Endothelial Cells in Health and Disease

Citation preview

Endothelial Cells in Health and Disease

Endothelial Cells in Health and Disease Edited by

William C.Aird Harvard Medical School, Cambridge, Massachusetts, U.S.A. Beth Israel Deaconess Medical Center, Boston, Massachusetts, U.S.A.

Boca Raton London New York Singapore

This edition published in the Taylor & Francis e-Library, 2005. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/. Cover Illustration: Steven Moskowitz Published in 2005 by Taylor & Francis Group 6000 Broken Sound Parkway NW Boca Raton, FL 33487–2742 © 2005 by Taylor & Francis Group, LLC No claim to original U.S. Government works ISBN 0-203-02595-4 Master e-book ISBN

International Standard Book Number-10: 0-8247-5424-7 (Print Edition)(Hardcover) International Standard Book Number-13: 978-0-8247-5424-2 (Print Edition) (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access http://www.copyright.com/ (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-forprofit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Catalog record is available from the Library of Congress

Taylor & Francis Group is the Academic Division of T&F Informa plc. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com

Contents Foreword Preface Contributors

1. The Endothelium as an Organ William C.Aird 2. Blood-Brain Barrier Eric V.Shusta 3. Lymphatic Endothelium Satoshi Hirakawa and Michael Detmar 4. High Endothelial Venules Jean-Marc Gauguet, Roberto Bonasio and Ulrich H.von Andrian 5. The Use of Proteomics to Map Phenotypic Heterogeneity of the Endothelium Johanna Lahdenranta, Wadih Arap and Renata Pasqualini 6. The Use of Genomics to Map Phenotypic Heterogeneity of the Endothelium Mary E.Gerritsen, Stuart Hwang, Constance Zlot, James Tomlinson and Michael Ziman 7. The Role of Genetic Predeterminants in Regulating the Phenotypic Heterogeneity of the Endothelium Brant M.Weinstein 8. The Use of Fate Mapping Studies to Follow Lineage Determination of the Endothelium David E.Reese and Takashi Mikawa 9. Oxygen Regulation of Endothelial Cell Phenotypes Yasushi Yoshikawa, Maksim Fedarau, Koichiro Iwanaga, Hiroaki Harada and David J.Pinsky 10. Fluid Mechanical Forces as Extrinsic Modifiers of Endothelial Function Johannes R.Kratz, Kush Parmar, Sripriya Natarajan and Guillermo GarcíaCardeña 11. Vascular Bed-Specific Signaling and Angiogenesis Napoleone Ferrara, Rui Lin and Jennifer LeCouter 12. Differential Regulation of Endothelial Cell Barrier Function Jeffrey R.Jacobson, Steven M.Dudek and Joe G.N.Garcia

viii x xv

1 38 74 87 119

134

150

168

187

212

228 243

13. Differential Regulation of Leukocyte-Endothelial Cell Interactions D.NeilGranger and Karen Y.Stokes 14. Vascular Biology of the Placenta Hartmut Weiler 15. Blood Endothelial Cells Robert P.Hebbel and Anna Solovey 16. Determination of Endothelial Heterogeneity by the Recruitment of Bone Marrow Derived Endothelial Progenitors Shahin Rafii and Jay Edelberg 17. Transcriptional Networks and Endothelial Lineage Peter Oettgen 18. The Diversity of Vascular Disease: A Clinician’s Perspective John P.Cooke 19. Molecular Targets of Tumor Vasculature Eleanor B.Carson-Walter and Brad St. Croix 20. The Role of the Endothelium in Severe Sepsis and Multiple Organ Dysfunction William C.Aird 21. The Hepatic Sinusoidal Endothelial Cell as a Primary Target of Disease Rimma Shaposhnikov and Laurie D.DeLeve 22. Endothelium and Hemostasis William C.Aird 23. Thrombotic Microangiopathies: Role of Microvascular Endothelium in Pathogenesis Thomas O.Daniel 24. Pulmonary Circulation and Pulmonary Hypertension Troy Stevens, Michael Kasper, Carlyne Cool and Norbert Voelkel 25. Endothelial Cell Phenotypes Associated with Organ Transplantation Simon C.Robson Index

257 276 301 320

335 356 383 403

420 437 451

476 506

551

Foreword As a clinical hematologist and vascular biologist, I have spent my career trying to understand the biology of the endothelium. In doing so, I have come to appreciate that this cell layer eludes traditional approaches to investigation, diagnosis, and treatment. The endothelium is distributed throughout the body and behaves as a multifunctional system—a communication network of sorts. The historical fragmentation of biomedicine into organ-specific disciplines has detracted from a full consideration and understanding of the endothelium. Simply put, the endothelium meets every criterion for an organ, and as such warrants attention in its own right. There exists a large gap between our knowledge of cultured endothelial cells and the intact endothelium, and an even wider gap between the bench and the bed-side. Future breakthroughs in endothelial cell biology will depend on the systematic integration of multiple disciplines, including—but certainly not limited to—cardiology, pulmonary medicine, nephrology, neurology, hematology, hepatology, molecular and cell biology, proteomics, genomics, complex systems biology, and molecular diagnostics. Endothelial Cells in Health and Disease represents a laudable first effort to synthesize—and indeed create—a new field in endothelial biomedicine. By weaving together a series of chapters from multiple disciplines, the Editor has succeeded in “bringing the endothelium to life.” After finishing the book, you will have a new-found appreciation for the robustness and versatility of this organ system and a much clearer insight into its important role in health and disease. This book marks the beginning of an exciting new era in “endotheliology.” Robert D.Rosenberg Department of Medicine, Division of Molecular Medicine and Hematology-Oncology, Beth Israel Deaconess Medical Center, Boston, Massachusetts, U.S.A. Department of Biology, Massachusetts Institute of Technology, Cambridge, U.S.A.

Preface Until the mid 19th century, the endothelium was considered to be little more than an inert layer of nucleated cellophane. In the 1950s and 1960s, the use of physiological assays and electron microscopy provided new insights into the role of the endothelium in inflammation. The development of cell culture techniques to harvest and grow endothelial cells from the human umbilical vein in the 1970s revolutionized the field of endothelial cell biology, leading to breathtaking advances in research and development. At the same time, there has been a growing awareness that the endothelium displays heterogeneous properties in vivo. Electron microscopy studies in the last century were the first to uncover site-specific differences in endothelial structure and function. Over the past two decades, additional avenues of research have uncovered previously unimaginable levels of complexity in the intact endothelium. It is now recognized that phenotypic heterogeneity is mediated, in large part, by the micro-environment. Thus, the very assay that helped to jumpstart the field of endothelial cell biology—namely, in vitro cell culture—has fallen short in providing insight into the spatial and temporal dynamics of this cell layer. This limitation has been circumvented by the recent development of novel assay systems to interrogate the endothelium in the context of its native environment, either in the living organism or in a reconstituted system in vitro. Such advances are beginning to paint a clearer picture of a cell layer that is teeming with life and every bit as active as other organ systems. When combined with imminent advances in diagnosis (e.g., molecular imaging and proteomics) and therapy (e.g., endothelial cell “attenuating” agents), the field of endothelial biomedicine has never looked so promising. Unfortunately, the exchange of information and ideas among vascular biologists has long been hampered by the existence of infrastructural barriers—intellectual, financial, cultural, and otherwise. Investigators who study endothelial cell biology in vascular beds outside the heart tend to interact with other members of their own organ-specific discipline—for example, blood-brain barrier experts with neuroscientists, pulmonary endothelium investigators with respiratory physiologists, molecular and cell biologists studying endothelium in sickle cell anemia with their hematology and oncology colleagues. There is little question that a full understanding of the endothelium in health and disease—including its potential as a therapeutic target—will require a more complete synthesis of the field. The purpose of this first edition of Endothelial Cells in Health and Disease is to begin the process of bringing together—for the first time—endothelial cell biologists from different disciplines to summarize recent progress in their respective fields. The book is divided broadly into five sections. The first section will provide an overview of the endothelium as an organ system followed by a consideration of historically recognized “specialized” endothelium, including the blood-brain barrier, lymphatics, and high endothelial venules. The second section highlights new and exciting proteomic and genomic techniques for mapping endothelial cell heterogeneity. The third

section is focused on epigenetic and environmental determinants of endothelial cell function, and includes chapters on genetic predetermination, fate mapping studies, oxygen tension, hemodynamics, and vascular bed-specific growth factor signaling. The fourth section covers a subset of phenotypes of the endothelium, including barrier function and leukocyte transmigration. The final section is devoted to a consideration of endothelium in different disease sates, including tumors, sepsis, sinusoidal obstruction syndrome, hypercoagulability, pulmonary hypertension, and transplantation. Additional chapters cover transcriptional networks, endothelial progenitors, circulating endothelial cells and, importantly, a clinician’s view of endothelial based-disease. A book such as this is necessarily limited in size and scope. As a result, several important topics have been excluded from the current edition. Perhaps the most conspicuous of these omissions is a discussion of coronary artery disease. This should not be construed as an oversight or lack of interest on my part. Rather, I have chosen to focus this book on other areas in vascular biology, with the assumption that those readers interested in pursuing the coronary bed have access to a wealth of excellent reviews in the literature. Many other important areas have been omitted, including a description of endothelium in the retina, the gastrointestinal tract, the skin, and the bone marrow; a consideration of angiogenesis; a summary of novel strategies for diagnosing endothelial-based diseases; and the application of evolutionary principles and complexity theory to an understanding of the endothelium. It is my goal to include these and related topics in subsequent editions of this book. I would like to thank all the contributors who worked hard to meet the deadlines and produced such excellent chapters. I am indebted to Mitchell Halperin, who gave me the courage to think and speak outside the box, to Bob Rosenberg who introduced me to the wonders of the endothelium, and to Michael Gimbrone who has been a constant source of inspiration. William C.Aird Department of Medicine, Beth Israel Deaconess Medical Center Harvard Medical School, Boston, Massachusetts, U.S.A.

Contributors William C.Aird Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, U.S.A. Wadih Arap The University of Texas, M.D.Anderson Cancer Center, Houston, Texas, U.S.A. Roberto Bonasio The CBR Institute for Biomedical Research, Inc. and Department of Pathology, Harvard Medical School, Boston, Massanchusetts, U.S.A. Eleanor B.Carson-Walter Department of Neurosurgery, University of PittsburghPittsburgh, Pittsburgh, Pennsylvania, U.S.A. John P.Cooke Stanford University School of Medicine, Falk CVRC, Stanford, California, U.S.A. Carlyne Cool Department of Pathology, University of Colorado Health Sciences Center, Denver, Colorado, U.S.A. Brad St. Croix Tumor Angiogenesis Laboratory, National Cancer Institute-Frederick, Frederick, Maryland, U.S.A. Thomas O.Daniel Ambrx, Inc., San Diego, California, U.S.A. Laurie D.DeLeve USC Research Center for Liver Diseases, Division of Gastrointestinal and Liver Diseases, Keck School of Medicine, University of Southern California, Los Angeles, California, U.S.A. Michael Detmar Cutaneous Biology Research Center and Department of Dermatology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, U.S.A. Steven M.Dudek Division of Pulmonary and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A. Jay Edelberg Cornell University Medical College, Ithaca, New York, U.S.A. Maksim Fedarau Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan, U.S.A. Napoleone Ferrara Department of Molecular Oncology, Genentech Inc., South San Francisco, California, U.S.A. Joe G.N.Garcia Division of Pulmonary and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A. Guillermo García-Cardeña Laboratory for Systems Biology, Center for Excellence in Vascular Biology, Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, U.S.A. Jean-Marc Gauguet The CBR Institute for Biomedical Research, Inc. and Department of Pathology, Harvard Medical School, Boston, Massachusetts, U.S.A. Mary E.Gerritsen Department of Vascular Biology, Millennium Pharmaceuticals, Inc., South San Francisco, California, U.S.A. D.Neil Granger Department of Molecular and Cellular Physiology, Louisiana State University Health Sciences Center, Shreveport, Louisiana, U.S.A.

Hiroaki Harada Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan, U.S.A. Robert P.Hebbel Department of Medicine, University of Minnesota Medical School, University of Minnesota, Minneapolis, Minnesota, U.S.A. Satoshi Hirakawa Cutaneous Biology Research Center and Department of Dermatology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, U.S.A. Stuart Hwang Department of Vascular Biology, Millennium Pharmaceuticals, Inc., South San Francisco, California, U.S.A. Koichiro Iwanaga Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan, U.S.A. Jeffrey R.Jacobson Division of Pulmonary and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A. Michael Kasper Department of Anatomy, Technische Hochschule Gustav Carus Universität, Dresden, Germany Johannes R.Kratz Laboratory for Systems Biology, Center for Excellence in Vascular Biology, Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, U.S.A. Johanna Lahdenranta The University of Texas, M.D.Anderson Cancer Center, Houston, Texas, U.S.A. Jennifer LeCouter Department of Molecular Oncology, Genentech Inc., South San Francisco, California, U.S.A. Rui Lin Department of Molecular Oncology, Genentech Inc., South San Francisco, California, U.S.A. Takashi Mikawa Department of Cell and Developmental Biology, Cornell University Medical College, Ithaca, New York, U.S.A. Sripriya Natarajan Laboratory for Systems Biology, Center for Excellence in Vascular Biology, Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, U.S.A. Peter Oettgen Division of Cardiology, Beth Israel Deaconess Medical Center, Boston, Massachusetts, U.S.A. Kush Parmar Laboratory for Systems Biology, Center for Excellence in Vascular Biology, Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, U.S.A. Renata Pasqualini The University of Texas, M.D.Anderson Cancer Center, Houston, Texas, U.S.A. David J.Pinsky Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan, U.S.A. Shahin Rafii Cornell University Medical College, Ithaca, New York, U.S.A. David E.Reese Department of Cell and Developmental Biology, Cornell University Medical College, Ithaca, New York, U.S.A. Simon C.Robson Liver Center, Research North, Beth Israel Deaconess Hospital, Boston, Massachusetts, U.S.A. Rimma Shaposhnikov Division of Gastrointestinal and Liver Diseases, Keck School of Medicine, University of Southern California, Los Angeles, California, U.S.A.

Eric V.Shusta Department of Chemical and Biological Engineering, University of Wisconsin-Madison, Madison, Wisconsin, U.S.A. Anna Solovey Department of Medicine, University of Minnesota Medical School, University of Minnesota, Minneapolis, Minnesota, U.S.A. Troy Stevens Department of Pharmacology, University of South Alabama, Mobile, Alabama, U.S.A. Karen Y.Stokes Department of Molecular and Cellular Physiology, Louisiana State University Health Sciences Center, Shreveport, Louisiana, U.S.A. James Tomlinson Department of Vascular Biology, Millennium Pharmaceuticals, Inc., South San Francisco, California, U.S.A. Norbert Voelkel Pulmonary Hypertension Center and Pulmonary and Critical Care Medicine Division, University of Colorado Health Sciences Center, Denver, Colorado, U.S.A. Ulrich H.von Andrian The CBR Institute for Biomedical Research, Inc. and Department of Pathology, Harvard Medical School, Boston, Massanchusetts, U.S.A. Hartmut Weiler The Blood Research Institute, Blood Center of Southeastern Wisconsin, Department of Physiology, Medical College of Wisconsin, Milwaukee, Wisconsin, U.S.A. Brant M.Weinstein Laboratory of Molecular Genetics, NICHD, NIH, Bethesda, Maryland, U.S.A. Yasushi Yoshikawa Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan, U.S.A. Michael Ziman Department of Vascular Biology, Millennium Pharmaceuticals, Inc., South San Francisco, California, U.S.A. Constance Zlot Department of Vascular Biology, Millennium Pharmaceuticals, Inc., South San Francisco, California, U.S.A.

1 The Endothelium as an Organ William C.Aird Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, U.S.A.

1. INTRODUCTION The endothelium, which lines the blood vessels of the vascular tree, is a truly pervasive cell layer, weighing 1 kg in an average-sized human and covering a total surface area of 4000–7000 m2 (1). Endothelial cells from a single human, when lined end-to-end, would wrap more than four times around the circumference of the earth. The endothelium is not inert, but rather is highly active, participating in several physiological processes, including the control of vasomotor tone, the trafficking of cells and nutrients, the maintenance of blood fluidity, the regulation of permeability, and the formation of new blood vessels (2). According to the American Heritage Dictionary, an organ is defined as “a differentiated part of an organism, such as an eye, wing, or leaf that performs a specific function.” The Webster’s Revised Unabridged Dictionary defines an organ as a “natural part or structure in an animal or a plant, capable of performing some special action (termed its function), which is essential to the life or well being of the whole.” While the endothelium surely meets these criteria for an organ, it has yet to be widely accepted on these terms. In this chapter, I will argue that its membership into the “organ club” is long overdue.

2. THE BENCH-TO-BEDSIDE DISCONNECT If one peruses the table of contents and indexes of the more popular medical texts or oncall references, one finds little or no mention of the endothelium. The terms “endothelial cells” and “endothelium” are missing not only from the Merk Manual Index, but also from the 64-page July 2003 index of Scientific American Medicine. In the 15th edition of Harrison’s Principles of Internal Medicine, the index refers only to “endothelial injury, in sclerosis,” and “endothelial cell(s), interactions with lymphocytes; vascular proliferation” (3). In contrast to virtually every other conceivable organ, the endothelium lacks formal representation and organized support, that is, there is no subspecialty training in endothelial biomedicine, nor are there national or international societies for the endothelium. As physicians, few of us are attuned

Endothelial cells in health and disease

2

Table 1 Pub Med References According to Keyword and Yeara Endothe Endothe Hepato Hepato Cardio Cardiology lium liology cytes logy myocytes 1984 1987 1990 1993 1996 1999 2001 Total a

1,307 1,973 3,150 4,221 5,012 6,083 10,203 84,798

— — — — — 1 — 1

1,236 1,302 1,608 1,728 1,710 1,738 3,440 36,309

87 95 119 173 206 287 444 3,863

52 61 115 190 346 464 825 5,078

247 407 607 723 833 1,226 1,965 14,162

Accessed august 2003

to the health of this cell layer as we interview and examine our patients. Diarrhea, syncope, or jaundice equivalents do not presently exist for the endothelium. There is no “endothelial box” to circle or check off as we move through the review of systems. Moreover, the endothelium is not amenable to the traditional maneuvers of inspection, palpation, percussion, and auscultation. When it comes to laboratory testing, while renal function is readily assayed with urea and creatinine; liver function with transaminases and/or bilirubin; and hematological function with a complete blood count and peripheral smear, there are no convenient and reliable markers for endothelial cell dysfunction. In a recent Pub Med search, the keywords “endothelial cells” and “endothelium” yielded a total of approximately 55,000–85,000 articles, respectively (Table 1); the term “endotheliology,” a total of 1. This of course is no surprise, since current medical lexicon does not include the term “endotheliology” (nor for that matter any analogous term that embraces an endothelial-centric clinical discipline). But that is precisely the point. Contrast the above ratio with that in the liver or heart fields and one begins to see a curious disconnect (Table 1). In other words, despite the exponential growth of (largely basic science) studies over the years, endothelial disease continues to fly well below the clinical “radar screen.” There are several explanations for this bench-to-bedside gap, three of which are discussed below.

2.1. Out of Sight; Out of Mind One explanation for the under-appreciation of the endothelium as an organ relates to its hidden and enigmatic nature. The endothelium rarely “shows its hand,” at least in the classic ways that we, as physicians, are trained to detect. Like the hematological system, the endothelium is highly diffuse and spatially distributed, extending to all reaches of the human body. Yet unlike blood cells, the endothelial lining is tethered to the blood vessel wall and therefore inaccessible and poorly amenable to study. Although assays do exist for circulating markers of “activated” endothelium, many of these lack specificity, and as

The endothelium as an organ

3

single markers provide little in the way of useful information. Pathological specimens of the endothelium are not routinely available and even if they were, the findings would not necessarily correlate with function.

2.2. Historical Legacies A second factor that has paradoxically contributed to an under-appreciation of the endothelium as an organ relates to the traditional link between vascular biology (and by extension, endothelial cell biology) and cardiology. The connection is steeped in history, its roots dating back to William Harvey’s discovery of the circulation in the early 1600s. Following Harvey’s seminal work, the prevailing view of the cardiovascular system was that of a closed circulatory loop consisting of a pump and series of conduit vessels, with the singular role of delivering oxygen and nutrients to the various tissues of the body. Over the next 400 years, clinical and basic research focused largely on the pump itself, namely on the coronary arteries, the contractile apparatus, the conduction system, and the conduit vessels vis-à-vis their impact on the function of the heart (e.g., hypertension). These developments contributed to and were reinforced by the founding of a large clinical discipline (cardiology), powerful societal infrastructure (American Heart Association), highly successful public awareness campaigns, generous private and public funding, and enormous progress inresearch and development. The importance of these milestones cannot beoverestimated. They have led to vastly improved detection, prevention, andtreatment of coronary artery disease. Over the past 40 years, however, two seminal observations have revolutionized the field of vascular biology and, when taken together, argue for a more complete synthesis of the field. First was the recognition that the endothelium is not an inert barrier, but rather a highly active cell layer that is involved in a wide variety of homeostatic processes. The second important observation was that the endothelium, in traversing each and every organ, establishes a dialogue that is unique to the underlying tissue—in effect marching to the tune of the local microenvironment. This endothelial-tissue interface plays an important role not only in maintaining health of the organism, but also in dictating the focal nature of vascular disease states. Viewed from this perspective, the study of the endothelium transcends all clinical disciplines. While 20 years ago, one was hard pressed to identify more than a small handful of disorders in which the endothelium played a prominent role, today it may be argued that virtually every disease state involves the endothelium, either as a primary determinant of disease or as a victim of collateral damage (Table 2).

2.3. Complexity A final consideration that helps explain the bench-to-bedside chasm relates to the complexity of the endothelium. It was not that long ago that many investigators subscribed to the one gene-one enzyme-single function hypothesis. The goal of the human genome project was to develop a blueprint for human health and disease and to establish a menu list for selective drug targeting. One of the surprises arising from this project was a discovery that human genome contains a mere 30,000 genes—compared with an estimated 14,000 genes in Drosophila. These results indicate that complexity

Endothelial cells in health and disease

4

does not scale with the number of genes. The discordance is explained in part by differences in alternative splicing and variation in posttranslational modification. More importantly, complexity arises from the connections between components, that is, the regulatory network of protein-protein, protein-RNA, and protein-DNA interactions. Stated another way, all biological systems—including the endothelium—are non-linear and display emergent properties. Most investigators in vascular biology (this author included) tend to focus on specific aspects of endothelial cell function using in vitro assays. In doing so, we may overlook critical levels of organization that are essential to a full understanding of the system (Fig. 1). Just as one could never map the human mind by studying a

Table 2 Role of Endothelium in Disease

Disease

Number of Cross-References of Disease Selected Terms with “Endothelium”/“Endothelial References Cells” (Based on Pub Med Search 8/30/03)

Hemaatology-oncology Cancer

74,75

6,133/9,487

—

—

SSD

76–79

237/157

Thalassemia

80,81

29/35

82

—

Myeloproliferative diseases

83,84

68/73

Bone marrow transplantation

85,86

169/189

Transfusion medicine

87–90

282/297

TTP/HUS

91,92

125/100

Coagulation

40,93

2622/2016

Infection

94–96

2820//3141

Sepsis

97,98

891/741

Atherosclerosis

99–104

7318/3727

Congestive heart failure

105–108

570//228

Valvular heart disease

109,110

158/82

Hemoglobinopathies

Hemachromatosis

Infectious disease

Cardiology

The endothelium as an organ

5

Pulmonary Asthma

111,112

251/285

COPD

113,114

55/39

Pulmonary hypertension

115–117

771/397

38,118

182/159

Acute renal failure

119–121

139/125

Chronic renal failure

122–124

269/152

Peptic ulcer disease

125,126

64/74

Inflammatory bowel disease

127,128

147/162

Hepatitis

129,130

197/315

Cirrhosis

131,132

389/461

Pancreatitis

133,134

85/92

Rheumatoid arthritis

135–137

436/600

Scleroderma

138–140

250/204

141,142

2900/1693

Stroke

143–145

824/469

Multiple sclerosis

146,147

186/207

148,149

437/327

ARDS Nephrology

Gastroenterology

Rheumatology

Endocrinology Diabetes Neurology

Other Preeclampsia

Endothelial cells in health and disease

6

Figure 1 Levels of organization. As with all biological systems, the endothelium displays emergent properties. While each level of organization offers a unique platform for investigation, it is important to recognize that microlevel properties do not necessarily predict for macrobehavior. Indeed, as shown in this schematic, certain properties of the endothelium and vasculature are expressed only at higher levels of

The endothelium as an organ

7

organization. Representative examples of level-specific properties are shown. cultured monolayer of neurons, one cannot rely solely upon in vitro systems to predict and model the behavior of the intact endothelium. An important goal for the future, which will be expanded on below, is to learn how to harness the strength of reductionist and holistic approaches to better understand the algorithms that link individual endothelial cells to blood vessels, blood vessels to organs, and organs to the whole organism.

3. FALLING THROUGH THE CRACKS—A REAL-WORLD EXAMPLE Let us consider one example of how the conceptual gap in endothelial-based disease has had an impact on patient care. The case study is severe sepsis, defined as systemic inflammatory response to infection with secondary organ failure. Patients with severe sepsis are typically admitted to the intensive care unit. Since the syndrome is complicated by organ dysfunction, medical care is provided not only by critical care physicians but also by a team of organ-specific consultants, including nephrologists, neurologists, hematologists, and cardiologists. The pathophysiology of sepsis is complex and includes a non-linear interplay between multiple cell types and soluble mediators, including components of the inflammatory and coagulation pathway (4) (see Chapter 20). Over the past decade, enormous resources have been expended on sepsis trials, with more than 10,000 patients enrolled in over 20 placebo-controlled, randomized Phase 3 clinical trials. The vast majority of these therapies have failed to improve survival in patients with severe sepsis. A notable exception is a recombinant form of the natural anticoagulant activated protein C (rhAPC), which was shown in the PROWESS study to reduce 28-day all cause mortality in this patient population (5). The results of the PROWESS trial have created an identity crisis for intensivists and hematologists alike. The critical care field has finally come across an agent that saves lives in severe sepsis. However, the biological plausibility of its mechanism of action is mired in a maze of inflammatory and coagulation pathways. Hematologists have had their own struggles, in this case to make sense of the fact that after 20 years of failed trials in sepsis, the critical care world has embraced a molecule that is near and dear to their hearts as a life-saving measure in a patient population with which they are barely familiar with, and through a mechanism certainly more complicated than simple anticoagulation. In fact, recent evidence suggests that the common underlying thread in severe sepsis is endothelial cell dysfunction and the efficacy of aPC may ultimately be explained by its attenuation of adverse endothelial cell processes, such as apoptosis (6,7). However, the extent to which this is true will remain unknown as long as the intact endothelium continues to defy diagnostic interrogation. At a clinical level, an under-appreciation both for the non-linear nature of the host response to infection and the importance of the endothelium as an organ system may have an impact on patient care. In some cases, physicians who lack a full understanding of

Endothelial cells in health and disease

8

sepsis pathophysiology, as well as the potential mechanisms of action and risk-benefit profile of rhAPC, may avoid prescribing the agent for fear of the unknown. In other cases, poorly informed clinicians may administer the agent, unprepared for its potential complications. Finally, there exist a growing number of physicians genuinely interested in broadening their understanding of the complexities of the host response, and learning more about the endothelium as a component of this response. Owing to the bench-tobedside gap in the endothelial field, these individuals currently have limited resources. As long as an understanding of the role of the endothelium eludes the clinical mainstream, the potential for developing a new class of sepsis drugs, capable of attenuating endothelial cell dysfunction, will remain unrealized. Abetter understanding of the endothelium in health and disease and the development of new tools to assay the endothelium in vivo should help redirect research along more productive lines.

4. HISTORICAL PERSPECTIVES A consideration of time scales and history provides a valuable conceptual framework for approaching the endothelium. In the hospital setting, physicians are focused on stabilizing and discharging their patients as soon as possible—they work on a scale of minutes to days. Primary care physicians are concerned with the life cycle of the patient. With continued breakthroughs in biomedicine, including advances in robotics and information technology, medicine continues to evolve further from a healing art to a preventative science. It is difficult to predict how such a transition will alter the time scale of patient care. Many scientists, including chemists, molecular biologists, and cell biologists, are more concerned with events that occur over nanoseconds to minutes. At the other end of the spectrum is geological time, dating back to earth’s formation some 4.6 billion years ago (Fig. 2). Why should we be concerned with deep time? One of the features that separate biological from non-biological systems is the capacity to adapt and evolve. For billions of years, every one of our ancestors—on both our mother’s and father’s side—was able to attract a mate, and successfully reproduce, thus withstanding the most pressing test of natural selection. Perhaps, we can gain insights into human health and disease by examining our origins. Microfossil records indicate that life first evolved on the earth approximately 3.5 billion years ago. The earliest form was a unicellular prokaryote, similar (in some ways) to today’s photosynthesizing cyanobacteria. Eukaryotes evolved ≈1.8 billion years ago. The next major transition was the evolution of metazoan (multicellular) organisms, beginning 600 million years ago. The earliest forms resembled today’s simple segmented invertebrates. Vertebrates arose from a common ancestor approximately 500 million years ago. The first land vertebrates date back to 400 million years ago. The following

The endothelium as an organ

9

Figure 2 Historical perspective. The earth formed 4.5 billion years ago. The first prokaryotic fossils date back to approximately 3.6 billion years (1). Shown are various milestones since that time drawn to scale. Eukaryotes appeared 1.8 billion years ago (2), multicellular organisms or metazoa 600 million years ago (3), and vertebrates 500 million years ago (4). The first land vertebrates evolved approximately 400 million years ago (not shown). The evolution of vertebrates marked the appearance of a closed circulation and endothelium (4). The human species is 150,000 years old (5). The transition from the Stone Age (Paleolithic) to the agricultural revolution (Neolithic) occurred a mere 10,000 years ago (6). The history of medicine (particularly as it relates to the cardiovascular system) is a mere blink of the eye in evolutionary terms (7).

Endothelial cells in health and disease

10

classes of animals appeared in relatively short order: amphibians, reptiles, birds, and mammals. The human species is relatively young, with its origins tracing back to 150,000 years ago. The endothelium appeared with the dawn of vertebrate evolution. Indeed, the vertebrate transition marked the appearance of several common design features, including the endothelium, a closed circulation, three types of circulating blood cells, a coagulation cascade, and acquired immunity. Remarkably, each of these systems evolved over a period of 50 million years, a mere “blink of the eye” in evolutionary terms. For 2.5 million years, our ancestors lived as hunter-gathers, using stones as tools, and building fires. Approximately 10,000 years ago, our ancestors had hunted most of the large species to extinction, and facing the law of diminishing returns, began to form communities, and to engage in farming and animal breeding. This Paleolithic-Neolithic boundary is perhaps the most important transition in the history of medicine and modern disease. The past 10,000 years—which span a mere 400 generations—have witnessed not only stunning advances in technology and civilization, but also the emergence of occupational, nutritional, and infectious epidemics. A consideration of time scales raises interesting questions about our future. It is sobering to consider that 99.99% of all species that have populated this planet are extinct and that the average metazoan survives only 2.5 million years. Regardless of our predictions about the future, as physicians and scientists, we are primarily focused on the present. Indeed, what separates us from every other species on earth is our consciousness and self-awareness. We are gifted with intelligence and innate curiosity and from this blend arises a desire to heal and alleviate human suffering. The history of “modern” western medicine is framed around three major figures. Hippocrates, a Greek physician, is known as the “Father of Medicine.” Hippocrates contributed to a large collection of medical works, termed the Hippocratic Corpus. Interestingly, Hippocrates and his contemporaries did not appreciate that the heart was a propulsive organ. The next major figure was Galen. Also a Greek physician, Galen was one of the most prolific investigators of all time, publishing some 2.5 million words in over 100 collections. Galen, too, did not recognize that blood circulated. He believed that blood was constantly produced in the liver and was distributed to the various tissues and organs via the veins, where it was consumed. Moreover, he hypothesized that the arteries carry a vital spirit from the lung to the peripheral tissues. Galen’s mistaken conclusions are not all that surprising when viewed in the context of the times. After all, arteries and veins differ in their thickness and pulse, arterial and venous blood differ in color, there is a time lag between the cardiac and arterial pulse, and the capillaries are invisible to the human eye. What is surprising is that 1500 years would pass before the fundamental errors of Galen’s theory were exposed. William Harvey, the third major figure in this discussion, discovered and reported that blood circulates in 1629 AD. Although he could not visualize the capillaries, Harvey did surmise their existence. In 1859, an upper-class English gentleman by the name of Charles Darwin made what may be one of the most important observations in the history of mankind. He proposed that all life evolved from a common beginning by the process of natural selection. Until that time, the prevailing view was that world was created 6000 years earlier. Virtually every intellectual scholar in biomedicine, including William Harvey—the discoverer of the circulation—was operating under the belief that they were studying God’s work.

The endothelium as an organ

11

The change in paradigm is nicely captured by the landscape metaphor, as originally described by Richard Dawkins in his book, Climbing Mount Improbable (8). According to this metaphor, complex design rests at the peak of a mountain (Fig. 3). Dawkins called the mountain Mount Improbable because the species or organ could not have reached the summit by chance alone. There are two sides to the mountain. For centuries, mankind recognized only the side with the cliff. Reaching the summit depended on giant leaps through divine intervention or single generation macromutations, a process referred to as saltation. Darwin exposed the other side of the mountain. He proposed that the gradual incline was surmountable by the cumulative selection of chance mutations, a mechanism that came to be known as natural selection (9). The incline may not be gradual as Darwin once believed, but rather punctuated by “waves of innovation.” Once at the summit, the product—whether a species or an organ—carries the illusion of design (10). The evolutionary history and comparative biology of the endothelium may provide important insights into the endothelium in health and disease. For example, properties of the endothelium that are invariant between vertebrate species are more likely to represent fixed or core design features, whereas those that are not conserved may reflect new contingencies and thus be more amenable to therapeutic retooling. Human evolution has been driven in large part by an arms race with pathogens. Many behavioral properties of the endothelium can be explained by this process. Path dependence (also termed historical constraint or phylogenetic inertia) refers to the notion that a subsystem such as the endothelium is a product of an unbroken lineage of intermediates, and that the continuum over evolutionary time results in maladaptive or jury-rigged structural (and functional) designs. Defining these design

Figure 3 The Landscape metaphor. Imagine that complex design—for example, a species or an organ system (shown is human)—rests at the peak of a summit. Prior to 1859, mankind (preDarwin observer) recognized only the side with the cliff, ascribing the ascent

Endothelial cells in health and disease

12

to divine intervention. Darwin exposed the other side of the mountain—the one with a gradual incline, surmountable through step-by-step modification (natural selection of random mutations). Along a second axis is the complexity scale of modern day organisms, marked by unicellular organisms, open circulation (most invertebrates), and closed circulation (all vertebrates). “flaws” would provide novel insight into therapy. Evolutionary legacy refers to a mutation or design that may have been beneficial in an earlier time, but is no longer apparently adaptive. For example, we are likely to have evolved specific adaptations to species of pathogens which are now extinct. Finally, it is important to recognize that the endothelium evolved to maximize fitness in a far earlier era, approximately 30,000 years ago. A consideration of environmental changes provides important insight into the vulnerability of the endotheliumto disease.

5. WEAVING AN EVOLUTIONARY TALE An important challenge in studying evolutionary biology is to reconstruct the past. There are several approaches, including the study of fossils, molecular phylogeny, and comparative biology (with the assumption that what works for modern day creatures may have worked for ancestral forms). Unfortunately, the cardiovascular system does not fossilize. Thus, any rendition of the evolutionary history of the endothelium is, at best, speculative. Nevertheless, a consideration of the ancient past provides a powerful conceptual framework for understanding the endothelium in modern times. Single cell organisms obtain their oxygen through simple diffusion, a process that is defined by Fick’s law (the flow of oxygen is directly proportional to the pressure difference and surface area, and inversely proportional to distance oxygen must travel). In multicellular organisms, the cardiovascular system provides a means of overcoming the time-distance constraints of diffusion. As a general rule, the body plan of most multicellular organisms can be simplified according to the scheme shown in Fig. 4 (there are some interesting exceptions to this body plan, including insects). There are four basic elements to the scheme: (1) convection or bulk flow of oxygen from the environment to a highly vascularized surface, (2) oxygen diffusion from environment to blood, (3) convection or bulk flow to the various tissues of the body, and (4) diffusion across into the mitochondrial “sink” of each and every cell. Note that the laws of nature have not been defined:

The endothelium as an organ

13

Figure 4 Body plan of multicellular organisms. This schematic simplifies oxygen transport according to four steps: (1) convection of oxygenated air or water from environment to a highly vascularized surface (skin, gills, or lungs), (2) diffusion of oxygen across the gas exchanger into the blood, (3) convection of oxygenated blood around to the various tissues of the body, and (4) diffusion of oxygen to the individual cells of the tissues. The circulation is closed in vertebrates, open in invertebrates. (Adapted from Ref. 73.) oxygen transport is still critically dependent on simple diffusion, both at the lung-blood interface and the blood-tissue interface. Oxygen is poorly soluble in water or plasma. Therefore, even in simple multicellular organisms, oxygen delivery is aided by the presence of a respiratory pigment, a molecule that essentially acts like a magnet to attract and carry oxygen. In invertebrates, the respiratory pigment (usually hemocyanin, but sometimes hemoglobin) circulates freely in solution. This observation provides the rationale for developing hemoglobin substitutes in transfusion medicine. In vertebrates, the hemoglobin is packaged inside red blood cells, where it is protected from oxidative stress of the environment and where oxygen binding may be finely regulated through a series of allosteric and cooperative interactions. Non-mammalian vertebrates (fish, amphibians, reptiles, and birds) contain nucleated red cells. In fact, the anulceate red blood cell is unique to mammals. It is interesting to speculate why the nucleus may have been “discarded” during recent evolution. There are many possible explanations. Perhaps the most compelling is that anucleate cells lack mitochondria and oxidative phosphorylation, and therefore do not consume oxygen. In this way, the mammalian red blood cell avoids the conflict of interest of being both a carrier and consumer of oxygen. Countering this argument is the observation that

Endothelial cells in health and disease

14

hummingbirds, which transport their oxygen in nucleated red cells, have a higher metabolic rate compared with most mammals. Whether from a 70-kg human, a 2-g shrew or a 150-ton blue whale, the mammalian red blood cell is remarkably similar in size and shape (there are some interesting exceptions which will not be discussed here). The high degree of invariance in cell size and shape is consistent with the notion that selection—acting upon the mammalian red blood cell—has led to the optimization of the variables in Fick’s equation. For example, the biconcave shape provides a high surface area-to-volume ratio and a short distance for oxygen to travel. During evolution, a solution to one problem tends to beget a new set of problems. As one example, the development of a cardiovascular system, while providing a means to overcome the time-distance constraints of diffusion (and thereby paving the way for the evolution of large animals), resulted in a highly pressurized system that is at risk for rupture and/or leakage, with potentially two life-threatening consequences: (1) exsanguination or loss of blood from the interior to the exterior (hence, the formation of the coagulation mechanism, consisting of “sticky” cells and a protein gel), and (2) the entry and dissemination of pathogens from the exterior to the interior (hence, the formation of the innate immune response, also composed of cells and a protein gel). In invertebrates, the heart pumps blood (termed hemolymph) into an open cavity (the hematocoele), which bathes the various tissues of the body. In vertebrates, blood is retained within a closed circulation, separated from the underlying tissues by an endothelial lining (11).

6. METAPHORICAL LEXICON An important challenge in promoting the endothelium as an organ is to describe what one cannot see, and to do so in a language that is understandable to the clinical mainstream. One approach is to fall back on the metaphor, otherwise described as a “rhetorical device for transporting knowledge by using a word that brings

Table 3 Metaphors for Describing the Endothelium Metaphor

Advantages

Disadvantages

Cellophane

Intact, delicate, thin; protective barrier property

Ignores living (active, non-inert) qualities; overlooks phenotypic heterogeneity

Gatekeeper

Active role in mediating selective transport Does not describe non-gatekeeping of gases, macromolecules, and white blood functions of the endothelium cells; especially applicable to postcapillary venules

Barometer

Senses biomechanical forces

Does not account for ability to sense biochemical signals; ignores capacity of endothelium to integrate and respond the signals

The endothelium as an organ

15

Biosensor

Senses biomechanical and biochemical forces

Ignores capacity of endothelium to integrate and respond to the signals

Input-output device

Senses and responds to extracellular signals

Does not describe the intracellular transduction apparatus; fails to emphasize the non-linearity of the device; the study of a single device (e.g., endothelial cell) does not predict global behavior

Circuit board Represents multiple (trillions) of inputoutput devices Ant colony

Implies that the system is hardwired and fixed

Sophisticated, bottom-up organizations; emergence of properties; local rules generate global behavior; spatially and temporally distributed; both colonies (ant and endothelium) are hidden from view, and require special excavation tools for study

connotations from one field into play in another field.”a Metaphors are not literal and therefore must be chosen and interpreted carefully. The intention of this section is not to entrap the reader within the frame of any single example, but rather to provoke speculation and “jumpstart” the recognition process. One of the metaphors is outdated, others are admittedly fanciful, and still others may prove to be closer to the truth (Table 3). For all we know, the endothelium may itself be used one day as a metaphor to describe other biological or non-biological systems.

6.1. Endothelium as Nucleated Cellophane Under low-resolution light microscopy, the endothelium has the appearance of a layer of nucleated cellophane. Indeed, in the absence of more sophisticated a

http://www.santafe.edu/sfi/publications/Bulletins/bulletinFall99/insideSfi/language.html

diagnostic and investigative tools, the endothelium was considered to be little more than an inert barrier, separating flowing blood from underlying tissue. Virchow described the endothelium as “a membrane as simple as any that is ever seen in the body.” Today, the cellophane wrapper metaphor should be viewed for what it is: a historical relic that, while understandable in the context of 19th century biomedicine, fails to capture the living essence of the endothelium.

6.2. Endothelium as Gatekeeper The use of electron microscopy in the 1950s and 1960s provided a powerful new window into the endothelium. These studies demonstrated that the endothelium acts as a selective barrier to macromolecules (12,13). The repertoire of endothelial cell functions was

Endothelial cells in health and disease

16

subsequently expanded to include regulation of hemostasis, vasomotor tone and leukocyte trafficking. For example, in 1970, the endothelium was shown to express plasminogen activator activity and a cellular inhibitor of fibrinolysis (14), an observation that was later confirmed in cell culture studies (15,16). In 1973, Eric Jaffe demonstrated that endothelial cells express the procoagulant, von Willebrand factor (vWF) (17). Michael Gimbrone first reported that endothelial cells express inducible levels of prostaglandin E, providing additional evidence for the role for the endothelium in regulating hemostasis (18). In 1980, Furchgott (19) demonstrated that the endothelium plays an obligatory role in acetylcholine-mediated vasomotor relaxation via the release of a highly labile diffusible factor, originally termed endothelial derived relaxing factor (EDRF), and later identified as nitric oxide. In a series of elegant studies in the mid1980s, Gimbrone’s group provided compelling evidence that the endothelium actively mediates leukocyte adhesion and transmigration (20–22).

6.3. Endothelium as a Barometer or Biosensor As we have refined our molecular and cellular tools, we have come to appreciate the endothelium more as a barometer or biosensor of the local microenvironment. The advantage of these two metaphors is that they describe the capacity of the endothelial cell to sense biomechanical and biochemical changes in the microenvironment. The disadvantage is that they ignore the capacity of the endothelium to integrate and respond, either adaptively or non-adaptively, to these inputs.

6.4. Endothelium as an Input-Output Device Much of endothelial cell biology can be understood—at least conceptually—by considering each and every endothelial cell in the body as an adaptive input-output device (Fig. 5). The input arises from the extracellular milieu and may include biochemical or biomechanical signals. The output is manifested as the cellular phenotype and includes a number of structural and functional properties. Some of these properties are expressed at the level of individual cells or cell culture (e.g., protein, mRNA, proliferation, apoptosis, migration, and permeability), whereas other properties are expressed at higher levels of organization—for example, the blood vessel (e.g., leukocyte trafficking, vasomotor tone, fibrin deposition) or organism (e.g., redistribution of blood flow, vascular bed-specific phenotypes). Each endothelial cell may have unique intrinsic properties, the so-called “set point.” Differences in set

The endothelium as an organ

17

Figure 5 Input-output device. Each endothelial cell may be considered an input-output device. Input arises from the extracellular environment and may include biochemical and biomechanical forces. Output is manifested by the cellular phenotype. This model provides a valuable framework for understanding the molecular basis of endothelial cell heterogeneity. If endothelial cells are intrinsically identical (top), then spatial and temporal differences in input signals (e.g., input A, B) will result in spatial and temporal differences in output (output A, B), resulting in heterogeneity. If endothelial cells are epigenetically modified (bottom), then they may display heterogeneous phenotypes (output A, C) at rest and/or in response an identical input (input A). Both mechanisms are operative in the intact organism and contribute to

Endothelial cells in health and disease

18

generation and maintenance of endothelial cell heterogeneity. point are brought about by environmentally induced epigenetic changes in the “hardwiring” of the cell. If one accepts the analogy of each endothelial cell representing its own input-output device, it is not a stretch to consider the endothelium as a circuit board—one that is hardwired (to some extent) to meet the demands of the tissue, and one that is highly vulnerable to short-circuiting as a mechanism of vasculopathic disease.

6.5. Endothelium as a Small-World, Scale-Free Network One way to approach the complexity of the host response is through the theoretical framework of a network. When we think of networks, we typically imagine neural connections within the brain. However, in a broader sense, networks may be viewed as any interconnected system in which constituent parts (called “nodes”) are linked to one another. Examples of networks that have been extensively studied and mapped include the web of Hollywood actors, the Internet, the World Wide Web, the structure of scientific collaborations, and certain cellular metabolic pathways (reviewed in Ref. 23). In each case, the topology reveals a scale-free behavior in which a relatively small number of nodes have an unusually high number of connections (called “hubs”), and in which no single node is typical of the others. Such topology predicts for two behavioral traits: (1) the network is resistant to accidental attacks—random removal of many nodes does not alter the overall architecture, since the hubs hold the network together, and (2) the network is vulnerable to coordinated attacks—provided that the hubs can be identified and destroyed. Although scale-free networks are pervasive and will likely prove ubiquitous in nature, the topology of most biological systems is not amenable to statistical analyses using current mathematical and molecular/cellular tools. Whether or not such networks actually exist in the endothelium is beside the point—network theory provides a valuable conceptual framework. For example, at the level of the individual endothelial cell, one might envision that various networks are at play, including metabolic and signaling pathways. At the level of the whole body, the endothelium is linked to multiple nodes including (but certainly not limited to) other endothelial cells, circulating cells, abluminal cells, and soluble mediators. In pathophysiology, there may be a dynamic change in the number and nature of both the nodes and links. For example, in sepsis there is induction of signal intermediate and transcription factor activity (nodes) while at a larger scale there is an increase in the interaction between the endothelium with leukocytes, platelets, and soluble mediators (links). An interesting feature of scale-free networks is what is called a “small-world” property—the distance between any two nodes is short. The concept is embodied in the popular expression “six degrees of separation,” a reference to studies showing that any two people in the world are only five handshakes away from each other. In sepsis, the

The endothelium as an organ

19

increased number of links brought about by an “activation state” might be predicted to reduce the network diameter and increase robustness. A consideration of these principles may have a therapeutic payoff. For example, a robust and error-free sepsis network will be generally resistant to treatment. This of course is the case in sepsis—virtually all single-mediator trials (aimed at one or another “nodes”) have failed to put a dent in sepsis mortality. Based on network principles, there are two approaches to overcome the inherent resistance of the system. One is to destroy a sufficient number of nodes so as to induce collapse of the network (multimodality therapy or cluster bomb approach). The other is to identify and target the hubs in the network (smart bomb approach). The p53 transcription factor has been postulated to represent an intracellular hub in the context of cancer (24). In sepsis, the NF-κB transcription factor may be predicted to play a similar role.

6.6. Endothelium as a Social Colony Studies of ant behavior (entomology) and city life (sociology) provide the most unlikely, yet in many ways the most powerful, of metaphors for the endothelium. Both are examples of self-organizing complexity. The ant colony is a highly sophisticated, extraordinarily robust, and well-adapted biological system, which displays classic properties of emergence, that is, its behavior is understood at the level of the colony and not at the level of the individual ants. Ant colonies and the endothelium not only share this phenomenon of emergence, but also demonstrate a similar propensity to specialization—the ant colony is composed of soldiers, the queen, workers, and is geographically separated into distinct quarters, while the endothelium is similarly specialized in space and time. Both the ant colony and the endothelium represent a bottom-up society; there is no commander or pacemaker driving the system (contrary to popular belief, the queen ant is not in charge of the colony). Each member of society, whether an ant or an endothelial cell, goes about its business fully ignorant of the larger order, following local rules to generate complex global behavior. In reading the following passage from Deborah Gordon’s book, “Ants at Work,” one cannot help draw a comparison with the endothelium: The basic mystery about ant colonies is that there is no management…. There is no central control. No insect issues commands to another or instructs it to do things in a certain way. No individual is aware of what must be done to complete and colony task. Each ant scratches and prods its way through the tiny ant world of its immediate surroundings. Ants meet each other, separate, go about their business. Somehow these small events create a pattern that drives the coordinated behavior of colonies…. (25) A consideration of urban life provides striking parallels to the endothelium. Jacobs (1916–), who has no college degree and no formal training in urban planning, singlehandedly revolutionized the way people think about cities. She argued that variety, diversity, growth, and activity are key to their survival. Although cities are unlike ant colonies in that they involve some degree of central planning, there are certain features

Endothelial cells in health and disease

20

that capture the essence of spontaneous emergence and nonlinear dynamics, akin to ant colonies and the endothelium: Under the seeming disorder of the old city, wherever the old city is working successfully, is a marvelous order for maintaining the safety of the streets and the freedom of the city. It is a complex order. Its essence is intimacy of sidewalk use, bringing with it a constant succession of eyes. The order is all composed of movement and change, and although it is life, not art, we may fancifully call it the art form of the city and liken it to the dance-not to a simple-minded precision dance with everyone kicking up at the same time, twirling in unison and bowing off en masse, but to an intricate ballet in which the individual dancers and ensembles all have distinctive parts which miraculously reinforce each other and compose an orderly whole. (26) What is interesting about Jacob’s description of city life is that she recognized the importance of interactions, diversity, complexity, and emergent order. The (somewhat paradoxical) notion that a bustling sidewalk is good for the health of society is akin to the notion that a healthy endothelium is not quiescent but rather is highly active, feeding off the dynamic enterprise of cellular and soluble signals (27). Dampen the interactions (for example, remove all the circulating platelets, or reduce blood flow to a crawl) and the endothelium is no longer healthy.

7. ENDOTHELIAL CELL HETEROGENEITY The notion that endothelial cells are heterogeneous is by no means new. In the 1950s, several investigators—including Lord Florey, Guido Majno, George Palade, and Maia and Nicolae Simionescu—employed electron microscopy to demonstrate structural differences between capillaries in different organs. In 1958, Hibbs et al. wrote: Some variation in the structure of capillaries and arterioles normally occurs from one organ to another, and even among vessels of the same organ. (28) In 1961, Majno et al. (13) employed a “vascular labeling” technique in rat cremesteric muscle—involving systemic administration of black colloidal particles and transillumination under a light microscope—to show that histamine or serotonin resulted in differential leakage and deposit of black colloid particles on the venular side of the circulation. The growing recognition that not all endothelial cells are identical was articulated by Florey in 1966:

The endothelium as an organ

21

Now it is recognized that there are many kinds of endothelial cells which differ from one another substantially in structure, and to some extent in function. (29) In 1967, Reese and Karnovsky (30) were the first to prove the existence of a functional blood brain barrier. In electron microscopic studies of mouse tissues, they demonstrated that exogenously administered horseradish peroxidase readily penetrated the endothelium of the heart, but not the brain. Significant breakthroughs in endothelial cell biology would wait until the early 1970s, when Jaffe and Gimbrone (31–33) independently reported the first successful isolation and primary culture of human endothelial cells from the umbilical vein. The cells, which were obtained by collagenase digestion, could be maintained in culture for weeks to months, and were identified as endothelium by the presence of Weibel-Palade bodies and vWF (VIII-associated antigen). Shortly thereafter, Gimbrone and Cotran (34) successfully isolated vascular smooth muscle cells from umbilical cords, demonstrating a clear distinction between these two cell types. These seminal findings provided the research community with a powerful new tool for dissecting endothelial cell biology and paved the way for breathtaking advances in the field. Indeed, most of our present-day knowledge about the endothelium—from cell surface receptors to signaling pathways, transcriptional networks, cytoskeleton, and cellular function—is directly attributable to our capacity to study endothelial cells in culture. While the cultured endothelial cell became a focal point for research in vascular biology, increasing evidence was pointing to the highly complex topology of the intact endothelium. In 1980s, several groups carried out systematic immunohisto-chemical analyses of the endothelium in various organs (35–37). These studies, which expanded on the earlier results of electron microscopy, revealed differential expression of lectins and antigens in vivo. In 1990s, and more recently in the new millennium, the use of novel genomic and proteomic techniques has uncovered a large array of site-specific properties of the endothelium, providing credence to the analogy of the endothelium as a circuit board, and supporting Gimbrone’s characterization of the endothelium as a dynamically modulatable multifunctional organ (2,38–40). The study of static cultures of one or another endothelial cell type has advantages and disadvantages. On the one hand, such an approach is essential for dissecting, mapping, and/or cloning certain properties of the cell. On the other hand, these studies fail to provide insight into the molecular basis and scope of phenotypic heterogeneity. This limitation was described by Auerbach in 1985: The concept that vascular endothelial cells are not all alike is not a new one to either morphologists or physiologists. Yet laboratory experiments almost always employ endothelial cells from large vessels such as the human umbilical vein or the bovine dorsal aorta, since these are easy to obtain and can be readily isolated and grown in culture. The tacit assumption has been that the basic properties of all endothelial cells are similar enough to warrant the use of the cells as in vitro correlates of endothelial cell activities in vivo. (41)

Endothelial cells in health and disease

22

According to Auerbach, a key to understanding structural and functional heterogeneity was to isolate and study microvascular endothelial cells from different organs. Unfortunately, site-specific properties of endothelial cells are not always retained in culture. Indeed, when removed from their native microenvironment, endothelial cells are uncoupled from critical extracellular cues and undergo phenotypic drift. While endothelial cells cultured from different sites of the vasculature have been shown to express different properties in vitro, the extent to which these in vitro phenotypes are representative of their in vivo counterparts remains largely unexplored. Michael Gimbrone not only described the reproducible isolation and culture of endothelial cells, thus setting the stage for virtually every breakthrough in endothelial cell biology over the next 30 years, but he was among the first to recognize the limitation associated with static cell cultures. Indeed, Gimbrone carried out many of his studies in more than one type of endothelial cell. More importantly, he pioneered approaches for manipulating the biochemical and biomechanical milieu of the endothelial cell in vitro as a means to study the spatial and temporal regulation of endothelial cell phenotypes (21,42−46). To return to an earlier analogy, endothelial cell heterogeneity may be readily understood from the perspective of an input-output device (Fig. 5B). At any given point in time, the net input of biochemical and biomechanical signals is certain to vary between endothelial cells—between and within different organs. Moreover, for any given endothelial cell, the input will vary from one moment to the next. The spatial and temporal variation in input is sufficient to explain structural and functional differences in properties (output). However, there is also evidence that some site-specific properties of the endothelium are mitotically hereditable and thus “locked in.” Epigenetic modification presumably occurs during development and/or in the postnatal period (and perhaps in disease and aging). Thus, endothelial cell heterogeneity arises from a combination of sitespecific differences in the microenvironment (“nature”) and epigenetic imprinting (“nuture”).

8. ENDOTHELIAL CELL ACTIVATION AND DYSFUNCTION When considering the role of the endothelium in disease, the two most common terms that are used are endothelial cell activation and endothelial cell dysfunction. Each of these terms is discussed below and qualified based on recent advances in the field. Early support for the role of the endothelium in pathophysiology is found in studies of inflammation. In the late 19th century, Cohnheim (47) described in detail the changes seen after injury to the tongue of the frog. He demonstrated that leukocytes adhere to the blood vessel wall of venules (so-called pavementing of leukocytes), many of which passed through the wall into the extravascular tissues (leukocyte emigration). These observations were later confirmed in mammalian species. Studies in an ear chamber model demonstrated that leukocytes adhere to the damaged side of a blood vessel, suggesting that the blood vessel wall—as distinct from the leukocyte—is primarily responsible for mediating adhesion (48). However, the mechanisms underlying inflammation-induced leukocyte adhesion remained elusive for decades. According to

The endothelium as an organ

23

one theory, the endothelium secreted a gelatinous substance that traps leukocytes (49). Others claimed that electrostatic forces were responsible for mediating the endothelialleukocyte interactions (50). Like so many other aspects of endothelial cell biology, the elucidation of the molecular basis of leukocyte trafficking would wait until the successful culture of endothelial cells in the early 1970s. Pober and Gimbrone (51) were the first to demonstrate that a well-defined stimulus (lectin phytohemagglutinin) could induce the expression of an endothelial cell marker (Ia-like antigen). Through a series of elegant biochemical, molecular, and cellular studies, Gimbrone et al. (52–54) identified the first inducible endothelial cell-specific leukocyte adhesion molecule (ELAM-1; later designated E-selectin). Gimbrone’s group went on to show that numerous inflammatory mediators, including endotoxin, TNF-α and IL-1, induced the expression of new antigens (so-called “activation antigens”) on the surface of HUVEC, an effect that was correlated with the expression of proadhesive, antigenpresenting and procoagulant activities (21,22,43,55,56). Similar findings were reported by other labs (57,58). Pober and Cotran (59) proposed that “activation” reflects the capacity of endothelial cells to perform new functions without evidence of cell injury or cell division. In 1986, Cotran et al. (60) described activation of the endothelium in vivo. In the latter study, a murine monoclonal antibody, which had been shown to bind an antigen in IL-1stimulated HUVEC (52) (later identified as endothelial-leukocyte adhesion molecule (ELAM)-1 or E-selectin), was found to bind to the microvascular endothelium of human skin in delayed hypersensitivity reactions (DHR), but not normal skin (60). Contran concluded: To our knowledge, this is the only reported endothelial-specific mAB that fails to react with normal endothelium, but identifies activated endothelium in vivo… although evidence for active role of the endothelium in inflammation and the importance of inducible endothelium functions has come from recent work (in vitro studies— author’s italics), the notion of endothelial cell activation is a relatively old one. Early light and electron microscopic studies of DHR described plump, hypertrophied endothelium and increased numbers of intracellular organelles, and it was suggested that the endothelium was activated in these reactions. (Reference to Ref. 61). It now seems likely that these changes of activation reflect some of the alterations in structure, function, and growth induced in endothelium by specific monokines and lymphokines. (60) Though not intended as such, the initial observations by Cotran, Gimbrone and others have given way over the years to the notion of the endothelium as a toggle switch. According to this view, quiescent endothelial cells express an anticoagulant, antiadhesive and vasodilatory phenotype, whereas activated endothelial cells express procoagulant, proadhesive and vasoconstricting properties. However, the notion that endothelial cell activation is an all-or-none phenomenon is an over-simplification. For one, the phenotypic spectrum of an endothelial cell follows a continuum (one only has to look at dose-response studies to appreciate this point). Thus the endothelium is

Endothelial cells in health and disease

24

more analog in its behavior than it is digital.b Second, what constitutes activation for one cell type at a particular snapshot in time may not meet the definition of activation at another site or another moment in time. Third, as initially pointed out by Pober et al. (62), not all inflammatory mediators or endothelial cell activators are created equal. Commonly studied mediators such as TNF-α, thrombin, and lipopolysaccharide have overlapping, yet distinct effects on endothelial cell phenotypes (63). Finally, the terms “activation” and “activity” are not b

At the level of the genome, information is digital (not binary—0 and 1—like a computer but quaternary, represented by four nucleotides: A, C, G, and T). However, the more one moves through hierarchical levels of organization (gene→cell→organ→organism), the more one appreciates graded or analogue-like behavior. synonymous. Normal endothelium is by its very nature highly active—constantly sensing and responding to alterations in the local extracellular environment, as might occur in the setting of transient bacteremia, minor trauma, and other common daily stresses, most of which we are not consciously aware. Therefore, endothelial cell activation is not an allor-nothing response, nor is it necessarily linked to disease. Instead, endothelial cell activation represents a spectrum of response and occurs under both physiological and pathophysiological conditions. Early descriptions of endothelial cell dysfunction focused on structural changes or loss of anatomical integrity, particularly in the context of atherosclerosis. In 1966, Florey suggested that: …consideration of endothelial permeability may be of importance in elucidating the initial phases of the development of atherosclerosis. (64) In 1973, Ross and Glomset proposed a response-to-injury hypothesis to explain the lesions of atherosclerosis: The intact arterial endothelium normally acts as a barrier to some substance or substances present in plasma which upon exposure to vascular smooth muscle promote cell proliferation…the major effect of hemodynamic or other factors that injure the endothelium is to decrease this barrier. (65) Subsequent to Ross’s hypothesis, there was a growing appreciation that the intact endothelium may actively contribute to disease initiation and/or progression (66). The term endothelial cell dysfunction was first coined by Gimbrone (67) in 1980 to describe hyper-adhesiveness of the endothelium to platelets. In 1986, Ganz and colleagues demonstrated paradoxical vasoconstriction of coronary arteries induced by acetylcholine in early and advanced human atherosclerosis, suggesting that abnormal vascular response to acetylcholine may represent a defect in endothelial vasodilator function (68). Cybulsky and Gimbrone (69) were the first to hypothesize a pathophysiological link between inducible endothelial-leukocyte adhesion molecules and atherosclerosis (socalled athero-ELAMs). Using a combination of in vitro cell culture and monoclonal

The endothelium as an organ

25

antibody strategy, they identified an inducible endothelial cell-specific antigen that binds predominantly to monocytes. Peptide sequencing revealed homology to the predicted sequence of human VCAM-1, which had been previously cloned as a cytokine-inducible protein in endothelial cells (70). In support of their hypothesis, Cybulsky and Gimbrone (69) localized VCAM-1 to the endothelium overlying atherosclerotic lesions in a hyperlipidemic rabbit model. These latter observations not only emphasized the role of endothelial dysfunction as a primary determinant of atherosclerosis, but also helped refocus research and development on the inflammatory nature of this disease process. Based on their findings, Cybulsky and Gimbrone amended the definition of endothelial cell dysfunction as follows: Endothelial cell dysfunction has been implicated in the vasospastic and thrombotic complications that are evident in advanced atherosclerosis. Induction of an adhesion molecule early in atherogenesis may also be considered a manifestation of endothelial dysfunction, in that it results in an abnormally hyperadhesive EC surface. (69) Given that the endothelium is multifunctional and highly distributed in space, it is safe to assume that endothelial cell dysfunction is not restricted anatomically to the heart, nor is it limited in disease scope to atherosclerosis. Endothelial cells residing in arteries, capillaries, and veins of every tissue and organ are prone to dysfunction. Gimbrone described endothelial cell dysfunction as: …nonadaptive changes in endothelial structure and function, provoked by pathophysiological stimuli, (resulting in) localized, acute and chronic alterations in the interactions with the cellular and macromolecular components of circulating blood and the blood vessel wall. (71) Indeed, the term endothelial cell dysfunction may be broadly applied to states in which the endothelial phenotype—whether or not it meets the definition of activation—poses a net liability to the host. Assigning liability scores is of course a subjective exercise. An evolutionary biologist might argue that endothelial cell dysfunction is most relevant in its effect on an individual’s reproductive capacity. A physician would surely expand the meaning of dysfunction to include a far broader spectrum of morbidity. An investigator interested in applying evolutionary principles to an understanding of endothelium in health and disease would point out that the endothelium evolved to a state of maximal fitness in the early ancestral environment, and is not adapted to withstand the rigors of high fat diet, epidemics associated with high density populations, sedentary lifestyle or old age.

9. THERAPEUTIC IMPLICATIONS The endothelium is an attractive therapeutic target for several reasons. First, it is strategically located between the blood and tissue; it is rapidly and preferentially exposed

Endothelial cells in health and disease

26

to systemically administered agents. Second, it is highly malleable and thus amenable to therapeutic modulation. Third, in establishing a dialogue with the underlying tissue, the endothelium provides the pharmacotherapist with a direct line of communication to the various organs of the body. Finally, if one believes, as I do, that the endothelium is involved in many if not most disease states, then the odds of endothelial-based treatments having a significant impact on human health and disease are high. When applying the concepts of endothelial cell activation and endothelial cell dysfunction to a consideration of therapeutics, it is important to recognize that endothelial cells may be activated—for example, they may express a phenotype that is characteristic of an inflammatory response—without being dysfunctional. Indeed, there are many instances in which endothelial cell activation is a welcome response, whether in wound healing, physiological angiogenesis, local defense against pathogens, and foreign bodies. Therapy is perhaps best reserved for cases in which the phenotype of the endothelium (whether or not it meets the definition of activation) represents a net liability to the host. The notion that endothelial cells resemble input-output devices and that their behavior is not binary, but continuous, has important therapeutic implications. The goal in treating the endothelium is not to reset the switch, but rather to fine-tune and recalibrate the cell, nudging it back to its ideal state. An important challenge is to learn how to determine the nature of that ideal state. Endothelial cell dysfunction usually arises from otherwise adaptive responses (or at least ones that were adaptive in the ancestral environment) that are now excessive, sustained, or spatially and/or temporally misplaced. The transition between endothelial cell function and dysfunction is not always clear. As more effective treatments become available for attenuating dysfunctional endothelium, it will be important to avoid overshooting the desired effect or “lobotomizing” the cells. In this respect, it will serve us well to remember that an active endothelium is a healthy endothelium. Finally, given that endothelial cell phenotypes vary according to time and location in the vascular tree—in both health and disease—it will be essential to target therapy to specific vascular beds.

10. REDUCTIONISM VS. HOLISM There is no question that future advances will rely heavily on continued studies at the cellular and molecular level. But it is equally important that we keep our eye on the larger perspective. Such considerations are by no means to unique to the endothelial cell biology or even to medicine at large. Deborah Gordon, a biologist from Stanford University, studies ants, because she is “interested in linking levels of organization.” The bottom-up emergence of ant colonies, embodied in following passages from her book, Ants at Work, is uncannily reminiscent of the endothelium: The most direct way to investigate animal behavior is to try to see it as a complete pattern, not to take it apart. The more we take it apart, the more work we have to do to put back together the conditions and other kinds of behavior it belongs with…there is some balance between the paralyzing contemplation of the complexity of everything, and a focus on

The endothelium as an organ

27

components that can each be understood separately but are so isolated that they cannot be traced back to see how they fit into the whole system. (25) The emergence of complex systems poses a common set of challenges in many areas of investigation. How does one leverage the advantages that are inherent in each level of study (molecular, cellular, whole organisms, and colonies) for mechanistic, diagnostic, and therapeutic gain? In the case of the endothelium, advances at the reductionist (basic science) end of the spectrum have far outstripped those at the holistic (clinical) end. Indeed, as long as the field continues to elude the clinical mainstream and/or a platform for bridging the bench-to-bedside gap, it will remain entrenched in reductionism, its potential largely untapped, and unrealized. And herein lies perhaps the most powerful argument for recognizing the endothelium as a bona fide organ.

11. FUTURE DIRECTIONS 11.1. New Questions If we consider the endothelium as a dynamic organ, we may begin to ask new questions, which have gone largely unexplored. For example, if we accept that the endothelium displays emergent properties, what are the simple local rules that govern complex behavior? What determines task allocation in the “endothelial colony”? When viewed from the perspective of a network, the endothelium is a series of nodes that are connected to each other and to other nodes, including non-endothelial cells, soluble mediators, and extracellular matrix. Do these networks display scale-free properties? If so, which are the highly connected nodes (hubs)? And are these hubs vulnerable to attack from a therapeutic standpoint? How does one assign connection weights to the various nodes? How do the nodes and links change over time, and how do they differ in health and disease? As a non-linear system, does the endothelium display chaotic behavior? If so, can the butterfly effect explain the results of certain clinical trials and justify ongoing (largely fruitless) research and development based on linear reasoning? These and many other questions will be important to address as we move forward in understanding the endothelial organ.

11.2. Bridging the Bench-to-Bedside Gap Clinical progress in endothelial cell biology will depend on several factors. First, clinicians should begin to recognize the endothelium as an organ or subsystem—one that has pathophysiological, diagnostic, and therapeutic elements. Second, there must be coordinated efforts to teach and educate physicians about the endothelium, not simply to “bring them up to snuff” but also to train the next generation to develop and/or implement new diagnostic/therapeutic tools. Finally, there is an urgent need for improved technology for observing and tracking the endothelium. Several new diagnostic strategies may ultimately bear fruit in the field of endothelial biomedicine. First, the measurement of a panel of activation markers—as distinct from a

Endothelial cells in health and disease

28

single mediator—may yield previously unappreciated patterns in response that will aid in diagnosis and/or therapeutic monitoring. Further advances in proteomics will provide new technology platforms (e.g., protein-based chips) to simultaneously monitor dozens of biomarkers. One of the limitations of assaying blood from a peripheral vein or artery is that the sample represents the sum average of activity from multiple vascular beds, and may therefore overlook localized “hot spots” in specific sites of the vasculature. Thus, the use of catheters to sample blood from one or another vascular bed might provide a diagnostic window into these lesions. Refined protocols to isolate and interrogate the phenotype of circulating endothelial cells may yield insight into the function of their vascular-bed-of-origin. Perhaps a more comprehensive and systematic analysis of pathological specimens (e.g., skin biopsies) will provide data that correlate—at least to some extent—with endothelial cell function. Finally, imaging will play an increasingly important role in the clinic. Doppler measurements of blood flow, magnetic resonance angiography, and CT scanning are widely available in clinical practice and should be readily applicable to the study of the endothelial (dys) function in many disease states. Molecular imaging, which combines the power of proteomics and advanced labeling techniques, promises to revolutionize the diagnosis of endothelial-based disorders.

11.3. Should We Create a New Discipline? While we are standing on the cusp of a golden age in vascular biology—one that should see the endothelium promoted as a newly recognized organ—medicine, as it is currently structured, is poorly qualified to carry the field into the 21st century. In his book, “From Chaos to Care,” David Lawrence states the following: Strip away the professionals, the treatments, the equipment, and the institutions of modern medicine, and one finds that care is organized for an earlier, far different period of history…. Medicine today involves a vast, fragmented often isolated array of human, technical and institutional resources. (72) Lawrence is pointing out the limitations of the current infrastructure as an argument for overhauling the medical care system. However, the comments are also relevant to the present discussion. Medicine consists of series of highly fortified, largely historical, and in many cases outdated disciplines, which like cardiology have been cemented over the years by intellectual, financial, and administrative forces, not to mention fierce pride and loyalty. The American Board of Medical Specialties is an organization of 24 approved medical specialty boards. One of these, the American Board of Internal Medicine recognizes nine subspecialties, and provides certificates of added qualification for an additional seven disciplines—none of which systematically embraces endothelial-based diseases. Each of these fields has its own College, Association or Society. In many cases, basic and/or clinical advances are published in organ-specific biomedical journals. The resulting barriers discourage communication between disciplines and by way of mutual reinforcement create a culture that is highly resistant to change. The net result: there is little willingness to embrace the endothelium for what it is, namely a cell layer that is

The endothelium as an organ

29

teeming with life, and every bit as active (if not more so) than most other organs in the body. There are at least two ways to begin moving the field of endothelial cell biology into the clinic at a pace that is commensurate with advances at the bench. One is to design a new clinical discipline in endothelial medicine. Training could constitute an added qualification to such specialties as cardiology, pulmonary or hematology, just to name a few. It may be argued that a new discipline in “endothelial biomedicine” would only add to the fragmented state of medicine. Or that it is too early—that there are insufficient diagnostic and therapeutic tools—to justify a separate field. The counter-argument, of course, is that a new discipline would in fact represent a synthesis of an otherwise highly scattered field, and would provide a necessary framework for bridging the bench-tobedside gap. An alternative approach is to improve the communication between and within existing clinical and basic disciplines. Endothelial cell investigators from different disciplines tend to represent a minority in their respective fields and have little opportunity to interact with one another. For example, a researcher studying the blood-brain barrier may spend a great deal of time interacting with neurologists and neuroscientists, but remarkably little time with endothelial cell biologists whose work focuses on other vascular beds. As another example, a clinician-scientist in pulmonary medicine interested in understanding the molecular basis of pulmonary hypertension and improving treatment for this condition is unlikely to cross paths with a hematologist who studies the role of the endothelium in thrombotic thrombocytopenic purpura. Both investigators are studying a common cell type and stand to gain from one another’s knowledge. Not only are present divisions antiquated and artificial, they also dampen cross-fertilization and progress. Transcending these barriers will require concerted effort on the part of many by way of collaborative basic research, interinstitutional and industry consortiums for developing novel diagnostic and therapeutic tools, and multidisciplinary team approaches to patient care. These and other efforts to synthesize the field are important first step toward tapping the endothelium for its full potential.

ACKNOWLEDGMENTS I thank Guillermo Garcia-Cardena and Susan Glueck for their critical review of the chapter. I thank Michael Gimbrone for his helpful comments. I am indebted to students in the 2004 Harvard MIT Division of Health Sciences and Technology course HST-527 for their feedback and fresh perspective. This work was supported in part by National Institutes of Health grants HL63609, HL65216, and HL36028.

REFERENCES 1. Wolinsky H. A proposal linking clearance of circulating lipoproteins to tissue metabolic activity as a basis for understanding atherogenesis. Circ Res 1980; 47(3):301–11. 2. Cines DB, Pollak ES, Buck CA, et al. Endothelial cells in physiology and in the pathophysiology of vascular disorders. Blood 1998; 91(10):3527–3561.

Endothelial cells in health and disease

30

3. Braunwald E. Harrison’s Principles of Internal Medicine. 15th ed. New York: McGraw-Hill, 2001. 4. Aird WC. The hematologic system as a marker of organ dysfunction in sepsis. Mayo Clin Proc 2003; 78(7):869–881. 5. Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001; 344(10):699–709. 6. Esmon CT. New mechanisms for vascular control of inflammation mediated by natural anticoagulant proteins. J Exp Med 2002; 196(5):561–564. 7. Joyce DE, Gelbert L, Ciaccia A, DeHoff B, Grinnell BW. Gene expression profile of antithrombotic protein c defines new mechanisms modulating inflammation and apoptosis. J Biol Chem 2001; 276(14):11199–11203. 8. Dawkins R. Climbing Mount Improbable. 1st American ed. New York: Norton, 1996. 9. Darwin C. On the Origin of Species by Means of Natural Selection. London: J.Murray, 1859. 10. Aird WC. Hemostasis and irreducible complexity. J Thromb Haemost 2003; 1(2): 227–230. 11. Shigei T, Tsuru H, Ishikawa N, Yoshioka K. Absence of endothelium in invertebrate blood vessels: significance of endothelium and sympathetic nerve/medial smooth muscle in the vertebrate vascular system. Jpn J Pharmacol 2001; 87(4):253–260. 12. Bennett HS, Luft JH, Hampton JC. Morphological classifications of vertebrate blood capillaries. Am J Physiol 1959; 196(2):381–390. 13. Majno G, Palade GE, Schoefl GI. Studies on inflammation. II. The site of action of histamine and serotonin along the vascular tree: a topographic study. J Biophys Biochem Cytol 1961; 11:607–626. 14. Pugatch EM, Foster EA, Macfarlane DE, Poole JC. The extraction and separation of activators and inhibitors of fibrinolysis from bovine endothelium and mesothelium. Br J Haematol 1970; 18(6):669–681. 15. Levin EG, Loskutoff DJ. Comparative studies of the fibrinolytic activity of cultured vascular cells. Thromb Res 1979; 15(5–6):869–878. 16. Schleef RR, Bevilacqua MP, Sawdey M, Gimbrone MA Jr, Loskutoff DJ. Cytokine activation of vascular endothelium. Effects on tissue-type plasminogen activator and type 1 plasminogen activator inhibitor. J Biol Chem 1988; 263(12):5797–5803. 17. Jaffe EA, Hoyer LW, Nachman RL. Synthesis of antihemophilic factor antigen by cultured human endothelial cells. J Clin Invest 1973; 52(11):2757–2764. 18. Gimbrone MA Jr, Alexander RW. Angiotensin II stimulation of prostaglandin production in cultured human vascular endothelium. Science 1975; 189(4198):219–220. 19. Furchgott RF, Zawadzki JV. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 1980; 288(5789):373–376. 20. Pober JS, Collins T, Gimbrone MA Jr, et al. Lymphocytes recognize human vascular endothelial and dermal fibroblast Ia antigens induced by recombinant immune interferon. Nature 1983; 305(5936):726–729. 21. Bevilacqua MP, Pober JS, Wheeler ME, Cotran RS, Gimbrone MA Jr. Interleukin-1 activation of vascular endothelium. Effects on procoagulant activity and leukocyte adhesion. Am J Pathol 1985; 121(3):394–403. 22. Bevilacqua MP, Pober JS, Wheeler ME, Cotran RS, Gimbrone MA Jr. Interleukin 1 acts on cultured human vascular endothelium to increase the adhesion of polymorphonuclear leukocytes, monocytes, and related leukocyte cell lines. J Clin Invest 1985; 76(5):2003–2011. 23. Barabasi AL, Bonabeau E. Scale-free networks. Sci Am 2003; 288(5):60–69. 24. Vogelstein B, Lane D, Levine AJ. Surfing the p53 network. Nature 2000; 408(6810):307–310. 25. Gordon D. Ants at Work: How an Insect Society is Organized. New York: Free Press, 1999. 26. Jacobs J. The Death and Life of Great American Cities. New York: Random House, 1961. 27. Aird WC. Endothelial cell dynamics and complexity theory. Crit Care Med 2002; 30(Suppl 5):S180–S185.

The endothelium as an organ

31

28. Hibbs RG, Burch GE, Phillips JH. The fine structure of the small blood vessels of normal human dermis and subcutis. Am Heart J 1958; 56(5):662–670. 29. Florey L. The endothelial cell. Br Med J 1966; 2:487−490. 30. Reese TS, Karnovsky MJ. Fine structural localization of a blood-brain barrier to exogenous peroxidase. J Cell Biol 1967; 34(1):207–217. 31. Gimbrone MA Jr, Cotran RS, Folkman J. Endothelial regeneration: studies with human endothelial cells in culture. Ser Haematol 1973; 6(4):453–455. 32. Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J Clin Invest 1973; 52(11):2745–2756. 33. Gimbrone MA Jr, Cotran RS, Folkman J. Human vascular endothelial cells in culture. Growth and DNA synthesis. J Cell Biol 1974; 60(3):673–684. 34. Gimbrone MA Jr, Cotran RS. Human vascular smooth muscle in culture. Growth and ultrastructure. Lab Invest 1975; 33(1):16–27. 35. Kumar S, West DC, Ager A. Heterogeneity in endothelial cells from large vessels and microvessels. Differentiation 1987; 36(1):57–70. 36. Page C, Rose M, Yacoub M, Pigott R. Antigenic heterogeneity of vascular endothelium. Am J Pathol 1992; 141(3):673–683. 37. Turner RR, Beckstead JH, Warnke RA, Wood GS. Endothelial cell phenotypic diversity. In situ demonstration of immunologic and enzymatic heterogeneity that correlates with specific morphologic subtypes. Am J Clin Pathol 1987; 87(5):569–575. 38. Aird WC. Endothelial cell heterogeneity. Crit Care Med 2003; 31(Suppl 4):S221–S230. 39. Stevens T, Rosenberg R, Aird W, et al. NHLBI workshop report: endothelial cell phenotypes in heart, lung, and blood diseases. Am J Physiol Cell Physiol 2001; 281(5):C1422–C1433. 40. Rosenberg RD, Aird WC. Vascular-bed-specific hemostasis and hypercoagulable states. N Engl J Med 1999; 340(20):1555–1564. 41. Auerbach R, Alby L, Morrissey LW, Tu M, Joseph J. Expression of organ-specific antigens on capillary endothelial cells. Microvasc Res 1985; 29(3):401–411. 42. Dewey CF Jr, Bussolari SR, Gimbrone MA Jr, Davies PF. The dynamic response of vascular endothelial cells to fluid shear stress. J Biomech Eng 1981; 103(3):177–185. 43. Bevilacqua MP, Pober JS, Majeau GR, Cotran RS, Gimbrone MA Jr. Interleukin 1 (IL-1) induces biosynthesis and cell surface expression of procoagulant activity in human vascular endothelial cells. J Exp Med 1984; 160(2):618–623. 44. Davies PF, Dewey CF Jr, Bussolari SR, Gordon EJ, Gimbrone MA Jr. Influence of hemodynamic forces on vascular endothelial function. In vitro studies of shear stress and pinocytosis in bovine aortic cells. J Clin Invest 1984; 73(4):1121–1129. 45. Bevilacqua MP, Schleef RR, Gimbrone MA Jr, Loskutoff DJ. Regulation of the fibrinolytic system of cultured human vascular endothelium by interleukin 1. J Clin Invest 1986; 78(2):587– 591. 46. Bevilacqua MP, Pober JS, Majeau GR, Fiers W, Cotran RS, Gimbrone MA Jr. Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium: characterization and comparison with the actions of interleukin 1. Proc Natl Acad Sci USA 1986; 83(12):4533–4537. 47. Cohnheim J. Lectures on General Pathology. Vol.1. The New Sydenham Society, 1882. 48. Allison F Jr, Smith MR, Wood WB Jr. Studies on the pathogenesis of acute inflammation. I. The inflammatory reaction to thermal injury as observed in the rabbit ear chamber. J Exp Med 1955; 102(6):655–668. 49. Mc Govern. Reactions to injury of vascular endothelium with special reference to the problem of thrombosis. J Pathol Bacteriol 1955; 69(1–2):283–293. 50. Bangham AD. The adhesiveness of leukocytes with special references to zeta potential. Ann NY Acad Sci 1964; 116:945–949.

Endothelial cells in health and disease

32

51. Pober JS, Gimbrone MA Jr. Expression of Ia-like antigens by human vascular endothelial cells is inducible in vitro: demonstration by monoclonal antibody binding and immunoprecipitation. Proc Natl Acad Sci USA 1982; 79(21):6641–6645. 52. Pober JS, Bevilacqua MP, Mendrick DL, Lapierre LA, Fiers W, Gimbrone MA Jr. Two distinct monokines, interleukin 1 and tumor necrosis factor, each independently induce biosynthesis and transient expression of the same antigen on the surface of cultured human vascular endothelial cells. J Immunol 1986; 136(5):1680–1687. 53. Bevilacqua MP, Pober JS, Mendrick DL, Cotran RS, Gimbrone MA Jr. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc Natl Acad Sci USA 1987; 84(24):9238–9242. 54. Bevilacqua MP, Stengelin S, Gimbrone MA Jr, Seed B.Endothelial leukocyte adhesion molecule 1: an inducible receptor for neutrophils related to complement regulatory proteins and lectins. Science 1989; 243(4895):1160–1165. 55. Bevilacqua MP, Pober JS, Mendrick DL, Cotran RS, Gimbrone MA Jr. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc Natl Acad Sci USA 1987; 84(24):9238–9242. 56. Bevilacqua MP, Pober JS, Majeau GR, Fiers W, Cotran RS, Gimbrone MA Jr. Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium: characterization and comparison with the actions of interleukin 1. Proc Natl Acad Sci USA 1986; 83(12):4533–4537. 57. Schleimer RP, Rutledge BK. Cultured human vascular endothelial cells acquire adhesiveness for neutrophils after stimulation with interleukin 1, endotoxin, and tumor-promoting phorbol diesters. J Immunol 1986; 136(2):649–654. 58. Gamble JR, Harlan JM, Klebanoff SJ, Vadas MA. Stimulation of the adherence of neutrophils to umbilical vein endothelium by human recombinant tumor necrosis factor. Proc Natl Acad Sci USA 1985; 82(24):8667–8671. 59. Pober JS, Cotran RS. Cytokines and endothelial cell biology. Physiol Rev 1990; 70(2):427–451. 60. Cotran RS, Gimbrone MA Jr, Bevilacqua MP, Mendrick DL, Pober JS. Induction and detection of a human endothelial activation antigen in vivo. J Exp Med 1986; 164(2):661–666. 61. Willms-Kretschmer K, Flax MH, Cotran RS. The fine structure of the vascular response in hapten-specific delayed hypersensitivity and contact dermatitis. Lab Invest 1967; 17(3):334– 349. 62. Pober JS, Gimbrone MA Jr, Lapierre LA, et al. Overlapping patterns of activation of human endothelial cells by interleukin 1, tumor necrosis factor, and immune interferon. J Immunol 1986; 137(6):1893–1896. 63. Dekker RJ, van Soest S, Fontijn RD, et al. Prolonged fluid shear stress induces a distinct set of endothelial cell genes, most specifically lung Kruppel-like factor (KLF2). Blood 2002; 100(5):1689–1698. 64. Florey L. The endothelial cell. Br Med J 1966; 5512:487–490. 65. Ross R, Glomset JA. Atherosclerosis and the arterial smooth muscle cell: proliferation of smooth muscle is a key event in the genesis of the lesions of atherosclerosis. Science 1973; 180(93):1332–1339. 66. Schwartz SM, Gajdusek CM, Reidy MA, Selden SC 3rd, Haudenschild CC. Maintenance of integrity in aortic endothelium. Fed Proc 1980; 39(9):2618–2625. 67. Gimbrone MA Jr, ed. Endothelial Dysfunction and the Pathogenesis of Atherosclerosis. New York: Springer-Verlag, 1980. 68. Ludmer PL, Selwyn AP, Shook TL, et al. Paradoxical vasoconstriction induced by acetylcholine in atherosclerotic coronary arteries. N Engl J Med 1986; 315(17):1046–1051. 69. Cybulsky MI, Gimbrone MA Jr. Endothelial expression of a mononuclear leukocyte adhesion molecule during atherogenesis. Science 1991; 251(4995):788–791.

The endothelium as an organ

33

70. Osborn L, Hession C, Tizard R, et al. Direct expression cloning of vascular cell adhesion molecule 1, a cytokine-induced endothelial protein that binds to lymphocytes. Cell 1989; 59(6):1203–1211. 71. Haber E. Scientific American Molecular Cardiovascular Medicine. New York: Scientific American Medicine, 1995. 72. Lawrence D. From Chaos to Care: The Promise of Team-Based Medicine. Cambridge, MA: Perseus Pub, 2002. 73. Weibel ER. The pathway for oxygen: Structure and Function in the Mammalian Respiratory System. Cambridge, MA: Harvard University Press, 1984. 74. Ruoslahti E. Specialization of tumour vasculature. Nat Rev Cancer 2002; 2(2):83–90. 75. St Croix B, Rago C, Velculescu V, et al. Genes expressed in human tumor endothelium. Science 2000; 289(5482):1197–1202. 76. Turhan A, Weiss LA, Mohandas N, Coller BS, Frenette PS. Primary role for adherent leukocytes in sickle cell vascular occlusion: a new paradigm. Proc Natl Acad Sci USA 2002; 99(5):3047–3051. 77. Solovey AA, Solovey AN, Harkness J, Hebbel RP. Modulation of endothelial cell activation in sickle cell disease: a pilot study. Blood 2001; 97(7):1937–1941. 78. Hebbel RP, Vercellotti GM. The endothelial biology of sickle cell disease. J Lab Clin Med 1997; 129(3):288–293. 79. Saunthararajah Y, Hillery CA, Lavelle D, et al. Effects of 5-aza-2′-deoxycytidine on fetal hemoglobin levels, red cell adhesion, and hematopoietic differentiation in patients with sickle cell disease. Blood 2003; 102:3865–3870. 80. Cheung YF, Chan GC, Ha SY. Arterial stiffness and endothelial function in patients with betathalassemia major. Circulation 2002; 106(20):2561–2566. 81. Butthep P, Rummavas S, Wisedpanichkij R, Jindadamrongwech S, Fucharoen S, Bunyaratvej A. Increased circulating activated endothelial cells, vascular endothelial growth factor, and tumor necrosis factor in thalassemia. Am J Hematol 2002; 70(2):100–106. 82. Gaenzer H, Marschang P, Sturm W, et al. Association between increased iron stores and impaired endothelial function in patients with hereditary hemochromatosis. J Am Coll Cardiol 2002; 40(12):2189–2194. 83. Neunteufl T, Heher S, Stefenelli T, Pabinger I, Gisslinger H. Endothelial dysfunction in patients with polycythaemia vera. Br J Haematol 2001; 115(2):354–359. 84. Musolino C, Alonci A, Bellomo G, et al. Myeloproliferative disease: markers of endothelial and platelet status in patients with essential thrombocythemia and poly-cythemia vera. Hematol 2000; 4(5):397−402. 85. Takatsuka H, Wakae T, Mori A, Okada M, Okamoto T, Kakishita E. Effects of total body irradiation on the vascular endothelium. Clin Transplant 2002; 16(5): 374–377. 86. Paris F, Fuks Z, Kang A, et al. Endothelial apoptosis as the primary lesion initiating intestinal radiation damage in mice. Science 2001; 293(5528):293–297. 87. Nohe B, Kiefer RT, Ploppa A, Haeberle HA, Schroeder TH, Dieterich HJ. The effects of fresh frozen plasma on neutrophil-endothelial interactions. Anesth Analg 2003; 97(1):216–221. Table of contents. 88. Luk CS, Gray-Statchuk LA, Cepinkas G, Chin-Yee IH. WBC reduction reduces storageassociated RBC adhesion to human vascular endothelial cells under conditions of continuous flow in vitro. Transfusion 2003; 43(2):151–156. 89. Wyman TH, Bjornsen AJ, Elzi DJ, et al. A two-insult in vitro model of PMN-mediated pulmonary endothelial damage: requirements for adherence and chemokine release. Am J Physiol Cell Physiol 2002; 283(6):C1592–C1603. 90. Dry SM, Bechard KM, Milford EL, Churchill WH, Benjamin RJ. The pathology of transfusionrelated acute lung injury. Am J Clin Pathol 1999; 112(2):216–221.

Endothelial cells in health and disease

34

91. Mitra D, Jaffe EA, Weksler B, Hajjar KA, Soderland C, Laurence J. Thrombotic thrombocytopenic purpura and sporadic hemolytic-uremic syndrome plasmas induce apoptosis in restricted lineages of human microvascular endothelial cells. Blood 1997; 89(4):1224–1234. 92. Dong JF, Moake JL, Bernardo A, et al. ADAMTS-13 metalloprotease interacts with the endothelial cell-derived ultra-large von Willebrand factor. J Biol Chem 2003; 278(32):29633– 29639. 93. Aird WC. Vascular bed-specific hemostasis: role of endothelium in sepsis pathogenesis. Crit Care Med 2001; 29(7):S28–S35. 94. Peters CJ, Zaki SR. Role of the endothelium in viral hemorrhagic fevers. Crit Care Med 2002; 30(suppl 5):S268–S273. 95. Kim KS. Strategy of Escherichia coli for crossing the blood-brain barrier. J Infect Dis 2002; 186(suppl 2):S220–S224. 96. Hotchkiss RS, Tinsley KW, Swanson PE, Karl IE. Endothelial cell apoptosis in sepsis. Crit Care Med 2002; 30(suppl 5):S225–S228. 97. Aird WC. The role of the endothelium in severe sepsis and the multiple organ dysfunction syndrome. Blood 2003; 23:23. 98. Reinhart K, Bayer O, Brunkhorst F, Meisner M. Markers of endothelial damage in organ dysfunction and sepsis. Crit Care Med 2002; 30(suppl 5):S302–S312. 99. Bonetti PO, Lerman LO, Lerman A. Endothelial dysfunction: a marker of atherosclero tic risk. Arterioscler Thromb Vasc Biol 2003; 23(2):168–175. 100. Targonski PV, Bonetti PO, Pumper GM, Higano ST, Holmes DR Jr, Lerman A Coronary endothelial dysfunction is associated with an increased risk of cerebrovascular events. Circulation 2003; 107(22):2805–2809. 101. Quyyumi AA. Prognostic value of endothelial function. Am J Cardiol 2003; 91(12A):19H– 24H. 102. Stone PH, Coskun AU, Kinlay S, et al. Effect of endothelial shear stress on the progression of coronary artery disease, vascular remodeling, and in-stent restenosis in humans: in vivo 6-month follow-up study. Circulation 2003; 108(4):438–444. 103. Effect of nifedipine and cerivastatin on coronary endothelial function in patients with coronary artery disease: the ENCORE I investigators study (Evaluation of Nifedipine and Cerivastatin On Recovery of coronary Endothelial function). Circulation 2003; 107(3):422–428. 104. Chenevard R, Hurlimann D, Bechir M, et al. Selective COX-2 inhibition improves endothelial function in coronary artery disease. Circulation 2003; 107(3):405–409. 105. McNamara DM, Holubkov R, Postava L, et al. Effect of the Asp298 variant of endothelial nitric oxide synthase on survival for patients with congestive heart failure. Circulation 2003; 107(12):1598–1602. 106. Dixon LJ, Morgan DR, Hughes SM, et al. Functional consequences of endothelial nitric oxide synthase uncoupling in congestive cardiac failure. Circulation 2003; 107(13):1725–1728. 107. van den Berg BM, Vink H, Spaan JA. The endothelial glycocalyx protects against myocardial edema. Circ Res 2003; 92(6):592–594. 108. Ferrari R, Bachetti T, Agnoletti L, Comini L, Curello S. Endothelial function and dysfunction in heart failure. Eur Heart J 1998; 19(suppl G):G41–G47. 109. Leask RL, Jain N, Butany J. Endothelium and valvular diseases of the heart. Microsc Res Tech 2003; 60(2):129–137. 110. Poggianti E, Venneri L, Chubuchny V, Jambrik Z, Baroncini LA, Picano E. Aortic valve sclerosis is associated with systemic endothelial dysfunction. J Am Coll Cardiol 2003; 41(1):136–141. 111. Han B, Luo G, Shi ZZ, et al. Gamma-glutamyl leukotrienase, a novel endothelial membrane protein, is specifically responsible for leukotriene D(4) formation in vivo. Am J Pathol 2002; 161(2):481–490.

The endothelium as an organ

35

112. Ulfman LH, Joosten DP, van Aalst CW, et al. Platelets promote eosinophil adhesion of patients with asthma to endothelium under flow conditions. Am J Respir Cell Mol Biol 2003; 28(4):512–519. 113. Churg A, Wang RD, Tai H, et al. Macrophage metalloelastase mediates acute cigarette smokeinduced inflammation via tumor necrosis factor-alpha release. Am J Respir Crit Care Med 2003; 167(8):1083–1089. 114. Cella G, Sbarai A, Mazzaro G, et al. Plasma markers of endothelial dysfunction in chronic obstructive pulmonary disease. Clin Appl Thromb Hemost 2001; 7(3):205–208. 115. Nagaya N, Kangawa K, Kanda M, et al. Hybrid cell-gene therapy for pulmonary hypertension based on phagocytosing action of endothelial progenitor cells. Circulation 2003; 108(7):889– 895. 116. Ameshima S, Golpon H, Cool CD, et al. Peroxisome proliferator-activated receptor gamma (PPARgamma) expression is decreased in pulmonary hypertension and affects endothelial cell growth. Circ Res 2003; 92(10):1162–1169. 117. Yeager ME, Golpon HA, Voelkel NF, Tuder RM. Microsatellite mutational analysis of endothelial cells within plexiform lesions from patients with familial, pediatric, and sporadic pulmonary hypertension. Chest 2002; 121(suppl 3):61S. 118. Muller AM, Hermanns MI, Cronen C, Kirkpatrick CJ. Comparative study of adhesion molecule expression in cultured human macro- and microvascular endothelial cells. Exp Mol Pathol 2002; 73(3):171–180. 119. Sutton TA, Mang HE, Campos SB, Sandoval RM, Yoder MC, Molitoris BA. Injury of the renal microvascular endothelium alters barrier function after ischemia. Am J Physiol Renal Physiol 2003; 285(2):F191–F198. 120. Sutton TA, Fisher CJ, Molitoris BA. Microvascular endothelial injury and dysfunction during ischemic acute renal failure. Kidney Int 2002; 62(5):1539–1549. 121. Brodsky SV, Yamamoto T, Tada T, et al. Endothelial dysfunction in ischemic acute renal failure: rescue by transplanted endothelial cells. Am J Physiol Renal Physiol 2002; 282(6):F1140–F1149. 122. Bennett-Richards K, Kattenhorn M, Donald A, Oakley G, Varghese Z, Rees L, Deanfield JE. Does oral folic acid lower total homocysteine levels and improve endothelial function in children with chronic reral failure? Circulation 2002; 105:1810–1815. 123. Cottone S, Mule G, Amato F, et al. Amplified biochemical activation of endothelial function in hypertension associated with moderate to severe renal failure. J Nephrol 2002; 15(6):643– 648. 124. Jacobson SH, Egberg N, Hylander B, Lundahl J. Correlation between soluble markers of endothelial dysfunction in patients with renal failure. Am J Nephrol 2002; 22(1):42–47. 125. Ma L, Elliott SN, Cirino G, Buret A, Ignarro LJ, Wallace JL. Platelets modulate gastric ulcer healing: role of endostatin and vascular endothelial growth factor release. Proc Natl Acad Sci USA 2001; 98(11):6470–6475. 126. Ma L, Wallace JL. Endothelial nitric oxide synthase modulates gastric ulcer healing in rats. Am J Physiol Gastrointest Liver Physiol 2000; 279(2):G341–G346. 127. Danese S, de la Motte C, Sturm A, et al. Platelets trigger a CD40-dependent inflammatory response in the microvasculature of inflammatory bowel disease patients. Gastroenterology 2003; 124(5):1249–1264. 128. Rijcken E, Krieglstein CF, Anthoni C, et al. ICAM-1 and VCAM-1 antisense oligonucleotides attenuate in vivo leucocyte adherence and inflammation in rat inflammatory bowel disease. Gut 2002; 51(4):529–535. 129. Breiner KM, Schaller H, Knolle PA. Endothelial cell-mediated uptake of a hepatitis B virus: a new concept of liver targeting of hepatotropic microorganisms. Hepatology 2001; 34(4 Pt 1):803–808. 130. Cacoub P, Ghillani P, Revelen R, et al. Anti-endothelial cell auto-antibodies in hepatitis C virus mixed cryoglobulinemia. J Hepatol 1999; 31(4):598–603.

Endothelial cells in health and disease

36

131. Yokomori H, Oda M, Ogi M, Sakai K, Ishii H. Enhanced expression of endothelial nitric oxide synthase and caveolin-1 in human cirrhosis. Liver 2002; 22(2):150–158. 132. Mouta Carreira C, Nasser SM, di Tomaso E, et al. LYVE-1 is not restricted to the lymph vessels: expression in normal liver blood sinusoids and down-regulation in human liver cancer and cirrhosis. Cancer Res 2001; 61(22):8079–8084. 133. Chen HM, Sunamura M, Shibuya K, et al. Early microcirculatory derangement in mild and severe pancreatitis models in mice. Surg Today 2001; 31(7):634–642. 134. Masamune A, Shimosegawa T, Fujita M, Satoh A, Koizumi M, Toyota T. Ascites of severe acute pancreatitis in rats transcriptionally up-regulates expression of interleukin-6 and -8 in vascular endothelium and mononuclear leukocytes. Dig Dis Sci 2000; 45(2):429–437. 135. Pablos JL, Santiago B, Galindo M, et al. Synoviocyte-derived CXCL12 is displayed on endothelium and induces angiogenesis in rheumatoid arthritis. J Immunol 2003; 170(4):2147– 2152. 136. Ferrell WR, Lockhart JC, Kelso EB, et al. Essential role for proteinase-activated receptor-2 in arthritis. J Clin Invest 2003; 111(1):35–41. 137. Klimiuk PA, Sierakowski S, Latosiewicz R, et al. Soluble adhesion molecules (ICAM-1, VCAM-1, and E-selectin) and vascular endothelial growth factor (VEGF) in patients with distinct variants of rheumatoid synovitis. Ann Rheum Dis 2002; 61(9):804–809. 138. Apras S, Ertenli I, Ozbalkan Z, et al. Effects of oral cyclophosphamide and prednisolone therapy on the endothelial functions and clinical findings in patients with early diffuse systemic sclerosis. Arthritis Rheum 2003; 48(8):2256–2261. 139. Cerinic MM, Valentini G, Sorano GG, et al. Blood coagulation, fibrinolysis, and markers of endothelial dysfunction in systemic sclerosis. Semin Arthritis Rheum 2003; 32(5):285–295. 140. Marie I, Beny JL. Endothelial dysfunction in murine model of systemic sclerosis: tight-skin mice 1. J Invest Dermatol 2002; 119(6):1379–1387. 141. Wheatcroft SB, Williams IL, Shah AM, Kearney MT. Pathophysiological implications of insulin resistance on vascular endothelial function. Diabet Med 2003; 20(4):255–268. 142. Taylor AA. Pathophysiology of hypertension and endothelial dysfunction in patients with diabetes mellitus. Endocrinol Metab Clin North Am 2001; 30(4):983–997. 143. Cherian P, Hankey GJ, Eikelboom JW, et al. Endothelial and platelet activation in acute ischemic stroke and its etiological subtypes. Stroke 2003; 34:2132–2137. 144. Cheng T, Liu D, Griffin JH, et al. Activated protein C blocks p53-mediated apoptosis in ischemic human brain endothelium and is neuroprotective. Nat Med 2003; 9(3):338–342. 145. Brown RC, Davis TP. Calcium modulation of adherens and tight junction function: a potential mechanism for blood-brain barrier disruption after stroke. Stroke 2002; 33(6):1706–1711. 146. Greenwood J, Walters CE, Pryce G, et al. Lovastatin inhibits brain endothelial cell Rhomediated lymphocyte migration and attenuates experimental autoimmune encephalomyelitis. Faseb J 2003; 17(8):905–907. 147. Kuruganti PA, Hinojoza JR, Eaton MJ, Ehmann UK, Sobel RA. Interferon-beta counteracts inflammatory mediator-induced effects on brain endothelial cell tight junction molecules— implications for multiple sclerosis. J Neuropathol Exp Neurol 2002; 61(8):710–724. 148. Kauma S, Takacs P, Scordalakes C, Walsh S, Green K, Peng T. Increased endothelial monocyte chemoattractant protein-1 and interleukin-8 in preeclampsia. Obstet Gynecol 2002; 100(4):706–714. 149. Maynard SE, Min JY, Merchan J, et al. Excess placental soluble fms-like tyrosine kinase 1 (sFltl) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 2003; 111(5):649–658.

2 Blood-Brain Barrier Eric V.Shusta Department of Chemical and Biological Engineering, University of Wisconsin-Madison, Madison, Wisconsin, U.S.A.

1. INTRODUCTION AND BACKGROUND This chapter will describe the blood-brain barrier (BBB), a unique class of endothelium that separates the bloodstream from the neuronal milieu. The BBB has been studied quite extensively using classical pathology, physiology, cell biology, and biochemistry techniques. However, exciting opportunities afforded by genomics and proteomics technologies allow unparalleled access to the molecular constituents and mechanisms that endow the BBB with its functional characteristics. Combination of the various approaches mentioned above has led to the identification of functional and regulatory roles for the BBB in both health and central nervous system (CNS) disease. The BBB is functionally defined as the impermeable vasculature that is found throughout the entire brain. The large pial vessels associated with the meningeal membranes that envelop the brain have a lower permeability than that found in most peripheral vasculature but do not form an explicit barrier. However, advancing from the meninges into the brain matter, vascular permeability decreases rapidly and barrier-like impermeability (BBB) arises. The BBB impermeability is exhibited by brain vasculature of all sizes including, arterioles, capillaries, and venules as well as larger arteries and veins. Except for small regions of the brain, the impermeable vasculature extends throughout the entire brain (Sec. 2.1). The barrier is principally formed by the endothelial cell and functionally excludes blood-borne substances from entering the brain tissue. In addition to the endothelial barrier, several perivascular cell types on the brain side of the vessels intimately interact with the endothelial cell and help elicit the BBB phenotype. This endothelial cell-perivascular cell composite has been coined the “neurovascular unit” (Sec. 2). The perivascular cells include: vascular smooth muscle cells that line larger vessels, pericytes that share a basement membrane with endothelial cells, astrocytes which are glial cells that function in both endothelial cell differentiation and neuronal support, and neurons. The functional importance of these cell types on the neurovascular unit is discussed in detail throughout this chapter.

The endothelium as an organ

39

2. BBB STRUCTURE AND FUNCTION The BBB is comprised of a specialized class of endothelium that forms a cellular barrier between the bloodstream and the interstices of the brain. The brain microvasculature is distinct from non-BBB vascular beds found in the periphery because it acts as a barrier to ions, small molecule solutes, peptides, and proteins that can pass freely in many other vascular beds. As a consequence of this barrier function, the endothelial cell acts as a regulatory interface for signaling and transport between the blood and the brain. By restricting non-specific flux of blood-borne constituents, the BBB plays an important role in maintaining parenchymal homeostasis and protecting the neuronal environment of the CNS from fluctuations in blood composition. The relative lack of permeability of brain capillaries mandates the presence of nutrient transport systems, especially those required to supply the brain with energy. Finally, the BBB also participates in immune surveillance mechanisms since the circulating lymphocytes do not have free access to the interstitial space of the brain. The main phenotypic attribute promoting barrier function is the presence of intercellular tight junctions between neighboring endothelial cells. These epithelial-like tight junctions regulate the paracellular flux of both metabolites and cells, and in essence render the brain impermeable to most circulating substances. In addition, the BBB displays only low levels of pinocytotic transport. It follows that the BBB must rely on the presence of specialized transporters and carrier systems to regulate the bi-directional flux of molecules between the bloodstream and the brain. The brain is a highly vascularized organ, presumably reflecting its demands for highlevel aerobic metabolism. The human brain has been estimated to contain over 400 miles of capillaries that are spaced at distances of approximately 40 µm. This high vascular density and surface area permit rapid and efficient transport of selected substrates across the BBB (1). The capillaries themselves range from 10 µm in diameter at the precapillary arteriole to about 5 µm at their smallest diameter (Fig. 1) (2,3). Because of the similarity in size, isolated capillary preparations invariably include arteriole and venule components and are often referred to as brain microvasculature preparations. Endothelial cells of the BBB are elongated and thin, measuring a mere 100–300 nm between luminal (vessel) and abluminal (brain) compartments (3–5) yielding rapid molecular transport. Finally, by asymmetric segregation of the luminal membrane constituents from those located in the abluminal membrane, the tight junctions render the BBB endothelium highly polarized (6–8). Although the endothelium is the principal determinant of barrier function, perivascular non-endothelial cells have also been shown to make significant contributions (Fig. 2). Vascular smooth muscle cells line the precapillary arterioles (Fig. 1B), while pericytes dot the abluminal side of capillaries (Fig. 1B and C). Pericytes share a common basement membrane with the endothelial cells (Fig. 2). The end feet of astroglial cells form a cagelike network encasing a large fraction of the endothelial cell surface (9) (Fig. 1D–F) and have been shown to play very important roles in eliciting the barrier phenotype in endothelial cells (see below). This set of cells then interacts with the brain microvascular endothelium to form the specialized, impermeable neurovascular unit.

Endothelial cells in health and disease

40

2.1 Molecular Architecture of Tight Junctions The barrier function of brain endothelial cells was confirmed in classical experiments where it was noted that intravenously injected horse radish peroxidase (40,000 Da)

Figure 1 The intimate relationship between brain endothelial cells and perivascular cells. (A) Light microscopy of isolated bovine brain capillaries stained with o-toluidine blue. (B) Light microscopy of isolated bovine brain capillaries immunostained with an anti-smooth muscle actin antibody. Note that the precapillary arteriole stains due to the presence of a smooth muscle lining, while the capillaries do not exhibit any staining for smooth muscle actin.

The endothelium as an organ

41

Pericytes are indicated with arrowheads. (C) Light microscopy of isolated bovine brain capillaries stained with the endothelial cellspecific lectin, Griffonia simplicifolia agglutinin. The continuous staining of the microvessel is indicative of an antigen with endothelial cell origin, whereas the two pericytes (arrowheads) are unstained. Bars in A– C are 20 µm. (D and E) Confocal images of rat brain coronal sections immunostained with anti-GFAP antibodies. The rosette-like structure of the astrocyte endfeet form a cage around the blood vessels. Scale bars in D and E are 50 µm. L=lumen. (F) Drawing depicting interactions of astrocyte endfeet and the abluminal membrane of endothelial cells. (Panels D–F: From Ref. (9). Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc). was excluded from entering the brain parenchyma of the mouse (8). This was observed throughout the brain regardless of vessel size or presence of smooth muscle cells (8). It was subsequently observed that electron dense junctional contacts were present at each contact point between endothelial cells (7). Later, the barrier to low molecular weight substrates was confirmed as a microperoxidase, with a molecular weight of only 1800 Da, was also excluded from the brains of mice (6). These

Endothelial cells in health and disease

42

Figure 2 Schematic of brain capillary cross section. A=astrocyte foot process, N=nerve terminal, P=pericyte, EC=endothelial cell, TJ=tight junction. so-called tight junctions confer a high trans endothelial electrical resistance (TEER) of 500–3000 Ω cm2 (10,11). The barrier phenotype is present throughout the entire brain vasculature except in small regions near the ventricular system, known as circumventricular organs, that are perfused by capillaries without a BBB (12) and participate in neuroendocrine regulation. In a developing rodent, the brain vasculature is fenestrated until embryonic days E11–E13 during which time tight junctions develop as fenestrations typical of peripheral capillaries disappear (13,14). This process coincides with the impermeable phenotype found in the adult brain and restricts access to substances that normally diffuse freely into a developing embryonic brain.

The endothelium as an organ

43

The tight junctions found in brain capillaries are composed of both integral membrane proteins that link adjacent cells and peripheral membrane proteins that link the junctions to the cytoskeleton. The first transmembrane tight junctional component identified was occludin (15). Occludin is an integral membrane protein that mediates cell-cell connections (Fig. 3). The cytoplasmic domain of occludin is highly phosphorylated when located in tight junctions (16) and this phosphorylation can regulate tight junctional permeability (17). Although the exact role of tight junction occludin has not been fully determined, existing data support a regulatory role rather than a barrier-establishing role (18). In contrast to occludin, another class of

Figure 3 Drawing of brain capillary cell-cell junction. Paracellular flux is not allowed by the tight junction composite of occudin, claudin, and JAM proteins. Cell-cell junctions are stabilized by cadherins at adherens

Endothelial cells in health and disease

44

junctions. Cell accessory proteins ZO, cingulin, 7H6, and AF6 are involved in coordinating the linkage of the transmembrane junctional proteins to the cytoskeleton and other cellular mediators. transmembrane proteins known as claudins appears to play the major role in forming the seal that restricts paracellular transport. There are at least 24 isoforms of claudins and this broad class confers junctional specificity to different cell types (19,20). Claudins-1 (21), 3 (22), and -5 (22–24) have been identified at the BBB. Claudin-5 has recently been shown to selectively regulate paracellular transport as evidenced by leakage of small molecule tracers, but not of larger molecules (1.9 kDa or greater) in claudin-5 knockout mice (25). Junctional adhesion molecules (JAM) are another type of transmembrane protein that localize to tight junctions (26). These proteins are members of the immunoglobulin superfamily and have the ability to increase transcellular resistance in cells not normally forming tight junctions (26,27) in addition to promoting occludin localization at intercellular boundaries (28). The JAM proteins are also involved in the extravasation of monocytes and leukocytes in vitro and in vivo (26,29,30). Endothelial cell-selective adhesion molecule (ESAM) is yet another transmembrane protein that is localized to tight junctional regions (31) and has enriched gene expression at the BBB (32). Platelet endothelial cell adhesion molecule (PECAM-1/CD31) has also been observed to localize at endothelial junctions (33,34). However, it is not entirely clear if PECAM-1 localizes to tight junctions at the BBB (35,36). Adherens junctions are also important components of the brain endothelial cell junctions (Fig. 3), but will not be discussed in detail here. In order to link the cell-cell junctional proteins to the cytoskeleton, brain endothelial cells enlist a variety of accessory proteins in the intracellular compartment. Among this subset of junctional proteins are members of the membrane-associated guanylate kinases (MAGUKs) known as zonula occludens or ZO-1, -2, and -3. These proteins are important in the anchoring of transmembrane proteins to the cytoskeleton and act as a binding scaffold for signaling proteins. ZO proteins consists of three domains, an SH3 domain which commonly binds signaling and cytoskeletal proteins (37), a PDZ domain that binds to the claudins (38), and a guanylate kinase domain that binds to occludin (39,40). Other accessory proteins include cingulin, a myosin-like protein which has been demonstrated to interact with ZO-1, ZO-2, ZO-3, myosin (41), occludin (42), and JAM (28) at tight junctions. 7H6 antigen is also located at tight junctions and confers resistance to paracellular transport of macromolecules and metastatic cancer cells (43). The formation and maintenance of the tight junctions are regulated by a variety of signaling cascades (reviewed in Ref. 44), and this allows the BBB tight junctions to respond to various stimuli and pathological conditions in a dynamic fashion (see Sec. 3).

The endothelium as an organ

45

2.2. Molecular Transport Systems at the BBB To meet the high-energy demands of the brain, the relatively impermeable BBB modulates the influx and efflux of substances by way of highly regulated transport systems. Due to the polarized nature of the BBB endothelium, a transported molecule must travel serially through both the luminal and abluminal membranes. Thus, many transporters are located on both membranes and can demonstrate asymmetry, although some transport systems are localized exclusively at either the luminal or abluminal membrane. Transporter genes are estimated to occur at a frequency of approximately 3% in the human genome (45) and based on the predicted number of transcripts in the genome, this would yield ~1000 different transporters. Given the anatomical and functional topology of the brain microvasculature, one might expect the BBB to express an inordinately large number of transporter proteins. Indeed, recent genomics efforts indicate that transporters comprise between 10% and 15% of the expressed genome of the BBB (24,32,46). In many cases, the functional relevance of these transporters remains to be confirmed. One type of transport system, carrier-mediated transport, mediates the transBBB flux of small molecule energy sources, vitamins, and nutrients (Table 1). These carriers are highly stereospecific and operate on the millisecond timescale. The two

Table 1 Nutrient Carriers at the BBB Carrier

Example substrate BBB isoform Vmax (nmol/g min)

Hexose

Glucose

GLUT1

1420

Monocarboxylic acid

Lactate

MCT1

91

Large neutral amino acid Phenylalanine

LAT1

22

Basic amino acid

Arginine

CAT1

5

Amine

Choline

Unknown

11

Small neutral amino acid Alanine

Unknown

8

Nucleoside

Adenosine

Unknown

0.75

Acidic amino acid

Glutamic acid

Unknown

0.2

Adapted from Refs. 4 and 193 carrier-mediated facilitated transporters supporting the highest rate of saturable transport are the glucose (GLUT1) (47,48) and monocarboxylic acid (MCT1) transporters (49–51). Nearly all GLUT1 expression in the brain is restricted to the microvascular endothelium (48,52) and MCT1 expression is crucial for the transport of lactic acid and ketone bodies in the developing brain (50). Several nutrient transporters that import biosynthetic building blocks have also been localized to the BBB. The large neutral amino acid transporter type I isoform (LAT1) was demonstrated to be expressed at the BBB at 100 times the level found in any other tissue including the brain (53), in addition to having a much higher affinity than neutral amino acid carriers localized to other tissues (54). The

Endothelial cells in health and disease

46

basic amino acid transporter CAT1 has also been identified at the BBB (55) and in a recent genomics analysis was confirmed to be differentially expressed in the brain compared to liver and kidney tissues (32). Transport of choline as well as other molecules having quaternary ammonium groups occurs at the BBB (56). Adenosine transport, vitamin transport, and small neutral and acidic amino acid transport has been observed at the BBB, however, the genes for the proteins regulating this transport have yet to be cloned (Table 1). Receptor-mediated transport via the endosomal route is also prevalent at the BBB interface and occurs on the minute timescale. This transcytosis mechanism, analogous to that found at epithelial barriers, allows for the targeted uptake of circulating peptides and proteins, packaging into the endocytotic pathway, and deposition into the brain interstitium. Unlike endothelium elsewhere in the body, these proteinaceous substrates cannot be transported either by the paracellular route or by pinocytosis at the BBB and thus require the activity of the receptor-mediated transcytosis systems. Several receptormediated transcytosis systems have been identified at the brain microvasculature. The insulin receptor has been shown to mediate the receptor-mediated transport of insulin at the BBB and results in measurable brain insulin concentrations (57). Transferrin is also transported across the brain endothelial cells allowing coupled entry of iron into the brain (58). This process occurs via the transferrin receptor, which is highly expressed at the brain microvasculature (59). Leptin, a protein secreted by adipocytes, induces satiety when transported by an alternatively spliced short isoform of the leptin receptor present at the BBB (60). Similar to insulin, leptin is not produced in the brain (61) and must, therefore, enter from a peripheral source. In addition, the leptin receptor was specifically localized to the microvasculature in brain tissue (62). There is also evidence for saturable transcellular transport of low-density lipoprotein at the BBB (63,64). Another very interesting class of transporters at the BBB is the carrier-mediated efflux transporter. The protein product of the multidrug resistance gene (MDR1), or pglycoprotein, functions in the ATP-dependent efflux of many small molecule substrates and therapeutics (65) and has been localized to brain capillaries (66). This transport system has the ability to target lipid soluble drugs that diffuse through the luminal membrane and shuttle them back to the bloodstream (reviewed in Ref. 67). Evidence also suggests expression of other transporters in this family known as multidrug resistanceassociated proteins (MRP1 and MRP5) at the brain endothelium although the levels and species specificity is somewhat in question at this time (67). Ion transport and astrocyteregulated aquaporin-based transport of water at the BBB are important in maintaining homeostasis of the brain environment and have been reviewed elsewhere (68–70). Although the BBB could be described as a specialized transport interface, the number of cloned and characterized transporters is limited. Given the large percentage of transporter transcripts identified in functional genomics analyses as stated above, there is a large part of the BBB transport picture that remains to be resolved.

2.3. BBB Establishment and Maintenance—Role of the Microenvironment The endothelium of the BBB in vivo has unique characteristics when compared to the endothelium found in the periphery. The mechanism of specialized differentiation into

The endothelium as an organ

47

BBB endothelial phenotype could be either genetically or environmentally based. It could be argued that the brain endothelial cell originates from a special mesodermal lineage or that these cells are genetically predisposed to becoming brain endothelial cells. On the other hand, it is possible that an endothelial cell subjected to a brain microenvironment may, by association, become a specialized BBB endothelial cell. A battery of experiments has given credence to the hypothesis that the local microenvironment is responsible for the unique differentiation of brain endothelial cells. Factors in the microenvironment that contribute to BBB function include the extracellular matrix, neighboring brain cells, and flow. These factors will be discussed in detail below. Early studies indicated that the BBB is not genetically predetermined, but rather is induced by the brain microenvironment. When neural tissue was grafted to peripheral sites, vessels invading the neural graft developed BBB properties (71,72); however, when brain vessels invaded peripheral tissue that had been grafted into the brain, no barrier characteristics were observed (72). Ultimately, direct in vivo evidence for the effects of astrocytes confirmed the barrier-inducing properties of this perivascular cell. Grafts of astrocytes can induce BBB-like properties in peripheral endothelium (73). In addition, when glial fibrillary acidic protein (GFAP) positive astrocytes are selectively ablated at the site of invasive CNS injury, the BBB remains impaired and permeable. This effect can be countered by grafting normal astrocytes to the site of the injury that foster the repair and restoration of BBB impermeability (74). Transgenic mice with GFAP-deficient astrocytes have an impaired BBB (75) and astrocytes from this animal were not able to induce a functional BBB in vitro compared to control astrocytes (76). Based on these data and the fact that GFAP is upregulated after brain injury, it has been speculated that GFAP may be an important factor for BBB induction, repair, and restoration. However, when cultured in vitro, brain endothelial cells tend to de-differentiate, lose many of their specialized characteristics and in essence resemble peripheral endothelial cells. The de-differentiation effects are extremely problematic when analyzing the permeability of pharmaceuticals in vitro or when studying basic biochemical or immunological interactions at the BBB. The de-differentiation phenomenon suggests that the BBB phenotype is critically dependent on signals residing in the local microenvironment. Indeed as will be discussed below, recent studies have demonstrated a role for perivascular cells, growth substrate, and culture conditions in eliciting BBB properties. Brain endothelial cells and pericytes share a basement membrane that is composed of collagen, laminin, fibronectin, entactin, heparin sulfate proteoglycans, and chondroitin sulfate proteoglycans (Fig. 2). This basement membrane forms a robust protective sheath around the capillaries that actually allows them to be purified intact by mechanical homogenization for subsequent dissociation and growth in primary culture or use as intact “in vivo-like” models. Perhaps not surprisingly then, brain microvascular endothelial cells have improved growth characteristics and are more representative of the in vivo situation in terms of permeability and other BBB properties when cells are plated on collagen (77–80), fibronectin (81,82), pronectin F (83), and extracellular matrix secreted by astrocytes (81). Clearly, the presence of certain physical contacts with the basement membrane as well as contact with astrocyte-derived matrix proteins are keys in the development of the in vivo phenotype.

Endothelial cells in health and disease

48

The physical situation encompassed by the extracellular matrix influences the phenotype of the brain endothelial cells. However, as the transplantation experiments suggest, soluble mediators and physical contacts with perivascular cells such as astrocytes and neurons can also play an important role in the establishment and maintenance of BBB barrier properties. The most substantial improvements for in vitro BBB models arise when the endothelial cell monolayers are cultured with the perivascular cells that help regulate the functionality of the BBB in vivo. This has been studied extensively for astrocytes whose foot processes are invested in nearly the entire surface of the BBB in vivo (Fig. 1D–F). Culturing with astrocyte conditioned medium (84), astrocyte plasma membranes (85), and astrocytes both in direct contact (86) on opposite sides of a permeable filter (78,87–89) or in indirect contact through a diffusion apparatus (84,90) are critical for the permeability, morphological, and biochemical characteristics of the in vitro models. Astrocyte influences can induce γ-glutamyl transpeptidase (γGTP) (88,89,91) and alkaline phosphatase activities (87,92), induce glucose transporter (GLUT-1) expression (93), stimulate growth and glucose utilization (83), and influence the polarity and expression of p-glycoprotein (94). In addition, astrocytes increase the TEER of endothelial cell monolayers (88,95), increase the level of peripheral vs. distributed actin (82,84), contribute to the complexity and extent of tight junction morphology in vitro (87), and decrease the passive diffusion of impermeable molecules such as sucrose, fluorescein, horse radish peroxidase and inulin (87,88,90). Interestingly, it has been observed that neuronal processes directly contact the basement membranes of endothelial cells in the brain (96,97), yet co-culture with neurons has been studied to a much lesser extent. Neurons can induce BBB-specific biochemical activity in the form of γGTP enzymatic activity, Na+−K+ ATPase activity (85) and induce endothelial cells to synthesize and sort occludin (98). In addition, endothelial cells were able to support a serotonergic phenotype in neurons co-cultured with astrocytes and endothelial cells (99). The effects that pericytes have on endothelial cell phenotype have not been thoroughly investigated although these two cell types share a common basement membrane. The nature of direct contacts between pericytes and endothelial cells is an important aspect of the microenvironment and research is needed in this area. Oftentimes, in vitro BBB models are static models, yet under physiological conditions, the brain microvasculature is subject to laminar flow conditions and associated shear stresses. Recent progress in eliciting the in vivo phenotype in vitro has been accomplished by Janigro and colleagues as they have pioneered a dynamic model of the BBB that takes into account the presence of flow-induced shear stress. When grown under dynamic flow conditions on permeable three-dimensional poly-propylene capillaries coated with pronectin F, brain microvascular endothelial cells have improved permeability and electrophysiological resistance properties when compared to static models (83,95). However, since the polypropylene capillaries are of large diameter (~150–600 µm), it is difficult to completely reproduce the in vivo situation where the capillary diameter is on the order of 5 µm. The effects of the soluble mediators released by the perivascular cells are evident, but the identities of these components are largely unknown. One protein released from astrocytes that has been definitively implicated in the development of barrier function is glial cell line-derived neurotrophic factor (GDNF) (100). Also, the use of forskolin to elevate cellular cAMP levels increases the tight junctional complexity and P-face

The endothelium as an organ

49

association of the tight junction particles (101). Direct supplementation of cAMP has also been shown to increase barrier function (84) and alkaline phosphatase expression (92). Finally, the addition of hydrocortisone to cultures in the absence of serum has been demonstrated to increase the TEER by 10-fold and decrease the permeability to sucrose to near in vivo levels (102). Although the above discussion has focused on the effects of the microenvironment on endothelial BBB phenotype, cell-cell communication is likely to occur in both directions. However, at present, little is known about the effects that endothelial cells have on neighboring astrocytes, pericytes, and neurons.

2.4. Immunological Mechanisms at the BBB Historically, the brain has oftentimes been referred to as an immune-privileged organ. While it is true that levels of immunoglobulin and leukocytes are lower in brain parenchyma compared with other tissues, the brain is subject to immune surveillance, and the immune response may be upregulated during incidences of disease and inflammation. Given its strategic location, the BBB endothelium is likely to play a central role in regulating immune surveillance and leukocyte recruitment. The exact role of the BBB in regulating these responses is still in the process of being elucidated. However, significant evidence has been gathered that suggests an important immunoregulatory role for the BBB. In the absence of barrier disruption, blood to brain antibody transport does not occur to a significant extent (103). To date, no antibody transport systems have been demonstrated at the BBB that would facilitate import of these soluble immune components. Interestingly, however, rapid export of antibodies from the parenchyma to the bloodstream has been observed after intracerebral injection of IgG (104). Although preliminary evidence indicated the presence of an Fc receptor on brain endothelial cells (105), Fc receptors, FcR-I and FcR-II, are not located at the brain microvasculature while FcR-III is only expressed in some brain venular endothelium (106). Recently the neonatal Fc receptor, FcRn, was identified at the BBB of adult rats and has been implicated in IgG export from the brain (107). Therefore, under normal or pathological conditions, it is possible that the BBB may act in clearance of IgG from the brain in an attempt to attenuate local antibody-mediated immune response. Since the BBB is the first point of contact for circulating immune cells, it may be hypothesized that brain endothelial cells function as professional antigen presenting cells (APC) thereby facilitating T cell activation. Indeed, class II MHC can be induced on the surface of primary culture brain endothelial cells by inflammatory mediators such as interferon gamma (IFN-γ) (108,109). Treatment of endothelial cells with IFN-γ (109,110) or tumor necrosis factor alpha (TNFα) (110) in vitro can also stimulate the expression of T cell co-stimulatory molecule, B7. However, T cell proliferation assays with brain endothelium from a variety of species and under various pathophysiological conditions indicated that the endothelial cells are not able to fully activate proliferative T cell responses (111,112) and even the downregulation has been observed (109). While these data would argue against an antigen-presenting role for BBB endothelial cells, other studies have demonstrated an important role for the BBB in the regulation of leukocyte extravasation into the brain. Regulation of this process occurs via the expression of cell adhesion molecules in response to proinflammatory cytokines. For example, intercellular

Endothelial cells in health and disease

50

adhesion molecule-1 (ICAM-1), which serves as a docking receptor for circulating leukocytes, is upregulated on brain endothelium by IL-1α, IFN-γ, and TNF-α, and increases lymphocyte migration across brain endothelial cells (113). In addition to performing a basic tethering function, ICAM-1 binding can trigger a signaling cascade that may contribute to phenotypic modulation of the BBB and subsequent lymphocyte entry (reviewed in Ref. 114). TNF-α is a proinflammatory cytokine that is not normally produced in the CNS but has pronounced effects on brain endothelial cells. TNF-α can be produced locally at the site of inflammation or can originate from systemic sources and has been shown to increase the permeability of an in vitro model BBB (115) as well as enhancing the adherence and transmigration of leukocytes (116). When administered in vivo, TNF-α can also increase the flux of radiolabeled albumin from the cerebrospinal fluid (CSF) to the blood. This would then presumably allow surveillance of CNS constituents by the peripheral immune machinery (117). Thus, TNF-α may serve to enhance the surveillance of the CNS by allowing more immune cells into the brain and more brain-derived antigens out. In contrast to TNF-α, transforming growth factor-β (TGF-β) is produced constitutively in the CNS and has been demonstrated to downregulate adhesion molecule expression (118) and diminish leukocyte migration across the BBB in vitro and in vivo (116). TGF-β is also a candidate for suppression of T cell proliferation in the CNS as it is a component of the CSF (119). Therefore, while TNF-α and other proinflammatory molecules appear to enhance surveillance, TGF-β may act as a delicate counterbalance to regulate the immune response in the CNS. In the process of immune surveillance, leukocytes can migrate across an intact BBB in the absence of barrier compromise. For example, transendothelial transport of neutrophils across IL-1 treated human umbilical vein endothelial cell monolayers treated with astrocyte conditioned medium (as an in vitro BBB model) can occur without losing the characteristic TEER or the endothelial ultrastructure (120). Under these conditions, the tight junctional proteins occludin, ZO-1, and ZO-2 do not degrade, but rather remain localized at cellular junctions (120,121). One possible explanation for these findings is that leukocytes migrate preferentially through tricellular corners where three endothelial cells come together and the tight junctions are inherently discontinuous, although thus far this has only been observed in vitro (121). T-lymphocytes have been shown to express occludin (122), which in turn has been found to modulate transmigration across an intact epithelial barrier. These findings raise the intriguing possibility that T cells and endothelial cells may interact via occludin links, thereby limiting breaches in the BBB upon transmigration (123). Although there is an extensive literature relating to the stepwise progression of cellular transmigration across endothelium, it should be emphasized that the exact mechanism of transmigration across the BBB is not completely understood. Under pathological conditions there can be increases in the numbers of activated neutrophils, lymphocytes, and monocytes that cross the BBB due to loss of barrier function. This is especially prevalent in multiple sclerosis (MS), stroke, HIV encephalitis, Alzheimer’s disease, and brain tumors. These effects will be discussed in greater detail below.

The endothelium as an organ

51

3. BBB INVOLVEMENT IN DISEASE 3.1. BBB Involvement in HIV Nearly two-third of patients with AIDS develop a neurological disorder such as HIV encephalitis or AIDS dementia complex (124). Progression of these diseases has been linked to migration of infected monocytes across the BBB into the brain (125,126). Evidence also supports the transendothelial passage of free HIV virus via adsorptive endocytosis (126). Regardless of the mode of entry, the incidence of HIV encephalitis and AIDS dementia correlates with the presence of monocyte infiltration and subsequent microglial activation rather than with the actual viral load (124). Vascular changes have also been noted in the brains of AIDS patients in the form of increased diameter of the cortical microvessels and thinning of the basal lamina of the blood vessels (127). In addition, HIV has been found to infect brain endothelial cells both in vivo (128,129) and in vitro (130,131). The microvascular endothelial cells of the brain are believed to play a role in the regulation of entry of the HIV virus and/or HIV infected monocytes into the parenchyma of the brain. HIV infected monocytes can stimulate the adhesion properties of the BBB through the release of cytokines and inflammatory mediators such as TNFα, IL-1β, and IFN-γ and can activate perivascular microglial cells (132–134). The activated endothelium exhibits upregulation of cell adhesion molecules E-selectin and vascular cell adhesion molecule-1, which can then support the binding and migration of HIV infected monocytes across the BBB into the brain (126,135). During the migration of these monocytes, HIV proteins stimulate Gelatinase B and promote the degradation of the basement membrane further compromising the integrity of the BBB (136). A coordinate decrease in the tight junction proteins of the BBB, occludin and ZO-1, is seen upon increased monocyte infiltration in AIDS patients (137,138). Once in the brain, the HIV virus and/or activated monocytes can initiate a cascade of detrimental responses through the release of cytokines and other toxic factors. Based on the mechanistic observations described above, the BBB appears to be intimately involved in the pathogenesis of neuroAIDS.

3.2 BBB Involvement in Stroke The integrity of the BBB is altered after stroke due to ischemic insult and hypoxia. Oxygen deprivation leads to a complicated series of events that ultimately results in increased BBB permeability, leakage of plasma components across the BBB and edema formation. Early ischemic events lead to the observed BBB dysfunction and include recruitment of leukocytes by increased expression of adhesion receptors E-selectin (139), P-selectin, and ICAM-1 (140). There is also an alteration in tight junctional content that contributes to the increase in permeability. In vitro BBB models composed of rat brain endothelial cells in co-culture with astrocytes were subjected to hypoxic conditions and a significant decrease in TEER of 25% after 4 hr of hypoxia and 95% after 8 hr was

Endothelial cells in health and disease

52

observed (141). This same set of experiments also demonstrated that the presence of astrocytes diminished the decrease in TEER after a 4 hr hypoxic insult (141) again implicating a role for this cell type in modulating BBB phenotype. Paracellular transport was also increased in bovine brain endothelial cells following long duration hypoxia (24– 48 hr) whether or not astrocytes were present (142,143). Tight junction occludin, ZO-1 and ZO-2 protein expression levels were unaltered but were relocated from cell junctions during hypoxia (143). Upon reoxygenation, a rapid increase in actin expression was observed along with increases in occludin, ZO-1, and ZO-2 protein and a coordinate decrease in the paracellular permeability. Finally, the basement membrane integrity is also impaired in ischemic conditions (144) and may be due in part to metalloproteinasebased degradation (reviewed in Ref. 145). Many of the pathological components of stroke are exacerbated by BBB opening. This important observation reinforces the need for molecular level understanding of this process so treatments that have the power to reverse BBB opening, and thus ameliorate the symptoms of stroke, can be developed.

3.3. BBB Involvement in MS and Experimental Autoimmune Encephalomyelitis MS is a chronic inflammatory and demyelinating disease of the CNS that is manifested by BBB disturbance, local edema, and demyelination (146). Damage to the BBB represents an early stage in lesion formation. Indeed, gadolinium MRI scans of MS patients have demonstrated that alterations in BBB permeability precede clinical onset of the disease (147,148). Much of our knowledge of the molecular processes underlying MS is based on an animal model for MS known as experimental autoimmune encephalomyelitis (EAE), which shares certain pathological and clinical features of the human condition. This disease can be induced by sensitization with CNS myelinassociated antigens and has been demonstrated to be T cell mediated (149). As in the pathogenesis of many other CNS diseases, ICAM-1 is upregulated at BBB endothelial sites in EAE and the levels of ICAM-1 correlate with disease activity (150). In addition to ICAM-1, the vascular cell adhesion molecule-1 (VCAM-1) was also observed to be expressed at high levels in chronic human MS lesions (151). These responses may be mediated by the release of proinflammatory cytokines (IFN-γ, TNF-α) from activated immune cells and as described earlier, the tight junctional integrity of the BBB can also be affected by cytokine release. As an example of the changes in tight junction integrity, abnormal levels and distributions of ZO-1 and occludin have been observed in brains of MS patients (152). Also in EAE models, claudin-3 expression is selectively downregulated in tight junctions compared to normal BBB (22). Such modulation of the BBB in EAE can lead to multiple modes of immune cell entry into the CNS including via the tight junction complex, via channels or pores formed in the tight junction complex, or by emperipolesis (153). Concomitant with the induction of adhesion and tight junction molecules, EAE progression correlates with an increase in vesicular trafficking that is more reminiscent of the peripheral vasculature and when the animals recover, capillary vesicular transport is reduced to near normal levels (154). Unexpectedly, recent genomics studies have identified expression of myelin basic protein at the BBB. Microvascular expression of myelin basic protein was confirmed by in situ hybridization with a myelin basic protein probe (155). Taken together, these findings

The endothelium as an organ

53

raise the possibility that the BBB may play an even more important role in MS pathogenesis that was previously anticipated. For example, in addition to regulating the transendothelial migration process, the BBB may contribute to recruitment of T cells.

3.4. BBB Modulation in Brain Tumors In general, the vascular environment of low-grade gliomas is very similar to that found in normal brain (156). However, high-grade gliomas and metastatic brain tumors have significantly altered angio-architecture and permeability characteristics. Whereas normal brain capillaries have diameters around 5 µm, tumor capillaries are on average much larger (up to 40 (µm) and in contrast to normal brain capillaries, the larger tumor vessels are very tortuous and irregularly spaced (157–159). Microvascular permeability and loss of BBB function can be attributed to changes in the tight junctions, presence of fenestrae, and increases in vesicular trafficking. It has been well established that the tight junctions in brain tumors are abnormal when compared to the intact BBB. The overlap between adjoining endothelial cells is decreased or the tight junctions are substantially widened, suggesting increases in paracellular transport (160,161). Along these lines, the molecular composition of tight junctions in high-grade tumors is altered implying a role for tight junction regulation in tumor malignancy. Studies of tight junctions in human brain tumors determined that claudin-1 expression was lost, while claudin-5 and occludin expression were downregulated in hyperplastic vessels only, and ZO-1 expression was unaffected (162,163). Fenestrations are readily identified in brain metastases but are an extremely rare occurrence in glial tumors. Vesicular trafficking characteristics of brain tumor endothelium are much less clear and have been described as increased, decreased, or unchanged. The net result on the integrity of the blood-tumor barrier can be demonstrated by studies that confirmed a higher level of leakage of serum components from tumor vessels. Fluoresceinated molecules ranging in size from 376 to 500,000 Da leaked through the blood-tumor barrier (158) as did bovine serum albumin (164). However, it is important to note that a higher level of leakage does not imply absence of a barrier as it has been demonstrated that size-dependent leakage kinetics are operating at the compromised BBB (158,165). Also in some tumors, fluorescein leakage (376 Da) does not occur and this is indicative of a tight blood-tumor barrier (164). Vascular endothelial growth factor (VEGF), initially identified as vascular permeability factor, is a likely candidate for the regulation of tumor vessel permeability. In high-grade gliomas, VEGF is greatly upregulated and its receptors VEGFR-1 and VEGFR-2 are coordinately upregulated on tumor endothelium (166–168). Normal astrocytes assist in barrier formation; while in contrast, high-grade astrocytoma cells secrete VEGF that stimulates angiogenesis, vascular permeability, and redistribution of tight junction occludin expression (166,169). Glioma vessels continue to express the BBB-specific glucose transporter, GLUT-1 (170), and p-glycoprotein expression likely limits the efficacy of many anticancer agents (171) as would also be expected from normal brain endothelium. Finally, the microenvironment of brain tumors is critical to the ultimate phenotype of the tumor endothelium. As described above, metastatic brain tumors have presence of fenestrae and lack BBB properties. This perhaps is not surprising given the importance of perivascular cell influence on BBB properties, and the lack of these inductive factors in

Endothelial cells in health and disease

54

tumor cells arising from peripheral tissues. Another example is given by the transplantation of C6 rat astroglioma tumors into brain and muscle. The resulting tumor vasculature is identical in both sites (172) again suggesting that the brain cells provide inductive factors that determine the ultimate vessel phenotype. Although speculative, this may also explain the fact that low-grade gliomas have operational BBB while more transformed cells comprising a high-grade gliomas may lack the necessary inductive machinery to yield a working BBB. The molecular level understanding of brain tumor vessel diversity is crucial in the proper design of brain tumor drugs as efficacy could vary widely depending on local permeability characteristics.

3.5. BBB Involvement in Other CNS Diseases As documented above, the brain microvasculature participates heavily in the progression of a variety of CNS diseases. This list was by no means exhaustive, as the BBB may also have important roles in Alzheimer’s disease (173), hypertension (174), and the class of CNS complications due to infections by bacterial, fungal, and viral pathogens (175) among others.

4. GENOMICS AND PROTEOMICS OF THE BBB One of the major difficulties in discerning all of the important contributions to the BBB phenotype in health and disease is that many of the cells and soluble influences act synergistically in a dynamic fashion. A single soluble factor or cell-cell contact may influence one or more pathways in multiple cell types. In addition, the temporal and spatial relationships of disease progression are poorly understood, yet they are crucial for the development of appropriately targeted therapies. Genomics and proteomics analyses provide an opportunity to perform system-wide discrimination of healthy and diseased tissues in order to deconvolute complicated mechanistic networks that are involved in angiogenesis, aberrant proliferation, and altered cellular behavior. The information obtained by these approaches may facilitate the rapid identification of therapeutic targets and the discovery of novel BBB-specific transport systems having utility in non-invasive drug delivery.

4.1. Genomics of the BBB Oftentimes studies with the BBB are constrained by examining the expression and behavior of just a few genes or proteins of interest. This includes analyses of BBB disease involvement, validation of in vitro models, and identification of molecular determinants of in vivo phenotype. However, based on many of the phenotypic attributes of the BBB described throughout this chapter, it is apparent that a few genes do not adequately describe the underlying phenomena. The regulation of tight junctions, transcellular transport, cytoskeletal rearrangement, and immune responses unequivocally relies on intertwined networks of proteins derived from multiple cellular sources.

The endothelium as an organ

55

Recently, BBB researchers have applied a variety of genomic techniques to the analysis of BBB function in health and disease. A major difficulty in performing genomics techniques using mRNA derived from total brain homogenates is that only the most highly expressed microvasculature transcripts would be identified above the background produced by non-endothelial cells. This is due to the fact that the brain blood vessels themselves constitute only about 1/1000 of the total brain volume (176) and coordinately a total brain mRNA sample would fractionally contain only a small amount of capillary mRNA. At the same time, microarray sensitivity can be in the neighborhood of 10−4 (177). Thus, a much more successful BBB profiling approach would consist of capillary purification (Fig. 1A) coupled with isolation of mRNA directly from this subset of cells. As mentioned earlier, this task can be readily accomplished using mechanical homogenization techniques due to the robust capillary basement membrane. Importantly when capillary isolation is performed within short times of sacrifice, this technique allows for recovery of in vivo-like mRNA samples with little change in the in vivo mRNA content (178). Armed with capillary-specific mRNA, genomics techniques such as gene microarray, subtractive suppression hybridization (SSH) (179), and serial analysis of gene expression (SAGE) (180) have been used to address questions regarding the functional attributes of the BBB in vitro and in vivo. Two main genomic strategies have been applied to the BBB in order to determine the gene expression profiles that contribute to its unique functional phenotype. The first methodology discussed takes advantage of the SSH technology that identifies differentially expressed BBB genes compared with peripheral tissue and the second is a comprehensive compilation (SAGE) of all expressed BBB transcripts (transcriptome). In order to determine the molecular origin of unique BBB functions, SSH has been applied to the analysis of both human and rat capillaries in a concurrent BBB genomics program (24,32,155). SSH is an excellent technique for determining genes that are differentially expressed among mRNA samples from different tissues (179). It is also a valuable technique for the analysis of low-abundance transcripts that may be present at the BBB and have important roles in regulation of BBB phenotype (179). In this study, genes common to the brain microvasculature, kidney tissue, and liver tissue were subtracted while identifying those genes that were by comparison upregulated at the BBB. Results from three rounds of SSH differential analysis [one round of human (24) and two rounds of rat (32,155)] indicate that between 51% and 55% of the differentially expressed genes at the BBB encode proteins with known function. Figure 4 is a compilation of these three studies with the genes encoding known proteins clustered by function. Of course, clustering into distinct categories is somewhat arbitrary given that many of these gene products have multiple functional roles in vivo. What immediately stands out is the high number of enriched genes that encode proteins involved in the control of angiogenesis, signaling, transport, and junctional structure. Importantly, the methodology proved to be effective in identifying genes previously implicated in mediating BBB phenotype. The balance of the genes identified in this study (45–49%) encode BBB-enriched proteins with unknown function. The elucidation of the functions of these proteins, though fraught with difficulty, will be critical to a full understanding of the BBB. Another genomics study utilized the SAGE method and resulted in a comprehensive analysis of gene expression at the rat brain microvasculature (46). The advantage of

Endothelial cells in health and disease

56

SAGE analysis compared with SSH or even cDNA microarray is that one can generate a

Figure 4 Functional clustering of BBB-enriched genes determined in a concerted genomics /proteomics program of the brain microvasculature. Transcripts identified in given genomics/proteomics studies are identified by: rat genomics study I denoted by * (155), rat genomics study II denoted by † (32), human genomics study denoted by ‡ (24), and bovine proteomic study denoted by # (190,191). PDGF-Rβ, platelet-derived growth factor receptor β subunit; IGF2, insulin-like growth factor 2; FGF19, fibroblast growth factor 19; HARP,

The endothelium as an organ

57

heparin affinity regulatory peptide; IGF-BP-3, insulin-like growth factor binding protein 3; Gab2, Grb-2associated binder-2; LaAUF-1, AUrich RNA binding factor; Rgs5, Gprotein signaling regulator-5; Ptdgs, prostaglandin D synthase; VESP14, vascular endothelial cell-specific protein 14; hbrm, human homolog of yeast SW12 transcription factor; PC3, B-cell translocation gene-2; oatp2, organic anion transporting peptide type 2; MCT1, monocarboxylate transporter 1; BSAT-1, BBB-specific anion transporter type 1; CAT1, cationic amino acid transporter 1; FXYD5, FXYD domain-containing ion transport regulator 5; TfR, transferrin receptor; Cpe, carboxypeptidase E; Pgsg, secretory granule proteoglycan core protein precursor; APLP2, amyloid precursor-like protein 2; YWK-II, sperm membrane protein related to A4 amyloid protein; Itm2a, integral membrane protein 2a; Spi4, serine protease inhibitor 4; tPA, tissue plasminogen activator; MBP, myelin basic protein; PZR related, protein zero-related protein 1; PLP-1, proteolipid protein; PLTP, phospholipid transfer protein; Scd2, stearoyl-CoA desaturase 2; Flt-1, vascular endothelial growth factor receptor type 1; HIF-2α, hypoxia inducible factor 2α, VE-PTP, vascular endothelial receptor-type protein tyrosine phosphatase; MLC20, regulatory myosin light chain isoform C; ESAM, endothelial cell-selective

Endothelial cells in health and disease

58

adhesion molecule; Ro52, 52 kDa ribonucleoprotein; PECAM-1 platelet/endothelial cell adhesion protein. complete blueprint of the cellular transcriptome [estimated to encode around 30−40,000 proteins in humans (45,181)], and not simply a subset of genes that may be differentially expressed or present on a fabricated microarray chip. The SAGE analysis yields both the identity and the relative quantity of all expressed genes. One current drawback is that such a readout in the absence of SAGE analyses from other tissues limits the ability to assess what genes are differentially expressed at the BBB. However, SAGE data are being generated for different tissue components at a rapid pace and will allow for interesting molecular level comparisons. The SAGE analysis of the BBB identified the presence of nearly 11,000 transcripts, of which only 17% matched genes with known functions (46). When compared with SAGE libraries generated from cortex and hippocampus, BBB-enriched genes were identified and clustered into groups of transporters (11%), receptors (10%), vesicle trafficking (7%), structural proteins (12%), and signal transduction (18%). The distribution of enriched genes agrees quite well with that determined in the SSH genomics program emphasizing the functioning BBB as a “molecular switchboard” between the blood and brain. In addition to defining the molecular functionality of the BBB under normal conditions, both the SSH and SAGE analyses described above could be applied to compare and contrast molecular level details in pathological states. Genomics approaches have also been used to address several specific questions regarding the characteristics of the brain microvasculature. Lippoldt and co-workers (182) used SSH in order to identify potential contributions to stroke in hypertensive rats. They compared the gene expression profiles of cerebral capillaries from stroke-prone spontaneously hypertensive rats to stroke-resistant spontaneously hypertensive rats and identified several genes that were either upregulated or downregulated in capillaries from stroke-prone rats, including two encoded proteins with unknown function. The rat sulfonylurea receptor 2B was upregulated, while the G-protein signaling 5 regulator was downregulated in stroke-prone rats. In another study, vessels from epileptic tissue were compared to non-epileptic blood vessels by cDNA microarray to determine the classes of genes responsible for resistance to antiepileptic drugs. It was discovered that MDR1, MRP1, MRP2, MRP5, and cisplatin resistance-associated protein were all overexpressed in epileptic vessels helping to explain the origin of drug resistance (183). The effects of flow-induced shear stress on in vitro brain endothelial cultures were determined by cDNA microarray that compared static and dynamic in vitro models. This study concluded that flow induces cytoskeletal genes and contributes to the development of the antioxidant capacity of endothelial cells (184). Finally, the in vitro phenotype of human brain endothelial cells (HBEC) was compared to HUVEC cells to help identify genes crucial in conferring the BBB phenotype. Thirty-five genes were preferentially expressed in HBEC and included vasculogenic factors, immunoregulators, and growth factors (185).

The endothelium as an organ

59

4.2. Proteomics of the BBB Although genomics approaches yield a wealth of molecular level information about what factors ultimately confer phenotype, it is important not to use these techniques exclusively. For one, the correlation between gene expression levels and protein expression levels is not necessary linear (186). Also, post-translational modifications including phosphorylation, glycosylation, and proteolysis that are critical to protein function are not addressed in genomics techniques. Thus, it is important to also understand the protein makeup of the BBB and generate proteomic profiles that can readily complement genomic profiles. The use of 2-dimensional gel electrophoresis in BBB research has been remarkably low to date. A recent study used 2-dimensional gel electrophoresis to analyze protein leakage from cells upon BBB damage. Damage was induced in vitro by physically distancing the astrocytes from endothelial cells in coculture. This condition upregulated the release of the protease inhibitor α2-macroglobulin from astrocytes and it was suggested this could be a protective mechanism to lessen BBB damage (187). In addition, 2-dimensional gel electrophoresis combined with mass spectrometry was used to study the protein expression profiles of rat brain endothelial cells during ischemic conditions (188). Total cellular protein was greatly reduced upon ischemia/reperfusion and around 30 different proteins were found to be differentially regulated. Many of the proteins that were upregulated were involved in the proteasome pathway whereas there was a coincident decrease in ribosomal proteins. These two factors were suggested as rationale for the observed decrease in total protein content. Enzymes involved in protection against reactive oxygen and nitrogen species as well as transcription factors involved in inflammation were also upregulated. Many of these alterations in protein expression were also confirmed by microarray experiments performed in parallel. Although 2-dimensional electrophoresis coupled with mass spectrometry is the current gold standard of proteomic technologies, it is difficult to analyze membrane proteins with this method due to solubility constraints (189). As described throughout this chapter, many of the unique features of the BBB can be ascribed to membrane proteins involved in signaling, junctions, adhesion, and transport. A novel proteomics methodology was developed to specifically analyze differential membrane protein expression at the BBB (190). This method, unlike gel electrophoresis, relies on probing protein expression in a native mammalian membrane environment so solubility is not an issue. In this method, a polyclonal antiserum raised against all bovine brain microvessel endothelial proteins is depleted with protein preparations derived from kidney and liver tissue. This depleted antiserum bereft of antibodies recognizing proteins common in other tissues was used as a probe for BBB-specific membrane proteins. The COS-1 cells transfected with a bovine BBB cDNA library were probed with the BBB-specific antiserum and those cells that expressed a BBB-specific protein were recovered and the identity of the protein was determined. The methodology was validated by identification of three membrane-bound proteins having enriched expression at the BBB. These included: Lutheran membrane glycoprotein (190), a basal cell adhesion molecule; the membrane cofactor protein CD46 (191), a complement regulator also found in astrocytes; and Ro52, an autoantigen implicated in Sjogren’s syndrome (192). The polyclonal antiserum also recognizes a host of other BBB-enriched proteins that will likely be identified in the future (190).

Endothelial cells in health and disease

60

The use of genomics and proteomics for the study of molecular level phenomena at the BBB is still in the early stages. However, as the preliminary studies above indicate, these tools should be integrated into existing BBB research programs. They clearly have the potential to add tremendous understanding to a variety of physiological processes that occur at the BBB in health and disease and will likely unearth additional clues on how to appropriately target therapeutic intervention to the brain.

5. CONCLUSION Throughout this chapter, I have attempted to demonstrate that the BBB has many specialized properties when compared to endothelium from other vascular beds. However, many questions remain to be answered. One particularly important goal is to determine whether there are BBB variations within the brain vascular bed itself. Many of the experiments described above focus on the vasculature-rich gray matter of the cerebral cortex. However, it is interesting to speculate about the potential diversity of BBB attributes in other brain regions such as the cerebellum and brain stem, or perhaps about local BBB variation within different sub-cortical domains. For example, are there BBB domains that exhibit locally unique function, or are there regions that could be specifically targeted in drug treatments? With advances in laser capture microdissection methods that allow for extraction of single blood vessels, proteomic and genomic analyses will likely be extended to address such questions on a molecular level. Although the BBB was referred to as a unique vascular bed throughout this chapter, it is important to acknowledge that other endothelial barriers exist in the body, and in general, these are much more poorly understood. The blood-retinal barrier of the vascularized inner retina of the eye has many properties similar to that of the BBB. A peripheral blood-nerve barrier exists although it is substantially more permeable than the BBB. The blood-testes barrier is endowed with impermeability characteristics by the Sertoli cells that surround a comparatively permeable endothelium. Finally, a blood-CSF barrier in the brain is clearly distinct from the BBB in its barrier and transport characteristics, and it is comprised of tightly apposed epithelial cells rather than endothelial cells. Research regarding blood barriers continues to progress, and development of drug targeting and delivery strategies that overcome these barriers is imperative for the effective treatment of disease.

ACKNOWLEDGMENTS I am grateful to colleagues Dr. Lester R.Drewes and Dr. Zsuzsanna Fabry for their critical review of this chapter. This chapter was supported in part by the National Institutes of Health (1R21-AA13834).

The endothelium as an organ

61

REFERENCES 1. Pardridge WM. Brain drug delivery and blood-brain barrier transport. Drug Delivery 1993; 1:83– 101. 2. Wiederhold KH, Bielser W Jr, Schulz U, Veteau MJ, Hunziker O. Three-dimensional reconstruction of brain capillaries from frozen serial sections. Microvasc Res 1976; 11(2):175– 180. 3. Bar T. The vascular system of the cerebral cortex. Adv Anat Embryol Cell Biol 1980; 59:I–VI, 1–62. 4. Drewes LR. Molecular architecture of the brain microvasculature: perspective on blood-brain transport. J Mol Neurosci, 2001; 16(2–3):93–98 (Discussion 151–157). 5. Bickel U, Kang YS, Yoshikawa T, Pardridge WM. In vivo demonstration of subcellular localization of anti-transferrin receptor monoclonal antibody-colloidal gold conjugate in brain capillary endothelium. J Histochem Cytochem 1994; 42(11):1493–1497 . 6. Reese TS, Feder N, Brightman MW. Electron microscopic study of the blood-brain and bloodcerebrospinal fluid barriers with microperoxidase. J Neuropathol Exp Neurol 1971; 30(1):137– 138. 7. Brightman MW, Reese TS. Junctions between intimately apposed cell membranes in the vertebrate brain. J Cell Biol 1969; 40(3):648–677. 8. Reese TS, Karnovsky MJ. Fine structural localization of a blood-brain barrier to exogenous peroxidase. J Cell Biol 1967; 34(1):207–217. 9. Kacem K, Lacombe P, Seylaz J, Bonvento G. Structural organization of the perivascular astrocyte endfeet and their relationship with the endothelial glucose transporter: a confocal microscopy study. Glia 1998; 23(1):1–10. 10. Butt AM, Jones HC, Abbott NJ. Electrical resistance across the blood-brain barrier in anaesthetized rats: a developmental study. J Physiol 1990; 429:47–62. 11. Crone C, Olesen SP. Electrical resistance of brain microvascular endothelium. Brain Res 1982; 241(1):49–55. 12. Brightman MW, Reese TS, Feder N. Assessment with the electron microscope of the permeability to peroxidase of cerebral endothelium and epithelium in mice and sharks. In: Crone C, Lassen NA, eds. Capillary Permeability. Copenhagen Munksgaard, 1970:463. 13. Stewart PA, Hayakawa K. Early ultrastructural changes in blood-brain barrier vessels of the rat embryo. Brain Res Dev Brain Res 1994; 78(1):25–34. 14. Bauer HC, Bauer H, Lametschwandtner A, Amberger A, Ruiz P, Steiner M. Neovascularization and the appearance of morphological characteristics of the blood-brain barrier in the embryonic mouse central nervous system. Brain Res Dev Brain Res 1993; 75(2):269–278. 15. Furuse M, Hirase T, Itoh M, Nagafuchi A, Yonemura S, Tsukita S.Occludin: a novel integral membrane protein localizing at tight junctions. J Cell Biol 1993; 123(6 Pt 2):1777–1788. 16. Sakakibara A, Furuse M, Saitou M, Ando-Akatsuka, Y, Tsukita, S. Possible involvement of phosphorylation of occludin in tight junction formation. J Cell Biol 1997; 137(6):1393–1401. 17. Hirase T, Kawashima S, Wong EY, Ueyama T, Rikitake Y, Tsukita S, Yokoyama M, Staddon JM. Regulation of tight junction permeability and occludin phosphorylation by Rhoap160ROCK-dependent and -independent mechanisms. J Biol Chem 2001; 276(13):10423– 10431. 18. Wolburg H, Lippoldt A. Tight junctions of the blood-brain barrier: development, composition and regulation. Vascul Pharmacol 2002; 38(6):323–337. 19. Furuse M, Sasaki H, Tsukita S. Manner of interaction of heterogeneous claudin species within and between tight junction strands. J Cell Biol 1999; 147(4):891–903.

Endothelial cells in health and disease

62

20. Tsukita S, Furuse M. Claudin-based barrier in simple and stratified cellular sheets. Curr Opin Cell Biol 2002; 14(5):531–536. 21. Huber JD, Witt KA, Hom S, Egleton RD, Mark KS, Davis TP. Inflammatory pain alters bloodbrain barrier permeability and tight junctional protein expression. Am J Physiol Heart Circ Physiol 2001; 280(3):H1241–H1248. 22. Wolburg H, Wolburg-Buchholz K, Kraus J, Rascher-Eggstein G, Liebner S, Hamm S, Duffner F, Grote EH, Risau W, Engelhardt B. Localization of claudin-3 in tight junctions of the bloodbrain barrier is selectively lost during experimental autoimmune encephalomyelitis and human glioblastoma multiforme. Acta Neuropathol (Berl) 2003; 105(6):586–592. 23. Morita K, Sasaki H, Furuse M, Tsukita S. Endothelial claudin: claudin-5/TMVCF constitutes tight junction strands in endothelial cells. J Cell Biol 1999; 147(1):185–194. 24. Shusta EV, Boado RJ, Mathern GW, Pardridge WM. Vascular genomics of the human brain. J Cereb Blood Flow Metab 2002; 22(3):245–252. 25. Nitta T, Hata M, Gotoh S, Seo Y, Sasaki H, Hashimoto N, Furuse M, Tsukita S. Size-selective loosening of the blood-brain barrier in claudin-5-deficient mice. J Cell Biol 2003; 161(3):653– 660. 26. Martin-Padura I, Lostaglio S, Schneemann M, Williams L, Romano M, Fruscella P, Panzeri C, Stoppacciaro A, Ruco L, Villa A, Simmons D, Dejana E. Junctional adhesion molecule, a novel member of the immunoglobulin superfamily that distributes at intercellular junctions and modulates monocyte transmigration. J Cell Biol 1998; 142(1):117–127. 27. Aurrand-Lions M, Duncan L, Ballestrem C, Imhof BA. JAM-2, a novel immunoglobulin superfamily molecule, expressed by endothelial and lymphatic cells. J Biol Chem 2001; 276(4):2733–2741. 28. Bazzoni G, Martinez-Estrada OM, Orsenigo F, Cordenonsi M, Citi S, Dejana E. Interaction of junctional adhesion molecule with the tight junction components ZO-1, cingulin, and occludin. J Biol Chem 2000; 275(27):20520–20526. 29. Del Maschio A, De Luigi A, Martin-Padura I, Brockhaus M, Bartfai T, Fruscella P, Adorini L, Martino G, Furlan R, De Simoni MG, Dejana E. Leukocyte recruitment in the cerebrospinal fluid of mice with experimental meningitis is inhibited by an antibody to junctional adhesion molecule (JAM). J Exp Med 1999; 190(9):1351–1356. 30. Palmeri D, van Zante A, Huang CC, Hemmerich S, Rosen SD. Vascular endothelial junctionassociated molecule, a novel member of the immunoglobulin superfamily, is localized to intercellular boundaries of endothelial cells. J Biol Chem 2000; 275(25):19139–19145. 31. Nasdala I, Wolburg-Buchholz K, Wolburg H, Kuhn A, Ebnet K, Brachtendorf G, Samulowitz U, Kuster B, Engelhardt B, Vestweber D, Butz S. A transmembrane tight junction protein selectively expressed on endothelial cells and platelets. J Biol Chem 2002; 277(18):16294– 16303. 32. Li JY, Boado RJ, Pardridge, WM. Rat blood-brain barrier genomics II. J Cereb Blood Flow Metab 2002; 22(11):1319–1326. 33. Ayalon O, Sabanai H, Lampugnani MG, Dejana E, Geiger B. Spatial and temporal relationships between cadherins and PECAM-1 in cell-cell junctions of human endothelial cells. J Cell Biol 1994; 126(1):247–258. 34. Albelda SM, Muller WA, Buck CA, Newman PJ. Molecular and cellular properties of PECAM1 (endoCAM/CD31): a novel vascular cell-cell adhesion molecule. J Cell Biol 1991; 114(5):1059–1068. 35. Williams KC, Zhao RW, Ueno K, Hickey WF. PECAM-1 (CD31) expression in the central nervous system and its role in experimental allergic encephalomyelitis in the rat. J Neurosci Res 1996; 45(6):747–757. 36. Lossinsky AS, Wisniewski HM. Immunoultrastructural expression of ICAM-1 and PECAM-1 occurs prior to structural maturity of the murine blood-brain barrier. Dev Neurosci 1998; 20(6):518–524.

The endothelium as an organ

63

37. Itoh M, Nagafuchi A, Yonemura S, Kitani-Yasuda T, Tsukita S. The 220-kD protein colocalizing with cadherins in non-epithelial cells is identical to ZO-1, a tight junctionassociated protein in epithelial cells: cDNA cloning and immunoelectron microscopy. J Cell Biol 1993; 121(3):491–502. 38. Itoh M, Furuse M, Morita K, Kubota K, Saitou M, Tsukita S. Direct binding of three tight junction-associated MAGUKs, ZO-1, ZO-2, and ZO-3, with the COOH termini of claudins. J Cell Biol 1999; 147(6):1351–1363. 39. Furuse M, Itoh M, Hirase T, Nagafuchi A, Yonemura S, Tsukita S. Direct association of occludin with ZO-1 and its possible involvement in the localization of occludin at tight junctions. J Cell Biol 1994; 127(6 Pt 1):1617–1626. 40. Fanning AS, Jameson BJ, Jesaitis LA, Anderson JM. The tight junction protein ZO-1 establishes a link between the transmembrane protein occludin and the actin cytoskeleton. J Biol Chem 1998; 273(45):29745–29753. 41. Cordenonsi M, D’Atri F, Hammar E, Parry DA, Kendrick-Jones J, Shore D, Citi S. Cingulin contains globular and coiled-coil domains and interacts with ZO-1, ZO-2, ZO-3, and myosin. J Cell Biol 1999; 147(7):1569–1582. 42. Cordenonsi M, Turco F, D’Atri F, Hammar E, Martinucci G, Meggio F, Citi S. Xenopus laevis occludin. Identification of in vitro phosphorylation sites by protein kinase CK2 and association with cingulin. Eur J Biochem 1999; 264(2):374–384. 43. Satoh H, Zhong Y, Isomura H, Saitoh M, Enomoto K, Sawada N, Mori M. Localization of 7H6 tight junction-associated antigen along the cell border of vascular endothelial cells correlates with paracellular barrier function against ions, large molecules, and cancer cells. Exp Cell Res 1996; 222(2):269–274. 44. Matter K, Balda MS. Signalling to and from tight junctions. Nat Rev Mol Cell Biol 2003; 4(3):225–236. 45. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W, Funke R, Gage D, Harris K, Heaford A, Howland, J, Kann L, Lehoczky J, LeVine R, McEwan P, McKernan K, Meldrim J, Mesirov JP, Miranda C, Morris W, Naylor J, Raymond C, Rosetti M, Santos R, Sheridan A, Sougnez C, Stange-Thomann N, Stojanovic N, Subramanian A, Wyman D, Rogers J, Sulston J, Ainscough R, Beck S, Bentley D, Burton J, Clee C, Carter N, Coulson A, Deadman R, Deloukas P, Dunham A, Dunham I, Durbin R, French L, Grafham D, Gregory S, Hubbard T, Humphray S, Hunt A, Jones M, Lloyd C, McMurray A, Matthews L, Mercer S, Milne S, Mullikin JC, Mungall A, Plumb R, Ross M, Shownkeen R, Sims S, Waterston RH, Wilson RK, Hillier LW, McPherson JD, Marra MA, Mardis ER, Fulton LA, Chinwalla AT, Pepin KH, Gish WR, Chissoe SL, Wendl MC, Delehaunty KD, Miner TL, Delehaunty A, Kramer JB, Cook LL, Fulton RS, Johnson DL, Minx PJ, Clifton SW, Hawkins T, Branscomb E, Predki P, Richardson P, Wenning S, Slezak T, Doggett N, Cheng JF, Olsen A, Lucas S, Elkin C, Uberbacher E, Frazier M, Gibbs RA, Muzny DM, Scherer SE, Bouck JB, Sodergren EJ, Worley KC, Rives CM, Gorrel, JH, Metzker ML, Naylor SL, Kucherlapati RS, Nelson DL, Weinstock GM, Sakaki Y, Fujiyama A, Hattori M, Yada T, Toyoda A, Itoh T, Kawagoe C, Watanabe H, Totoki Y, Taylor T, Weissenbach J, Heilig R, Saurin W, Artiguenave F, Brottier P, Bruls T, Pelletier E, Robert C, Wincker P, Smith DR, Doucette-Stamm L, Rubenfield M, Weinstock K, Lee HM, Dubois J, Rosenthal A, Platzer M, Nyakatura G, Taudien S, Rump A, Yang H, Yu J, Wang J, Huang G, Gu J, Hood L, Rowen L, Madan A, Qin S, Davis RW, Federspiel NA, Abola AP, Proctor MJ, Myers RM, Schmutz J, Dickson M, Grimwood J, Cox DR, Olson MV, Kaul R, Shimizu N, Kawasaki K, Minoshima S, Evans GA, Athanasiou M, Schultz R, Roe BA, Chen F, Pan H, Ramser J, Lehrach H, Reinhardt R, McCombie WR, de la Bastide M, Dedhia N, Blocker H, Hornischer K, Nordsiek G, Agarwala R, Aravind L, Bailey JA, Bateman A, Batzoglou S, Birney E, Bork P, Brown DG, Burge CB, Cerutti L, Chen HC, Church D, Clamp M, Copley RR, Doerks T, Eddy SR, Eichler EE, Furey TS, Galagan J, Gilbert JG, Harmon C, Hayashizaki Y, Haussler D, Hermjakob Hs, Hokamp K, Jang W, Johnson LS, Jones TA, Kasif S, Kaspryzk A, Kennedy S, Kent WJ, Kitts P,

Endothelial cells in health and disease

64

Koonin EV, Korf I, Kulp D, Lancet D, Lowe TM, McLysaght, A, Mikkelsen T, Moran JV, Mulder N, Pollara VJ, Ponting CP, Schuler G, Schultz J, Slater G, Smit AF, Stupka E, Szustakowski J, Thierry-Mieg D, Thierry-Mieg J, Wagner L, Wallis J, Wheeler R, Williams A, Wolf YI, Wolfe KH, Yang SP, Yeh RF, Collins F, Guyer MS, Peterson J, Felsenfeld A, Wetterstrand KA, Patrinos A, Morgan MJ, Szustakowki J, de Jong P, Catanese JJ, Osoegawa K, Shizuya H, Choi S, Chen YJ. Initial sequencing and analysis of the human genome. Nature 2001; 409(6822):860–921. 46. Enerson BE, Drewes LR. The rat blood-brain barrier transcriptome. Brain03, XXIst International Symposium on Cerebral Blood Flow, Metabolism, and Function, 2003. 47. Birnbaum MJ, Haspel HC, Rosen OM. Cloning and characterization of a cDNA encoding the rat brain glucose-transporter protein. Proc Natl Acad Sci USA 1986; 83(16):5784–5788. 48. Pardridge WM, Boado RJ, Farrell CR. Brain-type glucose transporter (GLUT-1) is selectively localized to the blood-brain barrier. Studies with quantitative western blotting and in situ hybridization. J Biol Chem 1990; 265(29):18035–18040. 49. Takanaga H, Tamai I, Inaba S, Sai Y, Higashida H, Yamamoto H, Tsuji A. cDNA cloning and functional characterization of rat intestinal monocarboxylate transporter. Biochem Biophys Res Commun 1995; 217(1):370–377. 50. Leino RL, Gerhart DZ, Drewes LR. Monocarboxylate transporter (MCT1) abundance in brains of suckling and adult rats: a quantitative electron microscopic immunogold study. Brain Res Dev Brain Res 1999; 113(1–2):47–54. 51. Gerhart DZ, Enerson BE, Zhdankina OY, Leino RL, Drewes LR. Expression of monocarboxylate transporter MCT1 by brain endothelium and glia in adult and suckling rats. Am J Physiol 1997; 273(1 Pt 1):E207–E213. 52. Boado RJ, Pardridge WM. The brain-type glucose transporter mRNA is specifically expressed at the blood-brain barrier. Biochem Biophys Res Commun 1990; 166(1):174–179. 53. Boado RJ, Li JY, Nagaya M, Zhang C, Pardridge WM. Selective expression of the large neutral amino acid transporter at the blood-brain barrier. Proc Natl Acad Sci USA 1999; 96(21):12079– 12084. 54. Pardridge WM. Kinetics of competitive inhibition of neutral amino acid transport across the blood-brain barrier. J Neurochem 1977; 28(1):103–108 . 55. Stoll J, Wadhwani KC, Smith QR. Identification of the cationic amino acid transporter (System y+) of the rat blood-brain barrier. J Neurochem 1993; 60(5): 1956–1959. 56. Cornford EM, Braun LD, Oldendorf WH. Carrier mediated blood-brain barrier transport of choline and certain choline analogs. J Neurochem 1978; 30(2):299−308. 57. Pardridge WM, Eisenberg J, Yang J. Human blood-brain barrier insulin receptor. J Neurochem 1985; 44(6):1771–1778. 58. Fishman JB, Rubin JB, Handrahan JV, Connor JR, Fine RE. Receptor-mediated transcytosis of transferrin across the blood-brain barrier. J Neurosci Res 1987; 18(2):299–304. 59. Jefferies WA, Brandon MR, Hunt SV, Williams AF, Gatter KC, Mason DY. Transferrin receptor on endothelium of brain capillaries. Nature 1984; 312(5990):162–163. 60. Boado RJ, Golden PL, Levin N, Pardridge WM. Up-regulation of blood-brain barrier shortform leptin receptor gene products in rats fed a high fat diet. J Neurochem 1998; 71(4):1761– 1764. 61. Masuzaki H, Ogawa Y, Isse N, Satoh N, Okazaki T, Shigemoto M, Mori K, Tamura N, Hosoda K, Yoshimasa Y, Jingami H, Kawada K, Nakao K. Human obese gene expression. Adipocytespecific expression and regional differences in the adipose tissue. Diabetes 1995; 44(7):855– 858. 62. Golden PL, Maccagnan TJ, Pardridge WM. Human blood-brain barrier leptin receptor. Binding and endocytosis in isolated human brain microvessels. J Clin Invest 1997; 99(1):14–18. 63. Meresse S, Delbart C, Fruchart JC, Cecchelli R. Low-density lipoprotein receptor on endothelium of brain capillaries. J Neurochem 1989; 53(2):340–345.

The endothelium as an organ

65

64. Dehouck B, Fenart L, Dehouck MP, Pierce, A, Torpier G, Cecchelli R. A new function for the LDL receptor: transcytosis of LDL across the blood-brain barrier. J Cell Biol 1997; 138(4):877– 889. 65. Kartner N, Evernden-Porelle D, Bradley G, Ling V. Detection of P-glycoprotein in multidrugresistant cell lines by monoclonal antibodies. Nature 1985; 316(6031): 820–823. 66. Beaulieu E, Demeule M, Ghitescu L, Beliveau R. P-glycoprotein is strongly expressed in the luminal membranes of the endothelium of blood vessels in the brain. Biochem J 1997; 326(Pt 2):539–544. 67. Sun H, Dai H, Shaik N, Elmquist WF. Drug efflux transporters in the CNS. Adv Drug Deliv Rev 2003; 55(1):83–105. 68. Keep RF, Ennis SR, Betz AL. Blood-brain barrier ion transport. In: Pardridge WM, ed. Introduction to the Blood-brain Barrier. Cambridge: Cambridge University Press, 1998:207– 213. 69. Badaut J, Lasbennes F, Magistretti PJ, Regli L. Aquaporins in brain: distribution, physiology, and pathophysiology. J Cereb Blood Flow Metab 2002; 22(4):367–378. 70. Davies DC. Blood-brain barrier breakdown in septic encephalopathy and brain tumours. J Anat 2002; 200(6):639–646. 71. Risau W, Hallmann R, Albrecht U, Henke-Fahle S. Brain induces the expression of an early cell surface marker for blood-brain barrier-specific endothelium. EMBO J 1986; 5(12):3179– 3183. 72. Stewart PA, Wiely MJ. Developing nervous tissue induces formation of blood-brain barrier characteristics in invading endothelial cells: a study using quail-chick transplantation chimeras. Dev Biol 1981; 84(1):183–192. 73. Janzer RC, Raff MC. Astrocytes induce blood-brain barrier properties in endothelial cells. Nature 1987; 325(6101):253–257. 74. Bush TG, Puvanachandra N, Horner CH, Polito A, Ostenfeld T, Svendsen CN, Mucke L, Johnson MH, Sofroniew MV. Leukocyte infiltration, neuronal degeneration, and neurite outgrowth after ablation of scar-forming, reactive astrocytes in adult transgenic mice. Neuron 1999; 23(2):297–308. 75. Liedtke W, Edelmann W, Bieri PL, Chiu FC, Cowan NJ, Kucherlapati R, Raine CS. GFAP is necessary for the integrity of CNS white matter architecture and long-term maintenance of myelination. Neuron 1996; 17(4):607–615. 76. Pekny M, Stanness KA, Eliasson C, Betsholtz C, Janigro D. Impaired induction of blood–brain barrier properties in aortic endothelial cells by astrocytes from GFAP-deficient mice. Glia 1998; 22(4):390−400. 77. Cecchelli R, Dehouck B, Descamps L, Fenart L, Buee-Scherrer VV, Duhem C, Lundquist S, Rentfel M, Torpier G, Dehouck MP. In vitro model for evaluating drug transport across the blood-brain barrier. Adv Drug Deliv Rev 1999; 36(2–3): 165–178. 78. Tan KH, Dobbie MS, Felix RA, Barrand MA, Hurst RD. A comparison of the induction of immortalized endothelial cell impermeability by astrocytes. Neuroreport 2001; 12(7):1329– 1334. 79. Brayton J, Qing Z, Hart MN, VanGilder JC, Fabry Z. Influence of adhesion molecule expression by human brain microvessel endothelium on cancer cell adhesion. J Neuroimmunol 1998; 89(1–2):104–112. 80. Gerhart DZ, Broderius MA, Drewes LR. Cultured human and canine endothelial cells from brain microvessels. Brain Res Bull 1988; 21(5):785–793. 81. Mizuguchi H, Utoguchi N, Mayumi T. Preparation of glial extracellular matrix: a novel method to analyze glial-endothelial cell interaction. Brain Res Brain Res Protoc 1997; 1(4):339–343. 82. Biegel D, Spencer DD, Pachter JS. Isolation and culture of human brain microvessel endothelial cells for the study of blood-brain barrier properties in vitro. Brain Res 1995; 692(1–2):183–189. 83. Stanness KA, Guatteo E, Janigro D. A dynamic model of the blood-brain barrier “in vitro”. Neurotoxicology 1996; 17(2):481−496.

Endothelial cells in health and disease

66

84. Rubin LL, Hall DE, Porter S, Barbu K, Cannon C, Horner HC, Janatpour M, Liaw CW, Manning K, Morales J. et al., A cell culture model of the blood-brain barrier. J Cell Biol 1991; 115(6):1725–1735. 85. Tontsch U, Bauer HC. Glial cells and neurons induce blood-brain barrier related enzymes in cultured cerebral endothelial cells. Brain Res 1991; 539(2): 247–253. 86. Duport S, Robert F, Muller D, Grau G, Parisi L, Stoppini L. An in vitro blood–brain barrier model: cocultures between endothelial cells and organotypic brain slice cultures. Proc Natl Acad Sci USA 1998; 95(4):1840–1845. 87. Beck DW, Vinters HV, Hart MN, Cancilla PA. Glial cells influence polarity of the blood-brain barrier. J Neuropathol Exp Neurol 1984; 43(3):219–224. 88. Dehouck MP, Meresse S, Delorme P, Fruchart JC, Cecchelli R. An easier, reproducible, and mass-production method to study the blood-brain barrier in vitro. J Neuro-chem 1990; 54(5):1798–1801. 89. DeBault LE. Gamma-glutamyltranspeptidase induction mediated by glial foot process-to endothelium contact in co-culture. Brain Res 1981; 220(2):432–435. 90. Raub TJ, Kuentzel SL, Sawada GA. Permeability of bovine brain microvessel endothelial cells in vitro: barrier tightening by a factor released from astroglioma cells. Exp Cell Res 1992; 199(2):330–340. 91. Roux F, Durieu-Trautmann O, Chaverot N, Claire M, Mailly P, Bourre JM, Strosberg AD, Couraud, PO. Regulation of gamma-glutamyl transpeptidase and alkaline phosphatase activities in immortalized rat brain microvessel endothelial cells. J Cell Physiol 1994; 159(1):101–113. 92. Beuckmann C, Hellwig S, Galla HJ. Induction of the blood/brain-barrier-associated enzyme alkaline phosphatase in endothelial cells from cerebral capillaries is mediated via cAMP. Eur J Biochem 1995; 229(3):641–644. 93. Takaknra Y, Trammel AM, Kuentzel SL, Raub TJ, Davies A, Baldwin SA, Borchardt RT. Hexose uptake in primary cultures of bovine brain microvessel endothelial cells. II. Effects of conditioned media from astroglial and glioma cells. Biochim Biophys Acta 1991; 1070(1):11– 19. 94. Fenart L, Buee-Scherrer V, Descamps L, Duhem C, Poullain MG, Cecchelli R, Dehouck MP. Inhibition of P-glycoprotein: rapid assessment of its implication in blood-brain barrier integrity and drug transport to the brain by an in vitro model of the blood-brain barrier. Pharm Res 1998; 15(7):993–1000. 95. Stanness KA, Westrum LE, Fornaciari E, Mascagni P, Nelson JA, Stenglein SG, Myers T, Janigro D. Morphological and functional characterization of an in vitro blood-brain barrier model. Brain Res 1997; 771(2):329–342. 96. Cohen Z, Ehret M, Maitre M, Hamel E. Ultrastructural analysis of tryptophan hydroxylase immunoreactive nerve terminals in the rat cerebral cortex and hippocampus: their associations with local blood vessels. Neuroscience 1995; 66(3): 555–569. 97. Estrada J, DeFelipe J. Nitric oxide-producing neurons in the neocortex: morphological and functional relationship with intraparenchymal microvasculature. Cereb Cortex 1998; 8(3):193– 203. 98. Cestelli A, Catania C, D’Agostino S, Di Liegro I, Licata L, Schiera G, Pitarresi GL, Savettieri G, De Caro V, Giandalia G, Giannola LI. Functional feature of a novel model of blood brain barrier: studies on permeation of test compounds. J Control Release 2001; 76(1–2):139–147. 99. Stanness KA, Neumaier JF, Sexton TJ, Grant GA, Emmi A, Maris DO, Janigro D. A new model of the blood-brain barrier: co-culture of neuronal, endothelial and glial cells under dynamic conditions. Neuroreport 1999; 10(18):3725–3731. 100. Igarashi Y, Utsumi H, Chiba H, Yamada-Sasamori Y, Tobioka H, Kamimura Y, Furuuchi K, Kokai Y, Nakagawa T, Mori M, Sawada N. Glial cell line-derived neurotrophic factor induces barrier function of endothelial cells forming the blood-brain barrier. Biochem Biophys Res Commun 1999; 261(1):108–112.

The endothelium as an organ

67

101. Wolburg H, Neuhaus J, Kniesel U, Krauss B, Schmid EM, Ocalan M, Farrell C, Risau W. Modulation of tight junction structure in blood-brain barrier endothelial cells. Effects of tissue culture, second messengers and cocultured astrocytes. J Cell Sci 1994; 107(Pt 5):1347–1357. 102. Hoheisel D, Nitz T, Franke H, Wegener J, Hakvoort A, Tilling T, Galla HJ. Hydrocortisone reinforces the blood-brain properties in a serum free cell culture system. Biochem Biophys Res Commun 1998; 247(2):312–315. 103. Pardridge WM, Buciak JL, Friden PM. Selective transport of an anti-transferrin receptor antibody through the blood-brain barrier in vivo. J Pharmacol Exp Ther 1991; 259(1):66–70. 104. Zhang Y, Pardridge WM. Mediated efflux of IgG molecules from brain to blood across the blood-brain barrier. J Neuroimmunol 2001; 114(1–2):168–172. 105. Nag S, Gupta S. Demonstration of Fc receptor on cerebral endothelium. J Neuropathol Exp Neurol 1981; 40:327. 106. Ulvestad E, Williams K, Vedeler C, Antel J, Nyland H, Mork S, Matre R. Reactive microglia in multiple sclerosis lesions have an increased expression of receptors for the Fc part of IgG. J Neurol Sci 1994; 121(2):125–131. 107. Schlachetzki F, Zhu C, Pardridge WM. Expression of the neonatal Fc receptor (FcRn) at the blood-brain barrier. J Neurochem 2002; 81(1):203–206. 108. Jemison LM, Williams SK, Lublin FD, Knobler RL, Korngold R. Interferon-gamma-inducible endothelial cell class II major histocompatibility complex expression correlates with strain- and site-specific susceptibility to experimental allergic encephalomyelitis. J Neuroimmunol 1993; 47(1):15–22. 109. Prat A, Biernacki K, Becher B, Antel JP. B7 expression and antigen presentation by human brain endothelial cells: requirement for proinflammatory cytokines. J Neuro-pathol Exp Neurol 2000; 59(2):129–136. 110. Hart MN, Fabry Z. CNS antigen presentation. Trends Neurosci 1995; 18(11): 475–481. 111. Fabry Z, Waldschmidt MM, Moore SA, Hart MN. Antigen presentation by brain microvessel smooth muscle endothelium. J Neuroimmunol 1990; 28(1):63–71. 112. Risau W, Engelhardt B, Wekerle H. Immune function of the blood-brain barrier: incomplete presentation of protein (auto-)antigens by rat brain microvascular endothelium in vitro. J Cell Biol 1990; 110(5):1757–1766. 113. Fabry Z, Waldschmidt MM, Hendrickson D, Keiner J, Love-Homan L, Takei F, Hart MN. Adhesion molecules on murine brain microvascular endothelial cells: expression and regulation of ICAM-1 and Lgp 55. J Neuroimmunol 1992; 36(1):1–11. 114. Greenwood J, Etienne-Manneville S, Adamson P, Couraud PO. Lymphocyte migration into the central nervous system: implication of ICAM-1 signalling at the blood-brain barrier. Vascul Pharmacol 2002; 38(6):315–322. 115. Mark KS, Miller DW. Increased permeability of primary cultured brain microvessel endothelial cell monolayers following TNF-alpha exposure. Life Sci 1999; 64(21):1941–1953. 116. Fabry Z, Topham DJ, Fee D, Herlein J, Carlino JA, Hart MN, Sriram S. TGF-beta 2 decreases migration of lymphocytes in vitro and homing of cells into the central nervous system in vivo. J Immunol 1995; 155(1):325–332. 117. Hickey WF. Basic principles of immunological surveillance of the normal central nervous system. Glia 2001; 36(2):118–124. 118. Defazio G, Livrea P, Giorelli M, Martino D, Roselli F, Ricchiuti F, Trojano M. Interferon beta-la downregulates TNFalpha-induced intercellular adhesion molecule 1 expression on brain microvascular endothelial cells through a tyrosine kinase-dependent pathway. Brain Res 2000; 881(2):227–230. 119. Taylor AW, Streilein JW. Inhibition of antigen-stimulated effector T cells by human cerebrospinal fluid. Neuroimmunomodulation 1996; 3(2–3):112–118. 120. Burns AR, Bowden RA, MacDonell SD, Walker DC, Odebunmi TO, Donnachie EM, Simon SI, Entman ML, Smith CW. Analysis of tight junctions during neutrophil transendothelial migration. J Cell Sci 2000; 113(Pt 1):45–57.

Endothelial cells in health and disease

68

121. Burns AR, Walker DC, Brown ES, Thurmon, LT, Bowden RA, Keese CR, Simon SI, Entman, ML, Smith CW. Neutrophil transendothelial migration is independent of tight junctions and occurs preferentially at tricellular corners. J Immunol 1997; 159(6): 2893–2903. 122. Alexander JS, Dayton T, Davis C, Hill S, Jackson TH, Blaschuk O, Symonds M, Okayama N, Kevil CG, Laroux FS, Berney SM, Kimpel D. Activated T-lymphocytes express occludin, a component of tight junctions. Inflammation 1998; 22(6):573–582. 123. Huber D, Balda MS, Matter K. Occludin modulates transepithelial migration of neutrophils. J Biol Chem 2000; 275(8):5773–5778. 124. Glass JD, Wesselingh SL, Selnes OA, McArthur JC. Clinical-neuropathologic correlation in HlV-associated dementia. Neurology 1993; 43(11):2230–2237. 125. Price RW, Brew B, Sidtis J, Rosenblum M, Scheck AC, Cleary P. The brain in AIDS: central nervous system HIV-1 infection and AIDS dementia complex. Science 1988; 239(4840):586– 592. 126. Nottet HS, Persidsky Y, Sasseville VG, Nukuna AN, Bock, P, Zhai QH, Sharer LR, McComb RD, Swindells S, Soderland C, Gendelman HE. Mechanisms for the transendothelial migration of HIV-1-infected monocytes into brain. J Immunol 1996; 156(3):1284–1295 . 127. Weis S, Haug H, Budka H. Vascular changes in the cerebral cortex in HIV-1 infection: I. A morphometric investigation by light and electron microscopy. Clin Neuropathol 1996; 5(6):361–366. 128. Bagasra O, Lavi E, Bobroski L, Khalili K, Pestaner JP, Tawadros R, Pomerantz RJ. Cellular reservoirs of HIV-1 in the central nervous system of infected individuals: identification by the combination of in situ polymerase chain reaction and immunohisto-chemistry. Aids 1996; 10(6):573–585. 129. An SF, Groves M, Gray F, Scaravilli F. Early entry and widespread cellular involvement of HIV-1 DNA in brains of HIV-1 positive asymptomatic individuals. J Neuropathol Exp Neurol 1999; 58(11):1156–1162. 130. Banks WA, Freed EO, Wolf KM, Robinson SM, Franko M, Kumar VB. Transport of human immunodeficiency virus type 1 pseudoviruses across the blood-brain barrier: role of envelope proteins and adsorptive endocytosis. J Virol 2001; 75(10):4681–4691. 131. Moses AV, Bloom FE, Pauza CD, Nelson JA. Human immunodeficiency virus infection of human brain capillary endothelial cells occurs via a CD4/galactosylceramideindependent mechanism. Proc Natl Acad Sci USA 1993; 90(22):10474–10478. 132. Wesselingh SL, Power C, Glass JD, Tyor WR, McArthur JC, Farber JM, Griffin JW, Griffin DE. Intracerebral cytokine messenger RNA expression in acquired immunodeficiency syndrome dementia. Ann Neurol 1993; 33(6):576–582. 133. Tyor WR, Glass JD, Griffin JW, Becker PS, McArthur JC, Bezman, L, Griffin DE. Cytokine expression in the brain during the acquired immunodeficiency syndrome. Ann Neurol 1992; 31(4):349–360. 134. Tyor WR, Glass JD, Baumrind N, McArthur JC, Griffin JW, Becker PS, Griffin DE. Cytokine expression of macrophages in HIV-1-associated vacuolar myelopathy. Neurology 1993; 43(5):1002–1009. 135. Carlos T, Kovach N, Schwartz B, Rosa M, Newman B, Wayner E, Benjamin C, Osborn L, Lobb R, Harlan J. Human monocytes bind to two cytokine-induced adhesive ligands on cultured human endothelial cells: endothelial-leukocyte adhesion molecule-1 and vascular cell adhesion molecule-1. Blood 1991; 77(10):2266–2271. 136. Dhawan S, Weeks BS, Soderland C, Schnaper HW, Toro LA, Asthana SP, Hewlett IK, Stetler-Stevenson WG, Yamada SS, Yamada KM, Meltzer MS. HIV-1 infection alters monocyte interactions with human microvascular endothelial cells. J Immunol 1995; 154(1):422–432. 137. Dallasta, LM, Pisarov LA, Esplen JE, Werley JV, Moses AV, Nelson JA, Achim CL. Bloodbrain barrier tight junction disruption in human immunodeficiency virus-1 encephalitis. Am J Pathol 1999; 155(6):1915–1927.

The endothelium as an organ

69

138. Boven LA, Middel J, Verhoef J, De Groot CJ, Nottet HS. Monocyte infiltration is highly associated with loss of the tight junction protein zonula occludens in HIV-1-associated dementia. Neuropathol Appl Neurobiol 2000; 26(4):356–360. 139. Haring HP, Berg EL, Tsurushita N, Tagaya M, del Zoppo GJ. E-selectin appears in nonischemic tissue during experimental focal cerebral ischemia. Stroke 1996; 27(8):1386–1391; discussion 1391–1392. 140. Okada Y, Copeland BR, Mori E, Tung, MM, Thomas WS, del Zoppo GJ. P-selectin and intercellular adhesion molecule-1 expression after focal brain ischemia and reperfusion. Stroke 1994; 25(1):202–211. 141. Kondo T, Kinouchi H, Kawase M, Yoshimoto T. Astroglial cells inhibit the increasing permeability of brain endothelial cell monolayer following hypoxia/reoxygenation. Neurosci Lett 1996; 208(2):101–104. 142. Abbruscato TJ, Davis TP. Combination of hypoxia/aglycemia compromises in vitro bloodbrain barrier integrity. J Pharmacol Exp Ther 1999; 289(2):668–675. 143. Mark KS, Davis TP. Cerebral microvascular changes in permeability and tight junctions induced by hypoxia-reoxygenation. Am J Physiol Heart Circ Physiol 2002; 282(4):H1485– H1494. 144. Hamann GF, Okada Y, Fitridge R, del Zoppo GJ. Microvascular basal lamina antigens disappear during cerebral ischemia and reperfusion. Stroke 1995; 26(11):2120–2126. 145. Petty MA, Wettstein JG. Elements of cerebral microvascular ischaemia. Brain Res Brain Res Rev 2001; 36(1):23–34. 146. Hemmer B, Archelos JJ, Hartung HP. New concepts in the immunopathogenesis of multiple sclerosis. Nat Rev Neurosci 2002; 3(4):291–301. 147. Davie CA, Hawkins CP, Barker GJ, Brennan A, Tofts PS, Miller DH, McDonald WI. Serial proton magnetic resonance spectroscopy in acute multiple sclerosis lesions. Brain 1994; 117(Pt 1):49–58. 148. Miller DH, Grossman RI, Reingold SC, McFarland HF. The role of magnetic resonance techniques in understanding and managing multiple sclerosis. Brain 1998; 121(Pt 1):3–24. 149. Wekerle H. Immune protection of the brain-efficient and delicate. J Infect Dis 2002; 186(suppl 2):S140–S144. 150. Wilcox CE, Ward AM, Evans A, Baker D, Rothlein R, Turk JL. Endothelial cell expression of the intercellular adhesion molecule-1 (ICAM-1) in the central nervous system of guinea pigs during acute and chronic relapsing experimental allergic encephalomyelitis. J Neuroimmunol 1990; 30(1):43–51. 151. Cannella B, Raine CS. The adhesion molecule and cytokine profile of multiple sclerosis lesions. Ann Neurol 1995; 37(4):424–435. 152. Plumb J, McQuaid S, Mirakhur M, Kirk J. Abnormal endothelial tight junctions in active lesions and normal-appearing white matter in multiple sclerosis. Brain Pathol 2002; 12(2):154– 169. 153. Brosnan CF, Claudio L. Brain microvasculature in multiple sclerosis. In: Pardridge WM, ed. Introduction to the Blood-brain Barrier. Cambridge: Cambridge University Press, 1998. 154. Claudio L, Kress Y, Norton WT, Brosnan CF. Increased vesicular transport and decreased mitochondrial content in blood-brain barrier endothelial cells during experimental autoimmune encephalomyelitis. Am J Pathol 1989; 135(6):1157–1168. 155. Li JY, Boado RJ, Pardridge WM. Blood-brain barrier genomics. J Cereb Blood Flow Metab 2001; 21(1):61–68. 156. Plate KH, Risau W. Angiogenesis in malignant gliomas. Glia 1995; 15(3):339–347. 157. Vajkoczy P, Menger MD. Vascular microenvironment in gliomas. J Neurooncol 2000; 50(1– 2):99–108. 158. Vajkoczy P, Schilling L, Ullrich A, Schmiedek P, Menger MD. Characterization of angiogenesis and microcirculation of high-grade glioma: an intravital multifluorescence

Endothelial cells in health and disease

70

microscopic approach in the athymic nude mouse. J Cereb Blood Flow Metab 1998; 18(5):510– 520. 159. Zama A, Tamura M, Inoue HK. Three-dimensional observations on microvascular growth in rat glioma using a vascular casting method. J Cancer Res Clin Oncol 1991; 117(5):396–402. 160. Hirano A, Matsui T. Vascular structures in brain tumors. Hum Pathol 1975; 6(5): 611–621. 161. Stewart PA, Hayakawa K, Hayakawa E, Farrell CL, Del Maestro RF. A quantitative study of blood-brain barrier permeability ultrastructure in a new rat glioma model. Acta Neuropathol (Berl) 1985; 67(1–2):96–102. 162. Liebner S, Fischmann A, Rascher G, Duffner F, Grote EH, Kalbacher H, Wolburg H. Claudin1 and claudin-5 expression and tight junction morphology are altered in blood vessels of human glioblastoma multiforme. Acta Neuropathol (Berl) 2000; 100(3): 323–331. 163. Papadopoulos MC, Saadoun S, Woodrow CJ, Davies DC, Costa-Martins P, Moss RF, Krishna S, Bell BA. Occludin expression in microvessels of neoplastic and non-neoplastic human brain. Neuropathol Appl Neurobiol 2001; 27(5):384–395. 164. Yuan F, Salehi HA, Boucher Y, Vasthare US, Tuma RF, Jain RK. Vascular permeability and microcirculation of gliomas and mammary carcinomas transplanted in rat and mouse cranial windows. Cancer Res 1994; 54(17):4564–4568. 165. Schmiedl UP, Kenney J, Maravilla KR. Dyke Award Paper. Kinetics of pathologic bloodbrain-barrier permeability in an astrocytic glioma using contrast-enhanced MR. AJNR Am J Neuroradiol 1992; 13(1):5–14. 166. Machein MR, Kullmer J, Fiebich BL, Plate KH, Warnke PC. Vascular endothelial growth factor expression, vascular volume, and, capillary permeability in human brain tumors. Neurosurgery 1999; 44(4):732–740; (Discussi on 740–741). 167. Plate KH, Breier G, Weich HA, Mennel HD, Risau W. Vascular endothelial growth factor and glioma angiogenesis: coordinate induction of VEGF receptors, distribution of VEGF protein and possible in vivo regulatory mechanisms. Int J Cancer 1994; 59(4):520–529. 168. Machein MR, Plate KH. VEGF in brain tumors. J Neurooncol 2000; 50(1–2): 109–120. 169. Kevil CG, Payne DK, Mire E, Alexander JS. Vascular permeability factor/vascular endothelial cell growth factor-mediated permeability occurs through disorganization of endothelial junctional proteins. J Biol Chem 1998; 273(24):15099–15103. 170. Guerin C, Laterra J, Hruban RH, Brem H, Drewes LR, Goldstein GW. The glucose transporter and blood-brain barrier of human brain tumors. Ann Neurol 1990; 28(6):758–765. 171. Takamiya Y, Abe Y, Tanaka Y, Tsugu A, Kazuno M, Oshika Y, Maruo K, Ohnishi Y, Sato O, Yamazaki H, Kijima H, Ueyama Y, Tamaoki N, Nakamura M. Murine P-gly-coprotein on stromal vessels mediates multidrug resistance in intracerebral human glioma xenografts. Br J Cancer 1997; 76(4):445–450. 172. Coomber BL, Stewart PA, Hayakawa EM, Farrell CL, Del Maestro RF. A quantitative assessment of microvessel ultrastructure in C6 astrocytoma spheroids transplanted to brain and to muscle. J Neuropathol Exp Neurol 1988; 47(1):29−40. 173. Jellinger KA. Alzheimer disease and cerebrovascular pathology: an update. J Neural Transm 2002; 109(5–6):813–836. 174. Johansson BB. Hypertension. In: Pardridge WM, ed. Introduction to the Blood-brain Barrier. Cambridge: Cambridge University Press, 1998:427–433. 175. Huang SH, Jong AY. Cellular mechanisms of microbial proteins contributing to invasion of the blood-brain barrier. Cell Microbiol 2001; 3(5):277–287. 176. Pardridge WM. CNS drug design based on principles of blood-brain barrier transport. J. Neurochem. 1998; 70:1781–1792. 177. Schena M, Shalon D, Davis RW, Brown PO. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 1995; 270(5235): 467–470. 178. Boado RJ, Pardridge MM. A one-step procedure for isolation of poly(A)+mRNA from isolated brain capillaries and endothelial cells in culture. J Neurochem 1991; 57(6):2136–2139.

The endothelium as an organ

71

179. Diatchenko L, Lau YF, Campbell AP, Chenchik A, Moqadam F, Huang B, Lukyanov S, Lukyanov K, Gurskaya N, Sverdlov ED, Siebert PD. Suppression subtractive hybridization: a method for generating differentially regulated or tissue-specific cDNA probes and libraries. Proc Natl Acad Sci USA 1996; 93(12):6025–6030. 180. Velculescu VE, Zhang L, Vogelstein B, Kinzler KW. Serial analysis of gene expression. Science 1995; 270(5235):484–487. 181. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M, Evans CA, Holt RA, Gocayne JD, Amanatides P, Ballew RM, Huson DH, Wort-man JR, Zhang Q, Kodira CD, Zheng XH, Chen L, Skupski M, Subramanian G, Thomas PD, Zhang J, Gabor Miklos GL, Nelson C, Broder S, Clark AG, Nadeau J, McKusick VA, Zinder N, Levine AJ, Roberts RJ, Simon M, Slayman C, Hunkapiller M, Bolanos R, Delcher A, Dew I, Fasulo D, Flanigan M, Florea L, Halpern A, Hannenhalli S, Kravitz S, Levy S, Mobarry C, Reinert K, Remington K, Abu-Threideh J, Beasley E, Biddick K, Bonazzi V, Brandon R, Cargill M, Chandramouliswaran I, Charlab R, Chaturvedi K, Deng Z, Di Francesco V, Dunn P, Eilbeck K, Evangelista C, Gabrielian AE, Gan W, Ge W, Gong F, Gu Z, Guan P, Heiman TJ, Higgins ME, Ji RR, Ke Z, Ketchum KA, Lai Z, Lei Y, Li Z, Li J, Liang Y, Lin X, Lu F, Merkulov GV, Milshina N, Moore HM, Naik AK, Narayan VA, Neelam B, Nusskern D, Rusch DB, Salzberg S, Shao W, Shue B, Sun J, Wang Z, Wang A, Wang X, Wang J, Wei M, Wides R, Xiao C, Yan C, Yao A, Ye J, Zhan M, Zhang W, Zhang H, Zhao Q, Zheng L, Zhong F, Zhong W, Zhu S, Zhao S, Gilbert D, Baumhueter S, Spier G, Carter C, Cravchik A, Woodage T, Ali F, An H, Awe A, Baldwin D, Baden H, Barnstead M, Barrow I, Beeson K, Busam D, Carver A, Center A, Cheng ML, Curry L, Danaher S, Davenport L, Desilets R, Dietz S, Dodson K, Doup L, Ferriera S, Garg N, Gluecksmann A, Hart B, Haynes J, Haynes C, Heiner C, Hladun S, Hostin D, Houck J, Howland T, Ibegwam C, Johnson J, Kalush F, Kline L, Koduru S, Love A, Mann F, May D, McCawley S, McIntosh T, McMullen I, Moy M, Moy L, Murphy B, Nelson K, Pfannkoch C, Pratts E, Puri V, Qureshi H, Reardon M, Rodriguez R, Rogers YH, Romblad D, Ruhfel B, Scott R, Sitter C, Smallwood M, Stewart E, Strong R, Suh E, Thomas R, Tint NN, Tse S, Vech C, Wang G, Wetter J, Williams S, Williams M, Windsor S, Winn-Deen E, Wolfe K, Zaveri J, Zaveri K, Abril JF, Guigo R, Campbell MJ, Sjolander KV, Karlak B, Kejariwal A, Mi H, Lazareva B, Hatton T, Narechania A, Diemer K, Muruganujan A, Guo N, Sato S, Bafna V, Istrail S, Lippert R, Schwartz R, Walenz B, Yooseph S, Allen D, Basu A, Baxendale J, Blick L, Caminha M, Carnes-Stine J, Caulk P, Chiang YH, Coyne M, Dahlke C, Mays A, Dombroski M, Donnelly M, Ely D, Esparham S, Fosler C, Gire H, Glanowski S, Glasser K, Glodek A, Gorokhov M, Graham K, Gropman B, Harris M, Heil J, Henderson S, Hoover J, Jennings D, Jordan C, Jordan J, Kasha J, Kagan L, Kraft C, Levitsky A, Lewis M, Liu X, Lopez J, Ma D, Majoros W, McDaniel J, Murphy S, Newman M, Nguyen T, Nguyen N, Nodell M, Pan S, Peck J, Peterson M, Rowe W, Sanders R, Scott J, Simpson M, Smith T, Sprague A, Stock-well T, Turner R, Venter E, Wang M, Wen M, Wu D, Wu M, Xia A, Zandieh A, Zhu X. The sequence of the human genome. Science 2001; 291(5507):1304–1351. 182. Kirsch T, Wellner M, Luft FC, Haller, H, Lippoldt, A. Altered gene expression in cerebral capillaries of stroke-prone spontaneously hypertensive rats. Brain Res 2001; 910(1–2):106–115. 183. Dombrowski SM, Desai SY, Marroni M, Cucullo L, Goodrich K, Bingaman W, Mayberg MR, Bengez L, Janigro D. Overexpression of multiple drug resistance genes in endothelial cells from patients with refractory epilepsy. Epilepsia 2001; 42(12): 1501–1506. 184. Desai SY, Marroni M, Cucullo L, Krizanac-Bengez L, Mayberg MR, Hossain MT, Grant GG, Janigro D. Mechanisms of endothelial survival under shear stress. Endothelium 2002; 9(2):89– 102. 185. Kallmann BA, Wagner S, Hummel V, Buttmann M, Bayas A, Tonn JC, Rieckmann P. Characteristic gene expression profile of primary human cerebral endothelial cells. FASEB J 2002; 16(6):589–591. 186. Anderson L, Seilhamer J. A comparison of selected mRNA and protein abundances in human liver. Electrophoresis 1997; 18(3−4):533−537.

Endothelial cells in health and disease

72

187. Cucullo L, Marchi N, Marroni M, Fazio V, Namura S, Janigro D. Blood-brain barrier damage induces release of alpha 2-macroglobulin. Mol Cell Proteomics 2003. 188. Haqqani AS, Twine SM, Ding W, Tessier L, Stanimirovic DD, Kelly JF. Examination of protein expression profiles using various mass spectrometric techniques in brain endothelial cells during ischemic stroke. J Am Soc Mass Spectrom 2003; 51:465 (Abstract). 189. Santoni V, Molloy M, Rabilloud T. Membrane proteins and proteomics: un amour impossible?. Electrophoresis 2000; 21(6):1054–1070. 190. Shusta EV, Boado RJ, Pardridge WM. Vascular proteomics and subtractive antibody expression cloning. Mol Cell Proteomics 2002; 1.1:75–82. 191. Shusta EV, Zhu C, Boado RJ, Pardridge WM. Subtractive expression cloning reveals high expression of CD46 at the blood-brain barrier. J Neuropathol Exp Neurol 2002; 61(7):597–604. 192. Shusta EV, Li JY, Boado RJ, Pardridge WM. The Ro52/SS-A autoantigen has elevated expression at the brain microvasculature. Neuroreport 2003; 14(14):1861–1865. 193. Pardridge, WM. Brain Drug Targeting; The Future of Brain Drug Development. Cambridge, U.K.: Cambridge University Press, 2001:1–370.

3 Lymphatic Endothelium Satoshi Hirakawa and Michael Detmar Cutaneous Biology Research Center and Department of Dermatology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, U.S.A.

1. INTRODUCTION The lymphatic vascular system consists of a network of thin-walled capillaries that drain protein-rich lymph from the extracellular spaces within most organs. Lymphatic capillaries are lined by a continuous single-cell layer of overlapping endothelial cell lines. They lack a continuous basement membrane and pericyte coverage, and are thus highly permeable. The larger lymphatic vessels also contain a muscular and adventitial layer. Lymphatic vessels are absent from avascular tissues such as the epidermis and the nails. Unlike blood vessels, lymphatics are also absent from the cartilage, the brain, and the retina. Lymphatics serve as a drainage system, returning interstitial fluid to the venous circulation via the larger lymphatic collecting vessels and the thoracic duct. Lymphatic vessels also play a major role in the afferent immune response by attracting anti-genpresenting cells—through secretion of chemokines such as secondary lymphoid chemokine—into the lymphatic vascular system and to the regional lymph nodes. Unfortunately, tumor cells can take advantage of these immune pathways to promote lymphatic tumor spread. Other components of the lymphatic system, including the lymph nodes, tonsils, Peyer’s patches, spleen, and thymus, play an important role in the immune response (1). Studies of the lymphatic system have been hampered by the inability to specifically stain lymphatic vessels and by the lack of known lymphatic-specific growth factors. The recent discovery of genes that specifically control lymphatic development, however, and the identification of lymphatic endothelium-specific markers have provided new insights into the molecular mechanisms that control lymphatic development and function (1,2). They have also enabled an improved understanding of the genetic causes of several hereditary diseases that are associated with lymphedema, and they have provided surprising evidence that malignant tumors can actively promote lymphangiogenesis and lymphatic metastasis (3).

Lymphatic endothelium

75

Figure 1 Schematic representation of specific growth factors and of their receptor expression by lymphatic and blood vascular endothelium.

2. LYMPHATIC GROWTH FACTORS Recently, new members of the vascular endothelial cell growth factor (VEGF) family have been discovered that predominantly promote the growth and maintenance of lymphatic vessels in the skin (Fig. 1). The first discovered member of this family, VEGF (also named VEGF-A or VPF, vascular permeability factor) binds to VEGFR-1 and VEGFR-2 predominantly on blood vascular endothelial cells and thereby preferentially promotes vascular endothelial cell proliferation, migration, and survival. However, under certain conditions, VEGF-A might also promote lymphangiogenesis via activation of VEGFR-2 on lymphatic endothelium (4). In contrast, placental growth factor/PlGF selectively acts on VEGFR-1 but not on VEGFR-2 and, because VEGFR-1 is absent from lymphatics, does not promote lymphatic vessel growth. VEGF-C and VEGF-D are the only known ligands of VEGFR-3 (also known as Flt4) that is exclusively expressed by lymphatic endothelium in normal tissues (Fig. 1). Transgenic mice with targeted overexpression of VEGF-C or VEGF-D in the epidermis show enhanced numbers of lymphatic vessels, confirming their role as lymphatic growth factors (5). Moreover, overexpression of a soluble VEGFR-3 in the skin of transgenic mice resulted in a dramatic reduction of cutaneous lymphatic vessels (6).

Endothelial cells in health and disease

76

3. LYMPHATIC-SPECIFIC MARKER GENES AND THEIR POTENTIAL FUNCTION Vascular endothelial growth factor receptor-3 is expressed in developing venous and lymphatic endothelia during early embryonic development; however, its expression becomes largely restricted to the lymphatic endothelium in adult organs (7). Recent studies have identified nonsense mutations in the VEGFR-3 gene in several patients with hereditary lymphedema that are characterized by dilated lymphatic capillaries and interstitial accumulation of lymph fluid (8). Vascular endothelial growth factor receptor-3 mutations have also been identified in Chy mutant mice, which are manifest by cutaneous lymphedema (9). These findings indicate that VEGFR-3 plays an essential role in the development and function of the lymphatic system (10). However, VEGFR-3 is also expressed by some blood capillaries during tumor neovascularization and in wound granulation tissue (11,12) and, therefore, VEGFR-3 alone might not be considered as a sufficiently specific marker for lymphatic vessels. The homeobox gene Prox1 represents the most specific marker for lymphatic endothelium at present (Table 1). Inactivation of Prox1 in mice results in embryonic lethality and completely prevents the development of the lymphatic vasculature (13,14), whereas heterozygote Prox1 deficient mice develop chylous ascites. Among endothelial cells, Prox1 is exclusively expressed in embryonic lymphatic endothelial cells and in lymphatic vessels of adult tissues and tumors (1). During embryonic development, Prox1 plays a key role in the formation of lymphatic progenitor cells from embryonic veins. Beginning at E9.5 of mouse development, Prox1 starts to become specifically expressed in a subpopulation of endothelial cells located on one side of the anterior cardinal vein (13). At this stage, the venous endothelium also expresses the hyaluronan receptor LYVE-1 and VEGFR-3; the expression of both of these receptors later becomes restricted to lymphatic endothelium (7) (Fig. 2). This is followed by polarized budding and migration of Prox1/LYVE-1/VEGFR-3-positive lymphatic progenitor cells (13) that progressively down-regulate the expression of blood vascular genes such as CD34 and laminin (14), and that express increasing levels of lymphatic markers such as VEGFR-3, LYVE-1, and secondary lymphoid chemokine (CCL21) (15). It remains at present unknown whether modulation of specific ephrins and Eph receptors is also involved in mediating the budding of lymphatic progenitor cells from embryonic veins. The lymphatic endothelium hyaluronan receptor LYVE-1, a CD44 homolog, has been identified as a specific cell surface protein of lymphatic endothelial cells and activated macrophages (16,17). LYVE-1 and the blood vascular markers PAL-E and CD34 exhibit mutually exclusive vascular expression patterns in the skin (18). The biological function of LYVE-1 at present remains unclear. Secondary lymphoid chemokine (SLC; also known as CCL21) is released by the lymphatic endothelium, but not by blood vascular endothelium, and interacts

Lymphatic endothelium

77

Table 1 Differential Expression of Established Vascular Markers in Lymphatics vs. Blood Vessels Marker

Blood Vessels

Lymphatics

CD 31 (PECAM-1)

++

+

CD34

++

−

PAL-E

++

−

VEGFR-1

+

−

VEGFR-2

+

+

VEGFR-3

−

+

Type IV collagen

++

(+)

Type XVIII collagen

++

(+)

Laminin

++

(+)

Prox1

−

++

LYVE-1

−

++

Podoplanin

−

++

SLC/CCL21

−

+

Figure 2 Proposed model for the steps involved in the embryonic development of the mammalian lymphatic system. (Adapted from Ref 1.)

Endothelial cells in health and disease

78

with the CC chemokine receptor 7 (CCR7) on mature dendritic cells, leading to their attraction toward the lymphatic vessels (15). Podoplanin, a mucin-type transmembrane glycoprotein, is expressed by lymphatic vasculature but not by blood vessels. Mice lacking podoplanin exhibit congenital lymphedema and impairment of lymphatic vessel formation and function (19). Podoplanin is also known as T1alpha, OTS-1 and PA2.26, and it is also expressed by some other cell types including lung alveolar type I cells, choroid plexus cells, osteocytes and kidney podocytes. Its function has remained unclear but preliminary evidence suggests a role in actin cytoskeleton reorganization. Desmoplakin mediates the attachment of intermediate filaments to the plasma membrane in epithelial cells and is also expressed by lymphatic endothelium, but not in blood vascular endothelium (20). Neuropilin-2 (NRP-2) is a coreceptor for VEGF-C (Fig. 1) and also mediates axonal guidance during neuronal development. Neuropilin-2 is expressed by lymphatic endothelium and neuropilin-2 deficient mice have reduced numbers of small lymphatic vessels and capillaries (21). Angiopoietin-2 is also required for the proper development of the lymphatic vasculature (22). It binds to the endothelialspecific Tie2 receptor and regulates vascular remodeling that is important for vessel sprouting and vessel regression. At present, it remains unclear whether different segments of the lymphatic system or lymphatics in different organs are characterized by expression of distinct sets of marker genes.

4. LINEAGE-SPECIFIC DIFFERENTIATION OF LYMPHATIC ENDOTHELIUM IS MAINTAINED IN VITRO Recent studies have shown that it is possible to selectively isolate and expand human microvascular lymphatic endothelial cells (LEC). These studies revealed that the widely used method for the isolation of human dermal microvascular endothelial cells (HDMEC) in fact yields mixed cell populations of blood vascular endothelial cells and lymphatic endothelial cells, and that commercially available HDMEC cultures also contain variable mixes of both cell types. This may explain some of the previous reports on the different molecular and cellular behavior of HDMEC as compared to human umbilical vein endothelial cells (HUVEC) that are of pure blood vascular origin. Three different approaches for LEC purification have been successfully used: LEC were isolated based on their expression of the lymphatic-specific glycoprotein podoplanin (23,24) or of the lymphatic-specific VEGFR-3 (25). Alternatively, LEC were directly purified from neonatal foreskin cell suspensions as CD34-negative/CD31-positive cells (26). In contrast, pure populations of blood vascular endothelial cells (BEC) were isolated as CD45-negative, podoplanin-negative and CD31-positive cells (23) or as CD45-negative, CD34-positive and CD31-positive cells (26). Importantly, isolated LEC maintain strong and specific expression of LYVE-1, podoplanin and Prox1, whereas BEC do not express significant amounts of podoplanin, LYVE-1 or Prox1 even after multiple passages in vitro. These results indicate that the phenotype of lymphatic endothelium is predominantly genetically driven and stable in vitro, and that the lineage-specific

Lymphatic endothelium

79

lymphatic differentiation is not dependent upon microenvironmental factors. It is of interest that LEC and BEC—when mixed in culture—stay separated and form capillary tubes that wind around each other, indicating that both vascular cell lineages possess mechanisms for lineage-specific cell recognition (23). Gene array analyses of LEC vs. BEC have been successfully used to identify previously unknown lineage-specific genes that are expressed either by lymphatic or by blood vascular endothelium in the skin (24,26). LEC-specific genes include macrophage mannose receptor, desmoplakin, adducin, plakoglobin and CCL20/ MIP3alpha, whereas BEC-specific genes include versican, N-cadherin, endoglin/CD105, integrin 5, ICAM-1, CD44, CXCR4 and VEGFR-1/Flt-1 (24,26); see Table 2. The molecules specifically expressed by LEC and BEC likely play important roles in the specific functional regulation and physiological maintenance of the two types of vasculature, and they also mediate the trafficking of leukocytes into and out of the skin, as well as the exit of tumor cells to form lymphatic or hematogeneous metastases.

5. NEW INSIGHTS INTO THE GENETIC BASIS OF LYMPHEDEMA Lymphedema is caused by insufficient lymph transport due to lymphatic hypoplasia, impaired lymphatic function, or obstruction of lymph flow (27). Primary lymphedema has been classified as Milroy disease when present at birth, or as Meige disease when it develops predominantly after puberty. Milroy disease is linked, in some families, to the VEGFR-3 locus on distal chromosome 5q (28,29), and missense mutations of the VEGFR-3 gene have been identified in several cases of hereditary, early-onset lymphedema (8). As discussed earlier, a gene mutation in the VEGFR-3 tyrosine kinase domain has also been identified in Chy mutant mice (9)—characterized by chronic lymphedema of the extremities and by hypoplastic lymphatic vessels in the skin. In lymphedema-distichiasis, an autosomal-dominant disorder with congenital lymphedema and double rows of eyelashes (distichiasis), inactivating mutations in the FOXC2 gene, a member of the forkhead/winged-helix family of transcription factors, were identified in several families (30). Mutations of the SOX18 gene on chromosome 20q13, an SRYrelated transcription factor, cause recessive and dominant forms of hypotrichosislymphedema-telangiectasia syndrome. Amino acid substitutions in the DNA-binding domain of SOX18 have been found in the recessive form of the disease whereas a heterozygous nonsense mutation

Table 2 Lineage-Specific Expression of Blood Vascular Endothelial Cell (BEC) and Lymphatic Endothelial Cell (LEC) Genes Revealed by Microarray Analysis Blood Vascular Endothelial Cells

Lymphatic Endothelial Cells

Endothelial cells in health and disease

80

Growth factors and chemokines

VEGF-C Placental growth factor

CCL20 (MIP-3alpha) Fibroblast growth factor-12

Receptors

GPR 39 Chemokine receptor X VEGFR-1/Flt1 CD44 CXCR4 Endoglin

Macrophage mannose receptor Membrane glycoprotein gp130 Coxsackievirus/adenovirus rec. Intracellular hyaluronic acid binding protein

Extracellular matrix molecules

Versican Type III collagen, α1 chain Type VI collagen, α1 and α3 chains SPARC/osteonectin

Reelin MFAP3

Adhesion molecules

Integrin alpha4 Integrin alpha5 Integrin beta3 N-cadherin ICAM-1

CEA-CAM Desmoplakin Galectin 8 Integrin alpha9 Plakoglobin

Miscellaneous

Endothelial cell-specific molecule 1 Interferon alpha-inducible protein 27 uPA

Intestinal trefoil factor Down syndrome critical region gene 2

of the transactivation domain causes the dominant hereditary form of the disease (31). SOX18 mutations have also been identified to be responsible for the phenotype of ragged (Ra) mice that show characteristic abnormalities of the hair coat and also develop lymphedema (31). Integrin 9 appears to be specifically expressed by some lymphatic endothelial cells, and mice lacking integrin α9 develop fatal bilateral chylothorax, lymphedema, and lymphocytic infiltration in the chest wall (32). Lymphedema has also been detected in a number of other genetic mouse models, revealing important roles of distinct genes in lymphatic development and function (Table 3). Based on its potent lymphangiogenic effect, VEGF-C has been tested for gene and protein therapy of lymphedema in animal models. Adeno-associated virus-mediated VEGF-C gene therapy promoted lymphatic vessel generation in the skin of Chy mice (9). Furthermore, VEGF-C156S, a mutant form of VEGF-C that selectively activates VEGFR-3, induced regeneration of cutaneous lymphatic vessels without blood vessel growth or vascular leakiness, side effects observed with VEGF-C gene therapy due to activation of VEGFR-2 (33). Successful regeneration of a lymphatic network was also achieved by injection of VEGF-C protein in a surgical lymphedema model in the rabbit ear, indicating the potential use of VEGF-C for the treatment of secondary lymphedema (34).

Lymphatic endothelium

81

Table 3 Genetic Mouse Models with Abnormalities of the Lymphatic Vascular System Gene

Model References

Phenotypes

Transcription factors Prox1

KO

13

No lymphatic vasculature

FOXC2

KO

43

Heterozygote mice exhibiting lymphatic hyperplasia

Net

KO

44

Chyle in the thoracic cage

SOX18

Ragged mice

45

Generalized edema and chyle in the peritoneum; cardiovascular and hair follicle defects

Angiopoietin-2 KO

22

Chylous ascites and peripheral edema; abnormal patterning

KO

46

Lack of sprouting of first lymphatic vessels from embryonic veins

TG

47

Hyperplastic lymphatic vessels

KO

48

Cardiovascular failure and defect remodeling of vascular networks

VEGFR-3

Chy mice

9

Lymphedema

Neuropilin-2

KO

21

Reduction of small lymphatic vessels and capillaries during development

32

Respiratory failure caused by pleural fluid

Growth factors/receptors

VEGF-C

Adhesion molecules Integrin alpha9 KO Miscellaneous Podoplanin

KO

19

Lymphedema, impaired lymphatic patterning and diminished lymphatic transport

SLP-79 and Syk

KO

49

Abnormal blood vessel-lymphatic connections

6. TUMOR LYMPHANGIOGENESIS Tumor metastasis to regional lymph nodes frequently represents the first step of tumor dissemination and serves as a major prognostic indicator for cancer progression. However, little is known about the mechanisms by which tumor cells gain entry into the lymphatic system. A widely held view has suggested that lymphatic endothelium only plays a passive role during this process (35) and that lymphatic invasion only occurs once stroma-infiltrating tumor cells happen upon preexisting peritumoral lymphatic vessels.

Endothelial cells in health and disease

82

The recent identification of lymphatic growth factors and receptors together with the discovery of lymphatic-specific markers have provided important new insights into the formation of tumor-associated lymphatic vessels and their active contribution to lymphatic tumor spread (3). Several studies in animal tumor models have now provided direct experimental evidence that increased levels of VEGF-C or VEGF-D promote tumor lymphangiogenesis and lymphatic tumor spread to regional lymph nodes, and that these effects can be suppressed by blocking VEGFR-3 signaling (36–40). A large number of clinicopathological studies found a direct correlation between tumor expression of the lymphangiogenesis factors VEGF-C or VEGF-D and metastatic tumor spread in many human cancers, providing indirect evidence for the involvement of lymphangiogenesis in tumor progression (41). Our recent studies in human cutaneous malignant melanomas demonstrated—for the first time—the presence of both intratumoral and peritumoral lymphangiogenesis in cutaneous melanoma (42). They also showed that primary melanomas that later metastasized were characterized by increased lymphangiogenesis—as compared to nonmetastatic tumors—and that the degree of tumor lymphangiogenesis can serve as a novel predictor of lymph node metastasis and overall patient survival, independently of tumor thickness (42). Further studies involving larger numbers of cases are needed to confirm these findings.

7. CONCLUSION Due to a number of recent discoveries in the field of vascular biology, some of the mechanisms controlling the normal and pathological development of the lymphatic vasculature are now being established, and several genetic defects have been identified in patients with lymphedema. The identification of specific markers and growth factors for lymphatic vessels and the establishment of cultured lymphatic endothelial cells have been instrumental in this advance. The recent concept of tumor lymphangiogenesis and its role in tumor metastasiss are of particular importance for the progression of malignant tumors. Further progress in this field will likely lead to better diagnosis and treatment of a variety of lymphatic disorders and of certain types of cancer.

REFERENCES 1. Oliver G, Detmar M. The rediscovery of the lymphatic system.Old and new insights into the development and biological function of lymphatic vascular system. Genes Dev 2002; 16:773– 783. 2. Alitalo K, Carmeliet P. Molecular mechanisms of lymphangiogenesis in health and disease. Cancer Cell 2002; 1:219–227. 3. Detmar M, Hirakawa S. The formation of lymphatic vessels and its importance in the setting of malignancy. J Exp Med 2002; 6:713–718. 4. Nagy JA, Vasile E, Feng D, Sundberg C, Brown LF, Detmar MJ, Lawitts JA, Benjamin L, Tan X, Manseau EJ, Dvorak AM, Dvorak HF. Vascular permeability factor/vascular endothelial growth factor induces lymphangiogenesis as well as angiogenesis. J Exp Med 2002; 196:1497– 1506.

Lymphatic endothelium

83

5. Jussila L, Alitalo K. Vascular growth factors and lymphangiogenesis. Physiol Rev 2002; 82:673– 700. 6. Maekinen T, Jussila L, Veikkola T, Karpanen T, Kettunen MI, Pulkkanen KJ, Kauppinen R, Jackson DG, Kubo H, Nishikawa S, Yla HS, Alitalo K. Inhibition of lymphangiogenesis with resulting lymphedema in transgenic mice expressing soluble VEGF receptor-3. Nat Med 2001; 7:199–205. 7. Kaipainen A, Korhonen J, Mustonen T, van HV, Fang GH, Dumont D, Breitman M, Alitalo K. Expression of the fms-like tyrosine kinase 4 gene becomes restricted to lymphatic endothelium during development. Proc Natl Acad Sci USA 1995; 92:3566–3570. 8. Karkkainen MJ, Ferrell RE, Lawrence EC, Kimak MA, Levinson KL, McTigue MA, Alitalo K, Finegold DN. Missense mutations interfere with VEGFR-3 signalling in primary lymphoedema. Nat Genet 2000; 25:153–159. 9. Karkkainen MJ, Saaristo A, Jussila L, Karila KA, Lawrence EC, Pajusola K, Bueler H, Eichmann A, Kauppinen R, Kettunen MI, Yla HS, Finegold DN, Ferrell RE, Alitalo K. A model for gene therapy of human hereditary lymphedema. Proc Natl Acad Sci USA 2001; 98:12677– 12682. 10. Kukk E, Lymboussaki A, Taira S, Kaipainen A, Jeltsch M, Joukov V, Alitalo K VEGF-C receptor binding and pattern of expression with VEGFR-3 suggests a role in lymphatic vascular development. Development 1996; 122:3829–3837. 11. Paavonen K, Puolakkainen P, Jussila L, Jahkola T, Alitalo K. Vascular endothelial growth factor receptor-3 in lymphangiogenesis in wound healing. Am J Pathol 2000; 156:1499–1504. 12. Valtola R, Salven P, Heikkila P, Taipale J, Joensuu H, Rehn M, Pihlajaniemi T, Weich H, deWaal R, Alitalo K. VEGFR-3 and its ligand VEGF-C are associated with angio-genesis in breast cancer. Am J Pathol 1999; 154:1381–1390. 13. Wigle JT, Oliver G. Prox 1 function is required for the development of the murine lymphatic system. Cell 1999; 98:769–778. 14. Wigle JT, Harvey N, Detmar M, Lagutina I, Grosveld G, Gunn MD, Jackson DG, Oliver G. An essential role for Prox1 in the induction of the lymphatic endothelial cell phenotype. EMBO J 2002; 21:1505–1513. 15. Gunn MD, Tangemann K, Tam C, Cyster JG, Rosen SD, Williams LT. A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc Natl Acad Sci USA 1998; 95:258–263. 16. Banerji S, Ni J, Wang SX, Clasper S, Su J, Tammi R, Jones M, Jackson DG. LYVE-1, a new homologue of the CD44 glycoprotein, is a lymph-specific receptor for hyaluronan. J Cell Biol 1999; 144:789–801. 17. Prevo R, Banerji S, Ferguson DJ, Clasper S, Jackson DG. Mouse LYVE-1 is an endocytic receptor for hyaluronan in lymphatic endothelium. J Biol Chem 2001; 276: 19420–19430. 18. Skobe M, Detmar M. Structure, function and molecular control of the skin lymphatic system. J Invest Dermatol Symp Proc 2000; 5:14–19. 19. Schacht V, Ramirez MI, Hong YK, Hirakawa S, Feng D, Harvey N, Williams M, Dvorak AM, Dvorak HF, Oliver G, Detmar M. T1alpha/podoplanin deficiency disrupts normal lymphatic vasculature formation and causes lymphedema. EMBO J 2003; 22:3546–3556. 20. Ebata N, Nodasaka Y, Sawa Y, Yamaoka Y, Makino S, Totsuka Y, Yoshida S. Desmoplakin as a specific marker of lymphatic vessels. Microvasc Res 2001; 61:40–48. 21. Yuan L, Moyon D, Pardanaud L, Breant C, Karkkainen MJ, Alitalo K, Eichmann A. Abnormal lymphatic vessel development in neuropilin 2 mutant mice. Development 2002; 129:4797–4806. 22. Gale NW, Thurston G, Hackett SF, Renard R, Wang Q, McClain C, Martin C, Witte C, Witte MH, Jackson D, Suri C, Campochiaro PA, Wiegand SJ, Yancopoulos GD. Angio-poietin-2 is required for postnatal angiogenesis and lymphatic patterning, and only the latter role is rescued by angiopoietin-1. Dev Cell 2002; 3:411−23.

Endothelial cells in health and disease

84

23. Kriehuber E, Breiteneder GS, Groeger M, Soleiman A, Schoppmann SF, Stingl G, Kerjaschki D, Maurer D. Isolation and characterization of dermal lymphatic and blood endothelial cells reveal stable and functionally specialized cell lineages. J Exp Med 2001; 194:797–808. 24. Petrova TV, Maekinen T, Maekelae TP, Saarela J, Virtanen I, Ferrell RE, Finegold DN, Kerjaschki D, Ylae-Herttuala S, Alitalo K. Lymphatic endothelial reprogramming of vascular endothelial cells by the Prox-1 homeobox transcription factor. EMBO J 2002; 21:4593–4599. 25. Maekinen T, Veikkola T, Mustjoki S, Karpanen T, Catimel B, Nice EC, Wise L, Mercer A, Kowalski H, Kerjaschki D, Stacker SA, Achen MG, Alitalo K. Isolated lymphatic endothelial cells transduce growth, survival and migratory signals via the VEGF-C/D receptor VEGFR-3. EMBO J 2001; 20:4762–4773. 26. Hirakawa S, Hong YK, Harvey N, Schacht V, Matsuda K, Libermann T, Detmar M. Identification of vascular lineage-specific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am J Pathol 2003; 162:575–586. 27. Witte MH, Bernas MJ, Martin CP, Witte CL. Lymphangiogenesis and lymphangiodysplasia: from molecular to clinical lymphology. Microsc Res Tech 2001; 55:122–145. 28. Ferrell RE, Levinson KL, Esman JH, Kimak MA, Lawrence EC, Barmada MM, Finegold DN. Hereditary lymphedema: evidence for linkage and genetic heterogeneity. Hum Mol Genet 1998; 7:2073–2078. 29. Evans AL, Brice G, Sotirova V, Mortimer P, Beninson J, Burnand K, Rosbotham J, Child A, Sarfarazi M. Mapping of primary congenital lymphedema to the 5q35.3 region. Am J Hum Genet 1999; 64:547–555. 30. Fang J, Dagenais SL, Erickson RP, Arlt MF, Glynn MW, Gorski JL, Seaver LH, Glover TW. Mutations in FOXC2 (MFH-1), a forkhead family transcription factor, are responsible for the hereditary lymphedema-distichiasis syndrome. Am J Hum Genet 2000; 67:1382–1388. 31. Irrthum A, Devriendt K, Chitayat D, Matthijs G, Glase C, Steijlen PM, Fryns JP, Van Steensel MA, Vikkula M. Mutations in the transcription factor gene SOX18 underlie recesive and dominant forms of hypothrichosis-lymphedema-teleangeactasia. Am J Hum Genet 2003; 72:1470–1478. 32. Huang XZ, Wu JF, Ferrando R, Lee JH, Wang YL, Farese RV Jr, Sheppard D. Fatal bilateral chylothorax in mice lacking the integrin alpha9beta1. Mol Cell Biol 2000; 20:5208–5215. 33. Saaristo A, Veikkola T, Tammela T, Enholm B, Karkkainen MJ, Pajusola K, Bueler H, YlaHerttuala S, Alitalo K. Lymphangiogenic gene therapy with minimal blood vascular sideeffects. J Exp Med 2002; 196:719–730. 34. Szuba A, Skobe M, Karkkainen MJ, Shin WS, Beynet DP, Rockson NB, Dakhil N, Spilman S, Goris ML, Strauss HW, Quertermous T, Alitalo K, Rockson SG. Therapeutic lymphangiogenesis with human recombinant VEGF-C. FASEB J 2002; 16:1985–1987. 35. Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases. Nature 2000; 407: 249–257. 36. Skobe M, Hawighorst T, Jackson DG, Prevo R, Janes L, Velasco P, Riccardi L, Alitalo K, Claffey K, Detmar M. Induction of tumor lymphangiogenesis by VEGF-C promotes breast cancer metastasis. Nat Med 2001; 7:192–198. 37. Stacker SA, Caesar C, Baldwin ME, Thornton GE, Williams RA, Prevo R, Jackson DG, Nishikawa S, Kubo H, Achen MG. VEGF-D promotes the metastatic spread of tumor cells via the lymphatics. Nat Med 2001; 7:186–191. 38. Mandriota SJ, Jussila L, Jeltsch M, Compagni A, Baetens D, Prevo R, Banerji S, Huarte J, Montesano R, Jackson DG, Orci L, Alitalo K, Christofori G, Pepper MS. Vascular endothelial growth factor-C-mediated lymphangiogenesis promotes tumour metastasis. EMBO J 2001; 20:672–682. 39. Karpanen T, Egeblad M, Karkkainen MJ, Kubo H, Yla-Herttuala S, Jaattela M, Alitalo K. Vascular endothelial growth factor C promotes tumor lymphangiogenesis and intralymphatic tumor growth. Cancer Res 2001; 61:1786–1790.

Lymphatic endothelium

85

40. He Y, Kozaki K, Karpanen T, Koshikawa K, Yla-Herttuala S, Takahashi T, Alitalo K. Suppression of tumor lymphangiogenesis and lymph node metastasis by blocking vascular endothelial growth factor receptor 3 signaling. J Natl Cancer Inst 2002; 94: 819–825. 41. Stacker SA, Baldwin ME, Achen MG. The role of tumor lymphangiogenesis in metastatic spread. FASEB J 2002; 16:922–934. 42. Dadras SS, Paul T, Bertoncini J, Brown LF, Muzikansky A, Jackson DG, Ellwanger U, Garbe C, Mihm MC, Detmar M. Tumor lymphangiogenesis: a novel prognostic indicator for cutaneous melanoma metastasis and survival. Am J Pathol 2003; 162:1951–1960. 43. Kriederman BM, Myloyde TL, Witte MH, Dagenais SL, Witte CL, Rennels M, Bernas MJ, Lynch MT, Erickson RP, Caulder MS, Miura N, Jackson D, Brooks BP, Glover TW. FOXC2 haploinsufficient mice are a model for human autosomal dominant lymphedema-distichiasis syndrome. Hum Mol Genet 2003; 12:1179–1185. 44. Ayadi A, Suelves M, Dolle P, Wasylyk B. Net-targeted mutant mice develop a vascular phenotype and up-regulate egr-1. EMBO J 2001; 20:5139–5152. 45. Pennisi D, Gardner J, Chambers D, Hosking B, Peters J, Muscat G, Abbott C, Koopman P. Mutations in Sox18 underlie cardiovascular and hair follicle defects in ragged mice. Nat Genet 2000; 24:434–437. 46. Karkkainen MJ, Haiko P, Sainio K, Partanen J, Taipale J, Petrova TV, Jeltsch M, Jackson DG, Talikka M, Rauvala H, Betsholtz C, Alitalo K. Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat Immunol 2004; 5:74–80. 47. Jeltsch M, Kaipainen A, Joukov V, Kukk E, Lymbousssaki A, MX, Lakso M, Alitalo K. Hyperplasia of lymphatic vessels in VEGF-C transgenic mice. Science 1997; 276: 1423–1425. 48. Dumont DJ, Jussila L, Taipale J, Lymboussaki A, Mustonen T, Pajusola K, Breitman M, Alitalo K. Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science 1998; 282:946–949. 49. Abtahian F, Guerriero A, Sebzda E, Lu MM, Zhou R, Mocsai A, Myers EE, Huang B, Jackson DG, Ferrari VA, Tybulewicz V, Lowell CA, Lepore JJ, Koretzky GA, Kahn ML. Regulation of blood and lymphatic vascular separation by signaling proteins SLP-76 and Syk. Science 2003; 299:247–251.

4 High Endothelial Venules Jean-Marc Gauguet, Roberto Bonasio, and Ulrich H.von Andrian The CBR Institute for Biomedical Research, Inc. and Department of Pathology, Harvard Medical School, Boston, Massachusetts, U.S.A.

1. INTRODUCTION Secondary lymphoid organs, including peripheral lymph nodes (PLNs), mesenteric lymph nodes (MLNs), Peyer’s patches (PPs), appendix, tonsils, and spleen are essential components of the immune system. These organs are specialized to collect antigen (Ag) and Ag presenting cells (APCs) from distinct anatomical regions. Their extravascular environment is designed to optimize lymphocyte recognition of and subsequent responses to cognate Ag. A remarkable property of secondary lymphoid organs is their ability to recruit vast numbers of circulating B and T lymphocytes. With the exception of the spleen, secondary lymphoid organs contain specialized post-capillary and small collecting venules, called high endothelial venules (HEVs), which serve as the principal site of lymphocyte entry from the blood (1,2). High endothelial venules express organspecific patterns of lymphocyte traffic molecules that are not found in other microvascular beds. These molecules coordinate the recruitment of circulating lymphocytes by promoting multi-step adhesion cascades involving selectins, chemokines, integrins, and their respective ligands (3,4). In this chapter, we will review our current understanding of the functional, structural, and molecular characteristics of HEVs that allow them to function as the gateway to secondary lymphoid organs. We will examine the development of HEVs in normal and pathologic settings and how these unique microvessels respond to environmental cues. In addition, we will discuss techniques to study these structures as well as emerging technologies that may pave the way for future discoveries.

2. HISTORICAL PERSPECTIVE High endothelial venules have been the subject of scientific scrutiny since the 19th century. Because of the unusual cuboidal shape of the endothelial lining, histological cross-sections of these microvessels resemble those of exocrine gland ducts (the misnomer “lymph gland” for LNs probably derived from erroneous interpretations of HEVs). During the following decades, their physiological function was a topic of

Endothelial cells in health and disease

88

controversy, even after it was firmly established that HEVs are part of the circulatory system where large numbers of lymphocytes traverse the vessel wall. The prevalent initial interpretation was that lymphocytes were generated in lymphoid organs and secreted via HEVs into the blood (5,6). A much clearer picture emerged only in the early 1960s when Gowans and Knight (7) demonstrated that small (i.e., naïve) lymphocytes recirculate from the blood into lymph nodes and via the thoracic duct back into the blood. These authors injected radio-labeled lymphocytes i.v. into rats and recorded the recovery of labeled cells in the thoracic duct lymph. The recirculating cells first appeared 2–4 hr after injection and peaked in the lymph after 24 hr. More than 90% of the injected cells were recovered from chronic thoracic duct fistulas within four days, indicating that virtually all lymphocytes migrate actively and continuously (7). The fundamental importance of HEVs in the recirculation process was revealed when Marchesi and Gowans (1) used electron microscopy to demonstrate that these vessels serve as the entry site for circulating lymphocytes into secondary lymphoid organs. In this landmark study, i.v. injected lymphocytes could be “seen penetrating the endothelium of the venules” in LNs; lymphocytes were found within the vessel lumen, between endothelial cells and at the basement membrane. Labeled cells that had reached the LN parenchyma were in close proximity to HEVs, suggesting that the new arrivals had just crossed the microvascular barrier. While in most mammals the chief direction of lymphocyte migration is from the blood across HEVs into lymphoid tissues, some cells may also migrate in the opposite direction. In sheep and rat LNs (8,9), lymphocytes use HEVs predominantly to migrate into lymphoid organs, while in pig MLNs, lymphocyte migrate from the lymphoid tissue into the bloodstream (10). Despite these apparent species differences, HEVs are thought to be essential vascular determinants of lymphocyte traffic in all mammals.

3. HOW TO STUDY HEVs The study of HEVs has relied heavily on microscopic examination of tissue sections and, as a result, knowledge of these structures has deepened with advances in microscopy technology. The use of electron microscopy (EM) allowed investigators to peer into high endothelial cells (HECs) to examine their intracellular components and architecture (11). Electron microscopy studies also proved to be pivotal for the first description of how lymphocytes use HEVs to gain entry into lymphoid organs (1). More recent studies have employed EM to examine the transport of soluble lymph-borne factors across HEVs to the vessel lumen (12). The first in vitro tool to study interactions between lymphocytes and HEVs was described by Stamper and Woodruff (13). In this assay, frozen sections of lymphoid organs are incubated with a lymphocyte suspension, gently washed, and fixed with glutaraldehyde. Using dark-field microscopy, it is then possible to visualize and enumerate lymphocytes that interact with HEVs. A later modification of the StamperWoodruff assay reduced the number of lymphocytes needed for this technique (14). Since lymphoid organs rapidly recruit and transiently sequester circulating lymphocytes, a homing assay can be used to measure the number of transfused lymphocytes (distinguished from endogenous lymphocytes by labeling with radioisotopes, fluorescent

High endothelial venules

89

dyes, or genetically encoded markers) that enter into a lymphoid organ within a certain period of time (typically 1–24 hr) (14). By combining homing experiments with the Stamper-Woodruff assay, it was discovered that the cells that showed tropism for secondary lymphoid organs in vivo, also adhered most avidly to HEVs in vitro, and that this binding required an active metabolism and a calcium flux (15). Investigators also used these assays to test the role of specific traffic molecules, e.g., by treating lymphocytes, tissue sections or animals with antibodies or with an enzyme such as neuraminidase (16–19). A more recent development is intravital microscopy, which uses epifluorescence techniques to visualize lymphocyte interactions with HEVs in a living animal. The first intravital analysis of lymphocyte interaction with HEVs was performed in murine PPs (20). More recently, intravital microscopy techniques to examine murine subiliac (also called inguinal) (21,22) and popliteal PLNs were developed (23). Immunohistochemistry and affinity chromatography with HEVs-specific monoclonal antibodies (mAbs) and in situ hybridization have been among the key technologies used to identify and characterize molecules produced by HEVs that are essential for lymphocyte trafficking. For example, mAb MECA-79 was critical for the identification of peripheral node addressin (PNAd), a group of sulfated glycans that are expressed in both human and mouse PLNs HEVs (18). In situ hybridization was instrumental to identify the chemokine CCL21 (SLC/TCA-4/6Ckine/exodus 2), which induces chemotaxis of naïve lymphocytes and is expressed in the T cell area of PLNs, especially in HEVs (24). In addition to studying the function of HECs in vivo, there have been several attempts to establish and study primary HEC-like cell lines (25,26). These lines preserve some of the properties of bona fide HECs, including a certain ability to bind lymphocytes (27), and expression of some proteins and antigens characteristic of primary HECs (28,29). However, none of the cultured HEVs lines express all of the traffic molecules that are found on HECs in vivo and, therefore, they do not fully recapitulate the unique biology of these cells (28,29).

4. MORPHOLOGY AND ANATOMY OF HEVs 4.1. Functional Anatomy of LNs and PPs Lymph nodes are strategically located along lymphatics, which drain peripheral tissues. The prototypical LN is a lentil- or bean-shaped parenchymatous organ, composed of loosely associated lymphoid and myeloid leukocytes and stromal cells, surrounded by a fibrous capsule that confers structural integrity. Peripheral lymph nodes are subdivided into two main areas: cortex and medulla (Fig. 1a). The cortex comprises the superficial B cell area that contains B follicles and germinal centers and the deeper T cell area or paracortex. The structural unit of the deep cortex is the paracortical cord, a column of T cells and antigen presenting cells, delimited by lymph-draining sinuses (30). At the center of each cord is an HEVs, which merge into incrementally larger collecting venules while draining blood toward the medulla. The medulla contains a highly developed system of sinuses that receive lymph fluid from the subcapsular and cortical sinuses and coalesce at the LN’s hilus into an efferent lymphatic vessel (31). The paracortical cords extend into

Endothelial cells in health and disease

90

the medulla, where they are referred to as medullary cords, which contain many plasma cells and macrophages (32). Peyer’s patches are organized lymphoid organs in the wall of the small intestine. In contrast to LNs, PPs do not receive afferent lymph. Their interface with the gut lumen is formed by M cells, a specialized type of epithelial cell, which continuously transcytose antigenic material from the intestinal cavity into the subepithelial

Figure 1 Functional anatomy of lymph nodes. (A) A schematic representation of a mouse LN is provided with relevant features highlighted. The feeding artery and the venous microcirculation are shown in red and blue, respectively. The large venule that drains blood from the medulla is marked as order I. To improve clarity, only a segment of the intranodal venular tree is shown. The lymphdraining compartments are shown in light green. Arrows indicate the direction of lymph flow. (B) Crosssection of an HEVs with FRC that drain lymph fluid and lymph-borne small molecules such as chemokines (shown as small red circles) from lymphatic sinuses toward the HEV.

High endothelial venules

91

Electron microscopy studies suggest that the FRC conduits drain lymph into a narrow Peri-Vascular Channel (PVC) surrounding each HEVs. Upon delivery to HEVs, chemokines are transcytosed for presentation in the lumen. An adherent lymphocyte is shown inside the HEVs. (C) Architecture of a venular tree showing the characteristic branching orders in mouse inguinal LNs. dome, an area that is rich in DCs (33,34). The PP parenchyma is dominated by large B follicles, whereas T cells are confined to the interfollicular area.

4.2. Microvascular Organization of PLNs and PP The microvasculature organization varies considerably between different PLNs, but has a number of shared features. We distinguish a main feeding artery and a main collecting vein, both of which access the organ at the hilum (Fig. 1A). The arterial tree is not specialized and does not support interactions with circulating leukocytes; on the other hand, the degree of organization on the venular side is remarkable. In mouse inguinal LNs, starting from the main collecting venule, a distinct hierarchy of branches can usually be identified (22) (Fig. 1C). The endothelium of the first and second branching orders in the medulla has a flat appearance similar to endothelium in non-lymphoid tissues. In contrast, endothelium in higher order venules in the paracortex and deep cortex has a cuboidal, cobble-stone like shape. Thus, only microvessels with a higher branching order are HEVs. Ultrastmctural studies indicate that endothelial cells (ECs) in the transition region between HEVs and flat-walled venules are often arranged in an overlapping plate pattern (35). However, in immunohistochemical terms, the transition from HEVs to non-HEVs is typically very abrupt, from one EC to the next. This has been

Endothelial cells in health and disease

92

Figure 2 MECA-79 antigen expression is restricted to HEVs in PLNs. (A) A false-color intravital micrograph shows MECA-79 coated fluorescent microspheres (yellow) bound to small diameter venules (order III–V) within the venular tree (blue) of a mouse inguinal LN. Beads do not accumulate in arterioles (red), large diameter venules (order I and order II) or capillaries. (B) Intravital micrograph showing the distribution of CFSElabeled MECA-79 after i.v. injection (top panel). A schematic diagram (lower panel) identifies the venular orders and the superficial epigastric artery (SEA) and vein (SEV) in this preparation. (C) Immunohistochemistry reveals MECA-79 staining of a PLN HEVs illustrating the presence of MECA-79 on the lumenal (Lu) and abluminal [basement membrane (BM)] surface of the HECs. (Panels (B) and (C): Modified from Refs. 50,97).

High endothelial venules

93

shown by staining with antibodies against HEVs-specific traffic molecules, such as the mAb MECA-79 against the peripheral node addressin or PNAd (Fig. 2A–C). This transition in immunoreactivity is located at the cortico-medullary boundary, and can be so accurate that a vessel running parallel to the juncture can have a HEVs phenotype on the side that faces the cortex and flat endothelium on the medullary side (36). While HEVs in PLNs are restricted to the T cell area, HEVs in PPs originate in B follicles. From here, they drain blood toward the interfollicular T cell zone where they merge into larger collecting venules. While there are no apparent structural differences between HEVs in PLNs and PPs (37), biochemical, immunohistochemical, and functional studies have uncovered a number of tissue-specific differences, which manifest in the selective recruitment of distinct lymphocyte subsets (see below).

4.3. Lymph Flow in PLNs Lymph is drained to LNs by a system of blind-ending vessels lined by specialized lymphatic endothelial cells. Afferent lymphatic vessels perforate the capsule of the LN on its convex aspect, distal to the hilum, and drain into the subcapsular sinus (Fig. 1A). Lymph-borne cells, particulate material, and soluble molecules have different fates after entering a LN. Dendritic cells (DCs) and effector memory lymphocytes invade the subcapsular sinus floor and migrate to their proper location in the cortex (38,39). Particulate material and large molecules are excluded from the parenchyma (40) and are phagocytosed by subcapsular phagocytes or drain to the

Figure 3 Characteristic ultrastructural and molecular features of HECs in PLN. The vessel lumen and stromal spaces are indicated in the figure. Posttranslational modifications of sialomucins within the prominent Golgi apparatus generates L-selectin ligands and the MECA-79 antigen. efferent lymphatic vessel through the sinus system. Low molecular weight molecules with a radius of less than 4 nm are channeled into the parenchyma along a network of

Endothelial cells in health and disease

94

collagen fibers (41,42) that are ensleeved by fibroblastic reticular cells (FRCs). This sizing column-like drainage system is called the “FRC conduit” (30,43). Fibro-blastic reticular cells also encircle HEVs, and EM studies have suggested the existence of a narrow perivascular gap between them, the perivenular channel (PVC) that is thought to receive lymph fluid via the FRC conduit (Fig. 1B).

4.4. Ultrastructure of High Endothelial Cells The most prominent ultrastructural feature of HECs, besides their characteristic shape, is a prominent Golgi apparatus, which is often oriented toward the luminal aspect of the cell (11) (Fig. 3). In addition, HECs have abundant mitochondria and ribosomes, with sparse rough endoplasmic reticulum and several multi-vesicular bodies (11,44,45).This ultrastructural appearance of HECs is atypical for conventional ECs and was interpreted as indicative of a cellular machinery with high metabolic activity geared toward the biosynthesis of glycoproteins (11). Further support for this notion came from metabolic studies showing that PLNs HECs incorporate substantial amounts of sulfate (46). The importance of these two observations will be examined below. The apical surface of HECs is dotted with shallow pits (9) and a glycocalyx that appears thicker than the glycocalyx on non-HEVs endothelium (47). The abluminal surface rests on a thin basement membrane and macular tight junctions join juxtaposed cells at their apical and basal surfaces (44). These discontinuous junctions, unlike the continuous tight junctions observed in flat endothelium, are thought to allow lymphocytes to squeeze between adjacent endothelial cells without causing vascular leakiness.

5. LEUKOCYTE RECRUITMENT VIA HEVs It is now widely accepted that tissue- and subset-specific leukocyte migration is governed by a sequence of molecularly distinct adhesion and signaling events (3,4,48). Adhesion cascades are initiated by a tethering step that allows leukocytes to bind loosely to endothelial cells. Once attachment has occurred, marginated cells are pushed forward by the blood stream resulting in a slow rolling motion along the vascular wall (step 1). Subsequently, rolling cells encounter chemotactic stimuli on or near the EC surface that can bind to specific leukocyte receptors (step 2). Chemoattractant binding, in turn, induces rapid intracellular signaling and triggers activation-dependent adhesion steps that allow leukocytes to stick firmly (step 3) and, eventually, to emigrate through the vessel wall. While microvessels in most non-lymphoid tissues can only support substantial leukocyte traffic upon exposure to inflammatory mediators, HEVs must recruit large numbers of lymphocytes in the absence of inflammation. Thus, HEVs constitutively express a unique pattern of traffic molecules that are fundamental for the normal function of the immune system.

High endothelial venules

95

5.1. The Multi-Step Adhesion Cascade in PLNs Detailed intravital microscopy analyses have defined the specific adhesion cascades that mediate T and B cell homing to LNs (Fig. 4A) (49–52). Tethering and rolling of both subsets are mediated by L-selectin (CD62L). The endothelial L-selectin ligand is the peripheral node addressin (PNAd), an O-linked carbohydrate moiety, the major components of which are recognized in humans and mice by the monoclonal antibody MECA-79 (18,53). The structure of the MECA-79 antigen and HEV-s-expressed Lselectin ligands will be discussed below. Firm arrest of rolling T cells in LN HEVs is mediated by LFA-1 (49,54,55). Lymphocytes deficient in LFA-1 home poorly to LNs, although the observed magnitude of the homing defect was somewhat variable between two independently generated knockout strains (55,56). In one strain, residual LFA-1-independent homing was found to be mediated by another integrin, α4β7, which was shown to interact with VCAM-1 in LN HEVs (56). The agent that activates integrins on naive and central memory T cells is the chemokine CCL21 (also called SLC, TCA-4, 6Ckine, or exodus 2) (50,51). CCL21 is constitutively expressed by HECs and binds to CCR7 (57–59). A second CCR7 agonist, CCL19 (also called ELC or Mip-3β), is expressed by lymphatic endothelium and interstitial cells within LNs, but not by HEVs. However, lymph-borne CCL19 can be transported to the luminal surface of HEVs where it induces integrin activation on rolling T cells (12). The physiologic role of CCL 19 in lymphocyte homing remains to be determined. Mice have two isoforms of CCL21; CCL21ser is expressed in HEV, whereas CCL2Leu is generated in lymphatic endothelium (60). A mutant strain called plt/plt (paucity in lymph node T cells) is deficient in CCL21ser and CCL19 (60,61). Lymph nodes of plt/plt mice and CCR7 deficient mice contain few naïve T cells, but the B cell compartment is much less affected and LNs in these mutant animals also contain substantial numbers of central memory T cells (62,63). This indicates that B cells and central memory T cells may respond not only to CCR7 agonists, but also to another integrin activating signal in HEVs. Indeed, a recent study has shown that rolling B cell can be stimulated to arrest in HEV by CXCL12 (SDF-1α), the ligand for CXCR4 (52). CCR7 and CXCR4 can independently maintain B cell homing (albeit at some-what lower levels than in wildtype animals where both function simultaneously). Interestingly, although CXCL12 potently induces integrin activation on rolling naïve T cells in vitro (58,64), it does so very poorly in vivo (65). Thus, despite low-level expression of CXCL12 in wildtype murine HEVs, naïve T cells require CCR7 signals, at least in the Balb/c and DDD/1 genetic background (50,52). On the other

Endothelial cells in health and disease

96

Figure 4 Multi-step adhesion cascades in HEVs. Schematic representations of the multi-step cascade in PLNs (A, B) and in PPs (C, D) are shown. In (A) a naïve T lymphocyte undergoes tethering and rolling via L-selectinPNAd (step 1); chemoattractant stimulation by CCR7-CCL21 (step 2); and firm adhesion via activated LFA-1 (step 3); eventually permitting the arrested cell to diapedese. The selectivity of this recruitment cascade is illustrated in (B); L-selectin negative cells (e.g., effector memory T cells) and CCR7 negative leukocytes (e.g., granulocytes) fail to undergo each of the prerequisite steps. Peyer’s patches display different “ZIP codes” in HEV segments within the interfollicular T cell area (C) and in upstream B

High endothelial venules

97

follicles (D). MAdCAM-1 is expressed in all PP HEVs and mediates rolling interactions with both L-selectin and α4β7 integrins. For step 2, HEVs in different areas display distinct chemokine patterns: T cells are activated by CCL21, while in HEVs associated with B follicles present CXCL12 and CXCL13 to stimulate B cell-expressed CXCR4 and CXCR5, respectively. Firm arrest of both B and T cells in PPs is thought to be mediated by activated α4ß7 and/or LFA-1 integrins. hand, CXCL12/CXCR4 has a modest role in naïve T cell homing in C57BL/6 mice, indicating that some chemokine pathways may be regulated in a strain-dependent manner (52). Despite the ability of B cells to respond to two distinct integrin activation signals in HEVs, homing of B cells to LNs is significantly less efficient than that of T cells. This is because the L-selectin levels on B cells are ~50% lower than on T cells (66). T cells from L-selectin+/− heterozygous mice express similar L-selectin levels as L-selectin+/+ B cells, and these two lymphocyte populations home equivalently to LNs (66). In experiments with pre-B cell clones stably transfected with human L-selectin, we observed that at least 50000 L-selectin molecules/cell were needed for efficient rolling in LN HEVs (21,67 and unpublished data). Thus, it is important to keep in mind that the mere presence of Lselectin on a leukocyte does not necessarily predict its potential to home to LNs, even when this cell expresses all other prerequisite traffic molecules (i.e., CCR7 and LFA-1). A leukocyte may be deemed L-selectin+ based on flow cytometric criteria, but this would be of little consequence if the expressed copy number is substantially lower than that on naïve T cells (~70 000–100 000 molecules/cell). The fact that a sequence of three distinct molecular steps must be successfully engaged in HEVs explains why some leukocytes home to PLNs, whereas others do not. For example, granulocytes express L-selectin and LFA-1, but not CCR7 or CXCR4; mature myeloid DCs express CCR7 and LFA-1, but not L-selectin; and effector CTL lose both L-selectin and CCR7. Consequently, granulocytes roll, but fail to arrest, whereas DCs and effector cells cannot tether or roll in LN HEVs (49,51,68) (Fig. 4B). L-selectin-independent homing via HEVs can be induced by i.v. injections of activated platelets (69,70). Circulating activated platelets express P-selectin on their surface, which mediates platelet binding to PNAd in HEVs and, simultaneously, to PSGL-1 on circulating leukocytes (69). This platelet bridge can transiently restore lymphocyte homing in L-selectin deficient mice. Indeed, L-selectin deficient animals mount poor cutaneous hypersensitivity (CHS) responses when they are sensitized by painting of a hapten antigen on the skin because naïve T cells do not migrate into skin draining LNs

Endothelial cells in health and disease

98

(71). When the homing cascade is restored during the sensitization phase by infusing activated platelets for several hours, the CHS response is restored (70). Of note, a small fraction of leukocytes roll in a P-selectin-dependent manner in PLNs HEVs of L-selectindeficient mice even when no platelets are being infused (70). However, it is not yet clear if this minor rolling pathway contributes to physiologic L-selectin-independent homing and whether the relevant source of P-selectin in this setting are circulating activated platelets or endothelial cells or both.

5.2. The Multi-Step Adhesion Cascade in PP and MLN The homing cascade discussed above applies primarily to skin-draining LNs. High endothelial venules in mucosa-associated LNs, such as MLNs, express not only PNAd, but also the mucosal addressin cell adhesion molecule (MAdCAM)-1, a ligand for the α4β7 integrin (17,72). The α4β7/MAdCAM-1 pathway can mediate selectin-independent lymphocyte tethering and rolling in HEVs (73). Thus, L-selectin deficiency compromises lymphocyte homing to PLNs more severely than to MLNs (74), whereas β7 integrin deficiency results in reduced homing to MLNs, but has no effect in PLNs (75). Payer’s patches HEVs express only MAdCAM-1, but not PNAd, on their luminal surface however, there is MECA-79-reactive material at the abluminal side of PP HEV (18). The binding site for α4β7 resides within the first Ig domain of MAdCAM-1 (76). In addition, MAdCAM-1 contains a mucin domain that can be decorated with L-selectin ligands. Indeed, affinity-purified MAdCAM-1 from MLNs is recognized by the mAb MECA-79 (77). By contrast, MAdCAM-1 in PP HEVs is not detected by MECA-79, but nevertheless supports lymphocyte tethering and rolling via both L-selectin and the α4β7 integrin (73). When α4β7 becomes functionally activated, it can also mediate firm arrest (78). However, naïve lymphocytes express relatively little α4β7 (79) and require additional engagement of LFA-1 for firm arrest, while α4β7 is primarily critical to slow the rolling cells (73) (Fig. 4C). Intravital microscopy experiments have shown that the integrin-activating chemokines in PP HEVs are segmentally presented (52,80). In B follicles, HEVs present CXCL12 and CXCL13, the ligands for CXCR4 and CXCR5, respectively (Fig. 4D). Signals through these two receptors induce integrin activation on rolling B cells, but not T cells (52). Conversely, as soon as HEVs enter into the interfollicular T cell area, they express CCL21 and promote preferential T cell arrest (80). Thus, HEVs are not only distinct between different lymphoid tissues, but there is even segmental specialization within individual microvessels. The factors that orchestrate this remarkable endothelial subspecialization remain to be identified.

5.3. Remote Control An additional mechanism for leukocyte recruitment to PLNs operates during inflammation (50,81). Chemokines, such as CCL2 (MCP-1), produced at a peripheral site of inflammation are drained via the lymph to the subcapsular sinus. Here they enter the FRC conduit and are channeled toward HEVs in the cortex (42). Chemokines that reach

High endothelial venules

99

the abluminal surface of HEVs are then transcytosed within vesicles to the luminal side and presented to rolling leukocytes (12,81). The nature of the chemokine transport vesicles in HEVs has not been determined, but might involve caveolae, which have been shown to mediate chemokine transcytosis in dermal microvessels (82). This mechanism, termed “remote control,” allows the rapid recruitment of circulating monocytes, which express CCR2, the receptor for CCL2. Monocytes do not express CCR7 or CXCR4, and are therefore excluded from resting PLNs. However, by discharging CCL2 into the lymph, inflamed peripheral tissues project a potent monocyte chemoattractant signal onto HEVs in draining LNs, which triggers integrin activation (81). Once in the lymph node, monocytes may differentiate into macrophages or dendritic cells to participate in the ensuing immune response. In addition, inflammation induces mRNA for CXCL9 (MIG) in draining LNs and presentation of this chemokine on a subset of HEVs, which support monocyte adhesion in vitro (83). It is not known whether CXCL9 is produced by the inflamed HEVs themselves or by other intra- (or extra-) nodal cells from where the chemokine might have been transported to the HEVs. The ability to remotely modulate the multi-step adhesion cascade in HEVs by discharging chemokines into the lymph enables peripheral tissues to control the composition and function of leukocytes in draining LNs. However, recent work suggests that there may be also counter-regulatory mechanisms. Lymphatic endothelial cells express the non-signaling serpentine receptor D6, which binds promiscuously to a number of chemokines, including CCL2 (84). Upon ligation, D6 triggers the rapid internalization and degradation of its cargo and may thus have a role as a gatekeeper to avoid uncontrolled seepage of chemokines into the lymph (85). Interestingly, HEVs express high levels of a similar non-signaling chemokine receptor, the Duffy antigenrelated receptor for chemokines (DARC) (86). It seems plausible that DARC, too, has a function in the transportation, presentation, or turnover of chemokines in HEVs. However, the mechanisms that control the expression and function of either D6 or DARC are still poorly understood.

5.4. Transendothelial Migration The last step of the adhesion cascade, extravasation, is also the least understood at the molecular level. Adhesion molecules that play a role in transendothelial migration of inflammatory cells in non-lymphoid tissues, such as PECAM-1 (CD31) and CD99, have not been shown to contribute to lymphocyte diapedesis across HEVs. One candidate molecule is JAM-C, an immunoglobulin superfamily member that is concentrated at intercellular junctions between HECs (87) and forms homoand heterotypic interactions with other JAM family members (88). JAM-B (VE-JAM) is also found at intercellular boundaries in HEVs and normal endothelium (89), and interacts with leukocytes (88), but its role in transmigration has not been demonstrated. The reader is referred to recent reviews for further details (90–92).

Endothelial cells in health and disease

100

6. HEV-SPECIFIC LYMPHOCYTE TRAFFIC MOLECULES 6.1. The MECA-79 Antigen in PLNs HEVs MAb MECA-79 immunoprecipitates a set of glycans on the surface of HEVs in both murine and human lymphoid tissues (18,53,93). Affinity-purified MECA-79-reactive glycoproteins support L-selectin binding in vitro (53), and MECA-79 blocks lymphocyte recruitment to PLNs (18,94) as well as rolling on HEVs in vivo (21,95). Immunohistochemical staining of tissue sections with MECA-79 reveals reactive material on both the lumenal and abluminal surface of PLNs HEVs (Fig. 2C), but only on the abluminal side of PP HEVs (18,96). The physiological role of abluminal MECA-79 antigen is not understood, but it is clear that it is biosynthetically distinct from the lumenal material (96). Intravital microscopy studies of the localization of fluorescent MECA-79 or MECA-79-conjugated fluorescent latex beads have shown that MECA-79 antigen is highly restricted to cortical HEVs within the PLNs vasculature (97,98) (Fig. 2A, B). In the HEV lumen, MECA-79 localizes to microvillouslike protrusions, which may facilitate L-selectin-dependent tethering of circulating lymphocytes (86). Several sialomucins have been identified that can be decorated with the MECA-79 epitope. These include GlyCAM-1 (Sgp50), a secreted glycoprotein (99,100); CD34 (Sgp90) (101); podocalyxin (102); and one (or more) large molecular species termed Sgp200, which remain(s) to be identified (103). Two additional proteins that may support lymphocyte recruitment to PLNs include endomucin (104) and endoglycan (105,106). Endomucin can carry the MECA-79 epitope and has been detected on HEVs, whereas endoglycan functions as an L-selectin ligand through tyrosine sulfation and glycosylation, but is not recognized by MECA-79. As far as we can tell, these glycoproteins probably serve as interchangeable scaffolds that are functionally redundant, at least in the context of L-selectin ligand presentation. For example, CD34 is a major component detected by MECA-79 in human tonsils (107), but CD34 deficient mice have normal lymphocyte migration to lymphoid tissues (Table 1) (108). GlyCAM-1 expression is restricted to HEVs and mammary epithelial cells (109), podocalyxin was originally described in the kidney, and CD34 is expressed on many different cells types including non-HEV endothelium and cells of hematopoietic origin (110). However, on non-HEV cell types, none of these molecules react with MECA-79 or support L-selectin binding. Indeed, L-selectin ligand presenting sialomucins undergo extensive HEV-specific post-translational modifications that are essential for their function in lymphocyte migration.

6.2. Post-translational Modification of HEV Proteins The importance of HEV-specific carbohydrate modifications for lymphocyte trafficking was first suggested by observations that specific mono- and polysaccharides

High endothelial venules

101

Table 1 Effect of Disrupting Genes Involved in Lymphocyte Recruitment to Lymphoid Organs and Expressed by HEV Gene

Leukocyte Trafficking in Knockout

Reference(s)

L-selectin (CD62L)

Hypocellular PLNs and impaired lymphocyte trafficking. Impaired DTH responses.

74,163

CD34

Defects in early hematopoietic progenitor development. No defects in lymphocyte trafficking.

108

CD43

Increased lymphocyte migration to secondary lymphoid organs.

97

CD44

Delayed or subtle defect in entry of CD44 deficient lymphocytes to PLNs.

164,165

LFA-1 (CD11aCD18/αLβ2)

Impaired lymphocyte trafficking to PLNs, MLNs, and PPs.

55,56

β7 Integrin

Impaired lymphocyte trafficking to MLNs and PPs and impaired adhesion to PP HEVs.

75

ICAM-1 (CD54)

Defects in DTH response and neutrophil recruitment 166,167 to inflammatory sites. No defect in lymphocyte recruitment to lymphoid organs.

ICAM-2 (CD 102)

No defect in lymphocyte recruitment to lymphoid organs.

168

CCR7

Impaired lymphocyte and DC trafficking to PLNs and PPs.

62

CCL21ser/CCL19(plt/plt mice)

Impaired T lymphocyte trafficking to PLNs and PPs. 50,52,80,169

CXCR5 CXCL13 (BLC)

Impaired B lymphocyte homing and migration to B follicles in Peyer’s patches and spleen.

52,170

CXCR4 CXCL12 (SDF1α)

Impaired T and B lymphocyte migration to lymphoid organs in the absence of CCR7 ligands.

52

DARC

Potential defect in inflammatory chemokine transport or presentation on endothelium.

135,171

Knockout Phenotype Lymphocyte Homing Enzymes

MECA-79 L-selectin IgM Staining of HEV Staining of Reference(s) HEV

C2GlcNAcT-Ia

Normal/decreased

Normal

Absent

127,127a,131.

HEC-GlcNAc6ST 50%

Reduction

Abluminal

Abluminal

95,96

Endothelial cells in health and disease

FucT-VII FucT-IV FucT-VII/FucT-IV

80% Reduction Increased

staining only

staining only

Normal

Absent

119

b

Normal

98,120

b

b

NA

>95% Reduction

102

NA

NA

120

a

Abbreviations: C2GlcNAcT-I, Core2 β-1,6-N-acetylglucosaminyltransferase-I; HECGlcNAc6ST, HEC- GlcNAc-6-sulfotransferase; FucT-VII, α(1,3)fucosyltransferase-VII; FucT-IV, α(1,3)fucosyltransferase-IV. b NA: Data not available. inhibit lymphocyte binding to LN HEVs (111–113). After it was discovered that Lselectin functions as a lectin-like adhesion molecule that binds the HEV-specific MECA79 antigen (53,93), detailed biochemical studies of affinity-purified MECA-79-reactive material (or PNAd) were performed (114–116). Today, we know that most HEVexpressed L-selectin ligands are sialylated, fucosylated and sulfated O-linked carbohydrates that decorate a select group of sialomucins in HEVs (Table 2).

6.2.1 Sialic Acid Rosen et al. (117) demonstrated that removal of sialic acid from PLNs sections with sialidase blocked lymphocyte binding to HEVs, and intravenous injection of sialidase inhibited lymphocyte homing to PLNs (19). Although the mAb MECA-79 does not require the presence of sialic acid for binding, the HEV-expressed glycoproteins that react with MECA-79 must be modified with sialic acid to function as L-selectin ligands (118). Indeed, the major capping group that binds L-selectin is a sulfated form of sialyl LewisX (sLeX) (i.e., Siaα2,3Galβ1,4-[Fucα1,3]GlcNAc) (115,116). Sialic acid is not essential for lymphocyte migration to PPs, since treatment of PP HEVs with sialidase does not affect lymphocyte binding or homing to PPs (19,117). However, treatment of purified MAdCAM-1 with sialidase blocked in vitro L-selectin binding to MAdCAM-1 (77), suggesting that the interaction of α4β7 integrins with MAdCAM-1 does not depend upon sialylation, while L-selectin binding to MAdCAM-1 requires it. Which of the many α2,3 sialyltransferases is responsible for the biosynthesis of L-selectin ligands in PLNs HEVs remains to be determined.

6.2.2. α(1,3)-Fucosylation The presence of α(1,3)-linked fucose is essential for the activity of virtually all physiologic selectin ligands, including those in HEVs. The two fucosyltransferases (FucT) implicated in the generation of selectin ligands on leukocytes and endothelial cells are FucT-IV and FucT-VII (Table 1) (119,120). FucT-VII deficient mice have ~80% reduced lymphocyte homing to PLNs (119). FucT-IV-VII double knockout mice have a significantly more complete defect in lymphocyte homing to PLNs than FucT-VII deficient mice, indicating that both FucTs can generate L-selectin ligands in HEVs (120). However, FucT-IV deficient mice have moderately increased lymphocyte homing to PLNs and MLNs (98).

High endothelial venules

103

To explain this counter-intuitive observation, one must consider the acceptor preferences of FucT-IV and FucT-VII (121,122). FucT-VII requires an α(2,3)-sialylated terminal lactosamine to generate sLeX-like selectin ligands. FucT-IV can also use the sialylated lactosamine substrate, albeit less efficiently than FucT-VII. However, unlike FucT-VII, FucT-IV additionally fucosylates internal GlcNAc residues in the polylactosamine (121) and uses a non-sialylated lactosamine to generate LewisX-like structures (120), which do not interact with selectins. Thus, FucT-IV may compete with α(2,3)-sialyltransferase(s) and/or FucT-VII for terminal lactosamine as a shared precursor substrate on O-linked glycans. Consequently, FucT-IV may divert some terminal lactosamine acceptor moieties toward the LewisX pathway that does not yield L-selectin ligands, and thus away from the α(2,3)sialyltransferase/FucT-VII synthetic route. In wildtype mice, FucT-IV thus attenuates FucT-VII-dependent production of sLeX-based L-selectin ligands. Without FucT-IV, increased acceptor availability for α(2,3)sialyltransferase/FucT-VII permits increased selectin ligand production, whereas FucT-VII deficiency has a lesser impact, since sLeX-based L-selectin ligands could still be generated if FucT-IV is abundantly expressed.

Table 2 Structure of Sialyl Lewisx Derivatives and Enzymatic Modification Carried Out by HEV Expressed Enzymes

Enzyme C1GlCNACT

Enzyme reaction a

C2GlcNAcT-I

HECGlcNAc6ST

Endothelial cells in health and disease

104

FucT-VII (FucT-IV)

FucT-IVb

Biantennary glycans arise from a corel branch at R1 or a core2 branch at R2. * 6-Sulfated corel glycans are sufficient to generate the MECA-79 epitope found on the lumenal surface of HEVs. a Abbreviations: C1G1cNAcT, Corel-β1,3-N-acetylglucosaminyltransferase; C2GlcNAcT-I, Core2 β-1, 6-N-acetylglucosaminyltransferase-I; HEC-GlcNAc6ST, HEC-GlcNAc-6-sulfotransferase; FucT-VII, α(1,3)fucosyltransferase-VII; FucT-IV, α(1,3)fucosyltransferase-IV. b Unlike FucT-VII, FucT-IV fucosylates non-sialyated carbohydrates and internal GlcNAc residues on lactosamines. In addition, FucT-IV fucosylates glycolipids (not depicted) (172,173). Experimental evidence in support of this scenario has been provided recently (98). This work has revealed additional complexity, because FucT-IV and FucT-VII are not uniformly expressed within the venular tree in PLNs. The larger collecting venules in the medulla of mouse subiliac LNs express a functional L-selectin ligand that is spatially and antigenically distinguishable from MECA-79-reactive PNAd (98), which is restricted to para-and subcortical HEVs and requires primarily FucT-VII (97). By contrast, the Lselectin ligand(s) in medullary venules are MECA-79⎯ and are regulated largely by FucTIV, which is more highly expressed in this microvascular bed than elsewhere in PLNs.

6.2.3. Sulfation The critical role of sulfation of HEV proteins was first suggested by the unique ability of HEVs amongst vascular endothelium to incorporate large amounts of sulfate (46). Indeed, metabolic inhibition of sulfation reduces L-selectin ligand activity in PLNs (114). There are two sulfated sLeX moieties in PLNs, one containing Gal-6-SO4 (6′-sulfo Lewisx) and the other containing GlcNAc-6-SO4 (6-sulfo Lewisx) (Table 2) (116), but only 6-sulfation of sLeX is thought to enhance L-selectin binding (123). The GlcNAc-6O-sulfotransferases responsible for the generation of L-selectin ligands in HEVs are GlcNAc6ST (124) and HEC-GlcNAc6ST/LSST (125,126). Mice deficient in HECGlcNAc6ST/LSST have hypocellular PLNs, impaired (~50%) lymphocyte homing to PLNs, reduced L-selectin binding to HEVs, and a near absence of MECA-79-reactive

High endothelial venules

105

material in the lumen of HEVs (96,127). The latter is not surprising, since the epitope recognized by MECA-79 requires 6-sulfation of sLex (103,128). However, in HEVs of HEC-GlcNAc6ST knockout mice, lumenal MECA-79 antigen is absent, yet lymphocytes still roll in an L-selectin-dependent manner, but at a higher velocity (95,96). These findings indicate that some of the carbohydrates recognized by L-selectin are distinct from that recognized by MECA-79. Moreover, abluminal MECA-79 staining is preserved around HEVs of HEC-GlcNAc6ST knockout mice, suggesting that additional sulfotransferases are involved.

6.2.4. O-linked Carbohydrates O-linked glycans in a core-2 linkage (i.e., G1cNAcβ1,6[Galβ1,3]-GalNac) represent a major component of L-selectin ligands in HEVs (129) (Table 2). Thus far, only the contribution of core2 β-1,6-N-acetylglucosaminyltransferase (C2GlcNAcT-I) (130) and corel-β1,3-N-acetylglucosaminyltransferase (C1G1cNAcT) (128) to the generation of Lselectin ligands has been studied. While C2GlcNAcT-I was found to be critical for the generation of P- and E-selectin ligands, no defect was observed in lymphocyte trafficking to peripheral lymphoid organs of mice deficient for this enzyme (131), although more recent data suggest that these animals have subtle defects in lymphocyte trafficking to PLNs (127,127a) (Table 1). Residual L-selectin ligand activity in C2GlcNAcT-1 deficient mice may be due to additional core2 β-1,6-N-acetylglucosaminyltransferases and/or the activity of C1G1cNAcT, which has been shown in vitro to be critical in the generation of MECA-79-reactive glycans and contribute to the synthesis of L-selectin ligands (128) (Table 2).

6.3. MAdCAM-1 MAdCAM-1 has two distal regions homologous to the immunoglobulin (Ig) domains found on other adhesion molecules of the immunoglobulin superfamily, including ICAM1 and VCAM-1 (72,76), and a mucin-like region rich in serine and threonine, which can bear O-linked carbohydrate modifications necessary for L-selectin binding (132). MECA79 can immunoprecipitate MAdCAM-1 from MLNs, but not PPs, and MAdCAM-1 from MLNs supports L-selectin binding in vitro (77). However, MAd-CAM-1 in PP HEVs can also present L-selectin ligands (that are presumably not recognized by MECA-79) and support L-selectin tethering and rolling in vivo (73). Therefore, MAdCAM-1 represents a unique multi-functional adhesion molecule capable of interacting with both α4β7 integrin and L-selectin.

6.4. Chemokine Presentation on HEV Chemokines presented on the endothelial surface induce integrin activation and firm arrest of rolling leukocytes. The immobilization and display of chemokines are thought to be mediated by glycosylaminoglycans (GAGs). Glycosylaminoglycans are proteoglycans that carry negatively charged sulfate and carboxyl groups, which permits electrostatic interactions with basic peptide motifs present in most chemokines (133). Heparin and

Endothelial cells in health and disease

106

heparan sulfate are highly expressed on endothelial cells and are the likely candidates to bind and present chemokines. The Duffy antigen-related chemokine receptor (DARC) is a non-signaling seventransmembrane receptor that binds both CC and CXC chemokines. The DARC is highly expressed in HEVs (86) and has been suggested to play a role in chemokine transport. The DARC has been identified in endothelial caveolae (134), which have been shown to function as a conduit for chemokine translocation from the abluminal to the lumenal surface of venules (82) (Fig. 3). Studies in knockout mice suggest that DARC does not appear to play a role in regulating constitutive lymphocyte trafficking to PLNs (135). However, DARC can scavenge inflammatory chemokines including CCL2, which suggests a role for DARC in inflammation, possibly during “remote control” of monocyte recruitment to PLNs (see above).

7. PLASTICITY OF ENDOTHELIUM AND HEV 7.1. Perinatal Switch In the developing mouse, the main addressin expressed in both LN and PP HEVs is MAdCAM-1 (136). It is thought that a population of CD4+CD3−IL7Rαhi cells is required to provide the signals necessary for the further development of LNs and PPs (137). These rare cells (1–2% of total PBLs in the mouse fetus) express the α4β7 integrin and must bind to MAdCAM-1 to migrate to LNs (136).On the other hand, studies in rats found almost no HEVs in LNs prenatally (138). Endothelial cells in cortical vessels acquired a cuboid morphology only gradually after birth, indicating that HEVs commitment may be differentially regulated in different species.

7.2. Lymph-borne HEV Differentiation Factors Once developed, the mature HEVs phenotype requires constant maintenance signals. Hendriks et al. (139) reported that HEVs from PLNs that had undergone occlusion of their afferent lymphatics became flat and lost their characteristic morphology. Subsequent studies showed that lymphocytes do not adhere to the lymph-deprived HEVs (140), and genes that contribute to the generation of L-selectin ligands are turned off (141,142). The nature of the HEV-sustaining lymph-borne signal is still mysterious. Lymph flow per se is probably not essential, since HEVs are also found in PPs, which do not possess afferent lymphatics.

7.3. Ectopic HEV Normal endothelium has the potential to differentiate into HEVs when given the appropriate stimuli. During chronic inflammation, such as in certain infections and autoimmune diseases, the cellular infiltrate organizes itself into lymph node-like

High endothelial venules

107

structures containing naïve T cells, B cell follicles and dendritic cells, in a process termed “lymphoid neogenesis” (reviewed in Refs. 143–145). Post-capillary venules in these aggregates assume an HEV-like morphology, express PNAd and/or MAd-CAM-1 as well as CCL21, and recruit naïve T cells from the blood (65). The ectopic expression of CCL19, CCL21 or CXCL13 is sufficient to cause lymphoid neogenesis, at least in certain organs such as the pancreas (146–148). How the recruitment of naïve lymphocytes induces the differentiation of ectopic HEVs is not yet clear. A likely player is lymphotoxin (LTα), a member of the TNF family that is expressed by activated CD4 and CD8T cells (149), and also by the above mentioned CD4+CD3−IL7Rαhi cells that seed nascent LNs. Indeed LTα and its receptors, coordinate the genesis of secondary lymphoid organs (150,151), and expression of LTα1β2 is sufficient and necessary to induce lymphoid neogenesis (145,147).

8. CONCLUSION AND FUTURE DIRECTIONS 8.1. Unanswered Questions Although the past two decades have seen considerable progress in our understanding of the functions and molecular and biochemical underpinnings of HEVs, many fundamental questions remain to be answered. What does it take to make an HEV? As we have seen, factors in lymph and at sites of chronic inflammation are required to maintain HEVs and to turn regular endothelial cells into HECs, respectively. What are the mechanisms that induce and sustain this conspicuous differentiation? Can we interfere with these signals for therapeutic purposes? Can we use them to grow and study HECs in vitro? How many different subtypes of HEVs are there and what are the signals and the transcriptional programs that control their distinct phenotype? Another gray area is the question of how lymphocyte transmigration across HEVs is regulated. Is leukocyte recruitment also regulated at this last step of the adhesion cascade, or is diapedesis a default process? To answer these questions, novel technologies have begun to be employed together with the established techniques. We will briefly discuss some of these promising new approaches in this last section.

8.2. Genomics The unique features of HEVs likely derive from a specific set of expressed genes. Since bona fide HECs cannot be grown in vitro, material for expression profiling must rely on the cumbersome purification of relatively few cells from single-cell suspensions of mammalian lymphoid tissues. For example, quantitative 3′-cDNA libraries have been generated for both murine PLNs (152) and PPs (153) HEVs. These studies have identified several known and unknown genes that are expressed in HEVs, but not in flat endothelial cells. Human tonsil HEVs have been compared to other human endothelial cells by subtractive hybridization and similar molecular techniques (86,154,155). This work has identified several HEV-specific surface molecules, enzymes, molecules involved in sulfate metabolism, and a nuclear transcription factor, NF-HEV, which is

Endothelial cells in health and disease

108

much more abundantly expressed in HEVs compared to other vascular beds (156). It remains to be determined whether and to what extent NF-HEV is necessary and/or sufficient to induce or regulate HEV function or phenotype.

8.3. Multi-Photon Microscopy Intravital microscopy has been useful to dissect the multi-step adhesion cascades for lymphocyte homing via HEVs (43). However, traditional epifluorescence-based intravital microscopy cannot generate three-dimensional images, which restricts its usefulnes for studies of post-adhesion events, such as diapedesis or migration within the parenchyma. Multi-photon microscopy (157) is beginning to revolutionize this field by allowing investigators to generate optical section deep within intact lymphoid tissues (158). Recent studies have analyzed the behavior of lymphocytes in intact, freshly excised LNs (159– 161); and, more recently, in intravital settings using anesthetized mice (23,162).

REFERENCES 1. Marchesi VT, Gowans JL. The migration of lymphocytes through the endothelium of venules in lymph nodes: an electron microscope study. Proc R Soc 1964; B 159:283–290. 2. Girard J-P, Springer TA. High endothelial venules (HEVs): Specialized endothelium for lymphocyte migration. Immunol Today 1995; 16:449–457. 3. Springer TA. Traffic signals for lymphocyte recirculation and leukocyte emigration: The multistep paradigm. Cell 1994; 76:301–314. 4. von Andrian UH, Mackay CR. T-cell function and migration. Two sides of the same coin. N Engl J Med 2000; 343:1020–1034. 5. Thomé R. Endothelien als Phagocyten (aus den Lymphdrüsen von Macacus cynomolgous). Archives Mikrosk Anat 1898; 52:820–842. 6. von Schumacher S. Über Phagocytose und die Abfuhrwege der Leukocyten in den Lymphdrüsen. Arch Mikroskop Anat 1899; 54. 7. Gowans JL, Knight EJ. The route of re-circulation of lymphocytes in the rat. Proc R Soc Lond B 1964; 159:257–282. 8. Hall JG, Morris B. The origin of the cells in the efferent lymph from a single lymph node. J Exp Med 1965; 121:901–911. 9. Anderson AO, Anderson ND. Lymphocyte emigration from high endothelial venules in rat lymph nodes. Immunology 1976; 31:731–748. 10. Pabst R, Binns RM. Heterogeneity of lymphocyte homing physiology: several mechanisms operate in the control of migration to lymphoid and non-lymphoid organs in vivo. Immunol Rev 1989; 108:83–109. 11. Claesson MH, Jorgensen O, Ropke C. Light and electron microscopic studies of the paracortical post-capillary high-endothelial venules. Z Zellforsch Mikrosk Anat 1971; 119:195–207. 12. Baekkevold ES, Yamanaka T, Palframan RT, Carlsen HS, Reinholt FP, von Andrian UH, Brandtzaeg P, Haraldsen G. The CCR7 ligand ELC (CCL19) is transcytosed in high endothelial venules and mediates T cell recruitment. J Exp Med 2001; 193:1105–1112. 13. Stamper HB Jr, Woodruff JJ. Lymphocyte homing into lymph nodes: in vitro demonstration of the selective affinity of recirculating lymphocytes for high-endothelial venules. J Exp Med 1976; 144:828.

High endothelial venules

109

14. Butcher EC, Scollay RG, Weissman IL. Organ specificity of lymphocyte migration: mediation by highly selective lymphocyte interaction with organ-specific determinants on high endothelial venules. Eur J Immunol 1980; 10:556–561. 15. Woodruff JJ, Katz IM, Lucas LE, Stamper HB Jr. An in vitro model of lymphocyte homing II. Membrane and cytoplasmic events involved in lymphocyte adherence to specialized highendothelial venules of lymph nodes. J Immunol 1977; 119:1603–1610. 16. Gallatin WM, Weissman IL, Butcher EC. A cell-surface molecule involved in organ-specific homing of lymphocytes. Nature 1983; 304:30–34. 17. Streeter PR, Lakey-Berg E, Rouse BTN, Bargatze RF, Butcher EC. A tissue-specific endothelial cell molecule involved in lymphocyte homing. Nature 1988; 331:41–46. 18. Streeter PR, Rouse BTN, Butcher EC. Immunohistologic and functional characterization of a vascular addressin involved in lymphocyte homing into peripheral lymph nodes. J Cell Biol 1988; 107:1853–1862. 19. Rosen SD, Chi SI, True DD, Singer MS, Yednock TA. Intraveneously injected sialidase inactivates attachment sites for lymphocytes on high endothelial venules. J Immunol 1989; 142:1895–1902. 20. Bjerknes M, Cheng H, Ottaway CA. Dynamics of lymphocyte-endothelial interactions in vivo. Science 1986; 231:402–405. 21. von Andrian UH. Intravital microscopy of the peripheral lymph node microcirculation in mice. Microcirculation 1996; 3:287–300. 22. von Andrian UH, M’Rini C. In situ analysis of lymphocyte migration to lymph nodes. Cell Adhes Commun 1998; 6:85–96. 23. Mempel TR, Henrickson SE, von Andrian UH. T cell priming by dendritic cells in lymph nodes occurs in three distinct phases. Nature 2004; 427:154–159. 24. Nagira M, Imai T, Yoshida R, Takagi S, Iwasaki M, Baba M, Tabira Y, Akagi J, Nomiyama H, Yoshie O. A lymphocyte-specific CC chemokine, secondary lymphoid tissue chemokine (SLC), is a highly efficient chemoattractant for B cells and activated T cells. Eur J Immunol 1998; 28:1516–1523. 25. Ager A. Isolation and culture of high endothelial cells from rat lymph nodes. J Cell Sci 1987; 87:133–144. 26. Ise Y, Yamaguchi K, Sato K, Yamamura Y, Kitamura F, Tamatani T, Miyasaka M. Molecular mechanisms underlying lymphocyte recirculation: I. Functional, phenotypical and morphological characterization of high endothelial cells cultured in vitro. Eur J Immunol 1988; 18:1235–1244. 27. Ager A, Mistry S. Interaction between lymphocytes and cultured high endothelial cells: an in vitro model of lymphocyte migration across high endothelial venule endothelium. Eur J Immunol 1990; 18:1265–1274. 28. Derry CJ, Faveeuw C, Mordsley KR, Ager A. Novel chondroitin sulfate-modified ligands for L-selectin on lymph node high endothelial venules. Eur J Immunol 1999; 29:419−430. 29. Baekkevold ES, Jahnsen FL, Johansen FE, Bakke O, Gaudernack G, Brandtzaeg P, Haraldsen G. Culture characterization of differentiated high endothelial venule cells from human tonsils. Lab Invest 1999; 79:327–336. 30. Gretz JE, Anderson AO, Shaw S. Cords, channels, corridors and conduits: Critical artichitectural elements facilitating cell interactions in the lymph node cortex. Immunol Rev 1997; 156:11–24. 31. Witmer MD, Steinman RM. The anatomy of peripheral lymphoid organs with emphasis on accessory cells: light-microscopic immunocytochemical studies of mouse spleen, lymph node, and Peyer’s patch. Am J Anat 1984; 170:465–481. 32. Krall WJ, Braun J. In situ lacZ retrovirus-marked lymphocytes define a B cell microenvironment in the lymph node medulla. New Biol 1992; 4:581–590.

Endothelial cells in health and disease

110

33. Bockman DE, Stevens W. Gut-associated lymphoepithelial tissue: bidirectional transport of tracer by specialized epithelial cells associated with lymphoid follicles. J Reticuloendothel Soc 1977; 21:245–254. 34. Mowat AM. Anatomical basis of tolerance and immunity to intestinal antigens. Nat Rev Immunol 2003; 3:331–341. 35. De Bruyn PPH, Cho Y. Structure and function of high endothelial postcapillary venules in lymphocyte recirculation. In: Grundman E, Vollmer T, eds. Reaction Patterns of the Lymph Node. Berlin: Springer-Verlag, 1990:85–103. 36. Sainte-Marie G, Peng FS. High endothelial venules of the rat lymph node A review and a question: is their activity antigen specific? Anat Rec 1996; 245:593–620. 37. Yamaguchi K, Schoefl GI. Blood vessels of the Peyer’s patch in the mouse: I. Topographic studies. Anat Rec 1983; 206:391–401. 38. Farr AG, Cho Y, De Bruyn PP. The structure of the sinus wall of the lymph node relative to its endocytic properties and transmural cell passage. Am J Anat 1980; 157: 265–284. 39. Bajenoff M, Granjeaud S, Guerder S. The strategy of T cell antigen-presenting cell encounter in antigen-draining lymph nodes revealed by imaging of initial T cell activation. J Exp Med 2003; 198:715–724. 40. Anderson AO, Anderson ND. Studies on the structure and permeability of the microvasculature in normal rat lymph nodes. Am J Pathol 1975; 30:387–418. 41. Clark SL Jr. The reticulum of lymph nodes in mice studied with the electron microscope. Am J Anat 1962; 110:217–258. 42. Gretz JE, Norbury CC, Anderson AO, Proudfoot AE, Shaw S. Lymph-borne chemokines and other low molecular weight molecules reach high endothelial venules via specialized conduits while a functional barrier limits access to the lymphocyte microenvironments in lymph node cortex. J Exp Med 2000; 192:1425–1440. 43. von Andrian UH, Mempel TR. Homing and cellular traffic in lymph nodes. Nat Rev Immunol 2003; 3:867–878. 44. Anderson ND, Anderson AO, Wyllie RG. Specialized structure and metabolic activities of high endothelial venules in rat lymphatic tissues. Immunology 1976; 31:455–473. 45. Freemont AJ, Jones CJP. Light microscopic, histochemical and ultrastructural studies of human lymph node paracortical venules. J Anat 1983; 136:349–362. 46. Andrews P, Milsom DW, Ford WL. Migration of lymphocytes across specialized vascular endothelium V. Production of a sulphated macromolecule by high endothelial cells in lymph nodes. J Cell Sci 1982; 57:277–292. 47. Anderson AO, Shaw S. T cell adhesion to endothelium: the FRC conduit system and other anatomic and molecular features which facilitate the adhesion cascade in lymph node. Semin Immunol 1993; 5:271–282. 48. Butcher EC, Picker LJ. Lymphocyte homing and homeostasis. Science 1996; 272:60–66. 49. Warnock RA, Askari S, Butcher EC, von Andrian UH. Molecular mechanisms of lymphocyte homing to peripheral lymph nodes. J Exp Med 1998; 187:205–216. 50. Stein JV, Rot A, Luo Y, Narasimhaswamy M, Nakano H, Gunn MD, Matsuzawa A, Quackenbush EJ, Dorf ME, von Andrian UH. The CC chemokine thymus-derived chemotactic agent 4 (TCA-4, secondary lymphoid tissue chemokine, 6Ckine, exodus-2) triggers lymphocyte function-associated antigen 1-mediated arrest of rolling T lymphocytes in peripheral lymph node high endothelial venules. J Exp Med 2000; 191:61–76. 51. Weninger W, Crowley MA, Manjunath N, von Andrian UH. Migratory properties of naive, effector, and memory CD8(+) T cells. J Exp Med 2001; 194:953–966. 52. Okada T, Ngo VN, Ekland EH, Forster R, Lipp M, Littman DR, Cyster JG. Chemokine requirements for B cell entry to lymph nodes and Peyer’s patches. J Exp Med 2002; 196:65–75. 53. Berg EL, Robinson MK, Warnock RA, Butcher EC. The human peripheral lymph node vascular addressin is a ligand for LECAM-1, the peripheral lymph node homing receptor. J Cell Biol 1991; 114:343–349.

High endothelial venules

111

54. Hamann A, Westrich DJ, Duijevstijn A, Butcher EC, Baisch H, Harder R, Thiele HG. Evidence for an accessory role of LFA-1 in lymphocyte-high endothelium interaction during homing. J Immunol 1988; 140:693–699. 55. Andrew DP, Spellberg JP, Takimoto H, Schmits R, Mak TW, Zukowski MM. Transendothelial migration and trafficking of leukocytes in LFA-1- deficient mice. Eur J Immunol 1998; 28:1959–1969. 56. Berlin-Rufenach C, Otto F, Mathies M, Westermann J, Owen MJ, Hamann A, Hogg N. Lymphocyte migration in lymphocyte function-associated antigen (LFA)-1- deficient mice. J Exp Med 1999; 189:1467–1478. 57. Gunn MD, Tangemann K, Tam C, Cyster JG, Rosen SD, Williams LT. A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc Natl Acad Sci USA 1998; 95:258–263. 58. Campbell JJ, Hedrick J, Zlotnik A, Siani MA, Thompson DA, Butcher EC. Chemokines and the arrest of lymphoyctes rolling under flow conditions. Science 1998; 279:381–384. 59. Campbell JJ, Bowman EP, Murphy K, Youngman KR, Siani MA, Thompson DA, Wu L, Zlotnik A, Butcher EC. 6-C-kine (SLC), a lymphocyte adhesion-triggering chemokine expressed by high endothelium, is an agonist for the MIP-3beta receptor CCR7. J Cell Biol 1998; 141:1053–1059. 60. Vassileva G, Soto H, Zlotnik A, Nakano H, Kakiuchi T, Hedrick JA, Lira SA. The reduced expression of 6Ckine in the plt mouse results from the deletion of one of two 6Ckine genes. J Exp Med 1999; 190:1183–1188. 61. Nakano H, Gunn MD. Gene duplications at the chemokine locus on mouse chromosome 4: multiple strain-specific haplotypes and the deletion of secondary lymphoidorgan chemokine and EBI-1 ligand chemokine genes in the plt mutation. J Immunol 2001; 166:361–369. 62. Forster R, Schubel A, Breitfeld D, Kremmer E, Renner-Muller I, Wolf E, Lipp M. CCR7 coordinates the primary immune response by establishing functional microenvironments in secondary lymphoid organs. Cell 1999; 99:23–33. 63. Gunn MD, Kyuwa S, Tam C, Kakiuchi T, Matsuzawa A, Williams LT, Nakano H. Mice lacking expression of secondary lymphoid organ chemokine have defects in lymphocyte homing and dendritic cell localization. J Exp Med 1999; 189:451–460. 64. Cinamon G, Shinder V, Alon R. Shear forces promote lymphocyte migration across vascular endothelium bearing apical chemokines. Nat Immunol 2001; 2:515–522. 65. Weninger W, Carlsen HS, Goodarzi M, Moazed F, Crowley MA, Baekkevold ES, Cavanagh LL, von Andrian UH. Naïve T cell recruitment to non-lymphoid tissues: a role for endotheliumexpressed CCL21 in autoimmune disease and lymphoid neogenesis. J Immunol 2003; 170:4638–4648. 66. Tang ML, Steeber DA, Zhang XQ, Tedder TF. Intrinsic differences in L-selectin expression levels affect T and B lymphocyte subset-specific recirculation pathways. J Immunol 1998; 160:5113–5121. 67. Stein JV, Cheng G, Stockton BM, Fors BP, Butcher EC, von Andrian UH. L-selectin-mediated leukocyte adhesion in vivo: microvillous distribution determines tethering efficiency, but not rolling velocity. J Exp Med 1999; 189:37–50. 68. Robert C, Klein C, Cheng G, Kogan A, Mulligan RC, von Andrian UH, Kupper TS. Gene therapy to target dendritic cells from blood to lymph nodes. Gene Ther 2003; 10:1479–1486. 69. Diacovo TG, Puri KD, Warnock RA, Springer TA, von Andrian UH. Plateletsmediated lymphocyte delivery to high endothelial venules. Science 1996; 273: 252–255. 70. Diacovo TG, Catalina MD, Siegelman MH, von Andrian UH. Circulating activated platelets reconstitute lymphocyte homing and immunity in L-selectin-deficient mice. J Exp Med 1998; 187:197–204. 71. Catalina MD, Carroll MC, Arizpe H, Takashima A, Estess P, Siegelman MH. The route of antigen entry determines the requirement for L-selectin during immune responses. J Exp Med 1996; 184:2341–2351.

Endothelial cells in health and disease

112

72. Berlin C, Berg EL, Briskin MJ, Andrew DP, Kilshaw PJ, Holzmann B, Weissman IL, Hamann A, Butcher EC. α4 β7 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAM-1. Cell 1993; 74:185–195. 73. Bargatze RF, Jutila MA, Butcher EC.Distinct roles of L-selectin and integrins α4β7 and LFA-1 in lymphocyte homing to Peyer’s patch-HEV in situ: the multistep model confirmed and refined. Immunity 1995; 3:99–108. 74. Arbones ML, Ord DC, Ley K, Ratech H, Maynard-Curry C, Otten G, Capon DJ, Tedder TF. Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin-deficient mice. Immunity 1994; 1:247–260. 75. Wagner N, Lohler J, Kunkel EJ, Ley K, Leung E, Krissansen G, Rajewsky K, Muller W. Critical role for β7 integrins in formation of the gut-associated lymphoid tissue. Nature 1996; 382:366–370. 76. Tan K, Casasnovas JM, Liu J-H, Briskin MJ, Springer TA, Wang J-H. The structure of immunoglobulin superfamily domains 1 and 2 of MAdCAM-1 reveals novel features important for integrin recognition. Structure 1998; 6:793–801. 77. Berg EL, McEvoy LM, Berlin C, Bargatze RF, Butcher EC. L-selectin-mediated lymphocyte rolling on MAdCAM-1. Nature 1993; 366:695–698. 78. Berlin C, Bargatze RF, von Andrian UH, Szabo MC, Hasslen SR, Nelson RD, Berg EL, Erlandsen SL, Butcher EC. α4 integrins mediate lymphocyte attachment and rolling under physiologic flow. Cell 1995; 80:413–422. 79. Hamann A, Andrew DP, Jablonski-Westrich D, Holzmann B, Butcher EC. Role of α4-integrins in lymphocyte homing to mucosal tissues in vivo. J Immunol 1994; 152: 3282–3293. 80. Warnock RA, Campbell JJ, Dorf ME, Matsuzawa A, McEvoy LM, Butcher EC. The role of chemokines in the microenvironmental control of T versus B cell arrest in Peyer’s patch high endothelial venules. J Exp Med 2000; 191:77–88. 81. Palframan RT, Jung S, Cheng G, Weninger W, Luo Y, Dorf M, Littman DR, Rollins BJ, Zweerink H, Rot A, von Andrian UH. Inflammatory chemokine transport and presentation in HEV: A remote control mechanism for monocyte recruitment to lymph nodes in inflamed tissues. J Exp Med 2001; 194:1361–1374. 82. Middleton J, Neil S, Wintle J, Clark-Lewis I, Moore H, Lam C, Auer M, Hub E, Rot A. Transcytosis and surface presentation of IL-8 by venular endothelial cells. Cell 1997; 91:1001– 1011. 83. Janatpour MJ, Hudak S, Sathe M, Sedgwick JD, McEvoy LM. Tumor necrosis factordependent segmental control of MIG expression by high endothelial venules in inflamed lymph nodes regulates monocyte recruitment. J Exp Med 2001; 194:1375–1384. 84. Nibbs RJ, Kriehuber E, Ponath PD, Parent D, Qin S, Campbell JD, Henderson A, Kerjaschki D, Maurer D, Graham GJ, Rot A. The beta-chemokine receptor D6 is expressed by lymphatic endothelium and a subset of vascular tumors. Am J Pathol 2001; 158:867–877. 85. Fra AM, Locati M, Otero K, Sironi M, Signorelli P, Massardi ML, Gobbi M, Vecchi A, Sozzani S, Mantovani A. Cutting edge: scavenging of inflammatory CC chemokines by the promiscuous putatively silent chemokine receptor D6. J Immunol 2003; 170:2279–2282. 86. Girard JP, Baekkevold ES, Yamanaka T, Haraldsen G, Brandtzaeg P, Amalric F. Heterogeneity of endothelial cells: the specialized phenotype of human high endothelial venules characterized by suppression subtractive hybridization. Am J Pathol 1999; 155:2043–2055. 87. Aurrand-Lions M, Duncan L, Ballestrem C, Imhof BA. JAM-2, a novel immunoglobulin superfamily molecule, expressed by endothelial and lymphatic cells. J Biol Chem 2001; 276:2733–2741. 88. Liang TW, Chiu HH, Gurney A, Sidle A, Tumas DB, Schow P, Foster J, Klassen T, Dennis K, DeMarco RA, Pham T, Frantz G, Fong S. Vascular endothelial-junctional adhesion molecule (VE-JAM)/JAM 2 interacts with T, NK, and dendritic cells through JAM 3. J Immunol 2002; 168:1618–1626.

High endothelial venules

113

89. Palmeri D, van Zante A, Huang CC, Hemmerich S, Rosen SD. Vascular endothelial junctionassociated molecule, a novel member of the immunoglobulin superfamily, is localized to intercellular boundaries of endothelial cells. J Biol Chem 2000; 275:19139–19145. 90. Muller WA. Leukocyte-endothelial-cell interactions in leukocyte transmigration and the inflammatory response. Trends Immunol 2003; 24:327–334. 91. Worthylake RA, Burridge K. Leukocyte transendothelial migration: orchestrating the underlying molecular machinery. Curr Opin Cell Biol 2001; 13:569–577. 92. Johnson-Leger C, Aurrand-Lions M, Imhof BA. The parting of the endothelium: miracle, or simply a junctional affair? J Cell Sci 2000; 113(Pt 6):921–933. 93. Imai Y, True DD, Singer MS, Rosen SD. Direct demonstration of the lectin activity of gp90 mel, a lymphocyte homing receptor. J Cell Biol 1990; 111:1225–1232. 94. Clark RA, Fuhlbrigge RC, Springer TA. L-selectin ligands that are O-glycoproteaseresistant and distinct from MECA-79 antigen are sufficient for tethering and rolling of lymphocytes on human high endothelial venules. J Cell Biol 1998; 140:721–731. 95. van Zante A, Gauguet J-M, Bistrup A, Tsay D, von Andrian UH, Rosen SD. Lymphocyte-HEV interactions in lymph nodes of a sulfotransferase-deficient mouse. J Exp Med 2003; 198:1289– 1300. 96. Hemmerich S, Bistrup A, Singer MS, van Zante A, Lee JK, Tsay D, Peters M, Carminati JL, Brennan TJ, Carver-Moore K, Leviten M, Fuentes ME, Ruddle NH, Rosen SD. Sulfation of Lselectin ligands by an HEV-restricted sulfotransferase regulates lymphocyte homing to lymph nodes. Immunity 2001; 15:237–247. 97. Stockton BM, Cheng G, Manjunath N, Ardman B, von Andrian UH. Negative regulation of T cell homing by CD43. Immunity 1998; 8:373–381. 98. M’Rini C, Cheng G, Schweitzer C, Cavanagh LL, Palframan RT, Mempel TR, Warnock RA, Lowe JB, Quackenbush EJ, von Andrian UH. A novel endothelial L-selectin ligand activity in lymph node medulla that is regulated by a (1,3)-fucosyltransferase-IV. J Exp Med 2003; 198:1301–1312. 99. Lasky LA, Singer MS, Dowbenko D, Imai Y, Henzel WJ, Grimley C, Fennie C, Gillett N, Watson SR, Rosen SD. An endothelial ligand for L-selectin is a novel mucin-like molecule. Cell 1992; 69:927–938. 100. Brustein M, Kraal G, Mebius RE, Watson SR. Identification of a soluble form of a ligand for the lymphocyte homing receptor. J Exp Med 1992; 176:1415–1419. 101. Baumhueter S, Singer MS, Henzel W, Hemmerich S, Renz M, Rosen SD, Lasky LA. Binding of L-selectin to the vascular sialomucin, CD34. Science 1993; 262:436–438. 102. Sassetti C, Tangemann K, Singer MS, Kershaw DB, Rosen SD. Identification of podocalyxinlike protein as a high endothelial venule ligand for L-selectin: parallels to CD34. J Exp Med 1998; 187:1965–1975. 103. Hemmerich S, Butcher EC, Rosen SD. Sulfation-dependent recognition of high endothelial venules (HEV)-ligands by L-selectin and MECA 79. J Exp Med 1994; 180:2219–2226. 104. Samulowitz U, Kuhn A, Brachtendorf G, Nawroth R, Braun A, Bankfalvi A, Bocker W, Vestweber D. Human endomucin: distribution pattern, expression on high endothelial venules, and decoration with the MECA-79 epitope. Am J Pathol 2002; 160: 1669–1681. 105. Sassetti C, Van Zante A, Rosen SD. Identification of endoglycan, a member of the CD34/podocalyxin family of sialomucins. J Biol Chem 2000; 275:9001–9010. 106. Fieger CB, Sassetti CM, Rosen SD. Endoglycan, a member of the CD34 family, functions as a L-selectin ligand through modification with tyrosine sulfation and sialyl Lewis x. J Biol Chem 2003; 278:27390–27398. 107. Puri KD, Finger EB, Gaudernack G, Springer TA. Sialomucin CD34 is the major L-selectin ligand in human tonsil high endothelial venules. J Cell Biol 1995; 131: 261–270. 108. Cheng J, Baumhueter S, Cacalano G, Carver-Moore K, Thibodeaux H, Thomas R, Broxmeyer HE, Cooper S, Hague N, Moore M, Lasky LA. Hematopoietic defects in mice lacking the sialomucin CD34. Blood 1996; 87:479–490.

Endothelial cells in health and disease

114

109. Dowbenko D, Kikuta A, Fennie C, Gillett N, Lasky LA. Glycosylation-dependent cell adhesion molecule 1 (GlyCAM-1) mucin is expressed by lactating mammary gland epithelial cells and is present in milk. J Clin Invest 1993; 92:952–960. 110. Baumhueter S, Dybdal N, Kyle C, Lasky LA. Global vascular expression of murine CD34, a sialomucin-like endothelial ligand for L-selectin. Blood 1994; 84:2554–2565. 111. Stoolman LM, Rosen SD. Possible role for cell-surface carbohydrate-binding molecules in lymphocyte recirculation. J Cell Biol 1983; 96:722–729. 112. Stoolman LM, Tenforde TS, Rosen SD. Phosphomannosyl receptors may participate in the adhesive interaction between lymphocytes and high endothelial venules. J Cell Biol 1984; 99:1535–1540. 113. Yednock TA, Butcher EC, Stoolman LM, Rosen SD. Receptors involved in lymphocyte homing: Relationship between a carbohydrate-binding receptor and the MEL-14 antigen. J Cell Biol 1987; 104:725–731. 114. Imai Y, Lasky LA, Rosen SD. Sulphation requirement for GlyCAM-1, an endothelial ligand for L-selectin. Nature 1993; 361:555–557. 115. Hemmerich S, Rosen SD. 6′-Sulfated sialyl Lewis x is a major capping group of GlyCAM-1. Biochemistry 1994; 33:4830–4835. 116. Hemmerich S, Bertozzi CR, Leffler H, Rosen SD. Identification of the sulfated monosaccharides of GlyCAM-1, an endothelial-derived ligand for L-selectin. Biochemistry 1994; 33:4820−4829. 117. Rosen SD, Singer MS, Yednock TA, Stoolman LM. Involvement of sialic acid on endothelial cells in organ-specific lymphocyte recirculation. Science 1985; 228:1005–1007. 118. Imai Y, Singer MS, Fennie C, Lasky LA, Rosen SD. Identification of a carbohydrate based endothelial ligand for a lymphocyte homing receptor. J Cell Biol 1991; 113:1213–1221. 119. Maly P, Thall AD, Petryniak B, Rogers CE, Smith PL, Marks RM, Kelly RJ, Gersten KM, Cheng G, Saunders TL, Camper SA, Camphausen RT, Sullivan FX, Isogai Y, Hindsgaul O, von Andrian UH, Lowe JB. The α(1,3) fucosyltransferase Fuc-TVII controls leukocyte trafficking through an essential role in L-, E-, and P-selectin ligand biosynthesis. Cell 1996; 86:643–653. 120. Homeister JW, Thall AD, Petryniak B, Maly P, Rogers CE, Smith PL, Kelly RJ, Gersten KM, Askari SW, Cheng G, Smithson G, Marks RM, Misra AK, Hindsgaul O, von Andrian UH, Lowe JB. The alpha(1, 3)fucosyltransferases FucT-IV and FucT-VII exert collaborative control over selectin-dependent leukocyte recruitment and lymphocyte homing. Immunity 2001; 15:115–126. 121. Niemela R, Natunen J, Majuri ML, Maaheimo H, Helin J, Lowe JB, Renkonen O, Renkonen R. Complementory acceptor and site specificities of Fuc-TIV and Fuc-TVII allow effective biosynthesis of sialyl-TriLex and related polylactosamines present on glycoprotein counterreceptors of selectins. J Biol Chem 1998; 273:4021–4026. 122. Lowe JB. Glycosylation, immunity, and autoimmunity. Cell 2001; 104:809–812. 123. Scudder PR, Shailubhai K, Duffin KL, Streeter PR, Jacob GS. Enzymatic synthesis of a 6’sulphated sialyl-Lewisx which is an inhibitor of L-selectin binding to peripheral addressin. Glycobiology 1994; 4:929–933. 124. Uchimura K, Muramatsu H, Kadomatsu K, Fan QW, Kurosawa N, Mitsuoka C, Kannagi R, Habuchi O, Muramatsu T. Molecular cloning and characterization of an N-acetylglucosamine-6O-sulfotransferase. J Biol Chem 1998; 273:22577–22583. 125. Bistrup A, Bhakta S, Lee JK, Belov YY, Gunn MD, Zuo FR, Huang C-C, Kannagi R, Rosen SD, Hemmerich S. Sulfotransferases of two specificities function in the reconstitution of high endothelial cell ligands for L-selectin. J Cell Biol 1999; 145:899–910. 126. Hiraoka N, Petryniak B, Nakayama J, Tsuboi S, Suzuki M, Yeh JC, Izawa D, Tanaka T, Miyasaka M, Lowe JB, Fukuda M. A novel, high endothelial venule-specific sulfotransferase expresses 6-sulfo sialyl Lewis(x), an L-selectin ligand displayed by CD34. Immunity 1999; 11:79–89.

High endothelial venules

115

127. Hiraoka N, Kawashima H, Petryniak B, Nakayama J, Mitoma J, Marth JD, Lowe JB, Fukuda M. Core 2 branching {beta}1, 6-N-acetylglucosaminyl transferase and high endothelial venulerestricted sulfotransferase collaboratively control lymphocyte homing. J Biol Chem 2004; 279:3058–3067. 127a. Gauguet JM, Rosen SD, Marth JD, von Andrian UH. Core 2 branching β1, 6-Nacetylglucosaminyltransferase and high endothelial cell N-acetylglucosamine-6-sulfotransferase exert differential control over B and T lymphocyte homing to peripheral lymph nodes. Blood 2004 (in press). 128. Yeh JC, Hiraoka N, Petryniak B, Nakayama J, Ellies LG, Rabuka D, Hindsgaul O, Marth JD, Lowe JB, Fukuda M. Novel sulfated lymphocyte homing receptors and their control by a Core1 extension beta 1, 3-N-acetylglucosaminyltransferase. Cell 2001; 105:957–969. 129. Hemmerich S, Leffler H, Rosen SD. Structure of the O-glycans in GlyCAM-1, an endothelialderived ligand for L-selectin. J Biol Chem 1995; 270:12035–12047. 130. Williams D, Schachter H. Mucin synthesis. I. Detection in canine submaxillary glands of an N-acetylglucosaminyltransferase which acts on mucin substrates. J Biol Chem 1980; 255:11247–11252. 131. Ellies LG, Tsuboi S, Petryniak B, Lowe JB, Fukuda M, Marth JD. Core 2 oligosaccharide biosynthesis distinguishes between selectin ligands essential for leukocyte homing and inflammation. Immunity 1998; 9:881–890. 132. Briskin MJ, McEvoy LM, Butcher EC. MAdCAM-1 has homology to immunoglobulin and mucin-like adhesion receptors and to IgA1. Nature 1993; 363:461–464. 133. Kuschert GS, Coulin F, Power CA, Proudfoot AE, Hubbard RE, Hoogewerf AJ, Wells TN. Glycosaminoglycans interact selectively with chemokines and modulate receptor binding and cellular responses. Biochemistry 1999; 38:12959–12968. 134. Chaudhuri A, Nielsen S, Elkjaer ML, Zbrzezna V, Fang F, Pogo AO. Detection of Duffy antigen in the plasma membranes and caveolae of vascular endothelial and epithelial cells of nonerythroid organs. Blood 1997; 89:701–712. 135. Kashiwazaki M, Tanaka T, Kanda H, Ebisuno Y, Izawa D, Fukuma N, Akimitsu N, Sekimizu K, Monden M, Miyasaka M. A high endothelial venule-expressing promiscuous chemokine receptor DARC can bind inflammatory, but not lymphoid, chemokines and is dispensable for lymphocyte homing under physiological conditions. Int Immunol 2003; 15:1219–1227. 136. Mebius RE, Streeter PR, Michie S, Butcher EC, Weissman IL. A developmental switch in lymphocyte homing receptor and endothelial vascular addressin expression regulates lymphocyte homing and permits CD4+CD3− cells to colonize lymph nodes. Proc Natl Acad Sci USA 1996; 93:11019–11024. 137. Fu YX, Chaplin DD. Development and maturation of secondary lymphoid tissues. Annu Rev Immunol 1999; 17:399−433. 138. Belisle C, Sainte-Marie G. Tridimensional study of the deep cortex of the rat lymph node. V: postnatal development of the deep cortex units. Anat Rec 1981; 200:207–220. 139. Hendriks HR, Eestermans IL, Hoefsmit EC. Depletion of macrophages and disappearance of postcapillary high endothelial venules in lymph nodes deprived of afferent lymphatic vessels. Cell Tissue Res 1980; 211:375–389. 140. Hendriks HR, Duijvestijn AM, Kraal G. Rapid decrease in lymphocyte adherence to high endothelial venules in lymph nodes deprived of afferent lymphatic vessels. Eur J Immunol 1987; 17:1691–1695. 141. Mebius RE, Dowbenko D, Williams A, Fennie C, Lasky LA, Watson SR. Expression of GlyCAM-1, an endothelial ligand for L-selectin, is affected by afferent lymphatic flow. J Immunol 1993; 151:6769–6776. 142. Swarte VV, Joziasse DH, Van den Eijnden DH, Petryniak B, Lowe JB, Kraal G, Mebius RE. Regulation of fucosyltransferase-VII expression in peripheral lymph node high endothelial venules. Eur J Immunol 1998; 28:3040–3047. 143. Mebius RE. Organogenesis of lymphoid tissues. Nat Rev Immunol 2003; 3:292–303.

Endothelial cells in health and disease

116

144. Ruddle NH. Lymphoid neo-organogenesis: lymphotoxin’s role in inflammation and development. Immunol Res 1999; 19:119–125. 145. Kratz A, Campos-Neto A, Hanson MS, Ruddle NH. Chronic inflammation caused by lymphotoxin is lymphoid neogenesis. J Exp Med 1996; 183:1461–1472. 146. Fan L, Reilly CR, Luo Y, Dorf ME, Lo D. Cutting edge: ectopic expression of the chemokine TCA4/SLC is sufficient to trigger lymphoid neogenesis. J Immunol 2000; 164:3955–3959. 147. Luther SA, Lopez T, Bai W, Hanahan D, Cyster JG. BLC expression in pancreatic islets causes B cell recruitment and lymphotoxin-dependent lymphoid neogenesis. Immunity 2000; 12:471–481. 148. Chen SC, Vassileva G, Kinsley D, Holzmann S, Manfra D, Wiekowski MT, Romani N, Lira SA. Ectopic expression of the murine chemokines CCL21a and CCL21b induces the formation of lymph node-like structures in pancreas, but not skin, of transgenic mice. J Immunol 2002; 168:1001–1008. 149. Millet I, Ruddle NH. Differential regulation of lymphotoxin (LT), lymphotoxin-β (LT-β), and TNF-α in murine T cell clones activated through the TCR. J Immunol 1994; 152:4336–4346. 150. De Togni P, Goellner J, Ruddle NH, Streeter PR, Fick A, Mariathasan S, Smith SC, Carlson R, Shornick LP, Strauss-Schoenberger J, Russell JH, Karr R, Chaplin DD. Abnormal development of peripheral lymphoid organs in mice deficient in lymphotoxin. Science 1994; 264:703–707. 151. Koni PA, Sacca R, Lawton P, Browning JL, Ruddle NH, Flavell RA. Distinct roles in lymphoid organogenesis for lymphotoxins alpha and beta revealed in lymphotoxin betadeficient mice. Immunity 1997; 6:491–500. 152. Izawa D, Tanaka T, Saito K, Ogihara H, Usui T, Kawamoto S, Matsubara K, Okubo K, Miyasaka M. Expression profile of active genes in mouse lymph node high endothelial cells. Int Immunol 1999; 11:1989–1998. 153. Saito K, Tanaka T, Kanda H, Ebisuno Y, Izawa D, Kawamoto S, Okubo K, Miyasaka M. Gene expression profiling of mucosal addressin cell adhesion molecule-1+high endothelial venule cells (HEV) and identification of a leucine-rich HEV glycoprotein as a HEV marker. J Immunol 2002; 168:1050–1059. 154. Girard J-P, Springer TA. Cloning from purified high endothelial venule cells of hevin, a close relative of the antiadhesive extracellular matrix protein SPARC. Immunity 1995; 2:113–122. 155. Girard JP, Baekkevold ES, Amalric F. Sulfation in high endothelial venules: cloning and expression of the human PAPS synthetase. FASEB J 1998; 12:603–612. 156. Baekkevold ES, Roussigne M, Yamanaka T, Johansen FE, Jahnsen FL, Amalric F, Brandtzaeg P, Erard M, Haraldsen G, Girard JP. Molecular characterization of NF-HEV, a nuclear factor preferentially expressed in human high endothelial venules. Am J Pathol 2003; 163:69–79. 157. Denk W, Strickler JH, Webb WW. Two-photon laser scanning fluorescence microscopy. Science 1990; 248:73–76. 158. Cahalan MD, Parker I, Wei SH, Miller MJ. Two-photon tissue imaging: seeing the immune system in a fresh light. Nat Rev Immunol 2002; 2:872–880. 159. Miller MJ, Wei SH, Parker I, Cahalan MD. Two-photon imaging of lymphocyte motility and antigen response in intact lymph node. Science 2002; 296:1869–1873. 160. Stoll S, Delon J, Brotz TM, Germain RN. Dynamic imaging of T cell-dendritic cell interactions in lymph nodes. Science 2002; 296:1873–1876. 161. Bousso P, Robey E. Dynamics of CD8(+) T cell priming by dendritic cells in intact lymph nodes. Nat Immunol 2003; 4:579–585. 162. Miller MJ, Wei SH, Cahalan MD, Parker I. Autonomous T cell trafficking examined in vivo with intravital two-photon microscopy. Proc Natl Acad Sci USA 2003; 100: 2604–2609. 163. Tedder TF, Steeber DA, Pizcueta P. L-selectin-deficient mice have impaired leukocyte recruitment into inflammatory sites. J Exp Med 1995; 181:2259–2264.

High endothelial venules

117

164. Protin U, Schweighoffer T, Jochum W, Hilberg F. CD44-deficient mice develop normally with changes in subpopulations and recirculation of lymphocyte subsets. J Immunol 1999; 163:4917–4923. 165. Stoop R, Gal I, Glant TT, McNeish JD, Mikecz K. Trafficking of CD44-deficient murine lymphocytes under normal and inflammatory conditions. Eur J Immunol 2002; 32:2532–2542. 166. Xu H, Gonzalo JA, St.Pierre Y, Williams IR, Kupper TS, Cotran RS, Springer TA, GutierrezRamos J-C. Leukocytosis and resistance to septic shock in intercellular adhesion molecule 1deficient mice. J Exp Med 1994; 180:95–109. 167. Sligh JE Jr, Ballantyne CM, Rich S, Hawkins HK, Smith CW, Bradley A, Beaudet AL. Inflammatory and immune responses are impaired in mice deficient in intercellular adhesion molecule 1. Proc Natl Acad Sci USA 1993; 90:8529–8533. 168. Gerwin N, Gonzalo JA, Lloyd C, Coyle AJ, Reiss Y, Banu N, Wang B, Xu H, Avraham H, Engelhardt B, Springer TA, Gutierrez-Ramos JC. Prolonged eosinophil accumulation in allergic lung interstitium of ICAM-2 deficient mice results in extended hyperresponsiveness. Immunity 1999; 10:9–19. 169. Nakano H, Tamura T, Yoshimoto T, Yagita H, Miyasaka M, Butcher EC, Nariuchi H, Kakiuchi T, Matsuzawa A. Genetic defect in T lymphocyte-specific homing into peripheral lymph nodes. Eur J Immunol 1997; 27:215–221. 170. Forster R, Mattis AE, Kremmer E, Wolf E, Brem G, Lipp M. A putative chemokine receptor, BLRI, directs B cell migration to defined lymphoid organs and specific anatomic compartments of the spleen. Cell 1996; 87:1037–1047. 171. Dawson TC, Lentsch AB, Wang Z, Cowhig JE, Rot A, Maeda N, Peiper SC. Exaggerated response to endotoxin in mice lacking the Duffy antigen/receptor for chemokines (DARC). Blood 2000; 96:1681–1684. 172. Huang MC, Laskowska A, Vestweber D, Wild MK. The alpha (1,3)-fucosyltransferase FucTIV, but not Fuc-TVII, generates sialyl Lewis X-like epitopes preferentially on glycolipids. J Biol Chem 2002; 277:47786–47795. 173. Alon R, Feizi T, Yuen C-T, Fuhlbrigge RC, Springer TA. Glycolipid ligands for selectins support leukocyte tethering and rolling under physiologic flow conditions. J Immunol 1995; 154:5356–5366.

5 The Use of Proteomics to Map Phenotypic Heterogeneity of the Endothelium Johanna Lahdenranta, Wadih Arap, and Renata Pasqualini The University of Texas, M.D.Anderson Cancer Center, Houston, Texas, U.S.A.

1. INTRODUCTION The inner lining of the vasculature consists of a heterogeneous population of endothelial cells. Phenotypes of these cells vary between different organs, between different parts of the vasculature in a given organ, and even between neighboring endothelial cells of the same organ and the same blood vessel type. Every single endothelial cell of the body is subjected to a seemingly infinite array of signals, including soluble factors, such as growth factors and chemokines, cell-cell and cell-basement membrane interactions, and other variables, such as pH, pO2, sheer stress from blood flow, stretch, and temperature, to name a few. All these variables in the endothelial cell microenvironment will influence the phenotype—thus function—of the cell, to the extent that its predetermined genetic makeup allows. Together, these diverse phenotypes (structural and functional) lead to vascular heterogeneity (e.g., at the level of the organ, tissue, and blood vessel). Hopefully, this phenomenon will become increasingly recognized in the clinical practice of medicine. Structural and functional heterogeneity of the endothelium has been a subject of study for decades, but only more recently has the focus of studies in endothelial cell biology shifted to the molecular heterogeneity of the endothelium. The exploration of the molecular diversity of blood vessels is a rapidly expanding research area that is driven by the vast potential of discoveries in molecular heterogeneity to contribute to the development of targeted diagnostics and therapeutics. It is now recognized that a complex system of ligand-receptor pairs exists within the vasculature of tissues. The expression levels and activation states of these addresses are modulated in blood vessels during tumor progression, and can also be altered in the context of other pathological conditions involving abnormal blood vessel development and function, such as retinopathies, inflammation, and atherosclerosis.

Endothelial cells in health and disease

120

Many therapeutic targets may be expressed in very restricted—but highly specific and accessible—locations in the vascular endothelium. High-throughput DNA sequencing or gene arrays are typically carried out on endothelial cells that have been removed from their tissue of origin, and in neglecting the anatomical and functional context of the endothelium, may readily overlook potential targets for intervention. Moreover, such approaches yield an immense amount of data, adding to the difficulty in interpreting all the information in a meaningful way. Instead, identification of vascular bed specific ligand-receptor pairs and knowledge about their cellular distribution and accessibility will be requisite for the development of endothelium-targeted therapies. Proteomics may be defined as the systematic analysis of the proteins in biological samples that aims to document the overall distribution of proteins in cells, identify and characterize individual proteins of interest, and ultimately to elucidate their relationships and functional roles. Vascular proteomics is the molecular phenotyping of cells forming blood vessels at the protein-protein interaction level. Exploiting the molecular diversity of cell surface receptors expressed in the human endothelium may lead to a ligandreceptor-based molecular map of the blood vessels in the body, a so-called “vascular map.” Standard proteomic tools—two-dimensional protein gels combined with mass spectrometry—have uncovered a large repertoire of differentially expressed proteins on endothelial cells. Two-dimensional protein gels have been used to separate out several thousand proteins from different endothelial cell lines of differentially treated/stimulated endothelial cells. Mass spectrometry has then been employed to identify the differentially expressed proteins. In our laboratory, we have been developing integrated, combinatorial library platform technologies whose goal is to enable the identification, validation, and prioritization of functional molecular targets in human blood vessels. This methodology will allow drug development based on targeting the differential protein expression in the vasculature associated with normal tissues or diseases with an angiogenesis component. These include cancer, arthritis, diabetes, and cardiovascular diseases. Our long-term goal is to translate a functional map of molecular targets and biomarkers into clinical applications.

2. LARGE-SCALE FUNCTIONAL VASCULAR PROTEOMICS A major goal in drug development has long been to develop technology for targeting therapeutics more effectively to their intended disease site and to improve their therapeutic index by limiting the systemic exposure of other tissues to untoward or toxic effects. The methods described here have two main applications. First, they may lead to identification of vascular targeting ligands. Second, they may enable the construction of a molecular map of human vascular receptors. In theory, targeted delivery of drugs, liposomes, peptide sequences, gene therapy vectors, and biological therapies can be achieved in clinical applications. Ultimately, it may be possible to guide imaging or therapeutic compounds to the target site in real clinical situations. In the future, the determination of molecular profiles of blood vessels in specific conditions may also lead to the discovery of disease-specific vascular targets. Early identification of targets, optimized regimens tailored to the molecular profile of individual patients, and

The use of proteomics to map phenotypic heterogeneity

121

identification of new vascular addresses may provide a rationale for revisiting or salvaging otherwise ineffective or toxic drug candidates. Below, we review several phage display targeting strategies that may enable the construction of a molecular map outlining vascular diversity in each organ, tissue, or pathological condition.

3. PHAGE DISPLAY TARGETING STRATEGIES Phage display technology allows presentation of large peptide and protein libraries on the surface of filamentous phage permitting the selection of peptides and proteins, including antibodies, with high affinity and specificity to almost any target. This technology has had a major influence on the work and discoveries made in the fields of immunology, cell biology, drug discovery, and pharmacology. The power of phage display lies in the ability to propagate selected peptide/protein ligands through multiple rounds of selection and the direct link of the phenotype of the antigen/receptor binding ligand to the genotype of the phage particle presenting the ligand. Phage display technology was first introduced as an expression vector (“fusion phage”) capable of presenting a foreign amino acid sequence accessible to binding antibody (1). Since then, large numbers of phage displayed peptide and protein libraries have been constructed (2–4, reviewed in Ref. 5), leading to various techniques for screening such libraries. Peptide display technology has since then been applied to a wide range of protein interaction studies with purified/recombinant proteins, cells, and intact tissues in situ as well as in vivo. A vast body of work has been done using phage displayed antibody libraries for diagnostic and therapeutic applications (reviewed in Refs. 5 and 6). Phage display involves genetically manipulating bacteriophage so that peptides or antibodies can be expressed on their surface. Random peptide libraries consist of large random collection of peptides, displayed as recombinant proteins on the surface of a filamentous bacteriophage. Random peptide libraries usually contain up to 109 individual phage clones. Here, we review the recent progress that has been made in using phage displayed random peptide libraries as a proteomics tool to map heterogeneity of the endothelium.

3.1. In Vitro Targeting 3.1.1. Cell-Free Screening on Isolated Receptors Phage display of random peptide libraries is a powerful system for obtaining small peptide ligands for virtually any protein of interest. A high proportion of isolated peptide ligands interact with the natural binding site of the target protein acting as antagonists or agonists of the natural protein functions (reviewed in Ref. 7). This is likely due to the hydrophilic nature of the protein surface, which itself has a role in prevention of

Endothelial cells in health and disease

122

unspecific interactions with other molecules. However, the binding sites of proteins will generally allow water molecules to be displaced by the binding of specific ligands, such as selected peptides (8). It is a common observation that the binding motif of a targeting peptide is a tripeptide motif appearing several times in different sequence contexts. A motif consisting of three amino acids seems to provide the minimal framework for structural formation and protein-protein interaction (9). Examples of tripeptide recognition units and receptor binding ligand motifs include RGD, LDV, and LLG to integrins (10,11), GFE to membrane dipeptidase (12,13) and NGR to CD13/aminopeptidase N (APN) (14). Considerable progress has been made in the construction of phage display random peptide libraries and in screening methods. Ligands can be selected and isolated by “biopanning,” a process in which phage that bind to a target molecule are eluted and amplified in a host bacteria. Besides proteins, peptides affecting biologically significant protein-DNA interactions (15), peptides binding to carbohydrates (16,17), carbon nanotubes (18), and small chemical compounds like taxol (19) have been isolated from phage display random peptide libraries. In general, the affinity selection of ligands from a phage display random peptide libraries involves the following fundamental steps: (i) preparation of a primary library or amplification of an existing library, (ii) exposure of the phage particles to a target (immobilized protein/cell surface protein/vascular endothelium) for which specific ligands are planned to be identified, (iii) removal of nonspecific binders (washing/perfusion), (iv) recovery of the target bound phage by elution or direct bacterial infection and amplification of the recovered phage, (v) repetition of the steps (i)–(iv), usually three to six rounds, until an enriched population of binders is obtained, and finally, (vi) sequencing the peptide inserts of the enriched phage clones. Enriched peptide inserts are then analyzed, and desired peptides can be synthesized as recombinant or synthetic peptides for further analysis of the ligand-target interaction. Phage displayed random peptide libraries were used early on in the mid-1980s to map antigen recognition sites of antibodies (1). However, such an approach has not been employed to probe antibodies associated with cancer. Cancer patients can mount a humoral immune response (i.e., make antibodies) against vascular receptors that are also found in or presented by the nonmalignant cells of tumor blood vessels. We have developed a phage display-based screening to select peptide sequences recognized by the repertoire of circulating tumor-associated antibodies. We isolated peptides recognized by antibodies purified from serum of cancer patients. Consensus peptide motifs showed marked selective binding to circulating antibodies from cancer patients over control antibodies from blood donors. We next validated such motifs by showing that serum reactivity to the peptide can be specifically linked to disease progression and to patient survival. Finally, we isolated a corresponding tumor antigen eliciting the immune response. These results show that it is possible to identify tumor antigens by fingerprinting the pool of antibodies from the serum of cancer patients. Exploiting the differential humoral response to cancer through such approach may lead to the discovery of new molecular addresses (20). In addition, targeting tumor antigens may lead to the development of improved vaccines against tumors.

The use of proteomics to map phenotypic heterogeneity

123

3.1.2. Screening the Molecular Diversity of Cell Surfaces Identification of the endothelial cell surface receptor fingerprint is required for the development of vascular targeted therapies. Several features of the biology of cell surface receptors point to rationality of receptor-ligand identification for intact receptor molecules embedded in cell membranes instead of isolated receptors. As opposed to purified receptors, membrane-bound proteins are more likely to preserve their active conformation, which can be lost upon purification and immobilization once proteins are removed from their natural environment on the cell membrane. In addition, many cell surface receptors are active as homodimers or heterodimers, whose formation may require the cell membrane environment; these interactions further contribute to the ligand specificities of some receptors. Combinatorial approaches for probing the molecular heterogeneity of cell surfaces allow the identification of cell membrane ligands in an unbiased functional assay and without any predetermined notions of the cell surface receptor repertoire; thus, unknown receptors can be targeted. Nonetheless, the great complexity of cell surface molecules still presents a challenge for the isolation of highly specific ligands for a given cell population. In recent years, a number of successful cell biopannings done with phage display have been reported. Examples include cells expressing the urokinase receptor and the melanocortin receptor, fibroblasts and myoblasts, endothelial cells, neutrophils, T cells, head and neck carcinoma cells, and others (21). This relative success notwithstanding, cell surface selection of phage libraries has been plagued by technical difficulties. First, a high number of nonbinder and nonspecific binder clones are recovered when phage libraries are incubated with cell suspensions or monolayers. Moreover, removal of the background by repeated washes is both labor-intensive and inefficient. Finally, cells and potential ligands are frequently lost during the many washing steps required. We have developed a new approach for the screening cell surface-binding peptides from phage libraries. To circumvent some of the practical difficulties in probing the cell surface, i.e., the recovery of nonspecific clones and the loss of cells and subsequent specific phage clones, we developed a method called Biopanning and Rapid Analysis of Selective Interactive Ligands (termed BRASIL) for probing the cell surface with phage display libraries (22). The BRASIL method is based on differential centrifugation in which cells of interest incubated with a phage library in an aqueous upper phase are centrifuged through an organic phase separating the unbound phage from the specific phage-cell complex. The BRASIL method is an efficient and convenient technique for the selection of phage that binds specifically to a given cell surface. Since multiple samples and several subtraction rounds can be done in a relatively short amount of time, applications for high-throughput screenings are apparent. We have recently performed a screening of the NCI 60-cell panel using the BRASIL method creating a ligand “fingerprint” for the NCI 60-cell panel demonstrating the applicability of the BRASIL method for high-throughput analysis (Bover et al., in progress). Furthermore, the BRASIL method is also more sensitive and more specific than phage selection techniques relying on washing of the cells, representing a significant improvement over conventional cell-panning methods. The use of the BRASIL method is not limited to random peptide

Endothelial cells in health and disease

124

libraries or mammalian cells, but can be used to screen antibody fragment displaying phage libraries and as well as other cell types. Since our longstanding interest has been targeting the vascular endothelium, we screened a phage display random peptide library on activated, VEGF165-stimulated endothelial cells after a library subtraction step with quiescent endothelial cells (22). We subsequently isolated a peptide ligand for VEGF receptor. This VEGF-receptor ligand appears to be a chimera between overlapping binding sites on different VEGF-B isoforms, since part of it resembles a neuropilin-1-binding site found in VEGF-B167 and a part of it resembles a neuropilin-1-binding epitope of VEGF-B186 (23). Our chimeric peptide ligand interacts specifically with VEGF receptors in a pattern consistent with VEGF-B-type ligands (24) as confirmed by binding assays with individual phage on a panel of purified targets. We further examined the ability of the synthetic VEGF-receptor ligand to block phage binding to VEGF receptors in vitro and found that the isolated peptide ligand is about 100-fold more efficient in blocking phage binding to VEGFR-1 than to neuropilin-1. Because of the observed differential interaction of the chimeric peptide ligand with its receptors, it is tempting to speculate that there are differences in the number of peptide-binding sites on VEGFR-1 and neuropilin-1, or alternatively in the affinity of the interaction at each binding site. The isolation and further elucidation of vascular receptor-ligand interaction attest to the belief that vascular targets can be found on endothelial cell membranes in vitro. Important application for the BRASIL method can be foreseen in both targeting and identification of ligand-receptor pairs in cell populations derived from patient samples. The method may be used in tandem with fluorescence-activated cell sorting of leukemic cells obtained from bone marrow aspirates from patients or even with circulating endothelial progenitor cells from periferal blood. Tumor- and inflammatory cells from ascites or fine-needle aspirates of solid tumors also seem like ideal clinical material for the identification of novel tumor antigens with the BRASIL method. Moreover, because multiple samples and several selection rounds can be performed in a short amount of time, automation for high-throughput clinical applications is likely to follow.

3.2. In Vivo Targeting 3.2.1. Vascular Targeting Technology We have developed a technology for the identification of protein-protein interactions specific to a given site and functional status of vascular endothelium. Vascular targeting technology is used to identify molecular differences in blood vessels of different organs and tissues, as well as differences between normal blood vessels and angiogenic blood vessels (12,25–27), and is based on the isolation and identification of peptides from random displayed phage display libraries homing to specific vascular beds after an intravenous administration of the library. The ability of individual peptides to target a tissue can also be analyzed by this method (12,25,26). This work has uncovered a ligandreceptor-based molecular signature of the vasculature that allows both tissue-specific targeting of normal blood vessels as well as targeting of angiogenic blood vessels in tumors and other pathological conditions.

The use of proteomics to map phenotypic heterogeneity

125

We have been using animal models of human cancer to discover addresses that are present in tumor blood vessels but not in blood vessels of normal tissues. Based on our previous work, we have found vascular receptors in tumor-bearing mice, and have been able to show that they are also present in human tumors by staining human tissues with antibodies against the molecules (14,25,28). Thus, it is possible to use animal models to find molecular markers of human blood vessels. However, it is unknown whether targeted delivery will always be achieved in humans by using mouse-derived probes. Extrapolation of the results from mouse experiments to human biology requires that the molecules of interest be expressed and regulated similarly in both species. It has recently become evident that this is not always the case, offering one likely explanation for the difficulties translating information derived from mouse models into clinical applications. Several examples of cross-species variation of gene expression patterns within the vascular network have recently surfaced. For example, the prostate-specific membrane antigen (PSMA) shows notably different expression pattern in human and mice. In prostate PSMA in human is expressed; in mouse on the other hand, the expression of PSMA is limited to the brain and kidney (29). Additionally, PSMA is an endothelial cell marker of human tumor blood vessels (30), whereas mouse tumor blood vessels do not have a detectable endothelial expression of PSMA (W.D.Heston, personal communication). Another example of cross-species variation is the TEM7 gene, which is highly and selectively expressed in the endothelium of human colorectal adenomas (31). In mouse, TEM7 gene is expressed in Purkinje cells of the cerebellum, while the tumor blood vessels show no mTEM7 expression (32). There are also species-specific differences in the induction of protein expression by cytokines. For example, tumor necrosis factor-α (TNF-α) and oncostatin M function cooperatively to induce vascular expression of P- and E-selectin in mice, but diverge significantly in their effects on expression of P- and E-selectin in humans or nonhuman primates (33). These prominent species-specific differences in protein expression patterns and ligand-receptor accessibility prompt us to carefully evaluate the information obtained from animal studies before directly applying it to clinical studies. Clearly, the construction of a human vascular map will be of essence in the successful translation of vascular targeting into clinical practice.

3.2.2. Vascular Addresses in Blood Vessels Since 1996, when in vivo phage display methodology was first described (26), numerous murine tissue-specific endothelial cells markers and their peptide ligands have been identified this way by our laboratory and others (reviewed in Refs. 21 and 34). These include peptide ligands targeting brain and kidney (26), lung, skin, pancreas, intestine, uterus, adrenal gland and retina (12), muscle (35), prostate (36), normal and malignant breast tissue (37), lymph nodes (38,70), and placenta (39). In vivo phage display approach has also revealed a vascular address system in tumor blood vessels (14,25,27,28,40–43) as well as in tumor lymphatic vessels (44). Furthermore, it has been possible to identify unique molecular signatures by in vivo phage display that are specifically expressed at different stages of tumor progression from a highly proliferative and angiogenic dysplastic lesion to invasive phase in transgenic mouse models (42,43). By using transgenic mouse models for multistage tumorigenesis involving the pancreatic

Endothelial cells in health and disease

126

islet of Langerhans, peptides discriminating between the vasculature of the premalignant angiogenic islets and fully developed tumors were identified (43). A similar study has also been carried out with a transgenic model of squamous cell carcinoma, further confirming the findings of a tumor stage-specific vascular signature (42). Peptide ligands homing to atherosclerotic regions of blood vessel walls have been identified by using in vivo phage display in mice deficient in low-density lipoprotein (45). For isolation of peptide ligands to receptors in human synovium vasculature, in vivo phage display has been used in conjunction with mice transplanted with human synovium tissue (46). Since the blood vessels from human synovium grafts form functional anastomosis with murine subdermal vessels and support the adhesion and extravasation of human leukocytes into the grafts (47), it is likely that the peptide ligands isolated from human synovium xenografts target human vasculature supporting human tissue (46). We have recently taken the first steps to construct the human molecular vascular map (48). A patient with Waldenström macroglobulinemia, who after massive intracranial bleeding remained comatose with progressive and irreversible loss of brainstem function until meeting the formal criteria for brain-based determination of death (49), received an intravenous infusion of a CX7C (C=cysteine, X=any amino acid) phage random peptide library. In order to recover phage from various tissues, samples were obtained from bone marrow, prostate, liver, fat-tissue, skeletal muscle, and skin. Phage isolated from each tissue were processed for sequencing of the peptide inserts. To analyze the distribution of inserts from the random peptide library, we designed a high-throughput pattern recognition software for the analysis of short amino acid residue sequences. This analysis was applied for phage recovered from each target tissue and for the unselected CX7C random phage display peptide library. Briefly, we compared the relative frequencies of every tripeptide motif from each target tissue (47, 160 tripeptide motifs in total) to those of the motifs from the

unselected library to test for randomness of distribution. Comparisons of the motiffrequencies in a given organ relative to those frequencies in the unselected libraryshowed a nonrandom nature of the peptide distribution; such a bias is remarkablegiven that only a single round of in vivo screening was performed. Of the tripeptide motifs recovered from tissues, some were preferentially found in a single site, whereas others were recovered from multiple sites. This indicates that some of the recovered peptides home in a tissue-specific manner binding to differentially expressed endothelial cell markers in a given tissue, whereas others bind to ubiquitous endothelial cell surface molecules. Further analysis of the original phage peptide inserts revealed four to six amino acid residue motifs that were shared among multiple peptides isolated from a given organ. Each of these motifs were searched for similarities to known proteins in online databases, and found that some of the enriched peptide motifs appeared within known human proteins. Since our screening method isolates ligands for differentially expressed vascular receptors, our recovered peptide motifs are likely to mimic epitopes present in circulating ligands interacting with endothelial cell surface molecules. These circulating ligands may be either secreted proteins or surface receptors present on circulating cells interacting with the target tissue. We were able to identify a panel of candidate human proteins potentially mimicked by selected peptide motifs. For

The use of proteomics to map phenotypic heterogeneity

127

example, one of the peptide motifs identified from the bone marrow is contained within bone morphogenetic protein-3B, which is a growth factor known to regulate bone development (50). It is reasonable to expect that the isolated peptide ligand mimics this protein. Similarly, we also identified interleukin 11 (IL-11) as a ligand mimicked by a peptide specifically enriched in the prostate tissue. Interleukin 11 has been previously shown to signal through the IL-11 receptors within endothelium and prostate epithelium (51,52). This IL-11 mimicking peptide specifically bound to the endothelium and to the epithelium of normal prostate in phage overlay assay with human tissue sections, on the other hand, IL-11 mimetope failed to bind other organs, such as skin. In contrast, a phage isolated from the skin did not bind to prostate or other tissues; instead, this phage specifically identified blood vessels in the skin. The binding of the IL-11 mimetope peptide to IL-11Rα was also verified in vitro. Validation of the ligand-receptor interaction has confirmed that our high-throughput identification of circulating peptide ligands does provide us with functional information in vascular biology in addition to organ homing ligands useful by themselves for vascular targeting. Additional support for the use of combinatorial screenings in patients (48) for the development of anticancer targeted therapies comes from our studies, where we show the potential of IL11Rα as a target for intervention in human prostate cancer (53). Expression of IL-11Rα was increased in primary and metastatic prostate cancer and its associated blood vessels in a stage-specific manner during disease progression. Furthermore, a peptide guided by the IL-11 mimetope peptide linked to the well-established proapoptotic peptide D(KLAKLAK)2 (54) was specifically targeted and internalized into prostate cancer cells resulting in apoptosis (53). Integration of the isolation of ligands from the in vivo screenings to proteomic strategies to identify the receptors for these ligands has produced an array of tissuespecific vascular receptors. Complementary genetic and biochemical approaches have been used to identify receptors for tissue homing peptides. Membrane dipeptidase on lung endothelium was identified as the receptor for GFE-peptide (13), aminopeptidase N on angiogenic vasculature as the receptor for the tumor homing NGR-peptide (14), aminopeptidase P on both normal and malignant breast tissue as the receptor for CPGPEGAGC-peptide (37) and FcRn/β2-microglobulin as a target for the TPKTSVTpeptide in placenta (39). Moreover, the ligands themselves may be used as either drug discovery leads or for therapeutic modulation of their corresponding receptor(s). Since the early days of mapping addresses in the blood vessels of normal and diseased organs (26), we have now moved to methodologies that enable us to identify more vascular addresses (48) in even more refined time and space restricted expression patterns. In order to proceed from the ligands identified by phage display to receptors in the blood vessels, sophisticated array technology and state-of-the-art proteomics are needed. Array technology allows the simultaneous analysis of thousands of molecular parameters within a single experiment. DNA array technologies evolved to allow smaller sample volumes, more efficient analyses and higher throughput. Since proteins are more complex and diverse biomolecules than nucleic acids, development of similar platforms in the protein field has proved more difficult. We plan to explore current techniques used in the generation and development of protein arrays (55) and their application in proteomics to identify receptors targeted by peptides and antibodies of interest in humans. Ultimately, our goal is to identify a large-scale panel of receptors in human

Endothelial cells in health and disease

128

blood vessels. In addition, antibodies raised against each target could be arrayed and used to profile the proteins present in, for example, a cell or tissue. Besides offering a way of identifying endothelial cell surface markers accessible to the circulation and providing novel tools for selective vascular targeting, in vivo phage display studies further our understanding of organ and tumor endothelium specificity and define the role that endothelial cell markers play in angiogenesis. It is recognized that a tumor cannot get larger than 1 mm in diameter without new blood vessel formation. The rules of the normal blood vessel formation do not seem to apply to tumor blood vessels and there is a tremendous amount of functional and structural irregularity and molecular heterogeneity in tumor/angiogenic blood vessels when compared with normal blood vessels in the same tissue environment. Some of the vascular markers found in tumor blood vessels are vascular proteases that not only serve as receptors for circulating ligands but also modulate angiogenesis (14,28,41). We have identified by in vivo phage display CD13/amiriopeptidase N (APN) and aminopeptidase A (APA) as functional targets in angiogenic vasculature that may contribute to an important regulatory proangiogenic pathway (14,28). These peptidases that are expressed in both the endothelial and periendothelial cell compartment of angiogenic blood vessels are accessible to circulating ligands. Their enzyme activities regulate the angiogenesis process, since either genetic or biochemical ablation of the activity of these enzymes significantly reduces the formation of new blood vessels in several pathological states, such as cancer and retinopathies. The function of CD13/APN appears to depend on the availability of its substrates, thus its location: in synaptic membranes, it metabolizes enkephalins and endorphins; in the intestinal brush border, it degrades regulatory peptides and scavenges amino acids; in lymphocytes, its activity is associated with mitotic activation, antigen processing, cell adhesion, and migration (reviewed in Ref. 56). Aminopeptidase A also appears to be a molecule that has different functions, according to the organ and time period examined. A broad spectrum of tissues expresses APA (57), but its only well understood role is the conversion of angiotensin II to angiotensin III in the rennin-angiotensin system (58). An intriguing feature of CD13/APN is that its expression and enzymatic activity can be physiologically regulated. Its activity and substrate specificity depend on conformational changes induced by various stimuli, including proliferative signals. Studies using monoclonal antibodies indicate that CD13/APN undergoes regulatory intramolecular alterations that result in exposure of cryptic sites and regulation of enzyme activity (59). The immunoreactivity patterns obtained with cultured cells and tissue sections from kidney, breast, and prostate carcinomas suggest that different CD13/APN forms are expressed in myeloid cells, epithelia, and tumor-associated blood vessels (60). Association with proteins or factors present only in the tumor microenvironment might cause differential reactivity or accessibility to different CD13/APN ligands, such as the tumor vasculature targeting NGR-peptide ligand. This example of CD13/APN clearly demonstrates the importance of utilizing a functional proteomics method, such as in vivo phage display, for the identification of ligands that target molecules with complex regulatory mechanisms and expression patterns in time and space. Following the identification of the CD13/APN peptide ligand homing to angiogenic vasculature, it has been successfully used to target an array of therapies to tumor vasculature ranging from cytotoxic drugs to proapoptotic moieties (see Sec. 3 of this chapter).

The use of proteomics to map phenotypic heterogeneity

129

4. TARGETED THERAPIES Many malignant, cardiovascular, and inflammatory diseases have a marked angiogenic component. In cancer, tumor vasculature is a suitable target for intervention because the vascular endothelium is composed of nonmalignant cells that are genetically stable but epigenetically diverse. Cancers appear to put an epigenetic molecular signature on their own blood vessels that targeted probes can use as a homing signal to deliver a drug into the vascular receptors. The biological basis for such molecular signature of the tumor vasculature is still largely unknown. However, peptides selected by homing to blood vessels have been used to guide the delivery of cytotoxic drugs (25), proapoptotic peptides (36,54), metalloprotease inhibitors (41), cytokines (61), genes (62,63, Hajitou et al., in progress), and liposomes (64) to receptors in the angiogenic vasculature showing marked therapeutic efficacy in tumor-bearing mouse models. Tumor targeting peptide ligands can also deliver imaging agents to tumor vasculature (65). Generally, coupling to homing peptides appears to yield a targeted compound with a better therapeutic index than the untargeted parental compound. Clearly, there is also an advantage in the pharmacokinetics of a drug being targeted to the vascular endothelium, which is directly accessible upon intravenous administration, rather than tumor cells. Caveolae has also surfaced as an interesting new target for vascular drug delivery. Caveolae is specialized distinct plasma membrane microdomains and the associated noncoated plasmalemmal vesicles (66,67) on several cell types, including the cells of the continuous microvascular endothelium. Many proteins have been found enriched in caveolae, including cell surface receptors such as platelet-derived growth factor receptors, epithelial growth factor receptors, basic fibroblast growth factor receptor, and endothelin receptors (reviewed in Ref. 68). Even though the molecular differences of caveolae between endothelial cells derived from different tissues remain unknown, recent data suggest that caveolae can contain tissue-specific cell surface molecules. A lung endothelial cell specific antibody targeting the caveolae has been generated using an antibody and subfractionation strategy. Upon intravenous administration, this antilung caveolae antibody localized to the microvascular endothelium of rat lungs. In addition, targeting caveolae increased trasendothelial transport of the anticaveolae antibody (69). These preclinical data suggest that targeting the vasculature of normal and pathological tissues may be the basis of a new pharmacology for the treatment of malignant and inflammatory diseases by delivering therapeutic agents to blood vessels.

5. CONCLUSION A major goal in drug development has long been to develop technology for targeting therapeutics more effectively to their intended disease site and to improve their therapeutic index by limiting the systemic exposure of other tissues to untoward or toxic effects. The technologies reviewed here have two main applications. First, they may identify vascular targeting ligands. Second, they may enable the construction of a molecular map of human vascular receptors. In theory, targeted delivery of drugs,

Endothelial cells in health and disease

130

liposomes, peptide sequences, gene therapy vectors, and biological therapies can be achieved in clinical applications. Ultimately, it may be possible to guide imaging or therapeutic compounds to the target site in real clinical situations. In the future, the determination of molecular profiles of blood vessels in specific conditions may also lead to vascular targets. Early identification of targets, optimized regimens tailored to molecular profile of individual patients, and identification of new vascular addresses may lead to revisiting or salvaging of ineffective or toxic drug candidates.

ACKNOWLEDGMENTS Our work has been funded in part by grants from NIH (CA088106, CA078512, CA090270, CA082976, CA051134 and R14101-7200003 to R.P.; CA0103042, CA090270, CA090810, DR06783, CA103030 and CA103086 to W.A.), Juvenile Diabetes Research Foundation (to W.A.), and awards from the Gilson-Longenbaugh Foundation and Angelworks (to R.P. and W.A.). J.L. received fellowships from the Susan G.Komen Breast Cancer Foundation and the Cancer Society of Finland.

REFERENCES 1. Smith GP. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science 1985; 228:1315–1317. 2. Barbas CF III, Kang AS, Lerner RA, Benkovic SJ. Assembly of combinatorial antibody libraries on phage surfaces: the gene III site. Proc Natl Acad Sci USA 1991; 88:7978–7982. 3. McCafferty J, Griffiths AD, Winter G, Chiswell DJ. Phage antibodies: filamentous phage displaying antibody variable domains. Nature 1990; 348:552–554. 4. Smith GP, Scott JK. Libraries of peptides and proteins displayed on filamentous phage. Methods Enzymol 1993; 217:228–257. 5. Hoogenboom HR. Overview of antibody phage-display technology and its applications. Methods Mol Biol 2002; 178:1–37. 6. Azzazy HM, Highsmith WE Jr. Phage display technology: clinical applications and recent innovations. Clin Biochem 2002; 35:425−445. 7. Kay BK, Hamilton PT. Identification of enzyme inhibitors from phage-displayed combinatorial peptide libraries. Comb Chem High Throughput Screen 2001; 4:535–543. 8. Ringe D, Mattos C. Analysis of the binding surfaces of proteins. Med Res Rev 1999; 19:321– 331. 9. Vendruscolo M, Paci E, Dobson CM, Karplus M. Three key residues form a critical contact network in a protein folding transition state. Nature 2001; 409:641–645. 10. Ruoslahti E. RGD and other recognition sequences for integrins. Annu Rev Cell Dev Biol 1996; 12:697–715. 11. Koivunen E, Ranta TM, Annila A, Taube S, Uppala A, Jokinen M, van Willigen G, Ihanus E, Gahmberg CG. Inhibition of beta(2) integrin-mediated leukocyte cell adhesion by leucineleucine-glycine motif-containing peptides. J Cell Biol 2001; 153:905–916. 12. Rajotte D, Arap W, Hagedorn M, Koivunen E, Pasqualini R, Ruoslahti E. Molecular heterogeneity of the vascular endothelium revealed by in vivo phage display. J Clin Invest 1998; 102:430–437.

The use of proteomics to map phenotypic heterogeneity

131

13. Rajotte D, Ruoslahti E. Membrane dipeptidase is the receptor for a lung-targeting peptide identified by in vivo phage display. J Biol Chem 1999; 274:11593–11598. 14. Pasqualini R, Koivunen E, Kain R, Lahdenranta J, Sakamoto M, Stryhn A, Ashmun RA, Shapiro LH, Arap W, Ruoslahti E. Aminopeptidase N is a receptor for tumor-homing peptides and a target for inhibiting angiogenesis. Cancer Res 2000; 60:722–727. 15. Cheng X, Kay BK, Juliano RL. Identification of a biologically significant DNA-binding peptide motif by use of a random phage display library. Gene 1996; 171:1–8. 16. Matsubara T, Ishikawa D, Taki T, Okahata Y, Sato T. Selection of ganglioside GM1-binding peptides by using a phage library. FEBS Lett 1999; 456:253–256. 17. Peletskaya EN, Glinsky G, Deutscher SL, Quinn TP. Identification of peptide sequences that bind the Thomsen-Friedenreich cancer-associated glycoantigen from bacteriophage peptide display libraries. Mol Divers 1996; 2:13–18. 18. Wang S, Humphreys ES, Chung SY, et al. Peptides with selective affinity for carbon nanotubes. Nat Mater 2003; 2:196–200. 19. Rodi DJ, Janes RW, Sanganee HJ, Holton RA, Wallace BA, Makowski L. Screening of a library of phage-displayed peptides identifies human bcl-2 as a taxol-binding protein. J Mol Biol 1999; 285:197–203. 20. Mintz PJ, Kim J, Do KA, et al. Fingerprinting the circulating repertoire of antibodies from cancer patients. Nat Biotechnol 2003; 21:57–63. 21. Kolonin M, Pasqualini R, Arap W. Molecular addresses in blood vessels as targets for therapy. Curr Opin Chem Biol 2001; 5:308–313. 22. Giordano RJ, Cardó-Vila M, Lahdenranta J, Pasqualini R, Arap W. Biopanning and rapid analysis of selective interactive ligands. Nat Med 2001; 7:1249–1253. 23. Mäkinen T, Olofsson B, Karpanen T, Hellman U, Soker S, Klagsbrun M, Eriksson U, Alitalo K. Differential binding of vascular endothelial growth factor B splice and proteo-lytic isoforms to neuropilin-1. J Biol Chem 1999; 274:21217–21222. 24. Olofsson B, Jeltsch M, Eriksson U, Alitalo K. Current biology of VEGF-B and VEGF-C. Curr Opin Biotechnol 1999; 10:528–535. 25. Arap W, Pasqualini R, Ruoslahti E. Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 1998; 279:377–380. 26. Pasqualini R, Ruoslahti E. Organ targeting in vivo using phage display peptide libraries. Nature 1996; 380:364–366. 27. Pasqualini R, Koivunen E, Ruoslahti E. Alpha v integrins as receptors for tumor targeting by circulating ligands. Nat Biotechnol 1997; 15:542–546. 28. Marchiò S, Lahdentranta J, Schlingemann RO, et al Aminopeptidase A is a functional target in angiogenic blood vessels. Cancel Cell 2004; 5:151–162. 29. Bacich DJ, Pinto JT, Tong WP, Heston WD. Cloning, expression, genomic localization, and enzymatic activities of the mouse homolog of prostate-specific membrane antigen/NAALADase/folate hydrolase. Mamm Genome 2001; 12:117–123. 30. Chang SS, Reuter VE, Heston WD, Bander NH, Grauer LS, Gaudin PB. Five different antiprostate-specific membrane antigen (PSMA) antibodies confirm PSMA expression in tumorassociated neovasculature. Cancer Res 1999; 59:3192–3198. 31. St Croix B, Rago C, Velculescu V, et al. Genes expressed in human tumor endothelium. Science 2000; 289:1197–1202. 32. Carson-Walter EB, Watkins DN, Nanda A, Vogelstein B, Kinzler KW, St Croix B. Cell surface tumor endothelial markers are conserved in mice and humans. Cancer Res 2001; 61:6649–6655. 33. Yao L, Setiadi H, Xia L, Laszik Z, Taylor FB, McEver RP. Divergent inducible expression of P-selectin and E-selectin in mice and primates. Blood 1999; 94:3820–3828. 34. George AJ, Lee L, Pitzalis C. Isolating ligands specific for human vasculature using in vivo phage selection. Trends Biotechnol 2003; 21:199–203. 35. Samoylova TI, Smith BF. Elucidation of muscle-binding peptides by phage display screening. Muscle Nerve 1999; 22:460–466.

Endothelial cells in health and disease

132

36. Arap W, Haedicke W, Bernasconi M, Kain R, Rajotte D, Krajewski S, Ellerby HM, Bredesen DE, Pasqualini R, Ruoslahti E. Targeting the prostate for destruction through a vascular address. Proc Natl Acad Sci USA 2002; 99:1527–1531. 37. Essler M, Ruoslahti E. Molecular specialization of breast vasculature: a breast-homing phagedisplayed peptide binds to aminopeptidase P in breast vasculature. Proc Natl Acad Sci USA 2002; 99:2252–2257. 38. Trepel M, Arap W, Pasqualini R. Modulation of the immune response by systemic targeting of antigens to lymph nodes. Cancer Res 2001; 61:8110–8112. 39. Kolonin MG, Pasqualini R, Arap W. Teratogenicity induced by targeting a placental immunoglobulin transporter. Proc Natl Acad Sci USA 2002; 99:13055–13060. 40. Burg MA, Pasqualini R, Arap W, Ruoslahti E, Stallcup WB. NG2 proteoglycan-binding peptides target tumor neovasculature. Cancer Res 1999; 59:2869–2874. 41. Koivunen E, Arap W, Valtanen H, et al. Tumor targeting with a selective gelatinase inhibitor. Nat Biotechnol 1999; 17:768–774. 42. Hoffman JA, Giraudo E, Singh M, Zhang L, Inoue M, Porkka K, Hanahan D, Ruoslahti E. Progressive vascular changes in a transgenic mouse model of squamous cell carcinoma. Cancer Cell 2003; 4:383–391. 43. Joyce JA, Laakkonen P, Bernasconi M, Bergers G, Ruoslahti E, Hanahan D. Stage-specific vascular markers revealed by phage display in a mouse model of pancreatic islet tumorigenesis. Cancer Cell 2003; 4:393–403. 44. Laakkonen P, Porkka K, Hoffman JA, Ruoslahti E. A tumor-homing peptide with a targeting specificity related to lymphatic vessels. Nat Med 2002; 8:751–755. 45. Houston P, Goodman J, Lewis A, Campbell CJ, Braddock M. Homing markers for atherosclerosis: applications for drug delivery, gene delivery and vascular imaging. FEBS Lett 2001; 492:73–77. 46. Lee L, Buckley C, Blades MC, Panayi G, George AJ, Pitzalis C. Identification of synoviumspecific homing peptides by in vivo phage display selection. Arthritis Rheum 2002; 46:2109– 2120. 47. Wahid S, Blades MC, De Lord D, Brown I, Blake G, Yanni G, Haskard DO, Panayi GS, Pitzalis C. Tumour necrosis factor-alpha (TNF-alpha) enhances lymphocyte migration into rheumatoid synovial tissue transplanted into severe combined immunodeficient (SCID) mice. Clin Exp Immunol 2000; 122:133–142. 48. Arap W, Kolonin MG, Trepel M, et al. Steps toward mapping the human vasculature by phage display. Nat Med 2002; 8:121–127. 49. Wijdicks EF. The diagnosis of brain death. N Engl J Med 2001; 344:1215–1221. 50. Daluiski A, Engstrand T, Bahamonde ME, Gamer LW, Agius E, Stevenson SL, Cox K, Rosen V, Lyons KM. Bone morphogenetic protein-3 is a negative regulator of bone density. Nat Genet 2001; 27:84–88. 51. Campbell CL, Jiang Z, Savarese DM, Savarese TM. Increased expression of the interleu-kin-11 receptor and evidence of STAT3 activation in prostate carcinoma. Am J Pathol 2001; 158:25– 32. 52. Mahboubi K, Biedermann BC, Carroll JM, Pober JS. IL-11 activates human endothelial cells to resist immune-mediated injury. J Immunol 2000; 164:3837–3846. 53. Zurita AJ, Troncoso P, Cardo-Vila M, Logothetis CJ, Pasqualini R, Arap W. Combina-torial screenings in patients: the interleukin-11 receptor alpha as a candidate target in the progression of human prostate cancer. Cancer Res 2004; 64:435–439. 54. Ellerby H.M, Arap W, Ellerby LM, et al. Anti-cancer activity of targeted pro-apoptotic peptides. Nat Med 1999; 5:1032–1038. 55. Cahill DJ, Nordhoff E. Protein arrays and their role in proteomics. Adv Biochem Eng Biotechnol 2003; 83:177–187. 56. Riemann D, Kehlen A, Langner J. CD13—not just a marker in leukemia typing. Immunol Today 1999; 20:83–88.

57. Li L, Wu Q, Wang J, Bucy RP, Cooper MD. Widespread tissue distribution of amino-peptidase A, an evolutionarily conserved ectoenzyme recognized by the BP-1 antibody. Tissue Antigens 1993; 42:488–496. 58. Jackson EK. Renin and angiotensin. In: Goodman and Gilman’s The Pharmacological Basis of Therapeutics. Hardman JG, Limbird LE, Goodman Gilman A, eds. McGraw-Hill Medical Publishing Division, 2001:809–841. 59. Xu Y, Wellner D, Scheinberg DA. Cryptic and regulatory epitopes in CD13/amino-peptidase N. Exp Hematol 1997; 25:521–529. 60. Curnis F, Arrigoni G, Sacchi A, Fischetti L, Arap W, Pasqualini R, Corti A. Differential binding of drugs containing the NGR motif to CD13 isoforms in tumor vessels, epithelia, and myeloid cells. Cancer Res 2002; 62:867–874. 61. Curnis F, Sacchi A, Borgna L, Magni F, Gasparri A, Corti A. Enhancement of tumor necrosis factor alpha antitumor immunotherapeutic properties by targeted delivery to aminopeptidase N (CD13). Nat Biotechnol 2000; 18:1185–1190. 62. Grifman M, Trepel M, Speece P, Gilbert LB, Arap W, Pasqualini R, Weitzman MD. Incorporation of tumor-targeting peptides into recombinant adeno-associated virus cap-sids. Mol Ther 2001; 3:964–975. 63. Muller OJ, Kaul F, Weitzman MD, Pasqualini R, Arap W, Kleinschmidt JA, Trepel M. Random peptide libraries displayed on adeno-associated virus to select for targeted gene therapy vectors. Nat Biotechnol 2003; 21:1040–1046. 64. Pastorino F, Brignole C, Marimpietri D, Cilli M, Gambini C, Ribatti D, Longhi R, Allen TM, Corti A, Ponzoni M. Vascular damage and anti-angiogenic effects of tumor vessel-targeted liposomal chemotherapy. Cancer Res 2003; 63:7400–7409. 65. Hong FD, Clayman GL. Isolation of a peptide for targeted drug delivery into human head and neck solid tumors. Cancer Res 2000; 60:6551–6556. 66. Palade GE. The fine structure of blood capillaries. J Appl Physiol 1953; 24:1424. 67. Peters KR, Carley WW, Palade GE. Endothelial plasmalemmal vesicles have a characteristic striped bipolar surface structure. J Cell Biol 1985; 101:2233–2238. 68. Zajchowski LD, Robbins SM. Lipid rafts and little caves. Compartmentalized signalling in membrane microdomains. Eur J Biochem 2002; 269:737–752. 69. McIntosh DP, Tan XY, Oh P, Schnitzer JE. Targeting endothelium and its dynamic caveolae for tissue-specific transcytosis in vivo: a pathway to overcome cell barriers to drug and gene delivery. Proc Natl Acad Sci USA 2002; 99:1996–2001. 70. Kolonin MG, Saha PK, Chan L, Pasqualini R, Avap W. Reveusal of obesity by targeted ablation of adipose tissue. Nat Med 2004; 10:625–632.

6 The Use of Genomics to Map Phenotypic Heterogeneity of the Endothelium Mary E.Gerritsen, Stuart Hwang, Constance Zlot, James Tomlinson, and Michael Ziman Department of Vascular Biology, Millennium Pharmaceuticals, Inc., South San Francisco, California, U.S.A.

1. INTRODUCTION The vascular endothelium lines the blood vessels of the body, and in man, the estimated number of individual cells comprising this line is somewhere in the order of 1–6×1013. The magnitude of the surface area of this lining is also impressive, in the order of 719m2 (1). Although scientists initially thought this cell layer was essentially an inert barrier between the blood and the tissue, this extensive “organ” is now known to carry out a diverse array of specialized functions. Moreover, these functions vary markedly from one vascular bed to another. Quiescent endothelium presents a non thrombogenic surface and plays an important role in the regulation of the transit of the solutes, proteins, and cellular components of the blood. Perturbation of the cells by various stimuli, be they physical, biochemical, or mechanical, can result in the marked shift in the endothelial phenotype; a transition achieved by numerous mechanisms including changes in gene expression, protein expression, and protein phosphorylation. It is important to recognize the fact that there is no “generic endothelial cell,” there are actually many different populations of cells with remarkable heterogeneity in structure and biology. At an ultrastructural level, early electron microscopic studies demonstrated remarkable differences in the anatomy of endothelial cells (2,3). As of Majno (2) so aptly stated “Electron microscopy has revealed that there are almost as many varieties of ‘capillaries’ as there are organs and tissues.” One of the more obvious differences is in the continuity of the endothelial cells of the capillaries, leading to the definition of several broad categories of endothelial cells: continuous, discontinuous (sinusoidal), and fenestrated. Endothelium of the continuous type is the most common, and can be found in the walls of arterioles, capillaries, and venules of skeletal, smooth, and cardiac muscle, the mesentery, skin, connective tissue, lung, brain and eye, as well as lining the major conduit vessels (large arteries, veins). These endothelial cells are characterized by

The use of genomics to map phenotypic heterogeneity

135

occluding (tight) junctions. However, although structurally similar, continuous endothelia exhibit remarkable differences in their relative permeability to different solutes. Fenestrated endothelium is found in the exchange vessels of secretory and excretory organs (i.e., exocrine and endocrine glands), the gastric mucosa, the kidneys (glomerulus and peritubular capillaries, synovium, and choroid plexus). Numerous small “windows” (i.e., small transcellular openings ranging from 50 to 80 nm in diameter, as the name implies) characterize fenestrated endothelia. At the ultrastructural level, differences in the fine structure of the fenestrae are revealed. In some organs, the fenestrae appear open; in others, the fenestrae are “closed” by thin membranous diaphragm like structures. Discontinuous endothelia may be thought of as an extreme form of “fenestrated endothelium” and can be found in the sinusoids of the liver, spleen, and bone marrow. These cells are found in thin-walled blood vessels with irregular outlines and calibers, and highly fenestrated structures. In the liver, the fenestrae are often clustered to form sieve plates, and none of the fenestrae are closed with diaphragms. There are additional endothelial phenotypes; for example, another highly specialized endothelial structure is found in the high endothelial venules of peripheral lymph nodes. These cells are characterized by a cuboidal morphology with particularly well-developed Golgi complex, rough endoplasmic reticulum, and poly-ribosomes. The cell-cell junctions are generally discontinuous with few and then poorly organized tight junctions where they do occur (reviewed in Ref. 4). At a functional level, heterogeneity in endothelial cell biology is now widely recognized. Different vascular beds demonstrate a wide range of solute and protein permeabilities; although some of this is dictated by the aforementioned structural features of the endothelium regional differences in the expression of plasmalemmal vesicles, different receptors, transporters, and basement membrane also contribute (5,6). Exquisite differences in the composition of the endothelial glycocalyx have been revealed by elegant lectin binding studies (7–11). Endothelial cells at different vascular sites also appear to have specialized responses to specific stimuli such as cytokines, shear forces, or growth factors adding an additional layer of complexity to the concept of endothelial heterogeneity. Indeed, the expression of unique endothelial membrane proteins under basal or perturbed conditions provides the opportunity for vascular targeting, i.e., the ability to target antibodies or drugs to specific vascular beds. Vascular targeting offers exciting new therapeutic opportunities for the treatment of cancer, chronic inflammatory diseases, and other disorders (12).

2. GENOMICS AND ENDOTHEUAL DIVERSITY What dictates the phenotype of an endothelial cell? At least part of the heterogeneity derives from interactions of the endothelium with the organ or tissue environment, through soluble factors, cell-cell interactions, or cell-matrix interactions. An interesting question, now addressable at a whole genome level, is whether or not there are differences in the “genetics” of endothelial cells. Various approaches to address this issue are available; laser-capture microdissection of endothelial cells in situ, or isolation and culture of the cells from different organs. These cells can be used to generate RNA

Endothelial cells in health and disease

136

samples that can be further used for transcriptional analysis. Another approach is to map proteomic mapping proteomic differences using phage display. When endothelial cells are removed from the tissue and grown in culture, some of the differentiated features appear to be lost. For example, endothelial cells derived from the cerebral cortex lose many of their blood-brain barrier properties (13). Endothelial cells from endocrine organs lose their fenestrations (14), and endothelial cells from peripheral lymphatic tissues lose their high endothelial morphology (15). However, other differentiated features appear to be well preserved. The endothelial markers VE-cadherin, CD31 and intercellular adhesion molecule (ICAM)-2 are expressed in virtually all endothelial cells in vivo, and are expressed in virtually all properly identified cultured endothelial cells (16). Like their in vivo counterparts, cultured endothelial cells also respond to cytokines such as interleukin (IL)-1, lipo-polysaccharide (LPS), and tumor necrosis factor (TNF)α, upregulating the expression of adhesion molecules such as vascular cell adhesion molecule (VCAM)-1 and ICAM-1. Various technologies have been used to identify vascular bed specific antigens or markers, including differential lectin binding, raising antibodies to endothelial cells or membranes from specific vascular beds, suppression subtractive hybridization, and differential display. Podograbinska et al. (17) utilized Affymetrix oligonucleotide arrays to probe the molecular features of cultured human lymphatic endothelial cells (LEC) and what these authors defined as blood microvascular endothelial cells (BECs) (note, these might actually be better defined as “nonlymphatic” microvascular endothelial cells to avoid confusion with blood-outgrowth endothelial cells or circulating progenitor cells). These authors purified lymphatic endothelial cells using a magnetic bead approach and an antibody to a lymphatic endothelial cell marker (lymphatic vessel hyaluronic receptor, LYVE-1). After depletion of the LEC, BECs were isolated from the remaining cells using the vascular endothelial cell marker, CD31. Although it is not clear from the methods of this study whether different isolates of LEC and BEC were profiled, or if the data are derived from a single comparison, these authors reported remarkable differences in the molecular signature of LECs vs. BECs. The LECs expressed high levels of genes implicated in protein metabolism, sorting and trafficking, including proteins of the SNARE family, rab GTPases, and sec-related proteins. Shusta et al. (18) used suppressive subtractive hybridization to evaluate differential gene expression within the human brain microvasculature. These authors used isolated microvessels, which are comprised of endothelial cells, pericytes, smooth muscle cells and astrocyte foot processes; thus, the genes identified in this approach are not from purified endothelial cells. The human brain capillary tester cDNA was subtracted with driver cDNA obtained from human liver and kidney RNA. This study identified 20 known genes, 12 genes encoding proteins of unknown function and five novel genes. Several genes participating in the regulation of the endothelial tight junction and cytoskeleton were identified, including claudin 5; additional genes encoding for nutrient or peptide transporters were also identified including ATPase subunit α2, a monomer in a family of subunits that form heterodimers to catalzye the active transport of Na+ and K+ across the cell membrane. Kallman et al. (19) compared the gene expression of primary human cerebral endothelial cells (HCEC) (predominantly microvascular in origin) with human umbilical vein endothelial cells (HUVEC) using a cDNA array of 375 genes. These authors identified 35 genes selectively expressed in the HCEC and 20 in the

The use of genomics to map phenotypic heterogeneity

137

HUVEC under basal conditions. Cerebral endothelial specific genes (basal conditions) included VEGF, insulin growth factor (IGF) binding proteins 1, 5, and 6, follistatin, decorin, a number of chemokines (macrophage migration inhibitory factor (MIF-1), epithelial-derived neutrophil-activating peptide 78 (ENA-78), growth regulated protein (GRO)-α, eotaxin), c-kit ligand, macrophage colony stimulating factor (M-CSF), pleiotrophin, interferon (IFN) -γ receptor 1, c-kit, oncostatin M, platelet derived growth factor receptor subunits α and β, erythroblastic leukemia viral homolog 1 (erbB1), acidic and basic fibroblast growth factor (FGF), FGF5, integrins α1 and (35, IL-1β, IL-6, glial derived neurotrophic factor (GFRα3), semaphorin F, brain derived neurotrophic factor (BDNF), nerve growth factor (NGF) receptor, tissue inhibitor of metalloproteinase (TIMP)-3 transforming growth factor (TGF) β2, activin like kinase (ALK) 3, osteoprotegerin, and HVEM (herpes virus entry mediator). HUVEC specific genes included cadherin 8, CD31, cadherin 5 (VE-Cadherin), ICAM-2, endothelial specific tyrosine kinase receptor 2 (Tie-2), angiopoietin-2 (Ang-2) kinase insert domain receptor (KDR), interleukin 1 receptor-like 1 precursor (ST2 protein), CXC chemokine receptor 4 (CXCR4), ephrin (Eph) A4, EphB2, integrin β8, midkine, endothelial nitric oxide synthase (eNOS), the chemokine orphan receptor RDC-1, matrix metalloproteinase (MMP) 10, MMP13, MMP8, and the TNF superfamily members, Tumor necrosis factor receptor superfamily member 11A precursor (Receptor activator of NF; RANK) and Tumor necrosis factor (ligand) superfamily, member 4 (OX40L). The protein expression of several of the genes identified as HCEC specific (VEGF, follistatin, supercoiling factor (SCF), TGFβ2, IL6, BDNF and monocyte chemoattractant protein (MCP) -1) was confirmed by ELISA. Immunoflourescence studies also confirmed the selective expression of decorin in the cerebral microvasculature in situ (compared to umbilical cord). Arap et al. (20) reported the first in vivo screening of a peptide library in a human patient, surveying 47, 160 motifs that localized to different organs. Peptides that home to specific vascular beds were selected after intravenous administration of a phage-display random peptide library, revealing a vascular address system that could allow tissuespecific targeting of normal blood vessels. These authors identified specific peptide homing motifs that were preferentially identified in bone marrow (GGG, GFS), fat (EGG, LSP), skeletal muscle (LVS), prostate (AGG), skin (GRR, GGH, GTV) as well as motifs that were found in multiple organs (GVL, EGR, FGV, FGG, GER, SGT). Although a lung “specific” receptor was isolated from similar phage studies in mice (membrane dipeptidase) (21), the specific receptors for the aforementioned peptides have not yet been elucidated from the human studies. Several groups have generated genetically modified mice using different promoters coupled to reporter genes, and have identified vascular bed specific expression of certain transgenes. Knock-in approaches allow the use of endogenous promoter and locus for a specific gene whereas transgenic approaches use predefined promoter length and a heterologous genomic locus. For example, Koop et al. (22) generated receptor protein tyrosine phosphatase mu (RPTPmu) -LacZ mice knock-in mice that expressed the beta galactosidase (LacZ) reporter gene under the control of the endogenous RPTPmu promoter. These authors found LacZ expression in endothelial cells of arteries and capillaries, but expression was virtually absent in endothelial cells of veins and in the discontinuous endothelial cells of the adult liver and spleen. Aird et al. (23) generated

Endothelial cells in health and disease

138

transgenic mice using a fragment of the von Willebrand factor (vWF) gene (2182 bp of 5′ flanking sequence, the first exon and first intron) coupled to the LacZ reporter genes. Under the control of this promoter, β-galactosidase expression was detected within endothelial cells of the heart, brain, and skeletal muscle, suggesting that the vWF gene is regulated by vascular bed specific pathways, in response to signals derived from the local milieu. Many other transgenic studies have been carried out with endothelial cell-specific promoters. In virtually every case (with the possible exception of a long fragment of the Tie-2 gene), the promoter was shown to direct expression in a unique and limited subset of endothelial cells.

3. A GENOME-WIDE APPROACH TO EVALUATE HETEROGENEITY OF CULTURED HUMAN ENDOTHELIAL CELLS Recent technological advances have made genome-wide comparisons of the cellular transcriptome feasible and rapid. We have used the Affymetrix oligonucleotide arrays to profile endothelial gene expression in response to various stimuli, and have validated the genes identified by multiple independent methods, demonstrating that this approach is reliable, and robust (24–27). Ideally, obtaining the transcriptome of endothelial cells in situ would be the most desirable. However, although individual cells can be isolated by laser dissection, considerable amplification (and the inherent potential errors of amplification) of the RNA is required to obtain sufficient material for transcriptional profiling. Another approach is to use isolated, cultured endothelial cells from different organs. To determine potential genetic heterogeneity of human endothelial cells, we selected a number of commercially available endothelial cells, all of the so-called “continuous” type. All cell types had been well characterized by the manufacturer (Clonetics); in addition, we confirmed endothelial identity in independent studies assessing the expression of the endothelial marker CD31 and uptake of Di-I-Ac-LDL (>99% purity). Two of the cell preparations were termed microvascular endothelium, and can be considered to contain a mixture of arterial, venous, and capillary endothelial cells (human lung microvessel endothelial cells, HLMVEC and human dermal microvessel endothelial cells, HDMVEC). Two of the cell preparations were derived from large conduit vessels of the arterial tree (human aortic endothelial cells, HAEC and human coronary artery endothelial cells, HCAEC). We also evaluated HUVEC, which, while not necessarily representative of a venous phenotype, are the most widely studied endothelial cell type, facilitating the comparison of our results with the literature. At least two independent cell “lines” (i.e., from different donors) were evaluated of each representative “endothelial cell type”; this approach helps to rule out “individual differences” in gene expression, which could be misinterpreted as heterogeneity per se. To the best of our knowledge, none of the different EC types came from the same donor. To improve the fidelity of the results, all cells were grown under identical conditions, with great care to maintain identical lots of serum, media, plasticware, growth factors, and cytokines. All cells were used at similar “passage” number (5 or 6) and RNA was

The use of genomics to map phenotypic heterogeneity

139

extracted and labeling performed under rigid and identical conditions. A summary of the cell preparation information is provided in Table 1. The profiling experiments were performed on nearly confluent (90–95%) cells that had been incubated in Medium 199 containing 1X ITS (Insulin-transferrin-Selenium-A), 2mM L-glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin (all components from Invitrogen Corp.; Carlsbad, CA) and 1% fetal bovine serum (FBS, Tissue Culture Biologicals; Tulare, CA) for 18 hr. Use of this “Basal Medium” enables the identification of cytokine or growth factor induced genes without the complication of high background caused by high concentrations of FBS. At the initiation of experiments, the media were removed and fresh Basal Medium added with or without (Basal) addition of TNFα (10 ng/mL) or VEGF (10 ng/mL). The cells were incubated for 4 hr. At the termination of experiments, media were removed

Table 1 Summary of Endothelial Cell Preparations Used in the Affymetrix Array Analysis of Gene Expression Description

Passage

Biowhittaker ID

Lot Number

HUVEC (pool of five individual isolates)a

6

CC-2519

2F0132

HUVEC (pool of five individual isolates)a

6

CC-2519

2F0237

HUVEC (pool of five individual isolates)a

6

CC-2519

2F0332

HDMVEC (single isolate)

5

CC-2543

2F0047

HDMVEC (single isolate)

5

CC-2543

1F0928

HLMVEC (single isolate)

6

CC-2527

1F1554

HLMVEC (single isolate)

6

CC-2527

1F1643

HAEC (single isolate)

5

CC-2535

F1335

HAEC (single isolate)

5

CC-2535

F1340

HCAEC (single isolate)

5

CC-2585

1F2058

HCAEC (single isolate)

5

CC-2585

2F0239

a None of the individual isolates in this sample are present in either of the other two HUVEC lots, thus providing sampling of 15 individual isolates across each HUVEC treatment.

and 15 mL of Tri reagent (Molecular Research Center Inc.; Cincinnati, OH) were added directly to the endothelium in each T75 flask; lysates were removed and stored frozen at −80°C. RNA was subsequently extracted following manufacturer’s instructions. Total RNA isolated this way was further purified using RNeasy Mini kits as described in the product manual (Qiagen Inc., Valencia, CA). DNAse treated total RNA, 5 µg, was

Endothelial cells in health and disease

140

converted to cRNA and fragmented cRNA hybridized to arrays (U133A) per manufacturer’s suggested protocol (Affymterix, Santa Clara, CA). Data were analyzed with the MASv5 (Affymetrix) and Rosetta Resolver (Rosetta Biosoftware, Bothwell, WA). Array results that met with manufacturer’s (Affymetrix) recommended quality criteria were imported into Rosetta Resolver (28) which uses an additive and multiplicative noise based error model to increase confidence in gene expression measurement and a statistical method for combining replicates (29). Figure 1 shows a ratio-analyzed agglomerative cluster of the transcriptosomes of the five endothelial cell types under basal, TNFα and VEGF stimulated conditions. We used a fold change cutoff of 1.5-fold, and a p-value of 0.05 and included sequences where at least one combined intensity experiment met the aforementioned thresholds. It is clear from this figure that there was considerable variability in the basal expression of the different mRNAs (Fig. 1A), but when the cells were stimulated with either TNFα (Fig. 1B), many of the same genes were up- and downregu-lated in all five types of cultured endothelia. The more highly upregulated, common genes include ephrin A1, interleukin 8, the adhesion molecules ICAM-1, VCAM-1, and E-selectin, the membrane bound chemokine fractalkine, and the anti-apoptosis gene, B Cell CLL/Lymphoma 2 (Bcl-2). Two of the common downregulated genes were CD97 and BDNF. VEGF stimulated responses were somewhat more variable, but there are clear clusters of commonly upand downregulated genes (Fig. 1C). Examples of common upregulated genes include follistatin, bone morphogenic protein 2, E-selectin, CD55, Dual specificity phosphatase 6 (DUSP6), IL-6, IL-8, the VEGF receptor fms like tyrosine kinase 1 (flt-1), stanniocalcin 1, and Ang-2. Common downregulated genes included death associated protein kinase 3, placental growth factor, Bcl-3 and TNF receptor associated factor (TRAF) 5. For both TNFα and VEGF, many, if not all of these aforementioned upregulated genes, have been

The use of genomics to map phenotypic heterogeneity

141

Figure 1 Agglomerative clustering of (A) basal-, (B) TNF-, (C) and VEGFinduced gene expression in different endothelial cells. HUVE human umbilical vein endothelial cells,

Endothelial cells in health and disease

142

HAEC, human aortic endothelial cells, HLMVEC human lung microvessel endothelial cells, HDMVEC human dermal microvessel endothelial cells, HCAEC human coronary artery endothelial cells, HDMVEC human dermal microvascular endothelial cells. Red: upregulated; green: downregulated; black: not detected or did not meet statistical significance. previously documented to be upregulated at both the protein and RNA level (27,30–48). To identify possible cell-type specific sequence, we applied the following criteria. A given sequence had to increase or decrease more than twofold (log ratio 0.3) at a p-value of 1 year), even well after the arteriogenic stimulus had dissipated. Other growth factors may also play a role in the reparative response following ischemic tissue injury. Emanueli et al. (25) reported that the neurotrophin nerve growth factor (NGF) participates in the functional neovascularization of tissue following injury. Supplementation with this growth factor appears to promote angiogenesis through a VEGF-Akt-NO-mediated signaling pathway. As these studies focused on graft survival, function, and rejection following ischemic preservation/transplantation, it is conceivable that these strategies might be able to improve current techniques for organ transplantation. For example, if a harvested organ can be maintained for a prolonged period in spite of hypoxic and ischemic conditions, current limitations in time and distance in transplantation will diminish in importance and potentially expand the donor pool by increasing the likely use of “marginal” donor organs. Another theoretical advantage of improving organ preservation revolves around the putative link between organ rejection and ischemia-reperfusion injury. Suppressing the latter may lead to reduced acute (and hence, chronic) rejection, which currently

Oxygen regulation of endothelial cell phenotypes

193

represents the main limitation to the survival of patients who receive a transplanted organ. Overall, it is apparent that elucidating the changes in the vasculature during hypoxia may reveal potential for improving survival in patients who suffer from many types of ischemic organ injury.

3. MODULATION OF ENDOTHELIAL BARRIER FUNCTION BY HYPOXIA Endothelial cells form a metabolically active and barrier separating the contents of the vascular lumen from surrounding tissue and organs. Under normal physiological conditions, the endothelium comprises a selectively permeable monolayer, allowing macromolecules of various sizes to pass through, depending on their Stokes’ atomic radii. This process is referred to as restricted diffusion, and means that solutes of smaller diameter can easily migrate through the endothelial barrier, whereas larger molecules, such as proteins and cellular elements of circulating blood are confined within the vascular lumen. After a protracted period of severe hypoxia, however, the phenotype of this selectively permeable barrier changes quite remarkably, as it becomes characterized by impairment of endothelial barrier function manifest as increased vascular permeability. The physiological basis for this dramatic change in endothelial phenotype in low oxygen environment has been largely elucidated. Endothelial cells, exposed to oxygen-deficient environment in vitro, develop changes in their actin-based cytoskeleton with subsequent formation of 1–3 µm intercellular gaps between adjacent cells (26). This is due to internal reorganization of the internal cytoskeleton, which is connected at adherens junctions with the endothelial plasma lemma, resulting in a retraction of the edges of apposing endothelial cells (26). With the appearance of intercellular gaps, the endothelial layer loses its ability to restrict passage of molecules based on size, and unrestricted diffusion or even mass movement of fluids and solutes can transpire. At a tissue level, these changes are manifest as increased vascular permeability with subsequent perivascular interstitial edema due to massive paracellular leakage of solutes, proteins, and cellular components (27). This condition is clinically recognized as the capillary leak syndrome, which has been commonly accepted as pathophysiological evidence of impaired vascular barrier function in response to different stimuli (28), including hypoxia. It can be observed in various clinical situations and pathological conditions, such as high altitude exposure or ischemic brain stroke, which can be accompanied by acute pulmonary and cerebral edema, respectively (29,30). As a model to study the effect of prolonged hypoxic exposure on endothelial barrier function, a common experimental method relies on growing endothelial cells to confluence on special filters which restrict the passage of macromolecules and lower molecular weight solutes (31). This experimental model has been successfully used to investigate the role of different receptor agonists, such as thrombin, histamine, or TNF-α on vascular barrier permeability, and reflects pathophysiological processes occurring in vivo, although cultured endothelial monolayers in this model are somewhat more permeable than vessels from which they are originally derived (32,33). The increased passage of tracer solutes, such as [3H]-insulin, across cultured pulmonary artery or aortic monolayers has been noticed after prolonged (24 hr) exposure to hypoxia (12–14 mm

Endothelial cells in health and disease

194

Hg). This hypoxia-induced increase of vascular barrier permeability occurs in a timedependent manner and was dependent on absolute level of hypoxia as well (26), but there is a restoration of normal endothelial function within 48 hr upon reoxygenation that points to the reversible character of the barrier function perturbation driven by hypoxic stress (26). What is the key regulatory link affected during hypoxia/ischemia which causes such remarkable perturbations in endothelial barrier function? Multiple studies show that the cAMP second messenger system is an important determinant of many critical vascular functions, including permeability, coagulation balance, and leukocyte/blood cell interaction (26,34,35). A decreased level of intracellular cAMP is one of the principal mechanisms driving vascular dysfunction under conditions of decreased oxygen tension. In endothelial cells subjected to hypoxia, the activity of adenylate cyclase, both basal and stimulated, is diminished, which contributes to the decline in intracellular cAMP (26,36). A similar decline in cAMP synthesis and increased permeability have been observed following exposure of endothelial monolayers to TNF-α (36), an inflammatory cytokine whose effects in large part mimic those seen with hypoxic exposure. In vascular smooth muscle cells subjected to hypoxia, the level of intracellular cAMP also drops precipitously, but in this case it is mainly due to specific increases of types III and IV phosphodiesterase activity (37). Under subsequent reoxygenation following the hypoxic period, the extensive formation of reactive oxygen species (ROS) leads to activation of phosphodiesterase types II, III, and IV in endothelium, thus further contributing to the rapid decline in intracellular cyclic nucleotide level (38,39). Administration of membrane permeable cAMP analogs such as dibutyryl (db)-cAMP has been shown to reverse the hypoxia-mediated enhancement of vascular permeability. These same effects were elicited by treatment of hypoxic endothelial cell cultures with stimulators of type I (8-[4chlorophenylthio] adenosine 3′,5′-phosphate) and type II (N6-benzoyladenosine 3′,5′cyclic monophosphate) protein kinase A (26), but not adenosine or adenosine monophosphate. Taken together, these studies show that a decline in intracellular cAMP content, driven by hypoxia and/or redox stress and brought about either by a reduction in cAMP synthesis or an increment in cAMP catabolism, can lead to reversible changes in cytoskeletal architecture and permeability characteristics. Another cyclic nucleotide second messenger pathway which is also quite important in maintaining homeostatic physiological and biochemical processes, including barrier properties of endothelium, and which is disrupted by hypoxic exposure, is the nitric oxide (NO)/cGMP system (40). Levels of NO are markedly decreased, especially during reoxygenation, because of the scavenging effects of ROS on NO bioavailability (41). Administration of the NO donor S-nitroso-N-acetylpenicillamine or superoxide dismutase (which increases bioavailable levels of NO) can effectively attenuate hypoxia-induced increases in vascular permeability (42). Taken together, these studies show not only the common signaling mechanisms involved in hypoxic regulation of endothelial permeability characteristics, but show that pharmacological modulation of these pathways can normalize barrier function in ischemic tissue.

Oxygen regulation of endothelial cell phenotypes

195

4. INFLAMMATION Of the many homeostatic properties that are modulated by hypoxia, one of the most important is inflammation. Hypoxia clearly evokes an inflammatory response, marked by initial endothelial cell activation, which triggers local inflammation in the context of recruitable effector leukocyte populations in the blood and the organismal level. Inflammatory reactions, initially triggered by hypoxic modulation of the endothelial phenotype, lead to self-propagating and auto-amplifying regulatory loops in which inflammation itself perpetuates various local changes of communication between endothelial cells and leukocytes.

4.1. Proinflammatory Cytokines Proinflammatory cytokines play an important role in mediating hypoxic or ischemicrelated tissue injury. During hypoxia, proinflammatory cytokines are induced and secreted by endothelial cells as well as leukocytes and subsequently influence both local vascular function as well as function and gene expression of tissues/organs in remote locations, to which these cytokines are carried by flowing blood. IL-1 and TNF-α are critical mediators of the inflammatory response during hypoxia stress. These cytokines are synthesized de novo within hypoxic endothelial cells and secreted extracellularly during hypoxic stress (or reoxygenation/reperfusion). Hypoxia increases IL-1 mRNA in cultured endothelial cells, which may then act in an autocrine manner to induce E-selectin and ICAM-1 expression (2). Another key mediator of inflammatory stress is IL-8, which is a member of the CXC family of chemokines originally identified as a neutrophil activating peptide and now recognized to be a potent neutrophil chemotactic factor. IL-8 is induced and secreted extracellularly from endothelial cells during hypoxia stress (43,44). IL-8 production is regulated at a transcriptional level and hypoxia-mediated induction of the IL-8 gene involves increased binding of nuclear factor κB (NF-κB) binding to the upstream promoter region. IL-8 also promulgates the general inflammatory milieu because it induces transcription of monocyte chemotactic protein (MCP-1) in endothelial cells during hypoxia. MCP-1, in turn, serves to recruit monocytes during hypoxic stress. IL-8 also increases expression of inducible protein-10 (IP-10), a CXC chemokine family member which is chemotactic for T cells (43,44). In human coronary sinus samples, an increase in IL-8 was detected following the period of aortic cross-clamping, with a direct correlation between IL-8 levels, cardiac ischemic duration, and release of myoglobin, a marker of myocyte injury (43). There is conflicting information as to the physiological or pathological role for another cytokine, IL-6, which is synthesized and released extracellularly from endothelial cells exposed to hypoxic stress (45). There are data to suggest that IL-6 work may actually act as an anti-inflammatory cytokine, to quell inflammation during the hypoxic reaction to stress. IL-6 has been shown to suppress production of IL-1 and TNF-α (46), and to induce expression of IL-1 receptor antagonist and TNF-α receptor antagonist (47).

Endothelial cells in health and disease

196

4.2. Adhesion Receptors Adhesion molecules are activated and expressed during hypoxia on both endothelial cells and leukocytes, increasing their cognate interactions and promoting inflammation. Leukoadhesion to endothelial cells is a multistep process that involves the following sequence: (1) initial rolling deceleration of the leukocyte along the activated endothelium; (2) more highly adhesive interactions with other adhesion receptors; (3) activation of endothelial cell and tight attachment, and (4) emigration. It is useful to consider the various stages of leukocyte adhesion, because each may be modulated in different ways by the hypoxic environment. The initial tethering and rolling of leukocytes on endothelial cells are mediated by carbohydrate-containing glycoprotein adhesion receptors called selectins. P-selectin, which exists embedded in the membranes of Weibel-Palade bodies, is rapidly translocated to the plasmalemma during the process of Weibel-Palade exocytosis due to increased calcium influx during hypoxic stress; (48). Using cycloheximide to block protein synthesis, these experiments showed that very little (if any) active protein synthesis was required for hypoxia to trigger surface expression of P-selectin. The other hypoxia-inducible selectin, E-selectin, requires transcription and translation of message de novo following hypoxic stress, a process that is dependent on NF-κB (49). The next phase of leukocyte adhesion to activated endothelium is mediated by members of the immunoglobulin family of adhesion receptors, of which ICAM-1 is a primary member. ICAM-1 expression is regulated through various signal transduction pathways that are coupled to activation of a transcription factor such as Egr-1 and NF-κB, and hence is both hypoxia- and redox-sensitive. Once translated and expressed on the endothelial surface, ICAM-1 tightly engages β2-integrin on the leukocyte surface. ICAM1 is also upregulated secondarily by inflammatory cytokines such as IL-1 and TNF-α, which are also induced during hypoxia stress. The ICAM-1 adhesion molecule plays a critical role in ischemia-induced inflammation during and after heart transplantation and lung transplantation, the reduced expression or blockade of which may serve as a useful target for therapeutic intervention to mitigate ischemia-induced tissue injury (3,50).

4.3. Activation of Endothelial Cells and Leukocyte Emigration Adhesion of leukocytes to the endothelial cell surface activates the endothelial cells themselves, potentially through conformational changes in ICAM-1 during receptor engagement which are transduced into the cell to affect signaling. Some of these signals may be mediated by an increase of intracellular calcium, phosphorylation of myosin light chain kinase, and F-actin structural changes (51). Emigration of leukocytes from the vascular lumen is accompanied by endothelial cell activation, consisting of discrete structural changes of the endothelial cell-cell junctions. These structural alterations facilitate leukocyte emigration. ICAM-1 and other integral membrane proteins such as PECAM-1 are important signal transduction intermediaries which elicit cytoskeletal and junctional changes in endothelial mono-layers which promote leukocyte diapedesis (51,52). ICAM-1, which induces calcium signaling through PKCs, mediates phosphorylation of actin-associated protein and cytoskeletal rearrangement in brain

Oxygen regulation of endothelial cell phenotypes

197

endothelial cells (53). E-selectin also associates with elements of the actin cytoskeleton leading to cytoskeletal reorganization. Concomitantly cytoplasmic domain of E-selectin leads MAPK activation and subsequent changes in gene transcription (54). By changing endothelial cytoskeletal properties and increasing expression of various membrane proteins, it can easily be seen how hypoxia or ischemia can facilitate the pathways leading to leukocyte emigration.

5. HYPOXIA-DRIVEN REGULATION OF GENE TRANSCRIPTION In order to survive in the absence of sufficient oxygen, cells have developed specific physiologic adaptations consisting of discrete metabolic alterations. In many cases, these adaptations are regulated at the level of transcription by transcription factors binding to the promoter regions of target genes. Several transcription factors, such as Egr-1 and HIF-1 are activated during hypoxic stress; other genes, such as NF-κB and activation protein-1 (AP-1), are activated in the presence of redox stress, such as occurs during reoxygenation accompanied by the formation of reactive oxygen intermediates. When these transcription factors are induced and translocate to the nucleus, they promote expression of many downstream target genes, some of which can propagate or exacerbate local tissue injury.

5.1. Egr-1 Egr-1 belongs to a family of related transcription factors (zinc finger proteins), and is induced rapidly following hypoxic stress. Activation of protein kinase C II-β (PKC-IIβ) represents a proximal trigger for the induction of Egr-1. Induction of Egr-1 increases transcription of downstream target genes, including proinflammatory cytokines such as IL-1, chemokines such as MIP-2 and MCP-1, adhesion receptors such as ICAM-1, and procoagulant molecules such as PAI-1 and TF. Egr-1 can, therefore, be viewed as a critical pathological trigger for hypoxia- or ischemic-related inflammation (8,55). (Fig. 1).

5.2. HIF-1 As an essential survival mechanism, mammalian cells have evolved the ability to sense changes in O2 concentration. The transcription factor HIF-1 represents one of the most important of several O2 sensing mechanisms. HIF-1 plays a pivotal role in cell adaptation to hypoxic stress, as it facilitates metabolic alterations during hypoxia. HIF-1 influences downstream target genes such as erythropoietin, VEGF, insulin-dependent glucose transporter (GLUT-1), Hmox-1, and the inducible NO synthase (NOS), iNOS (56,57). Quite recently, the mechanism by which mammalian cells “sense” O2 by upregulating HIF-1 activity has been elucidated. Under normoxic conditions, HIF-1 is continually degraded through inactivation of the HIF-1α subunit by prolyl-hydroxylase, an enzyme that catalyzes hydroxylation of proline residues using oxygen as a substrate. In contrast,

Endothelial cells in health and disease

198

under anaerobic conditions, HIF-1α remains undegraded because the prolyl-hydroxylase is inactive, and hence HIF-1 heterodimer can migrate to and accumulate within the nucleus where it activates promoters to elicit transcription of downstream target genes (58). Asparaginyl-hydroxylase also regulates activity of this important transcription factor. Under hypoxic conditions, hydroxylation of asparagine residues of the COOHterminal of HIF is abrogated, which ultimately results in activation of downstream target genes (59). Both prolylhydroxylase and asparaginyl-hydroxylase hence serve as direct sensors of intracellular oxygen tension.

Figure 1 Hypoxia increases Egr-1 mRNA expression in lungs of mice. H: hypoxia; N: normoxia. Panel A: Northern blot; panel B: Western blot. (From Ref. 55.)

6. COMPLEMENT The complement cascade plays an important role in hypoxia-reoxygenation injury, and this system too can be modulated by a hypoxic-reoxygenated environment. Hypoxia leads to complement activation and deposition of C3 on the endothelial cell surface, with reoxygenation further augmenting C3 deposition. This activation is mediated by both the classical and lectin pathways of complement activation (60–62). Moreover, there is further amplification of this cascade in that deposition of C5b-9 on endothelial cells elicits expression of other endothelial cell-leukocyte adhesion molecules such as ICAM-1 and VCAM-1, predominantly via an NO/cGMP-regulated NF-κB translocation mechanism (63). Brain ischemia induces accumulation of C1q on neurons, and this in turn influences endothelial-leukocyte-platelet interactions. Accumulated C1q activates GPIIb–IIIa fibrinogen binding sites and the expression of P-selectin, which is likely to contribute to thrombotic events and in turn influences EC-PLT axis (64). A bifunctional

Oxygen regulation of endothelial cell phenotypes

199

complement inhibitory pro tein such as sLex-glycosylated sCR1 has been shown to inhibit both leukocyte and platelet accumulation in stroke (65). These data suggest that ischemia-reperfusion triggers a number of inflammatory cascades at the endothelial surface, which are linked at multiple levels.

7. VASODILATATION AND VASOCONSTRICTION Hypoxia/ischemia strongly modulates vascular tonus. For example, under hypoxic conditions, vessels exhibit a rapid but transient vasoconstriction followed by relaxation during the first few minutes. If the hypoxic exposure is sustained, however, the vasoconstriction too becomes more sustained. De-endothelialization of blood vessels in an experimental apparatus abolishes hypoxia-induced vasoconstriction, while at the same time enhancing vasodilatation. Under basal conditions, there is a dynamic balance between vasoconstrictive and vasodilatatory substances produced by endothelial cells, which modulate each other through multiple feedback mechanisms (66,67). Some factors are very important for control of vasodilatation [NO and prostacyclin (PGI2)], whereas others modulate vasoconstriction (endothelin) (68).

7.1. Endogenous Vasodilators NO is formed from L-arginine by one of three NOS isoenzymes (69). Activity of the predominant endothelial isoform, NOS III (also called eNOS), depends on ambient levels of intracellular concentration of Ca2+ (70). Shear stress, cellular proliferation, and various receptor agonists (such as bradykinin, serotonin, adenosine, ADP/ATP, histamine, or thrombin) increase eNOS enzymatic activity by increasing intracellular calcium (71). Experimentally, the calcium ionophore A23187 similarly increases intracellular calcium and hence its application results in a pulse of NO liberation. During hypoxia there is a decrease in basal e-NOS gene expression, leading to a substantial decrease in basal endothelial NO production (72). As NO activates vascular soluble guanylate cyclase (sGC) to form intracellular cGMP in vascular smooth muscle cells, leading to their relaxation, diminished levels of NO are expected to promote vasoconstriction. In the pulmonary vascular bed, this decrease may contribute in part to the sustained vasoconstrictor response to hypoxia. During reperfusion, NO (and hence, cGMP) levels plummet due to the quenching of NO by superoxide (41), and hence, reperfusion is typically characterized by vasoconstriction. Other vasodilators also contribute to the quintessentially vasodilated state of normal vessels. These include PGI2, formed in endothelial cells by the action of cyclooxygenase (COX) and arachidonic acid, a process that is similar to NO formation in that its release is modulated by physical and humoral stimuli (73). Shear stress and many agonists, which are also agonists leading to production of NO, contribute to the release of PGI2 (74). Agonist-induced release of PGI2 seems to be predominantly regulated by Ca2+ release from intracellular storage pools (75). PGI2 causes relaxation of vascular smooth muscle by activating adenylate cyclase resulting in an increased production of cAMP

Endothelial cells in health and disease

200

(70). Therefore, even if not directly regulated by hypoxia, the recognized modulation of cAMP synthesis and catabolism by hypoxia makes it likely that the downstream PGI2 vasodilatory signal would be dampened under hypoxic or ischemic conditions owing to effects of hypoxia on VSMC cAMP. A third endogenous vasodilator, endothelial-derived hyperpolarizing factor (EDHF), is also an important physiological regulator of vascular tone. Electrophy-siological studies in various arterial preparations demonstrate that acetylcholine and other receptor agonists elicit an endothelial-dependent hyperpolarization (and subsequent vascular smooth muscle cell relaxation), which is due to a discrete, diffusible EDHF that is distinct from NO and PGI2. EDHF-induced hyperpolarization of smooth muscle is inhibited by blocking the Na/K ATPase, as well as by blocking the inwardly rectifying K+ channel currents (76). These data indicate that EDHF works by activating a K+ channel in vascular smooth muscle (77). There is little information available as to how or whether hypoxia or ischemia modulates EDHF activity. Another endogenous vasodilator, whose physiological importance is becoming increasingly clear, is CO, a byproduct of the action of the enzyme Hmox. Heme is cleaved by Hmox to yield equimolar amounts of ferric iron (Fe3+), CO, and biliverdin. Biliverdin is subsequently converted to bilirubin by biliverdin reductase (78). The Hmox system consists of three enzymatic isoforms. Hmox-1 is the stress-induced isoform; Hmox-2 is the constitutive isoform, which is the dominant isoform under physiological conditions (79); and Hmox-3, also a constitutive isoform closely related to Hmox-2 (80). Hmox-1 is an inducible isoform whose expression skyrockets in many cell types, including endothelial cells, in response to oxygen deprivation or oxidant stress (67). In a manner similar to NO, CO activates sGC, causing an increase in cGMP content within neighboring vascular smooth muscle cells (81). Induction of Hmox-1 protects tissues against inflammatory stress and leukocyte adhesion (79,82). CO also functions as an inhibitor of platelet aggregation by activating platelet cGMP levels (83), and a mitogenic inhibitor. CO decreases the expression of mitogens such as ET-1 and PDGF-B in response to hypoxia. The Hmox-1/CO pathway may serve as a back-up mechanism for vascular homeostasis under nonphysiological conditions, such as hypoxia or ischemiareperfusion (84).

7.2. Endogenous Vasoconstrictors Endothelial-derived contracting factors (EDCFs) fall into one of three major categories, including vasoconstrictor peptides, eicosanoid metabolites of arachidonic acid, and free radicals. Various agonists such as norepinephrine, acetylcholine, and serotonin acting on endothelial surface receptors stimulate the production of vasoconstrictor metabolites such as endothelin, thromboxane A2, leukotrienes, and superoxide anions (71). One of the most potent EDCFs is endothelin (ET), which consists of peptides comprised of 21 amino acids (85). There are three ET isoforms; ET-1 is synthesized predominantly in endothelial cells, where it can act in a paracrine manner on ET-A receptors located on the surface of vascular smooth muscle cells, resulting in vasoconstriction. ET-1 also activates endothelial ET-B receptors that subsequently elicit release of vasodilators such as NO and PGI2 (77). Shear stress, hypoxia, and ischemia profoundly stimulate endothelial production of endothelins (84). In addition to the depressed expression of vasodilators,

Oxygen regulation of endothelial cell phenotypes

201

the production of vasoconstrictors by hypoxia quite strongly tips the vascular phenotype towards a vasoconstricted state.

8. COAGULATION/FIBRINOLYSIS Because of its strategic location juxtaposed between flowing blood and surrounding tissues, the endothelium bears an essential responsibility for maintaining blood flow by preventing thrombus formation and promoting degradation/fibrinolysis of preformed thrombus. Pathologic states accompanied by decreased oxygen tension can distinctly modulate this function of endothelium to shift the natural anticoagulant/ procoagulant balance to favor activation of coagulation. This activation of coagulation is encountered in diverse clinical situations and syndromes, including cerebral, cardiac, pulmonary, and vascular ischemic disorders. Another clinically relevant problem is the preservation of donor organs for transplantation, which is likewise associated with severe hypoxemia within the preserved vascular bed (see Chapter 25). Postischemic restoration of flow (reperfusion) results in prominent fibrin deposition that induces vascular dysfunction. It is essential to understand the role played by oxygen deprivation in triggering the procoagulant endothelial phenotype, and to understand the exact mechanisms underlying pathogenesis of thrombus accrual in situations in which the vasculature is subjected to severe hypoxic stress. One of the mechanisms by which the anticoagulant phenotype of vascular endothelium is achieved involves a cell surface thrombin-binding protein, thrombo-modulin, an integral endothelial protein which greatly enhances the ability of thrombin to convert inactive protein C to activated protein C. Activated protein C then functions as an anticoagulant by proteolytic degradation of two critical coagulation cofactors, Va and VIIIa (86). When cultured bovine endothelial cells are subjected to hypoxia, consisting of a pO2 of approximately 12–14 mm Hg (similar to levels seen in ischemic tissue), there is a marked decrease in expression of the thrombomodulin gene (31,87). This downregulation of both mRNA and protein levels leads to a parallel suppression of thrombomodulin functional activity by approximately 80–90% over a 48–72 hr hypoxic exposure. A potential signaling mechanism which contributes to the downregulation of thrombomodulin levels is the hypoxia-mediated suppression of cAMP second messenger levels. Repletion of intracellular cAMP by the addition of membrane-permeable cAMP analogs (such as dibutyryl-cAMP and 8-bromo-cAMP) abrogates the hypoxia- or TNF-αinduced decline in endothelial cell thrombomodulin expression (36,38). Another mechanism likely to be important in hypoxia-mediated alteration of the anticoagulant phenotype of vascular endothelium is the observed upregulation of TF expression, which serves as one of the strongest prothrombotic stimuli and as such which is typically excluded from the intravascular space. Expression of TF, which interacts with factor VIIa to activate the extrinsic pathway of coagulation, increases markedly in vascular wall cells following hypoxic exposure. In in vivo studies of mice exposed to normobaric hypoxia (6–8% oxygen), there was a profound increase in TF expression along with fibrin accrual in the hypoxic vs. normoxic control lung tissue, shown by histological evidence of fibrin deposited in pulmonary microvessels (89). In this model, it was apparent that mononuclear phagocytes recruited to hypoxic vascular foci, rather than

Endothelial cells in health and disease

202

endothelial cells themselves, were largely responsible for fibrin accumulation. This observation is concordant with other studies which have shown that hypoxia per se does not induce the significant endothelial expression of TF in vitro (31). In other pathologic conditions leading to abrupt intravascular thrombosis, such as atherosclerosis or massive vascular injury, the immediate exposure of subendothelial TF to blood contents leads to rapid coagulation with massive thrombus formation. Together with hypoxia-induced lung injury, these conditions share a common pathological feature, namely the recruitment of mononuclear phagocytes as a rich synthetic pool of TF. Although theoretically, polymorphonuclear leukocytes could also contribute to the pathological accumulation of fibrin under ischemic conditions, there are contrary data in that immunodepletion of polymorphonuclear leukocytes had no effect on fibrin accumulation (89). In contrast, the analogous immunologic depletion of monocytes did have a considerable reducing effect on fibrin deposition in lungs subjected to hypoxia (89). The analysis of TF expression in hypoxic murine lungs revealed a ~20-fold increase of TF transcripts in hypoxic lungs compared to the normoxic controls (90). In parallel immunohistochemical studies showed colocalization of increased TF in hypoxic lung with mononuclear phagocytes (90). These data make it clear that monocytes serve as a primary source of TF expression in hypoxia-driven thrombosis, after becoming arrested and activated at sites of hypoxic vascular injury. One of the signals by which monocyte/macrophages are drawn to ischemic vasculature is the MCP-1 whose expression is also upregulated by hypoxic endothelium (44). More detailed exploration of molecular mechanisms underlying hypoxic induction of TF expression in monocytes led to the identification of a key role for Egr-1 transcription factor, which increases rapidly following mechanical vascular injury (91,92) and which appears to be the primary force driving TF expression in hypoxia. In vitro studies show hypoxic mononuclear phagocytes to exhibit substantial upregulation of Egr-1 expression associated with induction of TF (55). Further evidence that Egr-1 expression is the main trigger for hypoxia-driven TF expression is based on experimental data obtained from mice lacking the Egr-1 gene. These Egr-1−/− mice showed significantly decreased intravascular fibrin deposition after ischemia-reperfusion lung injury and a minimal increment in TF mRNA, with absence of actual changes in TF antigen activity (90). In contrast, wild-type mice subjected to an otherwise equivalent normoxic environment expressed TF and exhibited significant fibrin deposition. The discussion of vascular perturbations during the hypoxia would be incomplete without emphasizing the role of the fibrinolytic system in the net accrual of fibrin following hypoxic exposure. PAI-1 is a 52 kDa serine protease inhibitor which serves as the primary inhibitor of the fibrinolytic cascade, thus promoting the net accrual of fibrin by retarding its dissolution. Under hypoxic conditions, plasma levels of PAI-1 antigen increase as early as 4 hr after the start of hypoxic stress in hypoxic mononuclear phagocytes (93) (Fig. 2). In contrast, mRNA levels for both tissue-type plasminogen activator (t-PA) and urokinase-type plasminogen activator (u-PA) decrease under hypoxic conditions (93). One can envision how these two

Oxygen regulation of endothelial cell phenotypes

203

Figure 2 Hypoxia (H) increases expression of PAI-1 mRNA (A) and antigen (B) compared with normoxic (N) conditions. (From Ref. 93.) mechanisms—decreased expression of plasminogen activators and increased expression of a potent suppressor of plasminogen activation—will synergize to potently suppress fibrinolysis. Experiments using mice deficient in PAI-1, t-PA, or u-PA provide strong evidence for the relevance of hypoxia-induced suppression of fibrinolysis in mediating fibrin accrual. Mice lacking PAI-1 showed significantly decreased intravascular fibrin deposition in hypoxia-exposed lungs compared to wild type mice. In contrast, t-PA−/− and u-PA−/− mice demonstrated a strong potentiation of fibrin formation on exposure to hypoxia (Fig. 3). Immunohistochemical analysis of cells in hypoxic lungs once again identified mononuclear phagocytes as the prime source of increased PAI-1 expression under conditions of low oxygen tension. As PAI-1 expression is increased by Egr-1 transcription factor (8), these data support the conceptual framework indicating that Egr-1 acts as a master switch regulating a range of effector mechanisms underlying hypoxic

Endothelial cells in health and disease

204

stress. Together these data lend support to the notion that the net fibrinolytic activity during the hypoxia is decreased, and that PAI-1 overexpression is an important factor in orchestrating this process. There are also other mechanisms that can contribute to the prothrombotic diathesis of hypoxia. As previously described, hypoxia sets in motion a cascade of events leading to progressive decrease in cAMP content and increase in intracellular Ca2+ concentration. This triggers the release of von Willebrand factor (vWF) from preformed storage pools in endothelial Weibel-Palade bodies, via Ca2+-dependent exocytosis (48). vWF has an important role in coagulation because it fosters platelet adhesion and aggregation, especially under conditions of high shear stress; thus, its

Figure 3 Relative 125I fibrinogen deposition (A–C), a surrogate marker for fibrin formation, is increased in mice lacking key components of the fibrinolytic system (uPA or tPA deficient). Mice lacking the fibrinolytic inhibitor (PAI-1) exhibit diminished fibrin accrual. (D) Similar results are seen on an immunoblot, which detects a neoepitope revealed in cross-linked fibrin. (From Ref. 93.)

Oxygen regulation of endothelial cell phenotypes

205

release under conditions of hypoxia can be viewed as an additional mechanism promoting the procoagulant phenotype of the endothelium. Interestingly, administration of a NO donor (3-morpholinosydonimine, SIN-1), or the cGMP-analog 8-bromo-cGMP decreased the hypoxia-induced release of P-selectin and vWF (94) but inhibition of endotheliumderived NO has the opposite effect on Weibel-Palade bodies exocytosis (95). NO also has other very important anticoagulant functions which are perturbed in states of ischemia or reperfusion, in which its synthesis is suppressed or it rapidly dissipates in a superoxide-rich environment. NO maintains the anticoagulant endothelial phenotype by inhibiting platelet aggregation and adhesion (96–98), and by decreasing retraction of lateral margins of endothelial cells thereby preventing exposure of subendothelial TF and collagen to direct contact with blood contents. The procoagulant aspects of NO dissipation are particularly prominent during the postischemic period in which restoration of blood flow (reoxygenation) induces ROS production in neutrophils that have accumulated during the hypoxic period. These ROS are highly reactive and rapidly react with NO quenching it and decreasing the bioavailability of NO (99). It has been shown that supplementation with the NO precursor L-arginine improves vascular function in the setting of lung reperfusion (100), and cardiac preservation is likewise enhanced in a cardiac transplantation model by administration of the NO donor nitroglycerin (101). There is one additional mechanism whose perturbation may contribute to a generalized hypoxic prothrombotic diathesis. Endothelial CD39 (nucleoside triphosphate diphosphohydrolase 1, NTPDase-1) is a recently recognized transmembrane protein, whose extracellular portion exhibits ectoapyrase activity. It hydrolyzes ADP released from activated platelets, thus strongly inhibiting ADP-induced platelet aggregation (102). It has been shown that administration of soluble CD39 has a major influence on maintaining blood fluidity and reduces microvessels thrombosis in different models such as brain ischemia (18). The regulation of CD39 expression under hypoxic conditions is currently under study by several groups.

9. CONCLUSION Endothelial cells, as well as other cells of the vascular wall, including mononuclear phagocytes and vascular smooth muscle cells, are prime movers in the delicate balance that preserves nutritive flow while maintaining immune, barrier, and hemostatic functions. Powerful and redundant mechanisms have evolved to protect this critical homeostatic balance. For instance, several primordial and exquisitely sensitive subcellular oxygen-sensing mechanisms have evolved as adaptive or protective mechanisms. In the absence of oxygen or in the presence of ROS generated during reperfusion, these mechanisms are triggered to unleash a torrent of cascades designed to restore homeostasis. Depending on the exuberance of the response and the specific backdrop, these cascades are often, however, frankly pathological. Hypoxia thereby causes the endothelial phenotype to become procoagulant and proinflammatory, the barrier to lose its selectivity, and various mediators to be released, leading to vasoconstriction. Although in the setting of an infectious process or a traumatic wound, these mechanisms may be adaptive and lead to healing, in other settings these changes in endothelial phenotype can lead to vascular compromise and organismal demise. Understanding the molecular basis

Endothelial cells in health and disease

206

for the hypoxic modulation of the endothelial phenotype may ultimately lead to new treatments as myocardial infarction, stroke, and a litany of other ischemic disorders.

ACKNOWLEDGMENTS/DISCLOSURES This work was supported in part by grants from the National Institutes of Health; R01 NS41460, R01 HL59488, R01 HL55397, R01 HL69448, and R01 HL060900. Dr. Pinsky has served as a consultant for and received research funds from Aga-LindeHealthcare and has an equity position in St. Camillus Medical.

REFERENCES 1. Calvillo L, Latini R, Kajstura J, et al. Recombinant human erythropoietin protects the myocardium from ischemia-reperfusion injury and promotes beneficial remodeling. Proc Natl Acad Sci 2003; 100:4802–4806. 2. Shreeniwas R, Koga S, Karakurum M, et al. Hypoxia-mediated induction of endothelial cell interleukin-1α. An autocrine mechanism promoting expression of leukocyte adhesion molecules on the vessel surface. J Clin Invest 1992; 90:2333–2339. 3. Wang CY, Naka Y, Liao H, et al. Cardiac graft intercellular adhesion molecule-1 (ICAM-1) and interleukin-1 expression mediate primary isograft failure and induction of ICAM-1 in organs remote from the site of transplantation. Circ Res 1998; 82: 762–772. 4. Kocher A, Schuster M, Szaboles M, et al. Neovascularization of ischemic myocardium by human bone-marrow-derived angioblasts prevents cardiomyocyte apoptosis, reduces remodeling and improves cardiac function. Nat Med 2001; 4:430–436. 5. Torry R, Bai L, Miller S, Labarrere C, Nelson D, Torry D. Increased vascular endo thelial growth factor expression in human hearts with microvascular fibrin. J Mol Cell Cardiol 2001; 33:175–184. 6. Luttun A, Tjwa M, Moons L, et al. Revascularization of ischemic tissues by PIGF treatment, and inhibition of tumor angiogenesis, arthritis and atherosclerosis by anti-Flt1. Nat Med 2002; 8:831–840. 7. Abraham D, Taghavi S, Riml P, et al. VEFG-A and -C but not -B mediate increased vascular permeability in preserved lung grafts. Transplantation 2002; 73:1703–1706. 8. Yan SF, Fujita T, Lu J, et al. Egr-1, a master switch coordinating upregulation of divergent gene families underlying ischemic stress. Nat Med 2000; 6:1355–1361. 9. Okada M, Fujita T, Sakaguchi T, et al. Extinguishing Egr-1-dependent inflammatory and thrombotic cascades following lung transplantation. FASEB J 2001; 15(14): 2757–2759. 10. Okada M, Yan S, Pinsky D. Peroxisome proliferator-activated receptor-γ (PPAR-γ) activation suppresses ischemic induction of Egr-1 and its inflammatory gene targets. FASEB J 2002; 16:1861–1868. 11. Okada K, Fujita T, Minamoto K, Liao H, Naka Y, Pinsky DJ. Potentiation of endogenous fibrinolysis and rescue from lung ischemia/reperfusion injury in interleukin (IL)-10reconstituted IL-10 null mice. J Biol Chem 2000; 275:21468–21476. 12. Fujita T, Toda K, Karimova A, et al. Paradoxical rescue from ischemic lung injury by inhaled carbon monoxide driven by derepression of fibrinolysis. Nat Med 2001; 7: 598–604. 13. Bronner L, Kanter P, Manson J. Primary prevention of stroke. N Engl J Med 1995; 333:1392– 1400.

Oxygen regulation of endothelial cell phenotypes

207

14. Mocco J, Choudhri T, Huang J, et al. HuEP5C7 as a humanized monoclonal anti-E/P-selectin neurovascular protective strategy in a blinded placebo-controlled trial of nonhuman primate stroke. Circ Res 2002; 91:907–914. 15. Bernaudin M, Marti H, Roussel S, et al.A potential role for erythropoietin in focal permanent cerebral ischemia in mice. J Cereb Blood Flow Metab 1999; 19:643–651. 16. Sadamoto Y, Igase K, Sakanaka M, et al. Erythropoietin prevents place navigation disability and cortical infarction in rats with permanent occlusion of the middle cerebral artery. Biochem Biophys Res Commun 1998; 253:26–32. 17. Renner O, Tsimpas A, Kostin S, et al. Time- and cell type-specific induction of plateletderived growth factor receptor-beta during cerebral ischemia. Brain Res Mol Brain Res 2003; 113:44– 51. 18. Pinsky DJ, Broekman MJ, Peschon JJ, et al. Elucidation of the thromboregulatory role of CD39/ectoapyrase in the ischemic brain. J Clin Invest 2002; 109:1031–1040. 19. Mahmoudi M, Willgoss D, Cuttle L, et al. In vivo and in vitro models demonstrate a role for caveolin-1 in the pathogenesis of ischaemic acute renal failure. J Pathol 2003; 200:396–405. 20. Lario S, Mendes D, Bescos M, et al. Expression of transforming growth factor-β1 and hypoxiainducible factor-1α in an experimental model of kidney transplantation. Transplantation 2003; 75:1647–1654. 21. Chang D, Su H, Tang G, et al. Adeno-associated viral vector-mediated gene transfer of VEGF normalizes skeletal muscle oxygen tension and induces arteriogenesis in ischemic rat hindlimb. Mol Ther 2003; 7:44–51. 22. Annex BH, Davies MG, Fulton GJ, et al. Local delivery of a tissue factor antibody reduces early leukocyte infiltration but fails to limit intimal hyperplasia in experimental vein grafts. J Surg Res 1998; 80:164–170. 23. Lederman R, Mendelsohn F, Anderson R, et al. Therapeutic angiogenesis with recombinant fibroblast growth factor-2 for intermittent claudication (the TRAFFIC study): a randomised trial. Lancet 2002; 359:2053–2055. 24. Rajagopalan S, Mohler E, Lederman R, et al. Regional angiogenesis with vascular endothelial growth factor in peripheral arterial disease: a phase II randomized, double-blind, controlled study of adenoviral delivery of vascular endothelial growth factor 121 in patients with disabling intermittent claudication. Circulation 2003; 108: 1933–1938. 25. Emanueli C, Salis M, Pinna A, Graiani G, Manni L, Madeddu P. Nerve growth factor promotes angiogenesis and arteriogenesis in ischemic hindlimbs. Circulation 2002; 106:2257–2262. 26. Ogawa S, Koga S, Kuwabara K, et al. Hypoxia-induced increased permeability of endothelial monolayers occurs through lowering of cellular cAMP levels. Am J Physiol 1992; 262:C546– C554. 27. Pinsky DJ, Yan S-F, Lawson C, et al. Hypoxia and modification of the endothelium: implications for regulation of vascular homeostatic properties. Sem Cell Biol 1995; 6:283–294. 28. Tooke J, Hannemann M. Adverse endothelial function and the insulin resistance syndrome. J Intern Med 2000; 247:425–431. 29. Hultgren I. High altitude pulmonary edema. In: Staub N, ed. Lung Water and Solute Exchange. New York: Marcel Dekker Inc, 1978:437–469. 30. Olesen S. Rapid increase in blood-brain barrier permeability during severe hypoxia and metabolic inhibition. Brain Res 1986; 368:24–29. 31. Ogawa S, Gerlach H, Esposito C, Pasagian-Macaulay A, Brett J, Stern D. Hypoxia modulates the barrier and coagulant function of cultured bovine endothelium. Increased monolayer permeability and induction of procoagulant properties. J Clin Invest 1990; 85:1090–1098. 32. Del Vecchio P, Siflinger-Birnboim A, Shepard J, Bizios R, Cooper J, Malik A. Endothelial monolayer permeability to macromolecules. Fed Proc 1987; 46:2511–2515. 33. Albeda S, Sampson P, Haselton F, et al. Permeability characteristics of cultured endothelial cell monolayers. J Appl Physiol 1988; 64:308–322.

Endothelial cells in health and disease

208

34. Yan S, Lawson C, Stern D, Pinsky D. Hypoxia-mediated modulation of vascular function— implication for organ preservation and thrombogenesis: Roger S. Mitchell lecture. Chest 1998; 114:46S–50S. 35. Yan S, Ogawa S, Stern D, Pinsky D. Hypoxia-induced modulation of endothelial cell properties: regulation of barrier function and expression of interleukin-6. Kidney Int 1997; 51:419–425. 36. Koga S, Morris S, Ogawa S, et al. TNF modulate endothelial properties by decreasing cAMP. Am J Physiol 1995; 268:C1104–C1113. 37. Pinsky D, Oz M, Liao H, et al. Restoration of the cAMP second messenger pathway enhances cardiac preservation for transplantation in a heterotopic rat model. J Clin Invest 1993; 92:2994– 3002. 38. Suttorp N, Weber U, Weisch T, Schudt C. Role of phosphodiesterases in the regulation of endothelial permeability in vitro. J Clin Invest 1993; 91:1421–1428. 39. Suttorp N, Ehreiser P, Hippenstiel S, et al. Hyperpermeability of pulmonary endothelial monolayer: protective role of phosphodiesterase isoenzymes 3 and 4. Lung 1996; 174:181–194. 40. Fischer S, Clauss M, Wiesnet M, Renz D, Schaper W, Karlicyek GF. Hypoxia induces permeability in brain microvessel endothelial cells via VEGF and NO. Am J Physiol 1999; 276(4pt 1):C812–C820. 41. Pinsky DJ, Naka Y, Chowdhury NC, et al. The nitric oxide/cyclic GMP pathway in organ transplantation: cricital role in successful lung preservation. Proc Natl Acad Sci USA 1994; 91:12086–12090. 42. Utepbergenov D, Mertsch K, Sporbert A, et al. Nitric oxide protects blood–brain barrier in vitro from hypoxia/reoxygenation-mediated injury. FEBS 1998; 424:197–201. 43. Oz MC, Liao H, Naka Y, et al. Ischemia-induced interleukin-8 release after human heart transplantation. A potential role for endothelial cells. Circulation 1995; 92: II-428–II-432. 44. Karakurum M, Shreeniwas R, Chen J, et al. Hypoxic induction of interleukin-8 gene expression in human endothelial cells. J Clin Invest 1994; 93:1564–1570. 45. Yan S-F, Tritto I, Pinsky D, et al. Induction of interleukin-6 (IL-6) by hypoxia in vascular cells: central role of the binding site for nuclear factor-IL-6. J Biol Chem 1995; 270:11463–11471. 46. Schindler R, Mancilla J, Endres S, Ghorbani R, Clark S, Cinarello C. Correlations and interactions in the production of interleukin-6 (IL-6), IL-1, and tumor necrosis factor (TNF) in human blood mononuclear cells: IL-6 suppresses IL-1 and TNF. Blood 1990; 75:40–47. 47. Tilg H, Trehu E, Atkins M, Dinarello C, Mier J. Interleukin-6 (IL-6) as an anti-inflammatory cytokine: induction of circulating IL-1 receptor antagonist and soluble tumor necrosis factor receptor p55. Blood 1994; 83:113–118. 48. Pinsky DJ, Naka Y, Liao H, et al. Hypoxia-induced exocytosis of endothelial cell WeibelPalade bodies. A mechanism for rapid neutrophil recruitment after cardiac preservation. J Clin Invest 1996; 97:493–500. 49. Ichikawa H, Flores S, Kvietys P, et al. Molecular mechanisms of anoxia/reoxygenareoxygenation-induced neutrophil adherence to cultured endothelial cells. Circ Res 1997; 81:922–931. 50. Toda K, Kayano K, Karimova A, et al. Antisense intercellular adhesion molecule-1 (ICAM-1) oligodeoxyribonucleotide delivered during organ preservation inhibits post-transplant ICAM-1 expression and reduces primary lung isograft failure. Circ Res 2000; 86:166–174. 51. Wang Q, Doerschuk C. The p38 mitogen-activated protein kinase mediates cytoskeletal remodeling in pulmonary microvascular endothelial cells upon intracellular adhesion molecule1 ligation. J Immunol 2001; 166:6877–6884. 52. Greenwood J, Etienne-Manneville S, Adamson P, Couraud P. Lymphocyte migration into the central nervous system: implication of ICAM-1 signalling at the blood-brain barrier. Vascul Pharmacol 2002; 38:315–322. 53. Etienne-Manneville S, Manneville J, Adamson P, Wilbourn B, Greenwood J, Couraud P. ICAM-1-coupled cytoskeletal rearrangements and transendothelial lymphocyte migration

Oxygen regulation of endothelial cell phenotypes

209

involve intracellular calcium signaling in brain endothelial cell lines. J Immunol 2000; 165:3375–3383. 54. Kiely J, Hu Y, Garcia-Cardena G, Gimbrone MJ. Lipid raft localization of cell surface Eselectin is required for ligation-induced activation of phospholipase C gamma. J Immunol 2003; 171:3216–3224. 55. Yan S-F, Lu J, Zou YS, et al. Hypoxia-associated induction of early growth response-1 gene expression. J. Biol Chem 1999; 274:15030–15040. 56. Semenza GL. Hypoxia-inducible factor 1: master regulator of O2 homeostasis. Curr Opin Genet Dev 1998; 8:588–594. 57. Semenza G. Series introduction: tissue ischemia: pathophysiology and therapeutics. J Clin Invest 2000; 106:613–614. 58. Ivan M, Kondo K, Yang H, et al. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 2001; 292:464–468. 59. Lando D, Peet D, Whelan D, Gorman J, Whitelaw M. Asparagine hydroxylation of the HIF transactivation domain a hypoxic switch. Science 2002; 295:858–861. 60. Collard C, Vakeva A, Morrissey M, et al. Complement activation after oxidative stress: role of the lectin complement pathway. Am J Pathol 2000; 156:1549–1556. 61. Collard C, Agah A, Stahl G. Complement activation following reoxygenation of hypoxic human endothelial cells: role of intracellular reactive oxygen species, NF-kappaB and new protein synthesis. Immunopharmacology 1998; 39:39–50. 62. Collard C, Vakeva A, Bukusoglu C, et al. Reoxygenation of hypoxic human umbilical vein endothelial cell activates the classic complement pathway. Circulation 1997; 96: 326–333. 63. Collard C, Agah A, Reenstra W, Buras J, Stahl G. Endothelial nuclear factor-kappaB translocation and vascular complement adhesion molecule-1 induction by complement: inhibition with anti-human C5 therapy or cGMP analogues. Arterioscler Thromb Vasc Biol 1999; 19:2623–2629. 64. Peerschke E, Reid K, Ghebrehiwet B. Platelet activation by C1q results in the induction of alpha IIb/beta 3 integrins (GPIIb-IIIa) and the expression of P-selectin and procoagulant activity. J Exp Med 1993; 178:579–587. 65. Huang J, Kim LJ, Mealey R, et al. Neuronal protection in stroke by an sLex-glycosylated complement inhibitory protein. Science 1999; 285:595–599. 66. Luscher T, Barton M. Biology of the endothelium. Clin Cardiol 1997; 20:3–10. 67. Karimova A, Pinsky D. The endothelial response to oxygen deprivation: biology and clinical implications. Intensive Care Med 2001; 27:19–31. 68. Stoltz J, Muller S, Wang X, et al. Hemorheology and vascular endothelial cells. Clin Hemorheol Microcirc 1999; 20:127–139. 69. Moncada S. Nitric oxide: discovery and impact on clinical medicine. J R Soc Med 1999; 92:164–169. 70. Cahill P, Redmond E, Sitzmann J. Endothelial dysfunction in cirrhosis and portal hypertension. Pharmacol Ther 2001; 89:273–293. 71. Arnal J, Dinh-Xuan A, Pueyo M, Darblade B, Rami J. Endothelium-derived nitric oxide and vascular physiology and pathology. Cell Mol Life Sci 1999; 55:1078–1087. 72. McQuillan LP, Leung GK, Marsden PA, Kostyk SK, Kourembanas S. Hypoxia inhibits expression of eNOS via transcriptional and posttranscriptional mechanisms. Am J Physiol 1994; 267:H1921–H1927. 73. Marnett L, Rowlinson S, Goodwin D, Kalgutkar A, Lanzo C. Arachidonic acid oxygenation by COX-1 and COX-2. Mechanisms of catalysis and inhibition. J Biol Chem 1999; 274:22903– 22906. 74. Topper J, Cai J, Falb D, Gimbrone MJ. Identification of vascular endothelial genes differentially responsive to fluid mechanical stimuli: cyclooxygenase-2, manganese super-oxide dismutase, and endothelial cell nitric oxide synthase are selectively up-regulated by steady laminar shear stress. Proc Natl Acad Sci USA 1996; 93:10417–10422.

Endothelial cells in health and disease

210

75. Luckhoff A, Pohl U, Mulsch A, Busse R. Differential role of extra- and intracellular calcium in the release of EDRF and prostacyclin from cultured endothelial cells. Br J Pharmacol 1988; 95:189–196. 76. Edwards G, Dora K, Gardener M, Carland C, Weston A. K+ is an endothelium-derived hyperpolarizing factor in rat arteries. Nature 1998; 396:213, 215–216. 77. Rubanyi G. Endothelium-derived relaxing and contracting factors. J Cell Biochem 1991; 46:27–36. 78. Argarwal A, Nick H. Renal response to tissue injury: lessons from heme oxygenase -1 GeneAblation and expression. J Am Soc Nephrol 2000; 11:965–973. 79. Willis D, Moore AR, Frederick R, Willoughby DA. Heme oxygenase: a novel target for the modulation of the inflammatory response. Nat Med 1996; 2:87–90. 80. McCoubrey WK Jr, Huang TJ, Maines MD. Isolation and characterization of a cDNA from the rat brain that encodes hemoprotein heme oxygenase-3. Eur J Biochem 1997; 247:725–732. 81. Morita T, Perralla MA, Lee M-E, Kourembanas S. Smooth muscle cell-derived carbon monoxide is a regulator of vascular cGMP. Proc Natl Acad Sci (USA) 1995; 92: 1475–1479. 82. Hayashi S, Takamiya R, Yamaguchi T, et al. Induction of heme oxygenase-1 suppresses venular leukocyte adhesion elicited by oxidative stress: role of bilirubin generated by the enzyme. Circ Res 1999; 85:663–671. 83. Brune B, Ullrich V. Inhibition of platelet aggregation by carbon monoxide is mediated by activation of guanylate cyclase. Mol Pharmacol 1987; 32:497–504. 84. Morita T, Kourembanas S. Endothelial cell expression of vasoconstrictors and growth factors is regulated by smooth muscle cell-derived carbon monoxide. J Clin Invest 1995; 96:2676–2682. 85. Rubanyi G, Botelho L. Endothelins. FASEB J 1991; 12:2713–2720. 86. Esmon C. Regulation of blood coagulation. Biochim Biophys Acta 2000; 1477:349–360. 87. Ogawa S, Shreeniwas R, Brett J, Clauss M, Furie M, Stern DM. The effect of hypoxia on capillary endothelial cell function: modulation of barrier and coagulant function. Br J Haematol 1990; 75:517–524. 88. Archipoff G, Beretz A, Bartha K, et al. Role of cyclic AMP in promoting the thromboresistance of human endothelial cells by enhancing thrombomodulin and decreasing tissue factor activities. Br J Pharmacol 1993; 109:18–28. 89. Lawson CA, Yan S-D, Yan S-F, et al. Monocytes and tissue factor promote thrombosis in a murine model of oxygen deprivation. J Clin Invest 1997; 99:1729–1738. 90. Yan S-F, Zou YS, Gao Y, et al. Tissue factor transcription driven by egr-1 is a critical mechanism of murine pulmonary fibrin deposition in hypoxia. Proc Natl Acad Sci USA 1998; 95:8298–8303. 91. Khachigian L, Lindner V, Williams A, Collins T. Egr-1-induced endothelial gene expression: a common theme in vascular injury. Science 1996; 271:1427–1431. 92. Khachigian L, Williams A, Collins T. Interplay of Sp1 and Egr-1 in the proximal plateletderived growth factor A-chain promoter in cultured vascular endothelia cells. J Biol Chem 1995; 270:27679–27686. 93. Pinsky DJ, Liao H, Lawson CA, et al. Coordinated induction of plasminogen activator inhibitor-1 (PAI-1) and inhibition of plasminogen activator gene expression by hypoxia promotes pulmonary vascular fibrin deposition. J Clin Invest 1998; 102:919–928. 94. Liao H, Heath M, Pinsky DJ. Stimulation of the NO/cGMP pathway reduces hypoxia-induced endothelial P-selectin expression: a mechanism whereby nitrosovasodilators reduces neutrophil adhesion. Circlation 1996; 94. 95. Davenpeck KL, Gauthier TW, Lefer AM. Inhibition of endothelial-derived nitric oxide promotes P-selectin expression and actions in the rat microcirculation. Gastroenterology 1994; 107:1050–1058. 96. Radomski M, Palmer R, Moncada S. Modulation of platelet aggregation by an L-arginine-nitric oxide pathway. Trends Pharmacol Sci 1991; 12:87–88.

Oxygen regulation of endothelial cell phenotypes

211

97. Furlong B, Henderson A, Lewis M, Smith J. Endothelium-derived relaxing factor inhibits in vitro platelet aggregation. Br J Pharmacol 1987; 90:687–692. 98. Sneddon JM, Vane JR. Endothelium-derived relaxing factor reduces platelet adhesion to bovine endothelial cells. Proc Natl Acad Sci USA 1988; 85:2800–2804. 99. Zweier JL, Kuppusamy P, Lutty GA. Measurement of endothelial cell free radical generation: Evidence for a central mechanism of free radical injury in postischemic tissues. Proc Natl Acad Sci USA 1988; 85:4046–4050. 100. Shiraishi Y, Lee JR, Laks H, et al. L-arginine administration during reperfusion improves pulmonary function. Ann Thoracic Surg 1996; 62:1586–1587. 101. Pinsky DJ, Oz MC, Koga S, et al. Cardiac preservation is enhanced in a heterotopic rat transplant model by supplementing the nitric oxide pathway. J Clin Invest 1994; 93:2291–2297. 102. Marcus A, Broekman M, Drosopoulos J, et al. The endothelial cell ecto-ADPase responsible for inhibition of platelet function is CD39. J Clin Invest 1997; 99: 1351–1360.

10 Fluid Mechanical Forces as Extrinsic Modifiers of Endothelial Function Johannes R.Kratz, Kush Parmar, Sripriya Natarajan, and Guillermo García-Cardeña Laboratory for Systems Biology, Center for Excellence in Vascular Biology, Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, U.S.A.

1. INTRODUCTION From the moment the heart starts beating and blood flow is first established for first time in the developing vertebrate embryo, the cardiovascular system is constantly exposed to fluid mechanical forces. The pulsatile nature of blood flow generates a complex interplay of three distinct types of fluid mechanical forces: wall shear stresses, cyclic strains, and hydrostatic pressures (1). These hemodynamic factors act on the cells that comprise the vascular wall, in particular the endothelium, influencing their structure and function (2). There is increasing evidence that mechanical stimulation plays an important role in the development of the vasculature, the maintenance of vascular integrity and homeostasis, and the development of vascular diseases (2–5). The endothelial lining of the heart and vasculature comprise a dynamic interface with the blood and acts as an integrator and transducer of both humoral and mechanical stimuli (3). This single-cell-thick layer is able to rapidly sense changes in blood flow and respond by secreting or metabolizing potent vasoactive substances (e.g., nitric oxide) that contribute to pressure/flow homeostasis. In face of chronic flow changes, a more deliberate structural remodeling of the vessel wall also can occur via endothelium-dependent mechanisms (6,7). These adaptive responses reflect rapid changes in protein function, enzymatic activity, and transient or long-lasting effects on endothelial gene expression. Moreover, multiple studies in in vitro model systems have confirmed that fluid shear stresses, comparable to those generated by the frictional force of blood flow on the endothelial lining in vivo, can directly influence protein function, enzymatic activities (1), and transcriptional events in cultured endothelial monolayers influencing their functional phenotype (8–11). It remains a central question in the field of vascular biology how these mechanical forces are sensed by the cells of the blood vessel wall and then translated into pathophysiologically relevant phenotypic changes. Activation of various signaling cascades and transcription

Fluid mechanical forces as extrinsic modifiers

213

factor systems have helped to provide insight into the cellular mechanisms linking shear stress stimuli and genetic regulatory events. It is now clear that endothelial cells have the capacity not only to sense fluid mechanical forces, but also to discriminate among distinct types of forces (8,10) Collectively, these observations strongly suggest that fluid mechanical forces can act as local “extrinsic modifiers” of endothelial functions within the vascular tree. This chapter focuses on the emerging areas where the role of fluid mechanical forces, in particular shear stress, is being explored with exquisite experimental approaches that allow us to unveil the molecular links between mechanics and biology.

2. CHARACTERIZING THE FLUID MECHANICAL STIMULI AND THE ENDOTHELIAL MECHANOSENSING SYSTEM 2.1. Blood Flow Patterns in the Human Circulatory System Although one central organ, the human heart, generates the force to pump blood throughout the entire vascular network, the different location with respect to the heart, diameters, curvatures, and surface characteristics of blood vessels lead to distinct blood flow patterns in different regions of the human vasculature. As a result of these differences, a complex set of fluid mechanical forces (e.g., shear stress, pressure, cyclic strain) are continuously imposed on the endothelial lining of these vessels. Thus, a current challenge in this area is to characterize with high accuracy the distinct flood patterns experienced by the endothelium along the circulatory system in order to simulate these flow profiles in vitro, and to be able to link hemodynamics with in vivo measurements of endothelial function. For large and medium scale vessels, the cells in blood are so small relative to the vessel diameter that the blood can be approximated as a Newtonian fluid. Thus, meaningful measurements of blood flow velocities can be made, and the shear stress that the endothelial cells sense is linearly related to the change in velocity profile from the center to the wall of the vessel. The most common technique for measuring blood velocity is Doppler ultrasound. In this method, the change of frequency in sound waves reflecting from inside the vessel is used to determine the velocity of the blood flowing within (12). Magnetic resonance (MR) imaging can also be used to measure blood velocities (13). In addition, structural MR or CT images of the vessels can be combined with ultrasound measurements of blood velocities to mathematically compute the blood flow patterns as they change through the length of a vessel (14). These techniques have demonstrated several distinct patterns of blood flow. Along the straight portions of medium and large arteries, e.g., the internal carotid artery or the abdominal aorta, the direction of blood flow is parallel to the vessel wall. There is a pulse of fast flow (peak shear stress from 10–70dyn/cm2) (15), followed by a deceleration to a

Endothelial cells in health and disease

214

slower flow rate, and then a more gradual acceleration to a steady velocity that is maintained for the rest of the cardiac cycle (shear stress of 0–10 dyn/cm2). In some vessels, blood may transiently flow backwards during the deceleration phase. This pulsatile nature of the flow, characterized by the rate at which blood accelerates to its peak speed, is a hallmark feature of arterial flow. In large and medium-caliber straight veins (e.g., saphenous veins), blood flow is still predominantly one-dimensional, but is much slower (peak shear stress of approximately 1–2dyn/cm2). Blood travels at this speed in the forward direction for approximately twothirds of the cardiac cycle, and in the reverse direction, at a similar speed, for the rest of the time. The acceleration and deceleration of the blood to reach its peak velocities are much slower than they are for arterial flow. In the branch points and curvatures of vessels, these flow patterns no longer hold. In arteries, disturbed flow occurs at these points—three-dimensional whorls of flow that often have a very low time-average speed in any one given direction (14). The aortic arch, carotid bulb, and branch points of the aorta are all regions that experience disturbed flow. Although different flow patterns have been established for several vessels, there are still many more regions that remain to be well characterized. The coronary arteries in particular are of great clinical interest, due to the high prevalence of ather-osclerosis in these vessels, but their smaller diameter (in comparison to the aorta or carotid artery) and constant motion with the beating heart has made measuring the velocity within the coronary vasculature especially challenging. The shear stress that endothelial cells experience in capillaries has also been challenging to characterize since whole blood can no longer be considered a Newtonian fluid owing to the small diameter of vessel.

2.2. Modeling Blood Flow Patterns In Vitro Much insight into the responses endothelial cells display when exposed to flow comes from in vitro experiments, where the effect of flow can be studied in a well-defined and controlled fashion and in isolation from other hemodynamic factors, such as blood pressure and cyclic strain (1). Thus, modeling more realistic flow patterns seen that better approximate those seen in vivo has been a constant challenge of such experiments (Fig. 1). The first in vitro flow experiments exposed endothelial cells to a uniform laminar shear stress with a value equal to physiological time-average values. Such laminar flows can be generated using either parallel plate chambers (16,17) or a cone-plate apparatus (18). With the latter device, the rotation of the cone can be controlled to generate onedimensional waveforms that may be closer to the cyclic nature of physiological flows— sinusoidal waveforms or oscillating square waves (19). More recently, a cone-plate apparatus controlled by a computerized motor has been able to reproduce most onedimensional physiological flow waveforms as captured by Doppler ultrasound (20). Another approach to mimic more physiological flows has been to grow endothelial cells on small tubes and pump fluid through. Pumping fluid through these tubes in a cyclic fashion exposes the endothelial layer to a pulsatile flow pattern (21). However, the precise definition of the several hemodynamic components present in these systems remains to be defined.

Fluid mechanical forces as extrinsic modifiers

215

Mimicking the two-and three-dimensional disturbed flow patterns seen in sites of pathology—such as the atherosclerosis-prone carotid bulb—has been another important in vitro effort. Cone-plate devices with sufficiently large angles between the cone and plate will generate turbulent (i.e., randomly varying, both temporally and spatially) instead of laminar flow over endothelial cells (19). Another model places a square step in the lower plate of a parallel plate chamber, establishing disturbed (i.e., spatial-varying) flow patterns at the junction of the step and plate (22). All these models have been extremely helpful for increasing our understanding of the influence of shear stress on the structure and function of endothelial cells.

2.3. Endothelial Mechanotransduction The capacity displayed by endothelial cells to sense and discriminate distinct flowinduced shear stresses raises the question of how these cells sense these mechanical

Figure 1 Modeling blood flow patterns in vitro. Different characteristics of the complex flow patterns seen in the vasculature have been captured in flow patterns used for in vitro experiments. Typical flows have been modeled as steady laminar flow, where the shear stress is constant over time and maintained at a physiologic average level, or as oscillatory flow with either a square wave or sinusoidal patterns.

Endothelial cells in health and disease

216

Nonlaminar flows have been modeled as turbulent (varying randomly with time) or disturbed (varying over space). Recently Doppler ultrasound measurements of flow have been used to develop flow patterns that more closely mimic physiological patterns, such as one-dimensional abdominal aortic flow. forces. Although the true nature of the mechanosensing and mechanotransduction systems in endothelial cells remains poorly understood, several plausible and promising hypotheses have been put forward in recent years. Donald Ingber and colleagues have proposed that the main mechanism of force-sensing by cells comes from changes to the cytoskeletal structure as a cell is deformed by mechanical forces. According to this model (termed, tensegrity), flow applies a shear stress-derived force to the entire cytoskeleton and some transmembrane molecules (e.g., integrins) that are linked to the cytoskeleton. This cytoskeletal association allows such molecules to exert a force through cytoskeletal components, transducing the force to a different region of the cell—perhaps altering nuclear structure and affecting transcription, or changing the physical conformation of a protein to activate it. In support of such a theory, the stiffness of the cell cytoskeleton has been demonstrated to increase in response to selectively applying a shear force to certain cell surface integrin receptors. In addition, these forces immediately cause changes at a distance, in the arrangement of molecular assemblies in the cell nucleus (23). The tensegrity model does not exclude the possibility of additional discrete molecules that serve to sense shear stress and transduce a signal independently of the cell cytoskeleton. Mathematical models of a molecule comprised of viscoelastic elements show that such a sensor on the surface of an endothelial cell would deform drastically at the onset of laminar flow, then approach a steady state behavior, which is the pattern seen in many endothelial responses to flow (24). In addition, such an element’s deformation would be different upon exposure to distinct types of shear stress (25). In the past several years, a number of putative flow sensors have been proposed. Ion channels are a major class of molecules that may behave in such a manner. Within seconds after cessation of flow over pulmonary endothelial cells that have been exposed to flow for 24 hr, the cells experience a membrane depolarization of approximately 20 mV, which is mediated by an inwardly rectifying K+ ATP channel (26). This change in membrane potential can cause a Ca2+ flux into the cell, triggering a number of signaling events (26,27). Na+ and Cl−channels may also be involved in sensing flow-induced shear stress (28,29). Other cell surface molecules have also been suggested to function as endothelial flow sensors. For example, under flow-induced shear stress, the VEGF receptor 2 (VEGF-R2, Flk-1) can be activated independently of its ligand (30). VEGF-R2 has been shown under flow to form a complex with the endothelial adherens junction proteins VE-cadherin and β-catenin. In endothelial cells derived from VE-cadherin −/− mice, several flowdependent phosphorylation events downstream of VEGF-R2 activation and several gene regulation changes were absent (31). Furthermore, αvβ3 and α1 integrins have been

Fluid mechanical forces as extrinsic modifiers

217

shown to be necessary for the rapid tyrosine phosphorylation of the VEGF-R2 in response to flow (32). Blocking antibodies against another candidate flow sensor, caveolin-1, which is one of the major structural proteins of caveolae, prevent the phosphorylation of ERK-1/2 that is normally induced in endothelial cells by flow (33). Interestingly, bovine aortic endothelial cells exposed for 24–72 hr to flow-induced shear stress exhibit a shift of caveolin-1 localization from the Golgi complex to the cell surface and increase the number of their caveolae (34). These results are consistent with the hypothesis that caveolae may serve as flow sensing organelles (35) and support the concept that endothelial cells may increase the concentration of mechanosensitive receptors in response to flow in order to become more sensitive to this stimulus. Additional molecules have recently been shown to be more directly affected by flow. Glycosaminoglycans and the proteoglycans to which they bind are promising candidates as flow sensors. The molecules form an often-overlooked apical extracel-lular layer known as glycocalyx. Selective removal of the heparan sulfate component of glycocalyx from endothelial cell membranes greatly abrogates their NO production in response to flow. Meanwhile, NO production in response to bradykinin, histamine, or acetylcholine remains largely unaffected (36,37). NMR studies have proposed a mechanism for this mechanotransductive effect, showing that proteoheparan sulfate changes from a random coil to a filamentous form when exposed to flow, enabling it to bind Na+ and triggering vasodilation (38). Platelet endothelial cell adhesion molecule-1 (PECAM-1) is phosphorylated in response to flow-induced shear stress and in turn promotes ERK phosphorylation. Using antibody-coated magnetic beads, PECAM-1 was selectively exposed to a tangential force, triggering its phosphorylation (39). G-proteins have been implicated in many of the pathways triggered by flow-induced shear stress, but they have also been proposed as a sensor of the flow itself. The application of 0–30 dyn/cm2 of flow-induced shear stress to phospholipid vesicles containing purified G-proteins, activated the G-proteins, with increased activation at higher levels of shear stress (40). Although these approaches to selectively shearing or removing molecules do not entirely exclude cytoskeletal transmission of forces, they do indicate that heparan sulfate, PECAM-1, and G-proteins are highly plausible candidates as endothelial flow sensors. It is thus clear that several endothelial cell molecules respond to fluid mechanical forces, and these responses may or may not be directly related to cytoskeletal changes determined by cellular tensegrity. Future challenges will be not only to uncover the integration of these various mechanosensors at the cellular systems level, but also to probe the selective importance of these sensors for specific cellular processes, which will ultimately help us better understand their dysregulation in disease and develop targeted therapeutics.

Endothelial cells in health and disease

218

3. FLUID MECHANICAL FORCES AND VASCULAR DEVELOPMENT 3.1. Historical Perspective The modern era of cardiovascular developmental biology dates to the major experimental and conceptual breakthroughs of the latter part of the 19th century. The German embryologist Wilhelm Roux postulated vascular development to proceed in three phases: a stage of primary differentiation governed entirely by “hereditary principles,” followed by a transitional stage wherein the genetic element is gradually supplanted by functional adaptation, and finally a stage where further vascular development is completely regulated by mechanical forces acting through the circulation (41). As we hope to convey in this section, the experimental knowledge we have amassed to date appears to confirm and extend Roux’s visionary paradigm of vascular development. Near the end of the 19th century, Richard Thoma penned the three biomechanical laws that would shape the study of cardiovascular development for the decades to follow (42). These laws, translated into English by Bruce, state the following: (1) “The increase in size of the lumen of the vessel wall depends on the rate of the blood current”; (2) “The growth in thickness of the vessel wall is dependent on its tension”; and (3) “Increase in the blood-pressure in the capillary areas leads to new formation of capillaries.” These laws, made by careful observation and deductive logic, were to set the stage for contemporary experimental biologists in the field of cardiovascular development. Elegant work by Thoma and Sabin had already established the presence of an extensive vascular network at the time circulation begins in the chick embryo, including the paired dorsal aortae, the duct of Cuvier, and cardinal veins (42). Given the existence of this primitive vascular plexus, subsequent studies were the first to explore the interrelationship of cardiac function and cardio-vascular development. In his landmark paper, Chapman studied later development of the vascular system after “eliminating most of the mechanical factors mentioned by Thoma.” He did this by surgical removal of the entire heart before the onset of circulation, and noticed a marked failure to form the major omphalomesenteric arteries and veins (43). This result, he concluded, confirmed the existence of Roux’s second stage of development. Indeed, other studies by Loeb and Stockard using potassium chloride to abrogate cardiac contractility came to similar conclusions (44). A study by Knower noted the irregular course of vessels and impaired formation of glomeruli after removal of the heart in frog embryos (45). These latter studies suggested that later stages of vascular development were subject to regulation by blood flow. There were several drawbacks to the pioneering work described above, including relatively primitive experimental manipulation of living embryos and possibly confounding effects of hypoxia and nutrient delivery. The limitations notwith-standing, they were to foreshadow the current era of molecular underpinnings of biological processes as it applies to vascular development.

Fluid mechanical forces as extrinsic modifiers

219

3.2. Cardiovascular Development: Genetics and Epigenetics The interplay of genetic and epigenetic influences on cardiovascular development is still poorly understood, but experimental verification of the importance of this relationship has recently been very informative. One of the most important advances to enable the understanding of these processes has been the identification of molecular markers of distinct endothelial identities. For example, vascular structures have been physiologically distinguished as arteries or veins since the time of William Harvey, but the elucidation of a molecular distinction only came with the recent discovery of the expression pattern of Ephrin-B2 in mouse embryos (46). Ephrin-B2 is expressed specifically in arterial endothelium and smooth muscle cells (47). Mice homozygous null for the EphrinB2 gene display vascular defects specifically in the arterial vessels (46). Analogous studies the zebra fish Danio rerio first identified Hey2, a transcriptional target of the canonical Notch pathway, as exhibiting arterial-specific expression (48). Homozygous hey2 mutants completely fail to fashion the dorsal aorta and have hearts reminiscent of coarctation defects commonly seen in human congenital heart disease (49). The number of molecular markers of arteries and veins has grown since these first series of studies were reported. They include Notch family members, Ephrins and their receptors, Neuropilins, and Connexin 40 among others (see Chapter 7) (50). The identification of these markers, in conjunction with the development of the zebra fish as an experimental system, has allowed for some of the most revealing in vivo studies probing the role of fluid forces in cardiovascular development. The power of this model system to ask questions about development lies primarily in the fact that the zebra fish embryo is perfectly viable with oxygen delivered by diffusion only, with no need for erythrocytes for at least five days, well past the development of a complete cardiovascular system. The first study to identify the relationship between cardiac function and peripheral vascular developmental in zebra fish was through a mutagenesis screen for defects in kidney development (50). It was noted that most of the embryos with impaired renal development also exhibited severe cardiac defects. To ask whether there was a functional relationship, the group examined embryos mutant for a known cardiaospecific gene required for contractility. Indeed, the authors of this study discovered in these cardiac mutants, a marked impairment in formation of the glomerular capillary network, which norm ally migrates to form an intimate plexus with the primordial renal tubular network (51). It was found that the endothelium of these heart mutants had dramatically reduced levels of MMP-2 expression, an important metalloproteinase for glomerular development. Importantly, reduction of MMP-2 levels could phenocopy the renal defects of the heart mutants. Lastly, in a careful control for the confounding issue of tissue perfusion, complete replacement of blood with saline in normal embryos had no effect on formation of the glomeruli. Thus, the mechanical force of blood flow is critical for proper glomemlar capillary development in zebra fish, and this was related to an induced change in gene expression in glomerular endothelium. It is interesting to note that the study a century before by Knower came to this same conclusion in frog embryos (44). Like the endothelium of vessels, the endocardial lining of the heart readily senses shear stress, and the role of these forces during cardiogenesis has recently been elegantly

Endothelial cells in health and disease

220

probed in zebra fish. Hove et al. used in vivo imaging to reveal the flow vortices inside the atria and ventricle of the developing zebra fish heart (52). Importantly, calculated shear stresses from these data were substantial (~2.5dyn/cm2 at 37 hr, ~76 dyn/cm2 at 4.5 days), well within the limits detectable by non-endocardial endothelial cells (minimum of ~1 dyn/cm2). This indicates that, at least in zebra fish, endocardial cells could transduce this mechanical force into altered gene expression and consequent cardiac morphogenesis. Indeed, Hove et al. observed severely disrupted cardiogenesis after occlusion of either the sinus venosus or the outflow tract in the early embryo. Because these two manipulations would produce opposing effects on intracardiac pressure, it follows that the commonality of abrogated shear stress is the underlying factor critical for proper cardiogenesis. In these occluded hearts, Hove et al. observed failure to form the third chamber, the bulbus cordis, and also impaired cardiac looping, a conserved mechanism that shifts the atrio-ventricular alignment from cephalocaudal to lateral positions. Lastly, they noted the collapse of both the inflow and outflow tracts. Intracardiac fluid forces are thus an essential component of proper cardiac development, and could possibly underlie the large percentage of congenital heart diseases of idiopathic nature. The role of the fluid forces in the formation of the vascular network and the identity of arteries and veins during zebra fish development has also recently been examined. In silent heart embryos, which have no heart beat, endothelial-specific transgenic expression of GFP reveals that the overall pattern of the vascular system is intact, just as the embryologists mentioned above noticed in their primitive experiments. However, careful utilization of this experimental system revealed that the fluid forces of circulation are in fact necessary to determine the functional relationship of angiogenic sprouts and the original vessel tree (53). Fluid forces thus determine the arterial or venous fate of endothelium within the context of the fully developed circulatory plexus in the zebra fish trunk. The field of vascular development, at a time of sophisticated methods and complex hypotheses, thus comes full circle with the visionary postulates by Roux.

3.3. The Lymphatic System A far less studied aspect of vascular development is the lymphatic circulation. Despite the dearth of molecular insights into lymphatic endothelium, some important work has implicated lymphatic flow as a critical mediator of lymphangiogenesis. Boardman et al. (54) replaced a small band of dermis from mouse tails with a type I collagen gel, allowing for flow of interstitial fluid and unobstructed visualization of lymphangiogenesis. Importantly, flow of lymphatic fluid preceded formation of the lymphatic vessels, which were identified with surface markers Flt4 and LYVE-1 (lymphatic endothelial hyaluronan receptor). Fluid channels in the collagen gel were observed well before the lymphatic endothelial cells formed proper tubes, and appeared to guide the precise tracks taken subsequently by the migrating endothelial cells. Furthermore, an important link was found between the expression pattern of VEGF-C, a major lymphangiogenic factor, and the direction of growth of these vessels. The pattern of VEGF-C expression suggested a role for fluid flow in the expression and transport of this molecule, as it first appeared at the sites of fluid channels. The proposed model implicates a central role for interstitial fluid flow in the guidance of subsequent migration

Fluid mechanical forces as extrinsic modifiers

221

and lumenization of lymphatic endothelium. This is in contrast to the vasculature, where lumen patency precedes fluid flow, rendering the role of flow more limited to regulation of the endothelium in the already formed vessel. The above studies provide strong evidence that fluid mechanical forces are critical for a number of aspects of cardiovascular development. However, they leave us with little understanding of the molecular mechanisms underlying this fascinating paradigm. From the question of mechanosensing by the endothelium, to the translation of these signals to changes in gene expression, to cell-cell signaling within the microenvironment, there lays an enormous gap in understanding between fluid forces and ultimate organismal development. This evolutionary conserved role of fluid flow is probably specifically adapted to each process, but some commonality might also exist. Even from the cursory understanding we are privy to now, it is apparent that changes in MMP expression are implicated in the role of fluid forces in both glomerular and lymphatic development. Surely unbiased investigation of endothelial and endocardial genomic expression patterns in response to shear stresses will aid to bridge the gulf between mechanical forces and complex developmental processes. Furthermore, the genetic specialization of a particular endothelial bed (e.g., endocardial, glomerular, cerebral) will add a further level of complexity in the genomic response to fluid forces. It is this interaction of microenviromental epigenetic influences with classical genetic predetermination that must confer an endothelial phenotype befitting its context.

4. FLUID MECHANICAL FORCES AND DISEASE 4.1. Atherogenesis The involvement of vascular endothelium in disease processes such as atherosclerosis has been recognized since the time of Virchow (55). However, mechanistic insight into the pathobiology of this tissue has developed only recently, largely as a result of the application of modern cellular and molecular biological techniques (56). We now appreciate that the single-cell thick lining of the circulatory system is, in fact, a vital organ whose health is essential to normal vascular physiology and whose dysfunction can be a critical factor in the pathogenesis of vascular disease. For example, the nonrandom distribution of early lesions of atherosclerosis in human subjects and experimental animals remains one of the most consistent, intriguing, and incompletely understood aspects of this disease process. Lesions typically develop in the vicinity of branch points and areas of major curvature within the arterial vasculature. Physical and computational models have identified these vascular regions as having low time-average shear stress, a high oscillatory shear index, and steep temporal and spatial gradients in shear stress. In contrast, unbranched arterial geometries that are exposed to more uniform laminar flows appear relatively protected from lesion development (57,58). In an effort to unveil mechanistic links between hemodynamics and atherogenesis, research over the last 20 years has focused on emulating these flow characteristics in vitro (as described in Sec. 1 of this chapter) to delineate the signaling pathways and downstream changes at the level

Endothelial cells in health and disease

222

of gene expression and cell structure-/function. The development and use of these in vitro models in several laboratories has allowed investigators to demonstrate direct effects of biomechanical forces on the expression of pathophysiologically relevant endothelial genes (3). In addition, differential upregulation of certain “atheroprotective” genes (i.e., eNOS, Mn-SOD, and COX-2) in endothelial cells exposed to laminar shear stress, but not to nonlaminar shear stress has been documented (59). These observations led to a guiding hypothesis for the field developed by Gimbrone and colleagues (59), namely, that the steady laminar shear stresses characteristic of lesion-protected areas elicit expression of “ather-oprotective” genes, whereas the altered shear stresses generated by disturbed laminar flows in lesion-prone areas elicit the expression of “atheropathogenic” genes (and/or suppress “atheroprotective” genes). To test this hypothesis at a fundamental cellular level, the Gimbrone laboratory and others used transcriptional profiling to capture the gene expression programs of endothelial cells exposed to distinct types of biomechanical stimuli (8,10). Results from these experiments clearly demonstrated that endothelial cells can differentially sense and transduce distinct biomechanical input stimuli into different patterns of gene expression. Moreover, using these profiles as predictors of function, these groups were able to demonstrate that endothelial cells can translate these input stimuli into distinctive functional phenotypes. Critical testing of Gimbrone’s “atheroprotective gene hypothesis” will depend upon refinement of both in vitro and in vivo fluid mechanical models and a validation of candidate atheroprotective genes in the setting of human vascular pathobiology. The development of reliable methods for linear amplification of transcripts from small numbers of cells and their analysis by cDNA microarrays or analogous genome scale technologies holds much promise in this regard as recently demonstrated by Peter Davies’s group (60). Application of these comprehensive and relatively unbiased methods of molecular analysis to endothelial cells subjected to experimentally defined flow conditions will add significantly to our understanding of the dynamic range of biomechanically induced phenotypic modulation. Ultimately, the extension of this method of analysis to endothelial phenotype in the natural disease context should provide valuable new insights into the links between fluid mechanical forces and atherogenesis.

4.2. Coronary Artery Bypass and Graft Failure While several medical therapies have been developed to prevent coronary artery disease, such as antihypertensive, cholesterol-lowering, and diabetes-controlling agents, surgical and interventional therapy remain the definitive treatment for advanced disease. Today, coronary artery bypass graft (CABG) surgery is performed on over 400,000 patients each year in the United States with a significant reduction in morbidity and mortality. Unfortunately, within the first year after surgery, 15% of venous grafts occlude, and after 10 years, only 60% of vein grafts remain patent. Subsequent revascularization, either reoperative surgery or percutaneous intervention, is required in 4% of patients after 5 years and 19% of patients after 10 years (61). The clinical impact of saphenous vein graft disease is currently increasing, and efforts to reverse this trend mandate an improved understanding of the molecular mechanisms of venous arterialization and the pathogenesis of failed bypass grafts.

Fluid mechanical forces as extrinsic modifiers

223

The remodeling of vessels in response to changes in environmental cues (e.g., blood flow, oxygen tension) as occurs during CABG surgery may initially be adaptive, but eventually becomes pathologic with the development of thrombosis followed by a cascade of events leading to graft failure (62). Arterialized venous bypass grafts demonstrate accelerated arteriosclerosis compared to normal vessels in spite of the fact that veins in their endogenous sites rarely develop atherosclerotic lesions. Thus, the biomechanical environment of the arterial milieu seems to play a critical role in vessel remodeling and subsequent graft failure. The venous-to-arterial shift of the local environment engenders reorganization of the venous vascular architecture, such that over time, the venous graft acquires an arterylike structure. Early anatomical adaptive changes include an elevation in cellular mass mainly caused by an increase in the number of smooth muscle cells in the vessel wall (medial thickening). Several studies have demonstrated that cell proliferation and apoptosis play an important role in this process (63). Insights into the molecular mechanisms of venous arterialization have come from studies in which the expression of individual molecules was characterized. These studies have demonstrated the dynamic expression of genes involved in extracellular matrix formation and turnover in the course of the adaptive changes displayed by the vein during its arterialization (62,64,65). In addition, the expression of several growth factors has been shown to be modulated during the arterialization process including TGF-β, PDGF, and VEGF (62,63). Direct insights into the functional role of specific cells of the vessel wall in vascular remodeling have been mainly derived from studies of arterial remodeling. Studies conducted in models of arterial remodeling demonstrated the obligatory role of vascular endothelium in this process (7). More recently, specific endothelial derived molecules (e.g., nitric oxide) have been shown to be responsible for this endothelial-dependent response (66). These studies strongly suggest that the vascular endothelium plays a prominent role in vessel remodeling by sensing changes in the biomechanical environment and responding to them by changes in gene expression and the production of bioactive molecules. Clearly, the molecular mechanisms of venous arterialization remain to be elucidated. The recent development of a mouse model of venous arterialization in mice (67) should help us to define such mechanisms. The extent to which the phenotype of the endothelium changes in response to the new biomechanical environment (venous-to-arterial), and how these changes orchestrate the remodeling process in the context of venous arterialization in health and disease remain central questions in the field.

5. CONCLUSIONS Our ability to couple in vitro models of pathophysiologically relevant hemodynamic environments with systems biology approaches will continue to yield molecular insights into the mechanoactivated programs in vascular endothelium. Validation of these programs in the context of intact blood vessels should mechanistically link endothelial mechanobiology, vascular development, and cardiovascular disease.

Endothelial cells in health and disease

224

REFERENCES 1. Davies PF. Flow-mediated endothelial mechanotransduction. Physiol Rev 1995; 75: 519–560. 2. Davies PF, Barbee KA, Volin MV, et al. Spatial relationships in early signaling events of flowmediated endothelial mechanotransduction. Annu Rev Physiol 1997; 59: 527–549. 3. Gimbrone MA, Jr, Nagel T, Topper JN. Biomechanical activation: an emerging paradigm in endothelial adhesion biology. J Clin Invest 1997; 100:S61–S65. 4. Traub O, Berk BC. Laminar shear stress: mechanisms by which endothelial cells transduce an atheroprotective force. Arterioscler Thromb Vasc Biol 1998; 18:677–685. 5. Chien 8, Shyy JY. Effects of hemodynamic forces on gene expression and signal transduction in endothelial cells. Biol Bull 1998; 194:390–391 (discussion 392–393). 6. Gibbons GH, Dzau VJ. The emerging concept of vascular remodeling. N Engl J Med 1994; 330:1431–1438. 7. Langille BL, O’Donnell F. Reductions in arterial diameter produced by chronic decreases in blood flow are endothelium-dependent. Science 1986; 231:405–407. 8. Garcia-Cardena G, Comander J, Anderson KR, Blackman BR, Gimbrone MA Jr. Biomechanical activation of vascular endothelium as a determinant of its functional phenotype. Proc Natl Acad Sci USA 2001; 98:4478–4485. 9. Chen BP, Li YS, Zhao Y, et al. DNA microarray analysis of gene expression in endothelial cells in response to 24-h shear stress. Physiol Genomics 2001; 7:55–63. 10. Brooks AR, Lelkes PI, Rubanyi GM. Gene expression profiling of human aortic endothelial cells exposed to disturbed flow and steady laminar flow. Physiol Genomics 2002; 9:27–41. 11. McCormick SM, Eskin SG, McIntire LV, et al. DNA microarray reveals changes in gene expression of shear stressed human umbilical vein endothelial cells. Proc Natl Acad Sci USA 2001; 98:8955–8960. 12. Taylor KJ, Holland S. Doppler US. Part I. Basic principles, instrumentation, and pitfalls. Radiology 1990; 174:297–307. 13. Edelman RR. Basic principles of magnetic resonance angiography. Cardiovasc Intervent Radiol 1992; 15:3–13. 14. Younis HF, Kaazempur-Mofrad MR, Chung C, Chan RC, Kamm RD. Computational analysis of the effects of exercise on hemodynamics in the carotid bifurcation. Ann Biomed Eng 2003; 31:995–1006. 15. Ali MH, Schumacker PT. Endothelial responses to mechanical stress: where is the mechanosensor?. Crit Care Med 2002; 30:S198–S206. 16. Koslow AR, Stromberg RR, Friedman LI, Lutz RJ, Hilbert SL, Schuster P. A flow system for the study of shear forces upon cultured endothelial cells. J Biomech Eng 1986; 108:338–341. 17. Levesque MJ, Nerem RM. The elongation and orientation of cultured endothelial cells in response to shear stress. J Biomech Eng 1985; 107:341–347. 18. Dewey CF, Jr, Bussolari SR, Gimbrone MA, Jr, Davies PF. The dynamic response of vascular endothelial cells to fluid shear stress. J Biomech Eng 1981; 103:177–185. 19. Davies PF, Dewey CF, Jr, Bussolari SR, Gordon EJ, Gimbrone MA, Jr. Influence of hemodynamic forces on vascular endothelial function. In vitro studies of shear stress and pinocytosis in bovine aortic cells. J Clin Invest 1984; 73:1121–1129. 20. Blackman BR, Garcia-Cardena G, Gimbrone MA, Jr. A new in vitro model to evaluate differential responses of endothelial cells to simulated arterial shear stress waveforms. J Biomech Eng 2002; 124:397–407. 21. Moore JE, Jr, Burki E, Suciu A, et al. A device for subjecting vascular endothelial cells to both fluid shear stress and circumferential cyclic stretch. Ann Biomed Eng 1994; 22: 416–422. 22. Hsu PP, Li S, Li YS, et al. Effects of flow patterns on endothelial cell migration into a zone of mechanical denudation. Biochem Biophys Res Commun 2001; 285:751–759.

Fluid mechanical forces as extrinsic modifiers

225

23. Ingber DE. Tensegrity: the architectural basis of cellular mechanotransduction. Annu Rev Physiol 1997; 59:575–599. 24. Mazzag BM, Tamaresis JS, Barakat AI. A model for shear stress sensing and transmission in vascular endothelial cells. Biophys J 2003; 84:4087–4101. 25. Barakat AI. A model for shear stress-induced deformation of a flow sensor on the surface of vascular endothelial cells. J Theor Biol 2001; 210:221–236. 26. Fisher AB, Al-Mehdi AB, Manevich Y. Shear stress and endothelial cell activation. Crit Care Med 2002; 30:S192–S197. 27. Rubanyi GM. Ionic mechanisms involved in the flow- and pressure-sensor function of the endothelium. Z Kardiol 1991; 80(suppl 7):91–94. 28. Traub O, Ishida T, Ishida M, Tupper JC, Berk BC. Shear stress-mediated extracellular signalregulated kinase activation is regulated by sodium in endothelial cells. Potential role for a voltage-dependent sodium channel. J Biol Chem 1999; 274:20144–20150. 29. Barakat AI, Leaver EV, Pappone PA, Davies PF. A flow-activated chloride-selective membrane current in vascular endothelial cells. Circ Res 1999; 85:820–828. 30. Jin ZG, Ueba H, Tanimoto T, Lungu AO, Frame MD, Berk BC. Ligand-independent activation of vascular endothelial growth factor receptor 2 by fluid shear stress regulates activation of endothelial nitric oxide synthase. Circ Res 2003; 93:354–363. 31. Shay-Salit A, Shushy M, Wolfovitz E, et al. VEGF receptor 2 and the adherens junction as a mechanical transducer in vascular endothelial cells. Proc Natl Acad Sci USA 2002; 99:9462– 9467. 32. Wang Y, Miao H, Li S, et al. Interplay between integrins and FLK-1 in shear stress-induced signaling. Am J Physiol Cell Physiol 2002; 283:C1540–C1547. 33. Park H, Go YM, Darji R, et al. Caveolin-1 regulates shear stress-dependent activation of extracellular signal-regulated kinase. Am J Physiol Heart Circ Physiol 2000; 278: H1285– H1293. 34. Boyd NL, Park H, Yi H, et al. Chronic shear induces caveolae formation and alters ERK and Akt responses in endothelial cells. Am J Physiol Heart Circ Physiol 2003; 285: H1113–H1122. 35. Rizzo V, McIntosh DP, Oh P, Schnitzer JE. In situ flow activates endothelial nitric oxide synthase in luminal caveolae of endothelium with rapid caveolin dissociation and calmodulin association. J Biol Chem 1998; 273:34724–34729. 36. Mochizuki S, Vink H, Hiramatsu O, et al. Role of hyaluronic acid glycosaminoglycans in shearinduced endothelium-derived nitric oxide release. Am J Physiol Heart Circ Physiol 2003; 285:H722–H726. 37. Florian JA, Kosky JR, Ainslie K, Pang Z, Dull RO, Tarbell JM. Heparan sulfate proteoglycan is a mechanosensor on endothelial cells. Circ Res 2003; 93:e136–e142. 38. Siegel G, Malmsten M, Klussendorf D, Walter A, Schnalke F, Kauschmann A. Blood-flow sensing by anionic biopolymers. J Auton Nerv Syst 1996; 57:207–213. 39. Osawa M, Masuda M, Kusano K, Fujiwara K. Evidence for a role of platelet endothelial cell adhesion molecule-1 in endothelial cell mechanosignal transduction: is it a mechanoresponsive molecule?. J Cell Biol 2002; 158:773–785. 40. Gudi S, Nolan JP, Frangos JA. Modulation of GTPase activity of G proteins by fluid shear stress and phospholipid composition. Proc Natl Acad Sci USA 1998; 95: 2515–2519. 41. Roux W. Uber die Verzweigungen der Blutgefasse des Menschen. Jenaische Zeitschr f Naturw 1878; 12:205–206. 42. Thoma R. Textbook of General Pathology and Pathological Anatomy. London: Adam and Charles Black, 1896. 43. Chapman WB. The effect of the heart-beat upon the development of the vascular system in the chick. Am J Anat 1918; 23:175–203. 44. Loeb J. On some facts and principles of physiological morphology. Biological Lectures Delivered at the Marine Biol. Boston: Laboratory at Wood’s Hole, 1893:37–61.

Endothelial cells in health and disease

226

45. Knower HM. Effects of early removal of the heart and arrest of the circulation on the development of frog embryos. Anat Rec 1907; 7:161–165. 46. Wang HU, Chen ZF, Anderson DJ. Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4. Cell 1998; 93:741–753. 47. Shin D, Garcia-Cardena G, Hayashi S, 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 2001; 230:139–150. 48. Zhong TP, Rosenberg M, Mohideen MA, Weinstein B, Fishman MC. gridlock, an HLH gene required for assembly of the aorta in zebrafish. Science 2000; 287:1820–1824. 49. Zhong TP, Childs S, Leu JP, Fishman MC. Gridlock signalling pathway fashions the first embryonic artery. Nature 2001; 414:216–220. 50. Lawson ND, Weinstein BM. Arteries and veins: making a difference with zebrafish. Nat Rev Genet 2002; 3:674–682. 51. Serluca FC, Drummond IA, Fishman MC. Endothelial signaling in kidney morphogenesis: a role for hemodynamic forces. Curr Biol 2002; 12:492–497. 52. Hove JR, Koster RW, Forouhar AS, Acevedo-Bolton G, Fraser SE, Gharib M. Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis. Nature 2003; 421:172–177. 53. Isogai S, Lawson ND, Torrealday S, Horiguchi M, Weinstein BM. Angiogenic network formation in the developing vertebrate trunk. Development 2003; 130:5281–5290. 54. Boardman KC, Swartz MA. Interstitial flow as a guide for lymphangiogenesis. Circ Res 2003; 92:801–808. 55. Virchow R. Der ateromatose Prozess der Arterien. Wien Med Wochenshr 1856; 6: 825–841. 56. Gimbrone MA, Jr, Topper JN. Biology of the vessel wall: endothelium. In: Chien KR, ed. Molecular Basis of Heart Diseases. Troy, MO: Harcourt Brace & Co., 1998:331–348. 57. Cornhill JF, Roach MR. A quantitative study of the localization of atherosclerotic lesions in the rabbit aorta. Atherosclerosis 1976; 23:489–501. 58. Glagov S, Zarins C, Giddens DP, Ku DN. Hemodynamics and atherosclerosis. Insights and perspectives gained from studies of human arteries. Arch Pathol Lab Med 1988; 112:1018– 1031. 59. Topper JN, Cai J, Falb D, Gimbrone MA, Jr. Identification of vascular endothelial genes differentially responsive to fluid mechanical stimuli: cyclooxygenase-2, manganese super-oxide dismutase, and endothelial cell nitric oxide synthase are selectively up-regulated by steady laminar shear stress. Proc Natl Acad Sci USA 1996; 93:10417–10422. 60. Passerini AG, Polacek DC, Shi C, et al. Coexisting proinflammatory and antioxidative endothelial transcription profiles in a disturbed flow region of the adult porcine aorta. Proc Natl Acad Sci USA 2004; 101:2482–2487. 61. Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease: pathogenesis, predisposition, and prevention. Circulation 1998; 97:916–931. 62. Xu C, Lee S, Singh TM, et al. Molecular mechanisms of aortic wall remodeling in response to hypertension. J Vasc Surg 2001; 33:570–578. 63. Westerband A, Crouse D, Richter LC, et al. Vein adaptation to arterialization in an experimental model. J Vasc Surg 2001; 33:561–569. 64. Wallner K, Li C, Fishbein MC, Shah PK, Sharifi BG. Arterialization of human vein grafts is associated with tenascin-C expression. J Am Coll Cardiol 1999; 34:871–875. 65. Sho E, Sho M, Singh TM, et al. Arterial enlargement in response to high flow requires early expression of matrix metalloproteinases to degrade extracellular matrix. Exp Mol Pathol 2002; 73:142–153. 66. Rudic RD, Bucci M, Fulton D, Segal SS, Sessa WC. Temporal events underlying arterial remodeling after chronic flow reduction in mice: correlation of structural changes with a deficit in basal nitric oxide synthesis. Circ Res 2000; 86:1160–1166.

Fluid mechanical forces as extrinsic modifiers

227

67. Kwei S, Stavrakis G, Takahas M, et al. Early adaptive responses of the vascular wall during venous arterialization in mice. Am J Pathol 2004; 164:81–89.

11 Vascular Bed-Specific Signaling and Angiogenesis Napoleone Ferrara, Rui Lin, and Jennifer LeCouter Department of Molecular Oncology, Genentech Inc., South San Francisco, California, U.S.A.

1. INTRODUCTION The cardiovascular system is the first organ system to develop and reach a functional state in an embryo (1). “Vasculogenesis,” the in situ differentiation of endothelial cell precursors, and “angiogenesis,” the growth from the endothelium of existing vessels cooperate in achieving the mature vasculature (2). Recent studies have complemented this view, suggesting that incorporation of bone marrow-derived endothelial progenitor cells (EPC) into the growing vessel contributes to normal and abnormal angiogenesis (3– 5). Blood vessel growth is also implicated in the pathogenesis of a variety of proliferative disorders, including tumors, intraocular neovascular syndromes, rheumatoid arthritis, and psoriasis (6–8). The endothelial cells that comprise the vascular beds of specific tissues display unique phenotypes, growth properties, and functions (9). Recent studies have shown that endothelial cells induce differentiation of liver and pancreas (10,11), suggesting that gut endothelium may have unique paracrine properties, independent on their role in providing a blood supply. This diversity extends also to the vasculature of tumors. Some of the earliest pioneering work in the field of tumor angiogenesis resulted in the seminal observation that the tumor vascular architecture is different depending on the tumor type, leading to the far-reaching hypothesis that the microenvironment has a major influence on the growth and morphologic characteristics of tumor vessels (12). Although distinct endothelial features have been noted to reflect functional requirements imposed by the tissue context, the mechanisms that determine these properties have not been well defined. In vivo (13,14) and ex vivo (15) experiments that have assessed endothelial cell phenotypic and functional development have indicated the existence of endothelial cell-extrinsic signals. These may include secreted peptides, extracellular matrix, or cell membrane components. Recent work has assessed in vivo selective binding of peptides displayed on phage, or assignment of so-called “vascular address” (16). In the adult, angiogenesis is restricted to and required for reproductive function and wound healing. Several angiogenic factors with otherwise pleiotropic

Vascular bed-specific signaling

229

activities have been reported including angiogenin, hepatocyte growth factor (HGF), acidic and basic fibroblast growth factor (FGF), and interleukin (IL)-8 (17–20). The requirements for the endothelial cell specific, angio-genic factors, vascular endothelial growth factor (VEGF) (21) and the angiopoietins Ang-1 (22) and Ang-2 (23) and their cognate receptors, were demonstrated in a series of genetic studies in mouse (24,25). Although VEGF and the angiopoietins are largely selective for endothelial cells, they are widely expressed (20,26). Therefore, it has been difficult to reconcile endothelial cell phenotypic diversity with the action of these ubiquitous factors. However, recent studies have provided evidence for certain vascular bed-specific response to VEGF and other angiogenic factors such as bFGF. The physiological properties of vessels induced by these factors were mainly determined by the micro-environment (14). Furthermore, a recent study has shown that liver sinusoidal endothelial cells strongly up-regulate HGF in response to VEGFR-1 agonists, while this response is not observed in other endothelial cell types (27). Therefore, it is conceivable that morphological and functional diversity among endothelia is achieved by several mechanisms, including vascular bed-specific response to ubiquitous mediators and the existence of unique mitogens/differentiation factors with a tissue-restricted expression pattern.

2. EG-VEGF EG-VEGF was identified as a novel human endothelial cell mitogen, through a bioassay assessing the ability of a library of purified human secreted proteins to promote the growth of primary bovine adrenal cortex capillary endothelial (ACE) cells (28). EGVEGF does not belong to the VEGF family or other known families of endothelial mitogens but instead is a member of a structurally related class of peptides including the digestive enzyme colipase, the Xenopus head-organizer, dickkopf (29), venom protein A (VPRA) (30) or Mamba intestinal toxin-1, “MIT-1” (31), a nontoxic component of Dendroaspis polylepis polylepis venom, and the secreted proteins from Bombina variegata designated Bv8 (32). The distinguishing structural motif is 10 cysteine residues that form five disulfide bridges within a conserved span, designated a colipase-fold (33). EG-VEGF (80% homologous to VPRA) and VPRA are most closely related (83% and 79% homology, respectively) to the Bv8 peptide. Interestingly, Bv8 and EG-VEGF share the first four amino acids of the mature protein, the sequence “AVIT” that, therefore, appears to be distinctive feature of this protein family (34). Mouse and human orthologues of Bv8, also known as prokineticin-2 (35), have been recently described. Both VPRA/MIT-1 and Bv8 were shown to induce gastrointestinal motility (31,32). More recently, the human orthologues of these proteins were shown to have the same motility-enhancing activity and the denomination of “prokineticins” was proposed to designate this property (35).

Endothelial cells in health and disease

230

3. EFFECTS OF EG-VEGF ON THE VASCULAR ENDOTHELIUM Significantly, EG-VEGF selectively promoted proliferation, survival, and chemotaxis of endothelial cells isolated from steroidogenic tissues. The mitogenic and prosurvival activities of EG-VEGF correlate with the ability of this peptide to induce phosphorylation of the mitogen activated protein kinases, ERK1/2, and the Akt serine/threonine kinase of the PI3K survival pathway (36). Nitric oxide production influences vascular tone and permeability (37). Endothelial nitric oxide synthetase (eNOS) is a downstream effector of Akt (37). Lin et al. (36) demonstrated the sequential phosphorylation of Akt and eNOS in EG-VEGF-treated ACE cultures, indicating that EG-VEGF may influence vessel permeability in vivo. Indeed, exogenous EG-VEGF in the ovary (28) or testis (38) can dramatically affect vascular leakage. A specialized ultrastructural feature of the endothelial cells of certain capillary beds (e.g., liver sinusoids, and most endocrine glands) is the presence of small plasma membrane discontinuities, referred to as fenestrae (39). These 63–68 nm openings facilitate fluid and solute exchange. Like VEGF (40,41), EG-VEGF can induce the formation of fenestrae in ACE cells, and may in part be an important mediator of this endothelial phenotype in select sites, namely the steroidogenic tissues. In vivo, delivery of adenovirus encoding EG-VEGF resulted in tissue-specific angiogenesis. Although no response was elicited by the administration of EG-VEGF in the skin or skeletal muscle, a potent angiogenic response was apparent within the ovary (28). These data confirmed the tissue-specificity of EG-VEGF response and supported the existence of endothelial-specific receptors with a restricted capillary bed expression.

4. EG-VEGF G-PROTEIN COUPLED RECEPTORS Biochemical studies of ACE cells pretreated with pertussin toxin indicated that the EGVEGF interacts with a Gai-coupled type receptor (36). Two small, highly identical 80– 90% G-protein coupled receptors of the neuropeptide Y (NPY) receptor class have been identified as the cognate receptors for EG-VEGF and the related peptide, Bv8 (42,43). In this context, it is noteworthy that NPY has both neurotropic and angiogenic activities (44,45). These receptors were designated EG-VEGFR-1/ZAQ and EG-VEGFR2/GPCR73 (43) or prokineticin receptor-1 (PKR-1) and PKR-2 (42). Transcripts, and presumably both proteins, for EG-VEGFR-1 and EG-VEGFR-2 are expressed in ACE cultures (43). Importantly, we demonstrated restricted expression of these receptors in endothelial cells residing in the testis interstitial tissue (38). Therefore, similar to VEGF and its receptors, the EG-VEGF/EG-VEGFR system represents a paracrine system, in which the ligand is produced by nonendothelial cells, and the receptors are selectively expressed in the vascular endothelium.

Vascular bed-specific signaling

231

5. EXPRESSION OF HUMAN EG-VEGF IS PREDOMINANTLY IN STEROIDOGENIC GLANDS Northern blot analysis of a panel of RNAs from a variety of human tissues revealed EGVEGF expression in ovary, testis, adrenal, and placenta, while a consideraly lower signal was noted in a nonsteroidogenic organ like the prostate (28). No significant hybridization signal was detected in other organs by this technique, indicating that steroidogenic glands are the predominant site of EG-VEGF expression (28). In agreement with this conclusion, Zhang et al. have recently reported that, while a very low-abundance signal can be detected in some organs using the highly sensitive Taqman analysis, a dramatically higher EG-VEGF expression signal is measured in steroidogenic organs (46). Among these, the ovary and testis expresses the highest level of the EG-VEGF transcript. In situ hybridization analysis demonstrated that steroidogenic cells within these glands are the source of EG-VEGF (28). Within the testis, the hybridization signal was essentially restricted to the testosterone-producing Leydig cells. The ovary is noted for its dynamic, cyclical growth that is accompanied by a high rate of angiogenesis and thus it has been a major focus for our analyses.

6. DIFFERENTIAL EXPRESSION OF EGVEGF AND VEGF IN THE NORMAL HUMAN OVARY SUGGESTS COMPLEMENTARY FUNCTIONS IN FOLLICULAR AND LUTEAL ANGIOGENESIS We initially reported the expression of EG-VEGF in the ovarian stroma and in the early follicles, restricted to cells of the theca interna (28). Recently, we expanded our earlier analysis of EG-VEGF expression in human and primate ovarian follicles to include a wider range of human preovulatory and atretic follicular stages, and a range of corpus luteum (CL) stages (47). Expression of VEGF and EG-VEGF mRNA was detected by in situ hybridization in all of the specimens examined. Granulosa cells in primordial and primary follicles express EG-VEGF strongly, whereas VEGF expression is very weak or undetectable. VEGF expression is more uniformly detectable but still weak in secondary follicles with 2–3 layers of granulosa cells. As preovulatory follicles mature, VEGF expression appears to progressively increase, so that antral follicles show intense granulosa cell signal that is often associated with moderate or weak VEGF expression in the adjacent thecal layers. As the secondary follicle matures, EG-VEGF expression in granulosa cells declines. Atretic follicles at different stages of their evolution strongly expresses EG-VEGF in the residual thecal cells surrounding the dense hyaline remnant of the follicular basal lamina. VEGF is only weakly expressed in a subset of these cells immediately adjacent to the follicular basal lamina.

Endothelial cells in health and disease

232

These studies have revealed that the generally complementary expression pattern of EG-VEGF and VEGF also extends to the luteal phase (47). The CL is a hormoneregulated, transient endocrine gland that develops following ovulation and produces progesterone, required for the maintenance of early pregnancy. Much evidence indicates that the VEGF transcript is highly expressed, especially in the early CL (48–51). Administration of VEGF inhibitors early in the luteal phase dramatically suppress luteal angiogenesis (52–54). Such treatment may also delay follicular development (55) in rodents and primates. However, it is debated whether VEGF is highly expressed by midearly-late stage and some studies have pointed out that the VEGF signal is significantly reduced by mid-stage (56). VEGF immunoneutralization studies performed in primates at mid-stage have shown a significant reduction in progesterone levels, but the magnitude of the reduction was considerably smaller than that induced when the anti-VEGF treatment was initiated early in the luteal phase (57). Furthermore, VEGF blockade during early pregnancy in the marmoset reduced progesterone levels, but had little effect on pregnancy rates (58). These findings suggest that, while VEGF-dependent angiogenesis is a critical rate-limiting step for the development of an early capillary plexus, later events may be less dependent on VEGF, raising the possibility that additional factors are implicated. CL derived from ovulatory follicles mature in a canonical 14-day pattern (59). We examined EG-VEGF and VEGF expression in a series of CL representing time points approximately 2–14 days postovulation. At approximately 2–3 days postovulation (time points are inferred, according to the histological criteria of Corner (59)) the EG-VEGF and VEGF expression resembles the pattern seen in the late preovulatory follicle: granulosa cells are intensely VEGF-positive, but lack significant EG-VEGF expression. At approximately 5 days postovulation, both VEGF and EG-VEGF are strongly expressed in a portion of granulosa lutein cells (theca lutein cells are not clearly distinct histologically at this stage; they may also express EG-VEGF and VEGF). At approximately 8 days postovulation, EG-VEGF expression is intense in the theca lutein cells, while VEGF expression has diminished to the point where only weak signal remains in the peripheral thecal cells. The functional significance of “granulosa” versus “theca” lutein cells is still object of debate and may involve compartimentalization of steroidogenic activities (60). However, the finding that VEGF is predominantly associated with granulosa and EG-VEGF with theca lutein cells, suggests an even greater level of specialization than previously realized as well as a differential contribution, both spatially and temporally, to vascular remodeling events in the CL. In agreement with previous findings, peak VEGF expression was found at early-stage, associated with the initial development of a capillary plexus within the human CL. However, EG-VEGF expression was not detectable until early-mid stage CL but persisted throughout the mid luteal phase, at a time when VEGF expression was much reduced or undetectable (47). Taken together, these observations support the notion that VEGF activity is rate-limiting for the creation of the capillary plexus within the CL. Additionally, EG-VEGF may stimulate the angiogenesis that accompanies early-mid CL development and be especially important for the formation of a more mature vascular bed that includes arterioles and thus for the persistence and adequacy of luteal function (47). These hypotheses can be tested directly by assessing follicular, and CL, development and function in primates treated with selective EG-VEGF inhibitors.

Vascular bed-specific signaling

233

As previously noted, EG-VEGF expression is consistently detected in the ovarian stroma (28). Furthermore, a particularly high expression of EG-VEGF (but not VEGF) mRNA was demonstrated in “hilus” cells (47). These cells are thought to be the functional equivalent of Leydig cells in the ovary (61) and hyperplastic or neoplastic changes affecting them are known to result in a masculinizing syndrome (62,63). These findings corroborate the association of EG-VEGF mRNA expression with a steroidogenic phenotype. Also, the intimate relationship of hilus cells with blood vessels and nerve terminals was noted even in the earliest studies (62,63). Intriguingly, as noted later in this chapter, the EG-VEGF homologue Bv8 has been shown to have neurotrophic (64) and neuromodulator (65) functions. While Bv8 mRNA is undetectable in the human ovary, it is tempting to speculate that EG-VEGF may play both an angiotrophic and neurotrophic role in this context.

7. EXPRESSION OF EG-VEGF IN PCOS AND POTENTIAL ROLE IN ANGIOGENESIS ASSOCIATED WITH CHRONIC ANOVULATION Angiogenesis is also a prominent feature of the polycystic ovary syndrome (PCOS), a leading cause of infertility affecting as many as 5–10% of women of reproductive age. PCOS was originally described as a disorder characterized by the association of hirsutism, obesity, reduced fertility, and enlarged, polycystic, ovaries (66). Hyperplasia of the theca interna and stroma, with excessive production of androgens, are hallmarks of PCOS (for review see Ref. 67). Indeed, the ultrasonographic assessment of stromal area (68) and blood flow (69) is currently used as diagnostic test. Although PCOS was described over 50 years ago, its etiology has remained largely unclear. However, increased LH/FSH ratio, defective selection of a dominant follicle, and anovulation are considered to be key aspects of the pathogenesis. Recent evidence also indicates that PCOS is a part of a complex endocrine/metabolic disorder, where insulin resistance plays a major role (70). Both VEGF and EG-VEGF are expressed in all PCOS ovaries examined, but with an almost mutually exclusive expression pattern (47). Somewhat surprisingly, expression of VEGF mRNA is largely limited to the cyst walls, with little or no expression in the stroma. Cysts appear to express only VEGF, only EG-VEGF, or VEGF in an inner rim surrounded by an outer rim of EG-VEGF. Some VEGF expression was seen in theca interna, although not as consistently as in granulosa cells. EG-VEGF expression was strongest in theca interna of follicles in various stages of atresia. As above noted, intense signal, albeit of lower magnitude than that detected in the theca, occurs in the stroma. Importantly, thecal and stromal tissue expressing EG-VEGF maintain an abundant vascular supply, despite lacking significant VEGF expression. Endothelial immunostaining with anti-CD34 demonstrates persistent vascularity in these areas. Such a pattern is consistent with the establishment of a proangiogenic gradient directing new vessel growth toward the EG-VEGF expressing cells. Therefore, at least in terms of mRNA expression, EG-VEGF shows a particularly strong correlation with vascularity in

Endothelial cells in health and disease

234

PCOS specimens. These findings raise the possibility that, while VEGF is an essential player in normal cycling ovaries, EG-VEGF might be particularly significant for the acyclical angiogenesis occurring during chronic anovulation. Additional studies are clearly needed to verify this hypothesis.

8. REGULATION OF HUMAN EG-VEGF GENE EXPRESSION Initially, we characterized the hypoxic regulation of human EG-VEGF gene (28). Adrenal carcinoma cell line exposed to low oxygen displays a strong up-regulation of both VEGF and EG-VEGF genes (28). This EG-VEGF response is potentially mediated by HIF-1 (71). In agreement with this hypothesis, several putative hypoxia response elements are present within a 4kb region in the EG-VEGF promoter. Importantly, hypoxia-induced angiogenesis is primarily an adaptive physiological response to an increase in metabolic demands associated with proliferative processes, including those occurring during the cyclical ovarian changes (72). Transcriptional regulation of the EG-VEGF gene is also of particular interest, given the selective expression of EG-VEGF mRNA within steroidogenic cells. Pro minently, a consensus binding site for the NR5A1 orphan nuclear receptor is present within the human EG-VEGF promoter (73). NR5A1 is considered to be a key regulator of endocrine development and function. It regulates multiple target genes involved in gonadal and adrenal determination and development; steroidogenesis; and reproduction (for review see Ref. 74). Gene-targeting studies revealed a critical role for NR5A1 in adrenal and gonad development (75,76). In the adult ovary, NR5A1 protein expression is recognized at the onset of follicular development and is strongly induced in antral follicles in theca and granulosa cells (77). The expression profiles of NR5A1 and the highly related NR5A2, another orphan nuclear receptor, within the ovarian follicles and CL are an intriguing aspect of target gene regulation. Although NR5A1 expression is extinguished at the onset of the conversion to CL (77), NR5A2 is induced in the mid-early stage (78) and may target common genes. The parallels between the known expression and activities of NR5A1/2 and the expression profile of human EG-VEGF indicate that these factors are potentially important regulators of EG-VEGF transcription and provide a plausible explanation for the restricted expression pattern of EG-VEGF. Since EG-VEGF expression is not associated with ovarian epithelial cells, which give rise to the majority of malignancies, this factor is not expected to contribute to initial growth and progression of ovarian carcinomas. In agreement with this hypothesis, Zhang et al. (46) recently reported that EG-VEGF mRNA is not detectable in ovarian surface epithelium and in ovarian epithelial malignancies. However, EG-VEGF may potentially be a mediator of angiogenesis in ovarian or testis tumors that arise from steroidogenic cells, such as thecomas, granulosa cell or Leydig cell tumors.

Vascular bed-specific signaling

235

9. THE EG-VEGF MOUSE ORTHOLOGUE: A DIFFERENT EXPRESSION PATTERN BUT RELATED FUNCTIONS To further characterize the function and biology of EG-VEGF, the mouse orthologue was cloned and its expression pattern and activity examined (73). The predicted mature mouse EG-VEGF (mEG-VEGF) peptide is 88% identical to the human sequence. mEGVEGF maps to a region of chromosome 3 syntenic with human chromosome 1p13.1, the locus for hEG-VEGF, providing further evidence that these genes are orthologues. The gene organization is also highly conserved; however, the promoter sequences have diverged reflecting unique transcriptional regulation. For example, the putative NR5A1 consensus site present in the human is absent from the mEG-VEGF promoter sequence (73). Apparently, the divergence of promoter sequence between human and mouse accounts for the selective and unique expression patterns. Strikingly, mEG-VEGF is expressed in a distinct expression pattern from its human counterpart. The transcript is predominantly restricted to the liver and kidney. In the adult kidney, the specific hybridization signal is restricted to the epithelial tubule cells. The fetal liver expresses a very high level of EG-VEGF transcript, with signal restricted to hepatocytes. The EG-VEGF mRNA level is reduced in the adult liver. Interestingly, the sinusoidal endothelial cells within the liver (79) and peritubular capillary plexuses in the kidney (80) are similar to those of the endocrine glands in that they are fenestrated. Hence, similar to human EG-VEGF, mouse EG-VEGF may influence the phenotype and growth properties of endothelial cells in distinct tissue compartments enriched in fenestrated endothelium. The expression of EG-VEGFR-2 was associated with purified liver endothelial cells or the endothelial cell-enriched fraction derived from kidney. EGVEGFR-1 expression was undetectable in these samples. Conversely, the transcript for the ligand was detected in RNA samples derived from the hepatocyte fraction, or the endothelial cell-depleted kidney fraction. mEG-VEGF expression was virtually undetectable in purified endothelial cells. These findings are consistent with the hypothesis that mEG-VEGF functions as a paracrine growth and survival factor for liver and kidney endothelial cells. Recombinant mEG-VEGF peptide stimulated the growth and survival of primary mouse liver sinusoidal endothelial cells (73). Therefore, although mEG-VEGF is expressed at distinct tissue sites relative to the human orthologue, the biological activities appear to be analogous, due to the selective coexpression of an EG-VEGF receptor.

10. Bv8 IS A CLOSELY RELATED EG-VEGF HOMOLOGUE WITH A DISTINCT EXPRESSION PATTERN As above noted, EG-VEGF is highly related to a peptide that was designated Bv8. This molecule was originally purified from the skin secretions of the yellow-bellied toad, B. variegata. This amphibian peptide was initially characterized as an inducer of

Endothelial cells in health and disease

236

gastrointestinal motility and hyperalgesia in the rat (32). The function of Bv8 as a neuropeptide was further evaluated with the mouse peptide. In addition to localizing mouse Bv8 transcript in the CNS, Melchiorri et al. (64) reported that Bv8 stimulated neuronal survival in cultures of primary granular cells. Negri et al. (81) have also described the ability of the Bv8 peptide to induce nociceptive sensitization. Also, the denomination of “prokineticin-2” has also been used for its human orthologue (35). This protein has been proposed to represent an output signal from the suprachiasmatic nucleus (SCN) that regulates circadian rhythm activity. The Bv8 expression pattern within the SCN is rhythmic, and Bv8 transcription is perturbed in circadian mutant mouse lines (65). Exogenous Bv8 peptide delivered into the rat brain in the dark phase inhibited the normal nocturnal locomotor activity. Importantly, endogenous Bv8 levels within the SCN are lowest in the active, dark phase (65). The primary structures of predicted mouse and human Bv8 isoforms have been described (35,38,82). Several putative hypoxia regulatory elements exist in the Bv8 promoter regions, and similar to EG-VEGF (28), Bv8 transcript is induced by hypoxic treatment in cell culture (38). Overall, the Bv8 human and mouse promoter sequences are highly conserved, indicating a related transcriptional regulation in these species. Purified Bv8, like EG-VEGF, induced proliferation, migration, and survival of ACE cells (38). As previously mentioned, these activities are attributed to the ability of both Bv8 and EGVEGF to bind and activate the same G-protein coupled receptors (43,42). The predominant site of Bv8 expression in both human and mouse is the seminiferous tubules of the testis (38,82). Within the tubules, the Bv8 transcript is largely restricted to the primary spermatocytes. Bv8 mRNA is not detected at this tissue site prior to spermatocytic development. Within the testis, the EG-VEGF/Bv8 receptors are expressed in vascular endothelial cells within the interstitial space (38). Delivery of exogenous Bv8, EG-VEGF, or VEGF to the testis via the direct injection of recombinant adenoviruses resulted in a potent, indistinguishable angiogenic response. Taken together, these results suggested that Bv8 or EG-VEGF function as angiogenic mitogens/survival factors in the testis. Interestingly, the testis exhibits the highest endothelial cell turnover among the noncyclic tissues (83). Additionally, Bv8 and EG-VEGF might have functions unrelated to angiogenesis within the male reproductive tract. In this context, it is noteworthy that overexpression of VEGF in the mouse testis resulted in an arrest of spermatogenesis (84). It remains to be established whether disregulation of Bv8/EG-VEGF may also contribute to the pathophysiology of male infertility. The availability of specific antibodies and the Bv8 gene-targeted mouse should permit further studies of the molecular functi on in the testis and during sperm maturation and fertilization.

11. CONCLUSION The identification of EG-VEGF suggests a novel view of the regulation of angiogenesis. Steroidogenic glands appear to have developed highly specific local mechanisms, apparently to complement the action of the ubiquitous VEGF. Interestingly, such an acquisition seems to be, at least in part, a late event in evolution and may reflect a greater functional/morphological complexity of organs like the ovary. In this context, the species

Vascular bed-specific signaling

237

differences in the expression pattern of human and mouse EG-VEGF are particularly intriguing. The association of human EG-VEGF expression with steroidogenic cells is compelling. However, the mouse orthologue has a different expression pattern. Although rodents have served as models for endocrinology and ovarian physiology, clear differences exist between the rodent and human ovary. A fundamental difference is the selection and development of a single ovarian follicle in humans and other monovular species. The process of selection of the dominant follicle has been associated with angiogenesis, as there is evidence that selected follicles possess a more elaborate microvascular network (85). The length of the ovarian or luteal cycle also distinguishes the human or primate from the rodent. In humans, the cycle is 28 days and in rodents the cycle is completed every 4 days (86). The primate CL is functional for 2 weeks prior to its regression in the infertile cycle, whereas the rodent CL is active for less than a day (87). Both the selection of the dominant follicle and the length of the cycle impose distinct regulatory functions in the human (primate system), including a more complex regulation of growth and maintenance of the vascular endothelium. It is tempting to speculate that EG-VEGF represents one growth factor component that contributes to these evolutionary changes. While at an earlier point in evolution EG-VEGF performs analogous function but at distant sites, this gene appears to have been “coopted,” by virtue of a tissue-specific transcriptional regulation, into the ovary, coincident with a greater length and complexity of the ovarian cycle. In this context, it will be interesting to determine when, on an evolutionary scale, EG-VEGF expression first became associated with the ovary and other steroidogenic tissues. Furthermore, as previously observed, human EG-VEGF is highly expressed in PCOS (47). This disorder occurs in humans but not in rodents. While a focus of this chapter has been the potential role of EG-VEGF in the cyclic ovarian angiogenesis, this molecule might play an important role in the pathophysiology of other steroidogenic organs such as adrenal and placenta, which remains to be investigated. The existence of restricted, local mediators of angiogenesis and endothelial cell survival and function permit novel approaches to salvage or regenerate tissue. Several attempts have been made to promote new vessel growth using angiogenic factors such as VEGF and bFGF. However, the success of these efforts have been limited by systemic side-effects, including hypotension, edema, and accelerated atherosclerosis. Moreover, these efforts have generally not achieved functional and mature blood vessels. Mitogens selective for the endothelium of specific tissues like cardiac or skeletal muscle would potentially offer major advantages. A principle benefit of tissue-specific angiogenic therapeutics could be the reduction of undesired side-effects associated with the broadspectrum molecules. Endothelial cell stimulation, even in the absence of angiogenesis, promoted salvage and regeneration of the tissue parenchyma by stimulating the release of paracrine factors (27). Thus, the EG-VEGF/Bv8 family not only has therapeutic implications per se, but also raises the possibility that other selective regulators of angiogenesis within specific tissue or organs exist and therefore presents novel potential therapeutic targets.

Endothelial cells in health and disease

238

REFERENCES 1. Hamilton WJ, Boyd JD, Mossmann HW. Human Embryology. Baltimore: Wiliam & Wilkins, 1962. 2. Risau W, Flamme I. Vasculogenesis. Ann Rev Cell Dev Biol 1995; 11:73–91. 3. Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor endothelial cells for angiogenesis. Science 1997; 275:964–967. 4. Rafii S, Meeus S, Dias S, Hattori K, Heissig B, Shmelkov S, Rafii D, Lyden D. Contribution of marrow-derived progenitors to vascular and cardiac regeneration. Semin Cell Dev Biol 2002; 13:61–67. 5. Gerber HP, Ferrara N. The role of VEGF in normal and neoplastic hematopoiesis. J Mol Med 2003; 81:20–31. 6. Folkman J. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1995; 1:27–31. issn: 1078–8956. 7. Garner A. Vascular diseases. In: Garner A, Klintworth GK, eds. Pathobiology of Ocular Disease. 2nd ed. New York: Marcel Dekker, 1994:1625–1710. 8. Ferrara N, Alitalo K. Clinical applications of angiogenic growth factors and their inhibitors. Nat Med 1999; 5:1359–1364. 9. Simionescu N, Simionescu M. The cardiovascular system. In: Weiss L, ed. Cell and Tissue Biology. Baltimore: Urban & Schwarzemberg, 1988:355–398. 10. Matsumoto K, Yoshitomi H, Rossant J, Zaret KS. Liver organogenesis promoted by endothelial cells prior to vascular function. Science 2001; 294:559–563. 11. Lammert E, Cleaver O, Melton D. Induction of pancreatic differentiation by signals from blood vessels. Science 2001; 294:564–567. 12. Lewis WH. The vascular pattern of tumors. Johns Hopkins Hosp Bull 1927; 41: 156–162. 13. Aird WC, Edelberg JM, Weiler-Guettler H, Simmons WW, Smith TW, Rosenberg RD. Vascular bed-specific expression of an endothelial cell gene is programmed by the tissue microenvironment. J Cell Biol 1997; 138:1117–1124. 14. Dellian M, Witwer BP, Salehi HA, Yuan F, Jain RK. Quantitation and physiological characterization of angiogenic vessels in mice: effect of basic fibroblast growth factor, vascular endothelial growth factor/vascular permeability factor, and host microenvironment [see comments]. Am J Pathol 1996; 149:59–71. 15. Stewart PA, Wiley MJ. Developing nervous tissue induces formation of blood–brain barrier characteristics in invading endothelial cells: a study using quail-chick transplantation chimeras. Dev Biol (Orlando) 1981; 84:183–192. 16. Ruoslahti E, Rajotte D. An address system in the vasculature of normal tissues and tumors. Annu Rev Immunol 2000; 18:813–827. 17. Folkman J, Klagsbrun M. Angiogenic factors. Science 1987; 235:442–447. 18. Klagsbrun M, D’Amore PA. Regulators of angiogenesis. Annu Rev Physiol 1991; 53:217–239. 19. Folkman J, Shing Y. Angiogenesis. J Biol Chem 1992; 267:10931–10934. 20. Yancopoulos GD, Davis S, Gale NW, Rudge JS, SJ, W, J, H. Vascular-specific growth factors and blood vessel formation. Nature 2000; 407:242–248. 21. Ferrara N. VEGF and the quest for tumour angiogenesis factors. Nat Rev Cancer 2002; 2:795– 803. 22. Suri C, Jones PF, Patan S, Bartunkova S, Maisonpierre PC, Davis S, Sato TN, Yancopoulos GD. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis [see comments]. Cell 1996; 87:1171–1180.

Vascular bed-specific signaling

239

23. Maisonpierre PC, Suri C, Jones PF, Bartunkova S, Wiegend SJ, Radziejewski C, Compton D, McClain J, Aldrich TH, Papadopulos N, Daly TJ, Davis S, Sato TN, Yancopoulos GD. Angiopoietin-2, a natural antagonist for Tie-2 that disrupts in vivo angiogenesis. Science 1997; 277:55–60. 24. Ferrara N, Carver Moore K, Chen H, Dowd M, Lu L, O’Shea KS, Powell Braxton L, Hillan KJ, Moore MW. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 1996; 380:439–442. issn: 0028–0836. 25. Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L, Gertsenstein M, Fahrig M, Vandenhoeck A, Harpal K, Eberhardt C, Declercq C, Pawling J, Moons L, Collen D, Risau W, Nagy A. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 1996; 380:435–439 issn: 0028–0836. 26. Ferrara N, Davis-Smyth T. The biology of vascular endothelial growth factor. Endocr Rev 1997; 18:4–25. 27. LeCouter J, Moritz D R, Li B, Phillips GL, Liang X H, Gerber HP, Hillan KJ, Ferrara N. Angiogenesis-independent endothelial protection of liver: role of VEGFR-1. Science 2003; 299:890–893. 28. LeCouter J, Kowalski J, Foster J, Hass P, Zhang Z, Dillard-Telm L, Frantz G, Rangell L, DeGuzman L, Keller G-A, Peale F, Gurney A, Hillan KJ, Ferrara N. Identification of an angiogenic mitogen selective for endocrine gland endothelium. Nature 2001; 412: 877–884. 29. Glinka A, Wu W, Delius H, Monaghan AP, Blumenstock C, Niehrs C. Dickkopf-1 is a member of a new family of secreted proteins and functions in head induction. Nature 1998; 391:357– 362. 30. Joubert FJ, Strydom DJ. Snake venom. The amino acid sequence of protein A from Dendroaspis polylepis polylepis (black mamba) venom. Hoppe-Seylers Zeitschrift fur Physiologische Chemie 1980; 361:1787–1794. 31. Schweitz H, Pacaud P, Diochot S, Moinier D, Ladzunski M. MIT (1), a black mamba intestinal toxin with a new and highly potent activity on intestinal contraction. FEBS Lett 1999; 461:183– 188. 32. Mollay C, Wechselberger C, Mignogna G, Negri L, Melchiorri P, Barra D, Kreil G. Bv8, a small protein from frog skin and its homologue from snake venom induce hyperalgesia in rats. Eur J Pharmacol 1999; 374:189–196. 33. Aravind L, Koonin EV. A colipase fold in the carboxy-terminal domain of the Wnt antagonists—the Dickkopfs. Curr Biol 1998; 8:477–478. 34. Kaser A, Winklmayr M, Lepperdinger G, Kreil G. The AVIT protein family. EMBO Rep 2003; 4:469–473. 35. Li M, Bullock CM, Knauer DJ, Ehlert FJ, Zhou QY. Identification of two Prokineticin cDNAs: recombinant proteins potently contract gestrointestinal smooth muscle. Mol Pharmacol 2001; 59:692–698. 36. Lin R, LeCouter J, Kowalski J, Ferrara N. Characterization of EG-VEGF signaling in adrenal cortex capillary endothelial cells. J Biol Chem 2002; 277:8724–8729. 37. Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher AM. Activation of nitric oxide synthase in endothelial cells by Akt-dependent phosphorylation. Nature 1999; 399:601– 605. 38. LeCouter J, Lin R, Frantz G, Tejada M, Peale F, Hillan KJ, Ferrara N. The EG-VEGF homologue, Bv8, is a potent angiogenic factor. Localization of Bv8 receptors to endothelial cells. Proc Natl Acad Sci USA 2003; 100:2685–2690. 39. Palade GE, Simionescu M, Simionescu N. Structural aspects of the permeability of the microvascular endothelium. Acta Physiol Scand 1979; 463(suppl):11–32. 40. Esser S, Wolburg K, Wolburg H, Breier G, Kurzchalia T, Risau W. Vascular endothelial growth factor induces endothelial fenestrations in vitro. J Cell Biol 1998; 140:947–959. 41. Roberts WG, Palade GE. Neovasculature induced by vascular endothelial growth factor is fenestrated. Cancer Res 1997; 57:765–772.

Endothelial cells in health and disease

240

42. Lin DC, Bullock CM, Ehlert FJ, Chen JL, Tian H, Zhou QY. Identification and molecular characterization of two closely related G protein-coupled receptors activated by prokineticins/endocrine gland vascular endothelial growth factor. J Biol Chem 2002; 277:19276–19280. 43. Masuda Y, Takatsu Y, Terao Y, Kumano S, Ishibashi Y, Suenaga M, Abe M, Fukusumi S, Watanabe T, Shintani Y, Yamada T, Hinuma S, Inatomi N, Ohtaki T, Onda H, Fujino M. Isolation and identification of EG-VEGF/prokineticins as cognate ligands for two orphan Gprotein-coupled receptors. Biochem Biophys Res Commun 2002; 293:396–402. 44. Zukowska-Grojec Z, Karwatowska-Prokopczuk E, Rose W, Rone J, Movafagh S, Ji H, Yeh Y, Chen WT, Kleinman HK, Grouzmann E, Grant DS. Neuropeptide Y: a novel angiogenic factor from the sympathetic nerves and endothelium. Circ Res 1998; 83:187–195. 45. Lee EW, Michalkiewicz M, Kitlinska J, Kalezic I, Switalska H, Yoo P, Sangkharat A, Ji H, Li L, Michalkiewicz T, Ljubisavljevic M, Johansson H, Grant DS, Zukowska Z. Neuropeptide Y induces ischemic angiogenesis and restores function of ischemic skeletal muscles. J Clin Invest 2003; 111:1853–1862. 46. Zhang L, Yang N, Conejo-Garcia JR, Katsaros D., Mohamed-Hadley A, Fracchioli S, Schlienger K, Toll A, B, L, Rubin SC, Coukos G. Expression of endocrine gland-derived vascular endothelial growth factor in ovarian carcinoma. Clin Cancer Res 2003; 9:264–272. 47. Ferrara N, Frantz G, LeCouter J, Dillart-Telm L, Pham T, Draksharapu A, Giordano T, Peale F. Differential expression of the angiogenic factor genes VEGF and EG-VEGF in normal and polycystic human ovaries. Am J Pathol 2003; 162:1881–1893. 48. Phillips HS, Hains J, Leung DW, Ferrara N. Vascular endothelial growth factor is expressed in rat corpus luteum. Endocrinology 1990; 127:965–967. 49. Kamat BR, Brown LF, Manseau EJ, Senger DR, Dvorak HF. Expression of vascular permeability factor/vascular endothelial growth factor by human granulosa and theca lutein cells. Role in corpus luteum development. Am J Pathol 1995; 146:157–165. 50. Ravindranath N, Little-Ihrig L, Phillips HS, Ferrara N, Zeleznik AJ. Vascular endothelial growth factor messenger ribonucleic acid expression in the primate ovary. Endocrinology 1992; 131:254–260. 51. Goede V, Schmidt T, Kimmina S, Kozian D, Augustin HG. Analysis of blood vessel maturation processes during cyclic ovarian angiogenesis. Lab Invest 1998; 78:1385–1394. 52. Ferrara N, Chen H, Davis-Smyth T, Gerber H-P, Nguyen T-N, Peers D, Chisholm V, Hillan KJ, Schwall RH. Vascular endothelial growth factor is essential for corpus luteum angiogenesis. Nat Med 1998; 4:336–340. 53. Ryan AM, Eppler DB, Hagler KE, Bruner RH, Thomford PJ, Hall RL, Shopp GM, O’Neill CA. Preclinical safety evaluation of rhuMAbVEGF, an antiangiogenic humanized monoclonal antibody. Toxicol Pathol 1999; 27:78–86. 54. Fraser HM, Dickson SE, Lunn SF, Wulff C, Morris KD, Carroll VA, Bicknell R. Suppression of luteal angiogenesis in the primate after neutralization of vascular endothelial growth factor. Endocrinology 2000; 141:995–1000. 55. Zimmermann RC, Xiao E, Husami N, Sauer MV, Lobo R, Kitajewski J, Ferin M. Short-term administration of antivascular endothelial growth factor antibody in the late follicular phase delays follicular development in the rhesus monkey. J Clin Endocrinol Metab 2001; 86:768– 772. 56. Yamamoto S, Konishi I, Tsuruta Y, Nanbu K, Mandai M, Kuroda H, Matsushita K, Hamid AA, Yura Y, Mori T. Expression of vascular endothelial growth factor (VEGF) during folliculogenesis and corpus luteum formation in the human ovary. Gynecolo Endocrinol 1997; 11:371–381. 57. Dickson SE, Bicknell R, Fraser HM. Mid-luteal angiogenesis and function in the primate is dependent on vascular endothelial growth factor. J Endocrinol 2001; 168:409–416.

Vascular bed-specific signaling

241

58. Rowe AJ, Morris KD, Bicknell R, Fraser HM. Angiogenesis in the corpus luteum of early pregnancy in the marmoset and the effects of vascular endothelial growth factor immunoneutralization on establishment of pregnancy. Biol Reprod 2002; 67:1180–1188. 59. Corner GWJ. The histological dating of the human corpus luteum of menstruation. Am J Anat 1956; 9:377–401. 60. Stouffer RL. Corpus luteum formation and demise. In: Adashi EY, Rock JA, Rosenwaks Z, eds. Reproductive Endocrinology, Surgery, and Technology. Philadelphia: Lippincott-Raven, 1996. 61. Sternberg SS. ed. Histology for Pathologists. 2nd ed. Philadelphia: Lippincott, 1997. 62. Sternberg WH. The morphology, androgenic function, hyperplasia and tumors of the ovarian hilus cells. Am J Pathol 1949; 25:493–521. 63. Merrill JA. Ovarian hilus cells. Am J Obstet Gynecol 1959; 78:1258–1271. 64. Melchiorri D, Bruno V, Besong G, Ngomba RT, Cuomo L, De Blasi A, Copani A, Moschella C, Storto M, Nicoletti F, Lepperdinger G, Passarelli F. The mammalian homologue of the novel peptide Bv8 is expressed in the central nervous system and supports neuronal survival by activating the MAP kinase/PI-3-kinase pathways. Eur J Neurosci 2001; 13:1694–1702. 65. Cheng MY, Bullock CM, Li C, Lee AG, Bermak JC, Belluzzi J, Weaver DR, Leslie FM, Zhou QY. Prokineticin 2 transmits the behavioural circadian rhythm of the suprachiasmatic nucleus. Nature 2002; 417:405–410. 66. Goldziher JW, Green JA. The polycistic ovary. I. Clinical and histologic features. J Clin Endocrinol Metab 1962; 22:325–332. 67. Yen SSC. A contemporary overview. In: Adashi EY, Rock JA, Rosenwaks Z, eds. Reproductive Endocrinology, Surgery, and Technology. Philadelphia: LippincottRaven, 1996:1117–1126. 68. Dewailly D, Robert Y, Helin I, Ardaens Y, Thomas-Desrousseaux P, Lemaitre L, Fossati P. Ovarian stromal hypertrophy in hyperandrogenic women. Clin Endocrinol (Oxf) 1994; 41:557– 562. 69. Fulghesu AM, Ciampelli M, Belosi C, Apa R, Pavone V, Lanzone A. A new ultrasound criterion for the diagnosis of polycystic ovary syndrome: the ovarian stroma/total area ratio. Fertil Steril 2001; 76:326–331. 70. Dunaif A, Thomas A. Current concepts in the polycystic ovary syndrome. Annu Rev Med 2001; 52:401–419. 71. Semenza G. Signal transduction to hypoxia-inducible factor 1. Biochem Pharmacol 2002; 64:993–998. 72. Dor Y, Porat R, Keshet E. Vascular endothelial growth factor and vascular adjustments to perturbations in oxygen homeostasis. Am J Physiol 2001; 280:C1367–1374. 73. LeCouter J, Lin R, Frantz G, Zhang Z, Hillan K, Ferrara N. Mouse endocrine-gland derived vascular endothelial growth factor: a distinct expression pattern from its human ortholog suggests different roles as a regulator of organ-specific angiogenesis. Endocrinology 2003; 144:2606–2616. 74. Parker KL, Rice DA, Lala DS, Ikeda Y, Luo X, Wong M, Bakke M, Zhao L, Frigeri C, Hanley NA, Stallings N, Schimmer BP. Steroidogenic factor 1: an essential mediator of endocrine development. Recent Prog Horm Res 2002; 57:19–36. 75. Luo X, Ikeda Y, Parker KL. A cell-specific nuclear receptor is essential for adrenal and gonadal development and sexual differentiation. Cell 1994; 77:481–490. 76. Sadovsky Y, Crawford PA, Woodson KG, Polish JA, Clements MA, Tourtellotte LM, Simburger K, Milbrandt J. Mice deficient in the orphan receptor steroidogenic factor 1 lack adrenal glands and gonads but express P450 side-chain-cleavage enzyme in the placenta and have normal embryonic serum levels of corticosteroids. Proc Natl Acad Sci USA 1995; 92:10939–10943. 77. Takayama K, Sasano H, Fukaya T, Morohashi K, Suzuki T, Tamura M, Costa MJ, Yajima A. Immunohistochemical localization of Ad4-binding protein with correlation to steroidogenic

Endothelial cells in health and disease

242

enzyme expression in cycling human ovaries and sex cord stromal tumors. J Clin Endocrinol Metab 1995; 80:2815–2821. 78. Boerboom D, Pilon N, Behdjani R, Silversides DW, Sirois J. Expression and regulation of transcripts encoding two members of the NR5A nuclear receptor subfamily of orphan nuclear receptors, steroidogenic factor-1 and NR5A2, in equine ovarian cells during the ovulatory process. Endocrinology 2000; 141:4647–4656. 79. Braet F, Wisse E. Structural and functional aspects of liver sinusoidal endothelial cell fenestrae: a review. Comp Hepatol 2002; 1:1. 80. Bulger RE. The urinary system. In: Weiss L, ed. Cell and Tissue Biology. Baltimore: Urban & Schwarzemberg, 1988:817–849. 81. Negri L, Lattanzi R, Giannini E, Metere A, Colucci M, Barra D, Kreil G, Melchiorri P. Nociceptive sensitization by the secretory protein Bv8. Br J Pharmacol 2002; 137: 1147–1154. 82. Wechselberger C, Puglisi R, Engel E, Lepperdinger G, Boitani C, Kreil G. The mammalian homologues of frog Bv8 are mainly expressed in spermatocytes. FEBS Lett 1999; 462:177–181. 83. Collin O, Bergh A. Leydig cells secrete factors which increase vascular permeability and endothelial cell proliferation. Int J Androl 1996; 19:221–228. 84. Korpelainen EI, Karkkainen MJ, Tenhunen A, Lakso M, Rauvala H, Vierula M, Parvinen M, Alitalo K. Overexpression of VEGF in testis and epididymis causes infertility in transgenic mice: evidence for nonendothelial targets for VEGF. J Cell Biol 1998; 143:1705–1712. 85. Zeleznik, AJ, Schuler HM, Reichert LE Jr. Gonadotropin-binding sites in the rhesus monkey ovary: role of the vasculature in the selective distribution of human chorionic gonadotropin to the preovulatory follicle. Endocrinology 1981; 109:356–362. 86. Lasley B, Benirschke K. Strategies of ovarian function of importance to gynecologic investigations. J Soc Gynecol Invest 1994; 1:247–255. 87. Fraser HM, Lunn SF. Angiogenesis and its control in the female reproductive system. Br Med Bull 2000; 56:787–797.

12 Differential Regulation of Endothelial Cell Barrier Function Jeffrey R.Jacobson, Steven M.Dudek, and Joe G.N.Garcia Division of Pulmonary and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A.

1. INTRODUCTION The endothelium serves as a semipermeable barrier separating the circulation from the surrounding interstitium. Its luminal surface is coated with a negatively charged glycocalyx comprised of membrane-bound proteoglycans and glycoproteins. The tight apposition of individual endothelial cells (ECs) with neighboring cells via intercellular junctions acts as a significant determinant of basal endothelial barrier function. Separately, focal adhesions, the integrin-based linkages between the extracellular matrix and the endothelial cytoskeleton, provide strong tethering of the endothelium to the vessel wall, a process that also contributes to barrier integrity. Long thought to be a passive cellular barrier, the endothelium is now recognized to be highly dynamic and responsive to various effectors of both barrier enhancement and disruption. Perturbations of cell-cell and cell-matrix interactions can result in marked effects on barrier integrity and, subsequently, vascular permeability. In this respect, key elements of the cytoskeleton and its regulators have recently been identified and characteriz ed as important determinants of endothelial barrier function. In this chapter, we will discuss the mechanisms of endothelial barrier regulation. We will explore various effectors of endothelial barrier function, both biochemical and biophysical, focusing on differences in EC phenotype along the vascular tree leading to differential barrier regulation and its significance with respect to various disease states.

2. OVERVIEW OF ENDOTHELIAL CELL BARRIER REGULATION Derangements in endothelial cell barrier function are now recognized as a critical determinant of the morbidity and mortality associated with disease states characterized by

Endothelial cells in health and disease

244

inflammation and increased vascular permeability, including sepsis, acute lung injury, and acute respiratory distress syndrome (1).a More precisely, vascular permeability is defined by two general pathways that determine the movement of fluids and solutes from the vascular space to the surrounding environment. The transcellular pathway utilizes a tyrosine kinase-dependent, gp60-mediated transcytotic albumin route, which is thought to be a minor contributor to inflammatory vascular permeability although its regulation and function remain unclear (2,3). Conversely, the paracellular pathway, characterized by the formation of paracellular gaps in response to various inflammatory mediators, is acknowledged as the primary determinant of vascular permeability, and its regulation has been more fully detailed (4). Intercellular adhesion occurs via the participation of several proteins involved in both tight junction and adherens junction complexes. Via engagement of the actin cytoskeleton, these complexes promote mechanical stability and are involved in transduction of extracellular signals into the cell (5). Tight junctions consist of the transmembrane proteins occludins, claudins, and junctional adhesion molecules (JAMs) coupled to cytoplasmic proteins such as the zona occludens (ZO) family (6). Adherens junctions are composed of cadherins which link adjacent cells via homotypic interactions (7). These proteins attach to catenins (α, β, γ) intercellularly, which in turn anchor to the actin cytoskeleton. Importantly, the primary adhesive protein of EC adherens junctions is vascular endothelial (VE) cadherin (8). Although evidence for a significant functional role of tight junctions in paracellular permeability is limited, adherens junctions are recognized as playing a central role in this setting, largely via their association (and disassociation) with the actin cytoskeleton (9,10). Regulation of paracellular gap formation can be thought of as a balance of competing intracellular contractile forces and adhesive cell-cell and cell-matrix tethering forces (Fig. 1). These forces are regulated through dynamic activation of the actin-based cytoskeleton, the complexities of which have only recently been brought to light. The EC cytoskeleton is comprised of three key elements: actin microfilaments, intermediate filaments, and microtubules. While the roles of microtubules and intermediate filaments in EC barrier regulation remain to be fully defined, the critical importance of actin microfilaments is demonstrated by increased EC permeability in response to cytochalasin D (11), a known actin disrupter. Conversely, phallacidin, an actin stabilizer, decreases sensitivity to agonist-induced EC barrier disruption (12). Through multiple focal linkage sites to various membrane adhesive proteins, glycocalyx components, functional intercellular proteins of the zona occludens and zona adherens, and focal adhesion complex proteins, the actin microfilament system is fully engaged in determining barrier integrity. At the same time, however, actin is also largely responsible for the generation of tensile intracellular forces via an actomyosin motor. Dynamic actin rearrangement is stimulated by the coordinate activities of the Ca2+/calmodulin-dependent myosin light chain kinase (MLCK) (13) and the kinase effector of the small GTPase Rho (14,15), Rho kinase, which combine to elevate levels of myosin light chain (MLC) phosphorylation and subsequent stress fiber formation (Fig. 2). Focal or spatially defined variability in levels of MLC phosphorylation accounts for the phenotypic-specific contracted or relaxed state of the cell. Actin rearrangement is further dependent on numerous actin binding, capping, nucleating,

Differential regulation of endothelial cell barrier function

245

a

Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The acute respiratory distress syndrome network.

Figure 1 Regulation of endothelial cell barrier function. The endothelial cell (EC) monolayer forms a semipermeable barrier between the vasculature and underlying extracellular matrix. The barrier enhancing agent sphingosine 1phosphate (S1P) promotes EC cytoskeletal changes that strengthen cell-cell and cell-matrix interactions as outlined in the cells on the left. S1P binds the G-protein coupled Edg (endothelial differentiation gene), initiating downstream signaling including Rac activation. Rac stimulates PAK-dependent cortical actin ring formation and cortactin translocation to the actin ring where it interacts with MLCK. S1P also induces adheren junction protein assembly (cadherins, catenins, etc.)

Endothelial cells in health and disease

246

and focal adhesion rearrangement (see text) that strengthen cell-cell and cellmatrix contacts. Various inflammatory stimuli (e.g. thrombin) induce EC cytoskeletal changes resulting in intercellular gap formation and subsequent passage of luminal contents into the interstitium and surrounding tissues (e.g. produces alveolar edema in the lungs). As outlined in the cell on the right, thrombin ligation of the PAR receptor increases intracellular Ca2+, which through Ca2+/calmodulin interaction activates MLCK to phosphorylate myosin light chains, producing actomyosin interaction, actin stress fiber formation, and cell contraction. Thrombin also elevates Rho activity, which further promotes this sequence of events by inhibiting (through Rho kinase) myosin dephosphorylation. Furthermore, thrombin disrupts adherens junction contacts and rearranges focal adhesion interaction to sites at the ends of stress fibers. and severing proteins which contribute to the overall complexity of its regulation. For example, caldesmon is an actin-binding protein that, in its unphosphorylated state, inhibits cell contraction via inhibition of actomyosin ATPase (16,17). However, this effect of caldesmon is attenuated by the actin-severing protein gelsolin resulting in increased stress fiber-dependent contraction (18,19).

3. AGONIST-MEDIATED ENDOTHELIAL BARRIER REGULATION The mechanisms by which various agonists are able to induce either EC barrier disruption or enhancement have been the focus of extensive investigation. We have employed several models in our own investigations including barrier disruption

Differential regulation of endothelial cell barrier function

247

Figure 2 Transendothelial monolayer electrical resistance (TER) measurements represent a sensitive in vitro assay of barrier regulation. Human pulmonary artery ECs were grown to confluence on evaporated gold microelectrodes, and TER measurements were performed in the authors’ laboratory using an electrical cell-substrate impedance sensing system (ECIS) (Applied Biophysics, Troy, NY) as previously described (21). Panel A illustrates the barrier disrupting effects of thrombin (1 U/mL) on EC permeability (increased permeability is reflected in decreased electrical resistance across the monolayer), while panel B shows the barrier enhancing effects of S1P (1µM). Arrows indicate when agonists were added. induced by thrombin (20) as well as barrier enhancement evoked by sphingosine 1phosphate (S1P) (Fig. 2), a product of platelets (21), and by hepatocyte growth factor (HGF) (22). These and other agonists demonstrate unique effects on EC cytoskeletal rearrangement and permeability and have provided important insights into EC barrier regulation (Table 1). The serine protease thrombin produces numerous EC responses which regulate hemostasis, thrombosis, and vessel wall pathophysiology and is recognized as a potentially important mediator in the pathogenesis of acute lung injury. We have characterized the ability of thrombin to activate the endothelium directly and to increase albumin permeability across endothelial cell monolayers in vitro (23).

Endothelial cells in health and disease

248

Table 1 Agonist-mediated Endothelial Cell Cytoskeletal Rearrangement and Barrier Regulation. The effects of specific agonists may be categorized as either Ca2+-dependent or independent. Agonist-induced cytoskeletal rearrangements, characterized by alterations in stress fiber formation and cortical actin, are intimately linked to EC barrier function and permeability. Ca2+

EC Permeability

Thrombin (26,35)

↑

↑

↑

↓

HGF (22)

↑

↓

↓

↑

↑

↓

↓

↑

Sheat stress (42)

↑

↓

early ↑ / late ↓

↑

VEGF (52,61)

↑

↑

↑

↓

LPS (62,63)

—

↑

↑

↓

TNF-α (38)

↑

↑

↑

↓

Agonist

a

S1P (21) b

Stress Fibers Cortical Actin

a

Higher concentrations associated with barrier disruption (>1 µM). Produces increased actin stress fibers at early times points (15 min, 10 dynes/cm2) whereas few fibers are apparent late (24 h). b

Thrombin induced a concentration-dependent increase in I125-albumin clearance that was independent of its interaction with fibrinogen and appeared to be due to a reversible change in EC shape through the formation of intercellular gaps. This observation provided a blueprint for a mechanistic examination of EC barrier properties. We subsequently reported that the direct activation of EC by thrombin is dependent upon its ability to proteolytically cleave the extracellular NH2-terminal domain of the PAR-1 receptor, a member of the family of proteinase-activated receptors (PARs) (24– 27). The cleaved NH2-terminus, acting as a tethered ligand, then activates the receptor. This event initiates a number of downstream effects including the activation of phospholipases A2, C, and D, increased cytosolic Ca2+, and increased permeability (27,28,26,29–31). Activation of the EC thrombin receptor also induces the release of various products including von Willebrand factor, endothelin, NO, and PGI2 (32–34). Specific components of the contractile apparatus are now recognized as targets for thrombin-mediated barrier regulatory signaling pathways. For example, activation of the thrombin receptor induces rapid activation of a Gq protein-coupled phospholipase C leading to inositol 1,4,5-triphosphate3(IP3)-mediated increases in cyotsolic Ca2+. In turn, increased cytosolic Ca2+ leads to the coordinate activation of Rho and MLCK (15,35,13,36,37), with the latter protein directly phosophorylating MLC at Thr-18 and Ser-19. Activation of the thrombin receptor also induces G12/13-mediated Rho activation

Differential regulation of endothelial cell barrier function

249

which, via its target effector Rho kinase, inhibits MLC phosphatase activity by direct myosin phosphatase (MYPT1) phosphorylation thus attenuating dephosphorylation of MLC (35). This potent increase in MLCK/Rho kinase-mediated MLC phosphorylation results in a dramatic increase in intracellular force development and tension, a diminution of the cortical actin ring, and a prominent increase in F-actin stress fibers that traverse the cell (Fig. 1). These intracellular events correspond to increases in EC monolayer permeability and are associated with morphologic cellular changes including the formation of intercellular gaps with the disruption of adherens junctions and reorganization of focal adhesion plaques. Notably, investigations of other agonists have implicated mechanisms of EC barrier disruption independent of Rho Kinase and MLCK activation. For example, we have identified an important role for microtubule rearrangement in EC barrier dysfunction by tumor necrosis factor (TNF)-α (38). In contrast to thrombin’s barrier-disrupting effects, the platelet-derived phospholipid, S1P, produces significant EC barrier enhancement in vitro (21). Originally characterized as a potent angiogenic factor (39), S1P ligates G-protein coupled Edg receptors on the surface of EC to initiate a series of cytoskeletal and adhesive protein rearrangements that result in decreased EC permeability (Fig. 1). Activation of the small GTPase Rac and its downstream target, PAK, is critical for formation of a prominent cortical actin ring that accompanies S1P-induced EC barrier enhancement (21). Moreover, Rac activation stimulates translocation of the multidomain actin-binding protein cortactin to the cortical actin ring where its ability to activate and possibly inhibit other cytoskeletal proteins appears to be essential for maximal S1P-mediated barrier enhancement (S.Dudek, personal communication, 2003). In addition to inducing actin cytoskeletal rearrangement, S1P alters cell–cell and cellmatrix contacts in association with EC permeability reduction. S1P dramatically increases localization of VE-cadherin, α-, β-, and γ-catenin at EC cell-cell junctions (40) while increasing interaction of these adherens junction proteins with the cortical actinbased cytoskeleton (K.Schaphorst, personal communication, 2003). Whereas the barrier disrupting agent thrombin rearranges cell–matrix contacts so that focal adhesion proteins assemble at the ends of newly formed massive actin stress fibers, S1P induces focal adhesion protein rearrangement in association with cortical actin ring formation (41). Stimulation of FAK phosphorylation and p60src activation may partially account for the differential focal adhesion distribution observed in S1P compared to thrombin-treated EC (41). Continued mechanistic evaluation of these models of EC barrier disruption and enhancement hopefully will provide further insights into potential targets for therapeutic modulation of vascular permeability.

4. EFFECTS OF MECHANICAL FORCES ON ENDOTHELIAL BARRIER FUNCTION Experiments investigating the effects of physiologically relevant mechanical forces have yielded additional insights into endothelial barrier regulation. Two forces are of particular interest in endothelial barrier function: shear stress, resulting from the effects of blood flow across EC monolayers, and cyclic stretch, resulting from either the pulsatile

Endothelial cells in health and disease

250

distention of blood vessels or from mechanical ventilation, specifically with respect to the lung vasculature. We previously reported that exposure of lung EC to physiologic levels of laminar shear stress (10 dyn/cm2) results in rapid actin stress fiber formation associated with increased levels of MLC phosphorylation in the cortical actin ring and marked translocation of cortactin, an actin-binding protein involved in peripheral actin polymerization (42). This increase in MLC phosphorylation is dependent on both MLCK and Rho signaling. Moreover, both cortactin translocation and cytoskeletal rearrangement are prevented by inhibition of the small GTPase Rac, known to be involved in membrane ruffling and lamellapodia formation. Cytoskeletal rearrangement occurs within minutes and persists in the presence of sustained shear stress (24 hr). These cortical actin changes are consistent with those seen in other models of barrier enhancement, including S1P (21), and suggest that endothelial adaptation to physiologic shear stress represents a protective response. Separately, we have employed the Flexercell® Tension plus (FX-4000T ) system to study EC subjected to cyclic stretch. This apparatus allows for EC to be grown in a monolayer overlying a flexible substrate under which a vacuum can be applied thus inducing stretch proportional to the degree of vacuum pressure. We previously reported that ECs exposed to 18% radial stretch for 48 hr do not demonstrate any breach in monolayer integrity evaluated histologically or by measurements of transmonolayer electrical resistance under basal conditions (43). However, cyclic stretch-preconditioned EC demonstrated greater paracellular gap formation with increased gap surface area in response to thrombin that correlated with increased levels of MLC phosphorylation. Likely determinants of this heightened response include increased intracellular signaling events via mechanical transduction, priming of the activated cytoskeleton, and the resultant concurrent effects of increased contractile and decreased tethering forces. Additionally, our findings were associated with significant changes in the expression of genes related to regulation of the cytoskeleton (including MLCK and Rho family signaling genes) as determined by Affymetrix microarray experiments. We believe this model is particularly relevant to the clinical condition of ventilator-associated lung injury.

5. DIFFERENTIAL ENDOTHELIAL CELL PHENOTYPES AND BARRIER REGULATION Although the endothelium is often regarded as a singular, uniform entity, it is now abundantly clear that there are notable phenotypic and functional differences of EC at various sites along the vasculature. These differences likely account for the physiologic variability in vascular function appreciated in health as well as the heterogeneous nature of many diseases characterized by increased vascular permeability. Morphologic differences between rat pulmonary artery EC monolayers and rat lung microvascular ECs are appreciable by electron microscopy, with tighter inter-cellular connections and fewer visible gaps in microvascular cells (44). Furthermore, examination of cultured bovine lung EC from microvessels, pulmonary vein, and pulmonary artery reveals cell-specific attributes with respect to the expression of surface proteins, cell

Differential regulation of endothelial cell barrier function

251

morphology, and permeability (45). In particular, microvascular ECs have a ~4-fold larger surface area than pulmonary artery EC (46), while electron microscopy of microvascular ECs demonstrates increased plasmalemmal vesicles, thought to be involved in the transcelluar transport of macromolecules across the cell, as well as a significant increase in focal adhesion sites relative to pulmonary vein or pulmonary artery EC (45). Evidence of a more abundant array of intercellular junctional complexes in microvascular ECs, determinants of small solute transport across cell monolayers, correlates with the observed decreased permeability in these cells under basal conditions. Moreover, despite the increased plasmalemmal vesicles noted in microvascular ECs, there is evidence that they serve as a more restrictive barrier to macromolecules as well (40). In fact, cultured human microvascular ECs exhibit barrier integrity that is approximately 10-fold higher than human macrovascular ECs as measured by electrical resistance across monolayers (47). This likely represents the interaction of several factors including site-specific glycocalyx protein profiles, the extent of junctional protein expression, and alterations in the cytoskeletal profile in a cell-specific manner. In addition to the notable differences between micro- and macrovascular ECs from the lung, significant phenotypic variability also exists in blood vessels from different sites or organs. For example, morphologic differences are noted in rat EC derived from aorta relative to pulmonary artery (48). Similarly, rabbit inferior vena cava ECs have been described as larger than rabbit aorta ECs (49). With respect to cell surface glycoproteins, cell-cell interactions, and protein and mRNA expression, EC variability throughout the vasculature is the rule. The functional significance of this variability remains a highly important area of study. Although data relying on a direct comparison of the response of various EC phenotypes to specific agonists are limited, these investigations have yielded curious results. For the most part, microvascular ECs are more resistant to agonists that induce increased permeability. For example, cultured bovine and sheep pulmonary artery ECs have been characterized as more sensitive to endotoxin than lung micro vascular ECs in terms of cell contraction and loss of barrier function (50). Likewise, investigation of bovine EC has demonstrated increased cell detachment of pulmonary artery EC relative to microvascular ECs in response to TNF-α (51). However, a more dramatic effect of thrombin on bovine EC permeability has been described in microvascular ECs compared to pulmonary artery ECs (46). These findings suggest that this differential response is characterized by specific variations in intracellular signaling yet to be fully defined and is dictated, in part, by distinct receptor expression profiles. It is readily apparent that inflammation and coagulation are tightly linked which may explain high thrombomodulin expression in the vasculature. Importantly, however, the qualitative effects of various agonists with respect to barrier regulation and permeability have proven largely consistent across EC phenotypes. Our own experiments have characterized the effects of vascular endothelial growth factor (VEGF) on large and small-vessel bovine pulmonary endothelial cells (52). Although VEGF induces chemotaxis and barrier disruption in both cell types, a differential concentration dependence is seen. In particular, barrier disruption is more pronounced in macrovascular ECs with the administration of low concentrations of VEGF (1 ng/mL), whereas microvascular ECs are more responsive with respect to chemotaxis. These differences, however, are less apparent with higher concentrations of VEGF (10–100 ng/mL).

Endothelial cells in health and disease

252

As for potential mechanisms underlying differential EC barrier regulation, there is evidence to indicate that Ca2+-dependent events do not play as significant a role in microvascular ECs. Specifically, studies employing rat pulmonary EC demonstrate a blunted response by microvascular cells to increases in intracellular Ca2+ relative to macrovascular cells with respect to changes in permeability (44). Additionally, microvascular ECs are characterized by increased levels of intracellular cAMP, a determinant of the resting contractile state and focal adhesion complex formation, which is regulated by mechanisms distinct from that of macrovascular ECs (53). Specifically, despite increased cAMP, there is decreased ATP-to-cAMP conversion in microvascular ECs. Furthermore, these cells are less responsive to β-adrenergic stimulation—relied on in the clinical setting to increase cAMP and decrease vascular permeability—likely due to increased phosphodiesterase activity relative to macrovascular ECs. Finally, differences in junctional protein expression may account for some of the differential permeability observed in endothelial subpopulations. Microvascular ECs express more VE-cadherin than macrovascular cells, and perhaps as a result, infusion of anti-VEcadherin antibodies increases permeability primarily in alveolar capillaries (9).

6. ROLE OF DIFFERENTIAL ENDOTHELIAL BARRIER FUNCTION IN DISEASE Recognition of the unique response of the endothelium at different sites along the vasculature in various disease states is crucial to understanding their underlying pathophysiology. The observation that the vast endothelial network is comprised of cells with a spectrum of phenotypes accounts, in part, for this variability. Evidence of this is provided by the differential production of nitric oxide by rat lung microvascular and aorta EC in response to TNF-α and LPS, two clinically relevant agonists (54). Models of acute lung injury also support this hypothesis as the systemic administration of lipopolysaccharide (LPS) to mice is associated with differential expression of von Willebrand factor in different organs (55). Separately, our finding of a differential response to VEGF in bovine pulmonary artery and lung microvascular cells (52) is relevant to models of ischemic lung injury which have demonstrated increased VEGF expression in this setting (56,57). One further example is provided by evidence of notable differences in protein expression and permeability between coronary and pulmonary EC exposed to human sera from patients with thermal injuries (58). In addition to specific cell differences, however, variability in the local extracellular environment may also contribute to a differential endothelial response. This is certainly the case in focal disease processes such as lobar pneumonia, typically associated with increased lung vascular permeability in the area of involvement, but is also true for more diffuse processes, such as sepsis and acute lung injury, in which the concentrations of various cytokines and inflammatory mediators may vary widely at different vascular sites. Moreover, variability in the extracellular environment is in particular relevant to the effects of mechanical forces on the endothelium. For example, an appreciation of the effects of shear stress on the lung vasculature must take into account regional differences in blood flow, specifically a decrease at the apices relative to the bases. Further differences in regional blood flow, and thus the degree of shear stress, may be affected by

Differential regulation of endothelial cell barrier function

253

a multitude of clinically relevant factors including changes in blood volume, oxygenation, positional changes, as well as the administration of mechanical ventilation. In fact, it is precisely because of the potentially favorable effects on regional blood flow and lung stretch that prone positioning is used in some mechanically ventilated patients with acute respiratory distress syndrome (59,60).

7. CONCLUSION Differential endothelial barrier regulation continues to be a focus of active investigation. Although a complete picture remains elusive, the increasing availability of new technologies provides hope that significant advances in this area may be readily forthcoming. While better methods to quantitatively image the vascular barrier in a dynamic way are lacking, the application of both genomic and proteomic tools may yield invaluable insights into the intricacies of EC barrier regulation with respect to individual EC phenotypes and their specific mRNA and protein profiles. Ultimately, these findings may be critical to future investigations into novel and highly specific therapeutic targets with respect to various vascular pathobiologies.

ACKNOWLEDGMENTS This work acknowledges support by the Center for Translational Respiratory Medicine, grants from the NHLBI (HL 71411, HL, 70013, HL 58064, and HL 66583) and the Dr. David Marine Endowment.

REFERENCES 1. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The acute respiratory distress syndrome network. N Engl J Med 2000; 342:1301–1308. 2. Minshall RD, Tiruppathi C, Vogel SM, Niles WD, Gilchrist A, Hamm HE, Malik AB. Endothelial cell-surface gp60 activates vesicle formation and trafficking via G(i)-coupled src kinase signaling pathway. J Cell Biol 2000; 150:1057–1070. 3. Tiruppathi C, Song W, Bergenfeldt M, Sass P, Malik AB. Gp60 activation mediates albumin transcytosis in endothelial cells by tyrosine kinase-dependent pathway. J Biol Chem 1997; 272:25968–25975. 4. Hirata A, Baluk P, Fujiwara T, McDonald DM. Location of focal silver staining at endothelial gaps in inflamed venules examined by scanning electron microscopy. Am J Physiol 1995; 269:L403–L418. 5. Vestweber D. Molecular mechanisms that control endothelial cell contacts. J Pathol 2000; 190:281–291. 6. Mitic LL, Anderson JM. Molecular architecture of tight junctions. Annu Rev Physiol 1998; 60:121–142. 7. Kemler R. Classical cadherins. Semin Cell Biol 1992; 3:149–155.

Endothelial cells in health and disease

254

8. Dejana E, Bazzoni G, Lampugnani MG. Vascular endothelial (VE)-cadherin: only an intercellular glue? Exp Cell Res 1999; 252:13–19. 9. Corada M, Mariotti M, Thurston G, Smith K, Kunkel R, Brockhaus M, Lampugnani MG, Martin-Padura I, Stoppacciaro A, Ruco L, McDonald DM, Ward PA, Dejana E. Vascular endothelial-cadherin is an important determinant of microvascular integrity in vivo. Proc Natl Acad Sci USA 1999; 96:9815–9820. 10. Hordijk PL, Anthony E, Mul FP, Rientsma R, Oomen LC, Roos D. Vascular-endothelialcadherin modulates endothelial monolayer permeability. J Cell Sci 1999; 112(Pt 12):1915– 1923. 11. Shasby DM, Shasby SS, Sullivan JM, Peach MJ. Role of endothelial cell cytoskeleton in control of endothelial permeability. Circ Res 1982; 51:657–661. 12. Phillips PG, Lum H, Malik AB, Tsan MF. Phallacidin prevents thrombin-induced increases in endothelial permeability to albumin. Am J Physiol 1989; 257:C562–C567. 13. Goeckeler ZM, Wysolmerski RB. Myosin light chain kinase-regulated endothelial cell contraction: the relationship between isometric tension, actin polymerization, and myosin phosphorylation. J Cell Biol 1995; 130:613–627. 14. Chrzanowska-Wodnicka M, Burridge K. Rho-stimulated contractility drives the formation of stress fibers and focal adhesions. J Cell Biol 1996; 133:1403–1415. 15. Amano M, Chihara K, Kimura K, Fukata Y, Nakamura N, Matsuura Y, Kaibuchi K. Formation of actin stress fibers and focal adhesions enhanced by rho-kinase. Science 1997; 275:1308– 1311. 16. Vorotnikov AV, Gusev NB, Hua S, Collins JH, Redwood CS, Marston SB. Phosphorylation of aorta caldesmon by endogenous proteolytic fragments of protein kinase c. J Muscle Res Cell Motil 1994; 15:37–48. 17. Stasek JE Jr, Patterson CE, Garcia JG. Protein kinase c phosphorylates caldesmon 77 and vimentin and enhances albumin permeability across cultured bovine pulmonary artery endothelial cell monolayers. J Cell Physiol 1992; 153:62–75. 18. Arora PD, Janmey PA, McCulloch CA. A role for gelsolin in stress fiber-dependent cell contraction. Exp Cell Res 1999; 250:155–167. 19. Gusev NB, Pritchard K, Hodgkinson JL, Marston SB. Filamin and gelsolin influence Ca(2+)sensitivity of smooth muscle thin filaments. J Muscle Res Cell Motil 1994; 15:672–681. 20. Garcia JG, Pavalko FM, Patterson CE. Vascular endothelial cell activation and permeability responses to thrombin. Blood Coagul Fibrinolysis 1995; 6:609–626. 21. Garcia JG, Liu F, Verin AD, Birukova A, Dechert MA, Gerthoffer WT, Bamberg JR, English D. Sphingosine 1-phosphate promotes endothelial cell barrier integrity by edg-dependent cytoskeletal rearrangement. J Clin Invest 2001; 108:689–701. 22. Liu F, Schaphorst KL, Verin AD, Jacobs K, Birukova A, Day RM, Bogatcheva N, Bottaro DP, Garcia JG. Hepatocyte growth factor enhances endothelial cell barrier function and cortical cytoskeletal rearrangement: potential role of glycogen synthase kinase-3beta. FASEB J 2002; 16:950–962. 23. Garcia JG, Siflinger-Birnboim A, Bizios R, Del Vecchio PJ, Fenton JW II, Malik AB. Thrombin-induced increase in albumin permeability across the endothelium. J Cell Physiol 1986; 128:96–104. 24. Vogel SM, Gao X, Mehta D, Ye RD, John TA, Andrade-Gordon P, Tiruppathi C, Malik AB. Abrogation of thrombin-induced increase in pulmonary microvascular permeability in par-1 knockout mice. Physiol Genomics 2000; 4:137–145. 25. Vu TK, Hung DT, Wheaton VI, Coughlin SR. Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell 1991; 64:1057– 1068. 26. Garcia JG, Patterson C, Bahler C, Aschner J, Hart CM, English D. Thrombin receptor activating peptides induce Ca2+ mobilization, barrier dysfunction, prostaglandin synthesis, and

Differential regulation of endothelial cell barrier function

255

platelet-derived growth factor mrna expression in cultured endothelium. J Cell Physiol 1993; 156:541–549. 27. Garcia JG. Molecular mechanisms of thrombin-induced human and bovine endothelial cell activation. J Lab Clin Med 1992; 120:513–519. 28. Garcia JG, Fenton JW II, Natarajan V. Thrombin stimulation of human endothelial cell phospholipase d activity. Regulation by phospholipase c, protein kinase c, and cyclic adenosine 3'5'-monophosphate. Blood 1992; 79:2056–2067. 29. Hong SL, Deykin D. Activation of phospholipases a2 and c in pig aortic endothelial cells synthesizing prostacyclin. J Biol Chem 1982; 257:7151–7154. 30. Pollock WK, Wreggett KA, Irvine RF. Inositol phosphate production and Ca2+ mobilization in human umbilical-vein endothelial cells stimulated by thrombin and histamine. Biochem J 1988; 256:371–376. 31. Tiruppathi C, Lum H, Andersen TT, Fenton JW II, Malik AB. Thrombin receptor 14-amino acid peptide binds to endothelial cells and stimulates calcium transients. Am J Physiol 1992; 263:L595–L601. 32. Schini VB, Hendrickson H, Heublein DM, Burnett JC Jr, Vanhoutte PM. Thrombin enhances the release of endothelin from cultured porcine aortic endothelial cells. Eur J Pharmacol 1989; 165:333–334. 33. Garcia JG, Painter RG, Fenton JW II, English D, Callahan KS. Thrombin-induced prostacyclin biosynthesis in human endothelium: role of guanine nucleotide regulatory proteins in stimulus/coupling responses. J Cell Physiol 1990; 142:186–193. 34. Bartha K, Muller-Peddinghaus R, Van Rooijen LA. Bradykinin and thrombin effects on polyphosphoinositide hydrolysis and prostacyclin production in endothelial cells. Biochem J 1989; 263:149–155. 35. Essler M, Amano M, Kruse HJ, Kaibuchi K, Weber PC, Aepfelbacher M. Thrombin inactivates myosin light chain phosphatase via rho and its target rho kinase in human endothelial cells. J Biol Chem 1998; 273:21867–21874. 36. van Nieuw Amerongen GP, Draijer R, Vermeer MA, van Hinsbergh VW. Transient and prolonged increase in endothelial permeability induced by histamine and thrombin: role of protein kinases, calcium, and rhoA. Circ Res 1998; 83:1115–1123. 37. Vouret-Craviari V, Boquet P, Pouyssegur J, Van Obberghen-Schilling E. Regulation of the actin cytoskeleton by thrombin in human endothelial cells: role of rho proteins in endothelial barrier function. Mol Biol Cell 1998; 9:2639–2653. 38. Petrache I, Birukova A, Ramirez SI, Garcia JG, Verin AD. The role of the microtubules in tumor necrosis factor-alpha-induced endothelial cell permeability. Am J Respir Cell Mol Biol 2003; 28:574–581. 39. English D, Welch Z, Kovala AT, Harvey K, Volpert OV, Brindley DN, Garcia JG. Sphingosine 1-phosphate released from platelets during clotting accounts for the potent endothelial cell chemotactic activity of blood serum and provides a novel link between hemostasis and angiogenesis. FASEB J 2000; 14:2255–2265. 40. Lee MJ, Thangada S, Claffey KP, Ancellin N, Liu CH, Kluk M, Volpi M, Sha’afi RI, Hla T. Vascular endothelial cell adherens junction assembly and morphogenesis induced by sphingosine-1-phosphate. Cell 1999; 99:301–312. 41. Shikata Y, Birukov KG, Garcia JG. S1p induces fa remodeling in human pulmonary endothelial cells: role of rac, git1, fak, and paxillin. J Appl Physiol 2003; 94:1193–1203. 42. Birukov KG, Birukova AA, Dudek SM, Verin AD, Crow MT, Zhan X, DePaola N, Garcia JG. Shear stress-mediated cytoskeletal remodeling and cortactin translocation in pulmonary endothelial cells. Am J Respir Cell Mol Biol 2002; 26:453−464. 43. Birukov KG, Jacobson JR, Flores AA, Ye SQ, Birukova AA, Verin AD, Garcia JG. Magnitudedependent regulation of pulmonary endothelial cell barrier function by cyclic stretch. Am J Physiol Lung Cell Mol Physiol 2003.

Endothelial cells in health and disease

256

44. Kelly JJ, Moore TM, Babal P, Diwan AH, Stevens T, Thompson WJ. Pulmonary microvascular and macrovascular endothelial cells: differential regulation of Ca2+ and permeability. Am J Physiol 1998; 274:L810–819. 45. Schnitzer JE, Siflinger-Birnboim A, Del Vecchio PJ, Malik AB. Segmental differentiation of permeability, protein glycosylation, and morphology of cultured bovine lung vascular endothelium. Biochem Biophys Res Commun 1994; 199:11–19. 46. Schaeffer RC Jr, Bitrick MS Jr. Effects of human alpha-thrombin and 8bromo-cAMP on large and microvessel endothelial monolayer equivalent “pore” radii. Microvasc Res 1995; 49:364– 371. 47. Blum MS, Toninelli E, Anderson JM, Balda MS, Zhou J, O’Donnell L, Pardi R, Bender JR. Cytoskeletal rearrangement mediates human microvascular endothelial tight junction modulation by cytokines. Am J Physiol 1997; 273:H286–294. 48. Kibria G, Heath D, Smith P, Biggar R. Pulmonary endothelial pavement patterns. Thorax 1980; 35:186–191. 49. Silkworth JB, McLean B, Stehbens WE. The effect of hypercholesterolemia on aortic endothelium studied en face. Atherosclerosis 1975; 22:335–348. 50. Meyrick B, Hoover R, Jones MR, Berry LC Jr, Brigham KL. In vitro effects of endotoxin on bovine and sheep lung microvascular and pulmonary artery endothelial cells. J Cell Physiol 1989; 138:165–174. 51. Meyrick B, Christman B, Jesmok G. Effects of recombinant tumor necrosis factor-alpha on cultured pulmonary artery and lung microvascular endothelial monolayers. Am J Pathol 1991; 138:93–101. 52. Becker PM, Verin AD, Booth MA, Liu F, Birukova A, Garcia JG. Differential regulation of diverse physiological responses to VEGF in pulmonary endothelial cells. Am J Physiol Lung Cell Mol Physiol 2001; 281:L1500–L1511. 53. Stevens T, Creighton J, Thompson WJ. Control of cAMP in lung endothelial cell phenotypes. Implications for control of barrier function. Am J Physiol 1999; 277:L119–L126. 54. Geiger M, Stone A, Mason SN, Oldham KT, Guice KS. Differential nitric oxide production by microvascular and macrovascular endothelial cells. Am J Physiol 1997; 273:L275–L281. 55. Yamamoto K, de Waard V, Fearns C, Loskutoff DJ. Tissue distribution and regulation of murine von Willebrand factor gene expression in vivo. Blood 1998; 92:2791–2801. 56. Kazi AA, Lee WS, Wagner E, Becker PM. Vegf, fetal liver kinase-1, and permeability increase during unilateral lung ischemia. Am J Physiol Lung Cell Mol Physiol 2000; 279:L460–L467. 57. Geiger R, Berger RM, Hess J, Bogers AJ, Sharma HS, Mooi WJ. Enhanced expression of vascular endothelial growth factor in pulmonary plexogenic arteriopathy due to congenital heart disease. J Pathol 2000; 191:202–207. 58. Murphy JT, Duffy S, Purdue GF, Hunt JL. Differential activation of coronary and pulmonary endothelial cells by thermal injury. Shock 2001; 16:227–231. 59. Broccard AF, Shapiro RS, Schmitz LL, Ravenscraft SA, Marini JJ. Influence of prone position on the extent and distribution of lung injury in a high tidal volume oleic acid model of acute respiratory distress syndrome. Crit Care Med 1997; 25:16–27. 60. Pelosi P, Tubiolo D, Mascheroni D, Vicardi P, Crotti S, Valenza F, Gattinoni L. Effects of the prone position on respiratory mechanics and gas exchange during acute lung injury. Am J Respir Crit Care Med 1998; 157:387–393. 61. Faehling M, Kroll J, Fohr KJ, Fellbrich G, Mayr U, Trischler G, Waltenberger J. Essential role of calcium in vascular endothelial growth factor a-induced signaling: mechanism of the antiangiogenic effect of carboxyamidotriazole. FASEB J 2002; 16:1805–1807. 62. Bannerman DD, Goldblum SE. Direct effects of endotoxin on the endothelium: barrier function and injury. Lab Invest 1999; 79:1181–1199. 63. Chakravortty D, Koide N, Kato Y, Sugiyama T, Kawai M, Fukada M, Yoshida T, Yokochi T. Cytoskeletal alterations in lipopolysaccharide-induced bovine vascular endothelial cell injury and its prevention by sodium arsenite. Clin Diagn Lab Immunol 2000; 7:218–225.

13 Differential Regulation of Leukocyte-Endothelial Cell Interactions D.Neil Granger and Karen Y.Stokes Department of Molecular and Cellular Physiology, Louisiana State University Health Sciences Center, Shreveport, Louisiana, U.S.A.

1. INTRODUCTION The adhesion of leukocytes to vascular endothelium is a hallmark of the inflammatory process. The requirement for and participation of specific adhesion glycoproteins in the binding of leukocytes to endothelial cells has been elegantly demonstrated using a variety of experimental approaches (1,2). However, our current understanding of the molecular basis for leukocyte-endothelial cell adhesion is largely based on data generated from studies utilizing isolated leukocytes and monolayers of cultured endothelial cells. Much of this information was derived from relatively few endothelial cell models, most notably human umbilical cord endothelial cells (HUVEC). The importance of the HUVEC monolayer model to the field of inflammation research cannot be overstated (1,2). It led to the discovery and characterization of a number of key cell adhesion molecules (CAMs) expressed on endothelial cells and/or leukocytes that are now known to mediate adhesive interactions such as leukocyte rolling, firm adhesion, and emigration. This in vitro model has also provided quantitative insights into the modulating role of physical factors such as shear stress on leukocyte-endothelial cell adhesion. Similarities in the time-course and magnitude of CAM expression following activation of different endothelial cell populations grown in culture have led to the general perception that endothelial cells distributed throughout the body are relatively homogeneous in their responses and contributions to inflammatory stimuli. Significant progress in our understanding of regional differences in CAM expression and the subsequent recruitment of leukocytes has resulted from the development of technologies that allow for quantification of endothelial CAMs as well as leukocyteendothelial cell adhesion in a variety of vascular beds. Studies employing these methodologies have revealed considerable quantitative differences between regional vascular beds and within different segments of the same vascular bed. The overall objective of this chapter is to summarize evidence that addresses the issue of intra- and inter-organ variations in leukocyte-endothelial cell adhesion.

Endothelial cells in health and disease

258

2. OVERVIEW OF LEUKOCYTEENDOTHELIAL INTERACTIONS Leukocyte-endothelial interactions can be divided into three main steps: rolling, firm adhesion, and emigration (1,3). The initial tentative contact between circulating white blood cells and the vessel wall, known as rolling, is mediated by the binding of selectin molecules on the endothelium (E-selectin and P-selectin) or leukocytes (L-selectin) to their corresponding ligands (e.g., sialyl Lewis X and PSGL-1) on the other cell. This slows the leukocytes down, essentially increasing the likelihood that they will firmly adhere to the vessel wall via other adhesion molecules, immunoglobulins such as ICAM1 on the vascular endothelium, and integrins such as CD11/18 on the leukocytes. These two families of CAMs are also responsible for the subsequent transmigration of the leukocytes through the vessel wall into the interstitium where they can promote further tissue damage.

3. INTRA-ORGAN DIFFERENCES IN LEUKOCYTE-ENDOTHELIAL CELL ADHESION Direct visualization of the microvasculature in tissues that are either acutely or chronically inflamed has revealed that leukocytes selectively bind to endothelial cells that line postcapillary venules. While leukocyte recruitment into capillaries frequently accompanies inflammation of the lung and, to a lesser extent, other vascular beds, steric hindrance of less deformable circulating leukocytes within long, narrow capillaries, rather than cell-cell interactions mediated by CAMs, is often offered as an explanation for the leukocyte sequestration in this segment of the vasculature (4). Endothelial cell swelling within capillaries as well as compression of the capillary lumen by an elevated interstitial fluid pressure caused by accumulation of edema fluid have also been implicated in the entrapment of leukocytes within capillaries of inflamed tissues (5). In addition, endothelial cells lining venules, when compared to other microvascular segments, appear to sustain most of the leukocyte trafficking that occurs in inflamed tissues. For example, in the rat mesenteric microcirculation, 39% of all leukocytes passing through venules are rolling, while only 0.6% of leukocytes roll in the upstream arterioles (6). The basis for the preferential binding of leukocytes to venular endothelial cells during inflammation has been the focus of much speculation. Relatively higher shear rates in arterioles and higher endothelial CAM expression in venules are two explanations that are most often provided for the differences in leukocyte-endothelial cell adhesion between these two microvascular segments.

3.1. Shear Rates Shear rates generated by the movement of blood within the microvasculature are generally higher in arterioles than in downstream venules. Whether leukocytes adhere to

Differential regulation of leukocyte-endothelial cell interactions

259

vascular endothelium in microvessels depends on the balance between the proadhesive forces generated by adhesion glycoproteins expressed on the surface of leukocytes, endothelial cells, or both, and the anti-adhesive forces generated by hydrodynamic factors such as wall shear stress or shear rate (6). Hence, vessels exhibiting a low spontaneous shear rate (low blood flow) would tend to exhibit more leukocyte adhesion than vessels with high shear rates. In 1973, Atherton and Born (7) suggested that the reason leukocyte rolling and adhesion are rarely observed in arterioles is because the higher shear forces exceeded the adhesive forces in these vessels. Based on their proposal, one might predict that reductions in arteriolar shear

Figure 1 Relationship between number of adherent leukocytes and wall shear rate in venules and arterioles in cat mesentery (8). Over the same range of shear rates, leukocytes preferentially adhere in venules, with minimal adhesion noted in arterioles, even at low shear rates. rate to levels experienced by venules should promote leukocyte adhesion in arterioles. To address this issue, cat mesenteric arterioles and venules of the same size were exposed to

Endothelial cells in health and disease

260

the same range of shear rates (100–1250 sec−1) (8). Leukocyte rolling and adhesion were observed in arterioles but only at low (385 sec−1, whereas in venules, leukocyte adherence persisted at shear rates up to 900 sec−1. Based on these observations, it was concluded that hemodynamic differences between arterioles and venules cannot explain the predilection for leukocyte rolling and adherence in venules. Another approach to elucidate the reason for these observations has been to perform retrograde perfusion of the microcirculation. In the mesentery, this is associated with a reduced flux of rolling leukocytes in venules and increased leukocyte rolling in arterioles (9). Despite this change in responses, more leukocytes still rolled in venules during normograde perfusion than rolled in arterioles during retrograde flow. These findings, coupled to the evidence summarized above, suggest that a more probable explanation for the greater adhesive interactions between leukocytes and venular endothelium is that the counter-receptors (ligands) for leukocyte adhesion molecules are more densely concentrated on venular endothelium.

3.2. CAM Expression CAM expression has been evaluated in the microvasculature using immunohistochemical methods. This approach yields results that consistently show a preferential expression of endothelial CAMs in postcapillary venules. Endothelial cell adhesion molecules such as intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM1), E-selectin, and P-selectin can be detected on the surface of activated endothelial cells in arterioles and occasionally capillaries; however, the density of these adhesion molecules is far greater on venular endothelium (3). In the liver, for example, P-selectin and E-selectin are highly expressed on arterial and venular endothelium, but not on capillary (sinusoidal) endothelium during acute or chronic inflammation (10). While immunohistochemical localization of endothelial CAMs has clearly demonstrated preferential distribution of these adhesion glycoproteins in venules, this approach has yielded relatively little quantitative information concerning the density of CAM expression within vascular beds. Laser confocal microscopy has been employed to localize and quantify the constitutive expression of ICAM-1 in arterioles, capillaries, and venules in the mesentery and liver of rats, using an FITC-labelled antirat monoclonal antibody (11). In the mesentery, venules exhibited a 10-fold higher density of ICAM-1 than in the upstream arterioles and capillaries (Fig. 2A). While the level of ICAM-1 expression in liver venules was comparable.to that detected in mesenteric venules, there was a surprising lack of difference between ICAM-1 density for liver sinusoids (capillaries)

Differential regulation of leukocyte-endothelial cell interactions

Figure 2 Laser confocal microscopic determinations of endothelial ICAM-1 expression in different segments of the rat mesenteric and hepatic microcirculation (A), and in different sized venules of rat mesentery (B). Data are expressed as apparent concentration, which is derived from the intensity of the fluorescence and the predetermined binding ratio between the fluorescence and the

261

Endothelial cells in health and disease

262

immunoglobulin to which it is attached. (Based on data from Ref. 11.) vs. venules, suggesting the absence of a capillary-venule expression gradient of ICAM-1 within the liver microcirculation. However, the authors could not exclude the possibility that hepatocytes per se, rather than endothelial cells, express the high levels of ICAM-1 seen within the sinusoids (12), since the FITC-labelled ICAM-1 antibody can readily diffuse into the space of Disse that separates the porous sinusoidal wall and hepatocyte membrane. Laser confocal microscopic assessment of ICAM-1 expression in mesenteric venules has also revealed heterogeneity of ICAM-1 expression within different sized venules (Fig. 2B). Venules with diameters of 25 µm appear to exhibit the greatest density of ICAM-1 on the endothelial cell surface, while 15 and 35−40 µm diameter venules exhibit the lowest constitutive expression of the adhesion molecule (10). This venule sizedependent distribution of ICAM-1 is consistent with functional evidence demonstrating that 25–30 µm diameter venules sustain the most intense leukocyte adhesion responses to proinflammatory mediators (11). It remains unclear whether other endothelial CAMs exhibit a similar size-dependent distribution within postcapillary venules, or indeed whether ICAM-1 displays comparable size-dependent distribution in organs other than the mesentery. Endothelial cells isolated and cultured from major arterial and venous vessels have also been used to compare the levels of constitutive and induced CAM expression between the arterial and venous segments of the vasculature. Generally, these studies demonstrate that endothelial cells isolated from both ends of the vascular tree have the capacity to express the major CAMs that participate in the recruitment of leukocytes (14– 16). Some studies show that different cytokines and endotoxin elicit qualitatively and quantitatively similar responses in the expression of ICAM-1, VCAM-1, and E-selectin between cultured human arterial and venous endothelial cells, with stimulants like oxidized low density lipoprotein (oxLDL) serving as a more effective inducer of the endothelial CAMs on arterial endothelial cells (14). A limitation of some of these comparisons is that the arterial and venous endothelial cells were not derived from the same vascular bed or donor. When such an analysis is performed, clear differences between arterial and venous endothelial cells can be demonstrated. For example, a comparison of donor-matched human iliac venous and arterial endothelial cells has revealed that both arterial and venous endothelial cells express tumor necrosis factoralpha (TNF-α)-inducible ICAM-1 and E-selectin and that these molecules are functional in mediating leukocyte adhesion (16). VCAM-1, on the other hand, was only inducible on the cultured venous endothelial cells, where it was functional in binding leukocytes.

Differential regulation of leukocyte-endothelial cell interactions

263

4. INTER-ORGAN DIFFERENCES IN LEUKOCYTE-ENDOTHELIAL CELL ADHESION 4.1. In Vivo CAM Expression In vivo studies based on immunohistology have rather consistently demonstrated a localized expression of different endothelial CAMs in venules of normal and/or inflamed tissues and the intensity of the immunostaining for certain endothelial CAMs in these venules appears to be functionally correlated with the amount of leukocyte infiltration (3). Differences in tissue processing and immunostaining procedures, imprecise localization of antigen to the lumenal membrane, the subjective nature of scoring the intensity of antibody staining, and variations in the binding affinity of antibodies employed between laboratories make it difficult to draw conclusions about inter-organ differences in CAM expression. An approach that has been developed to overcome many of the limitations of immunohistochemistry for quantification of endothelial CAM expression involves monitoring the accumulation of a radiolabelled monoclonal antibody (mAb) that is directed against a specific endothelial CAM. This novel approach was originally developed by Haskard and associates (17) to quantify vascular lumen expression of E-selectin in porcine skin following systemic or local injection of interleukin-1. We have modified this radiolabelled mAb method to provide quantitative measurements of endothelial CAM expression in rat and mouse models of acute and chronic inflammation (18–20). Our dual radiolabelled mAb method determines the relative accumulation, in any regional vascular bed, of a binding mAb to a specific endothelial surface epitope (e.g., P-selectin or CD40) and an isotype-matched nonbinding mAb, the latter of which is used to compensate for non-specific accumulation of the binding mAb. We have successfully employed this method to quantify the expression of different endothelial CAMs [P-selectin, E-selectin, VCAM-1, ICAM-1, ICAM-2, mucosal addressin cell adhesion molecule-1 (MAdCAM-1), platelet-endothelial cell adhesion molecule-1 (PECAM-1)] in different vascular beds of the mouse (18–23). The method yields values that are expressed as either % injected dose (%ID) or nanograms of mAb per gram of tissue. In our previous studies employing the dual radiolabelled mAb technique, it was demonstrated that: (1) the radiolabelling procedure does not alter the ability of the mAb to block leukocyte-endothelial cell adhesion in postcapillary venules (18); (2) accumulation of the radiolabelled binding (but not the non-binding) mAb in a vascular bed can be inhibited in a dose-dependent fashion by progressively increasing concentrations of the cold (unlabelled) binding mAb (but not the non-binding mAb) (18); (3) expression of the targeted endothelial cell surface epitope on non-binding circulating cells (e.g., CD40 on T-cells) does not interfere with the assay since all residual blood is flushed from the vasculature prior to tissue sampling; and (4) application of this technique to the relevant adhesion molecule-deficient mice yields expression values that are essentially 0, both under basal and stimulated conditions (19,20). It has also been

Endothelial cells in health and disease

264

determined that the accumulation of binding mAb, relative to the non-binding mAb, is essentially unchanged after a 2–5 min intravascular mixing period (18–20). This is an advantage of the technique since the rapid accumulation of a mAb due to engagement of its ligand allows for an estimate of mAb binding before significant immunoglobulin extravasation can occur. The distributional half-lives of the binding and non-binding mAbs are on the order of several hours while the endothelial CAM measurement technique quantifies the mAb binding that occurs within a few minutes. Our initial studies of ICAM-1 expression using the dual radiolabelled mAb technique were undertaken in rats (18). Constitutive and endotoxin-induced ICAM-1 expression was monitored in different regional vascular beds. The binding of the radiolabelled ICAM-1 mAb varied widely among organs. The constitutive level of ICAM-1 expression appeared to be significant in all vascular beds, which is consistent with the effectiveness of ICAM-1 neutralizing mAbs in blocking the rapid leukocyte–endothelial cell adhesion elicited by certain mediators and conditions (3). Endotoxin induced an increased ICAM-1 expression in virtually every organ, with vascular beds normally exhibiting a low constitutive expression of ICAM-1 (heart and skeletal muscle) showing the greatest increases and lung (which has the highest constitutive expression of ICAM-1) showing the smallest increment. The profound differences in predicted expression of basal ICAM-1 between vascular beds like the lung, heart, and skeletal muscle may reflect corresponding differences in endothelial surface area in these tissues, rather than a higher density of ICAM-1 per endothelial cell. This possibility is supported by experiments showing that the relative accumulation of a mAb directed against angiotensin converting enzyme, which is constitutively expressed on the surface of endothelial cells, in different organs follows a pattern similar to that obtained basally with an anti-ICAM-1 mAb (lung>heart>skeletal muscle) and it is positively and significantly (r=0.98, p70 or 50–69 years old with diabetes and/or smoking history were screened for PAD using Doppler-derived measurement of ankle pressures. Of those patients that had coronary and/or carotid artery disease, 40% also had PAD (8). Nevertheless, it is clear that large numbers of our patients prominently manifest arterial occlusive disease to a greater extent in one circulation than in another. This intriguing heterogeneity in the geography of plaque distribution has not been satisfactorily explained. Epidemiological studies have revealed that there is a slightly increased prevalence of smokers, diabetics, and elderly individuals in the population of patients presenting with symptomatic PAD by comparison to patients presenting with symptomatic CAD (66). Cross-sectional case-control studies have suggested that elevated plasma levels of lipoprotein (a) and reduced levels of high density lipo-protein cholesterol (HDL-C) have slightly greater predictive power for PAD by comparison with CAD (67). However, there is considerable overlap in risk factors between those patients with PAD and those that present with disease in other circulations. It is likely that there are unknown environmental and genetic influences that determine the variability of plaque distribution and its severity.

4.4. The Diversity of Presentations of PAD Sadly, although PAD is common, it is commonly unrecognized, and often under-treated. The PARTNERS study documented that less than 50% of these patients are diagnosed, and that life-saving therapies were not fully deployed (2). In those individuals for whom therapy was indicated, a significant number were not prescribed medication for hypertension (20%) or dyslipidemia (40%). Tragically, half were not receiving antiplatelet medication. Each of these therapies reduces cardiovascular morbidity and

Endothelial cells in health and disease

372

mortality by 20–30% (in CAD and PAD). There are several factors responsible for such poor care including a health care system that rewards technological intervention and undervalues prevention; a misperception among primary care providers that there is little recourse but surgery or angioplasty; and a lack of public education regarding the disease. Although these factors are outside the scope of this discussion, a contributing factor to the problem is the diversity of presentations of PAD. The patient with PAD typically feels an aching, burning, cramping, or fatigue of the calf muscles with walking, that is relieved by standing still. This symptom is known as intermittent claudication (from the Latin verb claudus, to limp, relating to the Roman emperor Claudius who walked with a limp). This discomfort is due to obstructive lesions that limit blood flow, typically in the superficial femoral artery. If the obstructive lesions are more proximal (e.g., aorto-iliac region), the patient may also have thigh and buttock claudication. However, these classic symptoms are only present in 30% or fewer of patients (8,68). Many individuals are “asymptomatic” because they are sedentary, or because they attribute their aches and pains to growing old, or to their arthritis. Of course, musculoskeletal disorders and PAD are each common in the elderly, and when they coexist, it can be difficult to determine by history which is most responsible for the patient’s symptoms. Of more relevance to this textbook is the unexplained individual heterogeneity in the development of collaterals that affect the presentation. Some individuals with superficial femoral artery occlusions bilaterally may be severely limited. Others may have little or no symptoms due to the development of extensive collaterals. Although much is known about the determinants of angiogenesis (the sprouting of new capillaries) and arteriogenesis (the growth and remodeling of collateral channels), and the intriguing potential of endothelial precursor cells, there is no explanation for the individual heterogeneity in the development of collateral channels. An understanding of this heterogeneity in the angiogenic/arteriogenic response to ischemia would be certain to lead to new therapeutic avenues.

4.5. Endothelial Heterogeneity in Coagulation Inhomogeneity of plaque distribution could also be explained by local differences in the vascular expression of factors influencing blood fluidity. Plaque thrombogenicity plays an important role in the progression of atherosclerosis (69,70). Tissue factor is a smallmolecular-weight glycoprotein and is paradigmatic of a vascular-derived factor that influences blood fluidity. Tissue factor triggers activation of the extrinsic clotting cascade by forming a high-affinity TF/VIIa complex that activates factors IX and X, which in turn lead to thrombin generation. Tissue factor activity is inhibited by tissue factor pathway inhibitor (TFPI) in the vessel wall. In the coronary arteries, plaque disruption exposes the lipid core, which is highly thrombogenic due to its abundance of macrophage derived TF (70). By contrast, in PAD, there is very little evidence that plaque rupture is responsible for progression of disease. Instead, significant plaque stenosis combined with systemic hyperthrombogenicity of the systemic blood may be a more common mechanism of thrombotic arterial occlusion in the lower extremities. PAD commonly coexists with diabetes and cigarette smoking (71,72), which are known to contribute to significant hyperthrombogenicity of the blood. Furthermore, there is evidence for circulating TF

The diversity of vascular disease

373

activity in microparticles, possibly released by apoptotic macrophages (73). The systemic hyperthrombogenicity may be exacerbated by local imbalance in vascular TF and TFPI activities (74). Intriguingly, transcriptional profiling studies at Stanford have recently revealed differential expression of TFPI between human coronary and aortic endothelial cells (personal communication, Dr. Thomas Quertermous).

4.6. Heterogeneity of Vascular Function and Gene Expression Significant differences exist between the coronary and peripheral vessels in endothelial and vascular smooth muscle function and gene expression. As an example, endotheliumdependent vasodilation in the peripheral arteries of humans is quite different from that in central arteries (75). Whereas human limb and coronary arteries each manifest a cholinergic endothelium-dependent vasodilation, only the coronary arteries respond to serotonin with an endothelium-dependent vasodilation (75,76). In animal studies, there are significant differences between coronary and limb arteries in endothelium-dependent vasodilation, or smooth muscle vasoconstriction, to a variety of agonists (77). This difference in vascular reactivity between the peripheral and coronary arteries reflects a difference in the expression of endothelial/vascular smooth muscle receptors and cell signaling pathways. There are also well-established differences between vascular beds in the expression of endothelial chemokines and adhesion molecules. These geographic differences are known to play a physiological role in lymphocyte homing and other cellular trafficking in immune surveillance (78). Furthermore, vascular smooth muscle cells manifest differences in gene expression throughout the vasculature (79). It is likely that geographic differences in vascular gene expression and vascular function in normal vessels provide the anlagen for the nonuniform distribution of plaque. It seems likely that polymorphisms of genes that are differentially expressed in the vasculature may be more likely to influence the individual’s susceptibility to developing atherosclerosis in a specific segment of the arterial tree. Specifically, the predisposition for PAD in some individuals is due to an interaction of candidate gene polymorphisms with hemodynamic, humoral, and metabolic stimuli that predispose to atherosclerosis.

5. GENES AS RISK FACTORS Extensive epidemiological studies have defined the importance of environmental influences and systemic risk factors in the development of atherosclerosis. Those risk factors with the strongest epidemiological support include age, gender, hypertension, diabetes mellitus, hypercholesterolemia, tobacco use, and family history of premature atherosclerotic disease (80). Other risk factors or markers include C-reactive peptide, obesity, sedentary state, and insulin resistance (81). There are important interactions between environmental and genetic determinants in the expression of the systemic risk factors. For example, elevated levels of low density lipoprotein cholesterol (LDL-C) are determined by the interaction of diet with genetic variability in the expression of LDL-C

Endothelial cells in health and disease

374

and its receptor (82). The adverse effect of tobacco exposure on the vessel wall is in part determined by a genetic polymorphism of the NOS gene. Specifically, in young individuals with the NOS Glu298Asp polymorphism, exposure to tobacco smoke causes a severe impairment of endothelium-dependent vasodilation, whereas young smokers without this polymorphism are resistant to the adverse effects of tobacco on this vascular function (83). The role of heredity as a determinant of susceptibility to atherosclerosis was indicated by large epidemiological studies such as the Framingham studies, which consistently showed the predictive value of a family history of premature atherosclerotic disease (84). These observations were confirmed and extended by studies of identical twins separated at birth, which revealed a persistently similar risk profile despite exposure to different environments (85). Subsequently, elegant studies of the genetic determinants of lipoprotein metabolism provided further evidence for the role of heredity. A notable example of this work was the elucidation of mutations of the LDL-C receptor or its signaling pathway that result in familial hypercholes-terolemia and premature atherosclerosis of carotid, coronary, and peripheral circulations (86–88). Familial hypercholesterolemia is an autosomal dominant disorder due to mutations of the LDL receptor gene, which results in diminished LDL uptake and markedly elevated levels of circulating LDL. The elucidation of the genetic determinants of familial hypercholesterolemia illustrates the power of genetics to reveal mechanistic insights as well as provide diagnostic and potential therapeutic targets in cardiovascular diseases. However, this example represents a single gene defect resulting in a relatively extreme phenotype that is not representative of the great majority of patients at risk for atherosclerosis.

5.1. A New Approach Is Needed to Fully Understand Genes as Risk Factors While loss of function mutations at the LDL-R locus gives rise to a Mendelizing phenotype of hyperlipidemia, these mutations are quite rare in the general population and have low attributable risk. The simple Mendelian inheritance pattern of this disorder and the availability of large affected kindreds facilitated a classical linkage approach, which is not feasible for understanding complex polygenic diseases like atherosclerosis (89). Elucidation of the genetic determinants of atherosclerosis will require innovative new paradigms. It is likely that the susceptibility or resistance to developing atherosclerotic disease, and the severity and location of the disease, involves the interaction of multiple genetic loci and environmental risk factors, and that most of these factors are found at moderate to high frequency in the general population (ApoE polymorphism as an example). Evidence supporting this model comes from twin studies that demonstrate a very high monozygotic to dizygotic twin concordance ratios for cardiovascular disease mortality at early ages, consistent with the interaction of multiple genetic factors (90). Considerable prior effort has focused on characterizing the genetic contribution to lipid metabolism and its relationship to atherosclerosis, for example by identification of such loci as LDL-R, ApoE, ApoCII, ApoB, Lp1, and Lp(a). By contrast, relatively little effort has been placed on identifying the genes involved in vascular function and structure. It is likely that vascular wall loci will have an important impact on determining

The diversity of vascular disease

375

the severity and distribution of atherosclerosis. However, it will also be important to analyze how these newly identified loci interact with other known atherosclerosispredisposing genes, such as those involved in lipid metabolism, and how they interact with environmental influences.

5.2. Rationale for Focusing on Genes Expressed in the Vasculature The epidemiologic studies that have identified the well-known cardiovascular risk factors usually compared clinically normal subjects with patients who developed cardiovascular disease. These studies have not distinguished factors that preferentially confer risk for disease in the peripheral circulation. In general, the systemic processes that drive the formation of atherosclerotic plaque do not seem to be very different for CAD and PAD. Is it possible that the development of atherosclerosis in the limb arteries is due to the interaction of systemic risk factors with individual differences in vascular gene expression? For example, is it possible that a functional polymorphism of an endothelial adhesion molecule, in combination with traditional systemic risk factors, could predispose to a greater degree of monocyte adhesion in the femoral artery than in the coronary artery? This seems likely, given the well-established geographic differences in the expression of endothelial adhesion molecules. While many previous studies have investigated genetic determinants of systemic coronary risk factors, relatively few have examined genes active in the blood vessel wall. The few studies that are available suggest that polymorphisms of genes preferentially expressed in the vasculature may play an important role in determining the severity of vascular disease. The eNOS gene missense Glu298Asp mutation is associated with a 1.7 odds ratio for acute MI in one case-control study (91). This finding has been confirmed in two independent samples from the United Kingdom, which found an odds ratio of 4.2 for coronary disease and 2.5 for acute MI due to the Glu298Asp mutation (92). Polymorphisms in the promoter for stromelysin, a matrix metalloproteinase expressed in the vessel wall, have been associated with CAD progression (93,94) and acute MI (95) in small studies. The DD genotype of the gene for angiotensin converting enzyme has been associated with an odds ratio of 1.3 for developing MI in meta-analysis of 15 case-control studies comparing MI patients to controls with no clinical evidence of CAD (96); but this association has not been confirmed in two studies of a large cohort of Danish subjects (97), and in the ISIS-3 case-control study (98). This brief overview suggests that polymorphisms of genes preferentially expressed in the vasculature may indeed predispose to the development of atherosclerosis. Most previous studies, while provocative, are limited by one or more factors, including small sample size. Few if any of these studies controlled for the effects of clinical factors when assessing the possible effect of genetic factors. Finally, no study has attempted to define the genetic determinants for distribution of plaque.

Endothelial cells in health and disease

376

5.3. The Importance of Understanding the Genetic Determinants of PAD The focus of our current work is to identify single nucleotide polymorphisms (SNPs) of candidate genes that are associated with atherosclerosis in the peripheral arteries. These studies will provide us with SNPs that are potentially of functional importance, and worthy of additional investigation using cell and molecular biological approaches and transgenic technology. These studies are likely to provide novel insights into the pathophysiology of atherosclerosis. Furthermore, our genomic studies may provide new markers of risk for PAD, an underdiagnosed disease. These studies may provide potential new targets for therapy or prevention of PAD, an undertreated disorder. Genetic epidemiological studies suggest that atherosclerosis is not determined by variation in a single gene, but by DNA variants in a number of genes, any one of which is unlikely to explain more than 5% of the individual variation in the disease process (90). Although the technique of genetic linkage analysis is a powerful approach for identifying individual genes that play a major role in disease, this approach has limited power to identify individual genes that contribute as little as 5–10% of disease variation. In contrast, genetic association analysis has excellent power to identify genes contributing as little as 2% of the disease variation using case-control designs employing as few as 1000 cases and 1000 controls (89,99). We hypothesize that common single DNA variants, each accounting for at least 2% of the individual variation in the atherosclerotic disease process, are widely distributed in the human population and can be identified by using association analysis with a case-control study design involving 2000 individuals. Accordingly, we are currently initiating a case-control cross-sectional genomic study designed to identify SNPs of candidate genes that are predictive for the development of PAD. The methodology to select cases and controls must result in two sharply defined phenotypes (i.e., presence or absence of PAD) with minimal differences between the two populations in the distribution of known risk factors. This methodology will increase the contribution made by unknown genetic factors to the difference in phenotypes. Our underlying premise is that identifiable clinical, genetic, and environmental factors affect the distribution of plaque, and that the heterogeneity of disease presentation is not a random process, but rather one that is modulated and mediated by identifiable clinical and genetic factors.

6. CONCLUSION The inhomogeneity in the distribution of atheromatous plaque is a function of endothelial heterogeneity. A striking example of endothelial heterogeneity occurs at bends, branches, and bifurcations of the blood vessel. At these sites, the transition from laminar to disturbed flow causes morphological and functional changes in endothelial cells. These alterations include an acceleration of the aging process with reduced synthesis of vasoprotective factors, thereby increasing the local susceptibility to atherogenesis.

The diversity of vascular disease

377

In addition, there are regional differences in endothelial response that may modulate the resistance to vascular disease. For example, the elaboration of NO by endothelial cells of the internal mammary artery is much greater than that of the saphenous vein. This difference may contribute to the dramatic disparity in the long-term patency of these two conduits when they are surgically placed into the coronary circulation. Furthermore, when the vein is exposed to the arterial circulation, its “arterialization” is manifested by changes in vascular function and structure that include intimal smooth muscle proliferation and endothelial alterations that promote atherogenesis (100,101). Undefined regional differences in endothelial response are likely to account for the relative resistance to atherosclerosis of some arteries (e.g., the upper extremity vessels). Finally, individual variants in endothelial genes are likely to be responsible for the great variation in the presentation of atherosclerosis. An understanding of the mechanisms for this individual heterogeneity may lead to new therapeutic avenues in the treatment of atherosclerotic vascular diseases.

ACKNOWLEDGMENTS This study was supported by grants from the National Institute of Health (R01HL63685, R01 HL075774; P01 A150153); and M01 RR00070(General Clinical Research Center, Stanford University School of Medicine) and the California tobacco related disease research program of the University of Calofornia (11RT-0147).

REFERENCES 1. Criqui MH. Peripheral arterial disease—epidemiological aspects. Vasc Med 2001; 6(3 Suppl):3– 7. 2. Hirsch AT, Criqui MH, Treat-Jacobson D, Regensteiner JG, Creager MA, Olin JW, Krook SH, Hunninghake DB, Comerota AJ, Walsh ME, McDermott MM, Hiatt WR. Peripheral arterial disease detection, awareness, and treatment in primary care. JAMA 2001; 286(11):1317–1324. 3. Kannel WB. The Framingham study: its 50-year legacy and future promise. J Atheroscler Thromb 2000; 6(2):60–66. 4. Clarke R, Daly L, Robinson K, Naughten E, Cahalane S, Fowler B, Graham I. Hyperhomocysteinemia: an independent risk factor for vascular disease. N Engl J Med 1991; 324(17):1149–1155. 5. Widmann MD, Sumpio BE. Lipoprotein (a): a risk factor for peripheral vascular disease. Ann Vasc Surg 1993; 7:446–451. 6. Millonig G, Malcom GT, Wick G. Early inflammatory-immunological lesions in juvenile atherosclerosis from the Pathobiological Determinants of Atherosclerosis in Youth (PDAY)study. Atherosclerosis 2002; 160(2):441–448. 7. Szuba A, Cooke JP. Thromboangiitis obliterans. An update on Buerger’s disease. West J Med 1998; 168(4):255–260. 8. Adar R, Papa MZ, Halpern Z, Mozes M, Shoshan S, Sofer B, Zinger H, Dayan M, Mozes E. Cellular sensitivity to collagen in thromboangiitis obliterans. N Engl J Med 1983; 308(19):1113–1116. 9. Olin JW. Thromboangiitis obliterans (Buerger’s disease). N Engl J Med 2000; 343(12): 864– 869.

Endothelial cells in health and disease

378

10. Numano F, Kobayashi Y, Maruyama Y, Kakuta T, Miyata T, Kishi Y. Takayasu arteritis: clinical characteristics and the role of genetic factors in its pathogenesis. Vasc Med 1996; 1(3):227–233. 11. Kerr GS, Hallahan CW, Giordano J, Leavitt RY, Fauci AS, Rottem M, Hoffman GS. Takayasu arteritis. Ann Intern Med 1994; 120(11):919–929. 12. Ross R, Glomset J, Harker L. Response to injury and atherogenesis. Am J Pathol 1977; 86(3):675–684. 13. Cooke JP, Dzau VJ. Nitric oxide synthase: role in the genesis of vascular disease. Annu Rev Med 1997; 48:489–509. 14. Griendling KK, Alexander RW. Oxidative stress and cardiovascular disease. Circulation 1997; 96(10):3264–3265. 15. Tsao PS, Buitrago R, Chan JR, Cooke JP. Fluid flow inhibits endothelial adhesiveness: nitric oxide and transcriptional regulation of VCAM-1. Circulation 1996; 94(7): 1682–1689. 16. Tsao PS, Wang B, Buitrago R, Shyy JY, Cooke JP. Nitric oxide regulates monocyte chemotactic protein-1. Circulation 1997; 96(3):934–940. 17. Jawien A, Bowen-Pope DF, Lindner V, Schwartz SM, Clowes AW. Platelet-derived growth factor promotes smooth muscle migration and intimal thickening in a rat model of balloon angioplasty. J Clin Invest 1992; 89(2):507–511. 18. Casscells W, Hathorn B, David M, et al. Thermal detection of cellular infiltrates in living atherosclerotic plaques: possible implications for plaque rupture and thrombosis. Lancet 1996; 347:1447–1449. 19. Rauch U, Osende JI, Fuster V, Badimon JJ, Fayad Z, Chesebro JH. Thrombus formation on atherosclerotic plaques: pathogenesis and clinical consequences. Ann Intern Med 2001; 134(3):224–238. 20. Moulton KS, Heller E, Konerding MA, Flynn E, Palinski W, Folkman J. Angiogenesis inhibitors endostatin or TNP-470 reduce intimal neovascularization and plaque growth in apolipoprotein E-deficient mice. Circulation 1999; 99(13):1726–1732. 21. Moulton KS, Vakili K, Zurakowski D, Soliman M, Butterfield C, Sylvin E, Lo KM, Gillies S, Javaherian K, Folkman J. Inhibition of plaque neovascularization reduces macrophage accumulation and progression of advanced atherosclerosis. Proc Natl Acad Sci USA 2003; 100(8):4736–4741. 22. Heeschen C, Jang J, Hoai-Ky V, Kaji S, Yang P, Hu RS, Cooke JP. Nicotine is an agent of angiogenesis. Nicotine stimulates angiogenesis and promotes tumor growth and atherosclerosis. Nat Med 2001; 7(7):833–839. 23. Heeschen C, Weis M, Cooke JP. A novel angiogenic pathway mediated by non-neuronal nicotinic acetylcholine receptors. J Clin Invest 2002; 110(4):527–536. 24. Jacobi J, Jang JJ, Sundram U, Fajardo L, Cooke JP. Nicotine stimulates wound healing in diabetic mice. Am J Pathol 2002; 161(1):97–104. 25. Heeschen C, Weis M, Cooke JP. Nicotine promotes arteriogenesis. J Am Coll Cardiol 2003; 41(3):489–496. 26. Celletti FL, Waugh JM, Amabile PG, Brendolan A, Hilfiker PR, Dake MD. Vascular endothelial growth factor enhances atherosclerotic plaque progression. Nat Med 2001; 7(4):425−429. 27. Shah PK, Falk E, Badimon JJ, Fernandez-Ortiz A, Mailhac A, Villareal-Levy G, Fallon JT, Regnstrom J, Fuster V. Human monocyte-derived macrophages induce collagen breakdown in fibrous caps of atherosclerotic plaques: potential role of matrix-degrading metalloproteinases and implications for plaque rupture. Circulation 1995; 92:1565–1569. 28. Sukhova GK, Schonbeck U, Rabkin E, Schoen FJ, Poole AR, Billinghurst RC, Libby P. Evidence for increased collagenolysis by interstitial collagenases-1 and -3 in vulnerable human atheromatous plaques. Circulation 1999; 99(19):2503–2509. 29. Mann J, Davies MJ. Mechanisms of progression in native coronary artery disease: role of healed plaque disruption. Heart 1999; 82:265–268.

The diversity of vascular disease

379

30. Liao JK, Bettmann MA, Sandor T, Tucker JI, Coleman SM, Creager MA. Differential impairment of vasodilator responsiveness of peripheral resistance and conduit vessels in humans with atherosclerosis. Circ Res 1991; 68(4):1027–1034. 31. Hiatt WR. Pharmacologic therapy for peripheral arterial disease and claudication. J Vasc Surg 2002; 36(6):1283–1291. 32. American Heart Association. Heart Disease and Stroke Statistics—2003 Update. Dallas, TX. 33. Ouriel K. Acute limb ischemia. In: Rutherford RB, ed.Vascular Surgery. Philadelphia: W.B.Saunders Company, 2000:813–821. 34. Davies PF, Shi C, Depaola N, Helmke BP, Polacek DC. Hemodynamics and the focal origin of atherosclerosis: a spatial approach to endothelial structure, gene expression, and function. Ann N Y Acad Sci 2001; 947:7–16. 35. Weidinger FF, McLenachan JM, Cybulsky M, Gordon JM, Rennke HA, Hollenberg NK, Fallon JT, Ganz P, Cooke JP. Persistent dysfunction of regenerated endothelium following balloon angioplasty of rabbit iliac artery. Circulation 1990; 81(5): 1667–1679. 36. Chang E, Harley CB. Telomere length and replicative aging in human vascular tissues. Proc Natl Acad Sci USA 1995; 92(24):11190–11194. 37. Matsushita H, Chang E, Glassford AJ, Cooke JP, Chiu CP, Tsao PS. eNOS activity is reduced in senescent human endothelial cells: preservation by hTERT immortalization. Circ Res 2001; 89(9):793–798. 38. McLenachan JM, Vita J, Fish DR, Treasure CB, Cox DA, Ganz P, Selwyn AP. Early evidence of endothelial vasodilator dysfunction at coronary branch points. Circulation 1990; 82(4):1169– 1173. 39. Cooke JP, Dzau VJ. Nitric oxide synthase: role in the genesis of vascular disease. Annu Rev Med 1997; 48:489–509. 40. Stamler JS, Mendelsohn ME, Amarante P, Smick D, Andon N, Davies PF, Cooke JP, Loscalzo J. N-acetylcysteine potentates platelet inhibition by endothelium-derived relaxing factor. Circ Res 1989; 65(3):789–795. 41. Cooke JP, Singer AH, Tsao P, Zera P, Rowan RA, Billingham ME. Anti-atherogenic effects of L-arginine in the hypercholesterolemic rabbit. J Clin Invest 1992; 90(3): 1168–1172. 42. Cayatte AJ, Palacino JJ, Horten K, Cohen RA. Chronic inhibition of nitric oxide production accelerates neointima formation and impairs endothelial function in hypercholesterolemic rabbits. Arterioscler Thromb 1994; 14(5):753–759. 43. Kuhlencordt PJ, Gyurko R, Han F, Scherrer-Crosbie M, Aretz TH, Hajjar R, Picard MH, Huang PL. Accelerated atherosclerosis, aortic aneurysm formation, and ischemic heart disease in apolipoprotein E/endothelial nitric oxide synthase double-knockout mice. Circulation 2001; 104(4):448–454. 44. Shimasaki Y, Yasue H, Yoshimura M, et al. Association of the missense Glu298Asp variant of the endothelial nitric oxide synthase gene with myocardial infarction. J Am Coll Cardiol 1998; 31:1506–1510. 45. Hingorani AD, Liang CF, Fatibene J, et al. A common variant of the endothelial nitric oxide synthase (Glu298 → Asp) is a major risk factor for coronary artery disease in the UK. Circulation 1999; 100:1515–1520. 46. Cooke JP. Does ADMA cause endothelial dysfunction? Arterioscler Thromb Vasc Biol 2000; 20(9):2032–2037. 47. Boger R, Bode-Boger S, Szuba A, Tsao PS, Chan J, Tangphao O, Blaschke T, Cooke JP. Asymmetric dimethylarginine (ADMA): a novel risk factor for endothelial dysfunction: its role in hypercholesterolemia. Circulation 1998; 98(18):1842–1847. 48. Miyazaki H, Matsuoka H, Cooke JP, Usui M, Ueda S, Okuda S, Imaizumi T. Endogenous nitric oxide synthase inhibitor: a novel marker of atherosclerosis. Circulation 1999; 99(9):1141–1146. 49. Boger RH, Bode-Boger SM, Thiele W, Junker W, Alexander K, Frolich JC. Biochemical evidence for impaired nitric oxide synthesis in patients with peripheral arterial occlusive disease. Circulation 1997; 95(8):2068–2074.

Endothelial cells in health and disease

380

50. Valkonen VP, Paiva H, Salonen JT, Lakka TA, Lehtimaki T, Laakso J, Laaksonen R. Risk of acute coronary events and serum concentration of asymmetrical dimethylarginine. Lancet 2001; 358:2127–2128. 51. Zoccali C, Bode-Boger S, Mallamaci F, Benedetto F, Tripepi G, Malatino L, Cataliotti A, Bellanuova I, Fermo I, Frolich J, Boger R. Plasma concentration of asymmetrical dimethylarginine and mortality in patients with end-stage renal disease: a prospective study. Lancet 2001; 358(9299):2113–2117. 52. MacAllister RJ, Parry H, Kimoto M, Ogawa T, Russell RJ, Hodson H, Whitley GS, Vallance P. Regulation of nitric oxide synthesis by dimethylarginine dimethylaminohy-drolase. Br J Pharmacol 1996; 119(8):1533–1540. 53. Ito A, Tsao PS, Adimoolam S, Kimoto M, Ogawa T, Cooke JP. Novel mechanism for endothelial dysfunction: dysregulation of dimethylarginine dimethylaminohydrolase. Circulation 1999; 99(24):3092–3095. 54. Lundman P, Eriksson MJ, Stuhlinger M, Cooke JP, Hamsten H, Tornvall P. Mild to moderate hypertriglyceridemia in young men is associated with endothelial dysfunction and increased plasma concentrations of asymmetric dimethylarginine. J Am Coll Cardiol 2001; 38(1):111– 116. 55. Stuhlinger MC, Abbasi F, Chu JW, Lamendola C, McLaughlin TL, Cooke JP, Reaven GM, Tsao PS. Relationship between insulin resistance and an endogenous nitric oxide synthase inhibitor. J Am Med Assoc 2002; 287(11):1420–1426. 56. Stuhlinger MC, Tsao PS, Her J-H, Kimoto M, Balint RF, Cooke JP. Homocysteine impairs the NO synthase pathway—role of asymmetric dimethylarginine. Circulation 2001; 104(21):2569– 2575. 57. Lin KY, Ito A, Asagami T, Tsao PS, Adimoolam S, Kimoto M, Tsuji H, Ogawa T, Reaven G, Cooke JP. Dysregulation of dimethylarginine dimethylaminohydrolase: A mechanism for endothelial dysfunction in diabetes mellitus. Circulation 2002; 106(8):987–992. 58. Leiper JM, Santa Maria J, Chubb A, MacAllister RJ, Charles IG, Whitley GS, Vallance P. Identification of two human dimethylarginine dimethylaminohydrolases with distinct tissue distributions and homology with microbial arginine deiminases. Biochem J 1999; 343 Pt 1:209– 214. 59. Wilcox JN, Subramanian RR, Sundell CL, Tracey WR, Pollock JS, Harrison DG, Marsden PA. Expression of multiple isoforms of nitric oxide synthase in normal and atherosclerotic vessels. Arterioscler Thromb Vasc Biol 1997; 17(11):2479–2488. 60. Lytle BW, Loop FD, Cosgrove DM, Ratliff NB, Easley K, Taylor PC. Long-term (5 to 12 years) serial studies of internal mammary artery and saphenous vein coronary bypass grafts. J Thorac Cardiovasc Surg 1985; 89(2):248–258. 61. Luscher TF, Diederich D, Siebenmann R, Lehmann K, Stulz P, von Segesser L, Yang ZH, Turina M, Gradel E, Weber E, et al. Difference between endothelium-dependent relaxation in arterial and in venous coronary bypass grafts. N Engl J Med 1988; 319(8):462–467. 62. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ. Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med 1987; 316(22): 1371– 1375. 63. Hertzer NR. Basic data concerning associated coronary artery disease in peripheral vascular patients. Ann Vasc Surg 1987; 1:616–620. 64. Ness J, Aronow WS. Prevalence of coexistence of coronary artery disease, ischemic stroke, and peripheral arterial disease in older persons, mean age 80 years in an academic hospital-based geriatrics practice. J Am Geriatr Soc 1999; 47:1255–1256. 65. Wilt TJ, Davis BR, Meyers DG, Rouleau JL, Sacks FM. Prevalence and correlates of symptomatic peripheral atherosclerosis in individuals with coronary heart disease and cholesterol levels less than 240 mg/dL: baseline results from Cholesterol and Recurrent Events (CARE) study. Angiology 1996; 47:533–541.

The diversity of vascular disease

381

66. Violi F, Criqui M, Longoni A, Castiglioni C. Relation between risk factors and cardiovascular complications in patients with peripheral vascular disease. Results from the A.D.E.P. study. Atherosclerosis 1996; 120(1–2):25–35. 67. Cooke JP. The pathophysiology of peripheral arterial disease: rational targets for drug intervention. Vasc Med 1997; 2(3):227–230. 68. McDermott MM, Mehta S, Greenland P. Leg symptoms in peripheral arterial disease: associated clinical characteristics and functional impairment. JAMA 2001; 286(13):1599–1606. 69. Fuster V, Corti R, Badimon JJ. Therapeutic targets for the treatment of atherothrombosis in the new millennium—clinical frontiers in atherosclerosis research. Circ J 2002; 66(9):783–789. 70. Rauch U, Osende JI, Fuster V, Badimon JJ, Fayad Z, Chesebro JH. Thrombus formation on atherosclerotic plaques: pathogenesis and clinical consequences. Ann Intern Med 2001; 134(3):224–238. 71. Schnieder FA, Comerota AJ. Intermittent claudication: magnitude of the problem, patient evaluation, and therapeutic strategies. Am J Cardiol 2001; 87(suppl):30–130. 72. Dieter RS, Chu WW, Pacanowski JP, McBride PE, Tanke TE. The significance of lower extremity peripheral arterial disease. Clin Cardiol 2002; 25:3–10. 73. Mallat Z, Hugel B, Ohan J, Leseche G, Freyssinet J, Tedgui A. Shed membrane micro-particles with procoagulant potential in human atherosclerotic plaques: a role for apoptosis in plaque thrombogenicity. Circulation 1999; 99:348–353. 74. Badimon JJ, Lettino M, Toschi V, et al. Local inhibition of tissue factor reduces the thrombogenicity of disrupted human atherosclerotic plaques: effects of TFPI on plaque thrombogenicity under flow conditions. Circulation 1999; 14:1780–1787. 75. Luscher TF, Cooke JP, Houston DS, Neves RJ, Vanhoutte PM. Endothelium-dependent relaxation in human arteries. Mayo Clin Proc 1987; 62(7):601–606. 76. Cohen RA, Shepherd JT, Vanhoutte PM. Inhibitory role of the endothelium in the response of isolated coronary arteries to platelets. Science 1983; 221(4607):273–274. 77. Vanhoutte PM, Miller VM. Heterogeneity of endothelium-dependent responses in mammalian blood vessels. J Cardiovasc Pharmacol 1985; 7(suppl 3):S12–S23. 78. Kunkel EJ, Butcher EC. Chemokines and the tissue-specific migration of lymphocytes. Immunity 2002; 16(1):1–4. 79. Geary RL, Wong JM, Rossini A, Schwartz SM, Adams LD. Expression profiling identifies 147 genes contributing to a unique primate neointimal smooth muscle cell phenotype. Arterioscler Thromb Vasc Biol 2002; 22(12):2010–2016. 80. Wilson PWF, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998; 97:1837–1847. 81. Saito I, Folsom AR, Brancati FL, Duncan BB, Chambless LE, McGovern PG. Nontraditional risk factors for coronary heart disease incidence among persons with diabetes: the Atherosclerosis Risk in Communities (ARIC) Study. Ann Intern Med 2000; 133(2):81–91. 82. Breslow JL. Genetics of lipoprotein abnormalities associated with coronary artery disease susceptibility. Annu Rev Genet 2000; 34:233–254. 83. Leeson CP, Hingorani AD, Mullen MJ, Jeerooburkhan N, Kattenhorn M, Cole TJ, Muller DP, Lucas A, Humphries SE, Deanfield JE. Glu298Asp endothelial nitric oxide synthase gene polymorphism interacts with environmental and dietary factors to influence endothelial function. Circ Res 2002; 90(11):1153–1158. 84. Myers RH, Kiely DK, Cupples LA, Kannel WB. Parental history is an independent risk factor for coronary artery disease: the Framingham study. Am Heart J 1990; 120(4): 963–969. 85. Sorensen, TI, Nielsen GG, Andersen PK, Teasdale TW. Genetic and environmental influences on premature death in adult adoptees. N Engl J Med 1988; 318:727–732; Annu Rev Genet 2000; 34:233–254. 86. Hobbs HH, Brown MS, Goldstein JL. Molecular genetics of the LDL receptor gene in familial hypercholesterolemia. Hum Mutat 1992; 1(6):445–466.

Endothelial cells in health and disease

382

87. Kroon AA, Ajubi N, van Asten WN, Stalenhoef AF. The prevalence of peripheral vascular disease in familial hypercholesterolaemia. J Intern Med 1995; 238(5):451–459. 88. Brugger D, Schuster H, Zollner N. Familial hypercholesterolemia and familial defective apolipoprotein B-100: comparison of the phenotypic expression In 116 cases. Eur J Med Res 1996; 1(8):383–386. 89. Risch NJ. Searching for genetic determinants in the new millennium. Nature 2000; 405(6788):847–856. 90. Marenberg ME, Risch N, Berkman LF, Floderus B, de Faire U. Genetic susceptibility to death from coronary heart disease in a study of twins. N Engl J Med 1994; 330(15):1041–1046. 91. Shimasaki Y, Yasue H, Yoshimura M, et al. Association of the missense Glu298Asp variant of the endothelial nitric oxide synthase gene with myocardial infarction. J Am Coll Cardiol 1998; 31:1506–1510. 92. Hingorani AD, Liang CF, Fatibene J, et al. A common variant of the endothelial nitric oxide synthase (Glu298 → Asp) is a major risk factor for coronary artery disease in the UK. Circulation 1999; 100:1515–1520. 93. Ye S, Watts GF, Maldalia S, Humphries SE, Henney AM. Preliminary report: genetic variation in the human stromelysin promoter is associated with progression of coronary atherosclerosis. Br Heart J 1995; 73:209–215. 94. Humphries SE, Luong LA, Talmud PJ, et al. The 5A/6A polymorphism in the promoter of the stromelysin-1 (MMP-3) gene predicts progression of angiographically determined coronary artery disease in men in the LOCAT gemfibrozil study. Lipid Coronary Angiography Trial. Atherosclerosis 1998; 139:49–56. 95. Terashima M, Akita H, Kanazawa K, et al. Stromelysin promoter 5A/6A polymorphism is associated with acute myocardial infarction. Circulation 1999; 99:2717–2719. 96. Samani NJ, Thompson JR, O’Toole L, Channer K, Woods KL. A meta-analysis of the association of the deletion allele of the angiotensin-converting enzyme gene with myocardial infarction. Circulation 1996; 94:708–712. 97. Agerholm-Larsen B, Nordestgaard BG, Steffensen R, Sorensen TIA, Jensen G, TybjaergHansen A. ACE gene polymorphism: ischemic heart disease and longevity in 10 150 individuals: a case-referent and retrospective cohort study based on the Copenhagen City Heart Study. Circulation 1997; 95:2358–2367. 98. Keavney B, McKenzie C, Parish S, et al. Large-scale test of hypothesised associations between the angiotensin-converting-enzyme insertion/deletion polymorphism and myocardial infarction about 5000 cases and 6000 controls. Lancet 2000; 355:434–442. 99. Risch N, Merikangas K. The future of genetic studies of complex human diseases. Science 1996; 273:1516–1517. 100. Mann MJ, Gibbons GH, Tsao PS, von der Leyen H, Cooke JP, Buitrago R, Kernoff R, Dzau VJ. Cell cycle inhibition preserves endothelial function in genetically engineered rabbit vein grafts. J Clin Invest 1997; 99(6):1295–1301. 101. Mann MJ, Gibbons GH, Kernoff RS, Diet FP, Tsao PS, Cooke JP, Kaneda Y, Dzau VJ. Genetic engineering of vein grafts resistant to atherosclerosis. Proc Natl Acad Sci USA 1995; 92(10):4502–4506.

19 Molecular Targets of Tumor Vasculature Eleanor B.Carson-Walter Department of Neurosurgery, University of Pittsburgh—Pittsburgh, Pittsburgh, Pennsylvania, U.S.A.

Brad St. Croix Tumor Angiogenesis Laboratory, National Cancer Institute—Frederick, Frederick, Maryland, U.S.A.

1. INTRODUCTION Most solid tumors arise as small nodal growths and utilize preexisting surrounding local vasculature to acquire oxygen and nutrients. As tumors continue to grow, they outsize the available blood supply and become dormant, potentially remaining so for years. Further expansion is believed to require acquisition of a proangiogenic phenotype, referred to as the “angiogenic switch” (1). The hypothesis that tumors are angiogenesis-dependent was put forth by Folkman (2) over 30 years ago. Although at first greeted with some skepticism, a large body of evidence now supports this idea, including recent genetic experiments (3). The dependence of tumor cells on their vasculature led to the idea of targeting tumor endothelium as a novel anticancer strategy, an approach that has generated much excitement amongst cancer researchers and clinicians. Targeting tumor vessels is likely to have several advantages over conventional cytotoxic regimens aimed at the tumor cells themselves. First, endothelial cells are easily accessible through the bloodstream, eliminating many of the pharmacokinetic challenges associated with targeting tumor cells. Second, unlike tumor cells, endothelial cells are genetically stable, rendering them less likely to accumulate mutations which would make them resistant to therapy (4). Third, each endothelial cell is thought to support the growth of many tumor cells, so there is likely to be a substantial bystander effect. Fourth, although “cancer” is not a single disease, angiogenesis appears to be a common requirement across organ sites; thus neovasculature-targeted therapies may be applicable to a wide variety of solid tumor types. Fifth, because an angiogenic phenotype is thought to be rate limiting for metastasis, preventing angiogenesis may limit the ability of cancer cells to spread. Finally, side effects of angiogenesis-targeted therapy may be relatively limited, as angiogenesis is a tightly regulated process. Most physiologic angiogenesis occurs during embryonic development, active wound healing, or in naturally regenerating tissues such

Endothelial cells in health and disease

384

as the corpus luteum. Outside of these settings, endothelial cells seldom divide. It is primarily for these reasons that the tumor endothelium has become such an attractive target for anticancer therapy. Until recently, it was widely believed that angiogenesis, the formation of new capillaries from preexisting vasculature, was the primary mechanism by which expanding tumors acquired a blood supply. Recent work has shown, however, that circulating endothelial cells originating from the bone marrow can also home to and contribute to the newly formed vasculature through a process known as postnatal vasculogenesis (see Chapters 15 and 16). Although these studies have demonstrated a role for circulating endothelial progenitor cells in tumor angiogenesis (5), the relative contribution of local vs. circulating endothelial cells to newly formed tumor vessels remains unclear. Most adult endothelium is quiescent. In contrast, endothelial cells in tumors divide more frequently and are morphologically distinct from normal vessels. Pericytes which normally form a continuous sheath surrounding endothelial cells of normal capillaries are often loosely attached or even absent from tumor vessels (6). This, in conjunction with an altered basement membrane, makes the endothelium of tumors weaker than normal capillaries (7). Compared to normal vessels, tumor vessels are often tortuous and dilated, containing erratic branching patterns. Due to their abnormal structure, tumor vessels cannot be readily distinguished as arterioles, venules, or capillaries. In some tumors, such as human colorectal cancer, chords containing clusters of adjacent microvessels can be found. Both solitary vessels and those found in chords often contain abrupt dead ends. The disorganized, irregular vessels of tumors have been elegantly visualized through molecular casting techniques (8) and multiphoton laser scanning microscopy (9), among other approaches (10). Many of the observed morphological abnormalities of tumor vessels are the result of tumor and endothelial cell interactions within the local microenvironment. Tumor vessels are more permeable than their normal counterparts, leading to leakage of plasma. Dvorak (11) noticed over 15 years ago that tumors have an increased propensity for thrombosis, and likened them to “wounds that never heal.” As a tumor reaches the limits of growth which can be supported by the existing vasculature, it generates an environment of local hypoxia. This hypoxic setting can stimulate hypoxia regulated genes such as the transcription factor HIF-1α which, in turn, activates expression of endothelial growth promoting genes such as vascular endothelial growth factor (VEGF). Mutations leading to activation of oncogenes or loss of tumor suppressor genes can also lead to an increase in the level of tumor-derived VEGF (12). Continued tumor growth in the absence of functional lymphatics eventually leads to high interstitial pressure (13,14). This, along with the pressure that is created when cells continue to divide in a limited space (15), can lead to the collapse of tumor vessels. The resultant impaired blood flow generates increased hypoxia stimulating the production of VEGF and promoting further angiogenesis. All of these morphologic and environmental differences, many of which are not observed during physiological angiogenesis, suggest that tumor endothelium may contain molecules distinct from normal endothelium. Indeed, an intensive search for markers highly restricted to tumor endothelium has begun [see below and Refs. 16–19]. Many molecules, although not necessarily confined to tumor endothelium, have already been identified as key regulators of angiogenesis over the past 20 years. Some of most notable

Molecular targets of tumor vasculature

385

are mentioned briefly here, but readers are referred to the following reviews for a more comprehensive overview (20–23).

2. MOLECULAR MECHANISMS OF TUMOR ANGIOGENESIS Over the past 20 years, a large body of evidence has accumulated supporting a central role for VEGF in neovascularization. Vascular endothelial growth factor was first discovered during the characterization of transplanted tumors in syngeneic animals (24,25). Its importance in vessel formation is highlighted by the fact that loss of even a single allele in mice leads to severe vascular defects and embryonic lethality (26). VEGFA (or VEGF165) is a soluble form of the growth factor and has been the most intensively studied but other matrix-bound isoforms, generated by alternative splicing, also stimulate angiogenesis (27). All of the VEGF isoforms bind to specific cell surface tyrosine kinase receptors (VEGFRs). VEGF-A binds two receptors, VEGFR1 (flt1) and VEGFR2 (KDR in humans and Flk-1 in mice). Although VEGF-A binds VEGFR1 with higher affinity than VEGFR2, the tyrosine kinase domain of VEGFR1 does not seem to be necessary for its role in angiogenesis (28). Hence, VEGFR1 may play more of an ancillary role in angiogenesis, perhaps by regulating the amount of VEGF-A available for stimulation of VEGFR2. VEGFR2, on the other hand, shows a strong phosphorylation of its tyrosine kinase domain in response to VEGF-A. Tumor hypoxia causes an increase in hypoxiainducible transcription factor 1α (HIF-1α), which stimulates increased transcription of VEGF-A. In turn, VEGF-A binds to and activates VEGFR2, causing vasodilation, increased vascular permeability, and migration and proliferation of endothelial cells, thus promoting angiogenesis (29). The importance of VEGF in endothelial biology is also highlighted by the fact that expression of its receptors is highly restricted to endothelium. Early studies suggested that VEGFR2 mRNA was expressed predominantly on tumor endothelium with little, if any, expression on normal endothelium (30,31). However, more recent studies have shown clear expression of VEGFR2 mRNA in the normal quiescent endothelial cells of multiple normal organs and tissues (32,33). The more recent findings are likely the result of an increased sensitivity of the assays used to measure receptor expression. Presumably, a basal level of expression of VEGFR2 on the surface of normal endothelium is necessary to ensure that these cells are able to respond when stimulated by VEGF. Several studies have indicated that VEGF is required for tumor-induced angiogenesis. Antibodies and soluble receptors have been developed to block its function (34,35) and reduced tumorigenicity in the absence of tumor-derived VEGF has been demonstrated (36–38). Similarly, VEGFR has been targeted using antibodies (39) and low molecular weight kinase inhibitors such as SU5416 and SU6668 (40, 41). Phase III clinical trials have now shown that combination of Avastin, an antibody directed towards VEGF, with conventional chemotherapy increases survival of colorectal cancer patients when compared to conventional treatment alone (42). This important finding demonstrates the validity of targeting VEGF and tumor angiogenesis in patients. Activation of the VEGF signal transduction pathway initiates an immediate increase in vascular permeability and vasodilation. This is caused by a loosening of the pericytes

Endothelial cells in health and disease

386

covering the existing vessels and is supported by the binding of angiopoietin-2 (ANG2) to the tyrosine kinase receptor Tie2 (23,43). Under normal conditions, ANG1 binds Tie2, causing endothelial cells to become resistant to the effects of VEGF-A, thereby reducing leakiness and promoting the maturation of vessels (44,45). When ANG2 binds Tie2 and antagonizes the action of ANG1, endothelial cells become “destabilized” and either apoptose in the absence of proangiogenic factors or survive and begin to proliferate in the presence of VEGF (46). Furthermore, high levels of ANG2 alone may act as a survival factor for the activated, immature endothelium (47). Continued VEGF/ANG2 activity stimulates endothelial proliferation and migration within perivascular space and contributes to vessel remodeling and the sprouting of new vessel branches (48). As the tumor continues to grow, new regions of tissue become hypoxic and subsequently lead to the activation of VEGF, exacerbating the cycle. Interventional studies using soluble receptors to block Tie2 activation and signal transduction have demonstrated significant antiangiogenic and antitumor results in mice (49). Another family of molecules that has been implicated in angiogenesis is the integrins. Integrins are cell surface receptors that control cell–extracellular matrix adhesion and cellular migration. An initial report on αvβ3 and αvβ5 integrins suggested that these receptors may play a role in the survival of endothelial cells since antagonists of these molecules inhibited tumor angiogenesis (50,51). The αvβ3 integrin is a potential target of the naturally occurring antiangiogenic agents, endostatin, tumstatin, and angiostatin, although evidence also suggests that endostatin may exert its inhibitory effects through interaction with αvβ1 and α5β1 (52–56). However, mice lacking all alpha v integrins or beta 3 integrins are viable and undergo normal or even enhanced angiogenesis when challenged with tumors (57). This suggests that the normal function of these integrins may be to repress angiogenesis and that their role as proangiogenic factors needs to be reevaluated (58). Most of the integrins implicated in angiogenesis are expressed at readily detectable levels in normal endothelial cells and other cell types in the body. A possible exception is the integrin alphal, which, when examined in the current SAGE database, appears to be expressed predominantly in tumor endothelium (Table 1). Interestingly, integrin alphal knockout mice display reduced vascularization and tumorigenesis supporting a role for this integrin in tumor angiogenesis (59).

3. TUMOR TARGETING APPROACHES Two approaches can be envisioned to target the tumor vasculature (60). The first relies on the use of antiangiogenic agents. Antiangiogenic agents prevent further expansion of tumor vasculature by inhibiting the growth of new vessels. In some instances, newly formed tumor vessels may also be sensitive to antiangiogenic agents, a phenomenon referred to as “capillary drop out.” These agents, which are usually cytostatic or restricted in their toxicity to tumor vessels, tend to have relatively mild side effects, and may need to be administered continually. Virtually all clinical drugs currently used to target tumor endothelium fall into this category. Examples include endostatin, angiostatin, agents that target VEGF such as Avastin or the VEGF-Trap, VEGFR and PDGFR inhibitors such as SU6668, the EGFR inhibitor Erbitux, and others. Because the targets of most of these agents are present on normal vessels or other normal cell types, these agents, by

Molecular targets of tumor vasculature

387

necessity, have been designed to inhibit the physiological function of their target proteins and do not simply kill all target-expressing cells. Several antiangiogenic drugs are in various phases of clinical trials for a number of angiogenesis related diseases including cancer. Some of these studies are beginning to demonstrate efficacy (61). The second approach to inhibiting tumor vessels involves the use of drugs generally referred to as vascular targeting agents. These agents are designed to be cytotoxic to tumor endothelial cells much like conventional chemotherapeutic

Table 1 Transcripts Preferentially Expressed in Endothelial Cells from Human Colorectal Cancers Compared to Endothelium from Normal Colonic Mucosa No.

Tag Sequencea

Gene Description

NECs

TECs

Refs.b

1

GGGGCTGCCCA

TEM1 (endosialin)

0

28

17,32,85

2

GATCTCCGTGT

TEM2

0

25

17,32

3

CATTTTTATCT

TEM3

0

23

17,32

4

CTTTCTTTGAG

REIC (Dkk-3)

0

22

5

TATTAACTCTC

TEM4

0

21

17,32

6

CAGGAGACCCC

MMP-11(stromelysin 3)

0

16

82

7

GGAAATGTCAA

MMP-2 (gelatinase A)

1

31

80,81

8

CCTGGTTCAGT

HeyL

0

15

9

TTTTTAAGAAC

TEM5

0

14

17,32

10

TTTGGTTTTCC

Collagen, type I, alpha 2, transcript Ac

5

139

95,96

11

ATTTTGTATGA

Nidogen (entactin)

0

13

97

12

ACTTTAGATGG

Collagen, type VI, alpha 3

1

23

98,99

13

GAGTGAGACCC

Thy-1 cell surface antigen

3

63

100

14

GTACACACACC

Cystatin S/cystatin SA

0

10

15

CCACAGGGGAT

Collagen, type III, alpha 1

2

38

101–103

16

TTAAAAGTCAC

TEM6

1

19

17,32

17

ACAGACTGTTA

TEM7

4

74

17,32

18

CCACTGCAACC

Unknown

1

18

19

CTATAGGAGAC

TEM8 (ATR)

1

18

17,32,91

20

GTTCCACAGAA

Collagen, type I, alpha 2, transcript Bc

0

9

95,96

Endothelial cells in health and disease

388

21

TACCACCTCCC

Unknown

0

9

22

GCCCTTTCTCT

TEM9 (endo 180 lectin)

1

17

23

TTAAATAGCAC

Collagen, type I, alpha 1

2

33

24

AGACATACTGA

Tensin

1

16

25

TCCCCCAGGAG

Bone morphogenetic protein 1

1

16

26

AGCCCAAAGTG

Unknown

0

8

27

ACTACCATAAC

Slit (Drosophila) homolog 3 (MEGF5)

0

8

28

TACAAATCGTT

KIAA0672 gene product

0

8

29

TTGGGTGAAAA

PRL-3

0

8

30

CATTATCCAAA

Integrin, alpha 1

0

8

59

31

AGAAACCACGG

Collagen, type IV, alpha 1

0

8

105

32

ACCAAAACCAC

Unknown

0

8

33

TGAAATAAAC

Unknown

0

8

34

TTTGGTTTCC

Unknown

1

15

35

GTGGAGACGGA

Unknown (FLJ40955)

1

15

36

TTTGTGTTGTA

Collagen, typeXII, alpha 1

1

14

37

TTATGTTTAAT

Lumican

3

39

38

TGGAAATGACC

Collagen, type I, alpha 1

15

179

39

TGCCACACAGT

Transforming growth factor, beta 3

1

13

40

GATGAGGAGAC

Collagen, type I, alpha 2, transcript Cc

3

35

41

ATCAAAGGTTT

Unknown(LOC169611)

2

23

42

AGTCACATAGT

Unknown

1

11

43

TTCGGTTGGTC

Unknown

4

45

44

CCCCACACGGG

Unknown(FLJ11190)

2

21

45

GGCTTGCCTTT

Unknown

1

10

46

ATCCCTTCCCG

Peanut-like protein 1 (CDCrel-1)

1

10

a

95,104

104

95,96

The top 46 tags with the highest tumor EC (T-ECs) to normal EC (N-ECs) tag ratios are listed in descending order. b Published reports supporting a role for the gene in angiogenesis. c Multiple tags for this gene are due to alternative polyadenylation sites. (For details, see Ref. 17.)

Molecular targets of tumor vasculature

389

agents. Because of their toxic nature, these agents must be targeted specifically to tumor endothelium. Historically, a lack of tumor-specific endothelial markers has been a major limitation for implementing this type of an approach. The most widely studied group of vascular targeting agents are toxins conjugated to antibodies that recognize tumor endothelium. Another interesting approach involves the delivery of tissue factor/antibody conjugates to tumor endothelium (62,63). Induction of coagulation by tissue factor has been shown to induce thrombosis of tumor vessels, and was originally demonstrated using mice engineered to overexpress MHC class II receptors on tumor endothelium (64). This strategy has now been tested on several transplantable tumor models against the endogenous endothelial cell surface receptor VCAM-1, and the ED-B domain of fibronectin (65,66). The agents appear to work very well in these studies, even causing tumor regression. Although the ED-B domain has been reported to be overexpressed in tumor endothelium, the results obtained with VCAM are somewhat surprising given that VCAM-1 is readily detected on normal endothelium (67). One possible explanation for the lack of toxicity to normal tissues is that phosphatidylserine, found to be overexpressed on tumor vessels, is also required for tissue factor-mediated thrombosis (68,69). This highlights an important point—even if a single molecule with the desired specificity for tumor endothelium cannot be found, it may be possible to design therapies that require two or more targets for activation, the combined expression of which is confined to tumor endothelium. Other endothelial receptors, including VEGFR and endoglin, have also been targeted by active immunization or delivery of immunotoxins or ligand-toxin conjugates (70–74). Again, surprisingly little toxicity to normal tissues was observed in these studies despite readily detectable levels of receptor expression in normal tissues (32,33). The reasons for the lack of normal tissue toxicity are unclear. However, it is conceivable that a modest upregulation of these markers does occur on tumor endothelium, and this is sufficient to offer a therapeutic window. Compared to normal vessels, the abnormal vessels in tumors may also be more prone to undergoing apoptosis. In addition, rodents have an exceptionally high metabolic rate which may enable them to withstand the adverse effects caused by these toxic agents. Although these results are encouraging, there is still much room for improvement, and it seems likely that markers more restricted to tumor endothelium will be necessary before such an approach can be utilized clinically.

4. NOVEL MARKERS OF TUMOR ENDOTHELIAL CELLS New molecular markers of tumor endothelium would be useful not only for vascular targeting approaches, but could also provide new targets for antiangiogenic therapy and provide insight into mechanisms underlying angiogenesis. An ideal strategy for identifying such markers might entail a systematic comparison of cell surface proteins expressed on human normal- or tumor-derived endothelium. Although this goal still remains elusive, Ruoslahti and coworkers used phage display to identify peptides that can home specifically to tumor endothelium [see Ref. (75), for a full review of this area see Chapter 3]. These studies demonstrated the existence of specific molecular addresses on tumor endothelium (76). Unfortunately, the low affinity of peptides for antigens on the

Endothelial cells in health and disease

390

cell surface has made the identification of novel targets using this type of an approach difficult. The use of antibody phage display (77) or cDNA phage display (78) may help to overcome some of these limitations. Nevertheless, this approach has led to the identification of aminopeptidase N as a potential target of tumor endothelium (79). As an alternative approach to identifying tumor endothelial markers, we took advantage of recent advances in gene expression technology (17). First, however, it was necessary to develop new techniques to purify endothelial cells from tissues, since endothelial cells lining blood vessels represent a minor fraction of the total number of cells in a given population. To overcome this obstacle, we developed a protocol to isolate human endothelial cells from either normal colonic mucosa or colorectal cancer based on the cell surface marker CD146 (P1H12). Although we have found this marker to cross react with smooth muscle in certain human tissues, in colon tissue CD146 appears to be highly specific for endothelium. By using magnetic beads coupled to CD146, we were able to selectively purify endothelial cell away from other cell types following enzymatic dissociation of tissues (for a detailed protocol, see: http://www.sagenet.org/angio/index.html). To measure gene expression in the purified endothelial cells, we utilized Serial Analysis of Gene Expression (SAGE). SAGE is a technique designed to identify and quantify mRNA transcripts on the basis of a unique sequence “tag” derived from the 3′ terminus of the molecule. One advantage of SAGE is its quantitative nature; each tag represents an individual mRNA molecule and the frequency of a given tag allows measurement of that gene’s expression level within the isolated population of cells. SAGE tags provide an unbiased account of gene expression, independent of preexisting databases, allowing novel genes to be discovered and cloned. This technique can be used quantitatively on as few as 50,000 cells directly isolated from human tissues. Although genes found to be differentially expressed by this strategy ultimately need to be confirmed at the protein level, focusing on markers that demonstrate the greatest tag differential (i.e. mRNA on or off) increases the probability that differences in expression will be maintained at the protein level. Using SAGE technology, we identified 93 pan-endothelial markers (called PEMs) selectively expressed in endothelium derived from both normal and cancerous colorectal tissues (17). We also identified 46 tumor endothelial markers (TEMs) which were expressed at significantly higher levels (>10-fold) in tumor- vs. normal-derived endothelium (see Table 1) and 33 normal endothelial markers (NEMs) elevated in normal- vs. tumor-derived endothelium. The expression pattern of the TEMs was confirmed by RT-PCR and in situ hybridization. Several of the previously characterized genes in Table 1 have been shown to be involved with angiogenesis through gene knockout studies. For example, MMP2-deficient mice displayed reduced angiogenesis and tumor growth when implanted with various tumor types (80,81). Also, a lack of hostderived stromelysin-3 (MMP11) in knockout mice was shown to inhibit the formation of DMBA-induced carcinomas (82). Similarly, loss of integrin alphal caused reduced tumor growth and vascularization (59). Several of the other markers on the list have also been correlated with angiogenesis (see Table 1). The majority of the SAGE tags identified in tumor endothelium correspond to previously uncharacterized genes. In an attempt to identify these new potential targets, nine of the most differentially expressed novel genes, designated TEM1–TEM9, were cloned and sequenced (32). Analysis of the sequence revealed hydrophobic domains in

Molecular targets of tumor vasculature

391

four of the TEMs, TEM1, 5, 7 and 8, suggesting that these genes encoded cell surface transmembrane proteins. Cell surface proteins are attractive since they are directly accessible via the circulation, facilitating the therapeutic and diagnostic targeting of tumor vessels. Furthermore, it was of interest to see if the expression pattern of these TEMs was conserved across species, specifically between human and mouse, since model systems will be critical for the development and testing of new strategies aimed at targeting tumor vessels. Mouse orthologs of each of the four cell surface TEMs, mTEM1, mTEM5, mTEM7 and mTEM8, were cloned and sequenced (32). The mRNA expression patterns of both the human and mouse TEMs were examined by in situ hybridization. TEM1, 5, and 8 were preferentially expressed on human tumor endothelium and were only rarely detected on normal, non-proliferative endothelium. Importantly, antibodies against cell surface TEMs have confirmed their high level of expression in tumor endothelium at the protein level (see Ref. 83 and A.Nanda, personal communication). The mouse orthologs for these three genes were expressed in the developing embryonic endothelium as well as in syngeneic and transplanted human tumors grown in adult mice (for mTEM1, see Fig. 1). Interestingly, a unique pattern of expression for TEM8 emerged when these markers where evaluated for expression during normal physiological angiogenesis in human tissues. While TEM1 and TEM5 mRNA expression was readily detected in healing wounds and corpus luteum, TEM8 was undetectable in these tissues, suggesting its expression may be more restricted to tumor endothelium. TEM7 also displayed a unique mRNA expression pattern in mice. Unlike its human counterpart, mTEM7 was not detected in tumor endothelial cells but was instead found in the Purkinje cells of the cerebellum. What is known about each of these cell surface TEMs? TEM1 encodes a 757 amino acid, type 1 transmembrane protein. The long extracellular domain contains three EGF repeats, a C-lectin-like carbohydrate recognition domain and weak

Endothelial cells in health and disease

392

Figure 1 mTEM1 in B16 mouse melanoma tumor endothelial cells. Expression of mTEM1 mRNA was assessed using a highly sensitive nonradioactive in situ hybridization assay. Note that the staining (red) is localized to the endothelial cells. The section was counterstained with hematoxylin. (From Ref. 32.) homology to a Sushi/SCR/CCP domain. Recently, Christian et al. (84) identified TEM1 as the antigen detected by the FB5 antibody. FB5 was originally described in 1992 by Lloyd Old’s group (85). Using immunohistochemistry, the group demonstrated FB5 positive staining of tumor endothelium in 85 immunoreactive tumors of various histological origin, but not in a panel of normal tissues. The antigen recognized by the FB5 antibody, which the authors called endosialin, was found to have an apparent molecular weight of 165 kD due to abundant O-glycosylation. Additional experiments demonstrated that FB5 was endocytosed by TEM1/endosialin expressing endothelial cells. Capitalizing on this selective uptake provides an attractive strategy for targeting tumor endothelium. More recent SAGE studies of brain tumor endothelium demonstrated that TEM1 was upregulated sevenfold in endothelial cells isolated from glioblastoma multiform as compared to normal brain endothelium (S.L. Madden and K.A.Walter, personal

Molecular targets of tumor vasculature

393

communication). These data provide independent support for a conserved role for TEM1 in tumor angiogenesis. TEM5 was predicted to encode a seven-pass transmembrane protein with homology to the class II family of G-protein coupled receptors (GPCRs). The extracellular domain contains four leucine rich repeats, one carboxy-terminal type leucine rich repeat, an immunoglobulin-type domain, and a hormone-receptor domain. A putative GPCR proteolysis site is located just prior to the first transmembrane spanning region. This site, characteristic of class II GPCRs, may be required for endogenous proteolysis. Other class II GPCRs bind small peptide ligands such as secretin and calcitonin, thereby activating adenyl cyclase and inositol phosphate signaling cascades (86). Similarly, TEM5 may bind a ligand present in the blood-stream and transmit signals to the interior of the endothelial cell. If this is so, it may be possible to target TEM5 with a compound that mimics the binding action of its native ligand. Indeed, many GPCR family members have been successfully targeted by the pharmaceutical industry (87). TEM7 encodes a type I transmembrane protein. The extracellular region of TEM7 contains a plexin-like domain and weak homology to nidogen, an extracellular matrix protein. A homologue to TEM7, called TEM7R, was also cloned and characterized. Human TEM7, TEM7R, and mouse TEM7R (mTEM7R) all showed preferential expression in activated tumor endothelium, although mTEM7 did not. The reason for this discordance remains to be explained, but it is possible that the role of these genes diverged over time. While this difference in gene expression suggests that mTEM7 may not provide a useful resource for mouse models of tumor angiogenesis, TEM7 may still prove to be a valuable target in human disease. This discrepancy underscores the importance of improving the animal models used in clinical testing, and may help to explain why impressive antitumor results obtained in rodents often predate disappointing human cancer trials. TEM8 also encodes a type 1 transmembrane protein containing a large cytoplasmic tail with at least seven potential phosphorylation sites, supporting the hypothesis that it is a cell surface signal transduction molecule. The extracellular domain of TEM8 includes a vWF-A domain containing a metal ion-dependent adhesion motif. This domain is similar to that of the integrin αD protein, which interacts with vascular cell adhesion molecule. To date, TEM8 is the only TEM analyzed that was not expressed in the corpus luteum (see Table 2). Recently developed monoclonal antibodies against TEM8 also fail to stain corpus lutuem, while staining of tumor endothelium is readily observed (83). This remarkable expression pattern means that TEM8 might provide a valuable clinical target in adults highly specific

Table 2 Detection of TEM Transcripts in Various Tumor Types and Tissues by In Situ Hybridizationa Tumor Endothelium Humans

Miceb

C Li Lg B16 HCT Tumor Cells

Endothelial Cells Wounds

Corpus Luteum

Embryosa

Endothelial cells in health and disease

394

TEM1

+

+

+

+

+

−

+

+

+

TEM4

+

+

+

ND

ND

−

+

+

ND

TEM5

+

+

+

+

+

+

ND

+

+

TEM7

+

+

+

−

−

−

+/−

+

+/−

TEM7R

+

ND

+

−

ND

ND

+

TEM8

+

+

+

+

+

+

+/−

−

+

TEM9

+

+

+

ND

ND

−

ND

+

ND

ND ND

a

+ indicates the presence of strong positive staining of vessels by in situ hybridization;— indicates an undetectable signal by in situ hybridization; +/− indicates a very weak signal in a limited number of vessels by in situ hybridization; ND indicates not determined. b To perform in situ hybridization on mouse tissues, riboprobes were generated using sequences from the mouse orthologus of each of the indicated TEMs. C: colorectal tumor; Li: liver metastasis of colorectal cancer; Lg: lung cancer; B16: mouse melanoma tumor; HCT: HCT1 16 human colon carcinoma xenograph. to tumor angiogenesis with limited cross-reactivity to sites of physiologic angiogenesis. In certain tumor types, particularly melanoma, TEM8 has been found to be overexpressed in the tumor cells themselves (BSC, unpublished observations and Refs. 32,88). Although the mechanisms underlying this dramatic increase in select tumors remain to be determined, these results suggests that TEM8 targeted therapy could hit both the tumor cell and endothelial cell compartments simultaneously. Recently, data from work in an unrelated field have unexpectedly increased interest in TEM8 and its role in tumor endothelium. A timely report from Bradley and Young (89) identified TEM8 as a receptor for the anthrax toxin protective antigen. Protective antigen itself is non-toxic, but is the subunit responsible for binding the tripartite anthrax toxin complex to receptor bearing cells. The two other components of anthrax toxin are enzymatic proteins, called lethal factor (LF) and edema factor (EF), and are responsible for eliciting cellular toxicity. Using a genetic complementation screen, the receptor for protective antigen, which the authors called the anthrax toxin receptor (ATR), was found to be identical to the first 364 amino acids of TEM8. The C-termini were divergent, presumably as a result of alternative splicing. Nevertheless, anthrax toxin binds full length TEM8 with the same affinity as the shorter variant (90). Importantly TEM8/ATR was able to bind to protective antigen and restore sensitivity to the mutagenized, anthrax toxin-resistant cells. Furthermore, a mutant protective antigen which did not bind to TEM8/ATR could not infect cells. The receptor/ligand relationship between TEM8 and protective antigen clarified an additional intriguing experiment described earlier the same year. Duesbery et al. (91) had reported that anthrax lethal toxin (LeTx), comprised of lethal factor (LF) plus protective antigen, suppressed H-ras-mediated transformation and inhibited tumor growth and angiogenesis. In vitro, treatment with LeTx was cytostatic, caused H-ras transformed NIH3T3 cells to revert to a non-transformed, flattened morphology and prevented growth in soft agar. In vivo, ras-transformed NIH3T3 tumors were grown in both flanks of nude mice and were injected with LeTx on one side only. Importantly, LeTx was administered

Molecular targets of tumor vasculature

395

at doses that did not appear to have a detrimental effect on the health of the animals. Over a period of 20 days, LeTx-treated tumors demonstrated a significant growth delay, and in some cases completely regressed. Remarkably, the treated tumors contained extensive necrosis, an unexpected result since only a cytostatic effect had been observed in vitro. The effect was systemic as there was no significant size difference between the injected tumors and the uninjected, contralateral tumors. The pale yellow color of the LeTx treated tumors stimulated the investigators to analyze blood vessel numbers using the endothelial marker CD31. The histology demonstrated “dramatically reduced” staining and led the authors to propose a possible antiangiogenic mechanism (91). The subsequent identification of TEM8 as the anthrax toxin receptor helps to explain this previously enigmatic result and suggests that LeTx treatment may have potential therapeutic utility. Two subsequent studies have expanded this initial observation to a variety of tumor types (88,92). Further examination of the interaction between TEM8/ATR and protective antigen has demonstrated that the cytoplasmic tail of TEM8 is dispensable for protective antigen binding, processing, and cellular infection (90). Cells expressing soluble TEM8 did not bind protective antigen, but protective antigen binding could be restored by expressing extracellular TEM8 anchored by an artificial GPI tail on the cell surface. These results suggest that the anthrax protective antigen exploits the TEM8 receptor for cellular infection, but it may not utilize the TEM8 signaling pathway per se. This reasoning implicates another, as yet to be identified, protein as the functional ligand for TEM8. One interesting candidate for such a ligand is Collagen VI alpha3, an extracellular matrix molecule recently shown to bind TEM8 (83). The interaction was confirmed by coimmunoprecipitation, and the binding region mapped to the carboxyl-terminal C5 domain of collagen VI. Interestingly Collagen VI alpha3 was also one of most differentially expressed TEMs identified by SAGE (see Table 1, No. 12). Its upregulation in tumor endothelium was validated by in situ hybridization and was strikingly similar to that of TEM8. Recent studies have shown that the C5 domain of collagen VI alpha 3 is incorporated into newly formed collagen VI fibrils, but soon after secretion is cleaved and is not present in the mature collagen VI containing matrix (93). The C5 domain has been found in the pericellular region as well as the cytoplasm of cells actively synthesizing collagen VI (94). Thus, a potential physiological role of TEM8 is in the binding and removal of the C5 domain which may be necessary for the correct processing of newly formed collagen fibrils during angiogenesis.

5. CONCLUSIONS The therapeutic potential of vascular targeted therapy has generated substantial excitement in the field of cancer biology. In order to bring this theoretical potential to clinical fruition, it is critical to understand the biology of tumor endothelial cells. Signaling pathways involving VEGF and VEGFR2 play important roles in stimulating and maintaining newly formed tumor vessels. The newly formed vessels are capable of sustaining tumor growth, but are structurally and functionally abnormal. Recognition of an increasing number of gene products that correlate with the distinct phenotype of tumor endothelium offers potential new targets for both antiangiogenic and vascular targeting

Endothelial cells in health and disease

396

approaches. Promising new targets include the recently identified cell surface tumor endothelial markers (TEMs). These genes and other like them may also be exploited as clinical markers of tumor stage or as prognostic indicators. The identification and characterization of molecular markers preferentially expressed on tumor vasculature is still at an early stage. However, the preliminary results are encouraging and suggest that there may be many more targets of angiogenesis than previously envisioned. Continued efforts in this area will eventually lead to a more comprehensive understanding of the mechanisms underlying angiogenesis. Ultimately, this should lead to the development of new rationally designed weapons for combating cancer and other angiogenesis-dependent diseases.

REFERENCES 1. Bergers G, Benjamin LE. Tumorigenesis and the angiogenic switch. Nat Rev Cancer 2003; 3:401–410. 2. Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J Med 1971; 285: 1182–1186. 3. Folkman J. Angiogenesis. In: Braunwald E, Fauci AS, Kasper DL, Hauser SL, Longo DL, Jameson JL, eds. Harrison’s Textbook of Internal Medicine. New York: McGraw-Hill, 2001:517–530. 4. Kerbel RS. Inhibition of tumor angiogenesis as a strategy to circumvent acquired resistance to anti-cancer therapeutic agents. Bioessays 1991; 13:31–36. 5. Lyden D, Hattori K, Dias S, Costa C, Blaikie P, Butros L, Chadburn A, Heissig B, Marks W, Witte L, Wu Y, Hicklin D, Zhu Z, Hackett NR, Crystal RG, Moore MA, Hajjar KA, Manova K, Benezra R, Rafii S. Impaired recruitment of bone-marrow-derived endothelial and hematopoietic precursor cells blocks tumor angiogenesis and growth. Nat Med 2001; 7:1194– 1201. 6. Morikawa S, Baluk P, Kaidoh T, Haskell A, Jain RK, McDonald DM. Abnormalities in pericytes on blood vessels and endothelial sprouts in tumors. Am J Pathol 2002; 160:985–1000. 7. Kalluri R. Angiogenesis: basement membranes: structure, assembly and role in tumour angiogenesis. Nat Rev Cancer 2003; 3A22–433. 8. Skinner SA, Frydman GM, O’Brien PE. Microvascular structure of benign and malignant tumors of the colon in humans. Dig Dis Sci 1995; 40:373–384. 9. Brown EB, Campbell RB, Tsuzuki Y, Xu L, Carmeliet P, Fukumura D, Jain RK. In vivo measurement of gene expression, angiogenesis and physiological function in tumors using multiphoton laser scanning microscopy. Nat Med 2001; 7:864–868. 10. McDonald DM, Choyke PL. Imaging of angiogenesis: from microscope to clinic. Nat Med 2003; 9:713–725. 11. Dvorak HF. Tumors: wounds that do not heal. Similarities between tumor stroma generation and wound healing. N Engl J Med 1986; 315:1650–1659. 12. Rak J, Yu JL, Kerbel RS, Coomber BL. What do oncogenic mutations have to do with angiogenesis/vascular dependence of tumors?. Cancer Res 2002; 62:1931–1934. 13. Boucher Y, Jain RK. Microvascular pressure is the principal driving force for interstitial hypertension in solid tumors: implications for vascular collapse. Cancer Res 1992; 52:5110– 5114. 14. Padera TP, Kadambi A, di Tomaso E, Carreira CM, Brown EB, Boucher Y, Choi NC, Mathisen D, Wain J, Mark EJ, Munn LL, Jain RK. Lymphatic metastasis in the absence of functional intratumor lymphatics. Science 2002; 296:1883–1886. 15. Helmlinger G, Netti PA, Lichtenbeld HC, Melder RJ, Jain RK. Solid stress inhibits the growth of multicellular tumor spheroids. Nat Biotechnol 1997; 15:778–783.

Molecular targets of tumor vasculature

397

16. Ruoslahti E. Targeting tumor vasculature with homing peptides from phage display. Semin Cancer Biol 2000; 10:435–442. 17. St Croix B, Rago C, Velculescu V, Traverso G, Romans KE, Montgomery E, Lal A, Riggins GJ, Lengauer C, Vogelstein B, Kinzler KW. Genes expressed in human tumor endothelium. Science 2000; 289:1197–1202. 18. Huminiecki L, Bicknell R. In silico cloning of novel endothelial-specific genes. Genome Res 2000; 10:1796–1806. 19. Shih SC, Robinson GS, Perruzzi CA, Calvo A, Desai K, Green JE, Ali IU, Smith LE, Senger DR. Molecular profiling of angiogenesis markers. Am J Pathol 2002; 161:35–41. 20. Carmeliet P. Angiogenesis in health and disease. Nat Med 2003; 9:653–660. 21. Jain RK. Molecular regulation of vessel maturation. Nat Med 2003; 9:685–693. 22. Bikfalvi A, Bicknell R. Recent advances in angiogenesis, anti-angiogenesis and vascular targeting. Trends Pharmacol Sci 2002; 23:576–582. 23. Yancopoulos GD, Davis S, Gale NW, Rudge JS, Wiegand SJ, Holash J. Vascularspecific growth factors and blood vessel formation. Nature 2000; 407:242–248. 24. Dvorak HF, Dvorak AM, Manseau EJ, Wiberg L, Churchill WH. Fibrin gel investment associated with line 1 and line 10 solid tumor growth, angiogenesis, and fibroplasia in guinea pigs. Role of cellular immunity, myofibroblasts, microvascular damage, and infarction in line 1 tumor regression. J Natl Cancer Inst 1979; 62:1459–1472. 25. Senger DR, Galli SJ, Dvorak AM, Perruzzi CA, Harvey VS, Dvorak HF. Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science 1983; 219:983–985. 26. Ferrara N, Carver-Moore K, Chen H, Dowd M, Lu L, O’Shea KS, Powell-Braxton L, Hillan KJ, Moore MW. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 1996; 380:439–442. 27. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med 2003; 9:669–676. 28. Hiratsuka S, Minowa O, Kuno J, Noda T, Shibuya M. Flt-1 lacking the tyrosine kinase domain is sufficient for normal development and angiogenesis in mice. Proc Natl Acad Sci USA 1998; 95:9349–9354. 29. Pugh CW, Ratcliffe PJ. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med 2003; 9:677–684. 30. Brown LF, Berse B, Jackman RW, Tognazzi K, Manseau EJ, Dvorak HF, Senger DR. Increased expression of vascular permeability factor (vascular endothelial growth factor) and its receptors in kidney and bladder carcinomas. Am J Pathol 1993; 143:1255–1262. 31. Brown LF, Berse B, Jackman RW, Tognazzi K, Guidi AJ, Dvorak HF, Senger DR, Connolly JL, Schnitt SJ. Expression of vascular permeability factor (vascular endothelial growth factor) and its receptors in breast cancer. Hum Pathol 1995; 26:86–91. 32. Carson-Walter EB, Watkins DN, Nanda A, Vogelstein B, Kinzler KW, St Croix B. Cell surface tumor endothelial markers are conserved in mice and humans. Cancer Res 2001; 61:6649–6655. 33. Witmer AN, Dai J, Weich HA, Vrensen GF, Schlingemann RO. Expression of vascular endothelial growth factor receptors 1, 2, and 3 in quiescent endothelia. J Histochem Cytochem 2002; 50:767–777. 34. Huang J, Frischer JS, Serur A, Kadenhe A, Yokoi A, McCrudden KW, New T, O’Toole K, Zabski S, Rudge JS, Holash J, Yancopoulos GD, Yamashiro DJ, Kandel JJ. Regression of established tumors and metastases by potent vascular endothelial growth factor blockade. Proc Natl Acad Sci USA 2003; 100:7785–7790. 35. Kim KJ, Li B, Winer J, Armanini M, Gillett N, Phillips HS, Ferrara N. Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumour growth in vivo. Nature 1993; 362:841–844.

Endothelial cells in health and disease

398

36. Cheng SY, Huang HJ, Nagane M, Ji XD, Wang D, Shih CC, Arap W, Huang CM, Cavenee WK. Suppression of glioblastoma angiogenicity and tumorigenicity by inhibition of endogenous expression of vascular endothelial growth factor. Proc Natl Acad Sci USA 1996; 93:8502–8507. 37. Grunstein J, Roberts WG, Mathieu-Costello O, Hanahan D, Johnson RS. Tumor-derived expression of vascular endothelial growth factor is a critical factor in tumor expansion and vascular function. Cancer Res 1999; 59:1592–1598. 38. Viloria-Petit A, Miquerol L, Yu JL, Gertsenstein M, Sheehan C, May L, Henkin J, Lobe C, Nagy A, Kerbel RS, Rak J. Contrasting effects of VEGF gene disruption in embryonic stem cell-derived versus oncogene-induced tumors. EMBO J 2003; 22:4091–4102. 39. Rosen LS. Inhibitors of the vascular endothelial growth factor receptor. Hematol Oncol Clin North Am 2002; 16:1173–1187. 40. Huss WJ, Barrios RJ, Greenberg NM. SU5416 selectively impairs angiogenesis to induce prostate cancer-specific apoptosis. Mol Cancer Ther 2003; 2:611–616. 41. Boyer SJ. Small molecule inhibitors of KDR (VEGFR-2) kinase: an overview of structure activity relationships. Curr Top Med Chem 2002; 2:973–1000. 42. McCarthy M. Antiangiogenesis drug promising for metastatic colorectal cancer. Lancet 2003; 361:1959. 43. Carmeliet P. Mechanisms of angiogenesis and arteriogenesis. Nat Med 2000; 6:389–395. 44. Papapetropoulos A, Garcia-Cardena G, Dengler TJ, Maisonpierre PC, Yancopoulos GD, Sessa WC. Direct actions of angiopoietin-1 on human endothelium: evidence for network stabilization, cell survival, and interaction with other angiogenic growth factors. Lab Invest 1999; 79:213–223. 45. Thurston G, Suri C, Smith K, McClain J, Sato TN, Yancopoulos GD, McDonald DM. Leakageresistant blood vessels in mice transgenically overexpressing angiopoietin-1. Science 1999; 286:2511–2514. 46. Holash J, Maisonpierre PC, Compton D, Boland P, Alexander CR, Zagzag D, Yancopoulos GD, Wiegand SJ. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science 1999; 284:1994–1998. 47. Kim I, Kim JH, Moon SO, Kwak HJ, Kim NG, Koh GY. Angiopoietin-2 at high concentration can enhance endothelial cell survival through the phosphatidylinositol 3′-kinase/Akt signal transduction pathway. Oncogene 2000; 19:4549−4552. 48. Vajkoczy P, Farhadi M, Gaumann A, Heidenreich R, Erber R, Wunder A, Tonn JC, Menger MD, Breier G. Microtumor growth initiates angiogenic sprouting with simultaneous expression of VEGF, VEGF receptor-2, and angiopoietin-2. J Clin Invest 2002; 109:777–785. 49. Lin P, Polverini P, Dewhirst M, Shan S, Rao PS, Peters K. Inhibition of tumor angiogenesis using a soluble receptor establishes a role for Tie2 in pathologic vascular growth. J Clin Invest 1997; 100:2072–2078. 50. Brooks PC, Montgomery AM, Rosenfeld M, Reisfeld RA, Hu T, Klier G, Cheresh DA. Integrin alpha v beta 3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. Cell 1994; 79:1157–1164. 51. Friedlander M, Brooks PC, Shaffer RW, Kincaid CM, Varner JA, Cheresh DA. Definition of two angiogenic pathways by distinct alpha v integrins. Science 1995; 270: 1500–1502. 52. Sudhakar A, Sugimoto H, Yang C, Lively J, Zeisberg M, Kalluri R. Human tumstatin and human endostatin exhibit distinct antiangiogenic activities mediated by alpha v beta 3 and alpha 5 beta 1 integrins. Proc Natl Acad Sci USA 2003; 100:4766–4771. 53. Rehn M, Veikkola T, Kukk-Valdre E, Nakamura H, Ilmonen M, Lombardo C, Pihlajaniemi T, Alitalo K, Vuori K. Interaction of endostatin with integrins implicated in angiogenesis. Proc Natl Acad Sci USA 2001; 98:1024–1029. 54. Wickstrom SA, Alitalo K, Keski-Oja J. Endostatin associates with integrin alpha5beta1 and caveolin-1, and activates Src via a tyrosyl phosphatase-dependent pathway in human endothelial cells. Cancer Res 2002; 62:5580–5589.

Molecular targets of tumor vasculature

399

55. Maeshima Y, Yerramalla UL, Dhanabal M, Holthaus KA, Barbashov S, Kharbanda S, Reimer C, Manfredi M, Dickerson WM, Kalluri R. Extracellular matrix-derived peptide binds to alpha(v)beta(3) integrin and inhibits angiogenesis. J Biol Chem 2001; 276:31959–31968. 56. Tarui T, Miles LA, Takada Y. Specific interaction of angiostatin with integrin alpha(v)beta(3) in endothelial cells. J Biol Chem 2001; 276:39562–39568. 57. Reynolds LE, Wyder L, Lively JC, Taverna D, Robinson SD, Huang X, Sheppard D, Hynes RO, Hodivala-Dilke KM. Enhanced pathological angiogenesis in mice lacking beta3 integrin or beta3 and beta5 integrins. Nat Med 2002; 8:27–34. 58. Hynes RO. A reevaluation of integrins as regulators of angiogenesis. Nat Med 2002; 8:918– 921. 59. Pozzi A, Moberg PE, Miles LA, Wagner S, Soloway P, Gardner HA. Elevated matrix metalloprotease and angiostatin levels in integrin alpha 1 knockout mice cause reduced tumor vascularization. Proc Natl Acad Sci USA 2000; 97:2202–2207. 60. Kerbel RS. Clinical trials of antiangiogenic drugs: opportunities, problems, and assessment of initial results. J Clin Oncol 2001; 19:45S–51S. 61. Marx J. Angiogenesis, A boost for tumor starvation. Science 2003; 301:452–454. 62. Thorpe PE, Ran S. Tumor infarction by targeting tissue factor to tumor vasculature. Cancer J 2000; 6(suppl 3):S237–S244. 63. Neri D. Targeted delivery of tissue factor to tumor neovasculature for cancer therapy. Haemostasis 2001; 31(suppl 1):21–22. 64. Huang X, Molema G, King S, Watkins L, Edgington TS, Thorpe PE. Tumor infarction in mice by antibody-directed targeting of tissue factor to tumor vasculature. Science 1997; 275:547– 550. 65. Nilsson F, Kosmehl H, Zardi L, Neri D. Targeted delivery of tissue factor to the ED-B domain of fibronectin, a marker of angiogenesis, mediates the infarction of solid tumors in mice. Cancer Res 2001; 61:711–716. 66. Ran S, Gao B, Duffy S, Watkins L, Rote N, Thorpe PE. Infarction of solid Hodgkin’s tumors in mice by antibody-directed targeting of tissue factor to tumor vasculature. Cancer Res 1998; 58:4646–4653. 67. Langley RR, Russell J, Eppihimer MJ, Alexander SJ, Gerritsen M, Specian RD, Granger DN. Quantification of murine endothelial cell adhesion molecules in solid tumors. Am J Physiol 1999; 277:H1156–H1166. 68. Ran S, Downes A, Thorpe PE. Increased exposure of anionic phospholipids on the surface of tumor blood vessels. Cancer Res 2002; 62:6132–6140. 69. Ran S, Thorpe PE. Phosphatidylserine is a marker of tumor vasculature and a potential target for cancer imaging and therapy. Int J Radiat Oncol Biol Phys 2002; 54: 1479–1484. 70. Wild R, Dhanabal M, Olson TA, Ramakrishnan S. Inhibition of angiogenesis and tumour growth by VEGF121-toxin conjugate: differential effect on proliferating endothelial cells. Br J Cancer 2000; 83:1077–1083. 71. Seon BK, Matsuno F, Haruta Y, Kondo M, Barcos M. Long-lasting complete inhibition of human solid tumors in SCID mice by targeting endothelial cells of tumor vasculature with antihuman endoglin immunotoxin. Clin Cancer Res 1997; 3:1031–1044. 72. Matsuno F, Haruta Y, Kondo M, Tsai H, Barcos M, Seon BK. Induction of lasting complete regression of preformed distinct solid tumors by targeting the tumor vasculature using two new anti-endoglin monoclonal antibodies. Clin Cancer Res 1999; 5: 371–382. 73. Li Y, Wang MN, Li H, King KD, Bassi R, Sun H, Santiago A, Hooper AT, Bohlen P, Hicklin DJ. Active immunization against the vascular endothelial growth factor receptor flk1 inhibits tumor angiogenesis and metastasis. J Exp Med 2002; 195:1575–1584. 74. Niethammer AG, Xiang R, Becker JC, Wodrich H, Pertl U, Karsten G, Eliceiri BP, Reisfeld RA. A DNA vaccine against VEGF receptor 2 prevents effective angiogenesis and inhibits tumor growth. Nat Med 2002; 8:1369–1375.

Endothelial cells in health and disease

400

75. Arap W, Pasqualini R, Ruoslahti E. Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 1998; 279:377–380. 76. Ruoslahti E. Specialization of tumour vasculature. Nat Rev Cancer 2002; 2:83–90. 77. Ditzel HJ. Human antibodies in cancer and autoimmune disease. Immunol Res 2000; 21:185– 193. 78. Crameri R, Achatz G, Weichel M, Rhyner C. Direct selection of cDNAs by phage display. Methods Mol Biol 2002; 185:461–469. 79. Pasqualini R, Koivunen E, Kain R, Lahdenranta J, Sakamoto M, Stryhn A, Ashmun RA, Shapiro LH, Arap W, Ruoslahti E. Aminopeptidase N is a receptor for tumorhoming peptides and a target for inhibiting angiogenesis. Cancer Res 2000; 60:722–727. 80. Takahashi M, Fukami S, Iwata N, Inoue K, Itohara S, Itoh H, Haraoka J, Saido T. In vivo glioma growth requires host-derived matrix metalloproteinase 2 for maintenance of angioarchitecture. Pharmacol Res 2002; 46:155–163. 81. Itoh T, Tanioka M, Yoshida H, Yoshioka T, Nishimoto H, Itohara S. Reduced angiogenesis and tumor progression in gelatinase A-deficient mice. Cancer Res 1998; 58: 1048–1051. 82. Rio MC, Lefebvre O, Santavicca M, Noel A, Chenard MP, Anglard P, Byrne JA, Okada A, Regnier CH, Masson R, Bellocq JP, Basset P. Stromelysin-3 in the biology of the normal and neoplastic mammary gland. J Mammary Gland Biol Neoplasia 1996; 1:231–240. 83. Nanda A, Carson-Walter EB, Seaman S, Barber TD, Stampfl J, Singh S, Vogelstein B, Kinzler KW, St. Croix B. TEM8 interacts with the cleaved C5 domain of collagen α3(VI). Cancer Res 2004; 64:817–820. 84. Christian S, Ahorn H, Koehler A, Eisenhaber F, Rodi HP, Garin-Chesa P, Park JE, Rettig WJ, Lenter MC. Molecular cloning and characterization of endosialin, a C-type lectin-like cell surface receptor of tumor endothelium. J Biol Chem 2001; 276: 7408–7414. 85. Rettig WJ, Garin-Chesa P, Healey JH, Su SL, Jaffe EA, Old LJ. Identification of endosialin, a cell surface glycoprotein of vascular endothelial cells in human cancer. Proc Natl Acad Sci USA 1992; 89:10832–10836. 86. Marinissen MJ, Gutkind JS. G-protein-coupled receptors and signaling networks: emerging paradigms. Trends Pharmacol Sci 2001; 22:368–376. 87. Wilson S, Bergsma DJ, Chambers JK, Muir AI, Fantom KG, Ellis C, Murdock PR, Herrity NC, Stadel JM. Orphan G-protein-coupled receptors: the next generation of drug targets? Br J Pharmacol 1998; 125:1387–1392. 88. Koo HM, VanBrocklin M, McWilliams MJ, Leppla SH, Duesbery NS, Woude GF. Apoptosis and melanogenesis in human melanoma cells induced by anthrax lethal factor inactivation of mitogen-activated protein kinase kinase. Proc Natl Acad Sci USA 2002; 99:3052–3057. 89. Bradley KA, Young JA. Anthrax toxin receptor proteins. Biochem Pharmacol 2003; 65:309– 314. 90. Liu S, Leppla SH. Cell surface tumor endothelium marker 8 cytoplasmic tail-independent anthrax toxin binding, proteolytic processing, oligomer formation, and internalization. J Biol Chem 2003; 278:5227–5234. 91. Duesbery NS, Resau J, Webb CP, Koochekpour S, Koo HM, Leppla SH, Vande Woude GF. Suppression of ras-mediated transformation and inhibition of tumor growth and angiogenesis by anthrax lethal factor, a proteolytic inhibitor of multiple MEK pathways. Proc Natl Acad Sci USA 2001; 98:4089–4094. 92. Liu S, Aaronson H, Mitola DJ, Leppla SH, Bugge TH. Potent antitumor activity of a urokinaseactivated engineered anthrax toxin. Proc Natl Acad Sci USA 2003; 100: 657–662. 93. Aigner T, Hambach L, Soder S, Schlotzer-Schrehardt U, Poschl E. The C5 domain of Co16A3 is cleaved off from the Co16 fibrils immediately after secretion. Biochem Biophys Res Commun 2002; 290:743–748. 94. Mayer U, Poschl E, Nischt R, Specks U, Pan TC, Chu ML, Timpl R. Recombinant expression and properties of the Kunitz-type protease-inhibitor module from human type VI collagen alpha 3(VI) chain. Eur J Biochem 1994; 225:573–580.

Molecular targets of tumor vasculature

401

95. Iruela-Arispe ML, Hasselaar P, Sage H. Differential expression of extracellular proteins is correlated with angiogenesis in vitro. Lab Invest 1991; 64:174–186. 96. Iruela-Arispe ML, Diglio CA, Sage EH. Modulation of extracellular matrix proteins by endothelial cells undergoing angiogenesis in vitro. Arterioscler Thromb 1991; 11: 805–815. 97. Niquet J, Represa A. Entactin immunoreactivity in immature and adult rat brain. Brain Res Dev Brain Res 1996; 95:227–233. 98. Oono T, Specks U, Eckes B, Majewski S, Hunzelmann N, Timpl R, Krieg T. Expression of type VI collagen mRNA during wound healing. J Invest Dermatol 1993; 100: 329–334. 99. Niiyama T, Higuchi I, Hashiguchi T, Suehara M, Uchida Y, Horikiri T, Shiraishi T, Saitou A, Hu J, Nakagawa M, Arimura K, Osame M. Capillary changes in skeletal muscle of patients with Ullrich’s disease with collagen VI deficiency. Acta Neuropathol (Berl) 2003; 106:137–142. 100. Lee WS, Jain MK, Arkonac BM, Zhang D, Shaw SY, Kashiki S, Maemura K, Lee SL, Hollenberg NK, Lee ME, Haber E. Thy-1, a novel marker for angiogenesis upregulated by inflammatory cytokines. Circ Res 1998; 82:845–851. 101. Liu X, Wu H, Byrne M, Krane S, Jaenisch R. Type III collagen is crucial for collagen I fibrillogenesis and for normal cardiovascular development. Proc Natl Acad Sci USA 1997; 94:1852–1856. 102. Pope FM, Narcisi P, Nicholls AC, Germaine D, Pals G, Richards AJ. COL3A1 mutations cause variable clinical phenotypes including acrogeria and vascular rupture. Br J Dermatol 1996; 135:163–181. 103. Hopkinson I, Evans W, Chant D, Hiscox S, Berry D, Harding K. Reverse transcriptionpolymerase chain reaction detection of collagen transcripts in healing human wounds. Eur J Clin Invest 1995; 25:539–542. 104. Fouser L, Iruela-Arispe L, Bornstein P, Sage EH. Transcriptional activity of the alpha 1(I)collagen promoter is correlated with the formation of capillary-like structures by endothelial cells in vitro. J Biol Chem 1991; 266:18345–18351. 105. Hewitt RE, Powe DG, Carter GI, Turner DR, Price JE. Basement membrane collagen-IV synthesis in colorectal tumours. Int J Cancer 1992; 51:530–536.

20 The Role of the Endothelium in Severe Sepsis and Multiple Organ Dysfunction William C.Aird Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, U.S.A.

1. INTRODUCTION Sepsis is the leading cause of death among hospitalized patients in non-coronary intensive care units. An important goal is to develop improved therapeutic strategies that will impact favorably on patient outcome. Recent studies have pointed to a critical role for the endothelium in orchestrating the host response in severe sepsis. In this chapter, a conceptual framework for understanding the pathophysiology of sepsis is provided and the potential value of the endothelium as a target for sepsis therapy is emphasized.

2. DEFINITIONS, EPIDEMIOLOGY, AND CLINICAL MANIFESTATIONS Systemic inflammatory response syndrome (SIRS) is defined by the presence of more than one of the following: (1) body temperature >38°C or 90/min, (3) hyperventilation evidenced by respiratory rate >20/min or PaCO2 12,000/µL or 90% patients) (6,7).

3. EVOLUTIONARY PERSPECTIVES Theodosius Dobzhansky (1900–1975) once wrote: “Nothing in biology (and by extension medicine) makes sense except in the light of evolution.” The application of evolutionary principles to an understanding of health and disease represents the foundation of a nascent discipline, termed Darwinian Medicine, which was popularized by Nesse and Williams (8) in their trade book, “Why We Get Sick: The New Science of Darwinian Medicine” and a feature article in Scientific American entitled: “The Evolutionary Origins of Disease” (9). In their book, Nesse and Williams state the following: When evolution is included in medical school curricula, it will give students not only a new perspective on disease but also an integrating framework on which to hang a million otherwise arbitrary facts. Darwinian Medicine could bring intellectual coherence to the chaotic enterprise of medical education. (8) A consideration of evolutionary principles underlying sepsis does indeed provide useful insight. As outlined in Chapter 1, most multicellular organisms require a pump (heart) to overcome the time-distance constraints of diffusion (Fig. 1). By definition, the system is pressurized and hence at risk for rupture and leakage. Invertebrates have an open circulation, in which hemolymph circulates and directly bathes the various tissues of the body. The horseshoe crab, an ancient invertebrate that belongs to a subclass spanning over 500 million years, has a single circulating blood cell, termed the hemocyte. When activated by bacterial endotoxin, the hemocyte initiates a crude clotting cascade that results in a fibrin-like gel on the surface of the cell. The resulting glue serves to encapsulate the organisms, thereby aiding in their engulfment and disposal. In humans, the monocyte and neutrophil may interact with fibrin to carry out a similar function. This is the first hint of a functional connection between the inflammatory and coagulation systems. As discussed in Chapter 1, vertebrates have a closed circulation, in which blood is contained within a closed space. Vertebrate blood—which contains all three lineages of white blood cells, red blood cells, and platelets (or their nucleated counterpart, thrombocytes)—is separated from underlying tissue by the endothelium. Two additional features that separate vertebrates from invertebrates are the presence of a clotting cascade and acquired immunity (antibody production).a The multicellular organism is analogous to a furnace, burning nutrients, and oxygen to produce energy, all towards one end, namely reproduction and transmission of genes to the future—a wonderfully simple Darwinian view of life. What makes the sepsis dynamic so interesting and so full of intrigue is that two species (humans and pathogens) are engaged in an all out battle for survival in which they have the same, sometimes mutually exclusive, goal of passing their genes on to the next generation. This interaction has led to an evolutionary “arms race,” involving a sophisticated array of defenses, offenses, and

The role of the endothelium in severe sepsis

405

counter-defenses. Ideally, the battlefield is restricted to the extravascular space, where the host response prevails in containing and eliminating the threat (Fig. 1B). On occasion, however, an excessive or sustained host response may spill over into the systemic circulation (Fig. 1C), where it escapes the highly regulated control mechanisms of the local tissue environment. a

As a striking example of convergent evolution, some invertebrates, such as the horseshoe crab, have independently evolved their own clotting cascade. In past times, this turn of events often signaled the demise of the patient. However, with recent advances in critical care, patients who would otherwise have died from their disease are now being artificially supported. In effect, the “rules of engagement” between pathogens and humans have changed; the “battle lines” have been redrawn; and from this battlefield has emerged a new syndrome in the chronically ill, namely severe sepsis and MODS.

4. SEPSIS PATHOPHSYIOLOGY There are several basic themes that underlie the pathophysiology of sepsis. First, the host response rather than the type of pathogen is the most important determinant of patient outcome. Second, monocytes and endothelial cells play a central role in initiating and perpetuating the host response. Third, sepsis is associated with the systemic activation of the inflammatory and coagulation cascades. Fourth, the inflammatory and coagulation pathways interact with one another to amplify the host response. Finally, in a concerted effort to fend off and eliminate pathogens, the host response may inflict collateral damage on normal tissues, resulting in pathology that is not diffuse, but remarkably focal in its distribution. Each of these themes has been previously reviewed in detail (10). The pathophysiology of sepsis may be simplified according to the scheme shown in Fig. 2. The monocyte (or tissue macrophage) recognizes lipopolysaccharide (or other components of pathogens) via toll-like receptors, resulting in activation of the inflammatory and coagulation pathways. The monocyte (and to some extent the neutrophil and endothelial cell) is the cornerstone of the innate immune response, serving to separate the world into self and non-self based on physical properties This response is highly evolutionary conserved, and as such is fast, reliable, and durable. However, like any great weapon, the innate immune response may ultimately turn on its bearer and result in organ dysfunction. On the inflammatory side, the monocyte (and to some extent the endothelium) releases a number of inflammatory mediators that operate in an autocrine manner to further activate the monocyte or in a paracrine fashion to activate neighboring endothelial cells. On the coagulation side, activated monocytes express tissue factor on their cell surface, which then triggers the clotting cascade, resulting in thrombin generation and fibrin formation. There is cross talk between the inflammatory and coagulation cascades. As discussed above, inflammatory mediators induce expression of tissue factor on the surface of monocytes (and perhaps selected populations of endothelial cells). In the other direction, serine proteases that are generated in the clotting cascade are capable of binding to protease-activated receptors (PAR) present on

Endothelial cells in health and disease

406

the surface of endothelial cells and monocytes, resulting in a procoagulant and proadhesive phenotype.

5. ROLE OF THE ENDOTHELIUM IN ORCHESTRATING THE HOST RESPONSE IN SEPSIS 5.1. Primer in Endothelial Cell Biology Four basic principles in endothelial cell biology provide a foundation for understanding the role of the endothelium is sepsis pathophysiology. First, the endothelium in not inert, but rather is metabolically active. Second, the endothelium is an input-output device, sensing changes in the extracellular compartment, and responding in ways that are beneficial or at times harmful to the host. Third, endothelial cell phenotypes vary in space and time, giving rise to endothelial cell heterogeneity. Fourth, the endothelium displays non-linear dynamics and emergent properties and therefore can be fully understood only in the context of the whole organism. Each of these themes is discussed in detail in Chapter 1.

The role of the endothelium in severe sepsis

407

Figure 1 Vertebrate body plan. (A) Scheme of a normal system showing convection of air from environment to gas exchanger (skin, gills, or lungs),

Endothelial cells in health and disease

408

diffusion of oxygen across into the blood, convection of blood around to the tissue of the body, and diffusion of oxygen into the individual cells of the tissues. Blood is contained within a closed space and is circulated by way of a pressurized pump (heart). Vertebrate blood consists of platelets, red blood cells, and leukocytes (shown are representative red blood cells, monocytes, and neutrophils). (B) The innate immune response is activated when pathogens (represented by four small shapes) invade body tissues, and consists of a cellular and protein response. Ideally, the host prevails in containing and eliminating the threat in the extravascular space. (C) The host response may spill into the systemic circulation and become uncoupled from checks and balances that exist locally in the extravascular space. This stage correlates with the development of severe sepsis.

The role of the endothelium in severe sepsis

409

Figure 2 Sepsis pathophysiology. Shown is a blood vessel lumen lined by endothelium on top and bottom (C). The circulating monocyte (and tissue macrophage, not shown) binds to lipopolysaccharide (LPS), and initiates inflammatory (A) and coagulation (B) cascades. The inflammatory pathway feeds back to further activate the monocyte (1) and leads to paracrine activation of the endothelium (2). Serine proteases within the coagulation cascade bind to protease activated receptors on the surface of the endothelium (and other cell types) to promote a shift in the inflammatory balance. (3) TF, tissue factor.

5.2. Historical Biases in the Sepsis Field As is true with any area in biomedicine, the sepsis field is hampered by certain historical biases. First, there seems to be a tendency to depict the host response in “black-andwhite” terms—that is, to label the response as a terrible misunderstanding, an accident or freak of nature, rather than a subtle shift in the balance of power between two

Endothelial cells in health and disease

410

sophisticated species at war. As a result, one tends to overlook the continuum between adaptation and non-adaptation (or function and dysfunction), an oversight that may have important ramifications in terms of treatment. Second, there is a penchant for taking sides. For example, some like to point fingers at the monocyte, others like to lay blame with the endothelium. Perhaps in the spirit of compromise, one might decide to apportion blame; for example, to assign 43% of responsibility to the inflammatory pathways, and 57% to the clotting cascade. I am exaggerating to make a point, but it is an important one nevertheless, which is that Newton’s universe of linear cause-and-effect, while perhaps useful for describing the force of gravity, is poorly suited to modeling biological systems. In characterizing the host response to infection, the whole is far greater than the sum of the parts. These considerations lead to an important disclaimer: the endothelium is not the “Darth Vader” of sepsis. Indeed, if forced to take sides, my vote would be to award the endothelium with the Purple Heart for matching wits and going head to head with pathogens, day in and day out, chalking up silent victory after silent victory. While it is true that in the line of duty, a robust and overzealous endothelial response may overwhelm the host and thus contribute to sepsis pathogenesis, it is really just one component of a far more complex system.

5.3. Endothelial Cell Function vs Dysfunction As discussed in Chapter 1, endothelial cell dysfunction describes situations in which the behavior or response of the endothelium represents a net liability to the host. For example, when pathogens invade a tissue, endothelial cells are induced locally to release inflammatory mediators, to recruit leukocytes and to promote clotting as a means of walling off the infection. During this process, endothelial cells may undergo necrosis or apoptosis as tissue is reabsorbed and repaired. When viewed from the standpoint of the individual cell, necrosis and/or apoptosis are the ultimate expression of dysfunction. However, when considered in the larger context of host defense, the local loss of endothelium is part of a larger coordinated, adaptive response. Available evidence suggests that severe sepsis is associated with excessive, sustained, and generalized activation of the endothelium (see the next section). Without artificial organ support, virtually all patients with severe sepsis would die from their disease. In other words, most of these individuals have crossed the threshold from an adaptive to a maladaptive response. In so far as the endothelium contributes to the severe sepsis phenotype, its behavior may be characterized as dysfunctional. An important goal for the future will be to learn how to identify the transition from function to dysfunction, before the onset of significant (and perhaps irreversible) organ damage.

5.4. Endothelial Response in Severe Sepsis When considering the role of the endothelium in sepsis, it is helpful to return to the analogy between the endothelium and an input-output device (see Chapter 1)

The role of the endothelium in severe sepsis

411

Figure 3 Endothelial cell targets for sepsis therapy. Each endothelial cell may be viewed as an input-output device. Sepsis-associated changes in input, output, and coupling represent potential targets for therapy. The input for two endothelial cells (EC) is shown in the center, and includes biochemical (e.g., pH, temperature, cytokines, chemokines, growth factors, compliment), biomechanical input (e.g., flow), and interaction with other cell types (e.g., circulating blood cells, and underlying pericytes, vascular smooth muscle cells and parenchymal cells). A representative output or phenotype is shown on the right, and includes leukocyte adhesion and fibrin deposition. On the left, the inputoutput device has been “opened,” revealing a representative signaling pathway that couples protease

Endothelial cells in health and disease

412

activated receptor-1 (round shape) with expression of vascular cell adhesion molecule (VCAM)-1 (square shape). VSMC: vascular smooth muscle cell. (Fig. 3). Sepsis is associated with many changes in the input signals, including components of the bacterial wall, complement, cytokines, chemokines, serine proteases, fibrin, activated platelets and leukocytes, hyperglycemia, and/or changes in oxygenation or blood flow. Endothelial outputs include both structural alterations (e.g., nuclear vacuolization, cytoplasmic swelling, cytoplasmic fragmentation, denudation, and/or detachment) and functional changes (e.g., shifts in the hemostatic balance, increased cell adhesion and leukocyte trafficking, altered vasomotor tone, loss of barrier function, and programmed cell death). Finally, the “set point” of the endothelium—as determined by the influence of epigenetic processes, age, comorbidity, and genetic polymorphisms— may alter the phenotype and/or transduction capacity (input-output coupling) of the endothelial cell.

5.5. Link Between Endothelial Cell Dysfunction and MODS The host response to sepsis involves an elaborate array of cell types, including leukocytes, platelets, and endothelial cells, as well as soluble mediators, including components of the inflammatory and coagulation cascades. Normally, these mechanisms are highly coordinated with one another to defend host against pathogen. However, if the host response is disproportionate to the nature of the threat, that is, it is excessively sustained or poorly localized, then the balance of power shifts in favor of the pathogen, resulting in the sepsis phenotype—namely, dysfunction of subsets of organ systems. Despite a plethora of hypotheses—most of which are based on linear cause-and-effect models—the mechanism(s) by which sepsis results in organ dysfunction are not known. In truth, the host response is far more complex than we are willing to admit and until we come to understand the nature and impact of the interactions between the various cells and soluble mediators, we will remain largely in the dark about sepsis pathophysiology. Based on our current understanding, the role of the endothelium in sepsis pathophysiology may be summarized as follows: (1) the endothelium is not an innocent bystander in sepsis, but rather is responsible for its own actions, (2) the behavior of this cell layer should be adjudicated in an appropriate evolutionary context: the endothelial response evolved as a mechanism to protect host against pathogen, not to withstand the rigors of severe sepsis and artificial life support, (3) the endothelium is a critical, but not the sole, component of the host response to sepsis, (4) the endothelium is strategically located between blood and underlying tissue, (5) the endothelium is a highly malleable and flexible cell layer, and therefore, (6) the endothelium is a potentially valuable target for sepsis therapy.

The role of the endothelium in severe sepsis

413

6. THE ENDOTHELIUM AS A THERAPEUTIC TARGET 6.1. Therapeutic Perspectives Over the past decade, enormous resources have been expended on sepsis trials, with more than 10,000 patients enrolled in over 20 placebo-controlled, randomized phase 3 clinical trials (11,12). Most of these therapies have failed to reduce mortality in patients with severe sepsis, including anti-endotoxin, anti-cytokine, anti-prostaglandin, antibradykinin, and anti-platelet activating factor (PAF) strategies, antithrombin III (ATIII), and tissue factor pathway inhibitor (TFPI) (11,13,14). At the time of this writing, a total of five phase 3 clinical trials have demonstrated improved survival in critically ill patients or patients with severe sepsis. These include the use of low tidal volume ventilation (15), recombinant human activated protein C (rhAPC) (7), low-dose glucocorticoids (16), intensive insulin therapy (17), and early goal-directed therapy (18). The results of these clinical trials have provided the basis for overhauling, modifying or otherwise fine-tuning our models of sepsis pathophysiology. If any consensus has been reached based on the myriad clinical and preclinical trials in sepsis, it may be summarized as follows: 1. Anti-inflammatory therapy is ineffective in improving survival or organ dysfunction. Therapy directed towards endotoxin or one or another inflammatory mediator has invariably failed to improve survival in large phase 3 clinical trials (14,19,20). These data are consistent with the notion that the inflammatory cascade, while certainly an important contributor to sepsis morbidity and mortality, is sufficiently redundant, pleiotropic, and inter-dependent so as to preclude single modality therapy. 2. While the selective inhibition of thrombin generation reduces fibrin deposition, it has no effect on organ dysfunction or mortality (21,22). These find-ings, which are derived from preclinical studies, suggest that it takes more than activation of coagulation to deliver the “fatal blow” in sepsis. 3. Combination antiinflammatory and anticoagulant therapy may be beneficial. For example, TFPI, ATIII and rhAPC has each been shown to inhibit inflammation and coagulation in vitro and in vivo and to yield improved survival in non-human primate models of sepsis and phase 2 clinical trials of severe sepsis (23–29). Of these three agents, only rhAPC was shown to reduce mortality in phase 3 clinical studies (13,30– 32). One interpretation of these data is that all three agents are beneficial, yet the design of the TFPI and ATIII trials was flawed for one reason or another. An alternative explanation is that activated protein C has unique biological effects that set it apart from ATIII and TFPI in humans with severe sepsis (if not in the baboon model of sepsis). Indeed, while TFPI and ATIII are likely to exert their anti-inflammatory indirectly effect through PAR (to date, there is no evidence of an ATIII receptor), activated protein C binds directly to PAR via a unique co-receptor, the endothelial protein C receptors (EPCR), which is expressed on the surface of endothelial cells and possibly monocytes. The interaction between activated protein C and its receptors has been implicated in its profound antiinflammatory and anti-apoptotic functions (33).

Endothelial cells in health and disease

414

6.2. Back to the Drawing Board At face value the results of the above trials suggest that therapy aimed at multiple components of the sepsis cascade inflammation and coagulation holds more promise than targeting any single component inflammation or coagulation. However, simple interpretation has been called into question by two recent studies. First, Kerlin et al. (34) reported that FV Leiden confers a survival advantage in human and animal models of severe sepsis. The FV Leiden mutation renders the clotting cascade resistant to the inhibitory effects of activated protein C. Thus, the results imply not only that activated protein C exerts its beneficial effects through a non-anticoagulant mechanism (e.g., the endothelium), but also that modest increases in thrombin generation (as occurs in heterozygous FV Leiden) are actually protective in the setting of severe sepsis. One explanation for this paradox is that thrombin generation results in increased activation of endogenous protein C.b Second, a recent ad hoc analysis of the PROWESS trial failed to detect significant changes in circulating levels of inflammatory biomarkers in patients receiving rhAPC (35), arguing against a significant anti-flammatory effect of this drug. If rhAPC is not exerting its benefit through inhibition of the coagulation and inflammatory cascade, how is it working? The honest answer is, we do not know. However, an attractive hypothesis is that the rhAPC attenuates endothelial cell dysfunction and/or inhibits endothelial cell apoptosis, and that these effects elude current diagnostic detection. b

These data suggest FV Leiden mutation (which appeared approximately 20,000–30,000 years ago) may have evolved as a means of protecting not so much against the Saber Tooth tiger, but rather as a defensive weapon in the host-pathogen arms race.

6.3. Other Strategies for Targeting the Endothelium in Sepsis A common theme that ties together the five successful treatments in severe sepsis is their capacity to attenuate endothelial cell dysfunction. The effect of rhAPC on the endothelium was discussed above. Low volume ventilation would be expected to reduce barotraumas to the pulmonary endothelium. Low-dose glucocorticoids may reduce the activity of proinflammatory transcription factors in endothelial cells, while intensive insulin therapy may reduce the deleterious effects of high glucose on the endothelium. Finally, early goal-directed therapy is predicted to maintain flow and hence shear stress at the level of the blood vessel wall. The extent to which these therapies exert their benefit through the endothelium remains unknown. There are many other possible strategies for attenuating endothelial cell dysfunction (for a detailed account, the reader is referred to a previously published review) (Table 1) (10). One may target the input signals, the coupling mechanism inside the cell or the cellular phenotype (output). Examples of extracellular signals as targets include endotoxin, TNF-α, IL-1, and PAF. Examples of coupling mechanisms include receptors, such as cellular adhesion molecules and protease-activated

Table 1 Targeting the Endothelium in Sepsis

The role of the endothelium in severe sepsis

415

Input

Output

A. Biochemical

Apoptosis/cell survival

Growth factors

Proliferation

GM-CSF

Migration

VEGF

Inflammatory mediators

Chemokines

Leukocyte adhesion/transmigration

MCP-1

Hemostatic balance

Cytokines

Permeability

TNF-α

Vasomotor tone

IL-1

Metabolism

Complement

Antigen presentation

Prostaglandin

Cell-cell communication

LPS Other components of bacteria, fungi, or viruses

Input-output coupling

Nucleotides

Receptors

Serine proteases

Toll-like receptors

Fibrin

Protease activated receptors

Free oxygen radicals

IL-1 receptor

Hypoxia

TNF-α receptor

Temperature

Platelet-activating factor receptor

Electrolytes

EPCR

Hyperglycemia

Signal intermediates

Neural input

P38 MAPK

Cell-cell interactions

PKC

Platelets

Transcription factors

Monocytes

NF-κB

Neutrophils

API-1

Pericytes

GATA-2

Parenchymal cells

Ets factors

Extracellular matrix B. Biomechanical Hemodynamic forces

Endothelial cells in health and disease

416

receptors; signaling pathways such as p38 MAPK or novel/atypical PKC isoforms; and transcription factors, including NF-κB, GATA-2, and the Ets family of transacting proteins. Potential target outputs include endothelial control of hemostasis, inflammation, vasomotor tone, permeability, and leukocyte trafficking. Although anti-endotoxin or anti-TNF-α antibodies failed to improve mortality in patients with severe sepsis therapy in, these findings do not exclude a role for single target (“smart bomb”) therapy in the future. The host response, while unquestionably redundant and pleiotropic, is likely to contain certain component parts—whether an extracellular mediator, a cell surface receptor, a signal intermediate, or a transcription factor—that are so highly connected as to render that component (and the entire system) vulnerable to therapeutic targeting. A key challenge is to identify these so-called “hubs” in the sepsis cascade and to target those factors accordingly.

7. CONCLUSIONS Despite new information about the pathophysiology and treatment of severe sepsis, this disorder continues to be associated with an unacceptably high mortality rate. Future breakthroughs will require a conceptual shift that emphasizes relationships between the various mediators and cells involved in host response. The endothelium is key in initiating, perpetuating, and modulating the host response to infection. Future studies promise to provide new insight into the endothelium, not as an isolated mechanism of sepsis pathophysiology, but rather as the coordinator of a far more expansive, spatially and temporally orchestrated response.

REFERENCES 1. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29(7):1303–1310. 2. Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003; 348(16):1546–1554. 3. Damas P, Canivet JL, de Groote D, et al. Sepsis and serum cytokine concentrations. Crit Care Med 1997; 25(3):405–412. 4. Thijs LG, Hack CE. Time course of cytokine levels in sepsis. Intensive Care Med 1995; 21(Suppl 2):S258–S263. 5. Pinsky MR, Vincent JL, Deviere J, Alegre M, Kahn RJ, Dupont E. Serum cytokine levels in human septic shock. Relation to multiple-system organ failure and mortality. Chest 1993; 103(2):565–575. 6. Yan SB, Helterbrand JD, Hartman DL, Wright TJ, Bernard GR. Low levels of protein C are associated with poor outcome in severe sepsis. Chest 2001; 120(3):915–922. 7. Bernard GR, Ely EW, Wright TJ, et al. Safety and dose relationship of recombinant human activated protein C for coagulopathy in severe sepsis. Crit Care Med 2001; 29(11):2051–2059. 8. Nesse RM, Williams GC. Why We Get Sick: The New Science of Darwinian Medicine. 1st ed. New York: Times Books, 1994. 9. Nesse RM, Williams GC. Evolution and the origins of disease. Sci Am 1998; 279(5): 86–93.

The role of the endothelium in severe sepsis

417

10. Aird WC. The role of the endothelium in severe sepsis and the multiple organ dysfunction syndrome. Blood 2003; 101:3765–3777. 11. Cohen J, Guyatt G, Bernard GR, et al. New strategies for clinical trials in patients with sepsis and septic shock. Crit Care Med 2001; 29(4):880–886. 12. Eichacker PQ, Parent C, Kalil A, et al. Risk and the efficacy of antiinflammatory agents: retrospective and confirmatory studies of sepsis. Am J Respir Crit Care Med 2002; 166(9):1197–1205. 13. Warren BL, Eid A, Singer P, et al. Caring for the critically ill patient. High-dose antith-rombin III in severe sepsis: a randomized controlled trial. JAMA 2001; 286(15): 1869–1878. 14. Zeni F, Freeman B, Natanson C. Anti-inflammatory therapies to treat sepsis and septic shock: a reassessment. Crit Care Med 1997; 25(7):1095–1100. 15. ARDSNET.Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000; 342(18):1301–1308. 16. Annane D, Sebille V, Charpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA 2002; 288(7):862–871. 17. van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med 2001; 345(19):1359–1367. 18. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001; 345(19):1368–1377. 19. Natanson C, Esposito CJ, Banks SM. The sirens’ songs of confirmatory sepsis trials: selection bias and sampling error. Crit Care Med 1998; 26(12):1927–1931. 20. Marshall JC. Clinical trials of mediator-directed therapy in sepsis: what have we learned? Intensive Care Med 2000; 26(suppl 1):S75–S83. 21. Coalson JJ, Benjamin B, Archer LT, et al. Prolonged shock in the baboon subjected to infusion of E. coli endotoxin. Circ Shock 1978; 5(4):423–437. 22. Taylor FB Jr. Chang AC, Peer GT, et al. DEGR-factor Xa blocks disseminated intravascular coagulation initiated by Escherichia coli without preventing shock or organ damage. Blood 1991; 78(2):364–368. 23. Emerson TE Jr, Fournel MA, Redens TB, Taylor FB Jr. Efficacy of antithrombin III supplementation in animal models of fulminant Escherichia coli endotoxemia or bacter-emia. Am J Med 1989; 87(3B):27S–33S. 24. Taylor FB Jr, Chang A, Esmon CT, D’Angelo A, Vigano-D’Angelo S, Blick KE. Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. J Clin Invest 1987; 79(3):918–925. 25. Creasey AA, Chang AC, Feigen L, Wun TC, Taylor FB Jr, Hinshaw LB. Tissue factor pathway inhibitor reduces mortality from Escherichia coli septic shock. J Clin Invest 1993; 91(6):2850– 2856. 26. Welty-Wolf KE, Carraway MS, Miller DL, et al. Coagulation blockade prevents sepsis-induced respiratory and renal failure in baboons. Am J Respir Crit Care Med 2001; 164(10 Pt 1):1988– 1996. 27. Oelschlager C, Romisch J, Staubitz A, et al. Antithrombin III inhibits nuclear factor kappaB activation in human monocytes and vascular endothelial cells. Blood 2002; 99(11):4015–4020. 28. Grinnell BW, Joyce D. Recombinant human activated protein C: a system modulator of vascular function for treatment of severe sepsis. Crit Care Med 2001; 29(7 suppl):S53– S60. 29. Souter PJ, Thomas S, Hubbard AR, Poole S, Romisch J, Gray E. Antithrombin inhibits lipopolysaccharide-induced tissue factor and interleukin-6 production by mononuclear cells, human umbilical vein endothelial cells, and whole blood. Crit Care Med 2001; 29(1):134–139. 30. Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001; 344(10):699–709. 31. Crowther MA, Marshall JC. Continuing challenges of sepsis research. JAMA 2001; 286(15):1894–1896.

Endothelial cells in health and disease

418

32. Opal SM, Kessler CM, Roemisch J, Knaub S. Antithrombin, heparin, and heparan sulfate. Crit Care Med 2002; 30(5 suppl):S325–S331. 33. Joyce DE, Gelbert L, Ciaccia A, DeHoff B, Grinnell BW. Gene expression profile of antithrombotic protein c defines new mechanisms modulating inflammation and apoptosis. J Biol Chem 2001; 276(14):11199–11203. 34. Kerlin BA, Yan SB, Isermann BH, et al. Survival advantage associated with heterozygous factor V Leiden mutation in patients with severe sepsis and in mouse endotoxemia. Blood 2003; 102(9):3085–3092. 35. Dhainaut JF, Yan SB, Margolis BD, et al. Drotrecogin alfa (activated) (recombinant human activated protein C) reduces host coagulopathy response in patients with severe sepsis. Thromb Haemost 2003; 90(4):642–653. 36. Gallucci S, Lolkema M, Matzinger P. Natural adjuvants: endogenous activators of dendritic cells. Nat Med 1999; 5(11):1249–1255.

21 The Hepatic Sinusoidal Endothelial Cell as a Primary Target of Disease Rimma Shaposhnikov Division of Gastrointestinal and Liver Diseases, Keck School of Medicine, University of Southern California, Los Angeles, California, U.S.A.

Laurie D.DeLeve USC Research Center for Liver Diseases, Division of Gastrointestinal and Liver Diseases, Keck School of Medicine, University of Southern California, Los Angeles, California, U.S.A.

1. INTRODUCTION The afferent circulation to the liver consists of the hepatic artery, which branches off the aorta, and the portal vein, which collects nutrient-rich blood from the venous circulation of the stomach, small and large intestine, pancreas, and spleen. The micro-circulation of the liver is composed of two afferents: the hepatic arterioles and terminal portal venules, sinusoids that are the equivalent of capillaries in the liver, and the effluent terminal hepatic venules. The terminal hepatic venules, also referred to as central venules, collect into hepatic veins that drain into the inferior vena cava. The heterogeneity between large vessel endothelial cells and cells within the microcirculation is apparent by phenotypic features such as function, antigenic composition, metabolic properties and response to growth factors in different tissues, and has also been demonstrated by expression profiling (1). Even among the cells of the microcirculation, there is substantial heterogeneity. Morphologically, the microvascular endothelium can be divided into continuous, discontinuous, and fenestrated endothelial cells. Continuous endothelium has continuous cytoplasm and tight junctions. Both discontinuous and fenestrated endothelial cells have pores, but the pores of fenestrated endothelial cells are larger. Fenestrated endothelial cells are found in endocrine glands, the choroids plexus, the renal peritubular and glomerular capillaries, and sinusoids of the spleen, bone marrow, and liver. Fenestrae can be open or diaphragmed. Examples of endothelial cell types with non-diaphragmed (open) fenestrae include hepatic sinusoidal endothelial cells (SEC) and glomerular endothelial cells. A major distinction between these latter two cell types is that SEC, but not glomemlar endothelial cells, lack a distinct basement membrane. The SEC fenestrae are somewhat larger in the periportal region

The hepatic sinusoidal endothelial cell

421

than in the centrilobular region of the liver lobule (2) and are organized in groups called sieve plates.

Figure 1 Schematic of the hepatic sinusoid. Sinusoidal endothelial cells line the hepatic sinusoid (Fig. 1). On the lumenal side of the SEC are the resident macrophages, which are called Kupffer cells. The space of Disse is on the ablumenal side and contains extracellular matrix components. The resident pericytes within the space of Disse are called stellate cells. The stellate cells surround the SEC and are connected by cytoplasmic projections to hepatocytes on the other side of the space of Disse. Stellate cells are contractile and can regulate sinusoidal diameter. A variety of liver diseases are initiated by damage to the liver circulation (Table 1). Two liver diseases originate in the large vessels, Budd-Chiari syndrome and portal vein thrombosis. Budd-Chiari syndrome may be due to blockage of either the hepatic veins or the hepatic portion of the inferior vena cava. These large vessel diseases may occur in individuals with a systemic diathesis to thrombosis or due to local mechanical outflow obstruction (3). At this point, there is no evidence that changes in the hepatic endothelium per se are involved in the initiation of these diseases, although it has been hypothesized that the predilection for involvement

Table 1 Liver Disease of Vascular Origin Initiated in the large vessels Budd-Chiari syndrome (including obstruction of the inferior vena cava) Portal vein thrombosis Initiated in either large vessels or the microcirculation Nodular regenerative hyperplasia

Endothelial cells in health and disease

422

Initiated in the microcirculation Sinusoidal obstruction syndrome (SOS or veno-occlusive disease) Peliosis hepatis Cold preservation injury

Table 2 Processes with Injury or De-differentiation of SEC Fibrotic liver disease Aging Sinusoidal obstruction syndrome (veno-occlusive disease) Nodular regenerative hyperplasia Peliosis hepatis Acetaminophen toxicity Cold preservation injury

of certain vascular beds by systemic thrombotic diseases may be a function of the local endothelial characteristics (4). This chapter will focus on the disorders that are associated with changes in the hepatic microcirculation (Table 2).

2. CAPILLARIZATION The porosity of SEC and the lack of an organized basement membrane are important in the normal functioning of the liver. In fibrotic liver diseases, there is de-differentiation of SEC with loss of fenestration and formation of a basement membrane, so-called capillarization (5–9). In alcoholic liver disease, capillarization has been shown to precede fibrosis (5). In the cirrhotic liver, the loss of SEC porosity and the formation of a basement membrane form a barrier that reduces oxygen delivery to hepatocytes (10). This impediment to oxygen delivery has been shown to impair oxygen-dependent hepatocyte functions such as oxidative drug metabolism (11–13). The SEC fenestration filters chylomicron remnants. The size of chylomicron remnants that pass the SEC barrier and are cleared by hepatocytes is determined by the size of the SEC fenestrae, so that changes in the porosity of SEC may markedly alter lipoprotein homeostasis (14,15). With aging, SEC fenestration decreases and the basement membrane becomes more pronounced, a process termed “pseudo-capillarization” (10,16). It has been suggested that pseudo-capillarization may contribute to age-related decreases in lipid clearance by the liver and thereby be a factor in the propensity for atherosclerosis with aging (2,15,17).

The hepatic sinusoidal endothelial cell

423

3. DRUG-INDUCED TOXICITY IN SINUSOIDAL ENDOTHELIAL CELLS Sinusoidal endothelial cells are the initial target of certain forms of drug-induced injury for several reasons. First, SEC are exposed to elevated concentrations of orally ingested drugs and toxins that are present in portal venous blood. Second, SEC are metabolically active cells with P450 activity and can therefore activate substrates taken up from sinusoidal blood (18–20). Finally, SEC are at risk for drug toxicity because of their location adjacent to hepatocytes. Metabolites exported across the sinusoidal pole of hepatocytes pass through the space of Disse and are diluted once they enter the sinusoidal blood. Sinusoidal endothelial cells are exposed to high concentrations of these metabolites in the space of Disse. The gradient in concentration of toxic metabolites generated by hepatocytes explains why for certain drugs the SEC are targeted in mild disease, whereas involvement of hepatic venular endothelial cells and endothelial cells in distant organs occurs with a more marked toxic insult. Types of liver injury that may occur when SEC are targeted by drugs include sinusoidal obstruction syndrome (SOS or veno-occlusive disease), nodular regenerative hyperplasia (NRH), sinusoidal dilatation, and peliosis hepatis. All four of these diseases may involve damage to hepatic sinusoidal and/or venular endothelial cells. A number of drugs have been linked to two or more of these diseases and in some patients all four of these lesions have been described in the same liver. Drugs that have been linked to more than one of these diseases include azathioprine (all four lesions), 6-thioguanine (peliosis hepatis, NRH, SOS), urethane (peliosis hepatis, SOS), thorotrast (peliosis hepatis, NRH), oral contraceptives (sinusoidal dilatation, peliosis hepatis, and NRH), and anabolic steroids (peliosis hepatis, NRH) (21–24).

4. SINUSOIDAL OBSTRUCTION SYNDROME This disease was originally called hepatic veno-occlusive disease, based on lesions of the central veins that were apparent on light microscopy (25). The first description of the disease in humans came from South Africa, where individuals had ingested bread made from inadequately winnowed wheat contaminated by plants containing pyrrolizidine alkaloids (26). In non-Western nations, SOS is still seen in individuals who ingest pyrrolizidine alkaloids, either in the form of “bush teas” or as contaminants of the food supply. The major plant species implicated are Crotalaria, Heliotropium, Senecio, and Symphytum. In North America and Western Europe, SOS occurs as a sporadic complication of chemotherapy for malignancy or of long-term immunosuppression by azathioprine for kidney and liver transplantation (22,27– 31). Sinusoidal obstruction syndrome has been described in association with chemotherapy at conventional doses with drugs such as gemtuzumab ozagamicin, actinomycin D, dacarbazine, cytosine arabinoside, mitramycin, 6-thioguanine, and urethane (24,32,33). Treatment of Wilm’s tumors, and in particular right-sided Wilm’s tumors, with actinomycin D plus abdominal irradiation is a risk factor for SOS (34,35). The most common cause of SOS in Western nations is the myeloablative conditioning therapy used prior to bone marrow

Endothelial cells in health and disease

424

transplantation for malignancy. The incidence of SOS in patients undergoing myeloablative conditioning regimens varies from 0% to 50% (36–41). The wide range in incidence is due to differences in patient selection criteria and differences in conditioning regimens, i.e., irradiation and chemotherapy. Chemotherapy regimens that include cyclopho-sphamide and regimens with higher doses of total body irradiation are more likely to cause SOS. Case fatality rates also vary widely, ranging from 0% to 67%, likely due to differences in diagnostic criteria. In regimens containing cyclophosphamide, the case fatality rate seems to be around 30% and this may be higher than for regimens without cyclophosphamide (36,40,42). The diagnosis of the disease is usually based on the clinical features, which include painful hepatomegaly, weight gain, and hyperbilirubinemia (37,39). A more extensive discussion of the clinical features of the disease can be found in a recent review (43).

4.1. Mechanisms of Sinusoidal Obstruction Syndrome The circulatory origin of the disease can be inferred from the clinical presentation. In contrast to other intrinsic liver diseases, in SOS symptoms of circulatory disruption precede rather than follow the decline in liver function. Originally the assumption was made that the circulatory obstruction stemmed from the venous changes. However, clinical studies have made clear that involvement of the central veins is not essential to development of the signs and symptoms of SOS (44). As will be described in the paragraphs that follow, studies in an experimental model of SOS also demonstrate that changes within the sinusoid initiate the disease. Based on the clinical and experimental studies that support the sinusoidal origin of the disease, it was proposed to change the name from hepatic veno-occlusive disease to SOS (43). Clinical studies demonstrate that the veins are more frequently involved with more severe disease (44). This suggests that the more marked the insult, the more extensively the endothelium is involved. Thus, the multiorgan failure that is associated with severe SOS may partially reflect a more widespread involvement of endothelial cells elsewhere in the body.

4.1.1. In Vitro Studies Sinusoidal endothelial cells can be isolated by elutriation with ≥98% purity and ≥99% viability. Yields for rat liver are ~80×106 cells per liver and for mouse are ~10– 12×106/liver. A variety of drugs and toxins implicated in SOS have been examined in in vitro studies of SEC in primary culture that were assayed within one to two days of isolation. The common findings for each of these compounds are that they are selectively more toxic to SEC than to hepatocytes and are detoxified by glutathione (GSH). A brief review of the factors that determine selective toxicity to SEC, as demonstrated in the in vitro studies, follows. Monocrotaline is a pyrrolizidine alkaloid, one of the “bush tea toxins.” It is one of the best-characterized compounds that cause SOS in humans. Toxicity requires metabolic activation by P450 to monocrotaline pyrrole and activation occurs only in the liver. SEC can activate monocrotaline. Monocrotaline pyrrole is detoxified by GSH and in vitro studies have shown that profound GSH depletion precedes SEC cell death. The GSH precursors that maintain SEC GSH prevent cell death in vitro. Monocrotaline is more

The hepatic sinusoidal endothelial cell

425

toxic to SEC than to hepatocytes and depletion of hepatocytes GSH does not exacerbate hepatocyte toxicity (45). The selective toxicity of monocrotaline to SEC is presumably due to the ability of SEC to form the monocrotaline pyrrole. Monocrotaline pyrrole formed in the liver is transported by red blood cells to the lung, leading to lung toxicity. Dacarbazine is only activated by P450 in the liver and is metabolically activated by SEC. Dacarbazine is toxic to SEC, but not to hepatocytes. Dacarbazine is detoxified by GSH, but GSH depletion does not render hepatocytes susceptible to toxicity (20). Thus, as with monocrotaline, these findings suggest that the selective toxicity to SEC is due to increased metabolic activation. Cyclophosphamide is the chemotherapeutic drug most frequently associated with SOS after bone marrow transplantation. Cyclophosphamide is P450 activated to 4hydroxycyclophosphamide, which then spontaneously tautomerizes to acrolein and phosphoramide mustard. Cyclophosphamide is not metabolically activated by SEC and is therefore not toxic when added to SEC cultured alone. In coculture studies of SEC with hepatocytes, in which hepatocytes can metabolize the cyclopho-sphamide, cyclophosphamide is toxic to SEC. The toxicity to SEC occurs within the therapeutic range, but toxicity to hepatocytes requires concentrations that greatly exceed the therapeutic range (46). Acrolein is the metabolite responsible for toxicity to endothelial cells, including SEC, and this occurs through profound GSH depletion. Depletion of hepatocyte GSH enhances toxicity, suggesting that the relative protection of hepatocytes against cyclophosphamide is due to the greater GSH detoxification capacity. The selective susceptibility of SEC is likely due to high concentrations of metabolite in the space of Disse, whereas downstream endothelial cells would be exposed to lower concentrations. Azathioprine is a prodrug that requires glutathione S-transferase catalyzed conjugation to GSH to form 6-mercaptopurine. Cells rapidly take up azathioprine, so that concentrations of orally ingested azathioprine are highest in the gut and in the liver. In vitro studies have demonstrated that profound GSH depletion precedes azathioprineinduced cell death in SEC, which suggests that GSH depletion plays a role in toxicity to SEC (45). Toxicity to SEC is greater than to hepatocytes in vitro. However, depletion of hepatocyte GSH abolishes the difference in susceptibility, demonstrating that the relative resistance of hepatocytes is due to greater GSH detoxification capacity. 6-Thioguanine, a metabolite downstream of 6-mercaptopurine, also causes SOS, so that the mechanism of injury must involve more than the GSH depletion that occurs when azathioprine is metabolized to 6-mercaptopurine. In summary, in vitro studies have demonstrated that SEC are selectively susceptible to a variety of toxins implicated in SOS through varying mechanisms. Each of the toxins studied share common features. All of the toxins are GSH detoxified and toxicity to SEC in vitro does not occur until SEC GSH is depleted.

4.1.2. In Vivo Studies The morphological events that occur in SOS and the biochemical underpinnings of these changes have been elucidated in the monocrotaline-induced rat model of the disease (Fig. 3). The model requires a single gavage of monocrotaline and follows a highly reproducible course subsequent to administration of the toxin. The time-course of events

Endothelial cells in health and disease

426

in the monocrotaline model can be divided into pre-SOS (0–48 hr after monocrotaline), early SOS (days 3–5), and late SOS (days 6 and 7) (47). In pre-SOS, there are minimal light microscopic changes and none of the “clinical signs” of SOS. However, morphological changes can be detected by scanning electron microscopy, transmission electron microscopy, and in vivo microscopy within the first 12 hr. Evaluation at 12 hr demonstrates loss of SEC fenestrae, formation of gaps within and between SEC and swelling or rounding up of SEC. During the course of the first 48 hr, red blood cells penetrate beneath the swollen cells into the space of Disse (48). As the swollen SEC block the sinusoids, the space of Disse becomes the path of least resistance and blood begins to flow in the space of Disse, with dissection of the sinusoidal lining. The Kupffer cells, SEC, and stellate cells embolize into the sinusoid and block sinusoidal flow. In early SOS, days 3–5, the predominant histological features are the centrilobular necrosis and hemorrhage and the loss of SEC and venular endothelial cells (Fig. 2). The “clinical signs” are similar to those described in the human disease: hepatomegaly with a 24–80% increase in liver weight, ascites formation, and hyper-bilirubinemia with values as high as 10 mg/dL. In late SOS, days 6 and 7, the centrilobular necrosis resolves completely. However, there is now venular fibrosis, with persistent damage to SEC and venular endothelial cells, and hemorrhage. The number of sinusoids containing flow reaches a nadir by day 4 and remains low through day 10. Kupffer cells decrease in number by day 1, but there is a progressive influx of monocytes adherent to areas denuded of SEC and venular endothelial cells and these aggregates of monocytes contribute to the sinusoidal obstruction by day 4.

Figure 2 Sinusoidal obstruction syndrome. Left panel: A normal liver.

The hepatic sinusoidal endothelial cell

427

The arrows indicate endothelial cells lining the central vein and the arrowheads indicated normal sinusoidal endothelial cells. Right panel: The centrilobular region of the liver in early SOS. Features of early SOS that can be seen are loss of endothelial cells within the central vein, loss of sinusoidal endothelial cells, severe centrilobular necrosis, hemorrhage, and congestion. The nucleated cells within the central vein are monocytes. (×20 magnification). Biochemical changes that occur during pre-SOS are important to consider, since this is the optimal time to intervene with a preventative strategy. Matrix metal-loproteinases (MMPs) are enzymes that break down extracellular matrix and allow cells to separate from the underlying matrix. MMP-9 expression and activity in the liver are increased 12 hr after monocrotaline administration. This activity increases markedly in the first 48 hr and then continues to rise through day 4 (49). There is a later, lower magnitude elevation of MMP-2 in the liver. In vitro studies of SEC, hepatocytes, stellate cells, and Kupffer cells reveal that SEC are the major source of both basal and monocrotaline-induced MMP-9/MMP-2 activity (49). This increased expression and activity of MMPs coincides with the progressive denudation of SEC lining, suggesting that this may account for the dehiscence of the SEC from the space of Disse. Prophylactic administration of inhibitors of MMP-9 and MMP-2 prevent the histological changes and the “clinical signs” of SOS, which confirms the contribution of MMPs to the disease (49). Monocrotaline is metabolically activated in the SEC (45). The reactive metabolite of monocrotaline, monocrotaline pyrrole, binds covalently within the endothelial cell to Factin (50). This leads to depolymerization of F-actin in SEC (49). Blocking of monocrotaline-induced F-actin depolymerization prevents the increase in MMP activity in SEC (49). The link between disassembly of F-actin and increased activity of MMPs has been previously reported for MMP-2 (51–53). In vivo microscopy shows that red blood cells penetrate into the space of Disse under SEC that are rounded up (48). The actin cytoskeleton plays a major role in cell shape and depolymerization of F-actin therefore leads to rounding up of cells. Release of MMPs on the abluminal side would loosen the tethering of SEC by breakdown of extracellular matrix in the space of Disse. F-actin depolymerization and increased MMP activity of SEC would therefore account for the rounding up of SEC and the ability of red blood cells to penetrate beneath these rounded up SEC. Red blood cells that enter the space of Disse dissect the endothelium off of the extracellular matrix. Nitric oxide (NO) levels in the hepatic vein decrease on day 1, drop by 70% on day 3 after monocrotaline and remain low through day 8. The initial decline in the first 24 hr appears to be due to the early loss of Kupffer cells, whereas the subsequent decrease is

Endothelial cells in health and disease

428

associated with the loss of SEC along the sinusoids (54). Inhibition of NO synthase increases the severity and extent of SOS after a subtoxic dose of monocrotaline. Conversely the infusion of V-PYRRO/NO, a liver selective nitric oxide prodrug, ameliorates the severity of SOS in a dose-dependent fashion (54). V-PYRRO/NO prevents the rounding up of the SEC, thereby preserving an intact SEC lining and maintaining sinusoidal perfusion. Studies in other organs have demonstrated that inhibition of endogenous NO production enhances I1–1β-induced increases in synthesis of MMP 9, whereas administration of NO donors inhibits new synthesis of MMPs (55– 61). V-PYRRO/NO also prevents synthesis of MMP-9 in the monocrotaline model. Taken together these findings demonstrate that the decrease in nitric oxide production due to loss of viable Kupffer cells and SEC contributes to the changes in the SEC, which then plays a role in the development of the disease. The process seems to be a positive feedback loop, whereby NO production falls and ultimately causes a further decline in NO (Fig. 3). Monocrotaline toxicity causes a loss of Kupffer cells and SEC. Due to loss of these cells, NO production decreases, the tonic suppression of MMP-9 synthesis by NO is lost, and this enhances increased

Figure 3 Proposed mechanisms in sinusoidal obstruction syndrome. This scheme depicts the interactions between the morphological and biochemical changes. The right side of the scheme shows the proposed positive feedback loop. MMP: matrix metalloproteinase; NO: nitric oxide; RBC: red blood cell; SEC: sinusoidal endothelial cells.

The hepatic sinusoidal endothelial cell

429

MMP-9 synthesis in the remaining SEC in which F-actin is depolymerized. Increased MMP-9 activity allows SEC to be dissected off the matrix in the space of Disse, which results in loss of SEC from the sinusoid; and NO diminishes further.

5. NODULAR REGENERATIVE HYPERPLASIA Nodular regenerative hyperplasia (NRH) is most commonly an asymptomatic disorder that is detected as an incidental finding at autopsy. Based on large autopsy series, the prevalence is around 2.5% (21,62). Symptomatic disease presents with signs of portal hypertension, notably variceal hemorrhage, ascites, and splenomegaly, but only rarely as end-stage liver disease. It is not known what causes NRH, but it has been postulated that NRH is due to impaired perfusion of areas of the liver with reactive hyperplasia in parts of the liver where perfusion is maintained (63). In the areas of hypoperfusion, hepatocytes become apoptotic or atrophic (64). The risk factors for NRH are widely disparate in nature, but share a predisposition to impair portal venous or sinusoidal blood flow (Table 3). NRH may occur in patients given long-term azathioprine immunosuppression for kidney or liver transplantation, or in patients who receive chemotherapy for bone marrow transplantation. Both azathioprine (45) and high-dose chemotherapy for bone marrow transplantation (46) are toxic to SEC, and have been implicated in other lesions, in which the SEC are the putative target, such as peliosis hepatis and SOS (22,44,65). Thus NRH in renal, liver,

Table 3 Risk Factors for Nodular Regenerative Hyperplasia Autoimmune diseases Rheumatoid arthritis Systemic lupus erythematosus Antiphospholipid syndrome Polyarteritis nodosa Scleroderma Myasthenia gravis Hematologic disorders Polycythemia vera Essential thrombocythemia Agnogenic myeloid metaplasia Chronic myeloid leukemia Lymphoma

Endothelial cells in health and disease

430

Multiple myeloma Cryoglobulinemia Xenobiotic induced Anabolic steroids Azathioprine Oral contraceptives Chemotherapy for bone marrow transplantation Thoratrast 6-Thioguanine Toxic oil syndrome

or bone marrow transplantation seems to be due to SEC toxicity and the consequent regional impairment of the microcirculation.

6. PELIOSIS HEPATIS Peliosis is a rare liver disease, in which blood filled cavities develop throughout the liver. The peliotic lesions vary from less than 1 mm to several centimeters. Peliosis is most common in the liver, but may also involve the spleen, bone marrow, or abdominal lymph nodes. In addition to peliosis due to drugs such as azathioprine and anabolic steroids, peliosis may occur in patients with AIDS, tuberculosis, leukemia, lymphoma, multiple myeloma, and myeloproliferative diseases (Table 4). The most clear-cut evidence for the SEC origin of peliosis hepatis comes from studies in AIDS patients. In this population, peliosis is due to infection with Bartonella species bacilli. Electron microscopy studies have demonstrated the presence of Bartonella species in SEC (66). This leads to disruption of the SEC lining, sinusoidal dilatation and ultimately to formation of cavities that lack SEC lining (67,68). Later in the disease the endothelial lining of the cavities may be partially restored. When Bartonella involves the organs of the reticuloendothelial system, which have a discontinuous endothelium, peliotic cavities develop, whereas Bartonella infection in the skin, which has a continuous endothelial lining, causes bacillary angiomatosis (68).

7. ACETAMINOPHEN TOXICITY In the early stages after acetaminophen toxicity, there is extreme hepatic congestion in both humans (69–73) and experimental animals (74–79). In the mouse the congestion is so extreme, that up to half of the red blood cell volume can accumulate in the liver (78). Ultrastructural studies have demonstrated the appearance of large pores in the SEC, accumulation of red blood cells in the space of Disse, and partial separation of SEC from

The hepatic sinusoidal endothelial cell

431

the underlying hepatocytes but without complete dehiscence (80). These circulatory changes precede hepatocyte necrosis and likely contribute to the liver toxicity. In vitro studies in murine liver cells have shown that acetaminophen is more toxic to SEC than to hepatocytes (81) and this may account for the microcirculatory impairment.

Table 4 Risk Factors for Peliosis Hepatis Wasting Illnesses

Xenobiotic Induced

AIDS (Bartonella bacilli)

Anabolic steroids

Macroglobulinemia

Arsenic

Myeloproliferative diseases

Azathioprine

Multiple myeloma

Oral contraceptives

Leukemia

6-Thioguanine

Lymphoma

Thoratrast

Tuberculosis

Vinyl chloride

8. COLD PRESERVATION INJURY In liver transplantation, the donor liver is preserved and transported in a cold University of Wisconsin solution. A variable degree of injury may manifest itself when the liver is reperfused in the recipient of the transplant and this injury can lead to graft failure (see Chapter 25). The major target for reperfusion injury are the SEC (82–85), with cell death, partial denudation of the sinusoidal lining and deterioration of the liver microcirculation. Cold preservation leads to increased calpain protease activity in SEC (86–88), which leads to depolymerization of F-actin in SEC (89), and consequent upregulation of MMP9 and MMP-2 activity (90). Upregulation of MMPs likely accounts for the dehiscence of SEC from the space of Disse in this setting. One of the other consequences of F-actin depolymerization and increased MMP activity is increased platelet adherence to SEC (91). This accumulation of platelets and leukocytes contributes to SEC injury after cold preservation (91–93).

9. CONCLUSION Sinusoidal endothelial cells are small, flat cells that are difficult to visualize by light microscopy. The procedure for isolating pure SEC is relatively new and fairly complex. It is therefore not surprising that these cells were overlooked for so long as a target in liver injury. It seems likely that with the more widespread availability of improved imaging techniques and the increasing number of laboratories that isolate SEC, an ever increasing number of diseases will be found that target SEC.

Endothelial cells in health and disease

432

At this point, there are no therapies that are specifically geared towards the SEC. Matrix metalloproteinase inhibitors will need to be examined as prophylactic therapy for SOS and perhaps for cold preservation injury. As additional forms of liver injury are identified that target SEC, therapy may be able to utilize the ability of SEC to phagocytose small sized particles.

REFERENCES 1. Chi JT, Chang HY, Haraldsen G, Jahnsen FL, Troyanskaya OG, Chang DS, Wang Z, Rockson SG, van de Rijn M, Botstein D, Brown PO. Endothelial cell diversity revealed by global expression profiling. Proc Natl Acad Sci USA 2003; 100:10673–10628. 2. Wisse E, De Zanger RB, Charels K, van der Smissen P, McCuskey RS. The liver sieve: considerations concerning the structure and function of endothelial fenestra, the sinusoidal wall and the space of Disse. Hepatology 1985; 5:683–692. 3. DeLeve LD. Vascular diseases of the liver. In: Powell DW, ed. Textbook of Gastroenterology. Vol. 2. Philadelphia: Lippincott, Williams and Wilkins, 2003:2517–2525. 4. Aird WC. Vascular bed-specific hemostasis: role of endothelium in sepsis pathogenesis. Crit Care Med 2001; 29:S28–S34. 5. Schaffner F, Popper H. Capillarization of hepatic sinusoids in man. Gastroenterology 1963; 44:239–242. 6. Horn T, Christoffersen P, Henriksen JH. Alcoholic liver injury: defenestration in noncirrhotic livers—a scanning electron microscopic study. Hepatology 1987; 7:77–82. 7. Urashima S, Tsutsumi M, Nakase K, Wang JS, Takada A. Studies on capillarization of the hepatic sinusoids in alcoholic liver disease. Alcohol Alcohol 1993; S1B:77–84. 8. Mori T, Okanoue T, Sawa Y, Hori N, Ohta M, Kagawa K. Defenestration of the sinusoidal endothelial cell in a rat model of cirrhosis. Hepatology 1993; 17:891–897. 9. Bhunchet E, Fujieda K. Capillarization and venularization of hepatic sinusoids in porcine seruminduced rat liver fibrosis: a mechanism to maintain liver blood flow. Hepatology 1993; 18:1450–1458. 10. Le Couteur DG, Cogger VC, Markus AM, Harvey PJ, Yin ZL, Annselin AD, McLean AJ. Pseudocapillarization and associated energy limitation in the aged rat liver. Hepatology 2001; 33:537–543. 11. Hickey PL, Angus PW, McLean AJ, Morgan DJ. Oxygen supplementation restores theophylline clearance to normal in cirrhotic rats. Gastroenterology 1995; 108:1504–1509. 12. Le Couteur DG, Hickey H, Harvey PJ, Gready J, McLean AJ. Hepatic artery flow and propranolol metabolism in perfused cirrhotic rat liver. J Pharmacol Exp Ther 1999; 289:1553– 1558. 13. Froomes PRA, Morgan DJ, Smallwood RA, Angus PW. Comparative effects of oxygen supplementation on theophylline and acetaminophen clearance in human cirrhosis. Gastroenterology 1999; 116:915–920. 14. Wisse E, DeZanger RB, Jacobs R, McCuskey RS. Scanning electron microscopic observations on the structure of portal veins, sinusoids and central veins. Scanning Electron Microscopy 1983; III:1441–1452. 15. Fraser R, Dobbs BR, Rogers GW. Lipoproteins and the liver sieve: the role of the fenestrated sinusoidal endothelium in lipoprotein metabolism, atherosclerosis, and cirrhosis. Hepatology 1995; 21:863–874. 16. McLean AJ, Cogger VC, Chong GC, Warren A, Markus AM, Dahlstrom JE, Le Couteur DG. Age-related pseudocapillarization of the human liver. J Pathol 2003; 200: 112–117.

The hepatic sinusoidal endothelial cell

433

17. Le Couteur DG, Fraser R, Cogger VC, McLean AJ. Hepatic pseudocapillarisation and atherosclerosis in ageing. Lancet 2002; 359:1612–1615. 18. Steinberg P, Lafranconi WM, Wolf CR, Waxman DJ, Oesch F, Friedberg T. Xenobiotic metabolizing enzymes are not restricted to parenchymal cells in rat liver. Mol Pharmacol 1987; 32:463–470. 19. Steinberg P, Schlemper B, Molitor E, Platt KL, Seidel A, Oesch F. Rat liver endothelial and Kupffer cell-mediated mutagenicity of polycyclic aromatic hydrocarbons and aflatoxin B1. Environ Health Perspect 1990; 88:71–76. 20. DeLeve LD. Dacarbazine toxicity in murine liver cells: a novel model of hepatic endothelial injury and glutathione defense. J Pharmacol Exp Ther 1994; 268:1261–1270. 21. Wanless IR. Micronodular transformation (nodular regenerative hyperplasia) of the liver: a report of 64 cases among 2,500 autopsies and a new classification of benign hepatocellular nodules. Hepatology 1990; 11:787–797. 22. Haboubi NY, Ali HH, Whitwell HL, Ackrill P. Role of endothelial cell injury in the spectrum of azathioprine-induced liver disease after renal transplant: light microscopy and ultrastructural observations. Am J Gastroenterol 1988; 83:256–261. 23. Zafrani ES, Cazier A, Baudelot AM, Feldmann G. Ultrastructural lesions of the liver in human peliosis. A report of 12 cases. Am J Pathol 1984; 114:349–359. 24. DeLeve LD. Cancer chemotherapy. In: DeLeve LD, ed. Drug-induced Liver Disease. Chapter 24. New York: Marcel Dekker, 2002:593–632. 25. Bras G, Jeliffe DB, Stuart KL. Veno-occlusive disease of the liver with non-portal type of cirrhosis occurring in Jamaica. Arch Pathol 1954; 57:285–300. 26. Willmot FC, Robertson GW. Senecio disease, or cirrhosis of the liver due to senecio poisoning. Lancet 1920; 2:848–849. 27. Eisenhauer T, Hartmann H, Rumpf KW, Helmchen U, Scheler F, Creutzfeldt W. Favourable outcome of hepatic veno-occlusive disease in a renal transplant patient receiving azathioprine, treated by portacaval shunt. Report of a case and review of the literature. Digestion 1984; 30:185–190. 28. Katzka DA, Saul SH, Jorkasky D, Sigal H, Reynolds JC, Soloway RD. Azathioprine and hepatic venocclusive disease in renal transplant patients. Gastroenterology 1986; 90:446–454. 29. Read AE, Wiesner RH, LaBrecque DR, Tifft JG, Mullen KD, Sheer RL, Petrelli M, Ricanati ES, McCullough AJ. Hepatic veno-occlusive disease associated with renal transplantation and azathioprine therapy. Ann Intern Med 1986; 104:651–655. 30. Liano F, Moreno A, Matesanz R, Teruel JL, Redondo C, Garcia Martin F, Orte L, Ortuno J. Veno-occlusive hepatic disease of the liver in renal transplantation: is azathioprine the cause? [see comments]. Nephron 1989; 51:509–516. 31. Sterneck M, Wiesner R, Ascher N, Roberts J, Ferrell L, Ludwig J, Lake J. Azathioprine hepatotoxicity after liver transplantation. Hepatology 1991; 14:806–810. 32. Giles FJ, Kantarjian HM, Kornblau SM, Thomas DA, Garcia-Manero G, Waddelow TA, David CL, Phan AT, Colburn DE, Rashid A, Estey EH. Mylotarg (gemtuzumab ozogamicin) therapy is associated with hepatic venoocclusive disease in patients who have not received stem cell transplantation. Cancer 2001; 92:406–413. 33. Rajvanshi P, Shulman HM, Sievers EL, McDonald GB. Hepatic sinusoidal obstruction following Gemtuzumab Ozogamicin (Mylotarg®). Blood 2002:2310–2314. 34. Tornesello A, Piciacchia D, Mastrangelo S, Lasorella A, Mastrangelo R. Venoocclusive disease of the liver in right-sided Wilims’ tumours. Eur J Cancer 1998; 34:1220–1223. 35. Czauderna P, Katski K, Kowalczyk J, Kurylak A, Lopatka B, Skotnicka-Klonowicz G, SawiczBirkowska K, Godzinski J. Venoocclusive liver disease (VOD) as a complication of Wilms’ tumour management in the series of consecutive 206 patients. Eur J Pediatr Surg 2000; 10:300– 303.

Endothelial cells in health and disease

434

36. McDonald GB, Hinds MS, Fisher LD, Schoch HG, Wolford JL, Banaji M, Hardin BJ, Shulman HM, Clift RA. Veno-occlusive disease of the liver and multiorgan failure after bone marrow transplantation—a cohort study of 355 patients. Ann Intern Med 1993; 118:255–267. 37. McDonald GB, Sharma P, Matthews DE, Shulman HM, Thomas ED. Veno-occlusive disease of the liver after bone marrow transplantation: diagnosis, incidence and predisposing factors. Hepatology 1984; 4:116–122. 38. Ganem G, Saint-Marc Girardin MF, Kuentz M, Cordonnier C, Morinello G, Teboul C, Braconnier F, Vernant JP, Dhumeaux D, Le Bourgeois JP. Venoocclusive disease of the liver after allogeneic bone marrow transplantation in man. Int J Rad Biol Phys 1988; 14:879–884. 39. Jones RJ, Lee KSK, Beschorner WE, Vogel VG, Grochow LB, Vogelsang GB, Sensenbrenner LL, Santos GW, Saral R. Veno-occlusive disease of the liver following bone marrow transplantation. Transplantation 1987; 44:778–783. 40. Carreras E, Bertz H, Arcese W, Vernant JP, Tomas JF, Hagglund H, Bandini G, Esperou H, Russell J, de la Rubia J, Di Girolamo G, Demuynck H, Hartmann O, Clausen J, Ruutu T, Leblond V, Iriondo A, Bosi A, Ben-Bassat I, Koza V, Gratwohl A, Apperley JF. Incidence and outcome of hepatic veno-occlusive disease after blood or marrow transplantation: a prospective cohort study of the European Group for Blood and Marrow Transplantation. European Group for Blood and Marrow Transplantation Chronic Leukemia Working Party. Blood 1998; 92:3599–3604. 41. Hasegawa S, Horibe K, Kawabe T, Kato K, Kojima S, Matsuyama T, Hirabayashi N. Venoocclusive disease of the liver after allogeneic bone marrow transplantation in children with hematologic malignancies: incidence, onset time and risk factors. Bone Marrow Transplantation 1998; 22:1191–1197. 42. Lee JL, Gooley T, Bensinger W, Schiffman K, McDonald GB. Veno-occlusive disease of the liver after busulfan, melphalan, and thiotepa conditioning therapy: incidence, risk factors, and outcome. Biol Blood Marrow Transplant 1999; 5:306–315. 43. DeLeve LD, Shulman HM, McDonald GB. Toxic injury to hepatic sinusoids: sinusoidal obstruction syndrome (venoocclusive disease). Semin Liver Dis 2002; 22:623–638. 44. Shulman HM, Fisher LB, Schoch HG, Henne KW, McDonald GB. Venoocclusive disease of the liver after marrow transplantation: histological correlates of clinical signs and symptoms. Hepatology 1994; 19:1171–1180. 45. DeLeve LD, Wang X, Kuhlenkamp JF, Kaplowitz N. Toxicity of azathioprine and monocrotaline in murine sinusoidal endothelial cells and hepatocytes: the role of glutathione and relevance to hepatic venooclusive disease. Hepatology 1996; 23:589–599. 46. DeLeve LD. Cellular target of cyclophosphamide toxicity in the murine liver: role of glutathione and site of metabolic activation. Hepatology 1996; 24:830–837. 47. DeLeve LD, McCuskey RS, Wang X, Hu L, McCuskey MK, Epstein RB, Kanel G. Characterization of a reproducible rat model of hepatic veno-occlusive disease. Hepatology 1999; 29:1779–1791. 48. DeLeve LD, Ito I, Bethea NW, McCuskey MK, Wang X, McCuskey RS. Embolization by sinusoidal lining cell obstructs the microcirculation in rat sinusoidal obstruction syndrome. Am J Physiol-Gastrointestinal Liver Physiol 2003; 284:G1045–G1052. 49. DeLeve LD, Wang X, Tsai J, Kanel GC, Strasberg SM, Tokes ZA. Prevention of sinusoidal obstruction syndrome (hepatic venoocclusive disease) in the rat by matrix metal-loproteinase inhibitors. Gastroenterology 2003; 125:882–890. 50. Lamé MW, Jones AD, Wilson DW, Dunston SK, Segall HJ. Protein targets of monocro-taline pyrrole in pulmonary artery endothelial cells. J Biol Chem 2000; 275:29091–29099. 51. Werb Z, Hembry RM, Murphy G, Aggeler J. Commitment to expression of the metalloendopeptidases, collagenase and stromelysin: relationship of inducing events to changes in cytoskeletal architecture. J Cell Biol 1986; 102:697–702.

The hepatic sinusoidal endothelial cell

435

52. Allenberg M, Weinstein T, Li I, Silverman M. Activation of procollagenase IV by cytochalasin D and concanavalin A in cultured rat mesangial cells: linkage to cytoskeletal reorganization. J Am Soc Nephrol 1994; 4:1760–1770. 53. MacDougall JR, Kerbel RS. Constitutive production of 92-kDa gelatinase B can be suppressed by alterations in cell shape. Exp Cell Res 1995; 218:508–515. 54. DeLeve LD, Wang X, Kanel GC, Tokes ZA, Tsai J, Ito Y, Bethea NW, McCuskey MK, McCuskey RS. Decreased hepatic nitric oxide production contributes to the development of rat sinusoidal obstruction syndrome. Hepatology 2003; 38:900–908. 55. Eberhardt W, Beeg T, Beck KF, Walpen S, Gauer S, Bohles H, Pfeilschifter J. Nitric oxide modulates expression of matrix metalloproteinase-9 in rat mesangial cells. Kidney Int 2000; 57:59–69. 56. Upchurch GR Jr, Ford JW, Weiss SJ, Knipp BS, Peterson DA, Thompson RW, Eagleton MJ, Broady AJ, Proctor MC, Stanley JC. Nitric oxide inhibition increases matrix metalloproteinase9 expression by rat aortic smooth muscle cells in vitro. J Vasc Surg 2001; 34:76–83. 57. Matsunaga T, Weihrauch DW, Moniz MC, Tessmer J, Warltier DC, Chilian WM. Angiostatin inhibits coronary angiogenesis during impaired production of nitric oxide. Circulation 2002; 105:2185–2191. 58. Gurjar MV, DeLeon J, Sharma RV, Bhalla RC. Mechanism of inhibition of matrix metalloproteinase-9 induction by NO in vascular smooth muscle cells. J Appl Physiol 2001; 91:1380–1386. 59. Eagleton MJ, Peterson DA, Sullivan W, Roelofs KJ, Ford JA, Stanley JC, Upchurch GR Jr. Nitric oxide inhibition increases aortic wall matrix metalloproteinase-9 expression. J Surg Res 2002; 104:15–21. 60. Mujumdar VS, Aru GM, Tyagi SC. Induction of oxidative stress by homocyst(e)ine impairs endothelial function. J Cell Biochem 2001; 82:491–500. 61. Jurasz P, Sawicki G, Duszyk M, Sawicka J, Miranda C, Mayers I, Radomski MW. Matrix metalloproteinase 2 in tumor cell-induced platelet aggregation: regulation by nitric oxide. Cancer Res 2001; 61:376–382. 62. Nakanuma Y. Nodular regenerative hyperplasia of the liver: retrospective survey in autopsy series. J Clin Gastroenterol 1990; 12:460−465. 63. Wanless IR, Godwin TA, Allen F, Feder A. Nodular regenerative hyperplasia of the liver in hematologic disorders: a possible response to obliterative portal venopathy. A morphometric study of nine cases with an hypothesis on the pathogenesis. Medicine 1980; 59:367–379. 64. Shimamatsu K, Wanless IR. Role of ischemia in causing apoptosis, atrophy, and nodular hyperplasia in human liver. Hepatology 1997; 26:343–350. 65. Snover DC, Weisdorf S, Bloomer J, McGlave P, Weisdorf D. Nodular regenerative hyperplasia of the liver following bone marrow transplantation. Hepatology 1989; 9:443−448. 66. Leong SS, Cazen RA, Yu GS, LeFevre L, Carson JW. Abdominal visceral peliosis associated with bacillary angiomatosis. Ultrastructural evidence of endothelial destruction by bacilli. Arch Pathol Lab Med 1992; 116:866–871. 67. Scoazec JY, Marche C, Girard PM, Houtmann J, Durand-Schneider AM, Saimot AG, Benhamou JP, Feldmann G. Peliosis hepatis and sinusoidal dilation during infection by the human immunodeficiency virus (HIV). An ultrastructural study. Am J Pathol 1988; 131:38–47. 68. Goerdt S, Sorg C. Endothelial heterogeneity and the acquired immunodeficiency syndrome: a paradigm for the pathogenesis of vascular disorders. Clin Invest 1992; 70:89–98. 69. Rose PG. Paracetamol overdose and liver damage. Br Med J 1969; 1:381–382. 70. Thompson RPH, Clark R, Wilson RA, Borirakchanyavat V, Widdop B, Goulding R, Williams R. Hepatic damage from overdose of paracetamol. Gut 1972; 13:836. 71. Zimmerman HJ. Effects of aspirin and acetaminophen on the liver. Arch Intern Med 1981; 141:333–342. 72. Klatskin G, Conn HO. Histopathology of the Liver. Vol. 7. New York, Oxford: Oxford University Press. 1993:111–142.

Endothelial cells in health and disease

436

73. Zimmerman HJ. Syndromes of environmental hepatotoxins. Hepatotoxicity—the adverse effect of drugs and other chemicals on the liver. New York: Appleton-Century-Crofts, 1978:279–302. 74. Dixon MF, Nimmo J, Prescott LF. Experimental paracetamol-induced hepatic necrosis: a histopathological study. J Pathol 1971; 103:225–229. 75. Dixon MF, Dixon B, Aparicio SR, Loney DP. Experimental paracetamol-induced hepatic necrosis: a light- and electron-microscope, and histochemical study. J Pathol 1975; 116:17–29. 76. Miller DJ, Pichanick GG, Fiskerstrand C, Saunders S. Hepatic erythrocyte sequestration as a cause of acute anaemia. Am J Dig Dis 1977; 22:1055–1059. 77. Chiu S, Bhakthan NMG. Experimental acetaminophen-induced hepatic necrosis: biochemical and electron microscopic study of cyteamine protection. Lab Invest 1978; 39:193–203. 78. Walker RM, Massey TE, McElligott TF, Racz WJ. Acetaminophen-induced hypothermia, hepatic congestion, and modification by N-acetylcysteine in mice. Toxicol Appl Pharmacol 1981; 59:500–507. 79. Walker RM, Racz WJ, McElligott TF. Acetaminophen-induced hepatotoxic congestion in mice. Hepatology 1985; 5:233–240. 80. Walker RM, Racz WJ, McElligott TF. Scanning electron microscopic examination of acetaminophen-induced hepatotoxicity and congestion in mice. Am J Physiol 1983; 113:321– 330. 81. DeLeve LD, Wang X, Kaplowitz N, Shulman HM, Bart JA, van der Hoek A. Sinusoidal endothelial cells as a target for acetaminophen toxicity: direct action versus requirement for hepatocyte activation in different mouse strains. Biochem Pharmacol 1997; 53:1339–1345. 82. Caldwell-Kenkel C, Thurman RG, Lemasters JJ. Selective loss of nonparenchymal cell viability after cold ischemic storage of rat livers. Transplantation 1988; 45:834–837. 83. McKeown CMB, Edwards V, Phillips MJ, Harvey PRC, Petrunka CN, Strasberg SM. Sinusoidal lining cell damage: the critical injury in cold preservation of liver allografts in the rat. Transplantation 1988; 46:178–191. 84. Caldwell-Kenkel JC, Currin RT, Tanaka Y, Thurman RG, Lemasters JJ. Reperfusion injury to endothelial cells following cold ischemic storage of rat livers. Hepatology 1989; 10:292–299. 85. Imamura H, Brault A, Huet PM. Effects of extended cold preservation and transplantation on the rat liver microcirculation. Hepatology 1997; 25:664–671. 86. Aguilar HI, Steers JL, Wiesner RH, Krom RA, Gores GJ. Enhanced liver calpain protease activity is a risk factor for dysfunction of human liver allografts. Transplantation 1997; 63:612– 614. 87. Kohli V, Gao W, Camargo CA Jr, Clavien PA. Calpain is a mediator of preservationreperfusion injury in rat liver transplantation. Proc Natl Acad Sci USA 1997; 94:9354–9359. 88. Upadhya GA, Topp SA, Hotchkiss RS, Anagli J, Strasberg SM. Effect of cold preservation on intracellular calcium concentration and calpain activity in rat sinusoidal endothelial cells. Hepatology 2003; 37:313–323. 89. Upadhya GA, Strasberg SM. Evidence that actin disassembly is a requirement for matrix metalloproteinase secretion by sinusoidal endothelial cells during cold preservation in the rat. Hepatology 1999; 30:169–176. 90. Upadhya GA, Harvey RP, Howard TK, Lowell JA, Shenoy S, Strasberg SM. Evidence of a role for matrix metalloproteinases in cold preservation injury of the liver in humans and in the rat. Hepatology 1997; 26:922–928. 91. Upadhya GA, Strasberg SM. Platelet adherence to isolated rat hepatic sinusoidal endothelial cells after cold preservation. Transplantation 2002; 73:1764–1770. 92. Sindram D, Porte RJ, Hoffman MR, Bentley RC, Clavien PA. Synergism between platelets and leukocytes in inducing endothelial cell apoptosis in the cold ischemic rat liver: a Kupffer cellmediated injury. FASEB J 2001; 15:1230–1232. 93. Lasnier E, Blanc MC, Housset C, Rey C, Roch-Arveiller M, Vaubourdolle M. Cytotoxic response of sinusoidal endothelial cells to polymorphonuclear leukocytes and its potential implication in hypoxia-reoxygenation injury. Liver 2002; 22:495–500.

22 Endothelium and Hemostasis William C.Aird Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts, U.S.A.

1. INTRODUCTION Vascular thrombotic disorders are among the most common causes of morbidity and mortality in the Western world. A remarkable feature of these disorders is the focal nature of their distribution (Tables 1 and 2). For example, veno-occlusive disease of the liver, a relatively common complication of myeloablative treatment and bone marrow transplantation, affects the sinusoids of the liver. Hemolytic uremia syndrome and thrombotic thrombocytopenia purpura are part of a spectrum of micro-angiopathic hemolytic anemias that are characterized by pathology in virtually all microvascular beds with the notable exception of the liver and lung. The congenital hypercoagulable states, as exemplified by the factor V Leiden mutation, predispose patients to an increased risk of venous, but not arterial, thrombosis (1,2). Perhaps

Table 1 Hypercoagulable States Associated with Disorders of Primary Hemostasisa Disease/Disorder Acquired

Congenital a

Site of Thrombosis

TTP

All organs except lung and liver

HUS

Predominantly kidney

MPD

Portal/hepatic veins

PNH

Portal/hepatic veins

DIC

Microvessels; all organs are susceptible

HIT

Arteries, veins, often in unusual sites

APL

Arteries and veins

Atherosclerosis

Conduit arteries

Sickle cell disease

Microvessels, especially joints

Primary and secondary hemostasis are integrally linked; therefore most of these diseases are also associated with activation of the clotting cascade. This is particularly true in the

Endothelial cells in health and disease

438

case of DIC, HIT, and APL. TTP: thrombotic thrombocytopenic purpura; HUS: hemolytic uremic syndrome; MPS: myeloproliferative disease; PNH: paroxysmal nocturnal hemoglobinuria; DIC: disseminated intravascular coagulation; HIT: heparin induced thrombocytopenia; APL: antiphospholipid antibody syndrome.

Table 2 Hypercoagulable States Associated with Disorders of Secondary Hemostasis Disease/Disordera Acquired

Pregnancy

Site of Thrombosis

S/B

DVT

Immobilization

S

DVT

OCP

B

DVT

V/A catheters

T

Site of catheter

Trauma

T

Site of trauma

Surgery

S/T

Site of surgery, DVT

Sepsis

S/T/B

Multiple (but not all) organs

Congestive heart failure

S

DVT

DIC

B

Multiple (but not all) organs

HIT

T/B

Arteries and veins

APL

B

Arteries and veins

Cancer Atherosclerosis Organ transplantation

Congenital

Virchow’s Triadb

S/T/B T/B

DVT AMI, stroke, peripheral artery

S/T/B

Transplanted organ

Hyperhomocysteinemia

B

Venous and arterial

Nephrotic syndrome

B

DVT, renal vein thrombosis

ATIII deficiency

B

DVT

PC deficiency

B

DVT

PS deficiency

B

DVT

Prothrombin mutation

B

DVT

V Leiden

B

DVT

Fibrinolytic defects (e.g.

B

DVT

Endothelium and hemostasis

439

plasminogen deficiency) Dysfibrinogenemia

B

Sickle cell disease

S/T/B

Hyperhomocysteinemia

B

DVT Different organs Arteries and veins

a

OCP: oral contraceptive pill; V/A: venous or arterial; ATIII: antithrombin III; PC: protein C; PS: protein S; DVT: deep venous thrombosis; AMI: acute myocardial infarction. b Listed are the major contributors in Virchow’s triad (as originally defined). S: stasis; B: blood constituents; T: trauma to blood vessel wall. Virtually every disease has an endothelial component. the most striking example of the systemic disorder-local thrombosis paradox is warfarininduced skin necrosis. This rare complication of warfarin develops during a narrow window of time following the initiation of treatment. The syndrome is characterized by disproportionately low levels of functional protein C, compared with levels of other vitamin k-dependent factors. Although the imbalance in vitamin k-dependent factors is distributed throughout the circulation and bathes every vascular bed, fibrin deposition is remarkably limited to the postcapillary venules of the dermis, particularly in the areas of the breasts and buttocks (3,4). In the final analysis, there does not exist a single thrombotic diathesis that affects virtually every blood vessel type in the body (Tables 1 and 2) (2). This observation is supported by genetic mouse models, in which the deletion of one or another natural anticoagulant mechanisms results in vascular bed-specific fibrin deposition and thrombosis (5). An important question that arises from these clinical and animal studies is how a systemic imbalance in hemostasis is ultimately manifested by local rather than diffuse vasculopathic lesions. In fact, as will be discussed below, the endothelium provides an important clue to the answer.

2. CLASSIFICATIONS IN HEMOSTASIS Hemostasis is strictly defined as the arrest of bleeding. Coagulation is the transformation of a liquid into a semisolid or solid coherent mass. In day-to-day practice, the terms hemostasis and coagulation are used interchangeably to describe the physiological process by which blood is maintained in a fluid state within the closed circulation. Hemostasis may be classified according to several schemes, as outlined below.

2.1. Primary vs. Secondary Hemostasis Primary hemostasis refers to the cellular (platelet) response, whereas secondary hemostasis refers to the soluble circulating clotting factors that converge in a cascade of enzymatic reactions to generate the end product, fibrin. In truth, the cellular and protein components of hemostasis are highly coordinated and interdependent; they function in unison in both space and time. Nevertheless, a conceptual distinction between these two

Endothelial cells in health and disease

440

compartments is helpful when considering mechanistic, diagnostic, and therapeutic implications. Platelets arise through membrane budding of terminally differentiated precursor cells, termed megakaryocytes, which are located in the bone marrow. The megakaryocyte is one of the largest cells in the mammalian body and one of the few cells in the animal kingdom with a hyperdiploid or polyploid nucleus. Once released into the blood, platelets circulate freely as small discoidshaped anucleate cells. When activated, platelets undergo a change in shape, adhere, and aggregate. In the process, they present a newly activated cell surface for assembly of the clotting cascade—just one example of the cross talk that occurs between primary and secondary hemostasis. Assays for primary hemostasis include an inspection of the peripheral blood smear, platelet count, and a small number of ancillary tests, including platelet aggregation studies and heparininduced thrombocytopenia (HIT) antibodies. Secondary hemostasis—or the protein component of coagulation—consists of circulating proteins derived almost exclusively from liver hepatocytes (exceptions include factor VIII, tissue factor (TF), and some of the anticoagulant proteins). The clotting cascade, which is often depicted in medical textbooks as an intricate maze of seemingly arbitrary pathways, lies outside the “comfort zone” of most health care providers. Indeed, the complexity of the coagulation mechanism has been cited as evidence for the existence of Divine intervention, and has found itself of all places in the middle of a heated, yet somewhat amusing debate between evolutionary biologists and Christian fundamentalists (6). Some have argued that the coagulation mechanism is irreducibly complex, and could not possibly have arisen through step-by-step modification and natural selection. As untenable as this latter position is, it nevertheless illustrates the challenges in understanding and teaching a system as complex as coagulation. In approaching secondary hemostasis, there are several important themes to consider. First, it serves us well to recall the bottom line: conversion of fibrinogen to fibrin, a process that is mediated by the serine protease, thrombin. Fibrin is an insoluble glue or scaffold that strengthens the cellular clot. Fibrin may also play an important role in containing pathogens. Second, there are two pathways that promote thrombin generation: the extrinsic pathway, which is initiated by TF, and the intrinsic pathway. Third, the clotting cascade is initiated through TF-mediated activation of factor VII (extrinsic pathway) (Fig. 1). Tissue factor is a transmembrane protein that is expressed on the surface of activated monocytes, in the subendothelial

Endothelium and hemostasis

441

Figure 1 The coagulation pathway. The clotting cascade consists of an extrinsic pathway (EP; tissue factor, factor VII), an intrinsic pathway (IP; factors XI, IX, VIII), and a common pathway (CP; factors X, V, (pro)thrombin and fibrinogen). Factors VII, XI, IX, X, II (thrombin) are serine proteases; factors V and VIII are cofactors; fibrinogen is a structural protein. Shown are the four major classes of natural anticoagulants: antithrombin III (ATIII)-heparan (which inhibits the serine proteases of the clotting cascade), protein C (PC)/protein S (not shown) and thrombomodulin (TM) (which inhibits the cofactors of the clotting cascade), tissue factor pathway inhibitor (TFPI) (which inhibits the extrinsic pathway), and the fibrinolytic systems (plasmin degrades fibrin). The liver and

Endothelial cells in health and disease

442

endothelium (the cell lining at bottom) both contribute to the synthesis and release of hemostatic factors. Note that factor XII is not included in the scheme. This factor, which can activate factor XI in vitro, is important to consider when interpreting results of coagulation assays. However, it is not involved in mediating in vivo hemostasis. Two key links connect the extrinsic and common pathways with the intrinsic pathway. First, factor VIIa activates factor IX (cross talk). Second, thrombin activates factors XI and VIII (feedback). t-PA; tissue-type plasminogen activator. The activated form of the serine protease is indicated by the suffix “a.” layers of the blood vessel wall, and possibly in certain subsets of endothelial cells. Interestingly, the precise source of TF that is responsible for initiating coagulation in physiological states is still an open question—one that may ultimately be addressed by studying genetic mouse models lacking TF in one or another cell line-age. Fourth, the clotting cascade is amplified through the intrinsic pathway, by mechanisms that involve cross talk (factor VIIa of the extrinsic pathway activates factor IX of the intrinsic pathway) and feedback (thrombin activates factors XI and VIII). Fifth, the clotting cascade consists of a series of linked reactions in which a serine protease, once activated, is capable of activating its downstream substrate. Sixth, the enzymatic reactions are phospholipid-dependent and take place on activated cell surfaces. Seventh, some reactions are accelerated by the presence of cofactors, namely factors VIIIa and Va. Finally, for every procoagulant reaction, there is a natural anticoagulant response, giving rise to the so-called Yin and Yang metaphor of blood coagulation. To understand these anticoagulant mechanisms, one must consider the role of the endothelium. The endothelium expresses tissue factor pathway inhibitor (TFPI) (7), which forms a quaternary structure with TF, VIIa and Xa, thus inhibiting the extrinsic pathway. The endothelium synthesizes heparan, a cofactor for antithrombin III (ATIII), which neutralizes each of the serine proteases in the clotting cascade (8,9). Protein C is converted to activated protein C in the presence of endothelial membrane-bound thrombomodulin (TM) and endothelial protein C receptor (EPCR) (10). Once activated, protein C inactivates the cofactors of the clotting cascade (factors VIIIa and Va), a process that is accelerated by the cofactor, protein S. Finally, the fibrinolytic system may be thought of as a natural anticoagulant mechanism, in which plasmin degrades preformed fibrin.

Endothelium and hemostasis

443

2.2. Hemostasis as a Finely Tuned Balance Hemostasis may be viewed as a finely tuned balance between procoagulant and anticoagulant forces (Fig. 2) (1). On the procoagulant side is the monocyte, the platelet (primary hemostasis), and the clotting cascade (secondary hemostasis). The anticoagulant side includes the four mechanisms described in the previous section (TFPI, heparan/ATIII, protein C/protein S/TM, fibrinolysis). In addition, blood flow (or lack of stasis) promotes shear stress and provides a means to clear activated proteases. The maintenance of vascular integrity (intact endothelium) and the attenuation of negatively charged membrane surfaces limit the activation of primary and secondary hemostasis. Depending on which side the scale tips towards, abnormalities in hemostasis clinically manifest as either bleeding (hemorrhage) or thrombosis (clotting).

Figure 2 The hemostatic balance. On the procoagulant side is the clotting cascade and cells (platelets and monocytes). On the anticoagulant side are the four specific (protein) inhibitors of the clotting cascade, as well as blood flow and tight regulation of activated membrane surfaces. The list is not all-inclusive. For example, activators and inhibitors of platelets may be added, including von Willebrand factor, collagen, prostacylcin, nitric oxide, and

Endothelial cells in health and disease

444

nucleosides IP, intrinsic pathway EP, extrinsic pathway.

2.3. Congenital vs. Acquired Disorders Hematologists often classify hemostatic abnormalities (whether bleeding or thrombosis) into congenital and acquired etiologies. For purposes of this discussion, we will focus only on the thrombotic or hypercoagulable states. According to conventional wisdom, patients who present with idiopathic thrombosis at a young age, recurrent episodes of thrombosis and/or with a positive family history raise a red flag for congenital etiology. Congenital causes include hereditary deficiency in ATIII, protein C, or protein S. Alternatively, patients may inherit a mutation in the factor V gene that renders the activated factor resistant to the inhibitor effects of activated protein C—the so-called factor V Leiden mutation (11). This latter mutation is so common in the general population (between 5% and 15% prevalence) that it suggests a protective or adaptive function. One possibility is that the factor V Leiden mutation—which is believed to have appeared approximately 20–30,000 years ago (12)—evolved as a means to boost hemostasis in the hunter-gatherer, thus protecting against the bite of the Saber Tooth Tiger. Another possibility, which is favored by this author, is that the mutation arose as part of the arms race between humans and pathogens. Patients who carry the factor V Leiden mutation have increased thrombin generation, and have been reported to have lower mortality rates in severe sepsis (see Chapter 20). It is tempting to speculate that the sedentary lifestyle and increased life span associated with the agricultural and industrial revolution may have unmasked an otherwise hidden propensity to develop thrombosis.

2.4. Virchow’s Triad Hypercoagulable states may be approached from the perspective of Virchow’s triad. The triad consists of a change in the blood vessel wall (namely loss of vascular integrity), a reduction in blood flow (namely stasis) or an alteration in blood constituents. Impairment in vascular integrity and/or local alterations in blood flow—as occur, for example, following hip surgery—are sufficient to explain many causes of focal or site-specific thrombosis. It may be argued that low flow in the deep veins of the leg renders this site particularly vulnerable to thrombosis in patients with congenital hypercoagulable states. Moreover, patients with congenital hypercoagulable states have an increased risk of venous thromboembolism following trauma, surgery, or immobilization (13,14). However, Virchow’s triad—as it is usually interpreted—fails to explain the lack of correlation between congenital hypercoagulable states and arterial thrombosis. If such individuals have the same incidence of atherosclerosis and plaque rupture as the normal population, then Virchow’s triad would predict a higher rate of acute myocardial infarction. Another example worth considering is warfarin-induced skin necrosis. The fundamental abnormality is a disproportionate reduction in circulating levels of functional protein C. According to Virchow’s triad, the predilection for dermal clots must be explained by local stasis and/or damage in the microvasculature of the skin. There is no evidence to support either of these mechanisms.

Endothelium and hemostasis

445

When Virchow proposed his triad in 1845, there was virtually no understanding of the biology of the vessel wall or the nature of the blood constituents. However, as I will argue in the sections that follow, a modified version of Virchow’s triad, which takes into account the critical role of the endothelium in modulating hemostatic phenotypes, provides the best conceptual framework for considering the local nature of thrombotic disorders.

3. THE SPATIAL AND TEMPORAL DYNAMICS OF THE ENDOTHELIUM The endothelium is not inert, but rather is highly metabolically active. The endothelium is involved in multiple physiological processes, including the control of vasomotor tone, the trafficking of cells and nutrients, the regulation of permeability, and the formation of new blood vessels. In addition, the endothelium plays a critical role in mediating the hemostatic balance (1,15–17). For example, on the anticoagulant side, endothelial cells express TFPI, tissue-type plasminogen activator (t-PA), TM, EPCR, ecto-ADPase, prostacylcin, and nitric oxide; whereas on the procoagulant side, endothelial cells synthesize TF, plasminogen activator inhibitor (PAI)-1, von Willebrand factor (vWF), and protease activated receptors (1). In keeping with the major theme of this book, each of these proteins is expressed in ways that differ according to time and location within the vascular tree (Table 3). For example, TFPI is expressed predominantly in capillary endothelium, EPCR in large veins and arteries, eNOS on the arterial side of the circulation, vWF in veins, and TM in blood vessel types of every caliber in all organs except the brain (1,18–20). Tissue factor is not detectable in normal endothelium, whereas in a baboon model of sepsis, the gene is upregulated in a subset of endothelial cells in the marginal zone of splenic follicles (21). The picture that emerges is one of heterogeneity layered upon heterogeneity. Indeed, if one were to survey endothelial cells from different sites of the vascular tree, one would find that the hemostatic balance is governed by site-specific formulas (Fig. 3). These observations provide a strong foundation for an updated model of hemostasis, as described below.

4. AN UPDATED MODEL OF HEMOSTASIS The liver synthesizes a relatively constant amount of fibrinogen, serine proteases, cofactors, and anticoagulants (protein C, ATIII, protein S) (Fig. 4). The bone marrow produces and releases into the circulation a relatively fixed number of monocytes and platelets, cells that are capable of either expressing TF or promoting clotting reactions. This net output of liver-derived proteins and marrow-derived cells is systemically distributed, where it is integrated into the unique hemostatic balance of each and every vascular bed. Thus, when there is an alteration in the net output of proteins (e.g., protein C deficiency or factor V Leiden) or cells (e.g., sepsis), the changes will affect the hemostatic balance in ways that differ between sites of the vascular tree. To return to the example of warfarin-induced skin necrosis, it seems likely that the site-specific

Endothelial cells in health and disease

446

Table 3 Distribution of Representative EndothelialDerived Hemostatic Factors Factora

Distribution

vWF

Veins

eNOS

Arteries

TFPI

Capillaries

EPCR

Large veins and arteries

t-PA

Highest levels in brain; in lung in bronchial but not pulmonary circulation

TM

Absent in brain

a

vWF: von Willebrand factor; eNOS: endothelial nitric oxide synthase; TFPI: tissue factor pathway inhibitor; EPCR: endothelial protein C receptor; t-PA: tissue-type plasminogen activator; TM: thrombomodulin.

Figure 3 Site-specific hemostatic formulas. Each endothelial cell contributes to the hemo-static balance by expressing and/or secreting surface receptors and soluble mediators. Receptors include the proteaseactivated receptors (or TR, thrombin receptor), thrombomodulin, tissue factor (TF), ectoADPase (not shown). Soluble mediators include von

Endothelium and hemostasis

447

Willebrand factor (vWF), plasminogen activator inhibitor-1 (PAI-1), tissuetype plasminogen activator (t-PA), tissue factor pathway inhibitor (TFPI), and heparan. Each of these factors is differentially expressed from one site of the vascular tree to another. Thus, at any point of time, the hemo-static balance is regulated by vascular bedspecific “formulas.” Shown is a hypothetical example, in which an endothelial cell from a liver capillary relies more on vWF, PAI-1, and TFPI to balance hemostasis, whereas an endothelial cell from a lung capillary expresses more thrombin receptor, tPA, and heparan.

Figure 4 Integrated model of hemostasis. hemostatic balance of the postcapillary venular endothelium in the skin renders the dermal microvasculature particularly vulnerable to the systemic imbalance in vitamin-Kdependent factors, particularly protein C. Interestingly, a similar pattern of thrombosis is

Endothelial cells in health and disease

448

seen in purpura fulminans associated with congenital homozygous protein C deficiency or meningococcemia-induced acquired protein C deficiency (22). Adding another layer of complexity (and heterogeneity) to the model is the fact that the endothelium may be differentially activated between sites of the vasculature. For example, sepsis may be associated with the local accumulation of cytokines, and secondary changes in leukocyte adhesion, thrombin generation, barrier function, blood flow, and oxygenation, each of which may affect the local endothelial-derived hemostatic balance (see Chapter 20). This revised scheme offers certain advantages over older models. First, it is more inclusive in that it recognizes the involvement of four functionally linked organ systems in mediating coagulation, namely the liver, the bone marrow, the cardiovascular system, and the endothelium. In this way, we are reminded of the importance of the hepatocyte in synthesizing the serine proteases, the two cofactors, fibrinogen as well as the natural anticoagulants, ATIII, protein C and protein S; the critical role of the TF-expressing monocyte in initiating coagulation; the participation of the platelet in localizing and perpetuating the coagulation response, and the importance of the endothelial cell as an important manufacturer of hemostatic factors and regulator of the hemostatic balance. Second, the scheme incorporates both primary and secondary hemostasis. All too often, the cellular and soluble phases of coagulation are perceived as separate and independent entities that operate in series, when in fact they are highly integrated, parallel processes. Finally, the paradigm provides a useful conceptual framework for understanding the local nature of thrombotic diathesis. The very existence of vascular bed-specific phenotypes is enough to explain how a systemic imbalance in proteins and/or cells may be channeled into local clot formation. It is not difficult to reconcile this model with that of Virchow. Two principles have not changed over the past 150 years. First, stasis of flow (e.g. obesity, tumor, congestive heart failure, pregnancy, or immobility), when introduced into the system (Fig. 4), may lead to accumulation of activated clotting factors, reduction in protective hemodynamic forces and downstream hypoxia—all of which may tip the balance to the procoagulant side. Second, frank disruption or denudation of the endothelium (e.g. as occurs in trauma, surgery or catheter placement) may result in the exposure of blood to subendothelial adventitial TF and secondary thrombosis. However, two new observations warrant emphasis. First, when considering about blood constituents, we now appreciate the importance not only of the soluble clotting factors, but also of the cells—and not just the platelet, but also the monocyte, perhaps the single most important initiator of blood coagulation. Second, we now understand that the endothelium is far from a passive barrier; it is highly active, very much alive, rich in diversity and complexity, and as a result is a critical determinant of local hemostatic balance. Based on our new knowledge, I would propose that Virchow’s triad be modified—slightly—by emphasizing the importance not only of the structural but also the functional integrity of the vessel wall, particularly as it relates to the endothelium.

Endothelium and hemostasis

449

5. CONCLUSIONS Virchow’s triad has withstood the test of time and continues to provide a working foundation for approaching patients with thrombotic disorders. However, the strict application of 19th century principles to a modern-day understanding of thrombosis neglects important advances in the fields of hemostasis and vascular biology. In this review, I have proposed a revised model that honors the spirit of the original triad, while at the same time incorporating exciting and novel information about the coagulation mechanism.

REFERENCES 1. Aird WC. Vascular bed-specific hemostasis: role of endothelium in sepsis pathogenesis. Crit Care Med 2001; 29:S28–S35. 2. Rosenberg RD, Aird WC. Vascular-bed-specific hemostasis and hypercoagulable states. N Engl J Med 1999; 340:1555–1564. 3. Chan YC, Valenti D, Mansfield AO, Stansby G.Warfarin induced skin necrosis. Br J Surg 2000; 87:266–272. 4. Stewart AJ, Penman ID, Cook MK, Ludlam CA. Warfarin-induced skin necrosis. Postgrad Med J 1999; 75:233–235. 5. Weiler-Guettler H, Christie PD, Beeler DL, et al. A targeted point mutation in thrombomodulin generates viable mice with a prethrombotic state. J Clin Invest 1998; 101: 1983–1991. 6. Aird WC. Hemostasis and irreducible complexity. J Thromb Haemost 2003; 1:227–230. 7. Broze GJ Jr. Tissue factor pathway inhibitor. Thromb Haemost 1995; 74:90–93. 8. Bauer KA, Rosenberg RD. Role of antithrombin III as a regulator of in vivo coagulation. Semin Hematol 1991; 28:10–18. 9. Damus PS, Hicks M, Rosenberg RD. Anticoagulant action of heparin. Nature 1973; 246:355– 357. 10. Esmon CT. The protein C pathway. Chest 2003; 124:26S–32S. 11. Dahlback B. The discovery of activated protein C resistance. J Thromb Haemost 2003; 1:3–9. 12. Zivelin A, Rosenberg N, Faier S, et al. A single genetic origin for the common prothrombotic G20210A polymorphism in the prothrombin gene. Blood 1998; 92:1119–1124. 13. Sanson BJ, Simioni P, Tormene D, et al. The incidence of venous thromboembolism in asymptomatic carriers of a deficiency of antithrombin, protein C, or protein S: a prospective cohort study. Blood 1999; 94:3702–3706. 14. Simioni P, Sanson BJ, Prandoni P, et al. Incidence of venous thromboembolism in families with inherited thrombophilia. Thromb Haemost 1999; 81:198–202. 15. Cines DB, Pollak ES, Buck CA, et al. Endothelial cells in physiology and in the pathophysiology of vascular disorders. Blood 1998; 91:3527–3561. 16. Bombeli T, Mueller M, Haeberli A. Anticoagulant properties of the vascular endothelium. Thromb Haemost 1997; 77:408–423. 17. Gross PL, Aird WC. The endothelium and thrombosis. Semin Thromb Hemost 2000; 26:463−478. 18. Yamamoto K, de Waard V, Fearns C, Loskutoff DJ. Tissue distribution and regulation of murine von Willebrand factor gene expression in vivo. Blood 1998; 92:2791–2801.

Endothelial cells in health and disease

450

19. Osterud B, Bajaj MS, Bajaj SP. Sites of tissue factor pathway inhibitor (TFPI) and tissue factor expression under physiologic and pathologic conditions. On behalf of the Subcommittee on Tissue factor Pathway Inhibitor (TFPI) of the Scientific and Standardization Committee of the ISTH. Thromb Haemost 1995; 73:873–875. 20. Ishii H, Salem HH, Bell CE, Laposata EA, Majerus PW. Thrombomodulin, an endothelial anticoagulant protein, is absent from the human brain. Blood 1986; 67:362–365. 21. Drake TA, Cheng J, Chang A, Taylor FB Jr. Expression of tissue factor, thrombomodulin, and E-selectin in baboons with lethal Escherichia coli sepsis. Am J Pathol 1993; 142:1458–1470. 22. Wiss K. Clotting and thrombotic disorders of the skin in children. Curr Opin Pediatr 1993; 5:452−457.

23 Thrombotic Microangiopathies: Role of Microvascular Endothelium in Pathogenesis Thomas O.Daniel Ambrx, Inc., San Diego, California, U.S.A.

1. INTRODUCTION The term thrombotic microangiopathy (TM) describes a clinical syndrome that spans across a broad landscape of disease entities and clinical settings. The hallmark features reflect common pathological manifestations of microvascular compromise, including microangiopathic hemolysis, thrombocytopenia, and organ dysfunction, in the absence of disseminated intravascular coagulation (1). The central and unifying element in this disease landscape is regional and specific microvascular endothelial injury. The operative pathogenetic mechanisms differ between individual patients sharing common clinical features. The nomenclature applied within the TMs is tied to classical clinical descriptions of thrombotic thrombocytopenic purpura (TTP) (2) and hemolytic uremic syndrome (HUS) (3) (Fig. 1). Individual patients with components of the classical triad acquire one of these labels during a given episode based largely on clinical setting and the pattern of organ involvement (4). The clinical manifestations of TM may occur in specific bacterial and viral infections, therapy with selected drugs, bone marrow transplantation (BMT), peripartum states, and, notably, in familial recurrent forms, as well as in isolated, idiopathic settings. Similar manifestations of endothelial injury also accompany malignant hypertension, systemic lupus erythematosus (SLE), scleroderma, sepsis, and other diseases where endothelial participation is important, but, in these cases, the microangiopathy represents a secondary consequence of disease progression. Definition of common elements of the molecular pathogenesis in hereditary and acquired TTP has contributed to an improved classification system, based on molecular pathogenesis, summarized in Table 1. Coupled with better definition of endothelial sensitivity to the inciting toxin of enteropathic Escherichia coli associated with HUS, the new framework provides an approach to further define and segregate individual patients by diagnostic molecular surrogates of illness, to direct their therapy, and to predict their outcome.

Endothelial cells in health and disease

452

Figure 1 The term “thrombotic microangiopathies” encompasses a range of clinical disease entities that were previously grouped into TTP or HUS categories, based on the tissue bed sites most obviously affected. Modern dissection of the molecular mechanisms responsible has emphasized the role that tissue based endothelial heterogenity plays in defining which organs and tissues are most prominently affected. These mechanistic differences are typified by vWF UL multimers in TTP and Shigalike toxin (Stx) producing enteropathogenic E. coli infections in HUS, though other features are now coming to light, and are described in the accompanying sections of this chapter. Despite distinct patterns of organ involvement, overlap in clinical presentation and tissue site damaged occurs among individuals patients. Dissection of the pathogenetic mechanisms underlying TM provides a key element of support to an underlying premise of this volume, namely that the endothelium is a regionally specialized organ that is both target and participant in disease processes. While

Role of microvascular endothelium in pathogenesis

453

endothelial roles in initiation, progression and resolution of TM are incompletely understood, the centrality of that role is undisputed. The endothelium of specific tissues displays site-restricted susceptibility to insulting stimuli, and integrates responses to insult that mediate, extend, or repair the thrombotic microangiopathic consequences. The repeated demonstration that cultured microvascular endothelial cells derived from different tissues retain, at least to some extent, heterogeneous features characteristic of that vascular bed has provided a lens through which to define endothelial sensitivity and responses relevant to the pathogenesis of TM (5– 7). Despite considerable overlap in clinical presentation and outcome, a consideration of the molecular pathogenesis provides a framework for classifying individual patients that is more discriminatory and valuable than the historical segregation into either TTP or HUS categories based on differential organ involvement (Table 1). The ensuing discussion focuses on dissecting these mechanisms, focusing first on the role for von Willebrand factor (vWF) in the subclass of TMs most commonly labeled TTP.

2. THROMBOTIC THROMBOCYTOPENIC PURPURA 2.1. Clinical Features Initially described by Moschcowitz in 1924, TTP is a systemic illness characterized by prominent microvascular platelet aggregation with consequent ischemia of

Table 1 Clinical Settings and Pathogenetic Mechanisms. Clinical Setting

Defining Features

Pathogenetic Mechanisms

Response to Plasma Exchange

TTP Hereditary

↑UL vWF/↓ vWFCP

ADAMTS13 mutations

Effective

Acquired

↑UL vWF/↓ vWFCP

Neutralizing antibodies to ADAMTS13

Effective

Ticlopidine, Clopidigrel

Acute onset (