fixed and processed for fluorescence microscopy using previously described procedures [26]. The nuclei were stained with DAPI (blue). Bar, 5 μm. A slice through the cell is shown in the top half of the panel and a cross-section through the cell is shown at the bottom. not contain any previously characterised motifs such as the the LOX-1 cytoplasmic domain still promotes constitutive tyrosine-based motif YxxΦ or the di-leucine-based motifs. endocytosis. Endocytosis of this transferrin receptor-LOX- Alanine-scanning mutagenesis of the LOX-1 cytoplasmic 1 protein chimera was blocked by replacement of the DDL domain allowed the identification of a sequence of three motif with a triple alanine sequence, thus showing specificity contiguous residues (DDL) at position +4 to +6 which in this endocytosis [31]. regulate LOX-1 endocytosis [31]. This diacidic DDL motif After internalization at the plasma membrane, LOX-1 has thus defines a new class of novel endocytic motifs that been shown to be separated from OxLDL in endosomes [27]. mediate clathrin- and AP2-independent endocytosis at the Immunofluorescence studies demonstrate that after 15 min plasma membrane. This motif is transplantable, as replace- of internalization LOX-1 and OxLDL are colocalised in the ment of the transferrin receptors cytoplasmic domain with same endocytic compartment. From 30 min onwards after 4 Cardiology Research and Practice

Target: OxLDL ligand Target: soluble cleaved LOX-1 fragment Extracellular Intracellular

Target: scaffold/ adaptor proteins facilitating Target: LOX-1 endocytosis at a genetic level

OxLDL LOX-1

Figure 3: Schematic diagram showing potential routes towards blocking OxLDL and/or LOX-1 function and trafficking to attenuate the pathological process of atherosclerosis.

ligand binding, this codistribution disappears, indicating 4. Conclusion that LOX-1 and OxLDL have now been sorted into sepa- rate compartments. OxLDL is trafficked through the early Scavenger receptors are heavily implicated in the process of endosome and on to the lysosome for degradation. It is atherosclerotic plaque formation [38]. This review focuses believed that the majority of LOX-1 is targeted back to the on LOX-1; however, there is similar evidence for the plasma membrane by a recycling pathway from endosome- involvement of several other classes of scavenger receptor. For to-plasma membrane. Cytosolic factors that mediate recog- example, ApoE-null (−/−) mice lacking macrophage CD36 nition, endocytosis, and/or recycling of LOX-1 have yet had an 88% decrease in atherosclerotic lesion area in the to be identified. Targeting such factors using genetic or aorta compared to controls [39], despite the heterogeneity pharmacological approaches would be a potential means of amongst the scavenger receptor family and their multiligand blocking the proatherogenic function of LOX-1 in promoting capabilities, their targeting could prove fruitful in the atherosclerosis. moderation of vascular disease, especially considering the LOX-1 appears to be essential for the phagocytosis of limited success of current treatment modalities. However, the aged and apoptotic cells in endothelial cells [32]. LOX-1 most effective means of targeting and disrupting the OxLDL- was demonstrated to be necessary for phagocytosis of such LOX-1 endocytic process remains unclear (Figure 3). The cellular bodies in transfected Chinese hamster ovary cells, OxLDL ligand itself could be targeted; in a murine model, but this is blocked by OxLDL indicating a competition LOX-1 expressed ectopically in the liver via adenovirus between OxLDL and apoptotic/aged bodies for binding to administration reduced levels of circulating OxLDL and the same or adjacent site on LOX-1. Phosphatidylserine (PS) inhibited the formation of atherosclerotic lesions [40], gene recognition on the plasma membrane of aged/apoptotic cells therapy with possible genomic manipulation of scavenger has been shown to be important in their phagocytosis [33], receptor expression by delivery of transgenes or by blockade and LOX-1 is able to recognise PS on apoptotic cells in a of gene expression may be possible [41]. Further understand- calcium-dependent manner [34]. ing of the proteins that facilitate OxLDL transfer across the LOX-1 endocytosis is also potentially important in endothelium may allow the development of pharmaceutical immune surveillance as it has been shown to regulate agents that inhibit its endocytosis. Finally, the role of the antigen presentation by MHC class I and II molecules soluble fragment of the LOX-1 receptor that is shed into the on dendritic cells [35] and B cells [36]. LOX-1 is also circulation is unclear, but the study of its use as a biomarker essential for the endocytosis of various heat shock proteins in the treatment of cardiovascular disease is promising. (hsp’s) complexed to antigen-derived peptides [37]. LOX-1 Significant recent progress has been made in the elucida- is postulated to deliver the antigenic complex to endosome- tion of the pathways of internalization of the OxLDL-LOX-1 like compartments, where it is presumably loaded onto MHC complex; however, there remains much to be done to further molecules, which mediate antigen presentation to CD4+ and characterise the role of this multiligand scavenger receptor CD8+ T cells. in regulating human health and disease. The existence of Cardiology Research and Practice 5 several different subclasses of scavenger receptor capable oxidized low-density lipoprotein in human macrophages and of internalising OxLDL makes the development of targeted its potential role as a scavenger receptor,” Biochemical Journal, therapies a more complex issue. The molecular study of vol. 334, no. 1, pp. 9–13, 1998. scavenger receptors remains an exciting avenue in the search [14] N. Kume, S. Miyamoto, M. Minami et al., “Oxidized LDL for therapeutic agents to attenuate the atherosclerotic process modulates Bax/Bcl-2 through the lectinlike Ox-LDL receptor- in humans. 1 in vascular smooth muscle cells,” Arteriosclerosis, Thrombo- sis, and Vascular Biology, vol. 21, no. 6, pp. 955–960, 2001. [15] M. Chen, M. Kakutani, T. Naruko et al., “Activation- Acknowledgment dependent surface expression of LOX-1 in human platelets,” Biochemical and Biophysical Research Communications, vol. TheworkinourlaboratoriesweresupportedbyaPhD 282, no. 1, pp. 153–158, 2001. studentship from Heart Research UK (KF) and a British [16] S. Matsunaga, Q. Xie, M. Kumano et al., “Lectin-like oxi- Heart Foundation project grant (SP). dized low-density lipoprotein receptor (LOX-1) functions as an oligomer and oligomerization is dependent on receptor density,” Experimental Cell Research, vol. 313, no. 6, pp. 1203– References 1214, 2007. [17] J. L. Mehta, N. Sanada, C. P. Hu et al., “Deletion of LOX- [1] D. L. Bhatt, P. G. Steg, E. M. Ohman et al., “International 1 reduces atherogenesis in LDLR knockout mice fed high prevalence, recognition, and treatment of cardiovascular risk cholesterol diet,” Circulation Research, vol. 100, no. 11, pp. factors in outpatients with atherothrombosis,” Journal of the 1634–1642, 2007. American Medical Association, vol. 295, no. 2, pp. 180–189, [18] K. Inoue, Y. Arai, H. Kurihara, T. Kita, and T. Sawa- 2006. mura, “Overexpression of lectin-like oxidized low-density [2] S. Allender, P. Scarborough, V. Peto et al., European Cardiovas- lipoprotein receptor-1 induces intramyocardial vasculopathy cular Disease Statistics, European Heart Network, 2008. in apolipoprotein E-null mice,” Circulation Research, vol. 97, [3] J. L. Goldstein and M. S. Brown, “Lipoprotein receptors, no. 2, pp. 176–184, 2005. cholesterol metabolism, and atherosclerosis,” Archives of [19] K. Hayashida, N. Kume, T. Murase et al., “Serum soluble Pathology and Laboratory Medicine, vol. 99, no. 4, pp. 181–184, lectin-like oxidized low-density lipoprotein receptor-1 levels 1975. are elevated in acute coronary syndrome a novel marker for [4] S. K. Basu, M. S. Brown, Y. K. Ho, and J. L. Goldstein, “Degra- early diagnosis,” Circulation, vol. 112, no. 6, pp. 812–818, 2005. dation of low density lipoprotein-dextran sulfate complexes [20] K. C. B. Tan, S. W. M. Shiu, Y. Wong, L. Leng, and R. Bucala, associated with deposition of cholesteryl esters in mouse “Soluble lectin-like oxidized low density lipoprotein receptor- macrophages,” The Journal of Biological Chemistry, vol. 254, 1 in type 2 diabetes mellitus,” Journal of Lipid Research, vol. 49, no. 15, pp. 7141–7146, 1979. no. 7, pp. 1438–1444, 2008. [5] I. S. Young and J. McEneny, “Lipoprotein oxidation and [21] R. Mango, F. Clementi, P. Borgiani et al., “Association of single atherosclerosis,” Biochemical Society Transactions, vol. 29, no. nucleotide polymorphisms in the oxidised LDL receptor 1 2, pp. 358–362, 2001. (OLR1) gene in patients with acute myocardial infarction,” [6] W. Jessup, L. Kritharides, and R. Stocker, “Lipid oxidation in Journal of Medical Genetics, vol. 40, no. 12, pp. 933–936, 2003. atherogenesis: an overview,” Biochemical Society Transactions, [22] R. Mango, S. Biocca, F. Del Vecchio et al., “In vivo and in vol. 32, no. 1, pp. 134–138, 2004. vitro studies support that a new splicing isoform of OLR1 gene [7] S. Yla-Herttuala,¨ W. Palinski, M. E. Rosenfeld, D. Steinberg, is protective against acute myocardial infarction,” Circulation and J. L. Witztum, “Lipoproteins in normal and atheroscle- Research, vol. 97, no. 2, pp. 152–158, 2005. rotic aorta,” European Heart Journal, vol. 11, pp. 88–99, 1990. [23] M. Tatsuguchi, M. Furutani, J. I. Hinagata et al., “Oxidized [8] S. D. Cushing, J. A. Berliner, A. J. Valente et al., “Minimally LDL receptor gene (OLR1) is associated with the risk of modified low density lipoprotein induces monocyte chemo- myocardial infarction,” Biochemical and Biophysical Research tactic protein 1 in human endothelial cells and smooth muscle Communications, vol. 303, no. 1, pp. 247–250, 2003. cells,” Proceedings of the National Academy of Sciences of the [24] D. Li and J. L. Mehta, “Upregulation of endothelial receptor United States of America, vol. 87, no. 13, pp. 5134–5138, 1990. for oxidized LDL (LOX-1) by oxidized LDL and implications [9]T.B.Rajavashisth,A.Andalibi,M.C.Territoetal.,“Induction in apoptosis of coronary artery endothelial cells: evidence of endothelial cell expression of granulocyte and macrophage from use of antisense LOX-1 mRNA and chemical inhibitors,” colony-stimulating factors by modified low-density lipopro- Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 20, no. teins,” Nature, vol. 344, no. 6263, pp. 254–257, 1990. 4, pp. 1116–1122, 2000. [10]K.Kugiyama,S.A.Kerns,J.D.Morrisett,R.Roberts, [25]J.Chen,J.L.Mehta,N.Haider,X.Zhang,J.Narula,andD. and P. D. Henry, “Impairment of endothelium-dependent Li, “Role of caspases in Ox-LDL-induced apoptotic cascade arterial relaxation by lysolecithin in modified low-density in human coronary artery endothelial cells,” Circulation lipoproteins,” Nature, vol. 344, no. 6262, pp. 160–162, 1990. Research, vol. 94, no. 3, pp. 370–376, 2004. [11] G. Endemann, L. W. Stanton, K. S. Madden, C. M. Bryant, R. [26] J. E. Murphy, R. S. Vohra, S. Dunn et al., “Oxidised LDL T. White, and A. A. Protter, “CD36 is a receptor for oxidized internalisation by the LOX-1 scavenger receptor is dependent low density lipoprotein,” The Journal of Biological Chemistry, on a novel cytoplasmic motif and is regulated by dynamin-2,” vol. 268, no. 16, pp. 11811–11816, 1993. Journal of Cell Science, vol. 121, no. 13, pp. 2136–2147, 2008. [12] T. Sawamura, N. Kume, T. Aoyama et al., “An endothelial [27] D. Li, L. Liu, H. Chen, T. Sawamura, S. Ranganathan, and J. receptor for oxidized low-density lipoprotein,” Nature, vol. L. Mehta, “LOX-1 mediates oxidized low-density lipoprotein- 386, no. 6620, pp. 73–77, 1997. induced expression of matrix metalloproteinases in human [13] H. Yoshida, N. Kondratenko, S. Green, D. Steinberg, and O. coronary artery endothelial cells,” Circulation, vol. 107, no. 4, Quehenberger, “Identification of the lectin-like receptor for pp. 612–617, 2003. 6 Cardiology Research and Practice

[28] Q. Gong, C. Huntsman, and D. Ma, “Clathrin-independent internalization and recycling: membrane Trafficking Review Series,” Journal of Cellular and Molecular Medicine, vol. 12, no. 1, pp. 126–144, 2008. [29] S. W. Sun, X. Y. Zu, Q. H. Tuo et al., “Caveolae and caveolin-1 mediate endocytosis and transcytosis of oxidized low density lipoprotein in endothelial cells,” Acta Pharmacologica Sinica, vol. 31, no. 10, pp. 1336–1342, 2010. [30] J. S. Bonifacino and L. M. Traub, “Signals for sorting of transmembrane proteins to endosomes and lysosomes,” Annual Review of Biochemistry, vol. 72, pp. 395–447, 2003. [31] R. S. Vohra, J. H. Walker, G. J. Howell, S. Homer-Van- niasinkam, and S. Ponnambalam, “The LOX-1 scavenger receptor cytoplasmic domain contains a transplantable endo- cytic motif,” Biochemical and Biophysical Research Communi- cations, vol. 383, no. 2, pp. 269–274, 2009. [32] K. Oka, T. Sawamura, K. I. Kikuta et al., “Lectin-like oxidized low-density lipoprotein receptor 1 mediates phagocytosis of aged/apoptotic cells in endothelial cells,” Proceedings of the National Academy of Sciences of the United States of America, vol. 95, no. 16, pp. 9535–9540, 1998. [33] S. J. Martin, C. P. M. Reutelingsperger, A. J. McGahon et al., “Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: inhibition by overexpression of Bcl-2 and Abl,” Journal of Experimental Medicine, vol. 182, no. 5, pp. 1545– 1556, 1995. [34]J.E.Murphy,D.Tacon,P.R.Tedburyetal.,“LOX-1scavenger receptor mediates calcium-dependent recognition of phos- phatidylserine and apoptotic cells,” Biochemical Journal, vol. 393, no. 1, pp. 107–115, 2006. [35] Y. Delneste, G. Magistrelli, J. F. Gauchat et al., “Involvement of LOX-1 in dendritic cell-mediated antigen cross-presentation,” Immunity, vol. 17, no. 3, pp. 353–362, 2002. [36] T. Matsutake, T. Sawamura, and P. K. Srivastava, “High efficiency CD91- and LOX-1-mediated re-presentation of gp96-chaperoned peptides by MHC II molecules,” Cancer Immunity, vol. 10, 7 pages, 2010. [37] J. Xie, H. Zhu, L. Guo et al., “Lectin-like oxidized low-density lipoprotein receptor-1 delivers heat shock protein 60-fused antigen into the MHC class I presentation pathway,” Journal of Immunology, vol. 185, no. 4, pp. 2306–2313, 2010. [38] K. J. Moore and M. W. Freeman, “Scavenger receptors in atherosclerosis: beyond lipid uptake,” Arteriosclerosis, Throm- bosis, and Vascular Biology, vol. 26, no. 8, pp. 1702–1711, 2006. [39] M. Febbraio, E. Guy, and R. L. Silverstein, “Stem Cell transplantation reveals that absence of macrophage CD36 is protective against atherosclerosis,” Arteriosclerosis, Thrombo- sis, and Vascular Biology, vol. 24, no. 12, pp. 2333–2338, 2004. [40] Y. Ishigaki, H. Katagiri, J. Gao et al., “Impact of plasma oxidized low-density lipoprotein removal on atherosclerosis,” Circulation, vol. 118, no. 1, pp. 75–83, 2008. [41] S. L. Stephen, K. Freestone, S. Dunn et al., “Scavenger re- ceptors and their potential as therapeutic targets in the treatment of cardiovascular disease,” International Journal of Hypertension, vol. 2010, Article ID 646929, 21 pages, 2010. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 620681, 9 pages doi:10.1155/2012/620681

Review Article Clinical Applications of Remote Ischemic Preconditioning

Kristin Veighey1 and Raymond J. MacAllister2

1 University College London Centre for Nephrology, Royal Free Campus, Rowland Hill Street, London NW3 2PF, UK 2 University College London Centre for Clinical Pharmacology, Rayne Institute, 5 University Street, London WC1E 6JF, UK

Correspondence should be addressed to Kristin Veighey, [email protected]

Received 19 July 2011; Accepted 29 November 2011

Academic Editor: Janice Tsui

Copyright © 2012 K. Veighey and R. J. MacAllister. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Ischemia-reperfusion injury is a composite of damage accumulated during reduced perfusion of an organ or tissue and the additional insult sustained during reperfusion. Such injury occurs in a wide variety of clinically important syndromes, such as ischemic heart disease and stroke, which are responsible for a high degree of morbidity and mortality worldwide. Basic research has identified a number of interventions that stimulate innate resistance of tissues to ischemia-reperfusion injury. Here, we summarise the experimental and clinical trial data underpinning one of these “conditioning” strategies, the phenomenon of remote ischemic preconditioning.

1. Introduction applied to a distant organ or tissue, which then protects against index ischemia. For example, the preconditioning Ischemia-reperfusion injury underpins the damage of stimulus might be suprasystolic blood pressure inflations myocardial infarction, stroke, and other conditions compli- on an arm or leg, which then confer myocardial protection cated by interruption of the blood supply to tissues. Strate- against subsequent ischemia. Postconditioning occurs when gies to limit the duration of ischemia have achieved sub- there is staged reperfusion, for example, in the setting of stantial health gains in myocardial infarction and, to a lesser balloon angioplasty. Its variant perconditioning occurs when degree, stroke. However door-to-needle times have probably the conditioning stimulus is applied during ischemia. reached the minimum that is possible in many health-care delivery systems, so further reduction in morbidity and mortality from IR injury will require strategies to increase 3. Ischemic Preconditioning tissue tolerance to ischemia or reduce damage that occurs on Ischemic preconditioning was first described in 1986, when reperfusion. One such approach is ischemic preconditioning, Murry et al. demonstrated that in the dog, brief episodes and its variant remote ischemic preconditioning, the subject of ischemia (4 cycles of 5-minute occlusion followed by of this paper. reperfusion) of the circumflex artery reduced the extent of infarction induced by subsequent prolonged occlusion of 2. Types of Ischemic Preconditioning that vessel [1]. This protection expired after a few hours, but subsequent studies indicated that it recovered approximately Ischemic preconditioning (IPC) describes the phenomenon 24 hours later, this second phase of protection lasting for up whereby transient, brief periods of ischemia confer protec- to a further 72 hours [2–4]. Despite the length of time that tion against a subsequent prolonged and injurious period has elapsed since the discovery of IPC, detailed exposition of ischemia. There are a number of ways in which pre- of its mechanism, and evidence that the biological processes conditioning can be induced. Local preconditioning occurs operate in humans, IPC has never progressed to detailed when the preconditioning stimulus is applied to the same clinical investigation. This was largely due to the logistics organ or tissue that will subsequently sustain the ischemic of inducing preconditioning ischemia in vital organs (such injury. Remote ischemic preconditioning refers to a stimulus as the heart or brain) in advance of a more prolonged 2 Cardiology Research and Practice insult (e.g., that would lead to myocardial infarction or Triggers in the initial cascade recruit early mediators stroke). (such as protein kinase C (PKC), tyrosine kinase, phos- phatidylinositol 3-kinases (PI3K), protein kinase B (PKB or Akt) [15], mitogen-activated protein kinases (MAP1/2 4. Remote Ischemic Preconditioning or MEK1/2), extracellular signal-regulated kinases (Erk1/2), A major breakthrough in clinical applicability of precondi- and janus kinase (JAK)), which activate transcription factors tioning protection came with the discovery that ischemic (such as signal transducer and activator of transcription pro- teins (STAT1/3), nuclear factor kappa-light-chain-enhancer preconditioning also had a systemic protective phenotype. κ This facet, termed remote ischemic preconditioning (RIPC), of activated B cells (NF B), activator-protein-1 (AP-1), nuclear factor-like 2 (Nrf2), and hypoxia-inducible-factor- resulted in protection from ischemia-reperfusion injury at α α sites remote from those undergoing the preconditioning 1 (HIF-1 )). Later phase protection requires synthesis of stimulus. This was first described in the setting of exper- inducible nitric oxide synthase (iNOS), heat shock proteins imental coronary artery occlusion, where preconditioning (HSPs), or cyclooxygenase-2 (COX-2). These then act locally one vascular territory of the heart effected protection in via the mPTP or KATP channels to induce a state of adjacent tissue that had not undergone any preconditioning cardioprotection [16]. ischemia [5]. Interorgan protection was confirmed by the observation that preconditioning stimuli applied to the small 6. Mechanism of the Systemic bowel [6] or kidney [7] reduced infarct size in the heart. Spread of Protection As was the case for IPC, further studies established that the time course of protection caused by RIPC was also biphasic. Evidence for involvement of a humoral factor in mediating The demonstration that RIPC could be activated merely via systemic spread is supported by the observation that protec- brief periods of limb ischemia simplified the logistics of tion can be transferred by the transfusion of serum from a inducing ischemic preconditioning in animals and humans rabbit that has undergone ischemic preconditioning to one [8]. Moreover, RIPC activated by limb ischemia protected which has not [17, 18]. This factor appears to be heat stable from experimental IR injury in humans. Together these and has been shown to be dialysable and of less than 15 kD observations framed the conditions that have led to a large [19, 20]. In some studies it is blocked by opioid antagonists, number of clinical trials of RIPC in patients. including naloxone [21, 22]. Neurogenic mechanisms have also been explored using 5. Mechanisms of Tissue Protection of autonomic ganglionic blockade. In a rat myocardial infarc- tion model hexamethonium abolished protection by RIPC IPC and RIPC achieved by mesenteric artery occlusion (MAO) but had ff During ischemia, anaerobic metabolism predominates and no e ect on myocardial IPC. Cardioprotection was absent ATP production decreases. There is insufficient available when MAO was sustained throughout the study, indicating energy to maintain cell membrane pump activity, antioxi- that reperfusion in the small intestine was essential to acti- dant defences, pH and calcium homeostasis, and mitochon- vate the neurogenic pathway [6]. The autonomic ganglion drial integrity. These and other consequences of ischemia blocker trimethaphan has also been shown to inhibit remote inevitably lead to cell death, unless blood flow is restored. ischemic preconditioning in a human model [23]. In the Though reperfusion with oxygenated blood is essential for rabbit, sympathetic nerve activity increases when RIPC is any tissue salvage, the sudden influx of oxygen leads to the induced using renal ischemia, consistent with a particular formation of reactive oxygen species. A key event in cell death role for the adrenergic component of the autonomic system is mitochondrial permeability transition, a phenomenon that in this species [24]. occurs when the mitochondrial permeability transition pore Noradrenaline has been implicated; however, studies (mPTP) becomes permeable to molecules of 1500 kDa or are conflicting about its potential role. Administration of smaller. This leads to a rapid influx of small molecules, noradrenaline has been shown to induce preconditioning in mitochondrial swelling, and subsequent cell death [8]. animal models [25]; however, studies differ as to whether IPC activates three main salutatory pathways, the cyclic alpha-adrenoceptor blockers such as prazosin inhibit pre- guanosine monophosphate/cGMP-dependent protein kinase conditioning [25, 26]. Noradrenaline levels were not seen (cGMP/PKG) pathway [9], the reperfusion injury salvage to be increased in the serum of preconditioned rabbits, kinase (RISK) pathway [10], and the survivor activating which when transfused conferred preconditioning, leading factor enhancement (SAFE) pathway [11].Thereisadegree to doubts about a role in the transfer of protection [18]. of overlap, in particular where the pathways converge on Sensory nerves have also been implicated in spread of the mitochondrion [12]. Here, the potassium-dependent protection. Intramesenteric bradykinin has been demon- ATP (K ATP ) channel is activated with evidence that this strated in animal models to stimulate local sensory nerves, leads to closure of the mPTP. IPC also initiates a complex resulting in RIPC-like protection that is abolished by genomic and proteomic response that underpins the late hexamethonium [27]. This suggests a pathway involving phase of protection. This includes antiapoptotic and anti- sensory nerves and the autonomic nervous system. However, inflammatory gene transcription, likely to be responsible for a recent human healthy volunteer study demonstrated that the second window of protection [13, 14]. the bradykinin-2 inhibitor HOE-140 had no effect on RIPC Cardiology Research and Practice 3

Remote organ or tissue

RIPC by brief ischemic periods Activation of protein kinases p38/JNK/ErK/MAPK Neural Humoral Systemic Adenosine Adenosine Anti-inflammatory Bradykinin Bradykinin and antiapoptotic CGRP Opioids response and gene CGRP transcription Angiotensin I Endocannabinoids

STAT1/3, NFκB, AP-1, Nrf2, and α ? Circulating factor HIF-1

Target G-protein-coupled organ/tissue receptor

Kinase pathways RISK, SAFE, cGMP/PKG iNOS, HSP, COX-2 NO + 1st window of ROS protection Mitochondria 2nd window of protection

KATP mPTP

Figure 1: proposed mechanisms of remote ischemic preconditioning.

[28], so as ever mechanisms might be different in humans. 20 minutes, and this safely induces a transient period of Calcitonin-gene-related peptide (CGRP), a neurotransmitter endothelial dysfunction in the conduit and resistance vessels. in capsaicin-sensitive sensory nerves (CSSNs), has also been Endothelial assessment has been made using ultrasound implicated [29, 30]. CGRP has been reported to increase of conduit vessels to measure flow-mediated dilatation systemically after RIPC, and pretreatment with capsaicin (to or forearm plethysmography to characterize the response deplete CSSN) blocks RIPC [31]. of resistance vessels to endothelium-dependent agonists. The nonselective adenosine antagonist 8-(p-sulfophe- In this model, vascular smooth muscle function is pre- nyl)theophylline (8-SPT) has been shown in a rabbit [24, served. Endothelial dysfunction is largely prevented if the 32, 33]andrat[34] model to abolish the protective ischaemic period is preceded by brief, repeated periods effects of remote ischemic preconditioning. Adenosine might of ischaemia ipsilaterally (IPC) [36] and contralaterally therefore be one step in a complex pathway, though not itself (RIPC) [37, 38], with two phases of protection [38]. the circulating factor. Administration of KATP channel blockers prevents IPC in The humoral and neuronal pathways may work in series healthy volunteers, and IPC is mimicked by KATP channel to spread protection systemically. In the rat, release of opening drugs [39, 40]. A number of studies suggest the dialysable humoral factor is prevented by hindlimb that IR injury is dependent on increased oxidative stress denervation [35]. [41, 42], making it possible that IPC and RIPC stimulate The mechanisms of tissue protection and systemic spread antioxidant defences [43]. Regarding the systemic spread have been summarised in Figure 1. of protection, ganglionic blockade inhibits both phases of RIPC [6, 38], though as yet it is unclear which com- 7. Mechanism of IPC and RIPC in Humans ponent of the autonomic system is responsible. Dialysate of human plasma from volunteers who have undergone The development of a vascular model of human IR injury RIPC reduces IR injury in vitro in an opiate-dependent has facilitated the investigation of the mechanism of IPC manner, and this supports activation of opiate pathways and RIPC. In this model the arm is made ischemic for in humans [20]. However, the relative contributions of 4 Cardiology Research and Practice the neuronal and humoral pathways remain to be deter- The RICO trial (the Effect of Remote Ischemic Con- mined. ditioning on Atrial Fibrillation and Outcome) is ongoing to examine the effects of remote ischemic preconditioning, 8. Clinical Trials: Current Status remote ischemic postconditioning, or remote ischemic pre- and postconditioning on the development of atrial fibrilla- Although initial clinical trials in this area focussed on the tion on holter monitor within 72 hours after coronary artery application of remote ischemic preconditioning in ischemic bypass grafting [50]. This will help to define if there is any cardiac disease, interest has broadened to other areas of clinical benefit of combining preconditioning strategies. potential clinical benefit, including acute kidney injury, A current multicenter double-blind randomised con- stroke, and transplantation. trolled trial, “Effect of Remote Ischemic Preconditioning on A table summarising clinical trials in remote ischemic Clinical Outcomes in Patients Undergoing Coronary Artery preconditioning to date is shown (Table 1). Bypass Graft Surgery” (ERICCA), is currently underway to investigate whether RIPC improves one-year cardiovascular outcomes and reduces acute kidney injury (AKI) in the 8.1. Cardiac Surgery. The first clinical trial of remote ischem- setting of cold-blood cardioplegia CABG. This trial aims to ic preconditioning was in 2000, when 8 patients undergoing recruit 1610 patients, randomised to either RIPC or sham- coronary artery bypass grafting (CABG) were randomised RIPC. to receive either ischemic preconditioning (forearm cuff inflated to 300 mmHg for 2 cycles of 3 minutes) or control. 8.2. Percutaneous Coronary Intervention (PCI) for Acute The study demonstrated an increase in lactate dehydrogenase Myocardial Infarction (AMI). In 2006, Iliodromitis et al. (LDH) in the preconditioned group, which was attributed investigated whether RIPC by 3 cycles of 5-minute ischemia to an ability to maintain anaerobic metabolism in precon- applied to both arms would attenuate the inflammatory ditioned cells [44]. response in elective single vessel PCI with coronary stenting. Following this, a randomised controlled trial of RIPC They in fact demonstrated an increase in CK-MB, troponin in the setting of pediatric surgery (37 patients) for con- I, and CRP in the preconditioned group and postulated that genital cardiac defects demonstrated that 4 cycles of 5 RIPC increased the inflammatory response [51]. minutes of lower limb ischemia prior to surgery were Subsequently, Hoole et al., in 2009, in a study of 242 ff e ective in reducing troponin levels, postoperative inotropic patients undergoing elective PCI demonstrated that RIPC requirements, and airways resistance at 6 hours [45]. A (3 cycles of 5-minute forearm cuff inflation to 200 mmHg) study of arm preconditioning in simple congenital cardiac prior to PCI attenuated procedure-related cardiac troponin defects (ventricular septal defect repair) demonstrated an I (cTnI) release [52]. Diabetics and hypertensives were improvement in lung compliance and a decrease in cardiac included and also benefited. However, in a separate study the enzymes (LDH, CK, and troponin I) and cytokines (IL-6, same group showed that there was no beneficial effect on left α IL-8, IL-10, and TNF- ) in the preconditioned group. In ventricular dysfunction during coronary balloon occlusion this study, those randomised to the preconditioning group in single vessel coronary disease [53]. received preconditioning at 24 hours and 1 hour before Also of note, Bøtker et al. demonstrated the potential surgery, to utilise both early and late phases of protection for prehospital use of remote ischemic perconditioning in the [46]. setting of AMI (4 cycles of 5-minute forearm cuff inflation Subsequently in 2007, Hausenloy et al. demonstrated and deflation, delivered in the ambulance). In a trial of 333 a reduction in troponin T levels in patients randomised patients, they demonstrated an improvement in myocardial to receive ischemic preconditioning (3 cycles of 5-minute salvage index (%) at 20 days after primary PCI in the group forearm cuff inflation to 200 mmHg after induction of anaes- randomised to receive preconditioning [54]. In a substudy thesia) prior to coronary artery bypass grafting [47]. In both of the same patients, remote ischemic conditioning delivered these trials, the patients underwent cross-clamp fibrillation; before hospital seemed to result in modest improvement however, Venugopal et al., in 2009, also demonstrated a in LV function in high-risk patients prone to develop large reduction in troponin T following remote ischemic precon- myocardial infarcts [55]. ditioning in patients undergoing cold blood cardioplegia [48]. 8.3. Vascular Surgery. In open abdominal aortic aneurysm However, in 2010, Rahman et al. published a larger (AAA) repair, 82 patients were randomised to receive either single-centre double-blind randomised controlled trial in RIPC (two cycles of intermittent cross-clamping of the com- which 162 patients undergoing CABG were randomised mon iliac artery with 10-minute ischemia followed by 10- to receive either 3 × 5-minute cycles of upper limb cuff minute reperfusion) or none. RIPC reduced the absolute risk inflation to 200 mmHg (separated by 5-minute reperfusion) of myocardial injury, myocardial infarction, and renal injury or placebo (in which the cuff was inflated on a “dummy [56]. Another study in the same clinical scenario, in which arm”). In this study there was no difference in troponin 51 patients were randomised to sequential common iliac release between the 2 groups [49]. There was also no artery cross-clamping as the conditioning stimulus or none, difference in cardiac performance, inotrope requirement, did not demonstrate any improvement in renal outcome echocardiographic function, arrhythmia protection, or renal indices (urinary retinol binding protein (RBP) and albu- and lung outcomes. min : creatinine ratio (ACR)) [57].Thesamegroup,however, Cardiology Research and Practice 5

Table 1: Clinical trials to date in remote ischemic preconditioning.

Year n Author Clinical setting Conditioning protocol Results Coronary artery bypass grafting ↑Lactate dehydrogenase (LDH) in 2000 8 Gunyaydin¨ Arm 2 × 3min (CABG) preconditioned group Clipping of cerebral aneurysm Proximal artery 2 min catheter ↓tissue hypoxia (slower decline in 2005 12 Chan following subarachnoid occlusion, 30 min reperfusion pO2 and pH) haemorrhage Elective single vessel percutaneous Increased troponin I, CK-MB, and 2006 41 Iliodromitis coronary intervention (PCI) and Both arms 3 × 5min C-reactive protein (CRP) stenting Paediatric surgery for congenital ↓troponin T, ↓post-op inotrope 2006 37 Cheung Leg 4 × 5min cardiac defects requirements, ↓airway resistance ↓absolute risk of myocardial Elective open abdominal aortic Internal iliac cross-clamp 2 × 2007 82 Ali injury/infarction (troponin I) and aneurysm repair 10 min renal injury (creatinine) 2007 57 Hausenloy Elective CABG Arm 3 × 5min ↓troponin T Decreased Glasgow Coma Score Balloon occlusion during carotid Temporary balloon occlusion of (GCS) seen on initial balloon 2008 165 Faries angioplasty and stenting ICA inflation not observed when blood reinflated CABG (cold-blood cardioplegia) 2009 45 Venugopal Arm 3 × 5min ↓troponin T +/− valve replacement ↓troponin I, reduces ischaemic pain 2009 242 Hoole Elective PCI Arm 3 × 5min during PCI No effect on left ventricular Elective PCI for single vessel dysfunction (conductance 2009 42 Hoole Arm 3 × 5min coronary artery disease catheter/Doppler/dobutamine stress echocardiogram) Reduced urinary retinol binding 2009 40 Walsh Endovascular AAA repair Both legs 10 min ischaemia protein and albumin : creatinine ratio Fewer saccadic latency 2010 70 Walsh Carotid endarterectomy Both legs 10 min ischaemia deteriorations (did not reach statistical significance) Post hoc analysis of 2 previous studies. Decreased Acute Kidney 2010 78 Venugopal CABG Arm 3 × 5min Injury (AKI) on Acute Kidney Injury Network (AKIN) criteria No difference in trop T/cardiac performance/inotrope 2010 162 Rahman CABG (on-pump) Arm 3 × 5min requirement/echo function/arrthymias/renal or lung outcomes CABG with cold crystalloid ↓troponin I; tramadol 2010 120 Wagner Arm 3 × 5min cardioplegia administration increased troponin No differences in renal outcomes (urinary retinol binding protein 2010 51 Walsh Open infrarenal AAA repair Sequential common iliac clamping (RBP) and albumin : creatinine ratio (ACR)) Reduced cytokines and cardiac Arm 3 × 5 min; 24 h and 1 h before enzymes, upregulation of heat 2010 60 Wenwu Ventricular septal defect repair surgery shock protein (HSP) 70; improved lung compliance Reduced markers of ischaemic Elective decompression surgery for neuronal injury S-100B and 2010 40 Hu adult cervical spondylotic Arm 3 × 5min neuron-specific enolase; improved myelopathy recovery rate on functional score Cardiac surgery with Reduced relative risk of AKI on 2011 120 Zimmerman Leg 3 × 5min cardiopulmonary bypass AKIN criteria 6 Cardiology Research and Practice

Table 1: Continued. Year n Author Clinical setting Conditioning protocol Results Paediatric surgery for correction of No effect on development of AKI, 2011 113 Pedersen Leg 4 × 5min complex cardiac defects urine output, or urinary biomarkers No effect on creatinine, cystatin C, or neutrophil-gelatinase-associated Complex valvular heart surgery 2011 76 Choi Leg 3 × 5min lipocalin (NGAL), estimated with CPB glomerular filtration rate Decreased CK-MB at 24 hours Modest improvement in LV systolic 2011 242 Munk Primary PCI for acute MI Arm 4 × 5 min, prehospital function (not significant) demonstrated lower levels of RBP and albumin : creatinine AKI in nondiabetic patients randomised to RIPC (three 5- ratio following surgery in the preconditioned group in the minute cycles of forearm ischemia). However, the numbers setting of endovascular AAA repair (EVAR) [58]. were small, there were more concomitant aortic valve A UK-based trial of remote ischaemic preconditioning in replacements in the RIPC group, the analysis was post the setting of elective abdominal aortic aneurysm repair is hoc, and, although there were more episodes of AKI in currently underway in Bristol. the nonintervention group, this was all AKI stage 1. [61] In a separate study of lower limb preconditioning in the 8.4. Brain and Neurological Injury. A pilot clinical trial of setting of cardiac surgery requiring cardiopulmonary bypass, 70 patients undergoing carotid endarterectomy, randomised AKI was reduced in the RIPC group (47% versus 20%). to remote ischemic preconditioning with 10 minutes of All AKI was either stage 1 or 2 [62]. A recently published ischemic applied to each leg showed a numerical but not trial of leg preconditioning in children undergoing surgery statistical reduction in saccadic latency deteriorations (the for complex congenital cardiac disease found no evidence primary neurological outcome). There was no significant that preconditioning protected renal function. End points difference between cardiac endpoints [59]. were development of acute kidney injury, initiation of However, in a trial in 40 adult cervical spondylotic dialysis, plasma creatinine, estimated glomerular filtration myelopathy patients undergoing elective decompression rate, plasma cystatin C, plasma and urinary neutrophil surgery there was a significant reduction in markers of gelatinase-associated lipocalin, and urinary output [63]. A ischemic neurological injury (serum S-100B and neuron- similar study in adults again demonstrated no evidence specific enolase) in the RIPC group. Patients randomized to of benefit in renal protection following complex cardiac RIPC received 3 × 5-minute arm cycles. Postsurgical recovery surgery; however, a reduction in CK-MB was observed [64]. at 7 days, 1, and 3 months after surgery (evaluated using a Trials examining the effects of preconditioning on AKI in Japanese Orthopaedic Association scale) was higher in the the setting of AAA surgery have been detailed above. preconditioned group [60]. In stroke, the “New Acute Treatment for Stroke—The 8.5.1. Solid Organ Transplantation. The ultimate renal IR Effect of Remote Perconditioning” has recently closed to injury occurs in renal transplantation, and in the case of recruitment and results awaited. This trial has examined live donor transplantation this has a predictable time of the utility of remote ischemic preconditioning using arm onset. Loukogeorgakis et al. have previously demonstrated ischemia in the ambulance prior to hospital admission and an almost 20% improvement in eGFR (Modification of thrombolysis. The primary endpoint in this study is salvage Diet in Renal Disease (MDRD) formula) after 2 years index, measured on diffusion-weighted T2 MRI. in children who underwent RIPC before live donor renal transplantation. Both donor and recipient received 3 cycles 8.5. Acute Kidney Injury (AKI). IR injury contributes to the of 5-minute forearm ischemia-reperfusion, 24 hours prior majority of AKI. Common causes of AKI such as sepsis, to transplantation. This was a small randomised controlled surgery, or drugs, for example, NSAIDS, ACE inhibitors, trial in 20 patients; however, a larger, 400-pair multicen- are all essentially wholly or in part due to ischemia- tre randomised controlled trial (Renal Protection against reperfusion injury, secondary to hypotension or reduced Ischemia Reperfusion in Transplantation (REPAIR)) is cur- renal blood flow. AKI is responsible for morbidity and rently underway. A trial of RIPC in the setting of cadaveric increased mortality in hospital admissions; however, its onset renal transplantation is being undertaken by Bøtker et al. is often unpredictable and patients present when AKI is The RIPCOT trial (Remote Ischemic Preconditioning in already manifest. Therefore, use of RIPC to protect against Abdominal Organ Transplantation), which utilises lower acute kidney injury has most widely been documented in the limb remote ischemic preconditioning in the setting of setting of surgery. In this setting, however, the acute injury is deceased donor liver, kidney, or pancreas transplantation, complex and multifactorial in nature. and RIPC before abdominal surgery trial in those undergoing Secondary analysis from two trials in elective coronary abdominal, large bowel, pancreatic, and hepatic surgery are artery bypass grafting has demonstrated a reduction in also currently underway. Cardiology Research and Practice 7

9. Conclusion [13] I. E. Konstantinov, S. Arab, R. K. Kharbanda et al., “The remote ischemic preconditioning stimulus modifies inflam- Remote ischemic preconditioning harnesses a powerful matory gene expression in humans,” Physiological Genomics, innate protective mechanism against ischemic injury. The vol. 19, no. 1, pp. 143–150, 2005. mechanism is as yet not fully elucidated; however, it has [14]T.Murphy,P.M.Walsh,P.P.Doran,andK.J.Mulhall, shown promise in clinical trials. No trial has yet been large “Transcriptional responses in the adaptation to ischaemia- enough to demonstrate an effect of RIPC to reduce the reperfusion injury: a study of the effect of ischaemic pre- incidence or impact of clinically relevant consequences of IR conditioning in total knee arthroplasty patients,” Journal of injury. However, large adequately powered studies will report Translational Medicine, vol. 8, article 46, 2010. within the next 2-3 years. These trials will either endorse the [15] L. Breivik, E. Helgeland, E. K. Aarnes, J. Mrdalj, and A. K. clinical usefulness of ischemic preconditioning or consign it Jonassen, “Remote postconditioning by humoral factors in effluent from ischemic preconditioned rat hearts is mediated to the laboratory. via PI3K/Akt-dependent cell-survival signaling at reperfu- sion,” Basic Research in Cardiology, vol. 106, no. 1, pp. 135– References 145, 2010. [16] D. J. Hausenloy and D. M. Yellon, “The Second Window of [1] C. E. Murry, R. B. Jennings, and K. A. Reimer, “Precondi- Preconditioning (SWOP) where are we now?” Cardiovascular tioning with ischemia: a delay of lethal cell injury in ischemic Drugs and Therapy, vol. 24, no. 3, pp. 235–254, 2010. myocardium,” Circulation, vol. 74, no. 5, pp. 1124–1136, 1986. [17] E. W. Dickson, C. P. Reinhardt, F. P. Renzi, R. C. Becker, ff [2] T. Kuzuya, S. Hoshida, N. Yamashita et al., “Delayed e ects of W. A. Porcaro, and S. O. Heard, “Ischemic preconditioning sublethal ischemia on the acquisition of tolerance to ischemia,” may be transferable via whole blood transfusion: preliminary Circulation Research, vol. 72, no. 6, pp. 1293–1299, 1993. evidence,” Journal of Thrombosis and Thrombolysis, vol. 8, no. [3]M.S.Marber,D.S.Latchman,J.M.Walker,andD.M.Yellon, 2, pp. 123–129, 1999. “Cardiac stress protein elevation 24 hours after brief ischemia [18] E. W. Dickson, M. Lorbar, W. A. Porcaro et al., “Rabbit heart or heat stress is associated with resistance to myocardial can be ’preconditioned’ via transfer of coronary effluent,” infarction,” Circulation, vol. 88, no. 3, pp. 1264–1272, 1993. American Journal of Physiology, vol. 277, no. 6, pp. H2451– [4]G.F.Baxter,F.M.Goma,andD.M.Yellon,“Characterisation H2457, 1999. of the infarct-limiting effect of delayed preconditioning: time- [19] H. H. Patel, J. Moore, A. K. Hsu, and G. J. Gross, “Cardio- course and dose-dependency studies in rabbit myocardium,” protection at a distance: mesenteric artery occlusion protects Basic Research in Cardiology, vol. 92, no. 3, pp. 159–167, 1997. the myocardium via an opioid sensitive mechanism,” Journal [5]K.Przyklenk,B.Bauer,M.Ovize,R.A.Kloner,andP. of Molecular and Cellular Cardiology, vol. 34, no. 10, pp. 1317– Whittaker, “Regional ischemic “preconditioning” protects 1323, 2002. remote virgin myocardium from subsequent sustained coro- [20] M. Shimizu, M. Tropak, R. J. Diaz et al., “Transient limb nary occlusion,” Circulation, vol. 87, no. 3, pp. 893–899, 1993. ischaemia remotely preconditions through a humoral mech- [6]B.C.G.Gho,R.G.Schoemaker,M.A.VandenDoel,D.J. anism acting directly on the myocardium: evidence suggesting Duncker, and P. D. Verdouw, “Myocardial protection by brief cross-species protection,” Clinical Science, vol. 117, no. 5, pp. ischemia in noncardiac tissue,” Circulation,vol.94,no.9,pp. 191–200, 2009. 2193–2200, 1996. ff [7]S.C.Lang,A.Elsasser,¨ C. Scheler et al., “Myocardial pre- [21] F. Tomai, F. Crea, A. Gaspardone et al., “E ects of naloxone conditioning and remote renal preconditioning. Identifying a on myocardial ischemic preconditioning in humans,” Journal protective factor using proteomic methods,” Basic Research in of the American College of Cardiology, vol. 33, no. 7, pp. 1863– Cardiology, vol. 101, no. 2, pp. 149–158, 2006. 1869, 1999. [8] M. Cour, L. Gomez, N. Mewton, M. Ovize, and L. Argaud, [22] H. H. Patel, J. Moore, A. K. Hsu, and G. J. Gross, “Cardio- “Postconditioning: from the bench to bedside,” Journal of protection at a distance: mesenteric artery occlusion protects Cardiovascular Pharmacology and Therapeutics, vol. 16, no. 2, the myocardium via an opioid sensitive mechanism,” Journal pp. 117–130, 2011. of Molecular and Cellular Cardiology, vol. 34, no. 10, pp. 1317– [9] D. S. Burley, P. Ferdinandy, and G. F. Baxter, “Cyclic GMP 1323, 2002. and protein kinase-G in myocardial ischaemia-reperfusion: [23] S. P. Loukogeorgakis, A. T. Panagiotidou, M. W. Broadhead, opportunities and obstacles for survival signaling,” British A. Donald, J. E. Deanfield, and R. J. MacAllister, “Remote Journal of Pharmacology, vol. 152, no. 6, pp. 855–869, 2007. ischemic preconditioning provides early and late protection [10] D. J. Hausenloy and D. M. Yellon, “New directions for against endothelial ischemia-reperfusion injury in humans: protecting the heart against ischaemia-reperfusion injury: role of the autonomic nervous system,” Journal of the American targeting the Reperfusion Injury Salvage Kinase (RISK)- College of Cardiology, vol. 46, no. 3, pp. 450–456, 2005. pathway,” Cardiovascular Research, vol. 61, no. 3, pp. 448–460, [24]T.J.Pell,G.F.Baxter,D.M.Yellon,andG.M.Drew, 2004. “Renal ischemia preconditions myocardium: role of adenosine [11] S. Lecour, “Activation of the protective Survivor Activating receptors and ATP-sensitive potassium channels,” American Factor Enhancement (SAFE) pathway against reperfusion Journal of Physiology, vol. 275, no. 5, pp. H1542–H1547, 1998. injury: does it go beyond the RISK pathway?” Journal of [25] T. Oxman, M. Arad, R. Klein, N. Avazov, and B. Rabinowitz, Molecular and Cellular Cardiology, vol. 47, no. 1, pp. 32–40, “Limb ischemia preconditions the heart against reperfusion 2009. tachyarrhythmia,” American Journal of Physiology, vol. 273, no. [12] G. Heusch, K. Boengler, and R. Schulz, “Editorial: inhibition 4, pp. H1707–H1712, 1997. of mitochondrial permeability transition pore opening: the [26] A. Imani, M. Faghihi, S. S. Sadr, S. S. Niaraki, and A. M. holy grail of cardioprotection,” Basic Research in Cardiology, Alizadeh, “Noradrenaline protects in vivo rat heart against vol. 105, no. 2, pp. 151–154, 2010. infarction and ventricular arrhythmias via nitric oxide and 8 Cardiology Research and Practice

reactive oxygen species,” Journal of Surgical Research, vol. 169, [41]S.J.Clarke,I.Khaliulin,M.Das,J.E.Parker,K.J.Heesom, no. 1, pp. 9–15, 2010. and A. P. Halestrap, “Inhibition of mitochondrial permeabil- [27] R. G. Schoemaker and C. L. van Heijningen, “Bradykinin ity transition pore opening by ischemic preconditioning is mediates cardiac preconditioning at a distance,” American probably mediated by reduction of oxidative stress rather than JournalofPhysiology, vol. 278, no. 5, pp. H1571–H1576, 2000. mitochondrial protein phosphorylation,” Circulation Research, [28] C. M. Pedersen, M. R. Schmidt, G. Barnes et al., “Bradykinin vol. 102, no. 9, pp. 1082–1090, 2008. does not mediate remote ischaemic preconditioning or [42] C. Weinbrenner, F. Schulze, L. Sarv´ ary,´ and R. H. Strasser, ischaemia-reperfusion injury in vivo in man,” Heart, vol. 97, “Remote preconditioning by infrarenal aortic occlusion is δ no. 22, pp. 1857–1861, 2011. operative via 1-opioid receptors and free radicals in vivo in the rat heart,” Cardiovascular Research, vol. 61, no. 3, pp. 591– [29]Y.J.Li,Z.S.Xiao,C.F.Peng,andH.W.Deng,“Calci- 599, 2004. tonin gene-related peptide-induced preconditioning protects [43] D. M. Yellon and J. M. Downey, “Preconditioning the against ischemia-reperfusion injury in isolated rat hearts,” myocardium: from cellular physiology to clinical cardiology,” European Journal of Pharmacology, vol. 311, no. 2-3, pp. 163– Physiological Reviews, vol. 83, no. 4, pp. 1113–1151, 2003. 167, 1996. [44] B. Gunaydin,¨ I. C¸ akici, H. Soncul et al., “Does remote organ [30] S. Wolfrum, J. Nienstedt, M. Heidbreder, K. Schneider, P. ischaemia trigger cardiac preconditioning during coronary Dominiak, and A. Dendorfer, “Calcitonin gene related peptide artery surgery?” Pharmacological Research,vol.41,no.4,pp. mediates cardioprotection by remote preconditioning,” Regu- 493–496, 2000. latory Peptides, vol. 127, no. 1–3, pp. 217–224, 2005. [45] M. M. H. Cheung, R. K. Kharbanda, I. E. Konstantinov et [31]L.Xiao,R.Lu,C.P.Hu,H.W.Deng,andY.J.Li,“Delayed al., “Randomized controlled trial of the effects of remote cardioprotection by intestinal preconditioning is mediated by ischemic preconditioning on children undergoing cardiac calcitonin gene-related peptide,” European Journal of Pharma- surgery: first clinical application in humans,” Journal of the cology, vol. 427, no. 2, pp. 131–135, 2001. American College of Cardiology, vol. 47, no. 11, pp. 2277–2282, [32] A. Takaoka, I. Nakae, K. Mitsunami et al., “Renal 2006. ischemia/reperfusion remotely improves myocardial energy [46] Z. Wenwu, Z. Debing, C. Renwei et al., “Limb ischemic metabolism during myocardial ischemia via adenosine preconditioning reduces heart and lung injury after an open receptors in rabbits: effects of ’remote preconditioning’,” heart operation in infants,” Pediatric Cardiology, vol. 31, no. 1, Journal of the American College of Cardiology, vol. 33, no. 2, pp. 22–29, 2010. pp. 556–564, 1999. [47] D. J. Hausenloy, P. K. Mwamure, V. Venugopal et al., “Effect [33] Y. F. Ding, M. M. Zhang, and R. R. He, “Role of renal nerve in of remote ischaemic preconditioning on myocardial injury in cardioprotection provided by renal ischemic preconditioning patients undergoing coronary artery bypass graft surgery: a in anesthetized rabbits,” Acta Physiologica Sinica,vol.53,no.1, randomised controlled trial,” The Lancet, vol. 370, no. 9587, pp. 7–12, 2001. pp. 575–579, 2007. [34] D. A. Liem, P. D. Verdouw, H. Ploeg, S. Kazim, and D. [48] V. Venugopal, D. J. Hausenloy, A. Ludman et al., “Remote J. Duncker, “Sites of action of adenosine in interorgan ischaemic preconditioning reduces myocardial injury in preconditioning of the heart,” American Journal of Physiology, patients undergoing cardiac surgery with cold-blood cardio- vol. 283, no. 1, pp. H29–H37, 2002. plegia: a randomised controlled trial,” Heart, vol. 95, no. 19, [35]J.H.Dong,Y.X.Liu,E.S.Ji,andR.R.He,“Limbischemic pp. 1567–1571, 2009. preconditioning reduces infarct size following myocardial [49]I.A.Rahman,J.G.Mascaro,R.P.Steedsetal.,“Remote ischemia-reperfusion in rats,” Sheng li Xue Bao,vol.56,no.1, ischemic preconditioning in human coronary artery bypass pp. 41–46, 2004. surgery: from promise to disappointment?” Circulation, vol. 122, no. 11, pp. S53–S59, 2010. [36] R. K. Kharbanda, M. Peters, B. Walton et al., “Ischemic [50] D. Brevoord, M. W. Hollmann, S. G. De Hert et al., “Effect preconditioning prevents endothelial injury and systemic of remote ischemic conditioning on atrial fibrillation and neutrophil activation during ischemia-reperfusion in humans outcome after coronary artery bypass grafting (RICO-trial),” in vivo,” Circulation, vol. 103, no. 12, pp. 1624–1630, 2001. BMC Anesthesiology, vol. 11, article 11, 2011. [37] R. K. Kharbanda, U. M. Mortensen, P. A. White et al., “Tran- [51] E. K. Iliodromitis, S. Kyrzopoulos, I. A. Paraskevaidis et al., sient limb ischemia induces remote ischemic preconditioning “Increased C reactive protein and cardiac enzyme levels after in vivo,” Circulation, vol. 106, no. 23, pp. 2881–2883, 2002. coronary stent implantation. Is there protection by remote [38] S. P. Loukogeorgakis, A. T. Panagiotidou, M. W. Broadhead, ischaemic preconditioning?” Heart, vol. 92, no. 12, pp. 1821– A. Donald, J. E. Deanfield, and R. J. MacAllister, “Remote 1826, 2006. ischemic preconditioning provides early and late protection [52] S. P. Hoole, P. M. Heck, L. Sharples et al., “Cardiac remote against endothelial ischemia-reperfusion injury in humans: ischemic preconditioning in coronary stenting (CRISP stent) role of the autonomic nervous system,” Journal of the American study. A prospective, randomized control trial,” Circulation, College of Cardiology, vol. 46, no. 3, pp. 450–456, 2005. vol. 119, no. 6, pp. 820–827, 2009. [39] J. C. Cleveland, D. R. Meldrum, B. S. Cain, A. Banerjee, [53] S. P. Hoole, S. N. Khan, P. A. White et al., “Remote ischaemic and A. H. Harken, “Oral sulfonylurea hypoglycemic agents pre-conditioning does not attenuate ischaemic left ventricular prevent ischemic preconditioning in human myocardium: two dysfunction in humans,” European Journal of Heart Failure, paradoxes revisited,” Circulation, vol. 96, no. 1, pp. 29–32, vol. 11, no. 5, pp. 497–505, 2009. 1997. [54] H. E. Bøtker, R. Kharbanda, M. R. Schmidt et al., “Remote [40] H. Klepzig, G. Kober, C. Matter et al., “Sulfonylureas ischaemic conditioning before hospital admission, as a com- and ischaemic preconditioning. A double-blind, placebo- plement to angioplasty, and effect on myocardial salvage in controlled evaluation of glimepiride and glibenclamide,” patients with acute myocardial infarction: a randomised trial,” European Heart Journal, vol. 20, no. 6, pp. 439–446, 1999. The Lancet, vol. 375, no. 9716, pp. 727–734, 2010. Cardiology Research and Practice 9

[55] K. Munk, N. H. Andersen, M. R. Schmidt et al., “Remote ischemic conditioning in patients with myocardial infarction treated with primary angioplasty impact on left ventricular function assessed by comprehensive echocardiography and gated single-photon emission CT,” Circulation, vol. 3, no. 6, pp. 656–662, 2010. [56] Z. A. Ali, C. J. Callaghan, E. Lim et al., “Remote ischemic preconditioning reduces myocardial and renal injury after elective abdominal aortic aneurysm repair: a randomized controlled trial,” Circulation, vol. 116, no. 11, pp. I98–I105, 2007. [57] S. R. Walsh, U. Sadat, J. R. Boyle et al., “Remote ischemic preconditioning for renal protection during elective open infrarenal abdominal aortic aneurysm repair: randomized controlled trial,” Vascular and Endovascular Surgery, vol. 44, no. 5, pp. 334–340, 2010. [58] S. R. Walsh, J. R. Boyle, T. Y. Tang et al., “Remote ischemic preconditioning for renal and cardiac protection during endovascular aneurysm repair: a randomized controlled trial,” Journal of Endovascular Therapy, vol. 16, no. 6, pp. 680–689, 2009. [59] S. R. Walsh, S. A. Nouraei, T. Y. Tang, U. Sadat, R. H. Carpenter, and M. E. Gaunt, “Remote ischemic precondi- tioning for cerebral and cardiac protection during carotid endarterectomy: results from a pilot randomized clinical trial,” Vascular and Endovascular Surgery, vol. 44, no. 6, pp. 434–439, 2010. [60]S.Hu,H.L.Dong,Y.Z.Lietal.,“Effects of remote ischemic preconditioning on biochemical markers and neurologic out- comes in patients undergoing elective cervical decompression surgery: a prospective randomized controlled trial,” Journal of Neurosurgical Anesthesiology, vol. 22, no. 1, pp. 46–52, 2010. [61] V. Venugopal, C. M. Laing, A. Ludman, D. M. Yellon, and D. Hausenloy, “Effect of remote ischemic preconditioning on acute kidney injury in nondiabetic patients undergoing coronary artery bypass graft surgery: a secondary analysis of 2 small randomized trials,” American Journal of Kidney Diseases, vol. 56, no. 6, pp. 1043–1049, 2010. [62] R. F. Zimmerman, P. U. Ezeanuna, J. C. Kane et al., “Ischemic preconditioning at a remote site prevents acute kidney injury in patients following cardiac surgery,” Kidney International, vol. 80, no. 8, pp. 861–867, 2011. [63] K. R. Pedersen, H. B. Ravn, J. V. Povlsen, M. R. Schmidt, E. J. Erlandsen, and V. E. Hjortdal, “Failure of remote ischemic preconditioning to reduce the risk of postoperative acute kidney injury in children undergoing operation for complex congenital heart disease: a randomized single-center study,” Journal of Thoracic and Cardiovascular Surgery. In press. [64]Y.S.Choi,J.K.Shim,J.C.Kimetal.,“Effect of remote ischemic preconditioning on renal dysfunction after complex valvular heart surgery: a randomized controlled trial,” Journal of Thoracic and Cardiovascular Surgery, vol. 142, no. 1, pp. 148–154, 2011. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 121237, 13 pages doi:10.1155/2012/121237

Review Article Toll-Like Receptors in Ischaemia and Its Potential Role in the Pathophysiology of Muscle Damage in Critical Limb Ischaemia

Hemanshu Patel,1 Sidney G. Shaw,2 Xu Shi-Wen,3 David Abraham,3 Daryll M. Baker,1 and Janice C. S. Tsui1

1 Division of Surgery & Interventional Science, University College London, Royal Free Campus, London NW3 2QG, UK 2 Department of Clinical Experimental Research, University of Bern, 3010 Bern, Switzerland 3 Centre for Rheumatology and Connective Tissue Disease, University College London, Royal Free Campus, London, NW3 2QG, UK

Correspondence should be addressed to Hemanshu Patel, [email protected]

Received 15 July 2011; Accepted 4 October 2011

Academic Editor: Gavin W. Lambert

Copyright © 2012 Hemanshu Patel et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Toll-like receptors (TLRs) are key receptors of the innate immune system which are expressed on immune and nonimmune cells. They are activated by both pathogen-associated molecular patterns and endogenous ligands. Activation of TLRs culminates in the release of proinflammatory cytokines, chemokines, and apoptosis. Ischaemia and ischaemia/reperfusion (I/R) injury are associated with significant inflammation and tissue damage. There is emerging evidence to suggest that TLRs are involved in mediating ischaemia-induced damage in several organs. Critical limb ischaemia (CLI) is the most severe form of peripheral arterial disease (PAD) and is associated with skeletal muscle damage and tissue loss; however its pathophysiology is poorly understood. This paper will underline the evidence implicating TLRs in the pathophysiology of cerebral, renal, hepatic, myocardial, and skeletal muscle ischaemia and I/R injury and discuss preliminary data that alludes to the potential role of TLRs in the pathophysiology of skeletal muscle damage in CLI.

1. Introduction blood flow. However, even successful revascularisation is not associated with an improvement in the functional ability Critical limb ischaemia (CLI) is the most severe form of of patients with CLI [3], and most patients have persistent peripheral arterial disease (PAD). Whilst PAD describes or recurring symptoms requiring further treatment [4]. In stenotic or aneurysmal disease in any arterial bed except addition, a significant number of patients with CLI are not the coronary arteries, CLI generally describes advanced suitable for revascularisation and treatment is limited to atherosclerosis in the lower limb arteries leading to a pharmacological agents such as iloprost where outcomes reduced blood supply to the tissues of the lower limb have been unsuccessful or inconsistent [5] or amputation. resulting in rest pain and/or tissue loss. PAD affects 27 Whilst work has been carried out on the aetiology of PAD the million people in Western Europe and North America, and downstream effects of reduced blood flow to the main organ, approximately 1-2% of these patients will develop CLI [1]. that is, skeletal muscle are still poorly understood. A better Further, CLI is associated with significant morbidity and understanding of the pathophysiological processes occurring mortality: a large observational multicentre cohort study within the skeletal muscle in CLI may enable us to identify of CLI patients observed a 6-month amputation rate of potential therapeutic targets. Recent studies on ischaemia 12% and 1-year mortality rate of 19.1% [2]. However and ischaemia/reperfusion (I/R) injury of various organs despite the importance of this condition, management systems have identified the involvement of toll-like receptors of CLI patients continues to be challenging with limited (TLRs) in the pathogenesis of hypoxic/ischaemic injury [6]. treatment modalities available. Surgical or endovascular We aim to review this evidence for the role of TLRs in intervention remains the mainstay of therapy by improving ischaemia and ischaemia/reperfusion injury as well as discuss 2 Cardiology Research and Practice

Table 1: Exogenous/endogenous ligands and antagonists of TLRs; n.d.: not discovered.

TLR Microbial ligands Endogenous ligands Antagonists TLR1/TLR2 Triacyl lipopeptides (Pam3CSK4) n.d. n.d.

Diacyl lipopeptides (Pam2CSK4), zymosan, porins, bacterial TLR2/TLR6 HSP-60, HSP-70, HMGB-1 n.d. peptidoglycan, LPSs of gram positive bacteria TLR3 Ds RNA mRNA n.d. HSP-22, HSP-60, HSP-70, HSP-96, Eritoran (E5564, a lipid A derivative), TLR4 LPS fibrinogen, HMGB-1, hyaluronan TAK-242 fragments, fibronectin (extra domain A) TLR5 Flagellin n.d. n.d. TLR7 ssRNA (viral) ssRNA (immune complexes) CpG ODN, CpG 52364 TLR8 ssRNA (viral) ssRNA (immune complexes) CpG 52364 TLR9 DNA (bacterial/viral) DNA (immune complexes) CpG ODN, CpG 52364 TLR11 Toxoplasma gondii n.d. n.d. TLR10, 12, 13 n.d. n.d. n.d. the potential implication of TLRs in the pathophysiology of receptor dimerization: almost all the TLRs form homodimers skeletal muscle in CLI. except TLR 1, 2, and 6 whilst TLR 2 can heterodimerise with either TLR 1 or 6 depending upon the ligand that is presented [16]. The exogenous TLR ligands can be 2. Toll-Like Receptors subdivided into 3 groups (Table 1). The first group consists Toll-like receptors are key receptors of the innate immune of lipids which are recognised by TLR 1, 2, 4, and 6, the system as they recognise and respond to components second group are proteins which bind to TLR 5 and 11, of invading microorganisms termed pathogen-associated and the third group consists of nucleic acids which activate molecular patterns (PAMPs). PAMPs consist of lipids, the intracellular TLRs such as TLR 3, 7, 8, and 9. Recent lipopeptides, proteins, and nucleic acids [7], and upon studies have identified numerous host-derived ligands of binding to TLRs they lead to the activation of the TLR TLRs that are released under certain physiological and signalling pathway. This culminates in the release of various pathophysiological conditions. These are also summarised in cytokines, chemokines, and interferons that has implications Table 1, however particularly interesting endogenous ligands for both the innate and adaptive immune systems [8]. TLRs are heat shock proteins (HSP) [17], hyaluronic acid [18], are type 1 membrane glycoproteins that consist of a ligand- and HMGB-1. These ligands are secreted in states of shock binding external domain comprised of 19–25 leucine rich or tissue injury and are therefore involved in activating repeat (LRR) motifs and a cytoplasmic signalling domain TLRs in certain pathological conditions such as hypoxia and that is termed the toll/interleukin 1 (TIR) domain. So ischaemia. far 13 TLRs have been identified in mammals of which The emerging importance of TLR activation in the eleven functional TLRs (TLR 1–11) have been discovered pathogenesis of numerous conditions has led to the devel- in humans. These can be subdivided by their subcellular opment of a number of synthetic TLR antagonists. So localisation. TLR 1, 2, 4, 5, 6, and 10 are expressed on far these antagonists are structural analogues of agonists the cell surface whereas TLR 3, 7, 8, and 9 are located in which prevent the stimulating ligand binding to the receptor the intracellular compartments, typically in the endosomes [19]. Disease modulation by TLR antagonism is not only and the endoplasmic reticulum [9]. Toll-like receptors are being studied in animal models but a number of clinical expressed on both immune and nonimmune cells such as trials in humans are underway for the treatment of septic macrophages [10], neutrophils [10],Bcells[11]aswellas shock and autoimmune disorders. Nonetheless there are very epithelial cells [12], myocytes [13], and skeletal muscle [14]. limited TLR antagonists available at present, and a better understanding of TLR ligands will aid in the development of a broader spectrum of antagonists. In this respect, various 3. Toll-Like Receptor Ligands and Antagonists groups are using screening techniques such as systemic evolution of ligands by exponential enrichment (SELEX) and In addition to PAMPS, TLRs are stimulated by host- high-throughput screening (HTS) to identify and develop derived molecules such as high mobility group box 1 oligonucleotides and synthetic phospholipid compounds protein (HMGB-1) [15]. TLRs achieve ligand specificity by with the potential to inhibit TLR signalling. Cardiology Research and Practice 3

MyD88-dependent signalling pathway MyD88-independent signalling pathway

All TLRs except TLR3 TLR4 TLR3

Cytoplasm TIRAP TRAM TRIF MyD88 TRIF

IRAK4 IRAK1 TRIF

TAB1 TAB2 TAK1 TRAF6 TRAF3 P13K Uev1A UBC13 RIP-1

MAPKK IKKγ IKKγ

IKKα IKKβ IKKα IKKβ TBK1 IKKi P38 JNK

κ P IκB P I B AKT AP1 IRF3 P50 P65 P50 P65

Nucleus AP1 P50 P65 P50 P65 IRF3 IRF2

Figure 1: TLR signalling pathway. MyD88-dependent signalling pathway is used by all TLRs except TLR 3. Signalling through the MyD88- dependent pathway leads to the activation of MAPKK and IKK complex resulting in activation and nuclear translocation of AP-1 and NF-κB, respectively. TLR 4 is capable of signalling through the MyD88-independent pathway as well; however this is the sole signalling mechanism for TLR 3. TRIF is the main adaptor protein in the MyD88-independent pathway and can associate with TRAF6 to activate AP-1 and NF-κB. Alternatively it also activates NF-κB by interacting with RIP-1. TRIF can further interact with TRAF3 and the phosphatidylinositol 3-kinase (PI3K)-AKT pathway resulting in the nuclear translocation of IRF3 and IRF2, respectively.

4. TLR Signalling (IRAK) family of which four members have so far be identified: IRAK1, IRAK2, IRAK3, and IRAK-M. Two sepa- Stimulation of TLRs upon ligand recognition leads to rate signalling pathways exist for the signalling transduction the activation of its downstream signalling cascade which of TLRs: these are termed the MyD88-dependent pathway culminates in the activation of the transcription factors, of which MyD88 plays a central role and the MyD88- nuclear factor-κB(NF-κB), and activator protein 1 (AP- independent pathway which uses TRIF instead. All TLRs 1). This results in the release of various proinflammatory utilise the MyD88-dependent pathway except TLR3; interest- cytokines such as IL-6, IL-1, and TNF-α. The initiation ingly TLR4 is capable of using either signalling pathway. of this signalling cascade depends upon the binding of In the MyD88-dependent pathway (Figure 1), ligand the TIR domain of the intracellular portion of TLRs binding leads to the association of MyD88 to the TIR to the TIR domain-containing cytosolic adapter proteins. domain of the receptor. This leads to the phosphorylation The four main adapter proteins are myeloid differentia- of IRAK4 which in turn phosphorylates IRAK1. IRAK1 tion primary response protein 88 (MyD88), TIR domain- binds to and activates tumour necrosis factor receptor- containing adapter protein (TIRAP/Mal), TIR domain- associated factor 6 (TRAF6). The IRAK1/TRAF6 complex containing adapter-inducing IFNβ (TRIF, also known as dissociates from the intracellular receptor complex and TICAM1), and TRIF-related adapter molecule (TRAM, also forms a complex with the E2 ligases Ubc13 and Uev1A known as TICAM2) [20]. Also important for TLR signalling which have been shown to catalyse the synthesis of a Lys transduction is the interleukin-1 receptor-associated kinase 63-linked polyubiquitin chain of TRAF6. This complex 4 Cardiology Research and Practice

Table 2: Summary of the evidence for the role of TLRs in the pathogenesis of tissue damage in ischaemia and I/R injury.

Potential endogenous Organ ischaemia Upregulation of EvidenceforroleofTLRin Expression of TLRs ligand implicated in and I/R injury TLRs in ischaemia pathophysiology pathogenesis TLR 2 and 4 knockout mice have Glialcells:TLR1–9[32, 33] TLR 2, 4 and 9 Cerebral reduced infarct size following HSP 70 [34] Neurons: TLR 2 and 4 [34] [34, 35] ischaemia [34, 36–38] Hepatocytes: TLR 2, 3, 4 TLR 4 and 9 knockout mice are and 5 [39] HSP 72 [44]and Liver TLR 2 [40] protected against ischaemia-induced Non-parenchymal cells: HMGB-1 [42, 45] liver injury [41–43] TLR2,3,4,5,7and9[39] TLR 2 and 4 knockout mice are Parenchyma: TLRs 1–10 at HMGB-1, hyaluronan TLR 2 and 4 protected against renal I/R injury, and Renal varying detection levels and biglycan [47, 49] this is associated with a reduction in [46–48] [49, 51, 53] inflammatory cytokine levels [49–52] TLR 2- and 4-deficient mice show reduced myocardial infarct size [56–59]. TLR 4 antagonist eritoran Myocytes: TLR 2, 3, 4 and 6 Myocardial TLR 4 [55] leads to reduced infarct size, NF-κB HMGB-1 [61] [54] nuclear translocation, and proinflammatory cytokine expression [60] TLR 2 antagonsim reduces pro-inflammatory cytokine expression Skeletal muscle TLR 1–9 [62–64] TLR 2, 4, 6 [65] Under investigation in an in vitro model of skeletal muscle ischaemia [65] further associates with a member of the MAP kinase kinase TRIF leads to the phosphorylation and the consequent kinase family, transforming growth factor-β-activated kinase activation of IRF 3 and 7 leading to the induction of type 1 (TAK1) and the TAK1-binding proteins, TAB1 and TAB2. 1 IFNs. Two members of the IKK family IKKi (IKKε)and This complex formation subsequently activates TAK1 which TBK1 [TRAF family member-associated NF-κBactivator simultaneously activates the IκB kinases (IKK) complex and (TANK) binding kinase-1 or T2K or NAK] are however members of the MAP kinase family such as extracellular essential to this role [23, 24]. signal-regulated kinase (ERK), c-Jun kinase (JNK), and p38. α β The IKK complex consists of two kinases, IKK and IKK 5. Ischaemia and Ischaemia/Reperfusion Injury and a regulatory subunit (NEMO/IKKγ)[21]. IKK-activated complex subsequently phosphorylates the inhibitory IκB The pathophysiology of ischaemia and ischaemia/reperfu- proteins which normally sequester NF-κB in an inactive sion- (I/R-) induced injury has been extensively studied and form in the cytoplasm. IκB protein phosphorylation leads evaluated in various organ systems such as the kidneys, liver, to their polyubiquitylation and subsequent degradation with cardiovascular, and central nervous systems (CNS) (Table 2). the concomitant release and nuclear translocation of NF- There is growing evidence that TLRs play an important κB. The activation of members of the MAPK family such as role in the propagation of the tissue damage caused by JNK leads to the subsequent phosphorylation and activation ischaemia and I/R. However there is still a lack of significant of the transcription factor AP-1. Further, stimulation of the progress in understanding the pathophysiology of skeletal MyD88-dependent pathway in TLR 7, 8, and 9 can also lead muscle damage in CLI and the role TLRs play in this disease to the induction of type 1 interferons(IFN) possibly through process. Following ischaemia or I/R a number of cellular the activation of interferon regulatory factor 7 (IRF7) [22]. and biochemical changes occur both locally and systemically. The MyD88-independent pathway which is the sole This include the recruitment of activated neutrophils [25] signalling transduction system of TLR 3 and is an alternative and lymphocytes [26], production of reactive oxygen species signalling pathway for TLR 4 also culminates in the activation (ROS), release of cytokines [27] and chemokines [28], and of NF-κB and AP-1 as well as IRF3 leading to the induction of activation of the complement system [29]. The generation type 1 interferons. In the MyD88-independent pathway TRIF of ROS, ATP deletion, and activation of enzymes such as is the main adaptor protein. In TLR 3 signalling TRIF binds phospholipases and proteases lead to cell necrosis. Apoptosis directly to the cytoplasmic portion of the receptor; however has also been shown to occur following ischaemia [30], in TLR 4 signalling the adaptor protein TRAM acts as a and it is thought that mitochondrial dysfunction secondary bridge between TRIF and the TIR-containing domain. TRIF to ROS generation plays a role in inducing the apoptotic can interact with TRAF6 or receptor-interacting protein-1 mechanism. In addition, TLRs seem to play an important (RIP-1) leading to the activation of NF-κB and AP-1. Further role in mediating some of the ischaemia-induced injury. Cardiology Research and Practice 5

Endogenous ligands of TLRs such as fibrinogen, heparin had MCA occlusion suggesting actual clinical improvement sulphate, hyaluronan, HSP60, HSP70, and HMGB-1 are as a result of abolishing the effects of TLR 4. Interestingly, released by injured and necrotic cells. The subsequent studies have shown that LPS (TLR 4 ligand) pre-conditioning stimulation of TLRs by these ligands leads to the activation helps to protect against subsequent ischaemic damage; of transcription factors NF-κB and AP-1 with a consequent however the mechanism for this is unclear. Thus, there is release of proinflammatory cytokines. Further TLRs have now significant evidence that TLRs in particular TLR 2 been implicated in directly causing apoptosis via a pathway and 4 play a role in ischaemia-induced cerebral damage involving Fas-associated death domain protein (FADD) and but the exact mechanisms are yet to be elucidated. Apart caspase 8 [31]. from the increase in proinflammatory cytokines, activated glial TLRs also lead to the release of chemokines such as macrophage inflammatory protein-2 (MIP-2) and monocyte 6. The Role of Toll-Like Receptors in chemoattractant protein-1 (MCP-1) which attract peripheral the Pathophysiology of Cerebral Ischaemia immune cells into the brain parenchyma and may lead to and I/R Injury further exacerbation of ischaemic damage. Using a model of glucose deprivation as a form of energy deprivation, Tang et There is emerging evidence that TLRs play a role in neuronal al. [34] found that TLR2 and 4 are implicated in apoptotic damage secondary to cerebral ischaemia. TLRs are thought neuronal cell death via activation of the JNK-AP-1 pathway. to be expressed primarily by the glial cells (microglia, In addition they showed that the endogenous ligand of TLR astrocytes, and oligodendrocytes) in the brain. Human 2 and 4, HSP70, is upregulated thus implicating HSP70 and microglia are found to express TLRs 1–8 at detectable levels other endogenous ligands of TLRs in the pathogenesis of andTLR9atlowbutdetectablelevels[32]. The expression of cerebral ischaemia. TLRs in human astrocytes, however, is more restricted with only TLR 3 mRNA detected at intermediate levels; mRNA forTLRs1,2,4,5,and9weredetectablebutwerelow, 7. The Role of Toll-Like Receptors in andTLRs6,7,8,and10mRNAexpressionwasfoundto the Pathophysiology of Liver Ischaemia be rare to undetectable [33]. Very little is known about the and I/R Injury expression of TLRs in human oligodendrocytes, but Bsibsi et al. [32] have reported expression of TLRs 2 and 3 in these Liver ischaemia and reperfusion can occur during a variety cells. TLRs 2 and 4 have also been found to be expressed in of situations particularly during surgical procedures such mouse cerebral cortical neurons [34]. When microglia and as liver transplant, vascular reconstruction, liver trauma, astrocytes are exposed to TLR ligands such as peptidoglycan, and resection of large hepatic tumours. During the initial double-stranded RNA, lipopolysaccharide, and bacterial ischaemic period a certain level of cellular damage has been DNA there is a release of a wide range of proinflammatory shown to occur [68]; however this is further exacerbated cytokines (including TNF-α, IL-6, and IL-12), chemokines, following reperfusion [69]. The liver is primarily composed and reactive oxygen species. This suggests that the TLRs are of parenchymal cells (hepatocytes) constituting 65% of involved in protecting the CNS against microbial infections, the cells in the liver. The remaining part of the liver is although it is unclear whether this neuroinflammatory composed of nonparenchymal cells such as kupffer cells, response is beneficial or detrimental. Growing evidence sinusoidal endothelial cells, biliary epithelial cells, hepatic suggests that the TLRs expressed in the CNS also play an stellate, and dendritic cells. Studies have shown that both important role in tissue development, cellular migration and the parenchymal and nonparenchymal cells express a large differentiation, and in limiting inflammation [66]. repertoire of TLRs [39]. Acute inflammation exacerbates brain damage in cere- Kupffer cells which are the resident macrophages of the bral ischaemia, and activation of the innate immune system liver play a central role in hepatic I/R injury. Following is an important component of this process. There is now hepatic ischaemia kupffer cells are activated and lead to strong evidence suggesting that TLRs within the CNS play an the release of both proinflammatory cytokines (TNF-α, integral part in this inflammatory process. For example TLRs IL-6, and IL-1) as well as anti-inflammatory cytokines 2, 4, and 9 have been shown to be upregulated and activated (IL-10 and IL-13) and ROS [70]. Inflammatory cytokines in cerebral ischaemia models [34, 35]. Significantly a number both initiate and maintain the inflammatory response and, of groups have reported reduced infarct size in TLR 2 and 4 together with the activated complement system, result in the knockout mice that have been exposed to cerebral ischaemia migration and adhesion of leukocytes and recruitment of [34, 36–38]. Kilic et al. [36] performed intraluminal middle neutrophils within the sinusoids [71]. Activated neutrophils cerebral artery occlusion as a model of ischaemic stroke in further exacerbate liver damage induced by reperfusion, adult male C3H/HEJ TLR 4 knockout mice. They showed through the release of more ROS and proteases [72]. The that in the TLR4-deficient mice there was reduced ischaemic resulting increase in ROS causes oxidative stress and cell neuronal injury, and the mechanism of this neuroprotective death. Significantly, ROS especially hydrogen peroxide [73] effect was associated with deactivation of the MAP kinases activate NF-κB which has been found to play a significant ERK 1, ERK2, JNK1, JNK2, and P38 [36]. Further, Caso et role in liver I/R injury [74]. Cell death in liver I/R has al. [67] not only demonstrated reduced infarct size but also been shown to occur by both apoptosis [75] and necrosis recovery of neurological deficit in TLR 4 knockout mice that [76]. Inflammatory mediators such as TNF-α released by 6 Cardiology Research and Practice kupffer cells and neutrophils have been shown to activate hepatocytes. HMGB-1 levels increase during liver I/R, and pro-apoptotic proteins such as caspase-3 and caspase-8 inhibition of HMGB-1 with a neutralising antibody reduces further highlighting the role of these cells in ischaemic liver liver damage after I/R [42, 45]. Thus, numerous TLR damage [77]. endogenous ligands have been implicated in hepatic I/R Several groups have illustrated the importance of TLR injury; however it remains to be seen how much each of them signalling in hepatic I/R damage [40–42, 45]. In particular contributes to hepatic I/R injury. TLR 4 and 9 are specifically implicated in the pathological process. Tsung et al. [41] showed that chimeric mice lacking functional TLR 4 subjected to liver I/R were protected from 8. The Role of Toll-Like Receptors in tissue damage. Further they showed that this protective the Pathophysiology of Renal Ischaemia ff e ect was associated with a reduction in the activation of and I/R Injury JNK and NF-κB as well as a decrease in the expression of the proinflammatory cytokines TNF-α and IL-6 and the Renal ischaemia and I/R injury can occur during trans- adhesion molecule ICAM-1. Shen et al. [43] also showed plantation, partial nephrectomy, aortic cross-clamping, and that TLR 4 knockout mice were protected against hepatic following systemic hypotension and is a common cause of I/R injury, and this was associated with a reduction in local acute renal failure (ARF). The pathogenesis of ischaemia and systemic TNF-α levels as well as reduced neutrophil and I/R-induced renal injury is complex and incompletely infiltration. understood. However, it is not surprising that the innate Circulating levels of HSP72 have shown to be increased immune system plays a significant role in the injury during hepatic I/R injury [44, 78]. HSP72 stimulated both process. Ischaemia leads to the depletion of cellular ATP, TLR 2 and 4 in hepatocytes leading to the activation and this causes tubular epithelial cells to undergo necrosis of NF-κB and the subsequent production of macrophage or apoptosis [81]. In addition to the cytotoxic effects of inflammatory protein-2 (MIP-2) but not TNF-α or IL-6. hypoxia, I/R triggers numerous inflammatory events. The MIP-2 promotes neutrophil infiltration during hepatic I/R renal endothelial and epithelial cells release inflammatory injury, and blockade of MIP-2 has been shown to reduce cytokines (IL-1, IL-6 and TNF-α)[82] as well as chemokines hepatic I/R injury [79]. In TLR 2 and 4 knockout mice, MIP- and express adhesion molecules that activate lymphocytes. 2 production is reduced suggesting that both TLR 2 and 4 Further, the infiltrating leukocytes also generate cytokines contribute to hepatic I/R injury [78]. However, despite an and ROS that exacerbate cellular injury. Several studies increase in TLR 2 mRNA levels after hepatic I/R injury [40] have also highlighted the importance of activation of the it has been found that TLR 4 but not TLR 2 is required in complement system [83], neutrophils [84], B cells [85], and initiating the hepatic injury cascade as it was demonstrated T cells [86] in the development of renal I/R injury. that TLR 2 knockout mice and wild-type mice livers suffered TLR expression in the kidney has been studied by several comparable I/R injury [80]. Further, Zhai et al. [80]have groups and mRNA for almost all the TLRs have been detected suggested that TLR 4-induced hepatic damage in I/R is in human kidneys [46–48].HoweverTLRs1,2,3,4,5,and mediated through the MyD88-independent pathway rather 7 were found to be more abundantly expressed than TLRs than the MyD88-dependant pathway. They showed that IRF 6, 8, 9, and 10. TLR 2 and 4 are expressed in mouse renal 3-deficient mice protected their livers from I/R injury in cortex and medulla, specifically in the proximal and distal a similar fashion to TLR 4-deficient mice but in MyD88- tubules as well as in the epithelium of Bowman’s capsule deficient mice significant hepatic I/R injury still occurred. [47]. Recent studies have implicated TLR 2 and 4 in the This finding however cannot account for the increase in pathogenesis of renal ischaemia and I/R injury. It has been the proinflammatory cytokines which occurs during liver shown that TLR 2 and 4 mRNA is upregulated following ischaemia and may be explained by the work carried out I/R in the epithelial cells of the distal tubules, thin limb of by Bamboat et al. [42] demonstrating that TLR 9 may the loops of Henle, and collecting ducts. This upregulation also be involved in hepatic I/R injury. They reported that was shown to be mediated by IFN-γ and TNF-α [47]. It following I/R TLR 9 knockout mice showed minimal hepatic has been found that TLR 2 knockout mice are protected damage associated with reduced proinflammatory cytokine against renal I/R injury [50, 51]. Leemans et al. [51] used levels (TNF-α, IL-6, and MCP-1) compared to wild-type TLR 2 antisense oligonucleotide treatment to reduce TLR 2 mice which exhibited severe hepatocellular necrosis. TLR protein in mice kidney and showed that this also protected 9 blockade also reduced liver damage and cytokine levels the kidney against I/R injury by demonstrating reduced renal in wild-type mice that were subjected to I/R. This suggests dysfunction. The detrimental effects of TLR 2 signalling in that both TLR 4 and 9 play an important role in hepatic renal I/R injury were observed to be due to an increase damage secondary to ischaemia and I/R injury and that TLR in chemokine and cytokine (MIP-2, TNF-α,IL-6,IL-1β) 9 stimulation may mediate the increase in proinflammatory production, granulocyte and macrophage infiltration, as well cytokines. as tubular necrosis, and tubular epithelial cell apoptosis, In addition to HSP72, other endogenous TLR ligands which was mediated through nonhematopoietic cells of the may be involved in hepatic I/R. Bamboat et al. [42]have kidney. Interestingly Shigeoka et al. [87] have suggested a shown that DNA from necrotic hepatocytes stimulate TLR fascinating concept that renal damage in I/R due to TLR 9 leading to cytokine release. HMGB-1 is a DNA-binding 2 activation is mediated through both a MyD88-dependent protein which is released by necrotic cells including pathway as well as a TLR 2-dependent/MyD88-independent Cardiology Research and Practice 7 pathway.TLR 4 has also been shown to play a significant mice have reduced myocardial infarct size when compared role in renal I/R injury. Wu et al. [49] have demonstrated with control mice [57, 58, 95]. Studies have shown that increased TLR 4 expression following kidney ischaemia. As the reduction in myocardial infarct size in TLR 4-deficient demonstrated in TLR 2 knockout mice, it has also been mice is attributed to a reduction in neutrophil infiltration shown that TLR 4 knockout mice are protected against renal and reduced JNK, NF-κB, and AP-1 activation with a dysfunction following I/R injury [49, 50, 52]. Further TLR subsequent decrease in the levels of inflammatory cytokines 4 knockout mice subjected to I/R showed reduced tubular (IL-1β and Il-6) and monocyte chemotactic factor-1 [57, damage, neutrophil and macrophage accumulation, and 60, 95]. The P13K/AKT pathway has also been reported inflammatory cytokine and chemokine production.Invitro, to play a role in protecting against myocardial I/R injury wild-type kidney tubular epithelial cells (TECs) that were in TLR 4-deficient mice [96]. Mice pretreated with the subjected to ischaemia produced inflammatory cytokines TLR 4 antagonist eritoran prior to transient occlusion of and chemokines and underwent apoptosis. These effects the left anterior descending artery were shown to develop were attenuated in TLR4−/− and MyD88−/− TECs [49]. These significantly smaller infarcts compared to mice treated with results provide significant evidence for the role of TLR 2 vehicle alone. Further, eritoran pretreatment resulted in and 4 in the pathogenesis of ischaemia and I/R-mediated reduced JNK phosphorylation, NF-κB nuclear translocation, renal damage. A number of studies have demonstrated an and proinflammatory cytokine expression [59]. upregulation of several TLR 2 and 4 endogenous ligands There is also emerging evidence for the role of TLR 2 such as HMGB-1, hyaluronan and biglycan in the kidney in myocardial I/R Injury. TLR 2 knockout mice subjected during I/R injury [49, 51, 53] which may be involved in the to myocardial infarction have been shown to have a higher activation of these receptors and subsequent inflammatory survival rate than wild-type mice associated with a smaller response and tissue damage in renal I/R. infarct size, reduced ROS production, and leukocyte infil- tration [97, 98]. Further, both bone marrow chimeric mice developed by transplanting TLR 2 knockout bone marrow 9. The Role of Toll-Like Receptors in to WT mice or WT bone marrow to TLR 2 knockout mice the Pathophysiology of Myocardial submitted to I/R 5 weeks after transplant displayed similar −/− Ischaemia and I/R Injury protection as TLR2 mice against I/R-induced endothelial dysfunction, suggesting a role for TLR 2 expressed on both Myocardial ischaemia is most commonly due to occlusion non-bone marrow cells such as endothelial cells and/or of a major coronary artery. Coronary artery occlusion cardiomyocytes and cells of bone marrow origin such as and the consequent reduction in blood flow usually occur neutrophils [98]. TLR 2 deficiency also abolished increased due to fissuring or erosion of an atherosclerotic plaque IL-1β expression but did not affect TNF-α or IL-6 expression. with subsequent formation of thrombus. The pathogenesis These studies provide some insight into the role of TLR 2 of ischaemia-induced myocardial damage has been inten- and 4 in myocardial I/R injury and may reveal potential sively studied, and a number of biochemical and cellu- therapeutic targets. lar mechanisms have been discovered. Oxygen deficiency HMBG-1 may again act as an endogenous TLR ligand induces metabolic changes such as decreased ATP and pH in myocardial ischaemia. Andrassy et al. [99] demonstrated as well as lactate accumulation. The altered biochemical elevated levels of HMGB-1 following hypoxia in cardiomy- status leads to impaired membrane transport resulting in ocytes in vitro and in ischaemic injury of the heart in an imbalance in intracellular electrolytes and propagates vivo. They also reported that treatment with recombinant various other pathological metabolic changes resulting HMGB-1 worsens I/R injury, whereas treatment with the in cardiomyocyte death through necrosis and apoptosis. HMGB-1 antagonist HMGB-1 box A reduced infarct size and Irreversible damage occurs after approximately 30 min- markers of tissue damage. In addition their data suggested utes of coronary artery occlusion. Reperfusion exacerbates that HMGB-1-mediated myocardial damage involved JNK, myocardial damage which is predominantly the effect of ERK 1 and 2, and NF-κB activation. Several other known oxygen radicals, calcium loading, and neutrophil activation. HMGB-1-inhibiting agents (ethyl pyruvate, green tea, and Oxygen radicals cause further membrane damage, and neu- adrenomedullin) have also been shown to preserve cardiac trophils release inflammatory mediators and contribute to function following a myocardial ischaemic insult [54, 100, microvascular obstruction [88–90]. Inflammatory cytokines 101]. Further studies may reveal the involvement of other such as IL-6, TNF-α,IL-1β, and IL-8 have been shown potential endogenous ligands in the setting of myocardial to be upregulated during periods of myocardial ischaemia ischaemia. [91–93]. Most of the TLRs are expressed within the cardiovascular system. In particular, TLR 2, 3, 4, and 6 are expressed in 10. The Potential Role of Toll-Like rat cardiomyocytes [94] whilst both healthy and atheroscle- Receptors in the Pathophysiology of rotic arteries express TLRs 1–9 [55]. TLR 2 and 4 have Critical Limb Ischaemia been strongly implicated in myocardial damage following ischaemia and reperfusion. Murine TLR 4 expression has CLI is a severe form of PAD that is predominantly caused by been shown to be increased after myocardial infarction atherosclerosis in the peripheral arterial system. Whilst the [56], and various studies have shown that TLR 4-deficient detailed pathophysiology of CLI is not well understood, there 8 Cardiology Research and Practice is a basic understanding of the pathogenesis of skeletal mus- cle damage in peripheral arterial disease (PAD) and in I/R injury. A large proportion of mass in the limb is comprised TLR2 of skeletal muscle, and therefore it is affected significantly by ischaemia-induced tissue damage. Studies have shown that irreversible muscle damage begins to occur after 3 hours of ischaemia and is nearly complete after 6 hours [102]. Pipinos et al. [103] have proposed a potential pathway for TLR4 the pathogenesis of PAD-induced manifestations such as skeletal muscle damage and tissue loss, where inflammation, neutrophil activation and degranulation, and mitochondrial dysfunction with excessive production of ROS occur when blood supply is critically compromised. The consequences TLR6 of these pathological processes include reduced energy production and damage to muscle as well nerves, skin, and subcutaneous tissue. Indeed, Hayes et al. [104] found that after a sustained period of ischaemia, diminishing ATP levels Tubulin correlated closely with worsening muscle necrosis. Further, gastrocnemius muscle biopsies obtained from patients with PAD contain a higher number of apoptotic cells com- pared to controls as shown by Terminal Deoxynucleotidyl Transferase Biotin-dUTP Nick-end Labelling- (TUNEL-) Control CLI staining and increased Caspase-3 activity [105]. Raised (a) circulating levels of proinflammatory cytokines (TNF-α; 0.2 14.48 ng/ml versus 9.32 ng/ml and IL-6; 11.81 ng/ml versus ∗ 7.30 ng/ml), chemokines (vascular cell adhesion molecule- 0.15 1 (VCAM-1); 485.09 ng/ml versus 464.35 ng/ml, and inter- TLR 2 cellular adhesion molecule-1 (ICAM-1); 316.7 ng/ml versus 0.1 207.65 ng/ml) are found in the plasma of patients with 0.05 PAD [106]. Significantly, TNF-α andIL-6havebeenshown Densitometry image units to induce muscle proteolysis, and this in turn has been 0 associated with reduced muscle mass and strength [62, 107, Control CLI 108]. In addition TNF-α has also been reported to induce 0.12 ∗ apoptosis in skeletal myoblasts [63]. It is therefore possible 0.1 that the elevated levels of cytokines play a role in skeletal 0.08 muscle damage in CLI. TLR 4 TLRs 1–9 have been shown to be expressed in skeletal 0.06 muscle [61, 64, 109]. Further, there is evidence to suggest 0.04 that some of the TLRs are functional as lipopolysaccharide 0.02

(LPS) and a synthetic tripalmitoylated cysteine-, serine-, Densitometry image units 0 and lysine-containing peptide (Pam), TLR 4, and TLR 2 Control CLI ligands, respectively, induce TNF-α, IL-6, and IL-1β mRNA expression in gastrocnemius muscle and C2C12 myotubes 0.12 ∗ [64, 109]. Further IL-6 induction has been shown to be 0.1 κ 0.08 mediated via activation of NF- B[62]. Warren et al. [61] TLR 6 demonstrated that TLRs 2, 4, 6, 8, and 9 are upregu- 0.06 lated following freeze-induced skeletal muscle damage. TLR 0.04 endogenous ligands such as HMGB-1 and HSPs have been 0.02

shown to be expressed in skeletal muscle. HMGB-1 has been Densitometry image units 0 reported to induce skeletal muscle damage in inflammatory Control CLI myopathies, and the expression of HSPs is upregulated (b) in skeletal muscle following hypoxia and ATP depletion [110, 111]. We have recently found upregulation of TLRs Figure 2: Representative western blots showing (a) increased TLR 2, TLR 4, and TLR 6 protein expression in gastrocnemius muscle 2, 4, and 6 protein expression in muscle biopsies obtained biopsies obtained from patients with CLI compared to controls. from patients with CLI (Figure 2). This indicates that TLRs (b) Densitometric quantification of TLR 2, 4, and 6 levels in CLI are likely to be involved in the pathophysiology of CLI, muscle. ∗P<0.05 compared to control. The human experiments possibly by contributing to the tissue damage that occurs, were conducted in accordance with the Declaration of Helsinki and provides the rationale to further elucidate the role of (1964), and informed consent was obtained from patients with TLRs in the pathogenesis of skeletal muscle damage in CLI. prior approval from the local ethics committee. Cardiology Research and Practice 9

Skeletal muscle cell necrosis Ischaemia

TLR Endogenous ligands-HMGB-1, Induction and release of HSP 60, 70, hyaluronan proinflammatory cytokine, chemokines, and interferons

TRIF MyD88 Activation of TLR signalling pathway

Nucleus NF-κB AP1 IRF3 IRF2

Viable skeletal muscle cell Figure 3: Proposed pathophysiological mechanism of skeletal muscle damage in CLI. Skeletal muscle ischaemia initiates muscle cell apoptosis and necrosis leading to the release of endogenous ligands such as HMGB-1. Subsequently TLRs are activated in other viable muscle cells causing signalling through one or more TLR signalling pathways. This may lead to the activation of transcription factors such as NF-κB, AP-1, IRF 3, and 2. The consequent activation of transcription factors leads to the induction and release of proinflammatory cytokines, chemokines, and interferons that propagate the skeletal muscle damage.

11. Conclusion on the role of these two receptors in mediating ischaemic and I/R injury. However, other TLRs such as TLR 1 and 6 It can be concluded that following ischaemia and I/R signif- are promising targets as they are known to heterodimerise icant tissue damage occurs in a number of different organs. with TLR 2 [16]. It would also be essential to identify The intricate mechanisms involved may differ depending on the particular TLR signalling pathways and transcription the type and duration of insult. However it is clear that factors involved in ischaemic skeletal muscle. Further TLR inflammation following immune cell infiltration and ROS antagonists are under clinical development in the treatment generation play a significant role in mediating cytotoxicity. of a number of inflammatory and autoimmune diseases There is ample evidence that TLRs of both immune and [19], and TLR antagonists have also been shown to reduce nonimmune cell origin are upregulated and activated in ischaemia-induced injury. A better understanding of the ischaemia leading to the production of various proinflamma- pathophysiology of CLI with concomitant development of tory cytokines and chemokines. Both the MyD88-dependent TLR antagonists may identify treatment modalities that can and MyD88-independent TLR signalling pathways may be be translated into clinical benefit for patients with CLI. involved in mediating the ischaemic tissue damage, and the dominant signalling cascade used may be organ specific. TLRs have also been implicated in apoptotic cell death which References plays a large part in ischaemia-induced cell damage. Necrotic cell death has also been shown to occur following ischaemia, [1] V. N. Varu, M. E. Hogg, and M. R. Kibbe, “Critical limb and this is important as numerous TLR endogenous ligands ischemia,” Journal of Vascular Surgery, vol. 51, no. 1, pp. 230– 241, 2010. such as HSP60 and HMGB-1 have been shown to be released during this process. This phenomenon may explain [2] V. Bertele,` M. C. Roncaglioni, J. Pangrazzi, and E. Terzian, “Clinical outcome and its predictors in 1560 patients with why TLRs are activated in sterile inflammation during critical leg ischaemia,” European Journal of Vascular and ischaemia and provide an opportunity to manipulate the Endovascular Surgery, vol. 18, no. 5, pp. 401–410, 1999. pathophysiological process in order to reduce cell damage. [3] E. Cieri, M. Lenti, P. De Rango, G. Isernia, A. Marucchini, There is growing evidence to suggest that some of these and P. Cao, “Functional ability in patients with critical destructive processes are involved in the pathophysiology of limb ischaemia is unaffected by successful revascularisation,” skeletal muscle damage in CLI, and TLRs are implicated European Journal of Vascular and Endovascular Surgery, vol. in mediating this damage. However, further studies are 41, no. 2, pp. 256–263, 2011. required to elaborate on the preliminary data including the [4] A.D.Nicoloff,Jr.TaylorL.M.,R.B.McLafferty, G. L. Moneta, identification of the key TLRs involved in mediating skeletal J. M. Porter, and J. O. Menzoian, “Patient recovery after muscle damage in CLI (Figure 3). One can speculate that infrainguinal bypass grafting for limb salvage,” Journal of TLR 2 and TLR 4 may be important as there is extensive data Vascular Surgery, vol. 27, no. 2, pp. 256–266, 1998. 10 Cardiology Research and Practice

[5] J. A. Dormandy, “Two randomised and placebo-controlled pathway,” Nature Immunology, vol. 4, no. 5, pp. 491–496, studies of an oral prostacyclin analogue (Iloprost) in severe 2003. leg ischaemia,” European Journal of Vascular and Endovascular [24] T. Kawai and S. Akira, “TLR signaling,” Seminars in Immunol- Surgery, vol. 20, no. 4, pp. 358–362, 2000. ogy, vol. 19, no. 1, pp. 24–32, 2007. [6] T. V. Arumugam, E. Okun, S. C. Tang, J. Thundyil, S. M. [25] J. E. Jordan, Z. Q. Zhao, and J. Vinten-Johansen, “The role Taylor, and T. M. Woodruff, “Toll-like receptors in ischemia- of neutrophils in myocardial ischemia-reperfusion injury,” reperfusion injury,” Shock, vol. 32, no. 1, pp. 4–16, 2009. Cardiovascular Research, vol. 43, no. 4, pp. 860–878, 1999. [7] S. Akira and K. Takeda, “Toll-like receptor signalling,” Nature [26] G. Yilmaz, T. V. Arumugam, K. Y. Stokes, and D. N. Granger, Reviews Immunology, vol. 4, no. 7, pp. 499–511, 2004. “Role of T lymphocytes and interferon-γ in ischemic stroke,” [8] N. G. Frangogiannis, “Chemokines in ischemia and reperfu- Circulation, vol. 113, no. 17, pp. 2105–2112, 2006. sion,” Thrombosis and Haemostasis, vol. 97, no. 5, pp. 738– [27] J. M. Thurman, “Triggers of inflammation after renal 747, 2007. ischemia/reperfusion,” Clinical Immunology, vol. 123, no. 1, [9] K. Takeda, T. Kaisho, and S. Akira, “Toll-like receptors,” pp. 7–13, 2007. Annual Review of Immunology, vol. 21, pp. 335–376, 2003. [28] P. Boros and J. S. Bromberg, “New cellular and molecular [10] J. S. Hadley, J. E. Wang, L. C. Michaels et al., “Alterations immune pathways in ischemia/reperfusion injury,” American in inflammatory capacity and TLR expression on monocytes Journal of Transplantation, vol. 6, no. 4, pp. 652–658, 2006. and neutrophils after cardiopulmonary bypass,” Shock, vol. [29]C.D.Collard,A.Vakeva,M.A.Morrisseyetal.,“Com- 27, no. 5, pp. 466–473, 2007. plement activation after oxidative stress: role of the lectin [11] S. Gerondakis, R. J. Grumont, and A. Banerjee, “Regulating complement pathway,” American Journal of Pathology, vol. B-cell activation and survival in response to TLR signals,” 156, no. 5, pp. 1549–1556, 2000. Immunology and Cell Biology, vol. 85, no. 6, pp. 471–475, [30] B. R. S. Broughton, D. C. Reutens, and C. G. Sobey, 2007. “Apoptotic mechanisms after cerebral ischemia,” Stroke, vol. [12] J. W. Breslin, M. H. Wu, M. Guo, R. Reynoso, and S. 40, no. 5, pp. e331–e339, 2009. Y. Yuan, “Toll-like receptor 4 contributes to microvascular [31] A. O. Aliprantis, R. B. Yang, D. S. Weiss, P. Godowski, and inflammation and barrier dysfunction in thermal injury,” A. Zychlinsky, “The apoptotic signaling pathway activated Shock, vol. 29, no. 3, pp. 349–355, 2008. by Toll-like receptor-2,” EMBO Journal, vol. 19, no. 13, pp. [13] J. H. Boyd, S. Mathur, Y. Wang, R. M. Bateman, and K. 3325–3336, 2000. R. Walley, “Toll-like receptor stimulation in cardiomyoctes [32] M. Bsibsi, R. Ravid, D. Gveric, and J. M. Van Noort, decreases contractility and initiates an NF-κBdependent “Broad expression of Toll-like receptors in the human central inflammatory response,” Cardiovascular Research, vol. 72, no. nervous system,” Journal of Neuropathology and Experimental 3, pp. 384–393, 2006. Neurology, vol. 61, no. 11, pp. 1013–1021, 2002. [14] R. A. Frost, G. J. Nystrom, and C. H. Lang, “Multiple Toll- [33] C. S. Jack, N. Arbour, J. Manusow et al., “TLR signaling like receptor ligands induce an IL-6 transcriptional response tailors innate immune responses in human microglia and in skeletal myocytes,” American Journal of Physiology— astrocytes,” Journal of Immunology, vol. 175, no. 7, pp. 4320– Regulatory Integrative and Comparative Physiology, vol. 290, 4330, 2005. no. 3, pp. R773–R784, 2006. [34] S. C. Tang, T. V. Arumugam, X. Xu et al., “Pivotal role for [15] S. P. Jong, F. Gamboni-Robertson, Q. He et al., “High neuronal Toll-like receptors in ischemic brain injury and mobility group box 1 protein interacts with multiple Toll-like functional deficits,” Proceedings of the National Academy of receptors,” American Journal of Physiology—Cell Physiology, Sciences of the United States of America, vol. 104, no. 34, pp. vol. 290, no. 3, pp. C917–C924, 2006. 13798–13803, 2007. [16] Y. Nakao, K. Funami, S. Kikkawa et al., “Surface-expressed [35] S. L. Stevens, T. M. P. Ciesielski, B. J. Marsh et al., “Toll-like TLR6 participates in the recognition of diacylated lipopep- receptor 9: a new target of ischemic preconditioning in the tide and peptidoglycan in human cells,” Journal of Immunol- brain,” Journal of Cerebral Blood Flow and Metabolism, vol. ogy, vol. 174, no. 3, pp. 1566–1573, 2005. 28, no. 5, pp. 1040–1047, 2008. [17] M. F. Tsan and B. Gao, “Endogenous ligands of Toll-like [36] U. Kilic, E. Kilic, C. M. Matter, C. L. Bassetti, and D. M. receptors,” Journal of Leukocyte Biology, vol. 76, no. 3, pp. Hermann, “TLR-4 deficiency protects against focal cerebral 514–519, 2004. ischemia and axotomy-induced neurodegeneration,” Neuro- [18] C. Termeer, F. Benedix, J. Sleeman et al., “Oligosaccharides of biology of Disease, vol. 31, no. 1, pp. 33–40, 2008. hyaluronan activate dendritic cells via Toll-like receptor 4,” [37] S. Lehnardt, S. Lehmann, D. Kaul et al., “Toll-like receptor Journal of Experimental Medicine, vol. 195, no. 1, pp. 99–111, 2 mediates CNS injury in focal cerebral ischemia,” Journal of 2002. Neuroimmunology, vol. 190, no. 1-2, pp. 28–33, 2007. [19]H.Kanzler,F.J.Barrat,E.M.Hessel,andR.L.Coffman, [38] C. X. Cao, Q. W. Yang, F. L. Lv, J. Cui, H. B. Fu, and “Therapeutic targeting of innate immunity with Toll-like J. Z. Wang, “Reduced cerebral ischemia-reperfusion injury receptor agonists and antagonists,” Nature Medicine, vol. 13, in Toll-like receptor 4 deficient mice,” Biochemical and no. 5, pp. 552–559, 2007. Biophysical Research Communications, vol. 353, no. 2, pp. [20] S. Akira and K. Takeda, “Toll-like receptor signalling,” Nature 509–514, 2007. Reviews Immunology, vol. 4, no. 7, pp. 499–511, 2004. [39] E. Seki and D. A. Brenner, “Toll-like receptors and adaptor [21] M. Karin and Y. Ben-Neriah, “Phosphorylation meets ubiq- molecules in liver disease: update,” Hepatology, vol. 48, no. 1, uitination: the control of NF-κB activity,” Annual Review of pp. 322–335, 2008. Immunology, vol. 18, pp. 621–663, 2000. [40] J. Zhang, H. Wu, L. Wang, J. Zhang, H. Wang, and Q. Zheng, [22] T. Kawai and S. Akira, “TLR signaling,” Cell Death and “TLR2 mRNA upregulation in ischemic lobes in mouse Differentiation, vol. 13, no. 5, pp. 816–825, 2006. partial hepatic ischemia/reperfusion injury model,” Journal [23] K. A. Fitzgerald, S. M. McWhirter, K. L. Faia et al., “IKKE of Huazhong University of Science and Technology—Medical and TBKI are essential components of the IRF3 signalling Science, vol. 24, no. 2, pp. 144–146, 2004. Cardiology Research and Practice 11

[41]A.Tsung,R.A.Hoffman, K. Izuishi et al., “Hepatic ische- heart,” Journal of Thoracic and Cardiovascular Surgery, vol. mia/reperfusion injury involves functional TLR4 signaling in 128, no. 2, pp. 170–179, 2004. nonparenchymal cells,” Journal of Immunology, vol. 175, no. [58] S. C. Kim, A. Ghanem, H. Stapel et al., “Toll-like receptor 11, pp. 7661–7668, 2005. 4 deficiency: smaller infarcts, but nogain in function,” BMC [42] Z. M. Bamboat, V. P. Balachandran, L. M. Ocuin, H. Obaid, Physiology, vol. 7, article 5, 2007. G. Plitas, and R. P. DeMatteo, “Toll-like receptor 9 inhibition [59] A. Shimamoto, A. J. Chong, M. Yada et al., “Inhibition of toll- confers protection from liver ischemia-reperfusion injury,” like receptor 4 with eritoran attenuates myocardial ischemia- Hepatology, vol. 51, no. 2, pp. 621–632, 2010. reperfusion injury,” Circulation, vol. 114, no. 1, pp. I270– [43] X.-D. Shen, B. Ke, Y. Zhai et al., “Toll-like receptor and I274, 2006. heme oxygenase-1 signaling in hepatic ischemia/reperfusion [60] S. Frantz, J. Tillmanns, P. J. Kuhlencordt et al., “Tissue- injury,” American Journal of Transplantation,vol.5,no.8,pp. specific effects of the nuclear factor κB subunit p50 on my- 1793–1800, 2005. ocardial ischemia-reperfusion injury,” American Journal of [44] T. Okaya, J. Blanchard, R. Schuster et al., “Age-dependent Pathology, vol. 171, no. 2, pp. 507–512, 2007. responses to hepatic ischemia/reperfusion injury,” Shock, vol. [61] G. L. Warren, T. Hulderman, A. Liston, and P. P. Simeonova, 24, no. 5, pp. 421–427, 2005. “Toll-like and adenosine receptor expression in injured [45] A. Tsung, R. Sahai, H. Tanaka et al., “The nuclear factor skeletal muscle,” Muscle and Nerve, vol. 44, no. 1, pp. 85–92, HMGB1 mediates hepatic injury after murine liver ischemia- 2011. ff reperfusion,” Journal of Experimental Medicine, vol. 201, no. [62] M. Visser, M. Pahor, D. R. Taa e et al., “Relationship of α 7, pp. 1135–1143, 2005. interleukin-6 and tumor necrosis factor- with muscle mass [46] P. S. Patole, R. D. Pawar, M. Lech et al., “Expression and and muscle strength in elderly men and women: the health regulation of Toll-like receptors in lupus-like immune com- ABC study,” Journals of Gerontology—Series A Biological plex glomerulonephritis of MRL-Fas(lpr) mice,” Nephrology Sciences and Medical Sciences, vol. 57, no. 5, pp. M326–M332, Dialysis Transplantation, vol. 21, no. 11, pp. 3062–3073, 2006. 2002. [63] K. A. Meadows, J. M. P. Holly, and C. E. H. Stewart, “Tumor [47] T. G. A. M. Wolfs, W. A. Buurman, A. Van Schadewijk et α al., “In vivo expression of Toll-like receptor 2 and 4 by renal necrosis factor- -induced apoptosis is associated with sup- γ α pression of insulin-like growth factor binding protein- epithelial cells: IFN- and TNF- mediated up-regulation ff during inflammation,” Journal of Immunology, vol. 168, no. 5 secretion in di erentiating murine skeletal myoblasts,” 3, pp. 1286–1293, 2002. Journal of Cellular Physiology, vol. 183, no. 3, pp. 330–337, 2000. [48] B. SchroppelandJ.C.He,“ExpressionofToll-likereceptors¨ [64] R. A. Frost, G. J. Nystrom, and C. H. Lang, “Multiple Toll- in the kidney: their potential role beyond infection,” Kidney like receptor ligands induce an IL-6 transcriptional response International, vol. 69, no. 5, pp. 785–787, 2006. in skeletal myocytes,” American Journal of Physiology— [49] H. Wu, G. Chen, K. R. Wyburn et al., “TLR4 activation Regulatory Integrative and Comparative Physiology, vol. 290, mediates kidney ischemia/reperfusion injury,” Journal of no. 3, pp. R773–R784, 2006. Clinical Investigation, vol. 117, no. 10, pp. 2847–2859, 2007. [65] H. Patel, X. Shiwen et al., “Toll-like receptor 2 antagonism [50] K. Rusai, D. Sollinger, M. Baumann et al., “Toll-like receptors reduces inflammatory cytokine release in skeletal muscle in 2 and 4 in renal ischemia/reperfusion injury,” Pediatric vitro,” British Journal of Surgery, vol. 98, no. S3, p. 209, 2011. Nephrology, vol. 25, no. 5, pp. 853–860, 2010. [66] J. M. van Noort and M. Bsibsi, “Toll-like receptors in [51] J. C. Leemans, G. Stokman, N. Claessen et al., “Renal- the CNS: implications for neurodegeneration and repair,” associated TLR2 mediates ischemia/reperfusion injury in the Progress in Brain Research, vol. 175, pp. 139–148, 2009. kidney,” Journal of Clinical Investigation, vol. 115, no. 10, pp. [67]J.R.Caso,J.M.Pradillo,O.Hurtado,P.Lorenzo,M.A. 2894–2903, 2005. Moro, and I. Lizasoain, “Toll-like receptor 4 is involved in [52] W. P. Pulskens, G. J. Teske, L. M. Butter et al., “Toll-like brain damage and inflammation after experimental stroke,” receptor-4 coordinates the innate immune response of the Circulation, vol. 115, no. 12, pp. 1599–1608, 2007. kidney to renal ischemia/reperfusion injury,” PLoS ONE, vol. [68] Y. Shibayama, S. Asaka, and A. Nishijima, “Mechanism of 3, no. 10, Article ID e3596, 2008. liver injury following ischemia,” Experimental and Molecular [53] B. Kruger,¨ S. Krick, N. Dhillon et al., “Donor toll-like Pathology, vol. 55, no. 3, pp. 251–260, 1991. receptor 4 contributes to ischemia and reperfusion injury [69] E. E. Montalvo-Jave, T. Escalante-Tattersfield, J. A. Ortega- following human kidney transplantation,” Proceedings of the Salgado, E. Pina,˜ and D. A. Geller, “Factors in the patho- National Academy of Sciences of the United States of America, physiology of the liver ischemia-reperfusion injury,” Journal vol. 106, no. 9, pp. 3390–3395, 2009. of Surgical Research, vol. 147, no. 1, pp. 153–159, 2008. [54] Y. J. Woo, M. D. Taylor, J. E. Cohen et al., “Ethyl pyruvate [70] H. Tsukamoto, “Redox regulation of cytokine expression in preserves cardiac function and attenuates oxidative injury Kupffer cells,” Antioxidants and Redox Signaling, vol. 4, no. 5, after prolonged myocardial ischemia,” Journal of Thoracic and pp. 741–748, 2002. Cardiovascular Surgery, vol. 127, no. 5, pp. 1262–1269, 2004. [71]H.Jaeschke,A.Farhood,A.P.Bautista,Z.Spolarics,andJ.J. [55] D. de Kleijn and G. Pasterkamp, “Toll-like receptors in Spitzer, “Complement activates Kupffer cells and neutrophils cardiovascular diseases,” Cardiovascular Research, vol. 60, no. during reperfusion after hepatic ischemia,” American Journal 1, pp. 58–67, 2003. of Physiology—Gastrointestinal and Liver Physiology, vol. 264, [56] S. Frantz, L. Kobzik, Y. D. Kim et al., “Toll4 (TLR4) expres- no. 4, part 1, pp. G801–G809, 1993. sion in cardiac myocytes in normal and failing myocardium,” [72] R. Anaya-Prado, L. H. Toledo-Pereyra, A. B. Lentsch, and P. Journal of Clinical Investigation, vol. 104, no. 3, pp. 271–280, A. Ward, “Ischemia/reperfusion injury,” Journal of Surgical 1999. Research, vol. 105, no. 2, pp. 248–258, 2002. [57] A. J. Chong, A. Shimamoto, C. R. Hampton et al., “Toll- [73] Y. Li, W. Zhang, L. L. Mantell, J. A. Kazzaz, A. M. Fein, and like receptor 4 mediates ischemia/reperfusion injury of the S. Horowitz, “Nuclear factor-κBisactivatedbyhyperoxia 12 Cardiology Research and Practice

but does not protect from cell death,” Journal of Biological therapeutic options,” International Journal of Cardiology, vol. Chemistry, vol. 272, no. 33, pp. 20646–20649, 1997. 58, no. 2, pp. 95–117, 1997. [74] A. Wullaert, G. van Loo, K. Heyninck, and R. Beyaert, [89] L. M. Buja, “Myocardial ischemia and reperfusion injury,” “Hepatic tumor necrosis factor signaling and nuclear factor- Cardiovascular Pathology, vol. 14, no. 4, pp. 170–175, 2005. κB: effects on liver homeostasis and beyond,” Endocrine [90] H. Otani, “Ischemic preconditioning: from molecular mech- Reviews, vol. 28, no. 4, pp. 365–386, 2007. anisms to therapeutic opportunities,” Antioxidants and Redox [75] H. A. Rudiger¨ and P. Clavien, “Tumor necrosis factor α, Signaling, vol. 10, no. 2, pp. 207–247, 2008. but not Fas, mediates hepatocellular apoptosis in the murine [91] H. Kin, N. P. Wang, J. Mykytenko et al., “Inhibition of ischemic liver,” Gastroenterology, vol. 122, no. 1, pp. 202–210, myocardial apoptosis by postconditioning is associated with 2002. attenuation of oxidative stress-mediated nuclear factor-κB [76]J.S.Gujral,T.J.Bucci,A.Farhood,andH.Jaeschke, translocation and TNFα release,” Shock,vol.29,no.6,pp. “Mechanism of cell death during warm hepatic ischemia- 761–768, 2008. reperfusion in rats: apoptosis or necrosis?” Hepatology, vol. [92] I. Ikonomidis, F. Andreotti, E. Economou, C. Stefanadis, P. 33, no. 2, pp. 397–405, 2001. Toutouzas, and P. Nihoyannopoulos, “Increased proinflam- [77]T.V.Arumugam,I.A.Shiels,T.M.Woodruff,D.N.Granger, matory cytokines in patients with chronic stable angina and and S. M. Taylor, “The role of the complement system in their reduction by aspirin,” Circulation, vol. 100, no. 8, pp. ischemia-reperfusion injury,” Shock, vol. 21, no. 5, pp. 401– 793–798, 1999. 409, 2004. [93] H. A. Hennein, H. Ebba, J. L. Rodriguez et al., “Relationship [78] E. Galloway, T. Shin, N. Huber et al., “Activation of hep- of the proinflammatory cytokines to myocardial ischemia atocytes by extracellular heat shock protein 72,” American and dysfunction after uncomplicated coronary revascular- Journal of Physiology—Cell Physiology, vol. 295, no. 2, pp. ization,” Journal of Thoracic and Cardiovascular Surgery, vol. C514–C520, 2008. 108, no. 4, pp. 626–635, 1994. [79] A. B. Lentsch, H. Yoshidome, W. G. Cheadle, F. N. Miller, [94]S.Frantz,R.A.Kelly,andT.Bourcier,“RoleofTLR-2inthe and M. J. Edwards, “Chemokine involvement in hepatic activation of nuclear factor κB by oxidative stress in cardiac ischemia/reperfusion injury in mice: roles for macrophage myocytes,” Journal of Biological Chemistry, vol. 276, no. 7, pp. inflammatory protein-2 and KC,” Hepatology, vol. 27, no. 4, 5197–5203, 2001. p. 1172, 1998. [95] J. I. Oyama, C. Blais Jr., X. Liu et al., “Reduced myocardial [80] Y. Zhai, X. D. Shen, R. O’Connell et al., “Cutting edge: TLR4 ischemia-reperfusion injury in Toll-Like receptor 4-deficient activation mediates liver ischemia/reperfusion inflammatory mice,” Circulation, vol. 109, no. 6, pp. 784–789, 2004. response via IFN regulatory factor 3-dependent MyD88- [96] F. Hua, T. Ha, J. Ma et al., “Protection against myocar- independent pathway,” Journal of Immunology, vol. 173, no. dial ischemia/reperfusion injury in TLR4-deficient mice is 12, pp. 7115–7119, 2004. mediated through a phosphoinositide 3-kinase-dependent [81] W. Lieberthal, S. A. Menza, and J. S. Levine, “Graded ATP mechanism,” Journal of Immunology, vol. 178, no. 11, pp. depletion can cause necrosis or apoptosis of cultured mouse 7317–7324, 2007. proximal tubular cells,” American Journal of Physiology— [97] T. Shishido, N. Nozaki, S. Yamaguchi et al., “Toll-Like Renal Physiology, vol. 274, no. 2, part 2, pp. F315–F327, 1998. receptor-2 modulates ventricular remodeling after myocar- [82] M. Takada, K. C. Nadeau, G. D. Shaw, K. A. Marquette, and dial infarction,” Circulation, vol. 108, no. 23, pp. 2905–2910, N. L. Tilney, “The cytokine-adhesion molecule cascade in 2003. ischemia/reperfusion injury of the rat kidney. Inhibition by [98] J. Favre, P. Musette, V. Douin-Echinard et al., “Toll-like a soluble P-selectin ligand,” Journal of Clinical Investigation, receptors 2-deficient mice are protected against postischemic vol. 99, no. 11, pp. 2682–2690, 1997. coronary endothelial dysfunction,” Arteriosclerosis, Thrombo- [83] J. M. Thurman, D. Ljubanovic, C. L. Edelstein, G. S. Gilkeson, sis, and Vascular Biology, vol. 27, no. 5, pp. 1064–1071, 2007. and V. M. Holers, “Lack of a functional alternative com- [99] M. Andrassy, H. C. Volz, J. C. Igwe et al., “High-mobility plement pathway ameliorates ischemic acute renal failure in group box-1 in ischemia-reperfusion injury of the heart,” mice,” Journal of Immunology, vol. 170, no. 3, pp. 1517–1523, Circulation, vol. 117, no. 25, pp. 3216–3226, 2008. 2003. [100] T. O. Cheng, “All teas are not created equal: the Chinese [84] K. Singbartl, S. A. Green, and L. Klaus, “Blocking P-select- green tea and cardiovascular health,” International Journal of in protects from ischemia/reperfusion-induced acute renal Cardiology, vol. 108, no. 3, pp. 301–308, 2006. failure,” FASEB Journal, vol. 14, no. 1, pp. 48–54, 2000. [101] P. Wang, “Andrenomedullin and cardiovascular responses in [85] M. J. Burne-Taney, D. B. Ascon, F. Daniels, L. Racusen, W. sepsis,” Peptides, vol. 22, no. 11, pp. 1835–1840, 2001. Baldwin, and H. Rabb, “B cell deficiency confers protection [102] M. Belkin, R. D. Brown, J. G. Wright, W. W. LaMorte, and from renal ischemia reperfusion injury,” Journal of Immunol- R. W. Hobson, “A new quantitative spectrophotometric assay ogy, vol. 171, no. 6, pp. 3210–3215, 2003. of ischemia-reperfusion injury in skeletal muscle,” American [86]D.K.Ysebaert,K.E.DeGreef,A.DeBeufetal.,“Tcells Journal of Surgery, vol. 156, no. 2, pp. 83–86, 1988. as mediators in renal ischemia/reperfusion injury,” Kidney [103] I. I. Pipinos, A. R. Judge, J. T. Selsby et al., “The myopathy of International, vol. 66, no. 2, pp. 491–496, 2004. peripheral arterial occlusive disease: part 2. Oxidative stress, [87] A. A. Shigeoka, T. D. Holscher, A. J. King et al., “TLR2 is neuropathy, and shift in muscle fiber type,” Vascular and constitutively expressed within the kidney and participates in Endovascular Surgery, vol. 42, no. 2, pp. 101–112, 2008. ischemic renal injury through both myD88-dependent and - [104] G. Hayes, S. Liauw, A. Romaschin, and P. M. Walker, independent pathways,” Journal of Immunology, vol. 178, no. “Separation of reperfusion injury from ischemia-induced 10, pp. 6252–6258, 2007. necrosis,” Surgical Forum, vol. 39, pp. 306–308, 1988. [88] S. R. J. Maxwell and G. Y. H. Lip, “Reperfusion injury: a [105] R. G. Mitchell, B. D. Duscha, J. L. Robbins et al., “Increased review of the pathophysiology, clinical manifestations and levels of apoptosis in gastrocnemius skeletal muscle in Cardiology Research and Practice 13

patients with peripheral arterial disease,” Vascular Medicine, vol. 12, no. 4, pp. 285–290, 2007. [106] S. S. Signorelli, M. C. Mazzarino, L. Di Pino et al., “High circulating levels of cytokines (IL-6 and TNFα), adhesion molecules (VCAM-1 and ICAM-1) and selectins in patients with peripheral arterial disease at rest and after a treadmill test,” Vascular Medicine, vol. 8, no. 1, pp. 15–19, 2003. [107] M. N. Goodman, “Tumor necrosis factor induces skeletal muscle protein breakdown in rats,” American Journal of Physiology—Endocrinology and Metabolism, vol. 260, no. 5, part 1, pp. E727–E730, 1991. [108] M. N. Goodman, “Interleukin-6 induces skeletal muscle protein breakdown in rats,” Proceedings of the Society for Experimental Biology and Medicine, vol. 205, no. 2, pp. 182– 185, 1994. [109] C. H. Lang, C. Silvis, N. Deshpande, G. Nystrom, and R. A. Frost, “Endotoxin stimulates in vivo expression of inflam- matory cytokines tumor necrosis factor alpha, interleukin- 1beta, -6, and high-mobility-group protein-1 in skeletal muscle,” Shock, vol. 19, no. 6, pp. 538–546, 2003. [110] C. Grundtman, J. Bruton, T. Yamada et al., “Effects of HMGB1 on in vitro responses of isolated muscle fibers and functional aspects in skeletal muscles of idiopathic inflammatory myopathies,” FASEB Journal,vol.24,no.2,pp. 570–578, 2010. [111] Y. Liu and J. M. Steinacker, “Changes in skeletal muscle heat shock proteins: pathological significance,” Frontiers in Bioscience, vol. 6, pp. D12–D25, 2001. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 852829, 8 pages doi:10.1155/2012/852829

Review Article Genomic Research to Identify Novel Pathways in the Development of Abdominal Aortic Aneurysm

Seamus C. Harrison,1 Anastasia Z. Kalea,1 Michael V. Holmes,2 Obi Agu,3 and Steve E. Humphries1

1 Centre for Cardiovascular Genetics, BHF Laboratories, the Rayne Building, Department of Medicine, University College London (UCL), 5 University Street, London WC1E 6JF, UK 2 Genetic Epidemiology Group, Department of Epidemiology and Public Health, University College London, 1-19 Torrington Place, London WC1E 7HB, UK 3 Department of Vascular Surgery, University College London Hospital, London, NW1 2BU, UK

Correspondence should be addressed to Seamus C. Harrison, [email protected]

Received 29 July 2011; Accepted 27 October 2011

Academic Editor: Janice Tsui

Copyright © 2012 Seamus C. Harrison et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abdominal aortic aneurysm (AAA) is a common disease with a large heritable component. There is a need to improve our understanding of AAA pathogenesis in order to develop novel treatment paradigms. Genome wide association studies have revolutionized research into the genetic variants that underpin the development of many complex diseases including AAA. This article reviews the progress that has been made to date in this regard, including mechanisms by which loci identified by GWAS may contribute to the development of AAA. It also highlights potential post-GWAS analytical strategies to improve our understanding of the disease further.

1. Introduction architecture of the condition may provide a blueprint for uncovering novel pathobiological pathways and targets for Abdominal aortic aneurysm (AAA) is a common, late onset nonsurgical treatments. disease which, left untreated, can rupture with a high resul- The role that genetic factors play in the development tant mortality. Approximately 5% of Caucasian males aged of AAA has become increasingly prominent in recent 65–74 will harbor a AAA [1] and the major risk factors for years following Clifton’s initial observation that the disease the condition include male sex, cigarette smoking, a history appeared to run in families [10]. Family history of AAA is an of cardiovascular disease, and a family history of AAA [2, 3]. established risk factor for the disease, with male first-degree Currently, the best predictor of rupture is maximal aneurysm relatives of probands at approximately fourfold greater risk diameter and surgical repair is indicated in AAA greater than than the general population [11–13]. A twin-study of AAA 5.5 cm [4]. Population screening with abdominal ultrasound has estimated the heritability to be as high as 70% [14], and scans (US) reduces the burden of aneurysm related death familial studies have failed to demonstrate consistent modes [5, 6], but there is a lack of evidence to support any of inheritance, suggesting that it is likely to be a complex pharmacological therapies to attenuate AAA progression disease [13, 15], resulting from a complicated network of and/or rupture. The advent of endovascular aneurysm repair environmental and genetic risk factors. There has been some has reduced short-term perioperative mortality associated progress in discovery of rare monogenic cause of aneurysmal with AAA repair [7] but nationwide audits indicate that disease in thoracic aorta (Table 1) but in common with elective repair still carries a mortality risk in region of 1.5– other complex disorders, deciphering causal genetic variants 7% [8]. In patients deemed unfit for surgical repair ten- in AAA has proved a difficult task. Familial-based linkage year survival is less than 25% [9]. Understanding the genetic studies have identified areas of the genome that are strongly 2 Cardiology Research and Practice

Table 1: Monogenic causes of thoracic aortic diseases. In genomewide association studies (GWASs), a panel of common SNPs (minor allele frequency >5%) capturing Phenotype/syndrome Gene Reference common genetic variation across the entire genome is Marfan syndrome FBN1 [17] compared in groups of cases and controls. This approach Loeys-Doetz—ascending TGFBR1 and TGFBR2 [18] is “hypothesis-free” and therefore not subject to potential aortic aneurysm biases seen in candidate gene studies. Owing to the large Thoracic aortic aneurysm MYH11, ACTA2, SMAD3 [19–21] number of independent test in a single association study, there are many factors to consider when designing a GWAS. Most importantly, this multiple testing strategy results in a large number of potentially false positive associations. associated with the disease, but attempts to refine the signal To adjust for this, stringent criteria for “genome wide have so far been unsuccessful [15, 16]. significance” are applied and replication of findings in independent cohorts is required. A potential consequence 2. Genetic Studies of AAA of this is that many true-positive associations may be lost in the “statistical noise”. A second issue is that with only a 2.1. Candidate Gene Approaches. The “common-disease few exceptions, the effect size of common variants is small. common-variant” hypothesis poses that common complex Carriers of risk alleles are generally at 10–30% increased odds diseases arise from the accumulation of genetic variants, of disease compared to noncarriers. These characteristics each with a modest effect on risk (low penetrance) and necessitate extremely large sample sizes in order to have suf- environmental risk factors [22, 23]. It is this hypothesis that ficient statistical power, with recent publications combining has underpinned the developments of genetic association multiple studying hundreds of thousands of subjects at a studies, whereby the frequency of indexed genetic variants is time [14, 30]. As of June 2011, 951 GWASs have now been compared between cases and controls. published in a wide range of disorders and traits (http://www A number of candidate gene association studies for AAA .genome.gov/gwastudies/). have been published. Review of the literature, however, reveals that many studies were underpowered and gave 2.2. GWASs and AAA. In 2007, the field of genomic research inconsistent results, a problem shared by many other com- was ignited by simultaneous publication of three GWASs plex disorders [24]. Small studies with a low P value obtained of cardiovascular disease [31–33]. Each of the studies by chance have been more readily published than negative demonstrated a strong association between common SNPs findings (so-called publication bias), and the results are often on Chromosome 9p21.3 [34], in a gene desert (an area of the not replicated in larger studies with greater statistical power. genome with no known protein-coding genes). These data Despite this, meta-analysis of candidate gene studies suggests exemplified the power of GWAS, as this locus would not have that single nucleotide polymorphisms (SNPs) in genes of been given priority using a candidate gene approach. The the renin-angiotensin system and folate metabolism are limitations were, however, also highlighted as the functional consistently associated with an increased risk of developing significance of this locus was unclear. It has taken a further AAA (Table 2)[25, 26]. There has been considerable interest 3-4 years to understand the biological mechanisms by which in the role of polymorphisms in the TGF-β superfamily and these variants act, and the translational benefit of these risk of developing AAA as these genes have been causally discoveries is not likely to be realized in the near future. A implicated in aneurysmal disease affecting the thoracic aorta. year following publication of these GWAS, it was reported Baas et al. found association between SNPs in TGF-β receptor that SNPs at this locus were also strongly associated with 1and2(TGFBR1 and TGFBR2) and risk of AAA in a Dutch the presence of AAA [34].TheseSNPsarecommoninthe cohort [27], but these associations were not replicated in population (risk allele frequency ∼45–50%), and individuals two cohorts from New Zealand and Australia [28]. There carry 0, 1, or 2 risk alleles. The risk of developing AAA is have also been studies demonstrating suggestive associations increased by ∼30% per allele carried. The association with between SNPs in Latent Transforming Growth Factor 4 AAA has now been replicated in a number of well-powered (LTBP4) and expansion of AAAs, but again, this finding has case-control studies (Table 3)[35–37]. not been replicated in independent follow-up studies [29]. The first GWAS specifically of AAA was published in ff E orts by the Human HapMap consortium (http:// 2009 and identified association of on SNP on Chr3p12.3 with hapmap.ncbi.nlm.nih.gov/index.html.en/), the SNP consor- AAA [38]. This association did not meet conventional levels tium (http://www.ncbi.nlm.nih.gov/SNP/), and more re- of genomewide significance and has not been replicated in cently the 1000 genome project (http://www.1000genomes independent sample sets [39]. However, in 2010 a larger .org/) have uncovered much of the common variation seen GWAS with greater statistical power reported a novel asso- throughout the human genome. Linkage disequilibrium ciation with sequence variant in DAB2IP on Chr9q33 [40]. (LD: the nonrandom association of alleles at two or more The discovery phase included 1,292 individuals with AAA loci) means that only a fraction of all possible SNPs require (defined as an infrarenal aortic diameter >3 cm) and 30,530 genotyping, in order to impute information on nontyped unscreened controls (a small proportion of whom are likely genetic variation, and chips that simultaneously genotype up to harbor AAA), while follow-up replication studies included to 1 million variants at a time are commercially available. 3,297 cases and 7,451 controls (all cases and controls were Cardiology Research and Practice 3

Table 2: SNPs associated with AAA after meta-analysis of candidate gene studies [25, 26].

Gene/polymorphism Number of studies (total cases/controls) Effect size (OR and 95% CI) Angiotensin type 1 Receptor/A116C (rs5186) 1 study, 3 populations (1226/1712) 1.386 (1.2–1.601) Angiotensin converting Enzyme I/D (rs4646994) 4 (1657/2238) 1.238 (1.12–1.36) Methlyenetetrahydrofolate reductase +677C>T 5 (1086/895) 1.234 (1.020–1.494) Matrix metalloproteinase 9 (MMP9, 1562C>T ) 3 (848/802) 1.09 (1.01–1.18)

Table 3: Association with SNPs in the 9p21 locus with AAA.

Author Cases/Controls SNP OR (P-value) Helgadottir et al. [34] 2836/16732 rs10757278 1.31 (1.2E − 12) Bown et al. [35] 899/815 rs1333049 1.22 (0.004) Thompson et al. [36] 741/1366 rs10757278 1.38 (0.03) of European ancestry). The variant conferred a per allele MTAP (methylthioadenosine phosphorylase, an enzyme that odds ratio for AAA of 1.21, a smaller effect than that seen plays a major role in polyamine metabolism), and IFNA21 with the 9p21 variant. Interestingly, the investigators also (interferon alpha-21). It was also demonstrated that the found an association between this SNP and CHD, venous transcriptional control of the 9p21 enhancers was remodeled thromboembolism, and peripheral arterial disease and the with interferon-γ, providing evidence that genetic variation association with CHD has now been replicated in further is one factor that determines the response to inflammatory independent cohorts [41]. Further GWASs are expected in stimuli within the vasculature [46]. the future [30], and it is possible that meta-analyses of these The SNP in DAB2IP discovered by GWAS also associates datasets will uncover further variants associated with the with coronary artery disease, peripheral arterial disease, disease. venous thromboembolism, and pulmonary embolism but shows no association with any classical CHD risk factors 2.3. Functional Analysis of GWAS Loci to Uncover Novel Pa- [40, 41]. DAB2IP, located on Chromosome 9q33, is a GTPase thobiological Pathways. Initial excitement from three sepa- activating protein thought to play an important role in rate GWAS reporting robust associations between common prostate cancer metastasis [47].ASNPinthisgenehas risk variants on Chr9p21.3 and myocardial infarction was been associated with aggressive prostate cancer [48], while tempered by the fact that the functional significance of in vitro functional studies have demonstrated that loss of the locus was not immediately obvious. The lead SNP the protein leads to enhanced cell proliferation and reduced (or any in close LD with it) does not lie in a protein apoptosis, via the PI3-Akt pathway [49]. DAB2IP expression coding gene. It has, however, been identified that this risk is significantly reduced in AAA tissue compared to tissue variant overlaps with the recently annotated noncoding RNA from healthy controls [50], and this SNP did correlate with (ncRNA), ANRIL. NcRNAs can alter expression of protein reduced expression of the protein in aortic tissue (though coding genes by mechanisms such as gene silencing, DNA this was not reproduced in mammary artery tissue) [40]. methylation, chromatin remodeling, and RNA interference It is possible, therefore, that this variant also promotes [42]. Functional studies of this locus have demonstrated excessive vascular smooth muscle cell (VSMC) proliferation, that carriers of the risk variant have reduced expression of through reduced expression of DAB2IP in aortic tissue. ANRIL, along with other nearby genes such as CDKN2A and Interestingly, DAB2IP expression is modulated by EZH2, CDKN2B [43] which are inhibitors of cellular senescence a histone methyltransferase forms part of the polycomb involved on controlling cellular proliferation and apoptosis. repressor complex, and has been proposed as a potential Jarinova et al. found that the risk locus has enhancer activity drug target in prostate cancer [51, 52]. If, at a molecular in primary human aortic smooth muscle cells and that level, the link between genetic variation at this locus, DAB2IP pathways involved in cellular proliferation were upregulated expression, and vascular disease was uncovered, enzymes in risk allele carriers [44]. Visel et al. then demonstrated such as EZH2 could also be potential novel targets in that targeted deletion of this region in a mouse model leads pharmacological therapies to attenuate AAA formation. to increased expression of the CDKN2A and CDKN2B and Whilst it appears that the two SNPs discovered for AAA that aortic smooth muscle cells from these animals displayed may both be influencing a common disease pathway, there excessive proliferation and diminished senescence [45]. More was no evidence of epistatic interaction between the 9p21 recently, it was shown that the region at 9p21 is densely and DAB2IP SNP, with simply additive effects on AAA risk packed with enhancer sites that are capable of altering [40]. We have found the same with regard to risk of CHD; expression of both neighboring and distant genes by physical approximately 40% of the population who carry 2 or more interaction. Specifically, variants associated with CHD (and risk alleles at these loci have a hazard ratio for myocardial AAA) disrupt binding of the transcription factor STAT1, infarction of 1.7 compared to individuals carrying zero which results in altered expression CDKN2A and CDKN2B, risk alleles [41]. This suggests that accumulation of small 4 Cardiology Research and Practice disturbances in different elements of the VSMC proliferation of expansion have often been small and underpowered, pathway combines to increase the risk of both atherosclerosis with heterogeneity in the cohorts with regard to how the and AAA. phenotype is actually measured and modeled, which is a The small effect sizes seen with GWAS-identified variants major methodological concern. These problems will only do not preclude potential biological importance, as they be overcome by a large-scale collaborative effort to produce may highlight important pathways in disease [53]. For standardized methods for phenotype definitions and data- example, genes highlighted by GWAS of type 2 diabetes collection. mellitus (T2DM) are known targets for thiazolidinediones and sulphonylureas [54], drugs commonly used in this 3.2. Rare Variants/Exome Sequencing. Despite the discovery condition. For AAA, the genomewide data are pointing to of large numbers of variants associated with many complex pathways involved in promoting excessive VSMC prolifera- diseases by GWAS, the majority of observed heritability in tion. Cigarette smoking, a major environmental risk factor most of these diseases remains unexplained [62]. This has for both diseases, leads to increased levels of proliferation in prompted speculation that rare variants of large effect (which VSMCs [48, 55], whilst a role for excessive VSMC prolifera- are poorly covered on currently available gene chips) may tion in aneurysm formation elsewhere in the arterial tree has be important in development of common diseases [62]. been demonstrated—mutations in ACTA2 (smooth muscle Rare causative mutations have been identified by exome- actin alpha 2) and TGFBR2 (transforming growth factor sequencing experiments in a handful of single genedisorders beta receptor 2) promote excessive VSMC proliferation and [63, 64], whilst deep resequencing efforts have identified are causal for thoracic aneurismal disease [19, 56]. Indeed, rare variants of large effect at loci implicated by GWAS evidence from candidate gene studies also suggests a role in traits such as triglyceride levels [65]. Despite these for excessive VSMC proliferation. The Angiotensin II type successes, the whole genome/exome resequencing studies 1 receptor 1166C polymorphism has been associated with for common complex diseases remain limited by expense, AAA in three independent cohorts [26] (per allele odds ratio statistical power, and computational capacity [66]. 1.60, 95% CI 1.32–1.93, P = 1.1 × 10−6), and it has been shown that this polymorphism increased vascular response 3.3. MicroRNAs and AAA. MicroRNAs (miRNAs) are a class to circulating Angiotensin II [57], a potent stimulator of of endogenous noncoding single-stranded RNAs (19–24 VSMC proliferation and migration [58]. nucleotides) that are important regulators of gene expres- sion. miRNAs are transcribed as primary miRNAs (pri- 3. Future Directions for Genomics and miRNAs), processed to precursor miRNAs (pre-miRNAs), Pathobiology of AAA and then to mature miRNAs. It is estimated that more than 60% of protein-coding genes are regulated by these small 3.1. Study Design to Refine and Augment Signals. In genetic RNAs [67, 68], composing a new complicated regulatory studies a useful alternative to dichotomizing complex disor- network with a significant role in biological functions that ders is to consider the clinical end-point as a combination are frequently regulated cooperatively by large numbers of of different continuous traits [59]. Within the population, genes. infrarenal aortic diameter is a continuously distributed The VSMC is crucial to the progression of almost all phenotype (skewed to the right) [60], with AAA rupture vascular wall disorders including AAA [69, 70]. Liu et (the clinical end-point of interest) in aortas less than 4 cm al. studied the expression of miRNAs in an experimental almost unheard of. Rather than dichotomizing into AAA animal model of AAA and discovered a group of miRNAs versus no AAA (cut-off threshold 3 cm), another option as differentially expressed in AAA versus normal Sprague suggested by Plomin [59] would be to study the trait across Dawley rat aortas [71]. Bioinformatics analyses for predicted the range of variation in the population. This strategy has mRNA targets of differentially expressed miRNAs showed been used to great effect in other complex disorders; for enrichment for cell signaling pathways thought to play a role example, following discovery of loci for T2DM, a binary in human AAA development, such as the mitogen-activated outcome, signals have been refined by studying continuous protein kinase pathway. Recently Leeper et al. performed traits associated with the disease such as fasting glucose, an in vitro miRNA microarray analysis of human VSMC to insulin secretion, and obesity. Population-based studies identify 28-upregulated and 3-downregulated miRNAs that provide greater freedom from biases, better definition of were significantly and sustainably altered during the differ- environmental exposures before disease onset, and clearer entiation process [72]. Among the regulatory miRNAs for characterization of the evolution of traits over time [61]. VSMC,miRNA-26awasofparticularinterestasitappeared Another area that has received limited attention in the to serve as an inhibitor of VSMC differentiation by inhibiting literature to date is the discovery of variants that associate the effect on the signaling pathways downstream of the with rapid aneurysm expansion of small AAA. It is not clear TGF-β/BMP superfamily of growth factors. Cells deficient in whether this phenotype has a large heritable component or miRNA-26a lost their migratory phenotype toward a growth whether the genes that predispose to AAA are also those factor/serum gradient and displayed enhanced rates of that predispose to rapid expansion. For example, it does programmed cell death. These effects on the TGF-β pathway not appear that the 9p21 SNP associates with expansion and on VSMC proliferation, migration, and apoptosis in rates [36]. It should be noted, however, that genetic studies particular suggested that miRNA-26a could be important in Cardiology Research and Practice 5

but light is being shed on pathobiological pathways such Pre-miR-155 as those involved in excessive VSMC proliferation, which has potential implications for development of nonsurgical therapies. Further discoveries will rely upon collaboration of large research consortia as seen in other complex diseases and careful consideration of how information from genomewide miR-155 + RISC data could be harnessed to develop specific therapies and individualized preventative strategies.    3 5 3 5 T T  3   Acknowledgments 5 A 5 C 3 AGTR1 AGTR1 S. C. Harrison is supported by a BHF Clinical Training 1166 A 1166 C Fellowship (FS/11/16/28696). A. Z. Kalea is funded by the Translational No translational BHF as a chair scholar. S. E. Hurphries is funded by the repression repression British Heart Foundation RG2008/08. M. V. Holmes is Figure 1: Model of miR-155 binding to the 3UTR of AGTR1 funded by a Population Health Scientist Fellowship from the and association with the 1166A>C polymorphism (rs5186). The Medical Research Council (G0802432). hairpin-shaped precursor of miR-155 (pre-miR-155) after being exported in the cytoplasm is assembled into the RNA-induced References silencing (RISC) complex (mature miR-155), which transfers it to recognise specific mRNA targets. miR-155 binds with complete [1]R.A.P.Scott,H.A.Ashton,M.J.Buxtonetal.,“The sequence complementarity to its target mRNA seed site and induces Multicentre Aneurysm Screening Study (MASS) into the effect posttranscriptional gene silencing. In the presence of the −1166 C  of abdominal aortic aneurysm screening on mortality in men: allele in the 3 UTR of AGTR1, the compensatory base pairing is a randomised controlled trial,” Lancet, vol. 360, no. 9345, pp. ff a ected, and miR-155 binding is inhibited leading to increased 1531–1539, 2002. AGTR1 expression. [2] J. Cornuz, C. S. Pinto, H. Tevaearai, and M. Egger, “Risk factors for asymptomatic abdominal aortic aneurysm: sytem- atic review and meta-analysis of population-based screening AAA development. These results were confirmed using two studies,” European Journal of Public Health,vol.14,no.4,pp. independent animal models of AAA disease, where miRNA- 343–349, 2004. 26a expression coincides temporally with VSMC apoptosis [3] S. H. Forsdahl, K. Singh, S. Solberg, and B. K. Jacobsen, “Risk and cell loss. factors for abdominal aortic aneurysms: a 7-year prospective Another miRNA-related mechanism of diseases suscepti- study: the tromso study, 1994–2001,” Circulation, vol. 119, no. bility is where SNPs alter miRNA target sites. For example, 16, pp. 2202–2208, 2009. the 1166A>C polymorphism (rs5186) in the angiotensin [4] J. T. Powell, A. R. Brady, L. C. Brown et al., “Mortality receptor 1 (AGTR1) that has been associated with AAA results for randomised controlled trial of early elective surgery by candidate gene analysis appears to abolish miR-155- or ultrasonographic surveillance for small abdominal aortic mediated regulation of the AGTR1 gene [73, 74](Figure 1). aneurysms,” Lancet, vol. 352, no. 9141, pp. 1649–1655, 1998. Angiotensin II receptor type 1 (AT1) signaling stimulates [5] J. S. Lindholt and P. Norman, “Screening for Screening for proliferation of VSMC and vascular fibrosis [75], while the abdominal aortic aneurysm reduces overall mortality in men. AT1 receptor blocker Losartan has been shown to reduce A meta-analysis of the mid- and long-term effects of screening experimental aneurysm formation in mouse models of for abdominal aortic aneurysms,” European Journal of Vascular Marfan’s disease [76]. Daugherty et al. showed that selective and Endovascular Surgery, vol. 36, no. 2, pp. 167–171, blockade of AT1 signaling with Losartan attenuated AAA 2008. formation in the Angiotensin II-infused apoE−/− mouse [6] H. Takagi, S.-N. Goto, M. Matsui, H. Manabe, and T. Umem- model, but blockade of AT2 signalling resulted in more severe oto, “A further meta-analysis of population-based screening for abdominal aortic aneurysm,” Journal of Vascular Surgery, atherosclerosis and aneurysmal disease [77]. Taken together, vol. 52, no. 4, pp. 1103–1108, 2010. these data suggest that micro-RNAs are likely to play a role in [7] R. M. Greenhalgh, “Endovascular aneurysm repair versus the remodeling process seen during aneurysm development, open repair in patients with abdominal aortic aneurym (EVAR and further research of genetic variation in both microRNAs trial 1): randomised controlled trial,” Lancet, vol. 365, no. and their targets may uncover some novel insights. 9478, pp. 2179–2186, 2005. [8] http://www.esvs.org/files/About the Society/ESVS VASCU - 4. Conclusions NET REPORT 2008 BW.pdf. [9] R. M. Greenhalgh, D. J. Allison, P. R. F. Bell et al., “Endovas- Whilst GWASs are redefining our understanding of many cular repair of aortic aneurysm in patients physically ineligible complex diseases including AAA, it is clear that they repre- for open repair,” New England Journal of Medicine, vol. 362, no. sent only an early step in the process of genetic discovery. 20, pp. 1872–1880, 2010. It is too early to define specific translational roles for [10] M. A. Clifton, “Familial abdominal aortic aneurysms,” British any of the loci identified so far that associate with AAA, Journal of Surgery, vol. 64, no. 11, pp. 765–766, 1977. 6 Cardiology Research and Practice

[11] S. A. Badger, M. E. O’Donnell, C. S. Boyd et al., “The low European Journal of Human Genetics, vol. 18, no. 2, pp. 240– prevalence of abdominal aortic aneurysm in relatives in north- 244, 2010. ern ireland,” European Journal of Vascular and Endovascular [28] E. Biros, P. E. Norman, G. T. Jones et al., “Meta-analysis Surgery, vol. 34, no. 2, pp. 163–168, 2007. of the association between single nucleotide polymorphisms [12] J. A. Salo, S. Soisalon-Soininen, S. Bondestam, and P. S. in TGF-β receptor genes and abdominal aortic aneurysm,” Mattila, “Familial occurrence of abdominal aortic aneurysm,” Atherosclerosis, vol. 219, no. 1, pp. 218–223, 2011. Annals of Internal Medicine, vol. 130, no. 8, pp. 637–642, 1999. [29] A. R. Thompson, J. A. Cooper, G. T. Jones et al., “Assessment of [13] J. T. Powell and R. M. Greenhalgh, “Multifactorial inheritance the association between genetic polymorphisms in transform- of abdominal aortic aneurysm,” European Journal of Vascular ing growth factor beta, and its binding protein (LTBP), and Surgery, vol. 1, no. 1, pp. 29–31, 1987. the presence, and expansion, of Abdominal Aortic Aneurysm,” [14] C. M. Wahlgren, E. Larsson, P. K. E. Magnusson, R. Hultgren, Atherosclerosis, vol. 209, no. 2, pp. 367–373, 2010. and J. Swedenborg, “Genetic and environmental contributions [30] The Aneurysm Consortium, “Genome Wide Association to abdominal aortic aneurysm development in a twin popula- Studies: identifying the genes that determine the risk of tion,” Journal of Vascular Surgery, vol. 51, no. 1, pp. 3–7, 2010. abdominal aortic aneurysm,” European Journal of Vascular and [15] H. Kuivaniemi, H. Shibamura, C. Arthur et al., “Familial Endovascular Surgery, vol. 36, no. 4, pp. 395–396, 2008. abdominal aortic aneurysms: collection of 233 multiplex [31] R. McPherson, A. Pertsemlidis, N. Kavaslar et al., “A common families,” Journal of Vascular Surgery, vol. 37, no. 2, pp. 340– allele on chromosome 9 associated with coronary heart 345, 2003. disease,” Science, vol. 316, no. 5830, pp. 1488–1491, 2007. [16] A. F. Baas, J. Medic, R. Van’T Slot et al., “Association study [32] N. J. Samani, J. Erdmann, A. S. Hall et al., “Genomewide as- of single nucleotide polymorphisms on chromosome 19q13 sociation analysis of coronary artery disease,” New England with abdominal aortic aneurysm,” Angiology, vol. 61, no. 3, pp. Journal of Medicine, vol. 357, no. 5, pp. 443–453, 2007. 243–247, 2010. [33] A. Helgadottir, G. Thorleifsson, A. Manolescu et al., “A com- [17] B. Lee, M. Godfrey, E. Vitale et al., “Linkage of Marfan mon variant on chromosome 9p21 affects the risk of myocar- syndrome and a phenotypically related disorder to two dial infarction,” Science, vol. 316, no. 5830, pp. 1491–1493, different fibrillin genes,” Nature, vol. 352, no. 6333, pp. 330– 2007. 334, 1991. [34] A. Helgadottir, G. Thorleifsson, K. P. Magnusson et al., [18] B. L. Loeys, J. Chen, E. R. Neptune et al., “A syndrome “The same sequence variant on 9p21 associates with myocar- of altered cardiovascular, craniofacial, neurocognitive and dial infarction, abdominal aortic aneurysm and intracranial skeletal development caused by mutations in TGFBR1 or aneurysm,” Nature Genetics, vol. 40, no. 2, pp. 217–224, 2008. TGFBR2,” Nature Genetics, vol. 37, no. 3, pp. 275–281, 2005. [35] M. J. Bown, P. S. Braund, J. Thompson, N. J. London, [19] D. C. Guo, H. Pannu, V. Tran-Fadulu et al., “Mutations N. J. Samani, and R. D. Sayers, “Association between the in smooth muscle α-actin (ACTA2) lead to thoracic aortic coronary artery disease risk locus on chromosome 9p21.3 aneurysms and dissections,” Nature Genetics, vol. 39, no. 12, and abdominal aortic aneurysm,” Circulation. Cardiovascular pp. 1488–1493, 2007. Genetics, vol. 1, no. 1, pp. 39–42, 2008. [20] L. Zhu, R. Vranckx, P. K. Van Kien et al., “Mutations in myosin [36] A. R. Thompson, J. Golledge, J. A. Cooper, H. Hafez, P. heavy chain 11 cause a syndrome associating thoracic aortic E. Norman, and S. E. Humphries, “Sequence variant on aneurysm/aortic dissection and patent ductus arteriosus,” 9p21 is associated with the presence of abdominal aortic Nature Genetics, vol. 38, no. 3, pp. 343–349, 2006. aneurysm disease but does not have an impact on aneurysmal [21] I. M. B. H. van de Laar, R. A. Oldenburg, G. Pals et al., expansion,” European Journal of Human Genetics, vol. 17, no. “Mutations in SMAD3 cause a syndromic form of aortic 3, pp. 391–394, 2009. aneurysms and dissections with early-onset osteoarthritis,” [37]E.Biros,M.Cooper,L.J.Palmer,P.J.Walker,P.E.Norman, Nature Genetics, vol. 43, no. 2, pp. 121–126, 2011. and J. Golledge, “Association of an allele on chromosome 9 [22] D. E. Reich and E. S. Lander, “On the allelic spectrum of and abdominal aortic aneurysm,” Atherosclerosis, vol. 212, no. human disease,” Trends in Genetics, vol. 17, no. 9, pp. 502–510, 2, pp. 539–542, 2010. 2001. [38] J. R. Elmore, M. A. Obmann, H. Kuivaniemi et al., “Iden- [23] E. S. Lander, “The new genomics: global views of biology,” tification of a genetic variant associated with abdominal Science, vol. 274, no. 5287, pp. 536–539, 1996. aortic aneurysms on chromosome 3p12.3 by genome wide [24] J. P. A. Ioannidis, E. E. Ntzani, T. A. Trikalinos, and D. association,” Journal of Vascular Surgery,vol.49,no.6,pp. G. Contopoulos-Ioannidis, “Replication validity of genetic 1525–1531, 2009. association studies,” Nature Genetics, vol. 29, no. 3, pp. 306– [39] G. T. Jones and A. M. van Rij, “Regarding “Identification of 309, 2001. a genetic variant associated with abdominal aortic aneurysms [25] A. R. Thompson, F. Drenos, H. Hafez, and S. E. Humphries, on chromosome 3p12.3 by genome wide association”,” Journal “Candidate gene association studies in abdominal aortic of Vascular Surgery, vol. 50, no. 5, pp. 1246–1247, 2009. aneurysm disease: a review and meta-analysis,” European [40] S. Gretarsdottir, A. F. Baas, G. Thorleifsson et al., “Genome- Journal of Vascular and Endovascular Surgery,vol.35,no.1, wide association study identifies a sequence variant within pp. 19–30, 2008. the DAB2IP gene conferring susceptibility to abdominal aortic [26] G. T. Jones, A. R. Thompson, F. M. Van Bockxmeer et al., aneurysm,” Nature Genetics, vol. 42, no. 8, pp. 692–697, 2010. “Angiotensin II type 1 receptor 1166C polymorphism is asso- [41] S. C. Harrison, J. A. Cooper, K. Li et al., “Association of ciated with abdominal aortic aneurysm in three independent a sequence variant in dab2ip with coronary heart disease,” cohorts,” Arteriosclerosis, Thrombosis, and Vascular Biology, European Heart Journal. In press. vol. 28, no. 4, pp. 764–770, 2008. [42]P.P.Amaral,M.E.Dinger,T.R.Mercer,andJ.S.Mattick,“The [27]A.F.Baas,J.Medic,R.Van’TSlotetal.,“Associationof eukaryotic genome as an RNA machine,” Science, vol. 319, no. the TGF-β receptor genes with abdominal aortic aneurysm,” 5871, pp. 1787–1789, 2008. Cardiology Research and Practice 7

[43] Y. Liu, H. K. Sanoff, H. Cho et al., “INK4/ARF transcript [59] R. Plomin, C. M. A. Haworth, and O. S. P. Davis, “Common expression is associated with chromosome 9p21 variants disorders are quantitative traits,” Nature Reviews Genetics, vol. linked to atherosclerosis,” PLoS One,vol.4,no.4,ArticleID 10, no. 12, pp. 872–878, 2009. e5027, 2009. [60] P. Norman, M. Le, C. Pearce, and K. Jamrozik, “Infrarenal [44]O.Jarinova,A.F.R.Stewart,R.Robertsetal.,“Functional aortic diameter predicts all-cause mortality,” Arteriosclerosis, analysis of the chromosome 9p21.3 coronary artery disease Thrombosis, and Vascular Biology, vol. 24, no. 7, pp. 1278– risk locus,” Arteriosclerosis, Thrombosis, and Vascular Biology, 1282, 2004. vol. 29, no. 10, pp. 1671–1677, 2009. [61] T. A. Manolio, “Cohort studies and the genetics of complex [45] A. Visel, Y. Zhu, D. May et al., “Targeted deletion of the 9p21 disease,” Nature Genetics, vol. 41, no. 1, pp. 5–6, 2009. non-coding coronary artery disease risk interval in mice,” [62] T. A. Manolio, F. S. Collins, N. J. Cox et al., “Finding the Nature, vol. 464, no. 7287, pp. 409–412, 2010. missing heritability of complex diseases,” Nature, vol. 461, no. [46] O. Harismendy, D. Notani, X. Song et al., “9p21 DNA variants 7265, pp. 747–753, 2009. associated with coronary artery disease impair interferon-γ [63] C. A. Albers, A. Cvejic, R. Favier et al., “Exome sequencing signalling response,” Nature, vol. 470, no. 7333, pp. 264–270, identifies NBEAL2 as the causative gene for gray platelet 2011. syndrome,” Nature Genetics, vol. 43, no. 8, pp. 735–737, 2011. [47] D. Xie, C. Gore, J. Liu et al., “Role of DAB2IP in modulat- ing epithelial-to-mesenchymal transition and prostate cancer [64] A. Hoischen, B. W. M. Van Bon, B. Rodr´ıguez-Santiago et al., metastasis,” Proceedings of the National Academy of Sciences of “De novo nonsense mutations in ASXL1 cause Bohring-Opitz the United States of America, vol. 107, no. 6, pp. 2485–2490, syndrome,” Nature Genetics, vol. 43, no. 8, pp. 729–731, 2011. 2010. [65] C. T. Johansen, J. Wang, M. B. Lanktree et al., “Excess of rare [48]C.S.Carty,M.Huribal,B.U.Marsan,J.J.Ricotta,andM. variants in genes identified by genome-wide association study Dryjski, “Nicotine and its metabolite cotinine are mitogenic of hypertriglyceridemia,” Nature Genetics,vol.42,no.8,pp. for human vascular smooth muscle cells,” Journal of Vascular 684–687, 2010. Surgery, vol. 25, no. 4, pp. 682–688, 1997. [66] V. Bansal, O. Libiger, A. Torkamani, and N. J. Schork, [49] D. Xie, C. Gore, J. Zhou et al., “DAB2IP coordinates both “Statistical analysis strategies for association studies involving PI3K-Akt and ASK1 pathways for cell survival and apoptosis,” rare variants,” Nature Reviews Genetics, vol. 11, no. 11, pp. Proceedings of the National Academy of Sciences of the United 773–785, 2010. States of America, vol. 106, no. 47, pp. 19878–19883, 2009. [67] L. P. Lim, N. C. Lau, P. Garrett-Engele et al., “Microarray [50] G. M. Lenk, G. Tromp, S. Weinsheimer, Z. Gatalica, R. Berguer, analysis shows that some microRNAs downregulate large and H. Kuivaniemi, “Whole genome expression profiling numbers of-target mRNAs,” Nature, vol. 433, no. 7027, pp. reveals a significant role for immune function in human 769–773, 2005. abdominal aortic aneurysms,” BMC genomics, vol. 8, p. 237, [68] R. C. Friedman, K. K. H. Farh, C. B. Burge, and D. P. 2007. Bartel, “Most mammalian mRNAs are conserved targets of [51] J. A. Simon and C. A. Lange, “Roles of the EZH2 histone microRNAs,” Genome Research, vol. 19, no. 1, pp. 92–105, methyltransferase in cancer epigenetics,” Mutation Research, 2009. vol. 647, no. 1-2, pp. 21–29, 2008. [69] R. W. Thompson, S. Liao, and J. A. Curci, “Vascular smooth [52] J. Min, A. Zaslavsky, G. Fedele et al., “An oncogene-tumor muscle cell apoptosis in abdominal aortic aneurysms,” Coro- suppressor cascade drives metastatic prostate cancer by coor- nary Artery Disease, vol. 8, no. 10, pp. 623–631, 1997. dinately activating Ras and nuclear factor-B,” Nature Medicine, [70] A. C. Doran, N. Meller, and C. A. McNamara, “Role of vol. 16, no. 3, pp. 286–294, 2010. smooth muscle cells in the initiation and early progression [53] J. N. Hirschhorn, “Genomewide association studies— of atherosclerosis,” Arteriosclerosis, Thrombosis, and Vascular illuminating biologic pathways,” New England Journal of Biology, vol. 28, no. 5, pp. 812–819, 2008. Medicine, vol. 360, no. 17, pp. 1699–1701, 2009. [71] G. Liu, Y. Huang, X. Lu et al., “Identification and charac- [54] K. L. Mohlke, M. Boehnke, and G. R. Abecasis, “Metabolic teristics of microRNAs with altered expression patterns in a and cardiovascular traits: an abundance of recently identified rat model of abdominal aortic aneurysms,” Tohoku Journal of common genetic variants,” Human molecular genetics, vol. 17, Experimental Medicine, vol. 222, no. 3, pp. 187–193, 2010. no. 2, pp. R102–R108, 2008. [55] G. Di Luozzo, S. Pradhan, A. K. Dhadwal, A. Chen, H. Ueno, [72] N. J. Leeper, A. Raiesdana, Y. Kojima et al., “MicroRNA-26a and B. E. Sumpio, “Nicotine induces mitogen-activated pro- is a novel regulator of vascular smooth muscle cell function,” tein kinase dependent vascular smooth muscle cell migration,” Journal of Cellular Physiology, vol. 226, no. 4, pp. 1035–1043, Atherosclerosis, vol. 178, no. 2, pp. 271–277, 2005. 2011. [56] S. Inamoto, C. S. Kwartler, A. L. Lafont et al., “TGFBR2 [73]M.M.Martin,E.J.Lee,J.A.Buckenberger,T.D.Schmittgen, mutations alter smooth muscle cell phenotype and predispose and T. S. Elton, “MicroRNA-155 regulates human angiotensin to thoracic aortic aneurysms and dissections,” Cardiovascular II type 1 receptor expression in fibroblast,” Journal of Biological Research, vol. 88, no. 3, pp. 520–529, 2010. Chemistry, vol. 281, no. 27, pp. 18277–18284, 2006. [57] P. P. Van Geel, Y. M. Pinto, A. A. Voors et al., “Angiotensin II [74] M. M. Martin, J. A. Buckenberger, J. Jiang et al., “The human type 1 receptor A1166C gene polymorphism is associated with angiotensin II type 1 receptor +1166 A/C polymorphism an increased response to angiotensin II in human arteries,” attenuates microRNA-155 binding,” Journal of Biological Hypertension, vol. 35, no. 3, pp. 717–721, 2000. Chemistry, vol. 282, no. 33, pp. 24262–24269, 2007. [58]P.E.McEwan,G.A.Gray,L.Sherry,D.J.Webb,andC. [75] J. P. Habashi, D. P. Judge, T. M. Holm et al., “Losartan, an J. Kenyon, “Differential effects of angiotensin II on cardiac AT1 antagonist, prevents aortic aneurysm in a mouse model cell proliferation and intramyocardial perivascular fibrosis in of Marfan syndrome,” Science, vol. 312, no. 5770, pp. 117–121, vivo,” Circulation, vol. 98, no. 24, pp. 2765–2773, 1998. 2006. 8 Cardiology Research and Practice

[76] S. A. Jimenez´ and J. Rosenbloom, “Angiotensin II blockade in Marfan’s syndrome,” New England Journal of Medicine, vol. 359, no. 16, pp. 1732–1733, 2008. [77] A. Daugherty, M. W. Manning, and L. A. Cassis, “Antagonism of AT2 receptors augments Angiotensin II-induced abdominal aortic aneurysms and atherosclerosis,” British Journal of Pharmacology, vol. 134, no. 4, pp. 865–870, 2001. Hindawi Publishing Corporation Cardiology Research and Practice Volume 2012, Article ID 521958, 9 pages doi:10.1155/2012/521958

Review Article Modulation of Fibrosis in Systemic Sclerosis by Nitric Oxide and Antioxidants

Audrey Dooley,1 K. Richard Bruckdorfer,2 and David J. Abraham1

1 Centre for Rheumatology and Connective Tissue Disease, University College London Medical School, Royal Free Campus, London NW3 2PF, UK 2 Institute of Structural and Molecular Biology, University College London Medical School, Gower Street, London WC1E 6BT, UK

Correspondence should be addressed to Audrey Dooley, [email protected]

Received 29 July 2011; Accepted 11 September 2011

Academic Editor: Janice Tsui

Copyright © 2012 Audrey Dooley et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Systemic sclerosis (scleroderma: SSc) is a multisystem, connective tissue disease of unknown aetiology characterized by vascular dysfunction, autoimmunity, and enhanced fibroblast activity resulting in fibrosis of the skin, heart, and lungs, and ultimately internal organ failure, and death. One of the most important and early modulators of disease activity is thought to be oxidative stress. Evidence suggests that the free radical nitric oxide (NO), a key mediator of oxidative stress, can profoundly influence the early microvasculopathy, and possibly the ensuing fibrogenic response. Animal models and human studies have also identified dietary antioxidants, such as epigallocatechin-3-gallate (EGCG), to function as a protective system against oxidative stress and fibrosis. Hence, targeting EGCG may prove a possible candidate for therapeutic treatment aimed at reducing both oxidant stress and the fibrotic effects associated with SSc.

1. Introduction: Nitric Oxide and has been demonstrated to act as a signalling molecule in Systemic Sclerosis many other tissues and regulates physiological and cellular processes in a variety of pathologies such as hypertension, The free radical nitric oxide (NO) is an important physio- cancer, diabetes, and male impotence [1]. logical signalling molecule, potent vasodilator, and mediator In the disease scleroderma (SSc: systemic sclerosis), the of oxidative stress. Nitric oxide is synthesised from L-ar- metabolism of NO appears to be profoundly disturbed. ginine by NO synthase (NOS), and three main isoforms of Thereisconsiderableevidenceimplicatingoverproduction NOS have been identified with a constitutive expression in of NO [2–5] and reactive oxygen species (ROS) such .− − neuronal (nNOS or NOS 1), endothelial (eNOS or NOS 3), as superoxide anions (O2 ) and peroxynitrite (ONOO ) and several other cell types including fibroblasts. Further- [3, 4, 6–8] in the pathogenesis of SSc, an often fatal more, an inducible expression (iNOS or NOS 2) in response rheumatic disease of unknown aetiology. Regulation of NO to a variety of inflammatory stimuli is possible [1]. Although by endogenous levels of the NOS inhibitor asymmetric NO is a gas, it is a highly reactive, short-lived, molecule dimethylarginine (ADMA) has also recently been proposed able to rapidly diffuseacrosscellmembranes.Nitricoxide [9]. Major features of SSc are enhanced fibroblast activ- exerts its biological effects by the reaction of NO with a ity, collagen overproduction, autoimmunity, and vascular diverserangeoftargetssuchashaemgroups,ironand dysfunction [10–12]. There are several classified clinical zinc clusters, and cysteine residues. Since the discovery of subgroups, including limited (lSSc) and diffuse (dSSc) NO as a key endothelial-derived vasodilator molecule in cutaneous SSc, which reflect the nature of the disease in cardiovascular physiology and the award of the Nobel Prize their degree of skin sclerosis, immunological profile, and in 1998 to Robert F. Furchgott, Ferid Murad, and Louis microvascular dysfunction [13, 14]. Endothelial activation J. Ignarro, the field of NO research has rapidly expanded and damage are also an early part of this process [14–17]. The to encompass many biomedical areas. Subsequently, NO nature of the factors that induce endothelial dysfunction is 2 Cardiology Research and Practice still unclear; however, there are several serological biomark- matrix, leading to fibrosis. Early studies demonstrated that ers that reflect the vasculopathy of the disease. These include there is reduced antioxidant capacity in SSc; plasma ascorbic the vasoconstrictor endothelin [18], cell adhesion molecules acid (vitamin C), α-tocopherol, β-carotene, and selenium such as selectin [19], anti-endothelial antibodies [20], and were found to be lower in patients than in controls [30, the vasodilator nitric oxide [2, 3, 5]. While indeed an early 32, 38]. Additionally, the beneficial effects of the antioxidant modulator of disease activity is thought to be oxidative stress, probucol in patients with Raynaud’s phenomenon, a vascular the etiology of events and role of NO remain unknown. pathology which is associated with SSc, have been shown Recent reports suggest that the abnormal production of ROS [39]. Subsequently, antioxidant therapy was proposed as a is linked to fibroblast activation by the increased expression possible treatment in SSc [30, 38, 39] and also in other of stimulatory serum autoantibodies to the platelet-derived diseases [40, 41] with the focus being on the antioxidants growth factor receptor in SSc [21]. counteracting the ROS-induced endothelial damage and vasculopathy that occur. More recently, natural antioxidants such as polyphenols from green tea (Camellia sinensis) 2. Fibrosis and Oxidative Stress in extracts, a popular beverage consumed worldwide, have Systemic Sclerosis attracted attention due to their potent antioxidant effects, particularly that of one component, (-)-epigallocatechin-3- In SSc the excessive connective tissue fibrosis is the most gallate (EGCG). In this paper, we will discuss findings about characteristic pathological manifestation of the disease [10]. the mechanisms of NO-mediated modulation of fibrosis The fibrosis is especially prominent in the diffuse cutaneous as well as evidence suggesting that the dietary antioxidant form of SSc, where excessive connective tissue accumulation EGCG could be a therapeutic target for SSc. is due to overproduction of the extracellular matrix by fibroblasts and myofibroblasts, activated by soluble factors such as transforming growth factor beta (TGF-β)[22–26] 3. Nitric Oxide as a Regulator of the and connective tissue growth factor [27]. Myofibroblasts are Fibrogenic Response. adifferentiated and activated form of fibroblast which have been shown to persist in SSc fibroblast cultures and are The versatile free radical NO has been implicated in the pathogenesis of SSc. In most biological situations NO is responsible for increased collagen synthesis and deposition. − − The molecular mechanisms underlying the origin of the largely oxidized to nitrate (NO3 ) and nitrite (NO2 ), myofibroblast are complex; however, they play a crucial role with the measurement of total nitrate and nitrite NO(x) in wound healing and the development of fibrosis [28, 29]. production, as well as ADMA levels, seen as a reflection The fibrotic process is most prominent in the skin, lungs, of endothelial dysfunction in many diseases [1, 9, 42]. heart, gastrointestinal tract, kidney, tendons and ligaments, Early studies involving NO(x) production in SSc have shown and endocrine glands; widespread perivascular fibrosis also conflicting results [2, 3, 5, 43, 44]; however, the discrep- occurs. Fibrotic damage to these affectedorgansaccountsfor ancy in these results could be explained by differences in much of the morbidity and mortality associated with SSc. the degree of inflammatory disorder, disease subset, and The scleroderma phenotype at the cellular level has treatment of the patients. Furthermore, modifications in the also been shown to be characterized by oxidative stress, dietary NO(x) intake were not attempted in many of these an imbalance between the elevated level of ROS and/or studies. Later on other groups [2, 4, 5, 45, 46] demonstrated the impaired function of the antioxidant defence system evidence implicating overproduction of NO, with increased .− − − plasma NO3 /NO2 levels, together with elevated nitration [30]. ROS usually include O2 ,hydrogenperoxide(H2O2), − hydroxyl radicals ( .OH), and ONOO−. The sources of of plasma proteins, a marker of ONOO production [4]. oxidative stress in SSc are complex and interactive and Nitrated proteins [47, 48] are also found in the skin in likely to be due to ischaemic-reperfusion injury, dysregulated SSc associated with sites of inflammation [3, 4]. There is metabolism of the free radical NO [2–5], and generation a growing literature showing strong tissue localization of of ROS by fibroblasts [6] and activated leukocytes such nitrotyrosine staining or increased levels of free and protein- as monocytes via the NADPH oxidase system [31]. ROS bound nitrotyrosine in inflammatory diseases. For example, have been demonstrated to be cell transducers of fibroblast strong staining of nitrotyrosine has been found in lung proliferation [6], collagen-gene expression, and myofibrob- sections from patients with lung injury [49], while significant last phenotype conversion in SSc [6, 8, 21]. During the amounts of 3-nitrotyrosine have been reported from patients 1990s–early 2000s clear evidence for oxidative stress in SSc with rheumatoid arthritis [50] and chronic renal failure emerged by the enhanced oxidation of lipids and lipoproteins [51, 52]. (oxLDL) [32, 33], increased isoprostane production [34– Interestingly, in SSc there is the paradoxical situation 37], and also the presence of modified, nitrated proteins in in which NO production by eNOS in endothelial cells the plasma and skin [4, 7]. In SSc what is less certain is is decreased possibly due to the rapid reaction of NO .− − the exact stage at which increases in free radicals and other and O2 to generate the reactive intermediate ONOO reactive species occur, the potential counteractive role of [53, 54] or due to the presence in the circulation of the antioxidant defence system or how both elements are natural inhibitors of NOS activity such as ADMA [9, 42] linked to the main events of the disease such as the vascular (Figure 1). Therefore, in SSc reduced eNOS expression and abnormalities, and the increased synthesis of extracellular microcirculatory dysfunction are in part contributory to the Cardiology Research and Practice 3

TGFβ PDGF NADPH NOS TNF, IL -1 Matrix Collagen, CTGF, oxidase fibronectin, 3 2 SMA

Ras Caveolae

HO-1 NO ERK iNOS TNF Proteosome AKT degradation eNOS L-Arginine Smad P Hsp90 Ras ↑ ADMA JNK NO CaM

JNK Oxidant stress O − Smad 2 NO N DDAH − ONOO ROS OH 1 − Antioxidant defence O2 ↓ Bioavailable NO enzymes

SOD H2O2 4

+ Nrf2 Smad ARE NFκB AP-1 SP-1

Collagen type I gene expression Figure 1: Schematic diagram depicting the possible pathways in which NO modulates collagen type I gene expression to affect fibrosis. In .− the first hypothesis (1), the rapid reaction between NO and O2 leads to decreased NO bioavailability. NO regulation by ADMA may also occur. NO normally can directly activate transcription factors such as NFκB, SP-1, and AP-1 to inhibit collagen gene expression. The second possibility (2) is that NO normally by activating the protective stress enzyme HO-1 can negatively modulate the NADPH oxidase pathway. In fibrosis, activation of the NADPH oxidase pathway has been shown to increase collagen synthesis and myofibroblast differentiation. The third plausible pathway (3) is that there is signalling crosstalk following TGF-β binding to a receptor. Signal pathways potentially important here include the MAP kinase JNK. This would synergise with the Smad signalling pathway and decrease the activation of downstream TGF- β-dependent genes. Alternatively, NO could enhance the proteasomal degradation of SMAD. In the fourth pathway (4), NO indirectly exerts its effects by modulating oxidative stress through upregulation of antioxidant/redox defence genes such as Nrf2 leading to regulation of the extracellular matrix. associated Raynaud’s Phenomenon that is well described in loss of NO bioactivity, in eNOS knock-out mice, resulted these patients [3, 14–16, 55]. In contrast at inflammatory in prolonged fibrosis [58]. Furthermore, another study also .− sites the formation of NO, by iNOS, and O2 are increased demonstrated that overexpressing eNOS, using transgenic by the presence of inflammatory cells such as macrophages mice, reduced fibrotic content after bleomycin-induced fi- or activated fibroblasts [56]. Immunohistological studies of brosis [59]. In rats, the long-term inhibition of the inducible scleroderma skin also show that, as the disease progresses to form of NOS (iNOS) has also been shown to favour the de- the later fibrotic stages, the production of eNOS is down- velopment of fibrosis [60]. Additional studies assessing NO regulated, while iNOS is upregulated [3]. Furthermore, stud- metabolism in the tight-skin 1 (Tsk-1/+) mouse, which ies suggest that under conditions when NO overproduction is predisposed to SSc and often used as an experimental occurs, S-nitrosylation of the ADMA regulating enzyme animal model for fibrosis, reported that while type I collagen dimethylarginine dimethylaminohydrolase (DDAH) dimin- protein expression was elevated in Tsk-1/+ skin tissue, eNOS ishes DDAH activity, leading to an accumulation of ADMA. protein and gene expressions were reduced compared to Subsequently, NOS inhibition as a type of regulatory feed- wild-type controls [61]. Furthermore, there was decreased back mechanism may result [57](Figure 1). Indeed, there NOS activity in Tsk-1/+ skin tissue [61]. Correspondingly, is further evidence to indicate increased circulatory levels of the protective antioxidant enzyme haemoxygenase-1 (HO- ADMA in the serum of diffuse SSc patients, suggesting an 1) and the associated transcription factor nuclear factor NOS regulatory mechanism later on in the disease [4]. erythroid-2-related factor 2 (Nrf2) showed reduced protein NO has also been reported to act as an antifibrotic and gene expression levels in Tsk-1/+ skin, while there effector in animal models of experimental fibrosis [58, 59]. was also less total antioxidant activity [61]. The findings For example, in a murine model of pulmonary fibrosis, a suggested that there was also abnormal NO metabolism in 4 Cardiology Research and Practice the Tsk-1/+ mouse, particularly in the skin, while expression occur in a cGMP-dependent [69, 76] or independent manner and activity of protective antioxidants were reduced. even though addition of NO could increase intracellular Several studies have gained insights into the intracellular levels of cGMP [75, 82]. Furthermore other investigators mechanisms by which NO might further modulate a fibrotic using rat mesangial cells show that it is possible that the phenotype. Nitric oxide can induce HO-1, which is also suppression of collagen synthesis by NO could involve an induced by a variety of cellular stress, including free-radical- increase in MMP production and activity [80, 83]. Indeed, mediated stresses and oxygen deprivation [62]. HO-1 in NO regulation of prolyl hydroxylase has been postulated turn can also increase oxidative-stress-related transcription in lapine articular chondrocytes [75]. Prolyl hydroxylase factors such as Nrf2. Because of high similarity between the catalyses the formation of 4-hydroxyproline in collagens by Nrf2-binding sequence (NF-E2 motif) and the antioxidant hydroxylation of proline, and the reaction, in particular, response element (ARE) in regulatory regions of sev- requires Fe2+ and ascorbate and generates free radicals, eral phase 2 antioxidant enzymes, Nrf2 is a putative medi- all of which are sensitive to NO [84]. Under-hydroxylated ator of ARE responses. Consequently, other ARE-bearing procollagens do not form stable triple helices at body tem- protective phase 2 antioxidant/redox enzymes such as perature and, thus, remain partially unfolded, where they are glutathione peroxidase, and classical antioxidant enzymes presumably more susceptible to degradation by intracellular such as superoxide dismutase [SOD] and catalase as well collagenases. An additional number of intracellular targets as HO-1 induce Nrf2 gene expression (Figure 1). There for NO have been described [85], including NO regulation is a large body of evidence suggesting that HO-1 is a of NADPH oxidase [86], MAP kinases (e.g., JNK, ERK) [87], cytoprotective enzyme, and its induction in the setting SMADs [71], and transcription factors (e.g., NFκB, AP-1, of increased cellular stresses helps maintain physiological SP-1) [88] which are key signalling pathways of fibroblast homeostasis [63]. Thus, embryonic fibroblasts derived from −/− proliferation and collagen gene expression [6, 8, 25, 89, 90] HO-1 knock-out mice are significantly less resistant to (Figure 1). Taken together these studies strongly suggest a the cytotoxicity induced by H2O2 and paraquat than wild- definitive link between NO expression and modulation of type controls [64, 65]. Conversely, cells overexpressing HO- fibrosis. 1 have been reported to be more resistant to oxidant- induced toxicity than controls [64]. Indeed, other groups which examined Nrf2−/− knock-out mice reported increased 4. Antioxidants, (-)-Epigallocatechin-3-Gallate, bleomycin-induced pulmonary fibrosis [66] and expression and Systemic Sclerosis of extracellular matrix genes such as collagens after hyperoxic exposure [67] when compared to wild-type controls. It has, In systemic sclerosis clear evidence for oxidative stress has .− thus, been suggested that an increased oxidative burden by been shown by increased levels of O2 [6], antibodies against suppression of antioxidant defence mechanisms in Nrf2−/− oxLDL [91], enhanced lipid peroxidation [32, 33], increased mice secondarily triggers regulation of extracellular matrix F2-isoprostanes [35–37], and increased circulatory levels of genes for repair responses [67]. Interestingly, recently studies nitrotyrosine [4, 7]. Additionally, ROS-induced endothelial in SSc have also suggested Nrf2 as a target for antifibrotic damage occurs as well as the underlying vasculopathy therapy [68]. associated with SSc known as Raynaud’s Phenomenon. Other studies suggest that NO influences fibrosis through Raynaud’s Phenomenon is characterised by transient attacks modulatory effects on the TGF-β pathway [69–71], initiation of cold-induced digital ischaemia associated with intense of fibroblast apoptosis and myofibroblast differentiation vasospasm in the fingers [55, 92, 93]. It occurs as a primary [60, 72, 73], and/or neutralization of profibrotic ROS [53, condition and as secondary to SSc. Currently, the underlying 54](Figure 1). Additionally, NO can modulate collagen disorder of Raynaud’s Phenomenon is thought to be related synthesis; however, at present, the mechanisms and signalling to the abnormal regulation of peripheral vascular tone at the pathways of NO-mediated inhibition of collagen are not level of the digital microcirculation [55, 92, 93]. Although clear. NO inhibition of collagen in dermal SSc fibroblasts nearly all patients with SSc exhibit Raynaud’s Phenomenon, has been reported to be by cGMP-independent regulatory it is still uncertain whether their pathogenesis is identical to mechanisms and in part may be due to up-regulation of patients with Raynaud’s Phenomenon alone. Furthermore, matrix metalloproteinase-1 (MMP-1, an essential collage- in SSc, it is not only the digital and peripheral vessels nase involved in collagen degradation) protein and activity that exhibit vasospasm but also the vessels of the internal levels and/or inhibition of prolyl hydroxylase activity (an organs [13]. The pathophysiology underlying the cold- enzyme important in the posttranslational processing of induced vasospasm characteristic of Raynaud’s Phenomenon collagen) [74]. Similarly, further evidence has also confirmed remains confused [55, 92]. Initially, it was suggested that a role for NO in regulation of the extracellular matrix in α2-adrenoceptors accounted for the supersensitivity [94]. other fibrotic diseases and cell types [70, 73, 75–80]. It was Further studies found that α1-adrenoceptors appears to pre- initially discovered in vascular smooth muscle cells [77, 78] dominate in the physiological control of cold-induced digital and in mesangial cells [70, 80]. Recently, this downregulation vasoconstriction whereas both α1-andα2-adrenoceptors has also been found in human dermal [69, 74], intestinal playanequalroleinRaynaud’ssubjects[15]. However, [81], and rat cardiac [76] fibroblast cells. From the view selective antagonism of these receptor subtypes in both of the signalling pathway, it has been shown that NO normal and Raynaud’s subjects did not abolish vasoconstric- downregulation of collagen synthesis in other cell types can tion suggesting that nonadrenergic mechanisms may also Cardiology Research and Practice 5 contribute to this response. Other studies in the dorsal hand tumorigenesis in human and animal skin models [117, 118]. vein [95], digital arteries [96], as well as isolated gluteal Particularly useful will also be animal models of fibrosis subcutaneous resistance arteries [97]ofprimaryRaynaud’s where recently, in the case of bleomycin-induced pulmonary subjects show an impairment of endothelium-dependent fibrosis and carbon tetrachloride-induced hepatic fibrosis, relaxation. EGCG has been promisingly shown to exert anti-fibrotic Although the vasospastic condition in Raynaud’s Phe- effects [119–121]. nomenon may arise as a result of a functional disturbance In summary, it is clear that further studies are needed at the level of the vessel wall, it is possible that circulating to delineate the key NO-mediated signal transduction and factors in the blood that alter the release of endothelial transcription pathways that facilitate type I collagen produc- cell mediators, such as prostacyclin, NO, endothelin, or in- tion and fibrosis in the disease scleroderma. Studies such as crease ROS and oxidative stress may contribute to the dis- these will help define key targets and candidates for therapy. ease [93]. Early clinical trials, however, indicate limited Furthermore the dietary antioxidant EGCG, with its long success in treatment of patients with Raynaud’s Phenomenon history of safe beverage consumption in green tea together secondary to SSc with antioxidants such as α-tocopherol with its demonstrated potent antioxidant capability, is a or vitamin C, which did not decrease urinary markers good candidate for therapeutic treatment targeting oxidative of oxidative stress such as F(2)-isoprostanes nor improved stress and fibrogenesis in patients with SSc. Further clinical microvascular perfusion after cold exposure [98, 99]. More studies, to confirm its efficacy, determine optimal dosage and hopeful has been the use of the potent antioxidant N- duration of use, and treatment indicators are required. acetylcysteine which has been shown to improve the vascular symptoms of Raynaud’s Phenomenon in patients with SSc [100, 101]. The type of antioxidant used, phase of the References disease, and duration of use may be key factors in successful [1] R. Bruckdorfer, “The basics about nitric oxide,” Molecular treatment therapies involving antioxidants in both SSc and Aspects of Medicine, vol. 26, no. 1-2, pp. 3–31, 2005. the associated Raynaud’s Phenomenon. [2] G. N. Andersen, K. Caidahl, E. Kazzam et al., “Correlation Natural antioxidants, such as polyphenols from green between increased nitric oxide production and markers of tea extracts, are now being considered and investigated in endothelial activation in systemic sclerosis: findings with the particular (-)-epigallocatechin-3-gallate (EGCG). Other less soluble adhesion molecules e-selectin, intercellular adhesion active polyphenol constituents of green tea are believed to molecule 1, and vascular cell adhesion molecule 1,” Arthritis be (-)-epigallocatechin (EGC), (-)-epicatechin (EC), and (-)- and Rheumatism, vol. 43, no. 5, pp. 1085–1093, 2000. epicatechin-3-gallate (ECG). The prominent antioxidant ef- [3]S.A.Cotton,A.L.Herrick,M.I.V.Jayson,andA.J.Free- fects of EGCG derive from the phenol rings that can act as mont, “Endothelial expression of nitric oxide synthases and electron traps to scavenge free radicals, inhibit the formation nitrotyrosine in systemic sclerosis skin,” Journal of Pathology, .− − vol. 189, no. 2, pp. 273–278, 1999. of ROS such as O2 and ONOO , and reduce oxidative stress [102, 103]. EGCG has a higher potent antioxidant [4]A.Dooley,B.Gao,N.Bradleyetal.,“Abnormalnitricoxide capability than α-tocopherol or vitamin C [104], and has metabolism in systemic sclerosis: increased levels of nitrated ff proteins and asymmetric dimethylarginine,” Rheumatology, been demonstrated to be an e ective inhibitor of oxidative- vol. 45, no. 6, pp. 676–684, 2006. stress-induced protein tyrosine nitration during isolation of [5] T. Yamamoto, I. Katayama, and K. Nishioka, “Nitric oxide platelets [48, 105]. production and inducible nitric oxide synthase expression in Studies have also shown that EGCG may directly inhibit systemic sclerosis,” Journal of Rheumatology,vol.25,no.2,pp. molecular targets and regulate multiple signal transduction 314–317, 1998. pathways such as MAP kinases (PI3-kinase, ERK) and [6]P.Sambo,S.S.Baroni,M.Luchettietal.,“Oxidativestress transcription factors (Nrf2, NF-κB, AP-1) and/or induce in scleroderma: maintenance of scleroderma fibroblast phe- antioxidant enzymes such as HO-1 [106, 107]. In addition notype by the constitutive up-regulation of reactive oxygen to its antioxidant properties, it has been shown to possess species generation through the NADPH oxidase complex antifibrotic, anticancer, and anti-inflammatory activities reg- pathway,” Arthritis and Rheumatism, vol. 44, no. 11, pp. ulating both TGF-β and PDGF-induced α1(I) collagen, fi- 2653–2664, 2001. bronectin, α-smooth muscle actin (α-SMA), and prolifera- [7] K. Shimizu, F. Ogawa, E. Muroi et al., “Increased serum levels tion in activated human and rat hepatic stellate cells [108– of nitrotyrosine, a marker for peroxynitrite production, in 110], rat pancreatic cells [111, 112], human keloid fibroblasts systemic sclerosis,” Clinical and Experimental Rheumatology, [113], and SSc dermal fibroblasts [114]. Additionally, EGCG vol. 25, no. 2, pp. 281–286, 2007. can counteract TGFβ-induced ROS in human dermal fibrob- [8] S. Svegliati, R. Cancello, P. Sambo et al., “Platelet-derived growth factor and reactive oxygen species (ROS) regulate lasts from healthy controls, SSc patients, and in a dermal ff Ras protein levels in primary human fibroblasts via ERK1/2: fibroblast cell line, indicating its potential e ectiveness as amplification of ROS and Ras in systemic sclerosis fibrob- an antioxidant to reduce oxidant stress in the disease scle- lasts,” Journal of Biological Chemistry, vol. 280, no. 43, pp. roderma [114]. Interestingly, other studies have shown data 36474–36482, 2005. that EGCG can beneficially inhibit ROS through attenuating [9]R.H.Boger,P.Vallance,andJ.P.Cooke,“Asymmetricdim-¨ NADPH oxidase expression [115, 116]. Furthermore, topical ethylarginine (ADMA): a key regulator of nitric oxide syn- administration of EGCG has successfully been shown to thase,” Atherosclerosis Supplements, vol. 4, no. 4, pp. 1–3, inhibit ultraviolet radiation-induced oxidative stress and 2003. 6 Cardiology Research and Practice

[10] C. P. Denton, C. M. Black, and D. J. Abraham, “Mechanisms mouse model of scleroderma,” American Journal of Pathology, and consequences of fibrosis in systemic sclerosis,” Nature vol. 165, no. 1, pp. 203–217, 2004. Clinical Practice Rheumatology, vol. 2, no. 3, pp. 134–144, [25] A. Leask and D. J. Abraham, “TGF-β signaling and the fi- 2006. brotic response,” FASEB Journal, vol. 18, no. 7, pp. 816–827, [11] C. T. Derk and S. A. Jimenez, “Systemic sclerosis: current 2004. views of its pathogenesis,” Autoimmunity Reviews, vol. 2, no. [26] S. Liu, X. Shi-Wen, L. Kennedy et al., “FAK is required for 4, pp. 181–191, 2003. TGFβ-induced JNK phosphorylation in fibroblasts: impli- [12] J. Varga and D. Abraham, “Systemic sclerosis: a prototypic cations for acquisition of a matrix-remodeling phenotype,” multisystem fibrotic disorder,” Journal of Clinical Investiga- Molecular Biology of the Cell, vol. 18, no. 6, pp. 2169–2178, tion, vol. 117, no. 3, pp. 557–567, 2007. 2007. [13] C. M. Black and C. P. Denton, “Scleroderma and related [27] C. Fonseca, G. E. Lindahl, M. Ponticos et al., “A polymor- disorders in adults and children,” in Oxford Textbook of phism in the CTGF promoter region associated with systemic Rheumatology, P. J. Maddison, D. A. Isenberg, P. Woo, and D. sclerosis,” New England Journal of Medicine, vol. 357, no. 12, N. Glass, Eds., pp. 1217–1247, Oxford University Press, New pp. 1210–1220, 2007. York, NY, USA, 2nd edition, 1998. [28] B. Hinz, S. H. Phan, V. J. Thannickal, A. Galli, M. L. [14]R.J.Prescott,A.J.Freemont,C.J.P.Jones,J.Hoyland,andP. Bochaton-Piallat, and G. Gabbiani, “The myofibroblast: one Fielding, “Sequential dermal microvascular and perivascular function, multiple origins,” American Journal of Pathology, changes in the development of scleroderma,” Journal of vol. 170, no. 6, pp. 1807–1816, 2007. Pathology, vol. 166, no. 3, pp. 255–263, 1992. [29] D. J. Abraham, B. Eckes, V. Rajkumar, and T. Krieg, “New [15] R. R. Freedman, R. Girgis, and M. D. Mayes, “Endothelial developments in fibroblast and myofibroblast biology: impli- and adrenergic dysfunction in Raynaud’s phenomenon and cations for fibrosis and scleroderma,” Current Rheumatology scleroderma,” Journal of Rheumatology, vol. 26, no. 11, pp. Reports, vol. 9, no. 2, pp. 136–143, 2007. 2386–2388, 1999. [30] A. L. Herrick and M. Matucci Cerinic, “The emerging problem of oxidative stress and the role of antioxidants in [16] A. J. Freemont, J. Hoyland, P. Fielding, N. Hodson, and M. systemic sclerosis,” Clinical and Experimental Rheumatology, I. V. Jayson, “Studies of the microvascular endothelium in vol. 19, no. 1, pp. 4–8, 2001. uninvolved skin of patients with systemic sclerosis: direct evidence for a generalized microangiopathy,” British Journal [31] P. Sambo, L. Jannino, M. Candela et al., “Monocytes of of Dermatology, vol. 126, no. 6, pp. 561–568, 1992. patients with systemic sclerosis (scleroderma) spontaneously release in vitro increased amounts of superoxide anion,” [17] R. Sgonc, M. S. Gruschwitz, G. Boeck, N. Sepp, J. Gruber, Journal of Investigative Dermatology, vol. 112, no. 1, pp. 78– and G. Wick, “Endothelial cell apoptosis in systemic sclerosis 84, 1999. is induced by antibody-dependent cell-mediated cytotoxicity [32] K. R. Bruckdorfer, J. B. Hillary, T. Bunce, R. Vancheeswaran, via CD95,” Arthritis and Rheumatism, vol. 43, no. 11, pp. and C. M. Black, “Increased susceptibility to oxidation of 2550–2562, 2000. low-density lipoproteins isolated from patients with systemic [18] M. B. Kahaleh, “Endothelin, an endothelial-dependent vaso- sclerosis,” Arthritis and Rheumatism, vol. 38, no. 8, pp. 1060– constrictor in scleroderma: enhanced production and profi- 1067, 1995. brotic action,” Arthritis and Rheumatism,vol.34,no.8,pp. [33] M. Tikly, K. Channa, P. Theodorou, and M. Gulumian, 978–983, 1991. “Lipid peroxidation and trace elements in systemic sclerosis,” ff [19]C.P.Denton,M.C.M.Bickersta , X. Shiwen et al., “Serial Clinical Rheumatology, vol. 25, no. 3, pp. 320–324, 2006. circulating adhesion molecule levels reflect disease severity in [34] L. Casciola-Rosen, F. Wigley, and A. Rosen, “Scleroderma systemic sclerosis,” British Journal of Rheumatology, vol. 34, autoantigens are uniquely fragmented by metal-catalyzed no. 11, pp. 1048–1054, 1995. oxidation reactions: implications for pathogenesis,” Journal [20] A. Pignone, C. Scaletti, M. Matucci-Cerinic et al., “Anti- of Experimental Medicine, vol. 185, no. 1, pp. 71–79, 1997. endothelial cell antibodies in systemic sclerosis: significant [35] J. L. Cracowski, P. H. Carpentier, B. Imbert et al., “Increased association with vascular involvement and alveolo-capillary urinary F2-isoprostanes in systemic sclerosis, but not in impairment,” Clinical and Experimental Rheumatology, vol. primary Raynaud’s phenomenon: effect of cold exposure,” 16, no. 5, pp. 527–532, 1998. Arthritis and Rheumatism, vol. 46, no. 5, pp. 1319–1323, [21] S. S. Baroni, M. Santillo, F. Bevilacqua et al., “Stimulatory 2002. autoantibodies to the PDGF receptor in systemic sclerosis,” [36] C. M. Stein, S. B. Tanner, J. A. Awad, L. J. Roberts, and New England Journal of Medicine, vol. 354, no. 25, pp. 2667– J. D. Morrow, “Evidence of free radical-mediated injury 2676, 2006. (isoprostane overproduction) in scleroderma,” Arthritis and [22]Y.Chen,S.W.Xu,J.vanBeeketal.,“Matrixcontraction Rheumatism, vol. 39, no. 7, pp. 1146–1150, 1996. by dermal fibroblasts requires transforming growth factor- [37] F. Ogawa, K. Shimizu, E. Muroi et al., “Serum levels of β/activin-linked kinase 5, heparan sulfate-containing proteo- 8-isoprostane, a marker of oxidative stress, are elevated in glycans, and MEK/ERK: insights into pathological scarring in patients with systemic sclerosis,” Rheumatology, vol. 45, no. chronic fibrotic disease,” American Journal of Pathology, vol. 7, pp. 815–818, 2006. 167, no. 6, pp. 1699–1711, 2005. [38] S. Gabriele, P. Alberto, G. Sergio, F. Fernanda, and M. C. [23] Y. Chen, S. W. Xu, M. Eastwood et al., “Contribution of Marco, “Emerging potentials for an antioxidant therapy as activin receptor-like kinase 5 (transforming growth factor a new approach to the treatment of systemic sclerosis,” β receptor type I) signaling to the fibrotic phenotype of Toxicology, vol. 155, no. 1–3, pp. 1–15, 2000. scleroderma fibroblasts,” Arthritis and Rheumatism, vol. 54, [39] C. P. Denton, T. D. Bunce, M. B. Dorado et al., “Probu- no. 4, pp. 1309–1316, 2006. col improves symptoms and reduces lipoprotein oxidation [24] G. Lakos, S. Takagawa, S. J. Chen et al., “Targeted disruption susceptibility in patients with Raynaud’s phenomenon,” of TGF-β/Smad3 signaling modulates skin fibrosis in a Rheumatology, vol. 38, no. 4, pp. 309–315, 1999. Cardiology Research and Practice 7

[40] D. A. Lawlor, G. D. Smith, D. Kundu, K. R. Bruckdorfer, [56] M. Matucci Cerinic and M. B. Kahaleh, “Beauty and and S. Ebrahim, “Those confounded vitamins: what can the beast. The nitric oxide paradox in systemic sclerosis,” welearnfromthedifferences between observational versus Rheumatology, vol. 41, no. 8, pp. 843–847, 2002. randomised trial evidence?” Lancet, vol. 363, no. 9422, pp. [57] J. Leiper, J. Murray-Rust, N. McDonald, and P. Vallance, “S- 1724–1727, 2004. nitrosylation of dimethylarginine dimethylaminohydrolase [41] K. R. Bruckdorfer, “Antioxidants and CVD,” Proceedings of regulates enzyme activity: further interactions between nitric the Nutrition Society, vol. 67, no. 2, pp. 214–222, 2008. oxide synthase and dimethylarginine dimethylaminohydro- [42] P. Vallance and J. Leiper, “Cardiovascular biology of the lase,” Proceedings of the National Academy of Sciences of the asymmetric dimethylarginine:dimethylarginine dimethyl- United States of America, vol. 99, no. 21, pp. 13527–13532, aminohydrolase pathway,” Arteriosclerosis, Thrombosis, and 2002. Vascular Biology, vol. 24, no. 6, pp. 1023–1030, 2004. [58] M. P. Chung, M. M. Monick, N. Y. Hamzeh, N. S. Butler, [43] B. Kahaleh and P. S. Fan, “Down regulation of nitric oxide L. S. Powers, and G. W. Hunninghake, “Role of repeated synthase gene in microvascular endothelial cells from lesional lung injury and genetic background in bleomycin-induced scleroderma: assessment by quantitative RT-PCR and possi- fibrosis,” American Journal of Respiratory Cell and Molecular ble role for cytoxic T-cells,” Arthritis and Rheumatism, vol. 41, Biology, vol. 29, no. 3, part 1, pp. 375–380, 2003. supplement 9, p. S277, 1998. [59] S. Yoshimura, Y. Nishimura, T. Nishiuma, T. Yamashita, K. [44] B. Kahaleh, P. S. Fan, M. Matucci-Cerinic, M. Stefanovic- Kobayashi, and M. Yokoyama, “Overexpression of nitric ox- Racic, and L. Ignarro, “Study of endothelial dependent ide synthase by the endothelium attenuates bleomycin-in- relaxation in scleroderma,” Arthritis and Rheumatism, vol. 36, duced lung fibrosis and impairs MMP-9/TIMP-1 balance,” supplement 9, p. S180, 1993. Respirology, vol. 11, no. 5, pp. 546–556, 2006. [45] A. Sud, M. Khullar, A. Wanchu, and P. Bambery, “Increased [60] M. G. Ferrini, D. Vernet, T. R. Magee et al., “Antifibrotic role nitric oxide production in patients with systemic sclerosis,” of inducible nitric oxide synthase,” Nitric Oxide—Biology and Nitric Oxide—Biology and Chemistry, vol. 4, no. 6, pp. 615– Chemistry, vol. 6, no. 3, pp. 283–294, 2002. 619, 2000. [61] A. Dooley, S. Y. Low, A. Holmes et al., “Nitric oxide synthase [46] K. Takagi, Y. Kawaguchi, M. Hara, T. Sugiura, M. Harigai, and expression and activity in the tight-skin mouse model of N. Kamatani, “Serum nitric oxide (NO) levels in systemic fibrosis,” Rheumatology, vol. 47, no. 3, pp. 272–280, 2008. sclerosis patients: correlation between NO levels and clinical [62] R. Motterlini, R. Foresti, M. Intaglietta, and R. M. Winslow, features,” Clinical and Experimental Immunology, vol. 134, “NO-mediated activation of heme oxygenase: endogenous no. 3, pp. 538–544, 2003. cytoprotection against oxidative stress to endothelium,” [47] S. Y. Low, M. Sabetkar, K. R. Bruckdorfer, and K. M. Naseem, American Journal of Physiology—Heart and Circulatory Phys- “The role of protein nitration in the inhibition of platelet iology, vol. 270, no. 1, part 2, pp. H107–H114, 1996. activation by peroxynitrite,” FEBS Letters, vol. 511, no. 1–3, [63] R. Stocker and M. A. Perrella, “Heme oxygenase-1: a novel pp. 59–64, 2002. drug target for atherosclerotic diseases?” Circulation, vol. 114, [48]M.Sabetkar,S.Y.Low,K.M.Naseem,andK.R.Bruckdorfer, no. 20, pp. 2178–2189, 2006. “The nitration of proteins in platelets: significance in platelet [64] R. Hori, M. Kashiba, T. Toma et al., “Gene transfection function,” Free Radical Biology and Medicine, vol. 33, no. 6, of H25A mutant heme oxygenase-1 protects cells against pp. 728–736, 2002. hydroperoxide-induced cytotoxicity,” Journal of Biological [49] N. W. Kooy, J. A. Royall, Y. Z. Ye, D. R. Kelly, and J. S. Chemistry, vol. 277, no. 12, pp. 10712–10718, 2002. Beckman, “Evidence for in vivo peroxynitrite production in [65] K. D. Poss and S. Tonegawa, “Reduced stress defense in human acute lung injury,” American Journal of Respiratory heme oxygenase 1-deficient cells,” Proceedings of the National and Critical Care Medicine, vol. 151, no. 4, pp. 1250–1254, Academy of Sciences of the United States of America, vol. 94, 1995. no. 20, pp. 10925–10930, 1997. [50] P. I. Mapp, R. Klocke, D. A. Walsh et al., “Localization of 3- [66] H. Y. Cho, S. P. M. Reddy, M. Yamamoto, and S. R. Klee- nitrotyrosine to rheumatoid and normal synovium,” Arthritis berger, “The transcription factor NRF2 protects against pul- and Rheumatism, vol. 44, no. 7, pp. 1534–1539, 2001. monary fibrosis,” FASEB Journal, vol. 18, no. 11, pp. 1258– [51] K. Bennett-Richards, N. J. Bradley, D. Kundu, J. E. Deanfield, 1260, 2004. L. Rees, and K. R. A. Bruckdorfer, “A marked increase [67] H. Y. Cho, S. P. Reddy, A. DeBiase, M. Yamamoto, and S. in plasma nitrated proteins in children with chronic renal R. Kleeberger, “Gene expression profiling of NRF2-mediated failure,” Nitric Oxide—Biology and Chemistry, pp. 238–239, protection against oxidative injury,” Free Radical Biology and 2000. Medicine, vol. 38, no. 3, pp. 325–343, 2005. [52] N. Fukuyama, Y. Takebayashi, M. Hida, H. Ishida, K. Ichi- [68]K.Komura,J.Wei,G.Lord,andJ.Varga,“Thetriterpenoid mori, and H. Nakazawa, “Clinical evidence of peroxynitrite CDDO abrogates canonical TGF-beta signaling and fibrotic formation in chronic renal failure patients with septic shock,” responses in normal and scleroderma fibroblasts by stabiliz- Free Radical Biology and Medicine, vol. 22, no. 5, pp. 771–774, ing the anti-oxidant nuclear factor-like 2 (Nrf2),” Arthritis 1997. and Rheumatism, vol. 62, supplement 10, p. 617, 2010. [53] H. Cai and D. G. Harrison, “Endothelial dysfunction in [69] A. J. Chu and J. K. Prasad, “Up-regulation by human cardiovascular diseases: the role of oxidant stress,” Circulation recombinant transforming growth factor β-1 of collagen Research, vol. 87, no. 10, pp. 840–844, 2000. production in cultured dermal fibroblasts is mediated by the [54] J. S. Beckman and W. H. Koppenol, “Nitric oxide, superoxide, inhibition of nitric oxide signaling,” Journal of the American and peroxynitrite: the good, the bad, and the ugly,” American College of Surgeons, vol. 188, no. 3, pp. 271–280, 1999. Journal of Physiology—Cell Physiology, vol. 271, no. 5, part 1, [70] P. A. Craven, R. K. Studer, J. Felder, S. Phillips, and F. R. pp. C1424–C1437, 1996. DeRubertis, “Nitric oxide inhibition of transforming growth [55] A. L. Herrick, “Pathogenesis of Raynaud’s phenomenon,” factor-β and collagen synthesis in mesangial cells,” Diabetes, Rheumatology, vol. 44, no. 5, pp. 587–596, 2005. vol. 46, no. 4, pp. 671–681, 1997. 8 Cardiology Research and Practice

[71] M. Saura, C. Zaragoza, B. Herranz et al., “Nitric oxide reg- [85] J. Pfeilschifter, W. Eberhardt, and A. Huwiler, “Nitric oxide ulates transforming growth factor-β signaling in endothelial and mechanisms of redox signalling: matrix and matrix- cells,” Circulation Research, vol. 97, no. 11, pp. 1115–1123, metabolizing enzymes as prime nitric oxide targets,” Euro- 2005. pean Journal of Pharmacology, vol. 429, no. 1–3, pp. 279–286, [72] D. Vernet, M. G. Ferrini, E. G. Valente et al., “Effect of nitric 2001. oxide on the differentiation of fibroblasts into myofibroblasts [86] F. Jiang, S. J. Roberts, S. R. Datla, and G. J. Dusting, “NO in the Peyronie’s fibrotic plaque and in its rat model,” Nitric modulates NADPH oxidase function via heme oxygenase-1 Oxide—Biology and Chemistry, vol. 7, no. 4, pp. 262–276, in human endothelial cells,” Hypertension,vol.48,no.5,pp. 2002. 950–957, 2006. [73] E. G. A. Valente, D. Vernet, M. G. Ferrini, A. Qian, J. Rajfer, [87]H.Kim,J.Shim,P.L.Han,andE.J.Choi,“Nitricox- and N. F. Gonzalez-Cadavid, “L-Arginine and phosphodi- ide modulates the c-Jun N-terminal kinase/stress-activated esterase (PDE) inhibitors counteract fibrosis in the Pey- protein kinase activity through activating c-Jun N-terminal ronie’s fibrotic plaque and related fibroblast cultures,” Nitric kinase kinase,” Biochemistry, vol. 36, no. 44, pp. 13677– Oxide—Biology and Chemistry, vol. 9, no. 4, pp. 229–244, 13681, 1997. 2003. [88] C. Bogdan, “Nitric oxide and the regulation of gene expres- sion,” Trends in Cell Biology, vol. 11, no. 2, pp. 66–75, 2001. [74] A. Dooley, B. Gao, X. Shi-Wen et al., “Effect of nitric oxide [89] F. Verrecchia, E. F. Wagner, and A. Mauviel, “Distinct involve- and peroxynitrite on type I collagen synthesis in normal ment of the Jun-N-terminal kinase and NF-κBpathways and scleroderma dermal fibroblasts,” Free Radical Biology and in the repression of the human COL1A2 gene by TNF-α,” Medicine, vol. 43, no. 2, pp. 253–264, 2007. EMBO Reports, vol. 3, no. 11, pp. 1069–1074, 2002. [75] M. Cao, A. Westerhausen-Larson, C. Niyibizi et al., “Nitric [90] A. K. Ghosh, “Factors involved in the regulation of type I col- oxide inhibits the synthesis of type-II collagen without al- lagen gene expression: implication in fibrosis,” Experimental tering Col2A1 mRNA abundance: prolyl hydroxylase as a Biology and Medicine, vol. 227, no. 5, pp. 301–314, 2002. possible target,” Biochemical Journal, vol. 324, part 1, pp. [91] A. L. Herrick, K. J. Illingworth, S. Hollis, J. M. Gomez- 305–310, 1997. Zumaquero, and F. J. Tinahones, “Antibodies against oxi- [76] N. N. Kim, S. Villegas, S. R. Summerour, and F. J. Villarreal, dized low-density lipoproteins in systemic sclerosis,” Rheu- “Regulation of cardiac fibroblast extracellular matrix produc- matology, vol. 40, no. 4, pp. 401–405, 2001. tion by bradykinin and nitric oxide,” Journal of Molecular and [92]E.P.L.Turton,P.J.Kent,andR.C.Kester,“Theaetiologyof Cellular Cardiology, vol. 31, no. 2, pp. 457–466, 1999. Raynaud’s phenomenon,” Cardiovascular Surgery, vol. 6, no. [77] V. Kolpakov, D. Gordon, and T. J. Kulik, “Nitric oxide- 5, pp. 431–440, 1998. generating compounds inhibit total protein and collagen syn- [93] J. P. Cooke and J. M. Marshall, “Mechanisms of Raynaud’s thesis in cultured vascular smooth muscle cells,” Circulation disease,” Vascular Medicine, vol. 10, no. 4, pp. 293–307, 2005. Research, vol. 76, no. 2, pp. 305–309, 1995. [94] J. P. Cooke, S. J. Creager, K. M. Scales et al., “Role of [78] P. R. Myers and M. A. Tanner, “Vascular endothelial cell digital artery adrenoceptors in Raynaud’s disease,” Vascular regulation of extracellular matrix collagen: role of nitric Medicine, vol. 2, no. 1, pp. 1–7, 1997. oxide,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. [95] G. V. Bedarida, D. Kim, T. F. Blaschke, and B. B. Hoffman, 18, no. 5, pp. 717–722, 1998. “Venodilation in Raynaud’s disease,” Lancet, vol. 342, no. [79] M. W. Owens, S. A. Milligan, and M. B. Grisham, “Inhibition 8885, pp. 1451–1454, 1993. ff of pleural mesothelial cell collagen synthesis by nitric oxide,” [96] S. Singh, J. C. De Tra ord, P. A. Baskerville, and J. F. Martin, Free Radical Biology and Medicine, vol. 21, no. 5, pp. 601–607, “Response of digital arteries to endothelium dependent 1996. and independent vasodilators in patients with Raynaud’s [80] H. Trachtman, S. Futterweit, and P. Singhal, “Nitric oxide phenomenon,” European Journal of Clinical Investigation, vol. modulates the synthesis of extracellular matrix proteins in 25, no. 3, pp. 182–185, 1995. cultured rat mesangial cells,” Biochemical and Biophysical [97] P. J. W. Smith, C. J. Ferro, D. S. McQueen, and D. J. Research Communications, vol. 207, no. 1, pp. 120–125, 1995. Webb, “Impaired cholinergic dilator response of resistance arteries isolated from patients with Raynaud’s disease,” [81]D.ChakravorttyandK.S.N.Kumar,“Inductionofcell British Journal of Clinical Pharmacology, vol. 47, no. 5, pp. proliferation and collagen synthesis in human small intesti- 507–513, 1999. nal lamina propria fibroblasts by lipopolysaccharide: possible [98] J. L. Cracowski, S. Girolet, B. Imbert et al., “Effects of short- involvement of nitric oxide,” Biochemical and Biophysical term treatment with vitamin E in systemic sclerosis: a double Research Communications, vol. 240, no. 2, pp. 458–463, 1997. blind, randomized, controlled clinical trial of efficacy based [82] R. D. Curran, F. K. Ferrari, P. H. Kispert et al., “Nitric oxide on urinary isoprostane measurement,” Free Radical Biology and nitric oxide-generating compounds inhibit hepatocyte and Medicine, vol. 38, no. 1, pp. 98–103, 2005. protein synthesis,” FASEB Journal, vol. 5, no. 7, pp. 2085– [99] M. E. Mavrikakis, J. P. Lekakis, C. M. Papamichael, K. S. 2092, 1991. Stamatelopoulos, C. C. Kostopoulos, and S. F. Stamatelopou- [83] H. Trachtman, S. Futterweit, P. Garg, K. Reddy, and P. C. los, “Ascorbic acid does not improve endothelium-dependent Singhal, “Nitric oxide stimulates the activity of a 72 kDa flow-mediated dilatation of the brachial artery in patients neutral matrix metalloproteinase in cultured rat mesangial with Raynaud’s phenomenon secondary to systemic sclero- cells,” Biochemical and Biophysical Research Communications, sis,” International Journal for Vitamin and Nutrition Research, vol. 218, no. 3, pp. 704–708, 1996. vol. 73, no. 1, pp. 3–7, 2003. [84] K. I. Kivirikko, R. Myllyla,¨ and T. Pihlajaniemi, “Protein [100] F. Salsano, C. Letizia, M. Proietti et al., “Significant changes hydroxylation: prolyl 4-hydroxylase, an enzyme with four of peripheral perfusion and plasma adrenomedullin levels cosubstrates and a multifunctional subunit,” The FASEB in N-acetylcysteine long term treatment of patients with Journal, vol. 3, no. 5, pp. 1609–1617, 1989. sclerodermic Raynauds phenomenon,” International Journal Cardiology Research and Practice 9

of Immunopathology and Pharmacology, vol. 18, no. 4, pp. [115] H. L. Li, Y. Huang, C. N. Zhang et al., “Epigallocathechin-3 761–770, 2005. gallate inhibits cardiac hypertrophy through blocking reac- [101] P. Sambo, D. Amico, R. Giacomelli et al., “Intravenous N- tive oxidative species-dependent and -independent signal acetylcysteine for treatment of Raynaud’s phenomenon sec- pathways,” Free Radical Biology and Medicine, vol. 40, no. 10, ondary to systemic sclerosis: A pilot study,” Journal of Rheu- pp. 1756–1775, 2006. matology, vol. 28, no. 10, pp. 2257–2262, 2001. [116] H. Nishikawa, K. Wakano, and S. Kitani, “Inhibition of [102] G. L. Tipoe, T. M. Leung, M. W. Hung, and M. L. NADPH oxidase subunits translocation by tea catechin Fung, “Green tea polyphenols as an anti-oxidant and anti- EGCG in mast cell,” Biochemical and Biophysical Research inflammatory agent for cardiovascular protection,” Cardio- Communications, vol. 362, no. 2, pp. 504–509, 2007. vascular and Hematological Disorders—Drug Targets, vol. 7, [117] S. K. Katiyar, F. Afaq, A. Perez, and H. Mukhtar, “Green no. 2, pp. 135–144, 2007. tea polyphenol (-)-epigallocatechin-3-gallate treatment of [103] A. Pannala, C. A. Rice-Evans, B. Halliwell, and S. Singh, human skin inhibits ultraviolet radiation-induced oxidative “Inhibition of peroxynitrite-mediated tyrosine nitration by stress,” Carcinogenesis, vol. 22, no. 2, pp. 287–294, 2001. catechin polyphenols,” Biochemical and Biophysical Research [118] S. K. Katiyar, “UV-induced immune suppression and pho- Communications, vol. 232, no. 1, pp. 164–168, 1997. tocarcinogenesis: chemoprevention by dietary botanical [104] C. Rice-Evans, “Implications of the mechanisms of action of agents,” Cancer Letters, vol. 255, no. 1, pp. 1–11, 2007. tea polyphenols as antioxidants in vitro for chemoprevention [119] N. Sriram, S. Kalayarasan, and G. Sudhandiran, “Epigal- in humans,” Proceedings of the Society for Experimental locatechin-3-gallate exhibits anti-fibrotic effect by attenu- Biology and Medicine, vol. 220, no. 4, pp. 262–266, 1999. ating bleomycin-induced glycoconjugates, lysosomal hydro- [105] M. Sabetkar, S. Y. Low, N. J. Bradley, M. Jacobs, K. M. lases and ultrastructural changes in rat model pulmonary Naseem, and K. Richard Bruckdorfer, “The effect of epigal- fibrosis,” Chemico-Biological Interactions, vol. 180, no. 2, pp. locatechin-gallate on protein tyrosine nitration in platelets,” 271–280, 2009. Platelets, vol. 19, no. 4, pp. 282–292, 2008. [120] Y. Yasuda, M. Shimizu, H. Sakai et al., “(-)-Epigallocatechin [106] N. Khan, F. Afaq, M. Saleem, N. Ahmad, and H. Mukhtar, gallate prevents carbon tetrachloride-induced rat hepatic “Targeting multiple signaling pathways by green tea polyphe- fibrosis by inhibiting the expression of the PDGFRβ and IGF- nol (−)-epigallocatechin-3-gallate,” Cancer Research, vol. 66, 1R,” Chemico-Biological Interactions, vol. 182, no. 2-3, pp. no. 5, pp. 2500–2505, 2006. 159–164, 2009. [107] H. K. Na and Y. J. Surh, “Intracellular signaling network [121] M. C. Zhen, Q. Wang, X. H. Huang et al., “Green tea polyphe- as a prime chemopreventive target of (-)-epigallocatechin nol epigallocatechin-3-gallate inhibits oxidative damage and gallate,” Molecular Nutrition and Food Research, vol. 50, no. preventive effects on carbon tetrachloride-induced hepatic 2, pp. 152–159, 2006. fibrosis,” Journal of Nutritional Biochemistry, vol. 18, no. 12, [108] A. Chen, L. Zhang, J. Xu, and J. Tang, “The antioxidant pp. 795–805, 2007. (-)-epigallocatechin-3-gallate inhibits activated hepatic stel- late cell growth and suppresses acetaldehyde-induced gene expression,” Biochemical Journal, vol. 368, part 3, pp. 695– 704, 2002. [109] F. Yumei, Y. Zhou, S. Zheng, and A. Chen, “The antifibro- genic effect of (-)-epigallocatechin gallate results from the induction of de novo synthesis of glutathione in passaged rat hepatic stellate cells,” Laboratory Investigation, vol. 86, no. 7, pp. 697–709, 2006. [110] R. Sakata, T. Ueno, T. Nakamura, M. Sakamoto, T. Torimura, and M. Sata, “Green tea polyphenol epigallocatechin-3- gallate inhibits platelet-derived growth factor-induced pro- liferation of human hepatic stellate cell line LI90,” Journal of Hepatology, vol. 40, no. 1, pp. 52–59, 2004. [111] H. Asaumi, S. Watanabe, M. Taguchi et al., “Green tea pol- yphenol (-)-epigallocatechin-3-gallate inhibits ethanol-in- duced activation of pancreatic stellate cells,” European Journal of Clinical Investigation, vol. 36, no. 2, pp. 113–122, 2006. [112] M. Meng, Y. Q. Li, M. X. Yan, Y. Kou, and H. B. Ren, “Effects of epigallocatechin gallate on diethyldithiocarbamate-in- duced pancreatic fibrosis in rats,” Biological and Pharmaceu- tical Bulletin, vol. 30, no. 6, pp. 1091–1096, 2007. [113] Q. Zhang, A. P. Kelly, L. Wang et al., “Green tea extract and (−)-epigallocatechin-3-gallate inhibit mast cell-stimulated type I collagen expression in keloid fibroblasts via blocking PI-3K/Akt signaling pathways,” Journal of Investigative Der- matology, vol. 126, no. 12, pp. 2607–2613, 2006. [114] A. Dooley, X. Shi-Wen, N. Aden et al., “Modulation of col- lagen type I, fibronectin and dermal fibroblast function and activity, in systemic sclerosis by the antioxidant epi- gallocatechin-3-gallate,” Rheumatology, vol. 49, no. 11, pp. 2024–2036, 2010.