Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Endothelial cell adhesion molecules: expression, upregulation and modulation of human brain microvessel… Wong, Donald Chun Kit 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1995-98368x.pdf [ 19.21MB ]
Metadata
JSON: 831-1.0088815.json
JSON-LD: 831-1.0088815-ld.json
RDF/XML (Pretty): 831-1.0088815-rdf.xml
RDF/JSON: 831-1.0088815-rdf.json
Turtle: 831-1.0088815-turtle.txt
N-Triples: 831-1.0088815-rdf-ntriples.txt
Original Record: 831-1.0088815-source.json
Full Text
831-1.0088815-fulltext.txt
Citation
831-1.0088815.ris

Full Text

ENDOTHELIAL CELL ADHESION MOLECULES: EXPRESSION, UPREGULATION AND MODULATION OF HUMAN BRAIN MICROVESSEL ENDOTHELIAL CELL-LEUKOCYTE INTERACTIONS by DONALD CHUN KIT WONG B.Sc. (Honours), The University of British Columbia, 1986 M.Sc, The University of British Columbia, 1990 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Pathology)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA February 1995 © Donald Chun Kit Wong, 1995  In presenting this thesis in partial fulfilment  of  the requirements  for an  advanced  degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department  or  by  his  or  her  representatives.  It  is  understood  that  copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  R f l T H t>Lo < W  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  ABSTRACT Endothelial cells (EC) lining the cerebral blood vessels are responsible for the formation and maintenance of a unique structural and functional barrier, the blood-brain barrier (BBB), that normally prevents the movement of white blood cells from blood into brain. Inflammatory and infectious diseases of the central nervous system (CNS), including multiple sclerosis, encephalitis, meningitis and infarction, are characterized by increased BBB permeability and infiltration of the brain by leukocytes. The mechanisms that regulate the traffic of leukocytes across the BBB have not yet been elucidated. A group of EC surface molecules called adhesion molecules, have been implicated in interactions between inflammatory cells and extracerebral EC. They include E-selectin, vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1) and platelet/endothelial cell adhesion molecule-1 (PECAM-1). Blocking these molecules with antibodies (Abs) reduces leukocyte traffic across EC. One of these molecules, ICAM-1, has been demonstrated in inflammatory CNS lesions. The objective of this study was to investigate the expression and function of four EC adhesion molecules in an in vitro model of the human BBB. The results indicate that primary cultures of unstimulated human brain microvessel EC (HBMEC) form high resistance monolayers and express low levels of E-selectin, VCAM-1 and ICAM-1 that can be significantly upregulated following stimulation with cytokines and LPS. PECAM-1 is strongly expressed by all cells constitutively. Treatment of HBMEC with TNF increases monolayer permeability, and augments adhesion and migration of polymorphonuclear leukocytes (PMN) and T-lymphocytes across the monolayers. Adhesion of resting T-lymphocytes to activated endothelium is mediated by VCAM-1 and ICAM-1. ICAM-1, E-selectin and PECAM-1 participate in T-lymphocyte migration across HBMEC monolayers. Adhesion of PMN to activated HBMEC is mediated by ICAM-1 and E-selectin, while ICAM-1 is involved in the migration process. Anti-adhesion molecule Abs reduce adhesion and/or migration but have no effect on the  ii  increased endothelial permeability secondary to leukocyte migration. These studies indicate that EC adhesion molecules, induced by cytokines, mediate cerebral EC-leukocyte interactions which, in association with a concurrent increase in endothelial permeability, may facilitate leukocyte traffic across the BBB in the early stages of CNS inflammation.  iii  TABLE OF CONTENTS Abstract  ii  Table of Contents  iv  List of Figures  yii  List of Tables  xi  List of Abbreviations  xii  Acknowledgement  xiv  Chapter 1 Introduction 1.1 History of Inflammation 1.2 History of the microcirculation 1.3 Current concepts on the microcirculation and inflammation 1.3.1 The microcirculation 1.3.2 Inflammation 1.4 Endothelial cell adhesion molecules 1.4.1 E-selectin 1.4.2 Vascular cell adhesion molecule-1 (VCAM-1) 1.4.3 Intercellular adhesion molecule-1 (ICAM-1) 1.4.4 Platelet/endothelial cell adhesion molecule-1 (PECAM-1) 1.4.5 Vascular addressins 1.4.6 Other endothelial cell adhesion molecules 1.4.7 Adhesion molecule shedding 1.4.8 Leukocyte adhesion molecules 1.4.9 Adhesion molecule therapy 1.5 Lipopolysaccharide and cytokines 1.5.1 Lipopolysaccharide (LPS) 1.5.2 Tumor necrosis factor (TNF) 1.5.3 Interleukin-1 (IL-1) 1.5.4 Interferon-y (IFN-y) 1.6 Endothelial cell heterogeneity 1.7 The blood-brain barrier (BBB) 1.8 Inflammation in the central nervous system 1.8.1 General aspects 1.8.2 Adhesion molecules in central nervous system disorders 1.8.3 Cytokines in central nervous system disorders 1.9 Human brain microvessel endothelial cells (HBMEC) in vitro 1.10 General aim of this study 1.11 Specific objectives of this study 1.11.1 Phase I. Investigation of the expression of adhesion molecules by human brain microvessel endothelial cells in primary culture 1.11.2 Phase II. Investigation of the role of endothelial cell adhesion molecules in leukocyte adhesion and transmigration across human brain microvessel endothelial cell monolayers 1.11.3 Phase III. Investigation of the monolayer permeability after cytokine treatment and in the course of HBMEC-leukocyte interactions . . . . iv  1 1 3 .4 4 4 .6 .7 9 11 12 13 14 15 15 17 18 18 19 20 20 21 22 23 23 24 25 26 27 21 27 28 28  Chapter 2 Materials and Methods 29 2.1 Expression of adhesion molecules 29 2.1.1 Human brain microvessel endothelial cells 29 2.1.2 Adhesion molecule induction by lipopolysaccharide and cytokines . . .30 2.1.3 Antibodies 31 2.1.4 Immunogold silver staining (IGSS) for light microscopy 31 2.1.5 Enzyme-linked immunosorbent assay (ELISA) 32 2.1.6 Immunogold electron microscopy (IGEM) 33 2.1.7 Controls 34 2.1.8 Statistics 34 2.2 Role of adhesion molecules in PMN and T lymphocyte adhesion to and migration across human brain microvessel endothelial cell monolayers . . . .35 2.2.1 Human brain microvessel endothelial cells 35 2.2.2 Polymorphonuclear leukocyte (PMN) isolation 35 2.2.3 T lymphocyte isolation 35 2.2.4 Antibodies 36 2.2.5 Adhesion assay 37 2.2.6 Migration assay 38 2.2.7 Controls 39 2.2.8 Statistics 40 2.3 Determination of monolayer permeability following TNF-a treatment and in the course of leukocyte migration across the monolayers 40 2.3.1 Permeability assay 40 2.3.2 Controls .41 2.3.3 Statistics 41 Chapter 3 Results 42 3.1 Human brain microvessel endothelial cell culture (HBMEC) 42 3.2 Expression of adhesion molecules by HBMEC in primary culture and upregulation by LPS and cytokines 42 3.2.1 E-selectin 42 3.2.1.1 Immunogold silver staining (IGSS) 42 3.2.1.2 Enzyme-linked immunosorbent assay (ELISA) 44 3.2.1.3 Immunogold electron microscopy (IGEM) .44 3.2.2 Vascular cell adhesion molecule-1 (VCAM-1) 45 3.2.2.1 Immunogold silver staining (IGSS) 45 3.2.2.2 Enzyme-linked immunosorbent assay (ELISA) 47 3.2.2.3 Immunogold electron microscopy (IGEM) .47 3.2.3 Intercellular adhesion molecule-1 (ICAM-1) 48 3.2.3.1 Immunogold silver staining (IGSS) 48 3.2.3.2 Reversibility of ICAM-1 expression 49 3.2.3.3 Immunogold electron microscopy (IGEM) 50 3.2.4 Platelet/endothelial cell adhesion molecule-1 (PECAM-1) 50 3.2.4.1 Enzyme-linked immunosorbent assay (ELISA) 50 3.2.4.2 Immunogold electron microscopy (IGEM) 51 3.2.5 HECA-452 antigen 51 3.3 Role of adhesion molecules in PMN and T lymphocyte adhesion to and migration across human brain microvessel endothelial cell monolayers . . . .51 3.3.1 Adhesion of polymorphonuclear leukocytes to human brain microvessel endothelial cells 52 3.3.2 Polymorphonuclear leukocyte (PMN) migration 53 3.3.3 T lymphocyte adhesion 54 V  3.3.4 T lymphocyte migration 55 3.4 Determination of monolayer permeability following TNF-a treatment and in the course of leukocyte migration across the monolayers 56 Chapter 4 Discussion 58 4.1 Expression of adhesion molecules 58 4.1.1 E-selectin 58 4.1.2 Vascular cell adhesion molecule-1 (VCAM-1) 62 4.1.3 Intercellular adhesion molecule-1 (ICAM-1) 66 4.1.4 Platelet/endothelial cell adhesion molecule-1 (PECAM-1) 72 4.1.5 HECA-452 antigen 73 4.2 Role of adhesion molecules in PMN and T lymphocyte adhesion to and migration across human brain microvessel endothelial cell monolayers . . . .76 4.2.1 PMN adhesion 76 4.2.2 PMN migration 79 4.2.3 T lymphocyte adhesion 84 4.2.4 T lymphocyte migration 87 4.3 Determination of monolayer permeability following TNF-a treatment and in the course of leukocyte migration across the monolayers 95 Chapter 5 Conclusions 5.1 Summary and significance 5.2 Future prospects  101 101 104  References  255  vi  LIST OF FIGURES Figure 1 Confluent culture of human brain microvessel endothelial cells  106  Figure 2 Primary cultures of human brain microvessel endothelial cells show positive staining for Factor VIII antigen 108 Figure 3 Illustration of a Cellagen disc in a well of a 24 well plate  110  Figure 4 HBMEC expression of E-selectin without LPS or cytokine, detected by immunogold silver staining  112  Figure 5 Quantitation of E-selectin expression on HBMEC, after incubation with LPS, by (a) IGSS and (b) ELISA 114 Figure 6 Quantitation of E-selectin expression on HBMEC, after incubation with TNF-a, by (a) IGSS and (b) ELISA 116 Figure 7 Quantitation of E-selectin expression on HBMEC, after incubation with IL-ip, by (a) IGSS and (b) ELISA 118 Figure 8 Immunogold silver staining quantitation of E-selectin expression on HBMEC after incubation with IFN-y 120 Figure 9 HBMEC expression of E-selectin detected by immunogold EM  122  Figure 10 Immunogold electron microcopy quantitation of E-selectin expression of HBMEC  124  ;  Figure 11 Immunogold silver staining of HBMEC monolayers for VCAM-1. . . 126 Figure 12 VCAM-1 expression by HBMEC following incubation with LPS, quantitation by (a) IGSS and (b) ELISA Figure 13 VCAM-1 expression by HBMEC following incubation with TNF-a. Quantitation by (a) IGSS and (b) ELISA  130  Figure 14 VCAM-1 expression by HBMEC following incubation with IL-lp\ Quantitation by (a) IGSS and (b) ELISA  132  Figure 15 VCAM-1 expression by HBMEC following incubation with IFN-y. Quantitation by (a) IGSS and (b) ELISA  134  128  Figure 16 VCAM-1 expression by HBMEC following incubation with IL-ip and TNF-a. Quantitation by IGSS  136  Figure 17 Immunogold labeling for VCAM-1 in HBMEC  138  Figure 18 Quantitation of VCAM-1 expression on the apical and basal surfaces of unstimulated and TNF-a treated HBMEC following immunogold labeling . . . . 140 Figure 19 HBMEC expression of ICAM-1 detected by immunogold silver staining . .142 vii  Figure 20 HBMEC expression of ICAM-1 after incubation with LPS  144  Figure 21 HBMEC expression of ICAM-1 after incubation with IFN-y  146  Figure 22 HBMEC expression of ICAM-1 after incubation with TNF-a  148  Figure 23 HBMEC expression of ICAM-1 after incubation with IL-ip  150  Figure 24 Combined effects of IFN-y and TNF-a on the ICAM-1 expression on HBMEC  152  Figure 25 Reversibility of ICAM-1 expression on HBMEC induced by incubation with LPS 154 Figure 26 Reversibility of ICAM-1 expression on HBMEC induced by incubation with IFN-Y  156  Figure 27 Reversibility of ICAM-1 expression on HBMEC induced by incubation with TNF-a 158 Figure 28 Reversibility of ICAM-1 expression on HBMEC induced by incubation with IL-lp 160 Figure 29 HBMEC expression of ICAM-1 detected by immunogold EM  162  Figure 30 Immunogold electron microcopy quantitation of ICAM-1 expression of HBMEC  164  Figure 31 HBMEC expression of PECAM-1 detected by immunogold silver staining 166 Figure 32 ELISA quantitation of PECAM-1 expression on HBMEC after incubation with LPS 168 Figure 33 ELISA quantitation of PECAM-1 expression on HBMEC after incubation with TNF-a 170 Figure 34 ELISA quantitation of PECAM-1 expression on HBMEC after incubation with IL-lp 172 Figure 35 ELISA quantitation of PECAM-1 expression on HBMEC after incubation with IFN-y  174  Figure 36 HBMEC expression of PECAM-1 detected by immunogold EM  176  Figure 37 HBMEC expression of PECAM-1 detected by immunogold EM  178  Figure 38 Immunogold electron microcopy quantitation of PECAM-1 expression of HBMEC 180 Figure 39 Scanning electron micrographs of PMN adhesion to unstimulated or activated HBMEC 182 Figure 40 Light micrographs of PMN adhesion to unstimulated or activated HBMEC 184  viii  Figure 41 Light microscopy quantitation of PMN adhesion to TNF-a activated and untreated HBMEC  186  Figure 42 Effects of adhesion molecule blocking Ab treatment on PMN adhesion to TNF-a treated and untreated HBMEC 188 Figure 43 Effects of adhesion molecule blocking Ab treatment on PMN adhesion to TNF-a treated and untreated HBMEC 190 Figure 44 Effects of adhesion molecule blocking Ab treatment on PMN adhesion to TNF-a treated and untreated HBMEC 192 Figure 45 Light micrographs of PMN migration across TNF-a treated and untreated HBMEC 194 Figure 46 Light microscopy quantitation of PMN migration across TNF-a activated HBMEC 196 Figure 47 Light microscopy quantitation of PMN migration for 2h across TNF-a activated HBMEC treated with anti-adhesion molecule Abs  198  Figure 48 Scanning electron micrographs of PMN interaction with HBMEC during transendothelial migration 200 Figure 49 Transmission electron micrographs of PMN interactions with HBMEC during transendothelial migration 202 Figure 50 Transmission electron micrographs of PMN interactions with HBMEC during transendothelial migration 204 Figure 51 Scanning electron micrographs of T lymphocyte adhesion to unstimulated or activated HBMEC 206 Figure 52 Light micrographs of T lymphocyte adhesion to unstimulated or activated HBMEC 208 Figure 53 Light microscopy quantitation of T lymphocyte adhesion to TNF-a activated and untreated HBMEC 210 Figure 54 Effects of adhesion molecule blocking Ab treatment on T lymphocyte adhesion to TNF-a treated and untreated HBMEC  212  Figure 55 Effects of adhesion molecule blocking Ab treatment on T lymphocyte adhesion to TNF-a treated and untreated HBMEC  214  Figure 56 Effects of adhesion molecule blocking Ab treatment on T lymphocyte adhesion to TNF-a treated and untreated HBMEC  216  Figure 57 Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell 218 Figure 58 Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell 220  ix  Figure 59 Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell 222 Figure 60 Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell 224 Figure 61 Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell 226 Figure 62 Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell 228 Figure 63 Light micrographs of T lymphocyte migration across TNF-a treated and untreated HBMEC  230  Figure 64 Light microscopy quantitation of T lymphocyte migration across TNF-a activated HBMEC  232  Figure 65 Light microscopy quantitation of T lymphocyte migration for 30min across TNF-a activated HBMEC treated with anti-adhesion molecule Abs 234 Figure 66 Light microscopy quantitation of T lymphocyte migration for 60min across TNF-a activated HBMEC treated with anti-adhesion molecule Abs 236 Figure 67 Light microscopy quantitation of T lymphocyte migration for 180min across TNF-a activated HBMEC treated with anti-adhesion molecule Abs 238 Figure 68 Scanning electron micrographs of T lymphocyte interaction with HBMEC during transendothelial migration 240 Figure 69 Transmission electron micrographs of T lymphocyte interaction with HBMEC  242  Figure 70 Transmission electron micrographs of T lymphocyte interaction with HBMEC  244  Figure 71 Transmission electron micrographs of T lymphocyte interaction with HBMEC  246  Figure 72 Transmission electron micrographs of T lymphocyte migration through HBMEC  248  Figure 73 Transmission electron micrographs of T lymphocyte migrated underneath the HBMEC 250  X  LIST OF TABLES Table 1 Effect of TNF treatment of HBMEC on the permeability of the monolayer determined by electrical resistance measurements  252  Table 2 Effect of PMN migration with and without adhesion molecule blocking antibodies on the permeability of HBMEC cultures determined by electrical resistance measurements across the monolayer  253  Table 3 Effect of T lymphocyte migration with and without adhesion molecule blocking antibodies on the permeability of HBMEC cultures determined by electrical resistance measurements across the monolayers 254  XI  LIST OF ABBREVIATIONS Ab  antibody  AGEPC  acetyl-glyceryl-ether-phosphorylcholine  ANOVA  analysis of variance  BBB  blood brain barrier  BBMEC  bovine brain microvessel endothelial cells  CNS  .central nervous system  CSF  cerebrospinal fluid  DAB  3,3'-diaminobenzidine  EAE  .experimental allergic encephalomyelitis  EC  endothelial cell  EGF  epidermal growth factor  ELAM  endothelial leukocyte adhesion molecule  EM  electron microscopy  fMLP  formyl-methionyl-leucyl phenylalanine  HBMEC  human brain microvessel endothelial cell  HDMEC  human dermal microvessel endothelial cell  HECA  human endothelial cell antigen  HRP  horseradish peroxidase  HUVEC  human umbilical vein endothelial cell  ICAM  intercellular adhesion molecule  IFN  interferon  Ig  immunoglobulin  IGEM  immunogold electron microscopy  IGSS  immunogold silver staining  IL-1  interleukin-1  L-VAP  lymphocyte-vascular adhesion protein xii  LFA  Jeukocyte function-associated antigen  LPS  lipopolysaccharide  LTB4  leukotriene-B4  mAb  monoclonal antibody  MECA  mouse endothelial cell antigen  MHC  major histocompatibility complex  MS  multiple sclerosis  NK  natural killer  PBMC  peripheral blood mononuclear cells  PECAM  platelet/endothelial cell adhesion molecule  PMN  polymorphonuclear leukocyte  SEM  scanning electron microscopy  TEM  transmission electron microscopy  TNF  .tumor necrosis factor  VAP  .vascular adhesion protein  VCAM  vascular cell adhesion molecule  VIA  very late activation antigen  xiii  ACKNOWLEDGEMENT Foremost, I am grateful for the encouragement and understanding of my parents, brother and sister through these years of long working hours, coming home late at night, absence from home and family events, neglecting my chores,... Nothing would be possible without their backing. First and most importantly, I must express my gratitude to Drs. Zis and Zis who nurtured and supported me, like guardian angels attending to my needs. Much of these studies were performed with greater efficiency and success through the excellent technical aid of Mrs. R. Prameya and E. Chang in cell culture and isolation, sectioning, photographic printing, just to name a few. I am obliged to their assistance. No less important are the companionship and good cheer from Dr. A. Easton, Miss J. Shukaliak and the supervisory and technical members of the lab. A great deal of thanks are due to my supervisory committee members, Drs. D. Walker, W. Moore, S. Vincent and F. Takei, without whose critical comments and constructive suggestions, these studies would have been weakened considerably. I am thankful for the Junior Research Traineeship provided by the Heart and Stroke Foundation of Canada. Also vital is the skillful way Mrs. P. Woo managed all the program requirements for the graduate students in the department. In addition, the support staff at Vancouver Hospital, Dawn, Frances, Christine, Susan, Patti et al., have created a very hospitable community at VH. My hat off to them for creating an atmosphere that is so conducive to good work. Lastly, no thanks is too much for Hanh Huynh.  xiv  CHAPTER 1 INTRODUCTION Inflammation is a set of dynamic events that occurs in living tissue in response to injury or infection. The term "inflammation" is derived from the Greek words phlogosis and phlegmone, meaning "heat" and "the burning thing", respectively, and the Latin flamma meaning fire (Ryan and Majno, 1977; Rocha e Silva, 1978). This phenomenon has been recognized and studied from the earliest times.  1.1 HISTORY OF INFLAMMATION Inflammation is a concept known as early as the Egyptian times. At approximately 1650 B.C., the swnw (physician) described the treatment for a gaping wound using the term "inflammation". At the time of the Sumerians in Mesopotamia, the chief physician to King Esharaddon (680-669 B.C.) referred to "inflammation" several times in his writing (Majno, 1975). By the time of Hippocrates (460-380 B.C.), the Greeks had reasoned that a possible site of disease is the venules/small veins which could allow blood to escape by sheer fatigue or falling into a "spasm". The latter would tear the venule apart. These theories are not completely out of line with present observations. Electron microscopy shows that irritated venules do, in a sense, tear themselves apart. Chemical mediators cause endothelial cells (EC) lining these vessels to retract and pull away from one another, allowing fluid to escape (Majno, 1975). The clinical symptoms of inflammation were not formally defined until Cornelius Celsus (approximately 10-40 A.D.) stated in his book "De Medicina" that "the signs of an inflammation are four: rubor et tumor cum calore et dolore" (redness and swelling with heat and pain) (Majno, 1975). Long before that, the Hippocratic treatises mentioned that pain, local heat and swelling are characteristics of a particular type of pathologic alteration (Mettler, 1947).  1  The fifth cardinal sign of inflammation functio laesa or "disturbed function" is often attributed to Galenos (130 A.D.), although no such sign was mentioned in any of his works (Rather, 1971). It appears that the author of a German Handbook of Pathology incorrectly attributed this in 1919. This idea quickly became widely accepted and became general knowledge. It was, in fact, Rudolf Virchow who first added this sign in his book "Cellular Pathology" in 1858 (Majno, 1975). With the invention of the microscope, Herman Boerhaave (1668-1738) observed blood vessels in inflamed tissue and concluded that the constriction of the finest vessels led to the generation of local heat and fever through the friction between blood and the vessels. The dominant role of the finest blood vessels is emphasized here (Eisen, 1977). Later, others found that dilated rather than constricted microvessels (arterioles, venules and capillaries) were present in inflamed tissue (Eisen, 1977). In 1867, Samuel also concluded that inflammation centered around "a lesion of the vessels, by irritating agents, in such a way that the inflamed vessels were rendered more permeable to liquids and blood cells, in a passive way", thus, producing exudate that would be collected at the site of minor resistance, resulting in inflammatory tumor (Rocha e Silva, 1978). More observations on early vascular events were provided by Cohnheim and Metchnikoff. In addition to supporting the above view of the changes in the vessels, Cohnheim observed that the microvessels undergo a brief period of dilatation. The blood flow first accelerates, then slows down. Newly opened capillaries become visible. White blood cells then begin to line these vessels and eventually force themselves through the vessel wall into the extravascular spaces. In some vessels, flow stop completely and red blood cells are tightly packed together as if plasma has escaped because of the increased permeability of the vessels (Cohnheim, 1890). Further support for these views came from Weigert who confirmed the blood origin of the cells in the inflammatory exudate and Arnold who demonstrated the process of diapedesis (Rocha e Silva, 1978). Metchnikoff, on the other hand, saw inflammation 2  not as a disease of the blood vessels but as a protective mechanism to bring phagocytes to the injured areas to ingest and digest damaged cells, bacteria and other particles to be removed, a process he called "phagocytosis" (Rocha e Silva, 1978). Concurrently, the discovery of antibodies (Abs) brought another role to the inflammatory process - to allow Abs to flood the infected tissue for local defense. Thus, it is clear that the microcirculation is of paramount importance in inflammation.  1.2 HISTORY OF THE MICROCIRCULATION The idea of microvessels extends far back to the Egyptian times. At about 250 B.C., an anatomist, Erasistratos believed that blood moves from veins into the arteries through invisible "synanstomoses". It is these finest branches of arteries and veins along with nerves and parenchyma that make up tissues. In 1628, William Harvey discovered the circulation of blood. He reasoned that microvessels are present only in order to keep the circulation going. Malpighi was the first to directly visualize capillaries under the microscope. He was observing the alveoli of frog lungs when he saw the network of blood vessels and blood flowing from arteries to capillaries then to veins. About thirty years later, Antoni van Leeuwenhoek marveled at the circulation of the blood in the gills of tadpoles, made visible by the passage of globules (blood cells). Some vessels were so small that the globules could pass only in single file. He called these "blood capillaries". In the 1800's, von Recklinghausen showed that capillaries are tubes with a lining of cells by staining the cell borders of EC using silver nitrate (Majno, 1992).  3  1.3 CURRENT CONCEPTS ON THE MICROCIRCULATION AND INFLAMMATION  1.3.1 The microcirculation Blood enters the microcirculation through the arterioles, vessels with a diameter of about lOOum. It travels to the metarterioles and capillaries which have diameters of 5 to 10^m and exits through thin walled venules with a diameter of 200nm (Hurley, 1983; Junqueira et al., 1989; Gartner and Hiatt, 1990). Blood flows through capillaries in spurts because of the intermittent contraction of the metarterioles. Red blood cells traverse capillaries in single file due to the small diameter of these vessels (Hurley, 1983). The velocity of blood flow is very slow, in the order of l-2mm per second. Leukocytes, having a larger size than red blood cells, have a much longer transit time. They must first squeeze into the shape of a cylinder. This sometimes makes them slow down so much that they stop all together. In the lung, they have been observed to stop for 2sec to 20min (Majno, 1992). In arterioles and venules, blood travels by laminar flow. The velocity is greatest at the center of the vessel, decreasing progressively towards the vessel wall. This causes the blood cells to accumulate in the center of the stream of flowing blood, producing axial flow. Outside this central stream of closely packed blood cells is a peripheral zone of cell-free plasma (Hurley, 1983).  1.3.2 Inflammation At the earliest stage of inflammation, there is an initial constriction of the arterioles followed by dilatation of arterioles and venules. Many vessels that had received little blood now open up to increase the blood flow as much as ten fold. The initial rapid flow of blood through these vessels slows down gradually. The central stream of closely packed cells widens while the peripheral zone of plasma narrows. In some vessels, flow may cease completely. Even before this occurs, however, leukocytes 4  start to appear in the plasma zone as the flow initially slows. Some will adhere to the vessel wall and roll along, a portion will flow back into the blood stream, while others will become more firmly adherent. This process is called "margination". The luminal side of the vessel wall may eventually become covered with a layer of leukocytes. This is termed "pavementing". The adherent leukocytes extend pseudopodia into the vessel wall, soon breaking through into the extravascular space. A small protrusion of the leukocyte appears just outside the vessel and this grows larger as the whole leukocyte crosses the wall. Soon, the whole cell is outside the vessel and connected back by only a long thin process, which eventually also detaches. This emigration process is called "diapedesis". Adhesion always precedes movement across the vessel wall (Allison et al., 1955). Whether diapedesis occurs across junctions between EC or through the EC cytoplasm remains controversial. The endothelial gaps close up immediately behind emigrated leukocytes. Concurrently, an exudation of fluid occurs as the vascular permeability increases (Hurley, 1983). Observations in rat skin with carbon or ink particle labeling suggest that emigrating leukocytes do not contribute to this vascular leakage, showing how tightly leukocytes squeeze between EC. Other studies, however, have found the opposite (Ryan and Majno, 1977). Acute inflammation is characterized by a predominance of polymorphonuclear leukocytes (PMN). As the response becomes chronic, mononuclear leukocytes (macrophages and lymphocytes) become dominant. This is because of the short life span of PMN and a switch from chemotactic agents for PMN to those for mononuclear leukocytes. Chronic inflammation may evolve from acute inflammation or arise without an acute phase. The lesions of chronic inflammation have a predominance of mononuclear leukocytes as well as granulation tissue - connective tissue with an abundance of inflammatory cells, fibroblasts and newly formed capillaries that may be leaky (Ryan and Majno, 1977). Thus the early inflammatory response includes three processes: 5  1.  Blood vessel dilation and changes in blood flow, responsible for "rubor".  2.  Increased vascular permeability leading to edema with inflammatory exudates ("tumor").  3.  Movement of leukocytes from the blood through the vessel wall into the extravascular space. "Calor" is due to the great increase in blood flow and stimulation of heat sensitive  nerve endings in the skin. Release of endogenous chemicals, such as bradykinin, serotonin and some prostaglandins, in addition to a rise in tissue tension cause "dolor". "Functio laesa" results from reflex inhibition of muscular movement by the pain, limitation of movement by the swelling and death or damage of parenchymal cells (Hurley, 1983).  1.4 CELL ADHESION MOLECULES A major development over the last fifteen years has been the realization that EC lining the blood vessels do not merely constitute a passive barrier to the movement of solutes and cells across the vascular wall, but are active participants in inflammation. A number of cell surface molecules that mediate EC-leukocyte interactions have been found on the EC surface and they are of particular importance since the specificity of the transendothelial movement of leukocytes seems to reside within an altered endothelium. Activated, as well as nonactivated, leukocytes preferentially migrate across endothelium that has been activated and altered in inflammation (Ward and Marks, 1989). The EC adhesion molecules that have been characterized to date include E-selectin, vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1) and platelet/endothelial cells adhesion molecule-1 (PECAM-1).  6  1.4.1 E-selectin E-selectin (CD62E) is a surface glycoprotein of the selectin family. It is a 64kD core protein with eleven potential glycosylation sites to give 115kD. Three domains are present: a lectin-like binding domain, an epidermal growth factor-like domain, and six complement-regulatory protein regions. The cytoplasmic domain contains tyrosine residues that may mediate the internalization of this antigen. Its ligands bind the amino-terminal of the lectin domain and a portion of the epidermal growth factor-like domain (Pigott et al., 1991; Carlos and Harlan, 1994). E-selectin binds to ligands containing the Lewis x, sialyl-Lewis x (CD15S) or sialyl Lewis a blood group carbohydrates (Kerr and Stocks, 1992; Phillips et al., 1990; Foxall et al., 1992; Ohmori et al., 1993; Bevilacqua and Nelson, 1993; Nelson et al., 1993) on neutrophils, eosinophils, monocytes and lymphocytes (Shimizu et al., 1991; Picker et al., 1991a; Carlos et al., 1991; Walcheck et al., 1993; Postigo et al., 1992; Benjamin et al., 1990; Weller et al., 1991). L-selectin on leukocytes may be one of the ligands for E-selectin (Picker et al., 1991b). Analysis of how mutations at the lectin domain affect binding has further defined the region near the antiparallel beta sheets derived from the NH2 and COOH terminus of the lectin domain, and two adjacent loops to be critical for carbohydrate recognition (Erbe et al., 1992). A large body of evidence indicates that E-selectin is important in the initial rolling of PMN on the EC surface in the presence of shear stress. Under conditions of flow, the initial contact leading to the rolling of PMN on unstimulated human umbilical vein EC (HUVEC) at a rate of lOum/sec can be blocked up to 90% by an anti-E-selectin Ab, while a 30% inhibition can be achieved following activation of HUVEC (Abbassi et al., 1993). Rolling under conditions of flow does not involve the ICAM-1/CD18 pathway (Lawrence et al., 1990). Adhesion via E-selectin does not require active metabolism, unlike that via the integrins, but the strength of adhesion can be just as great 7  and much less time is required for adhesion strengthening (Lawrence and Springer, 1993). Binding of E-selectin also enhances the activity of the Mac-1 integrin (CDllb/CD18), leading to firm adhesion (Lo et al., 1991). Unstimulated HUVEC do not express E-selectin. Following activation with bacterial lipopolysaccharide (LPS), interleukin-1 (IL-1), or tumor necrosis factor-a (TNF-a), E-selectin expression reaches a maximal level after 4h, and declines back to unstimulated levels by 24-48h (Bevilacqua, 1993; Butcher, 1992; Springer, 1994). The adhesion of neutrophils, basophils, eosinophils, monocytes, a subset of lymphocytes and Plasmodium-infected erythrocytes is partially mediated by E-selectin since anti-E-selectin Abs can inhibit this interaction (Bochner et al., 1991; Graber et al., 1990; Walcheck et al., 1993; Postigo et al., 1992; Benjamin et al., 1990; Weller et al., 1991; Ockenhouse et al., 1992; Hakkert et al., 1991; Luscinskas et al., 1991; Kishimoto et al., 1991; Kyan-Aung et al., 1991b; Shimizu et al., 1991b; Picker et al., 1991a; Ockenhouse et al., 1992). Moreover, binding of E-selectin to its ligand on neutrophils seems to induce a physiologic upregulation of the adhesive activity of another group of adhesion molecules, the P2 integrins, on the neutrophils (Lo et al., 1991; Kuijpers et al., 1991). However, it has likewise been reported that E-selectin is not involved in leukocyte transmigration (Hakkert et al., 1991; Kishimoto et al., 1991), or eosinophil adhesion to EC (Dobrina et al., 1991). The adhesion of some tumors has also been found to be mediated partially through E-selectin. E-selectin expressing COS cells bind the human colon carcinoma cell line HT-29 (Hession et al., 1990). Melanoma (Rice and Bevilacqua, 1989), carcinoma (Dejana et al., 1992) and leukemia cells (Takada et al., 1993) bind to activated HUVEC, at least partially, through E-selectin. E-selectin has been found in vivo on the post-capillary endothelium of many inflammatory disorders, including systemic sclerosis (Sollberg et al., 1992), late phrase allergic reactions to intradermal allergen (Leung et al., 1991; Munro et al., 1991), 8  dermatitis (Groves et al., 1991; Munro et al., 1992; Cotran et al., 1986), psoriasis (Groves et al., 1991; Munro et al., 1992), dermal infiltrates associated with benign, premalignant and malignant keratinocyte proliferation (Groves et al., 1991), uveitis (Whitcup et al., 1992), peritonitis (Engelberts et al., 1992), appendicitis (Munro et al., 1992; Etzioni et al., 1992), insect bite (Munro et al., 1992), reactive lymph nodes (Munro et al., 1992), sarcoidosis (Munro et al., 1992), tonsilitis (Munro et al., 1992; Etzioni et al., 1992), gingivitis (Moughal et al., 1992), septic shock (Redl et al., 1991), atherosclerotic lesions (van der Wal et al., 1992), rheumatoid and osteoarthritic synovial tissues (Koch et al., 1991), cardiac allografts (Briscoe et al., 1991), kidney transplants (Fuggle et al., 1993), Hodgkin's and T cell lymphoma (Cotran et al., 1986; Etzioni et al., 1992), acute granulomatous lymphadenitis (Cotran et al., 1986), and thyroiditis (Cotran et al., 1986). Patients with leukocyte deficiency type 2, due to the absence of the sialyl-Lewis x ligands for E-selectin have recently been described. Affected individuals have recurrent episodes of bacterial infections, high leukocyte counts and marked defect in neutrophil motility (Etzioni et al., 1992). In contrast, E-selectin-deficient mice show no changes in neutrophil trafficking in inflammation. It is possible that another selectin takes the place of E-selectin in the necessary pathways in these mice (Labow et al., 1994).  1.4.2 Vascular cell adhesion molecule-1 (VCAM-1) VCAM-1 (CD106) is a member of the immunoglobulin (Ig) gene superfamily. It consists of two forms: one with six extracellular Ig-like domains (6D VCAM-1) and another with seven Ig-like domains (7D VCAM-1). The former is the result of the alternative splicing of the latter. EC express mostly 7D VCAM-1. A gene duplication of domains 1 to 3 or 4 to 6 seems to be involved in this form. A 81 kD core protein is glycosylated at seven potential sites to give a 110 kD surface protein. Its ligands bind to the amino-terminal first domain of 6D and 7D VCAM-1 as well as the fourth domain of 7D VCAM-1 (Cybulsky et al., 1991a, Polte et al., 1991; Hession et al., 1991; Cybulsky et 9  al., 1991b; Carlos and Harlan, 1994). VCAM-1 binds to the very late antigen 4 molecule (o^Pi integrin, VLA4, CD49d) (Elices et al., 1990), as well as the CX4P7 integrin (Ruegg et al., 1992; Chan et al., 1992) on lymphocytes and monocytes. VCAM-1 is minimally expressed by unstimulated HUVEC. IL-1, TNF-a, and LPS activated EC express VCAM-1 maximally after 4-6h, and its expression persists for 72h in the presence of the cytokines (Bevilacqua, 1993; Butcher, 1992; Springer, 1994). VCAM-1 is involved in the adhesion and migration of lymphocytes, monocytes, Plasmodium-infected erythrocytes, eosinophils and basophils but not neutrophils across the endothelium (Pober et al., 1990; Bochner et al., 1991; Carlos et al., 1991; Postigo et al., 1992; Weller et al., 1991; Ockenhouse et al., 1992; Ryan et al., 1991; Simmons et al., 1992; Kyan-Aung et al., 1991a). The adhesion of these cells can be partially inhibited by anti-VCAM-1 and anti-VLA4 Abs (Bochner et al., 1991; Sadahiro et al., 1993). It has also been suggested that additional receptors for VLA-4, other than VCAM-1, may be present on activated HUVEC (Vonderheide et al., 1992). Van Kooyk et al. (1993) found that binding of activated T lymphocytes to endothelium is mainly mediated through LFA-1. VLA-4 is used only when LFA-1 is not expressed or not functional as in some T cell leukemia cell lines. However, Baron et al. (1993) demonstrated that VLA-4 may be crucial in allowing activated T lymphocytes to leave the blood and enter the brain and other tissues to clear infections. T lymphocyte clones expressing a high concentration of VLA-4 entered the mouse brain parenchyma in abundance and caused experimental allergic encephalomyelitis, while clones expressing low levels of VLA-4 do not enter the brain and are not encephalitogenic. Expression of VCAM-1 by vascular endothelium has been observed in several pathological and experimental conditions: acute appendicitis, acute colonic diverticulitis, lymphadenitis, dermatoses (Rice et al., 1991; Rice et al., 1990; Briscoe et al., 1992; Orosz et al., 1992; Briscoe et al., 1991), cardiac allograft rejections (Carlos et al., 1992;  10  Taylor et al., 1992), liver and pancreas allograft rejection (Bacchi et al., 1993; Steinhoff et al., 1993; Cybulsky et al., 1991a; Koch et al., 1991).  1.4.3 Intercellular adhesion molecule-1 (ICAM-1) ICAM-1 (CD54) is a member of the immunoglobulin (Ig) gene superfamily. It is synthesized from a single gene on chromosome 19. A 55kD protein is made with five extracellular Ig-like domains with different amounts of N-linked glycosylation in these domains to give a total of 76 to 114 kD. Its ligands bind to the amino-terminal first domain and the third Ig-like domain. The latter is affected by the amount of glycosylation. An eight amino acid portion of the cytoplasmic domain binds to the cytoskeleton, such as actin via cc-actinin. This allows the localization of ICAM-1 to specific regions of the cell membrane (Carlos and Harlan, 1994). ICAM-1 binds to leukocyte function associated molecule 1 (LFA-1), a P2 integrin composed of two portions: CDlla and CD18, on the neutrophil, lymphocyte and monocyte surface, and Mac-1 (CDllb/CD18, Mo-1) on neutrophils, monocytes and NK cells (Arnaout et al., 1988). Fibrinogen (Languino et al., 1993), hyaluronan (Mccourt et al., 1994), CD43 - a leukocyte cell surface molecule also called sialophorin, leukosialin, large sialoglycoprotein or gpll5 (Rosenstein et al., 1991), Plasmodium falciparum infected erythrocytes (Berendt et al., 1989,1992) and rhinovirus (Staunton et al., 1989; Greve et al., 1989) also bind ICAM-1. ICAM-1 is expressed at low levels on unstimulated HUVEC. After activation by IL-1, TNF-a, LPS or IFN-y, ICAM-1 expression reaches maximal levels after 24h (Bevilacqua, 1993; Butcher, 1992; Springer, 1994). ICAM-1 mediates the adhesion and transendothelial migration of lymphocytes, monocytes, neutrophils, basophils and eosinophils (Albelda and Buck, 1990; Albelda, 1991; Bochner et al., 1991; Pober and Cotran, 1991; Shimizu et al., 1991a; Kyan-Aung et al., 1991b). Abs against ICAM-1 or its ligand, LFA-1, can block the adhesion of leukocytes in vitro (Bochner et al., 1991) and the extravasation of leukocytes in vivo (Albelda, 1991; 11  Issekutz et al., 1992; Sadahiro et al., 1993; Kurtel et al., 1992; Scheynius et al., 1993). Microspheres coated with recombinant soluble ICAM-1 (Welder et al., 1993) and synthetic peptides based on specific amino acid sequences of human ICAM-1 (Ross et al., 1992) can also inhibit lymphocyte adhesion in vitro. In fact, it has been observed that decreased leukocyte adhesion with anti-CD 18 Abs is mediated by receptor internalization. This is not associated with cell activation (Rubin et al., 1992). Mice deficient in ICAM-1 exhibit impaired neutrophil migration in inflammation and decreased contact hypersensitivity (Sligh et al., 1993). In another report, such animals were observed to have also elevated circulating leukocytes and diminished allogeneic T cell responses. Their resistance to the lethal effects of endotoxin and exotoxin is due to reduced leukocyte-EC interaction and decreased systemic release of inflammatory cytokines and hepatocyte necrosis (Xu et al., 1994). Recently, it was reported that ICAM-1-deficient mice have greater incidence of Haemophilus influenzae type b bacteremia but less mortality than wild-type mice. In contrast, mortality due to Streptococcus pneumoniae was greater in ICAM-1-deficient mice than wild type. Histology of the meninges was not different between the two types of mice (Tan et al., 1995).  1.4.4 Platelet/endothelial cell adhesion molecule-l (PECAM-1) PECAM-1 (CD31) is also a member of the immunoglobulin (Ig) gene family. It was independently discovered by several groups to be a glycoprotein expressed constitutively by all HUVEC (Simmons et al., 1990; van Mourik et al., 1985; Albelda et al., 1990a; Muller et al., 1989). This is a 130kD glycoprotein present on EC, platelets, most monocytes, some granulocytes and a subset of CD4+ T lymphocytes and CD8+ T lymphocytes, including all naive CD8+ T lymphocytes (Stockinger et al., 1990; Tanaka et al., 1992). The exons of its gene seem to have arisen through gene duplication and shuffling. Alternate splicing of the complex cytoplasmic domain appears to generate 12  multiple isoforms (Kirschbaum et al., 1994). A 80kD core protein has six extracellular Ig-like domains and nine possible N-linked glycosylation sites (Albelda et al., 1991; Carlos and Harlan, 1994). PECAM-1 binds to other molecules of PECAM-1 as well as other structures (Muller et al., 1992; DeLisser et al., 1993a, 1993b). The distribution of PECAM is restricted to the EC borders (Albelda et al., 1991) and has been suggested to be re-distributed in response to cytokines. It is highly expressed especially on endothelium-adherent lymphocytes undergoing transmigration (Bogen et al., 1992). This expression amplifies the leukocytes' adhesive properties by transiently inducing the adhesive function of p^ (VLA) and $2 (LFA-1, Mac-1, pl50,95) integrins (Tanaka et al., 1992). Thus, in collaboration with other adhesion molecules, PECAM-1 functions in an "adhesion cascade" by amplifying integrin-mediated adhesion of PECAM-1+ leukocytes to EC. Anti-PECAM-1 Abs can block neutrophil accumulation in inflammatory sites in vivo (Vaporciyan et al., 1993). In addition, PECAM-1 plays a role in the control of EC migration (Schimmenti et al., 1992).  1.4.5 Vascular addressins In lymphoid organs, such as lymph nodes, lymphocytes leave the blood through high endothelial venules. Lymphocytes bind to these high endothelial venules through "homing receptors" on the lymphocyte surface and "vascular addressins" on the high or flat endothelium in lymphoid organs, and chronically inflamed non-lymphoid tissues. HECA-452 antigen was originally suggested to be a potential vascular addressin in humans, and MECA-325 antigen is its counterpart in mice. HECA-452 antigen can be induced on HUVEC by IFN-y (Albelda and Buck, 1990). A note of unpublished observation indicated that the epitope bound by the only Ab available against this antigen may not be involved in leukocyte binding (Duijvestijin, 1988). HECA-452 antigen was not found in vessels with various dermal inflammatory diseases (Jalkanen et al., 1990),  13  but is present on T lymphocytes, PMN and monocytes at cutaneous sites (Picker et al., 1990; Davis and Smoller, 1992; Horst et al., 1988).  1.4.6 Other endothelial cell adhesion molecules ICAM-2 is another molecule in the Ig gene superfamily. It consists of a core protein of 29 kD with six N-linked glycosylation sites. ICAM-2 has only two Ig-like domains and has been suggested to bind LFA-1 but not Mac-1 (Carlos and Harlan, 1994). Its expression is not regulated by cytokines (Staunton et al., 1989; Nortamo et al., 1991). Another selectin - P-selectin (CD62P) - is also inducible on EC. Initially, this molecule was found on platelets and called platelet-activation-dependent granule-external membrane protein (PADGEM) or granule membrane protein-140 (GMP-140) (Carlos and Harlan, 1994). The core protein of 86 kD has twelve potential N-linked glycosylation sites. A secreted and a membrane associated form of the molecule are produced (Johnston et al., 1990). P-selectin is present in Wiebel-Palade bodies in large vessel EC and a-granules in platelets (McEver et al., 1989; Stenberg et al., 1985). After activation by thrombin or histamine, P-selectin is rapidly redistributed to the cell surface (Stenberg et al., 1985). This surface molecule can then be utilized by EC to bind sialylated molecules, including L-selectin (Carlos and Harlan, 1994) on neutrophils and monocytes (Hamburger and McEver, 1990). It has been found to mediate the initial interactions between EC and leukocytes under conditions of flow, where leukocytes roll on the endothelium (Lawrence and Springer, 1991). P-selectin deficient mice exhibit elevated numbers of circulating neutrophils, loss of leukocyte rolling in venules and delayed neutrophil recruitment in inflammation (Mayadas et al., 1993). It appears that an intact E-selectin pathway is insufficient to completely make up the function. New adhesion molecules continue to be discovered. Some examples are vascular adhesion protein-1 (VAP-1) (Salmi et al., 1992; Jalkanen and Salmi, 1993) and  14  lymphocyte-vascular adhesion protein-2 (L-VAP-2) (Airas et al., 1993). These molecules are beyond the scope of this study.  1.4.7 Adhesion molecule shedding Of interest are reports that several forms of E-selectin (Leeuwenberg et al., 1992; Pigott et al., 1992; Jakobsen et al., 1994; Newman et al., 1993), VCAM-1 (Pigott et al., 1992; Hahne et al., 1994; Jakobsen et al., 1994), ICAM-1 (Seth et al., 1991; Leeuwenberg et al., 1992; Pigott et al., 1992; Clark et al., 1993; Rieckmann et al., 1993; Sharief et al., 1993b; Pizzolo et al., 1994; Jakobsen et al., 1994) and PECAM-1 (Goldberger et al., 1994) can be found in soluble form in culture media after EC activation, and in the circulation during inflammatory and immune reactions. In multiple sclerosis (MS) patients, VCAM-1 and E-selectin in the cerebrospinal fluid, as well as ICAM-1 and E-selectin in the serum are elevated (Dore-Duffy et al., 1995). Soluble ICAM-1 has been found to decrease lymphocyte adhesion to activated HUVEC in vitro (Hashimoto et al., 1994). For some adhesion molecules, this may influence the dynamics of leukocyte adhesion and diapedesis in vivo.  1.4.8 Leukocyte adhesion molecules A number of receptors for endothelial cell adhesion molecules have been characterized on leukocytes. L-selectin (CD62L) is a member of the selectin family that is constitutively present on lymphocytes, neutrophils and monocytes (Bevilacqua, 1993). A 37 kD core protein is differentially glycosylated in different leukocytes at eight possible N-glycosylation sites (Carlos and Harlan, 1994). Like other members of the selectin family, L-selectin is employed during flow conditions in the rolling of leukocytes on EC. It binds sialylated structures, CD34 and the murine mucosal lymphoid addressin MAdCAM-1 (Carlos and Harlan, 1994). When leukocytes are activated, the activity of L-selectin is transiently upregulated before its shedding (Bevilacqua, 1993). 15  In L-selectin deficient mice, lymphocytes do not bind peripheral lymph node high endothelial venules and the number of lymphocytes are reduced in peripheral lymph nodes, mucosal lymph nodes, Peyer's patch and spleen. They also show defects in neutrophil migration to inflammatory sites and leukocyte rolling (Arbones et al., 1994). Any rolling present is mediated by P-selectin. No evidence of E-selectin involvement is observed (Ley et al., 1995). The integrins are another major group of adhesion molecules on leukocytes. Each integrin consists of a heterodimer of noncovalently linked a and |3 chains. Integrins are separated into groups by the p chain used. The (32 integrins, cxjj^ (CDlla/CD18, LFA-1), a M p 2 (CDllb/CD18, Mac-1) and perhaps a x p 2 (CDllc/CD18, P150,95) mediate binding to the ICAMs. LFA-1, on lymphocytes, neutrophils, monocytes and NK cells, binds ICAM-1 and ICAM-2. Mac-1, on neutrophils, monocytes and NK cells, binds ICAM-1, fibrinogen, factor X and complement fragments. P150,95, on neutrophils, monocytes and NK cells, binds complement fragments and plays only a minor role in leukocyte adhesion to EC. Mac-1 and pl50,95, but not LFA-1, are present in vesicles / granules in neutrophils and monocytes. Activation of leukocytes and binding of leukocytes to E-selectin increase the binding activity of these integrins (Carlos and Harlan, 1994). A lack of p 2 integrins (LFA-1, Mac-1, pl50,95) or nonfunctional p 2 integrins on leukocytes causes a human disorder called leukocyte adhesion deficiency. There is abnormal leukocyte aggregation, margination, chemotaxis and phagocytosis, leading to a decrease in inflammation, poor wound healing, absence of pus, recurrent bacterial and fungal infections and persistent neutrophilia (Anderson et al., 1987). The p-L integrin VLA-4 ( a 4 p 1 ; CD49d/CD29) and (37 integrin a 4 p 7 bind VCAM-1, although the latter plays a minor controversial role (Bevilacqua, 1993). VLA-4 is present on lymphocytes, monocytes, eosinophil, basophil and NK cells but not neutrophils, and a^ftj on lymphocytes. In addition to VCAM-1, VLA-4 bind the 16  bacterial outer membrane protein invasin, and extracellular matrix proteins fibronectin and thrombospondin. The binding site for VCAM-1 and fibronectin, at least, are different (Carlos and Harlan, 1994).  1.4.9 Adhesion molecule therapy Anti-adhesion molecule Abs hold great promise as a therapy for inflammatory conditions. In animal models, anti-ICAM-1 Abs inhibit experimental allergic encephalomyelitis (EAE, an animal model for MS) induced by active immunization with myelin but not passive EAE induced by transfer of myelin basic protein-specific CD4+ T lymphocytes (Archelos et al., 1993; Willenborg et al., 1993; Cannella et al., 1993). Anti-o^p^ integrin (VLA-4) Abs prevent passively transferred EAE (Yednock et al., 1992). CNS injury due to ischemia (Clark et al., 1991; Bowes et al., 1993) and dermal thermal injury (Mileski et al., 1992) are also reduced by anti-ICAM-1 Abs. Abs against ICAM-1 and its ligand LFA-1 can inhibit leukocyte accumulation in the postcapillary venules of rabbits treated with zymosan-activated serum (C5a source) to induce leukocyte adhesion (Argenbright et al., 1991). Anti-PECAM-1 Abs block leukocyte emigration in a murine model of peritonitis (Bogen et al., 1994). Anti-VCAM-1 Abs decrease the rejection of murine cardiac allografts (Pelletier et al., 1992) and, on occasion, induce sometimes indefinite cardiac allograft acceptance (Isobe et al., 1994). In addition to wound healing, a long list of diseases have been targeted, including inflammation, cancer, MS, viral infections, rheumatoid arthritis, osteoporosis, atherosclerosis and the common cold. The rhinovirus which causes 50% of all colds enters cells by binding to the adhesion molecule ICAM-1. Genetically engineered ICAM-1 fragments fused to antibody (Ab) fragments are being generated to block rhinovirus infections. ICAM-1, VCAM-1 and E-selectin, are being examined as potential treatments for ischemia and reperfusion injury and inflammation (Peake, 1993; Rothlein et al., 1993; Travis, 1993; McMillen et al., 1993; Seekamp and Ward, 1993; Seekamp et 17  al., 1994). Clinical trials with anti-ICAM-1 Abs have commenced. In cadaveric kidney transplantation, they decrease the incidences of delayed graft function and rejection (Haug et al., 1993). Marked or moderate improvements have also been seen in rheumatoid arthritis patients treated with anti-ICAM-1 Abs (Kavanaugh et al., 1992).  1.5LIPOPOLYSACCHARIDEAND CYTOKINES It is apparent from the discussion of adhesion molecule upregulation above that bacterial lipopolysaccharide (LPS) and cytokines, such as tumor necrosis factor-cc (TNF-a), interleukin-ip (IL-1J3) and interferon-y (IFN-y) can be very important in inflammation. Cytokines are defined as non-antibody proteins that act as intercellular mediators. In addition to their role in the upregulation of adhesion molecules and movement of leukocytes into inflammatory lesions, they exhibit various proinflammatory and pro-thrombotic activities.  1.5.1 Lipopolysaccharide (LPS) Gram negative bacteria produce a toxin called endotoxin which is a component of the bacterial cell wall and is released by living as well as dead or disintegrated bacteria. It has been defined and given the name lipopolysaccharide (LPS) (Abbas et al., 1991; Doran, 1992). LPS is composed of a variable region of repeating oligosaccharide units called o-specific polysaccharide; core polysaccharides made of an inner-core of D-glucose, D-galactose and N-acetyl-glucosamine, and an outer core of 2-keto-3-deoxyoxtonic acid, heptose, phosphate and phosphate-bound ethanolamine; and lipid A - the most conserved region made of long chain fatty acids attached to a phosphorylated glucosamine-disaccharide backbone (Doran, 1992). LPS activates EC to express adhesion molecules, produce cytokines and activates both the classical and alternate pathways of complement (actions of lipid A and the oligosaccharides, respectively) with generation of the anaphylotoxins C3a and C5a (Abbas et al., 1991; 18  Doran, 1992). However, most of the toxicity caused by LPS is due to the host's response. LPS stimulates T and B lymphocytes, mononuclear leukocytes, EC and fibroblasts to secrete colony stimulating factors that stimulate progenitors to differentiate into mature leukocytes. In addition, it stimulates monocytes / macrophages to secrete large amounts of TNF, IL-1, prostaglandins and leukotrienes which may be responsible for much of LPS's biological activities. Endotoxin has been found to occur at <25 to 260ug/ml in the plasma of meningitis patients, and 210 to 170,000ug/ml in septicemia, but tissue concentrations may be much higher (Brandtzaeg et al., 1989).  1.5.2 Tumor necrosis factor (TNF) There are two species of TNF: TNF-a and TNF-J3 (lymphotoxin). They share 30% homology and are encoded by different genes, but have the same receptor and action. TNF is produced by lymphocytes, mast cells, PMN, keratinocytes, astrocytes, microglial cells, smooth muscle cells, intestinal paneth cells and tumor cells. By far the largest amount is produced by monocytes / macrophages activated by LPS, bacteria, viruses, fungi, parasites or tumor cells (Jattela, 1991; Vassalli, 1992). The actions of TNF are due to the activation of several nuclear-binding proteins / transcription factors (NF). In lymphocytes and monocytes, at least three of these NF-KB-like DNA binding proteins interact with KB-like enhancement elements to synthesize IL-2 receptors and express class II major histocompatibility complex (MHC) proteins. Activation of the transcription factor IFN regulatory factor 1 induces the synthesis of IFN-p\ Lastly, activation of the jun/AP-1 transcription factor activates collagenase in fibroblasts and class I MHC genes (Jattela, 1991). TNF has many effects on many different cell types. Its most important action on EC is the induction/upregulation of adhesion molecules. In addition, it induces the production of IL-1, colony stimulating factors (to increase circulating leukocytes), 19  chemokines (IL-8, MCP-1), prostaglandins and class IMHC protein, and inhibits proliferation. Coagulation is also promoted by suppressing thrombomodulin synthesis to suppress EC surface anticoagulant activity, increasing tissue factor-like procoagulant activity, and decreasing plasminogen activator synthesis (Mantovani and Dejana, 1989; Vassalli, 1992). In Kawasaki disease, up to 68U TNF/ml have been found in the patient's serum (Matsubara et al., 1990). A much higher concentration of up to 500U TNF/ml was found in the serum of patients with Guillain-Barre syndrome (Sharief et al., 1993a). The concentration of TNF rose to 15U TNF/mg cortex homogenate after fluid-percussion trauma of rat brains (Taupin et al., 1993).  1.5.3 Interleukin-1 (IL-1) There are two species of IL-1: IL-la and IL-1 (3. They are produced by different genes but share a related tertiary structure, bind the same receptor and generate the same biological activities. IL-1 is produced mainly by mononuclear leukocytes (Epstein, 1993). Most actions of TNF are shared by IL-1, including induction of adhesion molecules, chemokines, colony stimulating factors, prostaglandin, procoagulant activity, suppression of fibrinolytic activity and EC growth. The exceptions are its lack of effect on class I MHC protein expression and lack of cytotoxicity (Mantovani and Dejana, 1989; Abbas et al., 1991). Mechanical trauma in rats caused up to 3U IL-l/ml to be released into the brain as measured by a microdialysis technique (Woodroofe et al., 1991). IL-1 levels reached 18U IL-l/mg cortex homogenate following fluid-percussion trauma of rat brains (Taupin et al., 1993).  1.5.4 Interferon-y (IFN-y) IFN-Y is a cytokine quite different from TNF and IL-1, and even from other interferons (a and P). Probably all cell types produce IFN-a/p, but IFN-y is only 20  synthesized by T lymphocytes and natural killer (NK) cells after antigen and IL-2 stimulation, respectively. The actions of IFN-y are mainly mediated by the activation of protein kinase C and intracellular calcium fluxes followed by de novo RNA and protein synthesis (Williams et al., 1993). IFN-y induces synthesis of ICAM-1, class I and II MHC molecules and IL-1 on EC. It inhibits growth most likely by downregulating fibroblast growth factor receptors. In addition, it exhibits antiviral and antiprotozoal actions (Sen and Lengyel, 1992). Up to 2.5U IFN-y/ml have been found in the serum of patients with Kawasaki disease (Matsubara et al., 1990) and in human T lymphotropic virus type I-associated myelopathy (Kuroda and Matsui, 1993), while up to 3.5U IFN-y/ml was found in the cerebrospinal fluid (CSF) of the latter.  1.6 ENDOTHELIAL CELL HETEROGENEITY Most of the observations previously reported on cytokines and EC adhesion molecules were made using HUVEC. It is not clear how this fetal vascular tissue applies to the adult vasculature and whether results can be extrapolated particularly to specialized endothelium in other organs such as the brain. It is now well accepted that EC of different species and vascular beds have distinct phenotypes, functions and antigenic properties. There is already ample evidence that EC lining the cerebral microvessels are different from their counterparts in other organs. Cerebral capillaries contain the decarboxylation enzyme that converts L-dopa to dopamine while peripheral capillaries do not (Owman and Hardebo, 1988). In vitro, cerebral EC have greater y-glutamyl transpeptidase and alkaline phosphatase activity than aortic EC (Gordon et al., 1991), and aortic and adrenal cortical EC have protein kinase C type III while cerebral EC have type II and III enzymes (Kobayashi et al., 1990). More importantly, lymphocyte adhesion to unstimulated cerebral endothelium is much lower than to aortic endothelium. IFN-y induces two to three fold more class II MHC molecules on cerebral than aortic EC (Male 21  et al., 1990a). Another very important property of cerebral EC is their ability to form the blood-brain barrier (BBB).  1.7 THE BLOOD-BRAIN BARRIER (BBB) The brain is vigorously maintained in a constant microenvironment for the proper functioning of the neurons by means of a structural and functional barrier, the bloodbrain barrier (BBB). Two morphological features of the cerebral endothelium are responsible for the formation and maintenance of the BBB. Firstly, the presence of tight junctions between adjacent EC prevents the movement of ions, proteins, large lipid-insoluble nonelectrolytes and circulating blood cells from the blood into the brain paracellularly. Tight junctions form continuous belts made up of interconnected loops of strands around EC. At a junction, the two apposed plasma membranes come together and the outer leaflets fuse, obliterating the intercellular space. It has been hypothesized that the complexity of the tight junctional complexes correlates with the degree of protein impermeability (Claude and Goodenough, 1973), although it is uncertain how the anatomical subunits of tight junctions relate to their barrier function. EC junctions in normal intact brains do not appear to open and close since molecular tracers, such as horseradish peroxidase, and ionic and colloidal lanthanum are not found in any of the lacunae - intervening pockets formed by tandem junctions (Dorovini-Zis et al., 1983; Brightman, 1977). A second method to examine the efficiency (tightness) of the barrier involves measuring the electrical resistance across the endothelium. The average resistance of cerebral capillaries is 1900Q-cm^ in frogs and 3000Q-cm^ in rats. This is in the same order of magnitude as the tightest epithelium (Bradbury, 1993). Secondly, the paucity of cytoplasmic vesicles restricts the transport of solutes across the EC cytoplasm. A distinction must be made between two types of cytoplasmic vesicles. Pinocytosis generates vesicles that fuse with lysosomes or multivesiculate 22  bodies, the contents are digested and the whole process ends inside the cell. In vesicular transport, vesicles transfer material from the fluid compartment at one side of the cell to the other side. These vesicles do not appear to any extent in cerebral microvessel EC, but are abundant in skeletal muscle and other EC (Brightman, 1977). Of the vesicles that are present in cerebral EC, most have been found not to be true vesicles but rather invaginations connected via narrow necks to the extracellular milieu or are clathrin-coated vesicles (Bradbury, 1993). A number of specific transporters present on the luminal or abluminal EC membranes ferry biologically important substances across this efficient barrier. Many of these mechanisms are specific to the cerebral EC, such as the Glut-1 glucose transporter, y-glutamyltranspeptidase, alkaline phosphatase and abluminal Na + -K + -ATPase (Janzer, 1993). Thus, in contrast to the extracerebral microvasculature, cerebral EC permit only a very specific, highly selective exchange between the brain and the rest of the body.  1.8 INFLAMMATION IN THE CENTRAL NERVOUS SYSTEM  1.8.1 General aspects Inflammation in the CNS is characterized by increase in BBB permeability and influx of inflammatory cells into the brain. Acute CNS inflammation is characterized by PMN infiltration while chronic inflammation by mononuclear inflammatory cells, especially lymphocytes and macrophages. In acute-types of inflammation, such as those caused by bacteria (abscess, meningitis), viruses (viral meningitis and encephalitis), and infarction, there is an initial (1-2 days) migration of PMN across dilated vessels. PMN surround these vessels to form "perivascular cuffs". This short lived process may be accompanied by edema, hemorrhage and necrosis in bacterial infections. PMN not only release hydrolytic enzymes and free radicals to combat invading organisms but also phagocytoze and 23  destroy them. Later, after 4-5 days, macrophages, lymphocytes and plasma cells replace PMN in the extravascular tissue. Plasma cells generate Abs that opsonize invading organisms. This facilitates phagocytosis by macrophages, which also present antigens of these invaders to CD4+ T lymphocytes. The latter secrete cytokines to aid the inflammatory and immune reactions, in response to antigen recognition. In infarcts, the initial infiltration by PMN is followed, 36 to 48h later, by macrophage infiltration. If necrosis occurs in the brain parenchyma, inflammatory cells will stream from the perivascular spaces to the necrotic brain tissue (Brierley and Graham, 1984; Harriman, 1984; Bromwell and Tomlinson, 1984). A more chronic-type of CNS inflammation occurs in MS. In acute MS, the entire lesion is intensely inflammatory. Small mononuclear cells (T lymphocytes and macrophages) form perivascular cuffs. In chronic MS, inflammatory cells are rare in inactive plaques, but in active plaques, lymphocytes, some macrophages and plasma cells form perivascular cuffs. Only a small number of PMN is present (Raine, 1991).  1.8.2 Adhesion molecules in the central nervous system The mechanisms responsible for the recruitment of leukocytes to sites of CNS inflammation are largely unknown. Several lines of investigation indicate that expression or upregulation of EC adhesion molecules may play an important role during leukocyte-EC interactions in inflammation. There have been a few reports on the presence of adhesion molecules in CNS disorders in vivo. Sobel et al. (1990) reported that microvessels along with glia are ICAM-1 positive in MS plaques, viral encephalitis lesions, infarcts and some controls. Lymphocytes from MS patients induce human brain EC to express and release high concentrations of ICAM-1 in vitro (Tsukada et al., 1994). HECA-452 antigen was found in cerebral postcapillary venules in the peri-plaque area in a rapidly evolving case of MS (Raine et al., 1990a). The attachment and infiltration of lymphocytes in EAE was found to correlate with the appearance of MECA-325 and 24  ICAM-1 on the CNS vessels (Raine et al., 1990b; Cannella et al., 1990,1991). Recently, we demonstrated that E-selectin and VCAM-1 are present in patients with encephalitis, CNS vasculitis, multiple sclerosis, meningitis, brain abscess and to a lesser extent, cerebral infarcts, but not in traumatic lesions or normal brain (Dorovini-Zis et al., 1992b; Dorovini-Zis et al., 1993). Positive staining of vessels is usually associated with leukocytes undergoing diapedesis and extensive inflammatory cell infiltration. ICAM-1 and VCAM-1 are also upregulated in the CNS of mice with EAE (Dopp et al., 1994; Barten et al., 1994; Steffen et al., 1994), while ICAM-1 is upregulated in baboons after focal brain ischemia and reperfusion (Okada et al., 1994). Akiyama et al. (1993) and Verbeek et al. (1994) found some intense vascular staining for ICAM-1 in Alzheimer's disease, especially in areas with a large number of senile plaques, and in amyotrophic lateral sclerosis.  1.8.3 Cytokines in central nervous system disorders The cytokines that upregulate these adhesion molecules on HUVEC have also been found in CNS lesions in a variety of disorders: TNF-a in MS, bacterial meningitis, HIV infections and traumatic injury; IL-1 in MS, Alzheimer's disease and traumatic injury; and IFN-y in MS (Benveniste, 1992; Morganti-Kossmann et al., 1992; Woodroofe and Cuzner, 1993). TNF expression is associated with astrocytes, macrophages, microglia, some EC and perivascular inflammatory cuffs in EAE (Renno et al., 1995), encephalitis, acute and chronic active MS lesions but not chronic inactive lesions (Hofman et al., 1989; Woodroofe and Cuzner, 1993; Rieckmann et al., 1995). Intracerebroventricular administration of LPS into rats also induces the expression of TNF in the CSF (Sanna et al., 1995). IL-1 is also present in perivascular cells in chronic active MS lesions (Hofman et al., 1989; Woodroofe and Cuzner, 1993), in macrophages and activated microglia throughout the course of EAE in mice and rats (Kennedy et al., 1992; Bauer et al., 1993), as well as in ischemic rat brain (Liu et al., 1993; Wie^ner C, 25  1993), and rat brain after intraperitoneal or intraventricular injection of LPS (Quan et al., 1994).  IFN-Y  is secreted by some lymphocytes, astrocytes and glia in EAE (Traugott and  Lebon, 1988; Kennedy et al., 1992, Morganti-Kossmann et al., 1992; Woodroofe and Cuzner, 1993). All of these cytokines are also found in the CSF and brain of patients with meningitis and mice with lymphocytic choriomeningitis (Moller et al., 1991; Campbell et al., 1994).  1.9 HUMAN BRAIN MICROVESSEL ENDOTHELIAL CELLS (HBMEC) IN VITRO There is increasing evidence that HBMEC and their interaction with cytokines and leukocytes play a key role in CNS inflammation. Experiments in vivo can document the site of the BBB breakdown, but cell-cell interactions are difficult to characterize because of the many complex factors involved. This has led to the establishment and use of in vitro models of the BBB that respond to experimental manipulation. Thus, cultures of bovine brain microvessel EC (BBMEC) share the morphological and barrier characteristics of their in vivo counterparts. Neither ionic lanthanum nor horseradish peroxidase penetrate the interendothelial tight junctions normally, but hyperosmotic arabinose and urea solutions open these junctions to let the tracers penetrate paracellularly (Dorovini-Zis et al., 1984,1987a, 1987b). PMN can be stimulated by a gradient of formyl-methionyl-leucyl phenylalanine (fMLP), leukotriene-B4 (LTB4) or acetyl-glyceryl-ether-phosphorylcholine (AGEPC) to adhere to and migrate across monolayers of BBMEC (Dorovini-Zis et al., 1992a). Isolation and culture of cerebral EC from human brains has permitted the study of these cells and their barrier function in vitro. HBMEC monolayers in primary culture retain essential features of the BBB: tight junctions between adjacent cells that restrict the paracellular movement of horseradish peroxidase and a paucity of cytoplasmic vesicles (Dorovini-Zis et al., 1991). Using this system, it has been demonstrated that IFN-y induces de novo expression of class II MHC molecules in these cells, changes their morphology and increases the monolayer 26  permeability to horseradish peroxidase (Huynh and Dorovini-Zis, 1993). IFN-p reverses these changes (Huynh, 1994). Expression of class II MHC molecules is associated with increase in the adhesion and migration of resting and anti-CD3 activated T lymphocytes across HBMEC (Huynh, 1994). In the present study, the expression of adhesion molecules by HBMEC and their role in HBMEC-leukocyte interactions were examined using this in vitro model of the human BBB.  1.10 GENERAL AIM OF THIS STUDY The overall aim of this study is to characterize (D the expression / upregulation of adhesion molecules by human brain microvessel EC, (D their role in the adhesion and migration of inflammatory cells across the endothelium and (D the effect of EC-leukocyte interactions on the permeability of the blood-brain barrier (BBB).  1.11 SPECIFIC OBJECTIVES OF THIS STUDY  1.11.1 Phase I. Investigation of the expression of adhesion molecules by human brain microvessel endothelial cells in primary culture 1. To study the upregulation of E-selectin on human brain microvessel EC by cytokines and LPS. 2. To study the expression of VCAM-1 by human brain microvessel EC and its upregulation by cytokines and LPS. 3. To study the expression of ICAM-1 by human brain microvessel EC and the kinetics of its upregulation by cytokines and LPS. 4. To study the expression and change in distribution of PECAM-1 by human brain microvessel EC following activation with cytokines and LPS.  27  5. To study the upregulation of the high endothelial venule antigen - HECA-452 antigen on human brain microvessel EC by cytokines.  1.11.2 Phase II. Investigation of the role of endothelial cell adhesion molecules in leukocyte adhesion and transmigration across human brain microvessel endothelial cell monolayers 1. To investigate the role of these adhesion molecules in the adhesion and transmigration of PMN. 2. To study the modulation of T lymphocyte adhesion and migration across HBMEC monolayers by EC adhesion molecules.  1.11.3 Phase III. Investigation of the monolayer permeability after cytokine treatment and in the course of HBMEC-Ieukocyte interactions 1. To determine the role of TNF, as well as PMN and T lymphocyte migration and adhesion molecule blocking Abs on the permeability of confluent HBMEC monolayers.  28  CHAPTER 2 MATERIALS AND METHODS  2.1 EXPRESSION OF ADHESION MOLECULES  2.1.1 Human brain microvessel endothelial cells Primary cultures of HBMEC were established by methods previously described (Dorovini-Zis et al., 1991). Briefly, cerebral cortex was obtained from temporal lobectomy specimens removed from patients with intractable seizure disorders and autopsy brains less than 12 hours postmortem. The clinical history at the time of postmortem examination was carefully screened to exclude brains with various neuropathological processes including inflammatory / infectious diseases, tumors, infarcts and cerebral anoxia. Cerebral cortices were transferred to the laboratory in medium 199 (M199, Gibco, Burlington, ON) with antibiotics at 4°C. Meninges and associated large vessels were peeled off before cutting the tissue into millimeter cubes with a pair of scalpels. This was then digested in 0.5% dispase (Boehringer Mannheim, Indianapolis, Indiana) in a shaking water bath at 37°C for 3h. Following centrifugation at lOOOx g for lOmin, pellets were centrifuged at 5800x g in 15% dextran (Sigma, St. Louis, Missouri) for lOmin. The pellet containing the microvessels was washed and incubated with lmg/ml collagenase/dispase (Boehringer Mannheim) at 37°C overnight (10-16h), in order to free the EC from pericytes and basement membrane. EC were separated from pericytes, erythrocytes and basement membrane by centrifugation in Percoll gradients (Sigma) for lOmin at lOOOx g. The band containing the EC was aspirated and washed. The isolated clumps of EC were seeded onto fibronectin-coated 24 or 96 well plates (Corning Plastics, Corning, N.Y.) and maintained in culture in M199 supplemented with 10% horse serum, endothelial cell growth supplement (20ng/ml), heparin (100u,g/ml) (both from Sigma) and antibiotics (Gibco). Culture media were 29  changed every other day. EC reached confluency 7 to 9 days after plating and were used for studies on or after the tenth day (Fig. 1). The endothelial origin of the cells was determined by their positive staining for Factor VIII antigen (Fig. 2) and binding of Ulex europaeus agglutinin as previously described (Dorovini-Zis et al., 1991). Several primary cultures were used for each experiment.  2.1.2 Adhesion molecule induction by lipopolysaccharide and cytokines Bacterial lipopolysaccharide (LPS from E. coli 055:B5, Sigma, St. Louis, MO), human recombinant interferon-y (IFN-y, Collaborative Research Inc., Bedford, MA), human recombinant tumor necrosis factor-a (TNF-a, Sigma) or human recombinant interleukin-ip (IL-lp, Boehringer Mannheim, Laval, PQ) were dissolved in complete growth media at a final concentration of 0.001, 0.01, 0.1, lug LPS/ml; 10, 50,100, 500 U IFN-y/ml; 1,10,100 U TNF-a/ml; or 0.1,1,10 U IL-lp/ml. Ten to eleven day old cultures were incubated with different concentrations of LPS and cytokines for 4h, 12h, 24h and 48h. In addition, separate wells were incubated with various concentrations of TNF-a and IL-ip (1 U TNF-a/ml + 0.1 U IL-lp7ml; 10 U TNF-a/ml + 0.1 U IL-lp/ml) or TNF-a and IFN-y (1,10, 100 U TNF-a/ml + 100, 500 U IFN-y/ml) to examine the combined effects of these cytokines. Duplicate or triplicate wells were used for each treatment. Several primary cultures derived from different autopsy brains were utilized in these experiments. The reversibility of ICAM-1 expression was studied by incubating confluent, eight to ten day old, monolayers with LPS (0.001, 0.01, 0.1, 1, 5ug/ml); IFN-y (100, 500 U/ml); TNF-a (1,10, 100 U/ml) or IL-ip (0.1,1,10 U/ml) for 48h. At the end of the incubation period, the cultures were washed three times with M199 to remove the cytokines and LPS, placed in complete media and then returned to the incubator for 3 days prior to ICAM-1 detection.  30  2.1.3 Antibodies Mouse monoclonal anti-human ICAM-1 OKT27, a gift from Dr. P. Rao (Ortho Diagnostics, Raritan, NJ) and RR1/1, a gift from Dr. T. Springer (Harvard Medical School, Boston, MA), anti-human VCAM-1 2G7 and anti-human E-selectin 3B7, gifts from Dr. W. Newman (Otsuka America Pharmaceutical Inc., Rockville, MD), anti-human PECAM-1 4G6, a gift from Dr. S. Albelda (University of Pennsylvania, Philadelphia, PA), and rat anti-human HECA-452 IgM, a gift from Dr. Duijvestijin (University of Limburg, Maastricht, The Netherlands) were used as the primary Ab. Anti-human pituitary follicle-stimulating hormone IgG (Biogenex Lab, CA) was used as an irrelevant Ab in place of primary Ab in immunogold silver staining. Goat anti-mouse IgG coupled to 5nm gold particles (Auroprobe ™ LM GAM IgG, Janssen / Cedarlane Labs Ltd., Hornby, ON) and goat anti-rat IgM coupled to 12nm gold particles (Jackson/ Bio|Can, Mississauga, ON) were used as the secondary Ab for immunogold silver staining and goat anti-mouse IgG coupled to 5nm gold particles (Auroprobe ™ GAM IgG G5, Janssen / Cedarlane Labs Ltd., Hornby, ON) was used as the secondary Ab for immunogold EM. Horseradish peroxidase-conjugated goat anti-mouse Ab (Jackson Immunoresearch Laboratories, West Grove, PA) was used for ELISA.  2.1.4 Immunogold silver staining (IGSS) for light microscopy Following treatment with cytokines or LPS, cultures were washed briefly with buffer containing phosphate buffered saline (PBS) with 1% bovine serum albumin (BSA), 1% normal goat serum (NGS) and 0.05% sodium azide (PBS/BSA/NGS) (all from Sigma) and then incubated for 40 minutes at room temperature with the primary Ab at a final concentration of 2ug/ml for anti-ICAM-1, 5ug/ml for anti-VCAM-1 and anti-E-selectin, 1:10 dilution for anti-PECAM-1 and 1:20 dilution for anti-HECA-452 antigen in carrier buffer containing PBS with 5% BSA, 4% NGS and 0.05% sodium azide. The sodium azide was added to prevent internalization of antigen-Ab complexes. 31  At the end of the incubation period, the monolayers were washed twice with PBS/BSA/NGS, and then incubated with the secondary Ab, Auroprobe™ LM GAM IgG, at 1:40 dilution or goat anti-rat IgM Ab at 1:20 dilution in carrier buffer for lh at room temperature. Subsequently, cultures were washed with PBS/BSA/NGS, fixed in 9.25% formaldehyde and 45% acetone in PBS for 30 seconds, washed with distilled deionized water and incubated in silver enhancing solution (IntenSE M™, Janssen / Cedarlane) for 22-26 minutes at 21°C. EC were counterstained with Giemsa. Except as noted above, all cultures were grown, fed, stained and counted the same way at the same time. Controls included cultures incubated in growth media in the absence of cytokines or LPS, and monolayers incubated with 1) normal mouse IgG (Cedarlane) at the same concentration as the primary Ab, 2) irrelevant Ab and 3) carrier buffer instead of the primary Ab. Immunostained cultures were examined under a Nikon Labophot light microscope. Quantitation of adhesion molecule expression was performed by counting cells as labeled or unlabeled in four peripheral and one central randomly selected fields with an ocular grid under 20X magnification. An average of 200 EC were present per optic field. All measurements were performed blindly by masking the identity of the wells prior to counting.  2.1.5 Enzyme-linked immunosorbent assay (ELISA) Untreated as well as LPS or cytokine-treated monolayers were washed with PBS and fixed in 0.025% glutaraldehyde for 10 minutes. Following a rinse with PBS and three washes with PBS/BSA/NGS, cells were incubated for 60 minutes at room temperature with the primary Abs at the same concentration as for IGSS in carrier buffer containing PBS with 5% BSA and 4% NGS. At the end of the incubation period, the monolayers were washed with PBS/BSA/NGS and incubated with the secondary Ab (horseradish peroxidase conjugated goat anti-mouse IgG) at 1:5000 dilution in carrier 32  buffer for lh. Subsequent washes in PBS were followed by incubation with 2mg/ml o-phenylenediamine with 0.015% hydrogen peroxide in 0.1M PBS for 45 minutes. The color development was stopped by the addition of 2M sulfuric acid or 3M HC1 and the absorbance read on a ELISA microtiter plate reader at 490nm. Except as noted above, all cultures were grown, fed, stained and assayed the same way at the same time. Controls included cultures incubated in growth media in the absence of cytokines and LPS, and monolayers incubated with normal mouse IgG (Cedarlane) at the same concentration as the primary Ab or carrier buffer instead of the primary Ab.  2.1.6 Immunogold electron microscopy (IGEM) Nine day old subconfluent HBMEC monolayers were treated with 100U TNF-a/ml for 4,18 or 24h for maximal upregulation of E-selectin, VCAM-1 and ICAM-1, respectively. For PECAM-1, EC were treated with 500U IFN-y/ml for 48h. Cytokine treated as well as control untreated cultures were washed briefly with cold PBS/BSA/NGS and then incubated with the appropriate primary Ab at the same final concentration as IGSS in cold carrier buffer for 30 min at room temperature. The carrier buffer had the same composition as the one used in IGSS. At the end of the incubation period, the monolayers were washed three times with cold PBS/BSA/NGS and then incubated with the secondary Ab (5nm gold-conjugated goat anti-mouse IgG) at a 1:10 dilution in cold carrier buffer for lh. Subsequently, cultures were washed with cold PBS/BSA/NGS, fixed in 2.5% glutaraldehyde and 2% paraformaldehyde in 0.2M sodium cacodylate buffer for lh at 4°C, washed with 0.2M cacodylate buffer, post fixed with 1% osmium tetroxide for lh at 4°C, and en bloc stained with uranyl magnesium acetate overnight at 4°C. Dehydration in graded series of methanol was followed by embedding in epon-araldite. Blocks cut from embedded cultures were re-embedded perpendicular to the growth area for cross sectioning. Controls included cultures grown in the absence of cytokine, and monolayers incubated with carrier buffer instead of primary Ab. Thin 33  sections were viewed unstained with a Philips EM 400. Except as noted above, all cultures were grown, fed, stained, processed and examined with the TEM the same way at the same time. Quantitation. One hundred each of treated and untreated cells were photographed at standard magnification and the number of gold particles bound to the apical and basal surfaces were counted. The actual length of the cell membrane on each surface was determined by measuring the length on the electron micrograph, taking into account the magnification. The number of gold particles per [xm of EC membrane length was calculated using the above measurements.  2.1.7 Controls Controls included cultures incubated in growth media in the absence of cytokines or LPS, and monolayers incubated with 1) normal mouse IgG (Cedarlane) at the same concentration as the primary Ab, 2) irrelevant Ab and 3) carrier buffer instead of the primary Ab.  2.1.8 Statistics Statistics were performed using the computer program SigmaStat (Jandel Scientific, San Rafael, CA). ANOVA was first performed. Where significant differences were found, Student's t-test was applied to the data to determine the significance of the differences between cultures treated with cytokine or LPS and those untreated in the IGSS and ELISA data. Mann-Whitney test was employed to determine the significance of these differences in the IGEM data since they were not normally distributed. P<0.05 was taken as statistically significant for all analyses.  34  2.2 ROLE OF ADHESION MOLECULES INPMNAND T LYMPHOCYTE ADHESION TO AND MIGRATION ACROSS HUMAN BRAIN MICROVESSEL ENDOTHELIAL CELL MONOLAYERS  2.2.1 Human brain microvessel endothelial cells Primary cultures of HBMEC were established from brains as described in section 2.1.1. For the adhesion assays, cells were grown on fibronectin coated 96 well plates, and for migration assays, scanning electron microscopy (SEM) and transmission electron microscopy (TEM) on permeable membranes of purified collagen (Cellagen discs, ICN, Cleveland, OH). Individual Cellagen discs were placed in separate wells of 24 well plates. Supports at the bottom of the Cellagen discs separate them from the wells so that the Cellagen discs form the upper chamber and the 24 well plate the lower chamber of a double chamber chemotaxis system (Fig. 3). HBMEC readily grow on collagen membranes and form confluent cultures. 9-10 day old confluent cultures were used. Several primary cultures from different autopsy brains were utilized.  2.2.2 Polymorphonuclear leukocyte (PMN) isolation PMN were isolated from anti-coagulated peripheral blood of several healthy donors by centrifugation in a gradient of Lympholyte-poly (Cedarlane). By this method, we routinely obtained cell fractions containing 99% PMN as determined on Giemsa-stained smears. Viability was 99% by trypan blue exclusion test.  2.2.3 T lymphocyte isolation Peripheral blood mononuclear cells (PBMC) were isolated from anti-coagulated peripheral blood of healthy volunteers by centrifugation in Histopaque (Sigma). T lymphocytes were separated by passage of PBMC through a "T cell recovery column" (Cedarlane). Monocytes and B lymphocytes were trapped in the column by size and by 35  the action of anti-immunoglobulin Abs on the column beads, respectively. By this method, greater than 90% of cells eluted were T lymphocytes by fluorescence-activated cell sorting (FACS) analysis and viability was 99% by trypan blue exclusion test. FACS analysis was carried out by incubating lymphocyte suspensions with mouse mAbs against CD 14, CD45, CD2 or CD20 to detect monocytes, all leukocytes, T lymphocytes plus NK cells, and B lymphocytes, respectively. The anti-CD14 and anti-CD2 Abs were conjugated to phycoerythrin (Coulter Corp., Hialeah, FL). The anti-CD45 and anti-CD20 Abs were conjugated to fluorescein isothiocyanate (Becton Dickenson, Mississauga, ON). After a 30min incubation at 4°C, cells were washed with 2% fetal calf serum in TC 199 (Gibco) and fixed in 1% paraformaldehyde. Fluorescence was read on an Epic Profile I (Coulter Corp). The quadrants were set by using isotypic controls and dead cells were gated out.  2.2.4 Antibodies The following Abs were used: Mouse anti-human E-selectin at 20ug/ml (3B7, a mouse IgG2a), and mouse anti-human VCAM-1 at 20^g/ml (2G7, a mouse IgGl) (gifts from Dr. W. Newman). Mouse anti-human ICAM-1 at 20pig/ml (RR1/1, a mouse IgGl, Biosource International / Immunocorp, Montreal, Quebec). Rabbit anti-human PECAM-1 at 50ug/ml (a polyclonal Ab from Dr. S. Albelda). All the above Abs have been shown to bind to EC and to be active in the inhibition of specific leukocyte-EC interactions in our system and other EC systems (Graber et al., 1990; Shimizu et al., 1991; Dustin and Springer, 1988; Liversidge et al., 1990; Muller et al., 1993). Isotype matched control Abs included: mouse anti-membrane co-factor protein at 20ng/ml (J4.48, a mouse IgGl), mouse anti-HLA-A,B,C at 20ug/ml (B9.12.1, a mouse IgG2a), and rabbit gammaglobulin at 50[xg/ml (all from Dimension lab, Mississauga, Ontario). All Abs were free of preservatives and stabilizers. The Ab concentration giving the highest reading by ELISA was taken to be the optimal concentration for the assays. 36  Mouse anti-human leukocyte common antigen-HRP (Dako / Dimension Lab) was used to label leukocytes in the adhesion assay.  2.2.5 Adhesion assay Each adhesion molecule was maximally upregulated on HBMEC by treatment with 100U TNF-a/ml for 4-24h (4h for E-selectin, 18h for VCAM-1, 24h for ICAM-1 and PECAM-1). Monolayers were then incubated with individual blocking Abs for 30min and PMN (4xl0 6 cells/ml) or T lymphocytes (2xl0 6 cells/ml) were added to the wells and incubated with HBMEC for 10,20 or 30 min for PMN and 20, 30 or 60 min for T lymphocytes at 37°C. At the end of the incubation period, the supernatants with the nonadherent leukocytes were removed and the monolayers were gently washed four times, once at each of the lateral walls of the wells. Adherent leukocytes were fixed in 1:1 acetone:ethanol for 7 min at 4°C, and stained with the immunoperoxidase technique for leukocyte common antigen as follows. After a brief wash, endogenous peroxidase was blocked by treatment with 2.5% hydrogen peroxide in 100% methanol. Monolayers were incubated with mouse anti-human leukocyte common antigen-HRP for lh, then with 0.05% 3,3'-diaminobenzidine (DAB) and counterstained with haematoxylin. Controls included untreated EC, EC treated with TNF but without anti-adhesion molecule Abs and EC incubated with isotype matched control Abs. Except as noted above, all cultures were grown, fed, processed and examined with the microscope the same way at the same time. Quantitation. The number of leukocytes bound to the monolayers was determined by counting the number of adherent leukocytes per mm 2 in one central and four peripheral randomly selected fields by light microscopy. The peripheral fields chosen were uniformly one field diameter (1.2mm) from the edge of the well. This negates any effect from the possible gradual change of leukocyte density between the center and the edge of the well. The average number of adherent PMN per field varied from less than 37  one for unstimulated EC to sixty for EC treated with TNF for 24h. The number of adherent T lymphocytes per field varied from eight for unstimulated EC to six hundred for EC treated with TNF for 24h. The identity of the wells was masked before counting so that all counts were performed blindly (Dorovini-Zis et al., 1992a). In addition, the percentage of adherent cells was calculated by dividing the number of cells adherent per well (calculated knowing the area of the well and number of cells adherent per mm^) by the number of cells added to each well and multiplying by one hundred. Monolayers grown on cellagen discs and treated as above were processed for scanning EM (SEM) as previously described (Dorovini-Zis et al., 1991). Briefly, cells were fixed with 2.5% glutaraldehyde in 0.05M sodium cacodylate for lh at 4°C, then in 1% OSO4 for lh and 1% tannic acid for lh. After dehydration through a graded series of ethanol, monolayers were stained with 2% uranyl acetate in 70% ethanol overnight. Subsequently, cells were washed five times in 70% ethanol, dehydrated through 100% ethanol, critical point dried and coated with gold. Specimens were examined with a Cambridge Sterescan 250T SEM.  2.2.6 Migration assay In order to distinguish between the process of adhesion and migration, leukocytes (5xl0 6 PMN/ml or 3xl0 6 T lymphocytes/ml) were allowed to bind to the HBMEC monolayer for 15min (PMN) or 30min (T lymphocytes) in the absence of Abs. This period of time is not long enough for migration to occur. Unbound cells were washed off and then blocking Abs were added. After a further 2h incubation for PMN and 30, 60 or 180min for T lymphocytes, the HBMEC with the migrated leukocytes were gently washed and then processed for EM as described for immunogold EM above. Controls include untreated HBMEC and EC treated with TNF without anti-adhesion molecule Abs. As an additional positive control for the PMN migration experiments, a standard 38  PMN chemotactic factor (10" 'M N-formyl-methionyl-leucyl-phenylalanine, fMLP, Sigma) was placed in the outer chamber under the HBMEC monolayers to establish a chemotactic gradient before PMN were added in the inner chamber over the EC. Except as noted above, all cultures were grown, fed, processed and examined under the microscope the same way at the same time. Quantitation. The number of leukocytes that crossed the monolayers was counted in one hundred and twenty lum thick sections taken 15 urn apart and stained with toluidine blue. The length of the monolayer in each thick section was measured and the number of migrated leukocytes per mm of the monolayer length was calculated by dividing the total number of migrated leukocytes by the monolayer length in each section. The identity of the sections was masked before counting so that all counts were performed blindly. In addition, in order to calculate the percentage of cells that adhered and subsequently migrated, the number of leukocytes present in the wash before the addition of the Abs (pre-Ab wash) and in the wash at the end of the migration period and before fixation (pre-fixation wash) was counted in a haemocytometer. The percentage equals: Cells at start of assay - (Cells in pre-Ab wash + Cells in pre-fixation wash) Cells at start of assay - Cells in pre-Ab wash Selected treatments were repeated for scanning electron microscopy. Processing for SEM was performed as described in the adhesion assay above.  2.2.7 Controls For all experiments, cells grown in the absence of TNF, monolayers incubated with TNF but without Abs and/or EC treated with isotype matched Abs served as controls.  39  2.2.8 Statistics Statistics were performed using the computer program SigmaStat (Jandel Scientific). ANOVA was performed on all data. Where significant difference was found, Student's t-test was applied to examine the results between cultures treated with Abs and those untreated for the adhesion assay. The Mann-Whitney test was applied to examine the results between cultures treated with Abs and those untreated for the migration assay since the data were not normally distributed. P<0.05 was taken as statistically significant for all analyses.  2.3 DETERMINATION OF MONOLAYER PERMEABILITY FOLLOWING TNF-a TREATMENT AND IN THE COURSE OF LEUKOCYTE MIGRATION ACROSS THE MONOLAYERS  2.3.1 Permeability assay The permeability of HBMEC cultures was determined by measuring the electrical resistance across confluent monolayers using a volt-ohm meter (World Precision Instruments, Sarasota, FL). HBMEC grown to confluence on Cellagen discs were transferred from the double chamber chemotaxis system to a chamber with an electrode firmly positioned in the center of the chamber. Another electrode was placed on top of the monolayers and resistance was measured. Resistance was measured at each media change and the experiments were performed when a high resistance (>90Q-cm^) was reached. To assess the effect due to TNF treatment, HBMEC were treated with 100U TNF/ml for 4,18 or 24h. The resistance was measured in these cultures as well as in unstimulated cultures. The effect of PMN and T lymphocyte migration was examined following leukocyte incubation with untreated or TNF treated HBMEC for 4, 18 and 24h and PMN migration in response to gradients of 10"'M fMLP. Resistance was measured at the end of TNF treatment prior to leukocyte addition, then again at the end of the 40  migration period. In addition, HBMEC treated with TNF as above were incubated with anti-adhesion molecule Ab before addition of leukocytes. Again, resistance was measured prior to leukocyte addition and then after leukocyte migration. Since the resistance is inversely proportional to the area of the tissue, it is standard practice to report the product of the resistance and the growth area so that comparisons with results in the literature can be made. This value is then independent of the area of the tissue. In this study, the resistance was multiplied by the growth area of the Cellagen disc (0.64cm^) before reporting. Except as noted above, all cultures were grown, fed and measured with the ohm-meter the same way at the same time. Controls included cultures grown without TNF, with TNF but without Abs and EC without the addition of PMN or T lymphocytes.  2.3.2 Controls For all experiments, cells grown in the absence of TNF, monolayers incubated with TNF but without Abs and EC without the addition of leukocytes served as controls.  2.3.3 Statistics Statistics were performed using the computer program SigmaStat (Jandel Scientific). ANOVA, followed by Student's t-test where there was significant difference were applied to the results from transendothelial resistance measurements. T-tests compared cultures activated with TNF to those unstimulated, and cultures treated with Ab to those untreated. P<0.05 was taken as statistically significant.  41  CHAPTER 3 RESULTS  3.1 HUMAN BRAIN MICROVESSEL ENDOTHELIAL CELL CULTURE (HBMEC) HBMEC in primary culture form confluent, contact-inhibited monolayers made up of elongated cells. There are no differences in cell growth or morphology between HBMEC grown on fibronectin coated plastic wells (Fig. la) and Cellagen discs (Fig. lb). The endothelial origin of the cells was demonstrated by the positive granular, perinuclear staining for Factor VIII related antigen (Fig. 2).  3.2 EXPRESSION OF ADHESION MOLECULES BYHBMEC IN PRIMARY CULTURE AND UPREGULATIONBYLPSAND CYTOKINES  3.2.1 E-selectin  3.2.1.1 Immunogold silver staining (IGSS) Both the unstimulated and stimulated expression of E-selectin were determined by counting the number of cells with positive and negative surface staining following immunogold silver staining. Only 7% of unstimulated cells were stained for E-selectin (Figs. 4-8). There was minimal variation in the unstimulated expression among different cultures. Cells expressing E-selectin displayed diffuse staining over the cell surface (Fig. 4a). Treatment with LPS, TNF-a or IL-ip generated a great rise in the percentage of E-selectin expressing cells (P<0.001 by AN OVA for all three treatments). The staining was evenly distributed on the cell surface in the form of fine black, granular deposits. Overall, the intensity of labeling was greater in treated than in unstimulated cells (Fig. 4). No staining was observed in control cultures incubated with normal mouse IgG, irrelevant Ab or buffer instead of the primary Ab. 42  LPS induced the greatest upregulation of E-selectin expression (Figs. 4b, 5a). The level was highest after 4h of treatment for all concentrations, then dropped to a level not statistically different from the unstimulated level by 24h for the lower concentrations or 48h for the higher concentrations in the continuous presence of LPS. Treatment with the three highest concentrations of LPS (0.01,0.1 and ljAg/ml) for 4h induced E-selectin expression in about 90% of cells, while a 80% expression was observed with O.OOlug LPS/ml. The subsequent drop in the number of immunostained cells to unstimulated levels by 48h was similar for the 0.1 and lug/ml concentrations. The decline to pre-stimulation levels occurred earlier with O.Olug LPS/ml and even more so with 0.001 ng/ml. TNF-a was similarly effective in upregulating E-selectin (Fig. 4c). 91% of the cells showed positive staining after 4h of treatment with the highest concentration lOOU/ml (Fig. 6a). Lower concentrations also induced a maximal rise at 4h, which decreased to very low levels by 48h. Following the initial rise in expression with 100U TNF-a/ml, E-selectin levels decreased considerably after 48h to levels still statistically different from the unstimulated level. 10U TNF-a/ml increased E-selectin expression to a lesser extent, while 1U TNF-a/ml induced even lower levels of E-selectin which actually became statistically no different from unstimulated levels by 48h. Upregulation of E-selectin following IL-ip treatment was less impressive than that with LPS and TNF-a (Fig. 4d). The expression was maximal by 4h then decreased to unstimulated levels by 24 to 48h, similar to the other treatments (Fig. 7a). However, even with the highest concentration (10U IL-lp7ml), only 58% of cells became E-selectin positive by 4h. 1U IL-ip/ml induced levels slightly below those of lOU/ml, while the expression with O.lU/ml was much lower than with the other two concentrations. Amongst all the treatments applied, IFN-y was the only one that was not effective in raising the level of E-selectin expression even at the high concentration of 500U/ml (P=0.548 by ANOVA) (Fig. 8a). 43  3.2.1.2 Enzyme-linked immunosorbent assay (ELISA) The overall expression of E-selectin in confluent HBMEC monolayers was quantitated by ELISA. The results are similar to the counts obtained by immunogold silver staining. No increase in absorbance was seen in control wells incubated with buffer or normal mouse IgG instead of primary Ab. A low level of E-selectin expression was observed with unstimulated HBMEC (Figs. 5b-8b). LPS, TNF-a and IL-lp all significantly upregulated E-selectin (P=0.014, <0.033 and <0.001, respectively by ANOVA). The treatment that generated the greatest E-selectin upregulation was, again, incubation with LPS (Fig. 5b). For all concentrations used, there was a great increase in E-selectin expression by 4h, the maximal levels being quite similar for all concentrations. With 0.01 and O.OOljxg LPS/ml, the E-selectin levels rose even slightly higher at 12h. There was a rapid drop by 24h, and by 48h, there was no statistically significant difference from the unstimulated levels at all concentrations (Fig. 5b). TNF-a, at all concentrations tested, induced a maximal E-selectin expression at 4h (Fig. 6b). This returned promptly to unstimulated levels by 24h for 1U TNF-a/ml, and 48h for 10 and 100U TNF-a/ml. The effectiveness of IL-ip was intermediate, while the pattern of E-selectin upregulation was similar to that of LPS and TNF-a (Fig. 7b). There was maximal expression at 4h, which then declined to levels not statistically different from the unstimulated ones by 12h for 0.1U IL-lp7ml, 24h for 1U IL-lp/ml, and 48h for 10U IL-ip/ml. E-selectin expression was not affected by IFN-y even at high concentrations (500U/ml, P=0.333 by ANOVA) (Fig. 8b).  3.2.1.3 Immunogold electron microscopy (IGEM) Ultrastructural localization of E-selectin was performed by immunogold labeling following maximal upregulation with TNF-a. Staining was present on the surface of 44  both unstimulated and TNF treated HBMEC (Figs. 9 and 10). A significantly greater amount of labeling was present on the apical surface compared to the basal surface of all EC. TNF treatment dramatically increased the number of gold particles on the apical surface and, to a lesser but still significant extent, on the basal surface of HBMEC (P<0.001) (Fig. 10). Gold particles were absent along intercellular contacts and at tight junctions. Finger-like cytoplasmic processes on the apical surface were more prominently decorated (Fig. 9c) than adjacent flat surfaces. A small number of gold particles was observed infrequently within pits /vesicles some of which appeared to be clathrin-coated invaginations of the apical cell membrane in TNF treated cells and rarely in unstimulated cells. No labeling was observed on cells incubated with buffer without primary Ab.  3.2.2 Vascular cell adhesion molecule-1 (VCAM-1)  3.2.2.1 Immunogold silver staining (IGSS) Control HBMEC monolayers cultivated in the absence of LPS and cytokines exhibited minimal expression of VCAM-1 (13-18%) (Figs. 11a, 12a-15a). The degree of VCAM-1 expression in unstimulated cells was constant in a given culture, although varying slightly between cultures originating from different individuals. Growth in the presence of LPS, TNF-a or IL-lp induced marked upregulation of VCAM-1 expression (P=0.008, <0.001, =0.022, respectively, by ANOVA) as indicated by the large number of VCAM-1 positive cells (Fig. llb-d). The intensity of labeling varied among individual cells of the same culture, the largest cells generally being more darkly stained. Cells expressing VCAM-1 exhibited distinct, finely granular, brown-black staining diffusely throughout the cell surface (Fig. llb-d). Unstimulated EC showed less intense surface staining with a tendency for localization at the periphery of the cell (Fig. 11a). Control  45  wells incubated with normal mouse IgG or buffer without primary Ab showed no labeling. The greatest upregulation of VCAM-1 was observed following treatment with LPS. Incubation with lug/ml was associated with an increase to 50% by 4h (Fig. 12a) and a maximal level of 83% by 12h that remained at a plateau for up to 24h and then declined to a level not statistically different from the unstimulated level by 48h. 0.1^g/ml elicited a similar pattern of VCAM-1 upregulation, though the levels of expression were slightly lower. The increase in VCAM-1 induced by 0.01ng/ml was similar up to 12h. However, VCAM-1 levels started to fall by 24h and by 48h, they were no longer statistically different from the unstimulated levels. With 0.001 u,g/ml, there was significant upregulation to 45% by 12h followed by a decline close to unstimulated levels by 24h. TNF-a caused the second highest level of VCAM-1 expression. Treatment with lOOU/ml induced a maximal VCAM-1 level of 74% at 24h, which declined to 53% by 48h (Fig. 13a). lOU/ml generated a similar increase up to 12h but the level then declined to 48% and plateaued. A modest increase was noted with lU/ml at 12h and 24h with levels remaining steady thereafter. The extent of VCAM-1 upregulation was relatively less impressive following treatment of the monolayers with IL-lp (Fig. 14a). lOU/ml raised the VCAM-1 level to 33% by 4h, and 60% by 12h. Longer periods of incubation resulted in declines to 49% and 36% at 24h and 48h, respectively. lU/ml was not associated with any statistically significant difference in VCAM-1 levels at 4h. The level increased dramatically to 60% by 12h and then declined at a much faster rate than in cultures treated with lOU/ml. Cultures exposed to O.lU/ml expressed VCAM-1 levels that paralleled those induced by lU/ml except that the levels of expression were lower. Coincubation with the lowest concentrations of TNF-a and IL-ip for 48h failed to upregulate VCAM-1 (P=0.812 by  46  ANOVA) (Fig. 16a). IFN-y did not increase VCAM-1 expression to any great extent even at the high concentration of 500U/ml (P=0.1 by ANOVA) (Fig. 15a).  3.2.2.2 Enzyme-linked immunosorbent assay (ELISA) Quantitation by ELISA generated results similar to those obtained by immunogold silver staining. The relative amount of VCAM-1 expressed was measured by this method. Control cultures incubated with normal mouse IgG or buffer only and no primary Ab showed no change in absorbance. Significant increases in expression were induced by LPS (P=0.02 by ANOVA), TNF (P=0.014) and IL-1 (P=0.013). LPS induced the highest levels of VCAM-1 expression, peaking at 12h at all concentrations tested and declining to a level not statistically different from the unstimulated expression (Fig. 12b). 1 and 0.1ng/ml elicited the greatest upregulation while 0.01 and O.OOljig/ml resulted in successively modest increase. TNF-a was second in effectiveness to LPS. Upregulation of VCAM-1 expression was maximal at 12h (Fig. 13b), followed by a less steep decline, resulting in higher levels of VCAM-1 expression at the end. IL-ip was less effective than TNF-a in upregulating VCAM-1 (Fig. 14b). Maximal expression was observed by 12 to 24h followed by a decline. The increase in expression caused by O.lU/ml was not statistically significant and did not peak until 24h. Treatment with IFN-Y was not associated with any significant changes in VCAM-1 expression (P=0.198 by ANOVA) (Fig. 15b).  3.2.2.3 Immunogold electron microscopy (IGEM) Immunogold electron microscopy revealed that VCAM-1 was readily detectable on the surface of the EC. Labeling usually occurred in clumps of 3 or more gold particles. Both unstimulated and TNF-a treated cells showed a statistically significant (p<0.05) increase in labeling on the apical surface compared to the basal surface (Figs. 17-18). Labeling was not observed in coated pits, along interendothelial contacts or at 47  tight junctions (Fig. 17d). There was a tendency for gold particles to localize on or near finger-like cytoplasmic projections rather than flat apical surfaces especially on TNF-a treated cells (Fig. 17c). Compared to unstimulated EC, cells incubated with TNF-a exhibited several folds more labeling on both the apical and basal surfaces (p<0.05), the apical surface being consistently more densely decorated with gold particles than the basal surface. Control cultures incubated with buffer instead of primary Ab were unlabeled.  3.2.3 Intercellular adhesion molecule-l (ICAM-1)  3.2.3.1 Immunogold silver staining (IGSS) In control cultures, 20-40% of EC expressed ICAM-1 (Figs. 19-23). Treatment of HBMEC with LPS, IFN-y, TNF-a, and IL-lp led to significant upregulation of ICAM-1 expression (P=0.001, =0.027, =0.002, <0.001, respectively by ANOVA). Positively stained cells displayed diffuse surface staining in the form of finely granular, dark-brown to black deposits (Fig. 19). There was some variability in the intensity of labeling among individual EC, the plumper cells being decorated with the most dense reaction product. The overall intensity of staining was higher in experimental than in control cultures (Fig. 19). LPS was the most effective inducer of ICAM-1 upregulation (Figs. 19b, 20). 5ug LPS/ml resulted in expression of ICAM-1 by 95% of the cells within 12h (p<0.05), and this level remained constant up to 72h in the continuous presence of LPS. The high level of ICAM-1 induced by lug and 0.1 ug LPS/ml seemed to decline slightly after 24h. O.Olug LPS/ml produced an intermediate rise in ICAM-1 which dropped slowly after 12h. A statistically significant (p<0.05) rise in ICAM-1 (40%) was not observed with 0.00lug LPS/ml until after 48h. In contrast to the large increase in ICAM-1 expression observed with LPS, treatment with IFN-y induced only a slight increase (Figs. 19c, 21), with maximal levels 48  of expression reaching only 58%. By 12h, a statistically significant level of upregulation to 47% was observed with all concentrations (p<0.05), but at longer incubation times, the relatively larger variance made any increases in expression not to be statistically significant. The extent of ICAM-1 upregulation by both TNF-a (Figs. 19d, 22) and IL-10 (Figs. 19e, 23) was intermediate between that of LPS and IFN-y. By 24h, 85% of cells showed positive staining (p<0.05), after which there was a general trend of slight decline. The increased staining induced by 10U and 100U TNF-a/ml was statistically significant. The increase caused by 1U TNF-a/ml was not statistically significant until 12h, and after incubation for 72h, the increase declined to non-significant levels. The upregulation of ICAM-1 elicited by IL-lp reached maximal levels at 24h (84% for 10U, 76% for 1U, and 67% for 0.1U, all p<0.05), and then declined. Incubation of EC with TNF-a and IFN-y combined resulted in marked increase in ICAM-1 expression that was even greater than that induced by LPS (Figs. 19f, 24). The number of immunostained cells ranged from 99 to 100% with 500U IFN-y/ml plus 1U to 100U TNF-a/ml and reached 97% with 100U IFN-y/ml plus 100U TNF-a/ml. No labeling was observed in cultures incubated with normal mouse IgG or buffer instead of primary Ab.  3.2.3.2 Reversibility of ICAM-1 expression Stimulation of HBMEC with LPS or cytokines for 2 days followed by withdrawal and culture of cells in regular growth media for 3 days was associated with variable degrees of reversal of ICAM-1 upregulation following LPS and TNF-a incubation, and lack of reversal following IL-lp and IFN-y treatment (none statistically significant). These differences were due to cytokine and LPS removal since cultures incubated with cytokine or LPS for 5 days expressed high levels of ICAM-1 (Figs. 25-28). There was a decrease (10-30%) in the number of cells expressing ICAM-1 at all concentrations of 49  LPS (Fig. 25). Reversibility of expression was minimal when 5jig/ml LPS was used. Reversibility of upregulation was not observed with IFN-y at lOOU/ml, while 500U/ml was associated with a 19% increase instead (Fig. 26). Activation followed by withdrawal of TNF-a resulted in decrease in ICAM-1 expression ranging from 6% to 23%. TNF-a concentration of lOOU/ml was associated with the smallest decrease while lU/ml induced the greatest (Fig. 27). Paradoxically, incubation of EC in IL-ip-free media for 3 days following activation resulted in increase in the number of labeled cells ranging from a low of 1% with 10U IL-lp/ml to 17% with O.lU/ml (Fig. 28).  3.2.3.3 Immunogold electron microscopy (IGEM) ICAM-1 was expressed on the surface of both unstimulated and TNF treated HBMEC. There was significantly more labeling on the apical surface compared to the basal surface (Fig. 29 and 30). Indeed, gold particles were rarely observed on the basal surface. Labeling of cytoplasmic pits and vesicles and intercellular contacts was not observed. Gold particles were especially concentrated on finger-like cytoplasmic projections on the apical cell surface. TNF treatment greatly increased the apical and to a lesser extent the basal expression of ICAM-1. Control cultures incubated without primary Ab were unlabeled.  3.2.4 Platelet/endothelial cell adhesion molecule-1 (PECAM-1)  3.2.4.1 Enzyme-linked immunosorbent assay (ELISA) PECAM-1 was constitutively expressed by HBMEC (Figs. 31-35). Treatment with 0.01-lng LPS/ml (Fig. 32), 1-100 U TNF-a/ml (Fig. 33), 0.1-10 U IL-lp7ml (Fig. 34) or 50-500 U IFN-y/ml (Fig. 35) for 4 to 48h did not increase its expression quantitatively (P=0.063, 0.091, 0.723, 0.211, respectively by ANOVA) or change its distribution (Figs. 32, 33). In both unstimulated and activated EC, PECAM-1 staining 50  appeared as fine black grains mostly concentrated at the borders between adjacent cells but also distributed diffusely, though less abundantly, over the cell surface (Fig. 31).  3.2.4.2 Immunogold electron microscopy (IGEM) The presence of PECAM-1 was readily detectable on the surface of EC in the form of gold particles occurring singly or in small groups. Labeling was most prominent at the apical surface of HBMEC and, unlike ICAM-1, VCAM-1 and E-selectin, gold particles did not have the tendency to concentrate on finger-like projections at the apical cell surface (Figs. 36-38). There was no apparent difference in labeling between untreated and IFN-y treated cells. Although gold particles were not observed along intercellular contacts, the most peripheral processes of the cytoplasm particularly in the vicinity of cell to cell contact, were constantly decorated with gold particles.  3.2.5 HECA-452 antigen The HECA-452 antigen was not expressed by unstimulated HBMEC and could not be induced de novo in cells treated with 5ug LPS/ml, 10-100 U TNF-cx/ml, 1-10 U IL-lp/ml or 100-400 U IFN-v/ml alone or in combination with IL-lp for 1, 2, 3, 4, 6, or 7 days. Positive staining of high endothelial venule cells in control frozen sections of tonsils confirmed that this Ab was functional.  3.3 ROLE OF ADHESION MOLECULES IN PMNAND T LYMPHOCYTE ADHESION TO AND MIGRATION ACROSS HUMAN BRAIN MICROVESSEL ENDOTHELIAL CELL MONOLAYERS  51  3.3.1 Adhesion of polymorphonuclear leukocytes to human brain microvessel endothelial cells A small number of PMN (approximately 0.06%) adhered to unstimulated HBMEC. Pretreatment of HBMEC with TNF-a for 4,18 or 24h progressively increased the adhesion of PMN to EC (P<0.001 by ANOVA). This increase was directly related to the length of TNF treatment and the duration of EC-PMN incubation (Figs. 39-41). Thus, 4h of TNF treatment increased adhesion to 0.1-0.2%, 18h to 0.2-0.4% and 24h to 0.4-0.7%. Similarly, the number of adherent PMN increased from 11/mm^ after 10 min incubation to 34/mmr after 30 min incubation with EC (Fig. 41). None of the anti-adhesion molecule mAbs decreased PMN adhesion to unstimulated HBMEC (P>0.124 by ANOVA) (Figs. 42-44). Adhesion to TNF stimulated HBMEC was inhibited by Abs (P^O.001 by ANOVA). After 10, 20 and 30 min incubation of PMN with TNF-treated HBMEC, the anti-E-selectin mAb significantly blocked adhesion by 54, 40 and 38%, respectively. This level of adhesion is not significantly different from that obtained with unstimulated HBMEC. Anti-ICAM-1 mAb blocked adhesion by 37% at lOmin, 47% at 20min and 45% at 30min (Figs. 42-44). The anti-PECAM-1 mAb decreased adhesion by 10 to 30%, but this was not large enough to be statistically significant. The anti-VCAM-1 mAb did not significantly decrease adhesion. Isotype matched control Abs did not have any effect. A preliminary experiment with the F(ab')2 fragments of the anti-VCAM-1 and anti-E-selectin Abs showed the same results as experiments with intact Abs. It appears that the Fc receptor on peripheral blood PMN do not bind IgG strongly. Thus, all subsequent experiments were performed with intact Abs. Scanning electron microscopy (SEM) showed that only a few PMN adhered to unstimulated EC (Fig. 39a). Pretreatment of the monolayers with TNF markedly increased the number of adherent PMN (Fig. 39b). PMN first adhered by extending pseudopodia to the surface of HBMEC while still retaining their spherical shape. They  52  tight junctional complexes between HBMEC overlying the migrated PMN appeared intact.  3.33 T lymphocyte adhesion A small number of T lymphocytes (2%) adhered to unstimulated HBMEC (Figs. 51-53). Pretreatment of EC with TNF-a significantly upregulated adhesion of T lymphocytes following 20,30 and 60 min incubation with HBMEC (16-23%, P<0.001 by ANOVA) (Figs. 51-53). Ab treatment decreased adhesion to unstimulated and activated EC (P<0.005 by ANOVA). The anti-ICAM-1 mAb was the only one that significantly blocked adhesion to unstimulated HBMEC by 60, 55 and 63% after 20, 30 and 60 min of T lymphocyte incubation, respectively (Figs. 54-56). However, when HBMEC were pretreated with TNF, both anti-VCAM-1 and anti-ICAM-1 mAbs significantly blocked adhesion: anti-VCAM-1 mAb by 32, 33 and 29%, and anti-ICAM-1 mAb by 47, 57 and 75% at 20, 30 and 60 min of T lymphocyte incubation, respectively (Figs. 54-56). The anti-E-selectin and anti-PECAM-1 Abs did not block adhesion (Figs. 54-56). Paradoxically, the anti-PECAM-1 Ab increased adhesion (20-100%) especially at 60 min (100% increase, Figs. 54-56). A 90% increase in adhesion was also observed when anti-PECAM-1 Ab was applied to unstimulated HBMEC prior to incubation with T lymphocytes for 60min (Fig. 56). Isotype matched control Abs did not have any effect. SEM showed that only few T lymphocytes adhered to untreated EC (Fig. 51a). HBMEC treatment with TNF resulted in significant increase in the number of adherent lymphocytes (Fig. 51b-d). Adherent T lymphocytes appeared as small, roughly spherical cells that established contact with EC by extending long pseudopodia to their apical surface (Figs. 51, 57-61). They adhered to the surface of HBMEC (Fig. 62) and also aligned themselves singly or in groups along adjacent EC (Fig. 51).  54  3.3.4 T lymphocyte migration Very few T lymphocytes migrated across unstimulated HBMEC even after 180min of incubation with EC (Figs. 63-64). Pretreatment of HBMEC with TNF significantly increased migration only after prolonged (180min) EC-lymphocyte incubation (P<0.001 by ANOVA) (Figs. 63-64). Thus, a 30 fold increase in transmigration was observed when T lymphocytes were incubated for 180min with EC pretreated with TNF for 24h, while only a modest 7 and 8 fold increase was recorded for the same incubation time but with shorter TNF pretreatment (4h and 18h, respectively). Anti-adhesion molecule Abs had significant effect on migration at prolonged (180min, P<;0.002 by ANOVA) but not short (30 or 60min, P>0.08 by ANOVA) incubations (Figs. 65-66). Thus, anti-PECAM-1 Ab blocked migration by 75% and anti-ICAM-1 mAb by 93% (Figs. 67). In fact, the anti-ICAM-1 mAb brought migration across activated HBMEC down to levels obtained with unstimulated HBMEC. Anti-E-selectin mAb blocked migration by 40%, while the 20% decrease with anti-VCAM-1 is not statistically significant. Examination by SEM showed that, following their initial adhesion to the apical surface of the endothelium (Fig. 68a), lymphocytes proceeded with crossing the monolayers by moving between adjacent EC or by direct penetration of the EC cytoplasm (Fig. 68b, c). In the initial stages of transcellular migration, T lymphocytes were seen within deep invaginations of the HBMEC surface and appeared to fit tightly the size of these invaginations (Fig. 68c). Examination by TEM showed that following the initial firm contact between pseudopodia of lymphocytes and the apical surface of the endothelium (adhesion), lymphocytes began inserting foot-like processes into the endothelial cytoplasm. One or more such areas of penetration by the same lymphocyte could be seen in one EC (Fig. 69-71). These foot-like processes of the lymphocytes were in tight contact with the EC cytoplasm. Occasionally, the most advancing tip was preceded by a variably sized cytoplasmic "opening" or gap (Figs. 69-70), which 55  progressively receded and enlarged to allow for the remaining cytoplasm and nucleus of the lymphocyte to go through the EC into the subendothelial region (Fig. 72). Once across the monolayers, lymphocytes became elongated and remained between the underlying collagen membrane and the overlying EC while maintaining close contact with the latter (Fig. 73). At the end of the migration period, the monolayers resumed their continuity and the junctional complexes between adjacent EC appeared intact.  3.4 DETERMINATION OF MONOLAYER PERMEABILITY FOLLOWING TNF-a TREATMENT AND IN THE COURSE OF LEUKOCYTE MIGRATION ACROSS THE MONOLAYERS Lastly, we examined the effect of TNF treatment, leukocyte migration, and anti-adhesion molecule Abs (that block adhesion and/or migration) on the barrier properties of HBMEC monolayers by measuring the electrical resistance across monolayers treated with different combinations of the above. A drop in resistance indicated an increase in permeability, thus a decrease in barrier function. Unstimulated HBMEC grown on Cellagen discs had an average resistance of 108Q-cm^ (Table 1). TNF treatment progressively decreased the resistance to 68Q-cm2 by 18h, then further down to 63Q-cm2 by 24h (P<0.001) (Table 1). An additional drop in resistance was observed following PMN migration across the monolayers. This was especially evident when fMLP was placed in the lower chamber to attract PMN to migrate (44Q-cm2) compared to monolayers without chemotactic gradient (99Q-cmZl). PMN migration across HBMEC pretreated with TNF for 4 and 18h decreased the resistance by 25% while migration across HBMEC preincubated with TNF for 24h decreased it by 43%. The presence of anti-adhesion molecule Abs did not prevent the drop in resistance following PMN migration (P=0.11 by AN OVA) (Table 2). T lymphocyte migration across 4,18 or 24h TNF pretreated HBMEC also decreased resistance by an average of about 25% (P=0.002). This drop in resistance was not prevented by the presence of 56  anti-adhesion molecule Ab (duplicate wells of one experiment, P=0.082 by ANOVA) (Table 3).  CHAPTER 4 DISCUSSION 4.1 EXPRESSION OF ADHESION MOLECULES 4.1.1 E-selectin In the present study, we found that 7% of untreated HBMEC express E-selectin. This expression can be upregulated on HBMEC by LPS, TNF-a and IL-lp\ Positively stained cells display an intense black granular reaction product evenly distributed throughout the cell surface. The maximal staining intensity is much greater in HBMEC than in HUVEC where the reaction product is concentrated mainly at the peri-nuclear region (Varani et al., 1992). E-selectin expression is maximal 4-6h after LPS or cytokine treatment of HBMEC, HDMEC (Sepp et al., 1994), HUVEC (Bevilacqua et al., 1989; Luscinskas et al., 1989; Pober et al., 1986a, 1986b; Leeuwenberg et al., 1989; Leeuwenberg et al., 1990a, 1990b; Bevilacqua et al., 1987; Graber et al., 1990; Wellicome et al., 1990; Doukas et al., 1990) and mouse endothelioma cells (Hahne et al., 1993). The concentrations required to generate a maximal expression of E-selectin on HUVEC are mostly higher than those for HBMEC (Wellicome et al., 1990; Luscinskas et al., 1989). IFN-y, at a concentration that upregulates ICAM-1, is not effective in either cell type (Graber et al., 1990). After maximal expression, E-selectin declines to unstimulated levels by 48h in all in vitro systems examined thus far including HBMEC, HDMEC (Sepp et al., 1994), HUVEC (Bevilacqua et al., 1989; Luscinskas et al., 1989; Pober et al., 1986a, 1986b; Leeuwenberg et al., 1989,1990a, 1990b; Bevilacqua et al., 1987; Graber et al., 1990; Wellicome et al., 1990; Doukas and Pober, 1990) and mouse endothelioma cells (Hahne et al., 1993). Such declines in staining after 48h does not seem to be due to cytotoxicity on HBMEC since no cell retraction or necrosis is seen under light microscopy, SEM or TEM after 48h of LPS or cytokine treatment. Indeed, EC are still able to maintain their 58  high ICAM-1 expression at this time. However, a factor recently found in human serum causes E-selectin and its mRNA to be persistently expressed, at least up to 48h (Sepp et al., 1994). The effectiveness of treatments that increase E-selectin expression is different between HBMEC and HUVEC. For HBMEC, LPS and TNF-a are most effective followed by IL-1J3, while for HUVEC, the order is reversed (Wellicome et al., 1990). Summary of E-selectin expression on EC HBMEC  HUVEC  HDMEC  7%  None  -  expression  4h  4-6h  4-6h  Concentration  0.1 ng LPS/ml  1 ug LPS/ml  -  giving maximal  10 U IL-ip/ml  10 U IL-l/ml  -  expression  100U TNF-a/ml  320U TNF/ml  -  Level post  Declines to  Declines to  Declines to  maximal increase  unstimulated  unstimulated  unstimulated  levels by 48h.  levels by 48h.  levels by 48h.  Unstimulated Expression (% cells labelled) Time of maximal  Relative effectiveness  LPS/TNF-a>IL-lp\ IL-lpVTNF/LPS  of cytokines/LPS  IFN-y not effective.  -  It is unlikely that difference in staining found are due to different culture conditions between the published reports and the present studies since reports from different laboratories with different growth conditions are quite consistent.  59  The greater unstimulated level of E-selectin expression and greater sensitivity to cytokine and LPS suggest that E-selectin may be utilized to a greater extent in the brain than extracerebral organs. It is used in both PMN adhesion and T lymphocyte migration across HBMEC. Although E-selectin expression declines to unstimulated levels by 48h, indicating its use is short lived, if a human serum factor could maintain a high level of E-selectin on HBMEC, then it could be used in the long term. If not, then there is also the possibility that after a refractory period, E-selectin may be induced on these EC again and HBMEC in areas around the center of the inflammatory site may become progressively activated and express E-selectin stepwise. The greater expression of E-selectin on the apical surface of HBMEC observed ultrastructurally may have functional significance regarding leukocyte adhesion to EC. Prominent expression at the finger-like projections of EC may facilitate initial interactions between leukocytes and EC at these points of contact during the "rolling" of leukocytes. The upregulation of the low level of E-selectin expression at the basal surface by TNF may point to participation of this molecule in leukocyte migration across the endothelium. The use of subconfluent monolayers ensured that the Abs still reach the basal EC surface from the edge of the colony or between non-contacting EC. E-selectin was observed within cytoplasmic vesicles or pits in HBMEC, although infrequently. This presence may permit adherent leukocyte pseudopods to begin penetration of the EC for transmigration. We have shown that E-selectin acts in T lymphocyte migration, perhaps in this way. The presence of E-selectin in cytoplasmic vesicles has been reported for HUVEC (von Asmuth et al., 1992; Kuijpers et al., 1994). It was suggested that this is a consequence of endocytosis of these receptors, in order to remove E-selectin as well as free forms of its ligand, L-selectin, that have been shed into the circulation with eventual degradation in the lysosomal compartment. E-selectin labeling has also been localized in multivesicular bodies - a member of this compartment (von Asmuth et al., 1992). 60  E-selectin, a member of the selectin family, is expressed on LPS or cytokine treated EC (Bevilacqua and Nelson, 1993). Molecular biologic analysis demonstrated that interaction of TNF with the well characterized 55KDa TNF receptor on HUVEC can induce the expression of E-selectin (Throp et al., 1992). The induction by TNF, IL-1 or LPS involves the PKC-independent activation of NF-KB binding to the E-selectin enhancer/promoter sequence in HUVEC (Montgomery et al., 1991; Whelan et al., 1991; Read et al., 1994). Protein synthesis inhibitors can augment this N F - K B binding activity and increase the mRNA level by stabilization (Ghersa et al., 1992). E-selectin expression can be conferred to nonmyeloid cell lines by transfection of a human fucosyltransferase cDNA. Thus, this enzyme may regulate adhesion of E-selectin by modulating the cell surface expression of sialyl-Lewis x (Lowe et al., 1990). The expression of E-selectin itself is inhibited by the binding of glucocorticoid to its receptor, giving rise to a decreased level of E-selectin mRNA eventually (Cronstein et al., 1992). Cyclic AMP has been suggested to be an inhibitor but not a physiological regulator (Pober et al., 1993). In vitro, E-selectin, is endocytozed by HUVEC during the induction of its expression (von Asmuth et al., 1992). However, soluble E-selectin has also been shown to be released by activated HUVEC (Leeuwenberg et al., 1992), and soluble E-selectin is found in the serum of patients with septic shock (Newman et al., 1993). Only one report has dealt with the expression of E-selectin in CNS disorders in vivo (Dorovini-Zis et al., 1992b). The findings include the demonstration of E-selectin on capillaries, arterioles and venules in all cases of encephalitis, meningitis and brain abscess studied. Focal staining was seen in cases of infarcts. Neither larger arteries and veins, nor endothelium of normal brain or traumatic lesions stained. A positive correlation between E-selectin expression and the degree of inflammation was observed. Studies on the expression of E-selectin in cultured EC monolayers have been performed mainly on HUVEC and human dermal microvessel EC (HDMEC). Unstimulated HUVEC do not express E-selectin or express it at minimal levels (Varani et 61  al., 1992; Bevilacqua et al., 1989; Luscinskas et al., 1989; Pober et al., 1986a, 1986b; Leeuwenberg et al., 1989; Leeuwenberg et al., 1990b; Bevilacqua et al., 1987; Graber et al., 1990; Wellicome et al., 1990; Doukas and Pober, 1990). There has only been one brief report with a mouse endothelioma cell line established from polyoma-induced hemangioma in the brain (Hahne et al., 1993) and one report dealing with induction by endothelin on human brain microvascular EC (McCarron et al., 1993c). Our results indicate that activation of HBMEC with LPS, TNF and IL-lp, but not IFN-Y,  greatly upregulate the expression of E-selectin on HBMEC. Maximal  upregulation is reached by 4h, followed by a rapid decline to unstimulated levels by 48h in the continuous presence of LPS or cytokines. Comparing our data with those reported on HUVEC by other investigators, it appears that HBMEC are more sensitive to these treatments than HUVEC with respect to the upregulation of E-selectin expression. E-selectin is largely expressed on the apical surface of HBMEC. TNF treatment increases its level of expression by many folds, especially at finger-like cytoplasmic projections. E-selectin is also occasionally present within cytoplasmic vesicles, but absent at tight junctions.  4.1.2 Vascular cell adhesion molecule-1 (VCAM-1) The present study demonstrates that HBMEC in primary culture express low levels of VCAM-1 that can be upregulated by activation with LPS, TNF-a and IL-ip in a time and concentration dependent manner. Previous in vitro studies have demonstrated a lack or minimal constitutive expression of VCAM-1 on passaged HUVEC and HDMEC (Wellicome et al., 1990; Swerlick et al., 1992; Graber et al., 1990). In the CNS, VCAM-1 has not been observed in neurons, glia or EC in the absence of inflammation (Rice et al., 1991), however Abs against its ligand, VLA-4, have been shown to inhibit lymphocytes and monocytes from binding to cryostat sections of rat brain with  62  experimental allergic encephalomyelitis (EAE) and prevent the development of EAE in vivo (Yednock et al., 1992). Similarly, VCAM-1 expression was not observed in microvessel preparations isolated from normal human brains at autopsy (Washington et al., 1994), but could be upregulated on human brain microvascular EC by endothelin (McCarron et al., 1993c). Previous studies on VCAM-1 induction on HUVEC have demonstrated that VCAM-1 mRNA is already present 2h following treatment with IL-1 or TNF-a, peaks at 4h and is maintained for at least 72h (Swerlick et al., 1992; Osborn et al., 1989). Surface expression of VCAM-1 peaks from 4 to 8h after cytokine treatment (Rice and Bevilacqua, 1989; Wellicome et al., 1990; Swerlick et al., 1992; Carlos et al., 1990), then either stays constant with TNF and LPS (Wellicome et al., 1990; Swerlick et al., 1992) or declines gradually to unstimulated or near unstimulated levels after 48h to 72h (Rice and Bevilacqua, 1989; Wellicome et al., 1990; Swerlick et al., 1992). Activation of HDMEC with TNF-a induces early increase in VCAM-1 mRNA after 4h followed by a decline to undetectable levels by 24h (Swerlick et al., 1992). Surface expression rises 16h after treatment with cytokines and declines rapidly by 24h. On a mouse endothelioma cell lines, VCAM-1 becomes maximally expressed 16h after TNF treatment, then drops slightly by 24h (Hahne et al., 1993). In contrast, surface expression of VCAM-1 on HBMEC becomes maximal 12-24h after cytokine or LPS treatment and drops back to nearly unstimulated levels by 48h.  63  Summary of VCAM-1 expression on EC HBMEC  HUVEC  HDMEC  15%  Low  None  expression  12h  4-8h  16h  Concentration  0.1 ng LPS/ml  0.1-1 ng LPS/ml  0.01 jxg LPS/ml  giving maximal  10 U IL-lp/ml  10 U IL-l/ml  -  expression  100U TNF-a/ml  500U TNF-a/ml  500U TNF/ml  Level post  Declines to  Plateaus or  Drops fast by  maximal increase  unstimulated  drops slowly.  24h, then  levels by 48h.  Does not reach  slowly, but not  unstimulated  to  levels.  unstimulated  Unstimulated Expression (% cells labelled) Time of maximal  levels by 72h. Relative  LPS>TNF-a  TNF>IL-lp>LPS  effectiveness  >IL-lp  IL-1 not  of cytokines/LPS  IFN-y not effective.  effective.  TNF>LPS  Although surface expression of VCAM-1 was induced by LPS, TNF-a and IL-lp\ upregulation was maximal with LPS, followed by TNF-a and IL-ip in that order. Regulation of VCAM-1 expression by cytokines on HBMEC is different from that previously reported for HUVEC (Swerlick et al., 1992; Graber et al., 1990) and HDMEC (Swerlick et al., 1992). IFN-y, at a concentration that upregulates ICAM-1, failed to  64  upregulate VCAM-1 on HBMEC. This is in contrast to dermal microvessel EC which show significant upregulation of VCAM-1 following intradermal injection of IFN-y (Groves et al., 1993). The concentrations of LPS, TNF-a and IL-1 required for induction of maximal VCAM-1 expression on HBMEC are comparable to those inducing maximal upregulation on HUVEC by some investigators (Wellicome et al., 1990) but not by others (Swerlick et al., 1992). In contrast, much higher concentrations of TNF-a and lower concentrations of LPS are required for maximal induction of VCAM-1 on HDMEC (Swerlick et al., 1992). Activated HUVEC release VCAM-1 in a soluble form (Pigott et al., 1992), but VCAM-1 is not found in endocytotic vesicles in EC (Kuijpers et al., 1994). These differences in the kinetics of VCAM-1 upregulation in the three in vitro systems may either reflect diverse culture and experimental conditions among different laboratories, or, possibly indicate inherent differences in cytokine responsiveness between EC derived from different vascular beds. The present findings correlate with previous in vitro immunocytochemical studies demonstrating preferential localization of VCAM-1 on the apical surface of IL-1 activated HUVEC, while expression was less prominent on the basal cell surface and absent from interendothelial contacts (Oppenheimer-Marks, 1991). The polarized expression of VCAM-1 may be of functional significance considering cerebral EC-leukocyte interactions. The fact that VCAM-1 is present predominantly on the apical surface of activated HBMEC and not along intercellular contacts may indicate a possible role of VCAM-1 primarily in EC-leukocyte adhesion rather than in transendothelial migration. Thus, VCAM-1 / VLA-4 binding could augment the adhesion of leukocytes to HBMEC prior to their subsequent transendothelial migration, since the increase in labeling is greater on the apical than the basal surface. The high concentration at or near thin cytoplasmic processes, the supposed points of contact with leukocytes, could facilitate adhesion. The minimal expression on the basal cell surface and its upregulation  65  by TNF-a suggest a possible utilization of VCAM-1 at the final steps of the migration process or during post-migration events. In the CNS, VCAM-1 has been reported in AIDS encephalitis induced in macaques by the Simian Immunodeficiency virus (Sasseveille et al., 1992), and, recently, in human cerebral microvessels isolated from postmortem brains of patients with MS (Washington et al., 1994). Recently, the mechanism of VCAM-1 expression on glial cells has been linked to the transcription factor  NF-KB  (Meynagh et al., 1994; Shu et al., 1993; Neish et al., 1992)  and PKC activation (Mattila et al., 1992; Deisher et al., 1993). Cyclic AMP does not appear to be a physiological regulator but rather an inhibitor of VCAM-1 expression (Pober et al., 1993). The present studies demonstrate that unstimulated HBMEC express low levels of VCAM-1 which are significantly increased following activation with LPS, TNF-a or IL-ip, but not IFN-y and return to unstimulated levels after 48h in the continuous presence of LPS or cytokines. Ultrastructurally, VCAM-1 is localized mostly on the apical surface of EC and is not present along intercellular contacts or at junctional complexes. The induced upregulation of VCAM-1 is associated with increased expression mostly at the apical surface.  4.1.3 Intercellular adhesion molecule-l (ICAM-1) Intact, unstimulated EC constitutively express low levels of ICAM-1. Incubation with LPS, TNF-a, IL-ip and to a lesser extent IFN-y, results in upregulation of ICAM-1 expression which persists in the continuous presence of mediators in the culture. The pattern of ICAM-1 expression in inflammatory disorders of the nervous system was recently investigated in guinea pigs and mice during the acute and chronic phases of EAE. O'Neill et al. (1991) observed the upregulation of MALA-2 (the murine 66  homologue of ICAM-1) on EC and mononuclear infiltrates in spinal cord lesions during the active phase of chronic relapsing EAE in the Biozzi AB/H mouse. Recent studies on adoptively transferred EAE in the SJL/J mouse have shown that attachment and extravasation of mostly LFA-1 positive lymphocytes correlated with the appearance of MALA-2 on CNS vessels and the onset of clinical signs (Raine et al., 1990b). In fact, the expression of MALA-2 fluctuated according to the clinical phase of EAE, with upregulation occurring with each relapse and down-regulation during remissions (Cannella et al., 1991). In both studies, endothelial and glial cells of normal mice showed low levels of ICAM-1. Similarly, the normal spinal cord vasculature of guinea pigs has only a few ICAM-1 positive EC, while ICAM-1 was present on the endothelium of both lesion- and non-lesion-associated blood vessels of animals during the acute phase of EAE (Wilcox et al., 1990). In the same study, cultured EC isolated from guinea pig brains did not express ICAM-1 unless incubated in a lymphocyte-conditioned medium. The above animal studies suggest that upregulation of ICAM-1 may be important in the initial influx of lymphocytes into the brain in autoimmune demyelinating CNS disorders. The role of ICAM-1 in orchestrating leukocyte traffic across the BBB in the course of inflammatory diseases of the human CNS has not yet been fully investigated. In a recent neuropathologic report, based on the examination of post-mortem material from normal and pathological brain tissue, only a small number of microvessels were ICAM-1 positive within the normal CNS. In contrast, numerous ICAM-1 positive vessels were observed in MS plaque edges, viral encephalitis lesions and infarcts (Sobel et al., 1990). In vitro, ICAM-1 is released in a soluble form by activated HUVEC (Leeuwenberg et al., 1992; Pigott et al., 1992) but not endocytozed during the induction of its expression (von Asmuth et al., 1992; Kuijpers et al., 1994).  67  Summary of ICAM-1 expression on EC HBMEC  HUVEC  HDMEC  20-40%  <11%  11-19%  expression  12-24h  lOh or 24h  24h-6d  Amount of maximal  95% (LPS)  -  -  expression  58% (IFN-Y)  -  93% (IFN-Y)  (% cells labelled)  85% (IL-lp)  -  -  85% (TNF-a)  -  100% (TNF-a)  99% (TNF+IFNy)  -  100%  Concentration  1 fig LPS/ml  0.1 ug LPS/ml  giving maximal  1-500U IFN-Y/ml  -  10 3 U IFNY/ml  expression  10 U IL-lp/ml  >10 U IL-l/ml  -  1-100U TNF-a/ml  100U TNF-a/ml  10 4 UTNF/ml  500U IFNY/ml +  1000U IFN-Y/ml + 100U IFNY/ml  100U TNF-a/ml  100U TNF-a/ml  +10 3 UTNF/ml  Level post  Plateaus for  Plateaus for  Plateaus for  maximal increase  at least 72h.  at least 24h.  at least 24-72h.  Relative  LPS  TNF/IL-1>LPS  TNF/IL-1>LPS  effectiveness  >TNF-a/IL-lp  of cytokines/LPS  >IFN-Y  Unstimulated Expression (% cells labelled) Time of maximal  (TNF+IFNY)  >IFN-Y  68  The upregulation of ICAM-1 induced by LPS and TNF-a on HBMEC is very drastic even by 4h, and is maximal by 12h, while IL-ip did not induce a maximal level until 24h. IFN-y led to a slight but statistically significant increase in expression. In HUVEC and HDMEC, LPS induced a rise in ICAM-1 that did not peak until 24h (Swerlick et al., 1991; Wellicome et al., 1990). Swerlick et al. (1991) and Pober et al. (1986b) observed a similar rise of ICAM-1 in HUVEC and HDMEC following treatment with TNF-a and IL-lp, in contrast to Wellicome et al. (1990) who showed that TNF and IL-1 induced a peak in ICAM-1 expression on HUVEC by lOh. The slight upregulation of ICAM-1 observed after treatment of HBMEC with IFN-y is in keeping with previous reports indicating that IFN-y is a weak inducer of ICAM-1 upregulation on extracerebral endothelium. Thus, IFN-y did not induce a noticeable change in ICAM-1 levels in HDMEC until 72h (Swerlick et al., 1991), and caused a "slower" increase than IL-1 and TNF in HUVEC (Pober et al., 1986b). It is evident that the highest levels of ICAM-1 induced by LPS, TNF-a and IFN-y appear earlier in HBMEC than in HDMEC and some HUVEC cultures. The time course of ICAM-1 expression generated by IL-1J3 is similar in all three culture systems. The optimal concentration of cytokines and LPS required for the upregulation of ICAM-1 is different in cerebral and extracerebral endothelial cultures. Thus, a higher concentration of LPS (Wellicome et al., 1990) and lower concentration of TNF-a (Detmer et al., 1990; Pober et al., 1987) and IFN-y (Ruszczak et al., 1990) are needed for HBMEC compared to HUVEC. Human retinal capillary EC, on the other hand, showed maximal induction of ICAM-1 upregulation with the low concentration of 5U IFN-y/ml (Liversidge et al., 1990). IL-1 was required in a concentration greater than lOU/ml for maximal effect on HUVEC (Pober et al., 1986b; Wellicome et al., 1990), while lOU/ml induced highest levels on HBMEC. HBMEC are thus more sensitive to cytokines and less sensitive to LPS than HUVEC and HDMEC with respect to ICAM-1 upregulation.  69  The staining intensity of HBMEC varied even within monolayers derived from a single isolation. As a rule, the labeling of larger EC was stronger than that of actively proliferating cells. Such heterogeneity in ICAM-1 expression has also been reported in cultured human saphenous vein EC and HUVEC (Dustin and Springer, 1988) and in cultured fibroblasts (Dustin et al., 1986). The unstimulated expression of ICAM-1 in HBMEC is higher than that in human retinal capillary EC (Liversidge et al., 1990) and in HDMEC (Ruszczak et al., 1990), which in turn are higher than the expression in HUVEC (Swerlick et al., 1991). The maximal levels of ICAM-1 expression on HBMEC and HDMEC (Detmer et al., 1990; Ruszczak et al., 1990) are also different. HBMEC express much lower levels of ICAM-1 than HDMEC when activated with LPS or cytokines. Hess et al. (1994) demonstrated by ELISA that ICAM-1 upregulation by LPS and cytokines is much greater on HBMEC than HUVEC. Following incubation of HBMEC with a combination of TNF-a and IFN-y, 99 to 100% of the cells became labeled for ICAM-1. This combined effect has also been observed in HDMEC (Detmer et al., 1990) and HUVEC (Doukas and Pober, 1990). In the latter study, it was also found that IFN-y does not significantly enhance the expression of ICAM-1 induced by IL-lp\ It has also been shown that endothelin and cytokines upregulate ICAM-1 on human brain microvascular EC (McCarron et al., 1993c). TNF, IL-1 and IFN increase ICAM-1 expression on murine cerebrovascular EC (McCarron et al., 1993b, 1993c). Shear stress has also been shown to upregulate ICAM-1 expression on HUVEC (Nagel et al., 1994). Other cell types within the brain respond much differently from EC to cytokine activation. ICAM-1 expression on oligodendrocytes and astrocytes was markedly increased by IFN-y, while TNF-a, IL-la and LPS were less effective (Satoh et al., 1991). One study with a mouse endothelioma cell line reported that ICAM-1 was only very weakly upregulated in these cells (Hahne et al., 1993).  70  We found that removal of cytokines and LPS from the culture media was associated with variable expression of ICAM-1 after 3 days. Thus, there seemed to be a slight decrease in ICAM-1 with LPS (10-30%) and TNF-ct (6-23%), and a slight increase with  IFN-Y  (0-19%) and IL-lp4 (1-17%). Dustin and Springer (1988) observed that  ICAM-1 expression by HUVEC increased following stimulation with IL-1 and TNF, but returned to unstimulated levels when the mediators were washed off. It would seem that HBMEC maintain their ICAM-1 expression much longer than HUVEC. Ultrastructurally, ICAM-1 is present on the luminal but not abluminal surface of high endothelial venules. High concentrations are found on the "microvilli" of these EC (Brown et al., 1993; Sasaki et al., 1993). Similarly, ICAM-1 is expressed mainly on the apical surface of HBMEC, especially on finger-like projections. TNF treatment significantly upregulates the expression of this molecule, however, little labeling is observed on the basal surface while intercellular contacts remain unlabeled. This may suggest a role of ICAM-1 in the firm adhesion and perhaps initial migration of leukocytes across these monolayers but little utilization once the barrier is crossed. In contrast to HBMEC and high endothelial venules, ICAM-1 is prominent on the apical and basal surfaces of HUVEC, with high staining at the apical surface. It is also present at the tight junctions and at contact sites between T lymphocytes and the apical and basal surfaces of HUVEC (Oppenheimer-Marks et al., 1991). Thus, ICAM-1 may function in adhesion and migration of leukocytes across extracerebral EC, but its role in the HBMEC may be more limited. The mechanism(s) of ICAM-1 induction by cytokines is not well defined. Lane et al. (1990) and Sung et al. (1994) concluded that IL-1, TNF-a and LPS activate HUVEC expression of ICAM-1 in part by activation of protein kinase C (PKC), since PKC inhibitors decrease subsequent ICAM-1 upregulation by these agents. Myers et al. (1992a, 1992b) observed that ICAM-1 upregulation by phorbol myristate acetate, but not by cytokines or LPS, involved PKC activation. On the other hand, Ritchie et al. (1991) 71  found that TNF does not activate HUVEC through the PKC pathway (Ritchie et al., 1991) and PKC activation is capable of but not essential for ICAM-1 induction by cytokines and LPS (Myers et al., 1992a, 1992b). Renkonen et al. (1990a) found that IL-1 operates via cAMP but not via PKC, and IFN-y functions through PKC and calcium dependent pathways. In a subsequent paper Renkonen et al. (1990b) showed that IFN-y induced a rapid activation of phospholipase C, leading to a breakdown of phosphoinositol diphosphate into diacylglycerol (DAG) and inositol triphosphate (IP3). DAG is a natural activator of the PKC pathway. PKC inhibitors abolished the IFN-y effect, while PKC stimulators mimicked the effects of IFN-y. Elevations of IP3 levels led to activation of the calcium-dependent pathway. A calmodulin inhibitor decreased the IFN-y induced ICAM-1 expression. Recently, the transcription factor  NF-KB  has also been suggested to  be involved in ICAM-1 expression by glial cells (Maynagh et al., 1994). Our results indicate that HBMEC, compared to other EC, tend to express a higher unstimulated level of ICAM-1, are more sensitive to cytokines, and less sensitive to LPS, express ICAM-1 earlier but at a lower level when induced, and fail to reverse its expression within a short time upon removal of these agents. ICAM-1 is expressed mainly on the apical surface of HBMEC and on finger-like projections, and is greatly upregulated by TNF. Taken together, the relatively high level of ICAM-1 expression on unstimulated HBMEC and its fast maximal upregulation with relatively low concentrations of cytokines, to levels lower than those obtained in other EC in vitro systems, indicate different kinetics of ICAM expression and upregulation in HBMEC which may be relevant to unique responses of these cells in CNS inflammation.  4.1.4 Platelet/endothelial cell adhesion molecule-l (PECAM-1) PECAM-1 is especially concentrated at the EC cell borders where adjacent cells contact each other and tight junctions are formed. Cytokines do not alter PECAM-1 expression on HUVEC but it was suggested that they induce its redistribution from the 72  cell border evenly throughout the cell surface (Romer et al., 1991). The presence of this molecule at the cell borders has led to speculations that it may be involved in EC-EC interactions and EC-leukocyte interactions after cytokine activation of EC (Albelda et al., 1990b; Romer et al., 1991). The constitutive expression of PECAM-1 and lack of change after EC activation on HBMEC are consistent with the reports on HUVEC. Interestingly, PECAM-1 was not redistributed on HBMEC by LPS or cytokines, unlike the reported changes in HUVEC. This suggests it may act mainly in the migration of leukocytes instead of adhesion, unlike HUVEC. Consistent with the observations of Leach et al. (1993) on human placental microvessels, we found that, on HBMEC, PECAM-1 is present mainly on the apical surface. Labeling was prominent at the thin cytoplasmic areas close to the tight junctions but not along intercellular contacts or in junctions themselves. This may be due either to the absence of this molecule from the junctions or restriction of the Ab from penetrating the tight junctions. Staining after the junctions are opened by hyperosmotic solutions or post-embedding staining would ensure that the Ab can reach all areas of the tight junctions. Using HUVEC exposed to buffer without divalent ions to retract EC, Muller et al. (1989) found that PECAM-1 staining is prominent in the intercellular junctions between adjacent EC. Of interest is the observation that in subconfluent HBMEC cultures where EC are separated by a short space, the EC surfaces that would eventually contact the adjacent EC are heavily labeled. It seems likely that leukocytes can especially use PECAM-1 in these areas to migrate across the endothelium.  4.1.5 HECA-452 antigen Horst et al. (1988) reported that 10-100 U IFN-y/ml induced HECA-452 antigen expression on over 80% of HUVEC. Labeling was apparent after 6h, became maximal after 24h and remained high after 48h. Only cytoplasmic staining was observed, unlike the in vivo situation where membranes were also labeled. In contrast, we have not been 73  able to induce HECA-452 antigen on HBMEC with any of the cytokines including IFN-y at 400U/ml for 1 to 7 days. Positive staining of high endothelial venule cells on control frozen sections of tonsils confirmed that this Ab is functional. Thus far, only one Ab, a mouse IgM, has been produced against this antigen and has been applied in all studies mentioned. In a strange twist of events, HECA-452 antigen, also known as cutaneous lymphocyte antigen, has been found to be the receptor for E-selectin on EC and may be a lymphocyte homing receptor for skin homing lymphocytes (Bos et al., 1993; Berg et al., 1991). 80-90% of T lymphocytes in inflammatory skin diseases are HECA-452+ and bind to E-selectin+ EC. Abs against either molecule can decrease adhesion (DeBoer et al., 1994). Although it is possible that HBMEC, being phenotypically distinct from EC of other vascular beds, do not express HECA-452 antigen, it is also possible that other treatments, not examined here, may induce this molecule. In addition, different epitopes of this antigen or different vascular addressins, not recognized by the present Ab, may be expressed by HBMEC. In any case, the expression of this antigen is controlled differently between HBMEC and extracerebral EC.  The current dogma states that leukocytes roll on EC using P- and E-selectin. This initial adhesion is followed by firm adhesion via ICAM-1 and VCAM-1. Migration follows using ICAM-1 and sometimes, VCAM-1. In this study, it was found that maximal E-selectin expression on HBMEC occurs at 4h, followed by a quick decrease. VCAM-1 and ICAM-1 are not maximally expressed until 12 to 24h. However, between 12 and 24h, all three molecules are expressed to a certain extent. Moreover, even at 4h, ICAM-1 and VCAM-1 levels have begun to rise quite rapidly. There may be a sufficient amount for both adhesion and migration. In addition, ICAM-2 and PECAM-1 are both constitutively expressed and may be utilized even when ICAM-1 and VCAM-1 are not yet upregulated.  74  Summary of E-selectin, VCAM-1 and ICAM-1 expression on HBMEC ICAM-1  VCAM-1  E-selectin  (% cells labelled)  20-40%  15%  7%  Location  Border  Border  Surface  expression  12h-24h  12h-24h  4h  Amount of maximal  95% (LPS)  83% (LPS)  90% (LPS)  expression  85% (IL-ip)  60% (IL-1P)  58% (IL-ip)  (% cells labelled)  85% (TNF-a)  62% (TNF-a)  91% (TNF-a)  Concentration for  lug LPS/ml  O.lug LPS/ml  O.lug LPS/ml  maximal expression  10U IL-ip/ml  10U IL-ip/ml  1U IL-ip/ml  100U TNF-a/ml  100U TNF-a/ml  100U TNF-a/ml  Relative effectiveness  LPS>  LPS>TNF-a>  LPS/TNF-a>  of cytokines/LPS  TNF-a/IL-lp>  IL-lp.  IL-lp>  IFN-Y  IFN-y -no effect.  IFN-y -no effect.  Level post  Plateaus at  Drops to unstimulated  maximal increase  high level.  levels by 48h.  Unstimulated expression  Time of maximal  Thus, these results lead to the possibility that early in the course of inflammation, initial rolling on E-selectin is followed by firm adhesion using the moderate amount of VCAM-1 and ICAM-1 upregulated and ICAM-2 that is constitutively expressed. Some time later, rolling may be decreased due to the lower level of E-selectin expression, but  the presence of leukocytes already firmly attached to the surface and / or transmigrating across the endothelium may allow these leukocytes to interact with nonadherent leukocytes to form aggregates, which may also arrest them in the blood stream, ready for firm adhesion and transmigration. It is also possible that, in the human body, some substance in the serum may cause EC to maintain a high expression of E-selectin once activated. The close proximity of VCAM-1 and ICAM-1 on the finger-like projections to E-selectin, which is also prominently localized in these structures, facilitates the proposed sequence of events from rolling to firm adhesion. PECAM-1 may not be involved in these events since it is concentrated at the cell borders. This strategic location could reflect its possible role in transmigration. The presence of other adhesion molecules in these areas also suggest their role in this movement across the endothelium. The lack of adhesion molecule expression on the basal surface of HBMEC may imply that leukocytes do not cross back from the brain to the blood, or use other molecules for this purpose.  4.2 ROLE OF ADHESION MOLECULES IN PMN AND T LYMPHOCYTE ADHESION TO AND MIGRATION ACROSS HUMAN BRAIN MICROVESSEL ENDOTHELIAL CELL MONOLAYERS  4.2.1 PMN adhesion Adhesion of PMN to intact, as well as cytokine-treated HBMEC is very low. Only 0.04-0.07% of PMN adhered to unstimulated HBMEC and 0.34-0.7% to activated HBMEC. PMN adhesion to unstimulated HUVEC is considerably higher - 1 to 5% in most reports (Wankowicz et al., 1988; Morzycki et al., 1990; Bevilacqua et al., 1985; Bochner et al., 1991), up to 12% (Hakkert et al., 1991) or 15% (Smith et al., 1988) in others. Cytokine activation of HUVEC raises the adhesion rate to 10-30% in most reports (Wankowicz et al., 1988; Morzycki et al., 1990; Bevilacqua et al., 1985; Bochner 76  90% decrease. Other investigators using different mAbs reported a 40 to 90% blocking of PMN adhesion to activated HUVEC (Kuijpers et al., 1991; Kishimoto et al., 1991; Benjamin et al., 1991; Leeuwenberg et al., 1990a; Luscinskas et al., 1989; Shimizu et al., 1991b; Bochner et al., 1991; Benjamin et al., 1990). There is a general consensus that E-selectin is involved in PMN adhesion. This is also supported by the work of Hahne et al. (1993) who observed a 20 to 30% decrease in PMN adhesion to an activated mouse brain endothelioma cell line with an anti-E-selectin Ab. Ab blocking of PMN adhesion across EC monolayers (%) Untreated  Activated  Untreated  Activated  Ab  HBMEC  HBMEC  HUVEC  HUVEC  anti-E-selectin Ab  0%  38-54%  -  40-90%  anti-VCAM-1 Ab  0%  0-20%  -  3-5%  anti-ICAM-1 Ab  0-10%  37-45%  70%  30-60%  anti-PECAM-1 Ab  0%  10-40%  -  -  Adhesion of PMN to activated HBMEC appears to be mediated, in addition, by ICAM-1, since it was significantly blocked (37 to 45%) by the anti-ICAM-1 mAb. Using the same anti-ICAM-1 Ab on cytokine activated HUVEC, Smith et al. (1988) found a 50% inhibition of adhesion. Other studies employing different Abs have reported a 30 to 60% decrease in PMN adhesion (Luscinskas et al., 1991; Bochner et al., 1991; Smith et al., 1988,1989). PECAM-1 does not appear to have any apparent role in PMN adhesion to activated HBMEC, since Abs against this adhesion molecule did not affect PMN adhesion. At present, there are no other studies on the role of PECAM-1 in PMN adhesion to endothelium, although one study found that neutrophils, blocked from migration by anti-PECAM-1 Abs, remain attached to the EC surface. This is suggestive  78  of a lack of utilization in adhesion but no quantitation or statistical comparison to untreated EC was performed (Muller et al., 1993). There have only been two reports on the role of VCAM-1 in PMN adhesion to activated EC. Activated HUVEC were used in both studies and both found that anti-VCAM-1 Abs had no effect on adhesion (Barber et al., 1990; Bochner et al., 1991). Our results with HBMEC are consistent with these findings. It is generally believed that PMN cannot utilize the VCAM-1 adhesion pathway because their expression of VLA-4 (the ligand for VCAM-1) in very low (Lund-Johansen and Terstappen, 1993). Although E-selectin has been proposed to function under conditions of flow, during ischemia and inflammation stasis occurs in the microvascular bed, therefore, the findings using the present model would be physiologically relevant. Examination of the role of these adhesion molecules under flow conditions is beyond the scope of this study. Our results show that TNF treatment upregulates PMN adhesion. None of the anti-adhesion molecule Abs studied block PMN adhesion to unstimulated HBMEC. Both E-selectin and ICAM-1 appear to mediate PMN adhesion to activated HBMEC, however, none of the other molecules can be entirely ruled out, since epitopes on these molecules, not blocked by the Abs used in the present experiments, may conceivably operate in these interactions.  4.2.2 PMN migration Although many published reports tend not to distinguish between adhesion and migration of leukocytes in the design of experiments, these are two separate processes. Adhesion is followed by migration, but adherent cells do not necessarily migrate. Moreover, different sets of adhesion molecules are employed. Thus, in these studies, equal amounts of adhesion was permitted in all wells and nonadherent cells aspirated before the addition of Abs to assess migration.  79  Migration of PMN across unstimulated HBMEC monolayers was minimal (0.2 PMN/mm length of monolayer). Treatment of HBMEC with 100U TNF/ml significantly upregulated migration. Following 2h incubation of PMN with HBMEC, migration across 4h TNF pretreated monolayers increased by 4 fold (0.9 PMN/mm), while 12h pretreatment led to 12 fold increase (2.5 PMN/mm) and 24h pretreatment induced a 15 fold increase (3.1 PMN/mm). However, when the number of adherent PMN is taken into account, a large percentage of adherent cells migrated across both unstimulated HBMEC (79 to 88%) and activated HBMEC (81 to 94%). Our findings are consistent with previous reports in which a 3 fold increase in PMN migration across HUVEC treated with TNF for l-2h was observed (Moser et al., 1989; Morzycki et al., 1990). Longer periods of TNF treatment (4h) increased PMN migration across HUVEC by 16 fold (Furie et al., 1989). Treatment of HUVEC with IL-lp for 4h similarly led to a 3 fold increase in PMN transmigration (Hakkert et al., 1991; Luscinskas et al., 1991). In an in vivo study, intracisternal, but not intravenous, administration of 800 U TNF-cc into the brain of rats, raised the number of leukocytes in the CSF maximally ten fold after 4h of treatment (Kim et al., 1992). The present study shows that none of the anti-adhesion molecule Abs inhibited PMN migration across resting HBMEC. No other studies have been reported on this subject using in vitro EC systems, though Furie et al. (1991) observed that PMN migration across unstimulated HUVEC in response to a chemotactic gradient is blocked by 55% by an anti-ICAM-1 Ab. It should be clarified that the majority of assays used to examine transendothelial migration actually involve both adhesion to as well as migration across EC. The anti-adhesion molecule Abs are added to EC monolayers first, then the PMN or other leukocytes are added and allowed to migrate. This way, the Abs can inhibit both adhesion and migration, so that the two processes cannot be clearly separated. Results interpreted as inhibition of migration may actually have been overinterpreted. 80  Ab blocking of PM N migration across EC mono ayers (%) Untreated  Activated  Untreated  Activated  Ab  HBMEC  HBMEC  HUVEC  HUVEC  anti-E-selectin Ab  33%  0%  -  0-90%  anti-VCAM-1 Ab  16%  4%  -  -  anti-ICAM-1 Ab  33%  75%  55%  47-97%  anti-PECAM-1 Ab  17%  15%  -  80%  The anti-E-selectin, anti-VCAM-1 and anti-PECAM-1 Abs had no effect on the migration of PMN across activated HBMEC. Two studies on activated HUVEC found a similar lack of effect of an anti-E-selectin Ab (Kishimoto et al., 1991; Hakkert et al., 1991), in contrast to the findings of Luscinskas et al. (1991), who, using the same mAb as in the present study, observed a 90% decrease in migration. However, considering the fact that the migration assay was performed in such a way that the mAbs actually inhibited both adhesion and migration, the degree of inhibition may not be so high. Thus far, there have been no reports on the role of VCAM-1 on PMN migration across activated EC, as it is generally believed that PMN do not employ the VCAM-l/VLA-4 pathway, since they do not bear the VLA-4 molecule (Lund-Johansen and Terstappen, 1993). Only one study is available on the effect of anti-PECAM-1 Abs on PMN migration. Muller et al. (1993), using the same mAb as the one in the present study, found a 80% blocking of PMN migration and similar results were obtained with soluble PECAM. A different anti-PECAM-1 Ab inhibited migration only by 40%. Here again, the assay involved inhibition of both adhesion and migration since the blocking Ab was added to HUVEC cultures at the same time with the PMN. The distribution of PECAM-1 on HBMEC is different from HUVEC as discussed in a previous section.  81  HBMEC PECAM-1 is not as concentrated at the periphery, where adjacent cells contact, as in HUVEC. Moreover, unlike HUVEC, the distribution of PECAM-1 does not change with EC activation. It is notable that "PECAM-1 is expressed in high density on all early myeloid cells but is down-regulated during postproliferative maturation" (Lund-Johansen and Terstappen, 1993). It has also been reported that activation of granulocytes leads to further downregulation of PECAM-1 expression (Stockinger et al., 1990). Taken together, the above observations suggest another reason for the lack of use of the PECAM-1 pathway in PMN adhesion and migration due to the small amount of PECAM-1 expressed by peripheral blood PMN. The anti-ICAM-1 Ab was the only one that significantly blocked PMN migration across activated HBMEC (75% decrease). A similarly significant blocking effect has also been observed with activated HUVEC (97% by Luscinskas et al., 1991 and 70% by Kishimoto et al., 1991). These figures probably include inhibition of adhesion also, as discussed above. In most chemotaxis experiments, a chemotactic gradient is used for induction of PMN transmigration. In the present study, a considerable number of PMN migrated across TNF treated HBMEC in the absence of a chemotactic gradient. Previous studies indicate that immediately upon contact with activated EC, PMN become activated, as measured by increases in intracellular calcium levels (Kuijpers et al., 1991). Possibly as a result of contact with chemoattractants bound to the EC surface. Contrary to these observations, Adams and Shaw (1994) reported that integrins on circulating PMN do not adhere well to the endothelium unless they are activated. It is likely, therefore, that, in our studies, only a subset of possibly activated PMN was able to adhere to cytokine stimulated HBMEC. It has been reported that IL-1 or TNF stimulate HUVEC generate the chemoattractants platelet-activating factor (PAF) and IL-8 (Kuijpers et al., 1992; Sticherling et al., 1993). Anti-IL-8 Abs inhibit migration (Smart et al., 1994), while simultaneous blocking of both pathways with antagonists and Abs can completely 82  prevent migration (Kuijpers et al., 1992). IL-8 and PAF present on the EC surface can transiently activate PMN integrins, especially Mac-1, thus promoting PMN adhesion (Rot, 1992; Huber et al., 1991; Tanaka et al., 1993). The increase in surface expression of integrins is not involved but rather a change in the conformation of integrins resulting in the "expression" of "activation epitopes" (Springer, 1990). It has been proposed that this dynamic regulation of integrin adhesiveness permits deadhesion and retraction of the trailing edge of migrating PMN (Springer, 1994). Binding of PMN to E-selectin also activates the activity of Mac-1 (Lo et al., 1991). Whether these mechanisms are functional in the present culture system and whether the adherent PMN were activated or not is not known since we did not conduct studies to address this issue. The scanning electron micrographs in the present study showed that PMN migrate between HBMEC through the tight junctions. Early studies indicated that all or most inflammatory cells migrate across the endothelium paracellularly, that is between EC and through the tight junctions. This included rat mesentery EC (Marchesi and Florey, 1960; Marchesi, 1961), rabbit ear EC (Florey and Grant, 1961), pig EC (Beesley et al., 1979) and cultured HUVEC (Moser et al., 1989). The present findings are in agreement with the work of Dorovini-Zis et al. (1992a) who showed by SEM and TEM that PMN migration across primary cultures of bovine brain microvessel EC occurs paracellularly. However, other studies suggest that not all PMN cross the endothelial barrier paracellularly. Some authors believe that PMN migration occurs transcellularly through the cytoplasm of cat brain EC (Faustman and Dermietzel, 1985), human, cat and rat lymph nodes (Azzali et al., 1990a) and other human and animal EC (Hammersen et al., 1988). The findings in this study show that none of the adhesion molecules studied regulate migration of PMN across unstimulated HBMEC. ICAM-1 mediates PMN migration across TNF-treated HBMEC.  83  4.2.3 T lymphocyte adhesion The discussion here will be restricted to peripheral blood resting T lymphocytes. The percentage of T lymphocytes that adhere to HBMEC monolayers is higher than that of PMN. 0.64 to 3% T lymphocytes adhered to unstimulated HBMEC, while 16 to 23% adhered to activated HBMEC. These percentages are slightly lower than those reported on mouse brain EC where 30 and 50% of T lymphocytes adhered to unstimulated or activated, respectively, murine cerebrovascular EC (McCarron et al., 1993b, 1993c). 4 to 10% of lymphocytes bind unstimulated, while 6 to 30% of lymphocytes adhere to activated rat brain EC (Hughes et al., 1988; Male et al., 1990b). In contrast, 50% of CD4 + T lymphocytes adhere to activated or unstimulated human retinal capillary EC (Liversidge et al., 1990). Adhesion to HUVEC is similarly high with 5-35% of T lymphocytes binding to unstimulated EC and 20-50% to activated HUVEC (Bereta et al., 1993; Yu et al., 1986; Cavender et al., 1986; Shimizu et al., 1991a; Yu et al., 1985). Treatment of HBMEC with TNF increased T lymphocyte adhesion by 300 to 450%. Longer treatments (18 and 24h) caused a slightly greater increase than shorter (4h) incubations, however, there is no actual difference in the degree of T lymphocyte adhesion between 18 and 24h treatments. Activation of rat and mouse brain EC with TNF, IL-1 or IFN-y produced much less dramatic results with 20 to 100% (Hughes et al., 1988; Male et al., 1990b) and 20 to 80% increases in lymphocyte adhesion (McCarron et al., 1993b, 1993b), respectively. Activation of bovine brain EC increases lymphocyte adhesion by 500% (de Vries et al., 1994), which is more comparable to HBMEC. Surprisingly, activation of human retinal capillary EC does not increase CD4 + T lymphocyte adhesion. The effect of HUVEC activation on T lymphocyte adhesion varies among different laboratories, ranging from only 4 to 28% increase after lh of cytokine treatment (Yu et al., 1986), to 100-400% increase after 3-6h (Bereta et al., 1993; Cavender et al., 1986; Shimizu et al., 1991a, Rice et al, 1990) and down to 60-100% after 24h of activation (Yu et al., 1985). 84  Ab blocking of T cell adhesion across EC monolayers (%) Untreated  Activated  Untreated  Activated  Ab  HBMEC  HBMEC  HUVEC  HUVEC  anti-E-selectin Ab  0%  0-10%  -  20-40%  anti-VCAM-1 Ab  0-30%  29-32%  0-9%  30-60%  anti-ICAM-1 Ab  55-63%  47-75%  50-62%  8-70%  anti-PECAM-1 Ab  0%  0%  -  -  With the exception of anti-ICAM-1, which blocked adhesion by 55 to 63%, none of the other anti-adhesion molecule Abs significantly decreased T lymphocyte adhesion to unstimulated HBMEC. These results are in keeping with three other studies in which the use of the same anti-ICAM-1 Ab decreased T lymphocyte adhesion to unstimulated HUVEC by 50 to 62% (Oppenheimer-Marks et al., 1991; Liversidge et al., 1990; Oppenheimer-Marks and Lipsky, 1993). Oppenheimer-Marks and Lipsky (1993) also found that an anti-VCAM-1 Ab had no effect on T lymphocyte adhesion to unstimulated HUVEC. Similarly, Bereta et al. (1993) found that an anti-VCAM-1 Ab had no effect on the adhesion of splenocytes to unstimulated murine cerebrovascular EC. T lymphocyte adhesion to TNF-treated HBMEC was not blocked by the anti-E-selectin mAb. Applying the same mAb, Graber et al. (1990) and Shimizu et al. (1991b) found a 20 to 40% inhibition of the adhesion of T lymphocytes to activated HUVEC. Analysis using subsets of T lymphocytes is beyond the scope of this study, so comparisons are not possible. The VCAM-1 blocking Ab significantly decreased adhesion of T lymphocytes to activated HUVEC by 29 to 32%. A similar rate of inhibition (34 to 60%) was also observed with activated HUVEC using the same Ab as in this study (Graber et al., 1990; Shimizu et al., 1991a; Oppenheimer-Makes et al., 1993)  85  as well as with another Ab (Oppenheimer-Marks et al., 1991). A VCAM-1 blocking Ab also decreased the adhesion of splenocytes to activated murine microvascular EC (Bereta et al., 1993). The effect of the anti-ICAM-1 Ab was even greater, resulting in 47 to 75% blocking of T lymphocyte adhesion to activated HBMEC. The mAb used in this study also blocked adhesion to activated HUVEC by 34% (Liversidge et al., 1990). Another anti-ICAM-1 Ab had an even greater effect (70% decrease) on activated HUVEC (Shimizu et al., 1991a). Oppenheimer-Marks et al. (1991) found only a 8% decrease, but the HUVEC used were stimulated for only 4h, a time when ICAM-1 expression is still low. In addition, Abs to CDlla, CD18 and VLA-4, ligands to ICAM-1 and VCAM-1, respectively, decrease the adhesion of peripheral blood lymphocytes to bovine brain EC (de Vries et al., 1994). An anti-ICAM-1 Ab decreases the adhesion of lymphocytes to rat high endothelial venule EC (Tamatani and Miyaska, 1990). In contrast, mononuclear cells from MS patients show enhanced adhesion to unstimulated and activated HBMEC that does not appear to be dependent upon ICAM-1 (Tsukada et al., 1993). There have been no reports on the effect of anti-PECAM-1 Abs on T lymphocyte adhesion. Following TNF activation of HBMEC, the presence of anti-PECAM-1 in the assay increased T lymphocyte adhesion by 14 to 123%. Actually, a 90% increase was also observed with unstimulated HBMEC, though the actual numbers are not statistically significant. This effect has been examined in several reports. Binding of PECAM-1 by anti-PECAM-1 Abs activates the pi integrin, especially VLA-4, the ligand for VCAM-1, and to a lesser extent the p2 integrin LFA-1, the ligand for ICAM-1 on lymphocytes, monocytes, neutrophils and hemopoietic progenitor cells (Tanaka et al., 1992; Piali et al., 1993; Leavesley et al., 1994; Berman et al., 1995). In the present study, when the anti-PECAM-1 Ab was removed prior to the addition of T lymphocytes, the number of adherent lymphocytes was not different from that observed when HBMEC were not 86  treated with the anti-PECAM-1 Ab. Thus, PECAM-1 is likely not involved in the adhesion of T lymphocytes. This study shows that none of the Abs used blocked any epitope utilized in adhesion of resting T lymphocytes to unstimulated HBMEC. TNF activation of EC upregulates this adhesion by a mechanism involving VCAM-1 and ICAM-1.  4.2.4 T lymphocyte migration Migration of T lymphocytes across unstimulated HBMEC monolayers was extremely low (0.02 T lymphocyte/mm length of monolayer). Taking into account the initial adhesion step, 4 to 10% of adherent T lymphocytes migrated across unstimulated HBMEC and 0.25 to 17.5% migrated across TNF-treated HBMEC. These data are consistent with those presented in a recent report on rat retinal microvessel EC showing that 0.8% of peripheral lymph node lymphocytes and 52% of soluble retinal antigen specific T lymphocytes migrated across rat retinal EC regardless of the activation status of the endothelium (Greenwood and Calder, 1993). TNF activation of HBMEC significantly increased T lymphocyte migration by 10 fold (0.2 T lymphocyte/mm) after 4 or 18h of pretreatment and 50 fold (1 T lymphocyte/mm) after 24h of pretreatment. In contrast, activation of rat retinal microvessel EC with IFN-y did not change the migration of lymphocytes or soluble retinal antigen specific T lymphocytes (Greenwood and Calder, 1993). This may be due to the low efficiency of IFN-y as an inducer of EC adhesion molecules, or an inherent difference between EC from different tissues. An increase of 0.2 to 0.7 fold in T lymphocyte migration has been reported after HUVEC activation with IFN-y (Oppenheimer-Marks and Ziff, 1988) and a 0.5 fold increase after HUVEC activation with IL-1 (Oppenheimer-Marks and Lipsky, 1993).  87  Ab blocking of T cell migration across EC mono ayers (%) Untreated  Activated  Untreated  Activated  Ab  HBMEC  HBMEC  HUVEC  HUVEC  anti-E-selectin Ab  0%  40%  -  -  anti-VCAM-1 Ab  0%  25%  0-4%  0-90%  anti-ICAM-1 Ab  0%  93%  50-60%  40-53%  anti-PECAM-1 Ab  0%  75%  -  -  None of the anti-adhesion molecule Abs significantly blocked the migration of T lymphocytes across unstimulated HBMEC. Similarly, anti-VCAM-1 Abs do not decrease T lymphocyte migration across unstimulated HUVEC (Oppenheimer-Marks et al., 1991; Oppenheimer-Marks and Lipsky, 1993). Surprisingly, an anti-ICAM-1 Ab reduced T lymphocyte migration across unstimulated HUVEC by 50-60% (Oppenheimer-Marks et al., 1991; Oppenheimer-Marks and Lipsky, 1993). Since the migration assay used in the present study is the same as in these reports, these differences may reflect inherent differences between these two types of EC. Muller et al. (1993) found that an anti-PECAM-1 Ab decreases peripheral blood mononuclear cell migration across unstimulated HUVEC by 80% in contrast to our results. However, Bird et al. (1993) has reported that PECAM-1 is not necessary for migration of peripheral blood lymphocytes across unstimulated HUVEC or rat high endothelial venule EC, in support of our results. At the present time, there are no studies on the effects of anti-E-selectin Ab on T lymphocyte migration across unstimulated EC. Following TNF activation of HBMEC, the anti-E-selectin Ab marginally decreased T lymphocyte migration by 40%. The anti-VCAM-1 Ab had no effect, in  88  keeping with results previously reported on activated HUVEC (Oppenheimer-Marks et al., 1991; Oppenheimer-Marks and Lipsky, 1993). The anti-ICAM-1 Ab significantly blocked transmigration by 93% unlike the 40 to 53% decrease reported with 4h activated HUVEC (Oppenheimer-Marks et al., 1991; Oppenheimer-Marks and Lipsky, 1993). Inadequate upregulation of ICAM-1 after only 4h treatment with cytokines would most likely account for this difference. Interestingly, both the anti-ICAM-1 and anti-VCAM-1 Abs have been found to be involved in the migration of T lymphocytes from peripheral lymph nodes across rat brain EC (Male et al., 1992). A 75% decrease in migration across activated HBMEC was obtained with the anti-PECAM-1 Ab. A similar (70-90%) blocking of monocyte migration with anti-PECAM-1 Abs has been observed across activated HUVEC (Muller et al., 1993). In a manner similar to PMN, integrins on circulating lymphocytes have low adherence unless they are activated (Adams and Shaw, 1994). Since 10% of peripheral blood T lymphocytes are activated (Hannet et al., 1992), it is conceivable that lymphocytes adhering to HBMEC in the present assays, were among this subset of activated T lymphocytes. Lymphocytes that adhere to EC are found to be activated within a short period of time. This involves G-protein associated receptors (Bargatze and Butcher, 1993). Chemokines, such as MIP-ip can activate the p i integrins on the lymphocyte surface (Tanaka et al., 1993), thus, contact with MIP-1|3 or other chemokines on the EC surface could activate lymphocytes and their integrins to permit greater adhesion and migration. Since expression of MIP-ip by cerebral EC has not yet been reported, this mechanism remains speculative. Earlier studies showed that inflammatory cells migrate across the endothelium between adjacent EC (Wisniewski and Lossinsky, 1991). Recent reports indicate that lymphocytes cross transcellularly, through the EC cytoplasm in the microvessels of mice with EAE (Lossinsky et al., 1991), mouse lung EC (Fujitsu et al., 1991), high endothelial venules (DeBruyn and Cho, 1990), human, cat and rat lymph vessel cells (Azzali et al., 89  1990), rat lacteal lymphatic EC (Azzali, 1990) and mouse mesenteric lymph nodes (Cho and DeBruyn, 1981). In the present study, it appeared that some lymphocytes migrated through the cytoplasm of HBMEC, while others preferred the paracellular route. These results along with those of PMN migration suggest that the route of transendothelial migration may depend both on the type of the inflammatory cell and the type of the microvascular bed. Cross and Raine (1991) found that "finger-like fronds" project from the luminal surface of the involved brain microvessels of mice with EAE. Projections arise at interendothelial junctions and in animals with acute lesions (less than 24 to 48h clinical signs). Lossinsky et al. (1991) also noted these projections and called them "microvillar projections". Such structures were frequently seen on the apical surface of our TNF treated HBMEC and direct contacts between these finger-like extensions of the EC cytoplasm and pseudopodia of PMN and lymphocytes were often observed. Cross and Raine (1991) also observed that PMN contacted these fronds, while small lymphocytes rarely acted this way. Approximately 10% of T lymphocytes in the peripheral blood are activated (Hannet et al., 1992). Presumably, this was the case with the T lymphocytes used in the present study. Activation of lymphocytes in vitro by pharmacological agents increases their adherence to HUVEC (Shimizu et al., 1991a; Greenwood and Calder, 1993; Dustin and Springer, 1988), rat brain microvessel EC (Hughes et al., 1988; Tanaka et al., 1993) and HBMEC (Huynh, 1994) by 2 to 12 fold. Their utilization of the endothelial adhesion molecules is also altered. Binding of activated T lymphocytes to unstimulated HUVEC is inhibited by an anti-ICAM-1 Ab to a small degree and does not involve VCAM-1. Binding to activated HUVEC is not inhibited by anti-VCAM-1 or anti-ICAM-1 Abs (Oppenheimer-Marks et al., 1991). This is in contrast to resting T lymphocytes which use ICAM-1 to adhere to unstimulated EC and VCAM-1 for adhesion to activated EC (Oppenheimer-Marks et al., 1991). Male et al. (1994) also found that anti-ICAM-1 Abs 90  do not decrease adhesion of either activated or resting lymphocytes to IFN-y activated rat brain microvessel EC. The adhesion of memory and activated synovial T lymphocytes from rheumatoid arthritis patients to recombinant VCAM-1 and E-selectin is greatly increased compared to peripheral blood T lymphocytes. Migration of memory T lymphocytes across HUVEC is also enhanced more than 100% over naive T lymphocytes (Oppenheimer-Marks and Lipsky, 1993). Other subsets of peripheral blood lymphocytes are the B lymphocytes and natural killer (NK) cells. Compared to peripheral blood T lymphocytes, the percentage of NK cells that adhere to and transmigrate across unstimulated HUVEC is 3 fold higher and 2 fold higher across activated HUVEC. Both the ICAM-1 and VCAM-1 pathways are employed in these processes (Bianchi et al., 1993). NK cells also preferentially bind to human foreskin microvascular EC (Bender et al., 1987). B lymphocytes bind more strongly to murine brain microvessel EC than T lymphocytes (Pryce et al., 1991). It is apparent that, in vitro, subsets of lymphocytes in peripheral blood possess different abilities to adhere to and migrate across vascular endothelium from different systems. Work done in vivo on mice or rats with EAE demonstrated that sialyl residues (ligands for E-selectin) are important in lymphocyte accumulation in CNS inflammation (Simmons and Cattle, 1992). The majority of T lymphocytes infiltrating the CNS are activated (Hickey et al., 1991; Skundric et al, 1994) and memory cells (Skundric et al., 1994). From studies in which immunostaining for activated T lymphocytes is performed on sections of CNS lesions, it is not possible to determine whether only activated cells enter the CNS, or resting T lymphocytes entered and became activated in the extravascular tissue. Hickey et al. (1991) showed that when T lymphocytes from one strain of rat were injected into another strain, ten times more T lymphocytes were found in the spinal cord if the T lymphocytes were first activated. However, conclusions drawn must be made with care since, normally, in the blood, there are ten times more resting T lymphocytes than activated (Hannet et al., 1992). Thus, the mathematics would show 91  that the same number of activated and resting T lymphocytes will eventually wind up in the CNS. CD4 + helper/inducer T lymphocytes are more abundant in inflammatory CNS infiltrates than CD8 + suppressor/cytotoxic T lymphocytes. Most of these cells are of the T H 1 subset and synthesize IFN-y and TNF-a (Skundric et al., 1994). Pryce et al. (1991) found that CD8 + T lymphocytes bind more abundantly than CD4 + T lymphocytes to murine brain microvessel EC in culture, but transendothelial migration was not assessed and may not necessarily follow the pattern of adhesion. EAE can be transferred from an affected animal to a syngeneic unaffected recipient by transferring helper T lymphocytes from the former to the latter. In the lesions that form in the recipient animal, the majority of infiltrating cells are derived from the recipient animal and are not CNS specific. The transferred CNS-specific helper T lymphocytes remain exclusively in perivascular cuffs and disappear from the CNS within two weeks of chronic illness (Cross et al., 1990). The results from this study show that the baseline, unstimulated migration of resting T lymphocytes across HBMEC is minimal and not mediated by any of the EC adhesion molecules studied. TNF activation of EC upregulates migration largely via the E-selectin, ICAM-1 and PECAM-1 pathways. It has generally been surmised that E-selectin is only involved in adhesion under conditions of flow. This is the first study in which a role for E-selectin in transmigration of T cells is demonstrated. Although the contribution of E-selectin is definite, it is not as prominent as that of ICAM-1 and PECAM-1. The current dogma on leukocyte-EC interactions, based on studies utilizing HUVEC, indicates that under conditions of flow, leukocytes initially roll along EC using the selectins. At the same time, leukocytes come in contact with chemoattractants and chemokines that activate these blood cells and the integrins on their surface. The leukocytes then firmly adhere to VCAM-1, ICAM-1 and PECAM-1 on the EC, move toward the tight junctions between EC and migrate across transcellularly or 92  paracellularly, along a chemotactic gradient, using these same molecules. The present results with activated HBMEC generally support this theory, except that PECAM-1 does not seem to be involved in the migration of PMN, or adhesion of PMN and T lymphocytes although it is involved in the migration of T cells across activated HBMEC. E-selectin now takes on a new role in facilitating T cell migration. In unstimulated HBMEC, ICAM-1 does not participate in the adhesion of PMN or the migration of PMN and T lymphocytes, and VCAM-1 is not used in the migration of T lymphocytes in contrast to HUVEC. This may account for the small number of leukocytes that eventually migrate across unstimulated HBMEC, consistent with the formation of a barrier under normal conditions.  Anti-adhesion molecule treatment has been studied in vivo during CNS inflammation. Administration of Abs against the a461 (VLA-4) integrin, the ligand for VCAM-1, prevents the accumulation of leukocytes in the CNS of rats and the induction of EAE in these animals (Yednock et al., 1992). Abs against other molecules, including the 62 integrins, are ineffective. However, an anti-ICAM-1 Ab has been found to inhibit active (inoculation with CNS myelin plus adjuvant) or adoptively transferred (injection of CNS-specific T lymphocytes) EAE in rats and reduce CNS inflammation (Archelos et al., 1993), as well as reduce ischemic injury in rabbits (Clark et al., 1991). In contrast, other authors, using the same Ab, observed no inhibitory activity in TNF induced CNS inflammation or adoptively transferred EAE, but some protection in active EAE in a small fraction of animals (Willenborg et al., 1993). The use of a different anti-ICAM-1 Ab in another study did not have a significant effect on mice with adoptively transferred EAE (Cannella et al., 1993). The reasons for this discrepancy are unknown. How this applies to the human demyelinating disease, MS, or other inflammatory CNS lesions, and how different Abs against human ICAM-1 or other adhesion molecules will fare, all await future investigations. 93  The present results indicate that depending on the type of inflammation, the use of different combinations of Abs may be necessary. For example, in the early stages of acute inflammation in meningitis and encephalitis, when PMN infiltrate the brain, anti-E-selectin, anti-ICAM-1 and perhaps anti-PECAM-1 may be used. Several days later, anti-ICAM-1, anti-VCAM-1 and perhaps anti-PECAM-1 and anti-E-selectin could be used to inhibit lymphocyte infiltration. In more chronic types of inflammation, such as MS, where mononuclear leukocytes are more prominent than PMN, a combination of all these Abs should be used. It may be necessary to administer these Abs during remission so that Abs are always available to bind adhesion molecules whenever they may be upregulated. Soon after the inflammatory reaction is in full swing during a relapse and a large number of lymphocytes have crossed the BBB, it would be too late to use anti-adhesion molecule Abs. Unless, perhaps, injected or otherwise brought into the brain to prevent leukocytes from binding to and adversely affecting glia or neurons in the parenchyma. Two clinical trials using anti-ICAM-1 Abs in kidney transplants and rheumatoid arthritis seem to point toward some beneficial effects in these conditions (Haug et al., 1993; Kavanaugh et al., 1992). Of concern are the high immunogenicity of murine Abs, low yield of human Abs and reduced antigen affinity as well as low yield of humanized murine Abs (Peake, 1993). Different Abs block different epitopes of the antigen to a different degree. It is known that several epitopes on each adhesion molecule may be used to different extents. It may be necessary to employ a large number of Abs in an attempt to maximally inhibit binding. This may also be the reason why studies using different Abs report contrasting results. The findings by Welder et al. (1993) hold great promise. It was demonstrated that soluble ICAM-1 coated polystyrene microspheres can competitively inhibit cell adhesion that utilizes the ICAM-1 - LFA-1 pathway, whereas simple monovalent soluble ICAM-1 is ineffective. Moreover, all the epitopes are  94  conceivably blocked with soluble ICAM-1. Also worth consideration would be drugs that pharmacologically inhibit the expression of these adhesion molecules. Molecules other than adhesion molecules are also involved in EC- leukocyte interactions. These include the class II major histocompatibility complex (MHC) molecules and their receptor, the T cell receptor. The inhibition of specific T cell receptors and class II MHC antigen have also been shown to reduce T lymphocyte migration across EC in vitro and in vivo (Huynh, 1994; Kuchroo et al., 1994).  4.3 DETERMINATION OF MONOLAYER PERMEABILITY FOLLOWING TNF-a TREATMENT AND IN THE COURSE OF LEUKOCYTE MIGRATION ACROSS THE MONOLAYERS The low permeability of the cerebral endothelium can be variably demonstrated by the low hydraulic conductivity, high electrical resistance and exclusion of tracer molecules (Abbott and Revest, 1991). The electrical resistance of cerebral microvessels has been previously determined in amphibians and rats. The average resistance of frog pial microvessels is 1870Q-cm , while that of rat pial vessels ranges from 918Q-cm^ to 1490Q-cm2 (Crone and Olesen, 1982; Abbot and Revest, 1991). These values are indicative of a tight endothelial barrier with virtually impermeable tight junctions. There are no reports on the in situ electrical resistance of the primate brain microvessels. Studies on brain EC isolated from human and bovine brains have reported electrical resistances across monolayers ranging between 10-50Q-cm2 (Rubin et al., 1991). Interestingly, the low resistance values could be increased up to an average 500Q-cm 2 when EC were cultured in the presence of agents that increase cyclic AMP levels in the cell (Rubin et al., 1991). Growth in the presence of astrocyte condition medium increased resistance only to 70-115Q-cm . However, serially passaged bovine brain EC cultured on collagen gel can attain resistances of 150 to 800Q-cm2 depending on the culture, in the absence of cAMP or astrocytes (Rutten et al., 1981). Endothelial 95  monolayers from other organs have significantly lower resistance. For example, bovine adrenal cortical microvessel EC have resistances of 70S2-cm^ (Furie et al., 1984), HUVEC (Burke-Gaffney and Keenan, 1993), bovine aortic endothelium (Rutten et al., 1987; Territo et al., 1984; Albelda et al., 1988) and rabbit aortic endothelium (Navab et al., 1986) have resistances between 13-18£2-cm . Bovine and porcine pulmonary artery EC have resistance from 4 to 6Q-cmr (Shasby and Shasby, 1986; Turner, 1992) and rat epididymal pad EC as low as 0.45Q-cmi (Rutten et al., 1987). Since voltage equals current divided by resistance, difference ohm-meters employing different current may slightly affect measurements made by different investigators. However, this may not extend beyond the background noise. Differences in culture conditions, medium, temperature, etc. may also account for a small part of the differences between HBMEC and HUVEC in published reports. The average resistance of HBMEC, grown in the absence of astrocytes, astrocyte conditioned media or cyclic AMP in the present study, was 108Q-cm^ and as high as 142Q-cm . These values are certainly higher than those reported in extracerebral EC, lower than some of the reported resistances of bovine brain EC, but considerably higher than the resistance of human and some bovine brain EC grown without cyclic AMP or astrocyte conditioned media. Although the factors required for the formation and maintenance of physiologically "tight" tight junctions are presently unknown, culture of HBMEC in artificial media and the presence, in a given culture, of EC from different segments of the microvascular bed, including venules that have much less "tight" tight junctions, would contribute, at least in part, to the present resistance values. In addition, it has been suggested that, sometimes large, ranges of resistance in monolayers of the same type of endothelium may be due to small (0.5-2um) gaps between 5 to 10% of EC (Albelda et al., 1988). It should be emphasized that the relevance of the resistance values obtained in our system would be more meaningful if compared to the in situ electrical resistance of the human brain microvessels which is presently unknown. 96  TNF treatment significantly decreased the electrical resistance across HBMEC monolayers by about 10% after 4h of treatment. The decrease was maximal by 18h with a 40% drop. Previous studies have shown that TNF treatment of HUVEC for 24h increases I-^-albumin clearance rate by 4 fold (Burke-Gaffney and Keenan, 1993). In contrast, no significant change in I-^-albumin clearance was observed following TNF treatment of HUVEC for 4h (Furie and McHugh, 1989; Moser et al., 1989). Treatment of bovine pulmonary microvessel EC with 100-10000 U TNF/ml for 24h or bovine pulmonary artery EC with 200 U TNF/ml for 18h increased permeability to I *  -albumin  by 3 to 7 fold (Partridge et al., 1992,1993; Wheatley et al., 1993). The mechanism of TNF-induced increased EC permeability seems to involve the degradation of some extracellular matrix components, such as fibronectin, laminin, collagen and gelatin, by a metalloproteinase (Partridge et al., 1993; Wheatley et al., 1993) leading to remodelling of the extracellular matrix and decrease EC-extracellular matrix contact (Partridge et al., 1992). Among other inflammatory mediators, LPS at lOug/ml causes a 25% increase in the rate of equilibration of ^H-water, 40% increase in I * -albumin clearance and a 25-50% increase in hydraulic conductance of bovine pulmonary artery endothelium with maximal effect by 3-4h (Meyrick et al., 1986). IFN-y treatment of epithelial cell cultures decreases the resistance maximally to less than 25% by day 3. However, permeability to horseradish peroxidase (HRP) did not change (Adams et al., 1993). An alteration in the tight junctions (paracellular movement) rather than cell necrosis or transcellular movement was observed. Results from these different systems seem to suggest that longer cytokine treatments are required to increase the permeability of tissues that form a tighter barrier, such as HBMEC and epithelium, than those that are leaky. The different nature of the tight junctions may be involved. In vivo, intracisternal administration of 800 U TNF into the rat CNS increased I1  -albumin penetration into the CSF, maximally (4 to 7 fold) by 3h, while intravenous  administration lacked any effect (Kim et al., 1992). LPS, IL-1 and to a lesser extent, 97  TNF, injected intradermally in rats, induce leakage of colloidal carbon from postcapillary venules maximally after 3 to 4h, (Yi and Ulich, 1992). However, when IL-la and p are injected intraventricular^ or intravenously in the mouse, the integrity of the BBB, as measured by I-^-albumin movement, is not affected after 2h. It is difficult to compare these results because of the different routes of cytokine administration and different concentrations injected. There have been very few reports on changes in endothelial permeability due to cytokines and leukocyte migration and none reporting the decrease in resistance due to both processes as shown in the present study for HBMEC. The small increase in permeability due to leukocyte migration across unstimulated HBMEC is greatly increased after TNF treatment which not only increases permeability on its own, but also upregulates adhesion molecules to significantly increase leukocyte migration. Both of these processes combine to generate a large increase in permeability. In vitro, it has been reported that the migration of PMN across unstimulated or TNF (200U for 1 to 6h) activated bovine pulmonary artery EC causes a 20% increase in albumin permeability (Gibbs et al., 1990). Migration of activated memory T lymphocytes but not naive T lymphocytes increases endothelial permeability to albumin (Dalme and Doyle, 1990). Whether this is due to the migratory process itself is unknown. In animal models, the leakage of the tracer HRP across rat and mice cerebral microvessels in EAE has been variably attributed to increased tight junctional permeability, increased interendothelial space, increased numbers of cytoplasmic vesicles and tubular profiles or leakage alongside migrating leukocytes (Claudio et al., 1990; Lossinsky et al., 1989). The migration of leukocytes across mesenteric venules of cats subjected to ischemia and reperfusion has also been found to contribute to the leakage of HRP (Oliver et al., 1991). Even though the addition of an anti-CD18 (ligand to ICAM-1) Ab inhibited the adhesion and migration of leukocytes in this model, some HRP leakage was still evident across the blood vessels (Oliver et al., 1991). These data are consistent with our observations. 98  Migration of PMN and T lymphocytes variably decreased the resistance of monolayers by 20 to 70%. The decrease in resistance of unstimulated HBMEC following migration of PMN in response to chemotactic gradients of fMLP is especially great. This may conceivably be due to the greater number of PMN migrating per mm of EC monolayer than T lymphocytes. Although certain anti-endothelial adhesion molecule Abs decreased the adhesion and migration of PMN and T lymphocytes, they did not reverse the drop in resistance. This may be due to the lack of complete inhibition by the Abs. The small number of leukocytes still able to migrate may be sufficient to increase the permeability. Alternatively, EC-leukocyte contact may induce the leukocyte to release some fast acting mediators) on the EC with increase in permeability by some unknown mechanism. It is clear that leukocyte migration can increase the permeability of the BBB. It seems unlikely that the increase in HBMEC permeability caused by TNF treatment affects migration by itself since blocking ICAM-1 by an Ab greatly decreases T lymphocyte migration, but does not affect the increase in permeability due to TNF treatment. Thus, although the use of anti-adhesion molecule Abs in vivo may suppress inflammatory cell infiltration, the presence of cytokine, such as TNF, will increase the BBB permeability independent of leukocyte migration. It would be necessary, therefore, to counteract the effect of these cytokines perhaps by using Abs directed against them or their synthetic soluble receptors (Mohler et al., 1993). The effect of leukocyte migration on epithelial permeability has been documented by several laboratories. Peritoneal exudate cells decrease epithelial resistance by 70% after 90min of incubation (Evans et al., 1983), while PMN cause a 80 to 100% drop after 30 to 60 min in one report (Nash et al., 1987,1988) and a 46% drop after 40 min in another (Milks et al., 1986). 10"'M fMLP was present under the epithelium in these experiments. Although PMN migration across HBMEC in response to a gradient of 10"'M fMLP was prominent, the resistance never dropped to such low levels. It is interesting to note that upon removal of PMN that are on top of the epithelium, the 99  resistance recovered by 20 to 60 min (Evans et al., 1983; Milks et al., 1986). The drop in epithelial resistance has been further dissected by Parkos et al. (1992) who noted that this is a biphasic response. The initial decrease in permeability, maximal 5 to 13min after transmigration begins, is due to chloride secretion by the epithelial cells and not to changes in tight junctional permeability. Further decrease in resistance after 60min of migration is due to increased junctional permeability. Whether this also occurs in HBMEC is not known. It should also be kept in mind that resistance measurements represent the movement of a current (electrons) across the tissue. Even though electrons may be able to penetrate a tight junction in some instances, other physiologically relevant molecules that are larger, such as ions (Na + , K + , CI"), water and proteins may not. Thus, it would be more informative to complement electrical resistance measurements with tracer studies utilizing tracers such as HRP, I 1  -albumin, 3 H-H20 and ionic lanthanum  to further dissect this event. Indeed, morphological techniques employing HRP, lanthanum and autoradiography can clearly distinguish between the transcellular and paracellular pathways of tracer leakage. The present study shows that HBMEC monolayers have relatively high electrical resistance which can be progressively decreased by TNF treatment. Transendothelial migration of PMN and T lymphocytes further decreases the resistance. Although anti-adhesion molecule Abs decrease PMN and T lymphocyte migration, they do not affect the drop in resistance due to this process.  100  CHAPTERS CONCLUSIONS  5.1 SUMMARY AND SIGNIFICANCE The aim of this study was to 1) examine the expression, distribution and upregulation of the adhesion molecules E-selectin, vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1), platelet/endothelial cell adhesion molecule-1 (PECAM-1) and HECA 452 antigen in primary cultures of HBMEC. 2) Elucidate their role in polymorphonuclear leukocyte (PMN) and T lymphocyte adhesion to and transmigration across the endothelium. 3) Determine the effects these events have on endothelial permeability. We hypothesized that in the initial stages of central nervous system (CNS) inflammation, lipopolysaccharide (LPS) and cytokines can upregulate and / or change the distribution of adhesion molecules on HBMEC. This "activation" of EC can then initiate leukocyte adhesion and transendothelial migration, with a concomitant increase in the permeability of the endothelial barrier. Use of antibodies against these adhesion molecules can counteract some of these processes. Our findings show that a small percentage of cells in primary cultures of HBMEC express low levels of ICAM-1, VCAM-1 and E-selectin, the percentage of expression decreasing in that order. PECAM-1 is constitutively expressed by all HBMEC with a preferential localization at the cell borders. Activation of HBMEC with LPS, tumor necrosis factor-a (TNF-a) and interleukin-ip (IL-ip) dramatically upregulates the expression of E-selectin, VCAM-1 and ICAM-1. LPS is most effective in this respect, followed by TNF, then IL-1. IFN-y weakly upregulates ICAM-1 but not the other adhesion molecules. Activation of EC bears no effect on the expression of PECAM-1. E-selectin is maximally upregulated 4h after activation. Its expression then drops rapidly and by 24 to 48h, is not different from that in unstimulated cultures. VCAM-1 becomes 101  maximally expressed 12 to 24h after activation and returns to unstimulated levels by 48h with most treatments. ICAM-1 expression becomes maximal 24h after HBMEC activation and remains high for at least 72h. Growth in the absence of LPS and cytokines for 3 days after maximal activation does not reduce the expression of ICAM-1. Ultrastructurally, E-selectin, VCAM-1 and ICAM-1 are expressed predominantly on the apical surface on unstimulated HBMEC. TNF treatment increases the apical expression to a greater extent than the basal expression. Labeling is particularly dense at finger-like cytoplasmic projections. Cytokine activation of HBMEC greatly increases the number of adherent and migrated PMN and T lymphocytes. The minimal unstimulated adhesion and transmigration of PMN and T lymphocytes across unstimulated HBMEC is not affected by adhesion molecule blocking Ab with the exception of anti-ICAM-1 mAb that significantly decreases T lymphocyte adhesion. It is possible that some epitope of ICAM-1 is involved in this adhesion. Blocking mAb, however, effectively block the augmented adhesion and migration of PMN and T lymphocytes across cytokine treated HBMEC monolayers. Thus, both E-selectin and ICAM-1 are involved in PMN adhesion and ICAM-1 is utilized during PMN migration. VCAM-1 and ICAM-1 are required for T lymphocyte adhesion. Lastly, E-selectin, ICAM-1 and PECAM-1 all play an active role in T lymphocyte migration. Unstimulated monolayers of HBMEC have a high electrical resistance indicative of their low permeability. Treatment with TNF increases the permeability progressively. Migration of PMN and T lymphocytes across TNF-treated HBMEC further increases the permeability. Although anti-adhesion molecule Abs effectively block migration of these leukocytes, they do not prevent the increase in permeability. In summary, based on our findings in primary cultures of HBMEC, we hypothesize that cerebral EC in situ constitutively express high levels of PECAM-1, relatively low levels of ICAM-1 and very low levels of E-selectin and VCAM-1. 102  ICAM-1 would be involved in the adhesion of an extremely small number of T lymphocytes to intact HBMEC under normal conditions. Viral infections would lead to release of viral proteins and secretion of cytokines by mononuclear phagocytes and T cells that would in turn upregulate the expression of E-selectin, VCAM-1 and ICAM-1. Bacterial infections would similarly upregulate adhesion molecule expression via secretion of bacterial products and LPS. Upregulation of these adhesion molecules would then facilitate leukocyte adhesion and transendothelial migration. It is evident from the present studies that PMN utilize E-selectin and ICAM-1 for initial adhesion and ICAM-1 for migration, while T lymphocytes use VCAM-1 and ICAM-1 for adhesion and E-selectin, ICAM-1 and PECAM-1 for migration. In the course of these early events in the evolution of the inflammatory response, the permeability of the BBB will be compromised both by the direct action of cytokines and transendothelial migration of leukocytes. Abs against EC adhesion molecules can, at least partially, block the corresponding interactions but cannot prevent the increase in endothelial permeability secondary to leukocyte migration. The significance of this study lies in the utilization of a new cell culture system, not previously accessible, that models the human blood-brain barrier (BBB). It is apparent from the findings presented herein that some conclusions drawn from other models, such as HUVEC and EAE are applicable to the human BBB while others are not. For example, ICAM-1 expression does not reverse as readily as in HUVEC, VCAM-1 expression drops more rapidly and PECAM-1 is not affected by cytokines unlike HUVEC. The adhesion molecules that regulate PMN and T lymphocyte adhesion and migration differ in some instances. For example, the anti-E-selectin Ab, unlike its effect on HBMEC, was found to inhibit the adhesion of a subset of T lymphocytes and the migration of PMN across HUVEC. Anti-PECAM-1 Ab blocked PMN migration across HUVEC but not across HBMEC. Some molecules have not been previously examined in certain interactions, such as E-selectin in T lymphocyte migration and PECAM-1 in 103  PMN and T lymphocyte adhesion. Similarly, permeability changes of the cerebral endothelium during leukocyte migration have not been previously studied. The results of these studies extend our understanding of the mechanisms involved in the inflammatory process in the CNS. Therapeutic interventions utilizing mAbs against adhesion molecules or even soluble forms of adhesion molecules may prove useful in altering the course and severity of inflammation in the CNS.  5.2 FUTURE PROSPECTS Three interesting projects are apparent from this immediate study. Firstly, expression of E-selectin and VCAM-1 return to unstimulated levels 24 to 48h after HBMEC activation. However, in patients, these molecules can be detected many days after the initiation of inflammation in the CNS. This may be because areas peripheral to the initial lesion are being activated in sequence, but Sepp et al. (1994) have also found a factor in human serum that prolongs the expression of E-selectin by HDMEC at least for 48h. It would be important to see if this is restricted to HDMEC and if this or other factors in human blood can otherwise modulate adhesion molecule expression on HBMEC. Secondly, as mentioned in the discussion, leukocytes may bind to several different epitopes on each adhesion molecule, while each mAb can only block one epitope. It cannot be derived that an adhesion molecule is not employed by simply using only one mAb. Since it is very difficult to produce EC that are genetically altered not to produce specific adhesion molecules, the soluble adhesion molecule coated microspheres that Welder et al. (1993) used would be ideal for experiments to block all the epitopes on any one molecule. Thirdly, the permeability experiments done in this study only measure the movement of current (electrons) across the monolayer. Movement of larger molecules such as ions, water and albumin also need to be examined. This can be done using 104  electron dense tracers such as ionic lanthanum and HRP for electron microscopy, and •^H-t^O and I*  -albumin in physiological experiments.  Other interesting aspects have not been mentioned during the discussion, but would be just as important to examine. Other than using Abs and soluble adhesion molecules, it may also be possible to inhibit the synthesis of adhesion molecules. Drugs that can perform this way would be very important in therapy. Bennett et al. (1994) produced oligonucleotides that inhibit the expression of ICAM-1, VCAM-1 and E-selectin on treated HUVEC. Other agents, such as nitric oxide, inhibit EC-leukocyte interactions by unknown mechanisms. These could be other avenues for therapy. It is also known that tumor cells adhere to blood vessel walls via adhesion molecules. Experiments with these cells along the same line as the leukocytes in this study could elucidate the mechanisms employed by tumor cells in metastasis. It would be especially important to examine the differences between EC of different sources to determine the differences that lead to preferential metastasis to specific organs, as well as differences between tumor and normal vessels. Therapy can be produced to inhibit adhesion of tumor cells to these organs and to encourage lymphocytes to infiltrate tumor vessels. Other than adhesion molecules, chemokines / chemoattractants are also important in the recruitment of leukocytes to inflammatory sites. Many chemokines such as monocyte chemoattractant protein 1 and interleukin-8 have not been examined in CNS inflammation. Literally a life-time's worth of studies remain to be done in this fast-paced field.  105  Figure 1.  Confluent culture of human brain microvessel endothelial cells grown on (a) fibronectin coated plastic wells for adhesion molecule expression assays and adhesion assays, (b) Cellagen discs for migration assays, scanning electron microscopy and electrical resistance measurements. X270  106  Figure 1.  KV  107  Figure 2.  Primary cultures of human brain microvessel endothelial cells show positive staining for Factor VIII antigen, confirming their endothelial origin. X280.  108  Figure 2.  109  Figure 3.  Illustration of a Cellagen disc in a well of a 24 well plate. Human brain microvessel endothelial cells are grown in this system for migration, scanning and transmission electron microscopy.  110  Figure 3.  Outer chamber Inner chamber  Lymphocyte Support  Cellagen membrane Primary culture of human brain microvessel endothelial cells  111  Figure 4.  (a) HBMEC expression of E-selectin without LPS or cytokines, detected by immunogold silver staining, (b) HBMEC expression of E-selectin after incubation with l[xg LPS/ml for 4h, detected by immunogold silver staining, (c) HBMEC expression of E-selectin after incubation with 100U TNF-a/ml for 4h, detected by immunogold silver staining, (d) HBMEC expression of E-selectin after incubation with 10U IL-lp7ml for 4h, detected by immunogold silver staining. X270.  112  Figure 4.  113  Figure 5.  Quantitation of E-selectin expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with LPS for 0 • , 4 a , 12 3 , 24 • , or 48h a . Values represent mean ± standard error of triplicate (IGSS) or duplicate wells (ELISA) from three experiments. p<0.05 compared to expression at Oh.  114  Figure 5.  .9 100  IGSS  0.001 0.01 0.1 1 Concentration of LPS (jig/ml)  ELISA B  O  ON  Xt-t—>  0) O  B «3 -O t-i  O CO  -D  <l 0.001 0.01 0.1 1 Concentration of LPS (jxg/ml)  115  Figure 6.  Quantitation of E-selectin expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with TNF-a for 0 • , 4 Q , 12 Q , 24 3 , or 48h 13. Values represent mean ± standard error of triplicate (IGSS) or duplicate wells (ELISA) from three experiments. jig  p<0.05 compared to expression at Oh.  116  Figure 6.  .5 100  IGSS  o  80  1/3 i  M-l  60 T3  40  13 JO  20 O  n  ^  1 10 100 Concentration of TNF-a(U/ml)  ELISA oa 4 -  O  a  P  •8 o  1 0  10 100 Concentration of TNF-a(U/ml)  117  Time Oh 4h 12h 24h 48h  Figure 7.  Quantitation of E-selectin expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with IL-ip for 0 Q , 4 S , 12 Q , 24 Q , or 48h 03 • Values represent mean ± standard error of triplicate (IGSS) or duplicate wells (ELISA) from three experiments. p<0.05 compared to expression at Oh.  118  Figure 7.  a 100  IGSS  a 0) CO I  W VH  cS T3  in CO  u 0.1 1 10 Concentration of IL-1J3 (U/ml) a  ELISA  Concentration of IL-1|3 (U/ml)  119  Figure 8.  Quantitation of E-selectin expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with IFN-y for 0 Q , 4 Q , 12 a , 24 0 , or 48h 13. Values represent mean ± standard error of triplicate (IGSS) or duplicate wells (ELISA) from three experiments. No statistically significant difference from expression at Oh was found by AN OVA.  120  Figure 8.  IGSS .5 100 o 13  80  co i  W  «s 60 CO  13  40 20 -  15 o  500 Concentration of IFN-Y (U/ml)  ELISA  a  8 a\  A4  n"cd 3 CO  o  I  2 1L  o < 1 -9  0  IN  500 Concentration of IFN-Y (U/ml)  121  Figure 9.  (a) HBMEC expression of E-selectin without LPS or cytokines detected by immunogold EM. (b) HBMEC expression of E-selectin after treatment with 100U TNF-a/ml for 4h detected by immunogold TEM. (c) Note the higher amount of gold particles at the finger-like projections. X38,000.  122  Figure 9.  ... .  m  /Us < ,  .#>'  :  S?S"  |I  #:.4.«t., * 5 ;  123  Figure 10.  Immunogold electron microcopy quantitation of E-selectin expression on HBMEC apical S and basal Q surfaces after incubation with 100U TNF-a/ml for 4h or left unstimulated. Values represent mean ± standard error of one hundred cells for each treatment.  124  Figure 10.  SggApical g%3Basal  Unstimulated 100U TNF-a/ ml X4h j  i  0 2 4 6 8 10 Number of gold particles / urn cell membrane  125  Figure 11.  Immunogold silver staining of HBMEC monolayers for VCAM-1. (a) Unstimulated expression of VCAM-1 is limited to a small number of cells in confluent cultures. Staining appears most dense at the periphery of the cell, (b-d) Deregulation of VCAM-1 expression following treatment with l\ig LPS/ml for 24h (b); 100U TNF-a/ml for 24h (c); 10U IL-lp/ml for 24h (d). a-d, X140.  126  Figure 12.  Quantitation of VCAM-1 expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with LPS for 0 • , 4 a , 12 a , 24 a , or 48h a . Values represent mean ± standard error of duplicate wells from three (IGSS) or two (ELISA) experiments. compared to expression at Oh.  128  p<0.05  Figure 12.  I  IGSS  100  80 > !-l  60  13  15 •9  40  T — 1  CO  20  r—H i—i  o ^  0  0.001 0.01 0.1 1 Concentration of LPS (j^g/ml)  ELISA  as o ON -i—>  2h  <D O  fl  a X> O on JO <  0  0.001 0.01 0.1 1 Concentration of LPS (\ig/m\)  129  Figure 13.  Quantitation of VCAM-1 expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with TNF-a for 0 Q , 4 SI, 12 Q , 24 a , or 48h £3 • Values represent mean ± standard error of duplicate wells from three (IGSS) or two (ELISA) experiments. •J;  p<0.05 compared to expression at Oh.  130  Figure 13.  100  u  IGSS  80  >  >H  £ T3  60  <U  40 cd  •—<  <u u ^  20  24h 48h  0  1 10 100 Concentration of TNF-a(U/ml)  ELISA  Concentration of TNF-a(U/ml)  131  Figure 14.  Quantitation of VCAM-1 expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with IL-lp for 0 Q, 4 13,12 Q , 24 a , or 48h 3 • Values represent mean ± standard error of duplicate wells from three (IGSS) or two (ELISA) experiments. p<0.05 compared to expression at Oh.  132  Figure 14.  IGSS 100  >  Time Oh 4h 12h 24h 48h  -a 13  13 o  a  0.1 1 10 Concentration of IL-ip (U/ml)  ELISA  Concentration of IL-1(3 (U/ml)  133  Figure 15.  Quantitation of VCAM-1 expression on HBMEC by (a) immunogold silver staining (IGSS) and (b) ELISA after incubation with IFN-y for 0 Q , 4 S , 12 Q , 24 a , or 48h S • Values represent mean ± standard error of duplicate wells from three (IGSS) or two (ELISA) experiments. No statistically significant difference from expression at Oh was found by AN OVA.  134  Figure 15.  IGSS 100 80 >  a  60  T3 <D  40  •  ci •—H 00  20  <D  m  a ^  0  500 Concentration of IFN-Y (U/ml)  a  ELISA  a o °^ ~ *• 2 •*->  «j  <u  o a  •S 1 o on  <  0  -  H  T  f|  U  500 Concentration of IFN-Y (U/ml)  135  Time Oh  4h  12h 24h 48h  Figure 16.  Quantitation of VCAM-1 expression on HBMEC by immunogold silver staining after incubation with O.lU/ml SI or OU/ml • IL-lp and lOU/ml, lU/ml and OU/ml TNF-a for 48h. Values represent mean ± standard error of duplicate wells from three experiments. No statistically significant difference from expression without (Oh) TNF or IL-1 was found by ANOVA.  136  Figure 16.  IGSS  u >  «S i—H  X)  [IL-1|3] 1 0 U/ml  (/J  o  0 1 10 Concentration of TNF- a(U/ml)  137  B&  1 0.1 U/ml  Figure 17.  Immunogold labeling for VCAM-1 in HBMEC. (a) Small number of gold particles indicating the presence of VCAM-1 decorate the apical cell surface of a small population of untreated HBMEC. Labeling is minimal or absent at the basal cell surface. X44,000. (b-d) Treatment of HBMEC monolayers with TNF-cc (lOOU/ml) for 18h significantly increases the number of gold particles at the apical and to a lesser extent the basal cell surfaces indicating upregulation of VCAM-1 expression. Labeling is most intense near or at finger-like cytoplasmic projections of the apical surface (c) and is not present along intercellular contacts (between arrow heads) or at tight junctions (arrows) between adjacent EC. b, X52,000; c, X50,000; d, X50,000.  138  Figure 11.  + S-  121  Figure 17.  139  Figure 18.  Quantitation of VCAM-1 expression on the apical S and basal a surfaces of unstimulated and TNF-a (lOOU/ml for 18h) treated HBMEC following immunogold labeling. Values represent mean ± standard error of one hundred unstimulated and one hundred TNF-a treated cells.  140  Figure 18.  Unstimulated 100U TNF-a/ ml X18h  0 5 10 15 20 Number of gold particles / urn cell membrane  141  Figure 19.  (a) HBMEC expression of ICAM-1 without LPS or cytokine, detected by immunogold silver staining, (b) HBMEC expression of ICAM-1 after incubation with 0.1 ^g LPS/ml for 24h, detected by immunogold silver staining, (c) HBMEC expression of ICAM-1 after incubation with 500U IFN-y/ml for 24h, detected by immunogold silver staining, (d) HBMEC expression of ICAM-1 after incubation with 10U TNF-a/ml for 24h, detected by immunogold silver staining, (e) HBMEC expression of ICAM-1 after incubation with 10U IL-lp7ml for 24h, detected by immunogold silver staining, (f) HBMEC expression of ICAM-1 after incubation with 100U TNF-a/ml and 500U IFN-y/ml for 24h, detected by immunogold silver staining. X250.  142  Figure 19.  ^ * *  «3»  143  Figure 20.  Quantitation of ICAM-1 expression on HBMEC by immunogold silver staining after incubation with LPS for 0 • , 4 a , 12 a , 24 a , 48 a , or 72h [0. Values represent mean ± standard error of results from three experiments. p<0.05 compared to expression at Oh.  144  GO  O 4^ |s> (-* -^ O H £3"* t-T* tr* tr* CD  ««««^«««««««««««^^h  % cells labeled for ICAM-1  e  "3  Figure 21.  Quantitation of ICAM-1 expression on HBMEC by immunogold silver staining after incubation with IFN-y for 0 • , 4 Q , 12 3 , 24 a , 48 gl, or 72h m. Values represent mean ± standard error of results from three experiments.  p<0.05 compared to expression at Oh.  146  Figure 21.  IGSS 100 r-—i  3 l-H 5-i  I i—H  o  10 50 100 500 Concentration of IFN-Y(U/ml)  147  Figure 22.  Quantitation of ICAM-1 expression on HBMEC by immunogold silver staining after incubation with TNF-a for 0 • , 4 Q , 12 Q , 24 a , 48 SI, or 72h H . Values represent mean ± standard error of results from three experiments.  p<0.05 compared to expression at Oh.  148  o  T-L  4^ O T  o o"T  00  to oo 4^ K> er tr tr tr tr er CD  <| 4^ IS) H^ 4^ O  ^azazza^zazgzzsgz  hi  v^^/^/^/^//^^ ELDH #•*  o  to  o o  % cells labeled for ICAM-1  GO CO  O  3 e  Figure 23.  Quantitation of ICAM-1 expression on HBMEC by immunogold silver staining after incubation with IL-ip for 0 • , 4 S , 12 3 , 24 a , 48 a , or 72h H . Values represent mean ± standard error of results from three experiments.  p<0.05 compared to expression at Oh.  150  8  t 3  .2P fa  00 00  o  o o  * i-ES  *  o oo  OxJ-HN^-h  <D 43 43 43 43 d 43 43 <N ^t 00 (N  .5  Hi  o  o  sipo %  o  * V3&2ZZ2ZZZZ2ZZZSZZ2Z22L  ^  o  I - J W D I JOJ p^q^i  O i  o o • 1—I  c3 \-<  53  • * - >  0)  o a o U  lO  Figure 24.  Quantitation of ICAM-1 expression on HBMEC by immunogold silver staining after incubation with 500U/ml 3 , lOOU/ml Q , or OU/ml • IFN-y and lOOU/ml, lOU/ml, or lU/ml TNF-a for 24h. Values represent mean ± standard error of duplicate wells in one experiment. No difference from cultures treated with only TNF-a without IFN-y w a s  found b  y ANOVA.  152  CO  o  O  ON  O  o  CO  o o  3  3 3  U» i-* O O O w  t-rj y  C ° m^m^^m^ma  o H  p  o  r-t-  P  i-t  <•*•  P  P O CD  o  n  to  % cells labeled for ICAM-1  c/3  O  e  Figure 25.  Reversibility of ICAM-1 expression on HBMEC induced by incubation with LPS for 2d then 3d without LPS • , with LPS for 2d SI, with LPS for 5 days a , or without LPS a . Values represent mean ± standard error of results from three experiments. No significant difference was found by statistics.  154  Ox  en  o T t,\\\\S\\\NH  _  o  00  mmmwmmmggfe*  H - . h - . ^ - t—-  iT  GO 00  00  °* •dr *a rc r - Pd  fj- tsrtr gf sr  Oi t o W bJ  £  MM „ CD  t  05,5 /////////////////////////////////7///A B  00~  o£  o  o\  ~k///W YZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ 1  o  p o •/////////////S;//'///M ~$77?r, £?. ° O ~  >-i  CD O \////////)////)////A-  O o  o  to  o o  % cells labeled for ICAM-1  GO  o GO  in  c  Figure 26.  Reversibility of ICAM-1 expression on HBMEC induced by incubation with IFN-y for 2d then 3d without IFN-y • , with IFN-y for 2d Q , with IFN-y for 5 days Q , or without IFN-y 01 • Values represent mean ± standard error of results from three experiments. No significant difference was found by statistics.  156  Figure 26.  IGSS 80 Time ; 2d with IFN, 3d without 2d with IFN 5d with IFN untreated 0 100 500 Concentration of IFN- Y (U/ml)  157  Figure 27.  Reversibility of ICAM-1 expression on HBMEC induced by incubation with TNF-a for 2d then 3d without TNF-a Q , with TNF-a for 2d S , with TNF-a for 5 days Q , or without TNF-a 12 • Values represent mean ± standard error of results from three experiments. No significant difference was found by statistics.  158  Figure 27.  IGSS  Time 2d with TNF, 3d without 2d with TNF 5d with TNF H H untreated 0 1 10 Concentration of TNF-a  159  Figure 28.  Reversibility of ICAM-1 expression on HBMEC induced by incubation with IL-ip for 2d then 3d without IL-lf$ • , with IL-lp for 2d Q , with IL-1J3 for 5 days Q , or without IL-ip a . Values represent mean ± standard error of results from three experiments. No significant difference was found by statistics.  160  Figure 28.  IGSS  Time 2d with IL-1, 3d without 2d with IL-1 5d with IL-1 untreated Concentration of IL-lp (U/ml)  161  Figure 29.  (a) HBMEC expression of ICAM-1 without LPS or cytokines detected by immunogold EM. (b, c) HBMEC expression of ICAM-1 after treatment with 100U TNF-a/ml for 24h detected by immunogold TEM. Note the presence of gold particles at the apical surface and its finger-like projections. X38,000.  162  Figure 29.  rk,  -f  * - '-  yjb  .  -lit?" > ^3s  ... «?*#*  ••MMM  • > # »  .'*..--;;. *''*%*£&?  • V a t - ^"^fBfcw^ * '-'•'Vi'.'t, •  <" " J " * ^  ^ i ^ ^ ^ L H  •«*M*  163  Figure 30.  Immunogold electron microcopy quantitation of ICAM-1 expression on HBMEC apical £1 and basal a surfaces after incubation with 100U TNF-a/ml for 24h or left unstimulated. Values represent mean ± standard error of one hundred cells for each treatment.  164  Figure 30.  ESJApical ^Basal  Unstimulated lOOUTNF-a/ml X24h  H^| | - ,  1  i§ itl!! 1  1  Hiih 1  1  0 5 10 15 20 25 Number of gold particles / u.m cell membrane  165  Figure 31.  (a) HBMEC expression of PECAM-1 without LPS or cytokine, detected by immunogold silver staining, (b) HBMEC expression of PECAM-1 after incubation with lug LPS/ml for 24h, detected by immunogold silver staining, (c) HBMEC expression of PECAM-1 after incubation with 100U TNF-a/ml for 24h, detected by immunogold silver staining, (d) HBMEC expression of PECAM-1 after incubation with 10U IL-lp/ml for 24h, detected by immunogold silver staining. X183.  166  Figure 31.  167  Figure 32.  Quantitation of PECAM-1 expression on HBMEC by ELISA after incubation with LPS for 48 EJ, 24 a , 12 a , 4 S , or Oh • . Values represent mean ± standard error of duplicate wells from three experiments. No statistically significant difference was found between unstimulated (Oh) and treated cultures.  168  Figure 32.  0  0.01 0.1 1 Concentration of LPS (\xg/ml)  169  Figure 33.  Quantitation of PECAM-1 expression on HBMEC by ELISA after incubation with TNF-a for 48 8 , 24 Q , 12 Q , 4 SI, or Oh • . Values represent mean ± standard error of duplicate wells from three experiments. No statistically significant difference was found between unstimulated (Oh) and treated cultures.  170  Figure 33.  a o ON  a  o a cd  o GO  0  1 10 100 Concentration of TNF-a(U/ml)  171  Figure 34.  Quantitation of PECAM-1 expression on HBMEC by ELISA after incubation with IL-10 for 48 (SI, 24 a , 12 a , 4 S , or Oh • . Values represent mean ± standard error of duplicate wells from three experiments. No statistically significant difference was found between unstimulated (Oh) and treated cultures.  172  Figure 34.  Concentration of IL-1(3 (U/ml)  173  Figure 35.  Quantitation of PECAM-1 expression on HBMEC by ELISA after incubation with IFN-y for 48 a , 24 a , 12 a , 413, or Oh • . Values represent mean ± standard error of duplicate wells from three experiments. No statistically significant difference was found between unstimulated (Oh) and treated cultures.  174  3  .SP to  Tt-  m  U  XJ J 3 X5  (Sj ^t-  TH  H  .a o ^ t  H\\\NSS5 «)»»»»»»»»»»»,.  rH  O  o  in  o  o  ^vyyyyyyyyyyy/yyyyyyyAi n H CM  a 5^ M  O  a o aj  •>-H -t—>  4->  a o a o U  Figure 36.  Expression of PECAM-1 on unstimulated HBMEC (a-c). PECAM-1 is expressed on the apical surface of HBMEC and to a much lesser extent on the basal surface (a). Gold particles are concentrated on flat surfaces rather than on finger-like cytoplasmic projections (b). Gold particles are present on the surface immediately adjacent to the intercellular contacts but are absent along intercellular contacts and tight junctions (arrows) (c). a-b, X53,000; c, X67,000.  176  Figure 36.  i  • *ap l Kg -  .ri  m  Figure 37.  Expression of PECAM-1 on unstimulated HBMEC (a-d). PECAM-1 is especially concentrated at the periphery of the EC. In these areas, prominent labeling at the basal surface can be seen (a, b). In subconfluent areas of the monolayers, the advancing thin process of EC are heavily decorated with gold particles (c, d). Labeling on HBMEC activated with 500U IFN-y/ml for 48h (e) is not significantly different from unstimulated EC. X53,000.  Figure 37.  '.'.'.'• "  179  *K  ^  .i  Figure 38.  Immunogold electron microcopy quantitation of PECAM-1 expression on HBMEC apical S and basal Q surfaces after incubation with 500U IFN-y/ml for 48h or left unstimulated. Values represent mean ± standard error of one hundred cells for each treatment.  180  Figure 38.  Unstimulated 500UIFN-Y/ml X48h  0 1 2 3 Number of gold particles / jim cell membrane  181  Figure 39.  Scanning electron micrographs of PMN adhesion to (a) unstimulated and (b) TNF-treated (lOOU/ml for 24h) HBMEC. No adhesion molecule blocking Ab was used. X670.  182  Figure 39.  183  Figure 40.  Light micrographs of PMN adhesion to unstimulated (a) and TNF-treated (lOOU/ml) HBMEC for 4h (b), 18h (c), 24h (d). No adhesion molecule blocking Ab was used. PMN were stained for leukocyte common antigen using the immunoperoxidase method. HBMEC were counter-stained with haemotoxylin. X630.  184  Figure 40.  .  i  a  4 0  185  Figure 41.  Light microscopy quantitation of PMN adhesion to TNF-a activated and untreated HBMEC. Values represent means±SEM of two experiments, each performed in triplicate wells.  Significant (p<0.05) increase in adhesion  compared to untreated HBMEC.  186  CD  Ow  r^Z2^H** ^ ^ H .  zalh  K»  CD 4^-  3  '//////////////////A—< i IIII i IIIII i  p  *  *  *  r r r i i r i 111 ITH  H  H  1  o  D5 w  3'  I—•  p  o  3" W tO H* o o o oc 3 B B a*  iL* o CD 5S '/////////Art 3rr * urn-  W  3 O  o  I—*.  r-t-  Pi  »-i  c  O  o  to  1  4± O  No. of PMNs adherent/mm 2  e  Figure 42.  Effects of adhesion molecule blocking Ab treatment on PMN adhesion to TNF-a treated and untreated HBMEC. PMN were incubated for lOmin with HBMEC. Values represent means±SEM of two experiments, each performed ale  in triplicate wells. Significant (p<0.05) decrease in adhesion compared to HBMEC not treated with Abs.  188  Figure 42.  40 35 30 25 20 15 10 5 0  | | NoAb KSNN Anti-E-selectin F ^ Anti-VCAM-l V7A Anti-ICAM-l 15551 Anti-PECAM-l  I  <L8:  I  ;24-:  Duration of EC treatment (h)  189  Figure 43.  Effects of adhesion molecule blocking Ab treatment on PMN adhesion to TNF-a treated and untreated HBMEC. PMN were incubated for 20min with HBMEC. Values represent means±SEM of two experiments, each performed in triplicate wells. Significant (p<0.05) decrease in adhesion compared to HBMEC not treated with Abs.  190  Figure 43.  40 35 30 25 20 15 10 5 0  | | NoAb |^<S^ Anti-E-selectin ( = ] Anti-VCAM-l 77A Anti-ICAM-l 15551 Anti-PECAM-l  I II  I  £1E l  I  =— 0 — > < 4 > -<8>- < 2 4 ^  Duration of EC treatment (h)  191  Figure 44.  Effects of adhesion molecule blocking Ab treatment on PMN adhesion to TNF-a treated and untreated HBMEC. PMN were incubated for 30min with HBMEC. Values represent means±SEM of two experiments, each performed in triplicate wells. Significant (p<0.05) decrease in adhesion compared to HBMEC not treated with Abs.  192  Figure 44.  40 35 30 25 20 15 10 5 0  | NoAb 1CSSM Anti-E-selectin f = | Anti-VCAM-1 Y7A Anti-ICAM-l ®$® Anti-PECAM-l  i x  I  1H £ 0  <18:  j  ;24^  Duration of EC treatment (h)  193  Figure 45.  Light micrographs of PMN migration across TNF-a treated and untreated HBMEC on lum thick plastic sections stained with toluidine blue, (a) Unstimulated EC, (b) 4h TNF, (c) 18h TNF, (d) 24h TNF. X980.  194  Figure 45.  195  Figure 46.  Quantitation, by light microscopy, of PMN migration across TNF-a activated HBMEC. Values represent means±SEM of 120 sections per treatment of two experiment.  Significant (p<0.05) increase in the number  of migrated PMN compared to untreated HBMEC.  196  Figure 46.  E  "  *  J10 T3  T  <D  -4—»  2  8  6 Z  6  bJO • 1-H  S 4 O  •  o 55  * *  T  T  2^  * ' ~ '  0  4  18  24  ^ p  Duration of EC treatment (h)  197  Figure 47.  Quantitation, by light microscopy, of PMN migration across TNF-a activated HBMEC treated with anti-adhesion molecule Abs following 2h incubation with EC. Values represent means±SEM of greater than 100 sections per treatment of two experiment.  Significant (p<0.05)  decrease in the number of migrated PMN compared to EC not treated with Abs.  198  Figure 47.  3.5 3.0 h  2.5  | |NoAb |^\H Anti-E-selectin r 3 Anti-VCAM-l Y77X Anti-ICAM-1 £555] Anti-PECAM-l  I I  2.0 1.5 1.0 0.5 0.0  fimrii?i^ 0 — > < 4 > <L8> -e24^  Duration of EC treatment (h)  199  Figure 48.  Scanning electron micrographs of PMN interactions with HBMEC during transendothelial migration. PMN cluster at the contact points between EC (a, X5,100), then squeeze into a flattened shape as they move between endothelial cells in order to cross the endothelium (b, X5,100; c, X3,200).  200  Figure 48.  201  Figure 49.  Transmission electron micrographs of PMN interactions with HBMEC during transendothelial migration. PMN migrate at the periphery of the cells, close to or at tight junctions. X21,000.  202  Figure 49.  203  Figure 50.  Transmission electron micrographs of PMN interactions with HBMEC during transendothelial migration, (a) PMN that have migrated under the EC monolayer appear slightly flattened. The overlying monolayer remains intact. X3,700. (b) Close contacts are present at several points between migrated PMN and HBMEC Note the intact tight junction (arrow heads) above the migrated PMN. X15,000.  204  Figure 50.  W„  205  Figure 51.  Scanning electron micrographs of T lymphocyte adhesion to unstimulated (a) or activated HBMEC, 100U TNF-a/ml for 4h (b), 18h (c), 24h (d). No adhesion molecule blocking Ab was used. X320.  206  Figure 51.  ^ET"!--.  a  w  fJP'  w  207  Figure 52.  Light micrographs of T lymphocyte adhesion to unstimulated (a) or activated HBMEC, 100U TNF-a/ml for 4h (b), 18h (c), 24h (d). No adhesion molecule blocking Ab was used. X550.  208  Figure 52.  X 4\  ^  • Iv.  VJ.  *  m B  "V  / •  im  •  w  209  < *  Figure 53.  Quantitation, by light microscopy, of T lymphocyte adhesion to TNF-a activated and untreated HBMEC. Values represent means±SEM of three experiments, each performed in triplicate wells.  Significant (p<0.05)  increase in adhesion compared to untreated HBMEC.  210  Figure 53.  (N  800  T cell incubation time 20min *  600 - gS3 30min 60min f^ *  * Hi  x»  400 § §  200 0  11  § ^  0  N  4  18  24  Duration of EC treatment (h)  211  Figure 54.  Effects of adhesion molecule blocking Ab treatment on T lymphocyte adhesion to TNF-a treated and untreated HBMEC. T lymphocytes were incubated for 20min on HBMEC. Values represent means±SEM of three experiments, each performed in triplicate wells. Significant (p<0.05) decrease in adhesion compared to HBMEC not treated with Abs.  212  Figure 54.  2000 1600 h 1200  NoAb l^vSH Anti-E-selectin F ^ Anti-VCAM-1 Y77X Anti-ICAM-l K%$ Anti-PECAM-l  800 h 400  0  n^R 0  jig  1  3- <4:^ -<8:  Duration of EC treatment (h)  213  Figure 55.  Effects of adhesion molecule blocking Ab treatment on T lymphocyte adhesion to TNF-a treated and untreated HBMEC. T lymphocytes were incubated for 30min on HBMEC. Values represent means±SEM of three experiments, each performed in triplicate wells. Significant (p<0.05) decrease in adhesion compared to HBMEC not treated with Abs.  214  Figure 55.  2000 NoAb  1600 1200 -  t^sSS Anti-E-selectin F = j Anti-VCAM-l 777X Anti-ICAM-l Igggj Anti-PECAM-l  800 400 0  X nNprnlo.  0  -> < 4 > <L8^ <24^  Duration of EC treatment (h)  215  Figure 56.  Effects of adhesion molecule blocking Ab treatment on T lymphocyte adhesion to TNF-a treated and untreated HBMEC. T lymphocytes were incubated for 60min on HBMEC. Values represent means±SEM of three experiments, each performed in triplicate wells. Significant (p<0.05) decrease in adhesion compared to HBMEC not treated with Abs.  216  Figure 56.  (N  «  2000 1600  M CD  1200 o "4H  o o  IXsS 1=1 [2ZI Igggl  NoAb Anti-E-selectin Anti-VCAM-1 Anti-ICAM-l Anti-PECAM-l  800  xft I*  I  400 0  X  X  <18:  -=24-:  Duration of EC treatment (h)  217  Figure 57.  Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell. Adherent T lymphocyte appear as a small roughly spherical cell. X3,400.  218  Figure 57.  219  Figure 58.  Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell. Adherent T lymphocyte appear as a small roughly spherical cell that established contact to EC by extending pseudopodia to the apical surface of EC. X18,000.  220  Figure 58.  221  Figure 59.  Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell. Adherent T lymphocyte appear as a small roughly spherical cell that established contact to EC by extending pseudopodia to the apical surface of EC. X22,000.  222  Figure 59.  ^Sf^ J * i 'x l  223  Figure 60.  Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell. Adherent T lymphocyte appear as a small semi-spherical cell. X18,000.  224  Figure 60.  225  Figure 61.  Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell. Adherent T lymphocyte appear as a small semi-spherical cell. X25,000.  226  Figure 61.  J ^ ^ ^ ^ M ^  m  A  i^-^.Jtffek  7 f  M'f: '• ... p*  i  j?gt -.in0  •.'«  •V^V;'..\ *fc •• - - » • - - ;• -'" • *  H^H^B*Hj^H^HEfc<. . U .  - -  . •»  '  "  •  .^3IP3B3B ••••*  :  •  *t'^:>'-"_'  227  '•V"*''«J«  ••  ^^V&MI  .-  • : 'T - T  ••••*;»  d  • \\j.*J  ^^^9E  r'*fojj$ S*<Ki^a  m  Figure 62.  Transmission electron micrograph of T lymphocyte adhering to a human brain microvessel endothelial cell. Adherent T lymphocyte form close contacts with the apical surface of HBMEC. X35,000.  228  Figure 62.  l>  '  f 'yi^m^SaP^^^%  ^»fc TgXe ^H  **  JR^I H  IKr ^ i • tv^Mh  1  |Bx.-'  1  ^ Mb--*-  ,/**  :  ^*wz  "••*•'.•-  1  :  •-•'*'-v,-^ ? •  ' gssg^^v'.*  %  .  P^ v  ^ ^ ^ ^ 1  >  * wm W^mf^^n^'^'t^^i^^/^  i - ^H  i-.-""*5*  229  Figure 63.  Light micrographs of T lymphocyte migration across TNF-a treated and untreated HBMEC. These ljim sections were stained with toluidine blue, (a) Unstimulated EC, (b) 4h TNF, (c) 18h TNF, (d) 24h TNF. Xl,500.  230  Figure 63.  231  Figure 64.  Quantitation, by light microscopy, of T lymphocyte migration across TNF-a activated HBMEC. Values represent means±SEM of greater than 100 sections per treatment of one experiment.  Significant (p<0.05)  increase in the number of migrated T lymphocytes compared to untreated HBMEC.  232  Figure 64.  1.2 1.0 0.8  T cell incubation time 30min  I  S§3 60min 180min  0.6 0.4 0.2 0.0  rH^Ti  0  ftfir^tf 4  18  24  Duration of EC treatment (h)  233  Figure 65.  Quantitation, by light microscopy, of T lymphocyte migration for 30min across TNF-a activated HBMEC treated with anti-adhesion molecule Abs. Values represent means±SEM of greater than 100 sections per treatment of one experiment.  234  Figure 65.  B 1.2  a  T3 <L>  1.0  +-> <T| i-i  •T—1  a </i  0.8  "  | | No Ab R s ^ Anti-E-selectin | 1 Anti-VCAM-l Y//A Anti-ICAM-l |$$$ Anti-PECAM-1  0.6  r—H  (U  u  H  0.4  <4-H  O  o  2  0.2 0.0 ^n«lFlff.l5l — 0 —s-  flfi  nn  nis  ^43- <L8> - < 2 4 ^  Duration of EC treatment (h)  235  Figure 66.  Quantitation, by light microscopy, of T lymphocyte migration for 60min across TNF-a activated HBMEC treated with anti-adhesion molecule Abs. Values represent means±SEM of greater than 100 sections per treatment of one experiment.  236  Figure 67.  Quantitation, by light microscopy, of T lymphocyte migration for 180min across TNF-a activated HBMEC treated with anti-adhesion molecule Abs. Values represent means±SEM of greater than 100 sections per treatment of one experiment.  Significant (p<0.05) decrease in the  number of migrated T lymphocytes compared to EC not treated with Abs.  238  Figure 61.  1.2 1.0 0.8 CO i—i  NoAb £\S^ Anti-E-selectin Anti-VCAM-1 777\ Anti-ICAM-l [5551 Anti-PECAM-l  0.6  <D  o H o o  0.4 0.2  *  X  0.0 nflflflifi ;24^ 0 ^ ^ 4 > <18: Duration of EC treatment (h)  239  Figure 68.  Scanning electron micrographs of T lymphocyte interactions with HBMEC during transendothelial migration. T lymphocytes gather at the periphery of EC before migration (a, X2,000). Most T lymphocytes squeeze into a flattened shape between endothelial cells in order to cross the endothelium (b, X6,800), others move through the cytoplasm of endothelial cells (c, X5,300).  240  Figure 68.  241  Figure 69.  Transmission electron micrographs of T lymphocyte interaction with HBMEC. Foot-like processes are inserted by the lymphocyte into the endothelial cytoplasm. X47,000.  242  Figure 69.  ?S*5* ^ i?  243  Figure 70.  Transmission electron micrographs of T lymphocyte interaction with HBMEC. Foot-like processes are inserted by the lymphocyte into the endothelial cytoplasm. X27,000.  244  Figure 70.  :  -*P?  n^pr  •J. * « -  4 '»  245  *.  Figure 71.  Transmission electron micrographs of T lymphocyte interaction with HBMEC. Foot-like processes are inserted by the lymphocyte into the endothelial cytoplasm. X43,000.  246  Figure 71.  r^  JI0  T  247  &  Figure 72.  Transmission electron micrographs of T lymphocyte migration through HBMEC. A cytoplasmic "opening" or gap in the HBMEC progressively enlarges to allow the lymphocyte to pass through. X13,000.  248  Figure 72.  <w  > * > v  ' ' «V-'  :  M>#  &wfcs  ^fN*  249  Figure 73.  Transmission electron micrographs of T lymphocyte migrated underneath the HBMEC. The lymphocyte become elongated and remain between the underlying collagen membrane and the overlying EC, maintaining close contact with the latter. X13,000.  250  Figure 73.  ; ^^^Wf  ( i ^  251  could be found throughout the surface of the HBMEC but usually aggregated at the areas of contact between EC (Fig. 39).  3.3.2 Polymorphonuclear leukocyte (PMN) migration Transendothelial migration of PMN across unstimulated HBMEC was minimal after 2h of incubation (Figs. 45-46). Pretreatment of EC with TNF-a increased the number of migrated PMN 2 to 10 fold (P<0.001 by ANOVA). The largest number of PMN under the monolayers was observed in cultures treated with TNF for 24h (Figs. 45-46). The increase in adhesion by TNF was not as dramatic as the one obtained in response to chemotactic gradients of 10"'M fMLP placed under the unstimulated monolayers when a 30 fold increase in migration was observed (Fig. 46). MAb had no effect on PMN migration across unstimulated HBMEC (Fig. 47). However, Ab treatment decreased migration across activated EC (P<0.001 by ANOVA). Preincubation with anti-ICAM-1 mAb significantly decreased the numbers of PMN migrating across TNF-treated cultures by 75% (Fig. 47). Other mAbs did not significantly block transmigration (Fig. 47). Thus, the anti-PECAM-1 Ab blocked adhesion by a modest 15% and anti-VCAM-1 by 4% only. MAb against E-selectin had no effect on PMN migration. Observations with the SEM demonstrated that, following adhesion, PMN positioned themselves between adjacent EC (Fig. 48a) and eventually proceeded to migrate by moving between EC. During this process, PMN became elongated and flattened in their attempt to "squeeze" between the closely associated HBMEC (Fig. 48b, c). Examination by TEM revealed that PMN migrated between thin processes of adjacent HBMEC, possibly at the site of tight junctions (Fig. 49). PMN that had migrated under the HBMEC monolayer appeared flattened and remained between overlying EC and the underlying collagen membrane in close contact with the EC at several points (Fig. 50). The overlying monolayers showed no signs of disruption and 53  et al., 1991; Hakkert et al., 1991). Much higher rates of 70% (Smith et al., 1988) and 90% (Kishimoto et al., 1991) have also been reported with similar activation regimes. Cytokine treatment of EC has consistently been found to increase PMN adhesion to HUVEC and a mouse brain endothelioma cell line. Treatment with TNF, IL-1 or LPS, seems to upregulate adhesion to a similar extent. TNF treatment of HBMEC upregulates PMN adhesion by 81-683%. Two to four hours of HUVEC activation increases PMN adhesion to a much greater extent than in HBMEC (Wankowicz et al., 1988; Luscinskas et al., 1989; Luscinskas et al., 1991). The large range of reported values is due to the variable duration of PMN incubation with the EC and perhaps dissimilar culture conditions. These differences may also account, to some extent, for the differences between HBMEC and HUVEC in the literature. One report indicated that 24h treatment of HUVEC increased PMN adhesion by 400%, only half of the increase following 4h of activation (Luscinskas et al., 1989). This increase is in line with the 443 to 683% found after 24h treatment of HBMEC with TNF. Treatment of a mouse brain endothelioma line with TNF raised PMN adhesion by 500 to 1000% (Hahne et al., 1993), which is considerably higher than the one observed in HBMEC. None of the blocking Abs used affected PMN adhesion to unstimulated HBMEC. In the only other similar study published, Smith et al. (1988) were able to inhibit PMN adhesion to unstimulated HUVEC with an anti-ICAM-1 blocking Ab by 70%. Their anti-ICAM-1 Ab was different from the one used in this study and the percentage of PMN adherent to HUVEC was 500 times greater than in our study. A possible explanation for this difference could be that an epitope, not blocked by the mAb used in the present study, may be involved in this binding. The anti-E-selectin mAb significantly blocked PMN adhesion to TNF treated HBMEC by 38 to 54%. Luscinskas et al. (1991), working with HUVEC and the same anti-E-selectin mAb as the one we used, found a similar rate of inhibition (50%). However, Graber et al. (1990), using the same mAb on activated HUVEC, found a 80 to 77  Figure 66.  1.2 ^  1.0  1 0.8 I 0.6  NoAb ^ \ S Anti-E-selectin Ff~j Anti-VCAM-1 Y77X Anti-ICAM-l R5551 Anti-PECAM-1  •1—I  0.4 0.2 0.0  f\ r^i F i ry, a  QJS  nfl  d  =— 0 — > • < 4 > <L8^ ^ 2 4 ^  Duration of EC treatment (h)  237  Table 1.  Effects of TNF treatment of HBMEC on the permeability of the monolayer determined by electrical resistance measurements across the monolayers.  *  Significant  (p<0.05) decrease in resistance compared to EC not treated with TNF.  HBMEC treatment  Resistance (Q-crn^)  No TNF  108 ± 14  100UTNF/ml,4h  98 ± 1  lOOUTNF/ml, 18h  68* ± 9  100UTNF/ml,24h  63 ± 4  *  *  252  Table 2.  Effect of PMN migration with and without adhesion molecule blocking antibodies on the permeability of HBMEC cultures determined by electrical resistance measurements across the monolayers.  % change in resistance HBMEC treatment  after PMN migration  No TNF, no Ab, no fMLP  -7  No TNF, no Ab, 10"7M fMLP  -56  4h TNF, no Ab  -26  4h TNF, a-E-selectin  -44  18hTNF,noAb  -41  18hTNF, a-VCAM  -50  24h TNF, no Ab  -43  24h TNF, a-ICAM-1  -73  24h TNF, a-PECAM  -44  253  Table 3.  Effect of T lymphocyte migration with and without adhesion molecule blocking antibodies on the permeability of HBMEC cultures determined by electrical resistance measurements across the monolayers.  % change in resistance after T lymphocyte migration  HBMEC treatment  4h TNF, no Ab  -25 ± 10  4h TNF, a-E-selectin  -69 ± 10  18h TNF, no Ab  -24 ± 3  18h TNF, a-VCAM  -25 ± 23  24h TNF, no Ab  -26 ± 10  24h TNF, a-ICAM-1  -32 ± 12  24h TNF, a-PECAM  -24 ± 7  254  REFERENCES Abbas AK, Lichtman AH and Pober JS. Cellular and Molecular Immunology. W.B. Saunders Co., Philadelphia. 1991. Abbassi O, Kishimoto TK, Mclntire LV, Anderson DC, Smith CW. E-selectin supports neutrophil rolling in vitro under conditions of flow. J. Clin. Invest. 92:2719-2730, 1993. Abbott NJ and Revest PA. Control of brain permeability. Cerebrovasc. Brain Metab. Rev. 3:39-72, 1991. Adams RB, Planchon SM and Roche JK. IFN-y modulation of epithelial barrier function. J. Immunol. 150:2356-2363,1993. Airas L, Salmi M and Jalkanen S. Lymphocyte adhesion protein-2 is a novel 70-kDa molecule involved in lymphocyte adhesion to vascular endothelium. J. Immunol. 151:4228-4238, 1993. Akiyama H, Kawamata T, Yamada T, Tooyama I, Ishii T and McGeer P. Expression of intercellular adhesion molecule (ICAM)-l by a subset of astrocytes in Alzheimer disease and some other neurological disorders. Acta Neuropathol. 85:628-634, 1993. Albelda SM, Sampson PM, Haselton FR, McNiff JM, Mueller SN, Williams SK, Fishman AP and Levine EM. Permeability characteristics of cultured endothelial ell monolayers. J. Appl. Physiol. 64:308-322,1988. Albelda SM and Buck CA. Integrins and other cell adhesion molecules. FASEB J. 4:2868-2880,1990a. Albelda SM, Oliver PD, Rower LH and Buck CA. EndoCAM: a novel endothelial cell-cell adhesion molecule. J. Cell Biol. 110:1227-1237,1990b. Albelda SM. Endothelial and epithelial cell adhesion molecules. Am. J. Respir. Cell. Mol. Biol. 4:195-205,1991. Albelda SM, Muller WA, Buck CA and Newman PJ. Molecular and cellular properties of PECAM-1 (endoCAM/CD31): a novel vascular cell-cell adhesion molecule. J. Cell Biol. 114:1059-1068,1991. Allison F Jr, Smith MR and Wood WB. Studies on the pathogenesis of acute inflammation. J. Exp. Med. 102:655-668,1955. Anderson DL and Springer TA. Leukocyte adhesion deficiency: an inherited defect in the Mac-1, LFA-1 andpl50,95 glycoproteins. Ann. Rev. Med. 38:175-194,1987. Arbones ML, Ord DC, Ley K, Ratech H, Maynardcurry C, Otten G, Capon DJ and Tedder TF. Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin deficient mice. Immunity 1:247-260, 1994.  255  Archelos JJ, Jung S, Maurer M, Schmied M, Lassmann H, Tamatani T, Miyasaka M, Toyka KV and Hartung H-P. Inhibition of experimental autoimmune encephalomyelitis by an antibody to the intercellular adhesion molecule ICAM-1. Ann. Neurol. 34:145-154,1993. Argenbright LW, Letts LG and Rothlein R. Monoclonal antibodies to the leukocyte membrane CD 18 glycoprotein complex and to intercellular adhesion molecule-1 inhibit leukocyte-endothelial adhesion in rabbits. J. Leukocyte Biol. 49:253-257, 1991. Arnaout MA, Lanier LL and Faller DV. Relative contribution of the leukocyte molecules Mo-1,1FA-1 and pl50,95 (Leu M5) in adhesion of granulocyte and monocyte to vascular endothelium is tissue and stimulus-specific. J. Cell. Physiol. 137:305-309, 1988. Azzali G. The passage of macrophages and lymphocytes from the interstitium across the lymphatic endothelium of rat lacteals. Cell Tissue Res. 262:191-193,1990. Azzali G, Orlandini G and Gatti R. The migration of lymphocytes and polymorphonuclear leukocytes across the endothelial wall of the absorbing peripheral lymphatic vessel. J. Submicrosc. Cytol. Pathol. 22:543-549,1990. Bacchi CE, Marsh CL, Perkins JD, Carithers RL Jr, McVicar JP, Hudkins KL, Benjamin CD, Harlan JM, Lobb R and Alpers CE. Expression of vascular cell adhesion molecule (VCAM-1) in liver and pancreas allograft rejection. Am. J. Path. 142:579-591,1993. Banks WA and Kastin AJ. The interleukin-la, -lp\ and -2 do not acutely disrupt the murine blood-brain barrier. Int. J. Immunopharmac. 14:629-636,1992. Bargatze RF and Butcher EC. Rapid G protein-regulated activation event involved in lymphocyte binding to high endothelial venules. J. Exp. Med. 178:367-372,1993. Baron JL, Madri JA, Ruddle NH, Hasim G and Janeway CA. Jr Surface expression of alpha 4 integrin by CD4 T cells is required for their entry into brain parenchyma. J. Exp. Med. 177:57-68, 1993. Barten DM and Ruddle NH. Vascular cell adhesion molecule-1 modulation by tumor necrosis factor in experimental allergic encephalomyelitis. J. Neuroimmunol. 51:123-133,1994. Bauer J, Berkenbosch F, Van Dam A-M and Dijksta CD. Demonstration of interleukin-ip in Lewis rat brain during experimental allergic encephalomyelitis by immunochemistry at the light and ultrastructural levels. J. Neuroimmunol. 48:13-22, 1993. Beesley JE, Pearson JD, Hutchings A, Carleton JS and Gordon JL. Granulocyte migration through endothelium in culture. J. Cell. Sci. 38:237-248,1979. Bender JR, Pardi R, Karasek MA and Engelman EG. Phenotypic and functional characterization of lymphocytes that bind human microvascular endothelial cells in vitro. J. CLin. Invest. 79:1679-1688,1987.  256  Benjamin C, Dougas I, Chi-Rosso G, Luhowskyj S, Rosa M, Newman B, Osborn L, Vassallo C, Hession C, Goelz S, McCarthy K and Lobb R. A blocking monoclonal antibody to endothelial-leukocyte adhesion molecule-1 (ELAM1). Biochem. Biophys. Res. Commun. 171:348-353,1990. Bennett CF, Condon TP, Grimm S, Chan H and Chiang M-Y. Inhibition of endothelial cell adhesion molecule expression with antisense oligonucleotides. J. Immunol. 152:3530-3540,1994. Benveniste EN. Inflammatory cytokines within the central nervous system: sources, functions and mechanism of action. Ann. J. Physiol. 263:C1-C16,1992. Berendt AR, Simmons DL, Tansey J, Newbold CI and Marsh K. Intercellular adhesion molecule-1 is an endothelial cell adhesion receptor for Plasmodium falciparum. Nature 341:57-59,1989. Berendt AR, NcDowall A, Craig AG, Bates PA, Sternberg MJE, Marsh K, Newbold CI and Hogg N. The binding site of ICAM-1 for Plasmodium falciparum - infected erythrocytes overlaps, but is distinct from, the LFA-1 binding site. Cell 68:71-81, 1992. Bereta J, Bereta M, Cohen S, Cohen MC. Regulation of VCAM-1 expression and involvement in cell adhesion to murine microvascular endothelium. Cell. Immunol. 147:313-330,1993. Berg EL, Yoshino T, Rott LS, Robinson MK, Warnock RA, Kishimoto TK, Picker LJ and Butcher EC. The cutaneous lymphocyte antigen is a skin lymphocyte homing receptor for the vascular lectin endothelial cell-leukocyte adhesion molecule-1. J. Exp. Med. 174:1461-1466,1991. Berman ME and Muller WA. Ligation of platelet/endothelial cell adhesion molecule 1 (PECAM-1/CD31) on monocytes and neutrophils in increase binding capacity of leukocyte CR3 (CDllb/CD18). J. Immunol. 154:299-307,1995. Bevilacqua MP, Pober JS, Wheeler ME, Cotran RS and Gimbrone MA Jr. Interleukin 1 acts on cultured vascular endothelium to increase the adhesion of polymorphonuclear leukocytes, monocytes and related leukcoyte cell lines. J. Clin. Invest. 76:2003-2011,1985. Bevilacqua MP, Pober JS, Mendrick DL, Cotran RS and Gimbrone MA Jr. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc. Natl. Acad. Sci. USA. 84:9238-9242, 1987. Bevilacqua MP, Stengelin S, Gimbrone Jr MA and Seed B. Endothelial leukocyte adhesion molecule 1: an inducible receptor for neutrophils related to complement regulatory proteins and lectins. Science 243:1160-1165,1989. Bevilacqua MP. Endothelial-leukocyte adhesion molecules. Annu. Rev. Immunol. 11:767-804,1993. Bevilacqua MP and Nelson RM. Selectins. J. Clin. Invest. 91:379-387,1993.  257  Bianchi G, Sironi M, Ghibaudi E, Selvaggini C, Elices M, Allavena P and Matovani A. Migration of natural killer cells across endothelial cell monolayers. J. Immunol. 151:5135-5144,1993. Bird IN, Spragg JH, Ager A, Matthews N. Studies of lymphocyte transendothelial migration: analysis of migrated cell phenotypes with regard to CD31 (PECAM-1), CD45RA and CD45RO. Immunol. 80:553-560,1993. Bochner BS, Luscinskas FW, Gimbrone Jr MA, Newman W, Sterbinsky SA, Derse-Anthony CP, Klunk D and Schleimer RP. Adhesion of human basophils, eosinophils, and neutrophils to interleukin 1-activated human vascular endothelial cells: contributions of endothelial cell adhesion molecules. J. Exp. Med. 173:1553-1556,1991. Bogen SA, Baldwin HS, Watkins SC, Albelda SM and Abbas AK. Association of murine CD31 with transmigrating lymphocytes following antigenic stimulation. Am. J. Path. 141:843-854,1992. Bogen S, Pak J, Garifallon M, Deng X and Muller WA. Monoclonal antibody to murine PECAM-1 (CD31) blocks acute inflammation in vivo. J. Exp. Med. 179:1059-1064,1994. Bos JD, deBoer OJ, Tibosch E, Das PK and Pals ST. Skin-homing T lymphocytes: detection of cutaneous lymphocyte-associated antigen (CLA) by HECA-452 in normal human skin. Arch. Dermatol. Res. 285:179-183,1993. Bowes MP, Zivin JA and Rothlein R. Monoclonal antibody to the ICAM-1 adhesion site reduces neurological damage in a rabbit cerebral embolism stroke model. Exp. Neurol. 119:215-219,1993. Bowman PD, duBois M, Shivers RR, Dorovini-Zis K Endothelial tight junctions. In Tight junctions. Cereijido M. [ed.] CRC Press, Boca Raton, 1992. Bradbury MWB. The blood-brain barrier. Exp. Physiol. 78:453-472,1993. Brandtzaeg P, Kierulf P, Gaustad P, Skulberg A, Bruun JN, Halvorsen S and Sorensen E. Plasma endotoxin as a predictor of multiple organ failure and death in systemic meningococcal disease. J. Infect. Dis. 159:195-204,1989. Brierley JB and Graham DI. Hypoxia and vascular disorders of the central nervous system. In Greenfield's neuropathology. Adams JH, Corsellis JAN and Duchen LW. [Eds.] John Wiley and Sons, New York, 4th edition. 1984. Brightman MW and Reese TS. Junctions between intimately apposed cell membranes in the vertebrate brain. J. Cell. Biol. 40:648-677,1969. Brightman MW. Morphology of blood-brain interfaces. Exp. Eye Res. 25 Suppl.:l-25, 1977. Briscoe DM, Schoen FJ, Rice GE, Bevilacqua MP, Ganz P and Pober JS. Induced expression of endothelial-leukocyte adhesion molecules in human cardiac allografts. Transplant. 51:537-547,1991.  258  Briscoe DM, Cotran RS and Pober JS. Effects of tumor necrosis factor, lipopolysaccharide, and IL-4 on the expression of vascular cell adhesion molecule-1 in vivo. Correlation with CD3+ T cell infiltration. J. Immunol. 149:2954-2960,1992. Brosnan CF, Selmej K and Raine CS. Hypothesis: A role of tumor necrosis factor in immune-mediated demyelination and its relevance to multiple sclerosis. J. Neuroimmunol. 18:87-94,1988. Brown KA, Perry ME, Mustapha Y, Rothlein R and Dumonde DC. Immuno-electron microscopic analysis of the distribution of ICAM-1 in human inflammatory tissue. Agents Actions 38:C35-C38,1993. Brownwell B and Tomlinson AH. Viral diseases of the central nervous system. In Greenfield's neuropathology. Adams JH, Corsellis JAN and Duchen LW. [Eds.] John Wiley and Sons, New York, 4th edition. 1984. Burke-Gaffney A and Keenan AK. Modulation of human endothelial cell permeability by combinations of the cytokines interleukin-la/p, tumor necrosis factor-a and interferon-y. Immunopharm. 25:1-9,1993. Butcher, EC. Leukocyte-endothelial cell adhesion as an active, multi-step process: a combinatorial mechanism for specificity and diversity in leukocyte targeting. Adv. Exp. Med. Biol. 323:181-194,1992. Campbell IL, Hobbs MV, Kemper P and Oldstone BA. Cerebral expression of multiple cytokine genes in mice with lymphocytic choriomeningitis. J. IMmunol. 152:716-723,1994. Cannella B, Cross AH and Raine CS. Upregulation and coexpression of adhesion molecules correlate with relapsing autoimmune demyelination in the central nervous system. J. Exp. Med. 172:1521-1524,1990. Cannella B, Cross AH and Raine CS. Adhesion-related molecules in the central nervous system. Up-regulation correlates with inflammatory cell influx during relapsing experimental autoimmune encephalomyelitis. Lab. Invest. 65:23-31, 1991. Cannella B, Cross AH and Raine CS. Anti-adhesion molecule therapy in experimental autoimmune encephalomyelitis. J. Neuroimmunol. 46:43-56,1993. Carlos TM, Schwartz BR, Kovach NL, Yee E, Rosso M, Osborn L, Chi-Rosso G, Newman B, Lobb R and Harlan JM. Vascular cell adhesion molecule-1 mediates lymphocyte adherence to cytokine-activated cultured human endothelial cells. Blood. 76:965-970, 1990. Carlos T, Kovach N, Schwartz B, Rosa M, Newman B, Wayner E, Benjamin C, Osborn L, Lobb R and 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 77:2266-2271, 1991. Carlos TC, Gordon D, Fishbein D, Himes VE, Coday A, Ross R and Allen MD. Vascular cell adhesion molecule-1 is induced on endothelium during acute rejection in human cardiac allografts. J. Heart Lung Transplant. 11:1103-1109,1992. 259  Carlos TM and Harlan JM. Leukocyte-endothelial adhesion molecules. Blood 84:2068-2101,1994. Cavender DE, Haskard DO, Joseph B and Ziff M. Interleukin 1 increases the binding of human B and T lymphocytes to endothelial cell monolayers. J. Immunol. 136:203-207,1986. Chan BM, Elices MJ, Murphy E and Hemler ME. Adhesion to vascular cell adhesion molecule 1 and fibronectin. Comparison of alpha 4 beta 1 (VLA-4) and alpha 4 beta 7 on the human B cell line JY. J. Biol. Chem. 267:8366-8370,1992. Cho Y and DeBruyn PH. Transcellular migration of lymphocytes through the walls of the smooth-surfaced squamous endothelial venules in the lymph node: evidence of the direct entry of lymphocytes into the blood circulation of the lymph node. J. Ultrastructural Res. 74:259-266,1981. Clark WM, Madden KP, Rothlein R, Zivin JA. Reduction of central nervous system ischemic injury by monoclonal antibody to intercellular adhesion molecule. J. Neurosurg. 75:623-627,1991. Clark WM, Coull BM, Brilet DP, Mainolfi E and Rothlein R. Circulating intercellular adhesion molecule-1 levels and neutrophil adhesion in stroke. J. Neuroimmunol. 44:123-126, 1993. Claude P and Goodenough DA. Fracture faces of zonulae occuldentes from "tight" and "leaky" epithelia. J. Cell. Biol. 58:390-400,1973. Claudio L, Kress Y, Factor J and Brosnan CF. Mechanisms of edema formation in experimental autoimmune encephalomyelitis. Am. J. Path. 137:1033-1045,1990. Cohnheim J. Cohnheim's lectures on general pathology. Vol. 1. The circulation. The New Sydenham Society, London. 1890. Cotran RS, Gimbrone Jr MA, Bevilacqua MP, Mendrick DL and Pober JS. Induction and detection of a human endothelial activation antigen in vivo. J. Exp. Med. 164:661-666, 1986. Crone C and Olsen SP. Electrical resistance of brain microvascular endothelium. Brain Res. 241:49-55,1982. Cronstein BN, Kimmel SC, Levin RI, Martiniuk F and Weissmann G. A mechanism, for the antiinflammatory effects of corticosteroids: the glucocorticoid receptor regulates leukocyte adhesion to endothelial cells and expression of endothelial-leukocyte adhesion molecule 1 and intercellular adhesion molecule 1. Proc. Natl. Acad. Sci. U.S.A. 89:9991-9995,1992. Cross AH, Cannella B, Brosnan CF, Raine CS. Homing to central nervous system vasculature by antigen-specific lymphocytes. I. Localization of 14C-labeled cells during acute, chronic and relapsing experimental allergic encephalomyelitis. 63:162-170,1990. Cross AH and Raine CS. Central nervous system endothelial cell-polymorphonuclear cell interactions during autoimmune demyelination. Am. J. Path. 139:1401-1409, 1991. 260  Cybulsky MI, Fries JWU, Williams AJ, Sultan P, Davis VM, Gimbrone MA Jr. and Colins T. Alternative splicing of human VCAM-1 in activated vascular endothelium. Am. J. Path. 138:815-820,1991a. Cybulsky MI, Fries JWU, Williams AJ, Sultan P, Eddy R, Byers M, Shows T, Gimbrone MA Jr and Collins T. Gene structure, chromosomal location and basis for alternative splicing of the human VCAM1 gene. Proc. Natl. Acad. Sci. USA. 88:7859-7863, 1991b. Dalme NK and Doyle LV. Ability of human T lymphocytes to adhere to vascular endothelial cells and to augment endothelial permeability to macromolecules is linked to their state of post-thymic maturation. J. Immunol. 144:1233-1240,1990. Davis RE and Smoller BR. T lymphocytes expressing HECA-452 epitope are present in cutaneous acute graft-versus-host disease and erythema multiforme, but not in acute graft-versus-host disease in gut organs. Am. J. Path. 141:691-698,1992. DeBoer OJ, Horst E, Plas ST, Bos JD and Das PK. Functional evidence that the HECA-452 antigen is involved in the adhesion of human neutrophils and lymphocytes to tumor necrosis-factor-alpha-stimulated endothelial cells. Immunology. 81:359-365,1994. DeBruyn PPH and Cho Y. Structure and function of high endothelial postcapillary venules in lymphocyte circulation. Curr. Topics Pathol. 84:85-101,1990. Deisher TA, Haddix TL, Montgomery KF, Pohlman TH, Kaushansky K and Harlan JM. The role of protein kinase C in the induction of VCAM-1 expression on human umbilical vein endothelial cells. FEBS Lett. 331:285-290,1993. Dejana E, I Martin-Padura, D Lauri, Bernasconi S, Bani MR, Garofalo A, Giavazzi R, Magnani J, Matovani A and Menard S. Endothelial leukocyte adhesion molecule-1 -dependent adhesion of colon carcinoma cells to vascular endothelium is inhibited by an antibody to Lewis fucosylated type I carbohydrate chain. Lab. Invest. 66:324-330,1992. DeLisser HM, Newman PJ, Albelda SM. Platelet endothelial cell adhesion molecule (CD31). Curr. Topics Microbiol. Immunol. 184:37-45,1993a. DeLisser HM, Yan HC, Newman PJ, Muller WA, Buck CA and Albelda SM. Platelet / endothelial cell adhesion molecule-1 (CD31)-mediated cellular aggregation involves cell surface glycosaminoglycans. J. Biol. Chem. 268:16037-16046, 1993b. Dermietzel R and Krause D. Molecular anatomy of the blood-brain barrier as defined by immunochemistry. Intl. Rev. Cytol. 127:57-109,1991. Detmer M, Imcke E, Ruszczak Z and Orfanos CE. Effects of recombinant tumor necrosis factor-alpha on cultured microvascular endothelial cells derived from human dermis. J. Invest. Dermatol. 95.-219S-222S, 1990. deVries HE, Moor ACE, Blom-Roosemalen MCM, deBoer AG, Breimer DD, van Berkel TJC and Kuiper J. Lymphocyte adhesion to brain capillary endothelial cells in vitro. J. Neuroimunol. 52:1-8,1994.  261  Dobrina A, Menegazzi R, Carols TM, Nardon E, Cramer R, Zacchi T, Harlan JM and Patriarca P. Mechanisms of eosinophil adherence to cultured vascular endothelial cells. Eosinophils bind to the cytokine-induced endothelial ligand vascular cell adhesion molecule-1 via the very late activation antigen-4 integrin receptor. J. Clin. Invest. 88:20-26,1991. Dopp JM, Breneman SM and Olschowka JA. Expression of ICAM-1, VCAM-1, L-selectin and leukosialin in the mouse central nervous system during the induction and remission stages of experimental allergic encephalomyelitis. J. Neruoimmunol. 54:129-144,1994. Doran JE. Biological effects of endotoxin. Curr. Stud. Hematol. Blood Transfusion 59:66-99,1992. Dore-Duffy P, Newman W, Balabanon R, Lisak RP, Mainolfi E, Rothlein R and Peterson M. Circulating soluble adhesion proteins in cerebrospinal fluid and serum of patients with multiple sclerosis: correlation with clinical activity. Ann. Neurol. 37:55-62, 1995. Dorovini-Zis K, Sato M, Goping G, Rapoport S and Brightman M. Ionic lanthanum passage across cerebral endothelium exposed to hyperosmotic arabinose. Acta Neuropathol. 60:49-60, 1983. Dorovini-Zis K, Bowman PS, Betz AL and Goldstein GW. Hyperosmotic arabinose solutions open the tight junctions between brain capillary endothelial cells in tissue culture. Brain Res. 302:383-386,1984. Dorovini-Zis K, Bowman PS, Betz AL and Goldstein GW. Formation of a barrier by brain microvessel endothelial cells in culture. Fed. Proc. 46:2521-2522,1987a. Dorovini-Zis K, Bowman PS, Betz AL and Goldstein GW. Hyperosmotic urea reversibly opens the tight junctions between brain capillary endothelial cells in cell culture. J. Neuropathol. Exp. Neurol. 46:130-140,1987b. Dorovini-Zis K, Prameya R and Bowman P. Culture and characterization of microvascular endothelial cells derived from human brain. Lab. Invest. 64:425-436,1991. Dorovini-Zis K, Bowman PD and Prameya R. Adhesion and migration of human polymorphonuclear leukocytes across cultured bovine brain microvessel endothelial cells. J. Neuropath. Exp. Neurol. 51:194-205,1992a. Dorovini-Zis, K, Wong D and Newman W. Endothelial leukocyte adhesion molecule-1 (ELAM-1) expression in inflammatory and reactive disorders of the human central nervous system. J. Neuropathol. Exp. Neurol. 51:378,1992b. (Abstract) Dorovini-Zis K, Wong D, Newman WA and Dupuis B. Expression of vascular cell adhesion molecule-1 in inflammatory disorders of the central nervous system. J. Neuropathol. Exp. Neurol. 52:316,1993. (Abstract) Doukas J and Pober JS. IFN-y enhances endothelial activation induced by tumor necrosis factor but not IL-1. J. Immunol. 145:1727-1733,1990.  262  Duijvestijn AM, Horst E, Pals ST, Rouse BN, Steere AC, Pisker LJ, Meijer CJLM and Butcher EC. High endothelial differentiation in human lymphoid and inflammatory tissues defined by monoclonal antibody HECA-452. Am. J. Path. 130:147-155,1988. Dustin ML, Rothlein R, Bhan AK, Dinarello CA and Springer TA. Induction by IL-1 and interferon, tissue distribution, biochemistry and function of a natural adherence molecule (ICAM-1). J. Immunol. 137:245-254,1986. Dustin ML and Springer TA. Lymphocyte function-associated antigen-1 (LFA-1) interaction with intercellular adhesion molecule-1 (ICAM-1) is one of at least three mechanisms for lymphocyte adhesion to cultured endothelial cells. J. Cell. Biol. 107:321-331,1988. Eisen V. Past and present views of inflammation. Agents Actions Suppl. 3:9-16,1977. Elices MJ, Osborn L, Takada Y, Crouse C, Luhowskyj Y, Crouse C, Luhowskyj S, Helmer ME and Lobb RR. VCAM-1 on activated endothelium interacts with the leukocyte-integrin VLA-4 at a site distinct from the VLA-4/fibronectin binding site. Cell 60:577-584,1990. Engelberts I, von Hoff SCJ, Samyo SK, Buurman WA and van der Linden CJ. Generalized inflammation during peritonitis evidenced by intracutaneous E-selectin expression. Clin. Immunol. Immunopathol. 65:330-334,1992. Epstein FH. The role of interleukin-1 in disease. New Engl. J. Med. 328:106-113,1993. Erbe DV, Wolitzky BA, Presta LG, Norton CR, Ramos RJ, Burns DK, Rumberger JM, Narasinga Rao BN, Foxall C, Brandley BK and Lasky LA. Identification of an E-selectin region critical for carbohydrate recognition and cell adhesion. J. Cell. Biol. 119:215-227,1992. Etzioni A, Frydman M, Pollack S, Avidor I, Phillips ML, Paulson JC and Gershoni-Baruch R. Recurrent severe infections caused by a novel leukocyte adhesion deficiency. New Engl. J. Med. 327:1789-1792,1992. Evans CW, Taylor JE, Walker JD and Simmons NL. Transepithelial chemotaxis of rat peritoneal exudate cells. Br. J. Exp. Path. 64:644-654,1983. Fabry Z, Waldschmidt MM, Hendrickson D, Keiner J, Love-Homan L, Takei F, Hart M. Adhesion molecules on murine brain microvascular endothelial cells: expression and regulation of ICAM-1 and Lgp55. J. Neuroimmunol. 36:1-11,1992. Faustmann PM and Dermietzel R. Extravasation of polymorphonuclear leukocytes from the cerebral microvasculature. Cell Tissue Res. 242:399-407,1985. Florey HW and Grant LH. Leucocyte migration from small blood vessels stimulated with ultraviolet light: an electron microscopic study. J. Path. Bact. 82:13-17,1961. Foxall C, Watson S, Dowbenko D, Fennie C, Lasky LA, Kiso M, Hasegawa A, Asa D and Brandley BBC The three members of the selectin receptor family recognize a common carbohydrate epitope, the Sialyl Lewis x oligosaccharide. J. Cell. Biol. 117:895-902,1992.  263  Fuggle SV, Sanderson JB, Gray DWR, Richardson A and Morris PJ. Variation in expression of endothelial adhesion molecules in pretransplant and transplanted kidneys - correlation with intragraft events. Transplant. 55:117-123,1993. Fujita S, Puri RK, Yu Z-X, Travis WD and Ferrans VJ. An ultrastructural study of in vivo interactions between lymphocytes and endothelial cells in the pathogenesis of the vascular leak syndrome induced by interleukin-2. Cancer 68:2169-2174,1991. Furie MB, Cramer EB, Naprstek BL and Silverstein SC. Cultured endothelial cell monolayers that restrict the transendothelial passage of macromolecules and electrical current. J. Cell. Biol. 98:1033-1041,1984. Furie MB and McHugh DD. Migration of neutrophils across endothelial monolayers is stimulated by treatment of the monolayers with interleukin-1 or tumor necrosis factor. J. Immmunol. 143:3309-3317,1989. Furie MB, Tancinco MCA and Smith CW. Monoclonal antibodies to leukocyte integrins CDlla/CD18 and CDllb/CD18 or intercellular adhesion molecule-1 inhibit chemoattractant-stimulated neutrophil migration in vitro. Blood 78:2089-2097, 1991. Gartner LP and Hiatt JL. Color Atlas of Histology. Williams and Wilkins, Baltimore. 1990. Ghersa P, von Huijsduijnen RH, Whelan J and DeLamarter JF. Labile proteins play a dual role in the control of endothelial leukocyte adhesion molecule-1 (ELAM-1) gene regulation. J. Biol. Chem. 267:19226-19232,1992. Gibbs LS, Lai L and Malik AB. Tumor necrosis factor enhances the neutrophil-dependent increase in endothelial permeability. J. Cell. Physiol. 145:496-500,1990. Goldberger A, Middleton KA, Oliver JA, Paddock C, Yan H-C, De Lisser HM, Albelda SM and Newman PJ. Biosynthesis and processing of the cell adhesion molecule PECAM-1 includes production of a soluble form. J. Biol. Chem. 269:17183-17191, 1994. Goldstein GW and Betz AL. The blood-brain barrier. Sci. Am. 254:74-83,1986. Gordon EC, Danielsson PE, Nguyen T-S and Winn HR. A comparison of primary cultures of rat cerebral microvascular endothelial cells to rat aortic endothelial cells. In Vitro Cell. Dev. Biol. 27A:312-326,1991. Graber N, Venkat Gopal T, Wilson D, Beall LD, Polte T and Newman W. T cells bind to cytokine-activated endothelial cells via a novel, inducible sialoglycoprotein and endothelial leukocyte adhesion molecule-1. J. Immunol. 145:819-830,1990. Greenwood J, Calder V L Lymphocyte migration through cultured endothelial cell monolayers derived from the blood-retinal barrier. Immunol. 80:401-406,1993. Greve JM, Davis G, Meyer AM, Forte CP, Yost CP, Marlor CW, Kamarck ME and McClelland A. The major human rhinovirus receptor is ICAM-1. Cell 56:839-847, 1989.  264  Groves RW, Allen MH, Barker JNWN, Haskard DO and MacDonald DM. Endothelial leukocyte adhesion molecule-1 (ELAM-1) expression in cutaneous inflammation. Br. J. Dermatol. 124:117-123,1991. Groves RW, Ross EL, Baker JNWN and MacDonald DM. Vascular cell adhesion molecule-1: Expression in normal and diseased skin and regulation in vivo by interferon gamma. J. Am. Acad. Dermatol. 29:67-72,1993. Hahne M, Jager U, Tsenmann S, Hallmann R and Vestweber D. Five tumor necrosis factor-inducible cell adhesion mechanisms on the surface of mouse endothelioma cells mediate the binding of leukocytes. J. Cell Biol. 121:655-664,1993. Hahne M, Lanter M, Jager U and Vestweber D. A novel soluble form of mouse VCAM-1 is generated from a glycolipid-anchored splicing variant. Eur. J. Immunol. 24:421-428,1994. Hakkert BC, Kuijpers TW, Leeuwenberg JFM, van Mourik JA and Roos D. Neutrophil and monocyte adherence to and migration across monolayers of cytokine-activated endothelial cells: the contribution of CD18, ELAM-1, and VLA-4. Blood 78:2721-2726,1991. Hakomori S-I. Le x and related structures as adhesion molecules. Histochem. J. 24:771-776, 1992. Hamburger SA and McEver RP. GMP-140 mediates adhesion of stimulated platelets to neutrophils. Blood 75:550-554,1990. Hammersen F, Unterberg A and Hammersen E. The ultrastructural basis of interactions between leukocytes and endothelium. Adv. Exp. Med. Biol. 242:77-84,1988. Hannet I, Erkeller-Yuksel F, Lydyard P, Deneezs V and DeBruyere M. Developmental and maturational changes in human blood lymphocyte subpopulations. Immunol. Today 13:215-218, 1992. Harriman DGF. Bacterial infections of the central nervous system. In Greenfield's neuropathology. Adams JH, Corsellis JAN and Duchen LW. [Eds.] John Wiley and Sons, New York, 4th edition. 1984. Hashimoto M, Shingu M, Ezaki I, Nobunaga M, Minamihara M, Kato K and Sumioki H. Production of soluble ICAM-1 from human endothelial cells induced by IL-lp and TNF-a. Inflammation 18:163-173,1994. Haug CE, Colnin RB, Delmonico FL, Auchincloss H Jr, Tolkoff-Rubin N, Preffer FL, Rothlein R, Norris S, Scharschmidt L and Cosimi AB. A phase I trial of immunosuppression with anti-ICAM-1 (CD54) monoclonal antibody in renal allograft recipients. Transplant. 55:766-773,1993. Hess DC, Bhutwala T, Sheppard JC, Zhao W and Smith J. ICAM-1 expression on human brain microvascular endothelial cells. Neurosci. Lett. 168:201-204,1994. Hession C, Osborn L, Goff D, Chi-Rosso G, Vassallo C, Pasek M, Pittack C, Tizard R, Goelz S, McCarthy K, Hopple S and Lobb R. Endothelial leukocyte adhesion molecule 1: direct expression cloning and functional interactions. Proc. Natl. Acad. Sci. U.S.A. 87:1673-1677,1990. 265  Hession C, Tizard R, Vassallo C, Schiffer SB, Goff D, Moy P, Chi-Rosso G, Luhowskuj S, Lobb R and Osborn L. Cloning an alternate form of vascular cell adhesion molecule-1 (VCAM-1). J. Biol. Chem. 266:6682-6685,1991. Hickey WF, Hsu BL, Kimura H. T-lymphocyte entry into the central nervous system. J. Neurosci. Res. 28:254-260,1991. Hofman FM, von Hanwhr RI, Dinarello CA, Mizel SB, Hinton D and Merrill JE. Immunoregulatory molecules and IL-2 receptors identified in multiple sclerosis brain. J. Immunol. 136:3239-3245,1986. Hofman FM, Hinton DR, Johnson K and Merrill JE. Tumor necrosis factor identified in multiple sclerosis brain. J. Exp. Med. 170:607-612,1989. Horst E, Pals ST, Duijvestijn AM, Mourik JV, Kraal G, Butcher E and Meijer JLM. Expression and regulation of an antigen specific for endothelium involved in human lymphocyte homing. Adv. Exp. Med. Biol. 237:477-484,1988. Huber AR, Kunkel SL, Todd RF III, Weiss SJ. Regulation of transendothelial neutrophil migration by endogenous interleukin-8. Science 254:99-102,1991. Hughes CCW, Male DK and Lantos PL. Adhesion of lymphocytes to cerebral microvascular cells: effects of interferon-y, tumor necrosis factor and interleukin-1. Immunol. 64:677-681,1988. Hurley JV. Acute inflammation. Churchill Livingston, Edinburgh. 1983. Huynh HK and Dorovini-Zis K. Effects of interferon-y on primary cultures of human brain microvessel endothelial cells. Am. J. Path. 142:1265-1278,1993. Huynh HK. In vitro modulation of class IIMHC antigen expression by human cerebral endothelium and endothelial cell-lymphocyte interactions by interferons y and p\ The University of British Columbia, Ph.D. thesis. 1994. Isobe M, Suzuki J-I, Yagita H, Okmura K, Yamazaki S, Nagai R, Yazaki Y and Sekiguchi M. Immunosuppression to cardiac allograft and soluble antigens by anti-vascular cell adhesion molecule-1 and anti-very late antigen-4 monoclonal antibodies. J. Immunol. 153:5810-5818,1994. Issekutz AC and Issekutz TB. The contribution of LFA-1 (CDlla/CD18) and MAC-1 (CD lib/CD 18) to the in vivo migration of polymorphonuclear leukocytes to inflammatory reactions in the rat. Immunology. 76:655-661,1992. Jaattela M. Biologic activities and mechanisms of action of tumor necrosis factor-a / cachectin. Lab. Invest. 64:724-742,1991. Jakobsen PH, Morris-Jones S, Ronn A, Hviid L, Theander TG, Elhassan IM, Bygbjerg IC and Greenwood BM. Increased plasma concentrations of sICAM-1, sVCAM-1 and sELAM-1 in patients with Plasmodium falciparum or P. Vivax malaria and association with disease severity. Immunol. 83:665-669,1994.  266  Jalkanen S, Saari S, Kalimo H, Lammintausta K, Vainio E, Limo R, Duijvestijn AM and Kalimo K. Lymphocyte migration into the skin: the role of lymphocyte homing receptor (CD44) and endothelial cell antigen (HECA-452). J. Invest. Dermatol. 94:786-792,1990. Jalkanen S and Salmi M. Vascular adhesion protein-1 (VAP-1) - a new adhesion molecule recruiting lymphocytes to sites of inflammation. Res. Immunol. 144:746-749,1993. Janzer RC. The blood-brain barrier: cellular basis. J. Inher. Metab. Dis. 16:639-647, 1993. Johnston GI, Bliss GA, Newman PJ and McEver RP. Structure of the human gene encoding granule membrane protein-140, a member of the selectin family of adhesion receptors for leukocytes. J. Biol. Chem. 265:21381-21385, 1990. Junqueira LC, Carneiro J and Kelley RO. Basic Histology. Appleton and Lange, Connecticut. 1989. Kavanaugh A, Nichols LA and Lipsky PE. Treatment of refractory rheumatoid arthritis with an anti-CD54 (intercellular adhesion molecule-1, ICAM-1) monoclonal antibody. Arthritis Rheum. 35:S43,1992. (Abstract) Kennedy MK, Torrance DS, Picha KS and Mohler KM. Analysis of cytokine mRNA expression in the central nervous system of mice with experimental autoimmune encephalomyelitis reveals that IL-10 mRNA expression correlates with recovery. J. Immunol. 149:2496-2505,1992. Kerr MA and Stocks SC. The role of CD15-(Lex)-related carbohydrates in neutrophil adhesion. Histochem. J. 24:811-826,1992. Kim KS, Wass CA, Cross AS and Opal SM. Modulation of blood-brain barrier permeability by tumor necrosis factor and antibody to tumor necrosis factor in the rat. Lymphokine and Cytokine Res. 11:293-298,1992. Kirschbaum NE, Gumina RJ and Newman PJ. Organization of the gene from human platelet/endothelial cell adhesion molecule-1 shows alternatively spliced isoforms and a functionally complex cytoplasmic domain. Blood 84:4028-4037,1994. Kishimoto TK, Warnock RA, Jutila MA, Butcher EC, Lane C, Anderson DC and Smith CW. Antibodies against human neutrophil LECAM-1 (LAM-l/Leu-8/DREG-56 antigen) and endothelial cell ELAM-1 inhibit a common CD18-independent adhesion pathway in vitro. Blood 78:805-811,1991. Kobayashi H, Mizuki T, Koda Y, Okazaki M, Yanagihara M, Wada A and Izumi F. Characterization of protein kinase C in endothelial cells of brain and peripheral vessels. Soc. Neurosci. Abstr. 16:46,1990. Koch AE, Burrows JC, Haines GK, Carlos TM, Harlan JM and Leibovich SJ. Immunolocalization of endothelial and leukocyte adhesion molecules in human rheumatoid and osteoarthritic synovial tissues. Lab. Invest. 64:313-320,1991.  267  Kuchroo VK, Greer JM, Kaul D, Ishioka G, Franco A, Sette A, Sobel RA and Lees MB. A single TcR antagonist peptide inhibits experimental allergic encephalomyelitis mediated by a diverse T cell repertoire. J. Immunol. 153:3326-3336,1994. Kuijpers TW, Hakkert BC, Hart MH and Roos D. Neutrophil migration across monolayers of cytokine-prestimulated endothelial cells: a role for platelet-activating factor and IL-8. J. Cell. Biol. 117:565-572,1992. Kuijpers TW, Hakkert BC, Hoogerwerf M, Leeuwenberg JFM and Roos D. Role of endothelial leukocyte adhesion molecule-1 and platelet-activating factor in neutrophil adherence to IL-1-prestimuIated endothelial cells. J. Immunol. 147:1369-1376,1991. Kuijpers TW, Raleigh M, Kavanagh T, Janssen H, Calafat J, Roos D, and Harlan JM. Cytokine-activated endothelial cells internalize E-selectin into a lysosomal compartment of vesiculotubular shape. A tubulin driven process. J. Immunol. 152:5060-5069,1994. Kuroda Y and Matsui M. Cerebrospinal fluid interferon-gamma is increased in HTLV-1 associated myelopathy. J. Neuroimmunol. 42:223-226,1993. Kurtel H, Tso P and Granger DN. Granulocyte accumulation in postischemic intestine: role of leukocyte adhesion glycoprotein CD11/CD18. Am. J. Physiol. 262:G878-G882,1992. Kyan-Aung U, Haskard DO and Lee TH. Vascular cell adhesion molecule-1 and eosinophil adhesion to cultured human umbilical vein endothelial cells in vitro. Am. J. Respir. Cell. Mol. Biol. 5:445-450,1991a. Kyan-Aung U, Haskard DO, Poston RN, Thronhill MH and Lee TH. Endothelial leukocyte adhesion molecule-1 and intercellular adhesion molecule-1 mediate the adhesion of eosinophils to endothelial cells in vitro and are expressed by endothelium in allergic cutaneous inflammation in vivo. J. Immunol. 146:521-528,1991b. Labow MA, Norton CR, Rumberger JM, Lombardgillooly KM, Shuster DJ, Hubbard J, Bertko R, Knaack PA, Terry RW, Harbison ML, Kontgen F, Stewart CL, Mclntyre KW, Will PC, Burns DK and Wolitzky BA. Characterization of E-selectin-deficient mice-demonstration of overlapping function of the endothelial selectins. Immunity 1:709-720,1994. Lane TA, Lamkin GE and Wancewicz EV. Protein kinase C inhibitors block the enhanced expression of intercellular adhesion molecule-1 on endothelial cells activated by interleukin-1, lipopolysaccharide and tumor necrosis factor. Biochem. Biophys. Res. Commun. 172:1273-1281,1990. Languino LR, Plescia J, Duperray A, Brian AA, Plow EF, Geltosky JE and Altieri DC. Fibrinogen mediates leukocyte adhesion to vascular endothelium through an ICAM-1-dependent pathway. Cell 73:1423-1434,1993. Lawrence MB, Smith CW, Eskin SG and Mclntire LV. Effect of venous shear stress on CD18-mediated neutrophil adhesion to cultured endothelium. Blood 75:227-237, 1990.  268  Lawrence MB and Springer TA. Leukocytes roll on a selectin at physiologic flow rates: distinction from and prerequisite for adhesion through integrins. Cell 65:859-873, 1991. Lawrence MB and Springer TA. Neutrophils roll on E-selectin. J. Immunol. 151:6338-6346,1993. Leach L, Clark P, Lampugnani M-G, Arroyo AG, Dejana E and Firth JA. Immunoelectron characterisation of the inter-endothelial junctions of human term placenta. J. Cell. Sci. 104:1073-1081,1993. Leavesley DI, Oliver JM, Swart BW, Berndt MC, Haylock DN and Simmons PJ. Signals from platelet/endothelial cell adhesion molecule enhance the adhesive activity of the very late antigen-4 integrin of human CD34+ hemopoietic progenitor cells. J. Immunol. 153:4673-4683,1994. Leeuwenberg JFM, Jeunhomme TMAA and Buurman WA. Induction of an activation antigen on human endothelial cells in vitro. Eur. J. Immunol. 19:715-720,1989. Leeuwenberg JFM, Jeunhomme GMAA and Buurman WA. Adhesion of polymorphonuclear cells to human endothelial cells. Adhesion-molecule-dependent, and Fc receptor-mediated adhesion-molecule-independent mechanisms. Clin. Exp. Immunol. 81:496-500, 1990a. Leeuwenberg JFM, von Asmuth EJU, Jeunhomme TMAA and Buurman WA. IFN-y regulates the expression of the adhesion molecule ELAM-1 and IL-6 production by human endothelial cells in vitro. J. Immunol. 145:2110-2114,1990b. Leeuwenberg JFM, Smeets EF, Neefjes JJ, Shaffer MA, Cinek T, Jeunhomme TM, Ahern TJ and Buurman WA. E-selectin and intercellular adhesion molecule-1 are released by activated human endothelial cells in vitro. Immunology 77:543-549, 1992. Leung DYM, Pober JS and Cotran RS. Expression of endothelial-leukocyte adhesion molecule-1 in elicited late phase allergic reactions. J. Clin. Invest. 87:1805-1809, 1991. Ley K, Bullard DC, Arbones ML, Basse R, Vestweber D, Tedder TF and Beaudet AL. Sequential contribution of L- and P-selectin to leukocyte rolling in vivo. J. Exp. Med. 181:669-675,1995. Liu T, McDonnell PC, Young PR, White RF, Siren AL, Hallenbeck JM, Barone FC and Feuerstein GZ. Interleukin-ip mRNA expression in ischemic rat cortex. Stroke 24:1746-1751,1993. Liversidge J, Sewell HF and Forrester JV. Interactions between lymphocytes and cells of the blood-retina barrier: mechanisms of T lymphocyte adhesion to human retinal capillary endothelial cells and retinal pigment epithelial cells in vitro. Immunology. 71:390-396,1990.  269  Lo SBC, Lee S, Ramos RA, Lobb R, Rosa M, Chi-Rosso G and Wright SD. Endothelial-leukocyte adhesion molecule 1 stimulates the adhesive activity of leukocyte integrin CR3 (CDllb/CD18, Mac-1, a m P2) on human neutrophils. J. Exp. Med. 173:1493-1500,1991. Lossinsky AS, Badmajew V, Robson JA, Moretz RC and Wisniewski HM. Sites of egress of inflammatory cells and horseradish peroxidase transport across the blood-brain barrier in a murine model of chronic relapsing experimental allergic encephalomyelitis. Acta Neuropath. 78:359-371,1989. Lossinsky AS, Pluta R, Song MJ, Badmajew V, Moretz RC an Wisniewski HM. Mechanisms of inflammatory cell attachment in chronic relapsing experimental allergic encephalomyelitis: a scanning and high-voltage electron microscopic study of the injured mouse blood-brain barrier. Microvasc. Res. 42:299-310, 1991. Lowe JB, Stoolman LM, Nair RP, Larsen RD, Berhend TL and Marks RM. ELAM-1-dependent cell adhesion to vascular endothelium determined by a transfected human fucosyltransferase cDNA. Cell 63:475-484,1990. Lund-Johansen F and Terstappen LWMM. Differential surface expression of cell adhesion molecules during granulocytes maturation. J. Leukocyte Biol. 54:47-55, 1993. Luscinskas FW, Brock AF, Arnaout MA and Gimbrone MA Jr. Endothelial-leukocyte adhesion molecule-1-dependent and leukocyte (CDll/CD18)-dependent mechanisms contribute to polymorphonuclear leukocyte adhesion to cytokine-activated human vascular endothelium. J. Immunol. 142:2257-2263, 1989. Luscinskas FW, Cybulsky MI, Kiely J-M, Peckins CS, Davis VM and Gimbrone MA Jr. Cytokine-activated human endothelial monolayers support enhanced neutrophil transmigration via a mechanism involving both endothelial-leukocyte adhesion molecule-1 and intercellular adhesion molecule-1. J. Immunol. 146:1617-1625, 1991. Mackay CR and Imhof BA. Cell adhesion in the immune system. Immunol. Today 14:99-102, 1993. Majno G. The healing hand. Man and wound in the ancient world. Harvard University Press, Cambridge. 1975. Majno G. The capillary then and now: an overview of capillary pathology. Modern Path. 5:9-22, 1992. Male D, Pryce G and Rahman J. Comparison of the immunological properties of rat cerebral and aortic endothelium. J. Neuroimmunol. 30:161-168,1990a. Male D, Pryce G, Hughes C and Lantos P. Lymphocyte migration into brain modelled in vitro: control by lymphocyte activation, cytokines and antigen. Cell. Immunol. 127:1-11, 1990b. Male D, Pryce G, Linke A., Rahman J. Lymphocyte migration into the CNS modelled in vitro. J. Neuroimmunol. 40:167-172,1992. 270  Male D, Rahman J, Pryce G, Tamatami T and Miyasaka M. Lymphocyte migration into the CNS modelled in vitro: roles of LFA-1, ICAM-1 and VLA-4. Immunol. 81:366-372,1994. Mantovani A and Dejana E. Cytokines as communication signals between leukocytes and endothelial cells. Immunol. Today 10:370-375,1989. Marchesi VT. The site of leucocyte migration during inflammation. Quart. J. Exp. Physiol. 46:115-118,1961. Marchesi VT and Florey HW. Electron microscopic observations on the emigration of leucocytes. Quart. J. Exp. Physiol. 45:343-348,1960. Matsubara T, Furukawa S and Yabuta K. Serum levels of tumor necrosis factor, Interleukin 2 receptor and interferon-y in Kawasaki disease involved coronary-artery lesions. Clin. Immunol. Immunopath.. 56:29-36,1990. Mattila P, Majori M-L, Mattila PS and Renkonen R. TNF-a-induced expression of endothelial adhesion molecules, ICAM-1 and VCAM-1, is linked to protein kinase C activation. Scan. J. Immunol. 36:159-165,1992. Mayadas TN, Johnson RC, Rayburn H, Hynes RO and Wagner DD. Leukocyte rolling and extravasation are severly compromised in P-selectin-deficient mice. Cell 74:541-554,1993. McCarron RM, Wang L, Racke MK, McFarlin DE and Spatz M. Effect of cytokines on ICAM expression and T cell adhesion to cerebrovascular endothelial cells. In: Frontiers in cerebral vascular biology: transport and its regulation. Drewes LR and Betz AL [Eds.] Plenum Press, New York. 1993a. McCarron RM, Wang L, Racke MK, McFarlin DE and Spatz M. Cytokine-regulated adhesion between encephalitogenic T lymphocytes and cerebrovascular endothelial cells. J. Neuroimmunol. 43:23-30,1993b. McCarron RM, Wang L, Stanimirovic DB, Spatz M. Endotheliun induction of adhesion molecule expression on human brain microvessel endothelial cells. Neurosci. Lett. 156:31-34,1993c. Mccourt PAG, Ek B, Forsberg N and Gustafson S. Intercellular adhesion molecule-1 is a cell surface receptor for hyaluronan. J. Biol. Chem. 269:30081-30084,1994. McEver RP, Beckstead JH, Moore KL, Marshall-Carlson L and Bainton DF. GMP-140, a platelet a-granule membrane protein, is also synthesized by vascular endothelial cells and is localized in Weibel-Palade bodies. J. Clin. Invest. 84:92-99, 1989. McEver, RP. Leukocyte-endothelial cell interactions. Curr. Opinion Cell Biol. 4:840-849, 1992. McMillen MA, Huribal M and Sumpio B. Common pathway of endothelial-leukocyte interaction in shock, ischemia, and reperfusion. Am. J. Surg. 166:557-562,1993. Mesri M, Liversidge J and Forrester JV. ICAM-l/LFA-1 interactions in T-lymphocyte activation and adhesion to cells of the blood-retina barrier in the rat. Immunol. 83:52-57, 1994. 271  Mettler CC. History of Medicine. Mettler FA [ed.]. The Blakiston Company, Philadelphia. 1947. Meyrick BO, Ryan US and Brigham KL. Direct effect of E. coli endotoxin on structure and permeability of pulmonary endothelial monolayers and the endothelial layers of intimal explants. Am. J. Path. 122:140-151,1986. Mileski W, Borgstrom D, Lightfoot E, Rothlein R, Fannes R, Lipsky P and Baxter C. Inhibition of leukocyte-endothelial adherence following thermal injury. J. Surg. Res. 52:334-339,1992. Milks LC, Conyers GP and Cramer EB. The effects of neutrophil migration on epithelial permeability. J. Cell Biol. 103:2729-2738,1986. Mohler KM, Torrance DS, Smith CA, Goodwin RG, Stremler KE, Fung VP, Madani H, Widmer MB. Soluble tumor necrosis factor (TNF) receptors are effective therapeutic agents in lethal endotoxemia and function simultaneously as both TNF carriers and TNF antagonists. J. Immunol. 151:1548-1561,1993. Moller B, Mogensen SC, Wendelboe P, Benstzen K and Petersen CM. Bioactive and inactive forms of tumor necrosis factor-a in spinal fluid from patients with meningitis. J. Infect. Dis. 886:89,1991. Montgomery KF, Osborn L, Hession C, Tizard R, Goff D, Vassallo C, Tarr PI, Bomsztyk K, Lobb R, Harlan JM and Pohlman TH. Activation of endothelial-leukocyte adhesion molecule 1 (ELAM-1) gene transcription. Proc. Natl. Acad. Sci. U.S.A. 88:6523-6527,1991. Morganti-Kossmann MC, Kossmann T and Wahl SM. Cytokines and neuropathology. Trends in Pharm. Sci. 13:286-291,1992. Morzycki W, Sadowska J and Issekutz AC. Interleukin-1 and tumor necrosis factor a induced polymorphonuclear leukocyte-endothelial cell adhesion and transendothelial migration in vitro: the effect of apical versus basal monolayer stimulation. Immunol. Lett. 25:331-340,1990. Moser R, Schleiffenbaum B, Groscurth P and Tehr J. Interleukin 1 and tumor necrosis factor stimulate human vascular endothelial cells to promote transendothelial neutrophil passage. J. Clin. Invest. 83:444-455,1989. Moughal NA, Adonogianaki E, ThronhiU MH and Kinane DF. Endothelial cell leukocyte adhesion molecule-1 (ELAM-1) and intercellular adhesion molecule-1 (ICAM-1) expression in gingival tissue during health and experimentally-induced gingivitis. J. Periodont. Res. 27:623-630,1992. Moynagh, PN, Williams DC and O'Neill LAJ. Activation of N F - K B and induction of vascular cell adhesion molecule-1 and intercellular adhesion molecule-1 expression in human glial cells by IL-1. Modulation by antioxidants. J. Immunol. 153:2681-2690,1994. Muller WA, Ratti CM, McDonnell SL and Cohn ZA. A human endothelial cell-restricted, externally disposed plasmalemmal protein in enriched in intercellular junctions. J. Exp. Med. 170:399-414,1989.  272  Muller WA, Berman ME, Newman PJ, DeLisser HM and Albelda SM. A heterophilic adhesion mechanism for platelet/endothelial cell adhesion molecule 1 (CD31). J. Exp. Med. 175:1401-1404,1992. Muller WA, Weigel SA, Deng X and Phillips DM. PECAM-1 is required for transendothelial migration of leukocytes. J. Exp. Med. 178:449-460,1993. Munro JM, Pober JS and Cotran RS. Recruitment of neutrophils in the local endotoxin response: association with de novo endothelial expression of endothelial leukocyte adhesion molecule-1. Lab. Invest. 64:295-299,1991. Munro JM, Lo SIC, Corless C, Robertson MJ, Lee NC, Barnhill RL, Weinberg DS and Bevilacqua MP. Expression of sialyl-Lewis X, and E-selectin ligand, in inflammation, immune processes, and lymphoid tissues. Am. J. Pathol. 141:1397-1408, 1992. Myers CL, Desai SN, Schembri-King J, Letts GL, Wallace RW. Discriminatory effects of protein kinase inhibitors and calcium ionophore on endothelial ICAM-1 induction. Am. J. Physiol. 262:C365-C373,1992a. Myers CL, Wertheimer SJ, Schembri-King J, Parks T and Wallace RW. Induction of ICAM-1 by TNF-a, IL-lp and LPS in human endothelial cells after downregulation of PKC. Am. J. Physiol. 263:C767-772,1992b. Nagel T, Resnick N, Atkinson WJ, Dewey CF Jr and Gimbrone MA Jr. Shear stress selectively upregulates intercellular adhesion molecule-1 expression in cultured human vascular endothelial cells. J. Clin. Invest. 94:885-891,1994. Nash S, Stafford J and Madara JL. Effects of polymorphonuclear leukocyte transmigration on the barrier function of cultured intestinal epithelial monolayers. J. Clin. Invest. 80:1104-1113,1987. Nash S, Stafford J and Madara JL. The selective and superoxide-independent disruption of intestinal epithelial tight junctions during leukocyte transmigration. Lab. Invest. 59:531-537,1988. Navah M, Hough GP, Berliner JA, Frank JA, Fogelman A-M, Haberland ME and Edwards PA. Rabbit beta-migrating very low density lipoprotein increases endothelial macromolecular transport without altering electrical resistance. J. Clin. Invest. 78:389-397,1986. Neish AS, Williams AJ, Palmer HJ, Whitley MZ and Collins T. Functional analysis of the human vascular cell adhesion molecule 1 promoter. J. Exp. Med. 176:1583-1593,1992. Nelson RM, Dolich S, Aruffo A, Cecconi O and Bevilacqua MP. Higher-affinity oligosaccharide ligands for E-selectin. J. Clin. Invest. 91:1157-1166,1993. Newman W, Beall LD, Carson CW, Hunder GG, Graber N, Randhawa ZI, Venkat Gopal T, Wiener-Kronick J and Matthay MA. Soluble E-selectin is found in supernatants of activated endothelial cells and is elevated in the serum of patients with septic shock. J. Immunol. 150:644-654,1993.  273  Nortamo P, Li R, Renkonen R, Timonen T, Prieto J, Patarroyo M, Gahmberg CG. The expression of human intercellular adhesion molecule-2 is refractory to inflammatory cytokines. Eur. J. Immunol. 21:2629-2632,1991. O'Neill JK, Butter C, Baker D, Gschmeissner SE, Kraal G, Butcher EC and Turk JL. Expression of vascular addressins and ICAM-1 by endothelial cells in the spinal cord during chronic relapsing experimental allergic encephalomyelitis in the Biozzi AB/H mouse. Immunology. 72:520-525,1991. Ockenhouse CF, Tegoshi T, Maeno Y, Benjamin C, Ho M, Kan KE, Thway Y, Win K, Aikawa M and Lobb RR. Human vascular endothelial cell adhesion receptors for Plasmodium falciparum-infected erythrocytes: roles for endothelial leukocyte adhesion molecule 1 and vascular cell adhesion molecule 1. J. Exp. Med. 176:1183-1189,1992. Ohmori K, Takada A, Yoneda T, Buma Y, Hirashima K, Tsuyuoka K, Hasegawa A and Kannagi R. Differentiation-dependent expression of Sialyl stage-specific embryonic antigen-1 and I-antigens on human lymphoid cells and its implications for carbohydrate-mediated adhesion to vascular endothelium. Bloos 81:101-111, 1993. Okada Y, Copeland BR, Mori E, Tung M-M, Thomas WS and del Zoppo GJ. P-selectin and intercellular adhesion molecule-1 expression after focal brain ischemia and reperfusion. Stroke 25:202-211, 1994. Oliver MG, Specian RD, Perry MA and Granger DN. Morphologic assessment of leukocyte-endothelial cell interactions in mesenteric venules subjected to ischemia and reperfusion. Inflammation 15:331-346,1991. Oppenheimer-Marks N and Ziff M. Migration of lymphocytes through endothelial cell monolayers: augmentation by interferon-y. Cell. Immunol. 114:307-323,1988. Oppenheimer-Marks N, Davis LS, Bogne DT, Ramberg J and Lipsky PE. Differential utilization of ICAM-1 and VCAM-1 during the adhesion and transendothelial migration of human T lymphocytes. J. Immunol. 147:2913-1921,1991. Oppenheimer-Marks N and Lipsky PE. The adhesion and transendothelial migration of human T lymphocytes subsets. Behring. Inst. Mitt. 92:44-50,1993. Orosz CG, van Buskirk A, Sedmak DD, Kincade P, Miyake K and Pelletier RP Role of the endothelial adhesion molecule VCAM in murine cardiac allograft rejection. Immunol. Letters. 32:7-12,1992. Osborn L, Hession C, Tizard R, Vassallo C, Luhowskyj S, Chi-Rosso, G and Lobb, R. Direct expression cloning of vascular cell adhesion molecule 1, a cytokine-induced endothelial protein that binds to lymphocytes. Cell. 59:1203-1211,1989. Owman C and Hardebo JE. Functional heterogeneity of the cerebrovascular endothelium. Brain Behav. Evol. 32:65-75,1988.  274  Parkos CA, Colgan GP, Delp C, Arnaout MA and Madara JL. Neutrophil migration across a cultured epithelial monolayer elicits a biphasic resistance response representing sequential effects on transcellular and paracellular pathways. J. Cell Biol. 117:757-764, 1992. Partridge CA, Horvath CJ, Del Vecchio PJ, Phillips PG and Malik AB. Influence of extracellular matrix in tumor necrosis factor-induced increase in endothelial permeability. Am. J. Physiol. 263:L627-L633,1992. Partridge CA, Jeffrey JJ and Malik AB. A 96-kDa gelatinase induced by TNF-a contributes to increased microvascular endothelial permeability. Am. J. Physiol. 265:L438-L447,1993. Peake S. Monoclonal antibodies - immunotherapy for the critically ill. Anaesth. Intens. Care 21:739-751,1993. Pelletier RP, Ohye RG, Vanbuskirk A, Sedmak DD, Kinade P, Ferguson RM and Orosz CG. Importance of endothelial VCAM-1 for inflammatory leukocytic infiltration in vivo. J. Immunol. 149:2473-2481,1992. Phillips ML, Nudelman E, Gaeta FCA, Perez M, Singhal AK, Hakomori S-I and Paulson JC. ELAM-1 mediates cell adhesion by recognition of a carbohydrate ligand, Sialyl-Lex. Science 250:1130-1135,1990. Piali L, Albelda SM, Baldwin HS, Hammel P, Gisler RH and Imhof BA. Murine platelet endothelial cell adhesion molecule (PECAM-1)/CD31 modulates (32 integrins on lymphokine-activated killer cells. Eur. J. Immunol. 23:2464-2471,1993. Picker LJ, Mickie SA, Rott LS and Butcher EC. A unique phenotype of skin-associated lymphocytes in humans. Preferential expression of the HECA-452 epitope by benign and malignant T cells at cutaneous sites. Am. J. Path. 136:1053-1068, 1990. Picker LJ, Kishimoto TK, Smith CW, Warnock RA and Butcher EC. ELAM-1 is an adhesion molecule for skin-homing T cells. Nature. 28:796-799,1991a. Picker LJ, Warnock RA, Burns AR, Doerschuk CM, Berg EL and Butcher EC. The neutrophil selectin LECAM-1 presents carbohydrate ligands to the vascular selectins ELAM-1 and GMP-140. Cell 66:921-933,1991b. Pigott R, Needham LA, Edwards RM, Walker C and Power C. Structural and functional studies of the endothelial activation antigen endothelial leucocyte adhesion molecule-1 using a panel of monoclonal antibodies. J. Immunol. 147:130-135, 1991. Pigott R, Dillon LP, Hemingway IH and Gearing AJH. Soluble forms of E-selectin, ICAM-1 and VCAM-1 are present in the supernatants of cytokine activated cultured endothelial cells. Biochem. Biophys. Res. Commun. 187:584-589,1992. Pizzolo G, Vinante F, Nadali G, Chilosi M and Semenzato G. Circulating soluble adhesion molecules: more observations on the increased levels in disease. Immunol. Today 15:140-141,1994.  275  Pober JS, Bevilacqua MP, Mendrick DL, Lapierre LA, Fiers W and 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. 136:1680-1687,1986a. Pober JS, Gimbrone MA Jr., Lapierre LA, Mandrick DL, Fiers W, Rothlein R and Springer TA. Overlapping patterns of activation of human endothelial cells by interleukin 1, tumor necrosis factor, and immune interferon. J. Immunol. 137:1893-1896,1986b. Pober JS, Lapierre LA, Stolpen AH, Brook TA, Springer TA, Fiers W, Bevilacqua MP, Mendrick DL and Gimbrone MA Jr. Activation of cultured human endothelial cells by recombinant lymphotoxin: comparison with tumor necrosis factor and interleukin 1 species. J. Immunol. 136:3319-3324,1987. Pober JS and Cotran RS. The role of endothelial cells in inflammation. Transplantation 50:537-544, 1990. Pober JS and Cotran RS. What can be learned from the expression of endothelial adhesion molecules in tissues? Lab. Invest. 64:301-305,1991. Pober JS, Slowik MR, DeLuca LG and Ritchie AJ. Elevated cyclic AMP inhibits endothelial cell synthesis and expression of TNF-induced endothelial leukocyte adhesion molecule-1 and vascular cell adhesion molecule-1, but not intercellular adhesion molecule-1. J. Immunol. 150:5114-5123,1993. Polte T, Newman W, Raghunathan G and Venkat Gopal T. Structural and functional studies of full-length vascular cell adhesion molecule-1: internal duplication and homology to several adhesion proteins. DNA Cell. Biol. 10:349-357,1991. Postigo AA, Garcia-Vicuna R, Diaz-Gonzalez, Arroya AG, De Landazuri MO, Chi-Rosso G, Lobb RR, Laffon A and Sanchez-Madrid F. Increased binding of synovial T lymphocytes from rheumatoid arthritis to endothelial-leukocyte adhesion molecule-1 (ELAM-1) and vascular cell adhesion molecule-1 (VCAM-1). J. Clin. Invest. 89:1445-1452,1992. Pryce G, Male DK and Sarkar C. Control of lymphocyte migration into brain: selective interactions of lymphocyte subpopulations with brain endothelium. Immunol. 72:393-398, 1991. Quan N, Sundar SK and Weiss JM. Induction of Interleukin-1 in various brain regions after peripheral and central injections of lipopolysaccharide. J. Neuroimmunol. 49:125-134,1994. Raine CS, Lee SC, Scheinberg LC, Duijvestijn AM and Cross AH. Adhesion molecules on endothelial cells in the central nervous system: an emerging area in the neuroimmunology of multiple sclerosis. Clin Immunol. Immunopathol. 57:173-187,1990a. Raine CS, Cannella B, Duijvestijn AM and Cross AH. Homing to central nervous system vasculature by antigen-specific lymphocytes. II. Lymphocyte / endothelial cell adhesion during the initial stages of autoimmune demyelination. Lab. Invest. 63:476-489, 1990b. 276  Raine CS. Demyelianting diseases. In Textbook of neuropathology. Davis RL and Robertson DM. [Eds.] William and Wllkins, Baltimore. 2nd edition, 1991. Rapoport SI. Blood-brain barrier in physiology and medicine. Raven Press, New York. 1976. Rather LJ. Disturbance of function (functio laesa): The legendary fifth cardinal sign of inflammation, added by Galen to the four cardinal signs of Celsus. Bull. N.Y. Acad. Med. 47:303-322,1971. Read MA, Whitley MZ, Williams AJ and Collins T. NF-KB and I^B™ an inducible regulatory system in endothelial activation. J. Exp. Med. 179:503-512,1994. Redl H, Dinges HP, Buurman WA, van der Linden CJ, Pober JS, Cotran RS and Schlag G. Expression of endothelial leukocyte adhesion molecule-1 in septic but not traumatic/hypovolemic shock in the baboon. Am. J. Pathol. 139:461-466,1991. Reese TS and Kamovsky MJ. Fine structural localization of a blood brain barrier to exogenous peroxidase. J. Cell. Biol. 34:207-217,1967. Renkonen R, Mennander A, Mattila P and Ustinov J. Signal transduction during in vitro lymphocyte homing. Human Immunol. 28:134-140,1990a. Renkonen R, Mennander A, Ustinov J and Mattila P. Activation of protein kinase C is crucial in the regulation of ICAM-1 expression on endothelial cells by interferon-y. Intl. Immunol. 2:719-724,1990b. Renno T, Krakowski M, Piccirillo C, Lin J-Y and Owens T. TNF-a expression by resident microglia and infiltrating leukocytes in the central nervous system of mice with experimental allergic encephalomyelitis. J. Immunol. 154:944-953, 1995. Rice GE and Bevilacqua MP. An inducible endothelial cell surface glycoprotein mediates melanoma adhesion. Science 246:1303-1306,1989. Rice GE, Munro JM and Bevilacqua MP. Inducible cell adhesion molecule 110 (INCAM-110) is an endothelial receptor for lymphocytes. A CDll/CD18-independent adhesion mechanism. J. Exp. Med. 171:1369-1374, 1990. Rice GE, Munro JM, Corless C and Bevilacqua MP Vascular and nonvascular expression of INCAM-110. A target for mononuclear leukocyte adhesion in normal and inflamed human tissues. Am. J. Path. 138:385-393,1991. Rieckmann P, Albrecht M, Kitze B, Weber T, Tumani H, Broocks A, Luer W, Helwig A and Poser S. Tumor necrosis factor-a messenger RNA expression in patients with relapsing-remitting multiple sclerosis is associated with disease activity. Ann. Neurol. 37:82-88,1995. Rieckmann P, Nunke K, Burchhardt M, Albrecht M, Wiltfang J, Ulrich M and Felgenhauer K. Soluble intercellular adhesion molecule-1 in cerebrospinal fluid: an indicator for the inflammatory impairment of the blood-cerebrospinal fluid barrier. J. Neuroimmunol. 47:133-140,1993.  277  Ritchie AJ, Johnson DR, Ewenstein BM and Pober JS. Tumor necrosis factor induction of endothelial cell surface antigen is independent of protein kinase C activation or inactivation. Studies with phorbol myristate acetate and staurosporine. J. Immunol. 146:3056-3062,1991. Roche e Silva M. A brief history of inflammation. In: Inflammation. Vane JR and Ferreira SH [Eds.]. Springer-Verlag, N.Y. 1978. Romer LH, Albelda SM, DeLisser H, Buck CA. PECAM expression on human endothelial cells is altered by TNFa and ylFN. J. Cell. Biol. 115:69a, 1991. Rosentein Y, Park JK, Hahn WC, Rosen FS, Bierer BE and Burakoff SJ. CD43, a molecule defective in Wiskott-Aldrich syndrome, binds ICAM-1. Nature 354:233-235,1991. Ross L, Hassman F and Malony L. Inhibition of Molt-4-endothelial adherence by synthetic peptides from the sequence of ICAM-1. J. Biol. Chem. 267:8537-8543, 1992. Rot A. Endothelial cell binding of NAP-l/TL-8: role in neutrophil emigration. Immunol. Today 13:291-294, 1992. Rothlein RR, Mainolfi EA and Kishimoto TK. Treatment of inflammation with anti-ICAM-1. Res. Immunol. 144:735-739,1993. Rubin LL, Hall DE, Porter S, Barbu K, Cannon C, Horner HC, Janatpour M, Liaw CW, Manning K, Morales J, Tanner LI, Tomaselli KJ and Bard F. A cell culture model of the blood-brain barrier. J Cell Biol. 115:1725-1735,1991. Rubin BB, Rotstein OD, Lukacs G, Bailey D, Romaschin A and Walker PM. Decreased leukocyte adhesion with anti-CD 18 monoclonal antibodies is mediated by receptor internalization. Surgery 112:263-168,1992. Rugg C, Postigo AA, Sikorski EE, Butcher EC, Pytela K and Erie DJ. Role of integrin ctyPy/o^Pp in lymphocyte adherence to fibronectin and VCAM-1 and in homotypkfcell clustering. J. Cell. Biol. 117:179-189,1992. Ruszczak Z, Detmer M, Imcke E and Orfanos CE. Effects of rIFN alpha, beta and gamma on the morphology, proliferation and cell surface antigen expression of human dermal microvascular endothelial cells in vitro. J. Invest. Dermatol. 95:693-699,1990. Rutten MJ, Hoover RL and Karnovsky MJ. Electrical resistance and macromolecular permeability of brain endothelial monolayer cultures. Brain Res. 425:301-310, 1987. Ryan GB and Majno G. Inflammation. Upjohn Company, Kalamazoo. 1977. Ryan DH, Nuccie BL, Abboud CN and Winslow JNM. Vascular cell adhesion molecule-1 and the integrin VLA-4 mediate adhesion of human B cell precursors to cultured bone marrow adherent cells. J. Clin. Invest. 88:995-1004,1991.  278  Sadahiro M, McDonald TO and Allen MD. Reduction in cellular and vascular rejection by blocking leukocyte adhesion molecule receptors. Am. J. Path. 142:675-683, 1993. Salmi M and Jalkanen S. A 90-kilodalton endothelial cell molecule mediating lymphocyte binding in humans. Science 257:1407-1409,1992. Sanna PP, Weiss F, Samson ME, Bloom FE and Pich EM. Rapid induction of tumor necrosis factor a in the cerebrospinal fluid after intra cerbroventricular injection of lipopolysaccharide revealed by a sensitive capture immuno-PCR assay. Proc. Natl. Acad. Sci. U.S.A. 92:272-275,1995. Sasaki K, Okouchi Y, Pabst R and Rothkotter H-J. Three-dimensional detection of the expression of intercellular adhesion molecule-1 (ICAM-1) in the high endothelial venule (HEY) of the rat lymph node. Microscopy Res. Technique 25:264-265, 1993. Sasseville VG, Newman WA, Lackner AA, Smith MO, Lausen NCG., Beall D and Ringler DJ. Elevated vascular cell adhesion molecule-1 in AIDS encephalitis induced by simian immunodeficiency virus. Am. J. Path. 141:1021-1030,1992. Satoh J-I, Kastrukoff LF and Kim SU. Cytokine-induced expression of intercellular adhesion molecule-1 (ICAM-1) in cultured human oligodendrocytes and astrocytes. J. Neuropath. Exp. Neurol. 3:215-226,1991. Scheynius A, Comp RL and Pure E. Reduced contact sensitivity reactions in mice treated with monoclonal antibodies to leukocyte function-associated molecule-1 and intercellular adhesion molecule-1. J. Immunol. 150:655-663,1993. Schimmenti LA, Yan HC, Madri JA and Albelda SM. Platelet endothelial cell adhesion molecule, PECAM-1, modulates cell migration. J. Cell. Physiol. 153:417-428, 1992. Seekamp A and Ward, PA. Ischemia-reperfusion injury. Agents and Actions Suppl. 41:137-152,1993. Seekamp A, Till GO, Mulligan MS, Paulson JC, Anderson DC, Miyasaka M, Ward PA. Role of selectins in local and remote tissue injury following ischemia and reperfusion. Am. J. Path. 144:592-598,1994. Sen GC and Lengyel P. The interferon system. J. Biol. Chem. 267:5017-5020,1992. Sepp NT, Gille J, Li LJ, Caughman W, Lawley TJ and Swerlick RA. A factor in human plasma permits persistent expression of E-selectin by human endothelial cells. J. Invest. Dermatol. 102:445-450,1994. Seth R, Raymond FD and Makgoba MW. Circulating ICAM-1 isoforms: diagnostic prospects for inflammatory and immune disorders. The Lancet. 338:83-84,1991. Sharief MK, McLean B and Thompson EJ. Elevated serum levels of tumor necrosis factor-a in Guillain-Barre syndrome. Ann. Neurol. 33:591-596,1993a.  279  Sharief MIC, Noori MA, Ciardi M, Cirelli A and Thompson EJ. Increased levels of circulating ICAM-1 in serum and cerebrospinal fluid of patients with active multiple sclerosis. Correlation with TNFa and blood-brain barrier damage. J. Neuroimmunol. 43:15-22,1993b. Shasby DM and Shasby SS. Effects of calcium on transendothelial albumin transfer and electrical resistance. J. Appl. Physiol. 60:71-79,1986. Shimizu Y, Newman W, Gopal TV, Horgan KJ, Graber N, Beall LD, van Seventer GA and Shaw S. Four molecular pathways of T cell adhesion to endothelial cells: roles of LFA-1, VCAM-1 and ELAM-1 and changes in pathway hierarchy under different activation conditions. J. Cell Biol. 113:1203-1212,1991a. Shimizu Y, Shaw S, Graber N, Venkat Gopal T, Horgan KJ, van Seventer GA and Newman W. Activation-independent binding of human memory T cells to adhesion molecule ELAM-1. Nature 349:799-802,1991b. Shu HB, Agranoff AB, Nabel EG, Leung K, Duckett CS, Neish AS, Collins T and Nabel GJ. Differential regulation of vascular cell adhesion molecule 1 gene expression by specific N F - K B subunits in endothelial and epithelial cells. Molec. Cell. Biol. 13:6283-6289, 1993. Simmons DL, Walker C, Power C and Pigott R. Molecular cloning of CD31, a putative intercellular adhesion molecule closely related to carcinoembryonic antigen. J. Exp. Med. 171:2147-2152,1990. Simmons PJ, Masinovsky B, Longenecker BM, Berebson R, Torok-Storb B and Gallatin WM. Vascular cell adhesion molecule-1 expressed by bone marrow stromal cells mediates the binding of hematopoietic progenitor cells. Blood 80:388-395,1992. Simmons RD and Cattle BA. Sialyl ligands facilitate lymphocyte accumulation during inflammation of the central nervous system. J. Neuroimmunol. 41:123-130,1992. Skundric DS, Houston K, Shaw M, Tse HY and Raine CS. Experimental allergic encephalomyelitis. T cell trafficking into the central nervous system in a resistant Thy-1 cogenic mouse strain. Lab. Invest. 71:671-679,1994. Sligh JE Jr, Ballantyne CM, Rich SS, Hawkins HL, Wayne Smith C, Bradley A and Beaudet AC. Inflammatory and immune responses are impaired in mice deficient in intercellular adhesion molecule 1. Proc. Natl. Acad. Sci. USA. 90:8529-8533, 1993. Smart SJ and Casale TB. TNF-a-induced transendothelial neutrophil migration is IL-8 dependent. Am. J. Physiol. 266:L238-L245,1994. Smith CW, Rothlein R, Hughes BJ, Mariscalco MM, Rudloff HE, Schmalstieg FC, Anderson DC. Recognition of an endothelial determinant for CD18-dependent human neutrophil adherence and trans endothelial migration. J. Clin. Invest. 82:1746-1756,1988. Smith CW, Marlin SD, Rothlein R, Toman C and Anderson DC. Cooperative interactions of LFA-1 and mac-1 with intercellular adhesion molecule-1 in facilitating adherence and transendothelial migration of human neutrophils in vitro. J. Clin. Invest. 83:2008-2017, 1989. 280  Sobel RA, Mitchell ME and Fondien G. Intercellular adhesion molecule-1 (ICAM-1) in cellular immune reactions in the human central nervous system. Am. J. Path. 136:1309-1316,1990. Sollberg S, Peltonen J, Uitto J and Jimenez SA. Elevated expression of p i and (32 integrins, intercellular adhesion molecule 1, and endothelial leukocyte adhesion molecule 1 in the skin of patients with systemic sclerosis of recent onset. Arth. Rheum. 35:290-298,1992. Springer TA. Adhesion receptor of the immune system. Nature. 346:425-434,1990. Springer TA. Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell. 76:301-314,1994. Staunton DE, Merluzzi VJ, Rothelin R, Barton R, Marlin SD and Springer TA. A cell adhesion molecule, ICAM-1, is the major surface receptor for rhinovirus. Cell 56:849-853, 1989. Staunton DE, Dustin ML, Springer TA. Functional cloning of ICAM-2, a cell adhesion ligand for LFA-1 homologous to ICAM-1. Nature 339:61-64,1989. Steffen BJ, Butcher EC and Engelhardt B. Evidence for involvement of ICAM-1 and VCAM-1 in lymphocyte interactions with endothelium in experimental autoimmune encephalomyelitis in the central nervous system in the SJL/J mouse. Am. J. Path. 145:189-201,1994. Steinhoff G, Behrend M, Schrader B, Duijvestijn AM and Wonigeit K. Expression patterns of leukocyte adhesion ligand molecules on human liver endothelia. Lack of ELAM-1 and CD62 inducibility on sinusoidal endothelia and distinct distribution of VCAM-1, ICAM-1, ICAM-2 and LFA-3. Am. J. Path. 142:481-488,1993. Stenberg PE, McEver RP, Shuman MA, Jacques YU and Bainton DF. A platelet a-granule membrane protein (GMP-140) is expressed on the plasma membrane after activation. J. Cell. Biol. 101:880-886,1985. Sticherling M, Hetzel F, Schroder JM and Christophers E. Time- and stimulus-dependent secretion of NAP-l/IL-8 by human fibronectin and endothelial cells. J. Invest. Dermatol. 101:573-576,1993. Stockinger H, Gadd SJ, Eher R, Majdic G, Schreiber W, Kasinrerk W, Strass B, SchnabI E and Knapp W. Molecular characterization and functional analysis of the leukocyte surface protein CD31. J. Immunol. 145:3889-3897,1990. Sung C-P, Arleth AJ, Nambi P. Evidence of invovlement of intercellular adhesion molecule-1 (ICAM-1) by human vascular endothelial cells. Pharmacology 48:143-146,1994. Swerlick RA, Garcia-Gonzalez E, Kubota Y, Xu Y and Lanky TJ. Studies of the modulation of MHC antigen and cell adhesion molecule expression on human dermal microvascular endothelial cells. J. Invest. Dermatol. 97:190-196,1991.  281  Swerlick RA, Lee KH, Li L-J, Sepp NT, Caughman SW. and Lawley, T.J. Regulation of vascular cell adhesion molecule 1 on human dermal microvascular endothelial cells. J. Immunol. 149:698-705,1992. Takada A, Ohmori K, Yoneda T, Tsuyuoka K, Hasegawa A, Kiso M and Kannagi R. Contribution of carbohydrate antigens sialyl Lewis A and sialyl Lewis X to adhesion of human cancer cells to vascular endothelium. Cancer Res. 53:354-361, 1993. Tamatani T and Miyasaka M. Identification of monoclonal antibodies reactive with the rat homolog ICAM-1 versus activated lymphocytes to high endothelial cells. Intl. Immunol. 2:165-171,1990. Tan TQ, Smith CW, Hawkins EP, Mason EO and Kaplan SL. Hematogenous bacterial meningitis in an intercellular adhesion molecule-1-deficient infant mouse model. J. Infect. Dis. 171:342-349,1995. Tanaka Y, Adams DH and Shaw S. Proteoglycan on endothelial cells present adhesion-inducing cytokines to leukocytes. Immunol. TOday 14:111-115,1993. Tanaka Y, Albelda SM, Horgan KJ, van Seventer GA, Shimizu Y, Newman W, Haliam J, Newman PJ, Buck CA and Shaw S. CD31 expressed on distinctive T cell subsets is preferential amplifier of pi integrin-mediated adhesion. J. Exp. Med. 176:245-253,1992. Tanaka M, Sato A, Makino M and Tabira T. Binding of an SJL T cell clone specific for myelin basic protein to SJL brain microvessel endothelial cells is inhibited by anti-VLA-4 or its ligand, anti-vascular cell adhesion molecule 1 antibody. J. Neuroimmunol. 46:253-258,1993. Taupin V, Toulmond S, Serrano A, Benavides J and Zavala F. Increase in IL-6, IL-1 and TNF levels in rat brain following traumatic lesion. J. Neuroimmunol. 42:177-186, 1993. Taylor PM, Rose ML, Yacoub MH and Pigott R. Induction of Vascular adhesion molecules during rejection of human cardiac allografts. Transplant. 54:451-457, 1992. Territo M, Berliner JA and Fogelman AM. Effect of monocyte migration on low density lipoprotein transport across aortic endothelial cell monolayers. J. Clin. Invest. 74:2279-2284,1984. Throp KM, Southern C, Bird IN and Matthews N. Tumor necrosis factor induction of ELAM-1 and ICAM-1 on human umbilical vein endothelial cells - analysis of tumor necrosis factor-receptor interactions. Cytokine 4:313-319,1992. Tomczok J, Sliwa-Tomczok W, Klein CL, Bittinger F and Kirkpatrick CJ. Application of immunogold labelling for light and electron microscopic localization of endothelial leukocyte adhesion molecule 1 (ELAM-1) on cultured human endothelial cells. Micron. 25:257-266,1994. Traugott U and Lebon P. Multiple sclerosis: involvement of interferons in lesion pathogenesis. Ann. Neurol. 24:243-251, 1988.  282  Travis J. Biotech gets a grip on cell adhesion. Science. 260:906-908,1993. Tsukada N, Matsuda M, Miyagi Kand Yanagisawa N. Adhesion of cerebral endothelial cells to lymphocytes from patients with multiple sclerosis. Autoimmunity 14:329-333,1993. Tsukada N, Matsuda M, Miyagi K and Yanagisawa N. In vitro intercellular adhesion molecule-1 expression on brain endothelial cells in multiple sclerosis. J. Neruoimmunol. 49:181-187,1994. Turner MR. Electrical resistances of cultured bovine arterial endothelium in solutions of various resistivities. Expt. Physiol. 77:741-748,1992. van der Wal AC, Das PK, Tigges AJ and Becker AE. Adhesion molecules on the endothelium and mononuclear cells in human atherosclerotic lesions. Am. J. Pathol. 141:1427-1433,1992. Van Kooyk Y, van de Wiel-van Kemenade E, Weder P, Huijbens RJF and Figdor CG. Lymphocyte function-associated antigen 1 dominates very late antigen 4 in binding of activated T cells to endothelium. J. Exp. Med. 177:185-190,1993. Van Mourik JA, Leeksma OC, Reinders JH, deGroot PG, Zandbergen-Spaargaren J. Vascular endothelial cells synthesize a plasma membrane protein indistinguishable from the platelet membrane glycoprotein Ha. J. Biol. Chem. 260:11300-11306,1985. Vaporciyan AA, DeLisser HM, Yan H-C, Mendiguren II, Thom SR, Jones ML, Ward PA and Albelda SM. Involvement of platelet-endothelial cell adhesion molecule-1 in neutrophil recruitment in vivo. Science 262:1580-1582,1993. Varani J, Dame MK, Gibbs DF, Taylor CG, Weinberg JM, Sheyevitz J and Ward PA. Human umbilical vein endothelial cell killing by activated neutrophils. Loss of sensitivity to injury is accompanied by decreased iron content during in vitro cultures and is restored with exogenous iron. Lab. Invest. 66:708-714,1992. Vassalli P. The pathophysiology of tumor necrosis factor. Ann. Rev. Immunol. 10:411-452,1992. Verbeek MM, Otte-Holler I, Westphal JR, Wesseling P, Ruite PJ and de Waal RMW. Accumulation of intercellular adhesion molecule-1 in senile plaques in brain tissue of patients with Alzheimer's disease. Am. J. Path. 144:104-106,1994. Von Asmuth EJU, Smeets EF, Ginsel LA, Onderwater JJM, Leeuwenberg JFM and Buurman WA. Evidence for endocytosis of E-selectin in human endothelial cells. Eur. J. Immunol. 22:2519-2526,1992. Vonderheide RH and Springer TA. Lymphocyte adhesion through very late antigen 4: evidence for a novel binding site in the alternatively spliced domain of vascular cell adhesion molecule 1 and an additional alpha 4 integrin counter-receptor on stimulated endothelium. J. Exp. Med. 175:1433-1442,1992. Walcheck B, Watts G and Jutila MA. Bovine y/6 T cells bind E-selectin via a novel glycoprotein receptor: first characterization of a lymphocyte / E-selectin interaction in an animal model. J. Exp. Med. 178:853-863,1993. 283  Walsh, GM, Mermod J-J, Hartnell A, Kay AB and Wardlaw AJ. Human eosinophil, but not neutrophil, adherence to IL-1-stimulated human umbilical vascular endothelial cells is o^pi (very late antigen-4) dependent. J. Immunol. 146:3419-3423,1991. Wankowicz Z, Megyeri P and Issekutz A. Synergy between tumor necrosis factor a and interleukin-1 in the induction of polymorphonuclear leukocyte migration during inflammation. J. Leukocyte Biol. 43:349-356,1988. Ward PA and Mark RM. The acute inflammatory reaction. Curr. Opinion in IMmunol. 2:5-9,1989. Washington R, Burton J, Todd the third RF, Newman W, Dravgovic L and Dore-Duffy P. Expression of immunologically relevant endothelial cell activation antigens on isolated central nervous system microvessels from patients with multiple sclerosis. Ann. Neurol. 35:89-97,1994. Welder CA, Lee DHS and Takei F. Inhibition of cell adhesion by microspheres coated with recombinant soluble intercellular adhesion molecule-1. J. Immunol. 150:2203-2210,1993. Weller PF, Rand TH, Goelz SE, Chi-Rosso G and Lobb RR. Human eosinophil adherence to vascular endothelium mediated by binding to vascular cell adhesion molecule 1 and endothelial leukocyte adhesion molecule 1. Proc. Natl. Acad. Sci. USA. 88:7430-7433, 1991. Wellicome SM, Thornhill MH, Pitzalis C, Thomas DS, Lanchbury JSS, Panayi GS and Haskard DO. A monoclonal antibody that detects a novel antigen on endothelial cells that is induced by tumor necrosis factor, IL-1, or lipopolysaccharide. J. Immunol. 144:2558-2565,1990. Wheatley EM, McKewon-Longo PJ, Vincent PA and Saba TM. Incorporation of fibronectin into matrix decreases TNF-induced increase in endothelial monolayer permeability. Am. J. Physiol. 265:L148-L157,1993. Whelan J, Ghersa P, von Huijsduijnen RH, Gray J, Chandra G, Talabot F and DeLamarter JF. An NF-icB-like factor is essential but not sufficient for cytokine induction of endothelial leukocyte adhesion molecule-1 (ELAM-1) gene transcription. Nucleic Acid Res. 19:2645-2653,1991. Whitcup SM, Wakefield D, Li Q, Nussenblatt RB and Chan C-C. Endothelial leukocyte adhesion molecule-1 in endotoxin-induced uveitis. Invest. Ophth. Vis. Sci. 33:2626-2630, 1992. Wieftaer C, Gehrmann J, Lindholm D, Topper R, Kreutzberg GW and Hossmann KA. Expression of transforming growth factor-pl and interleukin-ip mRNA in rat brain following transient forebrain ischemia. Acta Neuropathol. 86:439-446, 1993. Wilcox CE, Ward AM, Evans A, Baker D, Rothlein R and Turk JL. Endothelial cell expression of the intercellular adhesion molecule (ICAM-1) in the central nervous system of guinea pigs during acute and chronic relapsing experimental allergic encephalomyelitis. J. Neuroimmunol. 30:43-51,1990.  284  Willenborg DO, Simmons RD, Tamatani T and Miyasaka M. ICAM-1-dependent pathway is not critically involved in the inflammatory process of autoimmune encephalomyelitis or in cytokine-induced inflammation of the central nervous system. J. Neuroimmunol. 45:147-154,1993. Williams JC, Jurkovich GJ and Maier RV. Interferon-y: a key immunoregulatory lymphokine. J. SUrg. Res. 54:79-93,1993. Wisniewski HM and Lossinsky AS. Structural and functional aspects of the interaction of inflammatory cells with the blood-brain barrier in experimental brain inflammation. Brain Path. 1:89-96,1991. Woodroffe MN, Sarna GS, Washwa M, Hayes GM, Loughlin AJ, Tinker A and Cuzner ML. Detection of interleukin-1 and interleukin-6 in adult rat brain, following mechanical injury, by in vivo microdialysis: evidence of a role for microglia in cytokine production. J. Neuroimmunol. 33:227-236,1991. Woodroffe MN and Cuzner ML. Cytokine mRNA expression in inflammatory multiple sclerosis lesions: detection by non-radioactive in situ hybridization. Cytokine 5:583-588,1993. Yednock TA, Cannon C, Fritz LC, Sanchez-Madrid F, Steinman L and Karin N. Prevention of experimental autoimmune encephalomyelitis by antibodies against a4(31 integrin. Nature. 356:63-66,1992. Yi ES and Ulich TR. Endotoxin, interleukin-1 and tumor necrosis factor cause neutrophil-dependent microvascular leakage in postcapillary venules. Am. J. Path. 140:659-663,1992. Yu C-L, Haskard DO, Cavender D, Johnson AR and Ziff M. Human gamma interferon increases the binding of T lymphocytes to endothelial cells. Clin. Exp. Immunol. 62:554-560,1985. Yu C-L, Haskard D, Cavender D, Ziff M. Effect of bacterial lipopolysaccharide on the binding of lymphocytes to endothelial cell monolayers. J. Immunol. 136:569-573, 1986. Xu H, Gonzalo JA, Stpierre Y, Williams IR, Kupper TS, Cotran RS, Springer TA and Gutierrezramos JC. Leukocytosis and resistance to septic shock in intercellular adhesion molecule 1-deficient mice. J. Exp. Med. 180:95-109,1994.  285  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0088815/manifest

Comment

Related Items