Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

In vitro modulation of classic II MHC antigen expression by human cerebral endothelium and endothelial… Huynh, Hanh Kim 1994

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

Item Metadata


831-ubc_1994-894105.pdf [ 10.19MB ]
JSON: 831-1.0076827.json
JSON-LD: 831-1.0076827-ld.json
RDF/XML (Pretty): 831-1.0076827-rdf.xml
RDF/JSON: 831-1.0076827-rdf.json
Turtle: 831-1.0076827-turtle.txt
N-Triples: 831-1.0076827-rdf-ntriples.txt
Original Record: 831-1.0076827-source.json
Full Text

Full Text

IN VITRO MODULATION OF CLASS II MHC ANTIGEN EXPRESSION BYHUMAN CEREBRAL ENDOTHELIUM AND ENDOTHELIAL CELL -LYMPHOCYTE INTERACTIONS BY INTERFERONS y AND ibyHANH KIM HUYNHM.Sc., Cell Biology, The University of British Columbia, 1989A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFINTERDISCIPLINARY DOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Departments of Neuroscience and Pathology)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAApril 1994© Hanh Kim Huynh, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department5 of i6t S(:iFCEz/The University of British ColumbiaVancouver, CanadaDate /qDE-6 (2/88)11ABSTRACTIn autoimmune demyelinating disorders of the central nervous system (CNS), theblood-brain-barrier (BBB) becomes permeable to plasma proteins and circulating leukocytes.Studies on Multiple Sclerosis (MS) suggest the involvement of T4+ lymphocytes and Ia+macrophages in lesion extension and demyelination. Circulating lymphocytes recognizeantigen only when it is complexed with Ia molecules on the surface of antigen presenting cells.Recent studies aiming at defining cell populations in the CNS capable of expressing Ia antigen(Ia Ag) indicate Ia Ag expression by EC, astrocytes, microglia and macrophages in MSlesions, while in experimental allergic encephalomyelitis, Ia Ag expression by brainendothelium has been reported by some, but not other investigators. The role of cerebralendothelium in CNS inflammation, therefore, remains ill defined and rather controversial.The objective of this thesis was to investigate the: 1) Induction of Ia Ag on humanbrain microvessel endothelial cells (HBMEC) treated with interferons (IFN) y and , and 2)modulation of T-lymphocyte adhesion and migration across HBMEC monolayers by IFN-yand . To address these issues, an in vitro model of the human BBB was utilized. The resultsindicate that HBMEC do not constitutively express Ta Ag. Treatment with IFN-y, not 3, resultsin de novo, polarized expression of Ia Ag. Co-incubation with IFN-13 downregulates the IFNy-induced Ia Ag expression. Treatment with IFN-y, not 1, induces morphological andfunctional changes of HBMEC associated with increased permeability to macromolecules.IFN-y induced changes do not appear in cultures incubated with both cytokines.Treatment of HBMEC with IFN-y upregulates the adhesion and migration of restingand activated T lymphocytes across the monolayers. T-lymphocyte activation alone greatlyaugments both processes. Treatment with IFN-f3 has no effect on lymphocyteadhesion/migration, however, the IFN-y-mediated increase in adhesion/migration is111suppressed by IFN-3. MAb blocking studies suggest a direct role of Ta molecules inadhesion/migration.These studies indicate a potentially important role of HBMEC in CNS inflammationand increase our understanding of some of the factors involved in the recruitment oflymphocytes into chronic inflammatory sites in the CNS. Therapeutic interventions utilizingmAbs against HLA-DR molecules as well as the use of cytokines, such as TFN-3, that cansuppress TFN-y-mediated responses, may have considerable therapeutic potential.ivTABLE OF CONTENTSAbstract iiTable of Contents ivList of Table viiiList of Figures ixList of Abbreviations .xiiiAcknowledgements xviDedication xviiCHAPTER ONE INTRODUCTION1.1 Inflammation 11.1.1 Definition 11.1.2 General aspects 21.1.3 Role of EC 41.2 Interferons 71.2.1 IFN-y 81.2.2 IFN-3 91.3 Major histocompatibility molecules 91.3.1 Definition 91.3.2 Class IMHC 101.3.3 Class II MHC 10a) Role in autoimmune disorder 11b) Expression by EC 121.4 Central nervous system inflammation and autoimmunity. . . .151.4.1 Permeability of the Blood-Brain-Barrier 151.4.2 Lymphocyte infiltration 17a) Ia Ag association 17b) Lymphocyte-EC adhesion 18c) Transendothelial migration of lymphocytes 201.4.3 FVIIIR:Ag in primary cultures of HBMEC 211.5 Summary and objectives 24CHAPTER TWO MATERIALS AND METHODS2.1 Isolation and culture of HBMEC 272.2 Antibodies 282.3 Induction of Ia Ag expression on HBMEC by IFN-y and 3. . .292.3.1 Treatment of primary HBMEC cultures 29V2.3.2 Light microscopic immunocytochemicallocalization of Ia Ag 302.3.3 Immunoelectron microscopy 312.3.4 Enzyme linked immunosorbent assay (ELISA) 322.3.5 Quantitation of Ta Ag expression by HBMEC 322.4 Scanning electron microscopy 332.5 Permeability studies 332.6 Growth studies 342.7 Preparation of T lymphocytes 352.8 Activation of T cells 362.9 HBMEC-lymphocyte adhesion assay 372.10 HBMEC-lymphocyte migration assay 382.11 Monoclonal antibody-blocking studies 392.12 Quantitation of lymphocyte adhesion and migration 402.13 Transmission electron microscopy 402.14 Localization of FVIIIR:Ag in HBMEC 412.14.1 In vitro drug treatment of EC 412.14.2 Immunoelectron microscopy for FVIIIR:Ag 412.15 Statistical analysis 42CHAPTER THREE RESULTS3.1 Human brain microvessel endothelial cell 433.2 Immunocytochemical localization of FVIIIR:Ag 443.3 Induction of Ia Ag expression on primary cultures of HBMEC 453.3.1 Effects of recombinant human TFN-y 453.3.2 Effects of recombinant human IFN- 463.4 Kinetics of the downregulation of Ta Ag expression by IFN- .473.5 Effects of IFN-y and IFN-13 on cell morphology,organization and growth 493.5.1 IFN-y 493.5.2 IFN-13 503.6 Permeability of HBMEC monolayers 50vi3.7 Lymphocyte characterization 513.8 Human T lymphocyte adhesion to untreatedand cytokine-stimulated HBMEC .523.9 Adhesion of activated T-lymphocytes to untreatedand cytokine-stimulated HBMEC .533.10 Effects of blocking antibodies on lymphocyte adhesion 553.11 Transendothelial migration of resting T lymphocytes 553.12 Migration of activated T-lymphocytes acrossuntreated and cytokine treated HBMEC monolayers 573.13 Effects of blocking antibodies on lymphocyte migration. . . . 583.14 Effects of calcium ionophore A23 187, EGTA and IFN-y onthe constitutive pathway of factor VIIIR:Ag release 58CHAPTER FOUR DISCUSSION4.1 Influence of cytokines on Ia Ag expression on HBMEC. . . . 604.1.1 Human brain microvessel EC 604.1.2 Induction of Ia Ag expressionon primary cultures of HBMEC 604.1.3 Surface localization of Ia Ag on HBMEC 624.1.4 Effects of IFN-13 on Ia Ag expression by HBMEC. . .634.1.5 Regulatory mechanism of Ia Ag expression 644.1.6 Kinetic studies on the modulation of Ia Agexpression by interferons y and fE 644.2 Effects of interferons y and 3 on the morphologicalphenotype and growth of HBMEC, organization of themonolayers and permeability to macromolecules 664.2.1 Effects of IFN-y and IFN-f3 on HBMEC growth 674.2.2 Effects of interferons y and 1 on HBMECmorphology and organization of the EC monolayers. 674.2.3 Permeability of IFN-y treated HBMEC monolayersto macromolecules 694.3 Significance of Ia Ag expression on HBMEC 704.4 Adhesion of resting and anti-CD3 stimulated lymphocytes tountreated, IFN-y and/or IFN-13 treated HBMEC 724.4.1 Activation of lymphocytes 724.4.2 Adhesion of resting lymphocytes to untreated,IFN-y and/or IFN-f3 treated HBMEC 734.4.3 Adhesion of activated lymphocytes to untreated,IFN-y and/or IFN-13 treated HBMEC 78vii4.5 Migration of resting and anti-CD3 stimulated lymphocytesacross untreated and cytokine treated HBMEC 844.5.1 Migration of resting lymphocytes across untreated,IFN-y and/or IFN-3 treated HBMEC monolayers. . . . 844.5.2 Migration of anti-CD3 stimulated lymphocytes acrossuntreated, IFN-y and/or IFN-3 treated HBMECmonolayers 864.6 Effects of IFN-’y on the storage and releaseof FVIIIR:Ag from HBMEC in primary culture 914.6.1 Immunocytochemical localization of FVIIIR:Agin HBMEC 914.6.2 Effects of IFN-y on the storage and/or releaseof FVIIIR:Ag from HBMEC 93CHAPTER FIVE CONCLUSIONS5.1 Summary and conclusions 955.2 Future prospects 985.3 Significance of this thesis 101TABLE 103FIGURES 104REFERENCES 214vii’LIST OF TABLETable 1. Permeability of HBMEC Monolayers to HRP: Quantitation of endocytosisand tight junction permeability of untreated and IFN-y-treated HBMEC 103ixLIST OF FIGURESFigure 1. Diagram of the longitudinal section of the double chemotacticculturing chamber 104Figure 2. Primary cultures of HBMEC grown on plastic wells or cellagen discsform confluent monolayers 106Figure 3. Immunoperoxidase staining for FVIIIR:Ag and UEA I confirming theendothelial origin of primary cultures of HBMEC 108Figure 4. Ultrastructural studies demonstrating the elongated morphology of ECwith their overlapping processes 110Figure 5. Ultrastructural demonstration of the pentalaminar configurationcharacteristic of the tight junctions that are present in areas ofcell to cell contact (a-d) 112Figure 6. Cytoplasmic morphology of an untreated HBMEC (a, b) 117Figure 7. Immunocytochemical localization of FVIIIR:Ag in untreated EC (a-d) 120Figure 8. Time course of Ia Ag induction on HBMEC by IFN-y 122Figure 9. Dose - response of Ia Ag induction by IFN-y on HBMEC 124Figure 10. Ia antigen expression by HBMEC detected by immunogold silver stainingin untreated/IFN-y treated EC at various times and also in culturescoincubated with anti-IFN-y antibody 126Figure 11. Immunogold staining of HBMEC for the demonstration of Ia antigen inuntreated/IFN-y treated EC 128Figure 12. Ia antigen expression by HBMEC detected by immunogold silver stainingin cells treated with IFN-f3 alone or a combination of IFN-(3 and y 130Figure 13. Dose-response of Ia Ag expression by HBMEC treated withIFN-y and/or IFN-3 132xFigure 14. Effects of different treatments of IFN-y and 3 on Ia Agexpression by HBMEC 134Figure 15. Quantitation by ELISA of Ia Ag expression by HBMEC treated withIFN-y and/or IFN-f 136Figure 16. Phase contrast microscopic demonstration of morphological alterationinduced by IFN-y treatment of confluent cultures of HBMEC 138Figure 17. SEM demonstration of morphological alteration induced byIFN-y treatment of confluent cultures of HBMEC 140Figure 18. Effects of IFN-y and upon the growth of primary cultures of HBMEC. . . .142Figure 19. SEM of HBMEC grown in the presence of IFN-13 alone (a) or a combinationof IFN-(3 and y (b) 144Figure 20. HRP localization in untreated (A) and (B) and IFN-y-treated (C-F)confluent HBMEC monolayers 147Figure 21. Expression of IL-2R on resting and anti-CD3 stimulated T lymphocytes . . . 151Figure 22. Immunoperoxidase staining for the demonstration of resting T lymphocyteand untreated HBMEC adhesion 153Figure 23. Adhesion of resting lymphocytes to IFN-y treated HBMEC 155Figure 24. Adhesion of resting lymphocytes to IFN-f3 treated HBMEC 157Figure 25. Adhesion of resting lymphocytes to untreated andcytokine-treated HBMEC 159Figure 26. SEM of adhesion of resting T lymphocytes to untreated HBMEC 161Figure 27. SEM of adhesion of resting T lymphocytes to IFN-y treated HBMEC 166Figure 28. Adhesion of anti-CD3 stimulated T lymphocytes to untreated HBMEC. . . . 169xiFigure 29. Adhesion of anti-CD3 stimulated T lymphocytes to IFN-y treated HBMEC. .171Figure 30. Adhesion of anti-CD3 stimulated T lymphocytes toIFN-y and 3 treated HBMEC 173Figure 31. Summary bar graph of adhesion of anti-CD3 activated lymphocytesto untreated and cytokine-treated HBMEC 175Figure 32. SEM demonstration of adhesion of anti-CD3 activated lymphocytesto untreated and cytokine-treated HBMEC 177Figure 33. Adhesion of stimulated lymphocytes to untreated HBMEC 181Figure 34. Adhesion of stimulated T cells to untreated HBMEC 183Figure 35. Light micrograph of adhesion of resting lymphocytes toIFN-y and anti-human HLA-DR treated HBMEC 185Figure 36. Light micrograph of adhesion of anti-CD3 stimulated lymphocytes toIFN-y and anti-human HLA-DR treated HBMEC 187Figure 37. Migration of resting lymphocytes across untreated and cytokinetreated HBMEC 189Figure 38. Summary bar graph of migration of resting and anti-CD3 activatedT cells across untreated and IFN-y and/or IFN-13 treated HBMEC 191Figure 39. TEM examination of migration of resting T lymphocytes acrossuntreated HBMEC 193Figure 40. TEM examination of transendothelial migration of restingT lymphocytes across untreated HBMEC 198Figure 41. TEM examination of the integrity of the EC monolayers at the endof migration of resting lymphocytes across IFN-y treated HBMEC 200Figure 42. Light microscopic examination of migration of activated lymphocytesacross untreated HBMEC 203xiiFigure 43. TEM examination of migration of anti-CD3 stimulated lymphocytesacross untreated and IFN-y treated HBMEC 205Figure 44. TEM examination of migration of anti-CD3 stimulated lymphocytesacross IFN-y treated HBMEC 208Figure 45. Immunocytochemical localization of FVIIIR:Ag in chemical,cytokine-treated HBMEC 210Figure 46. Summary bar graph of the immunocytochemical localization of FVIIIR:Agin HBMEC 212xliiLIST OF ABBREVIATIONSAD Autoimmune DisorderBBB Blood-brain-barrierBSA Bovine Serum AlbumincAMP Cyclic Adenosine MonophosphateCD3 Cluster of Differentiation 3CNS Central Nervous Systemcon A Concanavalin ACSF Cerebrospinal FluidEAE Experimental Allergic/Autoimmune EncephalomyelitisEAN Experimental Autoimmune NeuritisEAU Experimental Autoimmune UveoretinitisEC Endothelial CellEGTA Ethyleneglycol-tetraacetic acidELISA Enzyme Linked Immunosorbent AssayFACS Fluorescence Activated Cell SorterFCS Fetal Calf SerumFVIIIR:Ag Factor VIII related antigenGAMIgG Goat anti-mouse immunoglobulin GHBMEC Human Brain Microvessel Endothelial CellxivHBSS Hanks’ Balanced Salt SolutionHDMEC Human Dermal Microvessel Endothelial CellHLA Human Leucocyte AntigenHRP Horseradish peroxidaseHS Horse SerumHUVEC Human Umbilical Vein Endothelial CellIa Ag Immune associated antigenICAM-1 Intercellular adhesion molecule-iIFN- Human recombinant interferon-betaIFN-y Human recombinant interferon-gammaIFNs InterferonsIgG Immunoglobulin G isotypeIL-i Interleukin-iIL-2 Interleukin-2IL-2R Interleukin-2 receptorLCA Leucocyte common antigenLFA- 1 Lymphocyte Function-associated Antigen-imAb monoclonal antibodyMHC Major Histocompatibility complexmRNA messenger ribonucleic acidMS Multiple SclerosisxvNGS Normal Goat SerumPBS Phosphate buffered salinePECAM-1 Platelet/endothelial cell adhesion molecule-iRNA Ribonucleic acidSEM Scanning Electron MicroscopyTCR T cell receptorTEM Transmission Electron MicroscopyTNF-a Tumor necrosis factor-alphaUEA-I Ulex Europaeus type IVCAM-i Vascular cell adhesion molecule-iVLA-4 Very late antigen-4xviACKNOWLEDGEMENTSI would like to thank my supervisor Dr. K. Dorovini-Zis for her guidance, support and interestin my thesis, Mrs. R. Prameya for plating the EC, Donald Wong for his computer advice, Ms.Vivian Wu and Yolanda Bouwman for helping with the lymphocyte characterization. Thankyou all very much.xviiDEDICATIONI would like to dedicate this thesis to my supervisor, my mentor, DR. K. DOROVINI-ZIS,for her guidance, supervision, encouragement and patience during my Ph.D. training inNeuropathology Research, and to my very dear Mom (Huynh Thi Ngoc Suong), Dad (HuynhVan Tu), Brother (Huynh Kim Huu and family) and Sisters (Huynh Thi Kim Lien, HuynhThi Tuyet Mai, Huynh Thi My Dung, Huynh Thi Hoang Yen, Huynh Thi Hoang Anh and theirfamily members) for their sacrifice, their support and their belief in me, and most importantly,to my loving Mom and Dad who always believe that education is the best gift that any parentscan give to their children, and last but not least to God for all of the blessings that I have beengiven. Thank you all for helping me through this.1INTRODUCTION1.1 INFLAMMATION1.1.1 DefinitionInflammation is a localized protective response which occurs as a defensivemechanism against the invasion of the host by foreign material, frequently microbial in nature.Responses to toxins, neoplasms, and mechanical trauma may also result in inflammatoryreactions. Inflammation serves to destroy, dilute, or wall-off both the injurious agent and theinjured tissue. The symptoms of inflammation include redness (rubor), swelling (tumor), pain(dolor), heat (calor), and loss of function (functio laesa).According to Julius Friedrich Cohnheim, a pathologist in the late 1800’s, the first foursymptoms (redness, swelling, pain and heat) represent the cardinal signs of acuteinflammation, while the functional loss (functio laesa) is in reality a resulting condition. Hedescribed “redness” as the overloading of all blood vessels, “swelling” due to the increasedvascular flilness, especially the great increase of transudation, “pain” due to the pressure on,and dragging of, the nerves of sensation by the overfilled vessels and abundant transudation,and finally “heat” resulting from a more than normal amount of heat supplied to the site fromwithin by the increased blood supply (1).Inflammatory processes play a central role in mediating immune host defense andwound healing, but unfortunately, they also participate in the pathogenesis of many diseases,e.g. allograft rejection. Information concerning the mechanisms whereby inflammatory cells2accumulate in tissues, as well as the mechanisms whereby such cells are stimulated to damagetissues, should provide better insight into the pathogenesis of human diseases and should alsoprovide clues for developing more rational forms of therapy (2, 3).1.1.2 General aspectsAs first witnessed by Cohnheim (1), and later by Clark in 1935 (4), using intravitalmicroscopy, the initial event in leukocyte localization to sites of inflammation is called“margination”, whereby leukocytes leave the central stream of blood flow in post-capillaryvenules. These leukocytes then interact with the endothelium lining the vessel wall by“rolling” along the luminal surface, a process which occurs within minutes of theinflammatory stimulus. As inflammation progresses, the number of rolling leukocytesincreases along with a decrease in their velocity, and the process finally comes to a halt.Consequently, “diapedesis” will take place, a process whereby the leukocytes migrate throughthe endothelial cell (EC) junctions along the vessel wall and into the tissues. The accumulationand subsequent activation of leukocytes are central events in the pathogenesis of virtually allforms of inflammation. The recruitment of humoral and cellular components of the immunesystem leads to the amplification and propagation of most forms of inflammation.Immunologically-mediated elimination of foreign material proceeds through a seriesof steps. Firstly, the material to be eliminated (i.e. antigen) is recognized as being “foreign”by either specific or non-specific means: a) Specific recognition is mediated byimmunoglobulins (i.e. antibodies) or by T cell receptors which bind to specific determinants3(i.e. epitopes) on the antigen. b) Non-specific forms of recognition, such as recognition ofdenatured proteins or endotoxins, can be mediated directly by the alternative complementpathway or by phagocytes. Secondly, the binding of a recognition component of the immunesystem to an antigen generally leads to activation of an amplification system, initiatingproduction of proinflammatory substances. These mediators then, in turn, will alter the bloodflow, increase vascular permeability, augment adherence of circulating leukocytes to vascularendothelium, promote migration of leukocytes into the tissues, and stimulate leukocytes todestroy the inciting agent. The production of inflammatory mediators which leads toalterations of the normal function of the blood vessels had been suggested by Cohnheim morethan one hundred years ago based on his research in inflammation. He speculated that ininflammation, a chemical change must occur in the vessels to induce the characteristiccirculatory disturbances that he observed (1).The actual destruction of antigens by immune mechanisms is mediated by phagocyticcells. These cells may migrate freely or may exist at the fixed tissue sites as components of themononuclear phagocytic system. Macrophages and related cells (e.g. Kupffer cells, type-Asynovial lining cells) are the central components of this system. Destruction of antigensoutside of the mononuclear phagocytic system generally takes place in tissue spaces and ismediated by polymorphonuclear leukocytes (neutrophils) or monocytes, which are recruitedfrom blood.41.1.3 Role of endothelial cellsEndothelial cells (EC) are strategically located between the intravascular elements andthe parenchyma of every organ. Therefore, it appears reasonable to assume that, in addition toforming this crucial boundary, EC may participate in a number of important physiologic roles.However, in the past, the role of EC in inflammatory processes had primarily been consideredto be passive. It was not until a series of breakthroughs occurred which allowed for theisolation and subsequent culture of EC in vitro that substantive questions could be addressedin a stepwise manner. Most of the information about the structure and function of human ECcomes from studies of human umbilical vein EC (HUVEC) because these cells are relativelyeasy to obtain, isolate and culture. These studies have provided convincing evidence that ECnot only provide a nonthrombogenic surface to the intravascular compartment but alsoperform a host of other functions. These include wound healing, angiogenesis, production ofclotting factors, tumor metastasis, cytokine production, leukocyte trafficking, vascular tone,and many others. Consequently, far from being either passive elements in physiologicprocesses or non-participants in pathologic processes, it is now realized that EC are intimatelyinvolved in inflammation and help to create, modulate, and terminate inflammation. In vitrostudies of cytokine effects on EC have provided much of the information regarding the role ofEC in inflammation (5).It is interesting to note that the observation made by Cohnheim (1) more than onehundred years ago led to his remark, ttwe have here to deal with a molecular change of thevessel walls”, and it took almost a century to establish the molecular basis for leukocyte5interaction with endothelium. An early event in inflammation involves the adhesion ofpolymorphonuclear leukocytes (i.e. neutrophils) to the blood vessel wall (i.e. EC) which is acrucial step leading to leukocyte accumulation within the inflammatory site. EC exposed toagents such as histamine or thrombin can express signals (e.g. granule membrane protein-140and platelet activating factor) which attract neutrophils and other blood leukocytes to leave themain vascular stream and marginate along the vessel wall (6). Within minutes of thistriggering, a phenomenon called 11rolling” which involves leucocytic attachment to, anddetachment from endothelium is initiated. During the initial stages of adhesion, the neutrophilscome in close contact with the endothelium and extend pseudopodia that attach to theendothelial surface or are directed toward interendothelial junctions. The dynamic interactionbetween different adhesion molecules expressed by endothelium and blood leukocytes playsan important role in leucocytic entry into tissues. The migration process starts by insertion ofa portion of the cytoplasm between two adjacent EC, followed by the movement of the entirecell across the monolayer (7). It is still unclear how exactly the neutrophils force theinterendothelial junctions apart and what molecular mechanisms are responsible for thedisassembly and resealing of the tight junctions. It has been shown that various cytokinesparticipate in this process, and interleukin-1 (IL-i), tumor necrosis factor-a (TNF-ct), andinterferon-y (IFN-y) are among the most potent inflammatory agents (8).Interleukin-1 and tumor necrosis factor appear to play pivotal roles in leukocyteendothelial adhesion. They act on neutrophils, rendering them more ??sticky?? than normal.Through separate sets of mechanisms, interleukin-1 and tumor necrosis factor also act on EC,6rendering these cells more adhesive for neutrophils, monocytes and lymphocytes (9 - 11). Theeffects of IL-i and TNF-ct on EC can be blocked by RNA and protein synthesis inhibitors (10,11).Interleukin-1 is synthesized by macrophages, microglia and astrocytes. Its effects includechemotaxis, induction of increased adherence, enhanced vascular permeability, stimulation ofthe release of platelet-activating factor and prostacyclin by EC (12, 13). Tumor necrosisfactor-a is synthesized predominantly by macrophages, but also by T cells, astrocytes andmicroglia cells. This cytokine has been linked to the inflammatory demyelinating process ofexperimental autoimmune neuritis (EAN), experimental autoimmune encephalomyelitis(EAE), and multiple sclerosis (MS) (8). MS is a chronic inflammatory disease involvingdemyelinization of the central nervous system, while EAE is an animal model for MS.Interferon-y, predominantly produced by CD4 T lymphocytes of the T helper inflammatoryphenotype, also exerts a multitude of inflammatory effects. This cytokine is the most potentinducer or upmodulator of the major histocompatibility class II molecules (MHC Class II), andit also enhances vascular permeability. The most convincing evidence that IFN-y plays animportant role in inflammatory demyelination arises from a clinical trial in which the systemicadministration of IFN-y to MS patients resulted in clinical exacerbations. These exacerbationswere accompanied by increased numbers of monocytes expressing Human Leucocyte Antigen(HLA), especially HLA-DR subtype, enhanced proliferative responses of peripheral blood Tcells and natural killer cell activity (15).Although many valuable observations have been made during in vitro studies of7human large-vessel EC, most physiologic and pathophysiologic events take place at the levelof the microvasculature. Furthermore, emerging evidence suggests that selected differencesexist between BC of the microvasculature and those that line the large blood vessels as well asbetween BC from different vascular beds. These include differences in morphology, insecreted products, in expression of cell adhesion molecules, in cytokine-induced regulation ofcommonly expressed cell adhesion molecules, and in response to BC injury (5, 16- 19).Subsequently, it is uncertain whether the observations on large vessel EC of other organs alsoapply to the human brain microvascular BC (HBMEC), so that the results should not beextrapolated from one system to the other.1.2 INTERFERONSThe interferons (IFNs) represent a group of glycoproteins discovered in 1957 asbiological agents interfering with the replication of viruses, hence the name “Interferon” (20).The IFNs can be produced by all nucleated cells and can be classified as cytokines. They aremultifunctional and are components of the host defenses against viral and parasitic infections(e.g. chronic infection with hepatitis B and C viruses), and also certain tumors (e.g. hairy cellleukemia, Kaposi’s sarcoma, non-Hodgkin’s lymphoma). They influence the functioning of theimmune system in various ways and also affect cell proliferation and differentiation. The IFNsexert their multiple activities primarily by inducing the synthesis of many proteins (21). TheIFNs were originally classified by their sources as leucocyte, fibroblast and immune IFNs.Leucocyte and fibroblast IFNs, together, were also categorized as type 1 IFNs and immune8IFN as type 2 IFN. For the time being, the nomenclature is based on sequencing data.Leucocyte IFNs as IFN-a and o, fibroblast IFN as IFN-f3, and immune IFN as IFN-y.IFN-y will be examined in this study because of its role in inducing Ia Ag expression(22), activating BC to bind T lymphocytes (23) and in enhancing the migration of lymphocytesacross EC monolayers (24). The effects of IFN-13 will also be determined because of its role indownregulating the effects seen with IFN-y (25 - 27) and more importantly, because of itspotential therapeutic application in treating autoimmune disorders of the CNS such as MS (28,282).1.2.1 IFN-yThe gene coding for IFN-y has been located on chromosome 12, and the matureIFN-y is made of 166 amino acids (29). IFN-y was initially considered to be a secreted productexclusively of T lymphocytes, especially of the T helper subset. However, this interpretationwas later recognized to be too restrictive. Many cytolytic T cells also release IFN-y uponcontact with the specific target, such as virally infected syngeneic cells (30, 31). Natural killercells have also been shown to be another source of IFN-y when they are exposed to targetcells, interleukin-2 (IL-2) (32), or hydrogen peroxide (33).It is now recognized that the most important immunomodulatory IFN is IFN-y. The proteinsinduced by IFN-y comprise those encoded by the major histocompatibility complex class I andespecially class II regions. These proteins are involved mainly in the processing, transport, andcell surface presentation of antigens, and also in cell to cell recognition (21).91.2.2 IFN-The gene coding for this cytokine has been located on the short arm of chromosome 9,and mature IFN-13 is comprised of approximately 165 - 172 amino acids long (29). IFN- isexpressed exclusively by fibroblasts and it is induced essentially by viruses or double-strandedRNAs (34). As mentioned previously, the IFNs (i.e. IFN-a, j3, and y) were originally identifiedas antiviral proteins. Accumulating evidence, however, indicates that IFN-13 also plays animportant role in the control of cell growth and differentiation (35, 36). Together with IFN-y,IFN-f3 produces synergistic and antiproliferative activities (37). In contrast, theimmunomodulatory effects of IFN-13 include its capability to inhibit or down-regulate IFN-yproduction in MS (38, 39), to antagonize the Ia-inducing effect of IFN-y in vitro (40, 41) andto augment suppressor cell function in MS patients (42).1.3 MAJOR HISTOCOMPATIBILITY MOLECULES1.3.1 DefinitionThe major histocompatibility complex (MHC) molecules are proteins discovered bythe British geneticist Peter Gorer and by George D. Snell of Jackson Laboratory in BarHarbor, ME, as the cause of graft rejection. Their long-winded names are derived from theGreek word for tissues (histo) and the ability to get along (compatibility). These MHCmolecules can be categorized into two classes: class I and class II MHC molecules (43).101.3.2 Class IMHCThese molecules can be found in almost all types of body cells. They are transportingproteins which are synthesized in the endoplasmic reticulum, and each MHC molecule has adeep groove into which a short peptide, or protein fragment, can bind. Class I MHC moleculesbind to peptides that orginate from proteins in the cytosolic compartment of the cell, however,they can only hold short peptides because their binding site is closed off. This class 1-peptidecomplex is then transported to the cell surface, and if the peptide is foreign to the cell, it willbe recognized by the passing T cells. These immune cells will release lymphotoxins or otherstructurally related molecules that destroy the cell presenting the peptide. These T cells arereferred to as killer T cells (43).1.3.3 Class IIMHCClass II MHC or human leucocyte antigens (e.g. HLA-DR, HLA-DQ, HLA-DP) aredistributed on the cell surface as c43 heterodimers which can be found primarily on specializedantigen presenting cells such as macrophages, dendritic cells and B lymphocytes (44). Theyplay an important role in the initiation of immune responses because the activation of T helperlymphocytes by an antigen is restricted to the presentation of the processed antigen togetherwith the homologous class II MHC molecule on macrophages and other accessory cells (45).It is now realized that class II MHC molecules, HLA-DR antigen in particular (commonlyreferred to as Ia Ag), are actually peptide transport proteins which bind, transport to, anddisplay at the cell membrane peptides that are derived from extracellular proteins. In contrast11to class I MHC molecules, class II MHC can bind to peptides of different lengths, because thebinding site is open at both ends. They present these peptides to helper T cells as part of themechanism for identifying foreign antigens and producing an immune response (46 - 48).a) Role in autoimmune disorderAutoimmune disorders (AD) affect 5% to 7% of the population. The AD are producedby autoimmunization secondary to a disruption of the normal self-tolerance which Erlichreferred to as “Horror autotoxicus” (49). The autoimmune process may be initiated by anenoneous antigen presentation by the MHC molecules. It is important to realize that AD is notthe result of a unique mutation of the HLA alleles exclusively found among the patients.Instead, the same sequence found among these patients can also be found in the healthycontrols, although with a different frequency.With the recent development of biochemical and molecular biological methods, it appears thatHLA class II molecules are important for at least some diseases because: i) the HLA class IIspecific associations are maintained in different ethnic groups (50), and ii) in some animalmodels, the induction of AD may be blocked by monoclonal antibodies directed against thealleged antigen, and the comparison of specific sequences between diseased and controlgroups has shown extremely strong associations with specific class II alleles (51).The immune recognition of normal self components which leads to tissue destructionand pathological abnormalities underlies a group of disparate diseases such as insulindependent-diabetes mellitus, rheumatoid arthritis, myasthenia gravis, and multiple sclerosis,12collectively known as autoimmune disorders. It has been shown that the susceptibility to thesediseases is strongly associated with particular class II alleles, suggesting that the class IIMHC-restricted binding and presentation of specific autoantigens may be involved in thedisease process (51 - 59). Class II MHC are also associated with a heterogeneous group ofleukoencephalopathy including cerebrovascular disease, Alzheimer disease and mixeddementia of Alzheimer type (60), IgA nephropathy (61), anterior uveitis (62), and lupuserythematosus disseminatus (63).b) Expression by ECDue to the strategic location of the endothelium of blood vessels, i.e. as the first celllayer that interacts with the blood constituents, the role of EC as active participants in immuneregulation has been extensively investigated in recent years. Thus, it has been shown that ECof extracerebral blood vessels can be induced to express class II molecules and that theimmunogenic capacity of the EC is directly proportional to the extent of class II MHC antigenexpression on the cell surface (64). There are conflicting reports as to whether the capillaryendothelium in brain expresses the HLA-DR antigen in normal or diseased states, e.g. theantigen could not be detected in rat brain microvessels in both control or EAE rats (65). Incontrast, the antigen is found in approximately 10% of guinea pig brain microvessels in thecontrol state and in 35% of guinea pig brain microvessels in EAE (66). In humans, themicrovascular DR-antigen has been reported to be rarely detectable in normal brain (67), butdiscontinuous vascular immunoreactivity for the antigen can be found in multiple sclerosis13(68). However, Pardridge et a!. have demonstrated that the DR-antigen can be detected in thebrain microvasculature of both normal subjects and subjects with neurologic disease (69). Invitro studies have shown that untreated primary cultures of rat brain endothelium areabsolutely negative for class II staining, but these EC can be induced to express Ia Ag in vitrowhen they are treated with JFN-y (70).The ability of IFN-y to induce Ta Ag expression has been well documented in manydifferent cell types including human and murine epithelial cells (71, 72), rat glomerularmesangial cells (73), murine macrophages, mast cells and myelomonocytic lines (74- 77),human melanoma cells (78), monocytes (78, 79), dermal fibroblasts (80), human myoblasts(81) and myotubes (82) in vitro. Moreover, IFN-y can induce class II MHC expression onextracerebral EC such as HUVEC (22, 83), and human dermal microvascular EC (HDMEC)(84) in cultures. These cells do not express class II MHC constitutively under standard cultureconditions. The effects of IFN-y on the aberrant expression of Ta Ag on cell types that do notnormally function as antigen presenting cells may lead to the presentation of cellular proteinsto the immune system which can then contribute to the induction of autoimmune diseases (85,86) or graft rejection (87). It has been reported that IFN-y upregulates class II MHC expressionon heart EC which plays a role in the organ allograft rejection (64). Subsequently, manystudies have focussed on determining the mechanisms by which Ta Ag expression can bedownregulated, especially in situations involving IFN-y upregulation.The IFNs have been suggested to have therapeutic potential in MS because of thehypothesis that a viral infection, an immunoregulatory defect, or both, may be implicated in14the disease process (88 - 90). Reports of deficient IFN-y production in MS patients (91, 92)further suggested the possibility of using IFN-y for therapeutic trial. In contrast, later studiesusing a sensitive solid phase radioimmunoassay to test for IFN-y levels reported that thelymphocytes isolated from untreated MS patients actually produced more IFN-y than thenormal controls (93, 94). Moreover, a small increase in cerebrospinal fluid IFN-y was alsoobserved in active MS patients (95).In the early 1980’s, when highly purified IFN preparations became available, it waspossible to demonstrate for the first time the specific binding of IFN to cellular binding sites.The binding studies established that IFN-y binds to a receptor different from that for IFN-3 (96- 100). However, several different groups have also reported that there is some cross-reactivitywith IFN-3 for the binding site for IFN-y (101 - 104). IFN-3 has been shown to be one of thesubstances that can antagonize the induction of Ia Ag expression by IFN-y. The antagonisticeffects of IFN-13 have been demonstrated in cultured adult human astrocytes (25), humanglioma cells (26), murine macrophages (27, 36), blood monocytes isolated from MS patients(28), astrocytoma cell lines (37) and also in other EC systems (105, 106).The ability of IFN-y and IFN- to induce and suppress Ta Ag expression on HBMEC,respectively, was examined in this study in an effort to provide better understanding of thepotential role of HBMEC as antigen presenting cells and to provide an in vitro system forscreening drug therapies which may show promising value in MS.151.4 CENTRAL NERVOUS SYSTEM INFLAMMATION AND AUTOIMMUNITY1.4.1 Permeability of the Blood-Brain-BarrierThe presence of tight junctions between adjacent EC, the absence of vesiculartransport system and the paucity of cytoplasmic vesicles are unique characteristics of brainmicrovascular EC, forming the basis of the blood-brain barrier (BBB) (107, 108). Undernormal physiological conditions, cerebral EC restrict the paracellular movement of proteins,ions, large lipid-insoluble nonelectrolytes, and the access of antibodies, complement moleculesand white blood cells from the blood to the brain parenchyma. Subsequently, the BBB controlsthe lymphocyte traffic into the CNS which is normally very limited.A variety of clinical and experimental conditions may modify or damage the BBB,leading to enhanced permeability to plasma proteins, ions, water and circulating white bloodcells (108, 109). The cerebrospinal fluid (CSF): serum albumin ratio indicates the level ofvascular permeability in the CNS. As long as the BBB is intact, the ratio between the tworemains relatively constant (110). Several studies have focussed on the mechanisms ofopening of the BBB under different experimental and clinical conditions. Transmissionelectron microscopic examination using horseradish peroxidase and lanthanum as electrondense tracers have demonstrated that these molecules may traverse the endothelial barrier viaopened tight junctions (111 - 114) or by means of accelerated vesicular transport (115 - 118).Moreover, the tight junctions between the EC also become permeable to the electron densetracers if a hyperosmotic solution is administered into the rat via intracarotid injection (113).It has been reported that in EAE, the experimental autoimmune demyelination of the CNS16mediated by T cells, increase in the BBB permeability to small molecules occurs early in thedisease process (119 - 124). Alterations in BBB permeability in MS have also been detectedusing gadolinium contrast enhancement (125). The mechanisms responsible for the increasedBBB permeability in disease are still unknown. MS is a demyelinating disorder of the CNSwhere some inflammation exists. It is thought to be immune mediated. The presence oflymphocytes on brain slides suggests that they have egressed from the vessels. The reason whythey accumulate in the cerebral parenchyma is probably due to an increased exit from thecirculation (i.e. by an increase in the permeability of the BBB). Interestingly, IFN-y has beenshown to induce alterations in the morphology and permeability properties of cultured EC(84, 126 - 128). Studies on the effects of different combinations of cytokines, e.g. IFN-y,tumor necrosis factor-a (TNF-a), and interleukin-1 a/13 (IL-i a/f3), on HUVEC monolayerpermeability have indicated that IFN-y plays a central role in increasing the permeability ofEC monolayer (128). These results support previous in vitro observations on IFN-y effectson HBMEC permeability (127). Moreover, an increase in vascular permeability was detectedwhen Wistar rats were injected intradermally with IFN-y (129). The effects of IFN-y on ECpermeability are further confirmed by studies demonstrating augmentation of lymphocytemigration across HUVEC monolayers by IFN-y, and the effect has been shown to be due to aselective action on EC (24). Furthermore, direct interaction between memory T cells andvascular EC in a noncytolytic manner augments the EC permeability to macromolecules. It issuggested that modulation of endothelial permeability may be a critical factor contributing tothe preferential migration and accumulation of lymphocytes at chronic inflammatory sites17(130).1.4.2 Lymphocyte infiltrationAutoimmune demyelinating disorders of the CNS such as MS are characterized bydemyelination and migration of acute and chronic inflammatory cells from the blood into thebrain parenchyma through the cerebral vasculature that normally excludes circulatingleukocytes from entering the brain. MS is a recurrent and progressive inflammation of theCNS, and it has been reported that the number of lymphocytes circulating through the cerebralmicrovasculature is always many folds higher than the disease-free control (110). Themechanism(s) leading to the migration of these cells from blood to brain through the highlyspecialized BBB are largely unknown.a) Ia Ag associationEpidemiological studies suggest that common viral-like infections, primarilyrespiratory infections, are temporally associated with exacerbations in MS (131). More recentstudies also show a close relationship between common febrile events and new MS diseaseactivity (132). Poser has suggested that a systemic viral infection may play a role in thedisease process of MS whereby activated T cells can secrete IFN-y in response to a viralinfection (133). As mentioned previously, IFN-y is the most potent inducer of class II MHCmolecules (Ia Ag), and its induction has been reported in extracerebral endothelium (22, 84,134) as well as cerebral endothelium (70, 127, 135, 136). The expression of Ia Ag maysubsequently allow EC to present antigen to T cells (137) with the consequent release of18cytokines leading to the amplification of the immune response. Immunohistochemical studiesin EAE and brain sections of MS lesions have demonstrated Ia Ag expression on EC,astrocytes and macrophages, while the antigen is not detected in the normal brain tissue (66,68, 138 - 141). Class II MHC (Ia Ag) expression appears 12 to 48 hours after exposure of ECto IFN-y in vitro, reaches a plateau by 4 to 6 days and is associated with the expression of allknown class II antigens (80, 142). The mechanism of action of IFN-y is by increasing mRNAlevels for class II molecules (80, 143). Moreover, untreated human EC have no detectablemRNA for class II antigen (80, 144).b) Lymphocyte - EC adhesionAdhesion of T lymphocytes to vascular endothelium is a necessary prerequisite tomigration of lymphocytes from the blood into chronic inflammatory sites. Studies oflymphocyte-EC adhesive mechanisms have shown that IFN-y plays an important role inaugmenting lymphocyte-EC adhesion (23, 24, 134, 135, 145). IFN-y-mediated increase inadhesion occurred when EC were preincubated with this cytokine. In contrast, preincubationof lymphocytes with IFN-y did not affect the adhesion, indicating that the action of cytokinewas mainly on the EC (23). Moreover, lymphocyte adhesion to EC and expression of HLADR antigens on EC are well correlated in terms of both kinetics and the dose-response patternof IFN-y (134). The extent of lymphocyte adhesion to EC appears to depend on the densityof HLA-DR antigens expressed on the EC. Blocking studies with monoclonal antibodiesdirected against HLA-DR or CD4 molecules significantly inhibit lymphocyte-EC adhesion inboth autologous and non-autologous syngeneic combinations, suggesting that in IFN-y19enhanced binding, Ia Ag plays a central role in the increased adhesion of T cells to HUVEC(134, 145). CD4 molecules are expressed in a subpopulation of lymphocytes, mainly T-helpercells, and are known to have an affinity for class II molecules (146, 147), providing strongsuggestion for the involvement of IFN-y-induced class II MHC as the correspondingendothelial adhesion molecule. To further support the above statements, in vivo studies onEAE mice injected with mAb against Ia molecules show a decrease in lymphocyteaccumulation in the CNS (148). Subsequently, it is speculated that the anti-Ia antibody musthave blocked the adhesion of Ia-reactive T cells to Ia-expressing EC, leading to inhibition of Tcell migration from blood into the CNS. Masuyama et al. (134) have suggested that T cellrecognition of HLA-DR antigen may represent the signal for the initiation of a subsequentadhesive processes whereby complementary adhesion surface molecules become engaged.These observations indicate that the release of IFN-y by activated lymphocytes in chronicinflammatory sites can upregulate the adhesion of circulating lymphocytes to the local EC.Since lymphocyte-EC adhesion represents the initial step of lymphocyte migration through themicrovessels (149), the IFN-y-mediated increase in lymphocyte adhesion on the surface ofmicrovessels may lead to further influx of lymphocytes into the site of chronic inflammation.In vitro studies have demonstrated that lymphocyte-EC adhesion can be significantlyupregulated by stimulating the T cells with phorbol esters (150- 153), con A (154 - 156), orIL-2 (157, 158). Significant increase in T-cell adhesion to purified ICAM-1 substrates hasalso been reported when T cells were pretreated with anti-CD3 IgG and anti-IgG (159).Preincubation studies with phorbol esters have shown that enhancement of the lymphocyte-EC20binding is entirely attributable to an effect on T cells, with no action on BC (150). Moreover,additive enhancement of adhesion can be detected when both lymphocyte and BC are activated(150, 154). More recent studies have demonstrated that IFN-y treatment of passaged culturesof human cerebral BC further enhances the lymphocyte-BC adhesion. The activatedlymphocytes were isolated from peripheral blood of acute relapsing MS patients duringexacerbations (160).c) Transendothelial migration of lymphocytesIn vivo observations on T-lymphocyte entry into the CNS have determined thatlymphocyte activation enhances their ability to cross the BBB and rapidly appear in the CNStissue, irrespective of their antigen specificity, MHC compatibility, T-cell phenotype, and Tcell receptor gene usage (161). Moreover, the level of lymphocyte infiltration into the tissue issignificantly upregulated in chronic inflammatory disorders, a process that may be modulatedby the enhancement of interactions between different adhesion molecules expressed on bothlymphocytes and vascular EC (150, 162). Studies of lymphocyte traffic through variousorgans (e.g. lungs, liver, lymph nodes) have reported differences in the migratory patterns ofresting and activated lymphocytes. Generally, resting lymphocytes move rapidly whileperipherally activated lymphocytes move much more slowly through the organs.T-lymphoblasts can readily leave the circulation and migrate into an inflammatory site (163).Migration of activated T cells from the peripheral blood to the CNS has been demonstrated inEAE (164) along with a significant decrease in their activated T-cell levels in the peripheralblood during exacerbations (165). A decreased proportion of activated T cells in the21peripheral blood of MS patients has also been reported (166, 167). Exacerbations in MS wereusually accompanied by further decreases in activated T cells in the peripheral blood (167).Moreover, lower percentages of activated T cells were consistently found in the CSF of MSpatients (168), suggesting that these cells may accumulate in the MS plaques. In vivotreatment of EAE rats with a monoclonal antibody specific for activated rat T cells hasdemonstrated significant reduction in inflammation (169).Subsequently, these observations suggest that activation of both lymphocytes and ECplays a central role in upregulating the adhesive and migratory interactions betweenlymphocytes and EC. It is notable that studies of lymphocyte migration across cultured ratretinal microvessel EC have shown that con A activation of T-cells did not augment themigratory process. Furthermore, IFN-y treatment of EC resulted in a slight, but notsignificant, increase in migration (170). The insignificant increase in migration oflymphocytes across rat retinal EC when both systems are activated supports the concept ofimmune privileged site of the eye due to the presence of the blood-retinal barrier. Theseresults further confirm the heterogeneity that exists among EC isolated from vascular beds ofdifferent organs and species with regard to immunological responses to cytokine treatment(19, 155).1.4.3 FVIIIR:Ag in primary cultures of HBMECFactor Vill-related antigen (FVIIIR:Ag) is a large multimeric glycoprotein synthesizedand released by EC of large and small blood vessels, and it is widely considered as the most22specific marker for cells of endothelial origin (171, 172). The protein has also been shown tobe synthesized in primary cultures of HBMEC as demonstrated by immunofluorescence andimmunoperoxidase techniques at the light microscopic level (107). When vascular injuryoccurs, FVIIIR:Ag facilitates the adhesion of platelets to the subendothelial matrix (173).Biochemical and immunohistochemical studies indicate that FVIIIR:Ag is generally stored inrod-shaped, membrane-bound cytoplasmic organelles called Weibel-Palade bodies (namedafter the investigators who first described them) which are found exclusively in EC (174 -177). However, these organelles are rarely found in EC lining the cerebral capillaries, andthey are absent in primary cultures of HBMEC when examined ultrastructurally (107).Weibel-Palade bodies are also absent in primary cultures of microvessel EC derived frommouse (178) and rat (179 - 181) brain. It has been recently reported that IFN-y treatment ofHUVEC causes a decrease in the release of FVIIIR:Ag from these EC (182). The mechanismof IFN-y-suppressed release of FVIIIR:Ag is still unknown at the present time. Tannenbaumand Gralnick speculated that IFN-y-mediated depression of FVIIIR:Ag release from the ECmay assist in maintaining blood fluidity during immune activation (182). Subsequently, thisstudy examined the distribution and fine structural localization of FVIIIR:Ag in primarycultures of HBMEC and also investigated the influence of IFN-y on the storage and release ofthis glycoprotein by HBMEC, considering the important role of IFN-y in autoimmunedisorders of the CNS.Consequently, the effects of IFN-y and IFN-f3 as potential factors regulatinglymphocyte entry into the CNS will be examined in this study. Their influences on Ia Ag23expression, cell proliferation, alteration of the morphology and permeability characteristics ofHBMEC in cultures will be investigated. Furthermore, using primary cultures of HBMEC as amodel of the BBB, these cytokines will also be tested for their effects on the adhesion andmigration of peripheral blood lymphocytes across HBMEC monolayers which may be relevantin understanding the in vivo immune reactions including increased vascular permeability andcell shape changes. Blocking experiments with mAbs directed against human HLA-DRmolecules will further determine the role of Ia Ag in the adhesive and migratory processes.The results of this study will emphasize the potentially critical role of HBMEC in the initiationof the immune reaction in autoimmune disorders of the CNS. Elucidation of the factorsmodulating the adhesion and migration of lymphocytes across microvascular EC can havegreat therapeutic potential. The information may assist in designing protocols that can preventfurther amplification of the unwanted inflammation or immune responses such as that in MS.In conclusion, understanding of the pathobiology of HBMEC, such as their responsesto IFN-y and IFN-13 treatments and their ability to mediate lymphocyte traffic into the CNS, iscentral to the understanding of CNS inflammation.241.5 SUMMARY AND OBJECTIVESIn summary, a large body of evidence indicates that Ia Ag expression can be inducedby IFN-y in many different cell types including EC. It has also been reported that IFN-3 iscapable of suppressing the IFN-y-induced Ia Ag expression. Furthermore, it has beendemonstrated that lymphocyte-EC adhesion and migration can be augmented by IFN-y.Monoclonal antibody blocking studies indicate that Ia Ag might play a critical role in theadhesive and migratory processes. Finally, IFN-y has been shown to alter the morphologyand permeability characteristics of the EC monolayers which may facilitate the influx oflymphocytes into chronic inflammatory sites. It is uncertain at the present time whether theobservations on EC of other organs and species also apply to the HBMEC. It is now wellaccepted that great heterogeneity exists among different types of EC with regard to antigenicdeterminants, cell surface molecules, metabolic properties, permeability functions andimmunological responses to cytokines. Cerebral EC are characterized by unique featuresresponsible for the formation of the BBB. How this barrier is breached in inflammation andautoimmune disorders of the CNS is still poorly understood.The main objective of this thesis is to determine the effects of IFN-y and IFN-f3 on IaAg expression, proliferation, and alteration of the morphology and permeability characteristicsof HBMEC monolayers. Furthermore, the effects of these cytokines on the adhesion andmigration of resting/anti-CD3 stimulated T lymphocytes across the EC monolayers will beinvestigated. Finally, the effects of IFN-y on the storage and release of FVIIIR:Ag byHBMEC will be examined.25My working hypothesis is that, in a chronic inflammatory site, some activated Tlymphocytes, mainly T helper/inducer phenotypes, release IFN-y locally. This cytokine canthen activate the local capillary EC by inducing Ia Ag expression and morphologicalalterations on the EC, and by increasing the vascular permeability. These changes willaugment the adhesion and migration of the lymphocytes across the EC, leading to an influx oflymphocytes into the chronic inflammatory sites. The administration of IFN- will play anantagonistic role on the IFN-y-mediated immunological responses. IFN-13 will suppress theIFN-y-induced Ia Ag expression and morphological alteration on EC and the IFN-y-mediatedincrease in adhesion and migration, resulting in the reduction of lymphocyte numbers in thechronic inflammatory sites. Finally, the ability of IFN-y to affect the storage and release ofFVIIIR:Ag in HBMEC may have some important implications in maintaining blood fluidityduring immune activation, considering the important role of IFN-y in autoimmune disordersof the CNS.The method of approach used in this study involved the successful application ofimmunogold labeling with silver enhancement technique for the detection of Ia Ag expressionat the light microscopic level. ELISA was further applied to confirm the IFN-3 effects insuppressing the IFN-y-induced Ta Ag expression. Kinetic studies of the cytokine effects on IaAg expression on HBMEC were also carried out in this study. Ultrastructural localization ofTa molecules on the surface of EC was determined using mAbs conjugated with colloidal goldmarkers. Electron dense tracer studies were applied in order to assess the permeability ofconfluent untreated or IFN-y treated HBMEC monolayers. Immunoperoxidase techniques26were used to study the IFN-y and IFN- effects on lymphocyte-EC adhesion and migration.Scanning and transmission EM studies further showed the effects of the cytokines on themorphological phenotype of HBMEC, and on lymphocyte-EC adhesion and migration.Finally, immunoelectron microscopy was performed for the ultrastructural localization ofFVIIIR:Ag in HBMEC, and the effect of IFN-y on its storage and release.The Specific Aims of this thesis are:1) To determine, in vitro, whether primary cultures of HBMEC constitutivelyexpress Ia antigen, and whether the expression can be modulated byinflammatory cytokines such as IFN-y and IFN-f3.2) To determine whether IFN-y and IFN- exert antiproliferative effects on primarycultures of HBMEC, whether these cytokines can induce alterations in themorphology and organization of the EC monolayers, and whether IFN-y caninfluence the permeability and endocytotic characteristics of HBMEC that wouldbe relevant to the in situ immune reaction.3) To determine whether IFN-y and IFN-j3 treatment of HBMEC can affect theadhesion of resting/activated lymphocytes to these untreated/cytokine-treated EC.4) To determine whether IFN-y and IFN-El treatment of HBMEC can influence themigration of resting/activated lymphocytes across untreated/cytokine-treated EC.5) To determine the influence of IFN-y on the storage and release of FVIIIR:Ag inprimary cultures of HBMEC which may reflect the maintainance of blood fluidityduring immune reactions in vivo.27MATERIALS and METHODS2.1 Isolation and culture of HBMECHBMEC were isolated from normal brains obtained at autopsy and from temporal lobectomyspecimens removed for intractable seizure disorders and cultured according to methodspreviously described (107). The time interval between death and removal of brain rangedfrom 3.5 hours to 15 hours. Pathologic abnormalities were not detected in any of the brainsused for isolation of cerebral microvessels. Large vessels and the leptomeninges wereremoved from the cerebral cortex with fine forceps, and the tissue was cut into 1 to 2 mmcubes. The tissue was then incubated for 3 hours at 37 °C in medium M199 (Gibco,Burlington, Ontario) containing 0.5% dispase (Boehringer Mannheim, Indianapolis, Indiana).Following centrifugation at 1000 x g for 10 minutes, the pellets were resuspended in mediacontaining 15% Dextran (Sigma, St. Louis, Missouri; average molecular weight, 70,000daltons) and centrifuged at 5,800 x g for 10 minutes in order to seperate the microvessels fromother brain components. The isolated microvessels were then incubated in M199 containing 1mg/mi of collagenase/dispase (Boehringer Mannheim) for 12 to 16 hours at 37 °C. Followingincubation, the microvessels were resuspended in M199 containing 5% horse serum (HS)(Hycione Laboratories, Logan, Utah), layered over Percoll (Sigma) gradients prepared asdescribed by Bowman et al. (183) and centrifuged at 1,000 x g for 10 minutes in order toseparate the EC from red blood cells, pericytes and cellular debris. The layer containing theEC was washed in 10% HS in M199, and collected by centrifugation for 10 minutes at 1,00028x g. Cell numbers were counted with a hemocytometer. Cell viability, as determined by thetrypan blue exclusion test, ranged between 85 and 90%.The isolated clumps of EC were seeded onto plastic wells (Coming Plastics, Coming, NY)previously coated with fibronectin (Sigma). The cultures were maintained in M199supplemented with 10% HS, 25 mM HEPES, 10 mM sodium bicarbonate, EC growthsupplement (20 ig/ml), heparin (100 tg/ml) (all from Sigma), and penicillin (100 tg/ml),streptomycin (100 tg/ml), and amphotericin B (2.5 tg/ml) (Gibco) at 37 °C in a humidified2.5% C02/97.5% air atmosphere. The culture media were changed every second or third day.The endothelial origin of the cells was confirmed by their intense, granular perinuclearstaining for Factor Vill-related antigen (FVIIIR:Ag) and their binding of Ulex Europaeus-Ilectin as previously reported (107). Confluent, contact-inhibiting monolayers were obtained 7to 9 days after plating. The monolayers were used once they reached confluency.2.2 AntibodiesMouse monoclonal antibody (mAb) against human recombinant interferon-y (KM48, IgG1)and mouse anti-human HLA-DR IgG (DK22, IgG, kappa) were obtained from DimensionLaboratories Inc., Missisauga, ONT. Goat anti-mouse IgG (GAMIgG) coupled to Snm goldparticles (Auroprobe LMGAM IgG) was obtained from Janssen/Cedarlane Labs Ltd., (Hornby,ONT), GAMIgG coupled to peroxidase from Jackson Immunoresearch Lab Inc., PA, andmouse anti-human pituitary follicle-stimulating hormone IgG from Biogenex Laboratories,CA. Mouse anti-human Leu-4 (CD3) mAb was purchased from Becton-Dickinson, and mouse29mAb to human Leukocyte common Ag (CD45) from Dimension Lab, Inc.,Mississauga, ONT.For the ultrastructural localization of FVIIIR:Ag, mouse mAb raised against human factor VIIIantigen (Cedarlane Laboratories, Hornby, ONT) was used as the primary antibody and goatanti-mouse IgG coupled to Snm gold particles (Janssen / Cedarlane) as the secondary antibody.For the light microscopy immunoperoxidase staining, a rabbit antiserum to factor VIII antigen(Dakopatts, Santa Barbara, CA) was used as the primary antibody and goat anti-rabbit IgGconjugated with HRP (Jackson Immunoresearch Laboratories, West Grove, PA) as thesecondary antibody.2.3 Induction of Ia Ag expression on HBMEC by human recombinant Interferon-gammaand beta2.3.1 Treatment of primary HBMEC culturesHuman recombinant interferon-y (IFN-y; Collaborative Research Inc., Bedford, MA) wasdiluted in complete media to a final concentration of 10, 20, 50, 100, 150 and 200 U/ml.Confluent monolayers of HBMEC, grown in replicate wells were incubated with differentconcentrations of IFN-y for 4 days and with 200 U/ml for 12 hours to 4 days at 37 °C.Cultures used for these experiments were derived from EC isolated from several differentautopsy brains. The specificity of Ia Ag induction by IFN-y was tested in cultures co-incubatedwith IFN-y (200 U/mI) and anti-IFN-y mAb (optimal concentration, 10 tg/ml) for 4 days. Inorder to study the reversibility of Ia Ag expression, monolayers previously treated with IFN-y30(200 U/mi) for 4 days were thoroughly washed with M199 to remove the cytokine, then placedin complete media and returned to the incubator for another 4 days prior to Ia Ag detection.For the IFN-f3 studies, recombinant human interferon-f3s r(Triton Biosciences Inc., Alameda,CA) was used at 100, 250, 500, 1,000, 2,000 and 6,000 U/mi. These units were determined bythe National Institutes of Health reference standard. This IFN- is the same cytokine that isused for MS therapeutical trial, and it is now available for patient use. Confluent cultures ofHBMEC grown in replicate wells were incubated with 100 U/ml IFN-y with or withoutvarious concentrations of IFN-j3 for 4 days. In separate experiments, replicate wells werepreincubated with IFN- (6,000 U/mi) or IFN-y (100 U/ml) for 2 days prior to coincubationwith IFN- and y for another 4 days.All experiments were performed in duplicate or triplicate wells.2.3.2 Light microscopic immunocytocLiemical localization of Ia AgFollowing cytokine treatment, the monolayers were washed 3 times with buffer containingphosphate buffered saline (PBS- 10 mM, pH 7.2), 1% bovine serum albumin (BSA) and 1%normal goat serum (PBS/BSA/NGS) and incubated for 40 minutes at room temperature withmouse anti-human HLA-DR mAb at 1:30 dilution in carrier buffer containing PBS, 5% BSAand 4% NGS. Following brief washing with PBS/BSA/NGS, the monolayers were incubatedwith the secondary antibody (Auroprobe LMGAM IgG coupled to 5 nm gold particles) diluted1:40 in carrier buffer for 60 minutes at room temperature. At the end of the incubation period,the cells were washed with PBS/BSA/NGS, fixed in fresly prepared buffered formaldehyde-31acetone fixative (20 mg Na2HPO4, 100 mg K112P04,30 ml distilled H20, 25 ml 37%formaldehyde and 45 ml acetone) for 30 seconds, washed with distilled H20, and incubatedin silver enhancing solution ( IntenseM, Janssen/ Cedarlane) for 25 - 35 minutes. Afterwashing with distilled H20, the monolayers were counterstained with Giemsa andcoverslipped using JB-4 plus (Polysciences/Analychem, Markham, ONT) as mountingmedium.Controls included untreated monolayers grown in the absence of IFN-y and IFN-13 and IFN-ytreated cultures incubated with 1) normal mouse IgG at the same concentration as theprimary antibody (5.9 tg/ml IgG), or 2) carrier buffer or 3) an irrelevant antibody (anti-human pituitary follicle-stimulating hormone IgG) instead of the primary antibody.2.3.3 Immunoelectron microscopyHBMBC monolayers treated with 200 U/ml IFN-y for 4 days were washed with buffercontaining PBS, 1% BSA and 0.2% NaN3 (PBS/BSA) and incubated with mouse anti-humanHLA-DR mAb at 1:30 dilution in carrier buffer containing PBS, 5% BSA and 4% NGS for 30minutes at room temperature. After washing with PBS/BSA, the cells were incubated with 5nm gold particle-conjugated secondary antibody (Auroprobe LMGAMIgG) at 1:40 dilution incarrier buffer for 45 minutes, washed, and fixed in periodate-lysine-paraformaldehyde fixative(184) overnight at 4 °C. Following fixation, the cells were washed in PBS, post fixed in 1%buffered 0s04, stained en bloc with uranyl magnesium acetate overnight at 4 °C, dehydratedin graded series of methanol, and embedded in Epon-Araldite. Blocks cut out from the32embedded cultures were re-embedded for cross-sectioning. Thin sections were examined in aPhilips EM400 without heavy metal staining. Controls consisted of cells maintained in IFN-yfree growth media and monolayers incubated with normal mouse IgG or carrier buffer insteadof the primary antibody.2.3.4 Enzyme linked immunosorbent assay (ELISA)Following incubation with IFN-y and/or IFN-13, the cells were washed 3 times with PBS andfixed with 0.025% glutaraldehyde in PBS for 10 minutes at room temperature. Themonolayers were thoroughly rinsed with PBS, washed 3 times with PBS/BSA/NGS andincubated with mouse anti-human HLA-DR mAb at 1:30 dilution in carrier buffer containingPBS, 5% BSA and 4% NGS for 60 minutes at room temperature. After brief washing withPBS/BSA/NGS, the cells were incubated with horseradish peroxidase-conjugated goat anti-mouse IgG diluted 1:5,000 in carrier buffer for 60 minutes at room temperature. Followingwashing in PBS, the cultures were incubated with o-Phenylenediamine (OPD) dihydrochloride(Sigma) (2 mg/ml) diluted in 0.1 M PBS containing 0.015% H20 for 45 minutes at roomtemperature. The reaction was stopped by the addition of 2 M H2S04.Absorbance wasmeasured at 490 nm on a Elisa Microtiter Plate Reader (Molecular Devices, CA). Allexperiments were performed in triplicate.2.3.5 Quantitation of Ia Ag expression by HBMECHBMEC monolayers stained for the light microscopic localization of Ta Ag were examined33under a Nikon Labophot light microscope. Quantitation of Ta Ag expression was performed bycounting one central and four peripheral randomly selected fields of each culture well with anocular grid under X200 magnification. All counts were performed blindly. Data are expressedas the mean ± standard error of the mean.2.4 Scanning electron microscopy (SEM)Confluent HBMEC monolayers treated with 200 U/ml IFN-y, 6,000 U/mi IFN-f3 or acombination of IFN-y (200 U/mi) and 1E (6,000 U/mi) for 3 to 4 days, monolayerscoincubated with 200 U/mi IFN-y and 10tg/ml anti-IFN-y monoclonal antibody for the sameperiod of time, as well as untreated control cultures were processed for SEM as described bySchroeter et al. (185). Briefly, the cultures were washed in Hank’s balanced salt solution andfixed in 2.5% glutaraldehyde in 0.05 M sodium cacodyiate buffer (pH 7.2) for 1 hour at 4°C. Following washing in cacodylate buffer, the cells were post fixed in buffered 1% 0s04for 1 hour, washed in buffer and treated with 1% Tannic acid for 1 hour. After further washingin cacodylate buffer, the monolayers were dehydrated in graded series of methanol up to 70%,and block stained with 4% uranyl acetate overnight at 4 °C. The cells were then dehydratedwith methanol up to 100%, critical point dried, gold coated and viewed with a CambridgeStereoscan 250T scanning electron microscope.2.5 Permeability studiesConfluent HBMEC monolayers treated with IFN-y (200 U/mi) for 4 days were washed with34serum-free aMEM and incubated in aMEM containing 1 mg/mi horseradish peroxidase (HRP;Sigma Type VI) for 5 - 10 minutes at 37 °C as previously described (186). At the end of theincubation period, the cells were fixed in 2.5% glutaraldehyde and 2% paraformaldehyde in0.1 M sodium cacodylate buffer (pH 7.4) for 1 hour at 4 °C. Following washing with buffer,the monolayers were incubated with 3,Y - diaminobenzidine (Sigma) for 1 hour at 4 °C,washed with cacodylate buffer, post-fixed in 1% buffered 0s04, stained en bloc with uranylmagnesium acetate, dehydrated with graded series of methanol, and embedded in EponAraldite. Thin plastic sections were examined in a Philips EM400 without heavy metalstaining. Controls consisted of identical age-matched primary cultures grown to confluence inIFN-y free media.Quantitation of the junctional permeability and pinocytotic activity of EC was performed bycounting the number of permeable and impermeable intercellular contacts and the number ofHRP labeled and unlabeled cytoplasmic vesicles in 100 IFN-y treated and 100 untreated cellsphotographed at standard EM magnification.2.6 Growth StudiesFreshly isolated BC were plated in replicate wells of Corning 24 - well plates at a density of 1X 10 cells/cm2on day 0. On day 1, all experimental wells were refed with complete mediumcontaining IFN-y (150 U/mi), IFN-3 (1,000 U/mi) or a combination of IFN-y and f3 (150 U/miand 1,000 U/mi, respectively). All experiments were carried out in duplicate and media werechanged every other day. Control cultures were maintained in growth media in the absence of35the interferons. The cells were viewed with a Nikon Diaphot TMD inverted microscope andphotographs of one central and four peripheral fields of each well were taken daily at loxmagnification. The number of cells in each photograph were counted and the data areexpressed as the mean ± standard error of the mean.2.7 Preparation of T lymphocytesHuman peripheral blood lymphocytes (PBL) were prepared from heparinized venous blood ofnormal adult volunteers by Ficoll-Hypaque density gradient centrifugation (Histopaque -1077, Sigma, St. Louis, MO, USA) at 500 x g for 30 minutes. The PBL were washed threetimes in PBS and resuspended in RPMI 1640 (Gibco) containing 10% fetal calf serum (FCS)(Gibco), 2 mM Glutamine, 100 U/mI penicillin, 100 pg/ml streptomycin, 0.25 ig/mlAmphotericin B, and 125 tg/ml gentamicin. By this procedure cell fractions containing 85 -89% lymphocytes were obtained. Viability was 95- 98% by the Trypan Blue exclusion test. Tlymphocytes were prepared by the nylon-wool separation technique (187). Sterile nylon-wool(Robbins Scientific Corporation, Sunnyvale, CA) was packed into 6 ml syringes at 500mg/syringe. The columns were washed 10 times alternately with distilled H20 and 0.02 N HC1and then with 15 ml of 1X Hanktsbalanced salt solution (with Ca2 and Mg2j, followed by15 ml of 10% Fetal Calf Serum in RPMI 1640 media (10% FCS/RPMI) per column. Thecolumns were used immediately or kept frozen at -20 °C for up to 2 months. Frozen columnswere thawed for 30 minutes at room temperature, followed by another 30 minutes at 37 °C.They were then washed with 10 ml of 10% FCS/RPMI, and compressed to lml volume prior36to cell incubation. One ml suspension of isolated PBL (1 x i0 cells/ml) was placed over eachcolumn dropwise and the columns were incubated for 45 minutes at 37 °C. Non-adherent Tlymphocytes were eluted with 12 ml 10% FCS/RPMI. By this method, we routinely obtained5 to 6 X io6 cells/ml. Phenotypic characterization of the lymphocytes in the eluate wascarried out according to the following protocol: aliquots of lymphocyte suspension (200 [Licontaining 2 x io6 celis/ml) were incubated for 30 mm at 4 °C with the following mousemAbs: anti CD2 (for T and Natural killer (NK) cells), anti CD4 (for T helper phenotype), antiCD8 (for T cytotoxic phenotype), anti CD16 and anti CD 56 (for resting NK cells), anti CD2O(for B cells), and anti-CD56 (for resting and activated NK cells) (Phycoerythrin-conjugatedanti CD2, CD8 and CD56 from Coulter Corporation, Hialeah, Fl., Fluorescein-conjugated antiCD4 and CD2O from Becton Dickenson, Mississauga, ONT). The tubes were then filled with2% FCS in TC 199 (Gibco) and centrifuged at 500 x g for 10 minutes. The supernatant wasdiscarded and 0.5 ml of 1% paraformaldehyde solution was slowly added while vortexing thetubes. Fluorescence was read on the Epic Profile I (Coulter Corporation, Hialeah, Fl.). Thequadrant was set by using isotypic control, and dead cells were gated out. CD4+ cellscomprised 57% of the eluate while CD8+ cells constituted 22% of the cell suspension. Therewas a small number of NK cells (<10%), and B cells (4%).2.8 Activation of T cellsFreshly isolated lymphocytes were cultured for 72 hours at 37 °C with mouse mAb to humanLeu 4 (CD3) (10 ng/ml; Becton Dickinson, Mississauga, ONY). Before their use, lymphocytes37were washed extensively and resuspended in 10% FCS/RPMI. Lymphocyte activation wasdetermined by the following protocol: aliquots of lymphocyte suspension (200 ti containing 2x io6 cells/mi) were incubated for 30 mm at 4 °C with fluorescein-conjugated mousemonoclonai anti CD25 antibody (for IL-2R, p55) (Becton Dickenson, Mississauga, ONT.).The cells were washed in PBS and centrifuged at 300 x g for 5 minutes. The pellet wasresuspended and fixed in 1% paraformaidehyde diluted in PBS for 30 mm at roomtemperature. Fluorescence was read at 580 nm. Approximately 2 fold increase in IL-2Rexpression was detected in anti CD3 stimulated lymphocytes in comparison to restinglymphocytes.2.9 HBMEC-lymphocyte adhesion assayConfluent monolayers of HBMEC were treated with recombinant human IFN-y (150 U/mi), orIFN- (2,000 U/mi) or a combination of IFN-f3 (2,000 U/mI) and y (150 U/mi) for 3 days.Prior to use, the cultures were washed extensively. 500 tl of freshly isolated resting T cells oractivated lymphocytes containing 2 x io6 cells/mi were incubated with the EC monolayers for1 hour at 37 °C. Following incubation, the monoiayers were washed thoroughly with warmedHankstbalanced salt solution (HBSS) and PBS (pH 7.2, 10mM) containing Ca2 and Mg2 toremove non adherent lymphocytes, fixed in Acetone: Ethanol (1:1 ratio) for 7 minutes at 4 °C,air dried and washed with Tris buffer. Blocking of the endogenous peroxidase was carried outby incubation with H20 and Methanol (1:4 ratio) for 30 minutes at room temperature. Afterseveral washes with Tris buffer, the cells were incubated for 90 minutes with mouse38monocional anti human leucocyte common antigen (LCA) (Dimension Laboratories Inc.,Mississauga, ONT.) at 1:100 dilution, and then with horseradish peroxidase conjugated goatanti mouse IgG at 1:200 dilution (Jackson Immunoresearch Laboratories Inc., West Grove,PA) for 120 minutes at room temperature. After washing with Tris buffer, the cells weretreated with3,3-diaminobenzidine (0.5 mg/mi; Sigma) and 0.01% H20 in Tris buffer (pH7.6) for 20 minutes at 4 °C, washed with H20, and counterstained with haematoxylin.Subsequently, the monolayers with adherent lymphocytes were covered with JB4 mountingmedium and the walls of the wells cut out with a hot scalpel blade prior to counting. Controlsconsisted of untreated HBMEC cultures incubated with resting or activated lymphocytes.For scanning EM studies of lymphocyte-EC adhesion, HBMEC were grown to confluency oncollagen discs (section 2.10) and then incubated with IFN-y (150 U/mi, 3 days) or leftuntreated (controls), followed by incubation with either resting or anti CD3 stimulated T cellsfor 2 hours at 37 °C. The cells were then processed for scanning EM as described above(section 2.4).2.10 HBMEC-iymphocyte migration assayFor the migration studies, HBMEC were grown on permeable collagen membranes made ofhighly purified, pyrogen-free, pepsin-solubilized collagen forming the floor of 14 mmdiameter wells (Cellagen Discs, ICN Biomedicals, Inc., Cleveland, OH) (Fig. 1). Immediatelybefore seeding the cells, collagen discs were placed inside 24-well culture plates andimmersed in M199. This double chemotactic chamber has been previously demonstrated to be39suitable for studying the adhesion and migration of human polymorphonuclear leukocytesacross cultured bovine brain microvessel EC (7). Plating of EC on collagen discs did notrequire prior coating with fibronectin. Isolated clumps of HBMEC were plated at 50,000cells/cm2.Media were changed every other day. Confluent monolayers of FVIIIR:Ag positiveEC were obtained by six to eight days in culture. Confluent cultures of HBMEC were treatedeither with IFN-y (150 U/mi), IFN-3 (2,000 U/mi) or a combination of IFN-13 (2,000 U/mi)and y (150 U/mi) for 3 days at 37°C. Before their use, HBMEC were rinsed three times andthe last wash was replaced by 200 d of lymphocyte suspension containing 2 X io6 cells/mi.Resting or activated lymphocytes were incubated with EC for 3 hours at 37 °C. Followingincubation, the monolayers were washed with warm M199 and then PBS to remove non-adherent lymphocytes and processed for transmission electron microscopy (section 2.16). Onetm thick and ultrathin sections were cut with an Ultracut E ultramicrotome (Reichert-Jung,Austria). One tm thick sections were stained with toluidine blue, coverslipped and used forquantitation of lymphocyte transendothelial migration by light microscopy (section 2.12). Thinsections were stained with uranyl acetate and lead citrate.2.11 Monoclonal antibody-blocking studiesThe effects of mAbs to IFN-y and HLA-DR on the adhesion and migration of resting T cellsand activated lymphocytes across HBMEC monolayers were examined in separateexperiments. Monolayers were coincubated with IFN-y (150 U/mi) and mouse monoclonaianti human IFN-y IgG (10 tg/ml) for 3 days at 37°C, or with IFN-y (150 U/mI) for 3 days at4037°C followed by incubation with mouse monoclonal anti human FILA-DR IgG (6 tg/m1) for2 hours prior to incubation with lymphocytes. Monolayers were thoroughly washed threetimes with M199 to remove the IFN-y and mAbs before the adhesion and migration assays.Both antibodies were purchased from Dimension Laboratories Inc., Mississauga, ONT.2.12 Quantitation of lymphocyte adhesion and migrationLymphocyte-EC adhesion was quantitated by counting the number of adherent lymphocytes toEC monolayers in one central and four peripheral randomly selected fields of each culture wellusing a bright field microscope equipped with a 1 cm2 ocular grid under X 200magnification. Adhesion index is expressed as the number of adherent T cells per mm2of themonolayer. Transendothelial migration was quantitated by counting the number oflymphocytes that migrated across the monolayers by light microscopy in 1 tm thick plasticsections stained with toluidine blue. For each treatment, 200 sections (40 tm apart), werecounted.2.13 Transmission electron microscopyEC monolayers were washed three times in serum-free media and fixed in 2.5%glutaraldehyde and 2% paraformaldehyde in 0.1M sodium cacodylate buffer (pH 7.35) for 1hour at 4°C. After washing with 0.2M cacodylate buffer for 30 minutes, the cells were postfixed in 1% 0s04 in 0.1M sodium cacodylate buffer for 1 hour at 4°C, block-stained withuranyl magnesium acetate overnight at 4°C, dehydrated and embedded in Epon-Araldite.41Ultrathin sections were stained with uranyl acetate and lead citrate and examined in a PhilipsEM 400.2.14 Localization of FVIIIR:Ag in untreated, cytokine/chemical-treated HBMEC2.14.1 In vitro drug treatment of ECThe effects of calcium ionophore A23 187, ethyleneglycol-tetraacetic acid (EGTA) andinterferon-y (IFN-y) on the release of FVIIIR:Ag were studied in confluent 8 day old cultures.The cells were incubated with growth medium containing: a) lOitM Ca2+ ionophore A23187(Sigma), diluted from a 10mM stock dissolved in dimethylsulfoxide, for 10 minutes; b) 1mMEGTA (Sigma) for 10 minutes and c) 200U/ml IFN-y (Collaborative Research Incorporation,Bedford, MA) for 24 hours. At the end of the incubation period the cultures were washed withPBS and processed for immunoelectron microscopy.2.14.2 Immunoelectron microscopy for FVIIIR:AgEndothelial monolayers were washed with phosphate buffered saline (PBS) with 0.lg/l CaCl2and fixed in freshly prepared periodate-lysine-paraformaldehyde (PLP) fixative (184)overnight at 4°C. After fixation the cultures were washed with PBS for 30 minutes andincubated in cold 10% sucrose solution in PBS for 4 hours, 15% sucrose for 4 hrs and 20%sucrose overnight. The cells were permeabilized by incubation in 0.01% to 0.05% Triton X100 (Sigma) in freshly prepared PLP for 10 minutes at room temperature, washed with PLPfor 20 minutes and then with 0.1M glycine in PBS for 30 minutes. The monolayers were42incubated with 5% normal goat serum (NGS) in 0.1% BSA-Tris buffer (BSA-Tris) for 20minutes and then with the primary antibody (mouse anti-human FVIII antigen) at 1:50 dilutionwith 1% NGS in BSA-Tris for 2.5 hours at room temperature. Following washing in BSATris, the cells were incubated with the secondary antibody (goat antimouse IgG) at 1:20dilution in BSA-Tris for 1.5 hours at room temperature. Cultures were further fixed in 1%glutaraldehyde in PBS containing 0.2% tannic acid for 40 minutes at room temperature,washed in PBS and post-fixed in 0.5% osmium tetroxide (0s04) in PBS for 15 minutes.Following post-fixation, the cells were washed in acetate buffer, block stained in uranylmagnesium acetate overnight at 4°C, dehydrated through graded series of methanol andembedded in Epon-Araldite. Blocks cut from the embedded cultures were re-embedded forcross-sectioning. Thin sections were examined in a Philips EM400 without heavy metalstaining.2.15 Statistical analysisStudent’s t-test, a procedure designed to test for differences in two groups, was used forstatistical evaluation of the data.Sigmastat program was used for this analysis.43RESULTS3.1 Human brain microvessel endothelial cellHBMEC grown on plastic wells or collagen membranes formed confluent monolayersby 7 to 10 days in culture. EC were elongated and grew in close association with each other.Upon reaching confluency, HBMEC exhibited density-dependent growth inhibition withsignificant decrease in mitotic activity and cellular proliferation. There was no difference inthe growth pattern between cells grown on plastic wells and those cultivated on collagen discs(Fig. 2 a, b).Immunofluorescence and immunoperoxidase staining for Factor VIII related antigen(FVIIIR:Ag) revealed strongly positive, perinuclear, granular staining of cells, thus confirmingtheir endothelial origin (Fig. 3a). Binding of Ulex europaeus type I (UEA I) lectin, a markerfor normal and neoplastic human endothelium (188, 189), by HBMEC, was demonstrated bytheir strongly positive immunoperoxidase staining for UEA-1 as previously described (Fig. 3b)(107). There were no other contaminating cells in any of the cultures used for theseexperiments.Ultrastructurally, EC were elongated with overlapping processes (Fig. 4). Junctionalcomplexes with the characteristic pentalaminar configuration of tight junctions were present inareas of cell to cell contact (Fig. 5 a, b, c, d). The cytoplasm was dense and containedprominent rough endoplasmic reticulum, frequent mitochondria and scattered 8 to lOnmintermediate filaments (Fig. 6a). Pinocytotic vesicles were infrequently seen. Rod-shaped,44membrane bound cytoplasmic organelles with parallel arrays of tubular structures (WeibelPalade bodies) were not observed in the cytoplasm of HBMEC. A constant finding in all cellsexamined was the presence of dilated cytoplasmic vesicles or vacuoles bound by a singlesmooth limiting membrane. These vesicular structures were invariably located in the vicinityof the nucleus, in close association with the cisternae of the rough endoplasmic reticulum andwere absent from the most peripheral portions of the cytoplasm and the cell processes. Theyvaried in size from 0.15 to 1.1tm and they appeared empty or contained small amounts ofamorphous material (Fig. 6b).3.2 Immunocytochemical Localization of FVIIIR:AgIn cultures stained for FVIIIR:Ag with the immunogold technique, Snm gold particleswere distinctly localized within the vesicular profiles immediately adjacent to the roughendoplasmic reticulum (Fig. 7a) and close to Golgi cisternae in sections where the Golgiapparatus was present (Fig. 7b) as previously mentioned. The endoplasmic reticulum waslargely unstained with the rare exception of single isolated particles within its cisternae.Occasionally, labeled vesicles communicated directly with cisternae containing gold particles(Fig. 7c). The number of labeled vesicles and the degree of labeling (number of gold particlesper vesicle) varied among different cells. This is largely attributed to variable leakage ofintracellular proteins following permeabilization or to insufficient permeabilization andpenetration of individual EC by the antibodies. Low (0.01%) and high (0.05%) concentrationsof Triton X-100 resulted in poor labeling while a concentration of 0.03% was associated with45frequent and denser labeling. There was no staining of the cytoplasmic membrane or thediscontinuous basement membrane-like material underlying the basal cell surface. The specificstaining was eliminated in control cultures incubated with normal mouse IgG or carrier buffer(Fig. 7d).3.3 Induction of Ia Ag expression on Primary cultures of HBMEC3.3.1 Effects of recombinant human IFN-yTreatment of cultures with IFN-y induced expression of Ia Ag by EC, which wasdependent upon the concentration and length of exposure to IFN-y. Surface labeling wasobserved as early as 12 hours following incubation with 200 U/mI in a small cell population(9.24 ± 0.99%), increased up to 88.35 ± 0.18% after 24 hours and reached 100% after 48 hours(Fig. 8). Ia Ag expression reached plateau levels after 2 days and persisted for 4 days in thecontinuous presence of the cytokine. Expression was maximal with 100 - 200 U/mI IFN-y(100% of cells) and minimal with 10 U/ml (25.76 + 7%) (Fig. 9). Incubation with 20 U/ml ofIFN-y induced Ia Ag expression in 68.73 ± 18.5% of cells, while 90.85 ± 5.5% of cells werelabeled after treatment with 50 U/ml. EC expressing Ia Ag showed diffuse surface staining inthe form of dark brown-black, granular deposits (Fig. bA). In marked contrast, untreated ECinvariably lacked Ia Ag expression as indicated by their consistently negative staining withimmunogold (Fig. lOB). The staining intensity varied with the concentration and length ofincubation with IFN-y. Thus, labeling was less intense in cells incubated with 10 to 20 U/mlfor 4 days or with 200 U/ml for 12 to 24 hours (Fig. bC), and most dense in cultures treated46with higher concentrations for 3 to 4 days (Fig. bA). Within the same culture, the larger cellswere usually stained most intensely. There were no differences in Ia Ag expression amongHBMEC monolayers originating from different individuals and subjected to identical cultureconditions and IFN-y treatment. Staining was not observed in control cultures incubated withnormal mouse IgG, carrier buffer or irrelevant antibody.In monolayers co-incubated with IFN-y and anti-IFN-y mAb for 4 days, induction of IaAg was completely abolished (Fig. 1OD) indicating that IFN-y specifically induces expressionof Class II MHC molecules on human brain EC. Treatment of cells with 200 U/mi IFN-yfollowed by withdrawal and culture in regular growth media resulted in complete reversal of IaAg expression and negative staining of the cultures.Ultrastructural examination following immunogold labeling showed that Ia Ag wasreadily detectable on the apical surface of EC. Gold particles were found at the cell membranewith a tendency to localize on or near thin cytoplasmic processes (Fig. ha). The basal cellsurface was not labeled. No labeling was seen in untreated control cultures (Fig. bib).3.3.2 Effects of recombinant human IFN-f3Treatment of primary HBMEC cultures with IFN-13 for 4 days at concentrations of 100- 6,000 U/mi failed to induce surface expression of Ia Ag as indicated by the negative stainingof EC (Fig. 12a). Coincubation of EC with IFN-y and 3 resulted in downregulation of Ia Agexpression in a dose-dependent fashion. Thus, expression was decreased by approximately40% in monolayers treated with 100 U/mi IFN-13 and 100 U/ml IFN-y for 4 days. Increase of47the IFN- concentration to 250 U/mi and 500 U/mi resulted in 62% and 79% suppression,respectively (Fig. 12b). Downregulation of Ta Ag expression was maximal (89%) followingcoincubation with 100 U/ml IFN-y and 2,000 U/mi IFN-; however, complete inhibition of IaAg expression was not observed even with IFN-3 concentrations as high as 6,000 U/mi (Fig.13).3.4 Kinetics of the downregulation of Ia Ag expression by IFN-In order to obtain greater insight into the temporal effects of IFN-3 on the induced IaAg expression by HBMEC, several treatment protocols were applied. Treatment of HBMECwith 100 U/ml IFN-y for 4 days induced Ia Ag expression in 86% of the EC. Incubation of themonoiayers with a combination of IFN-f3 and IFN-y (6,000 U/ml and 100 U/ml, respectively)for 4 days, significantly downregulated Ta Ag expression with positive immunogold staininglimited to <20% of EC (Fig. 14). Similar levels of downregulation were also achieved whenBC were pretreated with IFN- (6,000 U/mi) for 2 days, followed by a combination of IFN-13and IFN-y (6,000 U/mi and 100 U/mi respectively) for another 4 days (Fig. 14).Interestingly, a much less significant decrease in Ta Ag expression was noted when cultureswere pretreated for 2 days with IFN-y (100 U/ml), followed by a combination of IFN-13 and‘(6,000 U/ml and 100 U/ml respectively) for another 4 days. However, when cultures werepretreated for 2 days with a combination of IFN- (6,000 U/mi) and y (100 U/ml), followed bya 4 day incubation with 100 U/mI IFN-y, no suppression of Ia Ag expression was detected(Fig. 14).48These results indicate that downregulation of Ia Ag expression by IFN-13 is mosteffective when HBMEC monolayers are coincubated with both cytokines with or withoutpretreatment with IFN-13 alone. However, IFN-y-induced Ia Ag expression is not suppressedwhen treatment with IFN-3 and y is preceded or followed by incubation with IFN-y.The effects of IFN-13 on the IFN-y induced Ia Ag expression by HBMEC was furtherdetermined by Enzyme linked Immunosorbent Assay (ELISA) performed on primaryconfluent HBMEC cultures. Various concentrations of IFN- (100 to 6,000 U/ml) were usedin combination with an optimal concentration of IFN-y (100 U/ml) and different treatmentprotocols, similar to the ones used for immunohistochemistry were applied. The results weresimilar to those obtained by immunohistochemistry. Thus, treatment with IFN-13 failed toinduce expression of Ta Ag. Coincubation with IFN-f3 and y resulted in significantdownregulation of Ia Ag expression that was dependent upon the TFN-13 concentration (Fig.15). Significant reduction in Ia Ag expression was noted in cultures treated with 100 U/mIIFN-y in combination with 100 U/ml IFN-3 for 4 days (Fig. 15). Further suppression of Ta Agexpression was observed with higher concentrations of INF-3 (500 U/mI to 6,000 U/ml),however, complete inhibition of Ia Ag expression by IFN-13 was never achieved (Fig. 15).Downregulation of Ia Ag expression was maximal when EC were pretreated with IFN-f for 2days, followed by coincubation with IFN-3 and y for another 4 days. In monolayers incubatedwith IFN-y for 2 days followed by a combined treatment with IFN-y and i for 4 days, asimilarly significant suppression of Ta Ag expression was obtained.493.5 Effects of IFNj’ and 1FN43 on cell morphology, organization and growth3.5.1 IFN-yPrimary cultures of HBMEC grown in regular medium in the absence of IFN-y formedhighly ordered confluent monolayers of elongated, closely associated, contact inhibiting cells(Fig. 16a). EC treated with 200 U/mi IFN-y for 3 to 4 days acquired a spindle-like shape andlong attenuated processes. These markedly elongated cells frequently arranged themselves inill-defined whorls and exhibited prominent overlapping, thus contributing to a uniqueappearance of the monolayers (Fig. 16b).These changes were most conspicuous under SEM examination. Under normal cultureconditions, elongated HBMEC grow in close contact to each other and display distinctmarginal folds in areas of cell-to-cell contact (Fig. 17a). In contrast, EC treated with IFN-ybecame attenuated and their long, thin processes often extended over and covered adjacentcells (Fig. 17b). As a result of this rearrangement, intercellular contacts and marginal foldsbecame less prominent and the monolayers lost their highly organized appearance. The abovemorphological changes were reversed 4 days following withdrawal of the cytokine from theculture medium and were not observed in cultures co-incubated with IFN-y and anti-IFN-yantibody.The effect of IFN-y on the growth of HBMEC was less profound. Thus, the number ofcells in primary HBMEC cultures treated with 150 U/ml IFN-y from day 1, was slightly lessthan that in control cultures (Fig. 18). This slight inhibitory growth effect of IFN-y providesfurther support to the observation that re-arrangement and overlapping of HBMEC is the direct50effect of the cytokine and not the result of cell overgrowth.3.5.2 IFN-Primary cultures of HBMEC grown in media containing IFN- (6,000 U/ml) for 4days showed no changes in cell morphology and organization and were morphologicallyidentical to untreated cultures (Fig. 19a). When monolayers were coincubated with IFN-j3and IFN-y (6,000 U/mi and 200 U/mi, respectively) for 4 days, changes in cell shape andorganization of the monoiayers induced by IFN-y, were not detected (Fig. 19b). Similarly,scanning EM studies of EC treated with IFN-f3 alone, or a combination of IFN-3 and IFN-y,confirmed the normal morphology and growth pattern of HBMEC and the absence of themorphological phenotype and overlapping induced by IFN-y (Fig. 19, arrows).On the other hand, the effect of IFN- on the growth of HBMEC was profound:inhibition of growth was observed when HBMEC were treated from day 1 with IFN-3 alone(1,000 U/mi) or a combination of IFN-13 and IFN-y (1,000 U/mi and 100 U/mi, respectively)when compared to untreated cultures (Fig. 18).3.6 Permeability of HBMEC monolayersIn order to examine if IFN-y, in parallel to morphological changes, also inducedchanges in the permeability of the monolayers to macromolecules, confluent treated anduntreated cultures were incubated with HRP and the labeling of intercellular contacts andcytopiasmic vesicles was assessed ultrastructurally. Intercellular contacts that impeded the51tracer entirely or were penetrated only for a short distance from either the apical or basal cellsurface by HRP, were considered impermeable. In untreated cultures, 75.2% ofinterendothelial junctions prevented the passage of HRP, in contrast with 36.6% in culturesincubated with the cytokine for 4 days (Table 1). In untreated monolayers, EC formed a singlecell layer and were bound together by junctions, most of which were not labeled with thetracer (Fig. 20a). Focally, HRP penetrated an intercellular contact for a short distance from thebasal aspect of the monolayer before being arrested at a junctional complex of an otherwiseintact cleft (Fig. 20b). In treated cultures, interendothelial clefts were often penetrated by thetracer throughout their entire length (Fig. 20c). Overlapping of EC resulted in the formation of2 or more layers. Horseradish peroxidase often penetrated the intercellular clefts between ECat the top layer, and extensive deposits were found between adjacent cells at the lower layers(Fig. 20d). The number of cytoplasmic vesicles labeled with HRP was equally low in controland experimental cultures (Table 1), indicating that, contrary to the prominent conformationaland organizational changes, the pinocytotic activity of HBMEC is not affected by IFN-ytreatment.3.7 Lymphocyte characterizationThe different subsets of lymphocytes obtained by the nylon-wool separation techniquewere characterized by using mouse mAbs directed against human T cell surface molecules ofinterest. The results were determined by Coulter cytometry with the Epics Profile Analyzer.Greater than 90% of cells (91.3 ± 2.8%) recovered after nylon-wool seperation were T cells.52When peripheral blood lymphocytes were activated by incubation with mouse anti-humanLeu-4 (CD3) monoclonal antibody for 3 days at 37 °C, up to 91.5 ± 6% of cells were Tlymphocytes. Stimulation with anti-CD3 resulted in more than 2 fold increase in Interleukin2 receptor (IL-2R) expression as compared to resting T lymphocytes (Fig. 21). Scanning andtransmission EM studies showed that the surface membrane of activated T cells appearedruffled with numerous folds on the cell surface. In contrast, resting lymphocytes exhibited asmooth surface membrane. Subsequently, anti-CD3 treated T cells provided morphologicaland functional evidence of activation.3.8 Human T-iymphocyte adhesion to untreated and cytokine-stimulated HBMECConfluent cultures of HBMEC grown on plastic wells for 7 days were treated withIFN-y (150 U/mI) or IFN- (2,000 U/ml), or a combination of IFN-y (150 U/mi) and 1 (2,000U/mi) for 3 days. Controls consisted of monolayers grown to confluency in the absence ofcytokines.By light microscopy, following immunoperoxidase staining for leukocyte commonantigen, a small number of resting lymphocytes adhered to untreated HBMEC (50 ± 10 Tcells per mm2 of EC monolayers) (Fig. 22, 25). Treatment of HBMEC with IFN-y for 3 daysresulted in more than 3 fold increase in adhesion over control values (165 + 13 T cells permm2 of EC monolayers) (Figs 23, 25). Treatment of the monolayers with IFN-13 had no effecton the basal adhesion of lymphocytes to endothelium (p = 0.953) (Figs. 24, 25). In culturesincubated with both cytokines, adhesion was not significantly different from that observed in53control cultures (p = 0.004) (Fig. 25).Examination by SEM revealed that untreated HBMEC monolayers consisted of closelyassociated cells with marginal folds in areas of cell to cell contact. A similar morphology andgrowth pattern was observed in EC cultures treated with IFN-f3 or with a combination of IFN-yand Resting T lymphocytes first adhered to the endothelium by extending pseudopodia thatcontacted the endothelial surface (Fig. 26a). Eventually they positioned themselves betweenadjacent EC (Fig. 26b), and began migrating across the monolayer (Fig. 26c). Lessfrequently, lymphocytes were seen penetrating the apical EC plasma membrane and movingthrough the endothelial cytoplasm (Fig. 26d). Monolayers treated with IFN-y exhibitedelongation and overlapping of EC. A large number of lymphocytes established contact withthe endothelium via pseudopodia (Fig. 27a), and singly or in small aggregates alignedthemselves preferentially along the borders between adjacent EC in preparation for crossingthe monolayers (Fig. 27b). Adhesion and penetration of EC cytoplasm by lymphocytes wasencountered much less frequently. Ultrastructurally, lymphocytes first established contact withintact or cytokine treated EC by extending finger-like cytoplasmic processes to the surface ofthe endothelium. The two cell membranes became closely apposed.3.9 Adhesion of activated T-lymphocytes to untreated and cytokine stimulatedHBMECActivation of T-lymphocytes with anti-CD3 resulted in a three fold increase inadhesion to untreated EC over control values (202 + 36 activated T cells per mm2 of EC54monolayers) (p = 0.0 10) (Figs. 28, 31). Following pretreatment of EC with IFN-y, adhesion ofactivated T cells to endothelium was 2 fold greater than adhesion to untreated EC (403 ± 29activated T cells per mm2 of EC monolayers) (Figs. 29, 31) (p = 0.012). Pretreatment ofHBMEC with IFN-j3 had no effect on the adhesion of activated T cells to the endothelium.Preincubation of EC with a combination of IFN-y and i, however, resulted in levels ofadhesion comparable to those obtained when activated T cells were incubated with untreatedEC (Fig. 30, 31), indicating that IFN-3 downregulates the IFN-y-mediated increase in adhesion(218 activated T cells per mm2 of EC monolayers) (p 0.788) (Fig. 31). Examination bySEM revealed prominent changes in the morphology of activated T lymphocytes whichappeared larger and exhibited a ruffled cell membrane with numerous folds. Activatedlymphocytes adhered to the endothelium in large numbers (Fig. 32a-c). They appearedconsiderably larger than resting T cells and their surface was decorated with numerous foldsand cytoplasmic projections. As observed with resting T cells, activated lymphocytesestablished close contacts with EC by means of cytoplasmic projections and usually positionedthemselves along the borders between adjacent EC in both untreated and cytokine-treatedcultures in preparation for migration. Direct penetration of the endothelial cytoplasm byadherent lymphocytes was rarely observed. In such instances, a protuberance on the apicalsurface of the endothelium, having the size and shape of an activated T cell, indicatedmovement through the EC cytoplasm (Fig. 32, arrows). Ultrastructurally, activatedlymphocytes displayed abundant cytoplasm, increased numbers of mitochondria and variablenumbers of cytoplasmic vacuoles containing amorphous, flocculent material (Fig. 33). The55cell surface was extremely irregular due to the presence of numerous thin, finger-like, variablyundulating, cytoplasmic processes. Several points of close cell-to-cell contact betweenendothelium and processes of adherent lymphocytes were present (Figs. 33, 34).3.10 Effects of blocking antibodies on lymphocyte adhesionThe ability of mAbs to IFN-y and human HLA-DR to block the adhesion of restingand anti-CD3 activated T cells to HBMEC was examined. In monolayers coincubated withIFN-y and mAb to IFN-y for 3 days, adhesion of resting and activated T-lymphocytes wassignificantly decreased (p = 0.002 and p = 0.001, respectively) (Figs. 25, 31). When ECcultures were treated with IFN-y for 3 days, followed by incubation with mAb to humanHLA-DR for 2 hours prior to incubation of T cells with EC, marked suppression of IFN-yinduced adhesion of resting and activated T-cells was observed (p = 0.012 and p = 0.014,respectively) (Figs. 25, 31, 35, 36).3.11 Transendothelial migration of resting T lymphocytesTransendothelial migration of resting T lymphocytes across untreated HBMECmonolayers was minimal (Figs. 37a, 38). Significant increase in migration, up to 3 fold, wasobserved when cerebral EC were pretreated with an optimal concentration of IFN-y (150U/ml), known to induce maximal Ta Ag expression, for 3 days prior to incubation with thelymphocytes (p < 0.001) (Figs. 37b, 38). In contrast, IFN- treatment had no effect onmigration as the numbers of T cells detected underneath the EC monolayers were comparable56to those that migrated across untreated HBMEC (p = 0.304) (Fig. 38). Moreover, the INF-ymediated increase in migration was markedly suppressed when EC were preincubated for 3days with a combination of IFN-y and IFN- and then allowed to interact with the restinglymphocytes for 3 hours (Figs. 37c, 38). Adhesion of lymphocytes directly to collagenmembranes in the absence of EC was not observed. The results indicate that IFN-13significantly downregulated the IFN-y-induced increase in transendothelial migration (p <0.001).One tm thick, toluidine blue stained cross sections of the monolayers revealed that Tcells initially attached and subsequently moved across the endothelium. At the end of theirmigration, lymphocytes positioned themselves underneath the monolayer between EC and thecollagen membrane and assumed a flattened, elongated shape. The endothelial monolayersoverlying the migrated lymphocytes appeared to retain their continuity (Fig. 37a to c).Examination by TEM revealed that resting T-lymphocytes initiated their migrationacross the EC monolayers by directing one or more cytoplasmic processes between twoadjacent EC (Fig. 39a). Eventually, a small segment of the cytoplasm, without the nucleus,was inserted between the two EC and was followed by the remaining cytoplasm and nucleus(39 b, c). After passing between the EC, the lymphocytes became elongated and flattened andremained between the overlying EC and the underlying collagen membrane (39d).Throughout the migratory process, lymphocytes remained in close contact with the EC, theadjacent plasma membranes of the two cell types being tightly apposed (Figs. 39a to d).Infrequently, lymphocytes migrated by moving through the cytoplasm of EC. A lymphocyte57was considered moving through rather than between adjacent EC only when the cytoplasm ofthe EC completely surrounded the lymphocyte (Fig. 40). At the end of the migration period,EC monolayers rapidly assumed their continuity and appeared structurally intact. Culturestreated with IFN-y showed variable overlapping of EC. Lymphocytes that had completed theirmigration across adjacent EC of the top layer, would then proceed to migrate across the secondlayer of EC. The integrity of the monolayers was reestablished once resting T cells completelymigrated across the untreated/cytokine treated EC (Fig. 41 a, b).3.12 Migration of activated T lymphocytes across untreated and cytokine treatedHBMEC monolayersTo determine whether nonspecific activation of T lymphocytes had any effect onmigration, peripheral blood lymphocytes treated with anti-CD3 antibody for 3 days wereincubated with HBMEC for 3 hrs.Activation of T cells resulted in a four fold increase in the number of cells thatmigrated across untreated monolayers of HBMEC as compared with the migration of restinglymphocytes across untreated brain endothelium (p < 0.001) (Figs. 38, 42). Pretreatment ofEC with IFN-y further increased the migratory response by approximately 30% (p < 0.00 1)(Fig. 38). In contrast, IFN-13 had no effect on the basal level of migration of activated T cells(p = 0.341). Moreover, when HBMEC were preincubated with a combination of IFN-y andIFN-13, the level of migration was not different from that obtained when activated Tlymphocytes migrated across untreated EC monolayers (p = 0.268), indicating that IFN-1358downregulated the IFN-y-mediated increase in migration (Fig. 38).Ultrastructurally, large numbers of activated lymphocytes migrated across theendothelial monolayers. Although migration proceeded in a fashion similar to the oneobserved during migration of resting T cells (Figs. 43 a, b), crossing of the monolayers bymeans of moving through the cytoplasm of EC, was not observed in any of the materialexamined. Migration of activated lymphocytes was not associated with any apparentdisruption of the monolayers (Fig. 43b). Lymphocytes that had completed their migrationacross adjacent EC of the top layer, would then proceed to migrate across the next layer of EC(Fig. 44).3.13 Effects of blocking antibodies on lymphocyte migrationT lymphocyte migration across IFN-y-treated HBMEC monolayers was significantlyblocked by preincubation of EC with a mAb to human HLA-DR (p < 0.00 1) regardless of theactivation status of lymphocytes. The level of suppression approximated that obtained bytreating EC with a combination of IFN-y and IFN-f (Fig. 38). These results suggest that classII MHC molecules (Ia Ag) play a central role in the IFN-y-induced upregulation of Tlymphocyte migration across HBMEC, irrespective of the activation status of the lymphocytes.3.14 Effects of calcium ionophore A23187, EGTA and IFN-y on the constitutivepathway of factor VIIIR:Ag releaseTreatment of the monolayers with 10.tM Ca2+ ionophore A23187 for 10 minutes59resulted in almost complete loss of staining (Fig. 45a), while incubation with 1mM EGTA for10 minutes was associated with slightly increased numbers of labeled vesicles (Fig. 45b).When the monolayers were preincubated with 200U/ml IFN-y for 24 hours, there was asignificant increase in the number of immunostained vesicles over the untreated cultures (Fig.45c). In order to quantitate these findings, the number of labeled and unlabeled vesicles wascounted in 100 cells in each group of treated and in untreated cultures. Fig. 46 summarizesthese results. The differences reflect variations in the percentage of immunostained vesicles.There was no appreciable difference in the number of gold particles per vesicle betweentreated and untreated cells. The difference in the number of labeled vesicles between controlsand IFN-y treated EC was statistically significant (p = 0.000), while no significant differencewas found between controls and EGTA treated cultures (p = 0.21).60DISCUSSION4.1 INFLUENCE OF CYTOKINES ON Ia Ag EXPRESSION ON HBMECThe first specific aim of this thesis was to determine whether Ia Ag is constitutivelyexpressed in primary cultures of HBMEC and whether its expression can be induced andmodulated in vitro by the cytokines IFN-y and IFN-.4.1.1 Human brain microvessel ECHuman brain microvessel EC in primary culture form confluent contact-inhibitingmonolayers composed of elongated, closely associated cells. EC are uniformly positive forFactor VIIIR:Antigen, the most specific marker for cells of endothelial origin, and bind thelectin Ulex europaeus, a marker for human EC. Cultured HBMEC contain few pinocytoticvesicles and are bound together by tight junctional complexes that restrict the paracellularmovement of macromolecules. Primary cultures of HBMEC retain their human EC properties,exhibit morphological and permeability characteristics similar to cerebral endothelium in vivoand, therefore, provide a useful in vitro model for studying the biology and immunopathologyof these cells.4.1.2 Induction of Ia Ag expression on primary cultures of HBMECThe present studies demonstrate that human recombinant IFN-y induces de novo61expression of class II MHC antigen (Ta Ag) by HBMEC in primary culture in a time andconcentration - dependent manner. Unstimulated HBMEC grown under standard cultureconditions do not constitutively express Ta Ag as indicated by lack of immunogold staining onlight and electron microscopy. Previous in vivo immunohistochemical studies havedemonstrated absence of Ia Ag expression by EC within the normal human CNS with lowlevels of reactivity detected in blood vessels of patients with brain neoplasms, abscesses,autoimmune connective tissue disease, cerebral infarcts and in older patients withoutidentifiable CNS lesions (68, 190 - 192). Although the EC used in our studies were isolatedfrom normal brains of several donors with a wide age distribution, expression of Ta Ag was notobserved in any of the untreated cultures. A similar lack of constitutive expression of class TImolecules has been observed in primary cultures of rat brain endothelium (70), in freshlyisolated human umbilical vein EC (HUVEC) (22, 193), in serially passaged cultures of humancerebral vascular EC (136) and HUVEC (194), as well as in human glioblastoma multiformecells (195), and cultured adult human astrocytes (25) maintained under normal cultureconditions. Contrary to these reports, EC of normal guinea pig CNS display surface MHC invivo and in vitro (196) and minimal basal expression has been reported in primary cultures ofrhesus monkey cerebral endothelium (197), while cultured rat heart vascular EC constitutivelyexpress considerably higher levels of Ia Ag (64). It is apparent, from the above studies, that thepresence of Ia Ag on normal, unstimulated vascular endothelium may vary among differentspecies and vascular beds.Previous studies on Ta Ag induction by IFN-y on HUVEC report a rapid increase of62MHC class II mRNA that precedes surface expression by 1 - 2 days and rapidly declines toalmost undetectable levels following withdrawal of the cytokine, while surface expressiondeclines slowly after 4 days (80). In HBMEC, removal of IFN-y from the media results inuniform loss, rather than decrease to lower levels, of class II MHC surface expression after 4days. Rat heart endothelium, however, behaves in a much different way, since withdrawal ofIFN-y is not followed by return of the Ia Ag expression to basal levels after 3 days (64).4.1.3 Surface localization of Ia Ag on HBMECPrevious immunohistochemical studies on MS and EAE have demonstrated thatsurface expression of Ta Ag on EC is discontinuous along the microvessel lumen, so that Ia+cells are interposed between EC lacking Ia Ag expression (68, 139). A similarly variableexpression of Ta Ag was observed in vitro when HBMEC were treated with low concentrationsof IFN-y or with higher concentrations for less than 2 days. Taken together with the in vivostudies, these observations may indicate individual cell variation in the regulation of class IIMHC molecule expression.Induction of Ia Ag expression on HBMEC was restricted to the apical portion of thecell membrane. Immunogold particles were not identified on the lateral or basal cell surfaces.Our findings correlate with previous immunohistochemical studies in acute EAEdemonstrating Ta expression on the luminal but not abluminal surface of cerebral microvesselEC (139) and with similar observations in a variety of epithelial cells in mice treated withIFN-y (72). Although the mechanisms responsible for the asymmetrical presentation of Ta Ag63on the cell membrane are not known, polarization of expression is probably of functionalsignificance since it would enable circulating T lymphocytes to recognize antigen inassociation with class II MHC molecules on the luminal surface of the cerebral endotheliumand then migrate to sites of inflammation.4.1.4 Effects of IFN-j3 on Ia Ag expression by HBMECHuman recombinant IFN-3 failed to induce expression of Ia Ag on HBMEC at allconcentrations tested. A similar lack of Ia Ag expression has been previously reported incultured adult human astrocytes (25), human dermal microvascular EC (84), and humanglioblastoma multiforme cells (26) treated with IFN-3. In our studies, the results obtainedfrom the immunocytochemical staining and ELISA indicate that IFN- downregulates theIFN-y-induced Ia Ag expression in a concentration-dependent manner. Immunocytochemicallabeling indicates the total number of Ia-positive cells in the cultures, but provides noinformation on the membrane density of HLA-DR molecules per cell. ELISA providesrelative measurement of the total density of HLA-DR molecules within the culture but with noindication of the number of cells expressing Ia Ag. The suppressive effect of IFN-f3 on Ia Agexpression has been previously observed in other EC systems (105, 106). Downregulation ofthe IFN-y induced Ia Ag expression by IFN-13 has also been reported in cultured adult humanastrocytes (25), human glioma cells (26), murine macrophages (27, 40), blood monocytesisolated from MS patients (28), and in an astrocytoma cell line (41).644.1.5 Regulatory mechanism of Ta Ag expressionIt has been shown that the induction of Ta Ag on macrophages by IFN-y operates at thelevel of transcription and requires de novo synthesis of a new protein(s) (198). It has beenreported that the plateau values of HLA-DR mRNA content in HUVEC and human dermalfibroblasts treated with IFN-y precede maximal surface expression by 1 to 2 days (80). Thiscould explain the lag period of 12 to 24 hours between addition of IFN-y to the media anddetection of Ta Ag surface expression by immunohistochemistry on HBMEC (127).Interferons-13 and y bind to different receptors on the cell surface (21) and the inhibitory effectof TFN-13 on IFN-y induction of the Ia Ag genes is exerted at the transcriptional level (41, 75).It has been suggested that there is a complex interplay of trans-acting factors involved inmodulating the expression of the Ia genes product and the subsequent expression of theirpeptide products on the cell surface (41, 75). The fact that IFN- failed to completely inhibitthe induction of Ia Ag by IFN-y is not fully understood at the present time; further studies arerequired in order to elucidate the exact mechanism(s).4.1.6 Kinetic studies on the modulation of Ta Ag expression by interferons y and 13In our studies, the most significant downregulation of the IFN-y induced Ia Agexpression was found when HBMEC were either coincubated with the two cytokines for 4days or pretreated with IFN-13 for 2 days and then treated with a combination of IFN-13 and yfor 4 days (approximately 80% reduction). In contrast, significant decrease in Ta Agexpression was not observed (0% to 15% suppression) when EC were pretreated with TFN-1365and y for 2 days, followed by another 4 day treatment with IFN-y, or pretreated for 2 dayswith IFN-y followed by coincubation with IFN-3 and y for another 4 days. Similarobservations have been reported in cultured adult human astrocytes. The addition of IFN-ct or3 24 hours after incubation of astrocytic cultures with IFN-y did not significantly alter HLADR expression, while IFN-a or I added 24 hours before or at the initiation of incubation withsuboptimal concentrations of IFN-y reduced the extent of HLA-DR expression (25). Furtherwork using human astrocytoma cell lines demonstrated that the suppressive effect of IFN-3 onthe HLA-DR induction by IFN-y was relatively gene-specific since IFN-j3 could not impairthe induction of intercellular adhesion molecule-i (ICAM-i) expression by IFN-y in these celllines (4i). Similar results were observed in HDMEC, and the authors speculated that theeffect of IFN-y on HDMEC may be mediated through multiple distinct pathways which can beindependently regulated (i06). Consequently, the results could not be explained by IFN-f3downregulation of IFN-y receptors or defective receptor-linked signal transduction. Theinhibition was also suggested to be tissue-specific because IFN-.13 did not antagonize IFN-yinduction of HLA-DR expression in human monocytes (4i). The results of the present studiesindicate that in order to effectively suppress the induction of Ta Ag by IFN-y in vitro, IFN-13must be present continuously in the culture media. It is also shown that once the cells havebeen activated by IFN-y, downregulation of Ia Ag expression does not occur in the continuedpresence of IFN-j3 in the culture media. Taking into consideration that increased levels of IaAg have been associated with induction of autoimmune disorders of the CNS such as MS (68,i99), and that antibody blocking directed against class II MHC determinants can prevent the66induction of experimental autoimmune disease (200), these findings may partly explain thesignificant therapeutic potential of TFN-13 in MS (281). Inaba et al. have shown that there is acorrelation between downregulation of Ta Ag expression and reduced levels of antigenpresentation by macrophages in vitro (27). Together with the finding that IFN-y is unsuitablefor use as a therapeutic agent in MS (15), Joseph et al. suggested that administration of IFN-flin patients with MS could have beneficial effects if reduced Ta Ag expression occurs, and thereis reduced antigen presentation in the CNS (26). The expected results would be longerremission periods or fewer relapses in MS patients. Therapeutic application of recombinantinterferon beta-lb for the treatment of MS has recently reported that the cytokine is welltolerated and has a beneficial effect on the course of relapsing-remitting MS (281). Based onthe results obtained from our studies on HBMEC and others (26 - 28, 40), a repetitive dosingwith IFN-3 may be essential to effectively downregulate Ia Ag expression.4.2 EFFECTS OF INTERFERONS y AND f3 ON THE MORPHOLOGICALPHENOTYPE AND GROWTH OF HBMEC, ORGANIZATION OF THEMONOLAYERS AND PERMEABILITY TO MACROMOLECULESThe second specific aim of this thesis was to determine whether IFN-y and IFN-3 exertantiproliferative effects on HBMEC and whether treatment with these cytokines canmodulate the morphological phenotype and organization of the EC cultures, and alterthe permeability of the monolayers to macromolecules.674.2.1 Effects of IFN-y and IFN-3 on HBMEC growthThe anti-proliferative effect of IFN-y on primary cultures of HBMEC correlates withprevious studies demonstrating inhibition of cell growth by IFN-y induced on extracerebrallarge and small vessel endothelial cultures in a dose-dependent manner (84, 126, 201, 202) andpossibly through modulation of the EC growth factor receptors (202). Lower concentrations ofIFN-y (10 - 100 U/mi), however, appear to have a stimulating effect on cultured HUVEC bothin the absence and presence of EC growth factor (203). In addition, IFN-y significantly inhibitsformation of endothelial tubular structures in in vitro models of angiogenesis (204, 205).When the cells were treated with IFN-13 or a combination of IFN-3 and y, significantgrowth inhibition was detected. The antiproliferative effects of IFN-y and especially of IFN-f3on primary cultures of HBMEC correlate with previous studies demonstrating inhibition ofcell growth by these cytokines induced on human dermal microvascular EC (HDMEC) (84),human glioblastoma multiforme cells (195), cultured human brain tumors (206), and humanvascular smooth muscle cells in vitro (207).4.2.2 Effects of interferons y and on HBMEC morphology and organization of the ECmonolayersIFN-y-treated EC undergo unique changes in their morphology and organization, whichare associated with a considerable increase in the permeability of confluent cultures tomacromolecules. Treatment of HBMEC with IFN-y induces marked elongation of EC,prominent overlapping and frequent arrangement in a whorled pattern. A similar alteration of68the morphological phenotype and monolayer organization has been previously reported incultures of HUVEC (126, 202) and HDMEC (84) treated with IFN-y for 3 to 4 days and hasbeen shown to be associated with reorganization of the cytoskeletal filaments and considerableloss of the fibronectin matrix (126).It has been previously demonstrated that IFN-3 alters the morphology of culturedHDMEC. IFN-13 treated cells become spindle - shaped, an alteration which was also observedin cultures treated with IFN-y in comparison to untreated cells that showed the typicalmorphology of human EC (84). In contrast, morphologic changes were not observed inHUVEC treated with IFN- (126). Furthermore, treatment of HBMEC with IFN-3 failed toinduce structural or organizational alterations on the monolayers; in fact, it inhibited themorphological changes induced by IFN-y when the cells were incubated simultaneously with acombination of IFN- and y. These observations further emphasize the heterogeneity whichexists between EC derived from different organs or species (19).Our studies, therefore, demonstrate that IFN-3 downregulates the expression of Ia Aginduced by IFN-y on HBMEC and alone or in combination with IFN-y has greaterantiproliferative effect on these cells than IFN-y. In addition, IFN-13 downmodulates the IFNy-mediated changes in cell morphology and organization of the EC monolayer which may berelevant to the in vivo immune response. These findings and the work of other investigatorswould indicate that in situ vascular changes take place in response to cytokines generated atthe inflammatory site. Thus, IFN-y alone or in combination with other locally generatedcytokines, induces changes that may mimick immune regulatory events which signal69endothelial preparation for inflammatory cells to adhere and transmigrate, while IFN- mayplay a negative regulatory role in inflammation or disorders upmodulated by enhanced IFN-ysecretion. In fact, it has been reported that systemic administration of IFN-f3 to MS patientsinhibits endogenous IFN-y synthesis in their peripheral blood mononuclear cells (38). At thepresent time, it is not known whether cerebral EC in vivo undergo the same or a similarspectrum of changes in response to cytokines in inflammatory foci of the human CNS.4.2.3 Permeability of IFN-y treated HBMEC monolayers to macromoleculesHuman cerebral microvessel EC in primary culture are bound together by tightjunctions and have a paucity of cytoplasmic vesicles, two important morphologicalcharacteristics of their in vivo counterparts (107, 186). Under standard culture conditions, thegreat majority of interendothelial junctions restrict the passage of HRP. In cultures incubatedwith IFN-y, an increase in the permeability of the monolayers was observed that coincidedtemporally with changes in morphology and rearrangement of the cells. The number of labeledcytoplasmic vesicles was not increased in IFN-y treated monolayers indicating that increasedjunctional permeability is primarily responsible for the permeability changes of themonolayers. The mechanism(s) responsible for the increased junctional permeability are notknown at present. Recent in vitro studies have demonstrated that tumor necrosis factor (TNF)treated aortic EC cultures undergo prominent cytoskeletal changes similar to those induced byIFN-y alone or in combination with TNF, which are temporally related to increase in thepermeability of the monolayers to macromolecules and are regulated by G protein (208). The70fact that leakiness of intercellular contacts appears concomittantly with the morphologicalchanges of the endothelium following IFN-y treatment may indicate that physiologically??tightfl tight junctional complexes fail to form during the extensive rearrangement of the cellsand their cytoskeleton. However, other mechanisms, such as modulation of regulatory proteinsor cell surface molecules by IFN-y cannot be excluded. Disruption of the BBB has beenpreviously described as an early and critical event in the evolution of EAE (119, 120, 209,210). Recent electron microscopic studies indicate that increased junctional permeability aswell as increased interendothelial space and migration of inflammatory cells are primarilyresponsible for the increased permeability of the BBB to macromolecules in this disease (211).The functional significance of the in vitro morphological and permeability changes ofHBMEC, observed in this study, is presently unknown. If, however, similar changes areinduced in situ on cerebral EC by cytokines released locally by activated T-lymphocytes, theywould provide an additional mechanism for the opening of the BBB and could facilitate thetransmigration of inflammatory cells from blood into brain across the endothelial barrier.4.3 SIGNIFICANCE OF Ia Ag EXPRESSION BY HBMECExpression of Ia Ag in situ by cerebral vascular endothelium has been previouslydemonstrated in autoimmune demyelinating CNS disorders. Thus, class II MHC moleculeshave been localized on the surface of EC lining microvessels at the edge of demyelinatingplaques as well as within the adjacent white matter in acute, active and silent chronic MSlesions (68, 138). The presence of Ia positive EC has also been documented in acute EAE71(139, 140), while expression of Ia Ag by cerebral endothelium in chronic relapsing EAEappears to coincide with the appearance of inflammatory cell infiltrates and diminishes wheninflammation subsides (141). In addition, murine cerebral EC isolated from SJL mice withEAE are able to present antigen to sensitized syngeneic lymph node cells following incubationwith IFN-y in vitro (137). Contrary to these observations, cultured rat brain EC are noteffective at stimulating T-cell division and therefore, have not been considered important asantigen presenting cells (212). Recent studies using murine endothelial and fibroblast celllines to determine their capabilities in presenting antigen to helper T cells have reported thenecessary requirements for costimulatory signals (213 - 216). The lack of signals such as B7,CI’LA4 molecules on the surface of the potential antigen presenting cells can result in theinability of T helper cells to proliferate in response to a specific antigen. Consequently, theconclusions drawn by Pryce et al. (212) that rat brain EC are not important antigen presentingcells, prior to the realization that costimulatory signals may be required for the efficientfunction of antigen presenting cells, deserve further investigation. The present workdemonstrates that class II MHC molecules are not detectable on intact HBMEC isolated andcultured from normal human brain microvessels by the methods employed in our study, butcan be specifically induced in vitro by human recombinant IFN-y in association withprominent alterations in the morphology, organization and permeability of the monolayers tomacromolecules. Although the ability to present antigen by HBMEC has yet to beunequivocally proven, our findings indicate a possibly important role of the human cerebralendothelium in lymphocyte-endothelial interactions, lymphocyte recruitment and alteration of72blood-brain barrier permeability in immune - mediated CNS inflammation.4.4 ADHESION OF RESTING AND ANTI-CD3 STIMULATED LYMPHOCYTESTO UNTREATED, IFN-y and/or IFN- TREATED HBMECThe third specific aim of this thesis was to examine the effects of cytokine treatment ofHBMEC on the adhesion of resting and anti-CD3 stimulated lymphocytes to theendothelium.4.4.1 Activation of lymphocytesThe CD3-molecular complex is comprised of a series of noncovalently linkedpolypeptides (217). Monoclonal antibodies directed against the CD3 molecules have beenwidely used to study T cell activation (218, 219). The binding of anti-CD3 antibodies to Tcells leads to the rapid hydrolysis of phosphatidylinositols and results in an increase in freeintracellular calcium concentration, in generation of diacylglycerol, and activation of proteinkinase C (220, 221). It is notable that binding of mAbs to CD3 molecules clearly results in thegeneration of activation signals (220), but this situation has been reported to be insufficient toinitiate T cell proliferation (222). Moreover, Ledbetter et al. (223) have demonstrated thatbinding of anti-CD3 antibody to T cells results in a rapid rise (4 to 6 fold) in cyclic adenosinemonophosphate (cAMP), and high levels of cAMP are known to inhibit T cell growth. It hasalso been shown that other cAMP-elevating agents such as prostaglandin E2, cholera toxin,73and the cell-permeable analog 8-bromo-cAMP can inhibit nuclear IL-2 transcription anddecrease the stability of IL-2 mRNA (224, 225). Interestingly, other studies have shown that Tcells proliferate in response to soluble anti-CD3 mAb in the presence of IL-2, a lymphokinecentral to the mediation of antigen-activated T cell proliferation. IL-2 is produced by activatedT cells, and binds to the IL-2 receptor of secreting T cells and other antigen-stimulated cells(226). Since the expression of IL-2 receptor is reported to be increased when T lymphocytesare treated with anti-CD3 antibody (227, 228), activation of peripheral blood lymphocyteswith anti-CD3 mAb was confirmed in this study by the upregulation of IL-2 receptor ratherthan by the conventional tritiated thymidine uptake assay. The results obtained byFluorocytometry demonstrated more than 2 fold increase in IL-2R expression. By SEM,lymphocytes treated with CD3 mAb appeared larger than resting T cells and their surfacemembranes exhibited a ruffled appearance in contrast to the smooth membrane of resting Tcells, confirming their activated state. Moreover, transmission EM studies showed numerouscytoplasmic folds and finger-like projections at the surface of activated lymphocytes alongwith a significant number of vacuoles in the cell cytoplasm. Subsequently, anti-CD3 activatedT cells are functionally and morphologically different from resting T lymphocytes.4.4.2 Adhesion of resting lymphocytes to untreated, IFN-y and/or IFN-3 treated HBMECThe present studies demonstrate that lymphocyte adhesion to cultured HBMEC can bemodulated by treatment of the EC with IFN-y and/or IFN-13. A low basal level of adhesionbetween resting T cells and untreated EC was detected. Pretreatment of EC with an optimal74concentration of IFN-y (150 U/mi) known to induce Ia Ag expression (127), significantlyaugmented lymphocyte-EC adhesion. Scanning EM studies demonstrated that a small numberof resting lymphocytes, with relatively smooth surface membranes, adhered to confluentmonolayers of untreated HBMEC. The lymphocytes lined up on the borders between adjacentEC, and occasionally adhered to the apical surface of the endothelium. Significant increase inadhesion of resting lymphocytes to IFN-y-treated EC was noted by light microscopy and SEM.By SEM, increased numbers of resting T cells were detected along the overlapping processesof IFN-y-treated EC. The results suggest that the changes in the organization of themonolayers induced by IFN-y may further facilitate lymphocyte migration across these EC.Previous studies on HUVEC (23, 134), rat retinal endothelium (156), and rat (229) and mousebrain EC (135) have also demonstrated IFN-y-mediated increase in lymphocyte-endothelialadhesion. Moreover, it has been shown that treatment of rat brain endothelium withcycloheximide, an inhibitor of protein synthesis, inhibits the IFN-y-mediated increase inadhesion, implying the requirement for new protein synthesis (154). It is notable that theoptimal concentration of IFN-y required for maximal adhesive response varies with differentspecies of EC. Our results, therefore, provide additional evidence for the existence ofheterogeneity among EC of different organs or species with regard to cytokine responses.The adhesion of lymphocytes to mouse and rat brain EC has also been reported to increasewith the length of IFN-y treatment (from 4 hours to 2 days) when compared with the controls(135, 154), suggesting that different adhesion molecules with different kinetics of inductionmay participate in the lymphocyte-EC adhesive mechanisms over different time periods.75HBMEC treated for 3 days with IFN-j3 (2,000 U/mi) showed no increase inlymphocyte adhesion as compared with the baseline of adhesion in untreated control.Moreover, IFN- actually suppressed the increase in adhesion induced by IFN-y when brainEC were treated with a combination of IFN-y and IFN-f3. This observation indicates apotentially important role of IFN-f in downregulating immune responses mediated, at leastpartly, by IFN-y.The mAb blocking studies indicate that de novo expression of class II MHC Ag byHBMEC is largely responsible for the increased T lymphocyte-EC adhesion. Similar resultshave been reported in mouse brain EC (135). These investigators further confirmed theirobservations by transfecting a murine lung EC line with cDNA for the class II MHCmolecules in order to demonstrate the role of Ia Ag in lymphocyte-EC adhesion. The adhesiverole of EC is totally distinct from any antigen presenting function, as the lymphocytes are non-activated and there is no antigen present in either system (135). It is notable that completesuppression of IFN-y-enhanced adhesion by anti-human HLA-DR antibody was not achieved,indicating that other mechanisms, not related to DR antigens, operate during lymphocyte-ECbinding. MAb blocking studies in HUVEC have also reported similar observations (134).Working with retinal capillary EC, Liversidge et al. (230) have suggested that, if severaladhesion pathways are available for cellular interactions, then mAbs blocking one pathwaymay be ineffective in completely reducing the number of cells bound, since alternative ligandswould be utilized. MAb blocking studies with anti-human IFN-y indicate that the increase inadhesion observed with resting T cells is specifically mediated by IFN-y.76The results of the present study provide evidence that IFN-3, used at a concentrationknown to significantly suppress the IFN-y-induced Ia Ag expression (2,000 U/mi), candownregulate the increase in adhesion mediated by IFN-y. Taken with the IFN-13 effects onthe IFN-y induced Ia Ag expression, these results suggest that the IFN-13 suppression of theIFN-y-mediated increase in adhesion may be largely due to the downregulation of Class IIMHC molecules by IFN-f3.The role of class II MHC molecules in antigen presentation has been well documented(137); however, the role of class II MHC molecules in lymphocyte-EC adhesion remainscontroversial. Curtis (231) was the first to suggest that class II MHC molecules may functionin the adhesion of non-activated lymphocytes to endothelium and, together with Rooney (232),they state that these molecules may also participate in contact inhibition between epithelialcells, a process which partially involves cell adhesion. Doyle and Strominger report that Blymphocytes expressing class II MHC molecules could bind to CD4 transfected fibroblasts invitro and speculate that the interaction between these two molecules would cause cell-celladhesion independently of antigen presentation (147). They also suggest that at morephysiological levels of expression, it is possible that CD4 molecules and class II antigens helpto mediate low affinity, transient interactions among lymphocytes, and together with otherspecific and accessory adhesion molecules, functional cell-cell interactions can take place(147). Studies with HUVEC systems have also shown that HLA-DR molecules play animportant role in IFN-y-mediated HUVEC-lymphocyte adhesion (134). Furthermore, mAbblocking experiments implicate CD4-class II MHC interaction in IFN-y-induced endothelial77lymphocyte adhesion (145, 233). Interestingly, there is no significant difference in theadhesion between autologous and allogeneic assays which suggests that the interaction is notsimply alloreactive, but may be part of a physiological mechanism for the adhesion, migrationand accumulation of lymphocytes at sites of chronic inflammation (145). It has also beendemonstrated that experimental allergic encephalomyelitis (EAE), a model disease for MS invivo, can be prevented by administration of monoclonal antibodies to class II molecules(234). Alteration in the homing of lymphocytes to the brain in EAE has been implicated as thepossible mechanism for the prevention of disease development in anti-Ia antibody treatment(235). Finally, Ia Ag expression on EC in EAE has been shown to precede lymphocyteinfiltration (66). Taken together, these results suggest an important role for class II MHCmolecules in the interactions between lymphocytes and cerebral EC and in the development ofthe disease process.The adhesion of circulating lymphocytes to brain EC in vivo takes place in a dynamicrather than in a static system. Furthermore, the adhesive capacity of Ia Ag in situ would be ofa much lower affinity than observed in vitro, since the concentration of IFN-y and thesubsequent level of Ia Ag expression on individual EC and its distribution would probably belower. Nevertheless, if under the influence of focally increased concentrations of IFN-y in thecerebral microvessel microenvironment, a minimal level of initial adhesion could take placevia class II MHC molecules, followed by antigen presentation to specific T cells, leading tothe production of more cytokines including IFN-y, class II MHC expression can then beelevated and subsequently, enhance lymphocyte-EC adhesion. Working with HUVEC,78Masuyama and his coworkers (134) have suggested that T cell recognition of HLA-DRmolecules may be the signal for the initiation of subsequent adhesive processes in whichcomplementary adhesion surface molecules become engaged. McCarron et a!. (236) havefurther speculated that the CNS-immune cell interactions may be responsible for localizedalterations in the BBB permeability, resulting in the subsequent influx of non-specificinflammatory cells. In addition, the IFN-y-induced changes in EC morphology and monolayerpermeability, observed in our studies, may further facilitate the movement of lymphocytesacross the BBB.4.4.3 Adhesion of activated lymphocytes to untreated, IFN-y and/or IFN- treated HBMECLymphocyte activation results in great increase in adhesion of activated T cells tountreated HBMEC, indicating that the activation status of the lymphocytes plays an importantrole in lymphocyte-EC adhesion. The adhesive interaction is further augmented by treatingthe EC with IFN-y. Studies with HUVEC (150), human (230) and rat retinal EC (156), rataortic and brain microvascular endothelia (154, 155) and activated lymphocytes have alsoreported significant increase in lymphocyte-EC adhesion. These studies have shown that theadhesion between lymphocytes and EC is dependent on the state of cell activation: maximallevel of adhesion occurs when activated T cells interact with cytokine-stimulated EC.Scanning EM studies further confirm the light microscopic observations. Activatedlymphocytes preferentially adhere at the borders between adjacent and overlapping EC.MAb blocking studies with anti-human IFN-y indicate that the enhanced adhesion of activated79lymphocytes to IFN-y-treated EC is specifically mediated by IFN-y. It has been reported thatirrespective of the state of cell activation, the level of lymphocyte adhesion to CNS-derivedendothelium is generally lower than that reported for extracerebral large and small vessel EC(154 - 156, 237, 238). Subsequently, this low level of adhesion may account for the limitedlymphocyte traffic through the CNS of normal healthy individuals. Furthermore, it is notablethat irrespective of the mode of lymphocyte activation (150, 154 - 158, 230) (i.e. ConA,Phorbol ester, Phytohaemagglutinin, anti-CD3 antibody), adhesion of activated lymphocytes toEC is significantly upregulated in comparison to that of resting T cells.Significant increase in adhesion of anti-CD3 activated T cells to intercellular adhesionmolecule-i (ICAM-1) substrates has been previously reported (159). ICAM-i is a member ofthe immunoglobulin gene superfamily; it is expressed constitutively by HBMEC and can beupregulated by various cytokines such as IFN-y, TNF-cx and IL-13 (239). Its counter receptoris the lymphocyte function-associated antigen-i (LFA-i) which belongs to the integrin family;LFA-1 is expressed on T lymphocytes, not EC, and comprised of heterodimeric, divalentcation-dependent adhesion molecules (162). MAb blocking experiments directed againstICAM-i and the a and f3 subunits of LFA-1 molecules completely block the anti-CD3stimulated adhesion to purified ICAM-1 (159). Since there is no significant change in LFA-iexpression by T cells treated with anti-CD3 antibody versus resting lymphocytes, the authorsconclude that the stimulated increase in T-cell adhesion seems to be due mainly to an increasein LFA-i avidity (159). Fluorescence-activated cell sorter (FACS) studies with mAbs directedagainst both a and f3 subunits of LFA-1 molecules on the adhesion of resting or stimulated T80cells to HUVEC have also shown that the increased adhesion of stimulated lymphocytes to ECis possibly due to the altered function, not increase in expression, of the LFA-1 molecules(240).The binding of anti-CD3 antibodies to T cells triggers phosphatidylinositol turnoverand elevates cytoplasmic Ca2+ (220, 221). It has been demonstrated in rat cerebral and aorticendothelia that the removal of Ca2+ from the media can effectively inhibit lymphocyte-ECadhesion (155). Interestingly, earlier work indicates that aCa2+dependent epitope on LFA-1,termed L16, is a prerequisite for LFA-1 to mediate cell adhesion and may distinguish restinglymphocytes from activated lymphocytes (241). Subsequently, it has been shown that L16epitope is expressed when the lymphocytes are stimulated by phorbol ester or T cellreceptor/CD3 (TCR/CD3) complex. It has been suggested that there are possibly 3 distinctforms of LFA-1: a) an inactive form, partially exposed epitope, is present on resting T cells,b) an intermediate one can be found on mature or previously activated cells, and finally, c) anactive epitope, capable of high affinity ligand binding, can be demonstrated after TCR/CD3 orphorbol ester activation (241). Studies with human retinal pigment epithelial cells havedemonstrated that these cells constitutively express high levels of ICAM-1, and thesemolecules are functional in binding activated T lymphocytes but not resting T cells (230).LFA-1-dependent pathway has been implicated in the increased adhesion of activatedlymphocytes to vascular endothelium such as high EC (HEC) and HUVEC (151, 158, 240).Therefore, these results suggest that LFA-1/ICAM-1 interactions may also play a significantrole in the increased binding between anti-CD3 activated T cells and untreated HBMEC81observed in this study.Total inhibition of binding by monoclonal antibodies directed against LFA-i moleculeshas not been observed (151, 240). Approximately 20% to 40% of adhesion between severalT cell leukemia cell lines and HUVEC can not be suppressed by blocking with anti-LFA-1and anti-VLA-4 antibodies (242). In fact, these observations are in accordance with otherstudies indicating that there are at least three to four other mechanisms or pathwayscontrolling lymphocyte-EC adhesion (243, 244). Studies on adhesion of activated T cellleukemia cell lines to HUVEC have suggested that the LFA-1 adhesive mechanism dominatesthe interaction; however, very late antigen-4 (VLA-4) is used by T lymphocytes to bind ECwhen LFA-1 is not expressed or not functional to mediate adhesion (242), pointing to aselective use of different adhesion receptors by the T cells. In contrast, resting lymphocytesuse both LFA-1 and VLA-4 adhesion pathways (242). Like LFA-1, VLA-4 also belongs tothe integrin family of cell surface heterodimers; it is expressed by lymphocytes and caninteract with vascular cell adhesion molecule-i (VCAM-1) (245). VCAM-1 is a member ofthe immunoglobulin superfamily like ICAM-1 (246); however, VCAM-1 is not expressed byperipheral blood lymphocytes (247, 248). Primary cultures of HBMEC have been shown toexpress low levels of VCAM-1 constitutively (249). Blocking experiments with niAbsdirected against LFA-1, ICAM-1, VLA-4, and VCAM-i molecules on the adhesion of restingT cells to untreated/activated HUVEC have demonstrated that the VLA-4/VCAM-1 adhesivemechanism is largely responsible for the adhesion of T lymphocytes to cytokine-treated EC,while LFA-i/ICAM-i pathway mediates much of the binding of T cells to unstimulated EC.82Furthermore, the binding of activated lymphocytes is not blocked by antibodies to VLA-4or VCAM-1, irrespective of the activation status of the EC. However, antibodies to LFA-1 orICAM-1 can modestly inhibit the adhesion of activated T cells to HUVEC. It is notable thatcomplete inhibition of T ceil-EC interactions by these antibodies has never been detected(152). Other studies with high EC and rat cerebral EC have also indicated that VLA-4/VCAM-1 mechanisms do participate in lymphocyte-EC adhesion (158, 250). Finally, the participationof other pathways besides the LFA-1/ICAM-1 system in the T cell - EC binding has beenexamined using LFA-1-deficient T cell clones generated from a patient with leukocyteadhesion deficiency. The results of these experiments confirm previous observations by Dustinet a!. (243) and Shimizu et al. (244) stating that there are other pathways mediatinglymphocyte-EC adhesion in addition to LFA-1/ICAM-.1 mechanism. In fact, VLA-4/VCAM-1represent the alternate receptor/ligand pairs which mediate the binding of LFA-1-deficient Tcells to HUVEC (153).Treatment of HBMEC with IFN-y and IFN-3 further modulates the anti-CD3stimulated T lymphocyte-EC interactions. Activation of HBMEC with an optimalconcentration of IFN-y (150 U/mi), known to induce maximal Ia Ag expression and markedlyupregulate the binding of resting T cells to EC, results in a 2 fold increase in adhesion ofactivated lymphocytes to cytokine treated brain EC. Further enhancement of lymphocyte-ECinteractions when both systems are activated has been reported with rat retinal (156), aortic(155) and brain endothelia (154), and also HUVEC (150). Interestingly, the level of adhesionbetween activated T cells and IFN-f3 treated HBMEC is comparable to that of untreated EC,83indicating that IFN- alone has no influence on adhesion. Moreover, a combined treatment ofhuman brain EC with IFN-y and IFN-f3 in the present studies, demonstrated that IFN-j3actually inhibits the IF’N-y-mediated binding. These observations further support the negativeregulatory role of IFN-13 on changes induced by IFN-y. The IFN-y-mediated increase inadhesion of activated lymphocytes to HBMEC can also be suppressed by mAbs directedagainst human HLA-DR. In addition, blocking studies with anti-human IFN-y indicate thatIFN-y is responsible for the increased adhesion between activated T cells and IFN-y-treatedHBMEC most likely through induction of class II molecule expression by HBMEC. Theseresults indicate that class II MHC molecules play a central role in mediating the increasedadhesion of activated T cells to IFN-y-treated HBMEC. The fact that IFN-3 suppresses theIFN-y-induced Ia Ag expression on HBMEC also suggests that IFN-3 downregulates the IFNy-mediated binding via the Ia Ag mechanism, which is further supported by the resultsobtained from mAb blocking experiments. In this respect, IFN- downmodulates the IFN-yinduced increased binding between T lymphocytes and HBMEC regardless of the activationstatus of the lymphocytes. Studies on lymphocyte-EC binding with mAbs directed against aand 1EJ subunits of LFA-1 molecules have demonstrated significant inhibition of bindingbetween resting or stimulated T cells to untreated HUVEC; however, these antibodies have noinfluence on the adhesion of lymphocytes to cytokine-treated EC (240). These authors havesuggested that the mechanism of binding of T cells to unstimulated endothelia differs fromthat to stimulated endothelia, and the latter appears to be independent of LFA-1. Male et al.have also indicated in their work with rat cerebral EC that the control of basal binding and84binding to activated endothelia are regulated by different mechanisms. This system wouldallow brain endothelium to have low basal binding to minimize lymphocyte traffic into thebrain normally, while permitting rapid increase in traffic if the cerebral EC are stimulatedappropriately (155).4.5 MIGRATION OF RESTING and ANTI-CD3 STIMULATED LYMPHOCYTESACROSS UNTREATED, and CYTOKINE TREATED HBMECThe fourth specific aim of this thesis was to determine whether migration of resting andanti-CD3 stimulated lymphocytes across cerebral endothelium can be influenced bytreatment of HBMEC with interferons y and/or Migration of resting lymphocytes across untreated, IFN-y and/or IFN- treatedHBMEC monolayersIf the low basal level of migration of resting T lymphocytes across untreated HBMECmonolayers in vitro reflects the limited lymphocyte traffic into the CNS in vivo, it wouldcontribute to the relative immunological isolation of the brain under normal physiologicalconditions. As observed with adhesion, migration of resting lymphocytes was also regulatedby cytokine treatment of the cerebral endothelium in this study. Treatment of HBMEC withIFN-y results in a 3 fold increase in migration compared to that of untreated EC, suggestingthat IFN-y enhances the migration of T cells across the EC monolayers possibly by a direct85action on the endothelium. Light microscopic and TEM studies demonstrate large numbers ofmigrated lymphocytes underneath the monolayers of HBMEC previously treated with IFN-y,while fewer lymphocytes crossed untreated EC. Migration of resting T cells is not associatedwith damage to the integrity of the monolayers in either untreated or IFN-y treated monolayer.Lymphocyte migration usually takes place between adjacent EC. Migration through thecytoplasm of EC is a less common route of migration across the monolayers. IFN-y has beenpreviously reported to significantly upregulate the migration of lymphocytes across HUVEC(24) and rat retinal EC (170). Oppenheimer-Marks and Ziff (24) observed that the augmentingeffect of IFN-y on lymphocyte transendothelial migration is not dependent on the presence ofan exogenously added chemotactic factor below the EC monolayer. Using passaged culturesof rat cerebral EC, Male et al. (251) demonstrated that the activation status of the endotheliumhas no influence on the migration of activated T cells, however, they observed that the brain-specific surface phenotype of the cultured cells deteriorated after the first passage (251). Ourstudies, therefore, indicate that IFN-y upmodulates both adhesion and migration of resting Tcells across the cerebral endothelial banier. Whether the level of lymphocyte migration in thebrain is a reflection of the level of lymphocyte-EC adhesion or the two events arepathophysiologically different and under the control of distinct influences by IFNs and/orother cytokines remains to be further investigated.The level of migration of resting T cells across IFN-13 treated HBMEC is comparableto that of untreated endothelia, indicating that IFN-f3 has no direct effect on lymphocytemigration. Treatment of HBMEC with a combination of IFN-y and IFN-j3 results in significant86suppression of the IFN-y-mediated increased migration which further supports thedownregulatory role of IFN- Migration of anti-CD3 stimulated lymphocytes across untreated, IFN-y and/or IFN-treated HBMEC monolayersIn this study, nonspecific stimulation of T lymphocytes with anti-CD3 generates asignificant increase in migration across untreated monolayers of HBMEC as compared to themigration of resting lymphocytes across untreated EC. In accordance with the observations onresting T cells, the level of stimulated lymphocyte traffic in the CNS most likely reflects thelevel of adhesion of activated T cells to HBMEC: the rate of increase of lymphocyte migrationis comparable to that of adhesion. It has been previously demonstrated that lymphocyteactivation induces three to four fold increase in migration across HUVEC in vitro whencompared to resting T cells (151). MAb blocking studies directed against various adhesionmolecules and their ligands including LFA-1, ICAM-1, VLA-4 and VCAM-1 have reportedthat the LFA-1/ICAM-1 interaction plays an important role in transendothelial migration ofactivated lymphocytes through HUVEC (151, 152). In contrast, VCAM-1 has thus far notbeen found to be utilized during the migration process, regardless of the activation status of theT cells or EC (152). Migration of activated T cells is not entirely blocked by mAbs to LFA-1and ICAM-1, indicating that additional surface molecules are required for transendothelialmigration (151, 152). Furthermore, studies on patients with LFA-1 deficient leukocytes haveshown the presence of lymphocytes in inflammatory lesions of these patients, indicating that87lymphocyte migration into inflammatory foci is not entirely dependent upon the expression ofLFA-1 molecules (252). The fact that migration of activated lymphocytes across untreatedHBMEC is significantly greater than that of resting T cells through IFN-y treated cultures,indicates that antigen-nonspecific stimulation of lymphocytes plays a critical role in theiremigration from the blood into the perivascular tissue. In vivo observations in the rat havereported that activated lymphocytes can rapidly enter into the CNS tissue once they areintroduced into the circulation, irrespective of antigen specificity, MHC compatibility, T-cellphenotype or T-cell receptor gene usage (161). Furthermore, it has been shown that activatedlymphocytes can increase the levels of the enzyme heparan sulfate endoglycosidase (253), andsubstances that inhibit this enzymatic activity can prevent the development of EAE which isdependent upon T-cell entry into the CNS (254). These features may play some role in themigration of activated T cells, however, other mechanisms may also participate in thismigratory process.Immunohistochemical studies on the migration of T cells across HUVEC cultureshave shown the presence of ICAM-1 along the intercellular contacts between EC that are incontact with the migrating lymphocytes as well as on the basal membrane of the EC. Thepresence of ICAM-1 molecules at these sites as well as at sites of contact between the ECmembrane and the leading edge of migrating T cells suggests a critical role of ICAM-1 intransendothelial migration of T cells. The authors conclude that, as the T cells migrate acrossthe EC layer, migration proceeds by the successive formation of adhesive bonds betweenreceptors on T lymphocytes and their counter-receptors on EC, like a “zipper’t mechanism88(152). Recent studies have demonstrated that primary cultures of HBMEC express relativelyhigh levels of ICAM-1 (up to 40%) constitutively (239). As discussed previously, activationof lymphocytes may lead to an increase in avidity of LFA-1 molecules present on the T cells(159) which can further enhance the interaction of LFA-1 to its counter receptor, ICAM-1.MAb blocking studies directed against LFA-1 molecules have found that migration of restingT cells across HUVEC is not inhibited, while migration of activated T lymphocytes can besuppressed to a comparable level with resting T cells (151). Subsequently, the LFA-1/ICAM-1 dependent pathway may play a central role in the marked increase in migration of anti-CD3stimulated lymphocytes across untreated HBMEC.Studies on platelet/EC adhesion molecule 1 (PECAM-1) have recently indicated thatthis molecule may also play an important role in transendothelial migration of leukocytes(255). PECAM-1 is a member of the immunoglobulin gene superfamily (256), appears to beconcentrated at the junctions between EC (257) and is expressed on the surface of monocytes,neutrophils, and a small subset of lymphocytes (258 - 260). Muller et al. have suggestedseveral possible roles of PECAM-1 in transendothelial migration of leukocytes. The mostobvious role involves PECAM-1 as a direct adhesion molecule binding the leukocyte tightly tothe HUVEC during its passage through the junctions (255), since PECAM-1 has been shownto be concentrated at the intercellular junctions, with approximately 15% exposed to the apicalsurface (257). These authors suggest that an apical-basal gradient of PECAM-1 may existthrough the HUVEC junction which can act similarly to a surface-bound chemotactic gradientto produce directed migration of leukocytes through the junction. Another possible role is that89PECAM-1 may be ligated on the surface of leukocytes, which can then activate CD11/CD18binding activity, and this mechanism could apply as well if an apical-basal gradient ofPECAM-1 exists (255). Finally, induction of PECAM-1 has been reported on activated T cells(258), thus suggesting a possible role of this adhesion molecule in mediating migration ofactivated lymphocytes across untreated monolayers of HBMEC in addition to the LFA1/ICAM-1 mechanism.Migration of activated T lymphocytes across EC monolayers is further enhanced byIFN-y treatment of HBMEC. Similar observations have also been reported for the migration ofresting or activated lymphocytes across untreated and cytokine treated HUVEC (24, 151). Incontrast, treatment of rat retinal EC with IFN-y is associated with a small, but not significant,increase in the level of activated T-cell line lymphocyte migration (170). In addition, Male eta!. have reported that migration of activated lymphocytes across rat brain endothelia does notappear to depend on the activation state of the EC (251). A possible explanation for theseresults could be the relatively low concentrations of IFN-y used. Alternatively, inherentdifferences in the culture systems used could account for these discrepancies.Transmission EM studies show significant numbers of activated lymphocytesmigrating across untreated as well as IFN-y treated HBMEC. Activated T cells displayincreased size and altered appearance. Finger-like projections decorate the cytoplasmicmembrane and significant numbers of mitochondria and vacuoles occupy the cell cytoplasm.Migration of both resting and anti-CD3 stimulated lymphocytes across untreated and IFN-ytreated HBMEC is not associated with disruption of the monolayers. The integrity of the90monolayers is reestablished once lymphocyte migration is completed. Similarly, migration ofbovine peripheral blood lymphocytes across the endothelium of pulmonary artery intimalexplants has been shown to cause no damage to the continuity of the vascular endothelium(261).The optimal concentration of IFN-y used for the adhesion and migration assays hasbeen shown to induce maximal Ia Ag expression on HBMEC in primary culture (127). Weobserved no significant change in the degree of activated T cell migration across IFN-13treated EC as compared to controls, indicating that IFN-3 has no direct effect on the migrationprocess. However, the IFN-y-mediated increase in migration is downregulated by IFN-f3, sincemigration of activated lymphocytes across EC coincubated with IFN-y and IFN-3 iscomparable to that of controls. Blocking experiments with mAbs against human HLA-DR inIFN-y treated HBMEC show comparable levels of decrease in migration to those obtained bytreating HBMEC with both cytokines.Taken together, the results of our studies on the migration of resting andnonspecifically stimulated T lymphocytes across untreated and cytokine treated HBMECmonolayers, indicate that induction of class II molecules on the surface of HBMEC by IFN-yis, at least in part, responsible for the increased migration of T cells across the monolayers.IFN-f3 has no direct effect on lymphocyte-EC binding, but downregulates the IFN-y-mediatedincrease in transendothelial migration most likely through downregulation of the IFN-yinduced de novo expression of class II MHC molecules by HBMEC.914.6 EFFECTS OF IFN-y ON THE STORAGE AND RELEASE OF FVIIIR:AgFROM HBMEC IN PRIMARY CULTUREThe fifth and final specific aim of this thesis was to investigate the effects of IFN-y on thestorage and release of FVIIIR:Ag following its immunocytochemical localization inprimary cultures of HBMEC.4.6.1 Immunocytochemical localization of FVIIIR:Ag in HBMECHuman brain microvessel EC in primary culture synthesize FVIIIR:Ag as indicated bytheir positive, granular, perinuclear staining for FVIIIR:Ag with the immunoperoxidasetechnique. By immunoelectron microscopy FVIIIR:Ag is localized within cytoplasmic vesiclesclosely associated with the rough endoplasmic reticulum and Golgi apparatus in theperinuclear region. Treatment of EC with Ca2+ ionophore A23 187 results in marked reductionin labeled vesicles, while preincubation with IFN-y leads to increase of intracellular poois ofFVIIIR:Ag.EC lining large vessels and arterioles synthesize and secrete FVIIIR:Ag and store thenewly synthesized glycoprotein within cytoplasmic organelles unique to these cells, known asWeibel-Palade bodies (174 - 176). These rod-shaped structures are absent in primary culturesof microvessel EC derived from rat (179 - 180), mouse (178) and bovine cerebral cortex (262)and bovine retina (263), but have been reported to be present in EC derived from rat andbovine brain white matter (264). Weibel-Palade bodies are extremely rare or absent in normal92human cerebral capillaries (265, 266). They have been observed in the orbital cortex of normalaged humans (267) and with increased frequency in certain brain tumors (266, 268). HBMECin primary culture are similarly devoid of Weibel-Palade bodies. The perinuclear, granularstaining for FVIIIR:Ag with the immunoperoxidase technique corresponds to variably dilatedvesicular profiles within which deposits of colloidal gold were observed ultrastructurally. Thesingle limiting membrane of these vesicles is not decorated with ribosomes and therefore, it isunlikely that they represent dilated cisternae of rough endoplasmic reticulum. Their constantpresence near the Golgi apparatus and the rough endoplasmic reticulum suggests that theimmunolabeled vesicles belong to the polymorphous vacuoles that form part of the transmostGolgi section (269). These trans Golgi elements have been found to be part of the pathway ofnewly synthesized molecules (270). It is, therefore, likely that, following synthesis in theendoplasmic reticulum and extensive modification in the Golgi apparatus (271), the newlysynthesized FVIIIR:Ag is transported to the trans Golgi polymorphous vesicles where it isconcentrated. In the absence of Weibel-Palade bodies in cerebral microvessel EC, thesevesicular bodies most likely represent sites of short-term storage of FVIIIR:Ag prior to release.Previous in vivo studies on the localization of FVIIIR:Ag in vascular endothelium of normalhuman extracerebral tissues and one capillary hemangioma by immunoelectron microscopydemonstrated immunolabeling of endoplasmic reticulum and cytoplasmic vesicles andvacuoles in addition to Weibel-Palade bodies (176). These vesicular profiles strongly resemblethe ones observed in the present study. A similar localization of FVIIIR:Ag within cytoplasmicvesicles has been reported in EC lining the saphenous vein (172).934.6.2 Effects of IFN-y on the storage and/or release of FVIIIR:Ag from HBMECLarge vessel EC secrete the newly formed FVIIIR:Ag via two pathways (272, 273).The regulated pathway involves release of the large multimeric forms of the glycoprotein fromthe specific storage organelles, the Weibel-Palade bodies. In vitro studies on the secretion ofvon Willebrand factor by umbilical vein EC indicate that this pathway is highly polarized anddependent upon intact microtubular system, since microtubule-depolymerizing agents inhibitthe regulated release (274). The majority of FVIIIR:Ag synthesized by EC is secretedconstitutively in the form of small multimers. In contrast to the regulated pathway, constitutiverelease is not affected by microtubule-depolymerizing agents (275) and is not polarized. SinceHBMEC do not store FVIIIR:Ag in Weibel-Palade bodies, it is quite possible that they secretethe newly synthesized protein through the constitutive pathway only, with the Golgi-associatedcytoplasmic vesicles serving as temporary storage pools following multimerization and priorto release.A variety of stimuli can lead to increased release of FVIIIR:Ag from EC in vitro. Mostof these factors stimulate the regulated pathway of secretion. Thus, treatment of ECmonolayers with calcium ionophore A23 187, thrombin or phorbol-myristate-acetate results inrelease of the large multimeric forms of FVIIIR:Ag and a simultaneous disappearance ofWeibel Palade bodies from EC (272 - 274, 276) in association with a rise in the concentrationof intracellular calcium. The effect of calcium ionophore was inhibited by EGTA in a dosedependent manner (276). The basal secretion was apparently not affected by these treatments.In the present study short preincubation of HBMEC with calcium ionophore led to rapid94reduction in the number of labeled cytoplasmic vesicles. In contrast, addition of the calciumchelating agent EGTA to the culture media resulted in slight increase in immunostainedvacuoles, which was not statistically significant when compared to untreated cells. Loss ofstaining following calcium ionophore treatment may represent rapid release of FVIIIR:Agfrom intracellular pools, although other mechanisms, such as antigen degradation followingionophore - mediated protease activation, cannot be ruled out. These findings indicate that atleast some of the factors that stimulate the regulated pathway in large vessel EC, similarlyinfluence the release of FVIIIR:Ag from microvessel endothelium in the absence of WeibelPalade bodies. Whether the FVIIIR:Ag secreted by cerebral small vessel endothelium is in theform of small or large multimers, is presently unknown.Incubation of HBMEC with IFN-y for 24 hours resulted in significant increase in thenumber of immunostained vesicles suggesting that IFN-y interferes with the release and/orstorage of FVIIIR:Ag. Recent studies on the effect of cytokines on the release of vonWillebrand factor indicate that IFN-y decreases the constitutive and regulated release fromcultured HUVEC reversibly and in a time and dose-dependent manner (182). Although theexact mechanism of action is not presently known, it is possible that IFN-y exerts its effect bymodifying the concentration of intracellular calcium. It has been recently demonstrated thatIFN-y can activate the calcium-dependent pathway through activation of phospholipase C andinduce, in addition, a significant outflux of calcium ions from EC (277). Inhibition ofFVIIIR:Ag release by IFN-y may be important considering its pivotal role as mediator of thelocalized immune response in autoimmune diseases of the central nervous system.95CONCLUSIONS5.1 SUMMARY AND CONCLUSIONSThe main objective of this thesis was to examine the effects of IFN-y and IFN-13 on IaAg expression, cell proliferation, and alteration of the morphology and permeability propertiesof HBMEC using an in vitro model of the human BBB. In addition, the adhesion andmigration of resting and anti-CD3 stimulated T cells across untreated and cytokine-treatedHBMEC was studied. Finally, the effects of IFN-y on the storage and release of FVIIIR:Ag byHBMEC was determined.The working hypothesis of this thesis was that in chronic inflammation, some activatedT lymphocytes will release inflammatory cytokines including IFN-y. This cytokine can theninduce the local brain endothelia to express Ta Ag on the cell surface, to alter their morphologyand increase their permeability to macromolecules. These changes will facilitate the adhesionand migration of resting lymphocytes across the cerebral endothelial barrier. Lymphocyteactivation will further augment adhesion and migration. Finally, IFN-y by inhibiting therelease of FVIIIR:Ag from HBMEC will contribute to the maintenance of blood fluidityduring the immune reaction.In this study, IFN-y induces de novo expression of Ia Ag on HBMEC in a time anddose-dependent fashion. Primary cultures of HBMEC do not express Ia Ag constitutively.The expression of Ia molecules on HBMEC can be detected as early as 12 hours followingincubation with IFN-y and reaches plateau levels by 48 hours. Surface labeling for Ta Ag is96maximal with 100 to 200 U/mi IFN-y and minimal with 10 U/mi. In contrast, treatment ofHBMEC with IFN-13 has no influence on Ia Ag expression. Moreover, incubation of HBMECwith a combination of IFN-y and IFN-f3 results in downreguiation of Ia Ag expression. IFNsuppresses the IFN-y.-induced expression in a dose-dependent manner, however, completeinhibition was not detected. Kinetic studies on the effects of IFN-y and IFN-3 on Ia Agexpression indicate that administration of IFN-j3 prior to or simultaneously with IFN-ytreatment generates the most significant downregulation of IFN-’-induced Ia Ag expression.These observations may partly explain the results of recent therapeutic trials with IFN-13 inMS.Treatment of HBMEC with IFN-y results in changes in cell shape and organization ofthe EC monolayers. The IFN-y-treated endothelia acquire a spindle-like shape and longattenuated processes. Prominent overlapping and ill-defined whorls are unique features ofIFN-y-treated EC monolayer. The IFN-y-induced phenotypic alterations on HBMEC areinhibited when EC are incubated simultaneously with IFN-y and IFN-3; the monolayersresume their highly organized growth pattern with prominent marginal folds in the areas ofcell to cell contact. The morphological changes are associated with increased permeability ofconfluent monolayers to horseradish peroxidase as compared with untreated cultures. Thenumber of HRP labeled vesicles was not increased in IFN-y treated EC as compared tountreated EC.Lymphocyte-EC adhesion is significantly upregulated when HBMEC are pretreatedwith IFN-y, while IFN-f3 inhibits the IFN-y-enhanced adhesion when the brain endothelia are97incubated with a combination of IFN-y and IFN-3. IFN-3 alone has no effect on lymphocyteEC interactions since the level of adhesion is comparable to that of untreated EC. Nonspecificactivation of T lymphocyte causes a significant increase in lymphocyte-EC adhesion; in fact,the level of adhesion of activated lymphocytes to untreated EC is greater than that of resting Tcells to IFN-y treated EC, suggesting that lymphocyte activation plays an important role in Tcell-EC adhesion. Similar to the responses obtained with resting lymphocytes, IFN- inhibitsthe IFN-y-enhanced adhesion of activated T cells when EC are treated simultaneously withIFN-y and IFN-3. However, IFN- treatment alone has no effect on the adhesion of activatedT lymphocytes to EC. MAb blocking studies against IFN-y and human HLA-DR moleculesindicate that the enhanced binding is specifically induced by IFN-y and HLA-DR moleculesplay a central role in the IFN-y upregulated adhesion.Migration of resting lymphocytes is markedly augmented when HBMEC are treatedwith IFN-y, but not IFN-. Treatment of cerebral EC with a combination of IFN-y and IFN-13significantly downmodulates the IFN-y-mediated migration. Activation of lymphocytes isassociated with a dramatic increase in migration across untreated EC, and the level ofmigration is greater than that of resting T cells through IFN-y treated EC. IFN-y, but not IFN, further augments the migration of activated T cells. The IFN-y-enhanced migration issuppressed by IFN-13 treatment. Blocking studies with mAbs against HLA-DR moleculesindicate that Ta Ag plays a central role in the IFN-y mediated migration of both resting andanti-CD3 stimulated lymphocytes.Finally, it has been determined that IFN-y suppresses the release of FVIIIR:Ag from98HBMEC which suggests that IFN-y may also play a role in maintaining blood fluidity duringthe immune reaction.In conclusion, the results of these studies demonstrate the critical function of IFN-y inupregulating the immune response which plays an essential role in the host defensemechanism. Indeed, autoimmune disorders of the CNS such as MS may arise as the results ofunwanted inflammatory or immunological responses. Subsequently, the ability of cytokinessuch as IFN- to reduce or suppress the IFN-y-enhanced reactions may explain theirtherapeutic effect. The facts that IFN- is able to inhibit the IFN-y-induced Ia Ag expressionand to suppress the IFN-y-mediated increase in lymphocyte-EC adhesion and migrationindicate that Ia Ag plays a central role in these immunological responses. This statement isfurther supported by results obtained from mAb blocking studies directed against humanHLA-DR. Therefore, the results of this thesis demonstrate the important role of HBMEC inCNS inflammation and enhance our understanding of some of the factors involved in therecruitment of lymphocytes into chronic inflammatory sites in the CNS.5.2 FUTURE PROSPECTSThe results obtained from this study indicate that class II MHC participates in the IFNy-mediated increase in lymphocyte-HBMEC adhesion and migration. An avenue forimmediate future research would be to determine the role of HBMEC as antigen presentingcells in CNS inflammation using this in vitro BBB model. HBMEC can be induced to expressIa Ag by IFN-y and allowed to endocytose and process myelin basic protein (MBP), a protein99component of the myelin shealth. At the appropriate time, MBP-specific lymphocytes areincubated with these HBMEC, and lymphocyte proliferation can be determined with tritiatedthymidine assay.Since it has been demonstrated in this thesis that antigen-nonspecific stimulation ofperipheral blood lymphocytes plays a central role in markedly upregulating the lymphocyteEC adhesion and migration, another avenue for future research would be to examine themolecular mechanisms that are responsible for augmenting the adhesion and migration ofactivated lymphocytes to HBMEC. The application of mAbs directed against specificadhesion molecules such as ICAM-1, LFA-1, VCAM-1, VLA-4, that are present on thesurface of both lymphocyte and EC will help to determine the molecules responsible forupmodulating the immunological reactions.Finally, a concern that also needs to be addressed in future research in CNSinflammation is the question: What is the significance of peripherally activated Tlymphocytes in the development of immune reactions in the brain?In one version of rat EAR, autoimmune demyelination can be induced by immunization withmyelin basic protein, and systemic injection of a specific monoclonal antibody directed againstmyelin/oligodendrocyte glycoprotein can amplify demyelination. Immunotherapy of thisantibody-induced demyelination is possible with another specific mAb directed to an antigenon activated rat T cells, suggesting an important role of T lymphocyte activation in the diseasedevelopment (169). Studies on the migration of activated lymphocytes across rat brainendothelium lead Male and his coworkers (154) to speculate that lymphocyte activation in the100periphery may lead to increased traffic through the brain. This lymphocyte traffic can haveserious consequences if antigen specific interaction develops between circulating T cells andantigen presenting cells in the CNS. Amplification of the immune reaction can result fromcytokine release and local activation of the brain endothelium, causing further increases incellular migration. As suggested, this scenario is possible in Bordetella pertussis vaccinationin a small proportion of individuals. In accord with the above speculation, in vivo studies ofT-lymphocyte entry into the rat CNS by Hickey and his coworkers (161) have reported thatactivated T cells appear to enter the CNS in a random manner, irrespective of their antigenspecificity, MHC compatibility, T cell phenotype, and T-cell receptor gene usage. Theseauthors also showed that only T-lymphocytes that are able to recognize a specific antigen inthe CNS of the host remain beyond 72 hours in the target organ while non-specificallyactivated T-cells exit the CNS within 1 to 2 days.The above observations raise question(s) about the “immunological privilege” status ofthe CNS because any T-cell that is activated in the tissues of the immune system can possiblygain access to the brain in a random manner once it enters the circulation. Transplantation ofallogeneic tissue into the CNS is well accepted by the recipient; however, the graft is quicklyrejected when the same alloantigen is exposed to the host periphery (278). In concordancewith the results described by Hickey et al. (161), in vivo studies of mice EAE, Cross et al.(279) have shown that 14C-labeled CNS antigen-specific T cells home to the CNS endothelia24 hours prior to and during the initial clinical disease, but these cells always remain withinthe perivascular area. The antigen-specific lymphocytes only represent 1% to 4% of the101inflammatory cells that are present in the brain parenchyma during disease development.These investigators conclude that the inflammatory cells are predominantly of recipientderivation. More recently, Caspi et al. (280) have demonstrated that in experimentalautoimmune uveoretinitis (EAU), a T cell-mediated autoimmune disease in rat serving as amodel for a number of human blinding ocular diseases of a presumed autoimmune nature,only mild or essentially no disease can be induced by the CNS antigen-specific T cell lines inunreconstituted athymic rats. However, the situation can be significantly reversed by infusionof naive cell populations containing immunocompetent T cells. Subsequently, the recruitmentof naive T cells constitutes an amplification mechanism that is central to the expression andpathogenesis of uveitis. The phenomenon of recruitment can magnify the effect of a tinynumber of antigen-specific “pathogenic” T lymphocytes into a destructive inflammation.Consequently, the results that are available hitherto do support the significant role thatperipherally activated, antigen-nonspecific T cells can have in the development of immunereactions in the brain. The ability to suppress these activated lymphocytes, for example, withanti-LFA-1 monoclonal antibodies to block the LFA-1 dependent pathway of adhesion duringan unfavorable inflammation of the CNS may have great therapeutic potential.5.3 SIGNIFICANCE OF THIS THESISPrimary cultures of HBMEC provide a useful in vitro model for investigating theeffects of cytokines on the morphology and function of the cerebral endothelium and on thecomplex processes of lymphocyte adhesion and migration across the cerebral endothelium102barrier. These studies indicate that interferons y and 3 are important mediators of the localizedimmune response within the human CNS and that class II MHC molecules, induced de novoon HBMEC, play a pivotal role in lymphocyte-EC interactions in the CNS.As pointed out by Wekerle at al. (14), the detailed analyses of the cellular andmolecular mechanisms involved in the interaction between T cells and the BBB are oftremendous importance for various reasons: a) Such knowledge will give insight into thedevelopment of CNS disorders with putative autoimmune pathogenesis, e.g. multiple sclerosis.b) The information obtained will provide a better understanding of the mechanisms involve inphysiological immune surveillance, as they are relevant in the prevention and the control ofinfectious diseases within the CNS. c) On the basis of such knowledge, it may be possible todesign novel specific therapies of CNS (autoimmune) disease.103Table 1 Permeability of HBMEC monolayers to HRPNo. of labeled No. of interendothelialcytoplasmic vesicles * tight junctions* *Permeable ImpermeableControl 2.0 ± 1.7 vesicles/cell 24.8 ± 2.7% 75.2 ± 2.7%Experimental 2.4 ± 2.1 vesicles/cell 63.4 ± 5.2% 36.6 ± 5.2%* Numbers represent mean + SD of labeled vesicles in 100 control and 100 IFN-Y treated cellsfrom one experiment. P 5 0.05* * Numbers represent mean ± SD of 400 junctions (200 treated and 200 untreated) from twoexperiments using two different isolates. P < 0.05104Figure 1: Diagram of the double chamber chemotaxis system used to study themigration of lymphocytes across confluent HBMEC monolayers. Thecellagen membrane is a firm membrane, made up of solubilized collagen,that forms the bottom of 14 mm diameter wells. These wells are placedinside larger wells of 24-well plates. Four support feet separated the innerfrom the outer chamber. HBMEC are seeded onto the cellagen membranesand grown inside the inner chamber. Initial attachment of EC to themembranes does not require precoating with fibronectin. This system canbe used to study the interactions between EC and inflammatory cells suchas lymphocytes or polymorphonuclear leukocytes.S(ThCD0.‘C)—CD—‘CDL05C’.)CHC.)CDCii106Figure 2: Primary cultures of HBMEC grown on plastic wells (a) or cellagen membranes(b) and maintained under standard culture conditions form highly organized,confluent, contact inhibiting monolayers composed of elongated endothelialcells. Bars = 20 tm.107F. z108Figure 3: Intense, predominantly perinuclear cytoplasmic staining of cultured HBMEC forFVIIIR:Ag with the immunoperoxidase reaction (a). Lectin binding by HBMECis indicated by their positive immunoperoxidase staining for UEA 1(b). Bars =20 tm.109it,144 4 1.4 0,a’a a SF?. 3110Figure 4: Confluent monolayers of HBMEC (EC) grown on cellagen membrane (C).Elongated cells with focally evident finger-like cytoplasmic projections form acontinuous cell layer firmly attached to the substrate. Bars = 2 tm.Ill.112Figure 5: Primary cultures of HBMEC (EC) cultivated on cellagen membranes (C) (a - d).Intercellular contacts vary in length and complexity. Tight junctional complexes(arrows) with pentalaminar configuration are present in areas of cell to cellcontact. Bars = 0.2 tm.p4Cr’ 00cJ0’‘9[‘‘v’I*3‘V:%9*4*’’1wI117Figure 6: The cytoplasm of cerebral microvessel endothelial cells (EC) contains prominentrough endoplasmic reticulum (small arrows), small to large mitochondria(arrowheads), and a variable number of small and large vesicles (V) injuxtanuclear position. Small amounts of amorphous material are present,otherwise, the vacuoles are clear and bound by a single limiting membrane. Theendoplasmic reticulum is closely associated with the vesicular profiles. EC weregrown on cellagen membranes (a) or plastic wells (b). Weibel-Palade bodies arenot present. C = cellagen membrane; N = nucleus. Bars = 1 tm.£6ct-LI’1F. 61’120Figure 7: Immunogold staining of intact endothelial monolayers for FVIIIR:Ag. (A) Fivenm gold particles form dense aggregates within several cytoplasmic vesicles(large arrowheads). Other vesicles contain scant particles (small arrowheads),whereas still others are not stained. Labeled vesicles are located close to theGolgi apparatus (B) and the endoplasmic reticulum (A). (C) Occasionally, a fewgold particles localize within cisternae of endoplasmic reticulum (arrowhead)next to a labeled vesicle. (D) Staining is absent in control cultures incubated withcarrier buffer instead of primary antibody. N, nucleus; G, Golgi. Bars = 1 tm.I4122Figure 8: Time course of Ia antigen induction on human cerebral endothelium. ConfluentHBMEC cultures were incubated with 200 units/mi IFN-y for 0.5 to 4 days andthen stained with the immunogold technique for the immunohistochemicaldemonstration of Ia antigen. Results are expressed as percentage of labeled cellsin treated cultures. Untreated cells were not labeled. Bars represent the mean ±SEM of duplicate wells of two separate experiments.123C/)C)0ci)ci)0N00-Fig. 8 lime course of Ia Ag inductionon HBMEC by IFN-y100806040200 0.5 1 2 3 4Time (days)124Figure 9: Dose response of Ia antigen induction by IFN-y on HBMEC. Confluentmonolayers were incubated for 4 days with 10 to 200 units/mi IFN-y and thenimmunostained for the demonstration of Ia antigen. Results are expressed aspercentage of labeled cells in treated and untreated cultures. Bars represent themean ± SEM of duplicate wells of three separate experiments.125Fig. 9 Dose-response of Ia Ag inductionby IFN-y On HBMECC,)ci)C)ci)ci)-Q001008060402000 10 20 50 100 150 200IFN-’Y Concentration (U/mi)126Figure 10: Ia antigen expression by HBMEC detected by immunogold silver staining. A,endothelial cells incubated with 200 units/ml IFN-y for 4 days demonstratingintense granular surface staining for Ia antigen. B, control untreated monolayersnot expressing Ia antigen. C, endothelial cells treated with 200 units/ml IFN-y for24 hours exhibiting less dense labeling. Individual cell variation in stainingintensity is apparent in A and C. In D, cultures coincubated with anti-IFN-yantibody failed to label with the immunogold reagent. Bars = 20 tm.I<qI0Figure 11: Immunogold staining of HBMEC for the demonstration of Ta antigen. A,endothelial cells incubated with 200 units/mi IFN-y for 4 days. Fivenanometer gold particles focally decorate the apical surface of endothelialcells (arrowheads) with a tendency to localize close to fingerlike cytoplasmicfolds. The basal cell surface is not labeled. B, staining is absent in untreatedcells. Bars = 0.5 tm.12841I,•%111)4’4q“‘‘111*1I’1434b‘1‘“4tee14‘I14130Figure 12: Ia antigen expression by HBMEC detected by immunogold silver staining.a) endothelial cells incubated with 6,000 units/ml IFN-13 for 4 days failedto express of Ta Ag as indicated by the negative staining of EC. b) Incultures coincubated with IFN-y (100 units/mI) and IFN-3 (500 units/ml)expression of Ia Ag is limited to a small number of endothelial cellsdisplaying positive surface labeling with immunogold silver staining(arrows). Bars = 20 tm.rH,IId0aa132Figure 13: Dose response of Ia Ag expression by HBMEC treated with IFN-y and/orIFN-j3. Confluent monolayers were examined untreated, or followingtreatment with IFN-3 (6,000 units/mi) or IFN-y (100 units/mi) or with acombination of IFN-y (100 units/mi) and 13(100 to 6,000 units/mi) for 4days. At the end of the incubation period, monolayers were stained withthe immunogoid siiver staining technique for the surface detection of IaAg. Resuits are expressed as percentage of iabeied ceiis in treated anduntreated cultures. Bars represent the mean ± SEM of triplicate wells ofthree separate experiments.133Fig. 13 Dose-response of Ia Ag expressionby HBMEC treated with IFN-1 and/orIFN-p% labeled cells0 20 40 60 80 100untreatedIFN-6cJoIFN- 100c IFN-yioo/_________IFN- 13 100E IFN-yioo/___,__ IFN- 13250a,-FN- 135Y3F—IFN-y ico/FN-13ux0 IFN-’ lao!__FN- 132000 IIFN- y 100/IFN-13aOcxJ134Figure 14: Effects of different treatments of IFN-y and 3 on Ta Ag expression byHBMEC. Confluent monolayers were left untreated, or treated with TFN-13(6,000 units/mi) or TFN-y (100 units/mi) or a combination of IFN-y (100units/mi) and 1 (6,000 units/mi) for 4 days, or treated with TFN-y or IFN1 alone for 2 days, followed by a combination of TFN-y and for another 4days [TFN-y(2) or IFN-13(2)/IFN-+y(4)], or with a combination of IFN-j3(6,000 units/mi) and y (100 units/mi) for 2 days, followed by IFN-y alonefor another 4 days [IFN-3+y (2)/IFN-y (4)] and then immunostained forthe demonstration of Ta Ag. Results are expressed as percentage of labeledcells in treated and untreated cultures. Bars represent the mean ± SEM oftriplicate wells of two separate experiments.135Fig. 14 Effects of different treatmentsof IFN- and f on Ia Ag expressionby HBMEC% labeled cells0 20 40 60 80 100untreatedIFN-y icCaIFN-13÷yIFN-y(2)/IFN-13+y (4)FN-13 (2) /IFN-13+y (4)IFN-p+Y (2)/IFN-y (4)136Figure 15: Quantitation by ELISA of Ia Ag expression by HBMEC treated with IFN-yand/or IFN-. Confluent monolayers of HBMEC were left untreated, ortreated with IFN-y (100 units/ml), or IFN- (6,000 units/mi), or with acombination of JFN-y (100 units/mI) and 3 (100 to 6,000 units/mI) for 4days, or with IFN-i or y alone for 2 days, followed by a combinedtreatment with IFN-13 and y for another 4 days (132/y2 + f3y4). Valuesrepresent mean ± SEM of triplicate wells.Ia Ag expression was measured in triplicate wells of confluent HBMECcultures.137Fig. 15 Quantitation by ELISA of Ia Agexpression by HBMEC treated withIFN-y and/or IFN-pAbsothace at 490 nm0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8I I I I I I IUntreatedR1-y 100P14 6,000R1-y+ 100R1-y+P1-y+ 1,000IR1-y+ 6,000 I2+Y4 Iy2+y4138Figure 16: A, HBMEC maintained under standard culture conditions formed highlyorganized, confluent, contact inhibiting monolayers composed of elongatedcells. B, endothelial cells incubated with IFN-y (200 units/ml) for 4 dayshave become spindle shaped, overlap, and focally arrange themselves intowhorls. Bars = 10 tim.13917.140Figure 17: Scanning electron micrograph of HBMEC grown in the absence (A) orpresence (B) of IFN-y in the culture media. A, endothelial cells closelypacked without apparent intercellular spaces. Marginal folds (arrows) arepresent in areas of cell-cell contact. In B, incubation with IFN-y (200units/ml) for 3 days induces marked attenuation of cell cytoplasm anddisorganization of the monolayer due to the tendency of endothelial cellprocesses to extend over and under adjacent cells. Bars = 20 tm.141. . .. : . • • . . • •. • . .. . : .;;•wç. •• . . .—b- • —.--- : ::: ; E______--—— :—— —;i:__c—- .- -—‘-—N---N142Figure 18: Effects of IFN-y and f3 upon the growth of primary cultures of HBMEC.Cells were left untreated or treated from day 1 with IFN-y (150 units/mi) orIFN- (1,000 units/mi) alone, or with a combination of IFN-y and (100units/ml and 1,000 units/mi, respectively). Bars represent the mean ± SEMof triplicate wells.Cells/Well- 00oc,)(D:T 1 —aCl)r\)01001001000000C))0 00 :3UntreatedIFN-yIFN-FN-y+3144Figure 19: Scanning electron micrograph of HBMEC grown in the presence of IFN-falone (a) or IFN-3 and y (b) in the culture media. a) endothelial cells aremorphologically identical to untreated cultures, they are closely packedwithout apparent intercellular spaces. Marginal folds (arrows) are presentin areas of cell-cell contact. In b) incubation with a combination of IFN-13and ‘y (6,000 units/ml and 200 units/mi, respectively) prevents theoccurrence of the IFN-y induced changes in cell morphology andorganization of the monolayers. Arrows point to marginal folds in areas ofcell-cell contact. Bars = 50 tm.45&liiI Ii 6IL13147Figure 20: Permeability of untreated (A and B) and IFN-y-treated (C- F) confluentHBMEC monolayers to HRP. A, under standard culture conditions, tightjunctions at intercellular contacts (between arrowheads) impede thepassage of HRP. B, HRP penetrates a short segment of an intercellular cleftfrom the basal cell surface, forming small deposits at the basal aspect ofthe cleft (arrowheads) and stopping at a junctional complex (arrow). Theremaining interendothelial cleft is free of HRP. Untreated Cultures. In C,following 4 days’ incubation with IFN-y (200 units/ml), heavy deposits ofHRP are seen under the basal cell surface, and the tracer permeates theentire length of a long intercellular cleft. The proximal portion of the cleftis focally dilated (*). There is no increase in the pinocytotic activity of theendothelium. D, in monolayers treated with IFN-y, HRP penetrates theintercellular clefts and forms extensive deposits between the layers ofoverlapping EC. Bars = 0.5 tm. E, F, Permeable interendothelial clefts ofIFN-y treated monolayers exhibit HRP deposits throughout their length.Bars = 0.5 !m.-c0c0151Figure 21: Expression of IL-2R on resting T cells and anti-CD3 stimulated (a-CD3)lymphocytes. Approximately 2 fold increase in IL-2R expression isobserved when lymphocytes are stimulated with a-CD3 mAb for 3 days at37 CC.152Fig. 21 Expression of0C,)Anti-CD3 stimulated25lymphocytesIL-2R on Resting and2015-1050a-CD3 UntreatedTreatment153Figure 22: Adhesion of resting T lymphocytes to untreated confluent HBMECmonolayers. At the end of the incubation period, EC cultures with adherentlymphocytes were fixed and stained with the immunoperoxidase techniquefor leukocyte common antigen (LCA). Small number of LCA positivelymphocytes (L, arrowhead) adhere to the untreated endothelial cells (ec).Bar= 1Oim.4’1%S4.1 trD155Figure 23: Adhesion of resting T lymphocytes to IFN-y (150 units/mi) treatedHBMEC as demonstrated by immunoperoxidase staining for leukocytecommon antigen (LCA). Large numbers of LCA positive lymphocytes (L,anowhead) adhere to the IFN-y treated endothelial cells (ec). Bar = 10 tm.$S157Figure 24: Adhesion of resting T lymphocytes to IFN-13 (2,000 units/ml) treatedHBMEC as demonstrated by immunoperoxidase staining for leukocytecommon antigen (LCA). A small number of LCA positive lymphocytes (L,arrowhead) adhere to IFN- treated endothelial cells (ec). Bar = 10 tm.159Figure 25: Adhesion of resting lymphocytes to untreated and cytokine-treatedHBMEC. Confluent monolayers of EC were left untreated or treated withIFN-y (150 units/ml) or IFN-3 (2,000 units/mi), or with a combination ofIFN-y (150 units/mi) and 1E (2,000 units/mi) or IFN-y (150 units/mi) andanti-IFN-y antibody (ay - 10 fig/mi) for 3 days prior to incubation withresting T cells. For the mAb blocking studies, cuitures were treated withIFN-y for 3 days, followed by 2 hr incubation with anti-human HLA-DRmAb (ala) prior to incubation with resting T lymphocytes (T). Barsrepresent the mean ± SEM of triplicate weils of two separate experiments.160Fig. 25 Adhesion of resting Lymphocytesto untreated and cytokine-treatedHBMECCells/mm2EC+TEC+R4-y +TEC+FN-p +TEC+IRI-y+p +TEC+1R4-y+ay +TEC+1B4-y+ aIQ+T50 100 1500 200I-H161Figure 26: Scanning electron micrograph demonstrating the adhesion of resting Tlymphocytes to untreated HBMEC. a) Lymphocytes (L) first adhere to theendothelium (BC) by extending pseudopodia that contact the endothelialsurface (arrowhead). Marginal folds (arrow) are present in areas of cell-cellcontact. Bar = 20 tm. b) Lymphocytes eventually position themselvesbetween adjacent EC (bar = 4.5 tm), and c) begin migrating across themonolayer (bar = 1.8 tm). d) Lymphocytes were infrequently seenpenetrating the apical EC plasma membrane (arrowheads) and movingthrough the endothelial cytoplasm (bar = 4.36 tm).oQ6 aCc3p4166Figure 27: Scanning electron micrograph of the adhesion of resting T lymphocytes toIFN-y (150 units/ml) treated HBMEC. IFN-y treatment of the endothelialcells (EC) induces reorganization of the monolayer and a tendency of ECprocesses to overlap (arrow). a) Large numbers of lymphocytes (L)establish contact with the endothelium via pseudopodia (arrowheads). Bar= 50 m. b) Lymphocytes (L), singly or in small aggregates (arrowheads),align themselves along the borders between adjacent EC in preparation forcrossing the monolayers. Bar = 20 tm.Fig. ca169Figure 28: Adhesion of anti-CD3 stimulated T lymphocytes to untreated HBMEC asdemonstrated by immunoperoxidase staining for leukocyte commonantigen (LCA). Significant numbers of LCA positive activatedlymphocytes (L) adhere to untreated endothelial cells (EC). Activated Tcells are larger than resting lymphocytes and display irregular, folded, cellmembranes. Bar = 10 rim.j7OEC171Figure 29: Adhesion of anti-CD3 stimulated T lymphocytes to IFN-y (150 units/ml)treated HBMEC as demonstrated by immunoperoxidase staining forleukocyte common antigen (LCA). Large numbers of LCA positiveactivated lymphocytes (L) adhere to endothelial cells (EC). Focally,lymphocytes begin to migrate across the monolayer by extendingpseudopodia between EC (arrowheads). Bar 10 urn.‘USfr173Figure 30: Adhesion of anti-CD3 stimulated T lymphocytes (L) to IFN-y (150units/mi) and 1E (2,000 units/mi) treated HBMEC as demonstrated byimmunoperoxidase staining for ieukocyte common antigen (LCA).Leucocyte-EC adhesion is comparable to that observed between anti-CD3stimulated T cells and untreated EC. Bar = 10 tm.41:IE.S.4lbIIEC‘>4175Figure 31: Adhesion of anti-CD3 activated lymphocytes to untreated and cytokinetreated HBMEC. Confluent monolayers of EC were left untreated ortreated with IFN-y (150 units/mi) or IFN-3 (2,000 units/mi), or with acombination of IFN-y (150 units/mi) and f3 (2,000 units/mI) or IFN-y (150units/mi) and anti-IFN-y antibody (cry - 10 tg/m1) for 3 days prior toincubation with activated T ceiis. For the mAb blocking studies, cultureswere treated with IFN-y for 3 days, followed by 2 hr incubation with antihuman HLA-DR mAb (ala) prior to incubation with activated Tlymphocytes (Tcd3). Bars represent the mean ± SEM of triplicate wells oftwo separate experiments.176Fig. 31 Adhesion of anti-CD3 activatedlymphocytes to untreated andcytokine-treated HBMECCells/mm20 90 180 270EC+Tcd3EC+R1-y +Tcd3EC+FN-p +TcdSEC+IFN-y + p+Tcd3EC+FN-y+ a7 +Tcd3360 450HEC+R1-y+ ala+1177Figure 32: Scanning electron micrograph of the adhesion of activated T lymphocytesto untreated (a) and IFN-y (150 units/mI) treated HBMEC (b, c). IFN-ytreatment of endothelial cells (EC) induces overlapping of EC processes (*in b). Activated lymphocytes (L) adhere to the endothelium (EC) in largenumbers, and they appear enlarged and exhibit a ruffled cell membranewith numerous folds in comparison to the resting lymphocytes (r)(arrowheads). Activated lymphocytes establish close contact with EC bymeans of cytoplasmic projections (small arrows) and usually positionthemselves along the borders between adjacent EC in both untreated andIFN-y treated EC. Protuberances on the apical surface of the endothelium,having the size and shape of an activated T cell, indicate movementthrough the EC cytoplasm (large arrows) (a - c). Bars = 50 tm (a ,b) and19 tm (c).F. 30to—3eZCLgo181Figure 33: Adhesion anti-CD3 activated lymphocytes to untreated HBMEC. Activatedlymphocytes (L) display abundant cytoplasm, increased numbers ofmitochondria (m) and variable numbers of cytoplasmic vacuoles (V)containing amorphous, flocculent material. Several points of close cell tocell contact between endothelium (EC) and processes of adherentlymphocytes are present (arrows). C, cellagen membrane. Bar = 1’S----4t.J183Figure 34: Adhesion of anti-CD3 stimulated T cells to untreated HBMEC. The cellsurface of activated T lymphocytes (L) is extremely irregular due to thepresence of numerous thin, finger-like cytoplasmic processes (arrows).Variable numbers of cytoplasmic vacuoles (V) containing amorphousflocculent material are present in the lymphocyte cytoplasm. EC,endothelial cell; N, nucleus; C, cellagen membrane. Bar = 1 tim.:0::.-ijA185Figure 35: Adhesion of resting lymphocytes (L) to IFN-y (150 units/mi) and antihuman HLA-DR mAb treated HBMEC as demonstrated byimmunoperoxidase staining for leukocyte common antigen (LCA). FewLCA positive lymphocytes adhere to endothelial cells (EC). Bar = 10 tm.‘18I4a—ECa187Figure 36: Adhesion of anti-CD3 stimulated lymphocytes (L) to IFN-y (150 units/ml)and anti-human HLA-DR mAb treated HBMEC as demonstrated byimmunoperoxidase staining for leukocyte common antigen (LCA). Thenumber of lymphocytes adhering to endothelial cells (EC) is comparable tothat observed in the absence of pretreatment with mAb. Bar = 10 im.188189Figure 37: a) Transendothelial migration of resting T lymphocytes (arrow) acrossuntreated HBMEC (EC) monolayers is minimal. C, cellagen membrane. b)Significant increase in migration of lymphocytes (arrows) was observed inHBMEC pretreated with IFN-y (150 units/mi) for 3 days. c) Coincubationof EC with IFN-y (150 units/mi) and 13(2,000 units/mi) for 3 days,significantly suppresses the IFN-y-enhanced migration (arrow). Bars = 10tm (a-c).190‘3-f-191Figure 38: Migration of resting and anti-CD3 stimulated T lymphocytes acrossuntreated and IFN-y and/or IFN-3 treated HBMEC. Confluent monolayersof EC were left untreated or treated with IFN-y (150 units/mi) or IFN-(2,000 units/mI), or with a combination of IFN-y (150 units/mI) and 3(2,000 units/mi) for 3 days prior to incubation with resting (T) or anti-CD3stimulated lymphocytes (Tcd3) for 3 hours. For the blocking studies, ECwere treated for 3 days with IFN-y (150 units/mi), followed by 2 hours ofincubation with anti-human HLA-DR (ala) mAb before incubating with Tor Tcd3 for 3 hours. Bars represent the mean ± SEM of 200 differentlevels.192Fig. 38 Migration of resting and anti-CD3activated T cells across untreatedand IFN-y and/or IFN-p treatedHBMECI—+I-+Ia-i— .+ + + +. - ?-I—I— .+ + + +H”I00)04UUJ4D8J1193Figure 39: Migration of resting T lymphocytes across untreated HBMEC monolayers(a - d). a) A lymphocyte (L) initiates its migration across the endothelialcell (EC) monolayers by directing cytoplasmic processes between twoadjacent EC. N, nucleus; C, cellagen membrane. Bar = 1 m. b and c) Partof the cytoplasm and nucleus moves between two adjacent EC. The cellmembranes of the T cell and EC are closely apposed. Bars= 2 im for (b)and 1 tm for (c). d) At the end of the migration period, the lymphocytesposition themselves between the overlying EC and underlying cellagenmembrane and become elongated and flattened. The processes of theoverlying EC have been resealed (arrow). Bar = 1 im.-D cit198Figure 40: Transendothelial migration of a resting lymphocyte (L) across untreatedHBMEC (EC). This lymphocyte is considered moving through rather thanbetween adjacent EC because the EC cytoplasm surrounds the lymphocyte.C, collagen membrane. Bar = 1 tm.200Figure 41: The integrity of the EC monolayers is reestablished once restinglymphocytes (L) have completed their migration across the IFN-y (150units/mi, 3 days) treated EC (a, b). Bars = 2 rim.N 00203Figure 42: Nonspecific activation of T lymphocytes enhances their migration acrossuntreated HBMEC monolayers. Migrated T cells (arrows) remain betweenEC and cellagen membranes (C). Bar = 10 pm.204205Figure 43: Migration of activated lymphocytes across untreated (a) and IFN-y treatedHBMEC (b). a) An activated T lymphocyte (L) begins to migrate betweenadjacent endothelial cells (EC). Close contact between EC and thelymphocyte is maintained (arrows). C, collagen membrane. b) At the endof the migration period, monolayers resume their continuity over themigrated lymphocytes (L). EC, endothelial cells; C, collagen membrane.Bars = 2 m.0IzoqA208Figure 44: Migration of anti-CD3 stimulated lymphocytes across IFN-y treatedHBMEC. The lymphocytes (L) that have completed their migration acrossEC of the top layer, proceeded to migrate across the next layer of EC. Bar2 tm.0C..SLLZLr‘-.,,‘—C210Figure 45: Effect of (A) Ca2 ionophore, (B) EGTA, and (C) IFN-y on the release ofFVIIIR:Ag from human brain microvessel endothelial cells. After a 10 mmincubation with 10 pM Ca2+ ionophore (A), the cytoplasmic vesicles aredepleted of FVIIIR:Ag, as indicated by their lack of staining. N, nucleus.(B) After 10 mm treatment with 1 mM EGTA, most of the vesicles containgold particles in small aggregates. N, nucleus. (C) Incubation with 200U/mi IFN-y for 24 hours is associated with variable staining of a largenumber of vesicles. Bars: A = 1 aim; B,C = 1.5’•••‘••‘:s1•‘••. *•*•*J*•---•:—,•...--%••..--y:.:-‘i1--tvkt-•_;•—‘‘:‘tc‘-•-1,•is4’•*n’---L•§Ig1-%.•K-;*I4D:::e*-•w_*:--:•—,•-•*_+-**cst..%c%vtS.t4*It)3flt4—1-ici’ C (F’V4i-‘PY #r”$.%_*;*,.-Ar.7*4W--•-1-•.j:wc:”**-O-’•;,cNiiqØf’$I;c-,*egn.t.P;-v---i-•*•*It•..*Ifl•4*.*¾t%*4.j15•frç--*bc9•I-.,-.-t-$t)4-s-.,t4-½-çh—%:4”•:1--.“—---,----n!-,.1-.t-$Z*•e.•-•.-••-,•--4’3,.-.e:1,--i,,.*-c--.•.-cj•••:*T--.•-‘i:\4”_i%’—‘%‘-3***-*‘*42*-‘----*t$••2•‘.-•*•r(••.--,-•;4•.•43t<----“:--I*---4$.-r:H6—.•-S‘4id•-t1&i•-•-*---H-A‘4*-r-t$’----?-•--*-•---I—1’.--4%;%—.••A4‘¾•H’**-Ip-*:,7it%tt*E:‘•*---t.a212Figure 46: Percentage of immunostained vesicles in untreated and treated cultures.There is a slight increase in the number of labeled vesicles after incubationwith EGTA (55%) vs the untreated cultures (45%), p 0.21. The numberof stained vesicles increased significantly after pre-incubation with IFN-y(72%), p = 0.000. Staining was largely abolished after treatment withCa2+ ionophore. Bars represent the mean + SEM.213Fig. 46 Immunocytochemical Localizationof Factor Vill/Von WillebrandAntigen in Human BrainMicrovessel Endothelial Cells10080> 60________40C200——Untreated A23187 EGTA rIFN-7Treatment214REFERENCES1. Cohnheim J. Lectures on general pathology: A handbook for Practitioners andStudents. Section 1 - The pathology of the circulation. The New Sydenham Society,London, England 1889: 242 - 3822. Gallin J.I., Goldstein I.M., Snyderman R. Inflammation: Overview. In Gallin J.I.,Goldstein I.M., and Snyderman R. (eds). Inflammation: Basic Principles and ClinicalCorrelates. Raven Press, Ltd. New York 1992: 1 - 43. Weissman G. Inflammation: Historical Perspective. In Gallin J.I., Goldstein I.M., andSnyderman R. (eds). Inflammation: Basic Principles and Clinical Correlates. RavenPress, Ltd. New York 1992: 5 - 94. Clark E.R., Clark E.L. Observations on changes in blood vascular endothelium in theliving animal. Am. J. Anat. 1935, 57: 385 - 4385. Swerlick R.A., Lawley T.J. Role of microvascular EC in inflammation. J. Invest.Dermatol. 1993, 100: illS - 115S6. Lorant D.E., Patel K.D., McIntyre T.M., Mclver R.P., Prescott S.M., Zimmerman G.E.Coexpression of GMP-140 and PAF by Endothelium stimulated by histamine orthrombin: A juxtacrine system for adhesion and activation of neutrophils. J. Cell. Biol.Oct. 1991, 115: 223 - 2347. Dorovini-Zis K., Bowman P.D., Prameya R. Adhesion and Migration of humanpolymorphonuclear leukocytes across cultured bovine brain microvessel EC. J.Neuropath. Exp. Neurol. March 1992, 51: 194 - 2058. Hartung H.P., Jung S., Stoll G., Zielasek J., Schmidt B., Archelos J.J., Toyka K.V.Inflammatory mediators in demyelinating disorders of the CNS and PNS. J.Neuroimmunol. 1992, 40: 197 - 2109. Gamble J.R., Harlan J.M., Kiebanoff S.J., Vadas M.A. Stimulation of the adherence ofneutrophils to umbilical vein endothelium by human recombinant tumor necrosisfactor. Proc. Natl. Acad. Sci. U.S.A. 1985, 82: 8667 - 867110. Bevilacqua M.P., Pober J.S., Wheeler M.E., Cotran R.S., Gimbrone M.A. Jr.Interleukin- 1 acts on cultured human vascular endothelium to increase the adhesion ofpolymorphonuclear leukocytes, monocytes, and related leukocyte cell lines. J. Clin.Invest. 1985, 76: 2003 - 201111. Pober J.S., Gimbrone M.A. Jr., Lapierre l.A., Mendrick D.L., Fiers W., Rithlein R.,Springer T.A. Overlapping patterns of activation of human EC by interleukin-1, tumornecrosis factor, and immune interferon. J. Immunol. 1986, 137: 1893 - 189612. Hartung H.P., Schafer B., Heininger K., Toyka K.V. Recombinant interleukin-113stimulates eicosanoid production in rat primary culture astrocytes. Brain Res. 1989,215489: 113- 11913. Hartung H.P., Heininger K. Non-specific mechanisms of inflammation and tissuedamage in MS. Res. Immunol. 1989, 140: 226 -23314. Wekerle H., Engelhardt B., Risau W., Meyermann R. Interaction of T lymphocyteswith Cerebral EC in vitro. Brain Pathology 1991, 1: 107 - 11415. Panitch H.S., Hirsch R.K., Schindler J., Johnson K.P. Treatment of multiple sclerosiswith gamma interferon: exacerbation associated with activation of the immunesystemm. Neurology 1987, 37: 1097 - 110216. Page C., Rose M., Yacoub M., Pigott R. Antigenic Heterogeneity of VascularEndothelium. Am. J. Path. 1992, 141: 673 - 68317. Plendi J, Sinowatz F., Auerbach R. Heterogeneity of the vascular endothelium. Anat.Histol. Embryo!. 1992, 21: 256 - 26218. Fajardo L.F. The complexity of EC. Am. J. Clin. Path. 1989, 92: 241 - 25019. Zetter B.R. Endothelial heterogeneity: influence of vessel size, organ localization andspecies specificity on the properties of cultured EC. In Una S Ryan (ed) EC II. CRCPress, Inc. Florida 1988: 63- 7920. Isaacs A. and Lindenmann J. Virus interference. I The interferon. Proc. R. Soc. LondonB. Biol. Sci. 1957, 147: 258 - 26721. Sen G.C. and Lengyel P. The Interferon System. A bird’s eye view of its biochemistry.J. Biol. Chem. 1992, 267: 5017 - 502022. Pober J.S., Gimbrone M.A. Jr, Cotran R.S., Reiss C.S., Burakoff S.J., Fiers W., AultK.A. Ia expression by vascular endothelium is inducible by activated T cells and byhuman y interferon. J. Exp. Med. 1983, 157: 1339 - 135323. Yu C.L., Haskard D.O., Cavender D., Johnson A.R., Ziff M. Human gamma interferonincreases the binding of T lymphocytes to EC. Clin. Exp. Immuno!. 1985, 62: 554 -56024. Oppenheimer-Marks N., Ziff M. Migration of lymphocytes through EC monolayers:Augmentation by Interferon-y. Cell. Immunol. 1988, 114: 307 - 32325. Barna B.P., Chou S.M., Jacobs B., Yen-Lieberman B., Ransohoff R.M. Interferon betaimpairs induction of HLA-DR antigen expression in cultured adult human astrocytes. J.Neuroimmunol. 1989, 23: 45 - 5326. Joseph J., Knobler R.L., D’Imperio C., Lublin F.D. Down - regulation of interferongamma induced class II expression on human glioma cells by recombinant interferonbeta: effects of dosage treatment schedule. J. Neuroimmunol. 1988, 20: 39 - 4427. Inaba K., Kitaura M., Kato T., Watanabe Y., Kawade Y., Muramatsu S. Contrastingeffect of alpha/beta and gamma interferons on expression of macrophage Ia Ags. J.Exp. Med. 1986, 163: 1030- 103521628. Panitch H.S., Folus J.S., Johnson K.P. Beta interferon prevents HLA Class II antigeninduction by gamma interferon in MS. Neurology 1989, 39 (Suppl 1): 17129. Trent J.M., Olson S and Cawn R.M. Chromosomal localization of human leucocyte,fibroblast, and immune interferon genes by means of in situ hybridization. Proc. Natl.Acad. Sci. USA 1982, 79: 7809 - 781330. Morris A.G., Lin Y.L., Askonas B.A. Immune interferon release when a clonedcytotoxic T cell lime meets its correct influenza infected target cell. Nature 1982, 295:150 - 15231. Klein J.R., Raulet D.H., Pasternack M.S., Bevan M.J. Cytotoxic T lymphocytesproduce immune interferon in response to antigen or mitogen. J. Exp. Med. 1982, 155:1198 - 120332. Trinchieri G., Matsumoto-Kobayashi M., Clark S.V., Sheera J., London L., Perussia B.Response of resting human peripheral blood natural killer cells to interleukin-2. J. Exp.Med. 1984, 160: 1147 - 116933. Munakata T., Semba U., Shibaya Y., Kuwano K., Akagi M., Arai S. Induction ofinterferon-y production by human natural killer cells stimulated by hydrogen peroxide.J. Immunol. 1985, 134: 2449 - 245534. Taniguchi T. Regulation of Interferon-13 gene: Structure and Function of cis-elementsand trans-acting factors. J. mt. Res. 1989, 9: 633 - 64035. Lengyel P. Biochemistry of interferons and their actions. Annu. Rev. Biochem. 1982,51: 251 - 28236. Pestka S., Langer A.J., Zoon K., Samuel C. Interferons and their actions. Annu. Rev.Biochem. 1987, 56: 727 - 77737. Fieischmann W.R. Jr., Schwarz L.A. Demonstration of potentiation of the antiviraland antitumor actions of interferon. Methods Enzymol. 1981, 79: 432 - 44038. Panitch H.S., Folus J.S., Johnson K.P. Recombinant beta interferon inhibits gammainterferon production in multiple sclerosis. Ann. Neurol. 1987, 22: 13939. Noronha A., Toscas A., Jensen M.A. IFN-3 down-regulates IFN-y production byactivated T cells in MS. Neurology 1991, 41(Suppl.1): 21940. Ling P.D., Warren M.K., Vogel S.N. Antagonistic effect of IFN-1E on the IFN-yinduced expression of Ia Ag in murine macrophages. J. Immunol. 1985, 135: 1857 -186341. Ransohoff R.M., Devajyothi C., Estes M.L., Babcock G., Rudick R.A., Frohman E.M.,Barna B.P. IFN-3 specifically inhibits interferon-induced class II majorhistocompatibility complex gene transcription in a human astrocytoma cell line. J.Neuroimmunol. 1991, 33: 103 - 11242. Noronha A., Toscas A., Jensen M.A. Interferon beta augments suppressor cell function217in Multiple Sclerosis. Ann. Neurol. 1990, 27: 207 - 21043. Janeway C.A. Jr. How the Immune System Recognizes Invaders. ScientificAmerican 1993, 269: 73 - 7944. Benacerraf B., Dausset J., Snell G.D. The Nobel lectures in Immunology. Scand. J.Immunol. 1992, 36: 145 - 15745. Unanue E.R., Allen P.M. The basis for the immunoregulatory role of macrophages andother accessory cells. Science 1987, 236: 551 - 55746. Hirschberg H., Braathen L.R., Thorsby E. Antigen presentation by vascular EC andepidermal Langerhans cells. The role of HLA-DR. Immunol. Rev. 1982, 66: 57 - 7747. Brown J.H., Jardetzky T.S., Gorga J.C., Stern L.J., Urban R.G., Strominger J.L., WileyD.C. Three - dimensional structure of the human class II histocompatibility antigenHLA-DR1. Nature 1993, 364: 33 - 3948. Germain R.N. Antigen processing and presentation. AIDS Research and HumanRetrovirus 1992, 8: 769 - 77649. Erlich P., Morgenroth J. In: The collected papers of Paul Erlich. Edited by HimmelweitF., Marquardt M., Dale H.D.E. 1900. Pergamon, London.50. Nepom G.T., Hansen J.A., Nepom B.S. The molecular basis of HLA class IIassociations with rheumatoid arthritis. J. Clin. Immunol. 1987, 7: 1 - 751. Nepom G.T., Erlich H. MHC class-TI molecules and autoimmunity. Annu. Rev.Immunol. 1991, 9: 493 - 52552. Smilek D.E., Lock C.B., McDevitt H.O. Antigen recognition and peptide-mediatedimmunotherapy in autoimmune disease. Immunol. Rev. 1990, 118: 37 - 7153. Awata T., Kuzuya T., Matsuda A., Iwamoto Y., Kanazawa Y. Genetic analysis ofHLA class II alleles and susceptibility to type 1 (Insulin-dependent) diabetes mellitusin Japanese subjects. Diabetologia 1992, 35: 419- 42454. Stastny P. Association of the B cell alloantigen DRw4 with rheumatoid arthritis. N.Eng. J. Med. 1978, 298: 869 - 87155. Roudier J., Rhodes G., Petersen J., Vaughan J.H., Carson D.A. The Epstein-Barr virusglycoprotein gp 110, a molecular link between HLA DR4, HLA DR1, and rheumatoidarthritis. Scand. J. Immunol. 1988, 27: 367 - 37156. Carlsson B., Wallin J., Pirskanen R., Matell G., Smith C.I. Different HLA-DR-DQassociations in subgroups of idiopathic myasthenia gravis. Immunogenetics 1990, 31:285 - 29057. Spurkland A., Gilhus N.E., Ronningen K.S., Aarli J.A., Vartdal F. Myasthenia gravispatients with thymus hyperplasia and myasthenia gravis patients with thymoma displaydifferent HLA associations. Tissue Antigens 1991, 37: 90 - 9321858. Haegert D.G., Michaud M., Francis G.S. Multiple sclerosis in French Canadians:Evidence for HLA class II susceptibility and resistance genes. Can. J. Neurol. Sci.1990, 17: 382 - 38659. Spurkland A., Ronningen K.S., Vandvik B., Thorsby E., Vartdal F. HLA-DQ Al andHLA-DQB1 genes may jointly determine susceptibility to develop multiple sclerosis.Human Immunology 1991, 30: 69 - 7560. Tomimoto H., Akiguchi I., Akiyama H., Kimura J., Yanagihara T. T cell infiltrationand expression of MHC class II antigen by macrophages and microglia in aheterogenous group in Leukoencephalopathy. Am. J. Path. 1993, 143: 579 - 58661. Yokoyama H., Takaeda M., Wada T., Ogi M., Tomosugi N., Takabatake T., Abe T.,Yoshimura M., Kida H., Kobayashi K. Intraglomerular expression of MHC class II andKi-67 antigens and serum y-interferon levels in IgA Nephropathy. Nephron 1992, 62:169 - 17562. Helbig H., Gurley R.C., Reichl R.J., Mahdi R., Nussenblatt R.B., Palestine A.G.Induction of MHC class II antigen in cultured bovine ciliary epithelial cells. Graefe’sArch. Clin. Exp. Ophthalmol. 1990, 228: 556 - 56163. Sinha A.A., Lopez M.T., McDevitt H.O. Autoimmune disease: The failure of self-tolerance. Science 1990, 248: 1380 - 138864. Ferry B., Halttunen J., Leszczynski D., Schellekens H., Meide P.H., Hayry P. Impact ofclass II major histocompatibility complex antigen expression on the immunogenicpotential of isolated rat vascular EC. Transplantation 1987, 44: 499 - 50365. Vass K., Lassmann H., Wekerle H., Wisniewski H.M. The distribution of Ia-antigen inthe lesion of rat acute experimental allergic encephalomyelitis. Acta Neuropathol.1986, 70: 149 - 16066. Sobel R.A., Blanchette B.W., Bhan A.K., Colvin R.B. The immunopathology ofexperimental allergic encephalomyelitis. II EC Ia increases prior to inflammatory cellinflammation. J. Immunol. 1984, 132: 2402 - 240767. Lampson L.A. Molecular bases of the immune response to neural behaviors. TrendsNeurosci. 1987, 10: 211 - 21668. Traugott U., Scheinberg L.C., Raine C.S. On the presence of Ia-positive EC andastrocytes in multiple sclerosis lesions and its relevance to antigen presentation. J.Neuroimmunol. 1985, 8: 1 - 1469. Pardridge W.M., Yang J., Buciak J., Tourtellotte W.W. Human Brain MicrovascularDR-Antigen. J, Neurosci. Res. 1989, 23: 337- 34170. Male D.K., Pryce G., Hughes C.C.W. Antigen presentation in brain: MHC inductionon brain endothelium and astrocytes compared. Immunology 1987, 60: 453 - 45971. Berrih S., Arenzana-Seisdedos F., Cohen S., Devos R., Charron D., Virelizier J.L.Interferon-y modulates HLA class II antigen expression on cultured human thymicepithelial cells. J. Immunol. 1985, 135: 1165 - 117121972. Skoskiewicz M.J., Colvin R.B., Schneeberger E.E., Russel P.S. Widespread andselective induction of major histocompatibility complex-determined antigens in vivoby y-interferon. J. Exp. Med. 1985, 162: 1645 - 166473. Martin M., Schwinzer R., Schellekens H., Resch K. Glomerular mesangial cells inlocal inflammation. Induction of the expression of MHC class II antigens by IFN-y. J.Immunol. 1989, 142: 1887 - 189474. Steeg P.S., Moore R.N., Johnson H.M., Oppenheim J.J. Regulation of murinemacrophage Ia Ag expression by a lymphokine with immune interferon activity.J.Exp. Med. 1982, 156: 1780 - 179375. Fertsch-Ruggio D., Schoenberg D.R., Vogel S.N. Induction of macrophage Ia Ag byrIFN-y and down-regulation by IFN-ct/13 and dexamethasone are regulatedtranscriptionally. J. Immunol. 1988, 141: 1582- 158976. Wong G.H.W., Clark-Lewis I., McKimm-Breschkin J.L., Schrader J.W. Interferon-ylike molecule induces Ta Ags on cultured mast cell progenitors. Proc. Nat!. Acad. Sci.USA 1982, 79: 6989 - 699377. Morel P.A., Manolagas S.C., Provvedini D.M., Wegmann D.R., Chiller J.M.Interferon-y-induced Ia expression in WEHI-3 cells is enhanced by the presence of 1,25-dihydroxyvitamin D3. J. Immunol. 1986, 136: 2181 - 218678. Basham T.Y., Merigan T.C. Recombinant Tnterferon-y increases HLA-DR synthesisand expression. J. Immunol. 1983, 130: 1492 - 149479. Kelley V.E., Fiers W., Strom T.B. Cloned human interferon-gamma, but notinterferon-beta or -alpha, induces expression of HLA-DR determinants by fetalmonocytes and myeloid leukemic cell lines. J. Immunol. 1984, 132: 240 - 24580. Collins T., Korman A.J., Wake C.T., Boss J.M., Kappes D.J., Fiers W., Ault K.A.,Gimbrone M.A., Strominger J.L., Pober J.S. Immune interferon activates multipleclass II major histocompatibility complex genes and the associated invariant chaingene in human EC and dermal fibroblasts. Proc. Natl. Acad. Sci. USA 1984, 81:4917 - 492181. Mantegazza R., Hughes S.M., Mitchell D., Travis M., Blau H.M., Steinman L.Modulation of MHC class II antigen expression in human myoblasts after treatmentwith IFN-y. Neurology 1991, 41: 1128 - 113282. Michaelis D., Goebels N., Hohifeld R. Constitutive and cytokine-induced expressionof human leukocytes antigens and cell adhesion molecules by human myotubes. Am.J. Path. 1993, 143: 1142 - 114983. Doukas J., Pober J.S. Lymphocyte-mediated activation of cultured EC (EC). CD4 Tcells inhibit EC class II MHC and intercellular adhesion molecule-i expression. J.Immunol. 1990, 145: 1088 - 109884. Ruszczak Z., Detmar M., Imcke E., Orfanos C.E. Effects of rIFN Alpha, Beta, Gammaon the morphology, proliferation, and cell surface antigen expression of human dermal220microvascular EC in vitro. J. Invest. Dermatol. 1990, 95: 693 - 69985. Botazzo G.F., Pujol-Borrell R., Hanafusa T., Feldman M. Role of aberrant HLA-DRexpression and antigen presentation in the induction of endocrine autoimmunity.Lancet 1983, 2: 1115 - 111886. Unanue E.R. Antigen-presenting function of the macrophage. Annu. Rev. Immunol.1984, 2: 395 - 42887. de Waal R.M.W., Bogman M.J.J., Maass C.N., Cornelissen L.M.H., Tax W.J.M.,Koene R.A.P. Variable expression of Ia Ags on the vascular endothelium of mouseskin allografts. Nature 1983, 303: 426 - 42988. McFarlin D.E., McFarland H.F. Multiple Sclerosis. N. Engl. J. Med. 1982, 307: 1246 -125189. Oger J., Roos R., Antel J.P. Immunology of multiple sclerosis. In: Antel J.P., ed.Neurologic clinics. Philadelphia: WB Saunders, 1983: 655 - 67990. Waksman B.H., Reynolds W.E. Multiple sclerosis as a disease of immune regulation.Proc. Soc. Exp. Biol. Med. 1984, 175: 282- 29491. Neighbour P.A., Miller A.E., Bloom B.R. Interferon responses of leucocytes inmultiple sclerosis. Neurology 1981, 31: 561 - 56692. Vervliet G., Claeys H., Van Haver H., Carton H., Vermylen C., Meulepas E., BilliauA. Interferon production and natural killer cell (NK) activity in leucocyte cultures frommultiple sclerosis patients. J. Neurol. Sci. 1983, 60: 137 - 15093. Hirsch R.L., Panitch H.S., Johnson K.P. Lymphocytes from multiple sclerosis patientsproduce elevated levels of gamm interferon in vitro. J. Clin. Immunol. 1985, 5: 386 -38994. Goust J.M., Verselis S.J., Shums A. Interferon gamma and interleukin 2 (1L2)production in MS. Abstract. Neurology 1987, 37 (Suppl. 1): 28995. Sheremata W.A., Bohn R.N., Resnick L., Berger J.R., Sazant A. Cerebrospinal fluidgamma interferon is increased in active multiple sclerosis. Ann. Neurol. 1987, 22: 15396. Aguet M. High-affinity binding of 125-labeled mouse interferon to a specific cellsurface receptor. Nature 1980, 284: 459 - 46197. Branca A.A., Baglioni C. Evidence that type I and type II interferons have differentreceptors. Nature 1981, 294: 768 - 77098. kar F.H., Gupta S.L. Receptors for human y interferon: binding and cross-linking ofL I-labeled human recombinant y interferon to receptors on WISH cells. Proc. NatI.Acad. Sci. USA 1984, 81: 5160 - 516499. Littman S.J., Faltynek C.K., Baglioni C. Binding of human recombinant 125kinterferon-y to receptors on human cells. J. Biol. Chem. 1985, 260: 1191- 1195221100. Rashidbaigi A., Kung H., Ptka S. Characterization of receptors for immuneinterferon in U937 cells with3LPlabeled human recombinant immune interferon. J.Biol. Chem. 1985, 260: 8514 - 8519101. Anderson P., Yip Y.K., Vilcek J. Specific binding of 125-human interferon-y to highaffinity receptors on human fibroblasts. J. Biol. Chem. 1982, 257: 11301 - 11304102. Branca A.A., Faltynek C.R., D’Alessandro S.B., Baglioni C. Interaction of interferonwith cellular receptors. Internalization and degradation of cell-bound interferon. J.Biol. Chem. 1982, 257: 13291 - 13296103. Hannigan G.E., Gewert D.R., Williams B.R.G. Characterization and regulation of a-interferon receptor expression in interferon-sensitive and -resistant humanlymphoblastoid cells. J. Biol. Chem. 1984, 259: 9456 - 9460104. Thompson M.R., Zhang Z., Fournier A., Tan Y.H. Characterization of human Iinterferon-binding sites on human cells. J. Biol. Chem. 1985, 260: 563 - 567105. Lapierre L.A., Fiers W., Pober JS. Three distinct classes of regulatory cytokinescontrol EC MHC antigen expression. J. Exp. Med. 1988, 167: 794 - 804106. Swerlick R.A., Garcia-Gonzalez E., Kubota Y., Xu Y., Lawley T.J. Studies of themodulation of MHC antigen and cell adhesion molecule expression on human dermalmicrovascular EC. J. Invest. Dermatol. 1991, 97: 190 - 196107. Dorovini-Zis K., Prameya R., Bowman P.D. Culture and characterization ofmicrovascular EC derived from human brain. Lab. Invest. 1991, 64: 425 - 436108. Rapoport S.!. Blood-brain barrier in Physiology and Medicine. Raven Press, 1976109. Brightman M.W. Morphology of blood-brain interfaces. In Bito L.Z., Davson H.,Fenstermacher J.D. (eds). The ocular and cerebrospinal fluids. Exp. Eye Res. (suppl.)1977, 25: 1 - 25110. Leibowitz S., Hughes R.A.C. Immunology of the Nervous System. Edward Arnold,London, 1983111. Nagy Z., Pappius H.M., Mathieson G., Huttner I. Opening of tight junctions incerebral endothelium. 1. Effect of hyperosmolar mannitol infused through the internalcarotid artery. J. Comp. Neurol. 1979, 185: 569 - 578112. Brightman M.W., Hon M., Rapoport S.I., Reese T.S., Westergaard E. Osmoticopening of tight junctions in cerebral endothelium. J. Comp. Neurol. 1973, 152: 317 -326113. Dorovini-Zis K., Sato M., Goping G., Rapoport S.I., Brightman M.W. Ionic lanthanumpassage across cerebral endothelium exposed to hyperosmotic arabinose. ActaNeuropath. (Ben) 1983, 60: 49 - 60114. Nakagawa Y., Cenvos-Navarro J., Artigas J. Tracer study on a paracellular route inExperimental Hydrocephalus. Acta Neuropath. (Ben) 1985, 65: 247 - 254222115. Beggs J.L., Waggener J.D. Transendothelial vesicular transport of protein followingcompression injury to the spinal cord. Lab. Invest. 1976, 34: 428 - 439116. Petito C.K., Schaefer J.A., Plum F. Ultrastructural characteristics of the brain andblood-brain barrier in experimental seizures. Brain Res. 1977, 127: 251 - 267117. Petito C.K. Early and late mechanisms of increased vascular permeability followingexperimental cerebral infarction. J. Neuropath. Exp. Neurol. 1979, 38: 222 - 234118. Nag S., Robertson D.M., Dinsdale H.B. Cerebral cortical changes in acuteexperimental hypertension. Lab. Invest. 1977, 36: 150 - 161119. Juhier M., Barry D.I., Offner H., Konat G., Klinken L, Paulson O.B. Blood-brain andblood-spinal cord barrier permeability during the course of experimental allergicencephalomyelitis in the rat. Brain Res. 1984, 302: 347 - 355120. Vulpe M., Hawkins A., Rozdilsky B. Permeability of cerebral blood vessels in EAE asstudied by radioactive iodinated bovine albumin. Neurology 1960, 10: 171 - 177121. Kristensson K., Wisniewski H.M. Chronic relapsing experimental allergicencephalomyelitis: Studies in vascular permeability changes. Acta Neuropathol. 1977,39: 189 - 194122. Kato S., Nakamura H. Ultrastructural and ultracytochemical studies on the blood-brainbarrier in chronic relapsing experimental allergic encephalomyelitis. ActaNeuropathol. 1989, 77: 455 - 464123. Lossinsky A.S., Badmajew V., Robson J.A., Moretz R.C., Wisniewski H.M. Sites ofegress of inflammatory cells and horseradish peroxidase transport across the blood-brain barrier in a murine model of chronic relapsing experimental allergicencephalomyelitis. Acta Neuropathol. 1989, 78: 359 - 371124. Simmons R.D., Buzbee T.M., Linthicum D.C., Mandy W.J., Chen G., Wang C.Simultaneous visualization of vascular permeability change and leukocyte egress in thecentral nervous system during autoimmune encephalomyelitis. Acta Neuropathol.1987, 74: 191 - 193125. McDonald W.I., Barnes D. Lessons from magnetic resonance imaging in multiplesclerosis. Trends Neurosci. 1989, 12: 376 - 379126. Stolpen A.H., Guinan E.C., Fiers W., Pober J.S. Recombinant Tumor Necrosis Factorand Immune Interferon act singly and in combination to reorganize human vascular ECmonolayers. Am. J. Path. 1986, 123: 16 - 24127. Huynh H.K., Dorovini-Zis K. Effects of Interferon- gamma on primary cultures ofhuman brain microvessel EC. Am. 3. Path. 1993, 142: 1265 - 1278128. Burke-Gaffney A., Keenan A.K. Modulation of human EC permeability bycombinations of the cytokines interleukin-1 a/n, tumor necrosis factor-a andinterferon-y. Immunopharmacology 1993, 25: 1- 9129. Martin S., Maruta K., Burkart V., Gillis S., KoIb H. IL-i and IFN-y increase vascular223permeability. Immunology 1988, 64: 301 - 305130. Damle N.K., Doyle L.V. Ability of human T lymphocytes to adhere to vascular ECand to augment endothelial permeability to macromolecules is linked to their state ofpost-thymic maturation. J. Immunol. 1990, 144: 1233 - 1240131. Sibley W.A., Bamford C.R., Clark K. Clinical viral infections and multiple sclerosis.Lancet 1985, 1: 1313 - 1315132. Johnson K.P., Panitch H.S. Interferon therapy for multiple sclerosis. Maryland Med.J. 1992, 41: 601 - 603133. Poser C.M. Pathogenesis of multiple sclerosis. Acta Neuropathol. (Berl.) 1986, 71: 1 -10134. Masuyama J., Minato N., Kano S. Mechanisms of lymphocyte adhesion to humanvascular EC in culture. T lymphocyte adhesion to EC through endothelial HLA-DRantigens induced by gamma interferon. J. Clin. Invest. 1986, 77: 1596 - 1605135. Goodall C.A., Curtis A.S.G., Lang S.C. Modulation of adhesion of lymphocytes tomurine brain EC in vitro: relation to class II major histocompatibility complexexpression. J. Neuroimmunol. 1992, 37: 9 - 22136. McCarron R.M., Wang L., Cowan E.P., Spatz M. Class II MHC antigen expression bycultured human cerebral vascular EC. Brain Res. 1991, 566: 325 - 328137. McCarron R.M., Spatz M., Kempski 0., Hogan R.N., Muehl L., McFarlin D.E.Interaction between myelin basic protein sensitized T lymphocytes and murine cerebralvascular EC. J. Immunol. 1986, 137: 3428 - 3435138. Traugott U., Raine C.S. Evidence for antigen presentation in situ by EC andastrocytes. J. Neurol. Sci. 1985, 69: 365 - 370139. Sobel R.A., Natale J.M., Schneeberger E.E. The immunopathology of acuteexperimental allergic encephalomyelitis. J. Neuropath. Exp. Neurol. 1987, 46: 239 -249140. Craggs R.I., Webster H. deF. Ia Ags in the normal rat nervous system and in lesions ofexperimental allergic encephalomyelitis. Acta Neuropath. (Ben) 1985, 68: 263 - 272141. Sakai K., Tabira T., Endoh M., Steinman L. Ia expression in chronic relapsingexperimental allergic encephalomyelitis induced by long-term cultured T cell lines inmice. Lab. Invest. 1986, 54: 345 - 352142. Male D., Pryce G. Kinetics of MHC gene expression and mRNA synthesis in brainendothelium. Immunology 1988, 63: 37 -42143. Rosa F., Hatat D., Abadie A., Wallach D., Revel M., Fellows M. Human interferonsenhance HLA-DR mRNA. EMBO J. 1983, 2: 1585 - 1589144. Pober J.S., Collins T., Gimbrone M.A. Jr., Libby P., Reiss C.S. Inducible expressionof class II major histocompatibility complex antigens and the immunogenicity of224vascular endothelium. Transplant 1986, 41: 141 - 146145. Thornhill M.H., Williams D.M., Speight P.M. Enhanced adhesion of autologouslymphocytes to gamma-interferon-treated human EC in vitro. Br. J. Exp. Path. 1989,70: 59 - 64146. Royer H.D., Campen T.J., Ramarli D., Chang H.C., Acuto 0., Reinherz E.L.Molecular aspects of human T-lymphocyte antigen recognition. Transplantation 1985,39: 571 - 582147. Doyle C., Strominger J.L. Interaction between CD4 and class II MHC moleculesmediates cell adhesion. Nature 1987, 330: 256 - 259148. Steinman L., Solomon D., Lim M., Zamvil S., Sriram S. Prevention of experimentalallergic encephalitis with in vivo administration of anti Ia antibody. Decreasedaccumulation of radiolabeled lymph node cells in the central nervous system. J.Neuroimmunol. 1983, 5: 91 - 97149. Astrom K.E., Webster H. deF., Arnason B.G. The initial lesion in experimentalallergic neuritis. A phase and electron microscopic study. J. Exp. Med. 1968, 128:469 - 495150. Haskard D., Cavender D., Ziff M. Phorbol ester-stimulated T lymphocytes showenhanced adhesion to human EC monolayers. J. Immunol. 1986, 137: 1429- 1434151. Oppenheimer-Marks N., Davis L.S., Lipsky P.E. Human T lymphocyte adhesion toEC and transendothelial migration. Alteration of receptor use relates to the activationstatus of both the T cell and the EC. J. Immunol. 1990, 145: 140 - 148152. Oppenheimer-Marks N., Davis L.S., Bogue D.T., Ramberg J., Lipsky P.E. Differentialutilization of ICAM-1 and VCAM-1 during the adhesion and transendothelialmigration of human T lymphocytes. J. Immunol. 1991, 147: 2913 - 2921153. Kavanaugh A.F., Lightfoot E., Lipsky P.E., Oppenheimer-Marks N. Role ofCD11/CD18 in adhesion and transendothelial migration of T cells. Analysis utilizingCD18-deficient T cell clones. J. Immunol. 1991, 146: 4149 - 4156154. Male D., Pryce G., Hughes C., Lantos P. Lymphocyte migration into brain modelledin vitro: Control by lymphocyte activation, cytokines, and antigen. Cell. Immunol.1990, 127: 1- 11155. Male D., Pryce G., Rahman J. Comparison of the immunological properties of ratcerebral and aortic endothelium. J. Neuroimmunol. 1990, 30: 161 - 168156. Wang Y.F., Calder V.L., Greenwood J., Lightman S.L. Lymphocyte adhesion tocultured EC of the blood-retinal barrier. J. Neuroimmunol. 1993, 48: 161 - 168157. Damle N.K., Doyle L.V., Bender J.R., Bradley E.C. Interleukin-2 activated humanlymphocytes exhibit enhanced adhesion to normal vascular EC and cause their lysis. J.Immunol. 1987, 138: 1779 - 1785158. Pankonin G., Reipert B., Ager A. Interactions between interleukin-2-activated225lymphocytes and vascular endothelium: binding to and migration across specializedand non-specialized endothelia. Immunol. 1992, 77: 51 - 60159. Dustin M.L., Springer T.A. T cell receptor cross-linking transiently stimulatesadhesiveness through LFA-1. Nature 1989, 341: 619 - 624160. Tsukada N., Matsuda M., Miyagi K., Yanagisawa N. Adhesion of cerebral EC tolymphocytes from patients with multiple sclerosis. Autoimmunity 1993, 14: 329 - 333161. Hickey W.F., Hsu B.L., Kimura H. T-lymphocyte entry into the Central NervousSystem. J. Neurosci. Res. 1991, 28: 254 - 260162. Springer T.A. Adhesion receptors of the immune system. Nature 1990, 346: 425 -434163. Parrott D.M.V., Wilkinson P.C. Lymphocyte locomotion and migration. Prog.Allergy 1981, 28: 193 - 284164. Traugott U., Stone S.H., Raine C.S. Experimental allergic encephalomyelitis.Migration of early T cells from the circulation into the central nervous system. J.Neurol. Sci. 1978, 36: 55 - 61165. Traugott U., Stone S.H., Raine C.S. Chronic relapsing experimental allergicencephalomyelitis. J. Neurol. Sci. 1979, 41: 17 - 29166. Kateley J.R., Bazzell S.J. Immunological dysfunctions in multiple sclerosis. I.Diminution of “active” thymus-derived lymphocytes and presence ofimmunomodulating serum factors. Clin. Exp. Immunol. 1979, 35: 218 - 226167. Traugott U., Scheinberg L.C., Raine C.S. Multiple sclerosis: Circulating antigen-reactive lymphocytes. Ann. Neurol. 1979, 6: 425 - 429168. Kam-Hansen S. Reduced number of active T cells in cerebrospinal fluid in multiplesclerosis. Neurology 1979, 29: 897 - 899169. Schluesener H.J., Sobel R.A., Weiner H.L. Demyelinating experimental allergicencephalomyelitis (EAE) in the rat: treatment with a monoclonal antibody againstactivated T cells. J. Neuroimmunol. 1988, 18: 341 - 351170. Greenwood J., Calder V.L. Lymphocyte migration through cultured EC monolayersderived from the blood-retinal barrier. Immunology 1993, 80: 401 - 406171. Jaffe E.A. Synthesis of von Willebrand factor by EC. In Una S. Ryan, ed. EC. Vol. I.Boca Raton, Fl: CRC Press, 1988: 119 - 126172. Rand J.H., Gordon R.E., Sussman 1.1., Chu S.V., Solomon V. Electron microscopiclocalization of factor-Vill-related antigen in adult human blood vessels. Blood 1982,60: 627 - 634173. Sakariassen K.S., Bolhuis P.A., Sixma J.J. Human blood platelet adhesion to arterysubendothelium is mediated by Factor Vill-von Willebrand factor bound to thesubendothelium. Nature 1979, 279: 636 - 638226174. Reinders H.J., DeGroot P.G., Gonsalves M.D., Zandbergen J., Loesberg C., VanMourik J.A. Isolation of a storage and secretory organelle containing von Willebrandprotein from cultured human EC. Biochimica et Biophysica Acta 1984, 804: 361 - 369175. Wagner D.D., Olmsted J.B., Marder V.J. Immunolocalization of von Willebrandprotein in Weibel-Palade bodies of human EC. J. Cell. Biol. 1982, 95: 355 - 360176. Warhol M.J., Sweet J.M. The ultrastructural localization of von Willebrand factor inEC. Am. J. Pathol. 1984, 117: 310 - 315177. Weibel E.R., Palade G.E. New cytoplasmic components in arterial endothelia. J. Cell.Biol. 1964, 23: 101 - 112178. DeBault L.E., Henriquez E., Hart M.N., Cancilla P.A. Cerebral microvessels andderived cells in tissue culture: II. Establishment, identification and preliminarycharacterization of an EC line. In Vitro 1981, 17: 480 - 494179. Bowman P.D., Betz A.L., Ar D., Wolinsky J.S., Penney J.B., Shivers R.R., GoldsteinG.W. Primary culture of capillary endothelium from rat brain. In Vitro 1981, 17: 353 -362180. Diglio C., Grammas P., Giacomelli F., Wiener J. Primary culture of rat cerebralmicrovascular EC. Lab. Invest. 1982, 46: 554 - 563181. Spatz M., Bembry J., Dodson R.F., Hervonen H., Murray M.R. EC cultures derivedfrom isolated cerebral microvessels. Brain Res. 1980, 191: 577 - 582182. Tannenbaum S.H., Gralnick H.R. y-Interferon modulates von Willebrand factor releaseby cultured human EC. Blood 1990, 75: 2177 - 2184183. Bowman P.D., Betz A.L., Ar D., Wolinsky J.S., Shivers R.R., Goldstein G.W. Primaryculture of capillary endothelium from rat brain. In Vitro Cell Dev. Biol. 1981, 17: 353 -362184. McLean I.W., Nakane P.K. Periodate-Lysine-Paraformaldehyde fixative: A newfixative for immunoelectron microscopy. J. Histochem. Cytochem. 1974, 22: 1077 -1083185. Schroeter D., Spiess E., Paweletz N., Benke R. A procedure for rupture-freepreparation of confluently grown monolayer cells for scanning electron microscopy. J.Electron Microsc. Tech. 1984, 1: 219 - 225186. Dorovini-Zis K., Bowman P.D., Betz A.L., Goldstein G.W. Hyperosmotic ureareversibly opens the tight junctions between brain capillary EC in cell culture. J.Neuropath. Exp. Neurol. 1987, 46: 130 - 140187. Julius M.H., Simpson E., Hertzenberg L.A. A rapid method for the isolation offunctional thymus-derived murine lymphocytes. Eur. J. Immunol. 1973, 3: 645 - 649188. Hormia M., Lehto V.P., Virtanen I. Identification of UEA I binding surface227glycoproteins of cultured human EC. Cell Biol. mt. Rep. 1983, 7: 467 - 475189. Weber T., Seitz R.J., Liebert U.G., Gallasch B., Wechsler W. Affinity cytochemistry ofvascular endothelia in brain tumors by biotinylated Ulex europaeus type I lectin (UEAI). Acta Neuropathol. 1985, 67: 128 - 135190. Sobel R.A., Ames M.B. Major histocompatibility complex molecule expression in thehuman central nervous system: immunohistochemical analysis of 40 patients. J.Neuropath. Exp. Neurol. 1988, 47: 19 - 28191. Lampson L.A., Hickey W.F. Monoclonal antibody analysis of MHC expression inhuman brain biopsies: tissue ranging from “histologically normal” to that showingdifferent levels of glial tumor involvement. J. Immunol. 1986, 136: 4054 - 4062192. Frank E., Pulver M., de Tribolet N. Expression of Class II Major Histocompatibilityantigens on reactive astrocytes and EC within the gliosis surrounding metastases andabscesses. J. Neuroimmunol. 1986, 12: 29- 36193. Pober J.S., Gimbrone M.A. Jr. Expression of Ia-like antigens by human vascular EC isinducible in vitro: Demonstration by monoclonal antibody binding andimmunoprecipitation. Proc. Natl. Acad. Sci. USA 1982, 79: 6641 - 6645194. Wagner C.R., Vetto R.M., Burger D.R. Expression of I-region-associated antigen (Ia)and interleukin 1 by subcultured human BC. Cell. Immunol. 1985, 93: 91 - 104195. Reddy P.G., Graham G.M., Datta S., Guarini L., Moulton T.A., Jiang H., GottesmanM.M., Ferrone S., Fisher P.B. Effect of recombinant fibroblast interferon andrecombinant immune interferon on growth and the antigenic phenotype of multidrugresistant human glioblastoma multiforme cells. J. Natl. Cancer Inst. 1991, 83: 1307 -1315196. Wilcox C.E., Baker D., Butter C., Willoughby D.A., Turk J.L. Differential expressionof guinea pig class II major histocompatibility complex antigens on vascular EC invitro and in experimental allergic encephalomyelitis. Cell. Immunol. 1989, 120: 82 -91197. Beilke M.A., Riding In D., Hamilton R., Stone G.A., Jordan E.K., Brashears G.,Nusbaum W., Huddleston D., Gibbs C.J. Jr, Gravell M. HLA-DR expression inmacaque neuroEC in vitro and during SIV encephalitis. J. Neuroimmunol. 1991, 33:129 - 143198. Amaldi I., Reith W., Berte C., Mach B. Induction of HLA class II genes by IFNgamma is transcriptional and requires a trans acting protein. J. Immunol. 1989, 142:999 - 1004199. Traugott U. Multiple sclerosis: relevance of class I and class II MHC-expressing cellsto lesion development. J. Neuroimmunol. 1987, 16: 283 - 302200. Steinman L., Waldor M.K., Zamvil S.S., Lim M., Herzenberg L, McDevitt H.O.,Mitchell D., Sriram S. Therapy of autoimmune disease with antibody to immuneresponse gene products or to T-cell surface markers. Ann. N.Y. Acad. Sci. 1986, 475:274 - 284228201. Heyns A du P., Eldor A., Vlodavsky I., Kaiser N., Fridman R., Panet A. Theantiproliferative effect of interferon and the mitogenic activity of growth factors areindependent cell cycle events. Exp. Cell. Res. 1985, 161: 297 - 306202. Friesel R., Komoriya A., Maciag T. Inhibition of EC proliferation by gammaInterferon. J. Cell. Biol. 1987, 104: 689 - 696203. Saegusa Y., Ziff M., Welkovich L., Cavender D. Effect of inflammatory cytokines onhuman EC proliferation. J. Cell. Physiol. 1990, 142: 488 - 495204. Maheshwari R.K., Srikantan V., Bhartiya D., Kleinman H.K., Grant D.S. Differentialeffects of interferon gamma and alpha on in vitro model of angiogenesis. J. Cell.Physiol. 1991, 146: 164 - 169205. Tsuruoka N., Sugiyama M., Tawaragi Y., Tsujimoto M., Nishihara T., Goto T., Sato N.Inhibition of in vitro angiogenesis by lymphotoxin and interferon-y. Biochem. Biophys.Res. Commun. 1988, 155: 429 - 435206. Numa Y., Kawamoto K., Sakai N., Matsumura H. Flow cytometric analysis ofantineoplastic effects of interferon-ct, 1 and y labeled with fluorescein isothiocyanateon cultured brain tumors. J. Neuro-Oncology 1991, 11: 225 - 234207. Palmer H., Libby P. Interferon-3. A potential autocrine regulator of human vascularsmooth muscle cell growth. Lab. Invest. 1992, 66: 715 - 721208. Brett J., Gerlach H., Nawroth P., Steinberg S., Godman G., Stern D. Tumor necrosisfactor/cachectin increases permeability of EC monolayers by a mechanism involvingregulatory G proteins. J. Exp. Med. 1989, 169: 1977 - 1991209. Hirano A., Dembitzer H.M., Becker N.H., Levine S., Zimmerman H.M. Fine structuralalterations of the blood-brain barrier in experimental allergic encephalomyelitis. J.Neuropath. Exp. Neurol. 1970, 29: 432 - 440210. Kristensson K., Wisniewski H.M. Chronic relapsing experimental allergicencephalomyelitis. Studies in vascular permeability changes. Acta Neuropath. 1977,39: 189 - 194211. Claudio L., Kress Y., Factor J., Brosnan C.F. Mechanisms of edema formation inexperimental autoimmune encephalomyelitis. Am. J. Path. 1990, 137: 1033 - 1045212. Pryce G., Male D., Sedgwick J. Antigen presentation in brain: brain EC are poorstimulators of T - cell proliferation. Immunology 1989, 66: 207 - 212213. St. Louis J.D., Lederer J.A., Lichtman A.H. Costimulator deficient antigenpresentation by an EC line induces a nonproliferative T cell activation responsewithout anergy. J. Exp. Med. 1993, 178: 1597 - 1605214. Boussiotis V.A., Freeman G.J., Gray G., Gribben J., Nadler L.M. B7 but notIntercellular adhesion molecule-i costimulation prevents the induction of humanalloantigen-specific tolerance. J. Exp. Med. 1993, 178: 1753 - 1763215. Wu Y., Guo Y., Liu Y. A major costimulatory molecule on antigen-presenting cells,229CTLA4 ligand A, is distinct from B7. J. Exp. Med. 1993, 178: 1789 - 1793216. Lin H., Boiling S.F., Linsley P.S., Wei R., Gordon D., Thompson C.B., Turka L.A.Long-term acceptance of major histocompatibility complex mismatched cardiacallografts induced by C1’LA4Ig plus donor-specific transfusion. J. Exp. Med. 1993,178: 1801 - 1806217. Borst J., Alexander S., Elder J., Terhorst C. The T3 complex on human T lymphocytesinvolves four structurally distinct glycoproteins. J. Biol. Chem. 1983, 258: 5135 -5141218. Van Wauwe J.P., de Mey J.R., Goossens J.G. OKT3: A monoclonal anti-human Tlymphocyte antibody with potent mitogenic properties. J. Immunol. 1980, 124: 2708 -2713219. Chang T.W., Kung P.C., Gingras S.P., Goldstein S. Does OKT3 monoclonal antibodyreact with an antigen-recognition structure on human T cells? Proc. Natl. Acad. Sci.USA 1981, 78: 1805 - 1808220. Imboden J.B., Stobo J.D. Transmembrane signalling by the T-cell antigen receptor. J.Exp. Med. 1985, 161: 446 - 456221. Imboden J.B., Weiss A., Stobo J.D. Transmembrane signalling by the T3-antigenreceptor complex. Immunol. Today 1985, 6: 328 - 331222. Hara T., Fu S.M. Human T-cell activation. I. Monocyte-independent activation andproliferation induced by anti-T3 monoclonal antibodies in the presence of tumorpromoter 12-0-tetradecanoyl Phorbol-13-acetate. J. Exp. Med. 1985, 161: 641 - 656223. Ledbetter J.A., Parsons M., Martin P.J., Hansen J.A., Rabinovitch P.S., June C.H.Antibody binding to CD5 (Tp67) and Tp44 T-cell surface molecules: effects on cyclicnucleotides, cytoplasmic free calcium, and cAMP-mediated suppression. J. Immunol.1986, 137: 3299 - 3305224. Novak T.J., Rothenberg E.V. cAMP inhibits induction of interleukin-2 but not ofinterleukin-4 in T cells. Proc. Nat!. Acad. Sci. USA 1990, 87: 9353 - 9357225. Anastassiou E.D., Paliogianni F., Balow J.P., Yamada H., Boumpas D.T.Prostaglandin E2 and other cyclic AMP-elevating agents modulate IL-2 and IL-2Ragene expression at multiple levels. J. Immunol. 1992, 148: 2845 - 2852226. Keren D.F. Flow cytometry in clinical diagnosis. American Society of ClininicalPathologists Press, Chicago, 1989227. Platts K.E., Lawry J., Hancock B.W., Rees R.C. Phenotypic and cell cycle analysis ofhuman peripheral blood monocuclear cells activated with Interleukin-2 and/or OKT3.Exp. Cell Res. 1993, 208: 154 - 160228. Tsoukas C.D., Landgraf B., Bentin J., Valentine M., Lotz M., Vaughan J.H., CarsonD.A. Activation of resting T lymphocytes by anti-CD3 (T3) antibodies in the absenceof monocytes. J. Immunol. 1985, 135: 1719 - 1723230229. Hughes C.C.W., Male D.K., Lantos P.L. Adhesion of lymphocytes to cerebralmicrovascular cells: effects of Interferon-y, tumor necrosis factor and interleukin-1.Immunology 1988, 64: 677 - 681230. Liversidge J., Sewell H.F., Forrester J.V. Interactions between lymphocytes and cellsof the blood-retina barrier: mechanisms of T lymphocyte adhesion to human retinalcapillary EC and retinal pigment epithelial cells in vitro. Immunol. 1990, 71: 390 -396231. Curtis A.S.G. The H-2 histocompatibility system and lymphocyte adhesion.Interaction modulation factor involvement. J. Immunogenet. 1979, 6: 155 - 166232. Curtis A.S.G., Rooney P. H-2 restriction of contact inhibition of epithelial cells.Nature 1979, 281: 222 - 223233. Thornhill M.H., Speight P.M., Williams D.M. A monoclonal antibody to CD4 inhibitsinterferon-gamma (IFN-y) enhanced ahdesion of autologous lymphocytes to EC. J.Pathol. 1987, 151: 26A234. Sriram S., Topham D.J., Carroll L. Haplotype-specific suppression of experimentalallergic encephalomyelitis with anti-Ia antibodies. J. Immunol. 1987, 139: 1485 -1489235. Sriram S., Carroll L. Haplotype-specific inhibition of homing of radiolabeledlymphocytes in Experimental Allergic Encephalomyelitis following treatment withAnti-Ia Antibodies. Cell. Immunol. 1991, 135: 222 - 231236. McCarron R.M., Wang L., Racke M.K., McFarlin D.E., Spatz M. Cytokine-regulatedadhesion between encephalitogenic T lymphocytes and cerebrovascular EC. J.Neuroimmunol. 1993, 43: 23 - 30237. Cavender D.E., Haskard D.O., Joseph B., Ziff M. Interleukin-1 increases the bindingof human B and T lymphocytes to EC monolayers. J. Immunol. 1986, 136: 203 - 207238. Haskard D.O., Cavender D.E., Fleck R.M., Sontheimer R., Ziff M. Human dermalmicrovascular EC behave like umbilical vein EC in T cell adhesion studies. J. Invest.Dermatol. 1987, 88: 340 - 344239. Wong D., Dorovini-Zis K. Upregulation of intercellular adhesion molecule-i (ICAM1) expression in primary cultures of human brain microvessel EC by cytokines andlipopolysaccharide. J. Neuroimmunol. 1992, 39: 11 - 22240. Haskard D., Cavender D., Beatty P., Springer T., Ziff M. T lymphocyte adhesion toEC: Mechanisms demonstrated by anti-LFA-i monoclonal antibodies. J. Immunol.1986, 137: 2901 - 2906241. van Kooyk Y., Weder P., Hogervorst F., Verhoeven A.J., van Seventer G., te VedeA.A., Borst J., Keizer G.D., Figdor C.G. Activation of LFA-1 through a Ca -dependent epitope stimulates lymphocyte adhesion. J. Cell. Biol. 1991, 112: 345 - 354242. van Kooyk Y., van de Wiel-van Kemenade E., Weder P., Huijbens R.J.F., Figdor C.G.Lymphocyte function-associated antigen 1 dominates very late antigen-4 in binding of231activated T cells to endothelium. J. Exp. Med. 1993, 177: 185 - 190243. Dustin M.L., Springer T.A. Lymphocyte function-associated antigen-i (LFA-1)Interaction with Intercellular Adhesion Molecule-i (ICAM-i) is one of at least threemechanisms for lymphocyte adhesion to cultured EC. J. Cell. Biol. 1988, 107: 321 -331244. Shimizu Y., Newman W., Gopal T.V., Horgan K.J., Graber N., Beall L.D., vanSeventer G.A., Shaw S. Four molecular pathways of T cell adhesion to EC: Roles ofLFA-i, VCAM-i, and ELAM-1 and changes in pathway hierarchy under differentactivation conditions. J. Cell. Biol. 1991, 113: 1203 - 1212245. Elices M.J., Osbom L., Takada Y., Crouse C., Luhowskyj S., Hemler M.E., Lobb R.R.VCAM-i on activated endothelium interacts with the leukocyte integrin VLA-4 at asite distinct from the VLA-4/fibronectin binding site. Cell 1990, 60: 577 - 584246. Osborn L., Hession C., Tizard R., Vassallo C., Luhowskyj S., Chi-Rosso G., Lobb R.R.Direct expression cloning of vascular cell adhesion molecule-i (VCAM-i), acytokine-induced endothelial protein that binds to lymphocytes. Cell 1989, 59: 1203 -1211247. Wellicome S.M., Thornhill M.H., Pitzalis C., Thomas D.S., Lanchbury J.S.S., PanayiG.S., Haskard D.O. A monoclonal antibody that detects a novel antigen on EC that isinduced by tumor necrosis factor, IL-i or lipopolysaccharide. J. Immunol. 1990, 144:2558 - 2565248. Rice G.E., Munro J.M., Bevilacqua M.P. Inducible cell adhesion molecule 110(INCAM-ilO) is an endothelial receptor for lymphocytes: a CDii/CD18-independentadhesion mechanism. J. Exp. Med. 1990, 171: 1369 - 1374249. Wong D., Dorovini-Zis K. Expression of VCAM-1 and E-selectin by human brainmicrovessel EC in vivo and in vitro. Can. J. Neurol. Sci. 1993, 20: Si04250. Pryce G., Male D.K., Sarkar C. Control of lymphocyte migration into brain: selectiveinteractions of lymphocyte subpopulations with brain endothelium. Immunol. 1991,72: 393 - 398251. Male D., Pryce G., Linke A., Rahman J. Lymphocyte migration into the CNSmodelled in vitro. J. Neuroimmunol. 1992, 40: 167 - 172252. Haskard D.O., Strobel S., Thornhill M., Pitzalis C., Levinsky R.J. Mechanisms oflymphocyte adhesion to EC: studies using an LFA-i-deficient cell line. Immunol.1989, 66: 111 - 116153. Naparstek Y., Cohen I.R., Fuks Z., Vlodavsky I. Activated T lymphocytes produce amatrix degrading heparan sulphate endoglycosidase. Nature 1984, 310: 241 - 244254. Lider 0., Baharav E., Mekori Y.A., Miller T., Naparstek Y., Vlodavsky I., Cohen I.R.Suppression of experimental autoimmune diseases and prolongation of allograftsurvival by treatment of animals with low doses heparins. J. Clin. Invest. 1989, 83:752 - 756232255. Muller W.A., Weigi S.A., Deng X., Phillips D.M. PECAM-1 is required fortransendothelial migration of leukocytes. J. Exp. Med. 1993, 178: 449 - 460256. Newman P.J., Berndt M.C., Gorski J., White II G.C., Lyman S., Paddock C., MullerW.A. PECAM-1 (CD31) cloning and relation to adhesion molecules of theimmunoglobulin gene superfamily. Science (Wash. D.C.) 1990, 247: 1219 - 1222257. Muller W.A., Ratti C.M., McDonnell S.L., Cohn Z.A. A human EC-restricted,externally disposed plasmalemmal protein enriched in intercellular junctions. 3. Exp.Med. 1989, 170: 399 - 414258. Ohto H., Maeda H., Shibata Y., Chen R.F., Ozaki Y., Higashihara M., Takeuchi A.,Tohyama H. A novel leukocyte differentiation antigen: two monoclonal antibodiesTM2 and TM3 define a 120-kd molecule present on neutrophils, monocytes, platelets,and activated lymphoblasts. Blood 1985, 66: 873 - 881259. Goyert S.M., Ferrero E.M., Seremetics S.V., Winchester R.J., Silver J., Mattison A.C.Biochemistry and expression of myelomonocytic antigens. J. Immunol. 1986, 137:3909 - 3914260. Stockinger H., Gadd S.J., Eher R., Majdic 0., Schreiber W., Kasinrerk W., Strass B.,Schnabl E., Knapp W. Molecular characterization and functional analysis of theleukocyte surface protein CD 31. J. Immunol. 1990, 145: 3889 - 3897261. Perkett E.A., Disabato G., Brigham K.L., Meyrick B. Lymphocyte and granulocytemigration across the endothelial layer of bovine pulmonary artery intimal explantstowards lymphocyte conditioned medium. Tissue and Cell 1986, 18: 839 - 852262. Bowman P.D., du Bois M., Dorovini-Zis K., Shivers R.R. Microvascular EC frombrain. In H.M. Piper, ed. Cell culture techniques in heart and vessel research. Springer-Verlag, NY, 1990: 140 - 157263. Buzney S.M., Massicotte S.J. Retinal vessels: proliferation of endothelium in vitro.Invest. Opthalmol. Visual. Sci. 1979, 18: 1191 - 1195264. Phillips P., Kumar P., Kumar S., Waghe M. Isolation and characterization of EC fromrat and cow brain white matter. J. Anat. 1979, 129: 261 - 272265. Hirano A., Ghatak N.R., Becker N.H., Zimmerman H.M. A comparison of the finestructure of small blood vessels in intracranial and retroperitoneal malignantlymphomas. Acta. Neuropath. 1974, 27: 93 - 104266. Hirano A., Matsui T. Vascular structures in brain tumors. Human Path. 1975, 6: 611 -621267. Herrlinger H., Anzil A.P., Blinzinger K., Kronski D. Endothelial microtubular bodiesin human brain capillaries and venules. J. Anat. 1974, 118: 205 - 209268. Kumar P., Kumar S., Marsden H.B., Lynch P.G., Earnshaw E. Weibel-Palade bodies inEC as a marker for angiogenesis in brain tumors. Cancer Res. 1980, 40: 2010 - 2019269. Pavelka M. Functional morphology of the Golgi apparatus. Advances in anatomy,233embryology and cell biology 1987, 106: 1 -94270. Griffiths G., Pfeiffer S., Simons K., Matlin K. Exit of newly synthesized membraneproteins from the trans cisterna of the Golgi complex to the plasma membrane. J. Cell.Biol. 1985, 101: 949 - 964271. Handin R.I., Wagner D.D. The molecular and cellular biology of von Willebrandfactor. Prog. Hemostasis Thromb. 1989, 9: 233 - 259272. Loesberg C., Gonsalves M.D., Zandbergen J., Willems C., Van Aken W.G., Stel H.V.,Van Mourik J.A., DeGroot P.G. The effect of calcium on the secretion of factor-Villrelated antigen by cultured human EC. Biochim. Biophys. Acta 1983, 763: 160- 168273. Sporn L.A., Marder V.J., Wagner D.D. Inducible secretion of large, biologically potentvon Willebrand factor multimers. Cell 1986, 46: 185 - 190274. Sporn L.A., Marder V.J., Wagner D.D. Differing polarity of the constitutive andregulated secretory pathways for von Willebrand factor in EC. J. Cell. Biol. 1989, 108:1283 - 1289275. Sinha S., Wagner D.D. Intact microtubules are necessary for complete processing,storage and regulated secretion of von Willebrand factor by EC. Eur. J. Cell. Biol.1987, 43: 377 - 383276. Edgell C-JS., Haizlip J.E., Bagnell C.R., Packenham J.P., Harrison P., Wilbourn B.,Madden V.J. Endothelium specific Weibel-Palade bodies in a continuous human cellline, EA. hy 926. In Vitro Cell Dev. Biol. 1990, 26: 1167- 1172277. Renkonen R., Mennander A., Ustinov J., Mattila P. Activation of protein kinase C iscrucial in the regulation of ICAM-1 expression on EC by interferon-y. Internat.Immunol. 1990, 2: 719 - 724278. Mason D.W., Charleton H.M., Jones A.J., Lavy C.B.D., Puklavek M., Simmonds S.J.The fate of allogeneic and xenogeneic neuronal tissue transplanted into the thirdventricle of rodents. Neuroscience 1986, 19: 685 - 694279. Cross A.H., Cannella B., Brosnan C.F., Raine C.S. Homing to Centraj NervousSystem vasculature by antigen-specific lymphocytes. I. Localization of 1 C-labeledcells during acute, chronic, and relapsing experimental allergic encephalomyelitis.Lab. Invest. 1990, 63: 162 - 170280. Caspi R.R., Chan C., Fujino Y., Najafian F., Grover S., Hansen C.T., Wilder R.L.Recruitment of antigen-non specific cells plays a pivotal role in the pathogenesis of a Tcell-mediated organ-specific autoimmune disease, experimental autoimmuneuveoretinitis. J. Neuroimmunol. 1993, 47: 177- 188281. The IFN-3 Multiple Sclerosis Study Group. Interferon beta-lb is effective in relapsingremitting multiple sclerosis. I. Clinical results of a multicenter, randomized, doubleblind, placebo-controlled trial. Neurology 1993, 43: 655 - 661


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items