Retinal and Choroidal Angiogenesis

Retinal and Choroidal Angiogenesis

Retinal and Choroidal Angiogenesis Edited by

J.S. Penn Vanderbilt University School of Medicine, Nashville, TN, U.S.A.

Library of Congress Control Number: 2008920296

ISBN 978-1-4020-6779-2 (HB) ISBN 978-1-4020-6780-8 (e-book)

Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com

Printed on acid-free paper

All Rights Reserved © 2008 Springer Science+Business Media B.V. No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.

TABLE OF CONTENTS

Preface..................................................................................................ix Speaker Photo .....................................................................................xii Contributors ...................................................................................... xiii Introduction.........................................................................................xv Angiogenesis Study Models 1.

Cellular and Molecular Mechanisms of Retinal Angiogenesis M. A. Behzadian, M. Bartoli, A. B. El-Remessy, M. Al-Shabrawey, D. H. Platt, G. I. Liou, R. W. Caldwell, and R. B. Caldwell .................................................................................. 1

2.

Animal Models of Choroidal Neovascularization M. L. Clark, J. A. Fowler, and J. S. Penn ............................................. 41

3.

Rodent Models of Oxygen-Induced Retinopathy S. E. Yanni, G. W. McCollum, and J. S. Penn....................................... 57

4.

Animal Models of Diabetic Retinopathy T. S. Kern .............................................................................................. 81

5.

Neovascularization in Models of Branch Retinal Vein Occlusion R. P. Danis and D. P. Bingaman ........................................................ 103

Molecular Characterization 6.

Vasculogenesis and Angiogenesis in Formation of the Human Retinal Vasculature T. Chan-Ling....................................................................................... 119

v

vi

Table of Contents

7.

IGF-1 and Retinopathy L. E. H. Smith...................................................................................... 139

8.

Hypoxia and Retinal Neovascularization B. A. Berkowitz ................................................................................... 151

9.

Hypoxia Inducible Factor-1 and VEGF Induction A. Madan............................................................................................. 169

10. The Role of Protein Kinase C in Diabetic Retinal Vascular Abnormalities J. K. Sun and G. L. King ..................................................................... 187 11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis J. Chen, D. Brantley-Siders, and J. S. Penn ....................................... 203 12. Adenosine in Retinal Vasculogenesis and Angiogenesis in Oxygen-Induced Retinopathy G. A. Lutty and D. S. McLeod............................................................. 221 13. The Regulation of Retinal Angiogenesis by Cyclooxygenase and the Prostanoids G. W. McCollum and J. S. Penn ......................................................... 241 14. Extracellular Proteinases in Ocular Angiogenesis A. Das and P. G. McGuire.................................................................. 259 15. Oxygen-Independent Angiogenic Stimuli J. M. Holmes, D. A. Leske, and W. L. Lanier...................................... 279 16. Growth Factor Synergy in Angiogenesis A. V. Ljubimov .................................................................................... 289 17. Pigment Epithelium-Derived Factor and Angiogenesis J. Amaral and S. P. Becerra ............................................................... 311 18. Circulating Endothelial Progenitor Cells and Adult Vasculogenesis S. Caballero, N. Sengupta, L. C. Shaw, and M. B. Grant ................... 339 Applications to Clinical Conditions 19. Retinopathy of Prematurity D. L. Phelps ........................................................................................ 363 20. Angiogenesis in Sickle Cell Retinopathy G. A. Lutty and D. S. McLeod............................................................. 389 21. Diabetic Retinopathy R. N. Frank ......................................................................................... 407

Table of Contents

vii

22. Systems for Drug Delivery to the Posterior Segment of the Eye A. L. Weiner and D. A. Marsh ............................................................ 419 23. Novel Therapeutic Strategies for Posterior Segment Neovascularization D. P. Bingaman, X. Gu, A. M. Timmers, and A. Davis....................... 445 24. Choroidal Neovascularization in Age-Related Macular Degeneration—From Mice to Man L. Berglin ............................................................................................ 527

Glossary ............................................................................................545 Index..................................................................................................551

PREFACE

Eye diseases with retinal or choroidal angiogenesis as a critical pathological feature are responsible for the majority of all cases of blindness in developed countries. Thus, due to its profound impact, ocular angiogenesis is an intensely studied process, and the field is advancing at an astounding pace. The growing number of investigators interested in ocular angiogenesis has compounded the increasingly difficult task of managing all of the available information. We, therefore, thought that it was time to take stock of the collective research, to focus on its important and potentially beneficial aspects, and to summarize the progress to date. The contents of this book are based on the proceedings of the Retinal and Choroidal Angiogenesis Symposium, held at Vanderbilt University on October 15 and 16, 2004. The Symposium was generously sponsored by the National Eye Institute and a number of interested pharmaceutical companies, mentioned below. The primary goal of the Symposium was to promote the exchange of current information and ideas among basic and clinical scientists. It was our intention to foster a better understanding of the basic mechanisms underlying ocular angiogenesis and to advance the development of therapeutic interventions. To this end, we featured a collection of investigators from diverse research and clinical centers throughout the United States, ranging from cell and developmental biologists to clinicianscientists. Specifically, we wished to address three aims: (1) to facilitate scientific exchange and collaborative interaction among senior investigators in the field; (2) to create an opportunity for students, young researchers, and fellows to meet and interact with established investigators; and (3) to provide the impetus for this published work.

ix

x

Preface

This book encompasses a broad spectrum of topics related to angiogenesis within the eye. Topics include basic information on the cellular and molecular mechanisms of retinal and choroidal angiogenesis, animal models of ocular angiogenic conditions, novel therapeutic strategies for the treatment of these conditions, drug development efforts to address these novel strategies, and the application of new mechanistic theories to human disease pathogenesis. The book seeks to emphasize basic principles rather than specific experimental results, although contributors were encouraged to use recently acquired data to illustrate points of broader theoretical significance. I have attempted to arrange the chapters and their topics so that a progression exists, beginning with a description of research tools, model systems, and an examination of the molecular facets of the angiogenic cascade, and ending with the most recent efforts to translate these facets into molecular targets for drug development efforts. The target audience is the interested professional – basic scientist, clinician-scientist, or physician – whether involved in the field of ophthalmology or in other disciplines in which angiogenesis is important. That such a spectrum of topics on such a complicated subject could be encompassed in a single book may seem a daunting goal. Yet, I believe that we have met it. This is a tribute to the contributors’ command of their subjects, their range of interest, and the energy and enthusiasm that they brought to the task. And, it is clearly evident as one reads the chapters. I would like to thank all those who have participated as speakers and as authors. Without their willingness to attend the Symposium and to meet submission deadlines for their contributions, this book would not have been possible. Neither would it have been possible without the help, support, and encouragement of several others: Paul Sternberg, Jr., M.D., the chair of Ophthalmology and Visual Sciences at Vanderbilt University, who provided valuable advice and Department funding to get the project started; Melissa Stauffer, Ph.D., at Scientific Editing Solutions, who spent many hours poring over the chapters and providing other services related to the editing process; Yolanda Miller, who provided on-site support to the participants and attendees of the Symposium; Peter A. Dudley, Ph.D., of the National Eye Institute, who offered a number of suggestions that improved the Symposium and helped us to meet our aims; and finally, Kathy Haddix, who handled communication with Symposium participants, made sure that they were comfortable while in Nashville, and planned and hosted the meals and social functions. Her efforts were tireless and her positive influence was felt by every participant and attendee. The Symposium also received generous financial support from Pfizer Global, Alcon Laboratories, Eyetech Pharmaceuticals, and Genentech. In addition, I would like to thank the local

Preface

xi

attendees, faculty and students alike, who were present at the Symposium and who asked terrific questions and stimulated excellent discussion. It is my sincere hope that this volume will be useful as an introduction to angiogenesis in the posterior segment of the eye, and as a reference source for both established researchers and novices in the field. John S. Penn

1. Alexander V. Ljubimov 2. Stanley J. Wiegand 3. Karl G. Csaky 4. Gerard A. Lutty 5. Luyuan Y. Li 6. Timothy S. Kern 7. Arup Das 8. David P. Bingaman

9. Jonathan M. Holmes 10. Michael R. Niesman 11. Azza El-Remessy 12. Dale L. Phelps 13. Robert N. Frank 14. Bruce A. Berkowitz 15. Lois E. H. Smith 16. George L. King

17. John S. Penn 18. Asher Weiner 19. Martin Friedlander 20. Janet C. Blanks 21. Maria B. Grant 22. S. Patricia Becerra 23. Ruth B. Caldwell 24. Tailoi Chan-Ling

INTRODUCTION

I am honored to be asked to write the introduction to this book, Retinal and Choroidal Angiogenesis. Its publication is very timely because of rapid progress in the treatment of neovascular age-related macular degeneration by FDA-approved angiogenesis inhibitors1,2 and because of the initiation of clinical trials of this therapy for other types of ocular neovascularization. Professor Penn has organized a comprehensive and forward-looking set of central issues that inform the molecular basis of ocular neovascularization and its modern therapy. He has invited a distinguished group of authors to discuss these topics and to think about future directions. Taken together, the chapters in this book reveal certain principles of ocular angiogenesis that have emerged from the study of tumor angiogenesis. Endogenous inhibitors of angiogenesis are expressed by different types of cells in the eye and are stored in different matrix compartments. These inhibitors counterbalance pro-angiogenic molecules in the eye. Most neovascular diseases of the eye begin with a shift of the angiogenic balance to the pro-angiogenic phenotype, termed the “angiogenic switch” in cancer biology.3,4 Increased expression or mobilization of pro-angiogenic proteins, accompanied by decreased expression or deficiency of anti-angiogenic proteins, can be mediated or potentiated by hypoxia, infiltration of inflammatory cells or immune cells, accumulation of platelets and bone marrow-derived endothelial cells at an angiogenic site, changes in stromal fibroblast expression of angiogenesis inhibitors, and other events. A recent significant advance in treating diseases of ocular neovascularization by anti-angiogenic therapy is based on the development of drugs that neutralize a pro-angiogenic protein, vascular endothelial growth factor (VEGF). However, it took more than four decades for this xv

xvi

Introduction

angiogenic molecule to be identified and characterized as a target for antiangiogenic therapy of the eye. The journey was circuitous. In 1945, Algire et al. suggested that a diffusible “factor” could mediate tumor neovascularization.5 In 1948, Michaelson suggested that a diffusible “X-factor” could mediate neovascular retinopathies.6 Similar proposals of the existence of diffusible angiogenic factors were reported by others from experiments with tumors implanted in the anterior chamber of the guinea pig eye7 and from transfilter diffusion studies of tumors in the hamster check pouch.8,9 However, none of these experiments yielded a purified angiogenic molecule. In fact, efforts to completely purify a tumor-derived angiogenic factor were driven by a hypothesis that I published in 1971 that tumor growth is angiogenesis-dependent.10 This report also proposed that “anti-angiogenesis” could be a new therapeutic principle for cancer. This paper predicted the future discovery of angiogenesis inhibitors and that neutralization of a “tumor angiogenic factor” by an antibody could be therapeutic. Accordingly, we began to purify angiogenic activity from tumor extracts and to develop bioassays for angiogenesis.11 These bioassays included the implantation of tumors into a corneal micropocket in experimental animals12 and the development of sustained-release polymers that could be implanted into the corneal pocket to quantify the angiogenic activity of tumor-derived proteins.13 Throughout the 1970’s this hypothesis was widely ridiculed. However, when removal of pro-angiogenic sustained-release pellets from corneas was followed by complete regression of the induced neovascularization,14 our confidence was boosted, and we persisted in the purification of an angiogenic factor from a tumor. Since then, experimental ocular neovascularization has been essential for continued progress in the field of angiogenesis research. In 1984, we reported the complete purification by heparin-affinity chromatography of a capillary endothelial growth factor isolated from a tumor,15 and in 1985 its angiogenic activity.16 Subsequently, Esch et al. determined the amino acid sequence of a pituitary-derived protein, basic fibroblast growth factor (bFGF),17 previously isolated and partially purified from brain tissue by Gospodarowicz.18 In 1986, Klagsbrun in our laboratory determined that our capillary endothelial growth factor had the same sequence as bFGF.19 In 1983, Senger in Harold Dvorak’s lab reported that tumor cells secreted a vascular permeability factor (VPF), which promoted ascites.20 It was not known to be an endothelial growth factor at that time. By 1989, Rosalind Rosenthal in my laboratory employed heparin-affinity chromatography to purify to homogeneity a second endothelial growth factor. She had isolated this protein from sarcoma 180 cells, and it was not bFGF. We had set out to make sufficient quantities of the protein to

Introduction

xvii

determine its amino acid sequence when I received a call from Napoleone Ferrara of Genentech. He had heard that we had purified a new endothelial mitogen from a tumor. He had also purified a new endothelial mitogen from pituitary cells. He had already determined the amino acid sequence of his protein and offered to sequence our protein for comparison. This was opportune for us, because we faced at least another year of work to produce sufficient protein to sequence it ourselves. Ferrara determined the amino acid sequence of our protein and found it to be identical to his. Ferrara’s paper on this second endothelial growth factor appeared in 1989, entitled “Pituitary follicular cells secrete a novel heparin-binding growth factor specific for vascular endothelial cells.”21 Our paper appeared in 1990, entitled “Conditioned medium from mouse sarcoma 180 cells contains vascular endothelial growth factor” (VEGF).22 VPF also turned out to be identical to VEGF. Ferrara and Henzel were our co-authors because they had determined the amino acid sequence for us. I have outlined this history in more detail than is customary for an Introduction, because it reveals the importance of heparin affinity in purifying angiogenic proteins, and because it was the prelude to future collaborations between Ferrara’s lab and mine. We reported the first angiogenesis inhibitor in 1980 using the same bioassays, and reported eleven others over the next 25 years.2 Eight of these were endogenous angiogenesis inhibitors, including angiostatin and endostatin. However, by the early 1990’s, it was still not clear how retinal angiogenesis or other ocular neovascular pathologies were mediated. Was Michaelson’s “X-factor” bFGF, VEGF, or a different endothelial mitogen? Anthony Adamis in my laboratory began a formal exploration of this question. By 1993, Adamis et al. could report that human retinal pigment epithelial cells secreted VEGF.23 Patricia D’Amore, a co-author on this paper, was also a post-doctoral fellow in my lab. She is currently Professor of Ophthalmology at Harvard (Schepens Institute). During this period, Adamis also carried out the experiments that led to the development of pegaptanib (Macugen). In 1994, in collaboration with Joan Miller at the Massachusetts Eye and Ear Infirmary and Harold Dvorak at the Beth Israel Hospital, we reported that VEGF was significantly increased in the vitreous of monkey eyes when retinal neovascularization was induced by laser injury.24 We also reported at that time that samples of human vitreous obtained from diabetic eyes revealed very high levels of VEGF.25 In the same year, Lloyd Paul Aiello at the Joslin Clinic also reported high levels of VEGF in diabetic vitreous.26 By the following year, in collaboration with Napoleone Ferrara of Genentech, we had demonstrated that retinal cells subjected to hypoxia significantly increased their expression of VEGF, and that VEGF was the

xviii

Introduction

primary endothelial cell mitogen made by those cells.27 By 1996, we showed that iris neovascularization associated with retinal ischemia in monkeys was prevented by treatment with an antibody to VEGF, the precursor to bevacizumab (Avastin), given to us by Napoleone Ferrara.28 This became a seminal paper, because it proved that an antibody to VEGF could be used as a drug to treat ocular neovascularization in a non-human primate. It became the basis for (i) other experimental models of therapy of retinal neovascularization,29 (ii) anti-angiogenic therapy of human neovascular agerelated macular degeneration,30 and (iii) clinical trials of anti-angiogenic therapy for diabetic retinopathy. In December 2004, Macugen was approved by the FDA to treat neovascular age-related macular degeneration, and in June 2006, ranibizumab (Lucentis) was also approved for this indication.

FUTURE DIRECTIONS Long-term maintenance of angiostatic therapy. A recent report reveals that intravitreal endostatin is effective in treating experimental choroidal neovascularization in mice.31 These experiments suggest that endostatin may be used to treat neovascular age-related macular degeneration, analogous to its use as a “replacement therapy” in experimental atherosclerosis.32,33 Endostatin suppresses endothelial responsiveness to a wide spectrum of proangiogenic stimuli in pathological neovascularization,34,35 but not in reproduction or wound healing. Endostatin has shown no side effects in animals or during clinical trials. Therefore, it may also be useful for longterm “maintenance” therapy for patients with ocular neovascularization whose sight has been restored by intravitreal ranibizumab or bevacizumab. Endostatin could be administered subcutaneously or by intravitreal injection. Angiogenesis-based biomarkers in urine and blood. In the future, antiangiogenic maintenance therapy of ocular neovascularization could possibly be monitored by quantification of metalloproteinases in urine36 or by analysis of the platelet angiogenesis proteome.37,38 Microscopic tumors in mice can be detected by analysis of the platelet angiogenesis proteome because platelets sequester and accumulate VEGF and other angiogenesis regulatory proteins that these tumors release. It is possible that analysis of the platelet angiogenesis proteome could also be used to detect recurrence of choroidal vascular leakage or to detect an increase in choroidal neovascularization, long before detection by ophthalmoscopy. In other words, “ultra-early” prediction of patients at risk for ocular neovascularization may eventually be possible by quantification of angiogenesis-based biomarkers in blood or urine.

Introduction

xix

Betacellulin. Early clinical trials of ranibizumab in diabetic retinopathy reveal that visual acuity can be improved, but that higher doses, or more frequent dosing, may be required than are currently used for macular degeneration. It is possible that in addition to VEGF, there is another mediator(s) of angiogenesis in the diabetic retina, for example, betacellulin. We first isolated, purified, and determined the amino acid sequence of betacellulin from conditioned medium of proliferating neoplastic beta cells of murine pancreatic islets.39 Betacellulin is a 32-kD new member of the epidermal growth factor family with 50% homology to TGF-alpha. It is a mitogen for retinal pigment epithelial cells and for smooth muscle cells. We hypothesized that “regenerating beta cells in the diabetic pancreas may release excessive amounts of betacellulin.”40 Retinal pigment epithelial cells contain high concentrations of bFGF, which is a potent angiogenic peptide. Stimulation of retinal pigment epithelial cells by betacellulin could possibly initiate or potentiate neovascularization in the diabetic retina. This hypothesis could explain two well-known, but puzzling clinical observations: (i) Diabetic patients who receive a successful pancreas transplant that improves glucose metabolism and may free them from insulin-dependence rarely show improvement in their retinopathy or in their peripheral vascular disease. We speculate that the patient’s original pancreas continues to secrete betacellulin. (ii) Patients who undergo total pancreatectomy for cancer develop severe diabetes because of complete absence of insulin, but they rarely if ever develop diabetic retinopathy, even when they survive for more than 10-20 years. Thus, it is possible that excessive release of betacellulin may contribute to the vascular complications of diabetes. Recent experiments by Bela Anand-Apte of the Cole Eye Institute, Cleveland Clinic, in collaboration with my laboratory, show that in mice with diabetes induced by streptozotocin, intravitreal injection of betacellulin significantly increases vascular leakage in the retina (unpublished data). Betacellulin may provide a biochemical link between pancreatic islets and the microvasculature of the eye. It can be speculated that blockade of betacellulin, perhaps by an antibody, could ameliorate diabetic retinopathy and synergize anti-VEGF therapy.

SUMMARY Experimental models of ocular neovascularization in the early 1970’s made it possible to prove that tumors secreted specific pro-angiogenic proteins. These models also evolved into bioassays to identify novel angiogenesis inhibitors, both endogenous and synthetic. These angiogenesis inhibitors paved the way for the development of a new class of FDA-approved drugs

xx

Introduction

that have become a “fourth modality” for anti-cancer therapy. These same new drugs have more recently become a novel approach for the treatment of neovascular age-related macular degeneration. Current experiments in many laboratories indicate that in the future, other diseases of ocular neovascularization and/or vascular hyperpermeability may be treated by these angiogenesis inhibitors. Long-term maintenance of suppression of pathological ocular neovascularization may become possible. Angiogenesisbased biomarkers in the blood or urine may be employed to predict patients who are at risk for recurrence of ocular neovascularization, so that treatment can begin before detection by conventional methods. Finally, additional mediators of ocular neovascularization may exist, such as in diabetic retinopathy, where betacellulin is a candidate for study. Judah Folkman, MD

REFERENCES 1. E. M. Stone, A very effective treatment for neovascular macular degeneration. N. Engl. J. Med. 355, 1493-1495 (2006). 2. J. Folkman, Angiogenesis: an organizing principle for drug discovery? Nat. Rev. Drug Discov. 6, 273-286 (2007). 3. N. Almog, V. Henke, L. Flores, L. Hlatky, A. L. Kung, R. D. Wright, R. Berger, L. Hutchinson, E. Bender, E. Achilles, and J. Folkman, Prolonged dormancy of human liposarcoma is associated with impaired tumor angiogenesis, FASEB J. 20, 947-949 (2006). 4. G. N. Naumov, E. Bender, D. Zurakowski, S. Kang, D. Sampson, E. Flynn, R. S. Watnick, O. Straume, L. A. Akslen, J. Folkman, and N. Almog, A model of human tumor dormancy: an angiogenic switch from the nonangiogenic phenotype, J. Natl. Cancer Inst. 98, 316-325 (2006). 5. G. H. Algire, H. W. Chalkley, F. Y. Legallais, and H. Park, Vascular reactions of normal and malignant tumors in vivo. I. Vascular reactions of mice to wounds and to normal and neoplastic transplants, J. Natl. Cancer Inst. 6, 73-85 (1945). 6. I. C. Michaelson, Vascular morphogenesis in the retina of the cat. J. Anat. 82, 167-174 (1948). 7. H. S. N. Greene, Heterologous transplantation of mammalian tissues, J. Exper. Med. 73, 461, (1961). 8. M. Greenblatt and P. Shubik, Tumor angiogenesis: transfilter diffusion. Studies in the hamster by transparent chamber technique, J. Natl. Cancer Inst. 41, 1111, (1968). 9. R. L. Erhman and M. Knoth, Choriocarcioma transfilter stimulation of vasoproliferation in the hamster cheek pouch−studied by light and electron microscopy, J. Natl. Cancer Inst. 41, 329, (1968).

Introduction

xxi

10. J. Folkman, Tumor angiogenesis: therapeutic implications, N. Engl. J. Med. 285, 1182-1186 (1971). 11. J. Folkman, E. Merler, C. Abernathy, and G. Williams, Isolation of a tumor factor responsible for angiogenesis, J. Exp. Med. 133, 275-288 (1971). 12. M. A. Gimbrone, Jr., R. S. Cotran, S. B. Leapman, and J. Folkman, Tumor growth and neovascularization: an experimental model using rabbit cornea. J. Natl. Canc Inst. 52, 413-427 (1974). 13. R. Langer and J. Folkman, Polymers for the sustained release of proteins and other macromolecules, Nature 263, 797-800 (1976). 14. D. H. Ausprunk, K. Falterman, and J. Folkman, The sequence of events in the regression of corneal capillaries, Lab. Invest. 38, 284-294 (1978). 15. Y. Shing, G. Christofori, D. Hanahan, Y. Ono, R. Sasada, K. Igarashi, and J. Folkman, Heparin affinity: purification of a tumor-derived capillary endothelial cell growth factor, Science 223, 1296-1298 (1984). 16. Y. Shing, J. Folkman, C. Haudenschild, D. Lund, R. Crum, amd M. Klagsbrun, Angiogenesis is stimulated by a tumor-derived endothelial cell growth factor, J. Cell. Biochem. 29, 275-287 (1985). 17. F. Esch, A. Baird, N. Ling, N. Ueno, F. Hill, L. Denoroy, R. Klepper, D. Gospodarowicz, P. Bohlen, and R. Guillemin, Primary structure of bovine pituitary basic fibroblast growth factor (FGF) and comparison with the amino-terminal sequence of bovine brain acidic FGF, Proc. Natl. Acad. Sci. USA 82, 6507-6511 (1985). 18. D. Gospodarowicz, Purification of bovine fibroblast growth factor from pituitary, J. Biol. Chem. 250, 2515-2520 (1975). 19. M. Klagsbrun, J. Sasse, R. Sullivan, and J. A. Smith, Human tumor cells synthesize an endothelial cell growth factor that is structurally related to bFGF, Proc. Natl. Acad. Sci. USA 83, 2448-2452 (1986). 20. D. R. Senger, S. J. Galli, A. M. Dvorak, C. A. Perruzzi, V. S. Harvey, and H. F. Dvorak, Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid, Science 219, 983-985 (1983). 21. N. Ferrara and W. J. Henzel, Pituitary follicular cells secrete a novel heparin-binding growth factor specific for vascular endothelial cells, Biochem. Biophys. Res. Commun. 161, 851-858 (1989). 22. R. A. Rosenthal, J. F. Megyesi, W. J. Henzel, N. Ferrara, and J. Folkman, Conditioned medium from mouse sarcoma 180 cells contains vascular endothelial growth factor, Growth Factors 4, 53-59 (1990). 23. A. P. Adamis, D. T. Shima, K.-T. Yeo, T.-K. Yeo, L. F. Brown, B. Berse, P. A. D’Amore, and J. Folkman, Synthesis and secretion of vascular permeability factor/vascular endothelial growth factor by human retinal pigment epithelial cells, Biochem. Biophys. Res. Commun. 193, 631-638 (1993). 24. J. W. Miller, A. P. Adamis, D. T. Shima, P. A. D’Amore, R. S. Moulton, M. S. O’Reilly, J. Folkman, H. F. Dvorak, L. F. Brown, B. Berse, T.-K. Yeo, and K.-T. Yeo, Vascular endothelial growth factor/vascular permeability factor is temporally and spatially correlated with ocular angiogenesis in a primate model, Am. J. Pathol. 145, 574-584 (1994). 25. A. P. Adamis, J. W. Miller, M.-T. Bernal, D. J. D’Amico, J. Folkman, T.-K. Yeo, and K.-T. Yeo, Increased vascular endothelial growth factor levels in the vitreous of eyes with proliferative diabetic retinopathy, Am. J. Ophthalmol. 118, 445-450 (1994). 26. L. P. Aiello, R. L. Avery, P. G. Arrigg, B. A. Keyt, H. D. Jampel, S. T. Shah, L. R. Pasquale, H. Thieme, M. A. Iwamoto, J. E. Park, et al. Vascular endothelial growth

xxii

27.

28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

38. 39.

40.

Introduction

factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders, N. Engl. J. Med. 331, 1480-1487 (1994). D. T. Shima, A. P. Adamis, N. Ferrara, K.-T. Yeo, T.-K. Yeo, R. Allende, J. Folkman, and P. A. D’Amore, Hypoxic induction of endothelial cell growth factors in retinal cells: identification and characterization of vascular endothelial growth factor (VEGF) as the mitogen, Mol. Med. 1, 182-193 (1995). A. P. Adamis, D. T. Shima, M. J. Tolentino, E. A. Gragoudas, N. Ferrara, J. Folkman, P. A. D’Amore, and J. W. Miller, Inhibition of vascular endothelial growth factor prevents retinal ischemia-associated iris neovascularization in a nonhuman primate, Arch. Ophthalmol. 114, 66-71 (1996). J. S. Penn, V. S. Rajaratnam, R. J. Collier, and A. F. Clark, The effect of an angiostatic steroid on neovascularization in a rat model of retinopathy of prematurity, Invest. Ophthalmol. Vis. Sci. 42, 283-290 (2001). E. M. Stone, A very effective treatment for neovascular macular degeneration, N. Engl. J. Med. 355, 1493-1495 (2006). A. G. Marneros, H. She, H. Zambarakji, H. Hashizume, E. J. Connolly, I. Kim, E. S. Gragoudas, J. W. Miller, and B. R. Olsen, Endogenous endostatin inhibits choroidal neovascularization, FASEB J. 21: (published online, May 25, 2007). K. S. Moulton, E. Heller, M. A. Konerding, E. Flynn, W. Palinski, and J. Folkman, Angiogenesis inhibitors endostatin and TNP-470 reduce intimal neovascularization and plaque growth in Apolipoprotein E-deficient mice, Circulation 99, 1726-1732 (1999). K. S. Moulton, B. R. Olsen, S. Soon, N. Fukai, D. Zurakowski, and X. Zeng, Loss of collagen XVIII enhances neovascularization and vascular permeability in atherosclerosis, Circulation 110, 1330-1336 (2004). A. Abdollahi, P. Hahnfeldt, C. Maercker, H. J. Gröne, J. Debus, W. Ansorge, J. Folkman, L. Hlatky, and P. E. Huber, Endostatin’s antiangiogenic signaling network, Mol. Cell 13, 649-663, (2004). A. Abdollahi, C. Schwager, J. Kleeff, I. Esposito, S. Domhan, P. Peschke, K. Hauser, P. Hahnfeldt, L. Hlatky, J. Debus, J. M. Peters, H. Friess, J. Folkman, and P. E. Huber, A Transcriptional Network Governing the Angiogenic Switch- Evidence in Human Pancreatic Carcinoma, Proc. Natl. Acad. Sci. USA 2007, in press. R. Roy, U. M. Wewer, D. Zurakowski, S. E. Pories, and M. A. Moses, ADAM 12 cleaves extracellular matrix proteins and correlates with cancer status and stage, J. Biol. Chem. 279, 51323-51330 (2004). G. Klement, L. Kikuchi, M. Kieran, N. Almog, T. T. Yip, and J. Folkman, Early tumor detection using platelet uptake of angiogenesis regulators. Proc 47th American Society of Hematology. Blood 104, 239a Abs. #839 (2004). G. Klement, D. Cervi, T. Yip, J. Folkman, and J. Italiano, Platelet PF-4 Is an early marker of tumor angiogenesis, Blood 108, 426a, Abs. #147 (2006). Y. Shing, G. Christofori, D. Hanahan, Y. Ono, R. Sasada, K. Igarashi, and J. Folkman, Betacellulin: A novel mitogen from pancreatic beta tumor cells, Science 259, 1604-1607 (1993). Y. Shing and J. Folkman, Betacellulin, in: Human Cytokines, Handbook for Basic and Clinical Research, Vol II, edited by B. B. Aggarwal and J. U. Gutterman (Blackwell Scientific Publications, Inc., 1996) pp. 331-339.

ANGIOGENESIS STUDY MODELS

Chapter 1 CELLULAR AND MOLECULAR MECHANISMS OF RETINAL ANGIOGENESIS What have we learned from in vitro models? M. A. Behzadian, M. Bartoli, A. B. El-Remessy, M. Al-Shabrawey, D. H. Platt, G. I. Liou, R. W. Caldwell, and R. B. Caldwell Vascular Biology Center and the Departments of Pharmacology & Toxicology, Ophthalmology and Cellular Biology & Anatomy, Medical College of Georgia, Augusta, Georgia

Abstract:

Angiogenesis is a multi-factorial process that involves different cell types and a number of cytokines and growth factors. Physiological angiogenesis is characterized by the existence of a delicate balance between pro-angiogenic and anti-angiogenic factors. In an in vivo setting, pro-angiogenic stimuli such as endothelial-specific mitogenic factors and extracellular matrix (ECM)degrading enzymes must be tightly regulated and locally constrained. Differentiation factors, protease inhibitors, and the elements involved in reconstruction of ECM and recruitment of mural cells must be elicited in an appropriate temporal and spatial arrangement. Overexpression of angiogenesis-activating factors may cause hyper-vascularization. However, deficiency or disarray in expression of anti-angiogenic factors may result in leaky vessels, unstable capillaries, and formation of dysfunctional neovascular tufts as seen in retinopathy of prematurity, diabetic retinopathy, or other conditions of retinal neovascularization. In other words, pathological angiogenesis is characterized not only by excesses in pro-angiogenic factors but also an insufficiency in anti-angiogenic, pro-differentiation factors. To better understand pathological angiogenesis, our experimental models should be able to dissect the dissolution phase of the angiogenic process from the resolution phase. In an in vivo model of pathological angiogenesis, these two components occur in close spatial and temporal proximity and thus are difficult to dissect. By using in vitro models, it is possible to begin with the most basic elements in order to reconstitute physiological and pathological conditions and compare each step of the process. Retina explants, primary cultures of retinal vascular endothelial cells, and cocultures of endothelial and mural cells, together with gene transfer techniques, have enabled us to analyze the functional roles of cytokines, growth factors,

1 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 1–39. © Springer Science+Business Media B.V. 2008

2

M. A. Behzadian et al. and extracellular proteolytic enzymes involved in the angiogenic process and to develop assay systems for testing the efficacy of pharmaceutical reagents that specifically block intracellular signaling pathways and transcription factors. Finally, use of endothelial and mural cells isolated from transgenic animals in tissue culture models aids in defining gene functions and elucidating the mechanisms of their regulation.

1.

INTRODUCTION

Much of our knowledge about retinal neovascularization has been obtained from research done with animal models. In addition to retinopathies in naturally occurring rodent mutants, a number of pathologies representing human disease conditions, such as retinopathy of prematurity, age related macular degeneration, and diabetic retinopathy, can be modeled in normal and transgenic animal retinas to study morphological aspects of abnormal angiogenesis. Analysis of diseased retinas using advanced imaging techniques, confocal microscopy, immunohistochemistry, in situ hybridization, and laser-capture micro-dissection has uncovered cellular elements and biological factors involved in the initiation and progression of pathological angiogenesis. However, the precise cellular sources and targets of gene and protein expression, as well as their molecular regulation and mechanisms of action, are more readily identified using in vitro models. Anti-angiogenic or pro-angiogenic reagents are best characterized for target specificity, effective dose and treatment time, potential cell toxicity, and mechanism of action when tested on isolated cells or organ explants under well defined tissue culture conditions. Retinal neovascularization, the inappropriate proliferation of new vessels derived from preexisting vessels, is a major cause of blindness and a significant component of many ocular diseases. Different types of cells participate in retinal angiogenesis, including endothelial cells, pericytes, astrocytes and Muller glial cells (see Figure 1). Hypoxia and hyperglycemia have been shown to be important causes of retinal neovascularization. In both conditions, the balance between angiogenic and angiostatic growth factors, which usually serves to keep angiogenesis in check, is disturbed.

1. Mechanisms of Retinal Angiogenesis

3

Figure 1-1. Distribution of neurons, glia (Muller cells and astrocytes), and blood vessels within the retina. Color coding indicates the type of cell.

1.1

Angiogenic Process

The purpose of this introduction is to depict the multi-factorial/multi-cellular nature of the angiogenic process and show how in vitro models can be used to dissect and identify the individual elements involved and analyze their mechanism of action. Angiogenesis is defined as the formation of new blood vessels by sprouting from pre-existing capillaries. Physiological angiogenesis takes place during development and wound healing and in the female reproductive system. Pathological angiogenesis is manifested in proliferative retinopathy, hemangioma, psoriasis, and atherosclerosis and in the growth and metastasis of solid tumors. Angiogenesis is an intricate process that is regulated by many growth factors and cytokines. Angiogenesis-regulating factors can be classified into two groups based on whether they are involved in the activation or dissolution phase, which begins in the endothelium of pre-existing vessels, or the resolution phase, which results in differentiation of the newly formed capillaries. The dissolution phase of angiogenesis begins with expression or induction of proteolytic activities in otherwise quiescent endothelial cells, leading to breakdown of the cell junctions, increases in paracellular

4

M. A. Behzadian et al.

permeability, and degradation of the underlying basement membrane. The activated endothelial cells are thus set free to penetrate the surrounding tissue as they migrate and proliferate along a concentration gradient towards the source of angiogenic stimuli. Ischemia and hypoxia have been identified as major stimuli for angiogenesis in developing embryos, developing retinas, and growing tumors. A variety of angiogenic factors and cytokines including vascular endothelial growth factor (VEGF), angiopoietin-1, basic fibroblast growth factor (b-FGF), tumor necrosis factor alpha (TNF-alpha), transforming growth factor-beta (TGF-beta), and platelet derived growth factor (PDGF) have been identified in these tissues (for review, see1-4). In addition to the initial stimuli, which originate at distal sites, pro-angiogenic elements appear rapidly within the dissolution milieu. These include growth factors, pro-enzymes, and other serum proteins that leak from destabilized capillaries; extracellular matrix (ECM)-bound growth factors that are released or activated by proteases; and endothelial precursor cells that are recruited from the bone marrow or circulation. Depending on the tissue in which angiogenesis is taking place and whether the angiogenesis is physiological or pathological, other cellular elements in addition to endothelial cells may participate in the process. In the developing retina for example, astrocytes invade the retina from the optic nerve to guide the migrating endothelial cells.5-7 As endothelial cells elongate and assemble into a meshwork of interlaced cords, the resolution phase of angiogenesis begins with cessation of endothelial cell proliferation, induction of endothelial cell differentiation/lumen formation, and reconstruction of the vascular basement membrane. Redundant connections and excess branches are pruned, probably by a precisely localized apoptosis that involves leukocytes.8 The meshwork-like interconnecting tubes and tufts are reduced to a pattern of branching, bifurcated vessels with efficient directional blood flow. Critical to the stabilization of the newly formed capillaries is the recruitment of pericytes, which wrap around the vessels.9 A variety of angiogenic inhibitors have been implicated in the resolution phase of angiogenesis, including endostatin, angiostatin, thrombospondin, interleukin-1, interferon-beta, prostate-specific antigen, tissue inhibitors of metalloproteinases (TIMP), angiopoietin-2, pigment epithelial derived factor (PEDF), and TGF-beta. The role of TGF-beta in the resolution phase of angiogenesis has been studied extensively. In vitro, TGF-beta has been shown to inhibit endothelial cell proliferation and migration,10 whereas it promotes formation of capillary-like tubules in three-dimensional matrix gels.11,12 Studies by several groups have demonstrated that TGF-beta has a biphasic effect on endothelial cells. Depending on the concentration, it either inhibits or potentiates endothelial cell proliferation, invasion, or tube formation in three-dimensional collagen gels.13-15 Similarly, the dose-dependent,

1. Mechanisms of Retinal Angiogenesis

5

synergistic and/or antagonizing effects of angiopoietin (types 1 and 2) and VEGF have been the focus of numerous investigations.16-19 The seemingly paradoxical effects of the angiopoietins or TGF-beta observed in cultured endothelial cells are good testimony to the importance of in vitro approaches for understanding the mechanisms of cytokine actions in the angiogenesis process in vivo. For example, the biphasic function of TGFbeta in vitro suggests that it plays a dual role in vivo as well, acting as an activating factor to promote angiogenesis at the lower end of its concentration gradient near the sprouting vascular stalk and as a resolution factor to inhibit proliferation and promote differentiation at the upper end of its concentration spectrum near the site of ischemia/hypoxia. Conversely, a cytokine may act as an angiogenic activator at the sprouting site where it is released at high concentrations, but it may function as a ‘resolution’ factor at locations distal to the angiogenic sprout where it is present at lower concentrations. It is important to recognize that, in an in vivo setting, the dissolution/ activation phase of angiogenesis is closely followed by the resolution/ differentiation phase. This means that mitogenic and migratory factors, together with ECM-degrading enzymes, must be tightly regulated and locally constrained. The angiostatic, pro-differentiation factors, protease inhibitors, and other elements involved in reconstruction of ECM and recruitment of capillary-stabilizing mural cells should be able to function in close proximity to the activating ‘destabilizing’ factors. Much as the overexpression of activating factors may cause hyper-vascularization, deficiency (or mal-distribution) of resolution factors may be associated with leaky unstable capillaries and formation of dysfunctional vascular tufts in pathological situations. In other words, pathological angiogenesis can be characterized not only by excessive pro-angiogenic factors but also by an insufficient supply of anti-angiogenic, pro-differentiation factors.2-4 To better understand angiogenesis, our experimental models should enable us to dissect and separately manipulate the dissolution and resolution phases of this intricate process. In pathological angiogenesis, these two components occur in close spatial and temporal proximity, and, thus, are difficult to monitor. In order to model pathological angiogenesis, we should begin with the most basic elements and reconstitute the physiological and pathological conditions side-by-side and step-by-step. Endothelial cells are the essential element of angiogenesis. They form the lumen of blood vessels and function in a number of local roles, including control of vascular tone, provision of an anticoagulant surface, maintenance of the blood/tissue barrier and defense against inflammatory cells. Preparation of vascular

6

M. A. Behzadian et al.

endothelial cells in culture is the first step in developing the models necessary to characterize physiological and pathological angiogenesis.

1.2

Primary Cultures of Endothelial Cells

Jaffe and co-workers20 were probably the first to isolate endothelial cells on a large scale. They prepared and characterized human umbilical vein endothelial cell (HUVEC) cultures using a combination of morphologic, electron-microscopic, and functional assays. The cells were grown as a homogeneous population for up to 5 months or were subcultured for 10 serial passages, although cell growth rate was rather slow (doubling time of 92 hours). Using the same methods, Gimbrone’s group 21 prepared HUVEC cultures for studies of endothelial growth and migration and suggested the potential value of these cells as an in vitro model of angiogenesis. Large vessel endothelial cells from aorta or pulmonary arteries have been prepared in many laboratories using similar methods.22,23 The vessel can be filled with a solution of collagenase and incubated until the endothelial cells are released; alternatively, large vessels are cut open so that the interior endothelial layer can be removed by gentle scraping. Endothelial cells isolated from different areas of the vascular tree have diverse characteristics. Studies indicate that this diversity is due, in part, to micro-environmental influences. In culture, endothelial cells are capable of acquiring new properties depending on the characteristics of the plating surface and the culture medium.24,25 Milici et al. reported that bovine adrenal capillary endothelial cells cultured on plastic exhibited very low levels of diaphragmed fenestrations and almost no transendothelial channels as compared to cells grown on basal lamina.26 The ability of CNS endothelial cells to form a blood-brain barrier is thought to depend in large part on factors present in the CNS environment.27,28 That is, endothelial cells can be manipulated in culture to develop a specific phenotype and to satisfy the requirements of a particular functional assay. Despite this apparent plasticity of vascular endothelial cells, use of capillary endothelial cells is preferred for developing models of angiogenesis because the capillary vessel, not the large vessel, is the origin of new sprouts. Isolation of endothelial cells from capillaries has proven to be more difficult than isolation from large vessels. A large-scale preparation of capillary endothelial cells and pericytes from bovine cerebral cortex was reported first by Carson and colleagues in 1986.29 Microvascular endothelial cells were also isolated from rat epididymal fat pad and characterized for their growth on different substratum by Madri and Williams.30 Kern and colleagues isolated microvascular endothelial cells from human adipose tissue,31 and Marks and co-workers described an improved method of

1. Mechanisms of Retinal Angiogenesis

7

isolating human dermal microvascular endothelial cells from foreskin using percoll density gradient centrifugation. These authors examined the use of serum and other growth factor supplements for improving culture conditions.32 Mouse brain endothelial cells have been prepared by several groups.33,34 There are a number of good protocols for preparation of endothelial cells from retina. Buzney and Massicotte were probably the first to report that capillaries isolated from fetal calf retina give rise to endothelial cell colonies in culture.35 Frank and co-workers reported on growth of microvascular endothelial cells from kitten retina.36 Large-scale preparation of primary cultures of microvascular endothelial cells from bovine retina was reported by Bowman and co-workers37 and by Capetandes and Gerritsen.38 The latter group explained the advantage of using fibronectin coated dishes and plasma-derived serum supplement. The plasma-derived serum preparation, also designated as platelet-poor plasma, is a preferred medium supplement for endothelial cells because it does not support the growth of contaminating pericytes. Retinal endothelial cells have also been isolated from rhesus monkey,39 human,40 and mouse.41 Our laboratory has been using bovine retinal capillary endothelial (BRE) cells for several years.42-46 To isolate BRE cells, we follow the method of Bowman37 as improved by Capetandes38 and Laterra.47 The retinal tissue is homogenized, and small capillaries are collected over a nylon sieve (80 µm). The material retained by the sieve is briefly digested by collagenase and plated in dishes precoated with collagen and fibronectin. Using platelet-free serum favors endothelial cell growth over contaminating cells. However, if contaminating cells are overwhelming, individual endothelial colonies are isolated and pooled into a new sub-culture. In the initial culture, some colonies grow faster than the others. As a result, the subcultures may not represent the actual heterogeneity of the in vivo cell population. Nevertheless, depending on the purpose of experiment, one may compromise by using less homogenous cultures of early passage, or pure endothelial cells of later passage. Other parameters, such as cell density and the proliferative versus quiescent state of the cells, should be considered in such assays as cell proliferation, cell migration, or programmed cell death.

1.3

Naturally Transformed and Conditionally Immortalized Endothelial Cell Lines

Retinal endothelial cell cultures are most commonly prepared by isolation and enzymatic disruption of microvessels. The preparation is then cultured, and individual colonies of endothelial cells are isolated from the more

8

M. A. Behzadian et al.

aggressively growing pericytes and pooled into a pure endothelial subculture. Alternatively, the digested preparation is further enriched for endothelial cells by gradient centrifugation or selection by the use of endothelial cell surface-specific markers. Laboratory mice are a good source of retinal endothelial cells. However, isolation and maintenance of a primary culture of retinal or cerebral endothelial cells from small rodents is difficult due to the limited tissue supply. On the other hand, the wealth of information about the genetics and developmental biology of laboratory mice and the availability of transgenic mouse models provide a strong motivation for improving methods of preparing durable cultures of microvascular endothelial cells from the mouse. Genetically modified mouse strains can now be used to study the functional role of a particular gene (or gene mutation) in the manifestation of vascular disease in retina. Once such a causal relationship is established in vivo, cultured vascular cells from the transgenic mouse offer an excellent opportunity to analyze mechanisms of the gene’s regulation and function. Several laboratories have isolated clones of spontaneously transformed endothelial cells with a wide range of characteristics from mouse or rat retinas. Others have transformed primary endothelial cultures by introducing exogenous oncogenes. The advantage of the immortalized cell lines is that there is no need for a periodic preparation of primary cultures. The disadvantage is that some morphological or functional characteristics, such as reduced requirement for serum supplement and growth factors, lack of response to cytokines, or compromised barrier function, may render them unsuitable for a particular experimental protocol. For example, DeBault and co-workers cloned an endothelial cell line from mouse brain, designated ME-2, that retained many properties of primary cultures, including growth characteristics and specific cell surface antigens for up to 40 passages before becoming senescent.48 On the other hand, Robinson and co-workers isolated a spontaneously transformed clone of mouse brain endothelial cells, designated Ten, that exhibited characteristics of transformed cells, including growth in serum-free medium, anchorage-independent growth, tumorigenicity in nude mice, and lack of contact inhibition. Nevertheless, Ten cells maintained endothelial cell markers and responded to EGF and PDGF mitogenic activities.49 Comparing the preparations of DeBault and Robinson, it appears that there is neither a simple formula for preparation nor a standard inclusive definition of a cloned endothelial cell line that is qualified to serve as an experimental model for all kinds of assays. A cell line with reduced serum requirements is not the best choice for testing mitogenic factors, and a clone lacking contact inhibition would not serve well for studies of cell migration or permeability. One important criterion for the general suitability of cloned

1. Mechanisms of Retinal Angiogenesis

9

endothelial cultures would be their potential to senesce after a number of passages. Any cell preparation, whether it originated from a single colony or from several pooled colonies, can be considered as closer to primary when, over a number of passages, the growth rate declines and some large, multinucleated (fried-egg like) non-proliferating cells appear in the culture. Of course, mishandling cultures can also give rise to giant non-proliferating cells. Retinal endothelial cells fed on a regular basis and transferred at high density (at a ratio of 1:3 surface areas) can be maintained for up to 12-14 passages before senescent cells begin to appear in the culture. Moreover, depending on how successful the initial plating is, the culture may better represent the primary population of cells if only the contaminating cells are removed. On the other hand, naturally transformed endothelial cell colonies may be isolated and pooled from passages of the primary culture. These cells, while not representing the primary population of cells, may be used for a particular experiment if they are carefully characterized. Finally, one may modify the culture condition to optimize the use of transformed cell lines for a particular assay. The paracellular leakiness of transformed endothelial cells, which can be readily tested by measuring the transendothelial electrical resistance (TER), has been suppressed by the use of astrocytes as co-culture or astrocyte-conditioned medium as supplement.50,51 A number of brain capillary endothelial cell lines have been isolated from mouse or rat carrying the temperature sensitive version of the Simian Virus40 large tumor antigen (SV-40 Tag). SV-40 Tag is an oncogene that causes cells to grow continuously in the absence of growth factors. At the permissive temperature (33 °C), the cells show transformed characteristics and replicate continuously. When switched to 37-39 °C, the expression of SV-40 Tag is halted, and the cells exhibit characteristics of primary cultures. Using magnetic beads coated with anti-platelet/endothelial cell adhesion molecule-1 (PECAM-1) antibody, Su and co-workers isolated endothelial cells from wild-type and thrombospondin-1 deficient (TSP1-/-) mice carrying the SV-40 Tag gene (transgenic immorto-mouse).41 Both cell lines expressed endothelial cell markers, but TSP1-/- cells were deficient in their ability to form capillary-like networks on Matrigel, confirming the known antiangiogenic role of thrombospondin. This method would be extremely useful for comparing the physiological role of specific genes in endothelial cell function. Magnetic beads coated with PECAM-1 antibody have been used for isolating retinal endothelial cells from Lewis rats,52 and retinal pericyte cell lines expressing the SV-40 Tag have also been isolated from the SV-40 Tag rat strain.53 The cells grew exponentially at the permissive temperature of

10

M. A. Behzadian et al.

33 ºC, but became quiescent within 2 days when shifted to 37 ºC. Several immortalized brain capillary endothelial cell lines also have been established from transgenic mice harboring the SV-40 Tag.54,55 Reversion of immortalization seems to be necessary in some experimental models where the immortalizing gene may influence the experimental outcome.56 In H-2Kb-tsA58 transgenic mice, the temperature sensitive SV-40 Tag gene is controlled by the H-2k(b) class-I histocompatibility promoter, which is inducible by INF-gamma. Using this mouse, Lindington and co-workers established a cardiac endothelial cell line, which grew exponentially at the permissive temperature and in the presence of INF-gamma.57 At 38 ºC and in the absence of INF-gamma, the cells stopped growing and became responsive to basic FGF, VEGF, and EGF. By cross-breeding this same mouse with the uPAR-/- mouse, we have been able to isolate uPAR-/- brain endothelial cell lines; characterization of these cells is underway in our laboratory (Behzadian et al., unpublished). Urokinase plasminogen activator and its receptor (uPA/uPAR) have been implicated in the regulation of endothelial barrier function and endothelial cell migration. Isolation of mutant endothelial cells from transgenic uPAR-/- mice provides an important in vitro model for studying the function of the uPA/uPAR system in retinopathy.

1.4

Endothelial Precursor Cells

The process of angiogenesis, defined as formation of new blood vessels by sprouting from pre-existing blood vessels, has been classically distinguished from vasculogenesis, which is the mobilization and assembly of mesenchimal endothelial precursor cells into vascular structures. This distinction was based not only on differences in the way the two processes occur, but also on the notion that vessel growth during embryonic development mainly involves vasculogenesis, whereas postnatal neovascularization occurs mainly by angiogenesis. However, recent evidence indicates that vasculogenesis plays a significant role in postnatal neovascularization.58-60 The identification of endothelial precursor cells in the adult bone marrow and circulating blood61 led to the discovery of endothelial precursor cells in sites of neovascularization and changed our understanding of postnatal vessel growth and repair.62 Active recruitment of endothelial precursor cells has been demonstrated in the ischemic retina,63-67 suggesting that therapies targeting these precursor cells will help in blocking pathological neovascularization in retina. Studies of bone marrow-derived hematopoietic stem cells (HSCs) and endothelial progenitor cells are thoroughly discussed elsewhere in this volume. Our focus is on potential use of endothelial precursor cells for in

1. Mechanisms of Retinal Angiogenesis

11

vitro studies. Using cultured precursor cells, one can study the effects of chemokines, pharmaceutical reagents, and cellular mediators on precursor cell attachment, migration, differentiation, and resistance to stress conditions. This information will contribute to our understanding of the biology of precursor cells as well as their potential therapeutic use. Furthermore, genetic manipulation of endothelial precursor cells, as achieved by gene transduction, can be used to selectively promote the expression of pro- or anti-angiogenic factors at sites of tissue injury.68 Another important application for precursor cell cultures is to find new ways of enhancing their growth rate in order to obtain larger numbers of cells for transplantation in conditions where vascular growth is needed. Sources of endothelial precursor cells include bone marrow explants, umbilical cord, and peripheral blood. The cells can be isolated from the mononuclear cell fraction of peripheral blood by density gradients and then identified based on the expression of specific surface antigens including CD34, VEGF receptor-2 (VEGFR-2, Flk-1), and the orphan receptor AC133.69,70 This particular pattern of surface antigens is also present on HSCs, demonstrating that endothelial precursor cells are related to HSCs, with which they share a common progenitor, the hemangioblast.71 The expression of surface antigens is strictly dependent on the stage of endothelial precursor cell differentiation. For example, immature circulating endothelial precursor cells express AC133, but this antigen is not found on the surface of more mature “committed” endothelial precursor cells.72 The phenotypic switch of endothelial precursor cells to the mature, terminally differentiated endothelial cell phenotype can be monitored by the appearance of other specific surface antigens such as the von Willebrand factor.73 Endothelial precursor cells are specifically sensitive to bioactive peptides, including VEGF, insulin-like growth factor-1 (IGF-1), and hepatocyte growth factor (HGF), and an appropriate balance of different cytokines is critical for preventing their differentiation into mature endothelial cells. The maintenance of their undifferentiated state in vitro is essential to the preservation of their self-renewal abilities and stem cell-like properties. Specific genetic analysis of endothelial precursor cells has identified characteristic profiles of gene expression common to many stem cells; these are thought to be involved in the maintenance of their so-called “stemness.”74,75 Cultured endothelial precursor cells have been used in ex vivo models of hind limb and myocardial ischemia.59 Recent studies using these models have shown that endothelial precursor cells can display many of the morphological and functional characteristics of mature endothelial cells, such as formation of vascular-like structures in Matrigel.76 However their

12

M. A. Behzadian et al.

replicative capacity and their stress resistance are greater than mature endothelial cells. Endothelial precursor cells also appear to reach senescence at a much slower pace than endothelial cells, and this effect has been explained by enhanced antioxidant abilities as well as reduced telomerase activity in these cells.77,78 Drugs’ effects on endothelial precursor cells have also been studied in tissue culture models. For example, statins, which inhibit 3-hydroxy-3methylglutaryl coenzyme A (HMG-CoA) reductase, have been shown to extend the life span of endothelial precursor cells and to enhance their mobilization and incorporation at sites of vascular injury in a model of cardiovascular disease.78,79 Studies conducted in cultured endothelial precursor cells have clarified that statins’ effects are mediated by the activation of the PI3-kinase/Akt signaling pathway. Finally, the antibiotic rapamycin has been shown to induce endothelial precursor cells’ apoptosis and to inhibit their ability to differentiate as mature endothelial cells, partly explaining the anti-angiogenic properties of this drug and supporting its use in preventing pathological neovascularization.80

1.5

Pericytes

Blood vessels are formed by two cell types: the endothelial cells that line the vascular lumen and the mural cells (pericytes or smooth muscle cells) that wrap the endothelium on the abluminal side of the vessel wall and share the vascular basement membrane. The second phase of the angiogenic process (resolution) involves recruitment and proliferation of mural cells and their attachment to the newly formed capillaries, leading to stable mature vessels with proper directional blood flow.81-83 In retinal capillaries, the mural cells are pericytes. Therefore, substances that stimulate migration or proliferation of pericytes, such as TGF-beta, PDGF, and angiotensin, could also be potentially involved in regulating retinal neovascularization. Studies in the developing retina have shown that immature vessels that lack pericytes degenerate when exposed to hyperoxia, whereas mature blood vessels with pericytes are resistant to oxygen-induced degeneration.84-87 To explain the mechanism by which pericytes protect the retinal vasculature, Shih et al. showed that the TGF-beta expressing pericytes are specifically attached on vessels that are resistant to oxygen-induced dropout. Their in vitro studies show that TGF-beta induces VEGF and VEGF-receptor 1 (VEGFR-1, flt-1) expression in retinal endothelial cells.88 The VEGFR-1specific ligand, placental growth factor, acts as a survival factor for endothelial cells in culture. Understanding the mechanism of pericyte/endothelial-cell interaction is important in the context of oxygen-induced capillary dropout in retinopathy

1. Mechanisms of Retinal Angiogenesis

13

of prematurity (ROP). A number of in vitro studies have significantly advanced our knowledge of the mechanisms of cell-cell interactions and the factors involved in the ROP pathology. Orlidge and D’Amore used coculture models for a comprehensive in vitro study, which showed inhibition of endothelial cell growth by direct contact with pericytes or smooth muscle cells.89 Co-culture models can also be used to study the role of endothelial cell contact in differentiation of mesenchymal cells to pericytes/smooth muscle cells.90 Pericytes can be isolated by a method similar to that used for preparation of endothelial cells except that the microvessel preparation is directly plated on a gelatin-coated dish without any enzyme treatment.91 Pericytes are identified by their slow growth and robust cytoskeleton structure. Markers include alpha actin and a pericyte-specific ganglioside.92

2.

ASPECTS OF ENDOTHELIAL CELL FUNCTION STUDIED IN CULTURE

2.1

Permeability

In retina and brain, capillary endothelial cells form the inner blood-retina barrier (BRB) and blood-brain barrier (BBB), respectively, to regulate the exchange of molecules between blood and neural tissues. In both situations, tight junctions prevent the free diffusion of substances from the blood to neural tissues.93,94 It has been suggested that some mechanisms may function differently in the BRB to protect the retina from light-induced oxidative stress.95 Vascular endothelial cell dysfunction and breakdown of the BRB occur in a number of disease conditions including diabetic retinopathy, macular edema, hypertensive retinopathy, branch vein occlusion, and others.93 Vascular leakage contributes to disease progression by inducing edema and tissue damage. At the same time, vascular hyperpermeability is the critical first step in the angiogenic process in that extravasation of plasma proteins provides a milieu that favors neovascularization.96 Activation of plasminogen and matrix metalloproteinases (MMPs) plays a key role in this process by inducing degradation of the ECM and release of growth factors, which stimulates the migration and proliferation of endothelial cells.97 In vitro models have been used to study the regulation of vascular endothelial cell barrier function and to allow experimental manipulations and observations not possible with intact animals. In the Transwell dual chamber

14

M. A. Behzadian et al.

model, monolayers of endothelial cells grown on a porous membrane are situated in a culture dish such that two separate compartments are formed. The upper chamber represents the vascular luminal compartment, and the lower chamber represents the abluminal compartment. Usually, the flux of solutes of different sizes (e.g. sucrose, sodium fluorescein, fluoresceinlabeled albumin or dextran) from the upper to the lower chamber is monitored at timed intervals.98-102 This procedure has been used successfully to study permeability and transport mechanisms of retinal capillary endothelial cells103-105 and to analyze the effects of VEGF, hydrocortisone, or high glucose on the permeability of retinal endothelial cells.106-108 This model has also been used to investigate the functional role of extracellular proteinases such as MMPs and the urokinase/urokinase-receptor (uPA/uPAR) system as mediators of the TGF-beta or VEGF-induced breakdown of the BRB. In vitro studies have shown that astrocytes and Muller glia express TGF-beta in latent form and that TGF-beta becomes activated when the cells are incubated under hypoxia conditions.42,109 In separate experiments, retinal endothelial cells were found to express the basement membrane degrading gelatinase MMP-9 when treated with TGFbeta or cocultured with Muller glial cells or astrocytes. Both TGF-beta and MMP-9 increase retinal endothelial cell permeability, and anti-MMP-9 antibody or TGF-beta latency-associated peptide abrogated the TGF-beta effects.45 During retinal disease, glial cell production of active TGF-beta may contribute to breakdown of the blood-retina barrier by stimulating endothelial cell MMP-9 production. Another prominent extracellular proteinase system that works in concordance with the MMPs is the uPA/uPAR system. This system has also been shown to have a key role in triggering endothelial cell hyperpermeability. Studies of VEGF’s effects on permeability have shown that treatment with VEGF or uPA increases permeability of retinal endothelial cell monolayers, but with different kinetic response patterns.46 The uPA-induced permeability increase is rapid and stable for over six hours. In contrast, the permeability effect of VEGF is biphasic, with an early and transient permeability increase followed by a delayed and sustained permeability increase starting 4-6 hours post VEGF treatment and lasting for 24 hours. Moreover, this delayed phase is accompanied by a decline in transcellular electrical resistance (TER) of the monolayer, which was not seen with the initial permeability increase. It has been shown that the early permeability increase is transcellular and is mediated by cell membranederived caveolae.44 The late phase of the VEGF-induced permeability increase as well as the entire uPA-induced permeability response was shown to involve redistribution of junction proteins and, therefore, is very likely to involve alterations in paracellular permeability through the cell junctions.

1. Mechanisms of Retinal Angiogenesis

15

Measurement of TER across endothelial cell monolayers has been taken as an indicator of paracellular permeability barrier function and is usually done using hand-held chopstick electrodes. Recently, instrumentation referred to as ECIS (Electrical Cell-Substrate Impedance Sensing, Applied Biophysics, Troy, New York) has been introduced, which can measure both paracellular resistance and the average cell-substrate distance. In ECIS, cells are plated in special 8-well chamber slides equipped with gold plated electrode arrays. The electric current passing through the cell monolayer covering each electrode is measured independently in each chamber. The advantage of the ECIS method over previously used manual methods is that electrical resistance is monitored continuously and in real time, before, during, and after treatments applied to multiple independent chambers. Studies comparing ECIS measurements with those done using the Transwell model have shown that the resistance caused by cell-substrate contact substantially influences the TER data and that the extra resistance due to cell-substrate spaces depends on both cell type and properties of the polycarbonate filter system.110 Studies using the ECIS system should be useful in dissecting the potential contribution of such differences in cellsubstrate attachment to the TER alterations observed in studies of disease models where cell-substrate attachment is likely to be altered (such as diabetic retinopathy). The ECIS system can also be used for in vitro analysis of cell migration using a “wound-healing” assay in which cell migration is assayed following mechanical disruption. In manual assays, a scrape is made in the cell layer, and the advance of the cells into the wound is assessed by microscopy. Using the ECIS system, the wounding can be accomplished electrically by using high voltage to cause severe electroporation and death of the cells in direct contact with the electrode surface. After this treatment, the migration of the surrounding cells onto the electrode surface can be assayed in real time by charting the recovery of electrical impedance. This procedure has been shown to be highly reproducible and quantitative and to provide data similar to that acquired with traditional measurements.111 Studies in progress using this system with retinal endothelial cells indicate that high glucose increases endothelial cell permeability and migration through induction of uPA/uPAR activation in endothelial cells (Behzadian et al., unpublished).

16

2.2

M. A. Behzadian et al.

Extracellular Matrix Proteolysis

As has been explained above, extravasation of blood proteins from leaky vessels provides a favorable environment for the initiation of angiogenesis. In order for angiogenesis to occur, the activated endothelial cells must first detach from the vessel wall and penetrate their basement membranes and the surrounding ECM. The complex composition of the microvascular ECM implies that multiple highly specialized enzymes are required for its degradation. Proteolytic enzymes such as uPA and MMP collagenases and gelatinases112-115 are produced, bound, and activated by endothelial cells116-118 and mural cells.119,120 These enzymes degrade the basement membrane and the interstitial stroma of the surrounding tissue in the region of capillary sprouts. Because most of the matrix-degrading proteases are secreted as latent pro-enzymes, the physiological activation, rather than production of the enzyme, is the critical controlling point. Activation of pro-enzymes and zymogens occurs by a cascade of autocatalytic, reciprocal interactions. In the case of pro-MMP-9, this activation depends critically upon binding of uPA with uPAR at the cell membrane (see Figure 2).

Reciprocal Zymogen Activation uPA

Plasminogen

Pro-uPA + uPAR

Plasmin

Pro-MMP9

MMP9

Figure 1-2. Reciprocal zymogen activation. Binding of uPA with uPAR initiates the activation of a proteolytic cascade and focuses ECM proteolysis at the plasma membrane.

1. Mechanisms of Retinal Angiogenesis

17

The MMPs are a family of extracellular proteolytic enzymes that are mainly involved in tissue remodeling. MMP substrates include all forms of collagen and a variety of other ECM components, including ECM bound cytokines and growth factors. MMPs have been found in virtually every tissue of the body under conditions of both health and disease. In retina, MMP activity has been associated with numerous disease conditions, including age-related macular degeneration, proliferative diabetic retinopathy, glaucomatous optic nerve head damage, vitreoretinopathy, and others (for review see121,122). The role of the MMPs in angiogenesis has been investigated using an in vitro rat aortic ring model. Inhibition of microvessel outgrowth in this model by MMP inhibitors demonstrated the requirement of MMP activity for angiogenesis.123 These studies also showed that the profile of MMP expression depends both on matrix composition and exogenous growth factors. For example, the gelatinase MMP-2 and the stromelysin MMP-3 were present at high levels during vessel formation in fibrin matrix, whereas the stromelysin MMP-11 and membrane-type-1-MMP were expressed in collagen culture. Basic FGF induced upregulation of gelatinases (MMP-2 and MMP-9), stromelysins (MMP-3, MMP-10 and MMP-11), and the interstitial collagenase MMP-13, whereas VEGF induced expression of MMP-2 only. Such in vitro models provide a basis for developing MMP inhibitors for use as anti-angiogenic therapy. In vitro studies have indicated that most of the MMPs are induced in the same fashion and by a large number of cytokines or growth factors including IL-1-beta, TNF-alpha, PDGF, EGF, TGF-beta, NGF, and others. However, all MMPs are produced in latent pro-enzyme form. Multiple microenvironmental factors contribute to the temporal and spatial regulation of MMP activation, which is the rate-limiting step in their function. Plasmin and uPA have been implicated in physiological activation of many of the collagen-degrading MMPs, including MMP-9, which has been associated with pathological angiogenesis in retina.124,125 As has been explained above, activated endothelial cells express uPAR, which plays a key role in VEGFinduced increases in paracellular permeability.46 Since both plasminogen and pro-MMP-9 bind the cell membrane, uPA binding to uPAR provides a mechanism for the cell to focus a cascade of proteolytic activity at the cell surface. This localization is so precise that it may restrict the enzyme activity to the site of membrane contact with the ECM. Under normal physiological conditions such as wound healing and tissue remodeling, proteolytic activities are precisely controlled and localized on the cell surface. However, during conditions of pathological

18

M. A. Behzadian et al.

neovascularization, excessive proteolytic activities may contribute to the formation of disorganized, unstable and hyperpermeable vascular tufts, which fail to maintain appropriate levels of tissue oxygenation and nutrient delivery. In other words, excessive proteolysis is incompatible with normal capillary morphogenesis.126 Using a three-dimensional fibrin gel, Montesano and co-workers showed that neutralization of excess proteolytic activity plays a permissive role in angiogenesis and other invasive processes by preventing uncontrolled matrix degradation.127 A number of in vitro models have been devised to investigate production of proteolytic enzymes in vascular endothelial cells and to determine their role in cell migration. Gross et al. found that the capillary endothelial cells produced 5-13 times the basal levels of collagenase activity in response to tumor promoter 12-Otetradecanoyl phorbol-13-acetate (TPA), whereas aortic endothelial cells and fibroblasts showed a minimal response to TPA.128 Later, using crude angiogenesis stimulating factors, these authors confirmed that induction of plasminogen activator and collagenase activities are limited to capillary endothelial cells.129

2.3

Cell Migration

During angiogenesis, proteolysis of the ECM sets the stage for the directional migration of endothelial cells along a concentration gradient of pro-angiogenic factors towards the sites of tissue ischemia. Endothelial cell migration involves temporary attachment and detachment of cell surface adhesion molecules to the ECM. This process is influenced by a number of microenvironmental factors and is associated with changes in adhesion molecules on the cell membrane and rearrangement of cytoskeletal filaments within the cells. In the earliest study to distinguish between the migration and proliferation of endothelial cells during angiogenesis, Schoefl observed that endothelial cell migration was the initiating, and probably the ratelimiting, event in regeneration of capillaries after tissue injury.130 Similarly, Ausprunk and Folkman showed that migrating endothelial cells initiate the extension of capillary sprouts toward the source of the angiogenic factors.131 Ischemia/hypoxia is thought to be a major angiogenic stimulus. The chemotactic factors attracting endothelial cells toward the ischemic tissue have not been fully characterized, but are thought to include VEGF. In vitro models of cell migration allow for the study of the specific angiogenic factors and their interaction with cell adhesion molecules involved in cell migration. In the wound closure assay of cell migration, endothelial cells are grown to confluence in growth medium and then switched to serum-free medium prior to addition of the factor to be tested. The monolayer is wounded with

1. Mechanisms of Retinal Angiogenesis

19

an object such as a sterile wood stick or cell scraper. The culture is incubated and periodically monitored under a microscope as the cells move onto the denuded area.132 The effect of fibrin on the migration of bovine aortic endothelial cells was investigated by wounding the confluent monolayer and counting the number of cells crossing the wound border per unit time.133 To assay for migration independent of proliferation, wound-induced proliferation of endothelial cells is inhibited by mitomycin C. Directed migration (chemotaxis) can be assayed using modified Boyden chambers.134 For these studies, endothelial cells are seeded on porous polycarbonate membranes. Membranes are mounted on plastic rings to form small chambers and to fit the wells of a multi-well tissue culture plate. Test substances are added on the opposite side of the membrane. After a period of stimulation, cells on the attachment side of the membrane are scraped off, and membranes are stained for microscopic analysis. Cells that have migrated through the membrane are counted, and data are expressed as number of cells per high power field. Using this model, Nadal and coworkers showed that angiotensin II, via its AT-I receptor, acts as a chemotactic factor and stimulates migration of retinal microvascular pericytes.135

2.4

Cell Proliferation

Following detachment and migration of activated endothelial cells, cell proliferation is the next step in the angiogenic process. There are a number of well-established techniques for evaluating mitotic activity and proliferation of endothelial cells in culture. Some protocols call for synchronizing the cell population by serum starvation or by allowing the cells to grow to confluence in order to render them quiescent before they are treated with mitogens. The cells, arrested in G-0, are then stimulated to enter the S-phase, and they are monitored for DNA synthesis or mitochondrial enzyme activity. Alternatively, counting cells is a direct and simple method in which cells are seeded at low density in normal growth medium and then switched to a serum-free or low serum (0.1-0.5% FBS) medium with or without the regulatory growth factors. Representative sub-confluent cultures are counted at daily intervals. Cells are removed by trypsinization and counted with a hemocytometer or by using a coulter counter.136,137 Incorporation of thymidine into newly synthesized DNA can also be used to evaluate the mitogenic response of synchronized quiescent cells to various agents. After treatment with mitogen, cells are pulsed by exposure to [methyl-3H]-thymidine for 0.5 to 1 hour, washed thoroughly, and then incubated in normal medium for another 1-3 hours. The monolayer is then

20

M. A. Behzadian et al.

covered and rinsed with a chilled solution of trichloroacetic acid (TCA). The fraction of radiolabeled nucleotide that is incorporated into DNA is TCAinsoluble and remains in the dish. The TCA-insoluble material can be removed by NaOH and quantified by liquid scintillation counting.138 This method has been used to show that human retinal extracts stimulate thymidine uptake in bovine aortic endothelial cells139 and that human growth hormone stimulates thymidine uptake in human retinal microvascular endothelial cells.140 Thymidine incorporation has also been used to show the mitogenic effects of basic FGF and VEGF in retinal endothelial cells and pericytes under normal or hypoxic conditions.141 VEGF and basic FGF increased 3H-thymidine incorporation by both cell types, an effect that was more pronounced under hypoxic conditions. Moreover, it was found that the proliferation of pericytes was stimulated to a greater extent by basic FGF relative to VEGF. Incorporation of bromo-deoxyuridine (BrdU) is another method for measuring DNA replication in response to mitogenic stimuli. Cells are grown in 96-well microtiter plates to ~50% confluence. The cells are then exposed to particular test agents, and mitotic activity is quantified by using anti-BrdU antibody and a colorimetric substrate reaction. Another simple method of detecting cell proliferation is to determine cell density in culture using the DNA-enhanced fluorescence assay. Using microtiter plates, this method allows a large number of agents to be tested simultaneously for their effects on cell growth. However, when positive agents are identified by this method, the efficacy of the selected factors must be confirmed by a more direct method such as cell counting. Briefly, fixed cells are labeled with DNA stains such as 4’6-diamidino-2-phenylindole (DAPI), propidium iodide (PI), ethidium bromide, or Hoechst 33342. Then fluorescence intensity is quantified using a fluorometer. This procedure has been used for determining serum-stimulated growth of smooth muscle cells and mitogen-induced growth of endothelial cells.142 The advantage of this method is that the fixed cells can be stored for prolonged periods until all tests are completed, thus allowing time-course proliferation assays with minimal inter-assay variations. A colorimetric cell proliferation assay, referred to as MTS or MTT assay, is based on the ability of living cells to take up thiazolyl blue and convert it into dark blue formazan. The reaction is driven by mitochondrial succinate dehydrogenase activity, which can be correlated with cell density.143,144 The assay has been used to show the mitogenic effects of fibronectin fragments on human retinal endothelial cell proliferation.132,145

1. Mechanisms of Retinal Angiogenesis

2.5

21

Tube Formation

The resolution/differentiation phase of the angiogenic process is characterized by the cessation of endothelial cell proliferation, followed by alignment and differentiation into cords of lumenized vessels. This step of the angiogenic process can be modeled in vitro by monitoring endothelial cell alignment and lumenization in a three-dimensional ECM environment. For this “tube formation” assay, microvascular endothelial cells are grown on a three-dimensional gel consisting of type I collagen,146 Matrigel, or fibrin.147 Cells are grown for 2-5 days in the gels in basal medium and different stimuli are added. Cells invade the matrix and form cords and tubelike structures. Randomly selected fields are photographed by phase-contrast microscopy. The total length of sprouts and numbers of branches per field is measured. Lumenization of the tubes can be verified by microscopic analysis of tissue sections. The tube formation process can be induced by treatment of cells with tumor promoters, basic FGF,148,149 or other angiogenic factors. In a study of bovine retinal endothelial cells,150 cells plated within a collagen gel matrix self-associated to form three-dimensional meshworks. This morphogenesis was accomplished by cell migration and did not involve cell proliferation. By contrast, retinal pericytes and smooth muscle cells divided and remained homogeneously distributed when plated within a collagen gel matrix. Moreover, it was found that endothelial migration in collagen gels was induced more effectively by VEGF than by basic FGF and that VEGF and basic FGF have synergistic effects on cell invasion.151 The Matrigel tube formation assay has been used to compare the angiogenic properties of retinal endothelial cells isolated from wild-type and thrombospondin-1deficient mice. Retinal endothelial cells from wild-type mice formed capillary-like networks on Matrigel, whereas the ability of the retinal -/endothelial cells from TSP1 mice to form capillary-like networks was 41 severely compromised. Studies of tube formation in type I collagen have shown that a line of retinal endothelial cells formed capillary-like structures in response to apelin, an endogenous ligand for the orphan G proteincoupled receptor, whereas endothelial cells isolated from human umbilical vein (HUVECs) did not.152 During angiogenesis, the final pattern of blood vessel formation is governed by the concurrent and dual regulation of endothelial cell morphogenesis and regression.131 Endothelial apoptosis has been suggested as a major mediator of vascular regression during normal developmental or pathological vaso-obliteration.153-155 Studies using the tube formation assay have demonstrated that capillary morphogenesis in vitro is associated with apoptosis and that inhibition of TGF-beta signaling inhibits this process.156

22

M. A. Behzadian et al.

A variation of the tube formation assay for analysis of capillary regression was developed by Davis et al.157 The addition of plasminogen to threedimensional collagen matrices was found to result in activation of MMPs, collagen proteolysis, and capillary regression.

2.6

Retinal Explants

Organ cultures of intact retinas, partially dissected retinas and retinal slices have been used extensively for research on retinal differentiation, synaptic organization, cell and neurite outgrowth and cell-cell interactions between neurons and glial cells (for review see158). These explant models have the advantage of preserving near normal tissue architecture of the retina in situ. Although these differentiated cultures cannot be propagated in vitro, they can be maintained for periods of days or even weeks. The explant culture models have the disadvantages of greater experimental variability as compared with cultured cells and are difficult to use for some quantitative studies due to variations in tissue geometry and cellular composition. However, the cells in explanted tissues have the advantage of retaining to some extent in vivo histological and biochemical features that are commonly lost when isolated cells are propagated in culture. The use of retinal explants for studies of retinal angiogenesis has been limited. However, Knott and colleagues have described a human model of retinal angiogenesis159 based on refinement of a bovine retinal explant model developed previously by the same group.160 In this preparation, a 4 mm diameter disc of retinal tissue is placed within a fibrin matrix in medium containing 2.5% platelet-poor plasma and monitored by light microscopy for 1 to 14 days. Immunostaining analysis of tissue sections using antibodies against von Willebrand’s factor, glial fibrillary acidic protein, the macrophage/microglial cell marker (CD68), and the cell proliferation marker Ki-67 demonstrated that vascular growth during the in vitro incubation was correlated with activation of glial and microglial cells. Microscopic analysis of the intact tissue explants revealed obvious growth of vessels from the tissue into the surrounding matrix within 3 days of culture. Immunolocalization of von Willebrand’s factor showed an increase in the number and size of vessels within the inner nuclear layer of the tissue explants. Localized expression of endogenous VEGF was evident after 3 days in culture and was associated with angiogenic growth and glial cell activation as well as with the appearance of immunoreactivity for the monocarboxylate transporter-1 within the retinal endothelium. The expression of this transporter in the retinal endothelium could be attributed to the ability of the endothelium to respond to the demands of glucose metabolism and consequent lactate production in the ischemic retina.159

1. Mechanisms of Retinal Angiogenesis

23

Although levels of tissue oxygenation were not assessed in this study, it is likely that the increase in VEGF expression and retinal angiogenesis occurred secondary to a condition of relative hypoxia consequent to the thickness and lack of vascular perfusion in the explanted retina. Further study in this model may be useful for clarifying the combined effects of ischemia and fluctuating glucose concentrations on pathological retinal angiogenesis as seen in diabetes. This model may also be useful for comparing the specific patterns of vascular outgrowth and angiogenic sprouting that occur in retina with those that have been observed in other three-dimensional models of angiogenesis, such as the rat aortic ring model (see below). Such studies should help to answer the question of whether or not retinal glia and microglial cells have a specific role in retinal angiogenesis, as has been suggested based on results of previous studies using animal models in vivo and co-culture models in vitro.

2.7

Aortic Ring Explants

The rat aortic ring model has proven to be a highly useful method for analysis of the angiogenic process.161-163 In this assay, rat aortic rings embedded in collagen gels have been shown to give rise to a network of branching microvessels composed of a properly polarized monolayer of endothelial cells surrounded by a discontinuous coating of mural cells (pericytes/smooth muscle cells).162 The angiogenic response can be stimulated with angiogenic factors or blocked with angiogenic inhibitors. This model, which was first described in 1982164 and was later modified as a quantitative assay in 1990,161 is now widely accepted as a cost-effective and convenient method of assaying angiogenesis. While many studies have used the rat aortic ring model to test inducers and inhibitors of angiogenesis,165 its use in studies related to retinal angiogenesis has been limited. In a recent study of the role of VEGF in the guidance of angiogenic sprouting, Gerhardt and colleagues used the rat aortic ring model to show that stalks of vascular sprouts are composed of highly migratory cells and that the tips consist of specialized endothelial cells, which extend numerous filapodia in the direction of their migration.166 Time-lapse imaging revealed protrusion and retraction of lamellipodia from single endothelial cells at the tips of the angiogenic sprouts. The endothelial cells at the sprout tips were negative for mitogenic markers (Ki-67 or phospho-histone staining), whereas conspicuous cell proliferation was seen in the endothelial cells in the vascular stalk behind the advancing tips. This same pattern was evident in the sprouting vessels of the developing retina, indicating that angiogenesis in the aortic ring mimics that in the developing

24

M. A. Behzadian et al.

retina. Studies using this model should be helpful in defining the mechanisms that guide the misdirected cell migration during the angiogenic sprouting process that leads to subretinal and vitreoretinal neovascularization during age-related macular degeneration and diabetic retinopathy.

2.8

Molecular Strategies for Studies of Angiogenesis

Molecular approaches to identify and manipulate the expression of genes relevant to angiogenesis have taken good advantage of in vitro models. High-throughput screening technologies such as differential display, serial analysis of gene expression (SAGE), and microarray have helped to determine endothelial-specific genes and the mechanism of their regulation under various diseases conditions. Functional analysis of genes involves three primary approaches: gain of function, loss of function, and gene silencing. Gain of function studies use gene transfer by viral vectors that allow overexpression of the gene. Conversely, vectors overexpressing a dominant-negative gene can be used for loss of function studies. Delivery of inhibitory RNAs or antisense oligonucleotides can be used to silence gene expression and to determine the role of a specific gene product. In studies of retinal vascular development in mice, Adini and colleagues showed that the deletion of the small GTPase RhoB resulted in retarded retinal vascular development. Inhibition of RhoB in neonatal rat retina using farnesyl transferase induced vascular endothelial apoptosis. To confirm their in vivo data, the authors used both antisense oligonucleotides and dominant-negative RhoB expression to specifically reduce RhoB expression in a primary endothelial cell culture model. These treatments inhibited Akt survival signaling and tube formation and induced apoptosis, confirming a specific role of RhoB in endothelial cell survival.167 Recently, Oshima and co-workers demonstrated that chondromodulin (ChM-I), a cartilage-derived factor that inhibits angiogenesis, is expressed in both cartilage and eye. Others have discovered that tenomodulin (TeM), a protein homologous to ChM-I, is expressed in hypovascular tissues such as tendons and ligaments. To determine if TeM also has anti-angiogenic properties, adenoviral constructs expressing TeM were used to test the effects of TeM in cultured human retinal endothelial cells. It was found that TeM and ChM-I gene transfer inhibits cell proliferation and tube formation in retinal vascular endothelial cells.168 Reich and colleagues performed experiments to show that VEGF siRNA is effective in blocking VEGF expression in a human cell line in which hypoxia was chemically induced by desferrioxamine. The authors showed that VEGF siRNA treatment in vivo blocked hypoxia-induced increases in

1. Mechanisms of Retinal Angiogenesis

25

VEGF expression in mouse eyes and prevented choroidal neovascularization induced by laser photocoagulation.169 Work in our lab has employed an antisense olignucleotide strategy for blocking gene expression to demonstrate the role of the transcription regulator STAT3 (signal transducer and activator of transcription 3) in the autocrine expression of VEGF in bovine retinal endothelial cells.170 Studies using adenoviral vectors to overexpress a dominant-negative STAT3, which carries a DNA binding mutation, in retinal endothelial cells have demonstrated the specific involvement of STAT3 in high glucose-induced and peroxynitrite-induced VEGF function in these cells.171

2.9

Effects of Hypoxia on Vascular Cells

A decrease in tissue oxygen concentration has long been recognized as a primary cause of angiogenesis.172 However, the mechanisms underlying the induction of angiogenesis by hypoxia are still poorly understood. Chronic ischemia is clearly an important factor in induction of angiogenesis. For example, myocardial ischemia is known to result in collateral development and opening of preexisting vessels. Neovascularization also occurs in chronic inflammatory lesions and solid tumors, both of which are associated with tissue hypoxia.173-175 The retinal microcirculation develops late in fetal life and is strongly influenced by oxygen pressure. As the oxygen pressure rises, the progression of vascularization into the periphery of retina is decreased.176 Retinopathy of prematurity is caused by exposure of underdeveloped retinas to high oxygen at birth when the infant is placed in an oxygen incubator. This results in constriction and obliteration of retinal vessels, thus creating retinal hypoxia when infants are returned to ambient oxygen.177 The sudden hypoxic situation introduces an acute insult to undervascularized retina, leading to massive irregular growth of blood vessels, intra-ocular hemorrhage, degeneration of inner limiting membrane, and retinal detachment. In vitro models of hypoxia allow the study of these complicated events at the cellular level. Cultured vascular cells have been used to explore mechanisms that underlie hypoxia-induced proliferation and to characterize angiogenesisrelated factors released by endothelial, perivascular and glial cells that might play a key role in pathological neovascularization in retina. Lou et al. have shown an increase in DNA synthesis and proliferation of retinal endothelial cells in response to hypoxia exposure (2% oxygen for 4 days).178 A study using retinal endothelial cells maintained in 1% oxygen for 1 hour showed significant increases in the expression of VEGF and VEGF receptors VEGFR-1 and VEGFR-2.179 This study showed that expression of

26

M. A. Behzadian et al.

angiopoietin 1 was low as compared to angiopoietin 2 during normoxia, whereas hypoxia caused increases in angiopoietin 1 and its Tie-2 receptor while angiopoietin 2 was not altered. Work by Nomura and colleagues showed that relative hypoxia stimulated increases in VEGF expression and DNA synthesis in both endothelial cells and pericytes. Antisense oligonucleotides complementary to VEGF mRNAs efficiently inhibited DNA synthesis in endothelial cells cultured under hypoxic conditions, indicating that autocrine expression of VEGF is involved in hypoxia-induced proliferation of endothelial cells.180 Studies using in vitro models indicate that reduced oxygen causes endothelial cell proliferation via upregulation of VEGF. However, the involvement of other growth factors must also be taken into account when considering the processes that occur in vivo. In addition to VEGF, acidic and basic FGF, EGF, TGF-beta, and PDGF have been implicated in angiogenesis.174,175,181,182 The expression of basic FGF, which may be involved in all the steps of angiogenesis, has been shown not to be influenced by hypoxia.183 Expression of TGF-beta has also been shown to be unaffected by hypoxia.184 Instead, TGF-beta becomes activated under hypoxia conditions.42 PDGF-B is a major serum mitogen for mesenchymally derived cells. Since PDGF-B is released by platelets as well as by cells involved in inflammatory responses, it has been suggested to play a role in wound healing.185 Although PDGF-B was previously thought to be devoid of mitogenic activity on endothelial cells,186 functional PDGF-B receptors have been shown to be expressed on hyperplastic capillary endothelial cells in malignant glioma, suggesting that autocrine PDGF-B has a role in the proliferation of endothelial cells. Hypoxia-induced up-regulation of PDGF-B has also been reported.183 Available evidence thus suggests that the major autocrine/paracrine growth factors involved in the control of endothelial cell growth under normoxic conditions are basic FGF, VEGF, and PDGF-B. Under hypoxic conditions, induced VEGF and PDGF-B appear mainly responsible for the endothelial proliferation. Studies on the effect of hypoxia on other retinal cells, such as Muller cells, astrocytes, glial cells, and pericytes, are also important for our understanding of retinal angiogenesis. In vitro studies have shown that hypoxia stimulates release of angiogenesis-related factors in retinal Muller cells and in pericytes.42,180,187 We have demonstrated the effects of hypoxia on expression of VEGF and TGF-beta.42 Muller cells isolated from rat retina were incubated under normoxia or hypoxia and the resulting conditioned media were assayed for their effects on growth on BRE cells. Hypoxia was found to activate TGF-beta and to increase VEGF expression by Muller cells. Eichler et al. have shown that, under hypoxic conditions, Muller cells release not only VEGF but also TGF-beta, PEDF, and thrombospondin-1.187

1. Mechanisms of Retinal Angiogenesis

27

Studies have been done in cultured endothelial cells to explore the potential role of oxidative stress in hypoxia-induced upregulation of VEGF expression and retinal angiogenesis. This work was based on results of in vivo studies in the mouse model of retinopathy of prematurity, which showed that hypoxia-induced increases in VEGF expression and retinal angiogenesis are correlated with increases in superoxide production and upregulation of the NADPH oxidase catalytic subunit gp91phox.188 Moreover, inhibition of NADPH oxidase by apocynin blocked VEGF overexpression and retinal neovascularization. To examine the potential role of retinal endothelial cells in this process, retinal endothelial cells were exposed to hypoxia (1% oxygen, 6 hours), and the effects on expression and activity of NADPH oxidase and VEGF expression were determined. The results showed that hypoxia caused an increase in superoxide generation and VEGF autocrine expression and that both effects were blocked by inhibition of NADPH oxidase with apocynin or a gp91phox blocking peptide (gp91dstat). These observations indicate that NADPH oxidase is a critical source of oxidative stress and a key mediator of VEGF expression during hypoxia.

3.

CLOSING

The cultivation of human HeLa cells by George and Margaret Gey of Johns Hopkins University in 1951 was a milestone in the application of in vitro models to the field of biology. Samples of HeLa cells were soon distributed among laboratories throughout the world, and many scientists adopted the culture conditions to grow other human cell lines and to apply tissue culture models in studying virology, pharmacology, toxicology, and genetics. Tissue culture applications have come a long way in the past 55 years, not only by improvement of culture conditions and instrumentation, but also by development of a number of assay systems for monitoring cell behavior under normal and disease conditions. Large-scale culture of endothelial cells from human umbilical cords by Jaffe and coworkers in 1972 was a turning point for studies of angiogenesis and was soon followed by isolation of capillary endothelial cells from brain and retina. Assays for endothelial cell growth and proliferation, migration, apoptosis, barrier function, tube formation, and interaction with other mural cells have greatly advanced our understanding of retinal angiogenesis under physiological and pathological conditions. At present, we have not only learned how to handle a variety of cells and tissue explants in culture, but we are also more aware of the limitations of the systems, such as the relative advantages and disadvantages of using transformed cell lines vs. primary

28

M. A. Behzadian et al.

cultures for representing in vivo conditions of physiological or pathological angiogenesis. We are aware of the potential artifacts associated with the addition of serum or other crude supplements to the culture, and we are also attentive to practical limitations of gene transfer techniques in certain assay systems. In parallel, the use of laboratory mice as a model system for studying processes related to human health and disease has also expanded greatly. Naturally occurring mutant mice have been identified and are assisting us in understanding the functional role of particular genes. The advent of transgene mouse technology enables us to selectively manipulate the function of a specific gene and follow its effects on a disease process or to generate a desirable animal model for human disease. Pathologies representing human disease conditions, such as retinopathy of prematurity, age-related macular degeneration, and diabetic retinopathy, can be readily induced in normal and transgenic animals in order to understand gene function in the disease process. Tissue explants and conditionally immortalized cells isolated from such transgenic animals provide valuable in vitro models to complement in vivo studies of the molecular mechanisms of retinal disease.

ACKNOWLEDGMENTS This work was supported by the National Institutes of Health (NIH-EY04618 and NIH-EY11766), the American Heart Association, and the Juvenile Diabetes Foundation International.

REFERENCES 1. N. Ferrara, Vascular endothelial growth factor: basic science and clinical progress, Endocr. Rev. 25, 581-611 (2004). 2. J. H. Distler, A. Hirth, M. Kurowska-Stolarska, R. E. Gay, S. Gay, and O. Distler, Angiogenic and angiostatic factors in the molecular control of angiogenesis, Q. J. Nucl. Med. 47, 149-161 (2003). 3. J. Folkman, Role of angiogenesis in tumor growth and metastasis, Semin. Oncol. 29, 15-18 (2002). 4. L. L. Nguyen and P. A. D’Amore, Cellular interactions in vascular growth and differentiation, Int. Rev. Cytol. 204, 1-48 (2001). 5. B. Jiang, M. A. Behzadian, and R. B. Caldwell, Astrocytes modulate retinal vasculogenesis: Effects on endothelial cell differentiation, Glia 15, 1-10 (1995). 6. B. Jiang, G. I. Liou, M. A. Behzadian, and R. B. Caldwell, Astrocytes modulate retinal vasculogenesis: effects on fibronectin expression, J. Cell. Sci. 107, 2499-2508 (1994). 7. R. F. Gariano, Cellular mechanisms in retinal vascular development, Prog. Retin. Eye Res. 22, 295-306 (2003).

1. Mechanisms of Retinal Angiogenesis

29

8. S. Ishida, K. Yamashiro, T. Usui, Y. Kaji, Y. Ogura, T. Hida, Y. Honda, Y. Oguchi, and A. P. Adamis, Leukocytes mediate retinal vascular remodeling during development and vaso-obliteration in disease, Nat. Med. 9, 781-788 (2003), Epub 2003 May 05. 9. L. E. Benjamin, I. Hemo, and E. Keshet, A plasticity window for blood vessel remodelling is defined by pericyte coverage of the preformed endothelial network and is regulated by PDGF-B and VEGF, Development 125, 1591-1598 (1998). 10. M. Klagsbrun and P. A. D’Amore, Regulators of angiogenesis, Annu. Rev. Physiol. 53, 217-239 (1991). 11. J. A. Madri, B. M. Pratt, and A. M. Tucker, Phenotypic modulation of endothelial cells by transforming growth factor-beta depends upon the composition and organization of the extracellular matrix, J. Cell Biol. 106, 1375-84 (1988). 12. J. R. Merwin, J. M. Anderson, O. Kocher, C. M. Van Itallie, and J. A. Madri, Transforming growth factor beta 1 modulates extracellular matrix organization and cell-cell junctional complex formation during in vitro angiogenesis, J. Cell. Physiol. 142, 117-128 (1990). 13. J. R. Merwin, W. Newman, L. D. Beall, A. Tucker, and J. Madri, Vascular cells respond differentially to transforming growth factors beta 1 and beta 2 in vitro, Am. J. Pathol. 138, 37-51 (1991). 14. R. Flaumenhaft, M. Abe, P. Mignatti, and D. B. Rifkin, Basic fibroblast growth factorinduced activation of latent transforming growth factor beta in endothelial cells: regulation of plasminogen activator activity, J. Cell Biol. 118, 901-909 (1992). 15. M. S. Pepper, J. D. Vassalli, L. Orci, and R. Montesano, Biphasic effect of transforming growth factor-beta 1 on in vitro angiogenesis, Exp. Cell Res. 204, 356-363 (1993). 16. P. C. Maisonpierre, C. Suri, P. F. Jones, S. Bartunkova, S. J. Wiegand, C. Radziejewski, D. Compton, J. McClain, T. H. Aldrich, N. Papadopoulos, T. J. Daly, S. Davis, T. N. Sato, and G. D. Yancopoulos, Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis, Science 277, 55-60 (1997). 17. G. Thurston, C. Suri, K. Smith, J. McClain, T. N. Sato, G. D. Yancopoulos, and D. M. McDonald, Leakage-resistant blood vessels in mice transgenically overexpressing angiopoietin-1, Science 286, 2511-2514 (1999). 18. R. P. Visconti, C. D. Richardson, and T. N. Sato, Orchestration of angiogenesis and arteriovenous contribution by angiopoietins and vascular endothelial growth factor (VEGF), Proc. Natl. Acad. Sci. U. S. A. 99, 8219-8224 (2002), Epub 2002 Jun 04. 19. L. Zhou, W. Ma, Z. Yang, F. Zhang, L. Lu, Z. Ding, B. Ding, T. Ha, X. Gao, and C. Li, VEGF(165) and angiopoietin-1 decreased myocardium infarct size through phosphatidylinositol-3 kinase and Bcl-2 pathways, Gene Ther. 28, 28 (2004). 20. E. A. Jaffe, R. L. Nachman, C. G. Becker, and C. R. Minick, Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria, J. Clin. Invest. 52, 2745-2756 (1973). 21. M. A. Gimbrone, Jr., R. S. Cotran, and J. Folkman, Human vascular endothelial cells in culture. Growth and DNA synthesis, J. Cell Biol. 60, 673-684 (1974). 22. D. N. Slater and J. M. Sloan, The porcine endothelial cell in tissue culture, Atherosclerosis 21, 259-272 (1975). 23. A. N. Orekhov, A. V. Krushinsky, E. R. Andreeva, V. S. Repin, and V. N. Smirnov, Adult human aortic cells in primary culture: heterogeneity in shape, Heart Vessels 2, 193-201 (1986). 24. A. M. Schor, S. L. Schor, and T. D. Allen, Effects of culture conditions on the proliferation, morphology and migration of bovine aortic endothelial cells, J. Cell Sci. 62, 267-285 (1983). 25. M. E. Gerritsen, Functional heterogeneity of vascular endothelial cells, Biochem. Pharmacol. 36, 2701-2711 (1987).

30

M. A. Behzadian et al.

26. A. J. Milici, M. B. Furie, and W. W. Carley, The formation of fenestrations and channels by capillary endothelium in vitro, Proc. Natl. Acad. Sci. U. S. A. 82, 6181-6185 (1985). 27. P. A. Stewart and M. J. Wiley, Developing nervous tissue induces formation of bloodbrain barrier characteristics in invading endothelial cells: a study using quail-chick transplantation chimeras, Dev. Biol. 84, 183-192 (1981). 28. R. C. Janzer and M. C. Raff, Astrocytes induce blood-brain barrier properties in endothelial cells, Nature 325, 253-257 (1987). 29. M. P. Carson and C. C. Haudenschild, Microvascular endothelium and pericytes: high yield, low passage cultures, In Vitro Cell. Dev. Biol. 22, 344-354 (1986). 30. J. A. Madri and S. K. Williams, Capillary endothelial cell cultures: phenotypic modulation by matrix components, J. Cell Biol. 97, 153-165 (1983). 31. P. A. Kern, A. Knedler, and R. H. Eckel, Isolation and culture of microvascular endothelium from human adipose tissue, J. Clin. Invest. 71, 1822-1829 (1983). 32. R. M. Marks, M. Czerniecki, and R. Penny, Human dermal microvascular endothelial cells: an improved method for tissue culture and a description of some singular properties in culture, In Vitro Cell. Dev. Biol. 21, 627-635 (1985). 33. U. Tontsch and H. C. Bauer, Isolation, characterization, and long-term cultivation of porcine and murine cerebral capillary endothelial cells, Microvasc. Res. 37, 148-161 (1989). 34. B. V. Sapatino, C. J. Welsh, C. A. Smith, B. F. Bebo, and D. S. Linthicum, Cloned mouse cerebrovascular endothelial cells that maintain their differentiation markers for factor VIII, low density lipoprotein, and angiotensin-converting enzyme, In Vitro Cell. Dev. Biol. Anim. 29 (A), 923-928 (1993). 35. S. M. Buzney and S. J. Massicotte, Retinal vessels: proliferation of endothelium in vitro, Invest. Ophthalmol. Vis. Sci. 18, 1191-1195 (1979). 36. R. N. Frank, V. E. Kinsey, K. W. Frank, K. P. Mikus, and A. Randolph, In vitro proliferation of endothelial cells from kitten retinal capillaries, Invest. Ophthalmol. Vis. Sci. 18, 1195-1200 (1979). 37. P. D. Bowman, A. L. Betz, and G. W. Goldstein, Primary culture of microvascular endothelial cells from bovine retina: selective growth using fibronectin coated substrate and plasma derived serum, In Vitro 18, 626-632 (1982). 38. A. Capetandes and M. E. Gerritsen, Simplified methods for consistent and selective culture of bovine retinal endothelial cells and pericytes, Invest. Ophthalmol. Vis. Sci. 31, 1738-1744 (1990). 39. Q. Yan, R. B. Vernon, A. E. Hendrickson, and E. H. Sage, Primary culture and characterization of microvascular endothelial cells from Macaca monkey retina, Invest. Ophthalmol. Vis. Sci. 37, 2185-2194 (1996). 40. T. Su and M. C. Gillies, A simple method for the in vitro culture of human retinal capillary endothelial cells, Invest. Ophthalmol. Vis. Sci. 33, 2809-2813 (1992). 41. X. Su, C. M. Sorenson, and N. Sheibani, Isolation and characterization of murine retinal endothelial cells, Mol. Vis. 9:171-178 (2003). 42. M. A. Behzadian, X. L. Wang, M. Shabrawey, and R. B. Caldwell, Effects of hypoxia on glial cell expression of angiogenesis-regulating factors VEGF and TGF-beta, Glia 24, 216-225 (1998). 43. Y. Feng, V. J. Venema, R. C. Venema, N. Tsai, and R. B. Caldwell, VEGF induces nuclear translocation of Flk-1/KDR, endothelial nitric oxide synthase and Caveolin-1 in vascular endothelial cells, Biochem. Biophys. Res. Commun. 256, 192-197 (1999). 44. Y. Feng, V. J. Venema, R. C. Venema, N. Tsai, M. A. Behzadian, and R. B. Caldwell, VEGF-induced permeability increase is mediated by caveolae, Invest. Ophthalmol. Vis. Sci. 40, 157-167 (1999).

1. Mechanisms of Retinal Angiogenesis

31

45. M. A. Behzadian, X. L. Wang, L. J. Windson, N. Ghaly, and R. B. Caldwell, TGF-beta increases retinal endothelial cell permeability by increasing MMP-9, possible role of glial cells in endothelial barrier function, Invest. Ophthalmol. Vis. Sci. 42, 853-859 (2001). 46. M. A. Behzadian, L. J. Windsor, N. Ghaly, G. Liou, N. T. Tsai, and R. B. Caldwell, VEGF-induced paracellular permeability in cultured endothelial cells involves urokinase and its receptor, FASEB J. 19:19 (2003). 47. J. Laterra and G. W. Goldstein, Astroglial-induced in vitro angiogenesis: requirements for RNA and protein synthesis, J. Neurochem. 57, 1231-1239 (1991). 48. L. E. DeBault, L. E. Kahn, S. P. Frommes, and P. A. Cancilla, Cerebral microvessels and derived cells in tissue culture: isolation and preliminary characterization, In Vitro 15, 473-487 (1979). 49. R. A. Robinson, C. J. TenEyck, and M. N. Hart, Establishment and preliminary growth characteristics of a transformed mouse cerebral microvessel endothelial cell line, Lab. Invest. 54, 579-588 (1986). 50. S. Kuchler-Bopp, J. P. Delaunoy, J. C. Artault, M. Zaepfel, and J. B. Dietrich, Astrocytes induce several blood-brain barrier properties in non-neural endothelial cells, Neuroreport 10, 1347-1353 (1999). 51. I. Isobe, T. Watanabe, T. Yotsuyanagi, N. Hazemoto, K. Yamagata, T. Ueki, K. Nakanishi, K. Asai, and T. Kato, Astrocytic contributions to blood-brain barrier (BBB) formation by endothelial cells: a possible use of aortic endothelial cell for in vitro BBB model, Neurochem. Int. 28, 523-533 (1996). 52. T. A. Matsubara, T. A. Murata, G. S. Wu, E. A. Barron, and N. A. Rao, Isolation and culture of rat retinal microvessel endothelial cells using magnetic beads coated with antibodies to PECAM-1, Curr. Eye Res. 20, 1-7 (2000). 53. T. Kondo, K. Hosoya, S. Hori, M. Tomi, S. Ohtsuki, H. Takanaga, E. Nakashima, H. Iizasa, T. Asashima, M. Ueda, M. Obinata, and T. Terasaki, Establishment of conditionally immortalized rat retinal pericyte cell lines (TR-rPCT) and their application in a co-culture system using retinal capillary endothelial cell line (TR-iBRB2), Cell Struct. Funct. 28, 145-153 (2003). 54. K. Hosoya, K. Tetsuka, K. Nagase, M. Tomi, S. Saeki, S. Ohtsuki, H. Takanaga, N. Yanai, M. Obinata, A. Kikuchi, T. Okano, and T. Terasaki, Conditionally immortalized brain capillary endothelial cell lines established from a transgenic mouse harboring temperature-sensitive simian virus 40 large T-antigen gene, AAPS Pharm. Sci. 2, E27 (2000). 55. S. Kanda, E. Landgren, M. Ljungstrom, and L. Claesson-Welsh, Fibroblast growth factor receptor 1-induced differentiation of endothelial cell line established from tsA58 large T transgenic mice, Cell Growth Differ. 7, 383-395 (1996). 56. L. M. Postovit, P. Dutt, N. Dourdin, M. Park, P. A. Greer, C. H. Graham, and J. S. Elce, Calpain is required for MMP-2 and u-PA expression in SV40 large T-antigenimmortalized cells, Biochem. Biophys. Res. Commun. 297, 294-301 (2002). 57. E. A. Lidington, R. M. Rao, F. M. Marelli-Berg, P. S. Jat, D. O. Haskard, and J. C. Mason, Conditional immortalization of growth factor-responsive cardiac endothelial cells from H-2K(b)-tsA58 mice, Am. J. Physiol. Cell Physiol. 282, C67-74 (2002). 58. Y. Lin, D. J. Weisdorf, A. Solovey, and R. P. Hebbel, Origins of circulating endothelial cells and endothelial outgrowth from blood, J. Clin. Invest. 105, 71-77 (2000). 59. T. Asahara, H. Masuda, T. Takahashi, C. Kalka, C. Pastore, M. Silver, M. Kearne, M. Magner, and J. M. Isner, Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization, Circ. Res. 85, 221-228 (1999).

32

M. A. Behzadian et al.

60. J. R. Crosby, W. E. Kaminski,G. Schatteman, P. J. Martin, E. W. Raines, R. A. Seifert, and D. F. Bowen-Pope, Endothelial cells of hematopoietic origin make a significant contribution to adult blood vessel formation, Circ. Res. 87, 728-730 (2000). 61. T. Asahara, T. Murohara, A. Sullivan, M. Silver, R. van der Zee, T. Li, B. Witzenbichler, G. Schatteman, and J. M. Isner, Isolation of putative progenitor endothelial cells for angiogenesis, Science 275, 964-967 (1997). 62. J. J. Zwaginga and P. Doevendans, Stem cell-derived angiogenic/vasculogenic cells: possible therapies for tissue repair and tissue engineering, Clin. Exp. Pharmacol. Physiol. 30, 900-908 (2003). 63. N. Sengupta, S. Caballero, R. N. Mames, J. M. Butler, E. W. Scott, and M. B. Grant, The role of adult bone marrow-derived stem cells in choroidal neovascularization, Invest. Ophthalmol. Vis. Sci. 44, 4908-4913 (2003). 64. M. B. Grant, W. S. May, S. Caballero, G. A. Brown, S. M. Guthrie, R. N. Mames, B. J. Byrne, T. Vaught, P. E. Spoerri, A. B. Peck, and E. W. Scott, Adult hematopoietic stem cells provide functional hemangioblast activity during retinal neovascularization, Nat. Med. 8, 607-612 (2002). 65. M. B. Grant, S. Caballero, G. A. Brown, S. M. Guthrie, R. N. Mames, T. Vaught, and E. W. Scott, The contribution of adult hematopoietic stem cells to retinal neovascularization, Adv. Exp. Med. Biol. 522, 37-45 (2003). 66. A. Otani, K. Kinder, K. Ewalt, F. J. Otero, P. Schimmel, and M. Friedlander, Bone marrow-derived stem cells target retinal astrocytes and can promote or inhibit retinal angiogenesis, Nat. Med. 8, 1004-1010 (2002). 67. M. I. Dorrell, A. Otani, E. Aguilar, S. K. Moreno, and M. Friedlander, Adult bone marrow-derived stem cells use R-cadherin to target sites of neovascularization in the developing retina, Blood 103, 3420-3427 (2004). 68. S. Rafii, D. Lyden, R. Benezra, K. Hattori, and B. Heissig, Vascular and haematopoietic stem cells: novel targets for anti-angiogenesis therapy? Nat. Rev. Cancer 2, 826-835 (2002). 69. M. Vasa, S. Fichtlscherer, K. Adler, A. Aicher, H. Martin, A. M. Zeiher, and S. Dimmeler, Increase in circulating endothelial progenitor cells by statin therapy in patients with stable coronary artery disease, Circulation 103, 2885-2890 (2001). 70. M. Peichev, A. J. Naiyer, D. Pereira, Z. Zhu, W. J. Lane, M. Williams, M. C. Oz, D. J. Hicklin, L. Witte, M. A. Moore, and S. Rafii, Expression of VEGFR-2 and AC133 by circulating human CD34(+) cells identifies a population of functional endothelial precursors, Blood 95, 952-958 (2000). 71. E. Pelosi, M. Valtieri, S. Coppola, R. Botta, M. Gabbianelli, V. Lulli, G. Marziali, B. Masella, R. Muller, C. Sgadari, U. Testa, G. Bonanno, and C. Peschle, Identification of the hemangioblast in postnatal life, Blood 100, 3203-3208 (2002). 72. T. Asahara and A. Kawamoto, Endothelial progenitor cells for postnatal vasculogenesis. Am. J. Physiol. Cell Physiol. 287, C572-579 (2004). 73. C. Urbich and S. Dimmeler, Endothelial progenitor cells: characterization and role in vascular biology, Circ. Res. 95, 343-353 (2004). 74. A. V. Evsikov and D. Solter, Comment on “ ‘Stemness’: transcriptional profiling of embryonic and adult stem cells” and “a stem cell molecular signature”, Science 302, 393; author reply 393 (2003). 75. N. O. Fortunel, H. H. Otu, H. H. Ng, J. Chen, X. Mu, T. Chevassut, X. Li, M. Joseph, C. Bailey, J. A. Hatzfeld, A. Hatzfeld, F. Usta, V. B. Vega, P. M. Long, T. A. Libermann, and B. Lim, Comment on “ ‘Stemness’: transcriptional profiling of embryonic and adult stem cells” and “a stem cell molecular signature”, Science 302, 393; author reply 393 (2003).

1. Mechanisms of Retinal Angiogenesis

33

76. R. G. Bagley, J. Walter-Yohrling, X. Cao, W. Weber, B. Simons, B. P. Cook, S. D. Chartrand, C. Wang, S. L. Madden, and B. A. Teicher, Endothelial precursor cells as a model of tumor endothelium: characterization and comparison with mature endothelial cells, Cancer Res. 63, 5866-5873 (2003). 77. Z. He, C. Ehrnfelt, M. Kumagai-Braesch, K. B. Islam, and J. Holgersson, Aberrant expression of alpha-Gal on primary human endothelium does not confer susceptibility to NK cell cytotoxicity or increased NK cell adhesion, Eur. J. Immunol. 34, 1185-1195 (2004). 78. B. Assmus, C. Urbich, A. Aicher, W. K. Hofmann, J. Haendeler, L. Rossig, I. Spyridopoulos, A. M. Zeiher, and S. Dimmeler, HMG-CoA reductase inhibitors reduce senescence and increase proliferation of endothelial progenitor cells via regulation of cell cycle regulatory genes, Circ. Res. 92, 1049-1055 (2003). 79. D. H. Walter, K. Rittig, F. H. Bahlmann, R. Kirchmair, M. Silver, T. Murayama, H. Nishimura, D. W. Losordo, T. Asahara, and J. M. Isner, Statin therapy accelerates reendothelialization: a novel effect involving mobilization and incorporation of bone marrow-derived endothelial progenitor cells, Circulation 105, 3017-3024 (2002). 80. M. Butzal, S. Loges, M. Schweizer, U. Fischer, U. M. Gehling, D. K. Hossfeld, and W. Fiedler, Rapamycin inhibits proliferation and differentiation of human endothelial progenitor cells in vitro, Exp. Cell Res. 300, 65-71 (2004). 81. V. Goede, T. Schmidt, S. Kimmina, D. Kozian, and H. G. Augustin, Analysis of blood vessel maturation processes during cyclic ovarian angiogenesis, Lab. Invest. 78, 1385-1394 (1998). 82. M. Hellstrom, M. Kalen, P. Lindahl, A. Abramsson, and C. Betsholtz, Role of PDGF-B and PDGFR-beta in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse, Development 126, 3047-3055 (1999). 83. K. K. Hirschi and P. A. D’Amore, Control of angiogenesis by the pericyte: molecular mechanisms and significance, EXS 79, 419-428 (1997). 84. H. Yamada, E. Yamada, S. F. Hackett, H. Ozaki, N. Okamoto, and P. A. Campochiaro, Hyperoxia causes decreased expression of vascular endothelial growth factor and endothelial cell apoptosis in adult retina, J. Cell. Physiol. 179, 149-156 (1999). 85. X. Gu, S. Samuel, M. El-Shabrawey, R. B. Caldwell, M. Bartoli, D. M. Marcus, and S. E. Brooks, Effects of sustained hyperoxia on revascularization in experimental retinopathy of prematurity, Invest. Ophthalmol. Vis. Sci. 43, 496-502 (2002). 86. L. E. Benjamin, I. Hemo, and E. Keshet, A plasticity window for blood vessel remodelling is defined by pericyte coverage of the preformed endothelial network and is regulated by PDGF-B and VEGF, Development 125, 1591-1598 (1998). 87. H. P. Hammes, J. Lin, O. Renner, M. Shani, A. Lundqvist, C. Betsholtz, M. Brownlee, and U. Deutsch, Pericytes and the pathogenesis of diabetic retinopathy, Diabetes 51, 3107-3112 (2002). 88. S. C. Shih, M. Ju, N. Liu, J. R. Mo, J. J. Ney, and L. E. Smith, Transforming growth factor beta1 induction of vascular endothelial growth factor receptor 1: mechanism of pericyte-induced vascular survival in vivo, Proc. Natl. Acad. Sci. U. S. A. 100, 1585915864 (2003). 89. A. Orlidge and P. A. D’Amore, Inhibition of capillary endothelial cell growth by pericytes and smooth muscle cells, J. Cell Biol. 105, 1455-1462 (1987). 90. D. C. Darland, L. J. Massingham, S. R. Smith, E. Piek, M. Saint-Geniez, and P. A. D’Amore, Pericyte production of cell-associated VEGF is differentiation-dependent and is associated with endothelial survival, Dev. Biol. 264, 275-288 (2003).

34

M. A. Behzadian et al.

91. A. Kennedy, R. N. Frank, and S. D. Varma, Aldose reductase activity in retinal and cerebral microvessels and cultured vascular cells, Invest. Ophthalmol. Vis. Sci. 24, 1250-1258 (1983). 92. R. C. Nayak, A. B. Berman, K. L. George, G. S. Eisenbarth, and G. L. King, A monoclonal antibody (3G5)-defined ganglioside antigen is expressed on the cell surface of microvascular pericytes, J. Exp. Med. 167, 1003-1015 (1988). 93. J. G. Cunha-Vaz, The blood-ocular barriers, Survey Ophthalmol. 23, 279-296 (1979). 94. P. A. Stewart and U. I. Tuor, Blood-eye barriers in the rat: correlation of ultrastructure with function, J. Comp. Neurol. 340, 566-576 (1994). 95. M. Tomi, H. Abukawa, Y. Nagai, T. Hata, H. Takanaga, S. Ohtsuki, T. Terasaki, and K. Hosoya, Retinal selectivity of gene expression in rat retinal versus brain capillary endothelial cell lines by differential display analysis, Mol. Vis. 10, 537-543 (2004). 96. H. F. Dvorak, L. F. Brown, M. Detmar, and A. M. Dvorak, Vascular permeability factor/ vascular endothelial growth factor, microvascular hyperpermeability and angiogenesis, Am. J. Path. 146, 1029-1039 (1995). 97. M. S. Pepper, Role of the matrix metalloproteinase and plasminogen activator-plasmin systems in angiogenesis, Arterioscler. Thromb. Vasc. Biol. 21, 1104-1117 (2001). 98. W. Wang, M. J. Merrill, and R. T. Borchardt, Vascular endothelial growth factor affects permeability of brain microvessel endothelial cells in vitro, Am. J. Physiol. 271, C1973-1980 (1996). 99. C. G. Kevil, D. K. Payne, E. Mire, and J. S. Alexander, Vascular permeability factor/vascular endothelial cell growth factor-mediated permeability occurs through disorganization of endothelial junctional proteins, J. Biol. Chem. 273, 15099-15103 (1998). 100. B. A. Young, P. Wang, and S. E. Goldblum, The counteradhesive protein SPARC regulates an endothelial paracellular pathway through protein tyrosine phosphorylation, Biochem. Biophys. Res. Commun. 251, 320-327 (1998). 101. A. W. Cohen, J. M. Carbajal, and R. C. Schaeffer, Jr., VEGF stimulates tyrosine phosphorylation of beta-catenin and small-pore endothelial barrier dysfunction, Am. J. Physiol. 277, H2038-2049 (1999). 102. B. K. Lal, S. Varma, P. J. Pappas, R. W. Hobson, 2nd, and W. N. Duran, VEGF increases permeability of the endothelial cell monolayer by activation of PKB/akt, endothelial nitric-oxide synthase, and MAP kinase pathways, Microvasc. Res. 62, 252-262 (2001). 103. G. L. King, A. B. Berman, S. Bonner-Weir, and M. P. Carson, Regulation of vascular permeability in cell culture, Diabetes 36, 1460-1467 (1987). 104. M. C. Gillies, T. Su, and D. Naidoo, Electrical resistance and macromolecular permeability of retinal capillary endothelial cells in vitro, Curr. Eye Res. 14, 435-442 (1995). 105. J. R. Burdo, D. A. Antonetti, E. B. Wolpert, and J. R. Connor, Mechanisms and regulation of transferrin and iron transport in a model blood-brain barrier system, Neuroscience 121, 883-890 (2003). 106. S. Lakshminarayanan, D. A. Antonetti, T. W. Gardner, and J. M. Tarbell, Effect of VEGF on retinal microvascular endothelial hydraulic conductivity: the role of NO, Invest. Ophthalmol. Vis. Sci. 41, 4256-4261 (2000). 107. D. A. Antonetti, E. B. Wolpert, L. DeMaio, N. S. Harhaj, and R. C. Scaduto, Jr., Hydrocortisone decreases retinal endothelial cell water and solute flux coincident with increased content and decreased phosphorylation of occludin, J. Neurochem. 80, 667-677 (2002).

1. Mechanisms of Retinal Angiogenesis

35

108. M. C. Gillies, T. Su, J. Stayt, J. M. Simpson, D. Naidoo, and C. Salonikas, Effect of high glucose on permeability of retinal capillary endothelium in vitro. Invest. Ophthalmol. Vis. Sci. 38, 635-642 (1997). 109. M. A. Behzadian, X. L. Wang, B. Jiang, and R. B. Caldwell, Angiostatic role of astrocytes: suppression of vascular endothelial cell growth by TGF-beta and other inhibitory factor(s), Glia 15, 480-490 (1995). 110. C. M. Lo, C. R. Keese, and I. Giaever, Cell-substrate contact: another factor may influence transepithelial electrical resistance of cell layers cultured on permeable filters, Exp. Cell Res. 250, 576-580 (1999). 111. C. R. Keese, J. Wegener, S. R. Walker, and I. Giaever, Electrical wound-healing assay for cells in vitro, Proc. Natl. Acad. Sci. U. S. A. 101, 1554-1559 (2004). 112. J. L. Gross, M. N. Krupp, D. B. Rifkin, and M. D. Lane. Down-regulation of epidermal growth factor receptor correlates with plasminogen activator activity in human A431 epidermoid carcinoma cells, Proc. Natl. Acad. Sci. U. S. A. 80, 2276-2280 (1983). 113. P. Mignatti, E. Robbins, and D. B. Rifkin, Tumor invasion through the human amniotic membrane: requirement for a proteinase cascade, Cell 47, 487-498 (1986). 114. D. Moscatelli, M. Presta, and D. B. Rifkin, Purification of a factor from human placenta that stimulates capillary endothelial cell protease production, DNA synthesis, and migration, Proc. Natl. Acad. Sci. U. S. A. 83, 2091-2095 (1986). 115. M. Presta, D. Moscatelli, J. Joseph-Silverstein, and D. B. Rifkin, Purification from a human hepatoma cell line of a basic fibroblast growth factor-like molecule that stimulates capillary endothelial cell plasminogen activator production, DNA synthesis, and migration, Mol. Cell. Biol. 6, 4060-4066 (1986). 116. T. Kalebic, S. Garbisa, B. Glaser, and L. A. Liotta, Basement membrane collagen: degradation by migrating endothelial cells, Science 221, 281-283 (1983). 117. W. E. Laug, M. E. Weinblatt, and P. A. Jones, Endothelial cells degrade extracellular matrix proteins produced in vitro, Thromb. Haemost. 54, 498-502 (1985). 118. M. S. Pepper, J. D. Vassalli, R. Montesano, and L. Orci, Urokinase-type plasminogen activator is induced in migrating capillary endothelial cells, J. Cell Biol. 105, 2535-2541 (1987). 119. X. Houard, C. Monnot, V. Dive, P. Corvol, and M. Pagano, Vascular smooth muscle cells efficiently activate a new proteinase cascade involving plasminogen and fibronectin, J. Cell Biochem. 88, 1188-1201 (2003). 120. V. Ellis and S. A. Whawell, Vascular smooth muscle cells potentiate plasmin generation by both urokinase and tissue plasminogen activator-dependent mechanisms: evidence for a specific tissue-type plasminogen activator receptor on these cells, Blood 90, 2312-2322 (1997). 121. J. M. Sivak and M. E. Fini, MMPs in the eye: emerging roles for matrix metalloproteinases in ocular physiology, Prog. Retin. Eye Res. 21, 1-14 (2002). 122. A. Das and P. G. McGuire, Retinal and choroidal angiogenesis: pathophysiology and strategies for inhibition, Prog. Retin. Eye Res. 22, 721-748 (2003). 123. M. F. Burbridge, F. Coge, J. P. Galizzi, J. A. Boutin, D. C. West, and G. C. Tucker, The role of the matrix metalloproteinases during in vitro vessel formation, Angiogenesis 5, 215-226 (2002). 124. S. Majka, P. G. McGuire, and A. Das, Regulation of matrix metalloproteinase expression by tumor necrosis factor in a murine model of retinal neovascularization, Invest. Ophthalmol. Vis. Sci. 43, 260-266 (2002). 125. A. Das, A. McLamore, W. Song, and P. G. McGuire, Retinal neovascularization is suppressed with a matrix metalloproteinase inhibitor, Arch. Ophthalmol. 117, 498-503 (1999).

36

M. A. Behzadian et al.

126. M. S. Pepper and R. Montesano, Proteolytic balance and capillary morphogenesis, Cell Differ. Dev. 32, 319-327 (1990). 127. R. Montesano, M. S. Pepper, J. D. Vassalli, and L. Orci, Phorbol ester induces cultured endothelial cells to invade a fibrin matrix in the presence of fibrinolytic inhibitors, J. Cell. Physiol. 132, 509-516 (1987). 128. J. L. Gross, D. Moscatelli, E. A. Jaffe, and D. B. Rifkin, Plasminogen activator and collagenase production by cultured capillary endothelial cells, J. Cell Biol. 95, 974-981 (1982). 129. J. L. Gross, D. Moscatelli, and D. B. Rifkin, Increased capillary endothelial cell protease activity in response to angiogenic stimuli in vitro, Proc. Natl. Acad. Sci. U. S. A. 80, 2623-2627 (1983). 130. G. I. Schoefl, Electron microscopic observations on the regeneration of blood vessels after injury, Ann. N. Y. Acad. Sci. 116, 789-802 (1964). 131. D. H. Ausprunk and J. Folkman, Migration and proliferation of endothelial cells in preformed and newly formed blood vessels during tumor angiogenesis, Microvasc. Res. 14, 53-65 (1977). 132. R. Castellon, H. K. Hamdi, I. Sacerio, A. M. Aoki, M. C. Kenney, and A. V. Ljubimov, Effects of angiogenic growth factor combinations on retinal endothelial cells, Exp. Eye Res. 74, 523-535 (2002). 133. R. R. Schleef and C. R. Birdwell, The effect of fibrin on endothelial cell migration in vitro, Tissue Cell 14, 629-636 (1982). 134. A. E. Postlethwaite and A. H. Kang, Guinea pig lymphocyte-derived macrophage aggregation factor: its separation from macrophage migration inhibitory factor, J. Immunol. 117, 1651-1655 (1976). 135. J. A. Nadal, G. M. Scicli, L. A. Carbini, J. J. Nussbaum, and A. G. Scicli, Angiotensin II and retinal pericytes migration, Biochem. Biophys. Res. Commun. 266, 382-385 (1999). 136. M. Grant, S. Caballero, D. Bush, and P. Spoerri, Fibronectin fragments modulate human retinal capillary cell proliferation and migration, Diabetes 47, 1335-1340 (1998). 137. M. Ushio-Fukai, Y. Tang, T. Fukai, S. I. Dikalov, Y. Ma, M. Fujimoto, M. T. Quinn, P. J. Pagano, C. Johnson, and R. W. Alexander, Novel role of gp91(phox)-containing NAD(P)H oxidase in vascular endothelial growth factor-induced signaling and angiogenesis, Circ. Res. 91, 1160-1167 (2002). 138. R. Auerbach, R. Lewis, B. Shinners, L. Kubai, and N. Akhtar, Angiogenesis assays: a critical overview, Clin. Chem. 49, 32-40 (2003). 139. B. M. Glaser, P. A. D’Amore, R. G. Michels, A. Patz, and A. Fenselau, Demonstration of vasoproliferative activity from mammalian retina, J. Cell Biol. 84, 298-304 (1980). 140. Z. Rymaszewski, R. M. Cohen, and P. Chomczynski, Human growth hormone stimulates proliferation of human retinal microvascular endothelial cells in vitro, Proc. Natl. Acad. Sci. U. S. A. 88, 617-621 (1991). 141. K. Ohno-Matsui, I. Morita, J. Tombran-Tink, D. Mrazek, M. Onodera, T. Uetama, M. Hayano, S. I. Murota, and M. Mochizuki, Novel mechanism for age-related macular degeneration: an equilibrium shift between the angiogenesis factors VEGF and PEDF, J. Cell. Physiol. 189, 323-333 (2001). 142. T. A. McCaffrey, L. A. Agarwal, and B. B. Weksler, A rapid fluorometric DNA assay for the measurement of cell density and proliferation in vitro, In Vitro Cell. Dev. Biol. 24, 247-252 (1988). 143. R. S. Gieni, Y. Li, and K. T. HayGlass, Comparison of [3H]thymidine incorporation with MTT- and MTS-based bioassays for human and murine IL-2 and IL-4 analysis. Tetrazolium assays provide markedly enhanced sensitivity, J. Immunol. Methods 187, 85-93 (1995).

1. Mechanisms of Retinal Angiogenesis

37

144. M. B. Hansen, S. E. Nielsen, and K. Berg, Re-examination and further development of a precise and rapid dye method for measuring cell growth/cell kill, J. Immunol. Methods 119, 203-210 (1989). 145. P. E. Spoerri, S. Caballero, S. H. Wilson, L. C. Shaw, and M. B. Grant, Expression of IGFBP-3 by human retinal endothelial cell cultures: IGFBP-3 involvement in growth inhibition and apoptosis, Invest. Ophthalmol. Vis. Sci. 44, 365-369 (2003). 146. R. Montesano, L. Orci, and P. Vassalli, In vitro rapid organization of endothelial cells into capillary-like networks is promoted by collagen matrices, J. Cell Biol. 97, 1648-1652 (1983). 147. R. Montesano, M. S. Pepper, J. D. Vassalli, and L. Orci, Phorbol ester induces cultured endothelial cells to invade a fibrin matrix in the presence of fibrinolytic inhibitors, J. Cell. Physiol. 132, 509-516 (1987). 148. R. Montesano and L. Orci, Tumor-promoting phorbol esters induce angiogenesis in vitro. Cell. 42, 469-477 (1985). 149. R. Montesano, J. D. Vassalli, A. Baird, R. Guillemin, and L. Orci, Basic fibroblast growth factor induces angiogenesis in vitro, Proc. Natl. Acad. Sci. U. S. A. 83, 7297-7301 (1986). 150. A. M. Schor and S. L. Schor, The isolation and culture of endothelial cells and pericytes from the bovine retinal microvasculature: a comparative study with large vessel vascular cells, Microvasc. Res. 32, 21-38 (1986). 151. Q. Yan, Y. Li, A. Hendrickson, and E. H. Sage, Regulation of retinal capillary cells by basic fibroblast growth factor, vascular endothelial growth factor, and hypoxia, In Vitro Cell. Dev. Biol. Anim. 37, 45-49 (2001). 152. A. Kasai, N. Shintani, M. Oda, M. Kakuda, H. Hashimoto, T. Matsuda, S. Hinuma, and A. Baba, Apelin is a novel angiogenic factor in retinal endothelial cells, Biochem Biophys Res Commun. 325, 395-400 (2004). 153. R. S. Talhouk, M. J. Bissell, and Z. Werb, Coordinated expression of extracellular matrix-degrading proteinases and their inhibitors regulates mammary epithelial function during involution, J. Cell Biol. 118, 1271-1282 (1992). 154. T. Alon, I. Hemo, A. Itin, J. Pe’er, J. Stone, and E. Keshet, Vascular endothelial growth factor acts as a survival factor for newly formed retinal vessels and has implications for retinopathy of prematurity, Nat. Med. 1, 1024-1028 (1995). 155. S. Hughes and T. Chang-Ling, Roles of endothelial cell migration and apoptosis in vascular remodeling during development of the central nervous system, Microcirculation 7, 317-333 (2000). 156. M. E. Choi and B. J. Ballermann, Inhibition of capillary morphogenesis and associated apoptosis by dominant negative mutant transforming growth factor-beta receptors, J. Biol. Chem. 270, 21144-21150 (1995). 157. G. E. Davis, K. A. Pintar Allen, R. Salazar, and S. A. Maxwell, Matrix metalloproteinase-1 and -9 activation by plasmin regulates a novel endothelial cellmediated mechanism of collagen gel contraction and capillary tube regression in threedimensional collagen matrices, J. Cell Sci. 114, 917-930 (2001). 158. G. M. Seigel, The golden age of retinal cell culture, Mol. Vis. 5, 4 (1999). 159. R. M. Knott, M. Robertson, E. Muckersie, V. A. Folefac, F. E. Fairhurst, S. M. Wileman, and J. V. Forrester, A model system for the study of human retinal angiogenesis: activation of monocytes and endothelial cells and the association with the expression of the monocarboxylate transporter type 1 (MCT-1), Diabetologia 42, 870-877 (1999). 160. J. V. Forrester, A. Chapman, C. Kerr, J. Roberts, W. R. Lee, and J. M. Lackie, Bovine retinal explants cultured in collagen gels. A model system for the study of proliferative retinopathy, Arch. Ophthalmol. 108, 415-420 (1990).

38

M. A. Behzadian et al.

161. R. F. Nicosia and A. Ottinetti, Growth of microvessels in serum-free matrix culture of rat aorta. A quantitative assay of angiogenesis in vitro, Lab. Invest. 63, 115-122 (1990). 162. The rat aorta model of angiogenesis and its applications, edited by R. F. Nicosia (Birkhauser, Boston, 1998). 163. S. Blacher, L. Devy, M. F. Burbridge, G. Roland, G. Tucker, A. Noel, and J. M. Foidart, Improved quantification of angiogenesis in the rat aortic ring assay, Angiogenesis 4, 133-142 (2001). 164. R. F. Nicosia, R. Tchao, and J. Leighton, Histotypic angiogenesis in vitro: light microscopic, ultrastructural, and radioautographic studies, In Vitro 18, 538-549 (1982). 165. R. S. Go and W. G. Owen, The rat aortic ring assay for in vitro study of angiogenesis, Methods Mol. Med. 85, 59-64 (2003). 166. H. Gerhardt, M. Golding, M. Fruttiger, C. Ruhrberg, A. Lundkvist, A. Abramsson, M. Jeltsch, C. Mitchell, K. Alitalo, D. Shima, and C. Betsholtz, VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia, J. Cell Biol. 161, 1163-1177 (2003). Epub 2003 Jun 16. 167. I. Adini, I. Rabinovitz, J. F. Sun, G. C. Prendergast, and L. E. Benjamin, RhoB controls Akt trafficking and stage-specific survival of endothelial cells during vascular development, Genes Dev. 17, 2721-2732 (2003). 168. Y. Oshima, C. Shukunami, J. Honda, K. Nishida, F. Tashiro, J. Miyazaki, Y. Hiraki, and Y. Tano, Expression and localization of tenomodulin, a transmembrane type chondromodulin-I-related angiogenesis inhibitor, in mouse eyes, Invest. Ophthalmol. Vis. Sci. 44, 1814-1823 (2003). 169. S. J. Reich, J. Fosnot, A. Kuroki, W. Tang, X. Yang, A. M. Maguire, J. Bennett, and M. J. Tolentino, Small interfering RNA (siRNA) targeting VEGF effectively inhibits ocular neovascularization in a mouse model, Mol. Vis. 9, 210-216 (2003). 170. M. Bartoli, D. Platt, T. Lemtalsi, X. Gu, S. E. Brooks, M. B. Marrero, and R. B. Caldwell, VEGF differentially activates STAT3 in microvascular endothelial cells, FASEB J. 17, 1562-1564 (2003). 171. D. H. Platt, M. Bartoli, A. B. El-Remessy, M. Al-Shabrawey, T. Lemtalsi, D. Fulton, and R. B. Caldwell, Peroxynitrite induces VEGF transcription in vascular cells via Stat3, Free Radic. Biol. Med. 39 (10), 1353-1361 (2005). 172. A. Ladoux and C. Frelin, Hypoxia is a strong inducer of vascular endothelial growth factor mRNA expression in the heart, Biochem. Biophys. Res. Commun. 195, 1005-10 (1993). 173. J. Folkman, What is the role of angiogenesis in metastasis from cutaneous melanoma? Eur. J. Cancer Clin. Oncol. 23, 361-363 (1987). 174. J. Folkman and M. Klagsbrun, Angiogenic factors, Science 235, 442-447 (1987). 175. J. Folkman and M. Klagsbrun, Vascular physiology. A family of angiogenic peptides, Nature 329, 671-672 (1987). 176. D. L. Phelps, Oxygen and developmental retinal capillary remodeling in the kitten, Invest. Ophthalmol. Vis. Sci. 31, 2194-2200 (1990). 177. A. Patz, Studies on retinal neovascularization. Friedenwald Lecture, Invest. Ophthalmol. Vis. Sci. 19, 1133-1138 (1980). 178. Y. Lou, J. C. Oberpriller, and E. C. Carlson, Effect of hypoxia on the proliferation of retinal microvessel endothelial cells in culture, Anat. Rec. 248, 366-373 (1997). 179. E. Brylla, G. Tscheudschilsuren, A. N. Santos, K. Nieber, K. Spanel-Borowski, and G. Aust, Differences between retinal and choroidal microvascular endothelial cells (MVECs) under normal and hypoxic conditions, Exp. Eye Res. 77, 527-535 (2003). 180. M. Nomura, S. Yamagishi, S. Harada, Y. Hayashi, T. Yamashima, J. Yamashita, and H. Yamamoto, Possible participation of autocrine and paracrine vascular endothelial

1. Mechanisms of Retinal Angiogenesis

181. 182. 183.

184.

185. 186.

187. 188.

39

growth factors in hypoxia-induced proliferation of endothelial cells and pericytes, J. Biol. Chem. 270, 28316-28324 (1995). G. Carpenter and S. Cohen, Epidermal growth factor, J. Biol. Chem. 265, 7709-7712 (1990). J. Massague, Transforming growth factor-alpha. A model for membrane-anchored growth factors, J. Biol. Chem. 265, 21393-21396 (1990). S. Kourembanas, R. L. Hannan, and D. V. Faller, Oxygen tension regulates the expression of the platelet-derived growth factor-B chain gene in human endothelial cells, J. Clin. Invest. 86, 670-674 (1990). D. Gospodarowicz, N. Ferrara, L. Schweigerer, and G. Neufeld, Structural characterization and biological functions of fibroblast growth factor, Endocr. Rev. 8, 95-114 (1987). C. H. Heldin, Structural and functional studies on platelet-derived growth factor, EMBO J. 11, 4251-4259 (1992). K. Funa, V. Papanicolaou, C. Juhlin, J. Rastad, G. Akerstrom, C. H. Heldin, and K. Oberg, Expression of platelet-derived growth factor beta-receptors on stromal tissue cells in human carcinoid tumors, Cancer Res. 50, 748-753 (1990). W. Eichler, Y. Yafai, P. Wiedemann, and A. Reichenbach, Angiogenesis-related factors derived from retinal glial (Muller) cells in hypoxia, Neuroreport 15, 1633-1637 (2004). M. Al-Shabrawey, M. Bartoli, A. B. El-Remessy, D. H. Platt, S. Matragoon, M. A. Behzadian, R. W. Caldwell, and R. B. Caldwell, Inhibition of NAD(P)H oxidase activity blocks VEGF over-expression and neovascularization during ischemic retinopathy, Am. J. Pathol. 167 (2), 599-607 (Aug 2005).

Chapter 2 ANIMAL MODELS OF CHOROIDAL NEOVASCULARIZATION

Monika L. Clark,1 Jessica A. Fowler,2 and John S. Penn1,2

Departments of 1Cell & Developmental Biology and 2Ophthalmology & Visual Sciences, Vanderbilt University School of Medicine, Nashville, Tennessee

Abstract:

1.

Choroidal neovascularization (CNV) is a pathological condition in which proliferating choroidal blood vessels grow through Bruch’s membrane, penetrate the retinal pigment epithelium (RPE) and extend into the subretinal space. There, the blood vessels leak fluid, ultimately leading to serous retinal detachment. CNV associated with the wet form of age-related macular degeneration (AMD) is the major cause of vision loss in the elderly. However, in spite of its prevalence, relatively little is known concerning the pathogenesis of CNV. In order to better understand this disease process and explore therapies to treat it, several experimental animal models of CNV have been developed. The most widely used of these models is laser-induced CNV in primates and rodents, but several knockout and transgenic mouse models exist as well. The aim of this chapter is to explore the historical background and significance of these animal models of CNV.

BACKGROUND

There are many ocular conditions in which pathological angiogenesis is a key component. Ocular angiogenesis may occur as preretinal neovascularization, deep retinal neovascularization, or subretinal neovascularization (Figure 1). Subretinal neovascularization occurs in conditions such as age-related macular degeneration (AMD), Sorsby’s fundus dystrophy, Pseudoxanthoma Elasticum, ocular histoplasmosis and multifocal choroiditis. AMD is the leading cause of blindness in individuals 65 years or older in developed countries.1 Although it encompasses a wide range of pathologies, the disease is generally classified into two forms: “dry” and “wet.” Of these 41 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 41–56. © Springer Science+Business Media B.V. 2008

42

M. L. Clark et al.

two recognized forms, wet AMD is the more debilitating and life-changing in its progression. Fortunately, it is also the less prevalent form, affecting only about 10% of the general AMD population.2 Wet AMD typically becomes manifest with progressive choroidal neovascularization (CNV) characterized by abnormal blood vessel growth of the choriocapillaris through the retinal pigment epithelial (RPE) layer.1 Because the consequences of wet AMD can be so devastating, the vision research community has invested a tremendous amount of time and effort in its attempts to better understand the progression of this form of AMD. A major focus of this effort has involved the development of animal models of CNV, which are essential for identifying early diagnostic markers and for developming of better drug regimens.

Figure 2-1. Schematic diagram of pathological blood vessel growth within the posterior segment. A. Normal vasculature; B. Subretinal neovascularization (eg. age-related macular degeneration); C. Deep retinal neovascularization (eg. retinal angiomatous proliferation,3 Type II idiopathic juxtafoveolar telangiectasia4); D. Pre-retinal neovascularization (eg. retinopathy of prematurity, diabetic retinopathy).

2. Animal Models of Choroidal Neovascularization

2.

ANIMAL MODELS OF LASER-INDUCED CNV

2.1

Primate

2.1.1

Development of the primate model

43

The basis for the development of a laser-induced CNV model was the finding that argon laser photocoagulation, used clinically to obliterate neovascularization in the treatment of macular degeneration, could actually induce subretinal neovascularization.5 In 1982, Ryan and colleagues administered argon laser photocoagulation to primate eyes with the intention of inducing CNV rather than treating it, thus using criteria contradictory to what is used clinically.6 Specifically, following sedation and pupillary dilation, high intensity (600-900 mW), short duration (0.1 s) laser burns of a small exposure size (100 μm) were applied through a slit lamp and a Goldmann fundus contact lens to three distinct areas of the fundus of rhesus monkeys. Subretinal neovascularization, assessed by fluorescein leakage during angiography, was observed after three weeks in 39% of laser-induced lesions in the macular region and less than 3% of lesions located nasal to the optic nerve head and in the periphery. The increased incidence of subretinal neovascularization in the lesions of the macular region correlated well with the high predisposition of the human macula to develop CNV, an initial appeal of this new model. Closer inspection of the morphology of laser burn sites generated in the eyes of cynomolgus monkeys provided more direct evidence that laser photocoagulation can experimentally induce CNV, lending even more promise to the relevance of this model. Morphological assessment of cross sections of the laser lesions one day after laser treatment revealed that the choroid, Bruch’s membrane and RPE cells were disrupted or destroyed. By approximately one week following laser treatment, choroidal vessels had proliferated through the laser-induced breaks in Bruch’s membrane into the subretinal space and were observed overlying proliferating RPE cells.7,8 Histological examination of cross-sectioned lesions confirmed that all lesions exhibiting fluorescein leakage and pooling contained CNV, and interestingly, 80% of non-leaky lesions also contained subretinal vessels that morphologically had the potential to leak fluorescein. That is, like the leaky vessels, the walls of these non-leaky vessels contained diaphragmed fenestrations and intermediate interendothelial cell junctions. Fluorescein leakage was not observed in these vessels due to the absence of a fluid-filled space overlying the subretinal vessels that occurs as the result of serous

44

M. L. Clark et al.

retinal detachment.7,9 Thus, the lack of fluorescein leakage is not always an accurate representation of the absence of subretinal neovascularization. Together, the initial experiments demonstrated that high intensity laserinduced rupture of Bruch’s membrane is a highly effective and reproducible method for inducing CNV. 2.1.2

Advantages and disadvantages of the primate model

The primate model of laser-induced CNV has proven to be a valuable tool for investigating the pathogenesis of CNV, especially given that it is similar in many respects to the disease process in humans. As mentioned previously, clinical laser photocoagulation induces CNV in humans when Bruch’s membrane is ruptured, and the features of this photocoagulation-induced neovascularization are the most similar to those produced in the primate model. Laser-induced CNV mimics many features of CNV resulting from AMD as well. In both cases, new choroidal vessels migrate through holes in Bruch’s membrane into the subretinal space, where fluid accumulates. These new vessels contain fenestrations and interendothelial cell junctions characteristic of choroidal vessels.10,11 Also, in both laser-induced CNV and AMD, polymorphonuclear leukocytes and macrophages can be observed around the budding endothelial cells, and macrophages are also often found around thinned or ruptured areas of Bruch’s membrane, indicating that an intense inflammatory response is a feature of both forms of CNV.8,12-18 The growth factor profiles in CNV resulting from AMD and laser injury are comparable as well. For example, in both cases, immunohistochemistry has revealed that vascular endothelial growth factor (VEGF) is expressed in the RPE and leukocytes. VEGF receptors, basic fibroblast growth factor (bFGF), transforming growth factor-β (TGF-β) and tumor necrosis factorα (TNF-α) are likewise expressed in the same cell types in CNV stimulated by both processes.19 Its similarities to the human condition, as well as its high reproducibility, make the primate model an attractive and highly accepted one in which to study CNV. However, the laser injury model does not perfectly mimic the pathogenesis of CNV in human disease states. Laser-induced CNV is a wounding model, and consequently neovascularization occurs in a manner similar to that which occurs in the process of wound healing. Also, the CNV in the laser lesions regresses with time, or undergoes involution, as demonstrated by decreased vessel leakage in fluorescein angiograms. The cessation of leakage is due to RPE proliferation and subsequent envelopment of the new vessels.10,20 This aspect of laser-induced CNV contrasts with that of the human condition, where CNV is more chronic and leakage can

2. Animal Models of Choroidal Neovascularization

45

continue for years. Furthermore, the induction of CNV in this model, as well as in other experimental models, occurs in relatively young eyes, whereas AMD in humans occurs in the eyes of the elderly. Therefore, when using the laser-induced CNV as a model to study the pathogenesis of CNV as it occurs in humans, the findings must be interpreted with caution, keeping these differences in mind. 2.1.3

Knowledge gained from the primate model

In spite of the discrepancies between the pathogenesis of the CNV induced by laser photocoagulation and by the various human conditions, the primate model has been a valuable tool for increasing our understanding of CNV. From it, a great deal of knowledge has been gained regarding the natural progression of CNV.8-10 It has been used to define the roles of the RPE in reestablishing the blood-retina barrier, in the scarring process, and in involution of new subretinal vessels.10,20 The laser-induced model in primates has also been useful in investigating drug treatments and developing other therapeutic strategies to prevent CNV. For example, intravitreal administration of non-steroidal anti-inflammatory drugs has been shown to prevent angiographic leakage in this model for up to eight weeks,21 and photodynamic therapy using verteporfin prevented angiographic leakage for at least four weeks.22 Laser-induced CNV in primates is an excellent model in which to test the long-term effects of potential therapeutic strategies for AMD prior to the onset of clinical trials.

2.2

Rat

2.2.1

Development of the rat model

The primate model of laser-induced CNV provided the best method of the time for studying subretinal neovascularization. However, expense and availability limited its widespread use, and the need for a rodent model was evident. Pollack and colleagues reported several studies in which laser photocoagulation in rats produced CNV when Bruch’s membrane was breached.23-25 Subsequently, in 1989 a rat model of laser-induced CNV was developed by two different groups, Dobi and colleagues26 and Frank and colleagues,27 by administering krypton laser radiation between the major retinal vessels of the fundus. In this protocol, the animals were anesthetized and pupils dilated, and a handheld coverslip was used as a contact lens for the maintenance of corneal clarity during photocoagulation. The criteria for laser treatment included a small exposure size (100 or 500 μm), a power of

46

M. L. Clark et al.

50 to 100 mW, and an exposure duration of 0.02- 0.1s. Two weeks following this procedure, fluorescein leakage indicated the presence of CNV in 30% of laser-induced lesions, while histological examination revealed that CNV actually occurred in 60% of the laser lesions. Again, this discrepancy is due to the lack of fluid overlying the subretinal vessels in some eyes, a feature necessary for the pooling of dye. These lesions exhibited disruption of Bruch’s membrane and degeneration of RPE as well as the photoreceptors, outer nuclear layer, outer plexiform layer, and part of the inner nuclear layer of the retina. The choroidal capillaries proliferated through the break in Bruch’s membrane into these disrupted outer layers of the retina. Current rat models of laser-induced CNV have evolved from the earliest model. In addition to krypton, argon and diode laser photocoagulators are used, and there are some variations in the specific laser treatment criteria. However, the premise for all of these protocols is the same. A laser beam is focused on Bruch’s membrane with the intention of rupturing it, as evidenced by subretinal bubble formation with or without intraretinal or choroidal hemorrhage at the lesion site. 2.2.2

Advantages and disadvantages of the rat model

In comparison to the primate model, the rat model of laser-induced CNV is advantageous because of the high availability, low cost, and ease of maintenance of rats. The rat model is also more practical for investigating the efficacy of therapeutic strategies in prevention or treatment of CNV, where large sample sizes are beneficial. While it shares many of the benefits attributed to the primate model, the rat model also shares the drawback of producing CNV that, unlike in human disease states, regresses after a short time.28 Furthermore, the rat model is somewhat less ideal for studying CNV as it relates to humans since primate eyes are anatomically and functionally more analogous to human eyes. Nevertheless, the advantages conferred by the rat model far outweigh these disadvantages, causing it to be one of the most extensively used methods for studying CNV today. 2.2.3

Knowledge gained from the rat model

The rat model of laser-induced CNV has been used to obtain much information about the temporal and spatial expression patterns of various growth factors, such as VEGF29,30 and bFGF,31 during the progression of CNV. This knowledge is necessary for increasing our limited understanding of the pathogenesis of CNV. As mentioned previously, the rat model is valuable for investigating the modulation of CNV by various drugs or treatment strategies. It has been utilized to explore the efficacy of anti-

2. Animal Models of Choroidal Neovascularization

47

angiogenic agents administered orally,32 by intravitreal injection,33 and by intravitreal implants.34 The rat model is useful for evaluating the effect of such strategies on CNV before testing them in the primate model or human clinical trials. For example, the value of verteporfin photodynamic therapy was evaluated in this model.35

2.3

Mouse

2.3.1

Development of the mouse model

In 1998, Tobe and colleagues produced a murine model of laser-induced CNV.36 The method for inducing CNV in mice was similar to that which produces CNV in rats. Adult C57BL/6J mice were anesthetized and their pupils dilated. Krypton laser photocoagulation was administered to the posterior retina through a slit lamp using a cover slip as a contact lens. The laser treatment criteria consisted of a spot size of 50 μm, a power of 350-400 mW, and an exposure duration of 0.05 s. As evidenced by bubble formation, Bruch’s membrane was successfully ruptured in 87% of the laser burns. In addition to the disruption of Bruch’s membrane, all layers of the choroid were destroyed within the burn site and ablative damage occurred to the outer retina. Fluorescein leakage and histopathological evidence revealed that over 80% of the lesions contained CNV one week after laser treatment. These new vessels, characterized by large lumens and fenestrations, proliferated into the subretinal space where they were partially enveloped by the RPE. Presently, laser-induced CNV is produced in mice by methods similar to that published by Tobe et al. The laser treatment criteria might have slight variations, and krypton, argon, or diode laser photocoagulators may be used. 2.3.2

Advantages of the mouse model

The mouse model of laser-induced CNV possesses a distinct advantage over the primate and rat models, namely, that manipulation of gene expression is possible. The impact of specific genes on the development of CNV can be evaluated by observing the effects of laser photocoagulation administered to mice overexpressing or underexpressing these genes. The molecular mechanisms underlying the pathogenesis of CNV as well as anti-angiogenic approaches for therapy can thus more readily be explored using the mouse as an experimental animal model in laser-induced CNV.

48 2.3.3

M. L. Clark et al. Knowledge gained from the mouse model

Like the rat model, the mouse model of laser-induced CNV has been implemented to further define the role of various growth factors in the development of CNV. For instance, it has been used to show that FGF2 is not necessary for the occurrence of CNV,36 whereas VEGF is a major stimulator.37 The roles of cellular adhesion molecules have been explored in the mouse model,38 as well as the role of complement, since inflammation is thought to be an important part of the pathogenesis of CNV.39 Furthermore, this model has increasingly been used to test the efficacy of various potential anti-angiogenic treatments. The effect of a non-steroidal anti-inflammatory drug administered topically,40 a kinase inhibitor taken orally,41 and subretinal injection of siRNA targeting VEGF42 are a few examples of therapeutic strategies that have successfully inhibited neovascularization in the mouse model.

2.4

Evaluation of laser-induced CNV

When each of the experimental animal models of laser-induced CNV described above was first introduced, the primary methods for evaluating the extent of subretinal neovascularization were fluorescein angiography and histological examination of serial cross sections. While these methods are still widely used today, they are not without limitations. Leakage of fluorescein is not easily quantified and cannot always be directly correlated to the amount of CNV. Its use is limited in rodent eyes due to difficulty in performing fundus photography and a poor view of the periphery. Quantifying new vessels in histological sections requires thorough sampling of many sections and can be laborious. In 2000, Edelman and Castro introduced a new, high-resolution angiographic method to assess experimentally induced CNV using high molecular weight fluorescein isothiocyanate (FITC)-dextran.43 This method had been previously employed to examine neovascularization from oxygeninduced retinopathy in mouse retinal flatmounts.44 FITC-labeled two million molecular weight dextrans are retained in the blood vessels after fixation allowing the entire vasculature to be viewed by microscopy.45 To examine CNV, FITC-dextran is injected into the left ventricle of animals that have undergone laser photocoagulation. RPE-choroid-sclera flatmounts of the eyes must then be obtained. This is done by hemisecting the eye, peeling away the neural retina and making four incisions in the eyecup in order to flatten it on a microscope slide with the RPE on top. The entire choroid can be visualized by a fluorescence microscope, and whole mount images can be

2. Animal Models of Choroidal Neovascularization

49

captured and analyzed in order to obtain a precise measurement of the neovascular area (Figure 2).

Figure 2-2. FITC-dextran-perfused choroidal flatmount. A diode pumped solid state laser and slit lamp delivery system were used to deliver laser burns to the eyes of C57BL/6J mice. Laser parameters were 50 μm spot size, 0.10 s exposure time, and 150 mW power. Two weeks following laser treatment FITC-dextran (2 x 106 MW) in PBS solution was injected into the mouse via the tail vein. Eyes were enucleated, and choroidal flatmounts were obtained by removing the cornea, iris, and retina and peeling away the retinal pigment epithelium. CNV at one laser lesion site is shown. The green color demonstrates the extent of fluorescently tagged dextran accumulation within the subretinal space.

3.

OTHER ANIMAL MODELS OF CNV

While laser photocoagulation-induced CNV remains a widely utilized model, there are various animal models in which CNV develops spontaneously. These models include transgenic and knockout mice, as well as mice in which the retinas are transfected with relevant growth factors, such as VEGF, FGF, and others. The nature of the animal model depends on the methods used to develop CNV. For example, in several of the transgenic mouse models of CNV, photoreceptor degeneration or a breaching of Bruch’s membrane is necessary for initiation of abnormal vessel growth. Because the progression of CNV is largely dependent on the level of integrity in the barrier between the choroid and the retina, factors that

50

M. L. Clark et al.

contribute to loss of this integrity will have a major impact on the development of CNV.

3.1

Correlation between drusen and CNV

When cells of the RPE layer lose their ability to effectively remove waste produced by the photoreceptors during disk membrane turnover, these materials accumulate and ultimately form localized deposits between the basement membrane of the RPE and Bruch’s membrane. These deposits are commonly known as drusen (singular, druse).19 As drusen continue to accumulate in the subretinal spaces, RPE cell death can occur. This can lead to further photoreceptor damage, since the cells of the RPE layer are essential for filtering out debris to ensure healthy and functional photoreceptors.19 The presence of drusen constitutes a landmark feature of AMD, and in fact, there is a strong correlation between the number of drusen present and the rate of CNV progression. Although the presence of drusen correlates with CNV, it is not the only factor involved. Consequently, AMD animal models, such as the rhesus monkey, where age-related druse accumulation is the primary abnormality observed,46 are not adequate models of CNV. 3.1.1

Ceruloplasmin (Cp) and Hephaetin (Heph) deficient mice

Hahn and colleagues observed the accumulation of iron deposits in retinas and RPE of mice deficient in ceruloplasmin (Cp) and its homolog hephaestin (Heph).47 There is evidence that implicates these proteins in iron export from cells, explaining the increases in retinal iron in these knockout mice. In retinas of these mice, subretinal neovascularization was observed in areas of RPE hyperplasia and photoreceptor degeneration. The source of this neovascularization was not determined by the investigators. The dominating feature of this model is the accumulation of iron, otherwise considered drusen, which further constrains these animals as a model for drusen deposition rather than for CNV. 3.1.2

Ccl-2 and Ccr-2 deficient mice

Ambati and colleagues have recently generated a mouse model that spontaneously develops a clinical syndrome very similar to AMD.48 These mice are deficient in Ccl-2, a monocyte chemoattractant protein-1, and Ccr2, its C-C chemokine receptor-2. Because of their role in recruitment and accumulation of monocytes in various diseases, animals that are deficient in Ccl-2 and Ccr-2 are unable to recruit macrophages that subsequently

2. Animal Models of Choroidal Neovascularization

51

function in degradation and phagocytosis.49 Ambati’s group obtained histopathological sections from the eyes of mutant mice ranging in age from less than 12 months to greater than 24 and compared them to their agematched wild-type controls. The mutant mice exhibited a high frequency of protein complex deposits and pathologies similar to those found in AMD, with photoreceptor and RPE cell death attributed to the progressive subretinal accumulation of these deposits. Because AMD, and more specifically CNV, correlates with age, mice older than 9 months displayed clinical symptoms strikingly similar to those in AMD patients. However, despite the correlation between drusen and CNV progression, this transgenic Ccl-2/Ccr-2 mouse is more typical of a model of drusen deposition rather than of CNV.

3.2

Growth factor driven neovascularization

3.2.1

VEGF overexpression in photoreceptors

Vascular endothelial growth factor (VEGF) is produced by a variety of cells in the retina, including the RPE, and is implicated as a driving force in choroidal neovascularization. Its contribution to the development of CNV is supported by data showing an increase in VEGF mRNA levels in rat RPE following laser-induced CNV.29 To further elucidate the role of VEGF in the progression of CNV, transgenic mice have been generated that overexpress the growth factor in the photoreceptors under the control of the rhodopsin promoter.50 While VEGF overexpression results in retinal neovascularization,50 VEGF alone is not adequate for the induction and subsequent progression of CNV. Overexpression of VEGF must be coupled with photoreceptor cell death before CNV is observed in these animals.19 The usefulness of this model is further limited and complicated by the fact that both deep-retinal neovascularization and choroidal (subretinal) neovascularization can occur (Figure 1). 3.2.2

VEGF overexpression in RPE

Models of CNV involving VEGF overexpression in the RPE cells have also been developed. Campochiaro’s group created a transgenic mouse model where inducible VEGF overexpression in RPE cells is driven by the VMD2 promoter.51 However, these animals exhibited no signs of CNV unless an adenoviral vector containing an expression construct for angiopoietin-2 (Ang2) was injected into the subretinal space. It may be that this injection perturbed the

52

M. L. Clark et al.

RPE, thereby facilitating the occurrence of CNV. In an earlier study by Spilsbury and colleagues, injection of an adenovirus vector expressing VEGF164 cDNA into the subretinal space induced CNV in the rat eye.52 The compromise of the barrier between the retina and the choroid, caused by the needle puncture in this model, may be an important factor in producing CNV in these animals. 3.2.3

Subretinal injection of Matrigel

In a recent study by Qiu and colleagues,53 CNV was induced in rabbits via sub-retinal injection of VEGF-enriched Matrigel growth matrix. While CNV developed in the animals treated with VEGF-enriched matrix, it also developed in those injected with Matrigel alone. The Matrigel serves as a slow-release reservoir of growth factors and a scaffold for growth of subretinal neovascularization. In this model, the inflammatory response to the Matrigel plays a key role in development of CNV. 3.2.4

Prokineticin-1 expression in the retina

Recently, Tanaka and colleagues have produced transgenic mice that express the mitogen prokineticin-1 in the retina.54 Unlike transgenic animals that overexpress VEGF in the photoreceptors with subsequent retinal neovascularization, this model has the added benefit that the effects observed are specific to fenestrated vessels in the choroid. Because this mitogen is not normally expressed in the retina, the rhodopsin promoter was utilized to target its expression in the retina. The retinal vessels were not affected by this mutation, and the animals exhibited no disruption of Bruch’s membrane by choroidal vessels. In fact, the only pathological feature observed that was characteristic of AMD was a thickening of the choroid. Considering the absence of Bruch’s membrane penetration by the choroidal vessels, this transgenic mouse cannot be considered a successful model for CNV.

4.

CONCLUSION

Choroidal neovascularization is a pathological condition in which proliferating choroidal blood vessels grow through Bruch’s membrane, penetrate the RPE, and extend into the subretinal space. There, the blood vessels leak fluid through their fenestrations and interendothelial cell junctions, ultimately leading to serous retinal detachment. CNV associated with the wet form of age-related macular degeneration is the major cause of vision loss in the elderly1 and also plays a major role in other diseases such

2. Animal Models of Choroidal Neovascularization

53

as Sorsby’s fundus dystrophy, Pseudoxanthoma Elasticum, ocular histoplasmosis and multifocal choroiditis.2 However, in spite of its prevalence, relatively little is known concerning the pathogenesis of CNV. In order to better understand this disease process and explore therapies to treat it, several experimental animal models of CNV have been developed. The most widely used of these models is laser-induced CNV in primates and rodents, but several knockout and transgenic mouse models exist as well. While none of these models accurately reproduce all clinical aspects of CNV, they have been successfully implemented to vastly increase our knowledge of new subretinal choroidal vessel formation.

REFERENCES 1. R. Klein, B. E. Klein, and K. L. Linton, Prevalence of age-related maculopathy. The Beaver Dam Eye Study, Ophthalmology 99 (6), 933-943 (1992). 2. Subfoveal neovascular lesions in age-related macular degeneration. Guidelines for evaluation and treatment in the macular photocoagulation study, Macular Photocoagulation Study Group, Arch. Ophthalmol. 109 (9), 1242-1257 (1991). 3. L. A. Yannuzzi, S. Negrào, T. Iida, C. Carvalho, H. Rodriguez-Coleman, J. Slakter, K. B. Freund, J. Sorenson, D. Orlok, and N. Borodoker, Retinal angiomatous proliferation in age-related macular degeneration, Retina 21 (5), 416-434 (2001). 4. J. D. Gass, Chorioretinal anastomosis probably occurs infrequently in type 2A idiopathic juxtafoveolar retinal telangiectasis, Arch. Opthtalmol. 121 (9), 1345-1346 (2003). 5. J. Francois, J. J. De Laey, E. Cambie, M. Hanssens, and V. Victoria-Troncoso, Neovascularization after argon laser photocoagulation of macular lesions, Am. J. Ophthalmol. 79 (2), 206-210 (1975). 6. S. J. Ryan, Subretinal neovascularization. Natural history of an experimental model, Arch. Ophthalmol. 100 (11), 1804-1809 (1982). 7. H. Miller, B. Miller, and S. J. Ryan, Correlation of choroidal subretinal neovascularization with fluorescein angiography, Am. J. Ophthalmol. 99 (3), 263-271 (1985). 8. T. Ishibashi, H. Miller, G. Orr, N. Sorgente, and S. J. Ryan, Morphologic observations on experimental subretinal neovascularization in the monkey, Invest. Ophthalmol. Vis. Sci. 28 (7), 1116-1130 (1987). 9. H. Miller, B. Miller, and S. J. Ryan, Newly-formed subretinal vessels. Fine structure and fluorescein leakage, Invest. Ophthalmol. Vis. Sci. 27 (2), 204-213 (1986). 10. H. Miller, B. Miller, T. Ishibashi, and S. J. Ryan, Pathogenesis of laser-induced choroidal subretinal neovascularization, Invest. Ophthalmol. Vis. Sci. 31 (5), 899-908 (1990). 11. S. H. Sarks, D. Van Driel, L. Maxwell, and M. Killingsworth, Softening of drusen and subretinal neovascularization, Trans. Ophthalmol. Soc. U.K. 100 (3), 414-422 (1980). 12. M. C. Killingsworth, J. P. Sarks, and S. H. Sarks, Macrophages related to Bruch’s membrane in age-related macular degeneration, Eye 4 (Pt 4), 613-621 (1990). 13. T. L. van der Schaft, C. M. Mooy, W. C. de Bruijn, and P. T. de Jong, Early stages of age-related macular degeneration: an immunofluorescence and electron microscopy study, Br. J. Ophthalmol. 77 (10), 657-661 (1993). 14. H. E. Grossniklaus, K. A. Cingle, Y. D. Yoon, N. Ketkar, N. L’Hernault, and S. Brown, Correlation of histologic 2-dimensional reconstruction and confocal scanning laser

54

15. 16.

17.

18.

19.

20.

21

22.

23.

24.

25.

26. 27. 28.

29.

30.

M. L. Clark et al. microscopic imaging of choroidal neovascularization in eyes with age-related maculopathy, Arch. Ophthalmol. 118 (5), 625-629 (2000). T. Nishimura, R. Goodnight, R. A. Prendergast, and S. J. Ryan, Activated macrophages in experimental subretinal neovascularization, Ophthalmologica 200 (1), 39-44 (1990). H. Oh, H. Takagi, C. Takagi, K. Suzuma, A. Otani, K. Ishida, M. Matsumura, Y. Ogura, and Y. Honda, The potential angiogenic role of macrophages in the formation of choroidal neovascular membranes, Invest. Ophthalmol. Vis. Sci. 40 (9), 1891-1898 (1999). P. L. Penfold, J. M. Provis, and F. A. Billson, Age-related macular degeneration: ultrastructural studies of the relationship of leucocytes to angiogenesis, Graefes. Arch. Clin. Exp. Ophthalmol. 225 (1), 70-76 (1987). J. P. Sarks, S. H. Sarks, and M. C. Killingsworth, Morphology of early choroidal neovascularisation in age-related macular degeneration: correlation with activity, Eye 11 (Pt 4), 515-522 (1997). J. Ambati, B. K. Ambati, S. H. Yoo, S. Ianchulev, and A. P. Adamis, Age-related macular degeneration: etiology, pathogenesis, and therapeutic strategies, Surv. Ophthalmol. 48 (3), 257-293 (2003). H. Miller, B. Miller, and S. J. Ryan, The role of retinal pigment epithelium in the involution of subretinal neovascularization, Invest. Ophthalmol. Vis. Sci. 27 (11), 1644-1652 (1986). T. Sakamoto, D. Soriano, J. Nassaralla, T. L. Murphy, A. Oganesian, C. Spee, D. R. Hinton, and S. J. Ryan, Effect of intravitreal administration of indomethacin on experimental subretinal neovascularization in the subhuman primate, Arch. Ophthalmol. 113 (2), 222-226 (1995). D. Husain, M. Kramer, A. G. Kenny, N. Michaud, T. J. Flotte, E. S. Gragoudas, and J. W. Miller, Effects of photodynamic therapy using verteporfin on experimental choroidal neovascularization and normal retina and choroid up to 7 weeks after treatment, Invest. Ophthalmol. Vis. Sci. 40 (10), 2322-2331 (1999). A. Pollack, G. E. Korte, A. L. Weitzner, and P. Henkind, Ultrastructure of Bruch’s membrane after krypton laser photocoagulation. I. Breakdown of Bruch’s membrane, Arch. Ophthalmol. 104 (9), 1372-1376 (1986). A. Pollack, G. E. Korte, W. J. Heriot, and P. Henkind, Ultrastructure of Bruch’s membrane after krypton laser photocoagulation. II. Repair of Bruch’s membrane and the role of macrophages, Arch. Ophthalmol. 104 (9), 1377-1382 (1986). A. Pollack, W. J. Heriot, Fraco, Fracs, and P. Henkind, Cellular processes causing defects in Bruch’s membrane following krypton laser photocoagulation, Ophthalmology 93 (8), 1113-1119 (1986). E. T. Dobi, C. A. Puliafito, and M. Destro, A new model of experimental choroidal neovascularization in the rat, Arch. Ophthalmol. 107 (2), 264-269 (1989). R. N. Frank, A. Das, and M. L. Weber, A model of subretinal neovascularization in the pigmented rat, Curr. Eye. Res. 8 (3), 239-247 (1989). A. Pollack and G. E. Korte, Repair of retinal pigment epithelium and its relationship with capillary endothelium after krypton laser photocoagulation, Invest. Ophthalmol. Vis. Sci. 31 (5), 890-898 (1990). X. Yi, N. Ogata, M. Komada, C. Yamamoto, K. Takahashi, K. Omori, and M. Uyama, Vascular endothelial growth factor expression in choroidal neovascularization in rats, Graefes. Arch. Clin. Exp. Ophthalmol. 235 (5), 313-319 (1997). M. Wada, N. Ogata, T. Otsuji, and M. Uyama, Expression of vascular endothelial growth factor and its receptor (KDR/flk-1) mRNA in experimental choroidal neovascularization, Curr. Eye Res. 18 (3), 203-213 (1999).

2. Animal Models of Choroidal Neovascularization

55

31. N. L. Zhang, E. E. Samadani, and R. N. Frank, Mitogenesis and retinal pigment epithelial cell antigen expression in the rat after krypton laser photocoagulation, Invest. Ophthalmol. Vis. Sci. 34 (8), 2412-2424 (1993). 32. F. Kinose, G. Roscilli, S. Lamartina, K. D. Anderson, F. Bonelli, S. G. Spence, G. Ciliberto, T. F. Vogt, D. J. Holder, C. Toniatti, and C. J. Thut, Inhibition of retinal and choroidal neovascularization by a novel KDR kinase inhibitor, Mol. Vis. 11, 366-373 (2005). 33. M. El Bradey, L. Cheng, D. U. Bartsch, K. Appelt, N. Rodanant, G. Bergeron-Lynn, and W. R. Freeman, Preventive versus treatment effect of AG3340, a potent matrix metalloproteinase inhibitor in a rat model of choroidal neovascularization, J. Ocul. Pharmacol. Ther. 20 (3), 217-236 (2004). 34. M. R. Robinson, J. Baffi, P. Yuan, C. Sung, G. Byrnes, T. A. Cox, and K. G. Csaky, Safety and pharmacokinetics of intravitreal 2-methoxyestradiol implants in normal rabbit and pharmacodynamics in a rat model of choroidal neovascularization, Exp. Eye Res. 74 (2), 309-317 (2002). 35. D. N. Zacks, E. Ezra, Y. Terada, N. Michaud, E. Connolly, E. S. Gragoudas, and J. W. Miller, Verteporfin photodynamic therapy in the rat model of choroidal neovascularization: angiographic and histologic characterization, Invest. Ophthalmol. Vis. Sci. 43 (7), 2384-2391 (2002). 36. T. Tobe, S. Ortega, J. D. Luna, H. Ozaki, N. Okamoto, N. L. Derevjanik, S. A. Vinores, C. Basilico, and P. A. Campochiaro, Targeted disruption of the FGF2 gene does not prevent choroidal neovascularization in a murine model, Am. J. Pathol. 153 (5), 1641-1646 (1998). 37. N. Kwak, N. Okamoto, J. M. Wood, and P. A. Campochiaro, VEGF is major stimulator in model of choroidal neovascularization, Invest. Ophthalmol. Vis. Sci. 41 (10), 3158-3164 (2000). 38. E. Sakurai, H. Taguchi, A. Anand, B. K. Ambati, E. S. Gragoudas, J. W. Miller, A. P. Adamis, and J. Ambati, Targeted disruption of the CD18 or ICAM-1 gene inhibits choroidal neovascularization, Invest. Ophthalmol. Vis. Sci. 44 (6), 2743-2749 (2003). 39. P. S. Bora, J. H. Sohn, J. M. Cruz, P. Jha, H. Nishihori, Y. Wang, S. Kaliappan, H. J. Kaplan, and N. S. Bora, Role of complement and complement membrane attack complex in laserinduced choroidal neovascularization, J. Immunol. 174 (1), 491-497 (2005). 40. K. Takahashi, Y. Saishin, Y. Saishin, K. Mori, A. Ando, S. Yamamoto, Y. Oshima, H. Nambu, M. B. Melia, D. P. Bingaman, and P. A. Campochiaro, Topical nepafenac inhibits ocular neovascularization, Invest. Ophthalmol. Vis. Sci. 44 (1), 409-415 (2003). 41. M. S. Seo, N. Kwak, H. Ozaki, H. Yamada, N. Okamoto, E. Yamada, D. Fabbro, F. Hofmann, J. M. Wood, and P. A. Campochiaro, Dramatic inhibition of retinal and choroidal neovascularization by oral administration of a kinase inhibitor, Am. J. Pathol. 154 (6), 1743-53 (1999). 42. S. J. Reich, J. Fosnot, A. Kuroki, W. Tang, X. Yang, A. M. Maguire, J. Bennett, and M. J. Tolentino, Small interfering RNA (siRNA) targeting VEGF effectively inhibits ocular neovascularization in a mouse model, Mol. Vis. 9, 210-216 (2003). 43. J. L. Edelman and M. R. Castro, Quantitative image analysis of laser-induced choroidal neovascularization in rat, Exp. Eye Res. 71 (5), 523-533 (2000). 44. L. E. Smith, E. Wesolowski, A. McLellan, S. K. Kostyk, R. D’Amato, R. Sullivan, and P. A. D’Amore, Oxygen-induced retinopathy in the mouse, Invest. Ophthalmol. Vis. Sci. 35 (1), 101-111 (1994). 45. R. D’Amato, E. Wesolowski, and L. E. Smith, Microscopic visualization of the retina by angiography with high-molecular-weight fluorescein-labeled dextrans in the mouse, Microvasc. Res. 46 (2), 135-142 (1993).

56

M. L. Clark et al.

46. R. J. Ulshafer, H. M. Engel, W. W. Dawson, C. B. Allen, and M. J. Kessler, Macular degeneration in a community of rhesus monkeys. Ultrastructural observations, Retina 7 (3), 198-203 (1987). 47. P. Hahn, Y. Qian, T. Dentchev, L. Chen, J. Beard, Z. L. Harris, and J. L. Dunaief, Disruption of ceruloplasmin and hephaestin in mice causes retinal iron overload and retinal degeneration with features of age-related macular degeneration, Proc. Natl. Acad. Sci. U. S. A. 101 (38), 13850-13855 (2004). 48. J. Ambati, A. Anand, S. Fernandez, E. Sakurai, B. C. Lynn, W. A. Kuziel, B. J. Rollins, and B. K. Ambati, An animal model of age-related macular degeneration in senescent Ccl-2- or Ccr-2-deficient mice, Nat. Med. 9 (11), 1390-1397 (2003). 49. B. Sar, K. Oishi, A. Wada, T. Hirayama, K. Matsushima, and T. Nagatake, Induction of monocyte chemoattractant protein-1 (MCP-1) production by Pseudomonas nitrite reductase in human pulmonary type II epithelial-like cells, Microb. Pathog. 28 (1), 17-23 (2000). 50. N. Okamoto, T. Tobe, S. F. Hackett, H. Ozaki, M. A. Vinores, W. LaRochelle, D. J. Zacks, and P. A. Campochiaro, Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization, Am. J. Pathol. 151 (1), 281-291 (1997). 51. Y. Oshima, S. Oshima, H. Nambu, S. Kachi, S. F. Hackett, M. Melia, M. Kaleko, S. Connelly, N. Esumi, D. J. Zack, and P. A. Campochiaro, Increased expression of VEGF in retinal pigmented epithelial cells is not sufficient to cause choroidal neovascularization, J. Cell Physiol. 201 (3), 393-400 (2004). 52. K. Spilsbury, K. L. Garrett, W. Y. Shen, I. J. Constable, and P. E. Rakoczy, Overexpression of vascular endothelial growth factor (VEGF) in the retinal pigment epithelium leads to development of choroidal neovascularization, Am. J. Pathol. 157 (1), 135-144 (2000). 53. G. Qiu, J. M. Stewart, S. Sadda, R. Freda, S. Lee, D. Guven, E. de Juan, Jr., and S. E. Varner, A new model of experimental subretinal neovascularization in the rabbit, Exp. Eye Res. 83 (1), 141-152 (2006). 54. N. Tanaka, M. Ikawa, N. L. Mata, and I. M. Verma, Choroidal neovascularization in transgenic mice expressing prokineticin 1: an animal model for age-related macular degeneration, Mol. Ther. 13 (3), 609-616 (2006).

Chapter 3 RODENT MODELS OF OXYGEN-INDUCED RETINOPATHY

Susan E. Yanni,1 Gary W. McCollum,2 and John S. Penn1,2

Departments of 1Cell & Developmental Biology and 2Ophthalmology and Visual Sciences, Vanderbilt University School of Medicine, Nasvhille, Tennessee

Abstract:

1.

Retinopathy of prematurity (ROP), a condition affecting premature infants, is characterized by pathological ocular angiogenesis, or retinal neovasculariztion (NV). Much of what is known about the development of the retinal vasculature and the progression of ROP has been acquired through the use of animal models of oxygen-induced retinopathy (OIR), which approximate ROP. Animal models of OIR have provided a wealth of information regarding the cellular and molecular pathogenesis of ROP. This information has contributed to a better understanding of other, non-ocular, neovascular conditions. The aim of this chapter is to explore the significance of the two most prevalent animal models of OIR, the mouse and the rat.

BACKGROUND

In 1942, Terry first described ROP as a disease of prematurity, characterized by retinal neovascularization.1 An epidemic of ROP occurred during the 1950’s, exposing the need for research focused on the identification and characterization of the pathogenesis of ROP. In 1951, Campbell proposed that the incidence of ROP was linked to the supplemental oxygen administered to premature infants with under-developed pulmonary function.2 During the 1950’s, several convincing studies correlated the use of supplemental oxygen with the incidence and progression of ROP.3-7 This led to the rigorous monitoring of the oxygen being given to premature infants. Consequently, the percentage of blindness attributed to ROP dropped from 50% in 1950 to just 4% in 1965.8 The 1970’s and 1980’s saw 57 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 57–80. © Springer Science+Business Media B.V. 2008

58

S. E. Yanni et al.

an increased incidence of ROP,9 presumably from the increased survival of very low-birth-weight premature infants. According to the most recent estimates of the National Eye Institute, each year approximately 14,000-16,000 premature infants (classified as those weighing 1250 grams or less, and being born prior to 31 weeks’ gestation) develop some stage of ROP. Of these infants, 400-600 will suffer from ROPinduced blindness. ROP is the leading cause of childhood blindness in the developed world.10 For this reason, among others, research focused on understanding physiological and pathological retinal neovascularization is highly significant. Several animal models have been developed that approximate human ROP. To emphasize the differences between human ROP and experimentally induced retinopathy in animals, the term oxygen-induced retinopathy (OIR) is often used. Rodent models of OIR are widely used to study the cellular and molecular aspects of physiological and pathological retinal neovascularization.

1.1

Normal human retinal vascularization

The retina is one of the last organ systems of the developing fetus to undergo vascularization, beginning at approximately 16 weeks’ gestation. At this time, vasculogenesis (the de novo formation of blood vessels from mesodermal precursor cells) occurs, beginning in the most posterior region of the superficial retina (the optic disk) and proceeding to the periphery. At 25 weeks’ gestation, angiogenesis (the development of new capillaries from pre-existing vessels) begins, proceeding also from the optic disk in a peripheral wave, resulting in the development of a deeper (more sclerad) vessel network. It is believed that the hypoxic uterine environment (30 mm Hg) drives retinal vascularization during normal gestation. In utero, retinal hypoxia induces pro-angiogenic growth factors that stimulate the growth of retinal blood vessels. These blood vessels satisfy the increasing demands of the developing fetal retina for oxygen. Complete vascularization is attained at approximately 36-40 weeks’ gestation, and the relatively hyperoxic (55-80 mm Hg) postnatal environment effectively prevents further vasoproliferation.11,12

1.2

The pathogenesis of ROP

The pathogenesis of ROP is biphasic. The first phase of ROP results in the vasoattenuation and pruning of the existing vasculature. This is followed by the second, proliferative phase, characterized by retinal neovascularization.7,13-15 Vasoattenuation is a cessation of the retinal vascularization

3. Rodent Models of Oxygen-Induced Retinopathy

59

process, which occurs after the infant has been placed on supplemental oxygen therapy. At this time, the oxygen tension within the retina sufficiently inhibits the hypoxia-induced production and secretion of vascular growth factors. Diminished growth factor production results in an incompletely vascularized retinal periphery, a hallmark of ROP. Vasoattenuation results in retinal avascularity. Retinal avascularity results in retinal ischemia when oxygen supplementation to the infant is discontinued because the development and maturation of the neural components within the retina demand more oxygen than they are receiving. At this time, retinal hypoxia ensues. Retinal hypoxia induces the onset of the second, vasoproliferative phase of ROP. Vasoproliferation is best described as deregulated angiogenesis, resulting in the production of fragile, non-patent vascular structures that grow through the inner limiting membrane of the retina into the vitreous cavity. These abnormal vascular structures are often referred to as preretinal neovascular tufts, and they predispose affected infants to intravitreal hemorrhages, retinal detachment, and subsequent vision loss. The severity of ROP is inversely proportional to gestational age.16 Because retinal vascularization is completed at, or near, the time of birth, premature infants demonstrate an increased area of retinal avascularity. Placing these infants in a post-natal hyperoxic environment leads to vasoattenuation of the already sparse vasculature. Returning the infants to a hypoxic room air environment leads to retinal hypoxia and the subsequent development of ROP. The larger the avascular area at the time of birth, the more severe the retinal hypoxia upon return to room air, and hence, the more severe the ROP. Roughly half the infants that develop ROP do so while receiving supplemental oxygen therapy. Hypoxia or variable oxygen, therefore, is not the sole determinant in the pathogenesis of ROP. Developmental timing may regulate the responses of the immature retina to oxygen. ROP involves a complex sequence of pathological events.

2.

RODENT MODELS OF ROP

2.1

Mouse

2.1.1

Mouse vascular development

Unlike the human, whose retinal vasculature derives from spindle-shaped mesenchymal precursor cells of the hyaloid artery in a vasculogenic process, research has provided evidence that the retinal vasculature of the mouse

60

S. E. Yanni et al.

derives from immature retinal astrocytes in an angiogenic process.17 The contributions of vasculogenesis and angiogenesis to retinal vascularization may be species-specific.18 Regardless, the retinal vasculature of a newborn mouse is comparable to that of an infant at 25 weeks’ gestation who is at risk for developing ROP.19 For this reason, the retinal vasculature of the newborn mouse pup is an attractive model of the premature infant’s retinal vasculature. 2.1.2

Earliest mouse model

After the initial identification of ROP, experiments were conducted in both laboratory and clinical settings to ascertain the effect of oxygen therapy on retinal angiogenesis. In 1954, Gyllensten and Hellstrom exposed newborn mouse pups to 100% oxygen for 1-3 weeks. Ocular examination after oxygen withdrawal revealed that approximately one-third of the animals experienced hemorrhages in both the vitreous and the anterior chamber. It was further demonstrated that exposing the pups to 100% oxygen and subsequently removing the pups to room air for 5 days induced vasoproliferation of the retinal vessels, a hallmark of ROP.19 It should be noted that removal to room air was required for induction of the ROP-like vasoproliferative changes.20 2.1.3

Current mouse model

Gyllensten and Hellstrom provided the research community with a means to explore ROP in greater detail. Early studies were inconclusive, yielding highly varied results. One of the confounding factors in the early attempts to model ROP was the fact that hyperoxic exposure of newborn mice, followed by removal to room air, resulted in the proliferation and engorgement of the hyaloid.21,22 Reasoning that the hyaloidopathy might explain the observed variability in the early attempts to model ROP, Smith and colleagues proposed a novel method for inducing retinopathy in the mouse, a model that sought to minimize any hyaloidopathy.23 The Smith model allows a consistent and reliable reproduction of ROP. The widely used method involves exposing mice at postnatal day 7 (P7) to 75% oxygen for 5 days to induce vaso-attenuation and atrophy of the centralized portion of the retinal vascular bed (Figure 1). Removal of the mice to room air for variable lengths of time induces retinal vasoproliferation and revascularization of the central retina. At P17-P21, the eyes of the mice are analyzed for the presence of neovascularization (Figure 2).

3. Rodent Models of Oxygen-Induced Retinopathy

61

Figure 3-1. FITC-dextran infused mouse retinae at P12. Normoxia-raised mice (A) exhibit normal retinal vascular development. 5 days at 75% O2 (B) induces vaso-attenuation and atrophy of the central retinal bed, as well as substantial vascular leakage.

Since the advent and widespread use of the mouse model of retinopathy, extensive research has been conducted on the susceptibility of various inbred strains of mice to pathological retinal neovascularization. In order to evaluate genetic heterogeneity in angiogenic susceptibility, D’Amato and colleagues24 implanted a pellet containing an angiogenic protein, basic fibroblast growth factor (bFGF), into the corneas of 25 strains of mice. Normally, the blood vessels of the limbus do not grow into the avascular cornea. Strain differences in angiogenic response were observed by analyzing the growth of blood vessels into the cornea upon angiogenic stimulation. A 10-fold range of responsiveness was observed, with 129/SvImJ mice eliciting the most potent angiogenic response, while the commonly used C57BL/6J mice fell near the middle of the response profile. Subsequent studies revealed that vascular endothelial growth factor (VEGF) elicited a response profile that correlated closely with the bFGF response.25 Following D’Amato’s report, Hinton and colleagues analyzed strainrelated differences in retinal angiogenesis using the mouse OIR model.26 They demonstrated that the angiogenic response of the retina paralleled the results of the corneal assay. Analyzing retinal angiogenesis in mice with different genetic backgrounds has allowed for the identification of various pro- and anti-angiogenic factors (to be discussed in detail later) potentially involved in the pathogenesis of ROP.

62

S. E. Yanni et al.

Figure 3-2. P19 hematoxylin and eosin (H&E) stained retinal cross sections. In contrast to normoxia-raised mice (A), mice that have been exposed to hyperoxia (B) demonstrate a substantial number of retinal cell nuclei that penetrate the inner limiting membrane. These nuclei allow for retinal neovascularization to be quantified.

2.1.4

Disadvantages of the mouse model

Unfortunately, the retinal vascular pathology observed in mouse OIR is the opposite of that observed in human ROP. In the human condition, the central retina is vascularized, and the peripheral retina is avascular. In contrast, the mouse exhibits a central area of avascularity, and the peripheral retina is vascularized. These differences in patterning cannot be ascribed to any obvious differences between these two species. Claxton and Fruttiger27 studied the retinal vascular patterning in mice that had been exposed to hyperoxia. They hypothesized that because the retinal arteries and the hyaloidal blood supply pass through the optical nerve head, it is reasonable to suspect that the proximal retina has a relatively high oxygen tension. VEGF is a survival factor for endothelial cells and is downregulated in response to high oxygen tension, explaining the pruning of peri-arterial capillaries around the optic nerve head as well as around the central retinal arteries in the mouse. Hyperoxic exposure further increases the retinal oxygen tension, expanding regions of decreased retinal VEGF, inducing endothelial cell apoptosis and vascular atrophy, and resulting in the expansion of capillary-free zones within the retina. In addition, and in contrast to human ROP, the mouse OIR model does not lead to retinal detachment. The lens occupies 40% of the mouse eye,

3. Rodent Models of Oxygen-Induced Retinopathy

63

resulting in less tractional force on the retina compared to that observed in humans. 2.1.5

Advantages of the mouse model

The mouse model of ROP is the most commonly used model in studies of retinal angiogenesis. Mice reliably produce large litters, are relatively inexpensive to purchase and maintain, and consistently produce a neovascular response. The mouse model of ROP has provided much of what is currently known about the pathogenesis of retinopathy of prematurity, its progression, and potential means by which to prevent and/or ameliorate it. Importantly, the ability to manipulate the mouse genome has facilitated our understanding of the various genetic contributions and their interactions in producing the angiogenic phenotype.

2.2

Rat

2.2.1

Rat vascular development

As in the human, the retinal vasculature of the rat appears to derive from the adventitia of the hyaloid artery. Vascularization of a superficial network of arteries and veins occurs first, followed by the angiogenic vascularization of a deeper capillary network. In the human, retinal vascularization is usually complete at the time of birth. This process does not complete until postnatal day 15 (P15) in the rat. For this reason, the retinal vasculature of the newborn rat pup resembles that of a preterm infant and a newborn mouse— incomplete, largely avascular, and susceptible to OIR. 2.2.2

Current rat model

Early studies by Patz, Ashton, and Gole28-30 involved exposing newborn rat pups to a constant level of extreme hyperoxia. This resulted in substantial vasoattenuation but an inconsistent vasoproliferative response. Informative though these studies were, it was not until 1993 that Penn and colleagues developed a protocol that consistently induced proliferative retinopathy in the rat.31 Penn noted that variable oxygenation is more likely to produce retinal angiogenesis than is constant hyperoxia. This is because variable oxygenation more closely mimics the fluctuating lung function and

64

S. E. Yanni et al.

subsequent change in arterial blood oxygen partial pressure, PaO2, of a neonatal infant in the NICU, an infant likely to develop ROP. In Penn’s 1993 study, exposing rats to 80% oxygen throughout the course of treatment did not induce any pre-retinal neovascularization following an appropriate postexposure period. However, a variable oxygen exposure (cycling between 40% and 80% oxygen every 12 hours), in combination with a post-exposure period of return to room air, induced pre-retinal neovascularization in 66% of the rats. Subsequent experiments by Penn32 led to the rat OIR model that is used today. In this model, newborn rats are cycled between 10% and 50% oxygen every 24 hours for 14 days. This oxygen profile, which more accurately reflects the fluctuating lung function and PaO2 of a preterm infant in the NICU, resulted in a high incidence (97%) of retinopathy (Figure 3). Additionally, the angiogenic pattern seen in the rat mimics the pattern of ROP seen in the human. Both exhibit a peripheral region of avascularity and develop neovascularization at the boundary of vascular and avascular retina (Figure 4). Thus, the rat provides an extremely relevant model with which to address ROP-related questions.

Figure 3-3. FITC-dextran infused rat retinae at P20. Normoxia-raised rats (A) exhibit normal retinal vascular development. The rat OIR model causes avascularity of the peripheral retina (B).

3. Rodent Models of Oxygen-Induced Retinopathy

65

Figure 3-4. OIR-exposed rat retina at P20, stained with ADPase, an ablumenal enzyme marking the vascular endothelium. Neovascularization develops at the boundary of vascular and avascular retina, as in human ROP. Image reproduced with Permission from Investigative Ophthalmology and Visual Sciences.111

2.2.3

Disadvantages of the rat model

Like the mouse model, the rat OIR model is subject to both strain- and vendor-related differences in susceptibility to retinal neovascularization. Ma and colleagues33 compared the differential susceptibilities of two strains of rats, Brown Norway and Sprague Dawley, to ischemia-induced retinopathy. Using a modified constant oxygen exposure paradigm (developed by Smith and colleagues for the mouse), Ma found that at the time of removal to room air, the Brown Norway rats exhibited an avascular area approximately four times greater than that of the Sprague Dawley rats. The Brown Norway rats subsequently developed three times the amount of preretinal neovascularization. Later studies confirmed the above findings by demonstrating that Brown Norway rats exhibit an increased amount and duration of retinal vascular permeability relative to Sprague Dawley rats exposed to the same ROP paradigm.34 The difference between the two strains is likely due to differences in retinal expression of pro- and antiangiogenic factors, as has been demonstrated in the mouse.26,33 These two studies, though informative, were conducted under conditions of constant, extreme hyperoxia, instead of the more clinically relevant variable oxygen protocol. To address this issue, Holmes and colleagues used a modified protocol of cyclic hyperoxia and hypoxia. The Brown Norway strain again

66

S. E. Yanni et al.

demonstrated a higher incidence and severity of neovascularization than did the Sprague Dawley strain.35 Only a few Sprague Dawley rats, as opposed to all of the Brown Norway rats, developed neovascularization. There are also vendor-related differences in susceptibility within the same strain of rat. Penn (unpublished observations) was the first to identify differences in the pathological response of a single rat strain obtained from several different vendors. Sprague Dawley rats from Charles River (Charles River Laboratories, Wilmington, MA) produced a two-fold greater area of neovascularization than those from Zivic-Miller (Zivic Laboratories, Pittsburg, PA). Sprague Dawley rats obtained from Harlan (Harlan, Indianapolis, IN) and Hilltop (Hilltop Lab Animals, Scottdale, PA) demonstrated intermediate levels of pathology compared to Charles River and Zivic-Miller rats. Similarly, Holmes and colleagues tested the OIR response of Sprague Dawley rats from Harlan and Charles River. 36 Notably, the Charles River rats demonstrated a 62% greater susceptibility to and severity of oxygen-induced neovascularization. Thus, susceptibility to neovascularization depends on genetic variation, environment, and oxygen treatment paradigm. 2.2.4

Advantages of the rat model

The rat is an ideal model of retinopathy due to large litter sizes (typically twice the size of mouse litters) and relatively inexpensive maintenance costs. Most importantly, unlike the mouse, the rat model consistently produces human-like patterns of vasoattenuation and vasoproliferation. For this reason, the rat is an attractive model that is often used for testing the efficacy of anti-angiogenic compounds for application in both ocular and non-ocular pathologies. It should also be noted that the rat model of ROP was the first to utilize fluoroscein angiogram imaging in order to track the progression of the disease in real time,37-39 and the first to utilize computer assisted image analysis in order to improve the speed and objectivity of pathology assessment.40

3.

MOLECULAR MECHANISMS OF ANGIOGENESIS

Angiogenesis is the growth of new capillaries from pre-existing blood vessels. Angiogenesis involves complex interactions between cells, growth factors, cytokines and extracellular matrix (ECM) components. Ischemia is a feature common to virtually all retinal vasculopathies, and this observation formed the basis of early hypotheses suggesting the presence

3. Rodent Models of Oxygen-Induced Retinopathy

67

of hypoxia-induced retinal angiogenic factors. Beginning with ischemic insult, the angiogenic cascade involves: the hypoxia-induced expression of pro-angiogenic growth factors and cytokines; proteolytic degradation of the vascular basement membrane by endothelial cell-derived matrix metalloproteinases; proliferation and migration of invading endothelial cells to form and extend the new vasculature; and morphogenic stabilization involving the induction of vessel differentiation, matrix deposition, and mural cell recruitment. The molecular etiology of retinal angiogenesis will be discussed in the following sections.

3.1

Angiogenic factors

The growth of new blood vessels can be stimulated by a number of angiogenic factors including, but not limited to, vascular endothelial cell growth factor (VEGF), basic fibroblast growth factor (bFGF), hepatocyte growth factor (HGF), insulin-like growth factor (IGF), placental growth factor (PlGF), platelet-derived growth factor (PDGF), transforming growth factor-beta (TGFβ), tumor necrosis factor (TNF), nuclear factor-kappaB (NFκB), and interleukin-8 (IL-8).11 However, experimental evidence suggests that VEGF is the most important pro-angiogenic factor in the pathogenesis of vasoproliferative retinopathies. 3.1.1

Vascular endothelial growth factor

In 1954, Michaelson proposed the presence of a “vasoformative factor” produced in response to the retina’s metabolic needs. This factor was proposed to be involved in the physiological development of the retina, as well as the pathological angiogenesis that occurs, for example, in ROP.41 Thirty years later, the key vasoformative factor involved in retinal angiogenesis was discovered to be VEGF (formerly vascular permeability factor, or VPF).42,43 VEGF induces endothelial cell proliferation and tube formation in vitro and stimulates angiogenesis in vivo.44-47 VEGF binds to cell surface receptor tyrosine kinases VEGFR1/Flt-1 and VEGFR2/KDR/Flk-1, stimulating the angiogenic cascade.47 In fact, both animal models and human patients suffering from ocular vasoproliferative disorders exhibit elevated levels of VEGF within the eye.45,48-50 Pierce and colleagues investigated the effect of hyperoxia and hypoxia on VEGF in the mouse.51 At P7, VEGF mRNA was localized just anterior to the developing vasculature. Exposing the mice to 75% oxygen for just six hours led to substantial vasoattenuation with a reduction in VEGF mRNA as measured by in situ hybridization. Claxton and Fruttiger showed a similar reduction in VEGF expression following hyperoxic exposure.27 These studies

68

S. E. Yanni et al.

demonstrate that exposure to hyperoxia suppresses retinal VEGF mRNA expression and vascular growth. Conversely, VEGF is induced by hypoxia.45,52 Studies have shown that post-oxygen exposed mice show increased levels of VEGF mRNA. This increased expression is presumably induced by the onset of retinal hypoxia resulting from both the hyperoxiainduced vasoattenuation and the relatively hypoxic room air environment.53,54 Hypoxia-induced VEGF expression is a key event promoting vasoproliferation. Increased expression of VEGF in hypoxia appears to be mediated by the transcription factor hypoxia-inducible factor-1 (HIF-1). HIF-1 binds to the hypoxia response element (HRE) on several hypoxia-inducible genes. The promoter sequence of the VEGF gene contains several HREs. Ozaki et al. have demonstrated a functional link between levels of HIF-1 and VEGF transcription in the development of the mouse vasculature.55,56 Post-oxygen exposed mice, whose retinas are presumably hypoxic, demonstrate substantially increased HIF-1 levels that are temporally and spatially correlated with VEGF expression and retinal vasoproliferation. VEGF plays a prominent role in retinal vasoproliferation. For this reason, several groups have already begun to explore means by which to inhibit VEGF, as a prelude to developing human therapies. Studies have been conducted using antisense oligonucleotides, monoclonal antibodies, VEGF peptides, and chimeric proteins that inhibit VEGF binding to its receptor.57-60 These studies, conducted in the mouse and/or rat, have been effective at reducing retinal neovascularization. Inhibition of KDR and the use of soluble Flt-1 effectively reduces the severity of retinal neovascularization in the rat.61,62 Therapies directed against VEGF are being used to treat human patients suffering from other forms of ocular neovascularization.63,64 3.1.2

Insulin-like growth factor

VEGF is not the only growth factor involved in ocular neovascularization. In 1969, research demonstrated that removing the pituitary gland had a restorative effect on proliferative diabetic retinopathy.65 This finding led to the hypothesis that insulin-like growth factor-1 (IGF1), a product of the pituitary gland, must play a role in retinal neovascularization. Using the mouse OIR model, Smith and colleagues demonstrated that exogenous IGF1 induced retinal neovascularization when growth hormone (GH) was inhibited, and that an IGF-1 receptor antagonist suppressed retinal neovascularization.66,67 Moreover, knockout mice lacking the vascular endothelial cell IGF-1 receptor showed a 34% reduction in oxygen-induced retinal neovascularization.68 Clearly, IGF-1 signaling plays a mediating role in the pathogenesis of ROP.

3. Rodent Models of Oxygen-Induced Retinopathy

3.2

69

ECM breakdown

Binding of growth factor to its receptor leads to the degradation of the vascular basement membrane, permitting the extravasation and subsequent proliferation of endothelial cells. Matrix metalloproteinases (MMPs) are responsible for the proteolytic degradation of the vascular basement membrane, making proliferative neovascularization possible. Das and colleagues showed increased MMP-2 and MMP-9 mRNA expression in mice with retinal neovascularization. The increased mRNA expression correlated with protein level and proteolytic activity, as measured by zymographic analysis.69 Administering a nonselective MMP inhibitor to OIR-exposed mice leads to a reduction in MMP2 and MMP9 activity and a 72% reduction in retinal neovascularization.69

3.3

Cellular adhesion

Much of what is known about endothelial cell attachment and migration in pathological retinal neovascularization comes from studies investigating the mechanisms of normal retinal development in the rodent. Several classes of endothelial cell adhesion molecules bind to the ECM and initiate a number of endothelial cell-specific responses. The integrins are an example of such adhesion molecules, and studies have shown that integrins αvβ3 and αvβ5 are specifically involved in retinal neovascularization. A murine model of ischemia-induced retinopathy demonstrated an upregulation of integrin αvβ3 in endothelial cells undergoing neovascularization.70 Administering a cyclic RGD peptide that inhibits integrin αv activity effectively reduces hypoxiainduced retinal neovascularization by more than 70% in the mouse.71 Peptide antagonists of integrins αvβ3 and αvβ5 also reduce retinal neovascularization in the mouse model.72

3.4

Blood vessel remodeling

Following migration, adhesion, and proliferation, endothelial cells undergo several maturation processes that serve to stabilize the newly formed blood vessels. The endothelial cell surface receptor Tie2 and its ligands, the angiopoietins (Ang1 and Ang2), are involved in blood vessel maturation. Ang1 binds with high affinity to the Tie2 receptor, stimulating receptor phosphorylation that leads to downstream signaling events involved in vascular development.73 Ang1 stabilizes newly formed blood vessels by promoting an interaction between the endothelial cells and a network of support cells. Ang2 also binds to Tie2 with high affinity. However, Ang2 does not stimulate receptor phosphorylation.74 For this reason, Ang2 is

70

S. E. Yanni et al.

considered to be a competitive inhibitor of Ang1. As such, Ang2 has been shown to cooperate with VEGF to prevent stasis of the newly formed vasculature. Unstable vessels are more likely to respond to a VEGF signal with sprouting.75-78 Consistent with its role in vessel stabilization, over-expression of Ang1 in the mouse retina has been shown to decrease VEGF-induced leakiness of the vasculature. In addition, over-expressing Ang1 inhibits the initiation and progression of retinal neovascularization in the mouse OIR model.79,80 On the other hand, Ang2 mRNA expression is increased in rat pups exposed to a model of retinopathy in which animals are raised in 80% oxygen for the first 11 days of life and then removed to room air for 7 days.81 Mice lacking Ang2 exhibit abnormal vascular development, and Ang2-null mice are protected from retinal neovascularization upon oxygen exposure. These observations indicate a critical role for Ang2 in the development of the retinal vasculature and in neovascularization.82, 83

3.5

Additional proteins involved in blood vessel growth

3.5.1

Ephrins and EphRs

Recently, much attention has focused on the A and B classes of the Ephrin receptor tyrosine kinases (Eph RTKs) present on the surface of endothelial cells and their ligands, the ephrins. Ephrin-B2 is expressed in endothelial cells and has been detected in retinal endothelial cells isolated from patients suffering from either ROP or proliferative diabetic retinopathy.84 Reverse signaling through ephrin-B2 stimulates retinal endothelial cell proliferation and migration.85 Intravitreal injections of soluble ephrin-B2 or Eph-B4 are able to reduce the severity of retinal neovascularization in the mouse OIR model.86 Together, these data suggest that class B Eph RTKs and ephrin B ligands play a role in retinal neovascularization. Class A Eph RTKs and their ephrin A ligands also have a role in ocular neovascularization. Experiments performed with a soluble Eph-A2 chimeric protein suggest that the interaction between Eph-A2 and ephrin-A1 is required for maximal VEGF-induced neovascularization in a mouse corneal angiogenesis assay. The chimeric protein inhibited VEGF-induced endothelial cell survival, migration, sprouting, and corneal angiogenesis,87 suggesting a functional link between VEGF and the EphA RTKs. Cheng et al.88 sought to determine the effect of soluble Eph-A2 on retinal neovascularization using the rat model. Soluble Eph-A2 significantly lowered the severity of retinal neovascularization by 50%, hypothetically through competitive binding to available ephrin ligands. In addition, the

3. Rodent Models of Oxygen-Induced Retinopathy

71

soluble Eph-A2 receptor inhibited the migration and tube formation of retinal endothelial cells stimulated with either VEGF or ephrin-A1 ligand. 3.5.2

Cyclooxygenase and the prostaglandins

The cyclooxygenase enzymes (COX1 and COX2) and their products, the prostaglandins, have emerged as potential mediators of pathological angiogenesis. Patients who regularly take nonsteroidal anti-inflammatory drugs (NSAIDs) have a reduced incidence of and mortality from colorectal cancer.89 NSAIDs are compounds that inhibit the activity of the COX enzymes. This provides a functional link between COX and angiogenic diseases, such as cancers. The therapeutic potential of COX inhibition has been tested for efficacy at inhibiting ocular angiogenesis. The non-selective COX inhibitors indomethacin, ibuprofen, and nepafenac, and the COX2selective inhibitor, rofecoxib, reduce retinal neovascularization in the mouse OIR model. 90-93

3.6

Endogenous inhibitors of angiogenesis

Under physiological conditions, pro-angiogenic factors are counterbalanced by one or more endogenous anti-angiogenic factors. An increase in antiangiogenic factors may tip the scale in favor of vascular quiescence. Some of these factors are found within the eye, and targeting them for therapeutic application may lead to fewer side effects during treatment of retinal neovascularization. 3.6.1

Pigment epithelium-derived factor

Pigment epithelium-derived factor, PEDF, is a member of the serpin (serine protease inhibitor) family of proteins.94 Although PEDF lacks serine protease inhibition activity, it is one of the most potent endogenous angiostatic factors. PEDF is more potent than angiostatin, endostatin, or thrombospondin-1, and inhibits the VEGF-induced migration of endothelial cells.95 Animal models demonstrate that PEDF immunoreactivity is higher in control animals than it is in animals experiencing ocular angiogenesis, and that the addition of exogenous PEDF or viral delivery of the PEDF gene can inhibit retinal angiogenesis and induce microvascular endothelial cell apoptosis.33,96-100 Furthermore, research has shown that rodents receiving a penetrating ocular injury after oxygen exposure exhibit less retinal neovascularization. The angiostatic effect of the penetrating ocular injury is consistent with an increase in PEDF mRNA and protein expression.101,102 Administering recombinant PEDF leads to a significant inhibition of retinal

72

S. E. Yanni et al.

neovascularization in the mouse OIR model.97 In a recent phase I clinical trial, adenoviral-delivery of PEDF led to an inhibition of neovascular agerelated macular degeneration. The results of this phase I clinical trial suggest that ocular gene transfer is a rational approach for the treatment of ocular proliferative disorders.103 3.6.2

ECM-related inhibitors of angiogenesis

Endostatin is an endogenous fragment of collagen type XV and collagen type XVIII that has anti-angiogenic activity. Viral-mediated delivery of endostatin prevents retinal neovascularization in the mouse OIR model.98 May and colleagues have proposed that endostatin-like proteins (ELPs) may play a self-limiting role in ROP-associated neovascularization. ELPs are absent at the beginning of the neovascular response, but increase over time, persisting as the vessels regress.104 Angiostatin is an endogenous fragment of plasminogen that possesses anti-angiogenic activity. Treating mice with angiostatin leads to a significant reduction in the development of OIR.105,106 Thrombospondin-1 (TSP-1) is an anti-angiogenic ECM glycoprotein. If TSP-1 is injected immediately after oxygen treatment, rat retinal neovascularization is reduced by 48%.107 Similarly, synthetic peptides derived from TSP-1 that contain either the TGFβ activating domain or the heparin binding domain were shown to have an inhibitory effect on retinal neovascularization.107 In vivo, the TSP-1 peptide containing the heparin binding domain was shown to be the most potent inhibitor of neovascularization. VEGF presentation to the VEGFR is mediated by the binding of VEGF to heparin on the surface of endothelial cells.108 It is hypothesized that TSP-1, by binding heparin, effectively prevents the interaction between VEGF and heparin, serving to reduce VEGF’s activation of its receptor. 3.6.3

Proteolytic inhibitors of angiogenesis

Urokinase plasminogen activator (uPA) cleaves plasminogen, an inactive serine protease precursor, to yield the active protease plasmin. Plasmin has broad specificity and cleaves a variety of proteins, including several important ECM components.109 Plasmin is also able to activate several matrix metalloproteinases (MMPs),110 proteolytic enzymes responsible for the degradation of the vascular basement membrane. Therefore, uPA initiates a cascade culminating in the degradation of the basement membrane. This permits the extravasation, migration, proliferation, and tube formation of endothelial cells undergoing angiogenesis. Endogenous plasminogen activator inhibitor (PAI-1) suppresses uPA activity. In the rat

3. Rodent Models of Oxygen-Induced Retinopathy

73

OIR model, intravitreal injection of recombinant PAI-1 reduced retinal neovascularization by 52% at the highest dose tested.111 Endogenous tissue inhibitors of metalloproteinases (TIMPs) inhibit MMP activity. Targeting PAI-1 or the TIMPs, alone or in combination, offers an attractive antineovascular strategy, since these molecules are endogenous to the retina.

4.

CONCLUDING REMARKS

Retinopathy of prematurity is a disease characterized by abnormal retinal angiogenesis. The development and utility of the rodent OIR models have contributed much of the information currently known about physiological and pathological retinal capillary growth. Because of the ease of their manipulation, rodent OIR models have provided a commonly used means to study the angiogenic process. In addition, the rodent eye is readily accessible, and its vasculature is easy to visualize and assess. Rodent OIR models have shed light on the cellular and molecular pathogenesis and pharmacological treatment of ocular, as well as non-ocular, vasoproliferative disorders. The continued refinement of the models and the knowledge gained through their use will aid the development of therapies to alleviate neovascular diseases of the human eye.

REFERENCES 1. T. L. Terry, Extreme prematurity and fibroblastic overgrowth of persistent vascular sheath behind each crystalline lens: I, preliminary report, Am. J. Ophthalmol. 25, 203-204 (1942). 2. K. Campbell, Intensive oxygen therapy as a possible cause of retrolental fibroplasia; a clinical approach, Med. J. Austral. 2 (2), 48-50 (1951). 3. A. Patz, L. E. Hoeck, and E. De La Cruz, Studies on the effect of high oxygen administration in retrolental fibroplasia: I, nursery observations, Am. J. Ophthalmol. 35 (9), 1248-1253 (1952). 4. L. J. Gyllensten and B. E. Hellstrom, Retrolental fibroplasias: animal experiments, Acta. Paediatr. 41 (6), 577-582 (1952). 5. A. Patz, A. Eastham, D. Higginbotham, and T. Kleh, Oxygen studies in retrolental fibroplasia: II, the production of the microscopic changes of retrolental fibroplasia in experimental animals, Am. J. Ophthalmol. 36 (11), 1511-1522 (1953). 6. N. Ashton, B. Ward, and G. Serpell, Role of oxygen in the genesis of retrolental fibroplasia: a preliminary report, Br. J. Ophthalmol. 37 (9), 513-520 (1953). 7. N. Ashton, B. Ward, and G. Serpell, Effect of oxygen on developing retinal vessels with particular reference to the problem of retrolental fibroplasias, Br. J. Ophthalmol. 38 (7), 397-432 (1954). 8. E. M. Hatfield, Blindness in infants and young children, Sight Sav. Rev. 42 (2), 69-89 (1972).

74

S. E. Yanni et al.

9. D. L. Gibson, S. B. Sheps, M. T. Schechter, S. Wiggins, and A. Q. McCormick, Retinopathy of prematurity: a new epidemic, Pediatrics 83 (4), 486-492 (1989). 10. W. A. Silverman, Retrolental fibroplasia: a modern parable. New York: Grune and Stratton, 1980. 11. A. Madan, Angiogenesis and antiangiogenesis in the neonate: relevance to retinopathy, Neo. Reviews 4, 356-363 (2003). 12. A. Madan and J. S. Penn, Animal models of oxygen-induced retinopathy, Frontiers in Biosciences 8, 1030-1043 (2003). 13. A. M. Roth, Retinal vascular development in premature infants, Am. J. Ophthalmol. 84 (5), 636-640 (1977). 14. R. Foos and S. Kopelow, Development of retinal vasculature in paranatal infants, Surv. Ophthalmol. 18, 117-127 (1973). 15. I. Michaelson, The mode of development of the vascular system of the retina, with some observations in its significance for certain retinal diseases, Trans. Ophthalmol. Soc. U.K. 68, 137-180 (1948). 16. D. K. Coats, E. A. Paysse, and P. G. Steinkuller, Threshold retinopathy of prematurity in neonates less than 25 weeks’ estimated gestational age, J. AAPOS. 4 (3), 183-185 (2000). 17. M. Fruttiger, Development of the mouse retinal vasculature: angiogenesis versus vasculogenesis, Invest. Ophthalmol. Vis. Sci. 43 (2), 522-527 (2002). 18. T. Chan-Ling, D. S. McLeod, S. Hughes, L. Baxter, Y. Chu, T Hasegawa, and G. A. Lutty, Astrocyte-endothelial cell relationships during human retinal vascular development, Invest. Ophthalmol. Vis. Sci. 45 (6), 2020-2032 (2004). 19. L. Gyllensten and B. Hellstrom, Experimental approach to the pathogenesis of retrolental fibroplasias: I. Changes of the eyes induced by exposure of newborn mice to concentrated oxygen, Acta. Pediat. 43 (100), 131-148 (1954). 20. L. Gyllensten and B. Hellstrom, Experimental approach to the pathogenesis of retrolental fibroplasia II. The influence of the developmental maturity in oxygen-induced changes in the mouse eye, Am. J. Ophthalmol. 39 (4, Part 1), 475-488 (1955). 21. A. Patz, A. Eastham, D. H. Higgenbotham, and T. Kleh, Oxygen studies in retrolental fibroplasia II. The production of the microscopic changes of retrolental fibroplasia in experimental animals, Am. J. Ophthalmol. 36 (11), 1511-1522 (1953). 22. P. M. Bischoff, S. D. Wajer, and R. W. Flower, Scanning electron microscopic studies of the hyaloid vascular system in newborn mice exposed to O2 and CO2, Graefes Arch. Clin. Exp. Ophthalmol. 220 (6), 257-263 (1983). 23. L. E. Smith, E. Wesolowski, A. McLellan, S. K. Kostyk, R. D’Amato, R. Sullivan, and P. A. D’Amore, Oxygen-induced retinopathy in the mouse, Invest. Ophthalmol. Vis. Sci. 35 (1), 101-111 (1994). 24. R. M. Rohan, A. Fernandez, T. Udagawa, J. Yuan, and R. J. D’Amato, Genetic heterogeneity of angiogenesis in mice, FASEB J. 14 (7), 871-876 (2000). 25. M. S. Rogers, R. M. Rohan, A. E. Birsner, and R. J. D’Amato, Genetic loci that control vascular endothelial growth factor-induced angiogenesis, FASEB J. 17 (4), 2112-2114 (2003). 26. C. K. Chan, L. N. Pham, J. Zhou, C. Spee, S. J. Ryan, and D. R. Hinton, Differential expression of pro- and antiangiogenic factors in mouse strain-dependent hypoxia-induced retinal neovascularization, Lab. Invest. 85 (6), 721-733 (2005). 27. S. Claxton and M. Fruttiger, Role of arteries in oxygen induced vaso-obliteration, Exp. Eye Res 77 (3), 305-311 (2003). 28. G. A. Gole, Animal models of retinopathy of prematurity. In: Silverman WA, Flynn JT, eds. Contemporary Issues in Fetal and Neonatal Medicine 2: Retinopathy of Prematurity. Boston: Blackwell Scientific Publishers; 1985: 53-95.

3. Rodent Models of Oxygen-Induced Retinopathy

75

29. A. Patz, Oxygen studies in retrolental fibroplasia IV: clinical and experimental observations, Am. J. Ophthalmol. 38 (3), 291-308 (1954). 30. N. Ashton and R. Blach, Studies in developing retinal vessels VIII: effect of oxygen on the vessels of the ratling, Br. J. Ophthalmol. 45, 321-340 (1961). 31. J. S. Penn, B. L. Tolman, and L. A. Lowery, Variable oxygen exposure causes preretinal neovascularization in the newborn rat, Invest. Ophthalmol. Vis. Sci. 34 (3), 576-585 (1993). 32. J. S. Penn, M. M. Henry, and B. L. Tolman, Exposure to alternating hypoxia and hyperoxia causes severe proliferative retinopathy in the newborn rat, Pediatr. Res. 36 (6), 724-731 (1994). 33. G. Gao, Y. Li, J. Fant, C. E. Crosson, S. P. Becerra, and J. X. Ma, Difference in ischemic regulation of vascular endothelial growth factor and pigment epithelium-derived factor in brown norway and sprague dawley rats contributing to different susceptibilities to retinal neovascularization, Diabetes 51 (4), 1218-1225 (2002). 34. S. X. Zhang, J. X. Ma, J. Sima, Y. Chen, M. S. Hu, A. Ottlecz, and G. N. Lambrou, Genetic difference in susceptibility to the blood-retina barrier breakdown in diabetes and oxygen-induced retinopathy, Am. J. Pathol. 166 (1), 313-321 (2005). 35. B. N. Floyd, D. A. Leske, S. M. Wren, M. Mookadam, M. P. Fautsch, and J. M. Holmes, Differences between rat strains in models of retinopathy of prematurity, Mol. Vis. 11, 524-530 (2005). 36. A. Kitzmann, D. A. Leske, Y. Chen, A. Kendall, W. Lanier, and J. Holmes, Incidence and severity of neovascularization in oxygen-and metabolic acidosis-induced retinopathy depend on rat source, Curr. Eye Res. 25, 215-220 (2002). 37. L. I. Larrazabal, and J. S. Penn, Study of ocular vasculature in the newborn rat by fluorescein angiography, J. Ophthal. Phot. 11, 49-52 (1989). 38. L. I. Larrazabal and J. S. Penn, Fluorescein angiography in the newborn rat: Implications in oxygen-induced retinopathy, Invest. Ophthalmol. Vis. Sci. 31 (5), 810-818 (1990). 39. J. S. Penn and B. D. Johnson, Fluorescein angiography as a means of assessing retinal vascular pathology in oxygen-exposed newborn rats, Curr. Eye. Res. 12 (6), 561-570 (1993). 40. J. S. Penn, B. L. Tolman, L. A. Lowery, and C. A. Koutz, Oxygen-induced retinopathy in the rat: hemorrhages and dysplasias may lead to retinal detachment, Curr. Eye Res. 11 (10), 939-953 (1992). 41. I. C. Michaelson, N. Herz, E. Lewkowitz, and D. Kertesz, Effect of increased oxygen on the development of the retinal vessels, Brit. J. Ophthal. 38 (10), 577-587 (1954). 42. D. R. Senger, S. J. Galli, A. M. Dvorak, C. A. Perruzzi, V. S. Harvey, and H. F. Dvorak, Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid, Science 219 (4587), 983-985 (1983). 43. P. J. Keck, S. D. Hauser, G. Krivi, K. Sanzo, T. Warren, J. Feder, and D. T. Connolly, Vascular permeability factor, and endothelial cell mitogen related to PDGF, Science 246 (4935), 1309-1312 (1989). 44. D. W. Leung, G. Cachianes, W. J. Kuang, D. V. Goeddel, and N. Ferrara, Vascular endothelial growth factor is a secreted angiogenic mitogen, Science 246 (4935), 1306-1309 (1989). 45. N. Ferrara, K. Houck, L. Jakeman, and D. W. Leung, Molecular and biological properties of the vascular endothelial growth factor family of proteins, Endocr. Rev. 13 (1), 18-32 (1992). 46. K. J. Kim, B. Li, J. Winer J, M. Armanini, N. Gillett, H. S. Phillips, and N. Ferrara, Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumour growth in vivo, Nature 362 (6423), 841-844 (1993).

76

S. E. Yanni et al.

47. N. Ferrara and T. Davis-Smyth, The biology of vascular endothelial growth factor, Endocr. Rev. 18 (1), 4-25 (1997). 48. A. P. Adamis, J. W. Miller, M. T. Bernal, D. J. D’Amico, J. Folkman, T. K. Yeo, and K. T. Yeo, Increased vascular endothelial growth factor levels in the vitreous of eyes with proliferative diabetic retinopathy, Am. J. Ophthalmol. 118 (4), 445-450 (1994). 49. L. P. Aiello, R. L. Avery, P. G. Arrigg, B. A. Keyt, H. D. Jampel, S. T. Shah, L. R. Pasquale, H. Thieme, M. A. Iwamoto, and J. E. Park, Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders, N. Engl. J. Med. 331 (22), 1480-1487 (1994). 50. G. Gao, Y. Li, D. Zhang, S. Gee, C. Crosson, and J. Ma, Unbalanced expression of VEGF and PEDF in ischemia-induced retinal neovascularization, FEBS Lett. 489 (2-3), 270-276 (2001). 51. E. A. Pierce, E. D. Foley, and L. E. Smith, Regulation of vascular endothelial cell growth factor by oxygen in a model of retinopathy of prematurity, Arch. Ophthalmol. 114 (10), 1219-1228 (1996). 52. D. Shweiki, A. Itin, D. Soffer, and E. Keshet, Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis, Nature 359 (6398), 843-845 (1992). 53. E. A. Pierce, R. L. Avery, E. D. Foley, L. P. Aiello, and L. E. Smith, Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neovascularization, Proc. Natl. Acad. Sci. USA 92 (3), 905-909 (1995). 54. D. A. Simpson, G. M. Murphy, T. Bhaduri , T. A. Gardiner, D. B. Archer, and A. W. Stitt, Expression of the VEGF gene family during retinal vaso-obliteration and hypoxia, Biochem. Biophys. Res. Commun. 262 (2), 333-340 (1999). 55. H. Ozaki, H. Hayashi, S. A. Vinores, Y. Moromizato, P. A. Campochiaro, and K. Oshima, Intravitreal sustained release of VEGF causes retinal neovascularization in rabbits and breakdown of the blood-retinal barrier in rabbits and primates, Exp. Eye Res. 64 (4), 505-517 (1997). 56. H. Ozaki, A. Y. Yu, N. Della, K. Ozaki, J. D. Luna, H. Yamada, S. F. Hackett, N. Okamoto, D. J. Zack, G. L. Semenza, and P. A. Campochiaro, Hypoxia inducible factor-1alpha is increased in ischemic retina: temporal and spatial correlation with VEGF expression, Invest. Ophthalmol. Vis. Sci. 40 (1), 82-189 (1999). 57. L. P. Aiello, E. A. Pierce, E. D. Foley, H. Takagi, H. Chen, L. Riddle, N. Ferrara, G. L. King, and L. E. Smith, Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins, Proc. Natl. Acad. Sci. USA 92 (23), 10457-10461 (1995). 58. T. Tobe, N. Okamoto, M. A. Vinores, N. L. Derevjanik, S. A. Vinores, D. J. Zack, and P. A. Campochiaro, Evolution of neovascularization in mice with overexpresion of vascular endothelial growth factor in photoreceptors, Invest. Ophthalmol. Vis. Sci. 39 (1), 180-188 (1998). 59. W. T. Deng, Z. Yan, A. Dinculescu, J. Pang, J. T. Teusner, N. G. Cortez, K. I. Berns, and W. W. Hauswirth, Adeno-associated virus mediated expression of vascular endothelial growth factor peptides inhibits retinal neovascularization in a mouse model of oxygeninduced retinopathy, Hum. Gene Ther. 16, 1247-1254 (2005). 60. A. P. Adamis, D. T. Shima, M. J. Tolentino, E. S. Gragoudas, N. Ferrara, J. Folkman, P. A. D’Amore, and J. W. Miller, Inhibition of vascular endothelial growth factor prevents retinal ischemia-associated iris neovasculariztion in a nonhuman primate, Arch. Ophthalmol. 114 (1), 66-71 (1996).

3. Rodent Models of Oxygen-Induced Retinopathy

77

61. F. Kinose, G. Roscilli, S. Lamartina, K. D. Anderson, F. Bonelli, S. G. Spence, G. Ciliberto, T. F. Vogt, D. J. Holder, C. Toniatti, and C. J. Thut, Inhibition of retinal and choroidal neovascularization by a novel KDR kinase inhibitor, Mol. Vis. 11, 366-373 (2005). 62. R. Rota, T. Riccioni, M. Zaccarini, S. Lamartina, A. D. Gallo, A. Fusco, I. Kovesdi, E. Balestrazzi, D. C. Abeni, R. R. Ali, and M. C. Capogrossi, Marked inhibition of retinal neovascularization in rats following soluble-flt-1 gene transfer, J. Gene Med. 6 (9), 992-1002 (2004). 63. E. W. Ng, D. T. Shima, P. Calias, E. T. Cunningham, D. R. Guyer, and A. P. Adamis, Pegaptanib, a targeted anti-VEGF aptamer for ocular vascular disease, Nat. Rev. Drug Discov. 5 (2), 123-132 (2006). 64. R. M. Rich, P. J. Rosenfeld, C. A. Puliafito, S. R. Dubovy, J. L. Davis, H. W. Flynn, S. Gonzalez, W. J. Feuer, R. C. Lin, G. A. Lalwani, J. K. Nguyen, and G. Kumar, Shortterm safety and efficacy of intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration, Retina 26 (5), 495-511 (2006). 65. A. D. Wright, E. M. Kohner, N. W. Oakley, M. Hartog, G. F. Joplin, and T. R. Fraser, Serum growth hormone levels and the response of diabetic retinopathy to pituitary ablation, Br. Med. J. 2 (653), 346-348 (1969). 66. L. E. Smith, J. J. Kopchick, W. Chen, J. Knapp, F. Kinose, D. Daley, E. Foley, R. G. Smith, and J. M. Schaeffer, Essential role of growth hormone in ischemia-induced retinal neovascularization, Science 276 (5319), 1706-1709 (1997). 67. L. E. Smith, W. Shen, C. Perruzzi, S. Soker, F. Kinose, X. Xu, G. Rovinson, S. Driver, J. Bischoff, B. Zhang, J. M. Schaeffer, and D. R. Senger, Regulation of vascular endothelial growth factor-dependent retinal neovascularization by insulin-like growth factor-1 receptor, Nat. Med. 5 (12), 1390-1395 (1999). 68. T. Kondo, D. Vicent, K. Suzuma, M. Yanagisawa, G. L. King, M. Holzenberger, and C. R. Kahn, Knockout of insulin and IGF-1 receptors on vascular endothelial cells protects against retinal neovascularization, J. Clin. Invest. 111 (12), 1835-1842 (2003). 69. A. Das, A. McLamore, W. Song, and P. G. McGuire, Retinal neovascularization is suppressed with a matrix metalloproteinase inhibitor, Arch. Ophthalmol. 117 (4), 498-503 (1999). 70. J. Luna, T. Tobe, S. A. Mousa, T. M. Reilly, and P. A. Campochiaro, Antagonists of integrin alpha v beta 3 inhibit retinal neovascularization in a murine model, Lab. Invest. 75 (4), 563-573 (1996). 71. M. Friedlander, C. L. Theesfeld, M. Sugita, M. Fruttiger, M. A. Thomas, S. Chang, and D. A. Cheresh, Involvement of integrins alpha v beta 3 and alpha v beta 5 in ocular neovascular diseases, Proc. Natl. Acad. Sci. USA 93 (18), 9764-9769 (1996). 72. H. Hammes, M. Brownlee, A. Jonczyk, A. Sutter, and K. T. Preissner, Subcutaneous injection of a cyclic peptide antagonist of vitronectin receptor-type integrins inhibits retinal neovascularization, Nat. Med. 2 (5), 529-533 (1996). 73. S. Davis, T. H. Aldrich, and P. F. Jones, Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning, Cell 87 (7), 1161-1169 (1996). 74. P. C. Maisonpierre, C. Suri, P. F. Jones, S. Bartunkova, S. J. Wiegand, C. Radziejewski, D. Compton, J. McClain, T. H. Aldrich, N. Papadopoulos, T. J. Daly, S. Davis, T. N. Sato, G. D. Yancopoulos, Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis, Science 277, 55-60 (1997). 75. D. Hanahan, Signaling vascular morphogenesis and maintenance, Science 277 (5322), 48-50 (1997). 76. J. Folkman and P. A. D’Amore, Blood vessel formation: What is its molecular basis, Cell 87 (7), 1153-1155 (1996).

78

S. E. Yanni et al.

77. N. W. Gale and G. D. Yancopoulos, Growth factors acting via endothelial cell-specific receptor tyrosine kinases: VEGFs, angiopoietins, and ephrins in vascular development, Genes Dev. 13 (9), 1055-1066 (1999). 78. G. D. Yancoupolos, S. Davis, N. W. Gale, J. S. Rudge, S. J. Wiegand, and J. Holash, Vascular-specific growth factors and blood vessel formation, Nature 407 (6801), 242-248 (2000). 79. H. Nambu, R. Nambu, Y. Oshima, S. F. Hackett, G. Okoye, S. Wiegand, G. Yancopoulos, D. J. Zack, and P. A. Campochiaro, Angiopoietin 1 inhibits ocular neovascularization and breakdown of the blood-retinal barrier, Gene Therapy 11 (10), 865-873 (2004). 80. H. Nambu, N. Umeda, S. Kachi, Y. Oshima, H. Akiyama, R. Nambu, and P. A. Campochiaro, Angiopoietin 1 prevents retinal detachment in an aggressive model of proliferative retinopathy, but has no effect on established neovascularization, J. Cell Phys. 204 (1), 227-235 (2005). 81. S. Sarlos, B. Rizkalla, C. J. Moravski, Z. Cao, M. E. Cooper, and J. L. Wilkinson-Berka, Retinal angiogenesis is mediated by an interaction between the angiotensin type 2 receptor, VEGF, and angiopoietin, Am. J. Pathol. 163 (3), 879-887 (2003). 82. S. F. Hackett, H. Ozaki, R. W. Strauss, K. Wahlin, C. Suri, P. Maisonpierre, G. Yancopoulos, and P. A. Campochiaro, Angiopoietin 2 expression in the retina: up-regulation during physiologic and pathologic neovascularization, J. Cell Physiol. 184 (3), 275-284 (2000). 83. S. F. Hackett, S. Wiegand, G. Yancopoulos, and P. A. Campochiaro, Angiopoietin-2 plays an important role in retinal angiogenesis, J. Cell Physiol. 192 (2), 182-187 (2002). 84. N. Umeda, H. Ozaki, H. Hayashi, and K. Oshima, Expression of ephrinB2 and its receptors on fibroproliferative membranes in ocular angiogenic diseases, Am. J. Ophthalmol. 138 (2), 270-279 (2004). 85. J. J. Steinle, C. J. Meininger, U. Chowdhury, G. Wu, and H. J. Granger, Role of ephrin B2 in human retinal endothelial cell proliferation and migration, Cell Signal 15 (11), 1011-1017 (2003). 86. D. O. Zamora, M. H. Davies, S. R. Planck, J. T. Rosenbaum, and M. R. Powers, Soluble forms of ephrinB2 and EphB4 reduce retinal neovascularization in a model of proliferative retinopathy, Invest. Ophthalmol. Vis. Sci. 46 (6), 2175-2182 (2005). 87. N. Cheng, D. M. Brantley, H. Liu, W. Fanslow, D. P. Cerretti, K. N. Bussell, A. Reith, D. Jackson, and J. Chen, Blockade of EphA receptor tyrosine kinase activation inhibits VEGF-induced angiogenesis, Mol. Cancer Res. 1 (1), 2-11 (2002). 88. J. Chen, D. Hicks, D. Brantley-Sieders, N. Cheng, G. W. McCollum, X. Qi-Werdich, and J. Penn, Inhibition of retinal neovascularization by soluble EphA2 receptor, Exp. Eye Res. 82 (4), 664-673 (2006). 89. W. Smalley and R. N. DuBois, Colorectal cancer and nonsteroidal anti-inflammatory drugs, Adv. Pharmacol. 39, 1-20 (1997). 90. N. B. Nandgaonkar, T. Rotschild, K. Yu, and R. D. Higgins, Indomethacin improves oxygen-induced retinopathy in the mouse, Pediatr. Res. 46 (2), 184-188 (1999). 91. J. Sharma, S. M. Barr, Y. Geng, Y. Yun, and R. D. Higgins, Ibuprofen improves oxygeninduced retinopathy in a mouse model, Curr. Eye Res. 27 (5), 309-314 (2003). 92. K. Takahashi, Y. Saishin, Y. Saishin, K. Mori, A. Ando, S. Yamamoto, Y. Oshima, H. Nambu, M. B. Melia, D. P. Bingaman, and P. A. Campochiaro, Topical nepafenac inhibits ocular neovascularization, Invest. Ophthalmol. Vis. Sci. 44 (1), 409-415 (2003). 93. J. L. Wilkinson-Berka, N. S. Alousis, D. J. Kelly, and R. E. Gilbert, COX-2 inhibition and retinal angiogenesis in a mouse model of retinopathy of prematurity, Invest. Ophthalmol. Vis. Sci. 44 (3), 974-979 (2003).

3. Rodent Models of Oxygen-Induced Retinopathy

79

94. C. J. Barnstable and J. Tombran-Tink, Neuroprotective and antiangiogenic actions of PEDF in the eye: molecular targets and therapeutic potential, Prog. Retin. Eye Res. 23 (5), 561-577 (2004). 95. D. W. Dawson, O. V. Volpert, P. Gillis, S. E. Crawford, H. Xu, W. Benedict, and N. P. Bouck, Pigment epithelium-derived factor: a potent inhibitor of angiogenesis, Science 285 (5425), 245-248 (1999). 96. R. Z. Renno, A. I. Youssri, N. Michaud, E. S. Gragoudas, and J. W. Miller, Expression of pigment epithelium-derived factor in experimental choroidal neovascularization, Invest. Ophthalmol. Vis. Sci. 43 (5), 1574-1580 (2002). 97. E. J. Duh, H. S. Yang, I. Suzuma I, M. Miyagi, E. Youngman, K. Mori, M. Katai, L. Yan, K. Suzuma, K. West, S. Davarya, P. Tong, P. Gehlbach, J. Pearlman, J. W. Crabb, L. P. Aiello, P. A. Campochiaro, and D. J. Zack, Pigment epithelium-derived factor suppresses ischemia-induced retinal neovasculariztion and VEGF-induced migration and growth, Invest. Ophthalmol. Vis. Sci. 43 (3), 821-829 (2002). 98. A. Auricchio, K. C. Behling, A. M. Maguire, E. M. O’Connor, J. Bennett, J. M. Wilson, and M. J. Tolentino, Inhibition of retinal neovascularization by intraocular viralmediated delivery of anti-angiogenic agents, Mol. Ther. 6 (4), 490-494 (2002). 99. V. Stellmach, S. E. Crawford, W. Zhou, and N. Bouck, Prevention of ischemia-induced retinopathy by the natural ocular antiangiogenic agent pigment epithelium-derived factor, Proc. Natl. Acad. Sci. USA 98 (5), 2593-2597 (2001). 100. K. Mori, P. Gehlbach, A. Ando, D. McVey, L. Wei, and P. A. Campochiaro, Regression of ocular neovasculariztion in response to increased expression of pigment epitheliumderived factor, Invest. Ophthalmol. Vis. Sci. 43 (7), 2428-2434, (2002). 101. A. W. Stitt, D. Graham, and T. A. Gardiner, Ocular wounding prevents pre-retinal neovascularization and upregulated PEDF expression in the inner retina, Mol. Vis. 10, 432-438 (2004). 102. J. S. Penn, G. W. McCollum, J. M. Barnett JM, X. Q. Werdich, K. A. Koepke, and V. S. Rajaratnam, Angiostatic effect of penetrating ocular injury: role of pigment epithelium-derived factor, Invest. Ophthalmol. Vis. Sci. 47 (1), 405-414 (2006). 103. P.A. Campochiaro, Q. D. Nguyen, S. M. Shah, M. L. Klein, E. Holz, R. N. Frank, D. A. Saperstein, A. Gupta, J. T. Stout, J. Macko, R. DiBartolomeo, and L. L. Wei, Adenoviral vector-delivered pigment epithelium-derived factor for neovascular agerelated macular degeneration: results of a phase I clinical trial, Hum. Gene Ther. 17 (2), 167-176 (2006). 104. C. A. May, A. V. Ohlmann, H. Hammes, and U. H. Spandau, Proteins with an endostatin-like domain in a mouse model of oxygen-induced retinopathy, Exp. Eye Res. 82 (2), 341-348 (2006). 105. T. A. Drixler, I. H. Borel Rinkes, E. D. Ritchie, F. W. Treffers, T. J. van Vroonhoven, M. F. Gebbink, and E. E. Voest, Angiostatin inhibits pathological but not physiological retinal angiogenesis, Invest. Ophthalmol. Vis. Sci. 42 (13), 3325-3330 (2001). 106. T. Igarashi, K. Miyake, K. Kato, A Watanabe, M. Ishizaki, K. Ohara, and T. Shimada, Lenitivirus-mediated expression of angiostatin efficiently inhibits neovascularization in a murine proliferative retinopathy model, Gene Ther. 10 (3), 219-226 (2003). 107. A. Shafiee, J. S. Penn, H. C. Krutzsch, J. K. Inman, D. D. Roberts, D. A. Blake, Inhibition of retinal angiogenesis by peptides derived from thrombospondin-1, Invest. Ophthalmol. Vis. Sci. 41 (8), 2378-2388 (2000). 108. J. Schlessinger, I. Lax, and M. Lemmon, Regulation of growth factor activation by proteoglycans: What is the role of the low affinity receptors? Cell 83 (3), 357–360 (1995).

80

S. E. Yanni et al.

109. F. Blasi, J. D. Vassalli, and K. Dano, Urokinase-type plasminogen activator: proenzyme, receptor, and inhibitors, J. Cell. Biol. 104 (4), 801-804 (1987). 110. O. Saksela, Plasminogen activation and regulation of pericellular proteolysis, Biochim. Biophys. Acta. 823 (1), 35-65 (1985). 111. J. S. Penn and V. S. Rajaratnam, Inhibition of retinal neovascularization by intravitreal injection of human rPAI-1 in a rat model of retinopathy of prematurity, Invest. Ophthalmol. Vis. Sci. 44 (12), 5423-5429 (2003). 112. L. E. Bullard, X. Qi, and J.S. Penn, Role for extracellular signal-responsive kinase-1 and -2 in retinal angiogenesis, Invest. Ophthalmol. Vis. Sci. 44 (4), 1722-1731 (2003).

Chapter 4 ANIMAL MODELS OF DIABETIC RETINOPATHY

Timothy S. Kern Case Western Reserve University, Cleveland, Ohio

Abstract:

1.

If they are diabetic long enough, most or all species available for laboratory research will develop lesions characteristic of the early stages of diabetic retinopathy, including nonperfused (and acellular) capillaries and apoptotic loss of capillary cells. Although none of these animal models reliably proceed to preretinal neovascularization, they nevertheless provide valuable insight into the role of specific biochemical pathways and cell types in the early stages of retinopathy. An increasing number of therapeutic approaches have been identified that significantly inhibit the development of capillary obliteration in the retina. The challenge now is to integrate the results of these studies to identify the sequence of events that ultimately results in the characteristic histopathology in diabetes. Why diabetic animal models have not been found to develop the neovascular stages of diabetic retinopathy remains an important question, and one likely reason for this “failure” is that much less vasoobliteration develops in the retina of the diabetic animals during the short duration of their diabetes as compared to that of some diabetic patients who, over many years, develop extensive vaso-obliteration. Nevertheless, the models are still useful, because preventing progressive capillary obliteration from occurring in the retina is likely to be a more beneficial therapeutic goal than merely inhibiting neovascularization in an already damaged and ischemic retina.

INTRODUCTION

Diabetic retinopathy is a major complication of Type 1 and Type 2 diabetes mellitus, being observed in most patients after 15 years of diabetes, and increasing the risk of blindness 25-fold above normal.1,2 The natural history of clinically demonstrable retinopathy has been carefully documented, and 81 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 81–102. © Springer Science+Business Media B.V. 2008

82

T. S. Kern

important stages have been identified: vascular occlusion, formation of capillary microaneurysms, excessive vascular permeability, proliferation of new vessels and fibrous tissue, and contraction of the fibrovascular proliferations.3 This chapter will focus on the histological lesions that develop in animal models, and their relation to the lesions that develop in diabetic patients. Physiological abnormalities such as retinal blood flow and permeability have been reviewed elsewhere.4,5 The general picture that has emerged of the pathogenesis of vision loss in diabetic retinopathy focuses primarily on increased capillary permeability, which leads to retinal edema, and neovascularization. Retinal edema can result in appreciable visual impairment, presumably due to physical distortion of the retina. Neovascularization can prevent light from reaching the photoreceptors secondary to development of a fibrovascular membrane in front of the retina.

Figure 4-1. Simplified scheme postulated for the pathogenesis of diabetic retinopathy.

The nonproliferative stage of the retinopathy includes capillary cell death and capillary obliteration, microaneurysms, pericyte loss, and increased permeability. Pericyte loss was once believed to be the initial and most important lesion of the retinopathy, but it has since been demonstrated that both retinal endothelial cells and pericytes die at approximately the same rate

4. Animal Models of Diabetic Retinopathy

83

in diabetes. These “background” changes precede and are believed to be necessary for progression to the later neovascular changes (Figure 1). The early stages of the retinopathy (before microaneurysms are present) generally are not apparent clinically, even using sensitive techniques such as fluorescein angiography. At even earlier stages, however, these lesions are beginning to appear, and they can be studied histologically using eyes collected at autopsy or at surgery (Figure 2).

Figure 4-2. Microaneurysm (MA), acellular capillaries (long arrows), and pericyte ghosts (thick arrows) in dog diabetic for 5 years.

Nonperfused capillaries in diabetic retinopathy commonly lack endothelial cells or pericytes, and thus appear to be acellular. These “acellular capillaries” are the remnant basement membrane skeleton of a degenerate capillary from which all capillary cells have disappeared. Importantly, they are a histological marker of capillary nonperfusion, since acellular capillaries are not perfused.6 Thus, acellular capillaries are morphological lesions that have physiological significance and can be quantitated by light microscopy in animal studies of the retinopathy. Vaso-obliteration of retinal capillaries in diabetes begins in capillaries and can progress “upstream” to the arterioles and their side-branches. Potential causes of capillary occlusion include leukostasis, excessive platelet aggregation, endothelial swelling, endothelial death, and glial invasion of the capillary lumen.

84

2.

T. S. Kern

ANIMAL MODELS OF DIABETIC RETINOPATHY

Animal models of diabetic retinopathy have proven valuable in efforts to unravel the pathogenesis of retinopathy, and to identify therapies to inhibit it. Species used in studies of the effect of diabetes on the retina include spontaneously diabetic animals, including fish,7 mice,8,9 rats,10,11 cats,12 dogs,13-15 and apes,16 but these reports generally have been only descriptive in nature. Many investigations also have relied on experimental induction of diabetes with alloxan, streptozotocin, growth hormone, or by pancreatectomy. Early studies of animal models have been reviewed elsewhere,12,17-19 and the present review will focus on studies reported during the past decade.

2.1

Dogs and cats

The anatomical features of retinopathy in diabetic dogs have been shown repeatedly to be morphologically indistinguishable from those of background retinopathy seen in diabetic patients. They include capillary microaneurysms, acellular (and nonperfused) capillaries, pericyte ghosts, varicose and dilated capillaries (also called intraretinal microvascular abnormalities or IRMAs), and dot and blot hemorrhages.18,20-22 Arteriolar smooth muscle cell loss also has been observed in humans and dogs.23,24 The lesions in diabetic dogs are secondary to insulin deficiency, since they develop irrespective of how diabetes was induced (alloxan, growth hormone, pancreatectomy), and can be inhibited by strict regulation of glycemia with exogenous insulin.25 Microaneurysms, leukocyte and platelet plugging of aneurysms and venules, and degenerating endothelial cells likewise were observed in cats after several years of diabetes.26,27 These histological abnormalities were confined to small regions, and these animals developed hypoxia in at least some areas of retina early in the development of diabetic retinopathy, before capillary dropout was evident clinically. Hypoxia was correlated with endothelial cell death, leukocyte plugging of vessels, and microaneurysms. As is true in diabetic humans, there is a long interval before retinopathy becomes manifest in diabetic dogs or cats; capillary aneurysms usually begin to appear in these animals about 2 to 3 years after induction of elevated hexose levels. Likewise, after about 2 years of hyperglycemia in diabetic dogs, increasing numbers of retinal capillaries possess endothelial cells but few or no pericytes. Gradual obliteration of retinal vessels is apparent histologically from the increasing numbers of acellular capillaries that are scattered singly and in small groups about the retinal vasculature, especially

4. Animal Models of Diabetic Retinopathy

85

in temporal retina. Within 5 years of insulin-deficient diabetes, all dogs have a marked retinopathy. The reason for the prolonged interval before retinopathy develops is unknown, but any explanation of this latent period might offer valuable insight into the pathogenesis of the retinopathy. Improved glycemic control significantly inhibits the development and progression of retinopathy in diabetic dogs25,28 and in patients.29,30 The retinopathy that develops in cats and dogs more closely resembles lesions in human retinopathy than that of other species studied to date. Neovascularization has been observed to develop in diabetic dogs, albeit only within the retina (see Section 5). However, the cost, slow development of lesions, and lack of availability of antibodies or molecular biology techniques have made dog and cat models less used for the study of retinopathy in recent years.

2.2

Rats

During the past decade, streptozotocin-diabetic or alloxan-diabetic rats have been the primary model for research into the pathogenesis of the vascular lesions of diabetic retinopathy.31-50 Spontaneously diabetic BB rats and rats made diabetic with alloxan or streptozotocin exhibit similar retinal lesions: pericyte loss, basement membrane thickening, and an absence of microaneurysms after about 14 months of hyperglycemia.51 However, later stages of retinal microvascular disease do not develop reproducibly (microaneurysms) or at all (IRMA, hemorrhages, and neovascularization). As a model of diabetic retinopathy, the rat offers practical advantages over the dog and other large animals in terms of costs, housing requirements, and available reagents. Moreover, the early stages of retinopathy develop relatively quickly in the rat; pericyte loss and acellular capillaries are apparent after as little as 6 months of diabetes. A potential concern about this model is that the lens of rats has unusually high levels of aldose reductase compared to other species; 52 whether or not this is true in other tissues has not tissues has not been reported.

2.3

Mice

In the 1970’s and 80’s, there were a number of attempts to determine whether or not diabetic mice developed diabetic retinopathy, but the results were controversial.8,9,53-55 Since then, it is surprising that mice have been little studied with respect to diabetic retinopathy until recently, especially considering the widespread generation and use of genetically modified mice. Recent studies have begun to characterize the development of retinopathy in the streptozotocin-diabetic C57BL/6J mouse. This model develops the early

86

T. S. Kern

vascular pathology characteristic of diabetic retinopathy (acellular capillaries, pericyte loss, and capillary cell apoptosis) beginning at about 6 months of diabetes, and the acellular capillaries and pericyte ghosts become more numerous with increasing duration of diabetes (through 18 months of diabetes).56,57 These vascular abnormalities characteristic of retinopathy occurred despite the apparent lack of neuronal loss and Müller glial cell activation,57 although others have reported loss of cells in the ganglion cell layer in mice diabetic only 14 weeks58 (see section 2). Diabetes-induced retinal neovascularization has not been detected in any mouse model to date. Genetically modified mice are beginning to be used to explore the role of adhesion molecules and leukostasis in the pathogenesis of diabetes-induced retinal vascular disease.59 Mice deficient in the genes encoding adhesion molecules CD18 and ICAM-1 were made diabetic or experimentally galactosemic, and studied after durations of up to 11 months (diabetic) or 22 months (galactosemic). Wild-type diabetic or galactosemic animals developed acellular capillaries and pericyte loss, as well as associated abnormalities including leukostasis, increased capillary permeability, and capillary basement membrane thickening. In contrast, CD18-/- and ICAM-1-/mice developed significantly fewer of each of these abnormalities, thus providing novel insight that adhesion molecules play an important role in the pathogenesis of the retinopathy. There are both advantages and disadvantages to the use of mice as models of diabetic retinopathy. The principal advantages are cost, availability of reagents, and ability to generate (or availability of) genetically modified animals for study. The principal disadvantages are the small size of the retina (and consequently the quantitatively small number of lesions that can be detected per retina) and the extreme difficulty in unambiguously identifying pericyte ghosts for quantitation.

2.4

Lesser used models including primates

The early stages of diabetic retinopathy develop in all species that have diabetes for long durations. Thus, it seems unlikely that the ability to develop microvascular lesions of diabetic retinopathy is in any way unique. However, some laboratory species, such as guinea pigs and rabbits, are inherently of limited usefulness for the study of diabetic retinopathy. The guinea pig retina is avascular, as also is much of the rabbit retina, and the retinal vessels in rabbits are tortuous and limited chiefly to the most superficial inner layers of nasal and temporal retina. Diabetic hamsters develop the usual spectrum of lesions, including acellular capillaries, pericyte loss, and endothelial proliferation, but lack microaneurysms and neovascularization.60

4. Animal Models of Diabetic Retinopathy

87

In spite of many similarities between subhuman primates and humans, diabetic retinopathy has been little studied in primates. In the past decade, these studies have been limited to identification of microaneurysms and other lesions in aged, spontaneously diabetic monkeys.16

3.

DIABETES-INDUCED ABNORMALITIES IN NONVASCULAR CELLS OF THE RETINA

Several decades ago, damage to nonvascular cells of the retina (including ganglion cells) in diabetic humans was detected both ultrastructurally61 and functionally,62,63 and the possible role of neural disease in the pathogenesis of diabetic retinopathy was postulated.64,65 Recently, there has been renewed appreciation of diabetes-induced damage to nonvascular cells of the retina also in animals. Diabetic rats lose ganglion cells,44,58,66-75 and this neurodegeneration has been detected at as early as one month of diabetes.69 This nonvascular abnormality precedes the development of the vascular cell changes,69 raising the possibility that neurodegeneration might contribute to the pathogenesis of the vascular disease. This has yet to be conclusively studied. Retinal glial cells also undergo changes in diabetes in some species. Müller glial cells in diabetic rats became apoptotic in one study.68 In other studies, these cells changed from a quiescent to an injury-associated phenotype with high levels of expressed glial fibrillary acidic protein (GFAP)—a hallmark of glial cell activation—after a few months of diabetes.44,68,71,76-81 Alterations in GFAP expression patterns in Müller glial cells have also been observed in the human retina during early diabetes.77 In diabetic C57BL/6J mice, transient damage was noted in retinal ganglion cells (TUNEL-positive with activation of caspase 3) at about 4 weeks of diabetes. These abnormalities quickly returned to normal, however, and ultimately, no detectable loss of retinal ganglion cells or activation of Müller glial cells was noted in the retinas, even after one year of diabetes.44,57 In contrast, others have reported that 14 weeks of diabetes was sufficient to cause a 20-25% reduction in the number of cells in the ganglion cell layer compared to age-matched nondiabetic mice,58 and pro-apoptotic caspases were found to increase with increasing duration of diabetes.56 Likewise, C57BL/6J diabetic mice do not show GFAP activation in diabetes,44 other than a transient increase soon after induction of diabetes.57 Müller glial cells from mice show nuclear translocation of GAPDH in diabetes, a change that has been strongly linked to apoptosis.75 Ins2Akita diabetic mice also have increased retinal vascular permeability, greater than

88

T. S. Kern

normal numbers of caspase-3 positive cells, and unchanged GFAP immunoreactivity.82 Horizontal cells,71,74 amacrine cells, and photoreceptors74 also have been reported to undergo degeneration in diabetic rats. These changes are not known to be characteristic of retinal changes seen in diabetic patients, however, and the significance of these changes in animals remains to be learned.

4.

NONDIABETIC MODELS THAT DEVELOP A DIABETIC-LIKE RETINOPATHY

4.1

Galactose feeding

The importance of hyperglycemia per se in the pathogenesis of diabetic retinopathy was demonstrated a number of years ago by study of normal, nondiabetic dogs fed a galactose-rich diet.83,84 During the 3 to 5 years of study, normal dogs fed a diet enriched with 30% galactose developed a retinopathy that was indistinguishable from that of diabetic dogs and patients, including microaneurysms, vaso-obliteration, pericyte ghosts, and hemorrhages.22,35,83-93 Likewise, experimental galactosemia has also been shown to cause diabetic-like retinal lesions in rats and mice. Rats fed a 50% or 30% galactose diet for more than 1.5 years develop a significantly greater than normal prevalence of acellular capillaries and pericyte ghosts, excessive thickening of capillary basement membrane and, eventually, IRMAs.35,94-100 Mice fed 30% galactose also develop diabeticlike retinopathy, including rare but unmistakable saccular microaneurysms, as well as acellular capillaries, pericyte ghosts, and capillary basement membrane thickening.59,101 The galactose-retinopathy model has been utilized extensively for studies of the role of aldose reductase in the pathogenesis of “diabetic-like” retinopathy,22,35,85-98 but more recently the model has been used in studies of the role of leukostasis in retinopathy,59 and the ability of aminoguanidine, antioxidants, and antisense mRNA against fibronectin to inhibit retinopathy.99,100,102 As a means for producing a model of diabetic retinopathy in animals, experimental galactosemia can be advantageous because it is easily established and requires less nursing care than experimental diabetes. Not to be overlooked, however, is the expense of the galactose diet, which can be costly if animals are large or numerous. Moreover, the galactose-induced retinopathy has at least two important differences from that in diabetes. First,

4. Animal Models of Diabetic Retinopathy

89

it develops despite the absence of many of the systemic metabolic abnormalities that are characteristic of diabetes (such as those involving concentrations of glucose, insulin, fatty acids, etc.).84 This is valuable, in that it demonstrates that excessive blood hexose (either glucose or galactose) is important in the initiation of retinopathy. The second difference between the retinopathies induced by diabetes and galactose feeding is a different response to at least one therapy. Aminoguanidine has been shown several times to inhibit retinal microvascular disease in diabetic dogs and rats,99,103-105 but has not been found to do so in galactose-fed rats.99,106 Moreover, caspases activated in diabetic mice differ from those induced in galactose-fed mice.56 Thus, although the final histopathology induced by galactosemia seems morphologically identical to that in diabetes, the biochemical steps leading to that pathology apparently differ between the two models. The galactose model of retinopathy is a valuable source of comparison to diabetes, but it should not be assumed to respond to therapy like diabetic animals or patients would without comparing the two models first. Neurodegeneration has not yet been assessed in galactosemic models.

4.2

Sucrose or fructose feeding

Nondiabetic rats fed very high concentrations of sucrose or fructose (approximately 70% in the diet) also have been reported to develop retinal lesions, including loss of pericytes and endothelial cells, and formation of capillary strands,107,108 but these models have been used little in the past decade.

4.3

VEGF overexpression

Vascular endothelial growth factor (VEGF) was injected into the eyes of normal cynomolgus monkeys, and as a result, capillary nonperfusion and vessel dilation and tortuosity developed.109 Preretinal neovascularization was observed throughout peripheral retina, but not in the posterior pole. Arterioles demonstrated endothelial cell hyperplasia and microaneurysmal dilations. Thus, pharmacological doses of VEGF alone were able produce many features of nonproliferative and proliferative diabetic retinopathy.

4.4

IGF overexpression

Normoglycemic/normoinsulinemic transgenic mice overexpressing insulin-like growth factor-1 (IGF-1) in the retina developed several vascular alterations characteristic of diabetic retinopathy, including nonproliferative lesions (pericyte loss, thickened capillary basement

90

T. S. Kern

membrane, intraretinal microvascular abnormalities), proliferative retinopathy, and retinal detachment.110

4.5

Sympathetic denervation

Several retinal lesions consistent with diabetic retinopathy also have been detected after sympathectomy.111 Experimental elimination of sympathetic innervation to the eye by removal of the right superior cervical ganglion resulted in an increase in glial fibrillary acidic protein (GFAP) staining in Müller cells, reduced number of capillary pericytes, and alteration in expression of proteins found in basement membrane.

5.

RETINAL NEOVASCULARIZATION

To date, diabetic animal models (without other genetic modifications or experimental manipulations) have not been demonstrated to reproducibly progress to preretinal neovascularization, and some have criticized the available diabetic models for this failure. In fairness, however, most patients do not develop preretinal neovascularization even after many years of diabetes. Moreover, diabetic or galactosemic animals (at least dogs) have been demonstrated to develop new intraretinal vessels (the precursor to preretinal neovascularization). The new vessels, identified by their lack of basement membrane112 and their characteristic ‘chicken-wire’ pattern, developed within the retina in diabetic dogs and experimentally galactosemic dogs, but did not extend into the vitreous during the initial 5 years of study. New vessels extending into the vitreous have been reported in 2 of 9 dogs fed galactose for 6 to 7 years.89 The diabetic Ren-2 rat develops proliferation of retinal endothelial cells, but overt neovascularization has not yet been demonstrated. This endothelial proliferation can be attenuated by RAS blockade via VEGF-dependent pathways.113 In general, the overall evidence indicates that diabetic or galactosemic animals are not a good model for studies of preretinal angiogenesis. Although preretinal neovascularization has not been detected in diabetic rodents, diabetes is known to increase retinal concentrations of VEGF106,114-118 and other growth factors119,120 in these animals. Why diabetes or hyperglycemia in animal models fails to elicit preretinal neovascularization such as that which occurs in diabetic patients is an important question. Vaso-obliteration and subsequent retinal ischemia are believed to be major causes of neovascularization in the retina. Thus, one likely reason that the diabetic animal models do not develop preretinal neovascularization is that much less vaso-obliteration occurs in the retina of

4. Animal Models of Diabetic Retinopathy

91

the diabetic animals during the short duration in which they are studied (as compared to the more extensive vaso-obliteration that develops over many years in diabetic patients). In addition, perhaps the diabetes-induced increases in growth factor expression in the animals are not quantitatively great enough to stimulate the neovascular response, or other unknown factors also are required.

6.

NONDIABETIC MODELS OF NEOVASCULARIZATION

In the absence of diabetes-induced retinal neovascularization in animal models, investigators have utilized other experimental techniques to study the neovascular response. In support of this experimental approach, there is no evidence at present that diabetes-induced neovascularization of the retina involves signaling pathways distinct from those in the nondiabetic models listed below.

6.1

VEGF overexpression

Transgenic mice have been produced that overexpress VEGF in the photoreceptors under control of the rhodopsin promoter. In these animals, neovascularization has been observed to develop. These new vessels originate from the deep capillary bed and extend through the photoreceptor layer to form vascular complexes in the subretinal space (i.e., towards the choroid).121-124 This new vessel growth is in the opposite direction of that seen in diabetic retinopathy (the superficial capillary bed is not affected122). On the other hand, if VEGF is overexpressed in the front of the retina (in lens), abnormal new vessels develop on the surface of the retina.125 Intravitreal injection of VEGF into the eyes of normal cynomolgus monkeys also resulted in diabetic-like lesions, including areas of capillary nonperfusion, vessel dilation and tortuosity, endothelial cell hyperplasia, and preretinal neovascularization.126 The new vessels originated only from superficial veins and venules, and were observed throughout peripheral retina, but not in the posterior pole.

6.2

Overexpression of IGF

Normoglycemic/normoinsulinemic transgenic mice overexpressing IGF-1 in the retina developed many alterations characteristic of diabetic retinopathy, including loss of pericytes and thickening of basement membrane of retinal

92

T. S. Kern

capillaries. In mice 6 months and older, venule dilatation, IRMAs, and neovascularization of the retina and vitreous cavity were reported.110

6.3

Oxygen-induced retinopathy

Probably the most utilized animal model of preretinal neovascularization is the oxygen-induced retinopathy (OIR) model (see review by Madan and Penn127). In this model, exposure of neonatal animals to elevated concentrations of oxygen impairs development of the normal retinal vasculature, thus resulting in profound retinal ischemia and neovascularization when the animals are removed from the high-oxygen environment. The neovascularization in this model differs from diabetic retinopathy in that the neovascularization in the OIR model occurs acutely in a retina that is not fully differentiated, as compared to the progressive capillary obliteration that develops in the fully differentiated retina in diabetes. Whether or not this difference is important remains to be demonstrated.

6.4

Branch vein occlusion

This model differs from OIR in that the retina and the retinal vasculature are fully differentiated when retinal ischemia is induced by occluding some or all branch veins in the retina.128-136 This neovascular response differs from that in diabetes mainly in the acute nature of the ischemia induced by branch vein occlusion.

6.5

Koletsky rat

The obese SHR rat (koletsky rat; SHR-k; (f/f)) has a nonsense leptin receptor mutation, and the animals are obese, hyperphagic, hypertensive, hyperlipidemic, insulin-resistant, and infertile.137 These animals exhibit retinal vascular changes that include progressive retinal capillary dropout, increased capillary permeability, and in some animals, preretinal neovascularization.138

7.

THERAPIES THAT INHIBIT DIABETIC RETINOPATHY

Capillary occlusion begins early in the course of the retinopathy, and is a major contributor to the progressive retinal ischemia that is believed to drive retinal neovascularization in the proliferative stage. Thus, the increase in

4. Animal Models of Diabetic Retinopathy

93

numbers of acellular and nonperfused capillaries in diabetes likely is causally related to the development of retinal neovascularization in diabetes. Inhibition of the formation of acellular capillaries is expected to inhibit the development of retinal neovascularization and prevent consequent vision loss in diabetes. A variety of seemingly unrelated therapies have been found to inhibit formation of acellular capillaries and other lesions in diabetic animals (Table 1). The mechanism(s) by which these therapies inhibit retinopathy are not clear at present, due in large part to multiple actions of the various therapies. For example, aminoguanidine originally was viewed solely as an inhibitor of advanced glycation endproduct formation, but now is known also to be a potent inhibitor of iNOS activity. It seems unlikely that there are multiple independent biochemical abnormalities that lead to the development of morphologic lesions characteristic of diabetic retinopathy, so a simpler working hypothesis is that the different therapies are inhibiting a common pathway (albeit at different sites along that pathway) that leads to the lesions. A challenge over the coming years will be to see if this “final common pathway” can be identified. Table 4-1. Therapies or gene alterations reported to inhibit vascular lesions in diabetic or galactose-fed animals. 1. Insulin 2. Aminoguanidine 3. Aldose reductase inhibitors 4. Nerve growth factor 5. Antioxidants 6. Antisense oligos against fibronectin 7. High-dose aspirin 8. Pyridoxamine 9. Benfotiamine 10. Deletion of ICAM or CD-18 11. PARP inhibitor

Selected references 25, 28, 31 103-105, 99 94, 85, 86, 97, 98, 49 68, Kern unpublished 39, 42 102 105 43 46 59 141

A significant advance was made several years ago by the observation that retinal capillary cells and neuronal cells were dying by an apoptotic-like process.44,56,69,70,99,139,140 This cell death precedes the appearance of the classical lesions of diabetic retinopathy, and seems likely to play an important role in the development of the lesions.99 In several therapies studied to date, successful inhibition of capillary cell apoptosis led to inhibition of acellular capillaries and other lesions of the early stages of diabetic retinopathy.99,141 Conversely, failure to inhibit capillary cell apoptosis resulted in no inhibition of the retinopathy.99

94

7.1

T. S. Kern

Diabetic retinopathy as a chronic inflammatory disease

Novel insight into the pathogenesis of capillary obliteration and development of diabetic retinopathy recently has come from the recognition that retinas from diabetic animals exhibit biochemical and physiological abnormalities that, in composite, resemble inflammation. These abnormalities include leukostasis, increased expression of adhesion molecules, altered vascular permeability, and increased production of prostaglandins, nitric oxide, and cytokines.59,105,141-153 Assessing the role of inflammatory-like processes in the development of retinopathy, and the ability of anti-inflammatory agents to inhibit the retinopathy, is currently an exciting and rapidly moving area of research.

8.

FUTURE DIRECTIONS

One of the principal advantages of animal models of retinopathy is that they permit biochemical and physiological studies that are otherwise impractical with human subjects, including, for example, evaluation of potentially hazardous treatments. Upon discovery of each biochemical defect associated with retinopathy, new opportunities arise for screening pharmacological agents for their effects on the retinopathy. For screening, rat and mouse models offer the advantage of low cost and relatively rapid onset of significant anatomic criteria such as capillary cell loss. The mouse model can offer, in addition, unique opportunities for exploring the pathogenesis of retinopathy by genetic manipulation of metabolic pathways and pathophysiological syndromes. At present, these diabetic animal models have not been found to be useful models of preretinal neovascularization, but are excellent models to study the pathogenesis of capillary obliteration and cell death, likely the ultimate causes of the retinal neovascularization in diabetes. Preventing diabetes-induced capillary obliteration from ever occurring in the retina seems likely to be a more beneficial therapeutic goal than merely inhibiting neovascularization in an already damaged and ischemic retina.

ACKNOWLEDGMENTS This work was funded by PHS grants EY00300 and DK57733, the Medical Research Service of the Department of Veteran Affairs, and the Kristin C. Dietrich Diabetes Research Award.

4. Animal Models of Diabetic Retinopathy

95

REFERENCES 1. R. Klein, B. E. Klein, S. E. Moss, M. D. Davis, and D. D. DeMets, The Wisconsin Epidemiologic Sudy of Diabetic Retinopathy. I. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years, Arch. Ophthalmol. 102, 520-526 (1984). 2. R. Klein, B. E. Klein, S. E. Moss, M. D. Davis, and D. L. DeMets, The Wisconsin Epidemiologic Study of Diabetic Retinopathy. II. Prevalence and risk of diabetic retinopathy when age at diagnosis is 30 or more years, Arch. Ophthalmol. 102, 527-532 (1984). 3. M. D. Davis, Diabetic retinopathy. A clinical overview, Diabetes Care 15, 1844-1874 (1992). 4. D. A. Antonetti, E. Lieth, A.J. Barber, and T. W. Gardner, Molecular mechanisms of vascular permeability in diabetic retinopathy, Semin. Ophthalmol. 14, 240-248 (1999). 5. T. A. Ciulla et al., Ocular perfusion abnormalities in diabetes, Acta. Ophthalmol. Scand. 80, 468-477 (2002). 6. G. Bresnick, R. Engerman, M. D. Davis, G. de Venecia, and F. L. Myers, Patterns of ischemia in diabetic retinopathy, Trans. Am. Acad. Ophthalmol. Otolaryngol. 81, 694-709 (1976). 7. M. Yokote, in: Early Diabetes Adv. Metab. Disorders, (Academic Press, 1973) pp. 299-304. 8. M. Kaczurowski, Angiopathy of retinal vessels in diabetic mice, Arch. Ophthal. 84, 316-320 (1970). 9. J. Duhault, F. Lebon, and M. Boulanger, in: 7th Europ. Conf. Microcirculation, (Karger, Aberdeen, 1973) pp. 453-458. 10. A. A. F. Sima, R. Garcia-Salinas, and P. K. Basu, The BB Wistar rat: an experimental model for the study of diabetic retinopathy, Metabolism 32(Suppl. 1), 136-140 (1983). 11. S. Chakrabarti and A. A. F. Sima, Effect of aldose reductase inhibition and insulin treatment on retinal capillary basement membrane thickening in BB rats, Diabetes 38, 1181-1186 (1989). 12. D. Toussaint, Contribution a l’etude anatomique et clinque de la retinopathie diabetique chez l’homme et chez l’animal, Pathologia Europea, 108-148 (1968). 13. H. R. Hausler, T. M. Sibay, and J. Campbell, Retinopathy in a dog following diabetes induced by growth hormone, Diabetes 13, 122-126 (1964). 14. W. Gepts and D. Toussaint, Spontaneous diabetes in dogs and cats, Diabetologia 3, 249-264 (1967). 15. T. M. Sibay and H. R. Hausler, Eye findings in two spontaneously diabetic related dogs, Am. J. Ophthalmol. 63, 289-294 (1967). 16. N. Laver, W. G. Robison, Jr., and B. C. Hansen, Spontaneously diabetic monkeys as a model for diabetic retinopathy, Invest. Ophthalmol. Vis. Sci. 35(Suppl), 1733 (1994). 17. R. L. Engerman, R. K. Meyer, and J. A. Buesseler, Effects of alloxan diabetes and steroid hypertension on retinal vasculature, Am. J. Ophthalmol. 58, 965-978 (1964). 18. R. L. Engerman et al., Appropriate animal models for research on human diabetes mellitus and its complications. Ocular complications, Diabetes 31(Suppl. 1), 82-88 (1982). 19. R. L. Engerman and T. S. Kern, Retinopathy in animal models of diabetes, Diabetes/ Metabolism Rev. 11, 109-120 (1995). 20. R. L. Engerman and J. M. B. Bloodworth, Jr., Experimental diabetic retinopathy in dogs, Arch. Ophthalmol. 73, 205-210 (1965). 21. R. L. Engerman, Pathogenesis of diabetic retinopathy, Diabetes 38, 1203-1206 (1989). 22. R. L. Engerman and T. S. Kern, Aldose reductase inhibition fails to prevent retinopathy in diabetic and galactosemic dogs, Diabetes 42, 820-825 (1993).

96

T. S. Kern

23. N. Ashton, Arteriolar involvement in diabetic retinopathy, Brit. J. Ophthal. 37, 282-292 (1953). 24. T. A. Gardiner, A. W. Stitt, H. R. Anderson, and D. B. Archer, Selective loss of vascular smooth muscle cells in the retinal microcirculation of diabetic dogs, Brit. J. Ophthal. 78, 54-60 (1994). 25. R. L. Engerman, J. M. B. Bloodworth, Jr., and S. Nelson, Relationship of microvascular disease in diabetes to metabolic control, Diabetes 26, 760-769 (1977). 26. D. L. Hatchell, R. D. Braun, G. A. Lutty, D. S. McLeod, and C. A. Toth, Progression of diabetic retinopathy in a cat, Invest. Ophthalmol. Vis. Sci. 36, S1067 (1995). 27. R. A. Linsenmeier et al., Retinal hypoxia in long-term diabetic cats, Invest. Ophthalmol. Vis. Sci. 39, 1647-1657 (1998). 28. R. L. Engerman and T. S. Kern, Progression of incipient diabetic retinopathy during good glycemic control, Diabetes 36, 808-812 (1987). 29. Diabetes Control and Complications Trial Research Group, The effect of intensive treatment of diabetes on the development of long-term complications in insulin-dependent diabetes mellitus, N. Engl. J. Med. 329, 977-986 (1993). 30. United Kingdom Prospective Diabetes Study, Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes, Lancet 352, 837-853 (1998). 31. H. P. Hammes et al., Islet transplantation inhibits diabetic retinopathy in the sucrose-fed diabetic Cohen diabetic rat, Invest. Ophthalmol. Vis. Sci. 34, 2092-2096 (1993). 32. T. S. Kern and R. L. Engerman, Comparison of retinal lesions in alloxan-diabetic rats and galactose-fed rats, Curr. Eye Res. 13, 863-867 (1994). 33. H. P. Hammes et al., Aminoguanidine does not inhibit the initial phase of experimental diabetic retinopathy in rats, Diabetologia 38, 269-273 (1995). 34. H. P. Hammes et al., Secondary intervention with aminoguanidine retards the progression of diabetic retinopathy in rat model, Diabetologia 38, 656-660 (1995). 35. T. S. Kern and R. L. Engerman, Galactose-induced retinal microangiopathy in rats, Invest. Ophthalmol. Vis. Sci. 36, 490-496 (1995). 36. H. P. Hammes et al., Acceleration of experimental diabetic retinopathy in the rat by omega-3 fatty acids, Diabetologia 39, 251-255 (1996). 37. T. S. Kern, R. Kowluru, and R. L. Engerman, in: Lessons from Animal Diabetes, edited by E. Shafrir (Smith-Gordon, London, 1996) pp. 395-408. 38. C. D. Agardh, E. Agardh, H. Zhang, and C. G. Ostenson, Altered endothelial/pericyte ratio in Goto-Kakizaki rat retina, J. Diabetes Complications 11, 158-162 (1997). 39. H. P. Hammes, A. Bartmann, L. Engel, and P. Wulfroth, Antioxidant treatment of experimental diabetic retinopathy in rats with nicanartine, Diabetologia 40, 629-634 (1997). 40. S. H. Kim, Y. K. Chu, O. W. Kwon, S. A. McCune, and F. H. Davidorf, Morphologic studies of the retina in a new diabetic model; SHR/N:Mcc-cp rat, Yonsei Med. J. 39, 453-462 (1998). 41. N. Miyamura, I. A. Bhutto, and T. Amemiya, Retinal capillary changes in Otsuka LongEvans Tokushima fatty rats (spontaneously diabetic strain). Electron-microscopic study, Ophthalmic Res. 31, 358-366 (1999). 42. R. Kowluru, J. Tang, and T. S. Kern, Abnormalities of retinal metabolism in diabetes and galactosemia. VII. Effects of long-term administration of antioxidants on retinal oxidative stress and the development of retinopathy, Diabetes 50, 1938-1942 (2001). 43. A. Stitt et al., The AGE inhibitor pyridoxamine inhibits development of retinopathy in experimental diabetes, Diabetes 51, 2826-2832 (2002).

4. Animal Models of Diabetic Retinopathy

97

44. V. Asnaghi, C. Gerhardinger, T. Hoehn, A. Adeboje, and M. Lorenzi, A role for the polyol pathway in the early neuroretinal apoptosis and glial changes induced by diabetes in the rat, Diabetes 52, 506-511 (2003). 45. T. A. Gardiner, H. R. Anderson, and A. W. Stitt, Inhibition of advanced glycation endproducts protects against retinal capillary basement membrane expansion during longterm diabetes, J. Pathol. 201, 328-333 (2003). 46. H. P. Hammes et al., Benfotiamine blocks three major pathways of hyperglycemic damage and prevents experimental diabetic retinopathy, Nat. Med. 9, 294-299. (2003). 47. N. Kato, S. Yashima, T. Suzuki, Y. Nakayama, and T. Jomori, Long-term treatment with fidarestat suppresses the development of diabetic retinopathy in STZ-induced diabetic rats, J. Diabetes Complications 17, 374-379 (2003). 48. Z. Y. Lu, I. A. Bhutto, and T. Amemiya, Retinal changes in Otsuka long-evans Tokushima Fatty rats (spontaneously diabetic rat)--possibility of a new experimental model for diabetic retinopathy, Jpn. J. Ophthalmol. 47, 28-35 (2003). 49. Z. Dagher et al., Studies of rat and human retinas predict a role for the polyol pathway in human diabetic retinopathy, Diabetes 53, 2404-2411 (2004). 50. R. A. Kowluru, A. Kowluru, S. Chakrabarti, and Z. Khan, Potential contributory role of H-Ras, a small G-protein, in the development of retinopathy in diabetic rats, Diabetes 53, 775-783 (2004). 51. S. Chakrabarti, A. A. F. Sima, W. J. Tze, and J. Tai, Prevention of diabetic retinal capillary pericyte degeneration and loss by pancreatic islet allograft, Curr. Eye Res. 6, 649-658 (1987). 52. B. I. Gaynes, and J. B. Watkins, III, Comparison of glucose, sorbitol and fructose accumulation in lens and liver of diabetic and insulin-treated rats and mice, Comp. Biochem. Biophys. 92B, 685-690 (1989). 53. F. Morii, M. Hattori, E. Miyasaki, S. Fukuchi, and I. Tsukahara, Histological studies on congenitally diabetic KK mice, J. Ophthalmol. Soc. Japan 25, 372-389 (1974). 54. R. A. Cuthbertson and T. E. Mandel, The effect of murine fetal islet transplants on renal and retinal capillary basement membrane thickness, Transplant Proc. 19, 2919-2921 (1987). 55. A. Agren, G. Rehn, and P. Naeser, Morphology and enzyme activities of the retinal capillaries in streptozotocin-diabetic mice, Acta Ophthalmologica 57, 1065-1069 (1979). 56. S. Mohr, J. Tang, and T. S. Kern, Caspase activation in retinas of diabetic and galactosemic mice and diabetic patients, Diabetes 51, 1172-1179 (2002). 57. R. A. Feit-Leichman et al., The mouse model of diabetic retinopathy: Vascular damage without Müller glial cell activation and neuronal loss, Invest. Ophthalmol. Vis. Sci. 46, 4281-4287 (2004). 58. P. M. Martin, P. Roon, T. K. Van Ells, V. Ganapathy, and S. B. Smith, Death of retinal neurons in streptozotocin-induced diabetic mice, Invest. Ophthalmol. Vis. Sci. 45, 3330-3336 (2004). 59. A. M. Joussen et al., A central role for inflammation in the pathogenesis of diabetic retinopathy, Faseb J. 18, 1450-1452 (2004). 60. H. P. Hammes et al., The relationship of glycaemic level to advanced glycation endproduct (AGE) accumulation and retinal pathology in the spontaneous diabetic hamster, Diabetologia 41, 165-170 (1998). 61. J. M. Bloodworth, Jr. and D. L. Molitor, Ultrastructural aspects of human and canine diabetic retinopathy, Invest. Ophthalmol. 4, 1037-1048 (1965). 62. G. M. Bresnick and M. Palta, Predicting progression of severe proliferative diabetic retinopathy, Arch. Ophthalmol. 105, 810-814 (1987).

98

T. S. Kern

63. G. H. Bresnick and M. Palta, Oscillatory potential amplitudes: Relation to severity of diabetic retinopathy, Arch. Ophthal. 105, 929-933 (1987). 64. G. H. Bresnick, Excitotoxins: A possible new mechanism for the pathogenesis of ischemic retinal damage, Arch. Ophthalmol. 107, 339-341 (1989). 65. G. H. Bresnick, Diabetic retinopathy viewed as a neurosensory disorder, Arch. Ophthalmol. 104, 989-990 (1986). 66. A. A. Sima, W. X. Zhang, P. V. Cherian, and S. Chakrabarti, Impaired visual evoked potential and primary axonopathy of the optic nerve in the diabetic BB/W-rat, Diabetologia 35, 602-607 (1992). 67. M. Kamijo, P. V. Cherian, and A. A. F. Sima, The preventive effect of aldose reductase inhibition on diabetic optic neuropathy in the BB/W rat, Diabetologia 36, 893-898 (1993). 68. H. P. Hammes, H. J. Federoff, and M. Brownlee, Nerve growth factor prevents both neuroretinal programmed cell death and capillary pathology in experimental diabetes, Mol. Medicine 1, 527-534 (1995). 69. A. J. Barber et al., Neural apoptosis in the retina during experimental and human diabetes. Early onset and effect of insulin, J. Clin. Invest. 102, 783-791 (1998). 70. E. Lieth, T. W. Gardner, A. J. Barber, and D. A. Antonetti, Retinal neurodegeneration: early pathology in diabetes, Clin. Experiment Ophthalmol. 28, 3-8 (2000). 71. E. Agardh, A. Bruun, and C. D. Agardh, Retinal glial cell immunoreactivity and neuronal cell changes in rats with STZ-induced diabetes, Curr. Eye Res. 23, 276-284 (2001). 72. Y. Aizu, K. Oyanagi, J. Hu, and H. Nakagawa, Degeneration of retinal neuronal processes and pigment epithelium in the early stage of the streptozotocin-diabetic rats, Neuropathology 22, 161-170 (2002). 73. Y. Aizu et al., Topical instillation of ciliary neurotrophic factor inhibits retinal degeneration in streptozotocin-induced diabetic rats, Neuroreport 14, 2067-2071 (2003). 74. S. H. Park et al., Apoptotic death of photoreceptors in the streptozotocin-induced diabetic rat retina, Diabetologia 46, 1260-1268 (2003). 75. L. L. Kusner, V. P. Sarthy, and S. Mohr, Nuclear translocation of glyceraldehyde-3phosphate dehydrogenase: a role in high glucose-induced apoptosis in retinal Muller cells, Invest. Ophthalmol. Vis. Sci. 45, 1553-1561 (2004). 76. E. Lieth et al., Glial reactivity and impaired glutamate metabolism in short-term experimental diabetic retinopathy. The Penn State Retina Research Group, Diabetes 47, 815-820 (1998). 77. M. Mizutani, C. Gerhardinger, and M. Lorenzi, Muller cell changes in human diabetic retinopathy, Diabetes 47, 445-449 (1998). 78. A. J. Barber, D. A. Antonetti, and T. W. Gardner, Altered expression of retinal occludin and glial fibrillary acidic protein in experimental diabetes. The Penn State Retina Research Group, Invest. Ophthalmol. Vis. Sci. 41, 3561-3568 (2000). 79. X. X. Zeng, Y. K. Ng, and E. A. Ling, Neuronal and microglial response in the retina of streptozotocin-induced diabetic rats, Vis. Neurosci. 17, 463-471 (2000). 80. E. Rungger-Brandle, A. A. Dosso, and P. M. Leuenberger, Glial reactivity, an early feature of diabetic retinopathy, Invest. Ophthalmol. Vis. Sci. 41, 1971-1980 (2000). 81. Q. Li, E. Zemel, B. Miller, and I. Perlman, Early retinal damage in experimental diabetes: electroretinographical and morphological observations, Exp. Eye Res. 74, 615-625 (2002). 82. A. J. Barber et al., The Ins2Akita mouse as a model of early retinal complications in diabetes, Invest. Ophthalmol. Vis. Sci. 46, 2210-2218 (2005). 83. R. L. Engerman and T. S. Kern, Experimental galactosemia produces diabetic-like retinopathy, Diabetes 31(Suppl), 26A (1982).

4. Animal Models of Diabetic Retinopathy

99

84. R. L. Engerman and T. S. Kern, Experimental galactosemia produces diabetic-like retinopathy, Diabetes 33, 97-100 (1984). 85. P. F. Kador, Y. Akagi, H. Terubayashi, M. Wyman, and J. H. Kinoshita, Prevention of pericyte ghost formation in retinal capillaries of galactose-fed dogs by aldose reductase inhibitors, Arch. Ophthalmol. 106, 1099-1102 (1988). 86. P. F. Kador et al., Prevention of retinal vessel changes associated with diabetic retinopathy in galactose-fed dogs by aldose reductase inhibitors, Arch. Ophthalmol. 108, 1301-1309 (1990). 87. R. N. Frank, The galactosemic dog. A valid model for both early and late stages of diabetic retinopathy, Arch. Ophthalmol. 113, 275-276 (1995). 88. R. L. Engerman and T. S. Kern, Retinopathy in galactosemic dogs continues to progress after cessation of galactosemia, Arch. Ophthalmol. 113, 355-358 (1995). 89. P. F. Kador, Y. Takahashi, M. Wyman, and F. Ferris, III, Diabeteslike proliferative retinal changes in galactose-fed dogs, Arch. Ophthalmol. 113, 352-354 (1995). 90. T. S. Kern and R. L. Engerman, Vascular lesions in diabetes are distributed nonuniformly within the retina, Exp. Eye Res. 60, 545-549 (1995). 91. H. Neuenschwander, Y. Takahashi, and P. F. Kador, Dose-dependent reduction of retinal vessel changes associated with diabetic retinopathy in galactose-fed dogs by the aldose reductase inhibitor M79175, J. Ocul. Pharmacol. Ther. 13, 517-528 (1997). 92. T. Kobayashi et al., Retinal vessel changes in galactose-fed dogs, Arch. Ophthalmol. 116, 785-789 (1998). 93. P. F. Kador et al., Effect of galactose diet removal on the progression of retinal vessel changes in galactose-fed dogs, Invest. Ophthalmol. Vis. Sci. 43, 1916-1921 (2002). 94. W. G. Robison, Jr., M. Nagata, N. Laver, T. C. Hohman, and J. H. Kinoshita, Diabeticlike retinopathy in rats prevented with an aldose reductase inhibitor, Invest. Ophthalmol. Vis. Sci. 30, 2285-2292 (1989). 95. W. G. Robison, Jr., M. Nagata, T. N. Tillis, N. Laver, and J. H. Kinoshita, Aldose reductase and pericyte-endothelial cell contacts in retina and optic nerve, Invest. Ophthalmol. Vis. Sci. 30, 2293-2299 (1989). 96. W. G. Robison, Jr., Diabetic retinopathy: galactose-fed rat model, Invest. Ophthalmol. Vis. Sci. 36, 4A, 1743-1744 (1995). 97. W. G. Robison, Jr., N. M. Laver, J. L. Jacot, and J. P. Glover, Sorbinil prevention of diabetic-like retinopathy in the galactose-fed rat model, Invest. Ophthalmol. Vis. Sci. 36, 2368-2380 (1995). 98. W. G. Robison, Jr. et al., Diabetic-like retinopathy ameliorated with the aldose reductase inhibitor WAY-121,509, Invest. Ophthalmol. Vis. Sci. 37, 1149-1156 (1996). 99. T. S. Kern et al., Response of capillary cell death to aminoguanidine predicts the development of retinopathy: comparison of diabetes and galactosemia, Invest. Ophthalmol. Vis. Sci. 41, 3972-3978 (2000). 100. R. A. Kowluru, J. Tang, and T. S. Kern, Abnormalities of retinal metabolism in diabetes and experimental galactosemia. VII. Effect of long-term administration of antioxidants on the development of retinopathy, Diabetes 50, 1938-1942 (2001). 101. T. S. Kern and R. L. Engerman, A mouse model of diabetic retinopathy, Arch. Ophthalmol. 114, 986-990 (1996). 102. S. Roy, T. Sato, G. Paryani, and R. Kao, Downregulation of fibronectin overexpression reduces basement membrane thickening and vascular lesions in retinas of galactose-fed rats, Diabetes 52, 1229-1234 (2003). 103. H. P. Hammes, S. Martin, K. Federlin, K. Geisen, and M. Brownlee, Aminoguanidine treatment inhibits the development of experimental diabetic retinopathy, Proc. Natl. Acad. Sci. USA 88, 11555-11558 (1991).

100

T. S. Kern

104. H. P. Hammes et al., Aminoguanidine inhibits the development of accelerated diabetic retinopathy in the spontaneous hypertensive rat, Diabetologia 37, 32-35 (1994). 105. T. S. Kern and R. L. Engerman, Pharmacologic inhibition of diabetic retinopathy: Aminoguanidine and aspirin, Diabetes 50, 1636-1642 (2001). 106. R. N. Frank, R. Amin, A. Kennedy, and T. C. Hohman, An aldose reductase inhibitor and aminoguanidine prevent vascular endothelial growth factor expression in rats with long-term galactosemia, Arch. Ophthalmol. 115, 1036-1047 (1997). 107. L. Yanko, I. C. Michaelson, and A. M. Cohen, The retinopathy of sucrose-fed rats, Israel J. Med. Sci. 8, 1633-1636 (1972). 108. R. Boot-Handford and H. Heath, Identification of fructose as the retinopathic agent associated with the ingestion of sucrose-rich diets in the rat, Metab. 29, 1247-1252 (1980). 109. M. J. Tolentino et al., Intravitreous injections of vascular endothelial growth factor produce retinal ischemia and microangiopathy in an adult primate, Ophthalmology 103, 1820-1828 (1996). 110. J. Ruberte et al., Increased ocular levels of IGF-1 in transgenic mice lead to diabeteslike eye disease, J. Clin. Invest. 113, 1149-1157 (2004). 111. L. A. Wiley, G. R. Rupp, and J. J. Steinle, Sympathetic innervation regulates basement membrane thickening and pericyte number in rat retina, Invest. Ophthalmol. Vis. Sci. 46, 744-748 (2005). 112. I. H. Wallow and R. L. Engerman, Permeability and patency of retinal blood vessels in experimental diabetes, Invest. Ophthalmol. 16, 447-461 (1977). 113. C. J. Moravski et al., The renin-angiotensin system influences ocular endothelial cell proliferation in diabetes: transgenic and interventional studies, Am. J. Pathol. 162, 151-160 (2003). 114. T. Murata et al., The relation between expression of vascular endothelial growth factor and breakdown of the blood-retinal barrier in diabetic rat retinas, Lab. Invest. 74, 819-825 (1996). 115. H. Sone et al., Ocular vascular endothelial growth factor levels in diabetic rats are elevated before observable retinal proliferative changes, Diabetologia 40, 726-730 (1997). 116. H. P. Hammes, J. Lin, R. G. Bretzel, M. Brownlee, and G. Breier, Upregulation of the vascular endothelial growth factor/vascular endothelial growth factor receptor system in experimental background diabetic retinopathy of the rat, Diabetes 47, 401-406 (1998). 117. R. E. Gilbert et al., Vascular endothelial growth factor and its receptors in control and diabetic rat eyes, Lab. Invest. 78, 1017-1027 (1998). 118. Y. Segawa et al., Upregulation of retinal vascular endothelial growth factor mRNAs in spontaneously diabetic rats without ophthalmoscopic retinopathy. A possible participation of advanced glycation end products in the development of the early phase of diabetic retinopathy, Ophthalmic Res. 30, 333-339 (1998). 119. H. Kuang et al., The potential role of IGF-I receptor mRNA in rats with diabetic retinopathy, Chin. Med. J. (Engl) 116, 478-480 (2003). 120. V. Poulaki et al., Insulin-like growth factor-I plays a pathogenetic role in diabetic retinopathy, Am. J. Pathol. 165, 457-469 (2004). 121. N. Okamoto et al., Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization, Am. J. Pathol. 151, 281-291 (1997). 122. S. A. Vinores, N. L. Derevjanik, M. A. Vinores, N. Okamoto, and P. A. Campochiaro, Sensitivity of different vascular beds in the eye to neovascularization and blood-retinal barrier breakdown in VEGF transgenic mice, Adv. Exp. Med. Biol. 476, 129-138 (2000).

4. Animal Models of Diabetic Retinopathy

101

123. E. Yamada et al., TIMP-1 promotes VEGF-induced neovascularization in the retina, Histol. Histopathol. 16, 87-97 (2001). 124. K. Ohno-Matsui et al., Inducible expression of vascular endothelial growth factor in adult mice causes severe proliferative retinopathy and retinal detachment, Am. J. Pathol. 160, 711-719 (2002). 125. S. A. Vinores et al., Experimental models of growth factor-mediated angiogenesis and blood-retinal barrier breakdown, Gen. Pharmacol. 35, 233-239 (2000). 126. M. J. Tolentino et al., Pathologic features of vascular endothelial growth factor-induced retinopathy in the nonhuman primate, Am. J. Ophthalmol. 133, 373-385. (2002). 127. A. Madan and J. S. Penn, Animal models of oxygen-induced retinopathy, Front. Biosci. 8, d1030-d1043 (2003). 128. R. P. Danis and I. H. Wallow, Microvascular changes in experimental branch retinal vein occlusion, Ophthalmology 94, 1213-1221 (1987). 129. C. J. Pournaras, M. Tsacopoulos, K. Strommer, N. Gilodi, and P. M. Leuenberger, Experimental retinal branch vein occlusion in miniature pigs induces local tissue hypoxia and vasoproliferative microangiopathy, Ophthalmology 97, 1321-1328 (1990). 130. C. A. Wilson and D. L. Hatchell, Photodynamic retinal vascular thrombosis. Rate and duration of vascular occlusion, Invest. Ophthalmol. Vis. Sci. 32, 2357-2365 (1991). 131. R. P. Danis, Y. Yang, S. J. Massicotte, and H. C. Boldt, Preretinal and optic nerve head neovascularization induced by photodynamic venous thrombosis in domestic pigs, Arch. Ophthalmol. 111, 539-543 (1993). 132. M. Minamikawa, K. Yamamoto, and H. Okuma, H. [Experimental retinal branch vein occlusion. 4. Pathological changes in the middle and late stage]. Nippon Ganka Gakkai Zasshi 97, 920-927 (1993). 133. C. J. Pournaras, Retinal oxygen distribution: Its role in the physiopathology of vasoproliferative microangiopathies, Retina 15, 332-347 (1995). 134. R. P. Danis, D. P. Bingaman, Y. Yang, and B. Ladd, Inhibition of preretinal and optic nerve head neovascularization in pigs by intravitreal triamcinolone acetonide, Ophthalmology 103, 2099-2104 (1996). 135. R. Danis et al., Intravitreous anti-raf-1 kinase antisense oligonucleotide as an angioinhibitory agent in porcine preretinal neovascularization, Curr. Eye Res. 26, 45-54 (2003). 136. H. Akiyama et al., Inhibition of ocular angiogenesis by an adenovirus carrying the human von Hippel-Lindau tumor-suppressor gene in vivo, Invest. Ophthalmol. Vis. Sci. 45, 1289-1296 (2004). 137. R. J. Koletsky and P. Ernsberger, Obese SHR (Koletsky rat): a model for the interactions between hypertension and obesity, Genet. Hyperten. 218, 373-375 (1992). 138. S. S. Huang, S. A. Khosrof, R. J. Koletsky, B. A. Benetz, and P. Ernsberger, Characterization of retinal vascular abnormalities in lean and obese spontaneously hypertensive rats, Clin. Exp. Pharmacol. Physiol. 22(Suppl. 1), S129-S131 (1995). 139. M. Mizutani, T. S. Kern, and M. Lorenzi, Accelerated death of retinal microvascular cells in human and experimental diabetic retinopathy, J. Clin. Invest. 97, 2883-2890 (1996). 140. F. Podesta et al., Bax is increased in the retina of diabetic subjects and is associated with pericyte apoptosis in vivo and in vitro. Am. J. Pathol. 156, 1025-1032 (2000). 141. L. Zheng, S. Szabó, and T. Kern, Poly(ADP-ribose) polymerase is involved in the development of diabetic retinopathy via regulation of NF-6B, Diabetes 53, 2960-2967 (2004).

102

T. S. Kern

142. T. W. Gardner, D. A. Antonetti, A. J. Barber, K. F. LaNoue, and M. Nakamura, New insights into the pathophysiology of diabetic retinopathy: potential cell-specific therapeutic targets, Diabetes Technol. Ther. 2, 601-608 (2000). 143. A. M. Joussen et al., Leukocyte-mediated endothelial cell injury and death in the diabetic retina, Am. J. Pathol. 158, 147-152 (2001). 144. M. Lorenzi and C. Gerhardinger, Early cellular and molecular changes induced by diabetes in the retina, Diabetologia 44, 791-804 (2001). 145. A. P. Adamis, Is diabetic retinopathy an inflammatory disease? Br. J. Ophthalmol. 86, 363-365 (2002). 146. G. Romeo, W. H. Liu, V. Asnaghi, T. S. Kern, and M. Lorenzi, Activation of nuclear factor-kappaB induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes, Diabetes 51, 2241-2248. (2002). 147. Y. Du, M. A. Smith, C. M. Miller, and T. S. Kern, Diabetes-induced nitrative stress in the retina, and correction by aminoguanidine, J. Neurochem. 80, 771-779 (2002). 148. A. M. Joussen et al., Nonsteroidal anti-inflammatory drugs prevent early diabetic retinopathy via TNF-alpha suppression, Faseb J. 16, 438-440 (2002). 149. T. Abiko et al., Characterization of retinal leukostasis and hemodynamics in insulin resistance and diabetes: role of oxidants and protein kinase-C activation, Diabetes 52, 829-837 (2003). 150. K. Yamashiro et al., Platelets accumulate in the diabetic retinal vasculature following endothelial death and suppress blood-retinal barrier breakdown, Am. J. Pathol. 163, 253-259 (2003). 151. A. M. Joussen et al., Suppression of Fas-FasL-induced endothelial cell apoptosis prevents diabetic blood-retinal barrier breakdown in a model of streptozotocin-induced diabetes, Faseb J. 17, 76-78 (2003). 152. S. Mohr, Potential new strategies to prevent the development of diabetic retinopathy, Expert Opin. Investig. Drugs 13, 189-198 (2004). 153. Y. Du, V. Sarthy, and T. Kern, Interaction between NO and COX pathways in retinal cells exposed to elevated glucose and retina of diabetic rats, Am. J. Physiol. 287, R735-R741 (2004).

Chapter 5 NEOVASCULARIZATION IN MODELS OF BRANCH RETINAL VEIN OCCLUSION

Ronald P. Danis, MD,1 and David P. Bingaman, PhD, DVM2

1 Director of the Fundus Photograph Reading Center, Professor of Ophthalmology, Department of Ophthalmology and Visual Sciences, University of Wisconsin-Madison, and 2 Assistant Director, Ocular Angiogenesis & Diabetic Retinopathy Programs, Retina Discovery Unit, Alcon Research, Ltd., Fort Worth, Texas

Abstract:

1.

Branch retinal vein occlusion can be achieved in several species using laser photocoagulation with or without photodynamic agents. The neovascular response shows high variability within and between species. However, animal models of ischemia-associated intraocular neovascularization from branch retinal vein occlusion have been employed with success to demonstrate therapeutic effects of pharmaceutical agents and to study mechanisms of angiogenesis.

INTRODUCTION

Retinal ischemia is the primary cause of preretinal, optic nerve head, and iris neovascularization (NV) in human ocular disease. Causes of retinal vascular occlusion include diabetes mellitus, radiation, emboli, thrombosis, and inflammation. Occlusions, which can subsequently induce neovascularization, may affect either large or small caliber retinal vessels. Diabetic retinopathy and radiation retinopathy typically produce ischemia by affecting the microcirculation. Large vessel occlusions are identified by the primary site of obstruction: branch or central retinal vein occlusion, and branch or central retinal artery occlusion. NV, as a complication of retinal ischemia, is highly prevalent in human retinal disease. Diabetic retinopathy (DR) is one of the most common causes of acquired blindness in developed nations, causing about 12% of cases of new blindness in the U.S. annually. Diabetes mellitus afflicts nearly 103 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 103–117. © Springer Science+Business Media B.V. 2008

104

R. P. Danis and D. P. Bingaman

14 million Americans.1 Approximately 5% of all diabetic patients develop ocular neovascularization. Branch retinal vein occlusion (BVO) is the second most common retinal vascular disease;2 about 50% of large BVO cases have significant ischemia, and of these about 40% will develop neovascularization.3 Central retinal vein occlusions are also common clinical problems; approximately 30% of these cases have severe ischemia, and of these 40 to 60% may also suffer neovascular complications.4 Retinal ischemia can stimulate pathological angiogenesis in multiple ocular tissues, such as the posterior segment (at the optic nerve head or growing out of the retina) or the anterior segment on the anterior surface of the iris. Posterior segment NV requires vitreous to serve as a collagen scaffold for neovascular growth. Importantly, eyes with the vitreous removed by surgical vitrectomy do not develop retinal NV, except where there is persistent vitreous. Posterior segment NV can penetrate the internal limiting membrane to develop along the posterior vitreous face or emanate into the vitreous gel. Posterior segment NV is also associated with a highly variable fibrous component. Typically, the early clinical appearance is that of “naked” vessels growing out of the retina or optic nerve into the vitreous. While these vessels may appear as simple vascular proliferation to the clinician, histopathology invariably demonstrates a fibrous component. With continued proliferation, the fibrous component tends to become more evident as whitish tissue accompanying the vessels. In most cases, untreated preretinal and optic nerve head NV evolves with a time course of months to years. The vessels gradually accumulate fibrous extravascular tissue until the vascular component eventually begins to atrophy and the lesion involutes. During involution, the fibrous component predominates, and the vessels may become grossly unapparent on clinical examination, although histopathology often shows some perfused vasculature. Blinding complications occur as a consequence of the fibrous component of posterior segment NV. Optic nerve head NV may lead to bleeding into the vitreous cavity. Preretinal NV may also result in vitreous hemorrhage, but in addition, it may lead to a potentially more grave complication: tractional retinal detachment. Vitreous hemorrhage and retinal traction occur when the cellular component of the fibrovascular tissue contracts, causing rupture of the fragile new vessels and detachment of the retina. Vitreous hemorrhage, when mild, may clear spontaneously and cause only mild or transient visual impairment. Severe hemorrhage or tractional retinal detachment involving or threatening the macula can be blinding if left untreated and often requires surgical intervention. Iris NV, also known as rubeosis iridis, most often develops first as a lacy configuration of vessels around the pupil, on the iris surface, or as small tufts at the pupillary sphincter. In more severe cases, the NV grows across the

5. NV in Models of Branch Retinal Vein Occlusion

105

entire iris surface and across the iridocorneal angle at the base of the iris. As in posterior segment NV, a fibrous tissue component of the NV proliferates with the vessels, and this eventually leads to contractile changes within the membrane. Contraction on the iris surface may cause abnormal enlargement of the pupil (anisocoria) and eversion of some of the posterior iris tissue through the pupil (ectropion uvea). The iridocorneal angle includes the trabecular meshwork, which is the major pathway for egress of aqueous humor from the globe. Contraction of a fibrovascular membrane may cause obstruction of the trabecular meshwork and closure of the angle, which frequently results in a very dire complication: neovascular glaucoma. Neovascular glaucoma is characterized by a very high intraocular pressure that may not be responsive to medication. Patients with neovascular glaucoma may experience severe pain, severe optic nerve damage, and retinal infarction. Surgical intervention may salvage the eye in some cases. Another mechanism whereby iris NV may cause pressure elevation and vision loss is through bleeding into the anterior segment, causing hyphema. The blockade of the iridocorneal angle by red blood cells can produce ocular hypertension. Interestingly, the threshold for ischemia-induced injury to ocular tissues appears to vary with the underlying etiology. Iris NV is much more prevalent among cases of central retinal vein occlusion,4 whereas DR and BVO are more commonly associated with posterior segment NV. Both forms of intraocular NV can occur simultaneously in the same eye. Overall, iris NV is not observed as often as posterior segment NV in retinal clinical practice, because DR and BVO are more common diseases. Treatment of preretinal and optic nerve head NV with laser also inhibits iris NV, which limits its presentation in ischemic eyes that otherwise are at risk. The mechanisms by which ischemia leads to ocular NV are detailed elsewhere in this text. Briefly, NV is caused by a complex interplay of factors, including hypoxia-regulated soluble growth factors (VEGF, HGF, IGF-1, PEDF, etc.), extracellular matrix components found in the vitreous and fibrovascular tissue, and the influence of immune cells. Macrophages invariably occur in specimens of pathological NV of all types. Their importance is generally acknowledged, since macrophages are a prominent source of angiogenic growth factors as well as chemo-attractants. The potential for pharmacotherapeutic modulation of angiogenesis is the principal stimulus for the development of relevant animal models of ocular NV.

106

2.

R. P. Danis and D. P. Bingaman

ANIMAL MODELS OF OCULAR NEOVASCULARIZATION FROM BVO

The study of BVO in animal models has been attractive to researchers for decades. It is technically possible to occlude retinal veins by laser photocoagulation, a relatively non-invasive methodology. Animals with eye sizes comparable to man, such as pigs and primates, can therefore be studied clinically and histopathologically to monitor the evolution of retinal ischemia and neovascular responses. The relatively large eyes of primates and pigs allow surgical implants or other drug delivery devices not accommodated by smaller rodent eyes. The study of NV can be approached advantageously through ocular models approximating the human condition. Determining appropriate therapeutic targets is one of the challenges of studying posterior segment NV in animals. Another is the relative rarity with which this is created in laboratory animals. BVO models have been reported in mice, rats, rabbits, cats, dogs, various primate species, and pigs.5-12 Pigs provide the only BVO model where preretinal and optic nerve head (ONH) NV are consistently produced.13,14 The underlying causes of resistance of other species to the development of ischemia-induced retinal NV are unknown. However, inter-species (and even inter-strain) variations may affect the threshold for manifestation of ocular NV from the same stimulus.15 Another problematic feature of models of ocular NV due to BVO is the difficulty in quantifying the extent of NV. During efficacy studies, it is mandatory to have reproducible and standardized outcome measures. In BVO models, ordinal semi-quantitative grading systems have been established and applied in primate iris NV and pig retinal NV. However, these methods are dependent on subjective evaluation of photographic images or gross pathology with histopathological confirmation. More recently, histopathological assessment techniques have been employed to provide more reliable quantification.16 As noted, species differences in the neovascular proliferation following BVO are quite large. In monkeys, preretinal and optic nerve head NV are quite rare, whereas iris NV tends to be robust and quantifiable. Quantification of iris NV in the monkey model has relied exclusively on angiographic imaging.17-19 Iris angiography can be employed to determine the extent of the neovascular response in a masked manner. Since fundus features are not seen in the angiographic images, there is less potential for observer bias with this method.

5. NV in Models of Branch Retinal Vein Occlusion

3.

107

PORCINE MODELS OF BVO

BVOs in pigs were produced by intense argon laser photocoagulation by Kohner and colleagues more than 30 years ago.6 These researchers demonstrated many of the ultrastructural and clinical findings seen in human patients with BVO; however, they did not observe pathological NV despite long term follow-up. The studies of experimental BVO in pigs from Moorfield’s group in the 1970’s did not describe NV despite careful histopathological evaluation.6 It is not clear why a group in a different laboratory was later able to produce a 100% incidence of preretinal and optic nerve head NV from a similar stimulus.14 Pournaras and colleagues first noted neovascular changes following argon laser-induced BVO in miniature pigs.13 They reported a high rate of vitreous hemorrhage with their laser technique. NV was demonstrated histopathologically in about 50% of their animals. Although Pournaras’ group has not employed the model to assay NV per se, they have used their model of BVO in pigs to define the degree of hypoxia achieved using intraocular oxygen sensors. They have studied hypoxic reactions of the tissues, soluble growth factor regulation, and laser effects in ischemic retinas.13,20 Danis and colleagues standardized this model to reliably produce quantifiable NV in domestic pigs.14 An innovation was the use of intravenous Rose Bengal dye as a photodynamic agent. Rose Bengal is a phthalocyanine dye with structural similarities to fluorescein; however, it fluoresces and produces oxygen free radicals from hydrolysis at 555 nm. This absorption peak is close to the 514-nm light produced by the green argon laser. Consequently, use of this dye allows retinal branch veins to be reliably closed with one treatment session with minimal vitreous hemorrhage (Figure 1). When 50% or more of the retinal venous territory was occluded in this manner, a 100% incidence of preretinal and optic nerve head NV was observed (Figure 2). The procedure was effective for producing ocular NV in both domestic and miniature pigs.21 Vascular occlusion in this pig model is produced by combined thermal necrosis and photodynamic thrombosis. Intense thermal burns are produced despite low laser power due to the long duration required to block vessel perfusion. The thermal necrosis appears necessary to produce permanent occlusions. When branch points of the retinal veins are targeted, a focal constriction of the vasculature can be exploited to advantage in producing the occlusion. Immediately after occlusion, retinal edema and intraretinal hemorrhage are produced as clinical indications of increased intravascular pressure. When multiple branch veins are occluded, retinal detachment is commonly produced. A vitritis often develops that may include fibrin

108

R. P. Danis and D. P. Bingaman

Figure 5-1. Immediate post-treatment fundus photograph after photodynamic thrombosis of a superior branch retinal vein of a pig eye.

Figure 5-2. (A) Fundus photograph of a porcine optic nerve head prior to BRVO. (B) Fundus photograph of the same porcine optic nerve head with neovascular tissue in the center of the optic nerve head.

5. NV in Models of Branch Retinal Vein Occlusion

109

membranes. After 3 to 4 weeks, retinal traction, schisis cavities, retinal venous collaterals, and retinal atrophy become apparent. NV invariably develops at the optic nerve head and is usually visible clinically with indirect ophthalmoscopy by 6 to 8 weeks. Clinical observations in miniature pigs up to 6 months after laser suggest that the neovascular development plateaus around 3 months after BVO. Spontaneous involution is not apparent, perhaps in part because NV in pigs often includes a heavy fibrous matrix which resembles, from the beginning, involutional NV in human diabetics.21 NV in the porcine BVO model may appear as preretinal fibrovascular tufts (as noted above), but also as “naked” vessels extending into the vitreous cavity with very little perivascular tissue (Figure 3). The endothelial cells of even the naked vessels appear to have relatively few fenestrations, which correlates with the lesser degree of fluorescein leakage compared to human preretinal NV. Additional differences between porcine BVO and human BVO histopathology include a pronounced inflammatory infiltrate into the vitreous in pigs. Macrophages are routinely found accompanying the neovascular tissue in humans as well, but the inflammatory infiltration around the NV and in the vitreous is more dramatic in the pig BVO model. The pig also often features full thickness retinal necrosis of the ischemic area, with inner retinal damage probably mediated by infarction from vascular occlusion and outer retinal damage perhaps due to the exudative retinal detachment that accompanies BVO in pigs. Because the porcine retina lacks a macula lutea (it possesses an area of increased cone photoreceptors toward the “area centralis”), the study of one of the most prevalent complications of BVO in humans, macular edema, is problematic in this model. Quantifying the NV response in the porcine BVO model led to the development of a 5-step ordinal grading scheme. This scheme initially employed clinical grading of the fundus using stereoscopic color fundus photographs combined with histopathological confirmation of the NV.22 Later projects that involved intravitreal injections, which caused vitreous or lens opacity, underscored the weakness of relying on clinical grading and photography, and masked grading was then performed at the time of gross histology under a dissecting microscope.16 The confirmation of the clinical or gross histopathological grade with light microscopic study was necessary because avascular membranes are commonly produced in this model and are difficult to distinguish from neovascular membranes. In addition, the inner layer of schisis cavities may resemble neovascular vitreous membranes. Fluorescein angiography was thought to be unreliable in the detection of NV, because unlike in humans, preretinal NV in the pig does not always leak profusely.

110

R. P. Danis and D. P. Bingaman

Figure 5-3. Light photomicrograph demonstrating neovascularization of the retina (hematoxylin and PAS, original mag 40X.)

In the case of pig BVO, the gross histopathological grade is confirmed by microscopy, since the clinical grade is sometimes considered unreliable.14,16,22,23 Unlike the situation in oxygen-induced retinopathy models where histopathological features of preretinal NV are easily quantified from microscopic sections or whole mounts, pig BVO-induced retinal NV grows as a tuft into the vitreous. At present, this presentation of NV has not been adequately quantified from whole mounts (which distort and compress the NV) or from cross sections (which can confirm the presence of NV, but are not suitable for quantitation of an irregular endophytic mass). Since the fibrous proliferation that accompanies the intravitreal NV can mimic fibrinous nonvascular vitreous membranes and schisis cavities from vitreous traction, histopathology has been necessary to

5. NV in Models of Branch Retinal Vein Occlusion

111

confirm the clinical grading. Attempts to standardize this model have been published from a single laboratory and have relied upon a grading scheme employing combined clinical and histopathological endpoints. An inherent weakness of grading schemes that employ clinical grading is the difficulty in adequately masking if there are clinical signs of treatment. For instance, if there is cataract or material in the vitreous of experimental eyes that is not present in the control eyes, masking is impossible. Nevertheless, despite these limitations, this model has been employed to quantitatively describe pharmacotherapeutic effects of a variety of drugs. Despite the complicated grading scheme and the need for detailed masked histopathological analysis, this methodology has been successfully employed in several investigations of pharmacological efficacy. The pig model was first used to demonstrate the inhibition of neovascularization achieved from a single intravitreal injection of 4 mg triamcinolone.23 Another study investigating the intravitreal effects of an antisense oligonucleotide against RAF-1 kinase used a nearly identical design, except that the control eye received an intravitreal injection of vehicle only.16 In this study, drug-treated eyes showed a reduction in NV. In both of these studies, masking of clinical exams (and photographs) was difficult due to lens opacities in the antisense oligonucleotide study and vitreous material in the triamcinolone study. Assessments relied upon masked gross and histopathological observations. A study using systemic administration of a protein kinase C beta inhibitor, ruboxistaurin (LY333531), employed two groups of 10 animals with bilateral BVO.16 Since treatment groups could not be inferred from ocular signs associated with local administration, and the treatment drug and placebo control were coded, masking was ideal in this study. The median NV score was reduced by 65% in the PKC group vs. controls. Both triamcinolone and ruboxistaurin are being evaluated in human trials. As a model in which to study pathophysiological aspects of intraocular NV, the pig BVO model has developed a strong track record. Pournaras and colleagues have employed this model to demonstrate a reduction in preretinal oxygen tension in the area of ischemia and restoration of normal oxygen tension after scatter laser treatment.20 They have also used this model to investigate the effects of vasoactive agents on tissue oxygenation and blood flow and to characterize biochemical changes post-BVO.20 Danis and colleagues have assayed the levels of some soluble growth factors in the vitreous over the course of development of NV and demonstrated that worsening of NV can be detected with exogenous human recombinant insulin-like growth factor-1 administration.24 Advantages of the pig BVO model over some other angiogenesis models in testing pharmacotherapeutic interventions include (1) robust production of

112

R. P. Danis and D. P. Bingaman

NV, (2) ability to easily examine the eyes clinically and photographically, (3) the relatively large eyes, and (4) a less expensive study cost vs. primate models. Moreover, small pigs are easy to handle and unlikely to injure personnel. The major disadvantages of this model include difficulty in handling the animals after several months of growth, since they gain weight at 1-2 kg per week, and the lack of a fovea. Also, the assessment of NV is semi-quantitative and difficult and has only been published from one laboratory. Masking of treatment groups is problematic if there are local signs of ocular treatment.

4.

PRIMATE BVO MODELS

Laser-induced branch retinal vein occlusion was explored in macaques (rhesus and cynomolgus) by several groups decades ago.18,25 Hayreh and colleagues reliably produced iris NV in cynomolgus monkeys after occlusion of 3 of the 4 major branch veins with an argon laser.26 Several variants of the original model have emerged in order to enhance the neovascular response. A number of laboratories have employed this model of iris NV to investigate the pathophysiology of ocular angiogenesis as well as the pharmacological efficacy of potential therapeutic agents. NV has been demonstrated by clinical examination, fluorescein angiographic documentation (upon which later quantitative assessment was developed), and histopathological documentation.7,27,28 Preretinal and optic nerve head NV is only sporadically reported; thus, this model has not been advocated as a model of posterior segment angiogenesis. Acute venous obstruction is sometimes assisted by pretreatment with intravenous fluorescein29 or use of krypton yellow laser instead of argon green.17 Acute vascular closure is accompanied by retinal edema and intraretinal hemorrhage. Reopening of retinal veins soon after occlusion is common. Retreatment with additional laser may be performed to obtain permanent obstruction. Vitreous hemorrhage may occur during treatment, complicating later clinical observation.18 Retinal edema tends to resolve quickly, as observed by clinical examination and angiography, but histopathological evidence of subtle macular edema may persist.30 The clinical course of macular edema in this model differs markedly from human patients; consequently, this model has not been utilized as a model of macular edema. Permanent vascular occlusion often results in capillary closure and venous collateral development. Iris NV may be observed within the first week following BVO and tends to be maximal between one and two weeks.7 Iris NV varies in severity, ranging from subtle vessels to the most severe stage involving hyphema and

5. NV in Models of Branch Retinal Vein Occlusion

113

ectropion uvea. Variants of this model include performing lensectomy and vitrectomy at the time of vascular occlusion, which increases the NV response and may lead to neovascular glaucoma in a small percentage of cases.26 The increased NV is likely related to the surgical trauma in addition to the removal of the physiological barriers to the diffusion of angiogenic products into the anterior chamber. Another development has been to pass a full-thickness silk suture through the cornea to produce chronic aqueous leakage and hypotony.31 The resulting increase in NV is likely related to increased angiogenic products produced by the inflammation from the penetrating trauma and the serum products from the chronic vascular exudation produced in hypotonous globes. Given the variety of production methods utilized by different laboratories, the incidence of iris NV production ranges up to 100% in some laboratories. Histopathologically, new vessels may be observed on the anterior surface of the iris as seen in human disease. With more profuse proliferation, fibrous tissue also develops with the vessels to produce a neovascular membrane. This membrane may eventually contract and produce ectropion in some cases.27 If the proliferation extends into the trabecular meshwork and seals off the iridocorneal angle (termed peripheral anterior synechiae), neovascular glaucoma may result. Standardized assessments of iris NV in the primate model usually rely on fluorescein angiographic grading. Briefly, the angiographic extent of NV is categorized (sometimes with standard photographs for reference) by the density and extent of leakage from vascular abnormalities, with ectropion uvea and/or hyphema representing the most severe endpoint (see table from Miller et al., 1993).32 Use of this model has been employed extensively to investigate pharmacotherapeutic agents31 as well as to analyze elements of the angiogenesis cascade, particularly in regard to soluble growth factors such as VEGF. Notably, intravitreal injection of exogenous VEGF produces iris NV in monkey eyes in the absence of ischemia. Moreover, elevated VEGF levels have been documented in eyes with BVO, and inhibition of VEGF in eyes with BVO inhibits iris NV.17,31-33 Advantages of the primate BVO models include their wide acceptance and application, relatively easy production, large eyes with ease of surgical intervention and clinical evaluation, and relatively standardized assessment. Use of primates for preclinical testing is also commonly performed prior to clinical trial development. Disadvantages include the high cost of purchase and maintenance of large animals, the potential for transmission of communicable diseases and injuries, and lack of posterior segment NV, which would be needed to mimic the more common human diseases of interest.

114

5.

R. P. Danis and D. P. Bingaman

MODELS OF OCULAR NEOVASCULARIZATION FROM BVO IN OTHER ANIMALS

Many groups have described angiogenesis in the setting of BVO in species other than pigs and monkeys. This effort is likely inspired by the desire to avoid the expense and complications of working with the larger animal models. Iris NV from BVO has been produced in cats with a great deal of effort. Steffanson and colleagues produced a high proportion of iris NV in cats after surgical cautery and transaction of retinal veins followed by retinal detachment.34 Hjelmeland et al. also produced iris NV and ectropion uvea with a surgical technique employing lensectomy, vitrectomy, and retinal venous cautery and transaction.8 Preretinal or optic nerve head NV was not described in either model. Because of the technical difficulty and need for surgery to produce iris NV in cats, it is unlikely that this model will be widely employed for pharmacotherapeutic trials of angioinhibitors. Preretinal NV after venous occlusion in rats has been described by several groups. Saito et al. demonstrated convincing preretinal NV in pigmented rats after occlusion of all retinal veins with an argon blue-green laser and intravenous fluorescein pretreatment.35 This model featured extensive exudative retinal detachment and macrophage infiltration (also noted in the pig model) and resulted in identifiable NV in 70% of animals. Other groups have described preretinal NV due to BVO using argon green laser and Rose Bengal in rats, but based on angiographic interpretation without presenting histopathological data.11,36,37 To our knowledge, only one group has used this methodology for pharmacotherapeutic assessment with histopathology.38 The technique is not technically difficult, and the animals are inexpensive and easily maintained. Further work with this model appears in order.

6.

SUMMARY

Ischemia-induced ocular NV due to BVO in primates is a fairly standardized, reproducible model based on clinical and angiographic grading of iris NV. This model has been employed by many investigators to study the pathogenesis of ischemia-induced NV and potential therapeutic strategies for human use. Disadvantages of the primate model include the expense and difficulty of working with primates and the relatively obscure role of iris NV in human retinal disease. Preretinal NV can be reliably produced by BVO in pigs and more closely mimics the manifestations of ischemic retinal disease observed in humans. Quantification of NV in the pig model is difficult, but

5. NV in Models of Branch Retinal Vein Occlusion

115

has been applied with success during pathophysiological and pharmacotherapeutic investigations. Ocular NV due to BVO in other species has been described but has not been developed into standardized models or routinely employed in research.

REFERENCES 1. L. M. Aiello and J. Cavallerano, Diabetic retinopathy, Curr. Ther. Endocrinol. Metab. 5, 436-446 (1994). 2. D. H. Orth and A. Patz, Retinal branch vein occlusion, Surv. Ophthalmol. 22 (6), 357-376 (1978). 3. D. Finkelstein, Retinal branch vein occlusion, In: S.J. Ryan (Ed.) Retina, 2 (pp. 1387-1392). St. Louis: Mosby (1994). 4. J. Clarkson, Central retinal vein occlusion, In: S.J. Ryan (Ed.) Retina, 2 (p. 1382). St. Louis: Mosby (1994). 5. A. M. Hamilton, E. M. Kohner, D. Rosen, A. C. Bird, and C. T. Dollery, Experimental retinal branch vein occlusion in rhesus monkeys. I. Clinical appearances, Br. J. Ophthalmol. 63 (6), 377-387 (1979). 6. E. M. Kohner, C. T. Dollery, M. Shakib, P. Henkind, J. W. Paterson, L. N. De Oliveira, and C. J. Bulpitt, Experimental retinal branch vein occlusion, Am. J. Ophthalmol. 69 (5), 778-825 (1970). 7. S. S. Hayreh and G. F. Lata, Ocular neovascularization. Experimental animal model and studies on angiogenic factor(s), Int. Ophthalmol. 9 (2-3), 109-120 (1986). 8. L. M. Hjelmeland, M. W. Stewart, J. Li, C. A. Toth, M. S. Burns, and M. B. Landers, 3rd, An experimental model of ectropion uveae and iris neovascularization in the cat, Invest. Ophthalmol. Vis. Sci. 33 (5), 1796-1803 (1992). 9. F. P. Campbell, Retinal vein occlusion; an experimental study, Arch. Ophthalmol. 65, 2-10 (1961). 10. E. Okun and E. M. Collins, Histopathology of experimental photocoagulation in the dog eye. Iii. Microaneurysmlike formations following branch vein occlusion, Am. J. Ophthalmol. 56, 40-45 (1963). 11. W. Shen, S. He, S. Han, and Z. Ma, Preretinal neovascularisation induced by photodynamic venous thrombosis in pigmented rat, Aust. N Z J. Ophthalmol. 24 (2 Suppl), 50-52 (1996). 12. B. Becker and L. T. Post, Jr., Retinal vein occlusion. Clinical and experimental observations, Am. J. Ophthalmol. 34 (5:1), 677-686 (1951). 13. C. J. Pournaras, M. Tsacopoulos, K. Strommer, N. Gilodi, and P. M. Leuenberger, Experimental retinal branch vein occlusion in miniature pigs induces local tissue hypoxia and vasoproliferative microangiopathy, Ophthalmology 97 (10), 1321-1328 (1990). 14. R. P. Danis, Y. Yang, S. J. Massicotte, and H. C. Boldt, Preretinal and optic nerve head neovascularization induced by photodynamic venous thrombosis in domestic pigs, Arch. Ophthalmol. 111 (4), 539-543 (1993). 15. G. Gao, Y. Li, J. Fant, C. E. Crosson, S. P. Becerra, and J. X. Ma, Difference in ischemic regulation of vascular endothelial growth factor and pigment epithelium--derived factor in brown norway and sprague dawley rats contributing to different susceptibilities to retinal neovascularization, Diabetes 51 (4), 1218-1225 (2002).

116

R. P. Danis and D. P. Bingaman

16. R. Danis, M. Criswell, F. Orge, E. Wancewicz, K. Stecker, S. Henry, and B. Monia, Intravitreous anti-raf-1 kinase antisense oligonucleotide as an angioinhibitory agent in porcine preretinal neovascularization, Curr. Eye. Res. 26 (1), 45-54 (2003). 17. A. P. Adamis, D. T. Shima, M. J. Tolentino, E. S. Gragoudas, N. Ferrara, J. Folkman, P. A. D’amore, and J. W. Miller, Inhibition of vascular endothelial growth factor prevents retinal ischemia-associated iris neovascularization in a nonhuman primate, Arch. Ophthalmol. 114 (1), 66-71 (1996). 18. P. S. Virdi and S. S. Hayreh, Ocular neovascularization with retinal vascular occlusion. I. Association with experimental retinal vein occlusion, Arch. Ophthalmol. 100 (2), 331-341 (1982). 19. J. W. Miller, A. P. Adamis, D. T. Shima, P. A. D’amore, R. S. Moulton, M. S. O’reilly, J. Folkman, H. F. Dvorak, L. F. Brown, B. Berse, and Et Al., Vascular endothelial growth factor/vascular permeability factor is temporally and spatially correlated with ocular angiogenesis in a primate model, Am. J. Pathol. 145 (3), 574-584 (1994). 20. C. J. Pournaras, Retinal oxygen distribution. Its role in the physiopathology of vasoproliferative microangiopathies, Retina 15 (4), 332-347 (1995). 21. B. D. Danis Rp, Yang Y,, The long-term natural history of preretinal and optic nerve head neovascularization in miniature pigs, Veterinary & Comparative Ophthalmology 6 (4), 237-242 (1996). 22. R. P. Danis, D. P. Bingaman, M. Jirousek, and Y. Yang, Inhibition of intraocular neovascularization caused by retinal ischemia in pigs by pkcbeta inhibition with ly333531, Invest. Ophthalmol. Vis. Sci. 39 (1), 171-179 (1998). 23. R. P. Danis, D. P. Bingaman, Y. Yang, and B. Ladd, Inhibition of preretinal and optic nerve head neovascularization in pigs by intravitreal triamcinolone acetonide, Ophthalmology 103 (12), 2099-2104 (1996). 24. D. P. Bingaman, D. R.P., W. H. Lee, M. B. Grant, and W. S. Warren, Increased vegf levels precede igf1 system activation in the pig model of ocular angiogenesis induced via retinal ischemia. , Association for Research in Vision and Ophthalmology Annual Meeting (Fort Lauderdale, FL (1998). 25. A. M. Hamilton, J. Marshall, E. M. Kohner, and J. A. Bowbyes, Retinal new vessel formation following experimental vein occlusion, Exp. Eye. Res. 20 (6), 493-497 (1975). 26. A. J. Packer, X. Q. Gu, E. G. Servais, and S. S. Hayreh, Primate model of neovascular glaucoma, Int. Ophthalmol. 9 (2-3), 121-127 (1986). 27. C. P. Juarez, M. O. Tso, W. A. Van Heuven, M. S. Hayreh, and S. S. Hayreh, Experimental retinal vascular occlusion. Iii. An ultrastructural study of simultaneous occlusion of central retinal vein and artery, Int. Ophthalmol. 9 (2-3), 89-101 (1986). 28. C. P. Juarez, M. O. Tso, W. A. Van Heuven, M. S. Hayreh, and S. S. Hayreh, Experimental retinal vascular occlusion. Ii. A clinico-pathologic correlative study of simultaneous occlusion of central retinal vein and artery, Int. Ophthalmol. 9 (2-3), 77-87 (1986). 29. D. J. Hockley, R. C. Tripathi, and N. Ashton, Experimental retinal branch vein occlusion in the monkey. Histopathological and ultrastructural studies, Trans. Ophthalmol. Soc. UK 96 (2), 202-209 (1976). 30. I. H. Wallow, R. P. Danis, C. Bindley, and M. Neider, Cystoid macular degeneration in experimental branch retinal vein occlusion, Ophthalmology 95 (10), 1371-1379 (1988). 31. M. Genaidy, A. A. Kazi, G. A. Peyman, E. Passos-Machado, H. G. Farahat, J. I. Williams, K. J. Holroyd, and D. A. Blake, Effect of squalamine on iris neovascularization in monkeys, Retina 22 (6), 772-778 (2002). 32. J. W. Miller, W. G. Stinson, and J. Folkman, Regression of experimental iris neovascularization with systemic alpha-interferon, Ophthalmology 100 (1), 9-14 (1993).

5. NV in Models of Branch Retinal Vein Occlusion

117

33. M. J. Tolentino, J. W. Miller, E. S. Gragoudas, F. A. Jakobiec, E. Flynn, K. Chatzistefanou, N. Ferrara, and A. P. Adamis, Intravitreous injections of vascular endothelial growth factor produce retinal ischemia and microangiopathy in an adult primate, Ophthalmology 103 (11), 1820-1828 (1996). 34. E. Stefansson, M. B. Landers, 3rd, M. L. Wolbarsht, and G. K. Klintworth, Neovascularization of the iris: An experimental model in cats, Invest. Ophthalmol. Vis. Sci. 25 (3), 361-364 (1984). 35. Y. Saito, L. Park, S. A. Skolik, D. V. Alfaro, N. A. Chaudhry, C. J. Barnstable, and P. E. Liggett, Experimental preretinal neovascularization by laser-induced venous thrombosis in rats, Curr. Eye. Res. 16 (1), 26-33 (1997). 36. D. I. Ham, K. Chang, and H. Chung, Preretinal neovascularization induced by experimental retinal vein occlusion in albino rats, Korean J. Ophthalmol. 11 (1), 60-64 (1997). 37. S. G. Kang, H. Chung, and J. Y. Hyon, Experimental preretinal neovascularization by laser-induced thrombosis in albino rats, Korean J. Ophthalmol. 13 (2), 65-70 (1999). 38. C. C. Lai, W. C. Wu, S. L. Chen, M. H. Sun, X. Xiao, L. Ma, K. K. Lin, and Y. P. Tsao, Recombinant adeno-associated virus vector expressing angiostatin inhibits preretinal neovascularization in adult rats, Ophthalmic Res. 37 (1), 50-56 (2005).

MOLECULAR CHARACTERIZATION

Chapter 6 VASCULOGENESIS AND ANGIOGENESIS IN FORMATION OF THE HUMAN RETINAL VASCULATURE Cell-Cell Interactions and Molecular Cues Tailoi Chan-Ling Bosch Institute, Department of Anatomy, University of Sydney, Sydney, Australia

Abstract:

1.

Development of the human retinal vasculature takes place via two distinct cellular processes: angiogenesis and vasculogenesis. These processes are triggered by distinct molecular cues and proceed by distinct biological pathways, which offers the attractive possibility of using distinct inhibitory and stimulatory methods for intervention in retinal diseases. This chapter reviews what is known about human embryonic retinal development, focusing on the molecular, spatial, and temporal differences between vasculogenesis and angiogenesis.

INTRODUCTION

During early embryonic development, the human retina transforms from a single layer of undifferentiated neuro-epithelial cells to an organized stratified structure. Concomitant with the maturation of the neuronal elements, the retina’s vasculature develops to form an elaborate vascular tree that is well matched to the metabolic needs of the tissue. This formation of the intra-retinal vessels takes place via two distinct cellular processes under distinct molecular cues.1 Formation of the primordial superficial vessels of the central two-thirds of the human retina takes place via the process of vasculogenesis, the de novo formation of primitive vessels by differentiation from vascular precursor cells. Formation of the remaining retinal vessels takes place via angiogenesis, the process of new vessel formation by budding or intussusceptive growth from existing blood vessels. Vasculogenesis appears to take place independently of hypoxia-induced 119 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 119–138. © Springer Science+Business Media B.V. 2008

120

T. Chan-Ling

vascular endothelial growth factor (VEGF165).1 In contrast, angiogenesis is mediated via hypoxia-induced expression of VEGF165 by retinal glia (Müller cells and astrocytes)2 and pericytes.3 Retinal vascularization in human, primate, cat, dog, and rat supports the conclusion that both vasculogenesis and angiogenesis are involved. In contrast, compelling evidence for the existence of vascular precursor cells is not available in the mouse retina, though Ash, McLeod, and Lutty4 have preliminary data suggesting the presence of ADPase+ vascular precursor cells in advance of the forming vasculature in postnatal day 3 (P3) mice. This apparent species difference between the mechanism by which retinal vessels form in humans and mice necessitates caution when extrapolating directly from mouse studies to human application. This is particularly of relevance in the development of therapies for retinopathy of prematurity (ROP), age-related macular degeneration (ARMD), and diabetic retinopathy (DR), and could in part explain the failure of novel treatments, where successful pre-clinical trials would have predicted a more positive outcome. Failure to recognize key species differences could lead to ineffective clinical trials, worsened disease, or unexpected severe adverse events. Where the mouse model reproduces specific features of the histopathology and neurobiology of the human condition, its application is warranted and highly advantageous due to the availability of genetically modified animals and experimental reagents. However, mouse models apply only in the elucidation of the role of angiogenic processes and do not mimic fully human retinal and choroidal pathogenesis. There are limitations of various animal models of ROP, ARMD, and DR, but when used appropriately, animal models of various species continue to provide a crucial tool for improving the understanding and development of neovascularizing retinopathies. The marked species differences in the mechanism of retinal vascular formation reported to date point to the necessity to undertake studies on human tissues during normal development and in disease.

1.1

Three intra-retinal vascular plexuses in human retina: Superficial and deep vascular plexuses and radial peri-papillary capillaries (RPCs)

The human retina first appears as an undifferentiated neural epithelium early in embryonic development (Figure 1A). With further maturation, neuronal stratification produces an ordered multi-layered stratified structure. Concurrent with this neuronal maturation is the formation of three inherent retinal vascular plexuses (Figure 1B). A superficial vascular plexus is located in the ganglion cell and nerve fiber layers, and a second deep plexus is located at the junction of the inner

6. Formation of the Human Retinal Vasculature

121

nuclear and outer plexiform layers (Figure 1B). The superficial plexus contains arterioles, venules, capillaries, and post-capillary venules, while the deep vascular bed consists pre-dominantly of capillary-sized vessels. Both the superficial and deep retinal plexuses reach almost to the edge of the human retina, except for a small avascular rim, where the thinness of the human retina likely permits adequate retinal oxygenation via the choroidal vasculature.1,5 A third intraretinal plexus, the radial peri-papillary capillaries (RPCs), is also located in the nerve fiber layer in a small rim surrounding the optic disc (see Figure 10 D-F in 6; Figure 6 E-F in 1). These RPCs are located superficially in a small region surrounding the optic nerve head where the nerve fiber bundles are thickest prior to exiting the retina. Their superficial location and the fact that these vessels lack smooth muscle actin ensheathment (personal observation) could make the nerve fiber bundles nourished by these vessels uniquely prone to ischemic damage as a consequence of reduced blood flow during periods of raised intraocular pressure.

1.2

Two distinct mechanisms in the formation of the human retinal vasculature

Blood vessels in the human retina form by vasculogenesis, angiogenesis1,7-9 and intussusception.10 The term vasculogenesis describes the de novo formation of vessels from vascular precursor cells (VPCs), also called mesenchymal precursor cells,1,5,7 and angioblasts.9 Single spindle-shaped CD39+, Nissl stained VPCs (Figure 1C-F) stream in superficially from the optic nerve head into the avascular retina and differentiate at the location of future vessels, coalesce into cords (Figure 1F-H), differentiate into endothelial cells, and ultimately form patent vessels (Figure 1I-J and 1,6-9). These VPCs precede the leading edge of patent vessels by more than 1 mm (Figure 2A and 7). They differentiate to form a primordial vascular bed centered on the optic disk (Figure 1I-J). During human retinal vascular development, superficial inner retinal vessels form by vasculogenesis, starting at the optic nerve and developing along a gradient from the posterior to the anterior retina. Vasculogenesis is only responsible for formation of the primordial vessels that span the inner two-thirds of the superficial retinal plexus (Figure 2A).

122

T. Chan-Ling

6. Formation of the Human Retinal Vasculature

123

The remaining retinal vessels form via angiogenesis, which produces increasing capillary density in the central retina, formation of the peripheral blood vessels of the superficial retinal plexus, formation of the deep vascular plexus, and formation of the RPCs. The term angiogenesis describes a different process of blood vessel formation in which proliferating endothelial cells from pre-existing blood vessels extend the vascular network. Angiogenesis can take place via budding or intussusception. Budding angiogenesis involves filopodial extension by proliferating and migrating vascular endothelial cells (Figure 2C-D and 1,11). The term intussusception describes the remodeling and expansion of new vessels by the insertion of interstitial tissue columns into the lumen of pre-existing vessels.10

2.

TIMING AND TOPOGRAPHY OF HUMAN RETINAL VASCULAR FORMATION

The first event in retinal vascularization apparent before 14 weeks’ gestation (WG) is the appearance of large numbers of CD39+, Nissl stained VPCs centered around the optic nerve head (Figure 1C-H and 1,7). These VPCs are concentrated in 4 lobes of the future major artery-vein pairs of the human

Figure 6-1. (A-B) Developing human retina from an avascular undifferentiated neuroepithelium (A) to fully stratified adult retina (B). Two plexuses of retinal vessels are apparent. The superficial plexus is located predominantly in the ganglion cell and nerve fiber layers showing a range of vessel calibers from arterioles and venules to capillaries. The second deeper plexus is located at the junction of the inner nuclear layer and the outer plexiform layer with predominantly capillary-sized vessels. The formation and maturation of the human retinal vasculature is concurrent with neuronal differentiation and maturation. (CD) Low- and high-magnification views of a Nissl-stained human retinal whole mount at 14 to 15 weeks’ gestation (WG). C shows a region that is immediately adjacent to the optic nerve head (lower left-hand corner). Large numbers of spindle-shaped cells (arrowheads in C), which are interspersed among other somas, stream in superficially from the optic nerve head. In D, spindle-shaped, presumably vascular precursor, cells join to form vascular cords of cells (arrowheads at top right). (E) 16.5-WG human retina labeled with CD39. Spindleshaped CD39+ vascular precursor cells streaming in superficially from the optic nerve head. (F) Appearance of the edge of vasculature using Nissl-stained preparation from an 18-WG specimen. Nissl staining showed spindle-shaped cells in advance of the vasculature. (G-H) CD39+/CD34-/+ solid vascular cords at the leading edge of vessel formation at 17 WG. (I-J) Low- and high-magnification views of the first primordial vascular arcades labeled with CD34, evident in the region of the optic nerve head at 15 WG. These vascular arcades show the four-lobed topography of formation that is indicative of the future superior and inferior (temporal and nasal) artery vein pairs. Morphologically, they are straight and lack significant capillary density. (Modified from 1)

124

T. Chan-Ling

retina. Figure 2A shows the distribution of these cells at 14-15, 18 and 21 WG (modified from 1). Formation of the patent superficial vascular plexus begins by 14 to 15 WG. These primordial vessels are centered on the optic disc and show a four-lobed topography (Figure 2A). In the following weeks, the inner vascular plexus extends peripherally, curving around the location of the incipient fovea (Figure 2A). By 32 WG, the inner plexus reaches its outer limits, leaving a narrow rim of avascular tissue at the periphery of the retina. In contrast, the formation of the outer vascular plexus begins in the perifoveal region between 25 and 26 WG, coincident with the peak period of eye opening, when the visually evoked potential, indicative of a functional visual pathway and photoreceptor activity, is first detectable in the human infant.12 Formation of the deeper vascular plexus subsequently spreads with an elongated topography along the horizontal meridian (Figure 2B) and is centered around the fovea, rather than the optic disc, thus mimicking the topography of photoreceptor maturation.13 The timing and topography of formation of the deeper plexus supports the conclusion that angiogenesis is driven by increasing metabolic demand as a result of neuronal maturation. The outer plexus forms via extension of capillary-sized buds from the existing superficial vessels. This deeper plexus reaches the edge of the human retina by birth.

2.1

Angiogenesis results in increasing capillary density followed by vascular regression and remodeling

In addition to its contribution to the spread of vessels peripherally, angiogenesis also is responsible for increasing the vascular density of the primordial plexus formed by vasculogenesis. Initially, capillary networks are rare (Figure 1I-J). However, by 18 WG, regions of active sprouting begin to give rise to substantial capillary networks within the existing vascular tree (Figure 2C). As the retina increases in area and the radial vessels spread peripherally, the distance between these vessels becomes greater, and it is in these avascular spaces that sprouting angiogenesis is most pronounced. Filopodia extend, establish contact with other filopodia or vessels, and subsequently dilate to form vascular segments. Moreover, sprouting is evident even from the edges of larger preformed vessels and is especially marked near and along veins. By 21 WG, exuberant immature capillary plexuses are apparent throughout the vascular tree. Thus, angiogenesis augments the initial radial vessels by increasing capillary density.

6. Formation of the Human Retinal Vasculature

125

Figure 6-2. (A) The distribution of the spindle cells shown at 14.5, 18, and 21 WG. The stippled regions show the distribution of the Nissl stained spindle cells at each age; the white regions show the areas with vascular cords. Both the spindle-shaped vascular precursors and the vascular cords were more extended in the temporal and superior directions than in the nasal direction. With increasing maturity, the outer limit of the vascular cords expanded markedly, whereas that of the vascular precursor cells did not. At 21 WG, no spindle cells were evident in the retina. It is clear from these maps that the area formed by vasculogenesis is not circular in the developing human retina. The X indicates the location of the incipient fovea. (B) Maps of the outer limits of the inner and outer vascular plexuses as well as that of the RPCs at various times during development of the human fetal retina. (C-D) CD34-stained human retina from an 18 WG specimen. Red blood cells were apparent in patent vessels utilizing Normaski optics. Angiogenic budding as evidenced by filopodial extensions was abundant. (Modified from 1)

126

T. Chan-Ling

A remarkable feature of angiogenesis is the exuberance of the initial vessels during formation of the superficial human retinal plexus. Our earlier work in the rat retina has shown that this significant overproduction with subsequent vascular remodeling, or pruning, involves a combination of apoptosis of vascular endothelial cells and withdrawal of endothelial cells from excess vascular segments into neighboring vascular segments, where they are utilized to form new vessel segments.14

2.2

Formation of the Perifoveal and Temporal Raphe Vessels by Angiogenesis

The incipient fovea is avascular at 25 WG (see Figure 6A-B in 1). The avascular zone is oval in shape, with a diameter of 500 to 600 mm. Because no spindle cells were observed in the region of the temporal raphe or the perifoveal region of the human retina, these areas must be vascularized by angiogenesis alone.

2.3

Formation of the Deep Vascular Plexus by Angiogenesis

Capillary sprouting from the inner vascular plexus is first evident at the fovea between 25 and 26 WG. Capillary-sized buds descend into the inner nuclear and outer plexiform layers, giving rise to small vascular segments in a deeper plane (Figure 3B-C). With maturation, a confluent outer plexus becomes apparent.

2.4

Formation of RPCs by Angiogenesis

Fine RPCs were evident in the nerve fiber layer, extending from the inner vasculature, from 21 WG. RPCs in the region of the optic nerve head of 25and 26-WG retinas (see Figures 6E-F in 1) were located superficially in the nerve fiber layer and extended radially from the optic nerve head. The timing and extent of their formation suggest that their formation is driven by hypoxia-mediated VEGF resulting from the need to satisfy the metabolic requirements of the thick nerve fiber layer that surrounds the optic nerve head.

6. Formation of the Human Retinal Vasculature

3.

127

TWO DISTINCT PATTERNS OF MITOTIC ACTIVITY ASSOCIATED WITH RETINAL VASCULOGENESIS AND ANGIOGENESIS IN THE KITTEN AND HUMAN RETINA

Bromodeoxy-uridine (BrdU) is an analog of thymidine and can be used to identify proliferating cells. We have previously combined BrdU with endothelial cell-specific markers to demonstrate the distinct pattern of cell division associated with vasculogenesis and angiogenesis in the kitten retina.15,16 Vasculogenesis and angiogenesis are associated with markedly different patterns of mitotic activity. The formation of vessels by vasculogenesis is preceded by low mitotic activity among the vascular precursor cells some millimeters peripheral to the edge of the patent vessels (Figure 3A). More centrally, mitotic activity increases significantly during formation of patent vessels (Figure 3A), continues at a lower level as large vessels differentiate from the initial capillary plexus, and then falls to zero in the adult cat. Central to this leading edge of vessel formation, dividing vascular endothelial cells were frequently evident in the vascular tree as they remodeled and selected major channels. Possibly as a result of the higher oxygen tension in arteries than in veins, the density of mitotically active vascular cells was markedly higher in veins than in arteries.16 In contrast, the formation of vessels by angiogenesis is not preceded by migration and division of vascular precursor cells. During angiogenesis, division occurs only in the endothelial cells close to, but not at the tips of, growing vessels (Figure 3B-C and 15,16). A third vascular plexus that we have previously described as being formed by vascular budding or angiogenesis comprises the RPCs. These vessels are fine capillaries that radiate out from the region of the optic nerve head in the nerve fiber layer. Their formation is likely the result of the fact that the nerve fiber layer is thickest in this region, so that its metabolic requirements are not adequately satisfied by the superficial retinal plexus. Mitotic activity associated with the formation of RPCs is similar to that typical of angiogenesis (Figure 3D), with one or two mitotic nuclei evident close to the tip of the forming capillaries. We have recently developed a new in vitro technique for identifying vascular proliferation in human retina and choroid. Utilizing triple-label immunohistochemistry for CD39/CD34/BrdU, it was possible to demonstrate that both CD39+/BrdU+ vascular precursor cells and CD34+/BrdU+ vascular endothelial cells proliferate in situ in the human retina (Figure 4). A comparison of 4E and 4F shows that a higher proportion of CD30+ cells are proliferative than CD34+ cells.

128

T. Chan-Ling

Figure 6-3. (A-D) BrdU/Griffonia isolectin B4 double label histochemistry to visualize proliferating vascular endothelial cells in kitten retina. New vessel formation at the outer vascular plexus occurs by a budding process. (A) Small numbers of proliferating vascular precursor cells are evident preceding the leading edge of vessel formation, toward the right of the field of view. Large numbers of mitotic endothelial cells are evident within the leading edge of patent vessel formation. (B-C) Mitotic cells are evident close to the tips of angiogenic buds, but not at the tips of angiogenic buds (15, 16). Fields of view show new vessel growth in the outer vascular plexus of a P8 retina. Note the clear absence of mitotic vascular precursor cells in the region preceding the growing vessel bud. (D) Low numbers of mitotic cells associated with the formation of RPCs of a P28 kitten. (E-F) Desmin and Griffonia simplicifolia isolectin B4 double-labeled P8 rat retina showing desmin+ mural precursor cells are present concurrently with the presence of newly formed vascular segments. (Modified from 33)

6. Formation of the Human Retinal Vasculature

129

Figure 6-4. Vascular precursor cells and vascular endothelial cells in human retina proliferate in situ. (A-G) An 18-WG human retinal whole mount labeled with CD39 (Cy5), CD34 (Alexa 488) and BrdU (Cy3). Panel D shows overlapping expression between CD39+ and CD34+ cells as the cells mature along the vascular endothelial cell lineage. (E-F) Significant proportions of both CD39+ and CD34+ cells were also BrdU+. A comparison between E and F shows that a greater proportion of CD39+ vascular precursor cells are proliferative than the CD34+ vascular endothelial cells. (G) Triple labeling of 18-WG human retina showing CD39+/CD34+/BrdU+ vascular precursor cells (arrowheads), CD39-/CD34-/BrdU+ soma, likely an astrocyte or neuron (arrow), and nonproliferative vascular endothelial cells lining a lumen (double arrows). Scale bar = 50µm

130

3.1

T. Chan-Ling

Formation of retinal vessels via vasculogenesis is independent of VEGF165

A review of the literature and our own observations in the developing human retina led us to conclude that formation of retinal vessels via vasculogenesis is independent of metabolic demand and hypoxia-induced VEGF expression.1 Evidence for this conclusion includes the observations that (1) substantial vascularization in the human retina occurs prior to detection of VEGF mRNA,17 (2) vasculogenesis is well established by 14 to 15 WG, before the differentiation of most retinal neurons, and (3) the topography of formation of vessels by vasculogenesis does not correlate at all with the topography of neuronal density and maturation. Formation of the outer plexus begins around the incipient fovea between 25 and 26 WG, coincident with the peak period of eye opening and the first appearance of the visually evoked potential, indicative of a functional visual pathway and photoreceptor activity.1 Further, comparative analysis of retinal vascularization in other species has shown that VEGF expression, tissue oxygen levels, and vascularization are not always correlated.18 The guinea pig retina is virtually anoxic and yet remains avascular,19 whereas overexpression of VEGF in the avian retina did not induce vascularization.20 Further evidence of the independence of vasculogenesis from VEGF is provided by VEGF knockout mice. In these animals, in which not only paracrine but also autocrine VEGF production is lost, vessels still form by vasculogenesis but are highly abnormal.21 Reduced VEGF expression in mice heterozygous for the VEGF null mutation is associated with the formation of vessels in the forebrain mesenchyme but not in the neuroepithelium.22 Given that the formation of vessels in the forebrain mesenchyme is thought to occur by vasculogenesis, whereas that within the neuroepithelium is thought to take place by angiogenesis, these observations provide further evidence that vasculogenesis is not dependent on hypoxiainduced VEGF expression.

3.2

Retinal angiogenesis in human retina is mediated by hypoxia-induced VEGF165 expression by astrocytes, Müller cells and pericytes

In marked contrast to vasculogenesis, the timing and topography of angiogenesis in the human retina supports the conclusion that angiogenesis is induced by “physiological hypoxia,” a transient but physiological level of hypoxia induced by the increasing activity of retinal neurons.15,23 The formation of retinal vessels via angiogenesis is mediated by hypoxia-induced VEGF expression by astrocytes, Müller cells2,17 and pericytes.3 In retinal

6. Formation of the Human Retinal Vasculature

131

angiogenesis, hypoxia is the initial stimulus that causes the upregulation of growth factors, integrins and proteinases that results in endothelial cell proliferation and migration, essential steps in new vessel formation.24

4.

CELL-CELL INTERACTIONS IN THE FORMATION OF THE RETINAL VASCULATURE

The retina consists of three main cellular elements: the neurons, macroglia (including astrocytes and Müller cells), and the vasculature (including vascular endothelial cells, pericytes, and smooth muscle cells). Immune and phagocytic cells including retinal microglia, macrophages, and perivascular antigen presenting cells complete the cellular milieu. During the formation of the human retinal vasculature, these cellular elements interact in complex ways, resulting in the formation and then the remodeling of the vasculature to produce a vascular tree that is well matched to the metabolic needs of the tissue. Cells of the astrocytic and vascular lineage interact closely during formation of the mammalian retinal vasculature (Figure 5). In a comparative study, Schnitzer has shown that occurrence of astrocytes in mammalian retinae coincides with the presence of blood vessels.25 We have shown the close association of retinal astrocytes with the forming superficial vascular plexus in the developing kitten, rat and human retinae.1,7,26,27 More recent studies have shown that the superficial vascular network in the neonatal mouse retina forms according to a pre-existing astrocytic template, and both the superficial and deep vascular layers use R-cadherin cell adhesion molecules as guidance cues.28 In the human fetal retina, Pax-2 expression is restricted to cells of the astrocytic lineage.29 Pax-2 is a member of the Pax family of transcription factors. Each member of the Pax family is expressed in a spatially and temporally restricted manner, which suggests that these proteins contribute to the control of tissue morphogenesis and pattern formation. Pax-2+/GFAPastrocyte precursor cells (APCs) are first evident at the optic nerve head at 12 WG, preceding the appearance of Pax-2+/GFAP+ astrocytes. These immature astrocytes are seen immediately peripheral to the leading edge of vessel formation (approximately 20-40 μm) and at 18 WG loosely ensheath the newly formed vessels.7 With maturation, Pax2+/GFAP+ astrocytes extend toward the periphery, reaching the edge of the retina around 26 WG.29 The location of the astrocytes and APCs just ahead of the leading edge of vessel formation places them in an ideal position to mediate the angiogenic response to “physiological hypoxia” via upregulation of VEGF165 expression.17

132

T. Chan-Ling

Retinal vessels have blood-retina barrier (BRB) properties as soon as they become patent. Astrocytes have been shown to be responsible for inducing the blood-brain barrier properties in vascular endothelial cells30 and thus are thought to induce the BRB in the inner plexus. The processes of the Müller cells (the radial glia of the retina) ensheath the vessels of the outer plexus and are also capable of induction of the BRB.31 During normal human retinal vascularization, significant overproduction of vascular segments occurs, and the excess segments regress with maturation of the vasculature. Our earlier work has shown that endothelial cell apoptosis and macrophages do not initiate vessel retraction, but rather contribute to the removal of excess vascular endothelial cells throughout the immature retinal vasculature. Furthermore, our observations suggest that vessel retraction is mediated by endothelial cell migration and that endothelial cells derived from retracting vascular segments are re-deployed in the formation of new vessels.14 Mural cells (pericytes and smooth muscle cells) are an intrinsic part of blood vessel walls with broad functional activities, including blood flow regulation, and have been implicated in vessel stabilization.32 These cells are derived from a mural precursor cell which gives rise to pericytes on capillaries and smooth muscle cells on larger vessels.33 Immature mural cells, the ensheathing mural precursor cells, in cats, rats and mice, envelop newly formed vessels and have recently been shown to express VEGF165.3 The presence of these immature mural cells does not prevent vessel regression during normal development33 and hyperoxia-induced vessel regression.34 Because mature vasculatures with mature mural cells are considered stable, this suggests that mural cell maturation may be necessary for resistance to VEGF165 withdrawal-induced vessel regression. Macrophages are also part of the cellular milieu during formation of retinal vessels.35 Although their function is unclear, they are capable of expressing VEGF165 in the mouse model of hyperoxia-induced retinopathy36 and in ARMD.37 The endothelium in turn influences the development of astrocytes and mural cells. Retinal endothelial cells express PDGF-beta, which induces recruitment and proliferation of mural cells.38 It has also been shown that vascular endothelial cells can induce astrocyte differentiation.39,40 In addition, contact between mesenchymal precursor cells and vascular endothelial cells leads to mural cell differentiation in vitro.41 Taken together, these studies show that the formation and maturation of the human retinal vasculature is (1) concurrent with neuronal, astrocyte, and mural cell

6. Formation of the Human Retinal Vasculature

133

134

T. Chan-Ling

differentiation and maturation, and (2) the result of complex cellular interactions in which the vasculature both takes its developmental cues from and also influences its cellular environment. During normal development, the retinal vasculature is remodeled, resulting in a vascular pattern that is well matched to the metabolic demands of the tissue. The retinal vasculature is under the constant influence of its environment, where neovascularization or regression is determined by the equilibrium between pro-angiogenic and anti-angiogenic signals, proteases42 and other molecular cues.

Figure 6-5. Astrocyte/endothelial relationship during formation of the superficial human retinal vasculature. (A–C) Human fetal retinal whole mounts triple labeled with Pax2/GFAP/CD34 at 14 WG. (A) At 14 WG, Pax2+(TR)/GFAP-(FITC) astrocyte precursor cells (APCs) extended in advance of the leading edge of CD34(FITC) blood vessels by a small but distinct margin. Arrows: some neonatal astrocytes that were just starting to express GFAP. (B) Representative field of view from mid-retina. Arrows: Pax2+/GFAP- APCs. Most cells in this area were Pax2+/GFAP+ immature astrocytes. CD34+ blood vessels (FITC) were also clearly evident. (C) A region near the optic nerve head (ONH) of a 14-WG fetus. Arrows: Pax2+/GFAP- APCs. Most of cells in this area were Pax2+/GFAP+ immature astrocytes. (D) Map of the 14-WG retina shown in A–C. Red somas: Pax2+/GFAP- APCs; yellow somas: Pax2+/GFAP+ immature astrocytes; green: CD34+ blood vessels. White boxes are the representative areas of the fields of view seen in A-C. (Modified from 7)(E-L) Pericyte/endothelial relationship during formation of the superficial rat retinal vasculature. (E) Representative fields of view of an E20 preparation of retinal vasculature, triple labeled with anti-SMA (green), anti-desmin (red), and GS lectin (blue). (Note: desmin was not detected). (F) Postnatal day (P) 0 rat retinal vasculature triple labeled with anti-NG2 (green), antidesmin (red), and GS lectin (blue) showing ensheathing mural precursor cells (MPCs) with desmin filaments on undifferentiated vessels. (G) Embryonic day (E) 21 retinal vasculature triple labeled with anti-NG2 (green), anti-desmin (red), and GS lectin (blue) showing tips of vessels extending peripherally (Note: desmin was not detected in this field). (H) Capillaries in the adult retina triple labeled with GS lectin (blue), anti-desmin (red), and anti-NG2 (green). Shows an adult quiescent pericyte. (I) P0 preparation of rat retina, triple labeled with antiSMA (green), anti-desmin (red), and GS lectin (blue) showing differentiating central radial vessels immediately adjacent to the optic nerve head (Note: desmin was not detected in this field). (J) SMC differentiation in the rat retina. Central radial arteriole double labeled at P7 with anti-SMA (green), anti-desmin (red), and GS lectin (blue) showing an immature central radial arteriolar smooth muscle cell (SMC). (K) Central radial rat arterioles double labeled at P13 with anti-SMA (green), showing a juvenile central radial arteriolar SMC. (L) Adult radial and primary arterioles in the central retina double labeled with anti-SMA (green) and antidesmin (red). The abbreviated marker names are colored to represent the respective fluorescent dye in each micrograph. (Modified from 7,33)

6. Formation of the Human Retinal Vasculature

5.

135

CONCLUSION

Observations to date support the conclusion that the formation of primordial vessels of the superficial plexus in the central human retina is mediated by vasculogenesis, whereas angiogenesis is responsible for increasing vascular density and peripheral vascularization in the superficial retinal plexus. In contrast, the vessels in the perifoveal region and the deeper retinal plexus and the radial peripapillary capillaries are formed by angiogenesis only. Our understanding of retinal vessel formation is based on many sources, including the seminal works of Michaelson, Ashton, Dollery, Weiter, D’Amore, Schlingemann, Lutty, Penn, Das, Smith, Friedlander, Gariano, and our own review and analyses. The fact that the human retina is vascularized through two distinct pathways with distinct molecular cues and cellular processes as highlighted in this review offers the attractive possibility of using distinct inhibitory and stimulatory methods for intervention. With a clear understanding of the cellular and molecular cues that drive normal retinal vascularization, we can gain clues to the mechanism underlying neovascularization. These insights could be of relevance to neovascularizing retinopathies of infancy and adulthood.

ACKNOWLEDGMENT This contribution would not have been possible without the generous assistance provided by Ruth-Ann Sterling, Suzanne Hughes, and Louise Baxter. This work was supported by grant #153789 and #402824 from the National Health and Medical Research Council of Australia and the Financial Markets Foundation for Children.

REFERENCES 1. S. Hughes, H. Yang, T. and Chan-Ling, Vascularization of the human fetal retina: roles of vasculogenesis and angiogenesis. Invest. Ophthalmol. Vis. Sci. 41 (5), 1217-1228, (2000). 2. J. Stone, A. Itin, T. Alon, J. Pe’er, H. Gnessin, T. Chan-Ling, and E. Keshet, Development of retinal vasculature is mediated by hypoxia-induced vascular endothelial growth factor (VEGF) expression by neuroglia. J. Neurosci. 15 (7), 4738-4747, (1995). 3. D. C. Darland, L. J. Massingham, S. R. Smith, E. Piek, M. Saint-Geniez, and P. A. D’Amore, Pericyte production of cell-associated VEGF is differentiation-dependent and is associated with endothelial survival. Developmental Biology 264 (1) 275-288, (2003).

136

T. Chan-Ling

4. J. Ash, D. S. McLeod, and G. A. Lutty, Transgenic expression of leukemia inhibitory factor (LIF) blocks normal vascular development but not pathological neovascularization in the eye. Mol. Vis. 11 298-308, (2005). 5. T. Chan-Ling, P. Halasz, and J. Stone, Development of retinal vasculature in the cat: stages, topography and mechanisms. Curr. Eye Res. 9 (5), 459-478, (1990). 6. T. Chan-Ling, P. Halasz, and J. Stone, Development of retinal vasculature in the cat: processes and mechanisms. Curr. Eye Res. 9 459-478, (1990). 7. T. Chan-Ling, D. S. McLeod, S. Hughes, L. Baxter, Y. Chu, T. Hasegawa, and G. A. Lutty, Astrocyte-endothelial cell relationships during human retinal vascular development. Investigative Ophthalmology & Visual Science 45 (6), 2020-2032, (2004). 8. N. Ashton, Retinal angiogenesis in the human embryo. Brit. Med. Bull. 26 (2), 103-106, (1970). 9. G. A. Lutty and D. S. McLeod, A new technique for visualization of the human retinal vasculature. Arch. Ophthalmol. 110 (2), 267-276, (1992). 10. K. Gogat, L. Le Gat, L. Van Den Berghe, D. Marchant, A. Kobetz, S. Gadin, B. Gasser, I. Quere, M. Abitbol, and M. Menasche, VEGF and KDR gene expression during human embryonic and fetal eye development. Invest. Ophthalmol. Vis. Sci. 45 (1), 7-14, (2004). 11. H. Gerhardt, M. Golding, M. Fruttiger, C. Ruhrberg, A. Lundkvist, A. Abramsson, M. Jeltsch, C. Mitchell, K. Alitalo, D. Shima, and C. Betsholtz, VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J. Cell Biol. 161 (6), 1163-1177, (2003). 12. B. Dreher and S. R. Robinson, Development of the retinofugal pathway in birds and mammals: evidence for a common ‘timetable’. Brain Behav. Evol. 31 369-390, (1988). 13. E. M. Dorn, L. Hendrickson, and A. E. Hendrickson, The appearance of rod opsin during monkey retinal development. Investigative Ophthalmology & Visual Science 36 (13) 2634-2651, (1995). 14. S. Hughes and T. Chan-Ling, Roles of endothelial cell migration and apoptosis in vascular remodeling during development of the central nervous system. Microcirculation 7 (5) 317-333, (2000). 15. T. Chan-Ling, B. Gock, and J. Stone, The effect of oxygen on vasoformative cell division. Evidence that “Physiological Hypoxia” is the stimulus for normal retinal vasculogenesis. Invest. Opthalmol. Vis. Sci. 36 (7), 1201-1214, (1995). 16. T. Chan-Ling, Glial, vascular, and neuronal cytogenesis in whole-mounted cat retina. Microsc. Res. Tech. 36 (1), 1-16, (1997). 17. J. M. Provis, J. Leech, C. M. Diaz, P. L. Penfold, J. Stone, and E. Keshet, Development of the human retinal vasculature: cellular relations and VEGF expression. Experimental Eye Research 65 (4), 555-568, (1997). 18. H. Wolburg, S. Liebner, A. Reichenbach, and H. Gerhardt, The pecten oculi of the chicken: A model system for vascular differentiation and barrier maturation. Int. Rev. Cytol. 187 111-159, (1999). 19. D. Y. Yu, S. J. Cringle, V. A. Alder, E. N. Su, and P. K. Yu, Intraretinal oxygen distribution and choroidal regulation in the avascular retina of guinea pigs. Am. J. Physiol. 270 (3) PT 2, H965-973, (1996). 20. M. Schmidt and I. Flamme, The in vivo activity of vascular endothelial growth factor isoforms in the avian embryo. Growth Factors 15 (3), 183-197, (1998). 21. P. Carmeliet, V. Ferreira, G. Breier, S. Pollefeyt, L. Kieckens, M. Gertsenstein, M. Fahrig, A. Vandenhoeck, K. Harpal, C. Eberhardt, C. Declercq, J. Pawling, L. Moons, D. Collen, W. Risau, and A. Nagy, Abnormal blood vessel development and lethality in embroyis lacking a single VEGF allele. Nature 380 (6573), 435-439, (1996).

6. Formation of the Human Retinal Vasculature

137

22. N. Ferrara, K. Carver-Moore, H. Chen, M. Dowd, L. Lu, K. S. O’Shea, L. PowellBraxton, K. J. Hillan, and M. W. Moore, Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380 (6573), 439-442, (1996). 23. T. Chan-Ling, Glial, neuronal and vascular interactions in the mammalian retina. In: Progress in Retinal Research, edited by Osborne N and Chader G. Oxford: Pergamon Press, 1994, p. 357-389. 24. A. Das and P. G. McGuire, Retinal and choroidal angiogenesis: pathophysiology and strategies for inhibition. Progress in Retinal and Eye Research 22 (6), 721-748, (2003). 25. J. Schnitzer, Astrocytes in the guinea pig, horse, and monkey retina: their occurrence coincides with the presence of blood vessels. Glia 1 ,74-89, (1988). 26. T. L. Ling and J. Stone, The development of astrocytes in the cat retina: evidence of migration from the optic nerve. Developmental Brain Research 44 (1), 73-85, (1988). 27. T. Ling, J. Mitrofanis, and S. Stone, The origin of retinal astrocytes in the rat: Evidence of migration from the optic nerve. J. Comp. Neurol. 286, 345-352, (1989). 28. M. I. Dorrell, E. Aguilar, and M. Friedlander, Retinal vascular development is mediated by endothelial filopodia, a preexisting astrocytic template and specific R-cadherin adhesion. Investigative Ophthalmology & Visual Science 43 (11), 3500-3510, (2002). 29. Y. Chu, S. Hughes, T. and Chan-Ling, Differentiation and migration of astrocyte precursor cells and astrocytes in human fetal retina: relevance to optic nerve coloboma. FASEB Journal 15 (11), 2013-2015, (2001). 30. J. H. Tao-Cheng and M. W. Brightman, Development of membrane interactions between brain endothelial cells and astrocytes in vitro. Int. J. Dev. Neurosci. 6, 25-37, (1988). 31. S. Tout, T. Chan-Ling, H. Holländer, J. and Stone, The role of Müller cells in the formation of the blood-retinal barrier. Neuroscience 55 (1), 291-301, (1993). 32. L. E. Benjamin, I. Hemo, and E. Keshet, A plasticity window for blood vessel remodelling is defined by pericyte coverage of the preformed endothelial network and is regulated by PDGF-B and VEGF. Development 125 (9), 1591-1598, (1998). 33. S. Hughes and T. Chan-Ling, Characterization of smooth muscle cell and pericyte differentiation in the rat retina in vivo. Invest. Ophthalmol. Vis. Sci. 45 (8), 2795-2806, (2004). 34. T. Chan-Ling, M. P. Page, T. Gardiner, L. Baxter, E. Rosinova, and S. Hughes, Desmin ensheathment ratio as an indicator of vessel stability: evidence in normal development and in retinopathy of prematurity. Am. J. Pathol. 165 (4), 1301-1313, (2004). 35. P. L. Penfold, J. M. Provis, M. C. Madigan, D. van Driel, F. A. and Billson, Angiogenesis in normal human retinal development: the involvement of astrocytes and macrophages. Graefe’s Arch. Clin. Exp. Ophthalmol. 228, 255-263, (1990). 36. H. L. Naug, J. Browning, G. A. Gole, and G. Gobe, Vitreal macrophages express vascular endothelial growth factor in oxygen-induced retinopathy. Clinical & Experimental Ophthalmology 28 (1), 48-52, (2000). 37. J. Ambati, B. K. Ambati, S. H. Yoo, S. Ianchulev, and A. P. Adamis, Age-related macular degeneration: etiology, pathogenesis, and therapeutic strategies. Survey of Ophthalmology 48 (3), 257-293, (2003). 38. M. Hellstrom, M. Kalen, P. Lindahl, A. Abramsson, and C. Betsholtz, Role of PDGF-B and PDGFR-beta in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Development 126, 3047-3055, (1999). 39. K. K. Hirschi, S. A. Rohovsky, and P. A. D’Amore, PDGF, TGF-beta, and heterotypic cell-cell interactions mediate endothelial cell-induced recruitment of 10T1/2 cells and their differentiation to a smooth muscle fate. Journal of Cell Biology 141 (3), 805-841, (1998).

138

T. Chan-Ling

40. H. Mi, H. Haeberle, and B. A. Barres, Induction of astrocyte differentiation by endothelial cells. Journal of Neuroscience 21 (5), 1538-1547, (2001). 41. N. Ashton, B. Ward, and G. Serpell, Effect of oxygen on developing retinal vessels with particular reference to the problem of retrolental fibroplasia. Brit. J. Ophthalmol. 38, 397432, (1954). 42. A. Das, W. Fanslow, D. Cerretti, E. Warren, N. Talarico, and P. McGuire, Angiopoietin/ Tek interactions regulate mmp-9 expression and retinal neovascularization. Lab. Invest. 83 (11), 1637-1645, (2003).

Chapter 7 IGF-1 AND RETINOPATHY

Lois E. H. Smith, MD, PhD Department of Ophthalmology, Children’s Hospital, Harvard Medical School, Boston, Massachusetts

Abstract:

1.

Retinopathy continues to be a major cause of blindness in children (retinopathy of prematurity, ROP), in adults (diabetic retinopathy), and in the elderly (age-related macular degeneration), despite current therapy. Although ablation of the retina reduces the incidence of blindness by suppressing the neovascular phase of ROP and diabetic retinopathy, the visual outcomes after treatment are often poor. Preventive therapy is required and will likely come from a better understanding of the pathophysiology of the disease.

TWO PHASES OF ROP AND DIABETIC RETINOPATHY

Retinopathy of prematurity (ROP) was first recognized in the late 1940’s and was associated with excessive oxygen use.1 Despite controlled oxygen delivery, the number of infants with ROP has increased further, probably because of the increased survival of very low birth weight infants,2 indicating the likely association of ROP with both oxygen-related and non oxygen-related growth factors. Both ROP and diabetic retinopathy occur in two phases. In the first phase of ROP, there is cessation of the normal retinal vascular growth, which would normally occur in utero, as well as loss of some of the developed vessels. As the infant matures, the resulting non-vascularized retina becomes increasingly metabolically active and increasingly hypoxic. Similarly, the first phase of diabetic retinopathy consists of slow loss of capillaries associated most prominently with poor control of hyperglycemia. The second phase of ROP and diabetic retinopathy, retinal neovascularization or 139 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 139–149. © Springer Science+Business Media B.V. 2008

140

L. E. H. Smith

proliferative retinopathy, is hypoxia-induced.3 In ROP, the onset occurs at about 32 weeks post-menstrual age, and the progression of neovascularization is similar to that in adult diabetic retinopathy. There are no rodent models available for studying proliferative or neovascular diabetic retinopathy. We developed a mouse model of ROP to take advantage of the genetic manipulations possible in the murine system. The eyes of some animals, though they are born full-term, are incompletely vascularized at birth and resemble the retinal vascular development of premature infants. Exposure of these animals to hyperoxia causes vasoobliteration and cessation of normal retinal blood vessel development, which mimics Phase I of ROP.4-6 When mice return to room air, the nonperfused portions of the retina become hypoxic, which in turn causes retinal neovascularization similar to Phase II of ROP and of other retinopathies.

2.

VEGF AND PHASE II OF ROP

Because hypoxia is a driving force for retinal neovascularization or proliferative retinopathy, we first searched for a hypoxia-regulated factor during Phase II of ROP. Vascular endothelial growth factor (VEGF) is a hypoxia-inducible cytokine7 and is a vascular endothelial cell mitogen.8 In the mouse5, retinal hypoxia stimulates an increase in the expression of VEGF before the development of neovascularization.9 Furthermore, inhibition of VEGF decreases the neovascular response,10,11 indicating that VEGF is a critical factor in retinal neovascularization. Other investigators have also shown that VEGF is associated with ocular neovascularization in other animal models, confirming the central role of VEGF in neovascular eye disease.12-15 These results have been corroborated clinically. VEGF is elevated in the vitreous of patients with retinal neovascularization.16 In a patient with ROP, VEGF was found in the retina in a pattern consistent with mouse results.14

3.

VEGF AND PHASE I OF ROP

In animal models, the first phase of ROP is also VEGF-dependent. VEGF is required for normal blood vessel growth. VEGF is found anterior to the developing vasculature, in what has been described as a wave of physiological hypoxia that precedes vessel growth.17,18 As the retina develops anterior to the vasculature, there is increased oxygen demand, which creates localized hypoxia. VEGF is expressed in response to the hypoxia, and blood vessels grow toward the VEGF stimulus. As the hypoxia

7. IGF-1 and Retinopathy

141

is relieved by oxygen from the newly formed vessels, VEGF mRNA expression is suppressed, moving the wave forward. Supplemental oxygen interferes with normal retinal vascular development through suppression of VEGF mRNA (Figure 1). Furthermore, hyperoxia-induced vaso-obliteration is caused by apoptosis of vascular endothelial cells, and vaso-obliteration can be at least partially prevented by administration of exogenous VEGF18,19 and more specifically by placental growth factor-1 (PlGF-1), the specific agonist of VEGF receptor-1 (VEGFR1).20 This indicates that VEGFR-1 is required for maintenance of the immature retinal vasculature and explains at least in part the effect of hyperoxia on normal vessel development in ROP.

Ni vessel growth in retina ↓VEGF Resolution of VEGF ↑VEGF Proliferative retinopathy In utero

Premature birth

Maturing retina

Retinal

Normal vessel growth

Vessel growth stops

Hypoxia

Neovascularization

IGF-1 nl

↓IGF-1

slow ↑IGF-1

↑IGF-1 to “threshold”

VEGF nl

(↓VEGF)

↑↑VEGF

↑↑VEGF

Ni in utero oxygen 70% saturated

Relative hyperoxia

Retinal detachment VEGF IGF-1

Local hypoxia in areas of blood vessel loss

Figure 7-1. Onset of neovascularization in the retinas of premature infants. In utero (a), the infant is exposed to low oxygen, but upon premature birth (b), a state of relative hyperoxia is induced with exposure to room air or supplemental oxygen. In addition, preterm birth is associated with very low levels of IGF-1 due to loss from the placenta and inability to overcome this loss due to an immature liver. This causes blood vessel formation to stop, resulting in local areas of hypoxia (c). The biological response is then to promote neovascularization (d), in part by increasing the expression of VEGF. IGF-1 rises slowly from low levels after preterm birth so that VEGF can then activate Akt and MAPK.

4.

GH/IGF-1 IN PHASE II OF ROP

Although VEGF has an important role in the development of retinal blood vessels, it is clear that other biochemical mediators are also involved in the pathogenesis of ROP. Inhibition of VEGF does not completely inhibit hypoxia-induced retinal neovascularization in the second phase of ROP. Also, despite controlled use of supplemental oxygen, the disease persists as infants of ever lower gestational age are saved, suggesting that other factors

142

L. E. H. Smith

related to prematurity itself and growth and development are also at work. Because growth hormone (GH) has been implicated in proliferative diabetic retinopathy,21 we considered both GH and insulin-like growth factor 1 (IGF1), which mediates many of the mitogenic aspects of GH, as potential candidates for these factors. In transgenic mice expressing a GH receptor antagonist or in normal mice treated with a somatostatin analog that decreases GH release, there is a substantial reduction in the amount of proliferative retinopathy, the second phase of ROP.22 The effect of GH inhibition is mediated through an inhibition of IGF-1, because administration of exogenous IGF-1 completely restores the neovascularization seen in the control mice. The GH/IGF-1 inhibition occurs without diminishing hypoxia-induced VEGF production. Proof of the direct role of IGF-1 in the proliferative phase of ROP in mice was established with an IGF-1 receptor antagonist, which suppressed retinal neovascularization without altering the vigorous VEGF response induced in the mouse ROP model.23 IGF-1 regulation of retinal neovascularization is mediated at least in part through control of VEGF activation of p44/42 MAPK. IGF-1 acts permissively to allow maximum VEGF induction of new vessel growth. Inadequate levels of IGF-1 inhibit vessel growth despite the presence of VEGF (Figure 2).

Figure 7-2. The relationship between IGF-1 and VEGF and its effect of the growth of new blood vessels.

7. IGF-1 and Retinopathy

5.

143

LOW LEVELS OF IGF-1 AND PHASE I OF ROP

IGF-1 is also critical to the first phase of ROP24 and to the normal development of the retinal vessels. After birth, IGF-1 is not maintained at in utero levels due to the loss of IGF-1 provided by the placenta and the amniotic fluid. We hypothesized that IGF-1 is critical to normal retinal vascular development and that a lack of IGF-1 in the early neonatal period is associated with a lack of vascular growth and with subsequent proliferative ROP. To determine if IGF-1 is critical to normal blood vessel growth, retinal blood vessel development was examined in IGF-1 null mice. The retinal blood vessels grew more slowly in the IGF-1 null mice than in normal mice, a pattern very similar to that seen in premature babies with ROP. It was determined that IGF-1 controls maximum VEGF activation of the Akt endothelial cell survival pathway. This finding explains how loss of IGF-1 could cause ROP by preventing the normal survival of vascular endothelial cells. These findings were confirmed in premature infants, where the mean IGF-1 was significantly lower in babies with ROP than babies without ROP.24,25 These results suggest that replacement of IGF-1 to uterine levels might prevent ROP by allowing normal retinal vascular development. If phase I is aborted, the destructive second phase of vaso-proliferation will not occur.

6.

IGF-1 IN DIABETIC RETINOPATHY

6.1

Elevated IGF-1 levels

There is a long-standing (and complex) association between IGF-1 and diabetic retinopathy,26-28 with conflicting evidence that elevated levels of serum IGF-1 are associated with proliferative retinopathy (phase II). The hypothesis that elevated levels of IGF-1 cause proliferative retinopathy is based in part on observations that patients with neovascular disease have very high vitreous levels of IGF-1.29-33 IGF system components could accumulate in the vitreous because of local production.34 However, it is thought that diffusion from serum plays an important role,35 and it has been suggested that increased levels of IGF-1 in the vitreous are the result and not the cause of neovascularization. This is based on the well-established increased permeability of the blood–retina barrier in diabetic patients. Circulating IGF and IGF binding protein-3 (IGFBP-3) levels are 10–100 times higher than those measured in vitreous.35 Furthermore, patients with

144

L. E. H. Smith

proliferative diabetic retinopathy show a significant positive correlation between serum and vitreous levels of IGF-1,36 and the increase in vitreous levels of IGF-1, IGF-2, and IGFBP-3 parallels the increase in vitreous of liver-derived serum proteins. Thus, it is generally accepted that diffusion from serum plays a key role. This accumulation is caused by a non-specific increase in leakiness of the blood-retina barrier, since these same elevations are found in patients with non-diabetic causes of leaky retinal vasculature.31,35 Some longitudinal studies have shown that intensive insulin treatment in patients with poorly controlled hyperglycemia, which rapidly increases total serum IGF-1, is associated with accelerated diabetic retinopathy.37,38 However, the majority of investigations (cross-sectional as well as longitudinal) have found no significant correlation between circulating IGF1 and the development of proliferative diabetic retinopathy.39-43 An animal study of normoglycemic/normoinsulinemic transgenic mice overexpressing IGF-1 through an insulin promoter at supra-physiological levels in the retina developed loss of pericytes and thickening of basement membrane of retinal capillaries.44 In older transgenic mice over-expressing IGF-1, neovascularization of the retina and vitreous cavity were observed which was consistent with increased IGF-1 induction of VEGF expression45 in retinal cells. VEGF alone can cause these same effects.46 These accumulated findings suggest that once proliferative neovascular (and therefore leaky) vessels occur in the retina in phase II, leaked serum IGF-1 may further promote the proliferation of retinal vessels through stimulation of VEGF. However, it has not been established that serum IGF-1 in the absence of leaky vessels causes proliferative disease. In diabetic patients with acromegaly and elevated IGF-1 in serum and in vitreous, proliferative diabetic retinopathy is rare.47

6.2

Low IGF-1 levels

Although less attention has been paid to the study of the first phase of diabetic retinopathy, there is evidence that low IGF-1 is associated with vessel loss (phase I), which could then lead to phase II, proliferative retinopathy. There is a substantial body of work indicating that low IGF-1 is associated with the hyperglycemia of poorly controlled diabetes, which is in turn the strongest risk factor for diabetic complications. Hyperglycemia is associated with elevated GH secretion and reduced serum IGF-1 concentrations.48,49 Low portal insulin levels are thought to lead to decreased production of IGF-1 with subsequently increased GH and IGFBP-1 levels.50 The elevation in GH secretion, due to loss of feedback inhibition of IGF-1 as a result of the low portal insulin levels, may worsen hyperglycemia by

7. IGF-1 and Retinopathy

145

counteracting insulin action.51 Thus, restoration of normal IGF-1 levels in insulin-treated patients with recombinant human (rh) IGF-1 or IGF1/IGFBP-3 complex results in a concomitant reduction in GH secretion and insulin requirement to maintain euglycemia.52-55 A study in Laron dwarfs with diabetes and with very low levels of IGF-1 indicates that these patients undergo phase I and phase II of diabetic retinopathy, suggesting that low IGF-1 may be an important contributing factor to retinopathy.56 Low IGF-1 may also be involved in large vessel disease. Individuals with low circulating IGF-1 levels and high IGFBP-3 levels have a significantly increased risk of developing ischemic heart disease during a 15-year follow-up period.57 More recent evidence suggests that very low IGF-1 directly causes decreased vascular density.58 This accumulated evidence indicates that low IGF-1 is associated with vessel loss and may be detrimental in diabetes by contributing to early vessel degeneration in phase I. This vaso-obliteration sets the stage for hypoxia, leading later to neovascularization/proliferative retinopathy. Thus, treatment of diabetic patients with IGF-1 within the normal physiological range as an adjunct to insulin might prevent and not worsen the development of diabetic microvascular complications.59

7.

CLINICAL IMPLICATIONS

These studies suggest a number of ways to intervene medically in the development of retinopathy, but they also make clear that timing is critical to any intervention. Inhibition of either VEGF or IGF-1 early after birth can prevent normal blood vessel growth and precipitate ROP, whereas inhibition at the second neovascular phase might prevent destructive neovascularization. This also may be true in diabetic retinopathy. The choice of any intervention must be made to promote normal physiological development and survival of both blood vessels and other tissue. In particular, the proof that development of ROP is associated with low levels of IGF-1 after premature birth suggests that physiological replacement of IGF-1 to levels found in utero might prevent ROP by allowing normal vascular development.

146

L. E. H. Smith

REFERENCES 1. A. Patz, L. E. Hoeck, and E. DeLaCruz, Studies on the effect of high oxygen administration in retrolental fibroplasia: I. Nursery observations, Am. J. Ophthalmol. 35, 1248-1252 (1952). 2. J. T. Flynn, Acute proliferative retrolental fibroplasia: multivariate risk analysis, Transactions of the American Ophthalmological Society 81, 549-591 (1983). 3. I. Michaelson, The mode of development of the vascular system of the retina, with some observations in its significance for certain retinal diseases, Trans. Ophthalmol. Soc. UK 68, 137-180 (1948). 4. N. Ashton, Oxygen and the growth and development of retinal vessels. In vivo and in vitro studies. The XX Francis I. Proctor Lecture, Am. J. Ophthalmol. 62 (3), 412-435 (1966). 5. L. E. Smith, E. Wesolowski, A. McLellan, S. K. Kostyk, R. D’Amato, R. Sullivan, and P. A. D’Amore, Oxygen-induced retinopathy in the mouse, Invest. Ophthalmol. Vis. Sci. 35 (1), 101-111 (1994). 6. J. S. Penn, B. L. Tolman, and M. M. Henry, Oxygen-induced retinopathy in the rat: relationship of retinal nonperfusion to subsequent neovascularization, Invest. Ophthalmol. Vis. Sci. 35, 3429-3435 (1994). 7. K. H. Plate, G. Breier, H. A. Weich, and W. Risau, Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo, Nature 359 (6398), 845-848 (1992). 8. K. J. Kim, B. Li, J. Winer, M. Armanini, N. Gillett, H. S. Phillips, and N. Ferrara, Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumour growth in vivo, Nature 362 (6423), 841-844 (1993). 9. E. A. Pierce, R. L. Avery, E. D. Foley, L. P. Aiello, and L. E. Smith. Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neovascularization, Proc. Natl. Acad. Sci. U S A 92 (3), 905-909 (1995). 10. G. S. Robinson, E. A. Pierce, S. L. Rook, E. Foley, R. Webb, and L. E. Smith, Oligodeoxynucleotides inhibit retinal neovascularization in a murine model of proliferative retinopathy, Proc. Natl. Acad. Sci. U S A 93 (10), 4851-4856 (1996). 11. L. P. Aiello, E. A. Pierce, E. D. Foley, H. Takagi, H. Chen, L. Riddle, N. Ferrara, G. L. King, and L. E. Smith, Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins, Proc. Natl. Acad. Sci. U S A 92 (23), 10457-10461 (1995). 12. M. L. Donahue, D. L. Phelps, R. H. Watkins, M. B. LoMonaco, and S. Horowitz, Retinal vascular endothelial growth factor (VEGF) mRNA expression is altered in relation to neovascularization in oxygen induced retinopathy, Current Eye Research 15 (2), 175-184 (1996). 13. J. Stone, T. Chan-Ling, J. Pe’er, A. Itin, H. Gnessin, and E. Keshet, Roles of vascular endothelial growth factor and astrocyte degeneration in the genesis of retinopathy of prematurity, Invest. Ophthalmol. Vis. Sci. 37 (2), 290-299 (1996). 14. T. L. Young, D. C. Anthony, E. Pierce, E. Foley, and L. E. Smith, Histopathology and vascular endothelial growth factor in untreated and diode laser-treated retinopathy of prematurity, J. Aapos 1 (2), 105-110 (1997). 15. A. P. Adamis, D. T. Shima, M. J. Tolentino, E. S. Gragoudas, N. Ferrara, J. Folkman, P. A. D’Amore, and J. W. Miller, Inhibition of vascular endothelial growth factor prevents retinal ischemia-associated iris neovascularization in a nonhuman primate, Arch. Ophthalmol. 114 (1), 66-71 (1996).

7. IGF-1 and Retinopathy

147

16. L. P. Aiello, R. L. Avery, P. G. Arrigg, B. A. Keyt, H. D. Jampel, S. T. Shah, L. R. Pasquale, H. Thieme, M. A. Iwamoto, J. E. Park, et al., Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders [see comments], N. Engl. J. Med. 331 (22), 1480-1487 (1994). 17. J. Stone, A. Itin, T. Alon, J. Pe’er, H. Gnessin, T. Chan-Ling, and E. Keshet, Development of retinal vasculature is mediated by hypoxia-induced vascular endothelial growth factor (VEGF) expression by neuroglia, J. Neurosci. 15 (7 Pt 1), 4738-4747 (1995). 18. E. A. Pierce, E. D. Foley, and L. E. Smith, Regulation of vascular endothelial growth factor by oxygen in a model of retinopathy of prematurity [see comments] [published erratum appears in Arch. Ophthalmol. 115 (3), 427 (Mar 1997).], Arch. Ophthalmol. 114 (10), 1219-1228 (1996). 19. T. Alon, I. Hemo, A. Itin, J. Pe’er, J. Stone, and E. Keshet, Vascular endothelial growth factor acts as a survival factor for newly formed retinal vessels and has implications for retinopathy of prematurity, Nature Medicine 1 (10), 1024-1028 (1995). 20. S. C. Shih, M. Ju, N. Liu, J. R. Mo, J. Ney, and L. E. Smith, VEGFR-1 Prevents OxygenInduced Retinal Vascular Degeneration - Role of TGF-b1 and PlGF-1, J. Clin. Invest. 112 (1), 50-57 (2003). 21. P. S. Sharp, T. J. Fallon, O. J. Brazier, L. Sandler, G. F. Joplin, and E. M. Kohner, Longterm follow-up of patients who underwent yttrium-90 pituitary implantation for treatment of proliferative diabetic retinopathy, Diabetologia 30 (4), 199-207 (1987). 22. L. E. Smith, J. J. Kopchick, W. Chen, J. Knapp, F. Kinose, D. Daley, E. Foley, R. G. Smith, and J. M. Schaeffer, Essential role of growth hormone in ischemia-induced retinal neovascularization, Science 276 (5319), 1706-1709 (1997). 23. L. E. Smith, W. Shen, C. Perruzzi, S. Soker, F. Kinose, X. Xu, G. Robinson, S. Driver, J. Bischoff, B. Zhang, J. M. Schaeffer, and D. R. Senger, Regulation of vascular endothelial growth factor-dependent retinal neovascularization by insulin-like growth factor-1 receptor, Nature Medicine 5 (12), 1390-1395 (1999). 24. A. Hellstrom, C. Perruzzi, M. Ju, E. Engstrom, A. L. Hard, J. L. Liu, K. AlbertssonWikland, B. Carlsson, A. Niklasson, L. Sjodell, D. LeRoith, D. R. Senger, and L. E. Smith, Low IGF-I suppresses VEGF-survival signaling in retinal endothelial cells: direct correlation with clinical retinopathy of prematurity, Proc. Natl. Acad. Sci. U S A 98 (10), 5804-5808 (2001). 25. A. Hellstrom, E. Engstrom, A. L. Hard, K. Albertsson-Wikland, B. Carlsson, A. Niklasson, C. Lofqvist, E. Svensson, S. Holm, U. Ewald, G. Holmstrom, and L. E. Smith, Postnatal serum insulin-like growth factor I deficiency is associated with retinopathy of prematurity and other complications of premature birth, Pediatrics 112 (5), 1016-1020 (2003). 26. T. J. Merimee, J. Zapf, and E. R. Froesch, Insulin-like growth factors. Studies in diabetics with and without retinopathy, N. Engl. J. Med. 309 (9), 527-530 (1983). 27. D. G. Dills, S. E. Moss, R. Klein, B. E. Klein, and M. Davis, Is insulinlike growth factor I associated with diabetic retinopathy? Diabetes 39 (2), 191-195 (1990). 28. R. Simo, A. Lecube, R. M. Segura, J. Garcia Arumi, and C. Hernandez, Free insulin growth factor-I and vascular endothelial growth factor in the vitreous fluid of patients with proliferative diabetic retinopathy, Am. J. Ophthalmol. 134 (3), 376-382 (2002). 29. R. Burgos, C. Mateo, A. Canton, C. Hernandez, J. Mesa, and R. Simo, Vitreous levels of IGF-I, IGF binding protein 1, and IGF binding protein 3 in proliferative diabetic retinopathy: a case-control study, Diabetes Care 23 (1), 80-83 (2000). 30. R. Meyer-Schwickerath, A. Pfeiffer, W. F. Blum, H. Freyberger, M. Klein, C. Losche, R. Rollmann, and H. Schatz, Vitreous levels of the insulin-like growth factors I and II, and the insulin-like growth factor binding proteins 2 and 3, increase in neovascular eye

148

31.

32.

33.

34.

35.

36.

37.

38.

39.

40. 41.

42.

43.

44.

45.

L. E. H. Smith

disease. Studies in nondiabetic and diabetic subjects, J. Clin. Invest. 92 (6), 2620-2625 (1993). J. Spranger, J. Buhnen, V. Jansen, M. Krieg, R. Meyer-Schwickerath, W. F. Blum, H. Schatz, and A. F. Pfeiffer, Systemic levels contribute significantly to increased intraocular IGF-I, IGF-II and IGF-BP3 [correction of IFG-BP3] in proliferative diabetic retinopathy, Hormone & Metabolic Research 32 (5), 196-200 (2000). R. J. Waldbillig, B. E. Jones, T. J. Schoen, P. Moshayedi, S. Heidersbach, M. S. Bitar, F. J. van Kuijk, E. de Juan, P. Kador, and G. J. Chader, Vitreal insulin-like growth factor binding proteins (IGFBPs) are increased in human and animal diabetics, Curr. Eye Res. 13 (7), 539-546 (1994). M. Boulton, Z. Gregor, D. McLeod, D. Charteris, J. Jarvis-Evans, P. Moriarty, A. Khaliq, D. Foreman, D. Allamby, and B. Bardsley, Intravitreal growth factors in proliferative diabetic retinopathy: correlation with neovascular activity and glycaemic management, Br. J. Ophthalmol. 81 (3), 228-233 (1997). P. E. Spoerri, E. A. Ellis, R. W. Tarnuzzer, and M. B. Grant, Insulin-like growth factor: receptor and binding proteins in human retinal endothelial cell cultures of diabetic and non-diabetic origin, Growth Hormone & IGF Research 8 (2), 125-132 (1998). A. Pfeiffer, J. Spranger, R. Meyer-Schwickerath, and H. Schatz, Growth factor alterations in advanced diabetic retinopathy: a possible role of blood retina barrier breakdown, Diabetes 46 (Suppl 2), S26-S30 (1997). M. Grant, B. Russell, C. Fitzgerald, and T. J. Merimee, Insulin-like growth factors in vitreous. Studies in control and diabetic subjects with neovascularization, Diabetes 35 (4), 416-420 (1986). E. Chantelau, H. Eggert, T. Seppel, E. Schonau, and C. Althaus, Elevation of serum IGF-1 precedes proliferative diabetic retinopathy in Mauriac’s syndrome [letter], Br. J. Ophthalmol. 81 (2), 169-170 (1997). S. L. Hyer, P. S. Sharp, M. Sleightholm, J. M. Burrin, and E. M. Kohner, Progression of diabetic retinopathy and changes in serum insulin-like growth factor I (IGF I) during continuous subcutaneous insulin infusion (CSII), Horm. Metab. Res. 21 (1), 18-22 (1989). Q. Wang, D. G. Dills, R. Klein, B. E. Klein, and S. E. Moss, Does insulin-like growth factor I predict incidence and progression of diabetic retinopathy? Diabetes 44 (2), 161-164 (1995). P. S. Sharp, S. A. Beshyah, and D. G. Johnston, Growth hormone disorders and secondary diabetes, Baillieres Clin. Endocrinol. Metab. 6 (4), 819-828 (1992). J. Frystyk, T. Bek, A. Flyvbjerg, C. Skjaerbaek, and H. Orskov, The relationship between the circulating IGF system and the presence of retinopathy in Type 1 diabetic patients, Diabet. Med. 20 (4), 269-276 (2003). B. Feldmann, P. M. Jehle, S. Mohan, G. E. Lang, G. K. Lang, J. Brueckel, and B. O. Boehm, Diabetic retinopathy is associated with decreased serum levels of free IGF-I and changes of IGF-binding proteins, Growth Hormone & IGF Research 10 (1), 53-59 (2000). B. Feldmann, G. E. Lang, A. Arnavaz, P. M. Jehle, B. O. Bohm, and G. K. Lang, [Decreased serum level of free bioavailable IGF-I in patients with diabetic retinopathy]. Ophthalmologe 96 (5), 300-305 (1999). J. Ruberte, E. Ayuso, M. Navarro, A. Carretero, V. Nacher, V. Haurigot, M. George, C. Llombart, A. Casellas, C. Costa, A. Bosch, and F. Bosch, Increased ocular levels of IGF-1 in transgenic mice lead to diabetes-like eye disease, J. Clin. Invest. 113 (8), 1149-1157 (2004). R. S. Punglia, M. Lu, J. Hsu, M. Kuroki, M. J. Tolentino, K. Keough, A. P. Levy, N. S. Levy, M. A. Goldberg, R. J. D’Amato, and A. P. Adamis, Regulation of vascular endothelial

7. IGF-1 and Retinopathy

46.

47. 48.

49.

50. 51. 52.

53.

54.

55.

56. 57.

58.

59.

149

growth factor expression by insulin-like growth factor I, Diabetes 46 (10), 1619-1626 (1997). M. J. Tolentino, D. S. McLeod, M. Taomoto, T. Otsuji, A. P. Adamis, and G. A. Lutty, Pathologic features of vascular endothelial growth factor-induced retinopathy in the nonhuman primate, Am. J. Ophthalmol. 133 (3), 373-385 (2002). G. van Setten, K. Brismar, and P. Algvere, Elevated intraocular levels of insulin-like growth factor I in a diabetic patient with acromegaly, Orbit 21 (2), 161-167 (2002). D. B. Dunger, T. D. Cheetham, and E. C. Crowne, Insulin-like growth factors (IGFs) and IGF-I treatment in the adolescent with insulin-dependent diabetes mellitus, Metabolism 44 (Suppl 4), 119-123 (1995). J. A. Janssen, M. L. Jacobs, F. H. Derkx, R. F. Weber, A. J. van der Lely, and S. W. Lamberts, Free and total insulin-like growth factor I (IGF-I), IGF-binding protein-1 (IGFBP-1), and IGFBP-3 and their relationships to the presence of diabetic retinopathy and glomerular hyperfiltration in insulin-dependent diabetes mellitus [see comments], Journal of Clinical Endocrinology & Metabolism 82 (9), 2809-2815 (1997). N. Moller and H. Orskov, Does IGF-I therapy in insulin-dependent diabetes mellitus limit complications? Lancet 350 (9086), 1188-1189 (1997). J. M. Holly, S. A. Amiel, R. R. Sandhu, L. H. Rees, and J. A. Wass, The role of growth hormone in diabetes mellitus, J. Endocrinol. 118 (3), 353-364 (1988). T. D. Cheetham, M. Connors, K. Clayton, A. Watts, and D. B. Dunger, The relationship between overnight GH levels and insulin concentrations in adolescents with insulindependent diabetes mellitus (IDDM) and the impact of recombinant human insulin-like growth factor I (rhIGF-I), Clin. Endocrinol. (Oxf.) 46 (4), 415-424 (1997). A. C. Moses, S. C. Young, L. A. Morrow, M. O’Brien, and D. R. Clemmons, Recombinant human insulin-like growth factor I increases insulin sensitivity and improves glycemic control in type II diabetes, Diabetes 45 (1), 91-100 (1996). D. R. Clemmons, A. C. Moses, M. J. McKay, A. Sommer, D. M. Rosen, and J. Ruckle, The combination of insulin-like growth factor I and insulin-like growth factor-binding protein-3 reduces insulin requirements in insulin-dependent type 1 diabetes: evidence for in vivo biological activity, Journal of Clinical Endocrinology & Metabolism 85 (4), 1518-1524 (2000). T. O’Connell and D. R. Clemmons, IGF-I/IGF-binding protein-3 combination improves insulin resistance by GH-dependent and independent mechanisms, J. Clin. Endocrinol. Metab. 87 (9), 4356-4360 (2002). Z. Laron and D. Weinberger, Diabetic retinopathy in two patients with congenital IGF-I deficiency (Laron syndrome), Eur. J. Endocrinol. 151 (1), 103-106 (2004). A. Juul, T. Scheike, M. Davidsen, J. Gyllenborg, and T. Jorgensen, Low serum insulinlike growth factor I is associated with increased risk of ischemic heart disease: a population-based case-control study, Circulation 106 (8), 939-944 (2002). A. Hellstrom, B. Carlsson, A. Niklasson, K. Segnestam, M. Boguszewski, L. de Lacerda, M. Savage, E. Svensson, L. Smith, D. Weinberger, K. Albertsson-Wikland, and Z. Laron, IGF-I is critical for normal vascularization of the human retina, J. Clin. Endocrinol. Metab. 87 (7), 3413-3416 (2002). J. A. Janssen and S. W. Lamberts, Circulating IGF-I and its protective role in the pathogenesis of diabetic angiopathy, Clin. Endocrinol. (Oxf.) 52 (1), 1-9 (2000).

Chapter 8 HYPOXIA AND RETINAL NEOVASCULARIZATION

Bruce A. Berkowitz Departments of Anatomy and Cell Biology and Ophthalmology, Wayne State University, Detroit, Michigan

Abstract:

1.

For over 50 years, retinal hypoxia has been considered to be a major causative factor in the development of retinal neovascularization (NV), a condition associated with blindness and vision loss in a variety of retinopathies. Review of the existing literature and results of new experiments from our laboratory strongly suggest that the oxygen-based pathophysiology stimulating retinal NV is more complicated than previously thought. Our evidence identifies at least two independent conditions involved in the pathogenesis of retinal NV: hypoxia measured under steady-state conditions (i.e., static hypoxia) and found at the border of vascular and avascular retina, and subnormal oxygenation response measured during a provocation and found over both vascular and avascular retina. In practical terms, the identification of links between static hypoxia, oxygen supply dysfunction and NV may lead to improved therapeutic strategies for preventing vision loss and blindness from retinal NV.

INTRODUCTION

Normally, retinal vessels develop in utero by two mechanisms: vasculogenesis (formation of vessels from precursor cells) and angiogenesis (sprouting of vessels from the existing circulation). The inner (or superficial) circulation develops first, largely via vasculogenesis, and covers the retina by about week 26 post-conception. Outer (or deep net) vessel development lags behind that of the inner circulation and is mostly complete by birth. Under special circumstances, a third form of new retinal vessel growth also occurs. In this case, poorly formed blood vessels abnormally grow from the 151 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 151–168. © Springer Science+Business Media B.V. 2008

152

B. A. Berkowitz

existing circulation through the inner limiting membrane and into the vitreous and are subsequently associated with vision loss and blinding complications in retinopathies such as diabetic retinopathy and retinopathy of prematurity (ROP). This ocular pathological angiogenic process is termed neovascularization (NV). The development of all three forms of retinal vessels is commonly thought to occur when oxygen supply is inadequate to meet demand during resting conditions, or hypoxia. The retina is one of the most metabolically active tissues in the body, and it has a very high oxygen demand.1 Because oxygen is not stored within retinal tissue, a continuous supply of oxygen is necessary to maintain adequate retinal nutrition. Consequently, oxygen supply and demand must be precisely balanced through active regulation of nutrient delivery and waste removal to ensure the health of the retina. Retinal hypoxia can occur, for example, during a retinal ischemic event (e.g., branch retinal artery occlusion) in which oxygen supply stops but oxygen consumption is not downregulated. Historically, this hypoxia hypothesis (Figure 1) evolved from the initial work of Michaelson in 1948. He studied excised, india ink-injected pre- and postnatal retinas and noted that there were large capillary-free zones around arteries and that capillary growth tended to occur on the side of the vein farthest from the artery.2 Presumably, these avascular regions were receiving adequate amounts of oxygen from arteries relative to demand. Michaelson hypothesized that in the development of embryonic retinal vessels, and possibly for NV, oxygen concentration gradients from well to poorly oxygenated retina (e.g., from artery to vein) regulated a factor “X,” which in turn influenced new vessel development. Furthermore, subsequent work by others showed that as the inspired oxygen fraction increased or decreased, the size of the capillary-free zones around arteries widened or narrowed, respectively.3 In support of Michaelson’s hypothesis, Chan-Ling et al. examined normal retinal vessel development in kittens by measuring the extent of vasculogenic cell division.4 Cell division was found to be inversely proportional to the level of oxygen in the inspired gas mixture, and they speculated that a “physiological” level of hypoxia related to increased retinal neuronal demand stimulates vasculogenesis.4

8. Hypoxia and Retinal Neovascularization

153

Figure 8-1. Schematic of hypoxia hypothesis: a) during normoxia, no growth of new blood vessels is noted, b) during hypoxia, retina sends out biochemical signals which induce NV from existing circulation and, in turn, brings oxygen to relieve hypoxic region (and removes waste products).

To date, the hypoxia hypothesis has been applied to rationalize both normal retinal vessel growth and the development of retinal NV in diseases such as diabetic retinopathy, age-related macular degeneration, retinopathy of prematurity, and sickle cell retinopathy. However, it has remained unclear how hypoxia alone can be a necessary and sufficient condition (i.e., a cause) for the phenotype change to NV as well as normal vasculogenesis and angiogenesis. This chapter will critically examine the evidence for and against the hypoxia hypothesis. As will be seen, a causative link between hypoxia and NV is not strongly supported.

2.

HYPOXIA AND RETINAL NV: PROS AND CONS

In general, only indirect evidence (such as treatment response in patients and biochemical and oxygen measurements in animal models) has been used to support the hypothesis that hypoxia causes retinal NV. Our working definition of hypoxia is an inadequate supply of oxygen relative to demand.

154

B. A. Berkowitz

In this discussion, oxygen measurements are considered indirect because retinal oxygen demand (i.e., consumption) is not measured but is needed, according to our definition, to formally make an assessment of hypoxia. Oxygen consumption measurements in vivo are difficult, and it is often assumed that consumption changes remain small between control and experimental conditions.

3.

TREATMENT RESPONSE

If we assume for the moment that retinal hypoxia does cause NV, then one obvious treatment for minimizing NV would be to alleviate poor oxygenation by administering supplemental oxygen. In fact, experimental studies of proliferative retinopathy have demonstrated that constantly applied supplemental oxygen significantly reduces the risk of developing experimental retinal NV.5-7 These results helped motivate the National Eye Institue-sponsored multicenter clinical trial (STOP-ROP) to test the efficacy, safety, and costs of providing oxygenation in moderately severe prethreshold ROP.8 However, the STOP-ROP trial did not demonstrate that supplemental oxygen produced a significant reduction in the number of infants requiring peripheral ablative surgery for retinal NV compared with conventional oxygen exposure.9 In other words, the STOP-ROP trial did not demonstrate a beneficial effect of supplemental oxygen on NV. Although there are likely several reasons why the STOP-ROP results were not as expected, we wondered if one possibility was that administering oxygen 100% of the time (i.e., constantly) was impractical due to the daily care needs of at-risk neonates and that instead infants experienced a variable supplemental oxygen exposure. This hypothesis was tested in a simple model of variable supplemental oxygen in which oxygen was administered only 99% of the time.10 It was expected that supplemental oxygen administered either 100% or 99% of the time would reduce retinal hypoxia by similar degrees and thereby lessen retinal NV to comparable extents. Instead, we found that variable supplemental oxygen treatment was significantly less effective at reducing retinal NV than constantly applied supplemental oxygen.10 This outcome was somewhat surprising and raised the possibility that the beneficial effects of constantly applied supplemental oxygen on retinal NV are not entirely due to relieving retinal hypoxia. An alternative explanation may be that supplemental oxygen induces some degree of vasoconstriction, perhaps through the expression of endothelin-1, and this in turn alters retinal perfusion patterns.11 This hypothesis has not yet been tested because, to the best of our knowledge, it has been difficult to accurately quantitate retinal perfusion clinically or in

8. Hypoxia and Retinal Neovascularization

155

animal models of retinal NV during room air and supplemental oxygen breathing. Current noninvasive methods for measuring retinal perfusion are either not quantitative (e.g., fluorescein angiography), have limited spatial resolution and sensitivity (e.g., laser Doppler velocimetry), or are limited by media opacities such as cataracts or the presence of hyaloidal circulation (e.g., video fluorescein angiography). Nonetheless, the results from the supplemental oxygen studies do not appear to unequivocally demonstrate a link between hypoxia and NV. Laser treatment is a useful clinical procedure that has been found to minimize harmful visual consequences of retinal NV in, for example, patients with diabetic retinopathy, sickle cell retinopathy, and ROP. In principle, laser treatment, which destroys small retinal regions, can reduce oxygen consumption by the retinal pigment epithelium (RPE)-photoreceptor complex, thereby increasing oxygen availability from the choroidal circulation to the inner retina.12 However, data supporting an association between laser procedures and increased oxygen availability have been somewhat weak. Experimentally, laser treatment has been found to elevate retinal oxygen levels during room air breathing in nearly avascular retina (rabbit) but not in fully vascularized retina (cat).13,14 Zuckerman, Cheasty, and Wang measured increased inner retinal pO2 over laser burn in rats, but it was unclear whether their data were obtained during room air or 100% oxygen breathing.15 Clinically, some patients with diabetes, despite extensive laser treatment that would be expected to improve hypoxia, have NV that continues to develop and cause complications. Why laser treatment is beneficial in reducing the impact of retinal NV is unclear at present but may involve changes in retinal perfusion.16 For example, it has been reported that laser treatment produces a prolonged decrease in retinal perfusion,17,18 and this altered perfusion pattern might affect the NV outcome independently from whether hypoxia is present or not. In summary, adverse consequences of retinal NV can be reduced to a variable and somewhat unpredictable extent using laser treatment, but whether or not this benefit occurs by increasing oxygen availability and reducing a presumed hypoxia remains speculative. In addition, the reasons for the inconsistent improvement in outcome following these clinical approaches are unclear but may be related to non-hypoxic factors, such as changes in perfusion patterns17,18 or vasoreactivity (see below) and/or individual inflammatory response to the procedures.19 In any event, treatment responses are likely to result from a complex set of mechanisms, and their interpretation in terms of cause and effect between hypoxia and NV remains tentative.

156

4.

B. A. Berkowitz

BIOCHEMICAL EVIDENCE

The most likely candidate yet identified for Michaelson’s factor “X” is vascular endothelial growth factor (VEGF).20 VEGF is a potent hypoxiainducible mitogen found upregulated in cell cultures as well as in proliferative retinopathy. However, in enucleated eyes from patients with diabetes, evidence for retinal VEGF immunoreactivity has also been found before the appearance of gross retinal nonperfusion. While normal histology may suggest normal retinal oxygenation (i.e., normoxia), some caution is needed because retinal hypoxia was found in long-term diabetic cats that also presented with relatively mild histopathology.21 Nonetheless, the data from enucleated eyes at least supports the possibility that hypoxia may not be the sole stimulus for VEGF expression.22,23 Indeed, a variety of factors unrelated to hypoxia, but apparently related to cellular malnutrition, can also upregulate VEGF expression including, for example, increased insulin-like growth factor-1 (IGF-1) and oxidative stress.24,25 These observations raise some reservations about interpreting elevated VEGF levels solely in terms of hypoxia. The appearance of NV likely depends on more than an increased level of a single factor “X.” A minimum requirement for NV development may be that more angiogenic stimulators (e.g., VEGF) are present than angiogenic inhibitors (e.g., pigment epithelium-derived factor (PEDF)).26 Interestingly, hypoxia appears to suppress PEDF expression, and it has been suggested that PEDF downregulation is linked with the development of retinal NV, at least in animal models.27,28 However, this suggestion has not been supported by the available evidence since measures of PEDF levels in vitreous of patients with proliferative retinopathy have been reported as both higher and lower than normal.29 Changes in systemic biochemistry that are apparently unrelated to retinal hypoxia have also been strongly linked to the development of proliferative retinopathy. It has been suggested that metabolic acidosis is an independent risk factor associated with clinical ROP.30 Experimental support for this supposition has been established by the work of Holmes et al., who report that lowering systemic pH using either carbon dioxide, ammonium chloride, or acetazolamide, is a key factor in inducing retinal NV in newborn rats independently of hyperoxemia or hypoxemia.30-32 Other studies have recognized a link between aspects of thyroid function (as assessed by serum levels of thyroxine (T4), thyroid-stimulating hormone (TSH), and IGF-1) and retinal hemodynamics, endothelial cell barrier function, and NV formation.33-36 Neonates born prematurely (e.g., at week 25 or earlier) can present with very low birth weight, lower than normal T4 concentration, and an incompletely developed retinal circulation and large

8. Hypoxia and Retinal Neovascularization

157

avascular (ischemic) sections of peripheral retina. In addition, the development of ROP in very low birth weight infants has been associated with prolonged periods of low plasma IGF-1 levels.37 Lack of IGF-1 in knockout mice prevented normal retinal vascular growth.37 In our animal studies, methylimidazole-induced hypothyroidism in both control and a model of low retinal NV incidence newborn rats confirmed that thyroid function is linked with normal retinal vascular density and that hypothyroidism can play a permissive role in the development of retinal NV, if some risk of NV already exists.38 In summary, changes in retinal biochemistry alone have not unambiguously proven the hypoxia hypothesis.

5.

OXYGEN MEASUREMENTS

5.1

Methods

Several methods have been proposed to measure retinal oxygen tension including oxygen microelectrodes and spectroscopy (e.g., optical and 19F NMR). The most frequently used method of determining retinal oxygen tension (pO2) is through the use of oxygen-sensitive electrodes. This technique can be used to measure pO2 gradients within local regions of the vitreous and/or retina with high spatial and temporal resolution. Intravitreal and intraretinal oxygen electrode studies in vascularized retina, such as the cat or rat, have revealed that the vitreous (which is avascular) does not consume oxygen and that during normoxia, oxygen emanating from the choroidal circulation is effectively prevented from reaching the inner retina and vitreous by the high oxygen consumption rate of the photoreceptors.1,39 For this reason, and because there is a small diffusion distance between retinal vessels and vitreous/retina, oxygen electrode studies have shown that measurement of vitreous oxygen and its changes near the surface of the retina (i.e., preretinal) accurately mirror oxygenation of the most anterior portion of inner retina.1,39 Unfortunately, the invasive nature of this technique is a major limitation. To date, clinical measures of retinal O2 in the preretinal vitreous using an oxygen electrode have only been made on patients undergoing intraocular surgery and have not been useful in addressing questions of hypoxia and retinal NV.40,41 Less frequently used approaches to measuring retinal oxygen tension involve spectroscopic techniques. Several groups have developed methods for estimating retinal arterial and venous oxygen saturation that involve detecting the difference in light absorption between oxygenated and

158

B. A. Berkowitz

deoxygenated hemoglobin using multiple wavelength reflectance oximetry.42,43 This method is expected to be most useful for studying retinopathy associated with retinal hypoxia. However, a major technical limitation that remains to be overcome is the high degree of variability induced by backscattered light, which can confound data interpretation. Other groups have measured intravascular or perivascular pO2 in animal models using methods based on the quenching of phosphorescent or fluorescent dyes by oxygen.15,44,45 This method requires careful attention to light intensities, because at some light levels, the dyes used can produce toxic oxygen free radicals. Furthermore, additional work is needed to unambiguously separate the optical signals that are derived from the retinal and choroidal circulation. 19F magnetic resonance spectroscopy of a perfluorocarbon droplet also has been used to measure preretinal pO2, both in animal models and in a vitrectomized human eye.46-49 This approach is minimally invasive because the droplet is delivered via a 30 g needle, but the technique reflects the retinal oxygen tension with good spatial resolution (only where the droplet is). A major advantage of this approach is that it can measure preretinal oxygen tension under conditions that the previously mentioned methods cannot, such as in newborn rat eyes, mouse eyes, or eyes without a clear optical medium.49,50

5.2

Results

To the best of our knowledge, relatively few studies have attempted to test the hypoxia hypothesis by measuring retinal oxygen levels in models of proliferative retinopathy.46,49,51-53 Pournaras found, in an experimental retinal branch vein occlusion, achieved by using argon laser photocoagulation in miniature pigs, that the preretinal vitreous over all ischemic foci had subnormal pO2 (measured with an oxygen electrode) before the appearance of NV, but only approximately 45% of these ischemic retinas showed development of NV.51 He speculated that a critical level of hypoxia was needed for NV formation. An alternative possibility is that retinal hypoxia was a necessary but not sufficient factor leading to NV. In any case, it is clear that hypoxia alone was not correlated with NV growth in the pig occlusion model. Ernest and Goldstick measured preretinal vitreous oxygen tensions using microelectrodes in a kitten oxygen-induced retinopathy model.52 In this model, the retinal blood vessels of newborn kittens were largely obliterated by exposure to an atmosphere of 80% to 90% oxygen, producing what could be considered a vascular wound. Avascular retina was found to indeed have a lower pO2 than vascular retina at the optic nerve head.52 However, no spatial or temporal associations between this presumed hypoxia and NV

8. Hypoxia and Retinal Neovascularization

159

incidence and severity were provided, so the role of hypoxia in NV could not be determined from these studies. Macrophage infiltration has been linked to abnormal vessel growth in retinal NV models and in tumors.54,55 Thus, another possibility is that some combination of hypoxia and inflammation is involved in NV appearance. Studies using oxygen-induced retinopathy models have suggested the importance of macrophages and inflammatory factors in NV development.54,56 For example, angiogenic factors, such as tumor necrosis factor-alpha (TNF-alpha) and cyclooxygenase (COX-2), which are associated with inflammation, are also upregulated in a mouse oxygeninduced retinopathy model. In the branch retinal vein occlusion study discussed above, an inflammatory reaction to a laser burn could have been a confounding factor involved with NV growth.19 In models of diabetic retinopathy, it has been suggested that early activation of leukocytes leads to monocyte adhesion to the capillary endothelium (leukostasis) with subsequent decreases in retinal blood flow and, ultimately, retinal hypoxia.21,57 Linsenmeier et al. measured intraretinal oxygen profiles using electrodes in 3 long-term (> 6 years) diabetic cats and compared these data to prior results generated in their lab from control cats. They found evidence for retinal hypoxia, which appeared to be correlated with microaneurysms, leukocyte plugging of vessels, and/or endothelial cell death.21 Importantly, retinal NV was not found. This study clearly demonstrates that the link between hypoxia, NV, and inflammation is not well understood, since retinal hypoxia, which is found during chronic diabetes and expected to be linked with upregulated inflammatory factors, did not lead to NV. To also investigate these issues, Handa et al. tested the possibility that hypoxia and inflammation co-exist before retinal NV.46 In a fibroblastinjected eye of non-diabetic rabbits, there is cellular proliferation in the vitreous, NV, and retinal detachment. Antoszyk et al. suggested that NV growth in this model is partly attributable to inflammatory mediators such as macrophages.58 Handa et al. used 19F NMR of a small perfluorocarbon (PFC) droplet placed in the vitreous on the surface of the retina. Significantly lower than normal preretinal vitreous oxygen tensions were found from the first day after cell injection until the development of visible NV, without coexisting evidence for vascular occlusion or retinal detachment. These data support the suggested notion that some combination of hypoxia and inflammation can combine to generate NV. In what is perhaps the most comprehensive study to date, Zhang et al. measured preretinal vitreous pO2 using 19F NMR and a perfluorocarbon droplet in the newborn rat both during normal retinal vessel development and before and after appearance of retinal NV.49 The newborn rat model was

160

B. A. Berkowitz

chosen for a number of reasons. Rats have no retinal circulation at birth (P0). The retinal circulation grows during P0–P14 via vasculogenesis and angiogenesis in a pattern similar to that described above for humans.59,60 For example, coverage of the retina by the superficial vessels at P7 in the rat is similar in appearance to that at 18-20 weeks post-conception in the human. However, while normal human retinal vessels develop in utero at systemic arterial oxygen tensions < 30 mm Hg, the rat retinal circulation develops primarily after birth at arterial oxygen levels of about 100 mm Hg. Despite these differences, Penn et al. have shown that alternating arterial oxygen levels in newborn rats between P0 and P14 produces blood oxygen levels similar to those found in at-risk infants and results in retinal NV after removal to room air.61 For example, daily alterations between 50% and 10% oxygen between days P0 and P14 followed by recovery in room air between P14 and P20 (the “50/10” condition) result in 100% of the rat pups having retinal NV (i.e., 100% incidence) and, when graded by the number of clockhours involved, 6 clockhour severity. To determine clockhour, an analog clock face is mentally superimposed on the retinal surface and the number of clock hours (a score from 0 to 12) occupied by abnormal vessel growth determined. To the best of our knowledge, there have not been full studies to determine whether or not there is are inflammatory components in this newborn rat NV model, although a proinflammatory isoform of VEGF (VEGF-164) appears upregulated and associated only with NV but not normal retinal vessel development.62,63 Zhang et al. found evidence for retinal hypoxia at the border of the vascular and avascular retina during normal retinal vessel growth (P1–P10), but not after the retina had fully vascularized (at times > P14).49 This demonstrates that hypoxia is normally involved in vasculogenesis (supporting the “physiological hypoxia” theory) and implies that lower than normal oxygen tensions do not necessarily result in abnormal retinal vessel growth. Zhang et al. also reported retinal hypoxia before the appearance of NV, but not after NV was evident.49 Importantly, although all of the border between vascular and avascular retina was likely hypoxic, NV was only found in about 50% of the hypoxic regions.51 This result is similar to that reported by Pournaras in a pig model in which NV occurred in less than 50% of the hypoxic regions (see above). Because the presence of hypoxia was not correlated with NV occurrence during normal and before abnormal vessel development, Zhang et al. concluded that hypoxia does not seem to cause the phenotype change from normal to abnormal vessel development. In summary, when taken together, the above measurements of retinal oxygen levels strongly imply that hypoxia at the border of vascular and avascular retina is not a causative factor in the development of retinal NV.

8. Hypoxia and Retinal Neovascularization

6.

RETINAL OXYGENATION RESPONSE: AN ALTERNATIVE HYPOTHESIS

6.1

Rationale

161

Since it does not appear that hypoxia at the border of vascular and avascular retina per se is pathogenic (see above), perhaps changes in oxygenation ability may be linked with NV. There are clinically observable changes to the entire retinal circulation. For example, in ROP, retinal vessels posterior to the border between vascular and avascular retina can become dilated and tortuous (also known as plus disease), and this vascular abnormality has been linked with visual complications.64,65 In patients with diabetes, diffuse retinal edema (measured by leakage of fluorescein from extensive areas of posterior retinal capillary) has now been characterized and is associated with vision loss.66,67 How might panretinal oxygenation changes contribute to retinopathy? We first note that the retinovascular system is never at steady state and must constantly adapt. If the entire retinovascular system is unable to appropriately adjust, a panretinal dynamic mismatch between oxygen supply and demand can occur. This oxygen supply dysfunction can then increase the risk of retinopathy either developing or progressing. In other words, the presence of panretinal vascular abnormalities may prevent oxygen supply from adequately satisfying oxygen demand, not just during room air breathing but during conditions of normal retinovascular activity. One approach to determining if there is a defect in the retinovascular system’s regulatory response would be to use a provocation test. Such a test could also be envisioned as the foundation for a clinical test that predicts the course of diabetic retinopathy and its response to treatment. This approach can potentially produce important insights into the pathophysiological basis of the disease and reveal novel targets for therapy.

6.2

Method

We have developed a functional MRI method that detects a carbogeninduced increase in vitreous partial oxygen pressure over the room air value (ΔpO2) as an increase in the signal intensity.68-72 It is important to note that steady-state (room air) vitreous oxygen tension cannot be measured using MRI, because many factors (e.g., vitreous temperature and protein content) can unpredictably alter the baseline preretinal vitreous water signal and its relaxation properties. However, these factors are not likely to change on the short time scale between baseline and carbogen breathing. Thus, their

162

B. A. Berkowitz

contributions are expected to cancel and not contribute to the ΔpO2 measurement. In other words, an image of the eye obtained during room air breathing alone cannot be used to measure retinal oxygenation, but the images acquired before and during carbogen breathing can be compared to measure the change in pO2. The agreement between the MRI and oxygen electrode data support this interpretation.71 Carbogen is a gas mixture of carbon dioxide (5% CO2) and oxygen (95% O2) that has been used clinically, instead of 100% oxygen, to minimize the vasoconstrictive effects of pure O2 on retinal blood flow. We and others have confirmed that, in the rat, carbogen increases retinal oxygenation relative to O2 breathing.71,73 When this retinal oxygenation response (ROR or ΔpO2) is measured in the posterior vitreous (within 200 μm from the retina), we refer to it as a measure of inner retinal oxygenation.71 The functional MRI ΔpO2 measurement is particularly advantageous because (1) it is noninvasive, (2) it is applicable to a wide range of species including mice, rats, and humans,71,74-77 (3) it simultaneously measures the retinal oxygenation response from superior to inferior ora serrata, and (4) it is not affected by media opacities, such as cataract.68 There are currently no other techniques that can noninvasively measure the panretinal oxygenation response in rats, mice, and humans.

6.3

Results

To determine if ΔpO2 is sensitive to retinal NV, we measured the spatial and temporal ΔpO2 patterns associated with abnormal retinal vessel growth in the 50/10 newborn rat NV model described above. The small size of the newborn rodent eye and presence of hyaloidal circulation has made the measurement of retinal oxygenation and hemodynamic parameters in newborn rodent models difficult with other methods.78,79 In this model, a subnormal ΔpO2 was measured over avascular retina and, somewhat surprisingly, over vascular retina before (P14), during (P20), and after the appearance of retinal NV on P34.7,69 Subnormal ΔpO2 was clearly associated with NV histopathology in all cases. As discussed earlier, constantly applied supplemental oxygen treatment (SOT), unlike variable magnitude SOT, significantly reduces the risk of developing experimental retinal NV.5-7,10 We therefore investigated the consequences of constantly applied SOT on ΔpO2 as well as its association with NV incidence and severity. In the newborn rat 50/10 model, retinal ΔpO2 was measured during the appearance of NV (P20) in rats recovered in 28% supplemental oxygen instead of room air between P14 and P20.7 We found that, after 28% SOT, the expected decrease in NV incidence and severity occurred (P hypoxia response element--> VEGF cascade differentially regulates vascular response and growth rate in tumors, Cancer Res. 60 (22), 6248-6252 (2000).

Chapter 10 THE ROLE OF PROTEIN KINASE C IN DIABETIC RETINAL VASCULAR ABNORMALITIES

Jennifer K. Sun1 and George L. King2 1

Beetham Eye Institute and Eye Research Section, Joslin Diabetes Center, and the Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts; and 2 Section on Vascular Cell Biology and Complications, Joslin Diabetes Center, and the Department of Medicine, Harvard Medical School, Boston, Massachusetts

Abstract:

1.

There is an increasing preponderance of literature that suggests an important role for PKC in the development of diabetic retinopathy. Hyperglycemia activates the DAG-PKC pathway, which in turn regulates a number of vascular functions. Studies show that PKC has a direct effect on retinal blood flow and leukostasis, ECM deposition and basement membrane thickening, and vascular permeability and angiogenesis. Recent investigations have examined the potential role of PKC inhibitors in the treatment of diabetic retinopathy. This chapter outlines those investigations and discusses ongoing clinical trials in this area.

INTRODUCTION

Diabetes mellitus affects more than 16 million people in the United States,1 and over 171 million individuals worldwide.2 Its manifestations are both macro- and microvascular in nature; they include peripheral vascular disease, coronary artery disease, and atherosclerosis as well as retinopathy, nephropathy, and autonomic neuropathy. Diabetic retinopathy develops in almost all people with type 1 diabetes and in more than 60% of those with type 2 diabetes within the first 20 years of disease.3 It is the leading cause of new blindness in working age adults in the western world.4

187 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 187–202. © Springer Science+Business Media B.V. 2008

188

2.

J. K. Sun and G. L. King

PROGRESSION AND TREATMENT OF DIABETIC RETINOPATHY

Diabetes-associated changes are present in the retinal vasculature even before the first clinically recognized signs of diabetic retinopathy. These changes include decreases in retinal blood flow,5 the loss of retinal pericytes,6 and thickening of capillary basement membranes.7 The first standardized grading system for clinically recognizable lesions was created in 1968 at the Airlie House Convention in Alexandria, Virginia.8 The Airlie House Classification of Diabetic Retinopathy was subsequently modified by both the Diabetic Retinopathy Study (DRS)9 and the Early Treatment of Diabetic Retinopathy Study (ETDRS).10 Early signs of clinically apparent diabetic retinopathy include microaneurysms, intraretinal hemorrhages, and cotton wool spots. Retinal ischemia occurs later in the disease and is often accompanied by changes in venous caliber and intraretinal microvascular abnormalities (IRMAs). In some patients, advanced retinal ischemia leads to the formation of neovascular vessels on the optic disc or elsewhere in the retina. This stage is known as proliferative diabetic retinopathy (PDR). Development of neovascularization can lead to further complications such as vitreous hemorrhage or traction retinal detachments. Macular edema due to increased vascular permeability can present at any stage of the disease and is one of the most common causes for vision loss in diabetes.11 Large-scale studies, including the Diabetes Control and Complications Trial (DCCT)12 and the United Kingdom Prospective Diabetes Study (UKPDS),13 established a strong relationship between intensive glycemic control and a decreased risk of progression of diabetic retinopathy. Other studies, such as the DRS14 and the ETDRS,15 also proved the efficacy of laser-based treatment methods such as scatter (panretinal) photocoagulation and focal macular laser in managing high-risk PDR and clinically significant macular edema, respectively. Over the last half-century, treatment of diabetes and its attendant complications has improved significantly. However, a need still exists for more effective treatments. As understanding grows of the molecular mechanisms behind microvascular pathology, new treatments have been proposed that specifically target these mechanisms. This paper reviews the literature surrounding one signaling molecule of recent interest, protein kinase C (PKC). It discusses PKC’s role in the mechanisms underlying diabetic retinopathy as well as the development of PKC inhibitors to prevent retinopathy and its complications.

10. PKC in Diabetic Retinal Vascular Abnormalities

3.

189

MOLECULAR MECHANISMS UNDERLYING DIABETIC RETINOPATHY

Hyperglycemia appears to be the unifying etiologic factor that underlies the diverse vascular complications of diabetes. Multiple clinical studies have demonstrated a significant correlation between levels of glycosylated hemoglobin and the incidence and progression of diabetic retinopathy.13,16,17 The mechanisms by which hyperglycemia leads to retinal vascular endothelial damage have not yet been fully clarified. However, several molecular pathways involved in glucose metabolism have been elucidated. Elevations in blood glucose lead to an increased flux of glucose through glycolysis and affect the ratio of NAD to NADPH. Non-enzymatic reactions of glucose also result in the generation and accumulation of advanced glycation endproducts (AGE) and reactive oxygen species.18,19 In turn, these agents may activate the signal transduction pathway of diacylglycerol (DAG)-PKC (Figure 1). The DAG-PKC pathway acts upon functional enzymes, signaling proteins, cytokine expression, and cell cycle factors and transcription factors. Through these actions, it affects multiple facets of vascular function, including retinal hemodynamics, leukocyte adhesion, cell growth, extracellular matrix regulation, endothelial cell permeability, and angiogenesis.20 Through its effects on the microvasculature, PKC may play an important role in the two major causes of diabetic ocular morbidity: macular edema and proliferative retinopathy.

4.

CLASSIFICATION AND STRUCTURE OF PKC

The family of PKC is composed of at least 12 serine/threonine isoforms.21 These related enzymes serve as intracellular signaling systems for a number of growth factors, hormones, and cytokines. Each PKC molecule comprises a single polypeptide chain with an N-terminal regulatory domain and a Cterminal catalytic domain. Whereas the regulatory domain binds phospholipid cofactors and calcium, the catalytic domain is responsible for the enzyme’s kinase activity.22 The PKCs fall into three categories (classic, novel, and atypical) based on distinctive factors in their catalytic and regulatory domains.23 Classic PKCs are calcium dependent and are activated by both phosphatidylserine (PS) and DAG. Novel PKCs are calciumindependent, but are also regulated by PS and DAG. In contrast, atypical PKCs are calcium-independent and not regulated by DAG, but are sensitive to PS.24

190

J. K. Sun and G. L. King

Figure 10-1. The role of PKCβ in diabetic retinopathy.

5.

THE DAG-PKC PATHWAY

DAG can be derived from either the hydrolysis of phosphatidylinositides (PI) or from the metabolism of phosphatidylcholine (PC) by phospholipase C (PLC) or D (PLD).24,26 However, studies of glomerular mesangial cells and aortic smooth muscle cells reveal that hyperglycemia does not increase PI hydrolysis.27,28 It is more likely that increases in DAG result from de novo synthesis involving the metabolism of dihydroxyacetone phosphate into lysophosphatic acid and then phosphatidic acid (PA). Upregulation of DAG occurs both acutely and chronically in the hyperglycemic state. An increase in glucose levels from 5.5 mM to 22 mM caused increased levels of DAG and higher specific activity for PKC in rat retinal endothelial cells within 3-5 days.29 More persistent upregulation of DAG is seen in the aortic tissue of diabetic dogs up to five years after the onset of hyperglycemia.27 Work by Inoguchi et al. suggests that subsequent euglycemic control may not necessarily correct elevations in DAG levels. High DAG levels were sustained in the aortas but not the hearts of diabetic rats after islet cell transplantation.30

10. PKC in Diabetic Retinal Vascular Abnormalities

191

PKC isoforms are differentially localized and activated within mammalian tissues. Whereas PKCθ is found in skeletal muscle and haematopoietic tissue and PKCγ is seen primarily in the central nervous system, PKCβ is present in multiple vascularized organs, including the retina, kidney, and heart.20 Specific PKC isoforms that have been shown to be active in the diabetic rat retina include PKCα, β1, β2, and ε.29 Of all of these isoforms, the largest fraction is that of PKCβ2. PKCβ is a classic PKC isoform that requires DAG for activation. Additional immunoblotting studies suggest that in vascularized tissues in general, it is primarily PKCβ that is activated in states of hyperglycemia.30,31 Furthermore, hyperglycemia also activates PKCβ in monocytes and leukocytes.32 The preferential activation of certain isoforms of PKC in different cell types is incompletely understood. Three potential mechanisms for this activation specificity have been proposed. First, interactions between levels of DAG and calcium may allow specific activation of PKCβ over other PKC isoforms. It is possible that PKCβ is more sensitive to DAG elevations, especially in the presence of lower concentrations of calcium, than other classic or novel isoforms of PKC.33 Second, increasing DAG levels associated with glucose metabolism in the mitochondria and Golgi complex could cause preferential activation of PKCβ because of its intracellular rather than membrane-bound location. Finally, differential rates of synthesis and degradation of the PKC isoforms could also explain differing rates of isoform activity between different tissue types.22

6.

VASCULAR ALTERATIONS RELATED TO PKC ACTIVATION

Within the spectrum of changes due to diabetes, multiple vascular alterations on a cellular and functional level have been ascribed to the activation of PKC. The following section will review the literature regarding PKC activity and its effects on retinal hemodynamics and leukostasis; extracellular matrix (ECM) and basement membranes; and vascular permeability and angiogenesis, with special attention to mechanisms potentially related to diabetic retinopathy.

6.1

Retinal hemodynamics and leukostasis

It is well established that retinal circulatory changes are a hallmark of early diabetes in the eye, and that they can appear even before the onset of clinically recognized retinopathy. Decreases in retinal blood flow have been documented by video fluorescein angiography34 and laser Doppler35 in

192

J. K. Sun and G. L. King

patients with relatively short disease duration. Later in the disease, particularly after the development of PDR, retinal blood flow increases.36,37 Increased adherence of leukocytes and monocytes to the retinal endothelium is also present in subclinical diabetic retinopathy, but can occur in the presence of insulin resistance alone, without diabetes or overt hyperglycemia.38 Several studies suggest a connection between the DAG-PKC pathway and changes in retinal blood flow. The early impairment in retinal circulation can be mimicked by injection of a PKC activator, phorbol dibutyrate, into the vitreous cavity of normal rats. This injection causes a decrease in blood flow that is similar to that seen in 2-4 week diabetic rats.29 Intravitreal injection of a DAG kinase inhibitor (R59949) both increases retinal DAG levels and decreases retinal blood flow in a dose-dependent fashion.39 Furthermore, oral PKC inhibition can normalize retinal blood flow changes as well as changes in glomerular filtration rate and albumin excretion rate in diabetic rats.40 A possible mechanism by which PKC activity could alter retinal vasoreactivity is through affecting expression of vasoactive factors such as endothelin-1 (ET-1), a potent vasoconstrictor, or nitrous oxide (NO), which acts as a vasodilator. ET-1 is present in both retinal capillary endothelial cells and retinal pericytes,20 and increased levels of ET-1 mRNA have been found in retinal tissue from diabetic rats.41 Studies have shown that intravenous administration of ET-1 results in decreased retinal blood flow secondary to vasoconstriction in nondiabetic rats.42,43 Futhermore, treatment with either phosphoramidon, an endothelin-converting enzyme inhibitor, or the endothelin type A receptor antagonist BQ-123 inhibits the effects of ET1 and increases retinal blood flow.42 Clear links have been demonstrated between elevations in glucose levels, increased PKC activity, and higher levels of both endothelin-converting enzyme and ET-1 expression.44,45 The increase in ET-1 levels seen with increased glucose levels is reduced with a general PKC inhibitor, GF109203X.44 Recent studies suggest that PKC regulates ET-1 expression through increasing levels of platelet-derived growth factor (PDGF)-BB.46 High glucose concentrations also lead to overexpression of endothelial nitric oxide synthase (NOS) and subsequent decreased production of NO in cultured retinal endothelial cells. In one study, PKC inhibition partially reversed the effect of hyperglycemia on NO production.47 PKC activation also plays a role in the increased leukocyte and monocyte adhesion to retinal endothelial cells seen in early diabetes. Abiko et al. recently demonstrated that increased leukostasis alone probably does not suffice to explain diabetic decreases in retinal blood flow. However, it is still possible that leukostasis plays a contributing role in worsening diabetic

10. PKC in Diabetic Retinal Vascular Abnormalities

193

pathology in the retina. This study showed that although leukocyte/monocyte adhesion in the retina is related to oxidative stress rather than directly to hyperglycemia, it can be prevented using either a non-specific PKC inhibitor, d-α-tocopherol, or the PKCβ-specific inhibitor ruboxistaurin (RBX).38

6.2

Changes in basement membrane and ECM

Another change observed early in the course of diabetes is the thickening of capillary basement membranes.7 This thickening results from an increased deposition of ECM and leads to alterations in vascular permeability as well as cellular adhesion, proliferation, differentiation, and gene expression.22 Documented changes in diabetic basement membranes include increases in type IV and VI collagen and increases in fibronectin and laminin.48-51 PKC inhibitors, such as staurosporine and calphostin, act to prevent glucosestimulated transcription of collagen IV in cultured mesangial cells.52 Phorbol ester and other PKC agonists stimulate type IV collagen and fibronectin expression.52,53 PKC may mediate the glucose-induced overexpression of ECM components through its effects on transforming growth factor β (TGFβ) and angiotensin-II.54 It has been shown that the glucose-induced activation of PKC is a key component of the process by which TGFβ stimulates the production of type IV collagen, fibronectin, and laminin in cultured mesangial cells.55 A possible mechanism by which hyperglycemia increases TGFβ is via the regulatory action of PKC on the transcription factors c-fos and c-jun.56,57 These factors are proto-oncogenes that regulate gene transcription through the AP-1 binding site.58 Several studies have established that the AP-1 binding sequence is common to the promoter regions of TGF β1,59 fibronectin,60 and laminin.61

6.3

Vascular permeability and angiogenesis

Diabetic retinal and renal vessels are markedly more permeable than their nondiabetic counterparts to macromolecules such as albumin.62 Research suggests that PKC isoforms may play an important role in the mechanisms by which hyperglycemia leads to vascular permeability. It has been shown that phorbol esters increase cultured endothelial cell permeability through the activation of PKC.63 Increasing levels of the PKCβ1 isoform in dermal endothelial cells enhance the effect of phorbol esters on vascular permeability.64 PKCα is activated by hyperglycemia in porcine aortic endothelial cells and also serves to increase cell permeability.65 Furthermore, the permeability effects of high glucose and phorbol esters can be reduced in

194

J. K. Sun and G. L. King

rat skin tissue by PKC inhibition.66 One hypothesis is that PKC activation increases vascular permeability through its effects on phosphorylation of tight junction proteins, thereby affecting endothelial cell contractility. Specifically, PKC is known to regulate the phosphorylation of cytoskeletal proteins such as caldesmon, vimentin, talin, and vinculin.67-69 Phorbol esters also cause redistribution of cytoskeleton elements such as actin and vimentin.70 PKC isoforms also appear to play a role in the development of diabetesassociated neovascularization through their influence on the activity of growth factors such as vascular endothelial growth factor (VEGF).22 VEGF has mitogenic and pro-permeability effects on endothelial cells. A number of experimental and clinical studies have examined the role of VEGF in the pathogenesis of PDR. VEGF levels are significantly increased in the vitreous fluid and aqueous humor of patients with PDR.71,72 In human vascular smooth muscle cells, increased expression of VEGF due to hyperglycemia can be prevented by administration of PKC inhibitors.73 These properties are mediated via activation of PKCβ, through tyrosine phosphorylation of phospholipase Cγ.74 It has been found that inhibition of PKCβ by the selective inhibitor RBX blocks VEGF’s proliferative, angiogenic, and propermeability effects.40,75 Using a model of ischemia-induced proliferative retinopathy, there is a significant increase in VEGF-mediated retinal neovascularization in transgenic mice overexpressing the PKCβ2 isoform, and a corresponding decrease in angiogenic activity in PKCβ-null mice.76

7.

INHIBITION OF PKC

A number of studies have examined the effect of PKC inhibitors such as staurosporine, H-7, GF109203X, and chelerythrine.20 Their in vitro utilization has been effective in blocking PKC effects, thereby demonstrating an association between PKC activation and decreased retinal blood flow, thickening of basement membranes, and increased vascular permeability and angiogenesis. However, in vivo use of nonspecific inhibitors may be limited by their effects on PKC isoforms that perform vital, non-pathogenic functions throughout the body. Indeed, a recent trial of a non-selective PKC inhibitor, PKC412, as a therapeutic agent for diabetic macular edema resulted in approximately 10% of the patient population being withdrawn from the study due to systemic toxic effects, including gastrointestinal side effects and hepatotoxicity.77 Currently, interest in agents that may be used successfully in vivo has focused primarily on the PKCβ-specific agent, RBX. This PKC inhibitor has

10. PKC in Diabetic Retinal Vascular Abnormalities

195

a fifty-fold higher affinity for PKCβ1/2 than for other PKC isoforms (α, γ, δ, ε, η, and ζ).40 Perhaps because of this high selectivity for PKCβ, its safety profile appears to be better than that of previously tested, non-selective PKC inhibitors. Recent clinical studies have examined the efficacy of RBX in improving outcomes related to diabetic nephropathy78 and retinopathy.79 RBX is a bisindoylmaleimide compound that preferentially inhibits PKCβ1 and PKCβ2 over other PKC isoforms. When orally administered to diabetic rats, RBX successfully increases retinal blood flow as measured by mean circulation time and improves glomerular filtration rates and albumin excretion.40 Oral administration in diabetic rats also reduces microvasculature flow disturbances caused by leukocyte entrapment.80 Intravitreal injection of the compound in the same animal model has been found to decrease PKC activation and increase retinal blood flow.39 Consistent with the theoretical effects of inhibition of PKCβ, RBX has additional antipermeability and anti-angiogenic effects. It suppresses VEGFmediated vascular permeability in vivo75 and prevents the development of retinal neovascularization in a pig model of ischemic retinal disease.81 Clinical phase I and II trials demonstrated that oral RBX (Eli Lilly Co., Indianapolis, IN) is well tolerated for periods of up to a month in doses up to 32 mg a day. In these studies, a significant amelioration of retinal blood flow and mean circulation time was found in patients with no or minimal retinopathy and diabetes of duration less than 10 years.82,83 These improvements occurred despite a lack of change in either fasting blood glucose levels or hemoglobin A1c (HbA1c). Subsequent studies have revealed a favorable safety profile in patients taking 32 mg/day for up to 3 years.79 RBX also appears to have some efficacy in reducing the non-ophthalmic microvascular complications of diabetes. Although it did not improve sensory symptom scores in patients with peripheral diabetic neuropathy in phase III trials,84 a recent pilot study of RBX use in type 2 diabetic patients with proteinuria suggests that its ameliorating effects on renal function are additive to intensive glycemic control and blood pressure regulation by angiotensin inhibition.78 Phase III trials on the ophthalmic effects of RBX have focused on endpoints related to decreases in diabetic macular edema and neovascularization. One trial investigating RBX’s effect on diabetic macular edema found that it did not prevent progression of macular edema or decrease the need for macular grid/focal laser. However, subgroup analysis that excluded patients with poor glycemic control (HbA1c > 10) found a 31% risk reduction in progression of diabetic macular edema.85 Further analysis suggested that patients taking 32 mg/day of RBX were less likely to develop edema involving the central macula, and that when they did, their visual acuities were better than their placebo-taking counterparts.86 Additional trials

196

J. K. Sun and G. L. King

investigating the ability of RBX to ameliorate diabetic macular edema are ongoing. The Protein Kinase C (beta) Inhibitor Diabetic Retinopathy Study Group (PKC-DRS) reported initial results from its phase II/III clinical trial utilizing RBX in July of 2005.79 This study enrolled 252 patients with type 1 or type 2 diabetes and moderately severe to very severe nonproliferative diabetic retinopathy in at least 1 eye. Subjects were randomized to one of three dose levels of RBX, from 8 to 32 mg/day administered orally. The drug was found to have a favorable safety profile, with no clinically significant differences between treatment and placebo groups. Although the primary end point for the study (photographically determined progression of diabetic retinopathy or the use of laser photocoagulation) was not met after three years, there was a significant benefit in terms of decreased rates of moderate visual loss for patients treated with 32 mg/day of RBX.79 Based on the results of a second phase III trial that also demonstrated that RBX reduces sustained moderate vision loss in patients with diabetic retinopathy, FDA approval is currently being sought for the compound.84

8.

CONCLUSIONS

There is an increasing preponderance of literature that suggests an important role for PKC in the development of diabetic retinopathy. Hyperglycemia activates the DAG-PKC pathway, which in turn regulates a number of vascular functions. Studies show that PKC has a direct effect on retinal blood flow and leukostasis, ECM deposition and basement membrane thickening, and vascular permeability and angiogenesis. Recent investigations have examined the potential role of PKC inhibitors in the treatment of diabetic retinopathy. RBX, an oral PKCβ-selective inhibitor, significantly decreases the extent of visual loss in diabetic patients with no to minimal retinopathy over the course of three years. This drug will likely undergo evaluation for FDA approval for the treatment of diabetic retinopathy and possibly diabetic nephropathy in the near future. It is hoped that future clinical and experimental investigations will more clearly elucidate the potential of PKC inhibitors in the treatment of diabetic retinopathy.

10. PKC in Diabetic Retinal Vascular Abnormalities

197

ACKNOWLEDGMENTS This work was supported by the National Institutes of Health Grants K12 EY-16335 (J.K.S.) and DK-59725 (G.L.K.) and by the American Diabetes Association Grant 1-05-RA-61 (G.L.K.).

REFERENCES 1. J. S. Skyler and C. Oddo, Diabetes trends in the USA. Diabetes Metab. Res. Rev. 18(Suppl 3):S21-26, (2002). 2. S. Wild, G. Roglic, A. Green, R. Sicree, and H. King, Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 27, 1047-1053, (2004). 3. D. S. Fong, S. Aiello, T. W. Gardner, G. L. King, G. Blankenship, J. D. Cavallerano, F. L. Ferris, and R. Klein, Diabetic retinopathy. Diabetes Care 26 (Suppl 1): S99-S102, (2003). 4. National Diabetes Data Group. Diabetes in America, 2nd ed. NIH Publ. No 95-1468. (US Govt. Printing Office, Washington, 1995). 5. S. E. Bursell, A. C. Clermont, B. T. Kinsley, D. C. Simonson, L. M. Aiello, and H. A. Wolpert, Retinal blood flow changes in patients with insulin-dependent diabetes mellitus and no diabetic retinopathy. Invest. Ophthalmol. Vis. Sci. 37, 886-897, (1996). 6. D. G. Cogan, D. Toussaint, and T. Kuwabara, Retinal vascular patterns. IV. Diabetic retinopathy. Arch. Ophthalmol. 66,100-112, (1961). 7. J. R. Williamson and C. Kilo, Extracellular matrix changes in diabetes mellitus. In: Comparative pathobiology of major age-related diseases, edited by Scarpelli DG, Migahi G (Liss, New York, 1984), pp. 269-288. 8. L. M. Aiello, Perspectives on diabetic retinopathy. Am. J. Opthalmol. 136, 122-135, (2003). 9. Diabetic Retinopathy Study. Report number 6: design, methods, and baseline results. Report number 7: a modification of the Airlie House classification of diabetic retinopathy. Invest. Ophthalmol. Vis. Sci. 21, 1-226, (1981). 10. Early Treatment Diabetic Retinopathy Study Research Group. Grading diabetic retinopathy from stereoscopic color fundus photographs- an extension of the modified Airlie House classification. ETDRS report number 10. Ophthalmology 98(Suppl), 786-806, (1991). 11. L. P. Aiello, T. W. Gardner, G. L. King, G. Blankenship, J. Cavallerano, F. L. Ferris, and R. Klein, Diabetic retinopathy. Diabetes Care 21, 143-156, (1998). 12. Diabetes Control and Complications Trial Research Group. Progression of retinopathy with intensive versus conventional treatment in the Diabetes Control and Complications Trial. Ophthalmology 102, 647-661, (1995). 13. UK Prospective Diabetes Study Group. Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 352, 837-853, (1998). 14. The Diabetic Retinopathy Study Research Group. Photocoagulation treatment of proliferative diabetic retinopathy: the second report of Diabetic Retinopathy Study findings. Ophthalmology 85, 82-106, (1978).

198

J. K. Sun and G. L. King

15. Early Treatment Diabetic Retinopathy Study Research Group. Treatment techniques and clinical guidelines for photocoagulation of diabetic macular edema. Early Treatment Diabetic Retinopathy Study Report Number 2. Ophthalmology. 94, 761-774, (1987). 16. R. Klein, B. E. Klein, S. E. Moss, M. D. Davis, and D. L. DeMets, Glycosylated hemoglobin predicts the incidence and progression of diabetic retinopathy. JAMA 260, 2864-2871, (1988). 17. DCCT Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N. Engl. J. Med. 329, 977-986, (1993). 18. E. A. Friedman, Advanced glycosylated end products and hyperglycemia in the pathogenesis of diabetic complications. Diabetes Care 22(Suppl 2), B65-71, (1999). 19. D. Giugliano, A. Ceriello, and G. Paolisso, Oxidative stress and diabetic vascular complications. Diabetes Care 19, 257-267, (1996). 20. K. J. Way, N. Katai, and G. L. King, Protein kinase C and the development of diabetic vascular complications. Diabet. Med. 18, 949-959, (2001). 21. H. Mellor and P. J. Parker, The extended protein kinase C superfamily. Biochem. J. 332, 281-292, (1998). 22. M. Meier and G. L. King, Protein kinase C activation and its pharmacological inhibition in vascular disease. Vascular Med. 5, 173-185, (2000). 23. L. P. Aiello, The potential role of PKC β in diabetic retinopathy and macular edema. Surv. Ophthalmol. 47(Suppl 2), S263-269, (2002). 24. Y. Nishizuka, Intracellular signaling by hydrolysis of phospholipids and activation of protein kinase C. Science 258, 607-614, (1992). 25. Y. Nishizuka, Protein kinase C and lipid signaling for sustained cellular responses. FASEB 9, 484-496, (1995). 26. M. Liscovitch and L. C. Cantley, Lipid second messengers. Cell 77, 329-334, (1994). 27. P. Xia, T. Inoguchi, T. S. Kern, R. L. Engerman, P. J. Oates, and G. L. King, Characterization of the mechanism for the chronic activation of diacylglycerol-protein kinase C pathway in diabetes and hypergalactosemia. Diabetes 43, 1122-1129, (1994). 28. P. A. Craven and F. R. DeRubertis, Protein kinase C is activated in glomeruli from streptozotocin diabetic rats. J. Clin. Invest. 83, 1667-1675, (1989). 29. T. Shiba, T. Inoguchi, J. R. Sportsman, W. F. Heath, S. Bursell, and G. L. King, Correlation of diacylglycerol and protein kinase C activity in rat retina to retinal circulation. Am. J. Physiol. 265, E783-793, (1993). 30. T. Inoguchi, R. Battan, E. Handler, J. R. Sportsman, W. Heath, and G. L. King, Preferential elevation of protein kinase C isoform βII and diacylglycerol levels in the aorta and heart of diabetic rats: differential reversibility to glycemic control by islet cell transplantation. Proc. Natl. Acad. Sci. USA 89, 11059-11063, (1992). 31. D. Koya and G. L. King, Protein kinase C activation and the development of diabetic complications. Diabetes 47, 859-866, (1998). 32. G. Ceolotto, A. Gallo, M. Miola, M. Sartori, R. Trevisan, S. Del Prato, A. Semplicini, and A. Avogaro, Protein kinase C activity is acutely regulated by plasma glucose concentration in human monocytes in vivo. Diabetes 48, 1316-1322, (1999). 33. Y. Nishizuka, The molecular heterogeneity of protein kinase C and its implications for cellular regulation. Nature 334, 661-665, (1988). 34. S. E. Bursell, A. C. Clermont, B. T. Kinsley, D. C. Simonson, L. M. Aiello, and H. A. Wolpert, Retinal blood flow changes in patients with insulin-dependent diabetes mellitus and no diabetic retinopathy. Invest. Ophthalmol. Vis. Sci. 37, 886-897, (1996).

10. PKC in Diabetic Retinal Vascular Abnormalities

199

35. G. T. Feke, S. M. Buzney, H. Ogasawara, N. Fujio, D. G. Goger, N. P. Spack, and K. H. Gabbay, Retinal circulatory abnormalities in type 1 diabetes. Invest. Ophthalmol. Vis. Sci. 35, 2968-2975, (1994). 36. E. M. Kohner, V. Patel, and S. M. Rassam, Role of blood flow and impaired autoregulation in the pathogenesis of diabetic retinopathy. Diabetes 44, 603-607, (1995). 37. S. Konno, G. T. Feke, A. Yoshida, N. Fujio, D. G. Goger, and S. M. Buzney, Retinal blood flow changes in type 1 diabetes. A long-term follow-up study. Invest. Ophthalmol. Vis. Sci. 37, 1140-1148, (1996). 38. T. Abiko, A. Abiko, A. C. Clermont, B. Shoelson, N. Horio, J. Takahashi, A. P. Adamis, G. L. King, and S.-E. Bursell, Characterization of retinal leukostasis and hemodynamics in insulin resistance and diabetes. Diabetes 52, 829-837, (2003). 39. S. E. Bursell, C. Takagi, A. C. Clermont, H. Takagi, F. Mori, H. Ishii, and G. L. King, Specific retinal diacylglycerol and protein kinase C beta isoform modulation mimics abnormal retinal hemodynamics in diabetic rats. Invest. Ophthalmol. Vis. Sci. 38, 2711-2720, (1997). 40. H. Ishii, M. R. Jirousek, D. Koya, C. Takagi, P. Xia, A. Clermont, S. E. Bursell, T. S. Kern, L. M. Ballas, W. F. Heath, L. E. Stramm, E. P. Feener, and G. L. King, Amelioration of vascular dysfunctions in diabetic rats by an oral PKC β inhibitor. Science 272, 728-731, (1996). 41. D. Deng, T. Evans, K. Mukherjee, D. Downey, and S. Chakrabarti, Diabetes-induced vascular dysfunction in the retina: role of endothelins. Diabetologia 42, 1228-1234, (1999). 42. C. Takagi, S. E. Bursell, Y. W. Lin, H. Takagi, E. Duh, Z. Jiang, A. C. Clermont, and G. L. King, Regulation of retinal hemodynamics in diabetic rats by increased expression and action of endothelin-1. Invest. Ophthalmol. Vis. Sci. 37, 2504-2518, (1996). 43. S. E. Bursell, A. C. Clermont, B. Oren, and G. L. King, The in vivo effect of endothelins on retinal circulation in nondiabetic and diabetic rats. Invest. Ophthalmol. Vis. Sci. 36, 596-607, (1995). 44. J. Y. Park, N. Takahara, A. Gabriele, E. Chou, K. Naruse, K. Suzuma, T. Yamauchi, S. W. Ha, M. Meier, C. J. Rhodes, and G. L. King, Induction of endothelin-1 expression by glucose: an effect of protein kinase C activation. Diabetes 49, 1239-1248, (2000). 45. S. Keynan, M. Khamaisi, R. Dahan, K. Barnes, C. D. Jackson, A. J. Turner, and I. Raz, Increased expression of endothelin-converting enzyme-1c isoform in response to high glucose levels in endothelial cells. J. Vasc. Res. 41, 131-140, (2004). 46. T. Yokota, R. C. Ma, J. Y. Park, K. Isshiki, K. B. Sotiropoulos, R. K. Rauniyar, K. E. Bornfeldt, and G. L. King, Role of protein kinase C on the expression of plateletderived growth factor and endothelin-1 in the retina of diabetic rats and cultured retinal capillary pericytes. Diabetes 52, 838-45, (2003). 47. U. Chakravarthy, R. G. Hayes, A. W. Stitt, E. McAuley, and D. B. Archer, Constitutive nitric oxide synthase expression in retinal vascular endothelial cells is suppressed by high glucose and advanced glycation end products. Diabetes 47, 945-952, (1998). 48. E. Cagliero, T. Roth, S. Roy, and M. Lorenzi, Characteristics and mechanisms of highglucose-induced overexpression of basement membrane components in cultured endothelial cells. Diabetes 40, 102-110, (1991). 49. S. H. Ayo, R. A. Radnik, J. A. Garoni, W. F. Glass, and J. I. Kreisberg, High glucose causes an increase in extracellular matrix proteins in cultured mesangial cells. Am. J. Pathol. 136, 1339-1348, (1990). 50. G. Pugliese, F. Pricci, F. Pugliese, P. Mene, L. Lenti, D. Andreani, G. Galli, A. Casini, S. Bianchi, and C. M. Rotella, Mechanisms of glucose-enhanced extracellular matrix proteins in cultured mesangial cells. Diabetes 43, 478-490, (1990).

200

J. K. Sun and G. L. King

51. S. Roy, E. Cagliero, and M. Lorenzi, Fibronectin overexpression in retinal microvessels of patients with diabetes. Invest. Ophthalmol. Vis. Sci. 37, 258-266, (1996). 52. P. Fumo, G. S. Kuncio, and F. N. Ziyadeh, PKC and high glucose stimulate collagen α1 (IV) transcriptional activity in a reporter mesangial cell line. Am. J. Physiol. 267, F632-G638, (1994). 53. R. K. Studer, P. A. Craven, F. R. DeRubertis, Role for protein kinase C in the mediation of increased fibronectin accumulation by mesangial cells grown in high-glucose medium. Diabetes 42, 118-126, (1993). 54. K. Ikehara, H. Tada, K. Kuboki, and T. Inokuchi, Role of protein kinase C-angiotensin II pathway for extracellular matrix production in cultured human mesangial cells exposed to high glucose levels. Diabetes Res. Clin. Pract. 59, 25-30, (2003). 55. R. H. Weiss and A. Ramirez, TGF-beta and angiotensin-II-induced mesangial matrix protein secretion is mediated by protein kinase C. Nephrol. Dial. Transplant. 13, 2804-2813, (1998). 56. J. L. Messina, M. L. Standaert, T. Ishizuka, R. S. Weinstock, and R. V. Farese, Role of protein kinase C in insulin’s regulation of c-fos transcription. J. Biol. Chem. 267, 9223-9228, (1992). 57. J. Zhang, L. Wang, J. Petrin, W. R. Bishop, and R. W. Bond, Characterization of sitespecific mutants altered at protein kinase C β1 isoform autophosphorylation sites. Proc. Natl. Acad. Sci. USA 90, 6130-6134, (1993). 58. B. R. Franza, Jr., F. J. Rauscher III, S. F. Josephs, and T. Curran, The Fos complex and Fos-related antigens recognize sequence elements that contain AP-1 binding sites. Science 239, 1150-1153, (1988). 59. S.-J. Kim, P. Angel, R. Lafyatis, K. Hattori, K. Y. Kim, M. B. Sporn, M. Karin, and A. B. Roberts, Autoinduction of transforming growth factor β1 is mediated by the AP-1 complex. Mol. Cell. Biol. 10, 1492-1497, (1990). 60. D. C. Dean, J. J. McQuillian, and S. Weintraub, Serum stiumulation of fibronectin gene expression appears to result from rapid serum-induced binding of nuclear proteins to a camp response element. J. Biol. Chem. 265, 3522-3527, (1990). 61. R. Okano, T. Mita, and T. Matsui, Characterization of a novel promoter structure and its transcriptional regulation of the murine laminin B1 gene. Biochim. Biophys. Acta 1132, 49-57, (1992). 62. J. R. Williamson, K. Chang, R. G. Tilton, C. Prater, J. R. Jeffrey, C. Weigel, W. R. Sherman, D. M. Eades, and C. Kilo, Increased vascular permeability in spontaneously diabetic BB/W rats and in rats with mild versus severe streptozocin-induced diabetes. Prevention by aldose reductase inhibitors and castration. Diabetes 36, 813-821, (1987). 63. J. J. Lynch, T. J. Ferro, F. A. Blumenstock, A. M. Brockenauer, and A. B. Malik, Increased endothelial albumin permeability mediated by protein kinase C activation. J. Clin. Invest. 85, 1991-1998, (1990). 64. P. G. Nagpala, A. B. Malik, P. T. Vuong, and H. Lum, Protein kinase C beta 1 overexpression augments phorbol ester-induced increase in endothelial permeability. J. Cell. Physiol. 166, 249-255, (1996). 65. A. Hempel, C. Maasch, U. Heintze, C. Lindschau, R. Dietz, F. C. Luft, and H. Haller, High glucose concentrations increase endothelial cell permeability via activation of protein kinase C alpha. Circ. Res. 81, 363-371, (1997). 66. B. A. Wolf, J. R. Williamson, R. A. Easom, K. Chang, W. R. Sherman, and J. Turk, Diacylglycerol accumulation and microvascular abnormalities induced by elevated glucose levels. J. Clin. Invest. 87, 31-38, (1991).

10. PKC in Diabetic Retinal Vascular Abnormalities

201

67. J. E. Stasek, Jr., C. E. Patterson, and J. G. Garcia, Protein kinase C phosphorylates caldesmon 77 and vimentin and enhances albumin permeability across cultured bovine pulmonary artery endothelial cell monolayers. J. Cell. Physiol. 153, 62-75, (1992). 68. C. E. Turner, F. M. Pavalko, and K. Burridge, The role of phosphorylation and limited proteolytic cleavage of talin and vinculin in the disruption of focal adhesion integrity. J. Biol. Chem. 264, 11938-11944, (1989). 69. D. K. Werth, J. E. Niedel, and I. Pastan, Vinculin, a cytoskeletal substrate of protein kinase C. J. Biol. Chem. 258, 11423-11426, (1983). 70. M. Schliwa, T. Nakamura, K. R. Porter, U. Euteneuer, A tumor promoter induces rapid and coordinated reorganization of actin and vinculin in cultured cells. J. Cell. Biol. 99, 1045-1059, (1984). 71. L. P. Aiello, R. L. Avery, P. G. Arrigg, B. A. Keyt, H. D. Jampel, S. T. Shah, L. R. Pasquale, H. Thieme, M. A. Iwamoto, J. E. Park, H. V. Nguyen, L. M. Aiello, N. Ferrara, and G. L. King, Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N. Engl. J. Med. 331, 1480-1487, (1994). 72. H. Funatsu, H. Yamashita, Y. Nakanishi, and S. Hori, Angiotensin II and vascular endothelial growth factor in the vitreous fluid of patients with proliferative diabetic retinopathy. Br. J. Ophthalmol. 86, 311-315, (2002). 73. B. Williams, Factors regulating the expression of vascular permeability/vascular endothelial growth factor by human vascular tissues. Diabetologia 40, S118-S120, (1997). 74. P. Xia, L. P. Aiello, H. Ishii, Z. Y. Jiang, D. J. Park, G. S. Robinson, H. Takagi, W. P. Newsome, M. R. Jirousek, and G. L. King, Characterization of vascular endothelial growth factor’s effect on the activation of protein kinase C, its isoforms, and endothelial cell growth. J. Clin. Invest. 98, 2018-2026, (1996). 75. L. P. Aiello, S. E. Bursell, A. Clermont, E. Duh, H. Ishii, C. Takagi, F. Mori, T. A. Ciulla, K. Ways, M. Jirousek, L. E. Smith, and G. L. King, Vascular endothelial growth factorinduced retinal permeability is mediated by protein kinase C in vivo and suppressed by an orally effective beta-isoform-selective inhibitor. Diabetes 46, 1473-1480, (1997). 76. K. Suzuma, N. Takahara, I. Suzuma, K. Isshiki, K. Ueki, M. Leitges, L. P. Aiello, and G. L. King, Characterization of protein kinase C β isoform’s action on retinoblastoma protein phosphorylation, vascular endothelial growth factor-induced endothelial cell proliferation, and retinal neovascularization. Proc. Natl. Acad. Sci. USA 99, 721-726, (2002). 77. P. A. Campochiaro, Reduction of diabetic macular edema by oral administration of the kinase inhibitor PKC412. Invest. Ophthalmol. Vis. Sci. 45, 922-931, (2004). 78. M. E. Williams and K. R. Tuttle, The next generation of diabetic nephropathy therapies: an update. Adv. Chronic. Kidney Dis. 12, 212-22, (2005). 79. The PKC-DRS Study Group. The effect of ruboxistaurin on visual loss in patients with moderately severe to very severe nonproliferative diabetic retinopathy: initial results of the Protein Kinase C beta Inhibitor Diabetic Retinopathy Study (PKC-DRS) multicenter randomized clinical trial. Diabetes 54, 2188-97, (2005). 80. A. Nonaka, J. Kiryu, A. Tsujikawa, K. Yamashiro, K. Miyamoto, H. Nishiwaki, Y. Honda, and Y. Ogura, PKC-beta inhibitor (LY333531) attenuates leukocyte entrapment in retinal microcirculation of diabetic rats. Invest. Ophthalmol. Vis. Sci. 41, 2702-2706, (2000). 81. R. P. Danis, D. P. Bingaman, M. Jirousek, and Y. Yang, Inhibition of intraocular neovascularization caused by retinal ischemia in pigs by PKCbeta inhibition with LY333531. Invest. Ophthalmol. Vis. Sci. 38, 171-179, (1998).

202

J. K. Sun and G. L. King

82. L. P. Aiello, S. Bursell, and T. Devries, Protein kinase C beta selective inhibitor LY333531 ameliorates abnormal retinal haemodynamics in patients with diabetes. Diabetes 48, A19, (1999). 83. L. P. Aiello, S. E. Bursell, T. Devries, C. Alatorre, G. L. King, and D. K. Ways, Amelioration of abnormal retinal hemodynamics by a protein kinase C β selective inhibitor (LY333531) in patients with diabetes: results of a Phase 1 safety and pharmacodynamic clinical trial. IOVS 40, S192, (1999). 84. Eli Lilly and Co. Lilly will submit ruboxistaurin mesylate (Arxxant™) to FDA for treatment of diabetic retinopathy in 2005. Corporate News, August 2, 2005 (accessed at http://www.prnewswire.com/cgi-bin/micro_stories.pl?ACCT=916306&TICK=LLY&STORY=/ www/story/08-02-2005/0004080265&EDATE=Aug+2,+2005 on August 23, 2006). 85. L. P. Aiello, M. D. Davis, R. C. Milton, M. J. Sheetz, V. Arora, and L. Vignati, Initial results of the protein kinase C β inhibitor Diabetic Macular Edema Study (PKC-DMES). 18th International Diabetes Federation Congress. August 24-29, 2003. 86. Eli Lilly and Co. Data indicate that Lilly’s ruboxistaurin may have a potentially beneficial effect on diabetes-induced eye disease. Corporate news, Oct. 26, 2004 (accessed at http://www.prnewswire.com/cgi-bin/micro_stories.pl?ACCT=916306&TICK=LLY&STORY=/ www/story/10-26-2004/0002310518&EDATE=Oct+26,+2004 on August 23, 2006).

Chapter 11 EPH RECEPTOR TYROSINE KINASES: MODULATORS OF ANGIOGENESIS

Jin Chen,1,2,3 Dana Brantley-Siders,1 and John S. Penn 4 1

Department of Medicine, Division of Rheumatology, 2Department of Cancer Biology, Department of Cell and Developmental Biology, and 4Department of Ophthalmology and Visual Sciences, Vanderbilt University, Vanderbilt University School of Medicine, Nashville, Tennessee 3

Abstract:

1.

Angiogenesis, or the outgrowth of new sprouts from pre-existing vessels, involves a complex cascade of events. Among the diverse array of molecules involved in angiogenesis, receptor tyrosine kinases (RTKs) have emerged as critical mediators of neovascularization. This review will focus on the youngest family of essential vascular RTKs, the Eph receptors, and ephrins, their corresponding ligands. We will summarize our current understanding of Eph/ephrin function in vascular remodeling during embryogenesis and in neovascularization and tumorigenesis in adult tissues.

INTRODUCTION

Retinal neovascularization is the critical pathological component of a number of blinding conditions such as diabetes mellitus, retinopathy of prematurity, and age-related macular degeneration and is the leading cause of irreversible vision loss in developed countries.1 Although the present treatment, retinal laser photocoagulation, is partially effective, this procedure can destroy postmitotic retinal neurons and permanently affect visual function.2 During the last several years, a number of therapeutic agents have been developed, aiming at pharmacological inhibition of retinal angiogenesis. Angiogenesis, or the outgrowth of new sprouts from pre-existing vessels, involves a complex cascade of events (Reviewed by 3,4). First, the wall of the intact vessel loosens, reducing endothelial cell-smooth muscle cell 203 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 203–219. © Springer Science+Business Media B.V. 2008

204

J. Chen et al.

interaction, and endothelial cells are activated by angiogenic factors in response to hypoxia, ischemia, or developmental cues. The endothelial cells then invade the surrounding tissue through matrix degradation, proliferate and migrate toward the angiogenic stimulus, and coalesce into tubular structures. Finally, maturation of the new vessel is accomplished through recruitment of perivascular supporting cells and deposition of extracellular matrix. Each of these events is tightly regulated at the molecular level. In general, there is an upregulation of pro-angiogenic factors such as growth factors, integrins, and proteinases, and a downregulation of angiogenic inhibitors.5,6 Among the diverse array of molecules involved in angiogenesis, receptor tyrosine kinases (RTKs) have emerged as critical mediators of neovascularization.3,7 Vascular endothelial cell growth factor A (VEGF-A, hereafter referred to as VEGF), a RTK ligand, plays an essential role in hypoxia-induced proliferative retinopathy. VEGF expression is associated both temporally and spatially with the development of retinopathy in vivo.8-11 Suppression of VEGF activity by monoclonal antibodies, soluble receptor chimeric proteins, or antisense oligonucleotides inhibits angiogenesis in the retina.12,13 More recently, the angiopoietins and their Tie-2 receptors have been implicated in proliferative retinopathy.14,15 This review will focus on the youngest family of essential vascular RTKs, the Eph receptors, and their corresponding ligands. The present review aims at summarizing our current understanding of Eph/ephrin function in vascular remodeling during embryogenesis and neovascularization in adult tissues. Much of the data on post-natal angiogenesis has been obtained from tumor studies. Therefore, this review will provide an overview of Eph molecule function in tumor neovascularization and discuss the potential role of this family of RTKs in retinal angiogenesis.

2.

EPH RTKS AND EPHRIN LIGANDS

Eph receptors are unique RTKs that play critical roles in embryonic development and human diseases. First discovered in a human cDNA library screen for homologous sequences to the viral oncogene vfps, Eph receptors comprise the largest class of RTKs, and they display many unique features. The Eph family consists of at least 15 receptors and 9 ligands (Figure 1).16,17 The Eph receptors have been divided into two subclasses, A and B, according to sequence similarity and affinity to their ligands. In general, Eph class A (EphA) receptors bind to glycosylphosphatidyl inositol (GPI)anchored ephrin ligands (ephrin-A), while Eph class B receptors (EphB)

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

205

bind to ephrin ligands containing transmembrane domains (ephrin-B).17 Two exceptions have been found to these classes: EphA4 and EphB2. In addition to binding to ephrin-A ligands, EphA4 binds to ephrin-B2 and ephrin-B3.17 More recently, EphB2 was found to interact with ephrin-A5 in addition to ephrin-B ligands.18

A

B

Class B

Class A

PDZ

cell membrane P

P

GPI anchor Extracellular Space

ephrin

ephrin

Globular domain Cysteine rich region Fibronectin type III repeats P P

P P

Juxtamembrane domain

P P

P P

Kinase P

P

P

P

SAM domain PDZ

P

P

P

P

Eph

Figure 11-1. Domain structure and signaling of Eph receptors and ephrin ligands. Both GPIanchored ephrin-A and transmembrane ephrin-B ligands interact with the N-terminal globular domain of Eph receptors. The globular domain is followed by a cysteine-rich region and two fibronectin type III repeats, which contain a dimerization motif. Phosphorylated tyrosine residues provide docking sites for SH2 domain-containing signaling proteins. SAM domains form homodimers and may regulate receptor dimerization. Signaling proteins containing PDZ domains dock to the C-terminus of Eph receptors and ephrin-B ligands.

Consistent with other types of RTKs, both A and B Eph receptors contain a single transmembrane-spanning domain. The extracellular region of the Eph receptor is glycosylated and consists of a ligand binding domain containing immunoglobulin-like motifs, followed by a cysteine-rich domain and two fibronectin III-like repeats. The intracellular portion consists of a juxtamembrane region, a conserved tyrosine kinase domain, a sterile-α-motif

206

J. Chen et al.

(SAM) domain, and a PDZ binding motif (PSD-95 post synaptic density protein, Discs large, Zona occludens tight junction protein).19,20 The juxtamembrane region, kinase domain, and SAM domain contain several tyrosine residues, phosphorylation of which may create docking sites for interactions with signaling proteins containing SH2/SH3 (Src-Homology2/3) or PTB domains. The SAM domain has been implicated in mediating protein-protein interactions via the formation of homo- and hetero-typic oligomers. The PDZ binding motif binds to PDZ domain-containing proteins, which are thought to serve as scaffolds for the assembly of multiprotein signaling complexes at the membrane. Of the 9 known ligands, 6 are attached to the cell surface by a GPI linkage (class A) and 3 by transmembrane domains (class B). Analyses of amino acid sequences of ephrin ligands indicate that each ligand consists of a signal peptide at the amino-terminus, followed by a conserved receptorbinding region containing several cysteine residues and a spacer region. At the C-terminus, the A class ligands contain a hydrophobic region comprising the GPI linkage. In contrast, ephrin-B ligands contain a transmembrane domain and a cytoplasmic domain that, like the Eph receptors themselves, contains a PDZ-binding motif and conserved tyrosine residues that may be phosphorylated and serve as docking sites for proteins containing SH2/SH3 or PTB domains. These structural motifs control ligand attachment, receptor and ligand clustering, and regulate the binding of ephrins to specific Eph receptors to elicit distinct biological responses (Figure 1). Compared to other RTKs, Eph receptor signaling is unique due to the bidirectional signals activated by both receptor and ligand. Therefore, both the cell containing the receptor and the adjacent cell containing the ligand receive signals upon receptor-ligand binding. In the unbound state, the Eph tyrosine kinase domain is distorted by the helical conformation of the juxtamembrane region that renders the kinase domain inactive. Cell-cell contact allows Eph receptors to bind to their ligands and trigger a series of events that lead to receptor activation. The high-affinity binding of Eph-ephrin heterodimers causes them to cluster together on the cell surface.21 This heterodimeric conformation is thought to allow the transphosphorylation of the juxtamembrane domain, altering its inhibitory helical structure and allowing the kinase domain to be activated by phosphorylation of the activation loop. The newly phosphorylated tyrosines are then able to interact with various downstream signaling effectors. Eph receptor forward signaling and ephrin ligand reverse signaling will not be reviewed in detail here, but interested readers are referred to the excellent body of recent literature.22,23

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

3.

207

ROLE OF EPH/EPHRIN IN VASCULAR REMODELING DURING EMBRYONIC DEVELOPMENT

Many Eph RTK family members are expressed in the vasculature during embryogenesis. In mice, Xenopus, and chick, ephrin-B2 is expressed in endothelial cells in the extraembryonic yolk sac primary capillary plexus, in large arteries within the embryo, and in the endocardium of the developing heart.24-28 The principal receptor for ephrin-B2, EphB4, displays a reciprocal expression pattern in embryonic veins in the yolk sac, larger veins including the arterial cardinal vein and vitelline vein, and also in endocardium.24,26,28,29 These expression patterns provided the first evidence for a molecular distinction between arterial and venous endothelium. Although the Eph/ephrin system was originally identified in the nervous system, genetic studies using knockout mice have firmly established the role of these molecules in vascular development. Targeted disruption of either ephrin-B2 or EphB4 results in embryonic lethality at E11 or E9.5-10, respectively, due to strikingly similar defects in angiogenic remodeling of both arteries and veins as well as patterning defects in cardiac myocardium.28,29 Although vasculogenesis occurs normally in homozygous null embryos, with the formation of primitive capillary network structures, these networks fail to remodel and branch into more complex vascular networks. Similar angiogenic remodeling defects within intersomitic vessels are produced by overexpression of dominant-negative EphB4 in Xenopus embryos.26 Interestingly, overexpression of both full length ephrin-B ligands and their corresponding cytoplasmic domain deletion mutants recapitulates this phenotype, suggesting that remodeling of intersomitic vessels in the Xenopus trunk occurs through forward signaling rather than reverse.26 Reverse signaling through ephrin-B2, however, can affect angiogenesis in mice, as demonstrated by replacement of the endogenous ephrin-B2 gene with a cytoplasmic deletion mutant. Much like null mutants, ephrin-B2ΔC/ΔC “knockin” mutants display defects in remodeling of vessels in the yolk sac and in the embryo, as well as recapitulating heart defects.30 These data suggest that ephrins and Eph receptors have more complex functions in mammals than in lower vertebrates. Although vasculogenesis appears to occur normally in EphB4-deficient mice,29 a recent report suggests that EphB4 may still modulate differentiation of hemangioblasts in cooperation with other pro-angiogenic factors. Wang et al. reported that EphB4-deficient embryoid bodies display delayed expression of the hemangioblast marker VEGFR-2/Flk-1, as well as defective vascular morphogenesis, in response to VEGF and bFGF in vitro.31 These data suggest that EphB RTKs and ephrin signaling may

208

J. Chen et al.

modulate vasculogenesis and might explain why EphB4-deficient embryos die sooner than ephrin-B2-null mutants.28,29 Thus, ephrins might regulate sensitivity to early vascular developmental cues in addition to exerting direct effects on angiogenic remodeling of the embryonic vasculature. In addition to its expression in arterial endothelium, ephrin-B2 has been detected in mesenchyme surrounding some blood vessels, and it begins to be expressed in smooth muscle cells and pericytes surrounding vessels as development proceeds.26,29,32,33 Thus, mesenchymal expression of ephrin-B2 may also affect morphogenesis of EphB RTK-expressing endothelium. This hypothesis is supported by experiments in which ephrin-B2 is differentially overexpressed either ubiquitously or in endothelial cells specifically. Defective patterning of intersomitic vessels and defective outgrowth of venous vessels in the head region were observed in transgenic embryos overexpressing ephrin-B2 ubiquitously, but not in endothelial-specific, Tie-2 promoter-driven ephrin-B2 transgenics.34 The embryos ubiquitously overexpressing ephrin-B2 also display neonatal lethality due to aortic aneurysms that result from lack of vascular smooth muscle cell recruitment, and smooth muscle cell outgrowth and migration were impaired in ascending aortic explants relative to wild-type control littermates.34 Again, these defects were not observed in Tie-2p-ephrin-B2 transgenics overexpressing ephrin-B2 in endothelium. Tissue-specific deletion of ephrin-B2 in endothelium and endocardium, however, was sufficient to recapitulate angiogenic remodeling defects observed in conventional knockout animals.35 Since the full complement of vascular defects was produced by deletion of ephrin-B2 in endothelium, although mesenchymal expression remained intact, the authors argued that mesenchymal ephrin-B2 is not sufficient for vessel remodeling. Ephrin-B2 might, however, be necessary for proper remodeling, as demonstrated by several in vitro studies. Zhang et al. showed that mesenchymal expression of ephrin-B2 enhances differentiation of paraaortic splanchnopleuric mesoderm and endothelial precursor-enriched cell populations within this tissue into endothelium, whereas overexpression of EphB4 was inhibitory.36 Differentiation induced by mesenchymal ephrin-B2 was accompanied by morphogenesis into cordlike tubules and enhanced smooth muscle cell recruitment, demonstrating the importance of mesenchymal ephrin-B2 in vascular morphogenesis and maturation. Inhibition of vascular morphogenesis by mesenchymal EphB4 may be due to the anti-adhesive, repulsive effects of this receptor. Füller et al. recently reported that treatment with soluble ephrin-B2-Fc, a soluble form of the ephrin-B2 ectodomain, inhibited attachment of EphB4-positive endothelial cells, as well as induced detachment of three-dimensional spheroids and delamination of endothelial cells from umbilical vein explants.37 Treatment with EphB4-Fc had the opposite effect, inducing

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

209

migration and sprouting of endothelial cells. Similar anti-adhesive, antimigratory effects were reported in primary mouse microvascular endothelial cells in response to ephrin-B2-Fc.38 In addition, EphB4-positive endothelial cells co-cultured with endothelial cells expressing either full-length or a cytoplasmic deletion mutant of ephrin-B2 resulted in segregation of these cells. These data suggest that forward signaling through EphB4 restricts interaction with ephrin-B2 expressing cells, which could assist separation of arterial and venous domains in vascular morphogenesis.37 Ephrin-B2 and EphB4 are not the only Eph family members that regulate embryonic vessel patterning. Ephrin-A1 is expressed in the developing vasculature and promotes angiogenesis in vitro and in vivo.39-44 Though no data are yet available concerning the role of ephrin-A1 in embryonic angiogenesis, this ligand and its principal receptor, EphA2, are known to regulate post-natal angiogenesis as discussed in the next section. Ephrin-B1 is also expressed in the embryonic vasculature, in both arteries and veins, as is EphB3. In addition, EphB2 is expressed in vascular-associated mesenchyme.24 Although targeted disruption of EphB2 or EphB3 alone produced no discernable vascular phenotype, approximately 30% of double mutants die at E11 due to vascular remodeling defects in the head, heart, and intersomitic regions of the embryo, demonstrating that these EphB RTKs also participate in developmental angiogenesis.24 Although ephrin-B1 expression cannot compensate for the loss of ephrin-B2 in null mutants, in vitro studies have demonstrated that this ligand can induce angiogenic responses in cultured endothelial cells,24,45-47 as can reverse signaling through ephrin-B1.48 These data suggest that the ephrin-B1 ligand might also be necessary, though not sufficient, for vascular remodeling during embryogenesis. The vascular phenotypes of Eph/ephrin knockout and transgenic animals are summarized in Table 1.

210

J. Chen et al.

Table 11-1. Phenotypes in Eph/ephrin transgenic (Tg) or knockout mice Tg/knockout mice ephrinB2-/-

ephrinB2ΔC/ΔC EphB3-/EphB2-/EphB2-/-EphB3-/-

EphB4-/Tie2-Cre;ephrinB2fl/fl

CAGp-ephrinB2

Tie-2p-ephrinB2

EphA2-/-

4.

Phenotype Embryonic lethal, die at E10.5; Defective vessel remodeling and sprouting; Heart trabeculation defects; Defects in guidance of migrating cranial neural crest cells Defects in angiogenic remodeling similar to those observed in ephrinB2-/-; No guidance defects of neural crest cell migration No vascular defects; Defects in the formation of corpus callosum No vascular defects; Defects in pathfinding of commisural axons Embryonic lethal, die at E10.5 (~30%); Defects in neural development; Defective vessel remodeling, similar to that observed in ephrinB2-/Embryonic lethal, die at E10.5; Defective vessel remodeling, similar to that observed in ephrinB2-/Endothelial-specific ephrin-B2 knock out; Angiogenic remodeling defects identical to those in ephrinB2-/Ephrin-B2 Tg driven by CMV promoter; Neonatal lethality due to aortic aneurysms; Defective vascular patterning Endothelial-specific ephrin-B2 Tg; Aortic aneurysms or defective vascular patterning not detected; Some Tgs show intracerebral bleeding No overt defects during embryogenesis; Impairment in endothelial cell migration and assembly and tumor angiogenesis

Reference [28, 30]

[30]

[24, 73] [24, 74] [24]

[29]

[35]

[34]

[34]

[51]

ROLE OF EPH/EPHRIN IN POST-NATAL ANGIOGENESIS

Eph RTKs and ephrins also regulate post-natal angiogenic remodeling. Expression of ephrin-B2 persists in adult arterial endothelium and vascular smooth muscle cells surrounding arteries, and EphB4 expression persists in adult venous endothelium, suggesting that this ligand-receptor pair may regulate boundary maintenance and/or vascular remodeling in mature tissues.32,33 Indeed, in cultured endothelial cells, soluble ephrin-B2 facilitates adhesion and migration, processes critical for angiogenic remodeling.49

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

211

Eph RTKs and ephrins can induce post-natal vascular remodeling. For example, soluble ephrin-A1,40,43 ephrin-B2,50 and the ectodomain of EphB1 [48] induce corneal angiogenesis in adult mice, demonstrating that these mature endothelial cells have the capacity to respond to ephrin and Eph RTK signals. In addition, ephrin-B2 and ephrin-A1 can also induce an angiogenic response from subcutaneous vessels in vivo.50,51 Matrigel plugs harboring soluble ephrin-B2 induced an angiogenic response from subcutaneous host vessels when implanted into mice.50 Similarly, surgical sponges impregnated with soluble ephrin-A1 and implanted in the subcutaneous dorsal flank of wild-type mice induced sprouting of adjacent subcutaneous vessels and infiltration of new vessel sprouts into the sponges. When ephrin-A1containing sponges were introduced into EphA2-deficient mice, this angiogenic response was greatly diminished, suggesting that efficient vascular remodeling in response to ephrin-A1 requires EphA2 RTK.51 Lung microvascular endothelial cells isolated from adult mice can also respond to ephrin-A1, which induces assembly and migration in vitro.39,51 These processes are dependent upon expression of EphA2 RTK, since endothelial cells isolated from EphA2-deficient mice display impaired angiogenic responses to ephrin-A1 and these responses are rescued upon restoration of EphA2 expression.51 The female reproductive system is a major site for physiological angiogenesis. Ephrin-A1 expression has been observed in normal human endometrial epithelial cells.52 Given the pro-angiogenic activity of this ligand, it is possible that ephrin-A1 could contribute to normal endometrial angiogenesis. Ephrin-A1 expression is downregulated in biopsies from endometriosis tissues relative to normal endometrial tissue,53 however, it is unclear how this molecule might regulate pathogenic angiogenesis in this disease. B class Eph RTKs and ephrins may play a more direct role in endometrial disease, as EphB4 and ephrin-B2 overexpression was recently reported in human endometrial hyperplasias and carcinomas.54,55 Since B class ephrins, such as ephrin-B2, are expressed in vascular beds of mature tissues and in tumor xenografts,32 it is possible that EphB4 overexpression could affect neovascularization of endometrial cancer. Further investigation of this hypothesis could facilitate development of new therapies for endometrial cancer, since high microvascular density correlates with advanced disease and a poor patient survival rate.56 Cyclic vascular remodeling also occurs in the ovary during formation of the corpus luteum, a transient structure that provides progesterone to establish and sustain pregnancy. After ovulation, granulosa and theca cells of the developing follicle and steroid-producing cells in the corpus luteum produce pro-angiogenic factors that promote vascularization of the corpus luteum.57 Ephrin-B2 expression has been observed in murine corpus luteum

212

J. Chen et al.

vessels.32 Moreover, a recent report correlates expression of B class Eph RTKs and ephrins with corpus luteum formation in humans. EphB1, B2, and B4, as well as ephrin-B1 and -B2 mRNA expression was detected in human ovary samples.58 Interestingly, ephrin-B1 expression was localized to theca and granulosa cells during the window of corpus luteum formation and neovascularization. Given the pro-angiogenic functions of this molecule in vitro, it would be interesting to determine if ephrin-B1 regulates angiogenesis in the ovary, particularly since overexpression of ephrin-B1 has also been detected in human ovarian carcinomas.59

5.

ROLE OF EPH/EPHRIN IN TUMOR ANGIOGENESIS

Although it is now clear that Eph receptors and ephrin ligands play a critical role in vascular development during embryogenesis, the function of these molecules in pathological angiogenesis has just begun to be investigated. A survey of expression patterns of Eph molecules in tumor vasculature reveals that the ephrin-A1 and EphA2 ligand-receptor pair is consistently expressed in tumor-associated endothelium in a variety of tumors, including tumor xenografts (MDA-MB-435 human breast cancer and KS1767 human Kaposi’s sarcoma) and human tumor specimens (lung anaplastic adenocarcinoma and squamous carcinoma, gastric cancer, colon carcinoma, and kidney clear cell carcinoma).60 Expression patterns of ephrin-A1 and EphA2 in two murine tumor models that are angiogenesis-dependent, the RIP-Tag islet carcinoma transgenic model and the 4T1 transplantable metastatic mammary carcinoma model, have also been determined.39 EphrinA1 ligand was expressed predominantly in tumor tissue, and EphA2 receptor expression was mainly associated with the tumor vasculature. In addition, a soluble EphA2 receptor inhibited tumor neovascularization in a dorsal vascular window assay. These data suggest a role of ephrin-A/EphA molecules in promoting angiogenesis in tumors. The first functional evidence that Eph receptors regulate tumor angiogenesis came from studies using soluble EphA-Fc receptors. A soluble EphA2-Fc or EphA3-Fc receptor that blocks endogenous EphA receptor signaling inhibited tumor growth and angiogenesis in murine 4T1 malignant mammary carcinomas.39 In addition, local or systemic delivery of soluble EphA-Fc receptors also inhibited angiogenic islet formation and reduced tumor volume in multi-stage pancreatic carcinomas in RIP-Tag transgenic mice,61 indicating a possible relevance for EphA targeting in clinical cancer therapeutics. Because soluble EphA receptors can interact with multiple ephrin ligands and block multiple EphA receptor signaling pathways, the

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

213

precise A class receptor(s) that function in the tumor endothelium remains undefined. Since EphA2 is a major class A Eph receptor expressed in the adult endothelium (Bowen and Chen, unpublished results), EphA2-deficient mice were studied further. Gene disruption of EphA2 did not affect survival or major embryonic developmental events,51,62,63 reflecting the fact that the EphA2 receptor is not expressed in the embryonic vasculature.42 However, EphA2 receptors are expressed in adult endothelium and tumor neovasculature. Accordingly, EphA2-null endothelial cells failed to undergo cell migration and vascular assembly both in vitro and in vivo.51 In addition, tumor growth, angiogenesis, and metastasis were significantly inhibited in EphA2-deficient receipt mice.64 These data suggest that host EphA2 RTK function is required in the tumor microenvironment for tumor angiogenesis and metastatic progression. In addition to the function of EphA receptors in tumor neovascularization, ephrin-B2 expression has been observed in tumor arterioles infiltrating transplanted Lewis lung carcinomas and B16 melanomas in mice,33 suggesting a role in tumor angiogenesis. However, modulation of ephrin-B2 or EphB4 signaling separately has pleiotropic effects on tumor progression. For example, overexpression of ephrin-B2 in human colorectal cell increases tumor vessel density but suppresses tumor growth.65 Also, expression of a cytoplasmic truncation mutant of EphB4 in breast cancer cells promotes tumor angiogenesis and tumor growth.66 Nevertheless, blocking of bidirectional signaling between ephrin-B2 and EphB4 by a soluble EphB4 monomer effectively inhibited A375 melanoma growth and tumor angiogenesis.67 Taken together, these data suggest that ephrin-B2 or EphB4 could be targets for inhibition of tumor neovascularization, but blocking of bidirectional signaling would be a prerequisite of developing these targets for therapeutic intervention. The precise mechanism of how Eph/ephrin signaling regulates tumor angiogenesis is not known. However, from the available data it is conceivable that Eph/ephrin-dependent tumor neovascularization is mediated by the interplay of Eph receptors and ephrin ligands expressed by tumor cells and endothelial cells. During the initial phase of the angiogenic response, activation of EphA2 receptors on vascular smooth muscle induces retraction of perivascular supporting cells via inhibition of the Rac1 GTPase/PAK pathway,68 allowing endothelial cells to respond to angiogenic cues. In contrast, the signaling of EphA2 receptors on endothelial cells stimulates PI3 kinase-dependent Rac1 GTPase activation, promoting endothelial cell migration and vessel assembly.51 Ephrins expressed on the tumor cells may function as contact-dependent organizing molecules to guide incoming vessels that express EphA2 receptor. Alternatively, angiogenic factors such as VEGF or TNF-α in the tumor microenvironment may induce the expression

214

J. Chen et al.

and/or activation of ephrins in endothelial cells, as suggested in studies from cultured endothelial cells.40 The ephrins may then interact with Eph receptors on adjacent endothelial cells to promote endothelial cell sprouting, migration, and capillary tube formation through bidirectional signaling. Furthermore, the interactions between tumor cells and host blood vessels may provide a mechanism for the intravasation of tumor cells into the blood stream, facilitating tumor metastasis. Regardless of the mechanism, Eph receptors and ephrin ligands are attractive candidates for tumor prognostic markers and potential targets for therapeutic intervention in cancer.

6.

ROLE OF EPH/EPHRIN IN RETINAL ANGIOGENESIS

Retinopathy of prematurity, as well as diabetic retinopathy, neovascular glaucoma, and age-related macular degeneration, involves abnormal ocular angiogenesis. Large numbers of new blood vessels lead to blindness by disturbing the refractive surface of the ocular epithelium.69 Expression of ephrin-B2 and relatively lower levels of EphB4 has been observed in human retinal endothelial cells.70 In addition, reverse signaling through B class ephrins can affect retinal endothelial cells in culture, a fact made evident when treatment with the soluble EphB4-Fc receptor ectodomain induced retinal endothelial proliferation and migration via activation of PI3K, nitric oxide synthase, and ERK1/2.70 These in vitro cell culture data are consistent with recent findings in human retina with pathological angiogenesis.71 Ephrin-B2 was expressed in a significant percentage of the endothelial cells of fibroproliferative membranes that were obtained from patients with proliferative diabetic retinopathy (65%) and retinopathy of prematurity (25%). While the expression of EphB4 receptor was not detected in these clinical samples, EphB2 and EphB3 receptors were synthesized in these proliferative membranes. Given their functions in developmental angiogenesis, these data suggest that B class Eph RTKs and ephrins could contribute to pathogenic neovascularization in the retina. Recently, new functional data demonstrate that A class Eph RTKs and ephrins participate in retinopathy of prematurity.72 EphA2 is expressed in human retinal endothelial cells, and treatment with soluble EphA2-Fc, which inhibits activation of endogenous EphA RTKs in response to ephrin-A stimulation, reduces retinal neovascularization in a rat retinopathy of prematurity model. Inhibition of disease severity likely targets EphA2expressing retinal endothelium, since soluble EphA-Fc receptors have been shown to inhibit ephrin-induced migration and sprouting of endothelial cells in culture.39,40 In addition, soluble EphA receptors may also block the effects

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

215

of VEGF, which is known to regulate angiogenesis in retinopathy,69 since they inhibit VEGF-induced migration, sprouting, and corneal angiogenesis to a degree similar to ephrin-A1.40 In addition, EphA2-deficient endothelial cells display impaired VEGF-induced migration and vascular assembly in vitro. These data suggest that targeting EphA receptor function could provide a novel avenue for treatment of retinopathy. Investigation of Eph RTK function in diabetic retinopathy and age-related macular degeneration may also enhance our understanding of the molecular mechanisms underlying these diseases.

7.

PERSPECTIVES

Biochemical and genetic analyses of Eph RTK function have demonstrated that members of the Eph family are critical regulators of angiogenesis both during embryogenesis and under pathological conditions in adult tissues. However, bidirectional signaling through promiscuous receptor-ligand interactions complicates the dissection of mechanisms of Eph molecule action. A better understanding of Eph receptor and ephrin ligand expression profiles and their cis and trans interactions in vitro and in vivo would greatly advance the field. The angiogenic functions of Eph RTKs in disease make these molecules attractive targets for anti-angiogenic therapy. Data from animal models of retinopathy and cancer suggest that targeting EphA RTKs may reduce pathological angiogenesis associated with ocular diseases. Because soluble Eph receptors globally inhibit signaling through multiple Eph RTKs and ephrins, and because these reagents could inadvertently initiate reverse signaling through multiple ephrins, it will be necessary to target individual Eph family members to avoid any potential negative side effects that may occur with global inhibitors. Analysis of EphA2-deficient endothelial cells suggests that this particular RTK would make an excellent target.51,64 Further characterization of animals deficient in one or multiple members of the Eph family should provide valuable information on the therapeutic potential of these targets.

ACKNOWLEDGMENTS This work was supported by National Institutes of Health grant CA95004, Juvenile Diabetes Foundation grant 1-2001-519, and Department of Defense grant DAMD17-02-1-0604 to J. C.; American Heart Association postdoctoral

216

J. Chen et al.

fellowship 0120147B and Department of Defense postdoctoral fellowship DAMD17-03-1-0379 to D. B.-S.; and NIH grant EY07533 and the Lew R. Wasserman Merit Award for Research to Prevent Blindness to J. P.

REFERENCES 1. P. Lee, C. C. Wang, and A. P. Adamis, Ocular neovascularization: An epidemiological review, Surv. Ophthalmol. 43, 245-269 (1998). 2. M. A. Speicher et al., Pharmacologic therapy for diabetic retinopathy, Expert Opin. Emerg. Drugs 8, 239-250 (2003). 3. G. D. Yancopoulos et al., Vascular-specific growth factors and blood vessel formation, Nature 407, 242-248 (2000). 4. A. W. Griffioen and G. Molema, Angiogenesis: potentials for pharmacologic intervention in the treatment of cancer, cardiovascular diseases, and chronic inflammation, Pharmacol. Rev. 52 (2), 237-268 (2000). 5. G. Bergers and L. E. Benjamin, Tumorigenesis and the angiogenic switch, Nat. Rev. Cancer 3 (6), 401-410 (2003). 6. R. Kerbel and J. Folkman, Clinical translation of angiogenesis inhibitors, Nature Review/Cancer 2, 727-739 (2002). 7. N. W. Gale and G. D. Yancopoulos, Growth factors acting via endothelial cell-specific receptor tyrosine kinases: VEGFs, angiopoietins, and ephrins in vascular development, Genes & Development 13, 1055-1066 (1999). 8. M. Boulton et al., VEGF localisation in diabetic retinopathy, Br. J. Ophthalmol. 82, 561-568 (1998). 9. E. A. Pierce et al., Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neovascularization, Proc. Natl. Acad. Sci. USA 92, 905-909 (1995). 10. S. G. Robbins et al., Detection of vascular endothelial growth factor (VEGF) protein in vascular and non-vascular cells of the normal and oxygen-injured rat retina, Growth Factor 14, 279-295 (1997). 11. S. G. Robbins, V. S. Rajaratnam, and J. S. Penn, Evidence for upregulation and redistribution of vascular endothelial growth factor receptors flt-1 and flk-1 in the oxygen-injured rat retina, Growth Factors 16, 1-9 (1998). 12. L. P. Aiello et al., Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins, Proc. Natl. Acad. Sci. USA 92, 10457-10461 (1995). 13. G. S. Robinson et al., Oligodeoxynucleotides inhibit retinal neovascularization in a murine model of proliferative retinopathy, Proc. Natl. Acad. Sci. USA 93, 4851-4856 (1996). 14. M. Hangai et al., Systemically expressed soluble Tie2 inhibits intraocular neovascularization, Hum. Gene Ther. 12, 1311-1321 (2001). 15. H. Takagi et al., Potential role of the angiopoietin/tie2 system in ischemia-induced retinal neovascularization, Invest. Ophthalmol. Vis. Sci. 44, 393-402 (2003). 16. E. N. Committee, Unified nomenclature for Eph family receptors and their ligands, the Ephrins, Cell 90, 403-404 (1997). 17. N. W. Gale et al., Eph receptors and ligands comprise two major specificity subclasses and are reciprocally compartmentalized during embryogenesis, Neuron 17, 9-19 (1996).

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

217

18. J. P. Himanen et al., Repelling class discrimination: ephrin-A5 binds to and activates EphB2 receptor signaling, Nat. Neurosci. 7, 501-509 (2004). 19. K. Bruckner et al., EphrinB ligands recruit GRIP family PDZ adaptor proteins into raft membrane microdomains, Neuron 22, 511-524 (1999). 20. R. Torres, B. L. Firestein, H. Dong, J. Staudinger, E. N. Olson, R. L. Huganir, and G. D. Yancopoulos, PDZ proteins bind, cluster, and synaptically colocalize with Eph receptors and their ephrin ligands, Cell 21, 1453-1463 (1998). 21. J. P. Himanen et al., Crystal structure of an Eph receptor-ephrin complex, Nature 414, 933-938 (2001). 22. K. Kullander and R. Klein, Mechanisms and functions of Eph and ephrin signaling, Nat. Rev. Mol. Cell Biol. 3, 475 (2002). 23. N. K. Noren and E. B. Pasquale, Eph receptor-ephrin bidirectional signals that target Ras and Rho proteins, Cell Signal. 16, 655-666 (2004). 24. R. H. Adams et al., Roles of ephrinB ligands and EphB receptors in cardiovascular development: demarcation of arterial/venous domains, vascular morphogenesis, and sprouting angiogenesis, Genes & Development 3, 295-306 (1999). 25. R. K. Baker and P. B. Antin, Ephs and ephrins during early stages of chick embryogenesis, Dev. Dyn. 228 (1), 128-142 (2003). 26. P. M. Helbling, D. M. Saulnier, and A. W. Brandli, The receptor tyrosine kinase EphB4 and ephrin-B ligands restrict angiogenic growth of embryonic veins in Xenopus laevis, Development 127 (2), 269-278 (2000). 27. K. Othman-Hassan et al., Arterial identity of endothelial cells is controlled by local cues, Dev. Biol. 237 (2), 398-409 (2001). 28. H. U. Wang, Z. F. Chen, and D. J. Anderson, Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4, Cell 93, 741-753 (1998). 29. S. S. Gerety, et al., Symmetrical mutant phenotypes of the receptor EphB4 and its specific transmembrane ligand ephrin-B2 in cardiovascular development, Molecular Cell 4, 403-414 (1999). 30. R. H. Adams et al., The cytoplasmic domain of the ligand ephrinB2 is required for vascular morphogenesis but not cranial neural crest migration, Cell 104, 57-69 (2001). 31. Z. Wang et al., Ephrin receptor, EphB4, regulates ES cell differentiation of primitive mammalian hemangioblasts, blood, cardiomyocytes, and blood vessels, Blood 4, 4 (2003). 32. N. W. Gale, et al., Ephrin-B2 selectively marks arterial vessels and neovascularization sites in the adult, with expression in both endothelial and smooth-muscle cells, Dev. Biol. 230, 151-160 (2001). 33. D. Shin et al., Expression of ephrinB2 identifies a stable genetic difference between arterial and venous vascular smooth muscle as well as endothelial cells, and marks subsets of microvessels at sites of adult neovascularization, Dev. Biol. 230 (2), 139-150 (2001). 34. Y. Oike et al., Regulation of vasculogenesis and angiogenesis by EphB/ephrin-B2 signaling between endothelial cells and surrounding mesenchymal cells, Blood 100 (4), 1326-1333 (2002). 35. S. S. Gerety and D. J. Anderson, Cardiovascular ephrinB2 function is essential for embryonic angiogenesis, Development 129 (6), 1397-1410 (2002). 36. X. Q. Zhang et al., Stromal cells expressing ephrin-B2 promote the growth and sprouting of ephrin-B2(+) endothelial cells, Blood 98 (4), 1028-1037 (2001). 37. T. Fuller et al., Forward EphB4 signaling in endothelial cells controls cellular repulsion and segregation from ephrinB2 positive cells, J. Cell Sci. 116 (Pt 12), 2461-2470 (2003).

218

J. Chen et al.

38. K. Hamada et al., Distinct roles of ephrin-B2 forward and EphB4 reverse signaling in endothelial cells, Arterioscler. Thromb. Vasc. Biol. 23 (2), 190-197 (2003). 39. D. M. Brantley et al., Soluble EphA receptors inhibit tumor angiogenesis and progression in vivo, Oncogene 21, 7011-7026 (2002). 40. N. Cheng et al., Blockade of EphA receptor tyrosine kinase activation inhibits VEGFinduced angiogenesis, Mol. Cancer Res. (formerly Cell Growth and Differentiation) 1, 2-11 (2002). 41. T. O. Daniel et al., Elk and LERK-2 in developing kidney and microvascular endothelial assembly, Kidney Int. Suppl. 57, S73-S81 (1996). 42. J. L. McBride and J. C. Ruiz, ephrinA1 is expressed at sites of vascular development in the mouse, Mech. Dev. 77, 201-204 (1998). 43. A. Pandey et al., Role of B61, the ligand for the Eck receptor tyrosine kinase, in TNFalpha-induced angiogenesis, Science 268 (5210), 567-569 (1995). 44. H. Takahashi and T. Ikeda, Molecular cloning and expression of rat and mouse B61 gene: implications on organogenesis, Oncogene 11 (5), 879-883 (1995). 45. K. Nagashima et al., Adaptor protein Crk is required for ephrin-B1-induced membrane ruffling and focal complex assembly of human aortic endothelial cells, Mol. Biol. Cell 13 (12), 4231-4242 (2002). 46. Y. Sawai et al., Expression of ephrin-B1 in hepatocellular carcinoma: possible involvement in neovascularization, J. Hepatol. 39 (6), 991-996 (2003). 47. E. Stein et al., Eph receptors discriminate specific ligand oligomers to determine alternative signaling complexes, attachment, and assembly responses, Genes Dev. 12 (5), 667-678 (1998). 48. U. Huynh-Do et al., Ephrin-B1 transduces signals to activate integrin-mediated migration, attachment and angiogenesis, J. Cell Sci. 115, 3073-3081 (2002). 49. C. Vindis et al., EphB1 recruits c-Src and p52Shc to activate MAPK/ERK and promote chemotaxis, J. Cell Biol. 162 (4), 661-671 (2003). 50. H. Maekawa et al., Ephrin-B2 induces migration of endothelial cells through the phosphatidylinositol-3 kinase pathway and promotes angiogenesis in adult vasculature, Arterioscler. Thromb. Vasc. Biol. 23 (11), 2008-2014 (2003). 51. D. Brantley-Sieders et al., EphA2 receptor tyrosine kinase regulates endothelial cell migration and assembly through phosphoinositide 3-kinase-mediated Rac1 GTPase activation, J. Cell Sci. 117, 2037-2049 (2004). 52. H. Fujiwara et al., Human endometrial epithelial cells express ephrin A1: possible interaction between human blastocysts and endometrium via Eph-ephrin system, J. Clin. Endocrinol. Metab. 87 (12), 5801-5807 (2002). 53. L. C. Kao et al., Expression profiling of endometrium from women with endometriosis reveals candidate genes for disease-based implantation failure and infertility, Endocrinology 144 (7), 2870-2881 (2003). 54. G. Berclaz et al., Activation of the receptor protein tyrosine kinase EphB4 in endometrial hyperplasia and endometrial carcinoma, Ann. Oncol. 14, 220-226 (2003). 55. N. Takai et al., Expression of receptor tyrosine kinase EphB4 and its ligand ephrin-B2 is associated with malignant potential in endometrial cancer, Oncol. Rep. 8, 567-573 (2001). 56. S. Ozuysal et al., Angiogenesis in endometrial carcinoma: correlation with survival and clinicopathologic risk factors, Gynecol. Obstet. Invest. 55 (3), 173-177 (2003). 57. J. S. Davis, B. R. Rueda, and K. Spanel-Borowski, Microvascular endothelial cells of the corpus luteum, Reprod. Biol. Endocrinol. 1 (1), 89 (2003). 58. M. Egawa et al., Ephrin B1 is expressed on human luteinizing granulosa cells in corpora lutea of the early luteal phase: the possible involvement of the B class Eph-ephrin system during corpus luteum formation, J. Clin. Endocrinol. Metab. 88 (9), 4384-4392 (2003).

11. Eph Receptor Tyrosine Kinases: Modulators of Angiogenesis

219

59. M. E. Schaner et al., Gene expression patterns in ovarian carcinomas, Mol. Biol. Cell 14 (11), 4376-4386 (2003). 60. K. Ogawa et al., The ephrin-A1 ligand and its receptor, EphA2, are expressed during tumor neovascularization, Oncogene 19, 6043-6052 (2000). 61. N. Cheng et al., Inhibition of VEGF-dependent multi-stage carcinogenesis by soluble EphA receptors, Neoplasia 5, 445-456 (2003). 62. J. Chen et al., Germline inactivation of the murine Eck receptor tyrosine kinase by retroviral insertion, Oncogene 12, 979-988 (1996). 63. C. M. Naruse-Nakajima, M. Asano, and Y. Iwakura, Involvement of EphA2 in the formation of the tail notochord via interaction with ephrinA1, Mech. Dev. 102, 95-105 (2001). 64. D. M. Brantley-Sieders et al., Impaired tumor microenvironment in EphA2-deficient mice inhibits tumor angiogenesis and metastatic progression, FASEB J. 19, 1884-1886 (2005). 65. W. Liu et al., Effects of overexpression of ephrin-B2 on tumour growth in human colorectal cancer, Br. J. Cancer 90, 1620-1626 (2004). 66. N. K. Noren et al., Interplay between EphB4 on tumor cells and vascular ephrin-B2 regulates tumor growth, Proc. Natl. Acad. Sci. U S A. 101, 5583-5588 (2004). 67. G. Martiny-Baron et al., Inhibition of tumor growth and angiogenesis by soluble EphB4, Neoplasia 6, 248-257 (2004). 68. C. Deroanne et al., EphrinA1 inactivates integrin-mediated vascular smooth muscle cell spreading via the Rac/PAK pathway, J. Cell Sci. 116, 1367-1376 (2003). 69. A. P. Adamis, L. P. Aiello, and R. A. D’Amato, Angiogenesis and ophthalmic disease, Angiogenesis 3 (1), 9-14 (1999). 70. J. J. Steinle et al., Role of ephrin B2 in human retinal endothelial cell proliferation and migration, Cell. Signal. 15 (11), 1011-1017 (2003). 71. N. Umeda et al., Expression of ephrinB2 and its receptors on fibroproliferative membranes in ocular angiogenic diseases, Am. J. Ophthalmol. 138, 270-279 (2004). 72. J. Chen et al., Inhibition of retinal neovascularization by soluble EphA2 receptor, Exp. Eye Res. 82 (4), 664-673 (2006). 73. D. Orioli et al., Sek4 and Nuk receptors cooperate in guidance of commissural axons and in palate formation, EMBO J. 15, 6035-6049 (1996). 74. M. Henkemeyer et al., Nuk controls pathfinding of commissural axons in the mammalian central nervous system, Cell 86, 35-46 (1996).

Chapter 12 ADENOSINE IN RETINAL VASCULOGENESIS AND ANGIOGENESIS IN OXYGEN-INDUCED RETINOPATHY

Gerard A. Lutty, PhD, and D. Scott McLeod Wilmer Ophthalmological Institute, Johns Hopkins Hospital, Baltimore, Maryland

Abstract:

Adenosine is a ubiquitous molecule produced predominantly by catabolism of adenosine triphosphate. Levels of this nucleoside increase dramatically with ischemia and elevated tissue activity. Adenosine induces angiogenesis in tumors and wound healing and upregulates VEGF production in several cell types, including endothelial cells. The source of adenosine in most tissues appears to be the ectoenzyme 5’ nucleotidase, which is hypoxia inducible. 5’ nucleotidase expression is prominent during retinal vascular development in the innermost processes of Müller cells, and levels of its product, adenosine, are high in inner retina during retinal vascular development in postnatal dog. One of the adenosine receptors, A2A , is present on angioblasts and on endothelial cells of formed blood vessels during canine retinal vascular development. These observations suggest that adenosine is important in retinal vascular development. Oxygen-induced retinopathy (OIR) is a model for human retinopathy of prematurity (ROP). OIR is induced by exposure of the developing retina to high oxygen. Vascular development is halted, and over 60% of the retinal vasculature is lost during this stage, which is called vaso-obliteration. Expression of 5’ nucleotidase is dramatically reduced during vaso-obliteration, resulting in a sharp decline in adenosine. When animals are returned to room air, the retina is hypoxic because of the lack of blood vessels and increased oxygen consumption due to neuronal development. At this time, the vasoproliferative stage of OIR begins, and 5’ nucleotidase activity and adenosine levels become elevated well beyond normal. A2A -positive endothelial cell proliferation is also elevated compared to control animals. Florid preretinal neovascularization occurs and is characterized by high levels of adenosine and A2A receptors. Therefore, adenosine and its A2A receptor appear to be important in canine OIR. This has also been demonstrated in the mouse model of OIR. Systemically administered antagonists of the adenosine

221 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 221–240. © Springer Science+Business Media B.V. 2008

222

G. A. Lutty and D. S. McLeod A2B receptor significantly reduced retinal neovascularization in mice,1 as did cleavage of A2B by a ribozyme.2 These studies suggest that adenosine and its receptors are important in retinal vascular development and may be a therapeutic target in OIR.

1.

INTRODUCTION

Adenosine is a ubiquitous molecule that is produced predominantly by catabolism of adenosine triphosphate. Few signaling molecules have the ability to influence development like the nucleoside adenosine. Linden called adenosine a “primordial signaling molecule” that is present in and modulates physiological responses in all mammals.3 Adenosine levels in tissues change with increased tissue activity, hypoxia, or stress. Adenosine levels in normal tissues range from 1-50 nM and can rapidly climb to 1000 nM with ischemia and increased tissue activity.4 The action of adenosine is most prominent in tissues where oxygen demand is high, like retina. There are two major sources of adenosine: S-adenylsylhomocysteine and adenosine monophosphate (AMP). S-adenylsylhomocysteine is hydrolyzed to adenosine by S-adenylsylhomocysteine hydrolase, which is prominent in myocardium, but little is known about this source of adenosine in retina. AMP, however, is prominent in the eye, and it is processed, as in other tissues, primarily by 5’ nucleotidase (5’N, also known as CD73). 5’N (E.C. 3.1.3.5.) is an ectoenzyme that catalyzes the hydrolysis of purine, but not pyrimidine, nucleotide monophosphates to their corresponding nucleosides. Although 5’N can metabolize all purine monophosphates, the major product in ischemic dog hearts is adenosine.5 Braun et al. have demonstrated that 5’N expression is elevated during cerebral ischemia.6 In heart, 5’N is upregulated during hypoxia, and subsequently, adenosine levels increase 50fold.7 Synnestradt and associates recently demonstrated that 5’N expression is regulated by hypoxia-inducible factor-1 (HIF-1), explaining at least in part the increased production of adenosine in hypoxia.8 Adenosine is degraded to inosine by adenosine deaminase, which has been found in endothelial and smooth muscle cells, or adenosine kinase, which makes AMP from it.

2.

ADENOSINE RECEPTORS

The four recognized adenosine receptor subtypes are A1, A2A, A2B, and A3, and all are coupled to G-proteins. The A1 receptor has the highest affinity for adenosine and acts through Gi- and Go-proteins;9 therefore, binding the A1 receptor inhibits adenylate cyclase,10 activates phospholipase C,11 and opens

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

223

K+ATP channels.12 The A2 receptors are coupled to the Gs-protein,13 and their transduction systems involve stimulation of adenylate cyclase14 and 2+ activation of N-type Ca channels (Figure 1).15 A3 receptors are coupled to Gi- and Gq-proteins,16 and their transduction system includes the activation of phospholipase C/D17 and inhibition of adenylate cyclase.18,19

3.

EFFECTS OF ADENOSINE ON BLOOD VESSELS

Adenosine is a local regulator of blood flow in several organs.20-22 It may either contract or relax blood vessels, depending on the organ and which receptors are present.23 Vasodilation in response to adenosine may be modulated through nitric oxide (NO) production.24 Dusseau and Hutchins demonstrated that hypoxia-induced angiogenesis on the chorioallantoic membrane (CAM) in chicken was due to adenosine production and uptake.25 In vitro, adenosine is chemotactic and mitogenic for endothelial cells from large blood vessels.26,27 We have determined that adenosine does not stimulate proliferation of dog retinal microvascular endothelial cells but does stimulate endothelial cell migration and tube formation, two events that are critical in the development of the primary retinal vasculature in dog.28 Olanrewaju et al. found that human and porcine coronary artery endothelial cells have both A2A and A2B receptors.29 Feoktistov and associates have recently demonstrated a differential expression of adenosine receptors by large vessel versus microvascular endothelial cells: dermal microvascular endothelial cells expressed A2B receptors, and human umbilical vein endothelial cells (HUVECs) expressed A2A receptors.30 Dubey et al. observed that A2B agonists, and not A2A agonists, stimulated proliferation of porcine and rat arterial endothelial cells in culture.31 Grant and associates, using human retinal microvascular endothelial cells, found that A2B receptor stimulation caused increased proliferation, migration, tube formation, and extracellular signal-related kinase (ERK) activation (Figure 1).32

224

G. A. Lutty and D. S. McLeod

Figure 12-1. Relationship between adenosine (ADO) and its receptor signaling, hypoxia, and VEGF in retinal endothelial cells. Both adenosine and VEGF production are upregulated in hypoxia. For VEGF, this upregulation appears to be modulated through the transcription factor HIF-1α in endothelial cells,82,83 while adenosine upregulation is probably modulated through hypoxia induction of HIF-1α in Müller cells which upregulates adenosine nucleotide catabolism enzyme 5’ nucleotidase (5’N).8 There is evidence that adenosine, by signaling through the A2B receptor, can also upregulate VEGF production by endothelial cells in the absence of hypoxia, suggesting autocrine production by endothelial cells.52 Adenosine binding to either A2A or A2B receptor stimulates adenyl cyclase levels of cAMP. Grant and associates have demonstrated further that ligand binding to A2B activates extracellular signalrelated kinase (ERK) in a cAMP independent manner.32 Endothelial cell proliferation and migration and capillary tube formation are stimulated through this A2B/ERK pathway. Michiels et al. have suggested that ERK activation can result in HIF-1α activation84 but the complete pathway for this relationship has not been established. Therefore, autocrine production of VEGF by endothelial cells may be stimulated directly by hypoxia or indirectly through adenosine receptor signaling. The schematic was inspired by Grant et al.32,52

The angiogenic effect of adenosine may be more than a direct effect of adenosine binding its receptors and stimulating proliferation, migration, and tube formation of endothelial cells. Desai and associates found that adenosine binding the A2A receptor inhibited the secretion of thrombospondin-1, a potent inhibitor of angiogenesis.33 This would shift the balance between angiogenic and antiangiogenic agents toward angiogenesis.

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

4.

225

ADENOSINE IN THE RETINA

As mentioned previously, the major source of adenosine in most tissues is ecto-5’N (CD73). Enzyme and immunohistochemical studies have demonstrated that the glycoprotein 5’N is localized in certain domains of Müller cells in several adult mammalian species34 and is also present during development in murine35 and canine retinas.36 Adenosine, the major product of 5’N, has been proposed as an intercellular communication molecule in retina.37 Histochemical studies have demonstrated adenosine immunoreactivity in adult retinal neurons of several species.38,39 In retina, adenosine modulates blood flow in both adults and neonates40-42 and is released in response to ischemia.43-45 Larsen et al. and Li et al. demonstrated that A1 receptor activation protected against the detrimental effects of ischemia/reperfusion (I/R), and the latter group demonstrated further that A2A stimulation may exacerbate the effects of I/R.45,46 Ischemic preconditioning (brief periods of ischemia) induces a state of ischemic tolerance, and this protective effect is modulated through A1 and A2A receptors.47 Ghiardi and associates have recently reviewed these aspects of adenosine’s action in retina.48

5.

RELATIONSHIP BETWEEN ADENOSINE AND VEGF

Fisher et al. were the first group to demonstrate that adenosine stimulates production of vascular endothelial growth factor (VEGF).49 Takagi and associates then suggested that hypoxia-stimulated upregulation of VEGF mRNA happens via the cellular production of adenosine. They demonstrated that when adenosine agonists bind to A2A receptors on bovine retinal capillary endothelial cells, production of cAMP is elevated, activation of protein kinase A occurs, and then VEGF production is induced.50 Ironically, they have also found that binding of A2A receptor agonists inhibits the production of the VEGF receptor KDR.51 Grant and associates found that A2B activation specifically induced VEGF production.52 The nonselective agonist NECA (adenosine-5’Nethylcarboxamide) stimulated production of VEGF mRNA and protein in addition to increasing proliferation (Figure 1). When anti-VEGF antibody was included, the increased proliferation was inhibited. A2B, and not A2A or A1, antagonists inhibited the effects of NECA. Feoktistov and associates have recently demonstrated that microvascular endothelial cells have A2B receptors and that agonist stimulation of these receptors increases expression of interleukin 8 (IL-8), basic fibroblast growth factor (bFGF), and VEGF.30

226

G. A. Lutty and D. S. McLeod

This effect was modulated by coupling to Gq and possibly G12/13 proteins.30 Agonist stimulation of HUVECs did not increase expression of any of these growth factors.

6.

DEVELOPMENT OF THE RETINAL VASCULATURE

There is controversy over whether retinal vascular development occurs via vasculogenesis or angiogenesis. Vasculogenesis is the development of the vasculature by differentiation and organization of endothelial cell precursors, or angioblasts, while angiogenesis is the formation of blood vessels from an existing blood vessel by migration and proliferation of endothelial cells. In the dog, the retina is only 60% vascularized at birth, so development of the primary or inner retinal vasculature and secondary or deep vasculature can be observed.53 Our observations in dog support the view that the primary retinal vasculature forms by vasculogenesis. These observations were possible because angioblasts and endothelial cells have adenosine diphosphatase (ADPase, now known as CD39) activity54 and menadionedependent alpha glycerophosphate dehydrogenase activity.55 Both enzymes are exclusively found in angioblasts and endothelial cells in newly formed blood vessels in the neonatal canine retina. Angioblasts differentiate from a spherical morphology into spindle-shaped cells as they migrate anteriorly through the cell free spaces formed by the inner Müller cell processes.56 Angioblasts then organize to form cords and eventually lumens in the cell free spaces, essentially happening in the absence of proliferation, suggesting the primary or superficial retinal vasculature forms by vasculogenesis.56 The deep or secondary retinal vasculature starts forming in the dog by angiogenesis at about 15 days of age. We have recently demonstrated that human retinal angioblasts also express ADPase (CD39) and development of the initial human retinal vasculature happens in similar manner to dog.57

6.1

Adenosine and vascular development

Considering the integral morphological role of Müller cells during primary vasculogenesis and their capacity for producing vasogenic adenosine via 5’N, we examined the localization and relative levels of 5’N and adenosine during normal development of the retinal vasculature. The adenosine A2A receptor was also examined immunohistochemically, viable blood vessels were labeled with anti-von Willebrand’s factor (vWf) (Figure 2A), and angioblasts and newly formed blood vessels were labeled with αGPDH (Figure 2B). Microdensitometric image analysis was applied to provide

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

227

semiquantitative information about enzymes and their immunohistochemical reaction products36,58 and how they change with age and with oxygeninduced retinopathy. 5’N histochemical activity in inner Müller cell processes was elevated in areas of vasculogenesis and declined in periphery toward ora serrata at postnatal day 1 (Figure 2C, E). This activity was inhibited by α,β-methyleneadenosine 5’-diphosphate,36 a specific inhibitor of 5’N.59 Adenosine immunoreactivity was highest near the edge of the forming superficial vasculature (Figure 2D, F). A2A immunoreactivity was prominent in angioblasts and endothelial cells in forming blood vessels (data not shown).58 We have been unable to study A1, A2B, or A3 receptors as yet because the antibodies available do not react with dog. At postnatal day 5, 5’N activity and adenosine immunoreactivity are still greatest in inner retina, but adenosine is also elevated in the inner nuclear layer (Figure 3). 5’N activity is still most prominent in inner Müller cell processes, which surround developing blood vessels (Figure 3C). A2A receptor localization is also still confined to inner retina.58 By 15 days of age, inner retina still has the greatest level of 5’N activity and adenosine immunoreactivity (Figure 4). At day 22, when the inner retinal vasculature is complete and the secondary network continues to form in the inner nuclear layer, 5’N activity is greatest in the synaptic zones of the two plexiform layers. At this time (22-28 days postnatal), adenosine immunoreactivity is greatest in ganglion cells and also prominent in inner nuclear layer and photoreceptor inner segments (data not shown). This is the localization of adenosine that Braas and associates observed in adult retina.38 A2A immunoreactivity is still prominent in peripheral blood vessels in inner retina and in the secondary or deep retinal vasculature. However, A2A immunoreactivity is also now prominent in nerve fibers of inner retina and in optic nerve head.58 In summary, during development of the primary retinal vasculature in the nerve fiber layer, 5’N activity and adenosine and A2A immunoreactivity are highest in inner retina where vasculogenesis occurs. When the secondary or deep retinal vasculature forms, A2A immunoreactivity is also associated with the new deep capillaries, and adenosine immunoreactivity and 5’N activity shift toward outer retinal layers. As development of the retinal vasculature nears completion, 5’N in inner retina declines while levels in both plexiform layers increases. Coordinately, the level of adenosine in inner retina declines between day 15 and 28, except for that associated specifically with ganglion cell bodies, which remains elevated through adulthood. Levels in the developing photoreceptor inner segments and inner nuclear layer increase between 15 and 28 days of age. Therefore, 5’N at day 28 is most prominent in both plexiform layers, and adenosine is elevated in the inner nuclear layer, where A2A expression is increased at this time.

228

G. A. Lutty and D. S. McLeod

Figure 12-2. Relationship between developing blood vessels (A-B) and 5’N (C) and ADO (D) in a one-day-old dog retina. (A) Forming blood vessels in inner retina have vWf immunoreactivity. The edge of the forming blood vessels is indicated by an arrow in this and all other images at the internal limiting membrane of retina. (B) Alpha glycerophosphate dehydrogenase (αGPDH) enzyme histochemical reaction product is present in the forming blood vessels and in angioblasts (arrowheads). (C) 5’N enzyme histochemical activity is most prominent in inner Müller cell processes. (D) ADO immunoreactivity is most prominent in inner retina. (E) Relative grayscale values for 5’N in inner Müller cell processes are shown from central retina to ora serrata. The density from image analysis is greatest in areas with blood vessels and just in advance of the forming blood vessels and declines peripherally beyond the edge of the forming blood vessels. (F) ADO relative gray scale values are greatest just beyond the edge of the forming vasculature.

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

229

Figure 12-3. Blood vessels (vWf labeling), 5’N activity, and ADO in 5-day-old normal dog retina (A, C, E) and in 5-day-old oxygen-exposed animals (B, D, F). (A) At 5 days of age, only the superficial retinal vasculature is formed (arrows). (B) After exposure to hyperoxia, blood vessels are limited in inner retina and surviving channels are extremely constricted (arrows). (C) 5’N activity is greatest in the inner Müller cell processes. (D) Hyperoxia has yielded a severe reduction in 5’N activity. (E) ADO immunoreactivity in the air control is prominent in inner retina and in the anterior half of the neuroblastic layer. (F) ADO immunoreactivity is significantly lower in all areas of the hyperoxia-exposed retina. (Republished with permission of the Association for Research in Vision and Ophthalmology, Inc. Fig. 2 from Taomoto M, McLeod DS, Merges C, Lutty GA, Localization of adenosine A2a receptor during retinal vasculogenesis and oxygen-induced retinopathy. Invest. Opthalmol. Vis. Sci. 2000;41:230-243; permission conveyed through Copyright Clearance Center, Inc.)

230

7.

G. A. Lutty and D. S. McLeod

ADENOSINE IN OIR, A MODEL FOR RETINOPATHY OF PREMATURITY

Retinopathy of prematurity can be modeled in several species of neonatal animals by exposure of the neonates to high oxygen levels.60 The time of hyperoxic insult and level of oxygen used varies in the models of different species.61-65 The end result of hyperoxic insult, however, is that retinal vascular development ceases and many formed blood vessels degenerate, a process called vaso-obliteration or vaso-attenuation.54 It is of note that the choroidal vasculature, which is completely formed at birth, is unaffected by the hyperoxic insult.54 Once the animals are returned to room air, oxygen levels return to normal, but the retina becomes hypoxic because of its attenuated vasculature and because neurogenesis has progressed at some level, putting demands on the ambient retinal oxygen available.66-68 At this point, the vasoproliferative stage of the disease occurs, resulting in intraretinal and extraretinal neovascularization. Our studies have focused on the canine model of oxygen-induced retinopathy (OIR). One-day-old dogs are placed in 95-100% oxygen for four days and then removed to room air. The hyperoxic insult results in 70% decrease in capillary density of the forming retinal vasculature with little morphological effect on retinal neurons.54 The insult also results in a significant decline in levels of both 5’N activity and adenosine immunoreactivity (Figure 3D, F), which may be due to oxygen radical damage to 5’N during exposure to hyperoxia. This has already been demonstrated by Kitakaze in heart and in polymorphonuclear leukocytes.69,70 However, Chen and associates found that 5’N activity in kidney increased after exposure to superoxide.71 It is also possible that Müller cells, which are in contact with virtually every cell type in developing retina, act as sensors of retinal oxygen levels. During the initial hyperoxic insult in OIR, when it is thought that the inner retina becomes hyperoxygenated by diffusion of oxygen from the unaffected choriocapillaris,54 Müller cells may downregulate 5’N and, therefore, decrease adenosine concentrations, favoring vasoconstriction, which occurs during the first 24 hours of hyperoxia. The recent evidence identifying a HIF-1-dependent regulatory pathway for 5’N expression suggests the latter scenario.8

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

231

Figure 12-4. Serial sections of retina from two 15-day-old dogs, an air control animal (A, C, E, G) and an oxygen-exposed animal (B, D, F, H). (A) Blood vessels in the superficial retinal vasculature in inner retina are labeled with anti-vWf (arrows). (B) Blood vessels in inner retina (arrow) and in preretinal neovascularization (arrowheads) are labeled with anti-vWf antibody. (C) A2a receptors are present in inner retina (double arrow) and are associated with the superficial blood vessels. (D) A2a immunoreactivity is intense in inner retina and in preretinal neovascularization in OIR (arrowhead). (E) 5’N histochemical reaction is most intense in inner retina. (F) 5’N reaction product is greatly elevated in the OIR inner retina but not present in preretinal neovascularization. The stalk containing the feeder vessel for the neovascularization is indicated by a short arrow. (G) ADO immunoreactivity is most intense in inner retina. (H) ADO immunoreactivity is most intense in OIR retina in preretinal neovascularization and inner retina, where it is extremely elevated compared to the control retina. From Lutty and McLeod, Prog Ret Eye Res 2003;22:95-111.85

232

G. A. Lutty and D. S. McLeod

After three days return to room air, the level of 5’N increased well beyond normal throughout the retina, and the level of adenosine immunoreactivity was similarly elevated in inner retina where the vasculature is reforming.36 A2A receptor expression is also significantly elevated in inner retina where the retinal vasculature becomes dilated as it is reforming.58 Patz et al., Ashton, and more recently, Chan-Ling et al. have suggested that the retina is in a state of hypoxia after vaso-obliteration and during retinal vascular development.66-68 Adenosine levels increase after induction of retinal ischemia in other animal models,44 suggesting that Müller cells could upregulate 5’N production in ischemic environments, like retina during the vasoproliferative stage of OIR. It would, therefore, be logical for 5’N activity to be high, as we have shown, during vascular development and during the period of hypoxia that follows vasoobliteration,36 since 5’N expression is HIF-1-mediated. Furthermore, the levels of immunoreactive adenosine (a product of 5’N) were elevated at the same time that 5’N activity increased, as Kitakaze has shown in heart.70 By day 15, 10 days after hyperoxic insult and return to room air, preretinal neovascularization is prominent in canine OIR. At this time, 5’N activity is greatly elevated throughout retina but is not present in preretinal neovascularization (Figure 4). Adenosine and A2A receptor immunoreactivities are greatly elevated in both inner retina and preretinal neovascularization at this time. The preretinal neovascularization is still morphologically immature and consists of angioblastic-like masses with poorly defined lumens.72 Evidence for the immaturity of these formations is the high A2A immunoreactivity (Figure 4D) and high αGPDH enzyme activity.58 By days 22-28 in dog OIR, the neovascular formations have matured, and there is a 1:1 ratio of endothelial cells to pericytes in some preretinal vessels.63 The mature formations express less A2A receptor than the immature formations at 15 days of age, and proliferation in these formations is reduced (data not shown). 5’N activity and adenosine immunoreactivity remain elevated in retina at this time, and adenosine is still prominent in the preretinal neovasculature. During this period, astrogliosis has occured throughout the inner retina, even in avascular regions where astrocytes are not present during normal vasculogenesis.63 Our results suggest that A2A receptor is associated with normal vasculogenesis and angiogenesis in OIR in the dog. However, we have been unable to examine the localization of A2B receptors in dog because of the lack of an appropriate antibody. Mino and associates used a mouse model of OIR to study the effects of adenosine receptor antagonists on angiogenesis.1 Antagonists selective for A1, A2A, and A2B receptors were evaluated by

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

233

intraperitoneal administration of the antagonists daily for five days starting immediately after hyperoxia. Only the A2B antagonist significantly inhibited neovascularization in this model of OIR. The preretinal neovascularization is not only robust but persists up to P-45, 40 days after hyperoxia, and tractional retinal folds can form as early as P-28.63 Adenosine may also contribute to other aspects of OIR and retinopathy of prematurity. Vasodilation is prominent during the proliferative stage in canine OIR and during vascular development in dog. Adenosine and VEGF could be responsible for this as well. Adenosine is a potent vasodilator, and the A2A receptor, specifically, is an important modulator of vascular tone.74,75 Binding of the A2A receptor on endothelial cells and smooth muscle cells induces vasodilation by stimulating L-arginine transport and nitric oxide (NO) production.76,77 Gidday and Park have demonstrated that A2 receptors can specifically modulate vasodilation in the neonatal pig.42 Extreme vasodilation associated with increased adenosine and A2A receptors in oxygen-treated animals may contribute to the tortuousity of arteries and hemorrhage we have observed in the canine model of OIR.63 The importance of NO in OIR was recently demonstrated by Brooks et al.73 They found that vaso-obliteration in the mouse model of OIR was significantly reduced by inhibiting NO synthase with NG-nitro-L-arginine (L-NNA). Furthermore, both vaso-obliteration and vasoproliferation were significantly reduced in endothelial NO synthase (eNOS) knockout mice.

8.

CONCLUSIONS

There is an intimate relationship between retinal vascular development and adenosine. During vasculogenesis in inner retina, 5’N produces high levels of adenosine in the region where vasculogenesis is occurring (Figure 5). Both angioblasts and endothelial cells have adenosine A2A receptors in the dog. When inner retinal vasculature development is complete, 5’N activity and elevated adenosine are present in more posterior retina, where the secondary or deep capillary network is forming. Adenosine is also associated with angiogenesis in the canine model of OIR. After exposure to hyperoxia, retinal vascular development ceases and vaso-obliteration occurs. This is accompanied by significant reductions in 5’N activity and adenosine (Figure 5). When the vasoproliferative stage of OIR begins, 5’N activity and adenosine levels are elevated well beyond normal. Angiogenesis at this stage is accompanied by elevated levels of A2A receptors in the retinal vasculature and in preretinal neovascular formations. Astrogliosis also occurs in inner retina at this time, which may retard anterior vascular growth (Figure 5).

234

G. A. Lutty and D. S. McLeod

Figure 12-5. Schematic of the role of ADO in vasculogenesis in neonatal dog retina and vasoobliteration and vasoproliferation in the canine model of oxygen-induced retinopathy. Normal vasculogenesis: ADPase-positive angioblasts change from a round morphology to spindle-shaped morphology and express A2A receptors as they migrate through cell-free spaces formed by inner Müller cell processes. They organize into blood vessels in the anterior portion of these spaces. The inner Müller cell processes surrounding these spaces have high 5’N activity (black), which generates adenosine (red dots). Vaso-obliteration: Exposure to hyperoxia causes a significant decrease in 5’N activity and adenosine, severe vasoconstriction, and blood vessel degeneration while A2A levels do not change. Vasoproliferation: Following return to room air, angiogenesis occurs within retina and preretinal neovascularization forms. 5’N activity is greatly elevated in the inner Müller cell processes (black), resulting in high levels of adenosine (red) in inner retina. Astrogliosis occurs in inner retina during the vasoproliferative stage, and the cell-free spaces are filled with astrocytes (green cells). From Lutty and McLeod, Prog Ret Eye Res 2003;22:95-111.85

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

235

One of the A2 receptors may be a therapeutic target for angiogenesis in OIR. There may be a difference between species in terms of which A2 receptor is associated with neovascularization in the retina, based on comparisons of our studies and those of Maria Grant’s laboratory.1,58 The difference may not be important, however, because both A2 receptors act through Gs-proteins, resulting in the same stimulatory effect. Also, there may be redundancy in the system. For example, Morrison et al. recently demonstrated that, in A2A knockout mice, coronary function normally attributed to A2A was conferred when A2B agonists were administered.79 Although the A2 receptors are logical therapeutic targets for stopping retinal angiogenesis, therapy targeting A2 receptors may be dangerous unless delivered specifically to eye. Systemic administration of A2 antagonists could have serious negative effects on cardiac and central nervous system development and function.4,23,80 In any case, the duration of therapy and must be limited so as not to inhibit the neuromodulatory role of A2A in more mature retinas. Future therapies could include local delivery of A2 antagonists or 5’ nucleotidase inhibitors, perhaps by degradable polymers, to retina from vitreous. Unfortunately to date, the only potent inhibitor for 5’N, α,βmethylene adenosine 5’-diphosphate,59 is not tolerated well in the eye (unpublished data). Potent A2 antagonists have been developed recently that have greater water solubility than the original agents.3 It may also be possible to target the A2 receptors with antisense probes as was done successfully with VEGF. Grant and associates have evaluated a unique approach to targeting adenosine therapeutically: the use of a ribozyme. They have developed a ribozyme that degrades the mRNA for the A2B receptor and successfully inhibited angiogenesis in the mouse model of OIR by injecting it into vitreous.81 These potential therapies may have the positive effect of preventing adenosine action and, therefore, indirectly affecting production of VEGF as well.

ACKNOWLEDGMENTS The author acknowledges his collaborators Makoto Taomoto, M.D., and Carol Merges, M.A.S., who contributed substantially to the studies discussed in this manuscript, Andrew Newby, M.D., for his generosity in providing the antibody against adenosine, and Maria Grant, M.D., for helpful discussions. This work was supported by NIH grants EY 01765 (Wilmer Institute) and EY09357 (G. A. L.), the ROPARD Foundation (G. A. L.), Research to Prevent Blindness (Wilmer), and the Brownstein Foundation. G. A. L. is an

236

G. A. Lutty and D. S. McLeod

American Heart Association Established Investigator and the recipient of a Research to Prevent Blindness Lew Wasserman Merit Award.

REFERENCES 1. R. P. Mino, P. E. Spoerri, S. Caballero, D. Player, L. Belardinelli, I. Biaggioni, and M. B. Grant, Adenosine receptor antagonists and retinal neovascularization in vivo, Invest. Ophthalmol. Vis. Sci. 42 (13), 3320-3324 (2001). 2. A. Afzal, L. Shaw, S. Caballero, P. Spoerri, A. Lewin, D. Zeng, L. Belardinelli, and M. B. Grant, Reduction in preretinal neovascularization by ribozymes that cleave the A2b adenosine receptor mRNA, Circ. Res. 93, 500-506 (2003). 3. J. Linden, Molecular approach to adenosine receptors: receptor-mediated mechanisms of tissue protection, Annu. Rev. Pharmacol. Toxicol. 41, 775-787 (2001). 4. S. A. Rivkees, Z. Zhao, G. Porter, and C. Turner, Influences of adenosine on the fetus and newborn, Mol. Genet. Metab. 74 (1-2), 160-171 (2001). 5. G. Schopf, H. Rumpold, and M. M. Müller, Alterations of purine salvage pathways during differentiation of rat heart myoblasts toward myocytes, Biochimica et Biophysica Acta 884, 319-325 (1986). 6. N. Braun, C. Lenz, F. Gillardon, M. Zimmerman, and H. Zimmerman, Focal cerebral ischemia enhances glial expression of ecto-5’-nucleotidase, Brain Res. 766, 213-226 (1997). 7. M. Kitakaze, M. Hori, and T. Kamada, Role of adenosine and its interaction with alpha adrenoceptor activity in ischaemic and reperfusion injury of the myocardium, Cardiovascular Res. 27, 18-27 (1993). 8. K. Synnestvedt, G. T. Furuta, K. M. Comerford, N. Louis, J. Karhausen, H. K. Eltzschig, K. R. Hansen, L. F. Thompson, and S. P. Colgan, Ecto-5’-nucleotidase (CD73) regulation by hypoxia-inducible factor-1 mediates permeability changes in intestinal epithelia, J. Clin. Invest. 110 (7), 993-1002 (2002). 9. M. Freissmuth, W. Schutz, and M. E. Linder, Interactions of the bovine brain A1adenosine receptor with recombinant G protein a-subunits. Selectivity for rGioa-3, J. Biol. Chem. 266, 17778-17783 (1991). 10. C. Blazynski, Adenosine A1 receptor-mediated inhibition of adenylate cyclase in rabbit retina, J. Neuroscience 7, 2522-2528 (1987). 11. W. Abebe and S. J. Mustafa, A1 adenosine receptor mediated Ins(1,4,5)P3 generation in allergic rabbit airway smooth muscle, Am. J. Physiol. 275, L990-L997 (1998). 12. B. T. Liang, Direct preconditioning of cardiac ventricular myocytes via adensoine A1 and KATP channel, Am. J.Physiol. 271, H1769-H1777 (1996). 13. R. B. Marala and S. J. Mustafa, Immunological characterization of adenosine A2a receptors in human and porcine cardiovascular tissue, J. Pharmacol. Exp. Ther. 286, 1051-1057 (1998). 14. V. Stefanovic, P. Vlahovic, V. Savic, and R. Ardaillou, Adenosine stimulates 5’nucleotidase activity in rat mesangial cells via A2 receptor, FEBS Lett. 331, 96-100 (1993). 15. S. N. Li and P. T. Wong, The adenosine receptor agonist, APNEA, increases Ca++ influx into rat cortical synaptosomes through N-type channels associated with A2a receptors, Neurochem. Res. 25, 457-459 (2000).

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

237

16. T. M. Palmer, T. W. Gettys, and G. L. Stiles, Differentail interaction with and regulation of multiple G-proteins by the rat A3 adenossne receptor, J. Biol. Chem. 270, 16895-16902 (1995). 17. M. P. Abbracchio, R. Brambilla, S. Ceruti, H. O. Kim, D. K. von Lubitz, and K. A. Jacobsen, G protein-dependent activation of phospholipase C by adensoine A3 receptors in rat brain, Mol. Pharmacol. 48, 1038-1045 (1995). 18. Z. Zhao, C. E. Francis, and K. Ravid, An A3-subtype receptor is highly expressed in rat smooth muscle cells: its role in attenuating adensoine-induced increase in camp, Microvasc. Res. 54, 243-252 (1997). 19. Z. Zhao, K. Makaritsis, C. E. Francis, H. Gavris, and K. Ravid, A role for the A3 adensoine receptor in determining tissue levels of cAMP and blood pressure: studies in knockout mice, Biochem. Biophys. Acta 1500, 280-290 (2000). 20. R. Berne, R. Knabb, S. W. Ely, and R. Rubio, Adenosine in the local regulation of blood flow: a brief overview, Federation Proc. 42, 3136-3142 (1983). 21. J. W. Phillis, Adenosine in the control of cerebral circulation, Cerebrovascular and Brain Metabolism Reviews 1, 26-54 (1989). 22. H. Winn, S. Morii, and R. Berne, The role of adenosine in autoregulation of cerebral blood flow, Annals Biomed. Engineering 13, 321-328 (1985). 23. R. Tabrizchi and S. Bedi, Pharmacology of adenosine receptors in the vasculature, Pharmacol. Ther. 91 (2), 133-147 (2001). 24. H. A. Olanrewaju and S. J. Mustafa, Adenosine A(2A) and A(2B) receptors mediated nitric oxide production in coronary artery endothelial cells, Gen. Pharmacol. 35 (3), 171-177 (2000). 25. J. W. Dusseau and P. M. Hutchins, Hypoxia-induced angiogenesis in chick chorioallantoic membranes: a role for adenosine, Respir. Physiol. 71, 33-44 (1988). 26. C. J. Meininger, M. E. Schelling, and H. J. Granger, Adenosine and hypoxia stimulate proliferation and migration of endothelial cells, Am. J. Physiol. 255, H554-H562 (1988). 27. E. Teuscher and V. Weidlich, Adenosine nucleotides, adenosine and adenine as angiogenesis factors, Biomed. Biochim. Acta. 44, 493-495 (1985). 28. G. A. Lutty, M. K. Mathews, C. Merges, and D. S. McLeod, Adenosine stimulates canine retinal microvascular endothelial cell migration and tube formation, Curr. Eye Res. 17 (6), 594-607 (1998). 29. H. A. Olanrewaju, W. Qin, I. Feoktistov, J. L. Scemama, and S. J. Mustafa, Adenosine A(2A) and A(2B) receptors in cultured human and porcine coronary artery endothelial cells, Am. J. Physiol. Heart Circ. Physiol. 279 (2), H650-H656 (2000). 30. I. Feoktistov, A. E. Goldstein, S. Ryzhov, D. Zeng, L. Belardinelli, T. VoynoYasenetskaya, and I. Biaggioni, Differential expression of adenosine receptors in human endothelial cells: role of A2B receptors in angiogenic factor regulation, Circ. Res. 90 (5), 531-538 (2002). 31. R. K. Dubey, D. G. Gillespie, and E. K. Jackson, A(2B) adenosine receptors stimulate growth of porcine and rat arterial endothelial cells, Hypertension 39 (2 Pt 2), 530-535 (2002). 32. M. B. Grant, M. I. Davis, S. Caballero, I. Feoktistov, I. Biaggioni, and L. Belardinelli, Proliferation, migration, and ERK activation in human retinal endothelial cells through A(2B) adenosine receptor stimulation, Invest. Ophthalmol. Vis. Sci. 42 (9), 2068-2073 (2001). 33. A. Desai, C. Victor-Vega, S. Gadangi, M. C. Montesinos, C. C. Chu, and B. N. Cronstein, Adenosine A2A receptor stimulation increases angiogenesis by down-regulating production of the antiangiogenic matrix protein thrombospondin 1, Mol. Pharmacol. 67 (5), 1406-1413 (2005).

238

G. A. Lutty and D. S. McLeod

34. G. W. Kreutzberg and S. T. Hussain, Cytochemical heterogeneity of the glial plasma membrane: 5’-nucleotidase in retinal Müller cells, J. Neurocytol. 11, 53-64 (1982). 35. N. Braun, P. Brendel, and H. Zimmerman, Distribution of 5’-nucleotidase in the developing mouse retina, Brain Res. 88, 79-86 (1995). 36. G. A. Lutty, C. Merges, and D. S. McLeod, 5’ nucleotidase and adenosine during retinal vasculogenesis and oxygen- induced retinopathy, Invest. Ophthalmol. Vis. Sci. 41 (1), 218-229 (2000). 37. C. Blazynski and M. T. Perez, Neuroregulatory functions of adenosine in the retina, Prog. Retinal Res. 11, 293-332 (1992). 38. K. M. Braas, M. A. Zarbin, and S. H. Snyder, Endogenous adenosine and adenosine receptors localized to ganglion cells of the retina, Proc. Natl. Acad. Sci., USA 84, 3906-3910 (1987). 39. C. Blazynski, J. L. Mosinger, and A. I. Cohen, Comparison of adenosine uptake and endogenous adenosine-containing cells in mammalian retina, Vis. Neurosci. 2, 109-116 (1989). 40. P. Ostwald, S. S. Park, A. Y. Toledando, and S. Roth, Adenosine receptor blockade and nitric oxide synthase inhibition in the retina: Impact upon post-ischemic hyperemia and the electroretinogram, Vis. Res. 37, 3453-3461 (1997). 41. J. M. Gidday and T. S. Park, Adenosine-mediated autoregulation of retinal arteriolar tone in the piglet, Invest. Ophthalmol. Vis. Sci. 34, 2713-2719 (1993). 42. J. M. Gidday and T. S. Park, Microcirculatory responses to adenosine in the newborn pig, Pediatr. Res. 33, 620-627 (1993). 43. S. Roth, S. S. Park, C. W. Sikorski, J. Osinski, R. Chan, and K. Loomis, Concentrations of adenosine and its metabolites in the rat retina/choroid during reperfusion after ischemia, Curr. Eye Res. 16, 875-885 (1997). 44. S. Roth, P. S. Rosenbaum, J. Osinski, S. S. Park, A. Y. Toledano, B. Li, and A. A. Moshfeghi, Ischemia induces significant changes in purine nucleoside concentration in the retinachoroid in rats, Exp. Eye Res. 65, 771-779 (1997). 45. A. K. Larsen and N. N. Osbourne, Involvement of adenosine in retinal ischemia. Studies on rat, Invest. Ophthalmol. Vis. Sci. 37, 2603-2611 (1996). 46. B. Li, P. S. Rosenbaum, N. M. Jennings, K. A. Maxwell, and S. Roth, Differential roles of adenosine receptor subtypes in retinal ischemia-reperfusion injury in the rat, Exp. Eye Res. 68, 9-17 (1999). 47. B. Li and S. Roth, Retinal preconditioning in the rat: requirement for adenosine and repetitive induction, Invest. Ophthalmol. Vis. Sci. 40, 1200-1216 (1999). 48. G. J. Ghiardi, J. M. Gidday, and S. Roth, The purine nucleoside adenosine in retinal ischemia-reperfusion injury, Vision Res. 39, 2519-2535 (1999). 49. S. Fischer, H. S. Sharma, G. F. Kaliczek, and W. Schaper, Expression of vascular permeability factor/vascular endothelial growth factor in pig cerebral microvascular endothelial cells and its upregulation by adenosine, Mol. Brain Res. 28, 141-148 (1995). 50. H. Takagi, G. L. King, G. S. Robinson, N. Ferrara, and L. P. Aiello, Adenosine mediates hypoxic induction of vascular endothelial growth factor in retinal pericytes and endothelial cells, Invest. Ophthalmol. Vis. Sci. 37, 2165-2176 (1996). 51. H. Takagi, G. L. King, N. Ferrara, and L. P. Aiello, Hypoxia regulates vascular endothelial growth factor receptor KDR/Flk gene expression through adenosine A2 receptors in retinal capillary endothelial cells, Invest. Ophthalmol. Vis. Sci. 37, 1311-1321 (1996). 52. M. B. Grant, R. W. Tarnuzzer, S. Cabalerro, M. J. Ozeck, M. I. Davis, P. E. Spoerri, I. Feoktistov, I. Biaggioni, J. C. Shryock, and L. Belardinelli, Adenosine receptor activation

12. Adenosine in Retinal Vasculogenesis and Angiogenesis in OIR

53. 54. 55.

56. 57.

58.

59.

60.

61. 62.

63.

64.

65.

66.

67.

68. 69.

239

induces vascular endothelial growth factor in human endothelial cells, Circ. Res. 85, 699-706 (1999). R. W. Flower, D. S. McLeod, G. A. Lutty, B. Goldberg, and S. D. Wajer, Postnatal retinal vascular development of the puppy, Invest. Ophthal. Vis. Sci. 26, 957-968 (1985). D. S. McLeod, R. Brownstein, and G. A. Lutty, Vaso-obliteration in the canine model of oxygen-induced retinopathy, Invest. Ophthalmol. Vis. Sci. 37, 300-311 (1996). D. S. McLeod and G. A. Lutty, Menadione-dependent alpha glycerophosphate and succinate dehydrogenases in the developing canine retina, Curr. Eye Res. 14, 819-826 (1995). D. S. McLeod, G. A. Lutty, S. D. Wajer, and R. W. Flower, Visualization of a developing vasculature, Microvasc. Res. 33, 257-269 (1987). T. Chan-Ling, D. S. McLeod, S. Hughes, L. Baxter, Y. Chu, T. Hasegawa, and G. A. Lutty, Astrocyte-endothelial cell relationships during human retinal vascular development, Invest. Ophthalmol. Vis. Sci. 45, 2020-2032 (2004). M. Taomoto, D. S. McLeod, C. Merges, and G. A. Lutty, Localization of adenosine A2a receptor in retinal development and oxygen-induced retinopathy, Invest. Ophthalmol. Vis. Sci. 41 (1), 230-243 (2000). A. Van Waarde, M. E. Stromski, G. Thulin, K. M. Guadio, M. Kashgarian, R. G. Shulman, and N. J. Siegel, Protection of the kidney against ischemic injury by inhibition of 5’nucleotidase, Am. J. Physiol. 256, F298-F305 (1989). G. A. Gole, Animal models of retinopathy of prematurity, in: Retinopathy of prematurity, edited by W. A. Silverman and J. T. Flynn (Blackwell Scientific Publications, Boston, 1985) pp. 53-95. X. Reynaud and C. K. Dorey, Extraretinal neovascularization induced by hypoxic episodes in the neonatal rat, Invest. Ophthalmol. Vis. Sci. 35, 3169-3177 (1994). J. S. Penn, B. L. Tolman, and M. M. Henry, Oxygen-induced retinopathy in the rat: relationship of retinal nonperfusion to subsequent neovascularization, Invest. Ophthalmol. Vis. Sci. 35, 3429-3435 (1994). D. S. McLeod, S. A. D’Anna, and G. A. Lutty, Clinical and histopathologic features of canine oxygen-induced proliferative retinopathy, Invest. Ophthalmol. Vis. Sci. 39 (10), 1918-1932 (1998). L. E. H. Smith, E. Wesolowski, A. McLellan, S. K. Kostyk, R. D. D’Amato, R. Sullivan R, and P. A. D’Amore, Oxygen-induced retinopathy in the mouse, Invest. Ophthalmol. Vis. Sci. 35 (1), 101-111 (1994). T. Chan-Ling, S. Tout, H. Holländer, and J. Stone, Vascular changes and their mechanisms in the feline model of retinopathy of prematurity, Invest. Ophthalmol. Vis. Sci. 33 (7), 2128-2147 (1992). N. Ashton, B. Ward, and G. Serpell, Effect of oxygen on developing retinal vessels with particular reference to the problem of retrolental fibroplasias, Br. J. Ophthalmol. 38, 397-428 (1954). T. Chan-Ling, B. Gock, and J. Stone, The effect of oxygen on vasoformative cell division. Evidence that ‘physiological hypoxia’ is the stimulus for normal retinal vasculogenesis, Invest. Ophthalmol. Vis. Sci. 36, 1201-1214 (1995). A. Patz, Oxygen studies in retrolental fibroplasia:IV clinical and experimental observations, Am. J. Ophthalmol. 38, 291-307 (1954). M. Kitakaze, M. Hori, S. Takashima, K. Iwai, H. Sato, M. Inoue, A. Kitabatake, and T. Kamada, Superoxide dismutase enhances ischemia- induced reactive hyperemic flow and adenosine release in dogs, Circ. Res. 71, 558-566 (1992).

240

G. A. Lutty and D. S. McLeod

70. M. Kitakaze, M. Hori, T. Morioka, S. Takashima, T. Minamino, H. Sato, M. Inoue, and T. Kamada, Attenuation of ecto-5’-nucleotidase activity and adenosine release in activated human polymorphonuclear leukocytes, Circ. Res. 73, 524-533 (1993). 71. Y. F. Chen, P. L. Li, and A. P. Zou, Oxidative stress enhances the production and actions of adenosine in the kidney, Am. J. Physiol. Regulatory Integrative Comp. Physiol. 281, R1808-R1816 (2001). 72. D. S. McLeod, S. N. Crone, and G. A. Lutty, Vasoproliferation in the neonatal dog model of oxygen-induced retinopathy, Invest. Ophthalmol. Vis. Sci. 37 (7), 1322-1333 (1996). 73. S. E. Brooks, X. Gu, S. Samuel, D. M. Marcus, M. Bartoli, P. L. Huang, and R. B. Caldwell, Reduced severity of oxygen-induced retinopathy in eNOS-deficient mice, Invest. Ophthalmol. Vis. Sci. 42, 222-228 (2001). 74. C. D. Lewis, S. M. Hourani, C. J. Long, and M. G. Collis, Characterization of adenosine receptors in the rat isolated aorta, Gen. Pharmac. 25, 1381-1387 (1994). 75. S. M. Poucher, J. R. Keddie, R. Brooks, G. R. Shaw, and D. McKillup, Pharmacodynamics of ZM 241385, a potent A2a adenosine antagonist, after enteric administration in rat, cat and dog, J. Pharm. Pharmacol. 48, 601-606 (1996). 76. L. Sobrevia, D. L. Yudilevich, and G. E. Mann, Activation of A2-purinoceptors by adenosine stimulates L-arginine transport (system y+) and nitric oxide synthesis in fetal human endothelial cells, J. Physiol. 499, 135-140 (1997). 77. S. J. Mustafa and W. Abebe, Coronary vasodilation by adenosine-receptor subtypes and mechanism of action, Drug Development Res. 39, 308-313 (1996). 78. D. S. McLeod, M. Taomoto, J. Cao, Z. Zhu, L. Witte, and G. A. Lutty, Localization of VEGF receptor-2 (KDR/FLK-1) and effects of blocking it in oxygen-induced retinopathy, Invest. Ophthalmol. Vis. Sci. 43, 474-482 (2002). 79. R. R. Morrison, M. A. Talukder, C. Ledent, and S. J. Mustafa, Cardiac effects of adenosine in A(2A) receptor knockout hearts: uncovering A(2B) receptors, Am. J. Physiol. Heart Circ. Physiol. 282 (2), H437-H444 (2002). 80. J. L. Moreau and G. Huber, Central adenosine A(2A) receptors: an overview, Brain Res. Brain Res. Rev. 31, 65-82 (1999). 81. A. Afzal, L. C. Shaw, S. Caballero, E. A. Ellis, and M. B. Grant, The development of hammerhead ribozymes that specifically cleave the A2B receptor mRNA, Invest. Ophthalmol. Vis. Sci. 43, ARVO abstract #3711 (2002). 82. J. A. Forsythe, B. Jiang, N. V. Iyer, F. Agani, S. W. Leung, R. D. Koos, and G. L. Semenza, Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1, Mol. Cell. Biol. 16, 4604-4613 (1996). 83. P. H. Maxwell and P. J. Ratcliffe, Oxygen sensors and angiogenesis, Seminars in Cell & Developmental Biology 13 (1), 29-37 (2002). 84. C. Michiels, E. Minet, G. Michel, D. Mottet, J. Piret, and M. Raes, HIF-1 and AP-1 cooperate to increase gene expression in hypoxia: role of MAP kinases, IUBMB Life 52, 49-53 (2001). 85. G. A. Lutty and D. S. McLeod, Retinal vascular development and oxygen-induced retinopathy: a role for adenosine, Prog. Ret. Eye Res. 22, 95-111 (2003).

Chapter 13 THE REGULATION OF RETINAL ANGIOGENESIS BY CYCLOOXYGENASE AND THE PROSTANOIDS

Gary W. McCollum and John S. Penn Department of Ophthalmology and Visual Sciences, Vanderbilt University School of Medicine, Nashville, Tennessee

Abstract:

1.

Non-steroidal anti-inflammatory drugs (NSAIDs) inhibit cyclooxygenase and the formation of cyclooxygenase products, the prostanoids. Chronic use of NSAIDs has been associated with a reduced risk of colorectal cancer, which may be in part a consequence of reduced tumor-associated angiogenesis. These findings suggest that cyclooxygenase and the prostanoids may regulate angiogenesis. Several potentially blinding retinopathies have angiogenic components, and the putative roles of cyclooxgenase and the prostanoids in this context have been, and are currently, under investigation.

INTRODUCTION

Prostaglandins (PGs), prostacyclins (PGIs), and thromboxanes (TXs), collectively referred to as the prostanoids, are lipid-derived autocrine/paracrine signaling molecules that are involved in a wide range of physiological and pathophysiological processes.1 Since their discovery, the prostanoid literature has burgeoned into a wealth of experimental data suggesting complicated and often contradictory roles in, but by no means limited to: the immune response, inflammatory response, female reproductive biology, wound healing, arthritis, asthma, atherosclerosis, gastric ulcers, and cancer.1 Recent studies suggest roles for prostanoids in the context of ocular angiogenesis, and bear relevance to the pathological neovascular component of several potentially blinding conditions such as macular degeneration, diabetic retinopathy, retinal vein

241 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 241–257. © Springer Science+Business Media B.V. 2008

242

G. W. McCollum and J. S. Penn

occlusion and retinopathy of prematurity.2-5 Herein, these studies and their findings will be discussed.

2.

PROSTANOID SYNTHESIS

The discovery of the prostanoids resulted from the findings of two landmark studies conducted in the 1930s: the identification and characterization of the essential fatty acids, and isolation of a biological activity from human seminal fluid that would contract smooth muscle preparations.6-8 The prostaglandins E (PGE), F (PGF), and D (PGD) were the first to be characterized followed by the thromboxanes, prostacyclin and the leukotrienes. Subsequent studies demonstrated that essential fatty acids are converted to prostanoids by oxygenation pathways.9 Prostanoids are further classified into either series-1, -2 or -3 depending on which of three precursor essential fatty acids serves as the substrate for the oxygenation reactions (i.e., PGE1, PGE2, and PGE3). Series-1 and -3 are synthesized from γ-homolinolenic acid and eicosapentaenoic acid (20:5ω-3), respectively. Series-2 prostanoids are synthesized from arachidonic acid, the most abundant prostanoid precursor in humans.9,10 The initial step of series-2 prostanoid biosynthesis is arachidonic acid release from membrane phospholipids in a reaction catalyzed by phospholipase A2 (PLA2). There are at least 19 groups of PLA2s that are generally classified as cytosolic (cPLA2), secretory (sPLA2) or calcium-independent (iPLA2). PLA2 is activated in response to numerous stimuli including ischemia, oxidative stress, and cell signaling molecules.11 A cyclooxygenase (COX) enzyme catalyzes the reaction between two molecules of O2 and arachidonic acid. The catalytic domain of COX has two distinct active sites: the COX active site catalyzes the formation of endoperoxide and hydroperoxyl functionalities to produce prostaglandin G2 (PGG2), and the peroxidase active site catalyzes the reduction of the hydroperoxyl group to a hydroxyl group to form PGH2. Cell-specific synthases catalyze isomerization, oxidation, and reduction of PGH2 to yield the prostanoids (see Figure 1).12-14

13. The Regulation of Retinal Angiogenesis

243

cPLA2 COOH

Arachidonic Acid Cyclooxygenase 2O2 O

CO OH

COOH

O

PGH 2

OOH

PGD Synthase

HO

PGI Synthase O

COOH

O

COOH

O O

PGG2

OH

PGD 2

OH

OH

O

PGI 2 PGF Synthase

TX Synthase

PGE Synthase

HO

O COOH

COOH

COOH O O

HO

OH

PGF2α

HO

OH

PGE 2

TXA 2

OH

Figure 13-1. Oxygenation of arachidonic acid and subsequent conversion to prostanoids. In response to hormonal and/or stress-induced cues, arachidonic acid is released from the membranes of the nucleus and/or endoplasmic reticulum by enzymatic cleavage of phospholipids by phospholipase A2. Subsequent oxygenation and reduction of arachidonic acid by cyclooxygenase (COX) produces PGH2. PGH2 is converted to one or more of the prostanoids by cell-specific synthases.

3.

COX-1 AND COX-2

Early studies investigating the mitogen- and proinflammatory agentinduction of prostaglandin biosynthesis led researchers to postulate the existence of more than one form of COX.14 Platelet-derived growth factor (PDGF) stimulation of Swiss 3T3 cells revealed an initial induction of prostaglandin biosynthesis 10 minutes post-stimulation, followed by a second induction occurring 2-4 hours post-stimulation that depended on new protein synthesis. In 1989, Northern blotting with an ovine COX cDNA probe detected a 4.0-kb RNA in addition to a known 2.8-kb mRNA. The larger transcript was shown to be inducible and to parallel the induction of COX activity. In the late 1980s and early 1990s, studies involving gene upregulation by the v-src oncogene, phorbol esters and serum, and the

244

G. W. McCollum and J. S. Penn

identification of tetradecanoyl-13-phorbol acetate- and mitogen-inducible sequences in Swiss 3T3 cells, reported the upregulaion of COX DNA sequences. These data, along with the results of other studies, pointed to the existence of constitutive and inducible forms of COX referred to as COX-1 and COX-2, respectively.14

4.

PROSTANOID RECEPTORS

Bioassays performed in various tissues suggested that activity profiles of the prostanoids overlap to some degree but possess sufficient differences to allow distinction.15 These studies led researchers to propose the existence of multiple types of prostanoid receptors with cell-specific expression profiles that could perhaps explain the variety of actions and the sometimes opposing effects elicited by the prostanoids. Additional investigations linked prostanoid activities with the activation of intracellular second messenger systems such as phosphatidylinositol (PI) turnover, Ca2+ mobilization and changes in cAMP levels.15 These studies allowed functional correlation of cell or tissue binding activities to bioactivities or the activation of second messenger systems and led Coleman et al. to propose the existence of the prostanoid receptors and to classify them. Specific putative receptors for TX, PGI, PGE, PGF, and PGD were named TP, IP, EP, FP, and DP receptors, respectively. The EP receptor classification was further broken down into four subtypes, namely EP1, EP2, EP3, and EP4.15-17 Hirata et al. cloned the human TXA2 receptor in 1991,18 and homology-based screening of cDNA libraries from several species with probes based on this sequence were performed. All of the prostanoid receptors classified by previous pharmacological and biochemical studies were identified.15 These functional and genetic analyses have classified the prostanoid receptors into a subfamily of G-protein-coupled receptors with seven transmembrane domains belonging to the superfamily of the rhodopsin-type receptors.

5.

INHIBITION OF COX VIA NON-STEROIDAL ANTI-INFLAMMATORY DRUGS

In 1971, J. R. Vane and colleagues discovered that non-steroidal antiinflammatory drugs (NSAIDs) are potent inhibitors of COX.9 These seminal findings provided a powerful pharmacological tool to investigate the physiology and pathophysiology of COX-dependent processes. COX inhibition studies revealed that COX products are the mediators of pain, fever and inflammation. However, it is important to recognize that

13. The Regulation of Retinal Angiogenesis

245

COX-dependent processes are often integrated with other signaling cascades to produce a physiological outcome. For example, intradermal injection of a histamine-PGE2 mixture causes greater pain than if either of the compounds is administered alone. Furthermore, PGE2 will augment the effect of histamine at doses that produce no effect when administered alone.9

6.

COX-2-SELECTIVE INHIBITION

The discovery and characterization of the two COX isoforms (COX-1 and COX-2) led to the hypothesis that selective inhibition of COX-2 would alleviate pain and inflammation without the adverse side effects associated with COX-1 inhibition (e.g. gastrointestinal damage). The results of large clinical trials testing the COX-2 selective inhibitors, the coxibs, have shown this hypothesis to be true in a general sense.19 However, in the recent Adenomatous Polyp Prevention on Viox (APPOVe) study,20-22 the COX-2 selective inhibitor, rofecoxib, was linked to an increased risk of myocardial infarction, resulting in its withdrawal from the market. Other COX-2 selective inhibitors may also have a detrimental effect on the cardiovascular system.20,23,24

7.

ANGIOGENESIS

Angiogenesis, the formation of new capillaries from existing blood vessels, occurs in reproduction, growth and development, and wound healing.25-30 In normal physiological processes, angiogenesis is tightly regulated. However, in various pathologies such as arthritis, tumor growth and retinopathies, dysregulated and persistent angiogenesis occurs.30-32 Endothelial cells and pericytes are two prominent cell types found in microvessels, capillaries, and collecting venules. These cells are induced by angiogenic stimuli to proliferate and differentiate, ultimately leading to a capillary network.25-28,33 Endothelial cells within microvessels normally remain quiescent for several years under physiological conditions (except in female reproductive organs), maintained by an intricate balance of pro- and anti-angiogenic stimuli.26,27 In certain disease states, the balance is tipped in favor of angiogenesis, and the resting phenotype is converted to an angiogenic phenotype leading to the formation of new microvessels. 25-27,31,34-37 Angiogenesis consists of a cascade of carefully orchestrated events. Initially, there is the production of angiogenic growth factors such as vascular endothelial cell growth factor (VEGF) and basic fibroblast growth factor (bFGF) that may occur in response to tissue injury or ischemia. Extracellular

246

G. W. McCollum and J. S. Penn

proteinases degrade the microvessel basement membrane and remodel the extracellular matrix to allow migration of endothelial cells into the extravascular space. Endothelial cell proliferation and differentiation, resulting in tube formation, with subsequent anastomoses of the adjacent tubes, leads to a microvasculature that is stabilized by the attachment of supportive cells (e.g., pericytes).25-28,33,34

8.

OCULAR DISEASE AND ANGIOGENESIS

Retinopathy of prematurity (ROP), proliferative diabetic retinopathy (PDR) and age-related macular degeneration (ARMD) are vasoproliferative disorders that can lead to blindness in affected individuals. ROP occurs in premature infants, with PDR and ARMD primarily affecting working age individuals and the elderly, respectively.38-40 Pathological angiogenesis, common to each of these conditions and referred to as ocular neovascularization (NV), causes vascular permeability leading to retinal edema, the development of fragile vessels, and abnormal pre-retinal fibrovascular structures commonly referred to as neovascular tufts. These conditions predispose the affected individual to hemorrhage, tractional retinal detachment, and vision loss.41 Laser photocoagulation procedures are performed to treat ocular neovascular conditions; however, these procedures are plagued with undesirable side effects and do not target the underlying pro-angiogenic stimuli.42-46

9.

MECHANISMS OF OCULAR ANGIOGENESIS

Ischemia is common to retinal neovascular conditions and leads to retinal hypoxia that initiates the angiogenic cascade.47,48 In 1948, Michaelson proposed a link between retinal ischemia and retinal angiogenesis in terms of a diffusible pro-angiogenic factor that is synthesized and released in response to hypoxia. Since then, several pro-angiogenic factors have been identified including: fibroblast growth factor (FGF), insulin-like growth factor (IGF), vascular endothelial growth factor (VEGF), the angiopoietins, platelet-derived growth factor (PDGF), and tumor necrosis factor (TNF). 49,50 Several lines of evidence suggest that VEGF is the principal mediator of retinal angiogenesis.51 VEGF is a homodimeric glycoprotein that induces vasopermeability and angiogenic behaviors.52-55 There are five main homodimeric VEGF isoforms. The corresponding monomers have 121, 145, 165, 189 and 206 amino acids resulting from alternative splicing of a single VEGF transcript.56 VEGF165 and VEGF121 are diffusible isoforms, whereas

13. The Regulation of Retinal Angiogenesis

247

the VEGF145, VEGF189, and VEGF206 isoforms are bound to the heparincontaining proteoglycans of the extracellular matrix. VEGF receptor-1 (VEGFR1 or Flt-1) and VEGF receptor-2 (VEGFR2, KDR, or Flk-1) are two high-affinity plasma membrane receptors that bind VEGF and mediate its biological signals. They each have an extracellular VEGF-binding domain consisting of seven immunoglobulin–like domains, a transmembrane domain, and a cytoplasmic tyrosine kinase sequence interrupted by a kinase insert domain. Microvascular endothelial cells co-express these receptors, as do other endothelial cell types.57 Hypoxia induces VEGF synthesis in retinal cell types including endothelial cells, pericytes, retinal pigmented epithelial cells (RPE), Müller cells, and ganglion cells.58-62 Müller cells have been shown to be the principal source of VEGF in animal models of neovascular disease.60-62 The hypoxia-inducible transcription factor (HIF)-1, accumulates in response to hypoxia and stimulates transcription of the VEGF gene from a binding site at -975 in the human VEGF promoter.63-66 VEGF is also post-transcriptionally regulated by hypoxia.63,66 The observation that increased expression of VEGF correlates with retinal NV identifies VEGF as a major inducer of the angiogenic program. Subsequent investigations further support this notion and have shown that retinal NV is suppressed by agents that bind VEGF 59,67,68 and inhibitors of VEGF receptor tyrosine kinase activity.69,70

10.

COX-MEDIATED ANGIOGENESIS

Patients who take NSAIDs on a regular basis are less prone to the development of colorectal cancer.71 Colorectal tumors express high levels of COX, suggesting that PGs may influence the growth and development of tumors; and COX inhibitors may protect against tumorigenesis.72,73 It appears that PGs may help promote tumorigenesis by stimulating angiogenesis, because a pro-angiogenic PG effect has been noted in cancer models and other systems.74-76 On the other hand, COX inhibitors block angiogenesis in several experimental systems.77-83 The prostanoid effect is likely mediated through the stimulation of proangiogenic growth factor expression.73,79 In support of this notion, prostaglandin treatment of cells in vitro leads to increased levels of VEGF and bFGF.84,85 Furthermore, tumor viruses, such as the Epstein-Barr virus, induce VEGF expression in a COX-2-dependent manner.86 VEGF synthesis and release is decreased in wild-type fibroblasts treated with COX-2 inhibitors and COX-2-/- mouse fibroblasts, and COX-2 overexpression upregulates several angiogenic inducers73,79 in colon carcinoma cells. The

248

G. W. McCollum and J. S. Penn

prostanoid induction of angiogenesis may be amplified in some cases by an autocrine feedback loop. VEGF-induced COX-2 expression and activation of phospholipase A2 -mediated arachidonic acid release leads to enhanced prostaglandin synthesis and release, followed by binding of prostanoid receptors promoting enhanced VEGF expression.87,88 The influence of prostanoids on angiogenesis likely depends on the tissue, environmental and genetic background, and the mode of action (i.e. paracrine vs. autocrine). For example, TP receptor-specific agonists and antagonists have been shown to be involved in corneal and tumor angiogenesis.89,90 However, TP receptor agonists reverse angiogenesis in vitro.91

11.

INHIBITION OF COX-2 RESTRICTS ANGIOGENESIS

The anti-angiogenic function of NSAIDs has largely been attributed to the inhibition of COX-2, since selective inhibition of COX-1 fails to block NV.92-95 Reduced angiogenesis by NSAIDs may result, at least in part, from decreased prostanoid production, because in some cases NSAID suppression of angiogenesis is reversed by prostaglandins or prostanoid-receptor agonists.87,89,94 NSAIDs block the production of angiogenic factors by tumor cells and stromal fibroblasts and also inhibit pro-angiogenic signaling pathways in endothelial cells.73,79,81 It appears that the anti-angiogenic activity of NSAIDs has COX-dependent and –independent components. COXindependent effects that may block angiogenesis have been identified and include the inhibition of transcription factors nuclear factor kB (NF-kB) and activator protein-1 (AP-1). 95 Other effects that have been reported are the inhibition of the mitogen-activated protein kinase cascade80 and the suppression of GTPases, Cdc42 and Rac, via integrin αvβ3.96 These proteins are necessary for cell spreading and migration.97 However, it is not known whether these effects are COX-dependent.

12.

POTENTIAL ROLES OF COX AND THE PROSTANOIDS IN RETINAL ANGIOGENESIS

Animal models of oxygen-induced retinopathies (OIR) are crucial to understanding the pathogenesis of vasoproliferative retinopathies and have been used in studies investigating the role(s) of COX in retinal angiogenesis. A review of the development of these models is presented in Chapter 3 of this volume.

13. The Regulation of Retinal Angiogenesis

249

Convincing evidence exists that links tissue hypoxia to COX-2-mediated angiogenesis in tumors, suggesting the possibility that similar COXdependent mechanisms may exist for ischemic vasoproliferative retinopathies. Recent studies using animal models of OIR, choroidal NV (CNV), corneal NV and VEGF-induced vascular leakage2-5 have investigated COX-2-dependent mechanisms in ocular angiogenesis, with particular emphasis on hypoxia, the VEGF signaling cascade, and the inhibitory effects of NSAIDs. Wilkinson-Berka et al. tested the COX-2 selective inhibitor rofecoxib in a mouse model of OIR and observed a 37% reduction in pathological retinal angiogenesis in treated mice relative to untreated controls. Rofecoxib-treated mice maintained in room air had a 45% reduction in the formation of the inner retinal vasculature compared to untreated room air mice, suggesting a potential role for COX-2 in normal development. COX-2 immunoreactivity was observed in the ganglion cell layer and the blood vessels of the room air and OIR mice. COX-2 was also localized to pre-retinal blood vessels extending into the vitreous cavity in OIR mice. Takahashi et al. tested the effects of nepafenac in three murine models of ocular NV. 98 Nepafenac, the amide derivative of the COX-1 and -2 inhibitor amfenac, easily penetrates the cornea after topical administration and is readily deaminated to amfenac in vivo. OIR or CNV was induced in mice by standard protocols, and the mice were treated with 0.1%, 0.5% nepafenac or vehicle by topical administration. Nepafenac-treated OIR mice had significantly less ischemia-induced retinal NV than the corresponding vehicle-treated controls. To investigate the effects of nepafenac on VEGF expression in mouse OIR, semiquantitative RT-PCR analysis of retinal RNA showed a nepafenac-dependent decrease in VEGF mRNA levels, providing a plausible explanation for the observed reduction in ischemiainduced retinal NV. As previously discussed, Müller cells are a major source of VEGF in the hypoxic retina and play a key role in the pathogenesis of vasoproliferative retinopathies. COX-2 undergoes a dramatic upregulation when Müller cells are subjected to hypoxia. Furthermore, there is an approximate 3-fold increase in PGE2 synthase in hypoxic Müller cells relative to those maintained in normoxia (Penn, unpublished results). In vitro data have shown that amfenac dose-dependently inhibits hypoxia-induced VEGF production in Müller cells (Penn, unpublished results). It remains unclear if these observations are COX-dependent because COX-2-/- Müller cells showed significant hypoxia-induced VEGF expression (Penn, unpublished results). However, it has been demonstrated that PGE2 induces upregulation of VEGF and βFGF in Müller cells. Using selective inhibitors of protein kinase A, the authors inferred that EP2 and/or EP4 were responsible for

250

G. W. McCollum and J. S. Penn

VEGF induction.84 These data suggest the possibility of hypoxia-induced VEGF expression via a COX-2/PGE2 autocrine loop. Sennlaub et al. investigated the localization of COX-2 in human retinas from non-diabetic subjects and subjects with diabetic retinopathy, and in retinas of murine and rat models of OIR.3 COX-2 immunoreactivity was localized in RPE cells, the outer segment of the photoreceptors, and to some degree the inner plexiform layer. In all diabetic patients, COX-2 immunoreactivity was also detected in the nerve fiber layer, co-localizing to a significant extent with glial fibrillary acidic protein (GFAP). This suggests that significant COX-2 expression occurs in the retinal astrocytes of these diabetic patients. Immunolocalization of COX-2 in OIR mouse retinas was similar to that found in humans. Of particular interest is that COX-2 expression was detected in astrocytes (GFAP-positive cells) of the nerve fiber layer during the normoxic period following hyperoxic exposure, which is similar to the pattern observed in retinas from humans with diabetic retinopathy. In vitro experiments were performed with primary porcine retinal astrocyte cultures exposed to hypoxia (2% oxygen) for 24 hours. They revealed an 8-fold increase in COX-2 protein levels relative to normoxic controls as measured by western blot analysis. There was a concomitant increase in PGE2 synthesis that was significantly decreased by the COX-2 selective inhibitors APHS and etodolac, and the COX-1-selective inhibitor SC-560 resulted in only a small decrease. In the same study, APHS, etodolac, or SC-560 were tested in the murine and rat models of OIR by intravitreal injection. APHS showed a dose-dependent decrease in pre-retinal NV, and SC-560 had no effect, when both were tested in the murine model. The retinal PGE2 level in these mice was reduced by 65% 24 hours after APHS treatment. Intravitreal injection of PGE2 produced a small but significant increase in pre-retinal NV. In the rat OIR model, etodolac showed a decrease in pre-retinal NV when compared to vehicle-treated controls that was reversed by intravitreal injection of PGE2. The EP2- and EP3-specific agonists, butaprost and M&B28767 respectively, were tested in etodolac treated OIR rats. Butaprost exacerbated and M&B28767 partly reversed the inhibitory effects of etodolac. EP receptor protein expression profiles were determined in the rat OIR model during the course of oxygen treatment. During hyperoxic exposure, EP1 was not detected; EP4 was slightly decreased; EP2, and to a greater extent EP3, was decreased. After 24 hours at normoxia, there was no significant change in EP4; however, there were significant increases in EP2 and EP3. These data suggest that there is a COX2-dependent regulatory component of retinal NV in these models that is relayed by PGE2 through the EP2 and EP3 receptors. To probe for potential COX-2-dependent mechanisms of angiogenesis, the effect of COX-2 inhibition and EP3 stimulation on retinal pro-angiogenic

13. The Regulation of Retinal Angiogenesis

251

VEGF and VEGFR2 and anti-angiogenic Thrombospondin-1 (TSP-1) and its receptor (CD36) was investigated during the post-hyperoxic period in the rat OIR model. Exposure to normoxia for 24 hours after the hyperoxic insult resulted in increased TSP-1 protein expression. The COX-2 inhibitor etodolac induced a substantial increase in TSP-1 and CD36, and addition of M&B28767 reversed this effect, suggesting that EP3 stimulation inhibits the production of anti-angiogenic factors. This explains at least in part the antiangiogenic effect of COX-2 inhibition. VEGF expression was only marginally affected by etodolac and M&B28767, while VEGFR2 expression was not changed. According to these data, the influence of COX-2 on retinal NV could not be explained by modulation of VEGF protein levels. NSAIDs have been shown to inhibit endothelial cell angiogenic behaviors such as proliferation and tube formation, and hypoxia-induced VEGF expression in Müller cells. The role of COX remains unclear, because NSAIDs have non-specific activities that may contribute to the antiangiogenic effect observed in cultured cells and in animal models of neovascular disease.97,99 For example, amfenac inhibits the phosphorylation of Erk in human retinal microvascular endothelial cells (HRMEC), which is a major downstream signaling intermediate of VEGFR2 involved in cell proliferation (Penn, unpublished results). As a result, studies performed with NSAIDs must be interpreted with caution. To assess the role of COX in retinal angiogenesis, and, at the same time, avoid the complications associated with the non-specific effects of pharmacological COX-inhibitors, Cryan et al. investigated the effects of either COX-1 or COX-2 gene deletion in the mouse model of OIR.20 Histological analysis of retinas from wild-type, COX-1-/- and COX-2-/- mice raised in room air showed no differences in the development of the retinal vasculature. Pre-retinal NV, retinal vascular/avascular areas, and perfused retinal areas were measured in COX1-/-, COX-2-/-and wild type OIR mice. Interestingly, there was essentially no difference in pre-retinal NV for the COX-1-/- strain and a non-significant trend toward less pre-retinal NV for the COX-2-/- strain compared to the wild-type. Isolectin B4-staining of retinal vasculature, a technique that does not distinguish between perfused and nonperfused vessels, revealed similar percentages of capillary-free zones among these strains. As measured by fluorescein angiography, perfused retinal areas were reduced in COX-2-/mice compared to the other two strains. Immunohistochemical analysis showed increased fibrin deposits and thrombocyte staining in the retinas of COX-2-/- mice, suggesting that COX-2 protects against vascular obstruction (thrombosis). The authors postulate that the absence or reduction of COX-2derived PGI leaves the pro-thrombosis effects of COX-1-derived TX from platelets unbalanced, because PGI inhibits platelet activation and TX is a potent platelet activator. A substantial neovascular response occurred in the

252

G. W. McCollum and J. S. Penn

COX-2-/- mice. However, is there any evidence for a COX-2-dependent component? In rodent models of OIR, the size of the nonperfused retinal area frequently correlates with the severity of pre-retinal NV. Explaining this correlation, a commonly accepted hypothesis states that the level of proangiogenic stimulus depends on the level of tissue hypoxia, which is proportional to the size of the nonperfused retinal area. Although the nonperfused retinal area of the COX-2-/- mice was higher than the other two strains, there was a comparable neovascular response. Based on this observation, the authors speculate that a COX-2-dependent component of the neovascular response exists; however, the primary finding of increased retinal thrombosis in the COX-2-/- mice complicates the evaluation of COX2-dependent retinal pro-angiogenic mechanisms. The studies outlined above leave large gaps in our understanding of the COX-dependent mechanisms involved in ocular angiogenesis. No attempt has yet been made to examine systematically which PGs are important in neovascular eye pathology or to discern the COX-dependent and/or COXindependent effects of NSAIDs in animal models of proliferative retinopathy or the angiogenic endothelial cell behaviors. Thus, little in the way of mechanistic information has been uncovered. These questions are left to future studies.

REFERENCES 1. H. R. Herschman, Regulation of prostaglandin synthase-1 and prostaglandin synthase-2. Cancer Metastasis Rev. 3, 241-256 (1994). 2. B. N. Nandgaonkar, T. Rotschild, K. Yu, and R. D. Higgins, Indomethacin improves oxygen-induced retinopathy in the mouse. Pediatr. Res. 46, 184-188 (1999). 3. F. Sennlaub, F. Valamanesh, A. Vazquez-Tello, A. M. El-Asrar, D. Checchin, S. Brault, F. Gobeil, M. H. Beauchamp, B. Mwaikambo, Y. Courtois, K. Geboes, D. R. Varma, P. Lachapelle, H. Ong, F. Behar-Cohen, and S. Chemtob, Cyclooxygenase-2 in human and experimental ischemic proliferative retinopathy. Circulation 108, 198-204 (2003). 4. J. L. Wilkinson-Berka, N. S. Alousis, D. J. Kelly, and R. E. Gilbert, COX-2 inhibition and retinal angiogenesis in a mouse model of retinopathy of prematurity. Invest. Ophthalmol. Vis. Sci. 44, 974-979 (2003). 5. M. R. Castro, D. Lutz, and J. L. Edelman, Effect of COX inhibitors on VEGF-induced retinal vascular leakage and experimental corneal and choroidal neovascularization. Exp. Eye Res. 79, 275-285 (2004). 6. G. O. Burr and M. M. Burr, On the nature and role of the fatty acids essential in nutrition. J. Biol. Chem. 86, 587-621 (1930). 7. R. Kurzrok and C. C. Lieb, Biochemical studies of human semen. II. The action of semen on the human uterus. Proc. Soc. Exp. Biol. Med. 28, 268-722 (1930). 8. U. S. von Euler, An adrenal-like action in extracts from prostatic and related glands. J. Physiol. 81, 102-12 (1934). 9. A. R. Brash, (1990) Arachidonate Metabolites. Prostaglandins, Prostacyclin, Thromboxane and Leukotrienes. In: Drugs and Anesthesia: Pharmacology for

13. The Regulation of Retinal Angiogenesis

10. 11. 12.

13.

14. 15. 16.

17. 18.

19. 20.

21.

22. 23.

24.

25. 26. 27. 28.

253

Anesthesiologists. 2nd edition. pp 631-644. Wood, M. and Wood, A.J.J., Williams and Wilkins, Baltimore, Maryland. C. D. Funk, Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 294, 1871-1875 (2001). J. P. SanGiovanni and E. Y. Chew, The role of omega-3 long-chain polyunsaturated fatty acids in health and disease of the retina. Prog. Retin. Eye Res. 24, 87-138 (2005). C. A. Rouzer and L. J. Marnett, Structural and functional differences between cyclooxygenases: fatty acid oxygenases with a critical role in cell signaling. Biochem. Biophys. Res. Commun. 338, 34-44 (2005). L. J. Marnett, S. W. Rowlinson, D. C. Goodwin, A. S. Kalgutkar, and C. A. Lanzo, Arachidonic acid oxygenation by COX-1 and COX-2. Mechanisms of catalysis and inhibition. J. Biol. Chem. 274, 22903-22906 (1999). D. L. Simmons, R. M. Botting, and T. Hla, Cyclooxygenase isozymes: the biology of prostaglandin synthesis and inhibition. Pharmacol. Rev. 56, 387-437 (2004). S. Narumiya, Y. Sugimoto, and F. Ushikubi, Prostanoid receptors: structures, properties, and functions. Physiol. Rev. 79, 1193-1226 (1999). R. A. Coleman, I. Kennedy, P. P. A. Humphrey, K. Bunce, and P. Lumley, Prostanoids and their receptors. In: Comprehensive Medicinal Chemistry. Membranes and Receptors, edited by J.C. Emmett. Oxford, UK: Pergamon, 1990, vol. 3, p. 643-714. R. A. Coleman, S. P. Grix, S. A. Head, J. B. Louttit, A. Mallett, and R. L. G. Sheldrick, A novel inhibitory receptor in piglet saphenous vein. Prostaglandins 47, 151-168 (1994). M. Hirata, Y. Hayashi, F. Ushikubi, Y. Yokota, R. Kageyama, S. Nakanishi, and S. Narumiya, Cloning and expression of cDNA fro a human thromboxane A2 receptor. Nature 349, 617-620 (1991). B. N. Cronstein, Cyclooxygenase-2-selective inhibitors: translating pharmacology into clinical utility. Cleve. Clin. J. Med. 69, Suppl 1: SI13-19 (2002). L. M. Cryan, G. P. Pidgeon, D. J. Fitzgerald, and C. J. O’Brien, COX-2 protects against thrombosis of the retinal vasculature in a mouse model of proliferative retinopathy. Mol. Vis. 12, 405-414 (2006). R. S. Bresalier, R. S. Sandler, H. Quan, J. A. Bolognese, B. Oxenius, K. Horgan, C. Lines, R. Riddell, D. Morton, A. Lanas, M. A. Konstam, and J. A. Baron, Adenomatous Polyp Prevention on Vioxx (APPROVe) Trial Investigators. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N. Engl. J. Med. 352, 1092-1102 (2005). G. A. Fitzgerald, Coxibs and cardiovascular disease. N. Engl. J. Med. 351, 1709-1711 (2004). N. A. Nussmeier, A. A. Whelton, M. T. Brown, R. M. Langford, A. Hoeft, J. L. Parlow, S. W. Boyce, and K. M. Verburg, Complications of the COX-2 inhibitors parecoxib and valdecoxib after cardiac surgery. N. Engl. J. Med. 352, 1081-1091 (2005). S. D. Solomon, J. J. McMurray, M. A. Pfeffer, J. Wittes, R. Fowler, P. Finn, W. F. Anderson, A. Zauber, E. Hawk, and M. Bertagnolli, Adenoma Prevention with Celecoxib (APC) Study Investigators. Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N. Engl. J. Med. 352, 1071-1080 (2005). M. Klagsbrun, Regulators of angiogenesis: stimulators, inhibitors, and extracellular matrix. J. Cell. Biochem. 47, 199-200 (1991). J. Folkman and Y. Shin, Angiogenesis. J. Biol. Chem. 267, 10931-10934 (1992). J. Folkman and P. A. D’Amore, Blood vessel formation: what is its molecular basis? Cell 87, 1153-1155 (1996). W. Risau, Mechanisms of angiogenesis. Nature 386, 761-764 (1997).

254

G. W. McCollum and J. S. Penn

29. J. Li, Y. P. Zhang, and R. S. Kirsner, Angiogenesis in wound repair: angiogenic growth factors and the extracellular matrix. Microsc. Res. Tech. 60, 107-114 (2003). 30. J. Folkman, T. Browder, and J. Palmblad, Angiogenesis research: guidelines for translation to clinical application. Thromb. Haemost. 86, 23-33 (2001). 31. P. Carmeliet, Angiogenesis in health and disease. Nat. Med. 9, 653-660 (2003). 32. J. Folkman, Angiogenesis in cancer, vascular, rheumatoid and other diseases. Nat. Med. 1, 27-31 (1995). 33. R. K. Jain, Molecular regulation of vessel maturation. Nat. Med. 9, 685-693 (2003). 34. A. Tarnawski, D. Hollander, J. Stachura, et al. Role of angiogenesis in healing of experimental gastric ulcer. 1991. In: Halter F, Garner A, Tytgat GNJ (eds) Mechanisms of peptic ulcer healing. Kluwer, Dordrecht, p. 165-171. 35. M. K. Jones, R. M. Itani, H. Wang, M. Tomikawa, I. J. Sarfeh, S. Szabo, and A. S. Tarnawski, Activation of VEGF and Ras genes in gastric mucosa during angiogenic response to ethanol injury. Am. J. Physiol. 276, G1345-1355 (1999). 36. M. K. Jones, M. Tomikawa, B. Mohajer, and A. S. Tarnawski, Gastrointestinal mucosal regeneration: role of growth factors. Front. Biosci. 4, D303-309 (1999). 37. J. Constant, D. Suh, M. Hussain, et al. 1996. Wound healing angiogenesis: the metabolic basis of repair. In: Maragoudakis ME (ed) Molecular, cell, and clinical aspects of angiogenesis. Plenum, New York, p. 151-159. 38. P. G. Steinkuller, L. Du, C. Gilbert, et al. Childhood blindness. J. AAPOS 3, 26-32 (1999). 39. B. Rahmani, J. M. Tielsch, J. Katz, J. Gottsch, H. Quigley, J. Javitt, and A. Sommer, The cause-specific prevalence of visual impairment in an urban population. The Baltimore Eye Survey. Ophthalmology 103, 1721-1726 (1996). 40. N. M. Bressler and S. B. Bressler, Preventative ophthalmology. Age-related macular degeneration. Ophthalmology 102, 1206-1211 (1995). 41. P. Lee, C. C. Wang, and A. P. Adamis, Ocular neovascularization: an epidemiologic review. Surv. Ophthalmol. 43, 245-269 (1998). 42. L. P. Brodell, R. J. Olk, N. P. Arribas, E. Okun, G. P. Johnston, I. Boniuk, R. F. Escoffery, M. G. Grand, D. B. Burgess, and L. H. Schoch, Neovascular glaucoma: a retrospective analysis of treatment with peripheral panretinal cryotherapy. Ophthalmic Surg. 18, 200206 (1987). 43. G. C. Brown, L. E. Magargal, A. Schachat, and H. Shah, Neovascular glaucoma. Etiologic considerations. Ophthalmology 91, 315-320 (1984). 44. A. Chandra, Health aspects of pregnancy and child birth: United States, 1982-88. Vital Health Stat. 18, 50 (1995). 45. P. Archambault and J. E. Gomolin, Incidence of retinopathy of prematurity among infants weighing 2000 g or less at birth. Can. J. Ophthalmol. 22, 218-220 (1987). 46. S. R. Chandra, E. S. Gragoudas, E. Friedman, E. M. Van Buskirk, and M. L. Klein, Natural history of disciform degeneration of the macula. Am. J. Ophthalmol. 78, 579-582 (1974). 47. I. C. Michaelson, The mode of development of the vascular system of the retina with some observations on its significance for certain retinal diseases. Trans. Ophthalmol. Soc. UK 68, 137-180 (1948). 48. N. Ashton, Retinal vascularization in health and disease. Am. J. Ophthalmol. 44, 7-17 (1957). 49. P. A. D’Amore, Mechanisms of retinal and choroidal angiogenesis. Invest. Ophthalmol. Vis. Sci. 35, 3974-3979 (1994). 50. R. Casey and W. W. Li, Factors controlling ocular angiogenesis. Am. J. Ophthalmol. 124, 521-529 (1997).

13. The Regulation of Retinal Angiogenesis

255

51. L. P. Aiello, Vascular endothelial growth factor. 20th-century mechanisms, 21st-century therapies. Invest. Ophthalmol. Vis. Sci. 38, 1647-1652 (1997). 52. D. R. Senger, D. T. Connolly, L. Van de Water, J. Feder, and H. F. Dvorak, Purification and NH2-terminal amino acid sequence of guinea pig tumor-secreted vascular permeability factor. Cancer Res. 50, 1774-1778 (1990). 53. P. J. Keck, S. D. Hauser, G. Krivi, K. Sanzo, T. Warren, J. Feder, and D. T. Connolly, Vascular permeability factor, an endothelial cell mitogen related to PDGF. Science 246, 1309-1312 (1989). 54. N. Ferrara, K. A. Houck, L. B. Jakeman, J. Winer, and D. W. Leung, The vascular endothelial growth factor family of polypeptides. J. Cell. Biochem. 47, 211-218 (1991). 55. K. H. Plate, G. Breier, H. A. Weich, W. Risau, Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo. Nature 359, 845-848 (1992). 56. N. Ferrara, H.-P. Gerber, and J. LeCourter, The biology of VEGF and its receptors. Nat. Med. 9, 669-676 (2003). 57. H. Thieme, L. P. Aiello, H. Takagi, N. Ferrara, and G. L. King, Comparative analysis of vascular endothelial growth factor receptors on retinal and aortic vascular endothelial cells. Diabetes 44, 98-103 (1995). 58. L. P. Aiello, J. M. Northrup, B. A. Keyt, H. Takagi, and M. A. Iwamoto, Hypoxic regulation of vascular endothelial growth factor in retinal cells. Arch. Ophthalmol. 113, 1538-1544 (1995). 59. L. P. Aiello, E. A. Pierce, E. D. Foley, H. Takagi, H. Chen, L. Riddle, N. Ferrara, G. L. King, and L. E. Smith, Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins. Proc. Natl. Acad. Sci. USA 92, 10457-10461 (1995). 60. E. A. Pierce, R. L. Avery, E. D. Foley, L. P. Aiello, and L. E. Smith, Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neovascularization. Proc. Natl. Acad. Sci. USA 92, 905-909 (1995). 61. S. G. Robbins, J. R. Conaway, B. L. Ford, K. A. Roberto, and J. S. Penn, Detection of vascular endothelial growth factor (VEGF) protein in vascular and non-vascular cells of the normal and oxygen-injured rat retina. Growth Factors 14, 229-241 (1997). 62. S. G. Robbins, V. S. Rajaratnam, and J. S. Penn, Evidence for upregulation and redistribution of vascular endothelial growth factor (VEGF) receptors flt-1 and flk-1 in the oxygen-injured rat retina. Growth Factors 16, 1-9 (1998). 63. A. P. Levy, N. S. Levy, and M. A. Goldberg, Post-transcriptional regulation of vascular endothelial growth factor by hypoxia. J. Biol. Chem. 271, 2746-2753 (1996). 64. H. P. Gerber, F. Condorelli, J. Park, and N. Ferrara, Differential transcriptional regulation of the two vascular endothelial growth factor receptor genes. Flt-1, but not Flk-1/KDR, is up-regulated by hypoxia. J. Biol. Chem. 272, 23659-23667 (1997). 65. J. A. Forsythe, B.-H. Jiang, E. A. Rue, et al. Hypoxia-inducible factor 1 is a basic-helixloop-helix-PAS heterodimer regulated by cellular O2 tension. Proc. Natl. Acad. Sci. USA 92, 5510-5514 (1995). 66. C. W. Pugh and P. J. Ratcliffe, Regulation of angiogenesis by hypoxia: role of the HIF system. Nat. Med. 9, 677-684 (2003). 67. A. P. Adamis, D. T. Shima, M. J. Tolentino, E. S. Gragoudas, N. Ferrara, J. Folkman, P. A. D’Amore, and J. W. Miller, Inhibition of vascular endothelial growth factor prevents retinal ischemia-associated iris neovascularization in a nonhuman primate. Arch. Ophthalmol. 114, 66-71 (1996). 68. Y. Saishin, Y. Saishin, K. Takahashi, R. Lima e Silva, D. Hylton, J. S. Rudge, S. J. Wiegand, and P. A. Campochiaro, VEGF-TRAP(R1R2) suppresses choroidal neovascularization

256

69.

70.

71. 72.

73. 74. 75. 76. 77. 78. 79. 80.

81.

82.

83. 84.

85.

86.

G. W. McCollum and J. S. Penn

and VEGF-induced breakdown of the blood-retinal barrier. J. Cell. Physiol. 195, 241-248 (2003). M. S. Seo, N. Kwak, H. Ozaki, H. Yamada, N. Okamoto, E. Yamada, D. Fabbro, F. Hofmann, J. M. Wood, and P. A. Campochiaro, Dramatic inhibition of retinal and choroidal neovascularization by oral administration of a kinase inhibitor. Am. J. Pathol. 154, 1743-1753 (1999). H. Ozaki, M. S. Seo, K. Ozaki, H. Yamada, E. Yamada, N. Okamoto, F. Hofmann, J. M. Wood, and P. A. Campochiaro, Blockade of vascular endothelial cell growth factor receptor signaling is sufficient to completely prevent retinal neovascularization. Am. J. Pathol. 156, 697-707 (2000). W. E. Smalley and R. N. DuBois, Colorectal cancer and nonsteroidal anti-inflammatory drugs. Adv. Pharmacol. 39, 1-20 (1997). T. Kawamori, C. V. Rao, K. Seibert, and B. S. Reddy, Chemopreventive activity of celecoxib, a specific cyclooxygenase-2 inhibitor, against colon carcinogenesis. Cancer Res. 58, 409-412 (1998). C. S. Williams, M. Mann, and R. N. DuBois, The role of cyclooxygenases in inflammation, cancer, and development. Oncogene 18, 7908-7916 (1999). D. M. Form and R. Auerback, PGE2 and angiogenesis. Proc. Soc. Exp. Biol. Med. 172, 214-218 (1983). M. Ziche, J. Jones, and P. M. Gullino, Role of prostaglandin E1 and copper in angiogenesis. J. Natl. Cancer. Inst. 69, 475-482 (1982). L. Diaz-Flores, R. Gutierrez, and H. Varela, Angiogenesis: an update. Histol. Histopathol. 9, 807-843 (1994). T. A. Deutsch and W. F. Hughes, Suppressive effects of indomethacin on thermally induced neovascularization of rabbit corneas. Am. J. Ophthalmol. 87, 536-540 (1979). L. E. Davel, M. M. Miguez, and E. S. de Lustig, Evidence that indomethacin inhibits lymphocyte-induced angiogenesis. Transplantation 39, 564-565 (1985). M. Tsujii, S. Kawano, S. Tsuji, H. Sawaoka, M. Hori, and R. N. DuBois, Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 93, 705-716 (1998). M. K. Jones, H. Wang, B. M. Peskar, E. Levin, R. M. Itani, I. J. Sarfeh, and A. S. Tarnawski, Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: insight into mechanisms and implications for cancer growth and ulcer healing. Nat. Med. 5, 14181423 (1999). H. Sawaoka, S. Tsuji, M. Tsujii, E. S. Gunawan, Y. Sasaki, S. Kawano, M. and Hori, Cyclooxygenase inhibitors suppress angiogenesis and reduce tumor growth in vivo. Lab. Invest. 79, 1469-1477 (1999). M. Yamada, M. Kawai, Y. Kawai, and Y. Mashima, The effect of selective cyclooxygenase-2 inhibitor on corneal angiogenesis in the rat. Curr. Eye Res. 19, 300-304 (1999). J. Masferrer, Approach to angiogenesis inhibition based on cyclooxygenase-2. Cancer J. 7, S144-150 (2001). T. Cheng, W. Cao, R. Wen, R. H. Steinberg, and M. M. LaVail, Prostaglandin E2 induces vascular endothelial growth factor and basic fibroblast growth factor mRNA expression in cultured rat Müller cells. Invest. Ophthalmol. Vis. Sci. 39, 581-591 (1998). R. Pai, I. L. Szabo, B. A. Soreghan, S. Atay, H. Kawanaka, and A. S. Tarnawski, PGE(2) stimulates VEGF expression in endothelial cells via ERK2/JNK1 signaling pathways. Biochem. Biophys. Res. Commun. 286, 923-928 (2001). S. Murono, H. Inoue, T. Tanabe, I. Joab, T. Yoshizaki, M. Furukawa, and J. S. Pagano, Induction of cyclooxygenase-2 by Epstein-Barr virus latent membrane protein 1 is

13. The Regulation of Retinal Angiogenesis

87.

88.

89.

90.

91.

92.

93. 94.

95. 96.

97. 98.

99.

257

involved in vascular endothelial growth factor production in nasopharyngeal carcinoma cells. Proc. Natl. Acad. Sci. USA 98, 6905-6910 (2001). G. L. Hernandez, O. V. Volpert, M. A. Iniguez, E. Lorenzo, S. Martinez-Martinez, R. Grau, M. Fresno, and J. M. Redondo, Selective inhibition of vascular endothelial growth factor-mediated angiogenesis by cyclosporin A: roles of the nuclear factor of activated T cells and cyclooxygenase 2. J. Exp. Med. 193, 607-620 (2001). C. Wheeler-Jones, R. Abu-Ghazaleh, R. Cospedal, R. A. Houliston, J. Martin, and I. Zachary, Vascular endothelial growth factor stimulates prostacyclin production and activation of cytosolic phospholipase A2 in endothelial cells via p42/p44 mitogenactivated protein kinase. FEBS Lett. 420, 28-32 (1997). T. O. Daniel, H. Liu, J. D. Morrow, B. C. Crews, and L. J. Marnett, Thromboxane A2 is a mediator of cyclooxygenase-2-dependent endothelial migration and angiogenesis. Cancer Res. 59, 4574-4577 (1999). P. Pradono, R. Tazawa, M. Maemondo, M. Tanaka, K. Usui, Y. Saijo, K. Hagiwara, and T. Nukiwa, Gene transfer of thromboxane A(2) synthase and prostaglandin I(2) synthase antithetically altered tumor angiogenesis and tumor growth. Cancer Res. 62, 63-66 (2002). Y. Gao, R. Yokota, S. Tang, A. W. Ashton, and J. A. Ware, Reversal of angiogenesis in vitro, induction of apoptosis, and inhibition of AKT phosphorylation in endothelial cells by thromboxane A(2). Circ. Res. 87, 739-745 (2000). J. L. Masferrer, K. M. Leahy, A. T. Koki, B. S. Zweifel, S. L. Settle, B. M. Woerner, D. A. Edwards, A. G. Flickinger, R. J. Moore, and K. Seibert, Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res. 60, 1306-1311 (2000). K. M. Leahy, A. T. Koki, and J. L. Masferrer, Role of cyclooxygenases in angiogenesis. Curr. Med. Chem. 7, 1163-1170 (2000). M. K. Jones, H. Wang, B. M. Peskar, E. Levin, R. M. Itani, I. J. Sarfeh, and A. S. Tarnawski, Inhibition of angiogenesis by nonsteroidal anti-inflammatory drugs: insight into mechanisms and implications for cancer growth and ulcer healing. Nat. Med. 5, 14181423 (1999). I. Tegeder, J. Pfeilschifter, and G. Geisslinger, Cyclooxygenase-independent actions of cyclooxygenase inhibitors. FASEB J. 15, 2057-2072 (2001). O. Dormond, A. Foletti, C. Paroz, and C. Ruegg, NSAIDs inhibit alpha V beta 3 integrinmediated and Cdc42/Rac-dependent endothelial-cell spreading, migration and angiogenesis. Nat. Med. 7, 1041-1047 (2001). L. S. Price, J. Leng, M. A. Schwartz, et al. Activation of Rac and Cdc42 by integrins mediates cell-spreading. Mol. Biol. Cell. 9, 1863-1871 (1998). K. Takahashi, Y. Saishin, Y. Saishin, K. Mori, A. Ando, S. Yamamoto, Y. Oshima, H. Nambu, M. B. Melia, D. P. Bingaman, and P. A. Campochiaro, Topical nepafenac inhibits ocular neovascularization. Invest. Ophthalmol. Vis. Sci. 44, 409-415 (2003). K. Kashfi and B. Rigas, Non-COX-2 targets and cancer: expanding the molecular target repertoire of chemoprevention. Biochem. Pharmacol. 70, 969-986 (2005).

Chapter 14 EXTRACELLULAR PROTEINASES IN OCULAR ANGIOGENESIS Arup Das and Paul G. McGuire Division of Ophthalmology, Department of Surgery, and Department of Cell Biology & Physiology, University of New Mexico School of Medicine, and New Mexico VA Health Care System, Albuquerque, New Mexico

Abstract:

1.

The process of angiogenesis comprises several phases including upregulation of angiogenic factors and increased expression of integrins and extracellular proteinases. The proteinases facilitate the breakdown of the basement membrane and extracellular matrix, allowing endothelial cells to migrate. The enzymes primarily involved in this process are the serine proteinase, urokinase plasminogen activator (uPA), and members of the matrix metalloproteinase (MMP) family. The interaction between uPA and its receptor, uPAR, and the activation of MMPs have been described in tumor angiogenesis. We have found increased expression of MMPs and uPA in retinas of animal models of retinal and choroidal neovascularization. Endogenous inhibitors like tissue inhibitor of matrix metalloproteinase (TIMP) also play an important role in pathological angiogenesis. Pre-clinical studies have indicated that proteinase inhibitors may have therapeutic potential in retinal and choroidal angiogenesis. Some of these inhibitors are being tested in clinical trials in ocular angiogenesis.

INTRODUCTION

The angiogenesis cascade consists of several phases; upregulation of angiogenic growth factors is followed by increased expression of specific integrins and extracellular proteinases. The invasive process of cell migration through the basement membrane and extracellular matrix (ECM) is facilitated by these proteinases. The phenotype of endothelial cells activated during the proliferative and invasive phases of angiogenesis includes increased expression of cell-substrate adhesion molecules and proteolytic enzymes. Their action facilitates the degradation of the capillary 259 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 259–277. © Springer Science+Business Media B.V. 2008

260

A. Das and P. G. McGuire

basement membrane and migration and subsequent invasion of activated endothelial cells into the surrounding tissues.1-4 The enzymes primarily involved in this process are the serine proteinase, urokinase plasminogen activator (uPA), and members of the matrix metalloproteinase (MMP) family. Upregulation of proteinases is a crucial event in tumor as well as ocular angiogenesis. Pharmacological intervention in this pathway has proved to be an alternative therapeutic approach in preclinical angiogenesis studies, and some of these drugs are now in various phases of clinical trials.

2.

UROKINASE (uPA-uPAR SYSTEM)

The proteolytically active urokinase on the cell surface is critical for cell migration. Urokinase is secreted as a single chain proenzyme that can be cleaved by plasmin. Urokinase is present in two molecular forms: a 54 kDa high molecular weight form and a 33 kDa low molecular weight form, which lacks the amino-terminal fragment (ATF) of the protein.5-7 The ATF plays a role in cell proliferation.8,9 The main function of the urokinase is to convert the inactive zymogen form of the enzyme plasminogen to plasmin, a broadspectrum proteinase, which can cleave a variety of ECM components including collagen IV, fibronectin, and elastin. The uPA localizes to the surface of endothelial cells by binding to the uPA receptor (uPAR). This interaction of uPA and uPAR facilitates cell migration through localized proteolytic and nonproteolytic regulation of cell-substrate adhesion.10-11 Recent studies emphasize that the uPAR plays the role of a “versatile orchestrator” and that uPAR, integrin, and very-low-density lipoprotein receptor (VLDLR) interact with each other, resulting in the cycled attachment, detachment, and reattachment of integrins that is necessary for cell migration.12 uPA also activates several MMPs, causing the release of growth factors.10 The uPA-uPAR system has also been implicated in the regulation of cell migration and matrix remodeling involved in angiogenesis both in normal development and in tumor progression and metastasis.10-12

3.

MATRIX METALLOPROTEINASES

The MMPs are a family of enzymes involved in the degradation of a variety of ECM components including the collagens, laminin, fibronectin, elastin, and the core protein of proteoglycans.13 Currently, at least 21 members of the MMP family have been identified.13 All the MMPs contain a zinc ion at the active site and show consistent structural and sequence homologies. All are

14. Extracellular Proteinases in Ocular Angiogenesis

261

secreted as latent pro-enzymes and are activated by partial proteolytic cleavage. The MMPs are divided into five subclasses based upon their substrate specificity: the interstitial collagenases (MMP-1, MMP-8 and MMP-13), the gelatinases (MMP-2 and MMP-9), the stromelysins (MMP-3, MMP-10, and MMP-11), the other MMPs (matrilysin or MMP-7; metalloelastase or MMP-12) and the membrane type MMPs (MT-MMP).14 These proteinases play important roles in a variety of cellular process including the regulation of cell migration, proliferation, and apoptosis. MMPs are involved in the normal physiological processes of embryonic development and wound healing and are overexpressed in diseases such as cancer and arthritis. MMPs are upregulated in most types of human cancer and are often associated with a poor prognosis. Some of the MMPs are expressed by tumor cells, while other MMPs are expressed by stromal cells, including endothelial cells, fibroblasts, myofibroblasts, and inflammatory cells.15 The roles of both uPA and MMPs in cell migration in angiogenesis are summarized in Figure 1. Role of Proteinases in Angiogenesis PAI-1

Cell Migration

Plasminogen Plasmin

uPA

1. Degradation of ECM (collagen, elastin, fibronectin)

uPAR Active MMPs

Cell

Pro-MMPs 2. Breakdown of cell-matrix adhesion 3. Breakdown of cell-cell adhesion

MT-MMP

4. Cryptic sites exposed Cell

5. Degradation of PEDF (endogenous inhibitor)

6. Release of VEGF from ECM stores

Figure 14-1. Flowchart describing the role of extracellular proteinases in angiogenesis. The urokinase (uPA) acts on the receptor, uPAR, and the activated uPA then converts plasminogen to plasmin, which can degrade the extracellular matrix (ECM) components, as well as activate the MMPs. The MMPs have several functions, including degradation of ECM, breakdown of cell-matrix adhesions and cell-cell adhesions, exposure of cryptic sites, release of VEGF from the ECM, and degradation of PEDF. The combined activity of these proteinases ultimately regulates the cell adhesion and migration necessary for angiogenesis.

262

4.

A. Das and P. G. McGuire

PROTEINASES IN OCULAR ANGIOGENESIS

Do proteinases play any role in ocular angiogenesis? We examined proteinases by zymography (a technique to quantify proteolytic activity) in retinal extracts from animals with retinal neovascularization (NV) on day 17 (the active angiogenic phase). Significant increases in the high (54 kDa) and low (32 kDa) molecular weight forms of uPA were observed in the retinas of animals with active NV.16 Similar increases were also found in the levels of the proenzyme and active forms of both MMP-2 (72 kDa and 62 kDa, respectively) and MMP-9 (92 kDa and 84 kDa, respectively) in animals with NV. These results suggest that the active phase of the angiogenic process is associated with the increased expression of uPA, MMP-2, and MMP-9. RTPCR studies in experimental animals with retinal NV also revealed increases in mRNA levels for MMP-2, MMP-9, and MT-MMP that correlated with changes in proteinase levels and proenzyme activation.16 MMP-2 is a substrate for MT-MMP, which may explain the presence of increased levels of the activated form of MMP-2. Analysis of mRNA did not detect expression of MMP-3 or MMP-7 in either control or experimental animals, confirming the results of zymographic analysis. Exposure of a collagen type IV cryptic epitope (a protein sequence that normally remains hidden) represents one of the earliest remodeling events required before vessel sprouting.17 Exposure of these cryptic sites has been found to be inhibited in MMP-9-deficient but not MMP-2-deficient mice, suggesting a role of MMP-9 in their exposure. This would be a novel mechanism in which MMP-9 facilitates angiogenesis by promoting retinal endothelial cell migration and angiogenesis. We also reported an upregulation of the urokinase receptor, uPAR, in the retinas of a murine model of retinal NV.18 The uPAR protein was localized to vessel profiles within the superficial portion of the retina and to vessels on the vitreal side of the inner limiting membrane. To determine whether uPAR is necessary for the development of retinal NV, we subjected uPAR knockout mice to the same oxygen protocol as used for the murine model of retinal NV and quantified the extent of NV. Retinal NV in uPAR knockout animals was reduced by 73% compared to normal mice,18 and these knockout mice showed normal retinal vascular development. Thus, increased expression of proteinases was observed in the retinas of an animal model with retinal NV, indicating an activation of the proteolytic cascade during angiogenesis. These animal results were found to correlate with results from a study of proteinases in epiretinal neovascular membranes removed surgically from patients with proliferative diabetic retinopathy.19 The levels of uPA, MMP-2,

14. Extracellular Proteinases in Ocular Angiogenesis

263

and MMP-9 were significantly elevated in the neovascular membranes compared to normal retinas. Does the upregulation of proteinases also happen in choroidal neovascularization (CNV)? RT-PCR studies have shown upregulation of uPA and uPAR in the choroidal tissues of mice with laser-induced CNV as well as in CNV membranes from patients with age-related macular degeneration.20,21 We also found that uPAR localized to the endothelial cells of the fibrovascular tissue within the CNV complex in the laser-induced NV model. Studies with single-gene-deficient mice have shown that the absence of uPA, tPA (tissue plasminogen activator), or plasminogen significantly decreased the development of experimental CNV and that this effect could be explained by a modulation of MMP activity in the laser-induced wounds.20 If proteinases have a role in angiogenesis, are they also involved in early diabetic retinopathy? One of the early features of diabetic retinopathy is the alteration of the blood-retina barrier (BRB), which may involve the breakdown of endothelial cell tight junctions. We investigated the role of extracellular proteinases during the early stages of diabetic retinopathy, especially in relation to the BRB. We have shown in streptozotocin (STZ) treated diabetic Sprague Dawley rats a 1.7-fold increase in retinal vascular permeability after 12 weeks of diabetes and upregulation of the levels of specific extracellular proteinases in the retina compared to non-diabetic controls.22 Using conventional semi-quantitative RT-PCR, MMP-2, MMP-9, and MMP-14 mRNA levels were found to be significantly elevated in the retinas of diabetic animals. Real time RT-PCR was utilized to quantify the mRNA levels for components of the urokinase system in diabetic retinas, and all components of this system were found to be significantly elevated in the 12 week diabetic rats when compared to non-diabetic controls. How do the proteinases function in early diabetic retinopathy? Is there any role of MMPs and urokinase in regulating tight junction functions? Retinal endothelial and pigment epithelial cells treated with purified MMP-2 or MMP-9 were found to have alterations of tight junction function as shown by decreased transepithelial electrical resistance (TER).22 Western blot analysis of cell extracts treated with MMP-2 or MMP-9 revealed specific degradation of the tight junction protein, occludin. A previous study reported that uPA can regulate the paracellular permeability pathway in VEGFtreated cultured endothelial cells.23 Increased vascular permeability in experimental diabetes is associated with reduced endothelial occludin content,24 and VEGF has been shown to cause rapid phosphorylation of occludin. 25 Thus, elevated expression of MMPs in the retinas of diabetic animals may facilitate an increased vascular permeability by a mechanism involving proteolytic degradation of the occludin followed by disruption of

264

A. Das and P. G. McGuire

the overall tight junction complex. A greater understanding of the role of proteinases in altering tight junction proteins may provide future targets for therapeutic intervention.

5.

ENDOGENOUS PROTEINASE INHIBITORS

The balance of proteinases and inhibitors has been shown to be a critical determinant of endothelial cell morphology and tube formation in vitro.

5.1

Tissue inhibitors of metalloproteinases

MMP activity is in part regulated by the tissue inhibitors of metalloproteinases (TIMPs), which bind the proteinases and inhibit their activity. Four TIMPs have been identified, and they share significant homology at the amino acid level, including 12 cysteine residues that form six disulfide bonds in each member of the family.14,26 Each TIMP is capable of inhibiting all metalloproteinases; however, preferential binding to specific MMPs has been reported.27,28 TIMP-1 primarily inhibits the activities of MMP-1, -3, and -9, whereas TIMP-2 inhibits MMP-2.27,29 TIMP-2 has also been shown to bind and stabilize MMP-2 by preventing autolytic degradation and by participating in its activation.30,31 TIMP-3 is localized exclusively to the ECM and is relatively insoluble, illustrating its potential to prevent matrix proteolysis and the release of growth factors sequestered in the ECM.27 TIMP-3 is present in Bruch's membrane of normal human eyes,32 and TIMP-3 mRNA has been localized to mouse and human retinal pigment epithelial cells.33,34 A point mutation in the TIMP-3 gene has been implicated in patients with Sorsby's fundus dystrophy, an autosomal dominant macular disease with earlier onset of symptoms similar to those of age-related macular degeneration and characterized by choroidal NV.35,36 The TIMP-3 content in Bruch's membrane of the macula shows a significant increase in eyes with age-related macular degeneration compared with age-matched normal eyes.37 We have found that the TIMP-2 message and protein levels in retinas of normal mice in room air increased steadily until day 17, whereas in animals with retinal NV, TIMP-2 mRNA and protein remained significantly lower than in control animals.38 We could not find any significant changes in TIMP-1 and TIMP-3 levels in retinas with NV. Thus, at least in an animal model of retinal NV, we have shown a temporal correlation between proteinases (MMP-2, MMP-9, and MT1-MMP) and TIMP-2 in response to hypoxic stimulation.

14. Extracellular Proteinases in Ocular Angiogenesis

5.2

265

Plasminogen activator inhibitors

The proteolytic activity of uPA is physiologically regulated by plasminogen activator inhibitors (PAI), which are members of the serine proteinase inhibitor (SERPIN) family. PAI-1 and PAI-2 have been shown to interact with urokinase in a 1:1 ratio to inhibit enzyme activity and cause enzyme/inhibitor internalization and turnover.39 A role for PAI in the regulation of tumor cell invasion and motility has also been suggested, and PAI appears to be a useful prognostic marker for a number of different cancers.40 We have found significant increases in the level of PAI-1 mRNA and protein in retinas during all stages of the angiogenic response in the oxygeninduced retinopathy model.41 The functional significance of PAI-1 in retinal NV was determined by subjecting PAI-1 knockout mice to the oxygen protocol. There was about an 80% decrease in the extent of NV in these PAI1 knockout mice.41 The pro-angiogenic effect of PAI-1 may be explained by the fact that PAI-1 protects the stroma from excessive proteolysis during endothelial cell invasion. Excessive proteolysis during angiogenesis may prevent the coordinated assembly of endothelial cells into mature capillary tubes. A precise balance between proteolytic enzymes and their inhibitors is essential for endothelial cell migration and differentiation into functional vessels. These observations explain, at least in part, the paradoxical finding of high PAI-1 levels in advanced cancer, and thus identify PAI-1 as a potential target for anti-angiogenic therapy. It has been proposed that at low doses, PAI-1 may promote tumor growth and angiogenesis, while at higher concentrations, it may act as an anti-angiogenic agent.42 Reduction of retinal NV has been shown in an animal model by intravitreal injection of human PAI-1.43 Also, studies in choroidal NV have shown that PAI-1 can exhibit both pro- and anti-angiogenic effects, depending on the dose.44 The expression of both urokinase and the MMPs is modulated at the level of gene transcription by a variety of factors, including oncogenes and growth factors. Both urokinase and MMPs are secreted in a latent form and require activation. The production of the active form of these enzymes is inhibited by specific proteinase inhibitors found in the ECM. Because of these multiple levels of control, it has now become clear that these enzymes are part of a "proteolytic cascade" that functions in the regulation of cell migration and invasion, the remodeling and turnover of the ECM, and the release and activation of specific growth factors that affect cell behavior.

266

A. Das and P. G. McGuire

6.

INTERACTION OF PROTEINASES WITH OTHER MOLECULES DURING OCULAR ANGIOGENESIS

6.1

TNF-alpha

In addition to VEGF, other factors, including tumor necrosis factor alpha (TNFa), are expressed in the retinas of humans with proliferative eye diseases.45-47 TNFa is a 26 kDa transmembrane protein that is processed by the TNF converting enzyme, TACE, to yield a 17 kDa soluble protein.48 TNFa functions through its binding to two receptors: p55, implicated in apoptosis and NFkB (Nuclear Factor kappa B) activation, and p75, involved in lymphocyte proliferation.49,50 The cytoplasmic domain of the TNFa receptor, p55, has an 80-residue "death domain" that can regulate the apoptotic pathway.51,52 An alternative response following p55 stimulation is the activation of NFkB, which may result in a variety of cellular responses including the transcriptional regulation of select members of the MMP family of proteinases.52-54 In an animal model of retinal NV, we found increases in TNFa mRNA in the retinas on days 13 and 15.55 Isolated retinal endothelial cells did not significantly increase MMP production directly in response to a hypoxic stimulus, but required the presence of exogenous TNFa. TNFa increased the expression of MMP-3, MMP-9 and MT1-MMP in these cells. The levels of TACE and p55, proteins important in mediating the response of cells to TNFa, were increased by the angiogenic protein, VEGF, which is elevated in the retinas during NV.55 These findings support the hypothesis that growth factors such as TNFa and VEGF have a role in the regulation of extracellular proteinase expression during retinal NV.

6.2

Angiopoietin

The angiopoietins are the known ligands for the Tie receptors, which are endothelial cell-specific tyrosinase kinase receptors implicated in vascular growth and development. There are four definitive members of the angiopoietin family.56 It has been hypothesized that angiopoietin 2 (Ang2) might provide a key destabilizing signal involved in initiating angiogenic remodeling. Destabilized vessels would be prone to regression in the absence of other growth factors; however, in the presence of VEGF, capillary endothelial cells are stimulated to proceed through angiogenesis.56 Increased expression of Ang2 mRNA has been shown in the retina during both normal development and NV in mice.57-59 Stimulation of cultured retinal endothelial

14. Extracellular Proteinases in Ocular Angiogenesis

267

cells with Ang1 and Ang2 resulted in increased expression of MMP-9.59 Inhibition of the binding activity of the angiopoietins in vivo suppressed retinal NV concomitant with a reduction in the expression of MMP-9. All this evidence points toward the upregulation of MMP-9 as an early response to angiopoietin/Tek interaction, causing the destabilization of blood vessels during retinal NV.59

7.

ANTI-PROTEINASE THERAPY IN TUMOR ANGIOGENESIS

Proteinase inhibitors have been used in several clinical trials in cancer because of their attractiveness as therapeutic targets. Because proteinases are expressed at the tumor site or in the surrounding stroma, the effect of these inhibitors would be localized to the tumor itself with minimal side effects.60,61 Results of several MMP inhibitors showed that although they are effective in some studies, these inhibitors work most effectively in earlystage cancer without metastasis. In fact, several clinical trials with MMP inhibitors in cancer have been terminated for lack of efficacy and the occurrence of side effects. The most severe side effect noted with MMP inhibitors is tendonitis affecting shoulder, hand, and knee joints (MMP activity is required for maintenance of adult healthy joints).60 So, what are the lessons from these cancer trials with MMP inhibitors? To design a new clinical trial with the MMP inhibitors, one needs to focus on the following questions: At what stage do the MMP inhibitors work most effectively? Which specific MMPs should be inhibited for optimal therapeutic effect? What is the role of MMPs in interaction with other proteinases, particularly uPA? Specific MMPs play roles in specific stages of tumor progression, depending on the tissue type. For example, MMP-11 and -14 are negative prognostic indicators for small-cell lung cancer, whereas this tumor type has undetectable expression of MMP-2. So, in selecting a specific anti-MMP therapy in this cancer, one needs to use Tanomastat, which targets MMP-11 and has very little activity against MMP-2. In a transgenic mouse model of pancreatic islet cell carcinogenesis, several anti-angiogenic agents (AGM1470, angiostatin, BB-94, and endostatin) were compared for their effects at three distinct stages of cancer.62 The MMP inhibitor, BB-94, had a distinct efficacy profile in this model. It produced 49% reduction in the angiogenic islands in the prevention trial and 83% reduction in tumor burden in the intervention trial. However, it had no effect on regression of large tumors and invasive carcinoma. Thus, anti-angiogenic drugs may prove most efficacious when they are targeted to specific stages of cancer.

268

A. Das and P. G. McGuire

Since uPAR and uPA have been implicated in tumor and pathological angiogenesis, several anti-urokinase approaches have been tested in preclinical models. Anti-uPAR antibodies and the amino-terminal fragment (ATF) of uPA have been reported to inhibit tumor angiogenesis. The majority of validation studies have focused on blocking the interaction between uPA and uPAR. An oral non-cytotoxic small molecule, WX-UK1 (Wilex, Munich), an inhibitor of the uPA system, is currently in a Phase I/II clinical trial in patients with breast cancer in combination with an oral chemotherapeutic agent, Capecitabine.

8.

ANTI-PROTEINASE THERAPY IN OCULAR ANGIOGENESIS

Because proteinases are attractive targets for therapy, we also tested the efficacy of several proteinase inhibitors in retinal and choroidal NV models.

8.1

Reduction of Retinal NV by Inhibition of the MMPs

We used a broad-spectrum matrix metalloproteinase inhibitor, BB-94 (British Biotech Pharmaceuticals, Oxford, UK), in the retinal NV mouse model. BB-94 contains both a peptide backbone that binds it to MMPs and a hydroxamic acid group that binds it to the catalytic zinc atom they contain. Intraperitoneal (IP) injections of BB-94 have been shown to inhibit the growth of human ovarian carcinoma xenografts and murine melanoma metastasis.63,64 Upon histological examination, counts of neovascular nuclei revealed a 72% reduction in retinal NV in animals receiving a 1 mg/kg dose of BB-94 compared to control animals receiving saline as a placebo.16 The retinas of BB-94 treated animals also showed a significant decrease in the levels of active forms of MMP-2 and MMP-9, indicating that the drug reached the retinal tissues at this concentration.

8.2

Reduction of Retinal NV by Inhibition of the uPA/uPAR system

For these studies, we first systemically administered a urokinase inhibitor, BB-428 (4-substituted benzo-thiophene-2-carboxamidine), which has been shown to inhibit tumor growth and invasion in models of prostate cancer and mammary adenocarcinoma.65,66 In our laboratory, we found that BB-428 inhibited retinal NV in an animal model by 58% and significantly decreased the activity of uPA in retinas of treated animals.67

14. Extracellular Proteinases in Ocular Angiogenesis

269

A novel peptide, A6 (Angstrom Pharmaceuticals, San Diego, CA), derived from the receptor-binding region of urokinase, was also used in retinal and choroidal NV models. A6 inhibits the interaction of uPA with uPAR at the cell surface and has been shown to inhibit glioblastoma and breast cancer growth and metastasis without any direct cytotoxic effects.68,69 The anti-angiogenic activity of this peptide has been associated with a significant decrease in the density of blood vessels in these tissues. One mechanism of inhibition of new vessels may be a decrease in transforming growth factor beta activity and expression of the VEGF receptor Flk-1 as a direct or indirect result of the inhibition of the uPA-uPAR system.70 Alternatively, the uPA-uPAR interaction is required in the PAI-1 mediated recycling of uPAR and associated integrins that facilitate cell detachment from components of the ECM. Inhibition of the uPA-uPAR system by A6 might therefore be expected to disrupt this recycling process, causing increased cell-matrix adhesion and rendering the cells immobile. The amino-terminal fragment (ATF), an angiostatic molecule that targets the uPA/uPAR system and inhibits endothelial cell migration, was used in an animal model of oxygen-induced retinopathy.71 Intravitreal injection of an adenoviral vector carrying the murine ATF reduced retinal NV by 78% in this mouse model. We have injected the A6 peptide intraperitoneally at a dose of 5, 10, or 100 mg/kg once a day on days 12 to 16 in a model of oxygen-induced retinopathy.72 Histological analysis of mice treated with A6 peptide showed significant (63% at the highest dose) inhibition of retinal NV, and the response was dose-dependent. The reduction of NV by A6 was nearly equal to that seen in the uPAR knockout mice.

8.3

Reduction of CNV by Inhibition of the MMPs

An orally administered selective MMP inhibitor, N-biphenyl sulfonylphenylalanine hydroxamic acid (BPHA), has been shown to reduce laserinduced CNV.73 In a separate experiment in a laser-induced rat CNV model, a non-peptide, small molecular weight, synthetic MMP inhibitor, AG3340 (prinomastat), was injected intravitreally to treat choroidal NV.74 Prinomastat was found to be effective when given at the time of induction of CNV in the rat model (prevention study), whereas administration of prinomastat 2 weeks after induction was not effective (regression study).

270

8.4

A. Das and P. G. McGuire

Reduction of CNV by Inhibition of the uPA/uPAR system

We have found the expression of uPAR to be significantly elevated in the choroid of mice with laser-induced CNV.75 The expression of uPAR was localized specifically to the new vessels within the subretinal space associated with a disruption of Bruch’s membrane. Systemic administration of the uPA/uPAR peptide inhibitor, A6, resulted in a significant reduction of CNV (up to 94%), and the response was found to be frequency- and dosedependent (Figure 2).75 The inhibitory effects of A6 on CNV were confirmed by another group using a similar model.76 Taken together, studies on both retinal and choroidal NV demonstrate that the uPA/uPAR system is important in facilitating the development of abnormal new vessels in the retina, and thus the uPA/uPAR interaction may represent a new target for the development of anti-angiogenic therapies for ocular NV.

8.5

Reduction of Retinal NV by Inhibition of the Angiopoietin/Tek system

We, along with others, reported increased expression of Ang2 in retinas during NV.57-59 To determine whether an inhibitor of the Ang2/Tek system, muTekdeltaFc (Amgen Washington, Seattle, WA), can suppress retinal NV, we injected this inhibitor intraperitoneally into experimental mice once a day (40 or 80 mg/kg) during days 12 –16. Analysis of mice treated with the Tie2/Tek soluble antagonist showed a significant decrease in the numbers of capillary tufts on the vitreal side of the inner limiting membrane. Quantification of neovascular nuclei showed up to 87% inhibition of retinal NV in the animal model compared to the IgG-treated control animals.59 Furthermore, this response was found to be dose-dependent. Interestingly, RT-PCR analysis of the retinas from the Tek-treated animals showed a nearly 80% inhibition of MMP-9 expression.59 These data suggest that the upregulation of proteinases in microvascular endothelial cells by Ang2 may be an important early response during the development of retinal NV.

14. Extracellular Proteinases in Ocular Angiogenesis

271

Figure 14-2. A6 treatment prevents NV in the laser-induced model of choroidal neovascularization (CNV). Representative images of retinal pigment epithelium–choroid whole mounts infused with fluorescein isothiocyanate–conjugated dextran 14 days after laser induction of CNV. Images are from mice treated with phosphate-buffered saline or A6 peptide at differing frequencies. The red circle roughly outlines the area of the laser burn. The fluorescence in the center of the burn area demonstrates the extent of new vessel formation under the retina. The surrounding fluorescence represents the normal choroidal vasculature. A, Mouse treated with phosphate-buffered saline twice a day for 14 days. B, Mouse treated with 100 mg/kg A6 peptide twice a day once a week. C, Mouse treated with 100 mg/kg A6 twice a day every third day. D, Mouse treated with 100 mg/kg A6 twice a day every day. E, A higher magnification image of the central region in section D. Only a few fluorescein isothiocyanate–conjugated dextran-labeled blood vessels could be seen in this area. (Reproduced from Das et al., Arch Ophthalmol 122:1844-1849, 2005, American Medical Association).

272

9.

A. Das and P. G. McGuire

CONCLUSIONS

Because the upregulation and activation of proteinases and growth factors represents a final common pathway in the process of ocular NV, pharmacological intervention in these pathways may be useful as an alternative therapeutic approach to the treatment of proliferative retinopathy and exudative age-related macular degeneration.77 A clinical trial using an MMP inhibitor (AG3340, Agouron Pharmaceuticals) in patients with subfoveal CNV in age-related macular degeneration was recently terminated. We think such negative results with this MMP inhibitor do not mean that the drug is ineffective. It is possible that the drug did not reach the target tissues in effective concentrations and that the dose of the drug was insufficient. It can also be speculated that this MMP inhibitor is not effective at advanced stages of the angiogenic process and does not cause vessel regression. Based upon the pre-clinical results using A6 in our lab, two clinical trials (Phase I) have been completed, and the drug was found to be safe and well tolerated and did not trigger any immunogenic response.78,79 The A6 will be further tested in Phase II/III trials in patients with exudative age-related macular degeneration. As more and more novel anti-angiogenic drugs are being tested in clinical trials in patients with ocular angiogenesis, combination therapy with several anti-angiogenic drugs may be the ideal approach to completely inhibit NV, and proteinase inhibitors may play a significant role in this combination therapy.

ACKNOWLEDGMENTS This work was supported by grant RO1 12604-07 (to A.D.) from the National Eye Institute.

REFERENCES 1. M.S. Pepper and R. Montesano, Proteolytic balance and capillary morphogenesis, Cell Differentiation & Development 32(3), 319-327, (1990). 2. P. Mignatti and D.B. Rifkin, Plasminogen activators and matrix metalloproteinases in angiogenesis, Enzyme and Protein. 49, 117-137, (1996). 3. M.S. Pepper, D. Belin, R. Montesano, L. Orci, and J.D. Vassali, Transforming growth factor beta-1 modulates basic fibroblast growth factor-induced proteolytic and angiogenic properties of endothelial cells in vitro. J. Cell Biol. 111(2), 743-755, (1990). 4. M.S. Pepper, J.D. Vassali, J.W. Wilks, L. Schweigerer, L. Orci, and R. Montesano, Modulation of microvascular endothelial cells proteolytic properties by inhibitors of angiogenesis, J. Cellular Biochem. 55(4), 419-434, (1994).

14. Extracellular Proteinases in Ocular Angiogenesis

273

5. P.H.A. Quax, N. Pedersen, M.T. Masucci, E.J. Weening-Verhoeff, K. Dano, J.H. Verheijen, and F. Blasi, Complementation of urokinase and its receptor in extracellular matrix degradation, Cell Regulation 2, 793-803, (1991). 6. R.W. Stephens, J. Pollanen, H. Tapiovaara, K.C. Leung, P.S. Sim, E.M. Salonen, E. Ronne, N. Behrendt, K. Dano, and A. Vaheri, Activation of pro-urokinase and plasminogen on human sarcoma cells: a proteolytic system with surface-bound reactants, J. Cell Biol. 108, 1987-1995, (1989). 7. N. Manchanda and B. S. Schwartz, Single chain urokinase. Augmentation of enymatic activity upon binding to monocytes, J. Biol. Chem. 266, 14580-14584, (1991). 8. R. Blasi, Urokinase and urokinase receptor: a paracrine/autocrine system regulating cell migration and invasiveness, Bioassays 15, 105-111, (1993). 9. S.A. Rabbani, J. Desjardins, A.W. Bell, D. Banville, A. Mazar, J. Henkin, and D. Goltzman, An amino-terminal fragment of urokinase isolated from a prostate cancer cell line is mitogenic for osteoblast-like cells Biochem. Biophys. Res. Commun. 173, 1058-1064, (1990). 10. S. A. Rabbani and A. P. Mazar, The role of plasminogen activation system in angiogenesis and metastasis, Surg. Oncol. Clin. N. Amer. 10, 393-415, (2001). 11. Y. Wei, D. A. Waltz, N. Rao, R. J. Drummond, S. Rosenberg, and H. A. Chapman, Identification of the urokinase receptor as an adhesion receptor for vitronectin, J. Biol. Chem. 269, 32380-32388, (1994). 12. F. Blasi and P. Carmeliet, uPA: a versatile signaling orchestrator, Nat. Rev. Mol. Cell. Biol. 3, 931-943, (2002). 13. W. G. Stetler-Stevenson, The role of matrix metalloproteinases in tumor invasion, metastasis, and angiogenesis, Surg. Oncol. Clin. N. Amer. 10, 383-392, (2001). 14. M. S. Pepper, Role of the matrix metalloproteinase and plasminogen activator-plasmin systems in angiogenesis, Arterioscler. Thromb. Vasc. Biol. 21, 1104-1117, (2001). 15. M. Egeblad and Z. Werb, New functions for the matrix metalloproteinases in cancer progression, Nat. Rev. Cancer 2, 161-174, (2002). 16. A. Das, A. McLamore, W. M. Song, and P. G. McGuire, Retinal neovascularization is suppressed with a matrix metalloproteinase inhibitor. Arch. Ophthalmol. 117, 498-503, (1999). 17. M. Hangai, N. Kitaya, J. Xu, C. K. Chan, J. J. Kim, Z. Werb, S. J. Ryan, and P. C. Brooks, Matrix metalloproteinase-9-dependent exposure of a cryptic migratory control site in collagen is required before retinal angiogenesis, Am. J. Pathol. 161, 1429-1437, (2002). 18. P. G. McGuire, T. R. Jones, N. Talarico, E. Warren, and A. Das, The Urokinase/ Urokinase Receptor System in Retinal Neovascularization: Inhibition by Å6 Suggests A New Therapeutic Target. Invest. Ophthalmol. Vis. Sci. 44, 2736-2742, (2003). 19. A. Das, P. G. McGuire, C. Eriqat, R. R. Ober, E. DeJuan, G. A. Williams, A. McLamore, J. Biswas, and D. W. Johnson, Human diabetic neovascular membranes contain high levels of urokinase and metalloproteinase enzymes, Invest. Ophthalmol. Vis. Sci. 40, 809-813, (1999). 20. V. Lambert, C. Munaut, P. Carmeliet, R. D. Gerard, P. Declerck, A. Gils, C. Claes, J. M. Foidart, A. Noel, and J. M. Rakic, Dose-dependent modulation of choroidal neovascularization by plasminogen activator inhibitor type 1: Implications for clinical trials, Invest. Ophthalmol. Vis. Sci. 44, 2791-2797, (2003). 21. A. Das, N. Boyd, T. R. Jones, N. Talarico, and P. G. McGuire, Inhibition of Choroidal Neovascularization by a Peptide Inhibitor of the Urokinase Plasminogen Activator and Receptor System in a Mouse Model, Arch. Ophthalmol. 122, 1844-1849, (2004).

274

A. Das and P. G. McGuire

22. S. Giebel, G. Menucucci, P. McGuire, and A. Das, Matrix Metalloproteinases in Early Diabetic Retinopathy and Their Role in Alteration of the Blood-Retinal Barrier. Lab. Invest. 85, 597-607, (2005). 23. A. El-Remessy, M. A. Behzadian, G. Abou-Mohamed, T. Franklin, R. W. Caldwell, and R. B. Caldwell, Experimental diabetes causes breakdown of the blood-retinal barrier by a mechanism involving tyrosine nitration and increases in expression of vascular endothelial growth factor and urokinase plasminogen activator receptor. Am. J. Pathol. 162, 1995-2004 (2003). 24. D. A. Antonetti, A. J. Barber, S. Khin, E. Lieth, J. M. Tarbell, and T. W. Gardner, Vascular permeability in experimental diabetes is associated with reduced endothelial occludin content, Diabetes 47, 1953-1959, (1998). 25. D. A. Antonetti, A. J. Barber, L. A. Hollinger, E. B. Wolpert, and T. W. Gardner, Vascular endothelial growth factor induces rapid phosphorylation of tight junction proteins occludin and zonula occuden, 1. J. Biol. Chem. 274, 23463-23467, (1999). 26. Herron, M. J. Banda, E. J. Clark, J. Gavrilovic, and Z. Werb, Secretion of MMPs by stimulated capillary endothelial cells: expression of collagenases and stromelysin activities is regulated by endogenous inhibitors, J. Biol. Chem. 261, 2814-2818, (1986). 27. E. Baramova and J. M. Foidart, Matrix metalloproteinase family, Cell Biol. Int. 19, 239-242, (1995). 28. M. Moses, The regulation of neovascularization by MMPs and their inhibitors. Stem Cells 15, 180-189, (1996). 29. M. A. Moses and R. A. Langer, A metalloproteinase inhibitor as an inhibitor of neovascularization, J. Cell. Biochem. 47, 230-235, (1991). 30. P. D. Brown, D. E. Kleine, E. J. Unsworth, and W. G. Stetler-Stevenson, Cellular activation of the 72 kD type IV procollagenase:TIMP-2 complex, Kidney Int. 43, 163-168, (1993). 31. D. Kleiner, A. Tuuttila, K. Tryggvason, and W. Stetler-Stevenson, Stability analysis of latent and active 72 kD type IV collagenase:the role of tissue inhibitor of metalloproteinase-2 (TIMP-2), Biochemistry 32, 1583-1587, (1993). 32. R. N. Farris, S. S. Apte, B. R. Olsen, K. Iwata, and A. H. Milam, Tissue inhibitor of metalloproteinase-3 is a component of Bruch’s membrane of the eye, Am. J. Pathol. 150, 323-328, (1997). 33. N. Della, P. Campochiaro, and D. Zack, Localization of TIMP-3 mRNA to the retinal pigment epithelium, Invest. Ophthalmol. Vis. Sci. 37, 1921-1924, (1996). 34. A. Ruiz, P. Brett, and D. Bok, TIMP-3 is expressed in the human retinal pigment epithelium, Biochem. Biophys. Res. Commun. 226, 467-474, (1996). 35. B. H. F. Weber, G. Vogt, R. C. Pruett, H. Stohr, and U. Felbor, Mutations in the tissue inhibitor of metalloproteinase-3 (TIMP-3) in patients with Sorsby’s fundus dystrophy, Nat. Genet. 8, 352-356, (1994). 36. R. N. Farris, S. S. Apte, P. J. Luthert, A. Bird, and A. H. Milam, Accumulation of tissue inhibitor metalloproteinase-3 in human eyes with Sorsby’s fundus dystrophy or retinitis pigmentosa, Br. J. Ophthalmol. 82, 1329-1334, (1998). 37. M. Kamei and J. G. Hollyfield, TIMP-3 in Bruch’s membrane:changes during aging and age-related macular degeneration, Invest. Ophthalmol. Vis. Sci. 40, 2367-2375, (1999). 38. S. Majka, P. G. McGuire, S. Colombo, and A. Das, The balance between proteinases and inhibitors in a murine model of proliferative retinopathy. Invest. Ophthalmol. Vis. Sci. 42, 210-215, (2001).

14. Extracellular Proteinases in Ocular Angiogenesis

275

39. P. A. Andreasen, B. Georg, L. R. Lund, A. Riccio, and S. N. Stacey, Plasminogen activator inhibitors: hormonally regulated serpins, Mol. Cell. Endocrinol. 68, 1-19, (1990). 40. U. P. Thorgeirson, C. K. Lindsay, D. W. Cottam, and D. E. Gomez, Tumor invasion, proteolysis, angiogenesis, J. Neuro-Oncology 18, 89-103, (1994). 41. A. Das, G. Menicucci, S. Giebel, E. Colombo, and P. McGuire, Plasminogen activator inhibitor 1 (PAI-1) in early diabetic retinopathy and retinal neovascularization, ARVO Meeting abstract, (2005). 42. A. Noel, K. Bajou, V. Masson, L. Devy, F. Frankenne, J. M. Rakic, V. Lambert, P. Carmeliet, and J. M. Foidart, Regulation of cancer invasion and vascularization by plasminogen activator inhibitor-1, Fibrinolysis and Proteolysis 13, 220-225, (1999). 43. J. S. Penn and V. S. Rajaratnam, Inhibition of retinal neovascularization by intravitreal injection of human rPAI-1 in a rat model of retinopathy of prematurity, Invest. Ophthalmol. Vis. Sci. 44, 5423-5429, (2003). 44. V. Lambert, C. Munaut, P. Carmeliet, R. D. Gerard, P. Declerck, A. Gils, C. Claes, J. M. Foidart, A. Noel, and J. M. Rakic, Dose-dependent modulation of choroidal neovascularization by plasminogen activator inhibitor type 1: Implications for clinical trials, Invest. Ophthalmol. Vis. Sci. 44, 2791-2797, (2003). 45. G. A. Limb, A. H. Chignell, W. Green, F. LeRoy, and D. C. Dumonde, Distribution of TNF and its reactive vascular adhesion molecules in fibrovascular membranes. Br. J. Ophthalmol. 80, 168-173, (1996). 46. D. Armstrong, A. Augustin, R. Spengler, A. Al-Jada, T. Nickola, F. Grus, and F. Koch, Detection of VEGF and TNF alpha in epiretinal membranes of proliferative diabetic retinopathy, proliferative vitreo-retinopathy and macular pucker, Ophthalmologica 212, 410-414, (1998). 47. J. Pranger, R. Meyer, M. Klein, H. Schatz, and A. Pfeiffer, TNF in the vitreous body. Increase in neovascular eye diseases and proliferative diabetic retinopathy, Med. Klin. 90, 134-137, (1995). 48. R. A. Black, C. T. Rauch, C. J. Kozlosky, J. J. Peschon, J. L. Slack, M. F. Wolson, B. J. Castner, K. L. Stocking, P. Reddy, S. Srinivasan, N. Nelson, N. Boiani, K. A. Schooley, M. Gerhart, R. Davis, J. N. Fitzner, R. S. Johnson, R. J. Paxton, C. J. March, and D. P. Cerritti, A metalloproteinase disintegrin that releases tumor necrosis factor alpha from cells. Nature 385, 729-733, (1997). 49. D. Wallach, E. E. Varfolomeev, N. L. Malinin, Y. V. Goltsev, A. V. Kovalenko, M. P. Boldin, Tumor necrosis factor receptor and Fas signaling mechanism, Ann. Rev. Immunol. 17, 331-367, (1999). 50. E. C. Ledgerwood, J. S. Pober, and J. R. Bradley, Recent advances in the molecular basis of TNF signal transduction, Lab. Invest. 79, 1041-1050, (1999). 51. L. A. Tartaglia, T. M. Ayres, G. H. W. Wong, and D. V. Goeddel, A novel domain within the 55 kd TNF receptor signals cell death, Cell 74, 845-853, (1993). 52. M. Grell, Tumor necrosis factor receptors in cellular signaling of soluble and membraneexpressed TNF, J. Inflamm. 47, 8-17, (1996). 53. A. R. Farina, A. Tacconelli, A. Vacca, M. Maroder, A. Gulino, A. R. Mackay, Transcriptional upregulation of MMP-9 expression during spontaneous epithelial to neuroblast phenotype conversion by SK-N-SH neuroblastoma cells, involved in enhanced invasivity, depends upon a GT-box and nuclear factor kappa B elements, Cell. Growth Differ. 10, 353-367, (1999).

276

A. Das and P. G. McGuire

54. M. Bond, A. H. Baker, and A. C. Newby, Nuclear factor kappa B activity is essential for MMP-1 and –3 upregulation in rabbit dermal fibroblasts, Biochem. Biophys. Res. Commun. 264, 561-567, (1999). 55. S. Majka, P. G. McGuire, and A. Das, Regulation of matrix metalloproteinase expression by tumor necrosis factor in a murine model of retinal neovascularization, Invest. Ophthalmol. Vis. Sci. 43, 260-266, (2002). 56. G. D. Yancopoulos, S. Davis, N. W. Gale, J. S. Rudge, S. J. Wiegand, and J. Holash, Vascular-specific growth factors and blood vessel formation, Nature 407, 242-248, (2000). 57. H. Oh, H. Takagi, K. Suzuma, A. Otani, M. Matsumara, and Y. Honda, Hypoxia and vascular endothelial growth factor selectively upregulate angiopoietin-2 in bovine microvascular endothelial cells, J. Biol. Chem. 274, 15732-15739, (1999). 58. S. F. Hackett, H. Ozaki, R. W. Strauss, K. Wahlin, C. V. Suri, P. Maisonpierre, G. Yancopoulos, and P. Campochiaro, Angiopoietin2 expression in the retina: upregulation during physiologic and pathologic neovascularization, J. Cell. Physiol. 184, 275-284, (2000). 59. A. Das, W. Fanslow, D. Cerretti, E. Warren, N. Talarico, and P. McGuire, Angiopoietin/Tek Interactions Regulate MMP-9 Expression and Retinal Neovascularization, Lab. Invest. 83, 1637-1645, (2003). 60. L. J. McCawley and L. M. Matrisian, Matrix metalloproteinases: multifunctional contributors to tumor progression, Mol. Med. Today 6, 149-156, (2000). 61. M. Egeblad and Z. Werb, New functions for the matrix metalloproteinases in cancer progression. Nat. Rev. Cancer 2, 161-174, (2002). 62. G. Bergers, K. Javaherian, K. M. Lo, J. Folkman, and D. Hanahan, Effects of angiogenesis inhibition on multistage carcinogenesis in mice. Science 284, 808-812, (1999). 63. B. Davies, P. D. Brown, N. East, M. J. Crimmin, and F. R. Balkwill, A synthetic matrix metalloproteinase inhibitor decreases tumor burden and prolongs survival of mice bearing human ovarian carcinoma xenografts, Cancer Res. 53, 2087-2091, (1993). 64. R. G. S. Chirivi, A. Garofalo, M. J. Crimmin, L. J. Bawden, A. Stoppacciaro, P. D. Brown, and R. Giavazzi, Inhibition of the metastatic spread and growth of B 16-BL6 murine melanoma by a synthetic matrix metalloproteinase inhibitor, Int. J. Cancer 58, 460-464, (1994). 65. S. A. Rabbani, P. Harakidas, D. J. Davidson, J. Henkin, and A. P. Mazar, Prevention of prostate cancer metastasis in vivo by a novel synthetic inhibitor of urokinase-type plasminogen activator, Int. J. Cancer 63, 840-845, (1995). 66. D. F. Alonso, D. F. Farias, V. Ladea, L. Davel, L. Puricelli, and E. Bal de Kier Joffe, Effects of synthetic urokinase inhibitors on local invasion and metastasis in a murine mammary tumor model, Breast Cancer Res. Treat. 40, 209-223, (1996). 67. A. Das, P. McGuire, and L. Xu, Retinal neovascularization is suppressed with an inhibitor of proteinase enzymes, B-428. ARVO Meeting Abstract (1999). 68. Y. Guo, A. A. Higazi, A. Arakelian, B. S. Scahias, D. Cines, R. H. Goldfarb, T. R. Jones, H. Kwaan, A. P. Mazar, and S. A. Rabbani, A peptide derived from the nonreceptor binding region of urokinase plasminogen activator inhibits tumor progression and angiogenesis and induces tumor cell death in vivo, FASEB J. 14, 1400-1410, (2000). 69. K. Mishima, A. Mazar, A. Gown, M. Dskelly, X. D. Ji, X. D. Wang, T. R. Jones, W. K. Cavenee, and H. J. Huang, A peptide derived from the nonreceptor binding region of urokinase palsminogen activator inhibits glioblastoma growth and angiogenesis in vivo in combination with cisplastin, Proc. Natl. Acad. Sci. USA 97, 8484-8489, (2000).

14. Extracellular Proteinases in Ocular Angiogenesis

277

70. Y. J. Guo, A. P. Mazar, J.–J. Lebrun, and S. A. Rabbani, An antiangiogenic urokinasederived peptide combined with tamoxifen decreases tumor growth and metastasis in a syngeneic model of breast cancer, Cancer Res. 62, 4678-4684, (2002). 71. L. Le Gat, K. Gogat, C. Bouwuet, M. Saint-Geniez, D. Darland, L. Van Den Gerghe, D. Marchant, A. Provost, M. Perricaudet, M. Menasche, and M. Abitbol, In vivo adenovirus-mediated delivery of a uPA/uPAR antagonist reduces retinal neovascularization in a mouse model of retinopathy, Gene Ther. 10, 2098-2103, (2003). 72. P. G. McGuire, T. Jones, N. Talarico, E. Warren, and A. Das, The Urokinase/Urokinase Receptor System in Retinal Neovascularization: Inhibition by A6 Suggests a New Therapeutic Target, Invest. Ophthalmol. Vis. Sci. 44, 2736-2742, (2003). 73. T. Kohri, M. Moriwaki, M. Nakajima, H. Tabuchi, and K. Shiraki, Reduction of experimental laser-induced choroidal neovascularization by orally administered BPHA, a selective metalloproteinase inhibitor, Graefes Arch. Clin. Exp. Ophthalmol. 241, 943-952, (2003). 74. M. El Bradey, L. Cheng, D. U. Bartsch, K. Appelt, N. Rodanant, G. Bergeron-Lynn, and W. R. Freeman, Prevention versus treatment effect of AG3340, a potent matrix metalloproteinase inhibitor in a rat model choroidal neovascularization, J. Ocul. Pharmacol. Ther. 20, 217-236, (2004). 75. A. Das, N. Boyd, T. R. Jones, N. Talarico, and P. G. McGuire, Inhibition of Choroidal Neovascularization by a Peptide Inhibitor of the Urokinase Plasminogen Activator and Receptor System in a Mouse Mode, Arch. Ophthalmol. 122, 1844-1849, (2004). 76. H. J. Koh, K. Bessho, L. Cheng, D. U. Bartsch, T. R. Jones, G. Bereron-Lynn, and W. R. Freeman, Inhibition of choroidal neovascularization in rats by the urokinasederived peptide A6, Invest. Ophthalmol. Vis. Sci. 45, 635-640, (2004). 77. A. Das and P. G. McGuire, Retinal And Choroidal Angiogenesis: Pathophysiology & Strategies For Inhibition, Progress in Retinal and Eye Research 22, 721-748, (2003). 78. A. R. van Toostenburg, D. Lee, T. R. Jones, J. A. Dycj-Jones, M. H. Silverman, G. N. Lam, and S. J. Warrington, Safety, tolerability and pharmacokinetics of subcutaneous A6, an 8-amino acid peptide with anti-angiogenic properties, in healthy men, Int. J. Clin. Pharmacol. Ther. 42, 253-259, (2004). 79. A. Berkenblit, U. A. Matulonis, J. F. Kroener, B. J. Dezube, G. N. Lam, L. C. Cuasay, N. Brunner, T. R. Jones, M. H. Silverman, and M. A. Gold, A6, a urokinase plasminogen activator (uPA)-derived peptide in patients with advanced gynecologic cancer: a phase I trial, Gynecol. Oncol. 99, 50-57, (2005).

Chapter 15 OXYGEN-INDEPENDENT ANGIOGENIC STIMULI

Jonathan M. Holmes,1 David A. Leske,1 and William L. Lanier2

Departments of 1Ophthalmology and 2Anesthesiology, Mayo Clinic College of Medicine, Rochester, Minnesota

Abstract:

1.

Although much research has focused on the role of hypoxia and hyperoxia in preretinal neovascularization, there is growing evidence that other factors play a role. Carbon dioxide, acidosis, alkalosis, systemic infection, systemic growth retardation, and perturbations in the thyroxine and insulin-like growth factor (IGF-1) hormone axes all appear to be important risk factors in the pathogenesis of retinopathy of prematurity (ROP) and inducers of preretinal neovascularization in the immature retinae. Further advances in the prevention of ROP may require interventions directed at these oxygen-independent angiogenic stimuli.

INTRODUCTION

Retinopathy of prematurity (ROP) is a blinding disease of premature infants that, in its advanced stages, is characterized by preretinal neovascularization. Although excess inspired oxygen was identified as the primary risk factor for development of ROP almost 50 years ago,1,2 reduction of supplemental oxygen exposure for premature infants has failed to eliminate severe ROP.3,4 Multivariate analyses of retrospective clinical datasets have raised many alternative candidate risk factors in the pathogenesis of ROP, but such retrospective studies are limited by lack of independence of potential risk factors and incomplete data acquisition. Animal models of ROP provide an opportunity to study individual candidate risk factors, while allowing control of other potential confounders. The rat model for ROP has been described in previous chapters of this text. To briefly restate the critical features: the retinal vasculature of the neonatal 279 J.S. Penn (ed.), Retinal and Choroidal Angiogenesis, 279–288. © Springer Science+Business Media B.V. 2008

280

J. M. Holmes

rat is incompletely developed at birth, with a large avascular peripheral retina analogous to the premature human infant. Studies using this model make the assumption that exposing the neonatal rat retina to stimuli (e.g. hyperoxia) during the first few days of life is analogous to exposing premature human retina to those stimuli. Neovascularization in the rat model primarily develops at the junction of the vascular and avascular retinae, in the same way that stage 3 ROP develops in the retinae of human premature infants. Several laboratories have studied the role of fluctuating hyperoxia and hypoxia on the development of preretinal neovascularization in the rat model, a condition termed “oxygen-induced retinopathy” (OIR).5-10 In most OIR rat models, newborn pups are exposed to periods of hyperoxia, alternating with periods of absolute or relative hypoxia, for a total of 7 to 14 days, and then retinae are evaluated after a further period of room air recovery ranging from 0 to 6 or more days. In our laboratory, the period of oxygen exposure is 7 days, with 5 days of recovery, and analysis at day 13 using primarily ADPase staining methods11 and masked grading. We have primarily used an “expanded litter” design, where rats are raised in foster litters of 25 by one mother. Such expanded litters induce growth retardation,5,12 which we have found to be associated with increased incidence and severity of neovascularization.5 We believe that standardizing this growth retardation is important, since animals raised in different sized litters have different rates of vascular development.12 OIR in rats and mice13,14 appears to be mediated primarily by vascular endothelial growth factor (VEGF),9 analogous to ROP in premature infants. Nevertheless, other non-hypoxic, non-hyperoxic stimuli also appear to induce neovascularization in the neonatal rat, and many of these stimuli have clinical relevance to ROP in premature infants. In any discussion of “oxygen-independent” stimuli, a caveat is needed. To date, there is no direct evidence that the stimuli we describe are mediated secondarily by hypoxia or hyperoxia. Nevertheless, it is entirely possible that some of these stimuli might be acting through local changes in the oxygen environment. With that caveat, we will describe a number of oxygenindependent factors that induce preretinal neovascularization in neonatal rats, providing additional animal models of ROP.

2.

CARBON DIOXIDE

Premature infants who never experience hyperoxia, for example those with cyanotic congenital heart disease, may develop ROP.15 For those specific infants, and for premature infants in general, raised arterial carbon dioxide

15. Oxygen-Independent Angiogenic Stimuli

281

(termed hypercarbia or hypercapnia) has been suggested as a risk factor for ROP.16 In an initial study,17 we reported increased severity of OIR when 10% CO2 was added to the inspired fluctuating (80% to 10%) oxygen environment. We also reported that inspired 10% CO2 retarded normal retinal vasculogenesis in neonatal rats.18 We then studied whether CO2 alone would induce preretinal neovascularization. Complicating such a study is the effect of inspired CO2 on the arterial partial pressure of oxygen (PaO2).19 For each level of inspired O2, the PaO2 is higher in 10% CO2 compared to 0.2% CO2.19 We speculated that neonatal rats breathing a mixture of O2 and 10% CO2 become even more efficient at gas exchange (in part because their exaggerated respiration caused them to inspire gases that had not completely saturated with water vapor), accounting for the high PaO2 levels we observed. Regardless of the mechanism, studies that address increased inspired CO2 must account for increased PaO2 levels when breathing inspired CO2. In our study of the effects of CO2 on developing retina19 we created two experimental groups: (1) high inspired CO2 and (2) pure hypercarbia (where the inspired O2 was reduced to match normoxic PaO2 values), each followed by 5 days of room air recovery, analogous to our OIR models. We found that either high inspired CO2 or pure hypercarbia induced mild but distinct preretinal neovascularization at an incidence of 19% or 14%, respectively. No room air-exposed controls exhibited preretinal neovascularization. We termed this condition “carbon-dioxide-induced retinopathy” (CDIR).19 We have speculated19 that CDIR might in fact be mediated by increased retinal blood flow and therefore increased oxygen delivery to the local retinal environment, but further technological advances in measuring local oxygen concentrations in the retina are needed to confirm or refute this hypothesis. Alternatively, we also speculated19 that CDIR might be mediated by direct damage to the developing endothelium by acidosis, since raised PaCO2 is associated with reduced pH. This hypothesis led us to our next series of experiments on acidosis.

3.

ACIDOSIS

In order to render neonatal rats acidotic, we administered ammonium chloride (NH4Cl) by oro-gastric gavage.20 In a preliminary arterial blood gas study, we determined that a single dose of NH4Cl (10 mmol/kg) would induce maximum arterial blood acidosis to a pH of 7.10 at 3 hours following gavage, and that at 12 hours post-gavage the pH was still reduced at 7.23.20

282

J. M. Holmes

We then discovered that giving NH4Cl twice daily from days 2 to 7 of life, followed by 5 days of recovery, induced preretinal neovascularization in 36% of neonatal rats, compared to 5% of control animals receiving saline gavage.20 Neovascularization was confirmed in acidotic animals by crosssectional histology. It is unclear why a few saline gavaged control animals developed neovascularization. We speculated that twice-daily oro-gastric gavage may have reduced feeding and exacerbated growth retardation by inducing handling-related physiological stress. We termed the retinopathy induced by NH4Cl “metabolic acidosisinduced retinopathy” (MAIR),20 although we currently refer more generically to “acidosis-induced retinopathy” (AIR). We speculate20 that acidosis per se damages the developing retinal vasculature. Although there were minor changes in arterial PaO2 in acidotic animals, possibly due to compensatory hyperventilation, the PaO2 levels (108 mm Hg) were very near the normal range for rats of this age, in contrast to the levels encountered in OIR models (300 to 400 mm Hg). Nevertheless, analogous to CDIR, we have not ruled out local effects of acidosis, which might result in local vasodilation and increased local delivery of oxygen. To confirm the concept of an AIR in neonatal rats analogous to ROP, we then studied alternative pharmacological means of inducing a systemic acidosis in neonatal rats.21 Acetazolamide induces acidosis by inhibiting the ubiquitous enzyme carbonic anhydrase, resulting in bicarbonate loss from the kidney with subsequent systematic acidosis. This is in contrast to NH4Cl, which induces acidosis by providing a hydrogen ion load. In an initial arterial blood gas study,21 we selected two doses of intraperitoneal acetazolamide (50 mg/kg and 200 mg/kg), which induced moderate or severe acidosis, respectively. Studies of long-term arterial blood gases confirmed that the twice-daily dosing regime maintained a fairly stable level of acidosis over the period of drug exposure (pH 7.22 for the moderate dose and 7.13 for the high dose). Parallel studies confirmed that the high dose of intraperitoneal acetazolamide (200 mg/kg) and NH4Cl gavage (10 mmol/kg) induced similar severities of acidosis.21 Examining the retinae of rats who received these doses of acetazolamide for 7 days followed by 5 days of recovery, along with those of saline injected controls, revealed no preretinal neovascularization with the moderate dose but a 58% incidence in rats who received the high dose.21 These data confirm a dose-dependent and pH-dependent AIR in neonatal rats, regardless of the method of induction of acidosis. Again, we noted small increases in PaO2 with acetazolamide,21 but only 15 to 25 mm Hg above room air levels, and we speculated that these small changes were due to increased respiratory rate, changes in pulmonary artery pressure, and distribution of gas and blood flow within the lung. Although

15. Oxygen-Independent Angiogenic Stimuli

283

we believe that changes in arterial oxygen are not mediating the retinal neovascularization we observed, we cannot rule out local changes that might increase local delivery of oxygen. Following our initial studies with NH4Cl20 and acetazolamide21, we went on to investigate the relationship between duration of acidosis exposure and duration of recovery in the incidence and severity of neovascularization.22 We found that even one day of acidosis exposure was sufficient to induce mild preretinal neovascularization.22 We found that neovascularization appeared following 3 or 6 days of acidosis, even without a period of recovery, although it was maximal after 2 to 5 days of recovery.22 Longer-term followup of the rats revealed spontaneous resolution of preretinal neovascularization by day 20.22 In this respect, AIR shares similar features with clinical ROP; infants develop ROP during the period when they are still being exposed to the inducing factors, and at least 50% of stage 3 ROP (neovascularization) subsequently resolves spontaneously. The clinical relevance of AIR deserves comment. Premature infants have immature lungs, suffer episodes of apnea and bradycardia, and may have episodes of sepsis. As a result, the premature infant often experiences episodes of combined respiratory and metabolic acidosis. In addition, some neonatalogists are now advocating early weaning from ventilator support to reduce the incidence of barotrauma to the lung.23 Such an approach to ventilator management necessitates allowing the arterial blood CO2 to rise and the pH to fall: so-called “permissive hypercapnia.” Our studies on acidosis-induced retinopathy raise the issue of whether such an approach might be detrimental to the eyes of the developing infant, but any concern must be balanced against the welfare of the entire infant.

4.

BICARBONATE

Sodium bicarbonate is used intravenously in the neonatal intensive care unit as one possible treatment for severe acidosis.24 We conducted a series of experiments to investigate (1) whether bicarbonate would have a detrimental effect on the developing vasculature and (2) whether treatment of the underlying acidosis would prevent or reduce the severity of AIR.25 Initial arterial blood gas studies confirmed that bicarbonate oro-gastric gavage (15 mmol/kg twice daily) induced a systemic alkalosis (pH 7.55).25 Administering bicarbonate from days 2 to 7, followed by 5 days of recovery, induced a mild and somewhat rare retinopathy. Preretinal neovascularization was seen in 9% of rats treated with 15 mmol/kg bicarbonate twice daily and 8% of those treated with 20 mmol/kg once daily.25

284

J. M. Holmes

A further experiment was conducted where rats were made acidotic with intraperitoneal injections of acetazolamide (one of our AIR models described above).21 Rats were then given either bicarbonate (to partially normalize pH) or saline control via oro-gastric gavage. Unfortunately, the mortality rate in the animals that received both acetazolamide and bicarbonate was particularly high, but the incidence of neovascularization was reduced from 24% to 8% (albeit not statistically significant at the p