. 5,. mg {A ‘ . . i1: ,. z 43%, ; . .4 ‘ ‘ a. .5; a x45 a mum adv . i 3.. i 2:111 l liyr mawafiafiifigwg? E L BEARIES |H willlfii'lllllllMinimum 31293017721519 _ LIBRARY Michigan State Unlverslty This is to certify that the dissertation entitled THE EFFECT OF HUMAN HERPESVIRUS TYPE-6 ON THE HEMOSTATIC COMPONENTS OF ENDOTHELIAL CELLS presented by Walid T. Khalife has been accepted towards fulfillment of the requirements for _Eh_._D_.__ degree in MM Major professor Date M M; Ma? MS U i: an Affirmative Action/Equal Opportunity Institution 0- 12771 PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINE return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 1M m“ THE EFFECT OF HUMAN HERPESVIRUS TYPE-6 ON THE HEMOSTATIC COMPONENTS OF ENDOTHELIAL CELLS By Walid T. Khalife A DISSERTATION Submitted to Michigan State University in partial fulfillment ofthe requirements for the degree of DOCTOR OF PHILOSOPHY Department of Physiology 1998 ABSTRACT THE EFFECT OF HUMAN HERPESVIRUS TYPE-6 ON THE HEMOSTATIC COMPONENTS OF ENDOTHELIAL CELLS By Walid T. Khalife Herpesviruses have been proposed as potential initiators of vascular injury, based on studies performed in the late 1970's, when it was demonstrated that avian herpesvirus could induce atherosclerosis in chickens similar to human atherosclerosis (68). Since then, other animal models have been used. Herpesviruses have been linked to thrombosis and atherosclerosis through epidemiological data. Herpesviruses have been associated with accelerated atherosclerosis in heart transplant recipients and with restenosis after angioplasty. Atherosclerotic lesions have been reported to contain herpesviral genomic material and this latent viral infection may be an atherogenic trigger. In this thesis, it is shown that infection of human umbilical vein endothelial cells (HUVECs)with human herpesvirus type-6 (HHV-6) appears to modify the membrane conformation where the prothrombinase complex assembly is formed. In HHV-6 infected HUVECS, the prothrombinase complex formation is substantiated almost immediately by an increase in procoagulant activity. It is also supported by an increase in clotting time in factor X and factor V deficient plasma compared with the fresh frozen plasma (FFP) control. A higher procoagulant activity leads to an increase in thrombin generation that will induce platelet aggregation and adhesion on the surface of HUVECS. We have demonstrated that the percent of platelet binding in thrombin stimulated HUVECs is significantly higher in HHV- 6 infected cells than in non-infected endothelium. Using immunologic and functional assays, we have demonstrated that HHV-6 increased the antigen and activity level of plasminogen activator inhibitor-1 (PAI-l) in a dose-and -time dependent manner. But, it did not have any effect on tissue-plasminogen activator (t-PA). Northern blot analysis using PAI- 1 and t-PA cDNA probes indicated that HHV—6 infection of HUVECS increased the levels of the 3.2 and 2.3Kb PAI-l mRNA forms with a preferential increase in the 3.2 Kb form. We conclude that HHV-6 infection of endothelial cells shifts the property ofthe endothelium from anticoagulant to procoagulant, by enhancing thrombin generation and fibrin formation, increasing platelet binding, and by increasing PAI- 1 generation without any change in t-PA. These changes tend to shift the dynamic hemostatic balance toward a procoagulant condition that could lead to excess fibrin formation and deposits on the endothelial surface. These findings are consistent with the thrombogenic theory of atherogenesis and with the finding of accelerated atherosclerosis in herpesvirus infected chickens. To my mom Marie, my wife Vivian, my Sisters, and the rest ofthe family, with love. To the memory of my dad Tanios. iv ACKNOWLEDGMENTS I would like to acknowledge the advice, support, and scientific expertise of my advisor, Dr. Gerald Davis. Thank you Dr. Davis for the continued encouragement as well as the opportunity to do this research in your laboratory. I would like to express my sincere and profound gratitude to the members of my guidance committee, Dr. Douglas Estry, Dr. Seth Hootman, Dr. Donald Jump, Dr. Leonel Mendoza, and Dr. William Spielman for their support and input during my doctoral work. To Dr. Donald Jump, thank you for your support, guidance, and for having your laboratory available for me. To Dr. Annette Thelen, I am sincerely grateful for the input and knowledge that you have shared with me. I am indebted to Dr. Thelen and to Michelle Mater for helping me in Dr. Jump’s laboratory. I am very grateful for everyone at Saint Lawrence especially Kathleen for helping us. To my dad Tanios, you are always with us. We love you very much. To my mom Marie, you are one ofa kind. Your blessing and prayers helped me tremendously, especially when things got tough. To my wife Vivian, a special thank you for your understanding, care, love, and for helping me type this dissertation. To my sister Ghada, what can I say? Forever and ever, you have a precious place in my heart. To my sisters Natalie, Rona, and Nada, I am the luckiest for having sisters as wonderful as you. To my aunt Julia, thank you for your guidance and for having your house as a second home for us. To the rest ofthe family, your constant love and support helped me more than you can imagine. Thank you for believing in me. TABLE OF CONTENTS Page LIST OF TABLES ................................................................. viii LIST OF FIGURES ................................................................. ix INTRODUCTION ................................................................... 1 LITERATURE REVIEW ........................................................... 4 Role of Endothelium in Hemostasis Blood Coagulation and Fibrinolysis Inhibition of Platelet Aggregation and Adhesion Anticoagulant system Fibrinolytic system Human Herpesvirus Type-6 Effect on Viruses on Atherosclerosis and Thrombosis MATERIALS and METHODS .................................................. 50 RESULTS ............................................................................ 79 DISCUSSION ..................................................................... 143 SUMMARY and CONCLUSION .............................................. 163 REFERENCES .................................................................... 168 vii LIST OF TABLES Table Page 1. Endothelial Pro- and Anti-Coagulant Factors ofHemostasis ............................................................... 11 2. Receptors on Platelet Surface and Their Agonists ................. 12 3. HHV—6 Titration Using Karber Method ............................... S7 4. Human Herpes Virus Type-6 Titration ................................ 89 5. Adherence of Platelets to Non-infected and 18 Hour-Infected HUVECS ........................................ 104 6. Modulation of t-PA and PAI-l Antigens Secretion .............. 107 LIST OF FIGURES Figure Page 1. The Hemostatic Mechanism ............................................. 5 2. Role ofthe Endothelium in Hemostasis .............................. 8 3. Metabolism OfArachidonic Acid in Endothelial Cells .......... 18 4. The Fibrinolytic System ............................................... 29 5. Cultured Human Umbilical Vein Endothelial Cells Stained with Wright’s Stain .......................................... 80 6. Cultured Human Umbilical Vein Endothelial Cells Tested for Factor VIII Antigen ...................................... 82 7. Cord Blood Lymphocytes, Six Days Post Infection ............. 84 8. Cord Blood Lymphocytes, 12 Days Post Infection .............. 86 9. Immunostaining for HHV-6 ........................................... 91 10. Effect of Thrombin on Platelet Adherence to HUVEC Under Static and Rocking Conditions ................... 93 11. Effect of Thrombin Concentration on Platelet Adhesion to Endothelial Cells ........................................ 96 12. Adherence of Platelets to Non-infected and 18 Hour- infected Human Umbilical Vein Endothelial Cells ........................................................ 99 13. Platelet Binding to HUVECS ........................................ 101 14. T-PA Standard Curve for t-PA and PAI-l Activity ............ 109 15. Effect ofHHV-6 on PAI-l Activity ............................... 112 16. Kinetics of HHV-6 on PAI-l Activity ............................. 114 17 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. PAI-l Standard Curve (ELISA) for the Measurement ofPAI-l Antigen ....................................................... 116 Kinetics ofHHV-6 on PAI-l Antigen Concentration ......... 118 t-PA Standard Curve (ELISA) for the Measurement oft-PA Antigen (Free and Bound) ................................. 120 Kinetics of HHV—6 on t-PA Antigen Concentration ........... 122 Purification oft-PA and PAI-l Plasmid DNA ................... 126 Effect ofHHV-6 on PAI-l mRNA in HUVECS .................. 128 Effect ofHHV-6 on t-PA mRNA in HUVECS .................... 130 Effect of HHV-6 Concentration on Procoagulant Activity ................................................................... 133 Kinetics of HHV—6 on Procoagulant Activity ................... 135 Effect of Heating HHV-6 on HUVECS Procoagulant Activity ................................................. 139 Procoagulant activity of HUVEC ................................... 141 aPC ATIII BSA CaCl2 cAMP CMV CPE CPSR dDAVP DMSO DNA ECGF FCS FFP HHV-6 HIV HSV-l HUVECS IIa NO List of Abbreviations Activated protein C Antithrombin III Bovine serum albumin Calcium chloride Cyclic adenosine monophosphate Cytomegalovirus Cytopathic effect Controlled process serum replacement 1-Desamino-8-D-arginine vasopressin Dimethyl sulfoxide Deoxyribonucleic acid Endothelial cells growth factor Fetal calf serum Fresh frozen plasma Human herpesvirus type-6 Human immunodeficiency virus Human herpesvirus type-1 Human umbilical vein endothelial cells Thrombin Nitric oxide PAI-l PBS PC PCR PDGF PFU PGE, PGI2 PHA PNPP PPP PRP VWF Plasminogen activator inhibitor Phosphate buffered saline Protein C Polymerase chain reaction Platelet derived growth factor Plaque forming unit Prostaglandin E1 Prostacyclin Phytohemagglutinin-P p-Nitrophenyl Phosphate Platelet poor plasma Platelet rich plasma Room temperature Tissue plasminogen activator 50% tissue culture infective dose Tissue factor pathway inhibitor Thrombomodulin Von Willebrand factor xii INTRODUCTION Injury to the blood vessel’s endothelium has been implicated in the development ofa variety of vascular disorders including atherosclerosis (10, 75, 178), disseminated intravascular coagulation (38, 264), and immune vasculitis (105, 215, 223, 237). Mechanical trauma (135, 252), hyper-cholesterolemia, hemodynamic stress (250), endotoxin (172), virus (275), thrombin (80), interleukin-1 (IL-1)(18), tumor necrosis factor (TNF)(160) and other agents have also been suggested to cause vascular disorders. It has been shown that one mechanism of vasculitis induction is from direct invasion and replication of viruses in the vascular endothelium (105, 215). A second mechanism ofinduction is by the formation of circulating immune complexes and the expression och and complement receptors (235, 237,244). Recent studies have detected the presence of viral antigen and viral DNA in atherosclerotic lesions of chickens with hyper- cholesteremia (39, 146), and of humans with atherosclerosis (16, 36, 86, 142).. Several other studies have suggested that herpesviruses are potential initiators of vascular injury (243, 274, 275). This beliefis based on studies done in the late 19705. These studies have shown that fowl herpesvirus could induce atherosclerosis in chickens (68). This has been coupled with epidemiological evidence linking herpesvirus and atherosclerosis, especially accelerated atherosclerosis in patients with heart transplants, and with restenosis after angioplasty. Multiple studies have demonstrated herpesviral antigens and nucleic acids in the atherosclerotic lesion. Viruses have been shown to infect endothelial cells in vitro. Based on those studies, we hypothesized that human herpesvirus type-6 (HHV-6) can infect human umbilical vein endothelial cells (HUVECS) and alter the hemostatic balance of endothelial cells toward hypercoagulation. The objectives of this research project were to: 1. Harvest cord blood lymphocytes 2. Prepare cell-free virus from supernatant tissue fluid ofHHV-6 infected cord blood lymphocytes, and to determine HHV- 6 titer. 3. Harvest endothelial cells from human umbilical vein endothelium. 4. Determine the cytopathogenicity of HHV-6 on HUVEC. 5. Isolate platelets from human peripheral blood. 6. Evaluate qualitatively and quantitatively the adherence of platelets to infected HUVECS in the presence and absence of thrombin. Study the effect ofHHV-6 on the fibrinolytic system by measuring HUVEC tissue-plasminogen activator (t-PA) and plasminogen activator inhibitor-1 (PAI-l) activities, concentrations, and mRNAs. Study the effect of HHV—6 on the coagulation system by determining the procoagulant activity ofHHV-6 infected HUVECS LITERATURE REVIEW The fluidity of blood is controlled by the hemostatic system that , at the right time, converts the blood from a fluid to a solid and back to a fluid state at the site of an injury (79). We can divide the hemostatic system into four families of proteins that are in part linked to thrombin (Figure 1): the coagulant/ anticoagulant families regulate clot formation, and the fibrinolytic/ anti-fibrinolytic families regulate clot lysis (56, 66,157). The vascular endothelium is located at the interface between tissue and blood. The endothelium continuously resists blood-born insults, particularly lipids, immune complexes, and microorganisms. One ofits major functions is to maintain an antithrombotic surface. Under normal conditions, the endothelial cells lining the blood vessels are anti-thrombotic (66, 283). This anti-thrombotic characteristic is achieved by a number of different properties ofthe endothelium including: the production of nitric oxide, prostacyclin, ecto-adenosine diphosphatase (ADPase) and cAMP. These compounds prevent Figure 1. The Hemostatic Mechanism. The hemostatic system can be divided into four families of proteins. The procoagulant/ anticoagulant families control clot formation. The fibrinolytic/ antifibrinolytic families control clot dissolution. The activities and functions ofthese families are coordinated directly and indirectly by thrombin. Thrombin converts fibrinogen into fibrin monomers which polymerize non-enzymatically to form a gel that supports fibrinolysis. Thrombin activates factor XIII to XIIIa which crosslinks fibrin to resist fibrinolysis. Thrombin is able to induce platelet aggregation and modify platelet surface to accelerate coagulation. It can also modify the endothelial surface to inhibit coagulation i.e. endothelial associated heparin molecules, protein C and thrombomodulin, tissue factor pathway inhibitor, or to control fibrinolysis by regulating tissue plasminogen activator and plasminogen activator inhibitor activity and concentration. ‘ Thrombin ’ / \ i7, \ m... Endothelial / cells °°‘” mac it u . ¢ and mm°3°n Fibrinolytic Anticoagulant Procoagulant factors factors factors Fibvrin / v ( Antifibrinolytic ¢ Factors FDP Anticoagulation Coagulation Fibrinolysis Antifibr'molyaia CONTROL OF CLOT CONTROL OF CLOT FORMATION LYSIS platelet aggregation and adhesion (6, 50, 112, 171, 207). The anti- thrombotic activities also include the binding ofthrombin to endothelium- associated heparin like molecule (157); protein C activation by thrombin-thrombomodulin complex (64); and tissue plasminogen activator (t-PA)/ plasminogen activator inhibitor (PAI-l) expression. The expression oft-PA and PAI-l controls the dissolution ofthe formed fibrin clots (56, 82) (Figure 2). Mechanical, chemical, or biological injury to the endothelial cells is accompanied by changes in their properties from being anti- thrombotic to becoming pro-thrombotic by down-regulating the protective molecules and by expressing the procoagulant activities, the adhesive receptors and factors, and mitogenic factors. This transformation in properties can lead to development ofthrombosis, disseminated intravascular coagulation, and atherosclerosis (44, 100, 118, 180,211,). Figure 2. Role ofthe Endothelium in Hemostasis. This figure Shows the different components that give the endothelium its anti-thrombotic property. The upper left side ofthis figure Shows thrombin (IIa) binding to thrombomodulin (TM), an endothelial surface receptor, and activating protein C (PC). Activated protein C (aPC) in the presence of protein S (PS) inhibits factors VIIIa and Va. Another anticoagulant is antithrombin III which binds to its cofactor, heparin like receptor on the surface ofthe endothelium and inactivates thrombin and other procoagulant factors. This figure also shows, in the lower left Side, tissue factor (TF) expressed on the surface ofthe endothelial cells and serving as a cofactor for factor VIIa to activate factor X to Xa and IX to IX a. Tissue factor pathway inhibitor (TFPI) inhibits the TF-VIIa complex (89, 212). The upper right Side ofthis figure shows different inhibitors of platelet aggregation and adhesion. Finally, fibrinolysis is depicted by the activation ofendothelial cell tissue plasminogen activator (t-PA) and its inhibitor plasminogen activator inhibitor (PAI-l). ——* Activation -+i—> Inhibition ENDOTHELIAL CELLS ------ ---- -- --- Heparin like receptor TM-IIa l PC/‘P ij/PS Antithrombin III N 0 (EDRF) ,/ VIIIa & Va ' Ecto-ADPase XIa, IXa, Xa, III: II ’ Ila Fibrinogen W T \\\ :. { Xa \‘ Platelets V8 COAGULATION VIIIa Tm mm \EIBRINOLYSIS 1X11)“ X Plasmin Fr Plasminogen K t- A PAI-l F---VIIa T PI A - -- - - - - -- - -- - ENDOTHELIAL CELLS Endothelial cell injury or activation of circulating blood cells induces the desirable process of vascular constriction, blood clotting, and vascular repair (282). Injury, loss, and/or dysfunction of endothelial cells plays an important role in thrombotic disorders. Loss or dysfunction of the endothelial cells protective properties coupled with the pro-thrombotic expression on the subendothelial matrix tilts the hemostatic balance toward an undesirable hyper-coagulant event associated with thrombosis, DIC, and atherosclerosis. The initiation ofthrombus formation is mediated by endothelial, blood coagulation, anticoagulation and fibrinolytic systems (Table 1, Figure 2) (56, 157, 211) Circulating platelets have surface receptors that are expressed upon activation and can interact with different agonists, including extracellular matrix and plasma proteins(Table 2) (100). Receptor mediated interactions lead to platelet adhesion and aggregation in the vicinity ofthe vascular damage. One ofthe important players in platelet adherence to the endothelial damaged site is glycoprotein Ib- IX. This receptor is expressed on the platelet surface, and interacts with VWF on the endothelial surface (90, 159 ). Another important receptor on the platelet surface is the Glycoprotein Ia-IIa complex 10 Table l. Endothelial Pro- and Anti-coagulant Factors of Hemostasis W M91113! Surface Bound Factors Thrombomodulin Increases the afiinity of thrombin to protein C and promote its activation. Protein S Serves as a cofactor for activated protein C to inactivate factors Vflla and Va. Heparin like receptor Serves as a cofactor for antithrombin III to inactivate thrombin and other procoagulant factors. Ecto-ADPase Inhibits platelet aggregation. Plasma Factors Prostacyclin Inhibits platelet aggregation. NO (EDRF) Inhibits platelet aggregation. TFPI Inactivates factors Xa, TF and VIIa complex. t-PA Converts plasminogen into plasmin that lyses the fibrin clot. WES Surface Bound Factors Tissue Factor Cofactor for factor VIIa. Factor Va Serves as a cofactor for the prothrombinase complex. Plasma Factors PAI-l Inhibits t-PA. Von Willebrand factor promotes platelet adhesion. 11 Table 2: Receptors on Platelet Surface and their Agonists REQEHDRS AGQNISI CDMMENI Thrombin receptor thrombin constitutively expressed at all times. GP Ia-IIa collagen constitutively expressed at all times. GP IIb-IIa fibrinogen expressed only after platelet activation. Fibrinogen plays a role of a bridge between platelets which leads to platelet aggregation. Gpr-IX VWF constitutively expressed at all times. Important for platelet adhesion. 12 which interacts with the subendothelial collagen which, in turn, triggers the platelet activation cycle (39). Other receptors present on the platelet surface interact with several agonists such as thrombin, ADP, serotonin, AA and can trigger platelet activation, release, and aggregation (83). Platelet aggregation depends on the interaction ofthe glycoprotein IIb-IIIa complex on the platelet surface and fibrinogen in plasma (33). Platelet activation leads to accumulation of additional platelets, ventralization of heparin, and secretion of factor V and fibrinogen which affect clot formation (183, 230, 264). Finally, the platelet plug provides an anchoring place for the formation and activation ofthe coagulant serine protease complexes that lead to the formation of thrombin (257). Formation ofa clot is a host defense mechanism that runs in parallel with the repair and inflammatory mechanisms. Except for possibly factor VII, all proteins and cellular components involved in blood clotting exist under normal physiological conditions in inactive forms (zymogens). These zymogens are converted, by cleavage of one or two peptide bonds, to their active enzyme forms. The blood clotting system is a sequence of events that leads to the generation ofthrombin and the conversion of fibrinogen to fibrin. Precise regulation is required to keep the procoagulant activities within the area oftissue 13 injury (49, 79, 155). The key to stimulation ofthis sequence of events is tissue factor expression when the endothelium is injured (11, 39, 174, 213, 248). Factor VII binds to tissue factor and forms an activated complex ( TF-VIIa) that activates factors X and IX to Xa and IXa respectively (93, 194, 211, 289). Once the pathway is activated, a positive feedback occurs where factors VIIa and Xa further enhance factor VII activation (206). In addition to factor VII activation, thrombin along with factor XIa can activate factor XI to factor XIa on negatively charged surfaces (179). Factor XIa activates factor IX. Factor IXa forms a complex with factor VIIIa on phospholipid membranes in the presence of calcium and activates factor X on cell membrane surfaces (85, 113). The prothrombinase complex consists of factors Xa, Va and prothrombin on the membrane surface and the last gets activated to thrombin. Thrombin has numerous substrates in vivo. In thrombus formation, the most important targets for thrombin are platelets, fibrinogen, and factors V, VIII and XIII. Fibrinogen is a symmetrical six polypeptide chain plasma protein consisted of two Aa, two BB, and two Y chains. Thrombin converts fibrinogen to fibrin by cleaving off the A and B terminal peptides of the Au and BB chains. Thrombin also activates factors V, VII, VIII, XI. In addition, thrombin activates factor XIII to the active transglutaminase XIIIa which in the presence of calcium introduces covalent cross-links 14 between glutamyl and lysil side chains to polymerize and form the insoluble fibrin clot. The formation of blood clots is limited by antithrombin III, tissue factor pathway inhibitor (TFPI), and Protein C. Protein C controls the generation of thrombin by inhibiting factors Va and VIIIa. Antithrombin III inhibits most serine proteases ofthe coagulation cascade especially thrombin and Xa. This inhibitor is much more effective in the presence of heparin (183). TFPI is present at much lower concentration than ATIII. TFPI binds VIIa-TF-Xa ternary complex. While ATIII acts by removing excess generated serine proteases in plasma, TFPI acts by inhibiting the initiating event (228). Generated thrombin binds to thrombomodulin on the surface of the endothelial cells to activate Protein C. Protein C in the presence of its cofactor Protein S, inactivates Va and VIIIa. This inactivation depends on the presence ofa membrane surface (70, 116). The formation ofthe fibrin clot is a temporary measure to prevent immediate blood loss. This temporary structure must be degraded and replaced by more suitable vascular architecture where injury and clotting have occurred. This repair process involves the generation of new connective tissue and matrix ingredients derived from smooth muscles and endothelial cells. The clot lysis process is due to the action ofplasmin. Plasmin circulates in the plasma in an 15 inactive, precursor form, plasminogen. T-PA activates plasminogen to plasmin (143, 253). Plasminogen and t-PA interact with fibrin (143). They also interact with various receptors such as annexin II and mannose receptors, and with different matrix components such as osteonectin-collagen complex leading to localized plasminogen activation (60, 120, 143). These fibrinolytic enzymes are highly regulated by both plasma-derived and endothelial-derived inhibitors. PAI-l, an endothelial-derived inhibitor, inhibits t-PA. a2- antiplasmin, a plasma-derived inhibitor, inactivates plasmin (56). ,\:.i 0a. '6 ,, a. l i .kek'é':,.k A) PROSTACYCLIN Prostacyclin (PGIZ) is a potent inhibitor of platelet activation, adhesion, release and aggregation. The inhibitory activity of prostacyclin is mediated via guanosine nucleotide binding receptors and the activation of adenylate cyclase. This leads to an increase in cyclic adenosine monophosphate (cAMP) and inhibition of platelet activation (76, 171). Mechanical or chemical perturbation of endothelial cell membranes results in the formation and release of prostaglandins including prostacyclin. These prostaglandins are not 16 stored by cells (20, 201). A number of endogenous chemical stimulants like thrombin, histamine and bradykinin and those released by platelets like serotonin, platelet-derived growth factor, interleukin-1 and adenosine nucleotides can stimulate the production ofprostacyclin by activating the endothelial cells phospholipase A2 (PLA2) (73). In human endothelial cells prostacyclin synthase catalyzes the conversion ofPGH2 into PG12(282). The exact regulatory mechanisms are not entirely clear. Wu et a1. (1995) demonstrated that PGI2 synthesis is regulated at each enzymatic step and cyclooxygenase is the key enzyme (Figure 3)(283). Xu et al. and Sanduja et al. have demonstrated that the synthesis time OfPGI2 is short (about 20 minutes) and depends on the level ofcyclooxygenase which has a very short halflife (229,285). Cyclooxygenase has two isoforms identified in HUVECS. Prostaglandin H synthase isoform-1 (PGHS-l) is primarily responsible for synthesizing prostacyclin under physiologic conditions. The concentration of PGHS-l increases after stimulation with cytokines and shear stress. Prostaglandin H synthase isoform-2(PGHS-2) is present in low concentration in resting cells. However, when the endothelium is insulted by cytokines and mitogenic factors, the level of PGHS-2 increases rapidly. Because PGHS-2 is present in inflammatory and neoplastic cells, Wu et al. suggested that this isoform has been involved primarily in inflammation and cell proliferation. The exact 17 "“93...“ Figure 3. Metabolism of Arachidonic Acid in Endothelial Cells. Prostacyclin and thromboxane A2 are synthesized from arachidonic acid with prostaglandin G2 and H 2 as intermediates. Prostacyclin and thromboxane A2, potent platelet inhibitor and aggregator respectively, are unstable and converted into the more stable and less active 6-keto-prostaglandin F,“ and thromboxane B2. Prostacyclin production is initiated by the enzyme phospholipase A2 which catalyzes the release of arachidonic acid from phosphatidyl choline. Cyclooxygenase, also called prostaglandin H synthase, catalyzes the oxygenation ofarachidonic acid to form prostaglandin G2. Then peroxidase catalyzes the reduction ofprostaglandin G2 into prostaglandin H2. Prostaglandin H 2is a precursor for prostaglandins, thromboxane A2 and prostacyclin. Glucocorticoids and aspirin inhibit phospholipase A2 and cyclooxygenase respectively and reduce the synthesis ofprostacyclin and thromboxane A2. Lipid peroxides inhibit prostacyclin synthase and decrease prostacyclin production whereas Dazoxiben inhibits thromboxane synthase and decreases the synthesis of thromboxane A2 and increase the concentration of prostacyclin in endothelial cells. 18 Membrane Phospholipids Phospholipase A2 Glueocortico ids _._. if) Arachinodie acid —“‘——> lS-lipoxygenase Cyclooxygenase ¢ Eicosanoids Aspirin-like drugs 4H} Changes in Prostaglandin 6, “km Peroxidase ENDOTHELIAL CELLS PLATELETS Prostaglandin H2 Prostacyclin synthase Thrombome synthase Lipid peroxides ———fi—> \‘E‘i— Dm‘fl’“ Prostacyclin Thronboxane A2 Nonenzyme hydrolysis Nonenzyme hydrolysis 6-keto-Prostaglandin F1 3 Thromboxane 82 19 mechanism by which PGHS-l and 2 are regulated is not known(283). Wu et al. and DeWitt et al. demonstrated that prostacyclin synthase and PGHS-l are auto-inactivated during catalysis. Prostacyclin synthase, like PGHS-l, plays a major role in controlling PGI2 production (55, 283). It has been demonstrated that PGH2 , produced in activated platelets, can enter the endothelial cells and be converted into PGI2 by the endothelial prostacyclin synthase (282). B) NITRIC OXIDE Nitric Oxide (NO), the major component of endothelium dependent relaxing factor (EDRF), is a mediator ofvasorelaxation, neurotransmission, cytotoxicity, immunomodulation and platelet inhibition(78,l68). Nitric oxide synthase (NOS) catalyzes the synthesis ofNO from L-arginine in endothelial cells. The synthesized NO can activate smooth muscle guanylyl cyclase and increase cyclic guanosine monophosphate (cGMP) or, can diffuse into the blood vessel lumen where it enters platelets and inhibits platelet activation, adhesion and aggregation by increasing cGMP (207). When endothelial cells are activated by physiologic agonists, calcium concentrations increase. This elevated calcium binds to Calmodulin (CaM) to form a complex. This complex binds to the NOS-III (isoenzyme III) binding Site and the enzyme is activated. 20 Under normal conditions, the endothelial cells produce a constant level of PGI2 (71). Evidence for a constant basal level of NO is not clear. Thrombin, histamine, mechanical force, lipid and shear stress are responsible for maintaining a basal level of PGI2 and NO. An increase in calcium in endothelial cells leads to an increase in PGI2 and NO (71, 283). C) ECTO-ADPase Ecto-ADPase, an endothelial surface bound enzyme, inhibits platelet aggregation by degrading platelet ADP into AMP. Marcus et al demonstrated that ADP, released from activated platelets, plays the role of an important mediator for amplifying platelet aggregation whereas Ecto-ADPase plays an important physiologic role in limiting platelet aggregation. HUVEC Ecto-ADPase has not been purified(158). 21 A) ENDOTHELIUM ASSOCIATED HEPARIN-LIKE MOLECULES Rosenberg et al. demonstrated the presence ofa cofactor for Antithrombin III(ATIII) on the surface of the endothelial cells. These heparin-like molecules (heparin sulfate proteoglycan) were isolated from cloned endothelial cells and, when treated with heparinase, the anticoagulant activity was abolished (37, 223). ATIII, a major serine protease inhibitor, forms a 1:1 stoichiometric complex with thrombin and other serine proteases such as Xa, IXa, XIIa at a slow rate. Heparin accelerates the interaction between thrombin and ATIII by enhancing the interaction of arginine residues on ATIII with aspartic acid residues on the serine protease. A small concentration of plasma ATIII is bound to the endothelial surface heparin-like receptors to inactivate thrombin locally. When the endothelial cells are injured, a much larger quantity of the sub-endothelial heparin-like receptors become available (157, 223, 262). B) PROTEIN C AND THROMBOMODULIN Thrombomodulin (TM) is an integral membrane protein on the endothelial cell surface. By competing for thrombin and forming a complex, thrombomodulin inhibits the procoagulant activity of 22 thrombin. Thompson et al. demonstrated that in addition to thrombin inhibition, thrombomodulin binds factor Xa and inhibits its activation ofprothrombin (261). Thompson et al. have also shown that the thrombin-TM complex is internalized by endocytosis and, as a result, thrombin is degraded and thrombomodulin is recycled to the endothelial surface. A soluble form ofthrombomodulin is present in the plasma and has been used as a marker of endothelial cell activation (261). Pressner et al. and Esmon et al. showed that in HUVEC, thrombomodulin synthesis is down-regulated by tumor necrosis factor (TNF), interleukin-1 (IL-1) and endotoxin resulting in an increase in fibrin formation (65, 66,) . Esmon et al. demonstrated that Protein C is synthesized in the liver and when converted into activated protein C, it inactivates Va and VIIIa (64, 66) . Thrombin is the only known serine protease that can activate protein C. Thrombin undergoes conformational changes when it binds to thrombomodulin on the surface ofthe endothelial cells resulting in increased affinity for Protein C. Protein S, a cofactor for Protein C, is a vitamin K-dependent glycoprotein synthesized by the endothelium, liver and megakaryocytes. Dahlback et al. showed that about 50-60% of protein S is bound to C4b-binding protein and is inactive. Protein 8 binds to both the endothelial surface and activated Protein C to form a complex, enhancing the inactivation of Va and 23 VIIIa by activated Protein C (77). C) TISSUE FACTOR PATHWAY INHIBITOR Tissue factor (TF), a receptor and cofactor for VII and VIIa, is the primary cellular initiator ofthe coagulation cascade (182). Tissue factor is a potent initiator of coagulation and has a very critical function in hemostasis and thrombogenesis (181, 218, 280). TFPI (tissue factor pathway inhibitor), known also as extrinsic pathway inhibitor is a serine protease inhibitor present in low concentration in plasma. TFPI inhibits the VIIa-TF complex in the presence of Xa(89, 212). This inhibitor is synthesized in the liver and endothelial cells. It was reported that TFPI level is normal during liver disease but low in DIC, suggesting that endothelial cells may be the primary source of TFPI(22,98,176) A) TISSUE PLASMINOGEN ACTIVATOR Tissue-plasminogen activator (t-PA), a serine protease, catalyzes the conversion of plasminogen to plasmin by hydrolyzing the peptide bond between Arg560 and ValS6l in the precursor 24 plasminogen. T-PA, a 68-KDa glycoprotein, has 5 domains: a finger domain (homologous to a similar structure in fibronectin), a growth factor domain (homologous to epidermal growth factor), two Kringle domains (originally discovered in thrombin), and a catalytic domain (resembling trypsin) (12, 198). Studies have shown that the finger domain and Kringle 2 are involved in fibrin binding which stimulates t- PA/plasminogen reaction (130, 198). Another specific interaction occurs between t-PA and plasminogen activator inhibitor-1 (PAI-l). The loop between Kringles 2 and the catalytic domain is involved with the binding to PAI— 1. Degen et al. (53) showed that the t-PA gene is located on chromosome 8. They also demonstrated that the size ofthis gene from the site ofinitiation of transcription to the polyadenylation site is 32,720bp. Degen et al. (53) also reported that there is 81% homology of the genomic DNA of the t-PA coding strand and PAI-l noncoding strand . Based on this homology, the evolutionary relationship oft-PA and PAI-l appears to be close (53, 190). Studies done by Kooistra et al. (129) demonstrated that PKC, cAMP, and steroids regulate t-PA synthesis (129). This regulation is coordinated with PAI-l synthesis. Under physiological conditions, the concentration oft-PA in the plasma is 4-6 ng/ml. The level oft-PA increases with age and it is higher in males than females (125, 209, 251). The concentration oft- 25 PA decreases in healthy individuals who are on oral contraceptives or anabolic steroids (84, 272). T-PA is produced by all endothelial cells. However, the level of t-PA and the ratio oft-PA/mRNA are different depending on the source of the endothelial cells (Figure 4) (269, 271). Keber et al. showed that the continuous production oft-PA in vivo varies along the vascular system and the plasma t-PA concentration is the average of all the t-PA produced from different vascular endothelial sites and the clearance by the liver (119). A variety of Stimuli, in vivo and in vitro, leads to an increase in t- PA release. Exercise (59, 259), acidosis and increased venous pressure (110, 240, 268, 277) and l-Desamino-8-D-arginine vasopressin (dDAVP) (28, 31, 196) increase the level oft-PA in blood. Using a rat hindleg perfusion system, Emeis et al. (61) demonstrated that reaction products from platelets and the coagulation cascade stimulate an acute release of t-PA. They also investigated the involvement of calcium in the release oft-PA using the rat hindleg perfusion system. Their results suggested that calcium influx is essential for the release oft- PA. They also showed that phospholipase C and lipoxygenase- dependent mechanisms but not phospholipase A are involved in the release oft-PA (61, 265). Recent data have shown that protein synthesis is not required for either release or storage of t-PA because 26 there is a significant intracellular pool present in tissues (8, 217, 265). In addition to the intracellular pool oft-PA, Barnathan et al. (14) showed that there is an extracellular pool for t-PA especially two distinct binding sites on the endothelium (14, 15). In cultured cells, receptors and binding sites for t-PA have been identified in monocytes and endothelial cells. However, there is a contradiction about the effect Of receptor binding on t-PA activity. Some studies demonstrated that these sites protect t-PA from inhibition by PAI-l and t-PA can be dissociated from these receptors in an active form. Other reports disagreed with these findings (165). It has been demonstrated that during the formation ofa fibrin clot, t-PA becomes almost entirely incorporated into the clot in a short period oftime. But, in the presence of an existing clot, t-PA slowly binds and penetrates. So, the difference in t-PA effectiveness as an activator ofplasminogen depends on the status ofthe clot. Furthermore, plasmin bound to fibrin may be protected from inactivation by az-antiplasmin (29, 74). B) PLASMINOGEN ACTIVATOR INHIBITOR I PAI-l is the major physiological inhibitor oft-PA in plasma (7, 148). PAI-l is a single chain glycoprotein of MW 50,000 that contains 379 amino acids. PAI-l inhibits t-PA by forming a covalent bond. This 27 inhibition occurs rapidly and in stoichiometric manner. During the inhibition, PAI-l is consumed (40, 101, 148). Almost 75% of plasma t-PA is in complex with PAI-l. Altered PAI-l levels in blood correlate with either hemorrhagic or thrombotic problems (249). PAI-l has been detected in different types of cultured cells including HUVECS (40, 202), bovine aortic endothelium (Figure 4) (202), vascular smooth muscle cells (127), mesothelial cells (216), granulosa cells (190), human lung fibroblasts (151) and other cell lines. PAI-l is released from cultured endothelial cells in an active form. This inhibitor is not stable in solution because ofthe lack of cysteine residues (136). The free PAI-l in the fluid phase rapidly decays into the latent form possibly because of conformational changes. This latent fOrm of PAI-l can be converted to the active form in conditioned medium by exposing it to SDS and other denaturants (102,129,140, 148, 249). The binding of PAI-l to vitronectin in plasma and in the extracellular matrix stabilizes PAI-l in its active form (52, 58, 236, 281). While the majority ofPAI-l in blood is active, the PAH in platelets is latent (51, 52, 281). This latent PAI-l can be activated in vivo by conformational changes in the tertiary structure(52). The human PAI-l gene, 12.2 kilobase pairs in length, contains 9 28 Figure 4. The Fibrinolytic System. In endothelial cells, the t- PA mediated fibrinolysis is controlled by the expression oft-PA and PAI-l and the release/ uptake ofthese two proteins. In blood, four levels of control of fibrinolysis are present: 1) the fibrin—stimulated activation of fibrinolysis 2) the degradation reaction of fibrin by plasmin 3) t-PA inhibition by PAI—l in blood 4) az-antiplasmin inhibition of fibrin degradation. The fibrin clot has an important role in fibrinolysis by stimulating the fibrinolytic process and by protecting plasmin from inhibition by aZ-antiplasmin. Fibrin has a high affinity for plasmin, and the plasmin that is bound within the fibrin clot is protected from inhibition by az-antiplasmin. 29 -////.'———_\\ / idiotic/1&1 cal/3' \\5 \\ // WM 11 11”“ PAI-l t-PA $¥QQ§ kl plasminogen ——> plasmin l X a, antiplasmin fibrinogen ——> fibrin ———) FDP thrombin A coagulation cascade 3O exons and 8 introns. This gene is located on chromosome 7 (26, 72, 123, 149). Two distinct transcripts, 2.3 and 3.2 Kb in size, formed by alternative polyadenylation, are colinear at the 5' end. The 3' end of the 3.2 Kb contains an adenosine-thymine rich sequence implicated in the mRNA stability (238). Studies regarding PAI-l gene expression have been mostly done in vitro. The synthesis of PAI-l can be induced by a variety of stimulants like cytokines, hormones, growth factors, and viruses (18, 19, 41, 42, 46, 216, 274, 275). An increase in blood PAI-l concentration has been associated with thrombosis, and atherosclerosis (188, 233). Individuals who have inadequate levels of PAI-l are at risk of hemorrhage due to excess fibrinolysis (57). The importance of PAI-l in regulating the balance of t-PA/PAI-l ratio is supported by a variety of clinical studies (57, 173, 188,233) Recurrent bleeding disorders have been reported in patients who have lower PAI-l concentration than normal, leading to an increase in fibrinolysis and altered hemostatic balance (57, 234). An increase in PAI-l has been reported in different conditions associated with thrombosis and DIC. In gram-negative bacteremia, the level of PAI-l activity increases tenfold and in DIC is associated with this condition (193). An elevated PAI-l has been reported in myocardial infarctions (97) which raises the possibility of PAI-l being a risk factor for 3] recurrent myocardial infarction. It has been shown that PAI-l activity increases rapidly in hepatitis, malignancies, post surgery, after trauma and deep venous thrombosis (115, 124, 134, 258). An elevated PAI-l has also been associated with non-insulin dependent diabetes, hyperinsulinaemia, and hypertriglyceridemia (115) which are high risk factors for the development of atherosclerosis (256). A number of observations have shown that the underlying mechanism responsible for the induction of PAI-l mRNA differ depending on the tissue source ofthe endothelial cells (232, 233). PAI-l levels have been shown to correlate with body weight. Elevated levels associated with obesity can be decreased by weight reduction (148). Studies of cells in cultures suggest that PAI-l is not stored within the cells except in platelets and megakaryocytes. These studies suggest that the increase in PAI-l reflects a change in gene expression (149, 167). But mRNA stability accounts for part of the increase in PAI-l mRNA measured by an increase in 3.2 Kb transcripts that are rich in AU (238). 32 Based on biological properties, the herpes viruses are classified into three subgroups: alpha-herpesviruses (herpes Simplex 1, herpes simplex 2 and varicella zoster) which establish latent infections of neural tissues; beta-herpesviruses (cytomegalovirus) which persist in reticulo-endothelial cells, and replicate very well in fibroblasts; and gamma-herpesviruses (Epstein-Barr Virus) which are lymphotropic viruses (221). In 1986, a new human herpesvirus was isolated from the peripheral blood mononuclear cells of patients with lymphoproliferative disorders; some ofthese patients were infected with the human immunodeficiency virus (HIV) (225). It was shown that this virus had the ultrastructure and morphogenetic properties that are characteristics of herpes viruses (147), but it was distinct from the five previously known human herpes viruses based on antigenic properties and the failure to show homologous hybridization with nucleic sequences from the other five human herpes viruses (114). Despite the lack of knowledge on the nature of the latent site of HHV- 6, it was suggested that this virus should be classified as a gamma- herpesvirus because it infects lymphocytes in vivo and in vitro. But analysis ofthe structure and sequence ofthe genome, reveals that the 33 greatest degree of similarity is shared with cytomegalovirus (CMV)(147). HHV-6 is antigenically different from other human herpesviruses. Its nucleotide sequencing has shown 66% DNA sequence homology with CMV (137). The antiviral susceptibilities of HHV-6 also resemble CMV’s (224). HHV-6 is an enveloped virion with an icosohedral nucleocapsid of 162 capsomers, and it contains a linear double stranded DNA genome (131). Virions ofHHV-6 have four main structural elements: an electron-dense core, a capsid with icosehedral symmetry, a tegument, and an outer envelope where virally encoded glyc0proteins and integral membrane proteins are embedded (220). Nucleocapsids are assembled in the nucleus (21). After capsid assembly, the tegument is acquired, while still in the nucleus, by a series of events which has not been fully characterized (187, 219). These tegumented cytoplasmic capsids appear to acquire their envelope by budding into cytoplasmic vesicles. Mature Virions are released by exocytosis (219). Seroepidemiologic studies show that HHV-6 infection is endemic in humans and is commonly acquired in early childhood (27). Seroconversion occurs by the age oftwo and seroprevalence in healthy population exceeds 75% (192). As other herpesviruses, HHV-6 can persist in a latent form after primary infection (132). Although the exact site oflatency ofHHV-6 is unknown, the virus antigen and DNA 34 have been isolated from lymphocytes (43), lymph nodes (141), endothelial cells (192), macrophages and monocytes (109, 128), bronchial and salivary glands (132, 191, 210), and CNS tissues (32, 152) HHV-6 infection has been shown to be the cause of roseola infantum (exanthem subitum, a febrile illness of early childhood (17). Infections associated with HHV-6 range from mild illnesses with no specific symptoms (184) to acute febrile illness with or without rash (19, 254), and persistent active infection (27). HHV-6 in children, has been associated with complications such as seizures (17), increased intracranial pressure (193), fulminant liver hepatitis (242), myaligic encephalitis (278), and thrombotic thrombocytopenic purpura (189). HHV-6 has been associated with Epstein-Barr virus-like mononucleosis syndrome, non-Hodgkin and Hodgkin lymphomas, necrotizing lymphadenitis and encephalitis (133, 162, 227, 287). HHV-6 has also been reported as an important pathogen in patients with HIV infection and has been proposed as a cofactor in the acquired immunodeficiency syndrome (AIDS) (5, 126, 153, 154). On the basis of genomic DNA sequences, cell tropism, and protein expression, two variants of HHV~6 have been described. These variants are designated as variant A and variant B. HHV-6A variant has been obtained primarily from adults (1, 2, 225). HHV-6B variant is 35 isolated primarely from children and associated with Exanthum Subitum and pediatric febrile illnesses (54, 197). Phinney et al. reported that fatal systemic HHV-6 infection in newborns has a terminal course dominated by Disseminated Intravascular Coagulation due to the exposure ofthe subendothelial collagen (199). The primary target cells of HHV-6 are CD4+ T lymphocytes. This characteristic is shared with HIV where HHV-6 causes cytopathic effect and cell death (153). Even though HHV-6 replicates most efficiently in CD4+ T lymphocytes, the cellular host range of HHV-6 is large and includes CD8+ T cells, macrophages, megakaryocytes, epithelial cells and endothelial cells (4, 192). In 1852, Rokitansky (222) discussed the “atheromatous process” and suggested that the deposit in the arteries consists of fibrin (222). In 1946, Duguid (58) demonstrated that thrombosis is a factor in the pathogenesis of coronary and aortic atherosclerosis, and that fibrin is present within and on the surface of fibrous atherosclerotic lesions (58). But, it was not until the 19805 that it was widely accepted that fibrinogen, as well as other coagulation factors, fibrin deposits, and 36 fibrinolysis may be involved in atherogenesis and vascular occlusion. (205,241,260) Past studies ofthe mechanism of atherosclerosis have focused mainly on three areas. First the role oflipids, particularly their oxidation products that are toxic for the vascular wall. The second focus was on the role ofinflammatory cells. In the beginning, researchers were most interested in the role of monocytes and macrophages because oftheir early accumulation at the site ofinjury and the formation of foam cells. Majno (156) demonstrated that neutrophils were actually the earliest cells to marginate at vascular surfaces destined later to become atherosclerotic (156). The third focus was on the anticoagulant properties ofthe vascular system. The atherosclerotic vascular surfaces tend to lose their natural anticoagulant properties and become prothrombotic(117, 121). Injury to the vascular endothelium is believed to initiate the atherosclerotic lesion. Several risk factors like smoking, hypercholesterolemia and obesity have been implicated in the development of atherosclerosis, but they are not present in all individuals who develop atherosclerosis. In 1978, based on research conducted in fowls, it was suggested that herpesviruses are potential initiators of vascular endothelial injury and atherosclerosis (68). They infected some chickens with an avian herpesvirus. These chickens 37 were fed a cholesterol-free diet, yet they developed atherosclerosis. Another group of herpes infected chickens was fed a cholesterol-rich diet and suffered even more clogging of heart arteries. Whereas, chickens vaccinated against the avian herpesvirus resisted development of atherosclerosis regardless ofthe concentration of cholesterol in their blood (68). In 1987, Hajjar et al. (95) called human herpes viruses “the missing link between thrombosis and atherosclerosis” (95). At the American Heart Association’s Science Writers Forum in Monterey, California. David Hajjar and other researchers stated that the herpes theory helps to explain those cases in which people suffer heart attacks despite low or normal cholesterol. Joseph Melnick (3) at Baylor College of Medicine in Houston, TX said: “Many people get infected with some type of herpes virus early in life. The virus then hides out in various cells, includingendothelial cells. The herpes-induced vascular damage would take place invisibly at first, setting the stage for heart disease later in life under the right conditions” (3, 164). A number of studies have demonstrated a relationship between herpesvirus infection and atherosclerosis. Nieto et al. (185) did a cohort study that was comprised of 150 individuals with elevated carotid intimal-medial thickness measured by B-mode ultrasound. The control group had 150 age and sex-matched individuals with low 38 carotid intimal-medial thickness. Case subjects had higher mean CMV antibody titers than control subjects. There was also evidence ofa direct relationship between intimal-medial thickening and CMV antibody titer (185). Sorlie et al. (243) conducted a similar study on 340 matched case-control pairs from the Atherosclerosis Risk in Communities (ARIC). The cases were defined by B-mode ultrasonography as individuals with thickened carotid artery walls consistent with early atherosclerosis, but without a history of cardiovascular disease. Control groups were defined as individuals without thickened walls or history of cardiovascular disease. The results suggested a positive association between CMV and asymptomatic carotid wall thickening consistent with early atherosclerosis P=0.03 (243). Similar results were seen by a study performed by Adam et al. (3) done on 157 Caucasian male patients undergoing vascular surgery for atherosclerosis and compared to a matched control group of patients with high cholesterol levels. The prevalence of CMV antibodies in the surgical group (90%) was higher than in the control group (74%) and a significantly higher percentage of surgical cases (57%) than controls (26%) had high CMV titers P<0.001. A smaller difference in HSV antibody prevalence was seen between the two groups. Similar studies performed by other scientists demonstrated a relationship between CMV, HSV, EBV and other 39 herpesviruses antibody titers and atherosclerosis (3, 177, 185). Loebe et al. (144) performed serologic tests for CMV after cardiac transplantation at six to eight week intervals on 50 patients. Seventy-seven percent of patients who developed coronary atherosclerosis had a four fold increase in CMV IgG titer or positive IgM compared with only 25% in patients without coronary disease (144). A similar study was conducted by Grattan et al. (91) on 301 patients from 1980-1989. Two hundred ten patients were negative for CMV. During this study, 91 patients became positive for CMV as indicated by a four fold increase in IgG titer, positive CMV antigen in the tissue or positive culture. They also observed that the rate of cardiac allograft rejection was significantly higher in the patients who were positive for CMV. Using pathologic studies and angiographic criteria, the presence of graft atherosclerosis was significantly more in the CMV group (91). Finally, the death caused by graft atherosclerosis was significantly higher in patients with CMV. The same outcome in post heart transplantation was reported by Mattila et al. where CMV was the most harmful agent and a risk factor for post-transplant coronary artery disease. The CMV infection was significantly more harmful in pre-operatively seronegative patients with seropositive donors than in patients who were seronegative pre-operatively (161). Blum et al. (23) studied the correlation between anti-CMV 40 antibody titer and coronary artery disease in 65 patients. All patients underwent balloon coronary angioplasty and were followed for 15 months post-angioplasty with a thallium perfusion scan. Patients who had recurrent chest pain and/or a positive thallium scan had another coronary angiography. Patients with high anti CMV titer had a higher prevalence of coronary artery disease P<0.001 and restenosis P<0.05 than with patients with low anti CMV titer. These findings support the infectious theory of atherosclerosis that high anti CMV titer may be a strong marker for coronary artery disease (23). Zhou et al. (288) also looked at the relationship between restenosis and CMV titer because CMV DNA has been found in restenotic lesions. They studied 75 consecutive patients undergoing directional coronary atherectomy. CMV was determined before atherectomy was performed. After six months, patients with positive CMV titer had a greater reduction in the luminal diameter resulting in a significantly higher rate of restenosis (43%) than patients negative for CMV (8%). There was no evidence of acute infection with CMV because the IgG titers did not increase over time and the CMV IgM antibodies were negative (288). The relationship between CMV infection and atherosclerosis was evaluated in patients with diabetes mellitus type I and II. CMV antibodies were significantly higher in diabetic patients with atherosclerosis (71%) than in patients without atherosclerosis (45%) 41 P=0.018. CMV IgG titers were twice as high in the diabetic patients with atherosclerosis compared to diabetes patients without atherosclerosis. This difference was most striking in the female population (276). The mechanism by which CMV infection ofthe vascular cells may increase the possibility of atherosclerosis development in arteries is unknown. Salomon et al. (226) suggested that this accelerated atherosclerosis represents a form of chronic immunologic reaction resembling a delayed-type hypersensitivity. This localized immune reaction may be initiated by histocompatibility antigens (HLA) class II (226). It has been demonstrated in cultured endothelial cells that CMV infection upregulates major histocompatibility antigen class II expression (266). In addition, co-culture of T-lymphocytes with human endothelial cells infected with CMV resulted in T-lymphocytes activation (279). A number of studies have demonstrated the presence of herpesvirus antigens, nucleic acid in normal and atherosclerotic vessels (16, 164). Melnick et al. (164) studied arterial wall specimens from 135 patients who underwent vascular surgery for atherosclerotic vessels using Polymerase Chain Reaction (PCR) for the presence of CMV nucleic acid. CMV nucleic acid was detected in 76% oftissue tested using late gene primer and in 90% oftissue tested using 42 immediate early gene primer. There was no significant difference in CMV detection in atherosclerotic tissue and an uninvolved aortic tissue from the same patients. Positive serum CMV titers were detected in 86% of the patients in whom tissue CMV antigens were detected (164). Hendrix et al. (104) demonstrated the presence of CMV nucleic acids in arterial walls of patients with atherosclerosis by dot blot in situ DNA hybridization, and polymerase chain reaction using primers derived from immediate early and late genomic regions. The presence of complete CMV genome was demonstrated by both dot blot CMV hybridization and PCR (104). Other laboratories reported the same findings of CMV nucleic acids in the arteries predominantly in endothelial cells of different areas ofthe body of patients with and without atherosclerosis by dot blot and in situ hybridization (103, 166). Chen et al. (34) and Yamashiroya et al. (286) demonstrated the presence ofHSV and CMV nucleic acid and/or antigen in carotid and coronary arteries and aorta. They found the viral DNA and antigens in the intact luminal surface cells and in focal clusters of spindle-shaped or foamy cells in the intimal layer. This finding supports the theory that herpesviruses participate in the initial stages of atherosclerosis (34, 286). Toyoda et al. (263) investigated the incidence of CMV infection in 40 cardiac and 25 renal allograft recipients using polymerase chain 43 reaction (PCR). They also determined the development of antibodies to endothelial cells in sera. Anti-endothelial cell antibodies were significantly higher in allograft patients who were positive for CMV than the CMV negative allograft patients. Serum anti-endothelial cell antibodies increased one to four weeks after CMV DNA detection and persisted for four weeks. The level of anti-endothelial cell antibodies correlated well with the level ofIL-2 (263). Histological biopsies from coronary arteries of patients undergoing coronary artery bypass grafting were examined by Raza-Ahmad et al. (214) for the presence of herpes virus using specific DNA probe. Forty-five % of 46 patients were positive for HSV herpes antigen. Significant correlation existed between the positive biopsies for HSV antigen and recent history of cold sores (214). Phinney et al. (199) showed that newborns with fatal systemic herpes infection had terminal courses dominated by DIC. Autopsy revealed fibrin deposition, and HSV was isolated from endothelial cells. These researchers speculated that viral induced vascular damage may have set the stage for DIC by exposing subendothelial antigens, activating platelets and triggering coagulation factors (199, 163). Animal models have been developed since 1978 to study herpes induced atherosclerosis. Fabricant et al. used chicken as a model (68). In 1986 Hajjar and Fabricant et al. (94) used the same model to study 44 the effect of chronic herpes infection on lipid metabolism and accumulation. A two to three fold increase in total aortic lipid accumulation characterized by significant increase in cholesterol, cholesteryl ester, triacylglycerol and phospholipid compared with aortas from uninfected chickens P<0.05 (68, 94). Span et al. (247) studied CMV induced vascular injury in normo—and hypercholesterolemic rats. They demonstrated that CMV infection in rats caused morphological alterations of the endothelial and subendothelial spaces ofthe large vessels in both groups. These alterations consisted of swollen endothelial cells with a surface showing blob and microvilli formation. Adhesion ofleukocytes to the endothelium and subendothelium were found. Foam cells and lipid accumulated in the subendothelial space (245, 246, 247). HEMOSTASIS AND ATHEROSCLEROSIS Data from different laboratories have demonstrated that herpesvirus infection of endothelial cells in vitro alters the normal anticoagulant surface ofthe endothelium by: 1) inducing the procoagulant activity, 2) inhibiting the anticoagulant activity, 3)modifying the fibrinolytic properties, 4) increasing the binding sites 45 for inflammatory cells. Two mechanisms are involved in altering the properties ofthe endothelium from an anti-thrombotic surface to one that is prothrombotic. The expression of thrombomodulin on the endothelial surface is reduced as a result of herpes infection. Key et al. (121) reported that there is a loss of thrombomodulin activity within four hours of HUVEC infection. Similar significant thrombomodulin downregulation was seen in bovine aortic endothelial cells. Decreased surface thrombomodulin antigen closely paralleled the loss of thrombomodulin activity. Using Northern blotting, this decrease in thrombomodulin reflected rapid loss (4 hours) in mRNA (121). Loss of thrombomodulin leads to reduction in thrombin-dependent protein C activation and reducing the ability of protein C to inactivate factors Va and VIIIa. Heparin sulfate proteoglycan is also reduced on the surface of HUVEC infected with herpes. Kaner et al. (117) demonstrated that herpes infection ofHUVECs markedly reduced proteoglycan synthesis and surface expression by endothelial cells (117). Heparin sulfate proteoglycan binds antithrombin III which is responsible for inactivating coagulation factors XII, XI, X, IX and thrombin. Another mechanism by which herpesviral infection can alter the vascular endothelium is by inducing the procoagulant activity. Key et al. (122) demonstrated that in vitro infection of HUVECS by HSV-1 46 resulted in the expression of tissue factor (122). Confluent monolayers of endothelial cells were infected with HSV-1. At appropriate time intervals, tissue factor procoagulant activity and antigen were assessed. The dependence ofthe clotting time on factor VIIa and the inhibition of clotting time by using specific blocking antibodies support the hypothesis that clotting depends on the tissue factor availability. Tissue factor activity was found to parallel the level of tissue factor antigen and there was a transient de novo expression of tissue factor. This activity did not depend upon replicative infection ofHSV-1 within the endothelial cells since similar induction of tissue factor was present in HUVECS that were exposed to ultraviolet inactivated virus (122). CMV also has the ability to induce the procoagulant activity on HUVECS. The cause ofthis enhanced coagulability may be due to the prothombinase complex assembly on the CMV infected endothelium (204). Span et al. (245) demonstrated that early infection of human endothelial cell monolayers with HSV-l or CMV resulted in an increased binding of monocytes and neutrophils but did not alter platelet adherence (245, 246). Similar results were obtained by other laboratories regarding the effect ofHSV-1 infection and CMV infection of endothelial cells on neutrophils and monocyte binding. The increased binding of granulocytes could not be induced by 47 supernatants from the CMV infected cells. This was in contrast to the increased binding seen in HSV-1 infected endothelium which is most likely due to factor(s) secreted into the medium (67, 92, 245). In contrast to what Span et al. reported, Visser et al. (274) showed that infection of endothelial cells with HSV-1 leads to an increase in platelet binding to virus infected HUVECS. They also reported that HSV-1 diminished prostacyclin secretion and enhanced thrombin generation (274). Bok et al (24) demonstrated that the infection of HUVECS with HSV-1 lead to a decrease in PAI-l activity and antigen. The activity and antigen level oft-PA was also diminished following infection with HSV-l. The loss of t-PA and PAL] appears to be secondary to a decreased synthesis at the mRNA level (24). In conclusion, viruses have been implicated in the pathogenesis of atherosclerosis and thrombosis on the basis of one or more of the following: First, widespread nature of some of these viruses in humans and the presence ofa latent form in the herpes family. Second, the demonstration of herpes viral antigen and DNA in atherosclerotic lesions of vessel walls. Third, the accumulation of cholesterol-ester on the surface of cells. Fourth, ability to grow viruses in HUVECS. Fifth, loss of anticoagulant property of the endothelium and expression ofprocoagulant activity. Sixth, generation of thrombin which has 48 multi-functional effect. Seventh, increase binding of platelets and inflammatory cells. Finally, the epidemiological relationship between herpesvirus infection and accelerated atherosclerosis especially in heart transplant patients and in restenosis after angioplasty. 49 MATERIALS AND METHODS ISOLATION OF HUMAN ENDOTHELIAL CELLS Endothelial cells were isolated from human umbilical cord vein by the method ofJaffe et al (108). All visual traces of blood on the outside ofthe cord were removed by washing it with HEPES 0.01 M buffered saline (4 mM KCl, 10 mM HEPES, 11 mM glucose, 0.15 M NaCl, pH 7.4). The ends ofthe cord and any large clots or tears were trimmed away. The umbilical vein of non-traumatized umbilical cord was cannulated at both ends and flushed twice with 0.01 M HEPES buffered saline. The endothelial cells were detached from the basement membrane by injecting 10 ml of0. 1% collagenase in HEPES buffered saline into the vein and incubating the cord in prewarmed HEPES buffer at 37° C for 10 minutes. The collagenase solution containing the detached endothelial cells was flushed out with 25 ml of HEPES into a sterile 50 ml centrifuge tube containing 10 ml of M—199 supplemented with 20% fetal calf serum (FCS). After centrifugation for 8 minutes at 200xg, the pellet was suspended in M-199 supplemented with 20% FCS, 200 U/ml penicillin, 200 ug/ml streptomycin, 150 U/ml amphotericin B, 2 mM L-glutamine, 8 ug/ml endothelial cell growth factor (ECGF), and 90 ug/ml heparin (69, 111). HUVECS from 2 to 4 cords were pooled. To reduce contamination with fibroblasts, the cell suspension was 50 incubated in culture plates coated with 2% gelatin at 37° C in 5% C02 incubator for two hours, then HUVECS were washed twice with PBS and supplemented with M-199 (108). Confluency was reached in 4 to 7 days. The endothelial cells were identified by the cobblestone morphology and the fact that they are homogenous, closely opposed, polygonal cells with a centrally located nucleus. HUVECS form a confluent monolayer without a definable whirling pattern (69). The presence of factor VIII antigen on the surface of HUVECS was measured to exclude contamination by other cell types. These endothelial cells were used at the second to the fifth passage. FACTOR VIII EXPRESSION ON HUVECS To test for HUVECS purity, the presence of factor VIII antigen on their surface was assessed. Factor VIII antigen is present on endothelial cells and absent on epithelial, smooth muscles, and other cells’surfaces (267). HUVECS were grown on eight-well slides. After four days, the cells on the slides were confluent. After removal ofthe culture medium, the HUVECS were washed twice with PBS. The slide was incubated for 15 minutes at -20°C in 100% methanol. The methanol was aspirated off and the fixed slide was air dried. Goat anti VIII antibody was added and the slide was incubated for 30 minutes at 37° C. To remove the free antibody, the slide was washed twice with PBS 51 containing 1% BSA. A rabbit anti goat antibody labeled with FITC was added and the slide was incubated for 30 minutes at 37° C. After incubation, the slide was washed twice with PB S/BSA. HUVECS were viewed using a fluorescent scope with a 495 nm filter. CULTURE OF ENDOTHELIAL CELLS HUVECS were passaged (1:3 split ratios) by washing the culture three times with PBS after removing the medium, and incubating the culture in PBS containing 2% EDTA for 10 minutes at 37° C. After incubation, the culture plates were tapped gently on the side and the rounded endothelial cells detached. The PBS-EDTA medium containing the endothelial cells was centrifuged at 200xg for 4 minutes. The pellet was resuspended in M-199 medium supplemented with 20 % FCS (fetal calf serum), 2 mM L-glutamine, 90 ug/ml sodium heparin, 2000 U/ml penicillin, 200 ug/ml streptomycin, 150 U/ ml amphotericin-B, 8 ug/ml ECGF. Cells were grown on culture plates coated with 2% gelatin. Cell viability was estimated with the use ofthe trypan blue exclusion test. HUVECS were frozen in 10% DMSO at -70°C or in liquid nitrogen. 52‘ FREEZING OF HUVECS HUVECS from passage one to five were stored in -80°C for nine months or less. For longer storage, the endothelial cells were frozen in liquid nitrogen. Briefly, the HUVECS were detached from plates by adding PBS-EDTA. These cells were centrifuged at 300xg for five minutes, and resuspended in 10% DMSO in M-199 at a concentration of 3 x 105 cells/ ml. HUVECS were aliquoted in freezing tubes and placed in the -80°C freezer. For longer storage, cells were transferred to liquid nitrogen after a 24 hour incubation at -80°C. CELL VIABILITY USING TRYPAN BLUE STAIN Trypan Blue was used in the dye exclusion procedure for the determination of viable cells. After the cells were detached, centrifuged and re-suspended, different dilutions (1:10 to 1:100 depending on the platelet concentration) were made with trypan blue. The dilution was incubated for 10 minutes at room temperature. Using a Newbauer hemocytometer, the number of stained (dead) and the unstained (viable) cells was counted. The percent ofliving cells and the number of cells/ ml were calculated. The cells were not allowed to be exposed to Trypan blue for more than 20 minutes because viable cells may begin to pick up the dye. 53 ISOLATION OF CORD BLOOD LYMPHOCYTES Using gradient centrifugation, lymphocytes were isolated using the method of Suga et al. (255) with slight modifications. Five ml of heparinized cord blood was layered over 6 ml of Histopaque 1077 in 15 ml conical tubes. The tubes were centrifuged at 400xg for 30 minutes. The upper plasma layer was removed and discarded. The interfaces that contain the lymphocytes were removed and pooled. The pool was diluted with half-equal volume of phosphate buffered saline (PBS) pH 7.4, 2% controlled process serum replacement (CPSR-4) and mixed. Five ml ofthe diluted lymphocytes were layered over 5 m1 Histopaque 1077 in 15 ml conical tube. After centrifugation at 400xg for 30 minutes, the interfaces were collected and pooled together in a 50 ml tube. The volume was brought up to 40 ml with PBS containing 2% CPSR-4. The suspension was centrifuged at 250g for 10 minutes and the pellet was resuspended in RPMI-l640 supplemented with 10% CPSR-4, 2 mM L-glutamine, 2000 U/ml penicillin, 200 ug/ml streptomycin, 150 U/ ml amphotericin-B, and 1 ug/ ml Phytohemagglutinin-P (PHA). The lymphocytes were counted using a hemocytometer, and the volume was adjusted to l x 10‘ cells/ ml. These cultures were incubated in T-75 flasks in the CO2 incubator at 37°C for 2 to 3 days (255). 54 VIRUS INFECTION OF CORD BLOOD LYMPHOCYTES Two to three days after PHA stimulation of lymphocytes, two thirds ofthe medium were removed and HHV-6 was added at 1% volume of uninfected culture. After two hours adsorption, fresh supplemented RPMI-1640 was added. The culture was monitored for cytopathic effect (CPE) as determined by the formation of macrocytic lymphocytes. Peak CPE occurred 10-14 days after infection. When 3+ or 4+ CPE was observed, the cell suspension was frozen-thawed twice to lyse the lymphocytes and centrifuged at 300xg for 10 minutes. The supernatant was mixed with skim milk, aliquoted, titrated and stored at -70°C (13). VIRUS TITRATION USING KARBER METHOD Using the isolation procedure for cord blood lymphocytes, the lymphocytes were diluted to a concentration ofl x 10‘ cells/ ml. One hundred ul was added per well ofa 96 well plate (Table 3) followed by 200ul of cord blood medium. The plate was incubated for 2 to 3 days at 37° C in the CO2 incubator. Ten fold serial dilutions ofHHV-6 (10" to 10 3 were prepared in RPMI-1640. Two hundred ul ofthe culture medium from each well was aspirated off and 100 ul of each viral dilution was added to six wells. The plate was incubated for two hours at 37° C for HHV-6 to adsorb to the lymphocytes. After the incubation 55 time, 100 ul ofthe supplemented medium was added to bring the volume to 300 ul (255). The plate was incubated for twelve days and the CPE in each well was graded for cell death using the following scale: 0 No damage i ? damage 1+ 25% damage 2+ 50% damage 3+ 75% damage 4+ 100% damage 0 to 1+ were considered negative scores 2+ to 4+ were considered positive scores The titer was calculated using the method of Karber where: Virus end point = L-[L,(S-0.5)] L = log oflowest dilution tested L, = log ofinterval between dilution steps (log of dilution) S = sum of culture mortalities The Infective Dose was calculated according to R. Dulbecco and H. S. Ginsburg (87). TCID50 = 50% tissue culture infective dose TCIDSO titer = 10 Virus end point one TCID50 is equal to 0.7 PFU (plaque forming unit) 56 Table 3. Virus Titration Using Kirber Method 96 WELL PLATE l 2 3 4 5 6 7 8 9 10 ll 12 A H20 --- --- --- --- --- --- --- --- --- --- ---> B C D E F G H H20 --- --- —-- --- --- --- --- -..- um um ---> Rows A and H contain water to avoid dryness ofthe wells. B1 to G12 wells were infected with different concentration of the virus. The wells (B to G) in each column have the same concentration of the virus. The concentration ofthe virus is different among columns by a factor often. Column six was used as a control and did not contain any virus. 57 VIRAL INFECTION OF ENDOTHELIAL CELLS MONOLAYER Confluent HUVECS from second to fifth passage were used for infection experiments. Dilutions were made with nutridoma to expose the endothelial cells to HHV-6. The HUVECS were incubated with the HHV-6 for 60 minutes (275). After the incubation, the cells were washed and the maintenance medium was replaced. Incubations with the HHV-6 were carried out at 37° C for periods ranging from 1 to 48 hours. Cells from the same pool of umbilical cords were used in all experiments where virus-infected and uninfected HUVECS were compared. HHV-6 infection did not influence HUVECS viability and integrity asjudged by Trypan blue exclusion. DETECTION OF HHV-6 IN ENDOTHELIAL CELLS The HUVECS were grown on eight well slides and infected with HHV-6 at different concentrations. After 17 hours ofincubation, the slides were fixed with methanol for 15 minutes, and then immuno- stained with FITC-conjugated monoclonal antibody against HHV-6. A fluorescent scope with a 495 nm filter was used to determine glycoprotein expression ofthe virus. 58 PLATELET ISOLATION Blood was drawn into 30 ml syringe containing 3.8 % tri-sodium citrate to give a final anticoagulant concentration of 0.38%. To obtain platelet rich plasma (PRP), the blood was centrifuged for 10 minutes at 200xg. The PRP was transferred to a plastic tube containing a final concentration of 1 uM PGEl. The tube was covered and the platelets were allowed to rest for 10 minutes at room temperature. To concentrate the platelets, the PRP was centrifuged at 700xg for 10 minutes. The pellet was resuspended in 1 ml onyrode’s buffer without calcium. One uM of PGE1 was added and the platelets were allowed to rest for an additional 10 minutes at RT. A 10ml column was prepared using Sepharose 48-200 and was equilibrated with Tyrode’s buffer. The platelet suspension was layered onto the surface ofthe column and gel filtered to remove residual proteins. The first 25 cloudy drops were collected. The platelets were counted using a hemocytometer and diluted to a desired concentration depending on the experiment conducted. The final platelet suspension was allowed to rest for 30 minutes before being used. 59 PLATELET-HUVEC ADHERENCE AND AGGREGATION ASSAY Platelet binding to endothelial cells was measured as described by Zwaginga et al. with some modification (291). A known concentration of platelets was added to HHV-6 infected and non-infected HUVECS grown to confluence. HUVECS were pretreated with indomethacin (20 uM) or buffer for 30 minutes, followed, if appropriate, by adding thrombin (0.2 units/ml) for 5 minutes. Diluted platelets were added to give a final ratio of platelets to endothelial cells 1000: 1. After 30 minutes ofincubation at 37°C, the supernatant was aspirated and 1 ul of PGE1 was added. The cells were washed twice with one ml of washing solution. The supernatant and the wash solution were mixed gently and the pooled platelets were left at room temperature for 15 minutes. The platelets in the pool were counted and the percent of adherent platelets was determined. The percent of platelet binding was measured under static and rocking conditions. Some slides were stained using Wright’s stain and evaluated by light microscopy for adherence and aggregation. Aggregation and adhesion in some experiments were examined using Scanning Electron Microscopy. 6O ELECTRON MICROSCOPY Using the method of Schroeter et al. gelatin coated grids were placed in 12-well plates. HUVECS were cultured on these grids at 37°C in the CO2 incubator. At confluency, the cells were washed. A known concentration of platelets was added. The HUVECS and platelets were fixed with 2% glutaraldehyde in 0.1 M Cacodylate. The cells were post- fixed with 0.2 M phosphate buffer and 0.1% osmium tetroxide. To dehydrate these cells, they were treated in series of graded alcohol solutions 20%, 40%, 60%, 80%, 90%, 100%, 100%. The samples were dried in CO2 at a temperature of 40°C and a pressure of 1,350 lb/in2 . The grids were glued on aluminum stubs, and gold coated. Platelet binding was observed using Scanning Electron Microscopy. TISSUE-PLASMINOGEN ACTIVATOR ACTIVITY ASSAY This assay is based on converting plasminogen to plasmin by t-PA (tissue-plasminogen activator) (138). The amount of plasmin generated is proportional to the concentration oft-PA present. Ninety- six well, fiat bottomed microtiter plates were used. To each well, 5 ul of samples or standards, 5 ul of 0.4 ug/ml plasminogen, 5 ul of cyanogen bromide cleaved fibrinogen, and 15 ul of 0.1 M glycine-tris buffer containing 0.5 mg/ml BSA were added, and the plate was incubated at 37°C for 1 hour to allow for the activation of plasminogen to plasmin. 61 After the incubation, 250 ul of 100 parts PBS, 1 part of 0.1 triton x-100, 1 part of 22 mM DTNB, and 1 part of 20 mM Z-lys-SBZL was added to each well and the plate was incubated for 30 minutes. The absorbance, which is proportional to the activity oft-PA, was measured at 410 nm. PLASMINOGEN ACTIVATOR INHIBITOR-l ACTIVITY ASSAY PAI-l activity was measured in the samples by its ability to inactivate t-PA (138). Five ul of 10 IU/ml t-PA was challenged with 5 ul of the samples in the absence of fibrinogen fragments. The mixture was incubated at 37° C for 10 minutes to allow the PAH in the samples to inhibit t-PA. Five ul plasminogen, 5 ul of cyanogen bromide cleaved fibrinogen, and 10 ul of0.1 M glycine-tris buffer containing 0.5 mg/ml BSA were added, and the plate was incubated at 37°C for 1 hour to allow for the activation of plasminogen into plasmin. The rest ofthe assay is similar to the t-PA assay. ELISA for t-PA and PAL! CONCENTRATION MEASUREMENT Flat bottom microtiter plates were coated using 100 ul/ well of 5 ug/ ml antibody (catcher) diluted in 0.05 M Na2C03, pH 9.6, 0.2% Na azide. They were then incubated overnight at 37°C. The following day the plates were washed three times with washing buffer (0.01 M NazHPOh PH 7.4, 0.015 M NaCl, and 0.01% tween 20). Two hundred ul 62 ofthe washing buffer with 1% BSA was added to each well and the plates were incubated for 30 minutes at 37°C to decrease non-specific binding. The plates were washed three times with PBS pH 7.4. After washing, 100 1.11 ofthe standards or samples were added to each well. After incubating the plates for one hour at 37°C, the wells were washed three times with PBS. Then, One hundred ul/ well of 5 ug/ ml biotinylated antibody (chaser) was added to each well and the plates were incubated for one hour at 37°C. After the wells were washed three times with PBS, 100 ul/ well of avidin was added, and the plates were incubated for one hour at 37°C. After the incubation, the wells were washed three times and 100 ul of alkaline phosphatase labeled biotin was added to each well, incubated for one hour at 37°C, and washed three times. P-Nitrophenyl Phosphate (PNPP) diluted in 100 mM Na2C03, 10 mM MgCl 2 pH 9.5 was added at a concentration of0.012 mg/ well. Five ug/ml of cycloheximide was added to certain wells three hours post treatment to inhibit protein synthesis. The plate was incubated for 45 minutes, and the absorbance was read at 410 nm. ANTIBODY LABELING WITH BIOTIN One ml of 1 mg/ml ofthe catcher (t-PA or PAI-l purified IgG) in 0.1 M NaHCO3 , pH 8.4 was mixed with 0.1 ml of 1.1 mg/ml N-OH succinimydil biotin in DMSO. This mixture was incubated for 2 hours at 63 RT. The reaction was stopped by adding 0.1 ml of 1 M glycine in 0.1 M NaHCO3, pH 8.4. To separate the free from the bound biotin, Centricon microconcentrator type-Y M was used. Up to 2 ml of the sample was added into the top reservoir. The reservoir was covered with a cone- shaped cup and centrifuged at 700xg for one hour in a fixed angle centrifuge. The fluid that remained on top ofthe membrane contained the biotinylated antibody. The top half was detached and inverted. The top half was centrifuged at 700xg for 15 minutes to spin concentrate the labeled antibody in the cone shaped cup. The labeled antibodies were stored at -80°C. MEASUREMENT OF CELLULAR PROCOAGULANT ACTIVITY This assay is based on the method described by Muller et al. (175) with some modifications. HUVECS were grown in 24 well plates coated with 2% gelatin, and the procoagulant activity on the endothelial cells surface was measured as described by Muller et al. with some modifications, and without detachment ofthe cells. After 17 hour incubation of the endothelial cells with or without HHV-6, the culture medium was aspirated off and the HUVECS were washed twice with PBS. Four hundred 111 of normal plasma or factor deficient plasma was added to each well and the endothelial cells were incubated for 15 minutes at 37°C. After the incubation, 200 ul of CaCl2 was added to 64 recalcify the plasma and start the coagulation process. A hook was used to detect fibrin formation. The time from the addition of CaCl2 to the detection ofthe first fibrin strand corresponds to the clotting time. PURIFICATION OF PLASMID DNA Digestion The plasmid clone PAI B6 was a kind gift from David Ginsburg, MD. This plasmid is a full length human PAI-l cDNA clone encoding the shorter mRNA species. This two kb insert is cloned into Eco RI site of pUC-l3. The plasmid clone human t-PA cDNA was a kind gift from Sandra J. Degen, Ph.D. This cDNA was cloned into the Pst I site of pBR322 by dC tailing. Eco RI and Bgl II were used to digest plasminogen activator inhibitor I (PAH) and tissue plasminogen activator (t-PA) genes cloned into plasmids respectively. To a tube, 100 ug of PAI-l or t-PA cloned plasmid DNA was added. Twenty ul ofthe corresponding enzyme and fifty ul of 10x REACT 3 buffer (500 mM Tri-HCl pH 8.0, 100 mM MgC12, 1M NaCl) were added. GD-H20 was added to bring the volume to 500 ul. The contents were mixed well by tapping the tube and 65 then spun down for one second in a microcentrifuge. Since incomplete mixing can reduce digestion, this process was repeated. The tubes were incubated for one hour at 37°C. Precipitation After the incubation, 480 ul of PAH or t-PA digested DNA was precipitated by adding 48 ul of 3M sodium-acetate pH5.0 and 960 ul of 100% ethanol. These mixtures were inverted, then incubated at -20°C. After 20 minutes incubation, the tubes were centrifuged for 15 minutes at 12,000xg. The supernatants were drained off and the sediments were washed with 0.5ml of 70% ethanol and then centrifuged at 12,000xg for 15 minutes. After the centrifugation, the ethanol was poured off and the sediments were air-dried at RT. The pellets were resuspended in 40 ul TE-8 (10ml of 1.0M Tris-Cl pH8.0, 5ml of 0.2M EDTA pH8.0, 985 ml Glass Distilled(GD)-H20). Electroelution into Dialysis Bag After the cloned plasmid DNAS were digested and concentrated, the fragments were separated by electrophoresis through agarose gels containing 0.5 ug/ml ethidium bromide. The bands ofinterest (1.5 kb t- PA cDNA and 2 kb PAI-l cDNA) were located using long-wavelength ultraviolet light. The gel was photographed. Using a sharp razor blade, 66 the smallest possible slice of agarose containing the band ofinterest (PAI-l or t-PA cDNA) was cutout. A piece of dialysis tubing was sealed at one end with a dialysis clip. The cut slice of agarose was placed in the dialysis bag containing just enough TE-8 to keep the gel slice in contact at all times with the buffer. The dialysis bag was clipped just above the gel slice without trapping air bubbles inside, and then it was submerged in 1x TAE (SOXTAE = 242g Tris base; 57.1ml glacial acetic acid; 100ml of 0.5M EDTA pH 8.1; water to one liter) in an electrophoresis chamber. An electric current (100 mV) was passed through the bag for 45 minutes and the DNA was electroeluted out of the agarose gel onto the inner wall of dialysis bag. The electroeluted DNA were checked with long-wavelength ultraviolet light. The bag was re-immersed into the leAE buffer in the electrophoresis chamber. To release the DNA from the inner wall ofthe dialysis bag, the polarity of the current was reversed for one minute. Then, the bag was removed and massaged gently. The buffer from the dialysis bag containing the DNA was transferred to a sterile tube. Extraction with Phenol Each electroeluate was extracted twice with 0.5 ml phenol: chloroform equilibrated with 10 mM sodium-acetate, pH 5.0, 0.1M NaCl and lmM EDTA. After centrifugation , the aqueous phase was 67 transferred to a fresh tube and 0.5ml chloroform: isoamyl alcohol (one to one ratio) was added. The tube was mixed and centrifuged to separate layers. The aqueous phase was transferred to a new tube. Sixty ul Na-acetate and lml of ethanol were added to the aqueous phase and this mixture was incubated at -20°C overnight. The next day , the DNA was recovered by centrifugation at 7K rpm for 20 minutes in a Sorvall centrifuge. After centrifugation, the supernatant was poured off and the pellet was rinsed with 70% ethanol. The specimen was recentrifuged for 15 minutes; the supernatant was poured off and the pellet then air-dried at RT. The DNA was redissolved in 50ul of TE-8. To check the amount and quality of the extracted DNA, 2ul ofthe DNA fragment final preparation was mixed with 8ul of TE-8 and 2ul gel loading buffer (0.75ml deionized formamide, 0.15ml 10xMOPS, 0.24ml formaldehyde, 0.1ml H20, 0.1m1 glycerol, 0.08m] 10% bromphenol blue). This mixture and appropriate markers were loaded in different wells and an agarose gel was run using 250ml of 1x TBE buffer (0.1M Tris base, 0.1M Boric acid, 0.2mM EDTA) containing 12.5 ul ethidium bromide. The gel was examined carefully for the presence of faint fluorescent bands that signified the presence of DNA contaminants. The amount of cDNA ofthe final preparation was estimated from the relative fluorescence intensities of the fragment and A digested with hind III. 68 DIRECT ISOLATION OF POLYADENYLATED RNA FROM HUVECS Using the Dynabeads mRNA kit (DYNAL), polyadenylated RNA was prepared directly from human umbilical vein endothelial cells. In order to condition the Dynabeads oligo(dT)2,, beads were thoroughly suspended before use, then, were transferred to a microcentrifuge tube placed in a Dynal MPC (Magnet). After 30 seconds, the supernatant was discarded and the tube was removed from the Dynal MPC. The beads were resuspended in 0.25ml oflysis/binding buffer. When the lysate was ready for combination with the heads, the magnet was applied and the supernatant was removed. HUVECS were washed twice with phosphate-buffered saline (PBS). PBS containing EDTA was used to detach the endothelial cells. A pellet was prepared by centrifuging the detached cells. The supernatant was discarded and replaced with 1.0 ml of the kit lysis/binding buffer. The buffer was aspirated in and out through a pipette tip in order to lyse the cells completely. The viscous lysate was combined with the Dynabeads and annealed by rotating for 5 minutes at room temperature. Then, the tube was placed in the Dynal MPC for 2 minutes and the supernatant was removed. The beads were washed twice with 0.5ml of kit washing buffer with LIDS (10 mM Tris/HCl PH 8.0, 0.15 M LiCl,1mM EDTA, 0.1% LIDS) and once with 0.5 ml of washing buffer (10 mM Tris/HCl PH 8.0, 0.15 M LiCl, 1 mM EDTA). 69 After removing the last washing buffer, the mRNA was eluted by adding 20 ul of elution solution (10 mM Tris-HCl PH 8.0) and incubated at 65°C for 2 minutes. After incubation, the tube was placed in the Dynal MPC and the supernatant was transferred to a new microcentrifuge tube. To avoid cross-contamination between specimens, the used Dynabeads Olio (dT)25 were resuspended in 200 111 of reconditioning solution (0.1M NaOH) to regenerate the beads for reuse. The suspension was transferred to a new tube mixed very well and incubated at 65°C for 2 minutes. After incubation, the tube was placed in the Dynal MPC for one minute and the supernatant was removed. The regeneration protocol minus the incubation at 65°C was repeated twice. The beads are then resuspended in 200 ul of storage buffer Olio (dT)25 (250mM Tris-HCl pH 8.0, 20mM EDTA, 0.1% Tween 20, 0.02% sodium azide). The last washing with the storage buffer was repeated twice and the beads were stored at 4°C ready for another mRNA isolation. ISOLATION OF TOTAL RNA One ml of TRI REAGENT (Molecular Research Center Inc.) was added to HUVECS grown in a 100 mm2 culture dish. A policeman was used to scrap the lysed cells. The endothelial cell lysate was passed several times through a pipette and the homogenate was transferred to a 70 microcentrifuge tube. Two hundred ul of chloroform was added to the homogenate and the tube was shaken vigorously for 15 seconds. After incubating the mixture for two minutes at RT, the tube was centrifuged at 12k g for 15 minutes. Following centrifugation, the upper aqueous phase, that contains the RNA, was transferred to a clean microcentrifuge tube. The RNA was precipitated by adding 0.5ml of isopropanol. The mixture was inverted for 10 seconds and incubated for 10 minutes at RT. Following incubation, a pellet was obtained by centrifuging the tube at RT for 10 minutes. The supernatant was removed and the pellet was washed with one ml of 75% ethanol by vortexing and Subsequent centrifugation at 12k g for 2 minutes. The supernatant was discarded and the RNA pellet was air-dried at RT. The total RNA was resuspended in TE-8. Using a spectrophotometer, the absorbance ratio at 260/280 was determined and the final concentration of total RNA was calculated. NORTHERN BLOTTING AND HYBRIDIZATION Gel Preparation and Electrophoresis One percent agarose [2.5 gm agarose, 25 ml 10X MAE (0.2 M MOPS, 50mM sodium acetate, 10mM EDTA, pH7.0), 212 ml GD-HZO 71 heated to dissolve and cooled to 50°C] was prepared and thoroughly, but gently, mixed with 13ml of37% deionized formaldehyde. The mixture was poured into a gel mold. The gel was allowed to sit for at least one hour before use. Ten ug oftotal RNA was mixed with 25 ul ofloading buffer (0.75ml deionized formamide, 0.15ml 10X MAE, 0.24m] deionized formaldehyde, 0.2ml of 50% glycerol, 0.08ml bromophenol blue). The mixture was heated at 65°C for eight minutes and chilled on ice. Two 111 of0.5 mg/ml ethidium bromide was added. The mixtures and a standard were loaded into different wells and electrophoresed in 1X MAE until the bromophenol blue reached near the bottom ofthe gel. The gel was photographed on a long wavelength UV source. Transfer Preparation The gel was soaked twice in GD-HZO to remove the formaldehyde. Using Turboblotter from Schleicher and Schuell, the nitrocellulose blot paper and 4 sheets of 3M Whatman papers were soaked in 10x SSC (1.5M NaCl, 0.30M sodium citrate, PH 7.0). The transfer was set up by the manufacturer’s instructions using 10x SSC as the transfer buffer. Paper towels, 3M Whatman papers and the nitrocellulose blot paper were cut as big as the gel. Almost 3 to 3.5cm of paper towels were placed in the stack tray of the transfer device. Four 3M Whatman 72 papers were placed on top ofthe paper towel stacks. Then, one prewet 3M Whatman paper was placed on top ofthe stack. The transfer membrane, without trapping any air bubbles, was placed on the stack. Then, the gel was placed on top ofthe nitrocellulose membrane, making sure that no air bubbles were trapped in between. Three prewet sheets of 3M Whatman were placed on top ofthe gel. The buffer tray ofthe transfer device was attached to the stack device and 10x SSC was added. Finally the transfer was started by connecting the gel stack with the buffer tray using a presoaked buffer wick. To prevent evaporation, the wick cover was placed on top ofthe stack. The transfer was completed in 3 hours. After the transfer, the nitrocellulose membrane was allowed to air-dry. Then, the RNA was fixed by baking the blot at 80°C in a vacuum oven for 45 minutes. RANDOM PRIMERS DNA LABELING Using the manufacturer’s instructions (Life Technologies, GIBCO BRL), 25 ng of cDNA was denatured in a microcentrifuge tube by heating for 5 minutes in a boiling water bath, then cooling immediately on ice. The following reagents were added on ice: 2ul dATP, 2ul dGTP, 2ul dTTP, 15ul random primers buffer mixture (0.67M HEPES, 0.17M Tris-HCl, l7mM MgC12, 33mM 2-mercaptoethanol, 1.33 mg/ml BSA, 180 D260 units/ml oligodeoxyribonucleotide primers 73 PH6.8), 5ul (approximately 50uCi) [a-P32]dCTP, GD-HZO to a total volume of 49ul. This cocktail was mixed briefly. One ul Klenow fragment was added . The solution was mixed gently, but thoroughly, then centrifuged briefly. After one hour incubation at RT, 5ul of stop buffer (0.2M NazEDTA PH 7.5) was added. The labeled probe was separated from the free nucleotides by chromatography using the TB Midi SELECT-D G50 microcentrifuge spin column. After brief centrifugation, the purified labeled cDNA was recovered from the column without a significant change in volume (5 Prime, Sephadex, Pharmacia Inc.) Two ul ofthe labeled probe was diluted with 498 1.11 of GD-HZO. Five ul of this dilution was spotted on a Whatman filter. The filter was washed three times with 50 ml ofice cold 10% TCA containing 1% sodium pyrophosphate and once with 50 ml of 95% ethyl alcohol at RT. The filter was dried under a lamp and the precipitable radioactivity was determined by liquid scintillation counting. Following the kit’s instruction, a reading ofX cpm corresponds to a total of 2750X cpm incorporated in the entire incubation mixture. The number 2750 is a constant given by the manufacturer. This dilution and the procedure were repeated without washing with TCA. A reading ofX cpm corresponds to 2750X cpm oftotal radioactivity in the entire reaction mixture. 74 Northern Prehyb/Hybridization and Autoradiography To prepare 50 ml of prehybridization/ hybridization solution, the following reagents were added to a sterile tube: 25 ml of 100% deionized formamide, 12.5 ml of 20x SSC, 2.5 ml of 1M sodium phosphate, 2.5 ml of 100x Denhardt’s reagent, 0.5 ml of10% SDS, 0.5 ml Of10 mg/ml complementary DNA, 1.25 ml of10 mg/ml tRNA, 5.25 ml of GD-HZO. The complementary DNA and tRNA were heated in a glass tube at 90°C for 5 minutes before adding to the above solution. For a standard Northern blot, 20 to 25 ml of prehybridization solution was used. The prehybridization was conducted at 42°C with shaking for two hours to overnight. Finally the probe was denatured by heating in a boiling waterbath for 5 minutes, cooled on ice, and added directly to the prehybridization solution. The hybridization was conducted for six hours to overnight. After hybridization, the blot was washed 3x for 30 minutes at 55°C to 60 °C in 0.1x SSC/ 0.1% SDS. An autoradiograph was established by exposing the filter for four hours to three days to X-ray film at -70°C with an intensifying screen. An automatic developer was used to develop the X-ray film. 75 STATISTICAL ANALYSIS Results are expressed as means :1: the standard error ofthe mean (SEM). Data were analyzed using StatMost 2.5 for Windows, DataMost Corporation, P.O.Box 65389, Salt Lake City, UT 84165. The effects of thrombin and/ or HHV-6 on endothelial cells were analyzed utilizing unpaired t-test when comparing two groups and ANOVA when comparing more than two groups. 76 REAGENTS/ SUPPLIES 8 well slides 96 well plates amphotericin Bgl II BSA Collagenase Cord blood CPSR Cyanogen bromide Cycloheximide DMSO DTNB Dynabeads mRNA Kit ECGF Eco RI Fetal calf serum Gelatin glutamine Heparin COMPANY Nunc Dynatech Sigma Gibco Sigma Gibco Saint Lawrence Hospital Sigma Sigma Sigma Sigma Sigma Dynal Sigma Gibco Intergen Sigma Gibco Sigma 77 REAGENTS/ SUPPLIES HHV-6 Histopaque-1077 M-l99 Microconcentrator Nutridoma PAI-l cDNA Penicillin Petri dishes PHA Plasminogen PNPP RPMI-1640 Sepharose 4B-200 streptomycin T-75 flasks T-PA cDNA Tri reagent Trypan blue Umbilical cords Z-lys-SBZL COMPANY Kind gift of Norma Barratt, PhD. Sigma Mediatech Amicon Bohringer Mannheim Kind gift of David Ginsburg, M.D. Sigma Corning Sigma Calbiochem Calbiochem Sigma Sigma Sigma Corning Kind gift of Sandra Degen, Ph.D. Molecular research center Sigma Saint Lawrence Hospital Calbiochem 78 Results Culture of Human Umbilical Vein Endothelial Cells After four to five days in culture, human umbilical vein endothelial cells grew to confluent monolayer without any demarcated whirling pattern. The endothelial cells were homogeneous, large, closely opposed, polygonal with a central nucleus and indistinct cell borders (Figure 5). The incubation times used in all experiments were associated with greater than 93% cell viability using trypan blue stain. Figure 6 represents pure culture ofendothelial cells stained with FITC labeled monoclonal antibodies to human factor VIII antigen. Factor VIII is present on endothelial cells and absent from the surface of epithelial, smooth, and other cells. Isolation and Infection of Cord Blood Lymphocytes Figure 7 shows established umbilical cord lymphocytes six days post-infection in cell culture. Lymphocyte clumps were seen due to the cell multiplication stimulated by the addition of PHA. Starting from day six post-infection with HHV-6, some lymphocytes became macrocytic. More and more lymphocytes became infected and kept on increasing in size until they burst. Figure 8 shows characteristic large cell formation of cord blood lymphocytes infected with HHV-6 after twelve days incubation. 79 Figure 5. Cultured human umbilical vein endothelial cells. Endothelial cells were isolated and cultured as described in “Materials and Methods”, stained using Wright’s stain and photographed after five days in culture. 80 81 Figure 6. Cultured human umbilical vein endothelial cells. Endothelial cells were isolated and cultured as described in “Materials and Methods”. FITC labeled monoclonal antibody to human factor VIII antigen was used to test for purity of cell cultures. Bottom picture is the same field as the top one using 495 nm fluorescent filter. 82 {-‘H' \‘I :‘J‘ _ I 3'- in ta 1%. ‘W at Jo’- 83 “9 1'1 Figure 7. Cord blood lymphocytes, six days post-infection. Blood lymphocytes six days post infection cultured in RPMI 1640 supplemented as described in “Materials and Methods” 84 85 Figure 8. Cord Blood Lymphocytes, 12 Days Post Infection. Lymphocytes cultured for 12 days in RPMI 1640 supplemented as described in “Materials and Methods” and showing a characteristic large balloon cell formation twelve days post infection. (40X magnification: tOp, 10x magnification: bottom) 86 87 Table 4 shows a 96 well plate used to titrate HHV-6. Using Karber method, each well was graded for cell death. TCIDso was equal to 10"‘3 (l.9x10° PFU/ml). Figure 9 shows that HHV-6 can infect HUVECS. Using a monoclonal antibody labeled with FITC against the early viral antigen, we were able to detect the virus in endothelial cells. This virus was still infectious demonstrated by infecting lymphocytes by HHV-6 recovered from HUVECS. Platelet Adherence and Aggregation to HUVEC When platelets in Tyrode’s buffer were incubated with endothelial cells, under static condition, in the presence of0.2 U/ml of thrombin for 5 minutes, 11% of the added platelets adhered and aggregated on the surface of the endothelial cells compared with 4% platelet adhesion and aggregation to the control endothelial cells in the absence ofthrombin (Figure 10). Similar results were seen using Hanks buffer alone or with thrombin. The results were the same regardless of whether the HUVECS were incubated with thrombin, then the thrombin was removed and HUVECS were washed before adding platelets, or the thrombin was left on HUVECS for five minutes then platelets were added or whether thrombin and platelets were added to HUVECS at the same time. The addition of calcium did not affect the outcome of the experiments. Figure 10 shows that pretreatment of human umbilical vein 88 Table 4. A ninety-six well plate was prepared as described under “Materials and Methods”. Rows A and H were filled with water. Lymphocytes were grown in rows B, C, D, E, F, and G. The wells in column 6 were not infected with HHV-6. The rest of columns were infected with different concentration ofthe virus. All the wells in a column were infected with the same concentration ofHHV-6. The virus end point titer is 10"”3 TCIDso lml (1.9x10° PFU/ml). Data are results of four experiments. 89 'ith La] :51 of [115. FCIDE Human Herpes Virus Type-6 Titration 1 2 3 4 5 7 8 9 10 11 12 A H10 "‘ -- " "' -- -- -- D- II- --> B 3+ 3+ 2+ 2+ 1+ 4+ 3+ 2+ 2+ 1+ + C 3+ 3+ 3+ 3+ 1+ 3+ 3+ 2+ 3+ 2+ 1+ D 3+ 3+ 3+ 3+ 1+ 3+ 3+ 3+ 2+ 1+ 1+ E 3+ 3+ 2+ 2+ 2+ 3+ 3+ 2+ 2+ 1+ + F 3+ 3+ 2+ 2+ 2+ 1+ 3+ 3+ 3+ 3+ 2+ 1+ G 3+ 3+ 2+ 2+ 1+ 3+ 3+ 3+ 2+ + + H H20 -- -- -- -- -- -- -- -- -- --> Virus Dilutions: 10‘ 10‘2 10" 10“ 10‘s 10‘1 10‘2 10“ 104 10" 10*5 Score: 6/6 6/6 6/6 6/6 2/6 0/6 6/6 6/6 6/6 6/6 2/6 0/6 90 Figure 9. Immunostaining for HHV-6. Human umbilical vein endothelial cells, grown on chamber slides coated with 2% gelatin, were infected with HHV-6. Seventeen hours post- infection, the slides were fixed with methanol for 15 minutes and then immunostained with FITC conjugated monoclonal antibody against HHV-6. A fluorescent scope with 495 nm filter was used to determine glycoprotein expression ofthe HHV-6. Infected endothelial cells appear greenish in color (40x magnification). 91 92 Figure 10. Effect of thrombin on platelet adherence to HUVEC under static and rocking conditions. Endothelial cells were incubated with Tyrode buffer or 0.2 U/ml ofthrombin. After five minutes ofincubation, platelets were added at a concentration of 100011 (platelets: endothelial cells). Data represent mean + SEM ofthree experiments with three replicates per experiment. 93 [:1 Thrombin -Control Static P<0.05 P<0.05 i 0 m w I02I¢IID< .PIJIP‘i—L at 94 endothelial cells with 0.2 U/ml of thrombin increased platelet adherence to the same degree with rocking or static conditions and there was no difference between these conditions after 30 minutes of platelet incubation (P>0.05). All further studies were conducted under static conditions. Figure 11 shows platelet adherence to HUVECS that were stimulated with thrombin for 5 minutes. The percent adherence was dependent on the dose ofthrombin used. An increase in platelet binding was seen with as little as one mU/ml of thrombin. The adherence increased gradually with increase in thrombin concentration to reach a maximum at 0.2 U/ml ofthrombin or higher (P<0.05). The percent of platelet binding to HUVECS reflects a balance between the direct effect ofthrombin on platelet adhesion and aggregation to endothelial cells and the thrombin induced production of prostacyclin from endothelial cells, which is an inhibitor of platelets. The preincubation of uninfected HUVECs with thrombin, followed by the addition of platelets resulted in only a 5% increase in platelet binding, unless prostaglandin synthesis in endothelial cells was inhibited by indomethacin. In the presence of indomethacin the percent of platelet adherence on thrombin stimulated endothelial cells increases from 5% to 37% (Figure 12, Table 5). When platelets were incubated in Tyrode’s buffer with HHV-6 95 Figure 11. Effect of thrombin concentration on platelet adhesion to endothelial cells. HUVECs were incubated with different concentrations ofthrombin for five minutes. The fluid was removed and the endothelial cells were washed twice with PBS. Platelets were added and incubated for 30 minutes. After incubation, the % platelet adherence was determined . Data represent mean + SEM ofthree experiments with three replicates per experiment. 96 COEOLUO u0_0um_n_ Axe 001 002 004 DJ 02 0001 Thrombin conc lU/mll 97 infected endothelial cells in the absence ofthrombin for 30 minutes, 16% ofthe added platelets adhered compared with 5% in the control endothelial cells. The addition of0.2 U/ml of thrombin to infected HUVECS induced a significant increase in platelet binding to the endothelial cells (70%) compared to only 10 % platelet adherence to non-infected HUVECS stimulated with thrombin. The addition of indomethacin resulted in a further increase in platelet adherence up to 80% (Figure 12). Using scanning electron microscopy or Wright’s stain, large aggregates of platelets were formed on the surface of the HHV-6 infected endothelium stimulated with thrombin (Figure 13) compared with smaller platelet aggregates formed on thrombin stimulated non- infected endothelium or on HHV-6 infected endothelium without thrombin. Whereas, on the control endothelium ( non-stimulated with thrombin and non-infected with HHV-6) there were relatively few adherent platelets Because adherent platelets are activated and aggregated, we decided to determine whether the increase in platelet adherence was a property of platelet aggregators. ADP, as an aggregator, added to HHV-6 infected endothelial cells before adding platelets or co- incubated with HHV-6 infected endothelial cells and platelets, did not cause an increase in platelet adherence to endothelial cells (Table 5). 98 Figure 12. Adherence of platelets to uninfected and 18 hour infected human umbilical vein endothelial cells. Non-infected and HHV-6 infected HUVECS were pretreated with 20uM of indomethacin or Tyrode buffer for 30 minutes followed if appropriate by adding 0.2 units/ ml ofthrombin for 5 minutes. Platelets were added, and after 30 minutes ofincubation, the % of adherent platelets was determined. Bars represent mean + SEM of five experiments with three replicates per experiment. 99 I" :ted % of 3M0f % of PLATELET ADHERENCE 100 888 —L CO 88888 - NOT INFECTED [:1 INFECTED NM NM P<0.05 H H f i BUFFER Ila lNDOllET+lla PRETREATMENT of HUVEC with HHV-6 100 Figure 13. Platelet Binding to HUVECS. Platelet aggregate adherent to non-infected and HHV-6 infected HUVECS according to “Materials and Methods”. Non-infected HUVECS stained with Wright’s stain , 10x magnification (Page 102 top). Platelet adherence to non-infected HUVECS stained with Wright’s stain, 40x magnification (Page102 bottom). Platelet adherence to HHV-6 infected HUVECS pretreated with thrombin stained with Wright’s stain, 40x magnification (Page 103 top), scanning electron micrograph (Page 103 bottom). 101 102 103 Table 5. Adherence of Platelets to Non-Infected and 18 Hour Infected HUVECs. Non-infected and HHV-6 infected endothelial cells were cultured as indicated in “Materials and Methods”. HUVECS were pretreated with either 20 uM of indomethacin or buffer for 30 minutes, followed if appropriate by adding 0.2 U/ml ofthrombin for 5 minutes. HUVECS were also stimulated using 2 uM of ADP for 5 minutes. Data represent means 4: SEM ofthree experiments with three replicates per experiment. 104 Adherence of Platelets to Non-Infected and 18 Hour Infected HUVECS NOT-INFECTED HHV-6 INFECTED Tyrode Buffer 5 :1: 1.5% 16 i 1.4% Thrombin 10:1: 1.3% 70:1:2.2% Indomethacin & 37 +2. 1% 80 :t 2.6% Thrombin ADP 10+].4% 21+1.3% 105 Effect of HHV-6 on Tissue-Plasminogen Activator and Plasminogen Activator Inhibitor- 1 Colorimetric and immunologic assays were used to quantify the effect ofHHV-6 on the secretion oft-PA and PAL] from HUVECS. The colorimetric chemical assay was used to measure the activity of free t- PA and free PAI-l. Where as the immunologic assay was used to measure the concentration ofthe total (free and bound) t-PA and PAI-l. Figures 14, 17, and 19 show t-PA and PAI-l standard curves for activities or concentrations and their corresponding R2 . The basal level oft-PA antigen and PAI-l in non-infected control human umbilical vein endothelial cells were 6.34 IUand 34 ng/ml respectively (Table 6). The activity oft-PA was zero because ofthe high concentration ofthe inhibitor which inhibited all ofthe t-PA in the culture media. Stimulation ofendothelial cells with one IU/ml of thrombin led to an increase in t-PA antigen ( 24 IU) and in PAI-l antigen ( 90 ng/ml) compared with the baseline control of 6.34 IU and 34 ng/ml respectively (Table 6). Infection ofHUVECs with HHV-6 resulted in an increase in PAI-l to 74 ng/ml (Figure 18) without any Significant change in t-PA antigen (Figure 20). Cycloheximide (5 ug/ml) did not affect endothelial cell viability as measured by Trypan blue exclusion test ablated the PAI-l response to HHV-6 (Table 6). 106 Table 6. Modulation of t-PA and PAI-l antigens secretion. T-PA and PAI-l secretion to uninfected and 18 hour post-infected human umbilical vein endothelial cells. HUVECS were pretreated with Nutridoma, one IU/ml thrombin, or 100 ul HHV-6 (1900CFU/ul) followed if appropriate by adding 5 ug/ml cycloheximide three hours post-treatment. All wells were incubated for 18 hours. Data represent mean :t SEM ofthree experiments with three replicates per experiment. 107 Modulation of t-PA and PAL] Antigens Secretion Stimulus t-PA antigen PAI-l antigen (IU/ml) (ng/ml) Baseline 6.3 + 0.7 34 +1.6 Thrombin 24 + 1.1 90 + 3.1 HHV-6 5.3 +0.4 74 +2.4 Thrombin + cycloheximide 2.3 + 0.2 10 + 1.3 HHV-6 + cycloheximide 5.1 + 0.3 6.4 + 0.8 108 Figure 14. T-PA standard curve for t-PA and PAI-l activity. T-PA standards were run in triplicate every time the assay was performed. Absorbance was plotted against the activity oft-PA. Standards and Specimens were treated exactly the same. Blank containing M199 medium and the rest ofthe reagents was used to zero the reader. Bars represent mean + SEM. (R2 = 0.998) 109 OD 1.25 LN 0.75 0.50 0.25 0.00 1000 mlU t-PA 110 1500 Figure 15 shows the effect of different concentration of HHV-6 on HUVECS after eighteen hours ofincubation. Changes in PAL] release from infected endothelial cells was Observed with 40 ul ofHHV- 6 (1900 CFU/ul) and was maximal with a concentration of 100 111 ofthe virus or greater. This increase in PAI-l activity is dose dependent. Figure 16 demonstrates that infection of HUVECS with HHV-6 leads to an increase in PAI-l activity compared to the non-infected endothelial cells. The HHV-6 mediated effect on PAI-l activity required three to four hours before changes in protein level could be detected. Four to six hours post infection, the activity increased rapidly. Then, from six hours to eighteen hours post infection, the PAI- 1 activity increased at a slower rate to reach a maximum of 4923 mU/ml compared to 2245 mU/ml in the non-infected control endothelial cells. To measure the concentration (free and bound) oft-PA and PAI-l in the culture supernatant, an ELISA, developed in our laboratory, was used(Figure 17, 19). Figure 18 demonstrates the effect of HHV-6 on PAI-l protein synthesis/ release from HUVECS. There was a significant time dependent increase in PAL] antigen. A slow and steady increase was seen in both the non-infected control and in the HHV-6 infected endothelial cells. The concentration of PAI-l was higher at any given time in HHV-6 infected HUVECS than in control (figure 18) 111 Figure 15. Effect of HHV-6 on PAI-l Activity. PAl-l activity after HHV-6 infection ofendothelial cells in vitro. Confluent monolayers of HUVECS were washed and incubated in nutridoma with increased concentration of HHV-6. After 18 hours of incubation, the media were harvested and assayed for PAI-l activity by a quantitative chemical assay. No significant changes were seen in heat inactivated HHV-6 or freezing medium. Data represent mean + SEM of five experiments with three replicates per experiment (P<0.05). 112 PAl-l ACTIVITY mlU 5200 4800 4400 4000 3000 “ 3200 _ 2800 2400 —— activity J 2000 L l l l l 1 so so 100 120 140 HHV-6 couc ul (1soocrulu1) 113 l 160 1 l 180 200 Figure 16. Kinetics of HHV-6 on PAI-l activity. HUVECS infected with HHV-6 were run in parallel with non-infected cells. Conditioned media were harvested at different time intervals after infection and assayed for PAI-l activity by a quantitative chemical assay. Data represent mean + SEM of five experiments with three replicates per experiment. (P<0.05) 114 PAl-I ACTIVITY mlU -— not infected ----- infected 42 5500 — 5000 — I--- IIIII 1 ........ E 4500 L r j- [I “‘7 TTTTTT ~+ 4000 ~ ,7 I 7777777 i I I 3500 . I/ l 3000 ~ ,I’ I 2500 h [I] M l I 2000 ,9 I I 1 m L 1 1 1 J a 12 18 24 30 36 TIME HOURS 115 Figure 17. PAI-l standard curve (ELISA) for the measurement of PAI-l antigen (free and bound).PAI-1 standards and specimens were run in triplicate every time the assay was performed. Absorbance was plotted against the concentration ofthe standards. Standards and specimens were treated exactly the same. A blank containing M199 medium and the rest ofthe reagents was used to zero the reader. Bars represent mean + SEM.(R2 = 0.999) 116 OD .40r .20L .00” .80i .60‘ .40” .20” l *— 20 MM rig/ml 117 30 40 Figure 18. Kinetics of HHV-6 on PAI-l antigen concentration. Confluent monolayers ofHUVECs were washed and incubated in nutridoma alone or infected with HHV-6. At the indicated time, the conditioned media were harvested and assayed by ELISA for the PAI-l antigen (free and bound).Data represent mean + SEM of five experiments with three replicates per experiment. (P<0.05) 118 no/ml PAI-‘l ANTIG EN “‘* NOT INFECTED ———— INFECTED 90. 80- ,L- / """"""" "'1 7o . \\%fi 60 ~ [I 50~ 40 ~ ,1 . L I 30 L /// //X’”7L 20~ 2,-.1’ 10— 0 / J 1 1 l 1 1 1 l l l %/rlé‘ o 2 4 6 s 101214161820222440 TIME HOURS 119 Figure 19. T-PA standard curve (ELISA) for the measurement of t-PA antigen (free and bound). T-PA standards and Specimens were run in triplicate every time the assay was performed. Absorbance was plotted against the concentration ofthe standards. Standards and specimens were treated exactly the same. A blank containing M199 medium and the rest ofthe reagents was used to zero the reader. 120 OD 2.00 1.50 1 .00 0.50 0.00 T + . l " T +/ //i/jl: J L l J 0 2 4 6 8 1 0 t-PA antigen (lUIml) 121 12 Figure 20. Kinetics of HHV-6 on t-PA antigen concentration. Confluent monolayers of HUVECS were washed and incubated in nutridoma alone or infected with HHV-6. At the indicated time, the conditioned media were harvested and assayed by ELISA for the t-PA antigen (free and bound). Data represent mean + SEM of three experiments with three replicates per experiment. (P >005) 122 E H N w 05 U1 0‘ ‘1 w I I I I t-PA ANTIGE N (:0 N CE NTRATIO N O — Non-lnhcied ----- Infected r // 0 2 4 6 8 10 12 14 16 18 20 22 24 26 TIME HOURS 123 This increase in PAI-l antigen reached a maximum of 74 ng/ml at 18 hours post-infection of HUVECS with HHV-6 compared to 34 ng/ml in the control. The changes in PAI-l antigen concentration were still apparent at 48 to 72 hours post-infection (Figure 18). The concentration of t-PA in the non-infected and infected endothelial cells increased to almost 6 IU/ml after 17 hours of incubation. But, there were no significant difference between the control and the HHV-6 infected HUVECS at any time during sampling (Figure 20). Figure 21 (bottom) shows the fragments of the cloned plasmid DNAs of t-PA and PAI-l after digestion. The bands ofinterest (1.5 kb t-PA and 2 kb PAI-l cDNA), located using long-wavelength ultraviolet light, were cut, electroeluted , concentrated and run on a gel Figure 21 (top ). Figure 22 shows the corresponding 28s and 185 r RNA for t-PA and PAI-l following loading of equal concentration of RNA in each well. The t-PA and PAI-l infection, collection and Northern blots were run at the same time and treated exactly the same. Two PAI-l mRNA forms (3.2 and 2.3kb) were detected in our control and HHV-6 infected extractions (Figure 22). The concentration of both mRNA forms was up-regulated post infection with HHV-6 compared to the control at a particular time suggesting that the increase in PAI-l protein in culture supernatant is due in large part to increase synthesis ofthe inhibitor. 124 There was a preferential increase in the 3.2 kb form. Northern blotting demonstrated that HHV-6 infection of human umbilical vein endothelial cells did not affect t- PA mRNA levels which correlates with the t-PA protein in the supernatant fluid measured by ELISA (Figure 23). t-PA and PAI-l mRNA were used as internal controls for each other. There was no change in t-PA mRNA and an increase in PAI-l mRNA. 125 Figure 21. Purification of t-PA and PAI-l plasmid DNA. Plasmids were digested, precipitated, and electrophoresed (bottom) according to “Materials and Methods”. Then, the agarose around the band ofinterest was cut, placed in a dialysis tubing, electroeluted, extracted, and run on a gel (Figure 21-top). Figures are negative images. 126 t-PA cDNA PAI-l cDNA is t-PA cDNA PAI-l cDNA rs Iopl 9300//// 6600 4800 230 2000 127 Figure 22. Effect of HHV-6 on PAI-l mRNA in HUVECS. Confluent human umbilical vein endothelial cells were incubated in nutridoma and infected with 100 ul of HHV-6 (1900 CFU/ml). (Top) After 2, 9 and 17 hours, the monolayers were washed with PBS, and RNA was extracted from the cells. RNA was denatured and subjected to agarose gel electrophoresis in the presence of formaldehyde. After blotting to nylon membrane, the mRNA was hybridized to 32P labeled PAI-l cDNA as described under “Materials and Methods”. (Bottom) The corresponding 28S and 18S rRNA ofthe Northern blot. Figures are negative images. 128 Baas o->:_._ .52 t veep—5&0: Soc .5 cause 0 $1152. a 69035-8: 59.. m Baas m->II 59. N 66892.75: Soc N cubued Fumu edmm nmmw nceof VAwu SS and g5. 129 liverstandard Figure 23. Effect of HHV-6 on t-PA mRNA in HUVECS. Confluent human umbilical vein endothelial cells were incubated in nutridoma and infected with 100 ul of HHV-6 (1900 CFU/ml). (Table) After 2, 9 and 17 hours, the monolayers were washed with PBS, and RNA was extracted from the cells. RNA was denatured and subjected to agarose gel electrophoresis in the presence of formaldehyde. After blotting to nylon membrane, the mRNA was hybridized to 32P labeled t-PA cDNA as described under “Materials and Methods” and the volume was measured using a photoimmager. (Bottom) The corresponding 283 and 188 rRNA ofthe Northern blot. 130 2 hr control 2 hr HHV-6 9 hr control 9 hr HHV-6 17 hr control 17 hr HHV-6 infected infected infected t-PA mRNA 29369 27535 26546 26245 22375 24075 liver standard ‘ Gui... 131 Effect of HHV-6 on the Procoagulant Activity of HUVECS To test the effect of HHV-6 concentration on the procoagulant activity of endothelial cells, different concentrations ofthe virus were used to infect HUVECS. After 17 hours ofincubation, fresh frozen plasma was added and a clotting time was performed as described in “Materials and Methods”. Figure 24 shows that the clotting time correlated well with the CFU of HHV-6 added to HUVECS. An increase in HHV-6 concentration resulted in a decrease in clotting time. The clotting time of non-infected control human endothelial cells was 213 + 16 seconds (mean + SEM). There was no significant change in clotting time after adding 20 or 40 ul ofthe virus. The clotting time decreased to 132 :1: 34 seconds after infecting HUVECS with 60 ul of HHV-6. The maximal effect on clotting time was attained at a concentration of 80 ul or higher ofthe virus. The objective ofthe experiment represented by figure 25 was to determine ifthe effect of HHV-6 on the procoagulant activity of human umbilical vein endothelial cells is immediate or delayed and the duration ofthe effect. Clotting time was determined at 15 minutes, 1, 2, 6, 9 and 17 hour post-infection with HHV—6. The effect of HHV-6 on Human umbilical vein endothelial cells was detectable by measuring the 132 Figure 24. Effect of HHV-6 concentration on procoagulant activity. Human umbilical vein endothelial cells were infected with HHV-6 at a range of 20 ul to 140 ul (1900 CFU/ul). HUVECS were incubated for 17 hours, and the clotting time was performed as described in “Materials and Methods”. Results represent mean + SEM ofthree experiments with three replicates per experiment. A significant decrease in clotting time was seen at a HHV-6 concentration of 60 ul or higher (P<0.05). 133 own NSC. 02:1..040 80 100 120 140 60 HHV-6 CON ul( 1900 ciulul) 20 conrrol 134 Figure 25. Kinetics of HHV-6 on procoagulant activity. Human umbilical vein endothelial cell monolayer were infected with 100 III of HHV-6 (1900 CFU/ul). At 15 minutes, and at l, 2, 6, 9, and 17 hours post-infection with the virus, the clotting time was performed on non-infected control and HHV-6 infected HUVECS. Results represent mean + SEM ofthree experiments with three replicates per experiment. The difference in clotting times between the non-infected and HHV- 6 infected HUVECS at each incubation time is statistically significant (P<0.05). 135 Clotting Time in Seoo nds 400 800 280 200 150 100 b 50. - Non-infected 1:1 HHV-i Infected 15 min 1111 2111 HI 9hr lncubatbn Tine Post Infection 136 17 III clotting time at all post-infection times examined. Figure 25 demonstrates that the overall cellular procoagulant activity increased rapidly as indicated by shortening of the clotting time. It was noticed in Figure 25 that incubation of HUVECS in culture medium or infection medium resulted in a slight but significant increase (P<0.05) in clotting time over the seventeen hour period. HHV-6 was heated at 60°C for different period oftime to kill the infectious virus. The procoagulant effects ofthe heated virus were compared to the non-heated HHV-6 and to the non-infected endothelial cells by performing clotting time. Heating HHV-6 before being used to infect HUVECS resulted in reduction in procoagulant activity that was mirrored by an increase in clotting time (Figure 26). However, this cellular procoagulant activity of HUVECs infected with heat- inactivated HHV-6 was still higher than the non-infected control endothelial cells. This was reflected by a decrease in clotting time. Even after heating the virus for 60 minutes, the clotting time was shorter than in the non-infected endothelial cells. The decrease in clotting time was smaller when a heat-inactivated virus preparation was used instead of the active virus. To investigate the procoagulant activity on the surface of HUVECS, endothelial cells infected with HHV-6 were compared to control cells by measuring the clotting time using normal and factor 137 deficient plasma. Figure 27 compares the effect of normal plasma and factor deficient plasma on clotting time on the surface of human umbilical vein endothelial cells. In normal plasma and factor VII, X, or V deficient plasma, the clotting time was shorter in the HHV-6 infected HUVECS than in the corresponding non-infected control. Whereas there was no significant difference in clotting time using factor II deficient plasma on non-infected and HHV-6 infected endothelial cells. The clotting time ofX or V deficient plasma on the non-infected control HUVECS was longer than the clotting time of the normal or factor VII deficient plasma on non-infected control cells. Similar results were obtained with factor X or V deficient plasma on HHV-6 infected HUVECS compared with normal or factor VII deficient plasma on HHV-6 infected cells. Finally, no significant changes in procoagulant activity were seen in factor VII deficient plasma on the surface HHV-6 infected endothelial cells compared with the control. 138 Figure 26. Effect of heating HHV-6 on HUVECS procoagulant activity. HHV-6 was heated at 60 °C for different period oftime before being used to infect HUVECS prior to measuring the clotting time. Results represent mean + SEM ofthree experiments with three replicates per experiment. The difference in clotting times between non-heated HHV-6 infected HUVECs and 10 minutes or longer heated HHV-6 infected HUVECS is statistically significant (P<0.05). 139 Clotting Time in seconds 150 100 80 - II IIV-8 Matt! [:I l-iHV-o notiisatad [:I naiinisctsd ’ .1 ~ 1 . 5 10 ‘ ' ' ' o 5““an “ea‘ed mm 30 mm 80 mm 120 mm “0‘ Heating time 140 Figure 27. Procoagulant activity of HUVECs. Clotting time was measured in normal plasma, factor 11, VII, X, and V deficient plasma comparing the procoagulant activity in non-infected control and HHV-6 infected HUVECS according to “Materials and Methods”. Results represent mean + SEM oftwo experiments with three replicates per experiment. 141 CLOTTING TIME (SEC) 400 _ 350 300 100 i 250 i 200 i 150 : 50: - IO'I INFECTED [:3 INFECTED IoIaai Piasaa 11 Del Plasaa VI 001 Plasma dei plasna 142 v dot plasma DISCUSSION Human Herpesvirus-6 Effect on Endothelial Cells HHV-6 is acquired in early infancy and is widespread in populations of apparently healthy adults (27). HHV-6 has been shown to have the ultrastructural and morphogenetic properties characteristic of herpesvirus. It is distinct from the other herpesviruses by its antigenic properties and by failure to Show homologous hybridization with nucleic acid sequences from other herpesviruses (114). HHV-6 closest phylogenetic relative is CMV (66% DNA sequence homology) (137). Although the pathogenesis of HHV-6 is not yet fully understood, this virus has been recently recognized as an opportunistic pathogen in transplant recipients (192, 187), and like Other herpesviruses, HHV-6 is maintained in a latent state after primary infection. Autopsy reports of AIDs patients revealed widespread systemic dissemination of HHV-6 (126, 203). HHV-6 is a lymphotropic virus. Therefore, lymphocytes circulating in the blood serve as a reservoir for the virus. In our laboratory , HHV-6 was cultivated in cord blood lymphocytes using the method of Ablashi et al. (1). The infected lymphocytes had distinctive CPEs characterized by pleomorphic, balloon-like large lymphocytes. This agrees with what Suga et al. (255) reported. 143 In this study we report that HHV-6 can infect endothelial cells. Using a monoclonal antibody labeled with FITC against the early viral antigen, the virus was detected in the endothelial cells 17 hours post infection. In May 1998, Wu et al. (284) published an article studying the chronic infection of HUVEC by HHV-6. Their study demonstrated that HHV-6 could be transmitted from T-lymphocytes to the endothelium by cell to cell contact in vitro. Their results demonstrated that the endothelial cells are susceptible to HHV-6 infection. They detected both HHV-6 early and late viral antigens in HUVEC cultures 24 hour post infection or later (284). Our data is in agreement with Wu et al. In our experiments, HUVECS were directly infected with cell-free HHV-6. After infection, the HHV-6 early antigens were detected in HUVECS. To determine if the virus was still infectious after being in HUVECs for one, two or five days, HHV-6 was recovered from endothelial cells by freezing/thawing two times. Using this recovered virus, we were able to infect lymphocytes as indicated by the formation of balloon lymphocytes. Our finding suggests that endothelial cells may serve as a reservoir for harboring HHV-6. This is also in agreement with Wu et al. who were able to recover HHV-6 from HUVEC that contained fewer than 1% of antigen positive endothelial cells (284). Such an interactive exchange between circulating blood 144 components and the endothelium has also been seen with CMV (279). Wu et al. also reported that they were able to maintain the HHV-6 in continually passaged HUVEC for almost a month. The virus could be recovered from the endothelial monolayers by co-cultivation with lymphocytes (284). Platelet Adherence and Aggregation to HUVECs Adhesion of platelets to the vascular wall is an initial event in the development ofintravascular thrombosis and may be important in the development of other vascular pathologies. In the absence of thrombin we observed little adherence of platelets to monolayers of HUVECs. This observation is in agreement with Booyse et al. (25) who reported no interaction between undamaged bovine aortic endothelial cells and bovine platelets (25). Our study is also in agreement with Czervionke et al. (47) and Visser et al. (274) who reported that there is no significant interaction between platelets and undamaged human umbilical vein endothelial cells (47, 274). Endothelial damage leads to expression of subendothelial components that increased platelet binding (25). Others have demonstrated that virally transformed endothelial cells expressed an adherent surface for platelets (95). Hoak et al. (106) and Fry et al. (76) demonstrated that thrombin-aggregated platelets added to the vascular 145 endothelium resulted in an increase in platelet binding to HUVECS. The same effect was seen when thrombin, platelets and endothelial cells were incubated together( 76, 106). Our data confirms Hoak’s and Fry’s reports. We saw an increase in platelet binding to HUVEC when thrombin, platelets and endothelial cells were incubated together or when thrombin aggregated platelets were added to the endothelial cells. Our work also shows that pretreatment of endothelial cells with thrombin for five minutes in the absence of platelets results in an increased binding of platelets to HUVECs even after washing and replacing the fluid-phase that contains thrombin with thrombin free media (P<0.05). It has been shown that the effect of a-thrombin on endothelial cells can be suppressed or eliminated by the addition of DIP-thrombin (inactive thrombin) or by the addition of a-thrombin inhibitors. These observations suggest that platelet adhesion is mediated by active or- thrombin present on the endothelial cell’s binding sites (118). Specific binding sites on endothelial cells for III-thrombin have been documented (107,145). Using SDS-polyacrylamide gel electrophoresis, Isaacs et al. (107) demonstrated that ”5 1- labeled thrombin binds to a specific site on the endothelial cells with formation of a 77,000 dalton complex. Binding of the ”5 I-labeled thrombin was blocked by 100-fold excess of unlabeled thrombin and by the thrombin inhibitor hirudin. Formation of 146 the complex could be detected as early as 15 seconds after stimulation. The complex formation increased rapidly over the next 20-30 minutes then continued at a slower rate for the next 2.5 hours. This thrombin endothelial cell complex is as distinct from the thrombin— antithrombin III complex (107). It has been reported that the thrombin that binds to the endothelium gets inactivated via one or more of several mechanisms. These include internalization and degradation (145, 231), binding to thrombomodulin (63, 195) and heparin proteoglycan (99, 239), as well as interaction with antithrombin III (99). Esmon et al. (63) demonstrated that when thrombin binds to thrombomodulin, it loses its procoagulant activities in that it no longer clots fibrinogen, activates factor V, nor induces platelets to aggregate. They also showed that when thrombomodulin is added after thrombin had bound to the platelets, the thrombin rapidly redistributes onto the thrombomodulin. These data suggest that thrombomodulin may inhibit or reverse platelet activation by thrombin (63). Kaplan et al. (118) demonstrated that a-thrombin binds bovine pulmonary artery endothelial cells with high affinity. A portion ofthe bound thrombin remains active with respect to platelet activation for some time. They stimulated the endothelial cells for 10 and 60 minutes. The increased platelet adherence was inhibited by treating with D-phenylalanyl-L- prolyl-L-arginine chloromethyl ketone (PPACK), a thrombin specific 147 inhibitor for five minutes before adding platelets (118). In our ‘ laboratory we found no significant difference in platelet binding when the platelets were added five or sixty minutes post stimulation with thrombin. Platelet adherence to endothelial cells has been reported under static and rocking conditions (76, 106). In our study we found no significant difference in platelet binding between static and rocking conditions after platelet incubation for 30 minutes. Stimulating the endothelial cells with thrombin increased the platelet adherence to the same extent with rocking and under static conditions. Hoak et al. (106) reported that thrombin induced prostacyclin release from human endothelial cells antagonizes thrombin ~induced platelet adherence and aggregation. They also demonstrated that aspirin pretreatment of endothelial cells inhibited prostacyclin production that resulted in increase in platelet binding (106). In contrast, Johnson et al. (112) reported that inhibition of bovine endothelial cells prostacyclin production using aspirin had no effect on platelet adherence to endothelial cells (112). Our results support the work of Hoaks et al. (106) and Czervionke et al. (47), and contradict the reports of Johnson and coworkers (112). Thrombin induces platelet aggregation and adhesion to endothelial cells as well as stimulating endothelial cell prostacyclin production. Prostacyclin from endothelial 148 cells is known to be an inhibitor of platelet adhesion and aggregation. Thrombin therefore appears to stimulate two mechanisms that work in opposition to each other. Thus, the degree ofthrombin-induced platelet binding to HUVECs reflects a balance between these opposing effects. The inhibition of prostacyclin by adding indomethacin resulted in an increase in platelet adhesion and aggregation. We suggest that the influence of prostacyclin synthesized in endothelial cells may be dependent on the tissue source ofthe endothelial cells and may explain the contradictory finding between our study and that of Johnson et al (112). They used bovine endothelial cells whereas in our studies we used HUVEC. This increase in adhesion and aggregation as a result of thrombin’s effect on HUVECs may be important in the development of thrombosis when pro-coagulation factors are activated on the surface ofthe endothelium. In our study the percent of platelet adherence was dependent on thrombin concentration . An increase in platelet binding was seen at thrombin concentration as low as 0.001U/ml. The percent of platelet adherence increased gradually with an increase in thrombin concentration. The percent platelet binding was maximum at 0.2U/ml of thrombin or higher. Our experiments were conducted using 0.2U/ml of thrombin to study platelet adherence. In the literature, cOncentrations ofthrombin ranging from 0.2U/ml to 2U/ml have been used to stimulate 149 endothelial cells. Thrombin is a very potent serine protease. We suggest that a concentration close to 0.2U/ml should be used to avoid damaging the cells. We compared the percent of platelet binding in HHV-6 infected cells with non-infected endothelial cells in buffer, in the presence of thrombin, and in the presence ofindomethacin and thrombin. In this study, HHV-6 infection ofHUVECs resulted in 11% increase in platelet adhesion and aggregation compared to the control noninfected cells (P < 0.05). The difference in platelet binding between HHV-6 infected and non-infected endothelial cells was greatest after stimulation with thrombin (P<0.05). Preincubation of uninfected endothelial cells with thrombin, followed by the addition of platelets, resulted in only a small increase in platelet binding to endothelial cells unless prostaglandin synthesis was inhibited. It was demonstrated by Alheid et al. that the endothelium can inhibit platelet aggregation by two completely separate mechanisms, one mediated by prostaglandin and cAMP, and the other by EDRF and GMP (6). In our study, the increase in percent ofthrombin-induced platelet binding to HHV-6 infected HUVECS is partially due to inhibition of prostacyclin synthesis. The evaluation of percent binding in HHV-6 infected endothelial cells can not be limited to only abolishing the effect “Prostacyclin because even after inhibiting prostacyclin, using 150 indomethacin, the percent of platelet binding was higher in HHV-6 infected HUVECS than in the non-infected control endothelial cells. Thrombin induced marked platelet binding to HHV-6 infected endothelial cells even in the absence ofindomethacin. When indomethacin was added, the increase in platelet adherence was greater than when thrombin was added alone. From our experiments, we could not prove or disprove the involvement of EDRF/GMP on platelet binding. We wanted to check ifthe increase in platelet binding was the result ofthrombin’s property as an aggregator or its effect on HUVECs. To determine the probable cause ofincreased platelet binding, ADP, as an activator of platelet aggregation, was used to stimulate endothelial cells for five minutes followed by the addition of platelets or by adding ADP and platelets to the endothelial cells. Our data showed that ADP did not increase platelet binding to HUVECS. This indicates that the increase in platelet binding to endothelial cells was not due to the effect ofthrombin as an aggregator but rather on its direct effect on the endothelial cells. The morphological data using electron microscopy or Wright’s stain and light microscopy suggest that HHV-6 infected endothelial cells retracted and that platelet binding occurred toward the endothelial cell edges. Platelet binding was also seen on endothelial cells and on 151 the exposed subendothelium. Our results are consistent with some reports and contradict others depending on the endothelial cells source. Kaplan et al. and Johnson et al. reported the lack of any visible gaps between bovine endothelial cells and reported that some, but not all of the adherence of platelets occurs toward the endothelial cell edges (112, 118). Garcia et al. reported that III-thrombin induced gaps in bovine pulmonary artery endothelial cells using approximately 200 times the concentration of a-thrombin utilized in our study (81). Chen et al. demonstrated platelet adherence to human umbilical veins pro- perfused with thrombin. Their morphological studies demonstrated retraction of endothelial cells and adherence of platelets to the underlying subendothelium (34). We conclude that HHV-6 infection shifts the properties of endothelial cells from anticoagulant to procoagulant by increasing platelet adherence to endothelial cells. HHV-6 infection of HUVEC induces changes in the endothelial membrane. The infection also induces retraction ofthe endothelial cells and probably decreases prostacyclin secretion and are potentially involved in atherosclerosis and tissue necrosis. 152 Effect of HHV-6 on Tissue- Plasminogen Activator and Plasminogen Activator Inhibitor-l The vascular endothelium plays an important role in the initiation and control of fibrinolysis by synthesizing and secreting t-PA and PAI- 1. Under normal conditions, a balance exists between the enzyme t-PA and its inhibitor PAI-l. Chemicals, microorganisms, and other agents, or conditions which alter this fibrinolytic balance, do so by changing the activity and/or concentration ofthe t-PA:PAI-1 ratio (62, 143, 253) Bok et a1. (24) studied the effect ofHSV on HUVECS and the production oft-PA and PAI-l. They reported that HSV infection of HUVECS causes a decrease in PAI-l activity and antigen levels from the extracellular matrix. T-PA synthesis is also decreased in HSV-infected endothelium (24). Villeda et al. (273) studied the role of fibrinolysis in the pathogenesis of the hemorrhagic syndrome produced by virulent isolates ofthe African swine fever virus. Plasma levels oft-PA activity in pigs infected with the highly virulent Malawi 83 virus were found to be significantly increased whereas PAI-l activity was markedly decreased. In complete contrast to the previous results, pigs infected with the moderately virulent DR-78 virus developed high plasma levels of PAI-l activity with a decrease in t-PA activity to undetectable levels 153 nine day post-infection. A plausible explanation that they offered for the increase in t-PA in the Malawi 83 virus infected pigs was the increase in fibrin deposits in the microcirculation that stimulates the fibrinolytic system (273). Our data shows that the effect of HHV-6 on PAI-l activity level from HUVECs is dose dependent. Changes in PAI-l release from HHV- 6 infected endothelial cells were seen with 40 ul ofHHV-6 (1900 CFU/ul) and was maximal with a concentration of 100 ul ofthe virus. No significant changes in PAI-l activity were seen at virus concentrations greater than 100 ul. Changes in PAI-l activity between the non-infected control and the HHV-6 infected HUVECS were not statistically significant until four hours post-infection. The PAI-l activity increased rapidly to six hours post-infection. From 6 to 18 hour post-infection, there was a slow increase in PAI-l activity. This increase was still apparent at 42 hour post-infection. Bok et al. (24) measured PAI-l activity by measuring urokinase inhibition whereas in our study we measured the inhibition oft-PA. They reported a decrease in PAI-l activity in the extracellular matrix of HSV infected endothelial cells. This decrease in PAI-l activity was observed at eight hours post infection and the lowest PAI-l activity was seen at 24 hour after infection (24). Our data differ from that of Van Dam-Mieras et al. (270) who reported that CMV infection of HUVECs had no significant 154 effect on PAI-l activity (270). HHV-6 infection of HUVECs lead to a time dependent increase in PAI-l antigen. We measured the concentration of PAI-l antigen at different times in non-infected and HHV-6 infected endothelial cells. In both the control uninfected and the virus infected endothelial cells, the slow and steady increase reached a maximum at 18 hour post infection. The amount of PAI-l antigen in the culture media after 24 and 48 hours ofincubation was not different from the level measured at 18 hour post- infection. The lack of change of PAH in the culture media from 18 to 48 hours could indicate that PAI-l was not being added to or removed from the culture media. This lack of change could also reflect a balance between secretion and removal. This finding is also in contrast to what Bok et al. demonstrated in HSV-infected HUVECs. They showed that the PAI-l antigen in the extracellular matrix and the supernatant decreased after infection of HUVECs with HSV. This decrease was maximal after 24 hours ofinfection (24). Our data also differ from that of Van Dam-mieras et al. Who reported that CMV infection of endothelial cells does not alter PAI-l antigen level (270). This difference in finding between our data and that reported by others might be explained by the use of different viruses to infect the endothelial cells. In our study we used HHV-6 whereas Bok et al. used HSV and Van Dam-mieras used CMV. 155 PAI-l activity and antigen levels in heat-inactivated HHV-6 infection of endothelial cells were not significantly different from control endothelial cells. This implies that an infectious HHV-6 is required to effect the endothelial synthesis/release of PAH. Our northern blot experiments showed that human umbilical vein endothelial cells produce two distinct forms of PAH mRNA (3.2 and 2.3kb). This is consistent with other reported studies by Ginsburg et al. and Loskutoff et al.(88, 149). These two distinct transcripts are colinear from the 5' end and differ only in the length oftheir 3' untranslated regions and appear to be the result of alternative polyadenylation (149). Our northern blots indicate that infection of HUVECs with HHV- 6 increases the level of PAH mRNA. Quantitation ofthe relative changes in the two PAI-l mRNA forms induced by HHV-6 infection revealed a greater increase in the 3 .2kb form. This preferential increase could reflect changes at the level oftranscription, and increased polyadenylation at the downstream region to generate the larger transcript, or it could be due to preferential stabilization ofthe larger form. The time lag required for the effect of HHV-6 on PAI-l suggests that there is de novo synthesis of the inhibitor. This is further supported by adding cycloheximide three hours after infection with HHV-6 which 156 inhibited protein synthesis reflected by a significant decrease in PAI-l concentration. Our data show that HHV-6 infection ofHUVECs did not lead to any significant change in t-PA antigen release or synthesis between non- infected and infected endothelial cells. Bok et al. demonstrated a time dependent decrease in t-PA in endothelial cells infected with HSV (24). CMV infection of endothelial cells does not alter t-PA antigen level (270). Fibrinolysis is the result ofthe action of the serine protease plasmin on fibrinogen. Plasmin circulates in blood as the proenzyme plasminogen and is converted to the active enzyme by limited proteolysis. The major activator of plasminogen in blood is t-PA which is controlled by its major inhibitor PAI-l. Precise regulation ofthese fibrinolytic proteins is critical for effective hemostasis. In our study, we demonstrated an increase in PAI-l synthesis/ release with no significant change in t-PA antigen in HUVEC infected with HHV-6. This variation in the t-PAzPAI-l ratio shifts the fibrinolytic balance toward hypofibrinolysis which could increase the potential for thrombosis and disseminated intravascular coagulation. 157 Effect of HHV-6 on the Procoagulant Activity of HUVECs Endothelial cells play an active role in the control ofthe procoagulant-anticoagulant balance in blood (79). Any shift from this equilibrium can lead to thrombotic or hemorrhagic complications. When this nonequilibrium state is shifted toward protrombotic over a long period oftime it may lead to atherogenesis (95). The role of the endothelium in atherogenesis has received a great deal of attention in the literature. It is tempting to speculate that the formation of fibrin deposits on the surface ofthe subendothelium or the endothelium as a consequence ofinfection by a virus, could favor atherogenesis. The fibrin mesh would constitute an area where different cells or chemicals, believed to be involved in the formation of an atherosclerotic lesion, could interact, and would be consistent with Duguid’s proposal that fibrin formation is a factor in atherogenesis (58). Our results demonstrate that HHV-6 infection of HUVECs promotes an increase in thrombin generation and fibrin formation on the endothelial surface. This is reflected by a decrease in clotting time. This increase in procoagulant activity would seem to reflect a facilitated assembly ofthe prothrombinase complex (Xa, Va, thrombin) on the surface of HHV-6 infected endothelial monolayer by an unknown mechanism. We prOpose that infection of the endothelium by HHV-6 alters the endothelial surface. In normal endothelial cells, 158 Phospholipids, in particular phosphatidyl serine, are not significantly present on the external bilayer of cells (290). When infected with a virus, these negatively charged phospholipids are increased on the cells surface. The increased exposure of these phospholipids could lead to accelerated binding of factor Va and the potential generation of thrombin and the formation of fibrin (264). It is important to realize that HHV-6 belongs to the herpesvirus family and it is an envelope virus. One can speculate that upon infection of the endothelial cells with HHV-6, the interaction of HHV-6 envelope with the endothelial membrane could produce conformational changes or membrane perturbations that facilitate the interaction of procoagulant factors with the endothelial membrane. In our study, the increase in procoagulant activity is reflected by a decrease in clotting time on the surface of endothelial cells. This decrease in clotting time was dose-dependent on the concentration of HHV-6. No significant decrease in clotting was seen among noninfected control, 20 and 40ul of HHV-6 (1900 cfu/ul). The maximal effect ofthe virus was attained at a concentration of 80u1 or higher of the virus. The clotting time induction correlated with the concentration ofthe virus. Our report is in agreement with van Dam-Mieras et al. who reported that CMV infection of HUVEC resulted in an increase in 159 procoagulant activity (270). Our report is also in agreement with Visser et al. who demonstrated that HUVEC infection with HSV induced an increase in thrombin generation (274). Our data contradict the reports of Mazure et al. who reported that CMV infection of HUVEC did not induce procoagulant activity at any time interval measured(162) We speculate that the contradiction in the procoagulant activity ofthe CMV infected HUVECs between van Dam- Mieras et al. and Mazure et al. is because different strains of CMV were used. It has been demonstrated that the expression oftissue factor on the endothelial surface as a consequence ofinflammatory mediators can be detected after two to four hours and becomes maximal after four to six hours (45,202). Our results demonstrate that the decrease in clotting time on HHV-6 infected HUVECs was seen almost immediately after infection ofthe cells. We propose that the procoagulant activity of HHV-6 infected HUVEC seems to reflect the formation ofthe prothrombinase complex on the perturbed membrane ofinfected cells. To characterize the HUVEC procoagulant response, clotting times using fresh frozen plasma (FFP) and factor deficient plasma were determined. The clotting times of FFP and factors VII, X and V deficient plasma after exposure to HHV-6 infected HUVECs was Shorter than when the plasma was added to non-infected cells. 160 Comparing the procoagulant activity of FFP and factor VII deficient plasma on non-infected HUVECs, the clotting times were not statistically different. The clotting times between FFP and factor VII deficient plasma on HHV-6 infected HUVECs were also not statistically significant. The clotting times using factor X and factor V deficient plasma on the non-infected and HHV-6 infected HUVEC were higher than the corresponding ones using FFP. These results, along with the time course of procoagulant activity, suggest that tissue factor expression cannot completely explain the procoagulant response. Our results suggest the facilitated formation ofthe prothrombinase complex on HUVECS. Our data are in agreement with Visser et al. who reported that HSV infection ofthe endothelium promotes the prothrombinase complex formation. Other viruses have been shown to induce the formation oftissue factor on HUVECS (162). Infection of the endothelial cells with heat inactivated HHV-6 resulted in a procoagulant response reflected by a decrease in clotting time. But, this decrease in clotting time is less pronounced than post infection with an active virus. The induced procoagulant activity by the heat inactivated HHV-6 cannot be explained by an incomplete inactivation during heating because no residual infectivity of HHV-6 was detected in lymphocytes. One can question ifthe extra thrombin generated on the surface of 161 HUVECs as shown in our experiments, will be neutralized in vivo by the natural thrombin inhibitors present in the plasma. Neutralization by antithrombin III seems to be unlikely because heparin is required as a cofactor for antithrombin III and it has been demonstrated that in herpes infected endothelium heparan proteoglycan synthesis is decreased (117). Protein C is also not available because it has been reported by Visser et al. (274) that HSV-infected endothelium is less able to catalyze protein C activation than uninfected endothelium. They reported that protein C activation is 30 to 40 % diminished by the presence of various concentrations ofthrombin (274). Our data suggest that HHV-6 infection of HUVECs shifts the dynamic balance of endothelial cells from anti-thrombotic to prothrombotic. This finding is consistent with the theory that viral infection of the endothelial cells contributes to the increased risk of thrombosis and atherosclerosis (95). 162 SUMMARY & CONCLUSION The vascular endothelium is strategically located at the interface between blood and tissue. One ofits functions is to protect against vascular injury and maintain the blood fluidity. Under normal conditions, the endothelium is antithrombotic. Injury to the blood vessels is accompanied by loss ofthe antithrombotic properties and by expression ofthe prothrombotic ones leading to the development of DIC, thrombosis and atherosclerosis. Viruses have been implicated in the pathogenesis of atherosclerosis and thrombosis based on one or more of the following: the epidemiological relationship between herpesvirus infection and accelerated atherosclerosis in heart transplant patients and in restenosis after angioplasty; the widespread nature ofthe herpesviruses in humans and the presence ofa latent form; the presence ofherpesviral antigen and genomic materials in atherosclerotic lesions. The main purpose ofthis work was to determine if HHV-6, a member ofthe herpesvirus family, can infect HUVEC and to determine the effect of HHV-6 on the hemostatic and fibrinolytic components of endothelial cells. 163 The summary of our results is as follows: 1. HHV-6 can infect human umbilical vein endothelial cells. This finding is based on the detection of the HHV-6 early antigens in endothelial cells and the ability of the recovered infectious virus from HUVECs to infect cord blood lymphocytes. Thrombin increases the adherence and aggregation of platelets on the surface of HUVECs (P<0.05) even after washing and replacing the fluid phase containing thrombin with thrombin-free media. This increase in platelet binding to endothelial cells is dose dependent. The difference in binding between static and rocking conditions is not statistically significant (P>0.05). The effect ofthrombin on the percent of platelet binding to HUVECS is due to the direct effect of thrombin on endothelial cells and not on the property ofthrombin as an aggregator. There is no significant difference in platelet binding if: 1) thrombin and platelets are added together to the endothelial cells. 2) platelets are aggregated by thrombin, then, added to endothelial cells. 3) HUVECs are stimulated by thrombin, then, platelets are added even after replacing the fluid phase that contains thrombin. HHV-6 infection of HUVECs results in an increase in percent of platelet aggregation and adhesion to endothelial cells (P<0.05). This significant increase in platelet binding is greatest after 164 stimulation of HUVECs with thrombin. The inhibition of prostacyclin with indomethacin leads to a significant increase in platelet binding in non-infected HUVECS (P<0.05). A further increase in percent binding is seen in HHV-6 infected endothelial cells. The difference in binding between non- infected and HHV-6 infected HUVECs is probably due to membrane perturbation ofthe endothelial cells. We cannot rule out the involvement of NO/GMP pathway from our study. HHV-6 infection of HUVECS leads to a time and dose dependent increase in PAI-l activity. The t-PA activity is non-detectable because of high PAI-l in the culture media that totally neutralizes all the t-PA in the culture media. HHV-6 infection of HUVECs leads to a slow and steady increase in the level of PAI-l antigen in the culture media. Northern blots indicate that infection ofthe endothelial cells with HHV-6 increases the level of mRNA with preferential increase in the 3.2kb form. HHV-6 infection of HUVECs does not have any significant effect on t-PA antigen concentration or mRNA. HHV-6 infection of HUVECs leads to a dose dependent increase in the procoagulant activity reflected by the decrease in clotting time. This increase in procoagulant activity reaches a maximum 165 immediately after infection of HUVECS. 10. The increase in procoagulant activity in HHV-6 infected HUVECS reflects an enhanced assembly ofthe prothrombinase complex and cannot be explained by the activation oftissue factor. The significance ofthis investigation is that we have demonstrated that HHV-6 infects HUVECs and changes the property of the endothelial cells from antithrombotic to prothrombotic. This shift towards a hypercoagulable condition is reflected by an increase in platelet adhesion and aggregation, an increase in PAI-l activity and antigen concentration without any change in t-PA antigen or activity level, and by an increase in thrombin generation and fibrin formation. The thrombogenic theory of atherosclerosis proposes that fibrin is continuously being deposited on and removed from the blood vessel walls. This continuous depositing and removal of fibrin depend on a dynamic balance between the coagulation and fibrinolytic systems. An imbalance towards increased coagulation or decreased fibrinolysis can lead to excess fibrin deposition. A layer of fibrin impairs oxygen diffusion into the vessel wall creating a hypoxic condition that can lead to cell death. While the thrombotic theory of atherosclerosis continues to be controversial, Roberts (217a) in an editorial stated that "Most serious students ofthe morphology ofthe arterial plaque presently 166 support in whole or part, the thrombogenic origin of atherosclerosis." A decreased fibrinolytic potential in atherosclerosis has been reported by many investigators (43a, 178a, 199a). Viral infections can cause endothelial injury and induce procoagulant changes (183a). Our findings are in agreement with the association of procoagulant changes in virally infected endothelial cells. The findings of an increased synthesis and release of PAH from the HHV-6 infected HUVECS without any change in tPA synthesis or release is consistent with a decreased fibrinolytic potential. In our study we have demonstrated an increase in the procoagulant properties ofthe HUVECS, increased platelet binding, and a decrease in the fibrinolytic potential following infection with the HHV-6 virus. Our findings are supportive of the proposed role of viral infection in atherogenesis and the thrombogenic theory of atherosclerosis (183a). 167 10. LIST OF REFERENCES Ablashi, D.V., S. Josephs, A. Buchbinder, K. Hellman, S. Nakamura, et al. Human B-lymphotropic virus (human herpesvirus- 6). J. Virol. Methods. 21: 29-48, 1988. Ablashi, D.V., P. Lusso, C.L. Hung, S.Z. Salahuddin, S.F. Josephs, et al. Utilization of human hematopoietic cell lines for the propagation and characterization of HBLV (human herpesvirus-6). Int. J. Cancer. 42: 787-791, 1988. Adam, E., J.L.Melnick, and ME. DeBakey. Cytomegalovirus infection and atherosclerosis. Cent. Eur.J. Public Health 5: 99-106, 1997'. Agut, H., D. Kerouedan, P.M. Pele, H. Collandre, C. Samba- Lefevre, et al. Co- cultivation of HHV-6 and HIV-1 from blood lymphocytes. Res. Virol. 140: 23-26, 1989. Agut, H. Pazzles concerning the pathogenicity of human herpesvirus 6 (Editorial). N. Engl. J. Med. 329: 203-204, 1993. Alheid, U., I. Reichwehr, and U. Forstermann. Human endothelial cells inhibit platelet aggregation by separately stimulating platelet cAMP and cGMP. Eur. J. Pharm. 164: 103-110, 1989. Andreasen, P.A., B. Georg, L.R. Lund, A. Roccio, and SN. Stacey. Plasminogen activator inhibitors: hormonally regulated serpins. Mol. Cell Endocrin. 68: 1-19, 1990. Aoki, N. Preparation of plasminogen activator from vascular trees ofhuman cadavers. J. Biochem. 75: 731-741, 1974. Asano, Y., S. Suga, T. Yoshikawa, A. Urisu, and T. Yazaki. HHV-6 infection (exanthem subitum) without fever. J. Pediatr. 115: 264- 265, 1989. Ashford, T.P., and DC. Freiman. Platelet aggregation at sites of minimal endothelial injury: An electron microscopic study. Am. J. Pathol. 53: 599-606, 1968. 168 ll. 12. l3. 14. 15. l6. l7. l8. 19. 20. 21. Bach, R., Y. Nemerson, and W. Konigsberg. Purification and characterization of bovine tissue factor. J. Biol. Chem. 256: 8324- 8331, 1981. Backmann, F., and E.K.O. Kruithof. Tissue plasminogen activation: chemical and physiological aspects. Semin. Thrombos. Haemost.10:6-17,1984. Balachandran, N., S. Tirawatnapong, B. Pfeiffer, D.V. Ablashi, and S.Z. Salahuddin. Electrophoretic Analysis of Human Herpesvirus 6 Polypeptides Immunoprecipitated from Infected Cells with Human Sera. J. Inf. Dis. 163: 29-34, 1991. Barnathan, E.S., A. Kuo, H. Van der Keyl, K.R. McCrae, G.R. Larsen, and DB. Cines. Tissue type plasminogen activator binding to human endothelial cells. Evidence for two distinct binding sites. J. Biol. Chem. 263: 7792-7799, 1988. Beebe, D.P., Binding of tissue plasminogen activator to human umbilical vein endothelial cells. Thromb. Res. 46: 241-254, 1987. Benditt, E.P., T. Barrett, and J.K. McDougall. Viruses in the etiology of atherosclerosis. Proc. Natl. Acad. Sci. USA 80: 6386- 6389, 1983. Berberg, W., 8. Wright, and CA. Janeway. Roseola infantum (exantum subitum). N. Engl. J. Med. 251: 253-254, 1949. Bevilacqua, M.P., J.S. Pober, G.R. Majeau, W. Fiers, R.S. Cotran, et al. Interleukin-l biosynthesis and cell surface expression of procoagulant activity in human vascular endothelial cells. J. Exp. Med. 60: 618-623, 1984. Bevilacqua, M.P., R.R. Schleef, M.A. Gimbrone, and DJ. Loskutoff. Regulation of the fibrinolytic system of cultured human vascular endothelium by IL-1. J. Clin. Invest. 78: 587-591, 1986. Bhagyalakshmi, A., and IA Frangos. Mechanism of shear-induced prostacyclin production in endothelial cells. Biochem.Biophys. Res. Commun. 158: 31-37, 1989. Biberfeld, P., B. Kramarsky, S.Z. Salahuddin, and RC. Gallo. Ultrastructural characterization ofa new human B-lymphotropic DNA virus (HBLV) isolated from patients with lymphoproliferative 169 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. disease. J. Natl. Cancer Inst. 79: 933-941,]987. Bick, R.L. Disseminated intravascular coagulation and related syndromes: a clinical review. Semi. Thromb. Hemost. 14: 299- 338, 1988. Blum, A., M. Giladi, M. Weinberg, G. Kaplan, H. Pasternack, et al. High anti-cytomegalovirus (CMV) IgG antibody titer is associated with coronary artery disease and may predict post-coronary balloon angioplasty restenosis.Am. J. Cardiol. 81: 866-868, 1998. Bok, R., H.S. Jacob, J. Balla, M. Juckett, T. Stella, et al. Herpes simplex virus decreases endothelial cell plasminogen activator inhibitor. Thromb. Haemost. 69: 253-258, 1993. Booyse, F.M., S. Bell, B. Sedlak, and M. E. Rafelson. Development of an in vitro vessel wall model for studying certain aspects of platelet-vessel (endothelial) interactions. Artery 1: 518-539, 1975. Bosma, P.J., E.A. van den Berg, T. Kooistra, D.R. Siemieniak, and J.L. Slighton. Human plasminogen activator inhibitor-1 gene. J. Biol. Chem. 263: 9129-9141, 1988. Briggs, M., J. Fox, and TS. Tedder. Age prevalence of antibody to human HHV-6. Lancet 1: 1058-1059, 1988. Brink, H.S., F.H. Derkx, F. Boomsa, and M.A. Schalekamp. 1- Desamino-8-D-arginine vasopressin (DDAVP) in patients with congenital nephrogenic diabetes insipidus. Neth. J. Med. 43: 5-12, 1993. Brommer, E.J.P. The level of extrinsic plasminogen activator (t- PA) during clotting as a determinant ofthe rate of fibrinolysis; inefficiency of activators added afterwards. Thromb. Res. 34: 109- 115, 1984. Busso, N., D. Belin, C. Failly-Crepin, and JD. Vassalli. Glucocorticoid modulation of plasminogen activators and one of their inhibitors in the human mammary carcinoma cell line MDA- MB-231. Cancer Res. 47: 364-370, 1987. Casonato, A., E. Pontara, A. Bertomaro, D. Damhauser, and M.T. Sartori. Demonstration that venous occlusion fails to release von Willebrand Multimers. Blood, Coagul. Fibrinolysis 6: 574-578, 170 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 1995. Challoner, P.B., D.T. Smith, J.D. Parker, D.L. Macleod, S.N. Coulter, et al. Plaque-associated expression of human herpesvirus 6 in multiple sclerosis. Proc. Natl. Acad. Sci. USA 92: 7440-7444, 1995. Charo, I.F. Platelet membrane glyc0proteins. In: Coleman, R.W., et al. Hemostasis and thrombosis: basic principles and clinical practice. 489-507, 1994. Chen,S., S. Barmatoski, and MI Barnhart. Effect of thrombin on platelet-vessel wasll interactions. Scanning Electron Microsc.3: 783-792, 1979. Chomczynski, P. and K. Mackey. Modification ofthe TRI reagent procedure for isolation of RNA from polysaccharide and proteoglycan-rich sources. Biotechniques 19: 942-945, 1995. Cines, D.B., A. Lyss, M. Bina, R. Corkey, N.A. Kefalides, and HM. Friedman. FC and C3 receptors induced by herpes simplex virus on cultured human endothelial cells. J. Clin. Invest. 69: 123- 128, 1982. Colburn, P., and V. Buonassisi. Anti-clotting activity of endothelial cell cultures and heparin sulfate proteoglycans. Biochem. Biophys. Res. Commun. 104: 220-227 1982. Coleman, R.W., S.J. Robboy, and JD. Minna. Disseminated intravascular coagulation (DIC): an approach. Am. J. Med. 52:679-684, 1972. Coleman, R.W. Mechanisms of platelet aggregation. In: Coleman, R.W. Hemostasis and thrombosis: basic principles and clinical practice. 508-523, 1994. Colucci, M., J.A. Paramo, and D. Collen. Inhibition of one-chain and two-chain forms of human tissue-type plasminogen activator by the fast-acting inhibitor of plasminogen activator in vitro and in vivo. J. Lab. Clin. Med. 108:53-59,1986. Colucci, M., J.A. Paramo, and D. Collen. Generation in plasma of a fast-acting inhibitor of plasminogen activator in response to endotoxin stimulation. J. Clin. Invest. 75: 818-824, 1985. 171 42. 43. 43a 44. 45. 46. 47. 48. 49. 50. Colucci, M., G. Balconi, R. Lorenzel, A. Pietra, D. Locati, et al. Cultured human endothelial cells generate tissue factor in response to endotoxins. J. Clin. Invest. 71: 1893-1896, 1983. Cone, R.W., M.L. Huang, R. Ashley, and L. Corey. Human herpesvirus 6 DNA in peripheral blood cells and saliva from immunocompetent individuals. J. Clin. Microbiol. 32: 2633-2638, 1994. Constantini, R., G. Hilbe, F. Spottl, and F. Holzknecht. The plasminogen activator content of the arterial wall in occlusive arterial diseases. Throm Diath Haemat0127z649-654, 1972 Conway, E.M., and RD. Rosenberg. TNF suppresses transcription ofthe thrombomodulin gene in endothelial cells. Mol. Cell Biol. 8: 5588-5592,1989. Cotran, R.S. New roles for the endothelium in inflammation and immunity. Am. J. Pathol. 129: 407-413, 1987. Crutchley, D.J., L.B. Conanan, and J.R. Maynard. Human fibroblasts produce inhibitor directed against plasminogen activator when treated with glucocorticoids. Ann. NY Acad. Sci. 370: 609-616, 1981. Czervionke, R.L., J.C. Hoak, and G.L. Fry. Effect of aspirin on thrombin-induced adherence of platelets to cultured cells from the blood vessel wall. J. Clin. Invest. 62: 847-856, 1978. Dahlback, B. Interaction between vitamin K-dependent protein S and complement protein, C4b-binding protein. Semin. Thromb. Haemost. 10: 139-145,1984. Davie, E.W.and O.D. Ratnoff. Waterfall sequence for intrinsic blood clotting. Science 145: 1310-1312, 1964. De Groot, P.G., J.H. Reinders, and J.J. Sixma. Perturbation of human endothelial cells by thrombin and PMA changes the reactivity of their ECM toward platelets. J. Cell. Biol. 194: 697- 704, 1987. 172 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. Declerck, P.J., M.C. Alessi, M. Verstreken, E.K. Kruithof, I. Juhan-Vague, and D. Collen. Measurement of plasminogen activator inhibitor 1 in biologic fluids with a murine monoclonal antibody-based enzyme-linked immunosorbent assay. Blood. 71: 220-225,1988. Declerck, P.J., M. De Mol, M.C.A1essi, S. Baudner, and E.P. Paques. Purification and characterization ofa plasminogen activator inhibitor 1 binding protein from human plasma. J. Biol. Chem. 263: 15454-15461, 1988. Degen, S.J.F., B. Rajput, and E. Reich. The human tissue plasminogen activator gene. J. Biol. Chem. 261: 6972-6985, 1986. Dewhurst, S., K. McIntyre, K. Schnabel, and CB. Hall. Human herpesvirus 6 (HHV-6) variant B accounts for the majority of symptomatic primary HHV-6 infections in a population of U. S. infants. J. Clin. Microbiol. 31: 416-418, 1993. DeWitt, D.L., and W.L. Smith. Purification of prostacyclin synthase from bovine aorta by immunoaffinity chromatography. J. Biol. Chem. 258: 3285-3293, 1983. Dicheck, D., and T. Quartermous. Thrombin regulation of mRNA levels oftissue plasminogen activator and PAI-l in cultured umbilical vein endothelial cells. Blood 56: 222-228,]989. Dieval, J., G. Nguyen, S. Gross, J. Delobel, and BK. Kruithof. A lifelong bleeding disorder associated with a deficiency of plasminogen activator inhibitor type 1. Blood 77: 528-532, 1991. Duguid J.B. Thrombosis as a factor in the pathogenesis of coronary atherosclerosis. J. Pathol. Bacteriol. 58: 207-215, 1946. el-Sayed, M.S. Effects of exercise on blood coagulation, fibrinolysis and platelet aggregation. Sports Med. 22: 282-298, 1996. Ellis, V., N. Behrendt, and K. Dano. Cellular receptor for urokinase-type plasminogen activator: function in cell-surface proteolysis. Methods Enzymol. 223: 223-233, 1993. 173 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. Emeis, J.J. Mechanisms involved in short-term changes in blood levels of t-PA. In: C. Kluft, ed. Tissue-type plasminogen activator (t-PA). Physiological and Clinical Aspects. 2: 21-35, 1988. Erickson, L.A., R.R.Schleef, T. Ny, and DJ. Loskutoff. The fibrinolytic system ofthe vascular wall. Clin. Haematol. 14: 513- 530, 1985. Esmon, N.L., R.C. Carroll, and CT. Esmon. Thrombomodulin blocks the ability of thrombin to activate platelets. J. Biol. Chem. 258:12238-12242,1983. Esmon, C.T. Molecular events that control the protein C anticoagulant pathway. Thromb. Haemost. 70: 29-35, 1993. Esmon, N.L., W. Owen, and C. Esmon. Isolation ofa membrane bound cofactor for thrombin-catalyzed activation of protein C. J. Biol. Chem. 257: 859-865, 1982. Esmon, CT. The regulation of natural anticoagulant pathways. Science 235: 1348-1352, 1987. Etingin, 0., R. Silverstein, and D. Hajjar. Von Willebrand factor mediates platelet adhesion to virally infected endothelial cells. Proc. Natl. Acad. Sci. USA. 90: 5153-5156, 1993. Fabricant, C.G., J. Fabricant, M.M. Litre’nta, and CR. Minick. Virus-induced atherosclerosis. J. Exp. Med. 148: 335-340, 1978. Fawcett, J., A.L. Harris, and R. Bicknell. Isolation and properties in culture of human adrenal capillary endothelial cells. Biochem. Biophys. Res. Commun. 174: 903-908, 1991. Fay, P.J., P.J. Haidaris, and TM. Smudzin. Human factor VIIIa subunit structure. Reconstitution of factor VIIIa from the isolated Al/A3-Cl-C2 dimer and A2 subunit. J. Biol. Chem. 266: 8957- 8962, 1991. Fitzgerald, G.A., A.K. Pedersen, and C. Patrono. Analysis of prostacyclin and thromboxane biosynthesis in cardiovascular disease. Circulation 67: 1174-1177, 1984. 174 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. Follo, M., and D. Ginsburg. Structure and expression ofthe human gene encoding plasminogen activator inhibitor, PAI-l. Gene 84: 447-453,1989. Forsberg, E.J., G. Feuerstein, E. Shohami, and HE. Pollard. Adenosine triphosphate stimulates inositol phospholipid metabolism and prostacyclin formation in adrenal medullary endothelial cells by means of Pz-purinergic receptors. Proc. Natl. Acad. Sci. USA. 84: 5630-5634, 1987. Fox, K.A.A., A.K. Robison, R.M. Knabb, T.L. Rosamond, B.E. Sobel, and SR. Bergmann. Prevention of coronary thrombosis with sub-thrombolytic doses of tissue-type plasminogen activator. Circulation 72: 1346-1354, 1985. French, J.E.. Atherosclerosis in relation to the structure and function of the arterial intima with reference to the endothelium. Int. Rev. exp. Pathol. 5:523-527, 1966. Fry, G.L., R.L. Czervionke, J.C. Hoak, J.B. Smith, and D.L. Haycraft. Platelet adherence to cultured vascular cells: influence ofprostacyclin. Blood 55: 271-275, 1980. Fukudome, K., and CT. Esmon. Molecular cloning and expression of murine and bovine endothelial cell protein Clactivated protein C receptor (EPCR). J. Biol. Chem. 270: 5571-5576, 1995. Furchgott, RP, and J.V. Zawadzki. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 288: 687-692, 1980. Furie, B., and BC. Furie. Molecular and cellular biology of blood coagulation. NEJM. 326: 800-806, 1992. Galdal, K.S., T. Lyberg, S.A. Evensen, E. Nilson, and H. Prydz. Thrombosis induces thromboplatin synthesis in cultured vascular endothelial cells. Thromb. Haemost. 54: 373- 376, 1985. Garcia, J.G.N., A. Siflinger-Birnboim, R. Bizios, P.J. DelVecchio, J.W. Fenton, and AB. Malik. Thrombin-induced increase in albumin permeability across the endothelium. J. Cell. Physiol. 128:96-104,1986. 175 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. Gelehrter, TD, and R. Senycer-Laszuk. Thrombin induction of plasminogen activator in cultured human endothelial cells. J. Clin. Invest. 77: 165-169, 1986. George, J.N., A.T. Nurden, and DR. Phillips. Molecular defects in interactions of platelets with the vessel wall. N. Engl. J. Med.311:1084-1098,1984. Gevers Leuven, J.A., C. Kluft, R.M. Bertina, and L.W. Hessel. Effects oftwo low-dose oral contraceptives on circulating components ofthe coagulation and fibrinolytic systems. J. Lab. Clin. Med. 109: 631-636, 1987. Gilbert, G.E., B.C. Furie, and B. Furie. Binding of human factor VIII to phospholipid vesicles. J. Biol. Chem. 265: 815-822, 1990. Gimbrone, M.A., R.S. Cotran, and J. Folkman. Human vascular endothelial cells: growth and DNA synthesis. J. Cell Biol. 60: 673- 684, 1974. Ginsberg, H.S.: Pathogenesis of viral infections, Microbiology. 3rd edition. Edited by BB. Davis et al. Philadelphia, Harper and Row. 1031-1046,1980. Ginsburg, D., R. Zeheb, A.Y. Yang, U.M. Rafferty, P.A. Andreasen, et al. cDNA cloning of human plasminogen activator- inhibitor from endothelial cells. J. Clin. Invest. 78: 1673-1680, 1986. Girard, T.J., and GI. Broze Jr. Tissue factor pathway inhibitor. Methods Enzymol. 222: 195-209, 1993. Granlick, J.R., and D. Ginsburg. von Willebrand disease. In: Beutler,E. et al. Williams hematology ,5th ed. 1458-1480, 1995. Grattan, M., and C. Moreno-Cabral. Cytomegalovirus infection is associated with cardiac allograft rejection and atherosclerosis. JAMA.261:3561-3566,1989. Guetta, E., V. Guetta, T. Shibutani, and SE. Epstein. Monocytes harboring cytomegalovirus: interactions with endothelial cells, smooth muscle cells, and oxidized low-density lipoprotein. Possible mechanisms for activating virus delivered by monocytes to sites ofvascular injury. Circ. Res. 81: 8-16, 1997. 176 93. 94. 95. 96. 97. 98. 99. 100. 101. Guha, A., R. Bach, W. Konigsberg, and Y. Nemerson. Affinity purification of human tissue factor: interaction of factor VII and tissue factor in detergent micelles. Proc. Natl. Acad. Sci. USA. 83: 299-302, 1986. Hajjar, D.P. C.G. Fabricant, C.R. Minick, and J. Fabricant. Virus- induced atherosclerosis. Herpesvirus infection alters aortic cholesterol metabolism and accumulation. AM. J. Pathol. 122: 62- 70, 1986. Hajjar, D.P., K.B. Pomerantz, D.J. Falcone, B.B. Weksler, and A. J. Grant. Herpes simplex virus infection in human arterial cells. Implications in atherosclerosis. J. Clin. Invest. 80: 1317-21, 1987. Hajjar, K.A., and NM. Hamel. Identification and characterization of human endothelial cell binding sites for tissue plasminogen activator and urokinase. J. Biol. Chem. 265:2908-2916, 1990. Hamsten, A., B. Wiman, U. De-faire, and M. Blomback. Increased plasma levels ofa rapid inhibitor of tissue plasminogen activator in young survivors of myocardial infarction. N. Engl. J. Med. 313: 1557-1563,1985. Haskel, E.F., S.R. Torr, K.C. Day, M.O. Palmier, T.C. Wun, and BE. Sobel. Prevention of arterial reocclusion after thrombolysis with recombinant lipoprotein-associated with coagulation inhibitor. Circulation 84: 821-827, 1991. Hatton, M.W.C., and ST. Moar. A role for pericellular proteoglycan in the binding ofthrombin or antithrombin by the blood vessel endothelium. The effects of proteoglycan-degrading enzymes and glysoaminoglycan-binding protein on 12’I-thrombin binding by the rabbit thoracic aorta in vitro. Thromb. Haemostasis. 54: 228-234, 1985. Heimark, R.L., and SM. Shwartz. Binding of coagulation factors IX and X to the endothelial cell surface. Biochem. Biophys. Res. Commun. 111:723-731, 1983. Hekman, C.M., and DJ. Loskutoff. Kinetic analysis of the interactions between plasminogen activator inhibitor 1 and both urokinase and tissue plasminogen activator. Arch. Biochem. Biophys. 262: 199-210, 1988. 177 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. Hekman, C.M., and DJ. Loskutoff. Endothelial cells produce a latent inhibitor of plasminogen activators that can be activated by denaturants. J. Biol. Chem. 260: 11581-11587, 1985. Hendrix, M.G., P.H. Dormans, P. Kctslaar, F. Bosman, and CA. Bruggeman. The presence ofcytomegalovirus nucleic acids in arterial walls of atherosclerotic and nonatherosclerotic patients. Am. J. Pathol. 134: 1151-1157, 1989. Hendrix, M.G., M.M. Salimans, C.P. van-Boven, and CA. Bruggeman. High prevalence oflatently present cytomegalovirus in arterial walls of patients suffering from grade III atherosclerosis. AM. J. Pathol. 136: 23-28, 1990. Henson, J.B., and TB. Crawford. The pathogenesis of virus induced arterial disease, Aleutian disease and equine viral arteritis. Ad Cardiol 13: 183-191, 1974. Hoak, J.C., R.L. Czervionke, G.L. Fry, and TB. Smith. Interaction ofthrombin and platelets with the vascular endothelium. Federation Proc. 39: 2606-2609, 1980. Isaacs, J.D., N. Savcor, D. Gospodarowicz, J.W. Fenton II, and M.A. Sherman. Covalent binding ofthrombin to specific sites on corneal endothelial cells. Biochemistry 20: 398-403, 1981. Jaffe, E.A., R.L. Nachman, C.G. Becker, and CR. Minick. Culture of human endothelial cells derived from umbilical veins. Identification by morphology and immunologic criteria. J. Clin. Invest. 52: 2745-2756, 1973. Jarrett, R.F., D.A. Clark, S.F. Josephs, and DE. Onions. Detection of human herpesvirus-6 DNA in peripheral blood and saliva. J. Med. Virol. 32: 73-76, 1990. Jern, S., L. Selin, A. Bergbrant, and C. Jern. Release oftissue- type plasminogen activator in response to muscarinic receptor stimulation in human forearm. Thromb. Haemost. 72: 588-594, 1994. Jerrell, 8., EM. Levine, S.Shapiro, S. Williams, A. Carabasi, et al. Human adult endothelial cell growth in culture. J. Vascular Surgery 1: 757-764, 1984. 178 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. Johnson, C.M., and SC. Helgeson. Platelet adherence to cardiac and non cardiac endothelial cells in culture: lack ofa prostacyclin effect. J. Lab. Clin. Med. 112:372-379,1988. Jones, M.E., M.J. Griffith, D.M. Monroe, H.R. Roberts, and BR. Lentz. Comparison oflipid binding and kinetic properties of normal, variant, and gamma-carboxyglutamic acid modified human factor IX and factor [Xa. Biochemistry 24:8064-8069, 1985. Joseph, SE, A. Buchbinder, and H.Z. Streicher. Detection of HHV-6 sequences in B-cell lymphoma tissues ofthree patients. Leukemia 2: 132-135, 1988. Juhan-Vague, I., B. Moerman, F. DeCock, M.F. Ailland, and D. Colleen. Plasma levels of a specific inhibitor oftissue-type plasminogen activator (and urokinase) in normal and pathological conditions. Thromb. Res. 33:523-530, 1984. Kalafatis, M., and K.G. Mann. Role of the membrane in the inactivation of factor Va by activated protein C. J. Biol. Chem. 268: 27246-27257, 1993. Kaner, R., R. 10220, Z. Ziaie, and Z. Kefalides. Inhibition of proteoglycan synthesis in human endothelial cells after infection with herpes simplex virus type 1 in vitro. Am. J. Respir. Cell Mol. Biol. 2: 423-431, 1990. Kaplan, J.B., D.G. Moon, L.K. Weston, FL. Minnear, P.J. Del- Vecchio, et al. Platelets adhere to thrombin-treated endothelial cells in vitro. Heart Circ. Physiol. 26: H423-H433, 1989. Keber, D., Mechanism of tissue plasminogen activator release during venous occlusion. Fibrinolysis 2: 96-103, 1988. Kelm, R.J., N.A. Swords, T. Orfeo, and K.G. Mann. Osteonectin in matrix remodeling: a plasminogen-osteonectin-collagen complex. J. Biol. Chem. 269: 30147-30152, 1994. Key, N., G. Vercellotti, J. Winkelmann, C. Moldow, J.L. Goodman, et al. Infection of vascular endothelial cells with herpes simplex virus enhances tissue factor activity and reduces thrombomodulin expression. Proc. Natl. Acad. Sci. USA. 87: 7095-7097, 1990. 179 122. 123. 124. 125. 126. 127. 128. 129. 130. Key, N.S., R.R. Bach, G.M. Vercellotti, and CF. Moldow. Herpes simplex virus type I does not require productive infection to induce tissue factor in human umbilical vein endothelial cells. Lab. Invest.68: 645-651, 1993. Klinger, K.W., R. Winqvist, A. Riccio, P.A. Andreasen, R. Surtorio, et al. Plasminogen activator inhibitor type 1 gene is located at region q21.3-q22 of chromosome 7 and genetically linked with cystic fibrosis. Proc. Natl. Acad. Sci. USA 84: 8548- 8552, 1987. Kluft, C., J.H. Verheijen, A. F. Jie, D.C. Rijken, F.E. Preston, et al. The postoperative fibrinolytic shutdown: a rapidly reverting acute phase pattern for the fast-acting inhibitor oftissue-type plasminogen activator after trauma. Scand. J. Clin. Lab. Invest. 45: 605-610, 1985. Kluft, C., A.F.Jie, D.C. Rijken, and J.H. Verheijen. Daytime fluctuations in blood of tissue-type plasminogen activator (t-PA) and its fast-acting inhibitor (PAI-l). Thromb. Haemost. 59: 329- 332, 1988. Knox, K.K., and DR. Carrigan. Disseminated active HHV-6 infections in patients with AIDS. Lancet. 343: 577-578, 1994. Knudsen, B.S., P.C. Harpel, and R.L. Nachman. Plasminogen activator inhibitor is associated with the extracellular matrix of cultured bovine smooth muscle cells. J. Clin. Invest. 80: 1082- 1089, 1987. Kondo, K., T. Kondo, T. Okuno, M. Takahashi, and K. Yamanishi. Latent human herpesvirus 6 infection of human monocytes/ macrophages. J. Gen. Virol. 72: 1401-1408, 1991. Kooistra, T., P.J. Bosma, J. Jespersen, and C. Kluft. Studies on the mechanism of action of oral contraceptives with regard to fibrinolytic variables. Am. J. Gynecol. Obstet. 163: 404-413, 1990. Krause, J., and P. Tanswell. Properties of molecular variants of tissue-type plasminogen activator. Arzneim Forsch/Drug Res. 39: 632-637,1989. 180 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. Krueger, G.R., U. Klueppelberg, A. Hoffman, and D.V. Ablashi. Clinical correlates ofinfection with human herpesvirus-6. In Vivo. 8: 457-85, 1994. Krueger, G.R., K. Wassermann, L.S. De-Clerck, W.J. Stevens, and N. Bourgeois. Latent herpesvirus-6 in salivary and bronchial glands. Lancet 336: 1255-1256,1990. Krueger, G.R.F., D.V. Ablashi, S.Z. Salahuddin, and SF. Joseph. Diagnosis and differential diagnosis of progressive lymphoproliferation and malignant lymphoma in persistent active herpesvirus infection. J. Virol. Methods 21: 255-264, 1988. Kruithof, E.K.O., A. Gudinchet, and F. Bachmann. Plasminogen activator inhibitor 1 and plasminogen activator inhibitor 2 in various disease states. Thromb. Haem. 59: 7-12, 1988. Lawrence, M.E., C.W. Smith, S.G. Eskin, and L.V. McIntire. Effect of venous shear stress on CD18-mediated neutrophil adhesion to cultures endothelium. Blood 75:227-237, 1990. Lawrence, D.A., and DJ. Loskutoff. Inactivation of plasminogen activator inhibitor by oxidants. Biochemistry 25: 6351-6355, 1986. Lawrence, G.L., M. Chee, M.A. Craxton, U.A. Gompels, R.W. Honess, and HG. Barrell. HHV-6 is closely related to human CMV. I. Vir. 64(1): 287-299, 1990. Leprince, P., Rogister, B. and G. Moonen. A colorimetric assay for the simultaneous measurement of plasminogen activators and plasminogen activator inhibitors in serum-free conditioned media from cultured cells. Analytical Biochemistry 177: 341-346, 1989. Levin, E.G., U. Marzec, J. Anderson, and LA. Harker. Thrombin stimulates tissue plasminogen activator release from cultured endothelial cells. J. Clin. Invest. 74: 1988-1995, 1984. Levin, E.G., and L. Santell. Conversion of active to latent plasminogen activator inhibitor from human endothelial cells. Blood 70: 1090-1098, 1987. Levine, P.H., N. Jahan, P. Murari, M. Manak, and ES. Jaffe. Detection of human herpesvirus 6 in tissues involved by sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman 181 disease). J. Infect. Dis. 166: 291-295, 1992. 142. Lieb, J. Immunopotentiation and inhibition of herpes virus activation during therapy with lithium carbonate. Med. Hypothesis 7: 885-890, 1981. 143. Lijnen, J.R., and D. Collen. Molecular interactions between tissue-type plasminogen activator and plasminogen. Methods Enzymol. 223: 197-206, 1993. 144. Loebe,M., S. Schuler, S. Spiegelsberger, H. Warnecke, E. Fleck, and R. Hetzer. Cytomegalovirus infection and coronary sclerosis after heart transplantation. Dtsch. Med. Wochenschr. 115: 1266- 1269, 1990. 145. Lollar, P., and W.G. Owen. Clearance ofthrombin from circulation in rabbit by high-affinity binding sites on endothelium. Possible role in the inactivation ofthrombin by antithrombin III. J. Clin. Invest. 66: 1222-1230, 1980. 146. London, F.S., Z. Ziaie, J.M. Brinker, and N.A. Kefalides. Suppression of host mRNA in human smooth muscle cells by a virion competent factor in herpes simplex virus type 1. Lab. Invest. 62: 189-196, 1990. 147. Lopez, C., P. Pellet, J. Stewart, C. Goldsmith, and K. Sanderlin. Characteristics ofHHV-6. J. Infect. Dis. 157: 1271-1273, 1988. 148. Loskutoff, D.J., M. Sawdey, and J. Mimuro. Type I plasminogen activator inhibitor. In: Progress in Hemostas Thromb. Coller,B. ed. 9:87-115, 1988. 149. Loskutoff, D.J., M. Linders, J. Keijer, H. Veerman, H. Van- Heerikhuizen, and H. Pannekoek. The structure of the human plasminogen activator inhibitor 1 gene: non-random distribution ofintrons. Biochem 26: 3763-3768, 1987. 150. Loskutoff, D.J., J.A. van Mourik, L.A. Erickson, and D. Lawrence. Detection of an unusually stable fibrinolytic inhibitor produced by bovine endothelial cells. Proc. Natl. Acad. Sci. USA 80: 2956-2960, 1983. 182 .1 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. Lund, LR, and A. Riccio. Transforming growth factor-B is a strong and fast acting positive regulator of the level of type-1 plasminogen activator inhibitor mRNA in WI-38 human lung fibroblasts. EMBO J. 6: 1281-1286,1987. Luppi, M., P. Barozzi, A. Maiorana, R. Marasca, R.Trovato, et al. Human herpesvirus-6: a survey of presence and distribution of genomic sequences in normal brain and neuroglial tumors. J. Med. Virol.47:105-111,1995. Lusso, P., B. Ensoli, P.D. Markham, D.V. Ablashi, S.Z. Salahuddin, et al. Productive dual infection of human CD4+ lymphocytes by HIV-1 and HHV-6. Nature 337: 370-373, 1989. Lusso, P., and RC. Gallo. Human herpesvirus 6 in AIDS. Immunol. Today. 16: 67-71, 1995. Macfarlane, R.G.. An enzyme cascade in the blood clotting mechanism and its function as a biochemical amplifier. Nature 202: 498-499, 1964. Majno, G. Personal communication, in expansion upon: Zand, T., et al. Endothelial adaptations in aortic stenosis. Correlation with flow parameters. Am. J. Pathol. 133: 407-418, 1988. Marcum, J., J. Mckenney, and R. Rosenberg. Acceleration of thrombin-antithrombin complex formation in rat hindquarters via heparin like molecules bound to the endothelium. J. Clin. Invest. 74:341-350,1984. Marcus, A.J., L.B. Safier, K.A. Hajjar, H.L. Ullman, N. Islam, et al . Inhibition of platelet function by an aspirin-insensitive endothelial cell AD-Pase. J. Clin. Invest. 88: 1690-1696, 1990. Mason, R.G., S.F. Mohammad, H.Y. Chuang, and PD. Richardson. The adhesion of platelets to subendothelium, collagen and artificial surfaces. Semin. Thromb. Hemost.3: 98-116, 1976. Mathison, J., E. Wolfson, and R. Ulevitch. Participation of tumor necrosis factor in the mediation of gram negative bacterial LPS- induced injury in rabbits. J. Clin. Invest. 81: 1925-1937, 1988. 183 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. Mattila, S., L. Heikkila, J. Sipponen, K. Verkkala, K. Kyosola, T. Mattila, A. Jarvinen, R. Luosto, L. Heikkinen, et al. Heart transplantation in Finland 1985-1995. Ann. Chir. Gynaecol.86: 113-120,1997. McCullers, J.A., F.D. Lakeman, and RJ. Whitley. Human herpesvirus 6 is associated with focal encephalitis. Clin. Infect. Dis. 21: 571-576, 1996. McSorley, J., L. Shapiro, M.H. Brownstein, and KC. Hsu. Herpes simplex and Varicella-Zoster: comparative histopathology of 77 cases. Int. J. Dermatol. 13: 69-75, 1974. Melnick, J.L., C. Hu, J. Burek, E. Adam, and ME. DeBakey. Cytomegalovirus DNA in arterial walls of patients with atherosclerosis. J. Med. Virol. 42: 170-174, 1994. Miles, L.A., E.G. Levin, J. Plescia, D. Collen, and BF. Plow. Plasminogen receptors, urokinase receptors and their modulation on human endothelial cells. Blood 72: 628- 634, 1988. Millett, R., T. Tomita, H.E. Marshall, L. Cohen, and H.Hannah. Cytomegalovirus endomyocarditis in a transplanted heart. A case report with in situ hybridization. Arch. Pathol. Lab. Med. 115: 511-515, 1991. Mimuro, J., M. Kaneko, M. Matruda, and Y. Sakata. The plasminogen activator inhibitor-1 binding site in the kringle-Z domain oftissue plasminogen activator. Biochem. Biophys. 178: 1160-1166, 1991. Moncada,S., R.M.J. Palmer, and EA. Higgs. Nitric oxide: physiology, pathophysiology and pharmacology. Pharmacol. Rev. 43:109-142,1991. Moncada, S., and J.R. Vane. Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2 and prostacyclin. Pharmacol. Rev. 30: 293-331, 1979. Moncada, S. Biological importance of prostacyclin. Br. J. Pharmacol. 76: 3-31, 1982. 184 171. 172. 173. 174. 175. 176. 177. 178. 178a 179. Moncada, S., R. Gryglewski, S. Bunting, and J.R. Vane. An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature 263: 663-665, 1976. Moore, K.L., N.U. Bang, C.T. Esmon, and N.L. Esmon. The thrombogenic endothelial cells. Two endotoxin-mediated mechanisms. Clin. Res. 33: 350a, 1985. Morrison, DC, and R.J. Ulevitch. The effects of bacterial endotoxins on host mediation systems. Am. J. Path. 93: 527-617, cloning of the cDNA for tissue factor, the cellular receptor for the initiation of the coagulation protease cascade. Cell 50: 129-135, 1987. 1978. Morrissey, J.H., Y. Fakhrai, and TS. Edgington. Molecular [I Muller, A.D., M. Van Dam-Micras, and H. Hemker. The measurement of macrophage cellular procoagulant activity. Haemostasis. 15: 108-113,1985. Muller-Berghaus, G. Pathophysiologic and biochemical events in disseminated intravascular coagulation: dysregulation of procoagulant and anticoagulant pathways. Semi. Thromb. Hemost. 15: 58-98, 1989. Musiani, M., M.L. Zerbini, A. Muscari, G.M. Puddu, G. Gentilomi, et al. Antibody patterns against cytomegalovirus and Epstein-Barr virus in human atherosclerosis. Microbiologica 13: 35-41, 1990. Myerson, D., R.C. Hackman, J. A. Nelson, D.C. Ward, and J.K. McDougall. Widespread presence of histologically occult cytomegalovirus. Hum. Pathol. 15: 430-439, 1984. Naimi, A., R. Goldstein and S. Proger. Studies of coagulation and fibrinolysis in the arterial and venous blood in normal subjects and patients with arthersclerosis. Circulation 27:904-918, 1963 Naito, K. and K. Fujikawa. Activation of human blood coagulation factor XI independent of factor XII: factor X1 is activated by thrombin and factor XIa in the presence of negatively charged surfaces. J. Biol. Chem. 266: 7353-7358, 1991. 185 180. 181. 182. 183. 183a 184. 185. 186. 187. 188. 189. Nawroth, P.P., D.A. Handly, C.T. Esmon and D.M. Stern. IL-l induces endothelial cell procoagulant while suppressing cell- surface anticoagulant activity. Proc. Natl. Acad. Sci. USA 83: 3460-3464, 1986. Nemerson, Y. Tissue factor and haemostasis. Blood 71: 1-8, 1988. Nemerson, Y., and R. Bach. Tissue factor revisited. Prog. Hemostasis 6: 237-261, 1982. Nesheim, M., M.N. Blackburn, C.M. Lawler, and K.G. Mann. Dependence of antithrombin III and thrombin binding stoichiometric and catalytic activity on the molecular weight of affinity-purified heparin. J. Biol. Chem. 261: 3214-3221, 1986. Nicholson, AC. and DP. Hajjar. Hyperviruses in atherosclerosis and thrombosis. Etiologic agents or ubiquitous bystanders? Arteroscler Thromb Vasc Biol 18:339-348, 1998. Niederman, J.C., M.H. Kaplan, C.P. Liu, and N.A. Brown. Clinical and serological features of HHV-6 infections in three adults. Lancet 2: 817-819, 1988. Nieto, F.J., E. Adam, P. Sorlie, H.Farzadegan, J.L. Melnick, et al. Cohort study of cytomegalovirus infection as a risk factor for carotid intimal-medial thickening, a measure of subclinical atherosclerosis. Circulation 94: 922-927, 1996. Niewiarowski, S. Secreted platelet proteins. In Bloom, A.L., et al. Haemostasis and thrombosis. 3rd ed. 167-181, 1994. Nii, S., M. Yoshida, F. Uno, T. Kurata, K. Ikuta, and K. Yamanishi. Replication of human herpesvirus 6 (HHV-6): morphological aspects. Adv. Exp. Med. Biol. 278: 19-28, 1990. Nilsson, I.M., and H. Ljunger. Two different mechanisms in patients with venous thrombosis and defective fibrinolysis: Low concentration of plasminogen activator or increased concentration of plasminogen activator inhibitor. Brit. Med. J. 290: 1453-1456, 1985. Nishimura, K., and M. Igarashi. Thrombocytopenic purpura associated with exanthem subitum. Pediatrics 60: 260-263, 1977. 186 190. Ny, T., M. Ohlsson, and L. Strandberg. The gene for t-PA. In: C. Kluft ed. Tissue-type plasminogen activator (t-PA). Physiological and Clinical Aspects 1: 83-100, 1988. 191. Okuno, T., T. Takahashi, K. Balachandra, K. Shiraki, K. Yamanishi, et al. Seroepidemiology of HHV-6 infection in normal children and adults. J. Clin. Microbiol. 27: 651-653, 1989. 192. Okuno, T., K. Higashi, K. Shiraki, K. Yamanishi, M. Takahashi, and Y. Kokado. Human herpesvirus 6 infection in renal transplantation. Transplantation 49: 519-522, 1990. 193. Oski, F.A. Roseola infantum: another cause of bulging fontanel. Am. J. Dis. Child. 101: 376-378, 1961. 194. Osterud B. and 8.1. Rapaport. Activation of factor IX by the reaction product of TF and factor VII: additional pathway for initiating blood coagulation. Proc. Natl. Acad. Sci. USA. 74: 5260-5264, 1977. 195. Owen, W.G., and CT. Esmon. Functional properites of an endothelial cell cofactor for thrombin-catalyzed activation of protein C]. Biol. Chem. 256: 5523-5535, 1982. 196. Pedersen, O.D., J. Gram., and J. Jespersen. Plasminogen activator inhibitor type-1 determines plasmin formation in patients with ischaemic heart disease. Thromb. Haemost. 73: 835-840, 1995. 197. Pellett, P.E., G.J. Lindquester, P. Feorino, and C. Lopez. Genomic heterogeneity of human herpesvirus 6 isolates. Adv. Exp. Med. Biol. 278: 9-18, 1990. 198. Petersen L.C., V. Linde, L.S. Nielson, and DC. Foster. Elimination ofthe CysSS8-CysS66 bond in Lys 78-plasminogen-- efect on activation and fibrin interaction. Eur. J. Biochem. 251: 472-479, 1998. 199. Phinney, P.R., S. Fligiel, Y.J. Bryson, and DD. Porter. Necrotizing vasculitis in a case of disseminated neonatal herpes simplex infection. Arch. Pathol. Lab. Med. 106: 64-67, 1982. 199a Peabody, R.A., M.J. Tsapogas, K.T. Wu, K.T. Deveraj, A.M. Karmody, and C. Eckert. Altered endogenous fibrinolysis and biochemical factors in atherosclerosis. Arch Surg 109:309-313, 187 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 1974. Pietroboni, G.R., G.B. Harnet, M.R. Burns, and R.W. Honess. Isolation of human herpesvirus 6 from saliva. Lancet 1: 1059, 1988. Piper, P., and J.R. Vane. The release of prostaglandins from lung and aother tissues. Ann. NY. Acad. Sci. 180: 363-385,1971. Podor, T.J., R. Schleef, and DJ. Loskutoff. Immunoelectron microscopic localization oftype l plasminogen activator inhibitor on the surface of activated endothelial cells. J. Cell Biol. 113: 1413-1423, 1991. Prezioso, P.J., J. Cangiarella, M. Lee, G.J. Nuovo, W. Borkowsky, et al. Fatal disseminated infection with human herpesvirus-6. J. Pediatr. 120: 921-923, 1991. Pryzdial, E., and J.Wright. Prothrombinase assembly on an enveloped virus: evidence that the cytomegalovirus surface contains procoagulant phospholipid. Blood 84: 3749-3757, 1994. Rabbani, E., and J. Loscalzo. Recent observations on the role of hemostatic determinants in the development ofthe atherothrombotic plaque. Atherosclerosis. 105: 1-9, 1994. Radcliffe, RD. and Y. Nemerson. Mechanism of activation of bovine factor VII: products of cleavage by factor Xa. J. Biol. Chem. 251: 4749-4802, 1976. Radomski, M.W., R.M. Palmer, and S. Moncada. The anti- aggregating properties of vascular endothelium: Interaction between prostacyclin and nitric oxide. Br. J. Pharmacol. 92: 639- 646, 1987. Radomski, M.W., R.M. Palmer, and S. Moncada. The role of nitric oxide and cGMP in platelet adhesion to vascular endothelium. Biochem. Biophys. Res. Commun. 148: 1482-1489, 1989. Ranby, M., N. Bergsdorf, T. Nilsson, G. Mellbring, B. Winblad, and G. Bucht. Age dependence oftissue plasminogen activator concentrations in plasma as studied by an improved enzyme-linked immunosorbent assay. Clin. Chem. 32: 2160-2165, 1986. 188 210. 211. 212. 213. 214. 215. 216. 217. 217a Ranger-Rogez, S., E. Vidal, F. Labrousse, A. Riche, J. Vidal, and M. Collineau. Large-scale study suggests no direct link between human herpesvirus-6 and primary Sjogren’s syndrome. J. Med. Virol. 47: 198-203, 1995. Rao, L.V.M., and SI. Rappaport. Activation of factor VII bound to TF (tissue factor). A key step in the TF pathway of coagulation. Proc. Natl. Acad. Sci. USA 85: 6687-6691, 1988. Rappaport, 8.1., and L.V. Rao. Initiation and regulation of tissue factor-dependent blood coagulation. Arterioscler. Thromb. 12: 1111-1121,1991. Ratnoff, OD. and IE. Colopy. A familial hemorrhagic trait associated with a deficiency ofa clot- promoting fraction of plasma. J. Clin. Invest. 34: 602-613, 1955. Raza, A.A., G.A. Klassen, D.A. Murphy, J.A.Sullivan, C.E. Kinley, et al. Evidence oftype 2 herpes simplex infection in human coronary arteries at the time of coronary artery bypass surgery. Can. J. Cardiol. 11: 1025-1029,1995. Reyes, M.G., R. Fresco, S. Chokroverty, and E.Q. Salud. Virus like particles in granulomatous angiitis ofthe central nervous system. Neurology 26:797-799, 1976. Rheinwald, J.G., J.L. Jorgensen, W.C. Hahn, A.J. Terpstra, T.M. O’Connell, and K.K. Plummer. Mesosecrin: A secreted glycoprotein produced in abundance by human mesothelial, endothelial and kidney epithelial cells in culture. J. Cell Biol. 104: 263-275,1987. Rijken, D.C., G. Wijngaards, and D. Collen. Tissue-type plasminogen activator from human tissue and cell cultures and its occurrence in plasma. In: D. Collen et al. . Thrombolysis: Biological and Therapeutic Properties of New Thrombolytic Agents 15-30, 1985. Roberts, W.C. Does thrombosis play a major role in the development of symptom-producing atherosclerotic plaques? Circulation 48: 1 161-1 166, 1972. 189 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. Rodgers, G.M. Hemostatic properties of normal and perturbed vascular cells. FASEB J. 1: 116-123, 1988. Roffman, E., J.P. Albert, J. P. Goff, and N. Frenkel. Putative site for the acquisition of human herpesvirus 6 virion tegument. J. Virol. 64: 6308-6313, 1990. Roizman, B., and AB. Sears. Herpes simplex viruses and their replication. In Fields, B.N., ct al.(ed). Fields virology, 3rd ed., vol. 2. 2231-2295, 1996. Roizman, B., The family Herpesviridae: general description, taxonomy, and classification: In Roizman (ed.). The herpesviruses, Plenum publishing. NY. 1: 1-23, 1982. Rokitansky, C. A manual of pathological anatomy. Vol IV, part II, chap. 3 - Abnormal conditions ofthe arteries. London, Sydenham Society. 1952. Rosenberg, R.D. Heparin-antithrombin system. Hemostasis and thrombosis: basic principles and clinical practice. J.B. Lippincott, Philadelphia, 837-962, 1982. Russler, S.K., M.A. Tapper, and DR. Carrigan. Susceptibility of human herpesvirus 6 to acyclovir and ganciclovir (Letter). Lancet. 2: 382, 1989. Salahuddin, S.Z., D.V. Ablashi, P.D. Markham, S.F. Josephs, S. Sturzenegger, et al. Isolation ofa new virus, HBLV, in patients with lymphoproliferative disorders. Science 234: 596-601, 1986. Salomon, R., C. Hughes, F. Schoen, D. Payne, J.S. Pober, and P. Libby. Human coronary transplantation-associated arteriosclerosis: evidence for a chronic immune reaction to activated graft endothelial cells. Am. J. Pathol. 138: 791-798, 1991. Sandhoff, T., J.P. Kleim, and KB. Schneweis. Latent HHV-6 DNA sparsely distributed in peripheral blood lymphocytes of healthy adults and patients with lymphocytic disorders. Med. Microbiol. Immunol. 80: 127-134, 1991. 190 228. 229. 230. 231. 232. 233. 234. 235. 236. Sandset, F.M., B.J. Warn Cramer, L.V. Rao, S.L. Maki, and SI Rapaport. Depletion of extrinsic pathway inhibitor (EPI) sensitizes rabbits to disseminated intravascular coagulation induced with tissue factor. Evidence supporting a physiologic role for EPI as a natural anticoagulant. Proc. Natl. Acad. Sci. USA 88: 708-712,1991. Sanduja, S.K., A.L. Tsai, N. Matijevic-Aleksic, and K.K. Wu. Kinetics of prostacyclin synthesis in PGHS-l over-expressed endothelial cells. Am. J. Physiol. 267: 1459-1466, 1994. Savage, B. ,and Z. M. Ruggeri. Selective recognition of adhesive sites in surface- bound fibrinogen by glycoprotein II on nonactivated platelets. J. Biol. Chem. 266: 11227-11233, 1991. Savion, N., J.D.Isaacs, D. Gospodarowicz, and M.A. Shuman. Internalization and degradation ofthrombin and up regulation of thrombin-binding sites in corneal endothelial cells. J. Biol. Chem. 256:4514-4519,1981. Sawdey, M., and T.J. Podor. Regulation oftype 1 plasminogen activator inhibitor gene expression in cultured bovine aortic endothelial cells: Induction by transforming growth factor-B, lipopolysaccharide, and tumor necrosis factor-a. J. Biol. Chem. 264: 10396-10401,1989. Schleef, RR, and MP. Bevilacqua. Cytokine activation of vascular endothelium: effects on tissue-type plasminogen activator and type 1 plasminogen activator inhibitor. J. Biol. Chem. 263: 5797-5803, 1988. Schleef, R.R., D.L. Higgins, E. Pillemer, and L.J. Levitt. Bleeding diathesis due to decreased functional activity of type 1 plasminogen activator inhibitor. J. Clin. Invest. 83: 1747-1752, 1989. Scott, D.G.I., and A. Blann. Activated, cytotoxic lymphocytes in systemic vasculitis. Rheumatol. Int. 11: 69-72, 1991. Seiffert, D., N.N. Wagner, D.J. Loskutoff. Serum-derived vitronectin influences the pericellular distribution oftype 1 plasminogen activator inhibitor. J.Cell Biol. 266: 2824-2830, 1990. 191 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. Sergent, J.S.. Vasculitis associated with viral infections. Clin. Rheum. Dis. 6:339-349, 1980. Shaw, G., and R. Kamen. A conserved AU sequence from the 3' untranslated region of GM-C SF mRNA mediates selective mRNA degradation. Cell 46: 659-667, 1986. Shimada, K., and T. Ozawa. Evidence that cell surface heparan sulfate is involved in the high affinity thrombin binding to cultured porcine aortic endothelial cells. J. Clin. Invest. 75: 1308-1316, 1985. Smith, D.H., N.D. Vaziri, R.L. Winer, J.M. Neutel, W.F. Graettinger, et al. The relationship of arterial compliance with endothelial-derived proteins ofthe hemolytic system. Life Sci. 52: 1005-1011, 1993. Smith, E.B. Fibrinogen, fibrin, and the arterial wall. Eur. Heart J. 16:11-15,1995. Sobue, R., H. Miyazaki, M. Okamoto, M. Hirano, T. Yoshikawa, et al. Fulminant hepatitis in primary HHV-6 infection, NEJM. 324: 1290-1293, 1991. Sorlie, RD, E. Adam, S.L. Melnick, A. Folsom, T. Skeleton, et al. Cytomegalovirus/herpesvirus and carotid atherosclerosis: the ARIC Study. J.Med. Virol. 42(1): 33-37, 1994. Soter N.A. Clinical presentations and mechanisms of necrotizing angiitis ofthe skin. J. Invest. Dermatol. 67: 354-359, 1976. Span, A.H., M.C. van-Dam-Mieras, W. Mullers, J. Endert, A.D. Muller, et al. The effect of virus infection on the adherence of leukocytes or platelets to endothelial cells. Eur. J. Clin. Invest. 21(3): 331-338, 1991. Span, A., J. Endert, C. van Boven, and C. Bruggeman. Virus induced adherence of monocytes to endothelial cells. FEMS Microbiol. Immunol. 47: 237-244, 1989. Span, A.H., P.M. Frederik, G. Grauls, G.P. Van-Boven, and CA. Bruggeman. CMV induced vascular injury: an electron- microscopic study in the rat. In Vivo 7(6A): 567-573, 1993. 192 248. 249. 250. 251. 252. 253. 254. 255. 256. 257. Spicer, E.K., R. Horton, L. Bloem, R. Bach, K.R. Williams, et al. Isolation of cDNA coding for human tissue factor/; primary structure ofthe protein and cDNA. Proc. Natl. Acad. Sci. USA 84: 5148-5152,1987. Sprengers, ED, and C. Kluft, Plasminogen activator inhibitors. Blood 69:381-387,1987. Stebbens, W.E. Hemodynamic production oflipid deposition, intimal tears, mural dissection and thrombosis of blood vessel walls. Proc. R. Soc. London Ser. B. 185: 357- 364, 1976. Stegnar, M., and D. Keber. Age and sex differences in resting and postocclusion values oftissue plasminogen activator in a healthy population. Fibrinolysis 2: 121-122, 1988. Stemerman, M., T.H. Spaet, F. Pitlick, J. Cintron, I. Lejnieks, and ML. Tiell. Intimal healing: the pattern of re-endothelialization and intimal thickening. Am. J. Pathol. 87:125-133, 1977. Stump, D., and K.G. Mann. Mechanisms ofthrombosis formation. Ann. Emerg. Med. 17: 1138-1147, 1988. Suga, S., T. Yoshikawa, Y. Asano, T. Yazaki, and S. Hirata. HHV- 6 infection (exanthem subitum) without rash. Pediatrics 83: 1003- 1006, 1989. Suga, 8., T. Yoshikawa, Y. Asano, T. Yazaki, and T. Ozaki. Neutralizing Antibody Assay for Human Herpesvirus-6. J. Med. Virol. 30: 14-19, 1990. Sundell, I.B., T.K. Nilsson, G. Hallmans, G. Hellsten, and G.H. Dahlen. Interrelationships between plasma levels of plasminogen activator inhibitor, tissue plasminogen activator, lipoprotein (a), and established cardiovascular risk factors in a north Swedish population. Atherosclerosis 80: 9-16, 1989. Swords, N.A., P.B. Tracy, and K.G. Mann. Intact platelet membranes, not platelet-released microvesicles, support the procoagulant activity of adherent platelets. Arterioscler. Thromb. 13: 1613-1622, 1993. 193 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. Tabernero, J.D., A. Estelles, V. Vicente, I. Alberca, and J. Aznar. Incidence ofincreased plasminogen activator inhibitor in patients with deep venous thrombosis and/or pulmonary embolism. Thromb. Res. 56: 565-570, 1989. Takashima, N., and T. Higashi. Change in fibrinolytic activity as a parameter for assessing local mechanical stimulation during physical exercise. Eur. J. Appl. Physiol. 68(5): 445-449, 1994. Tankaka, K., and K. Sueishi. Biology of disease: the coagulation and fibrinolysis systems and atherosclerosis. Lab. Invest. 69: 5- 13, 1993. Thompson, E.A., and H.H. Salem. Inhibition of human thrombomodulin of factor Xa mediated cleavage of prothrombin. J. Clin. Invest. 78: 13-17, 1986. Tollefsen, D.M., M.M. Maimone, E.A. McGuire, and ME. Peacock. Heparin cofactor II activation by dermatan sulfate. Ann. NY Acad. Sci. 556: 116-122, 1989. Toyoda, M., K. Galfayan, O.A. Galera, A. Petrosian, L.S. Czer, and SC. Jordan. Cytomegalovirus infection induces anti- endothelial cell antibodies in cardiac and renal allograft recipients. Transplant Immunology 5(2): 104-111, 1997. Tracy, P.B., L.L. Eide, E.J. Bowie, and K.G. Mann. Radioimmunoassay of factor V in human plasma and platelets. Blood 60: 59-63, 1982. Tranquille, N., and J.J. Emeis. Protein synthesis inhibition by cycloheximide does not affect the acute release of tissue-type plasminogen activator. Thromb. Haemost. 61: 442-447, 1989. Ustinov, J., R. Loginov, C. Bruggeman, P. van der Meide, P.J. Hayry, and LT. Lautenschlager. Cytomegalovirus induces class II expression in rat heart endothelial cells. J. Heart Lung Transplant.12: 644-651, 1993. Vaccaro, P.S., S.B. Joseph, L. Titterington, and RE. Stephans. Methods for initiation and maintenance of human endothelial cell culture. Vascular surgery, 1987. 194 268. 269. 270. 271. 272. 273. 274. 275. 276. Valen, G., E. Eriksson, B. Risberg, and J. Vaage. Reactive oxygen intermediates and ischemia-reperfusion injury release t-PA from isolated rat hearts. Thromb. Res. 71(2): 113-21, 1993. Van Zonneveld, A.J., G.T.G. Chang, J. Van den Berg, T. Kooistra, J.H. Verheijen, et al. Quantification oftissue-type plasminogen activator (t-PA) mRNA in human endothelial-cell cultures by hybridization with a t-PA cDNA probe. Biochem. J. 235: 385-390, 1986. Van Dam-Mieras, M.C.E., C.A. Bruggeman, A.D. Muller, W.H. debie, and RF. Zwaal. Induction of endothelial cell procoagulant activity by cytomegalovirus infection. Thromb. Res. 47: 69-87, 1987. Van Hinsburgh, V.W.M. Regulation ofthe synthesis and secretion ofplasminogen activators by endothelial cells. Haemostasis 18: 307-327,1988. Verheijen, J.H., D.C. Rijken, G.T. Chang, F.E. Preston, and C. Kluft. Modulation of rapid plasminogen activator inhibitor in plasma by stanozolol. Thromb. Haemost. 51: 396-397, 1984. Villeda, C.J., J.C. Gomez-Villalmandos, S.M. Williams, J. Hervas, P.J. Wilkinson, and E. Vinuela. The role of fibrinolysis in the pathogenesis ofthe haemorrhagic syndrome produced by virulent isolates of african swine fever virus. Thromb. Haemost. 73(1): 112-117,1995. Visser, M.R., P.B. Tracy, G.M. Vercellotti, J.L. Goodman, J.G. White, and HS. Jacob. Enhanced thrombin generation and platelet binding on herpes simplex virus-infected endothelium. Proc. Natl. Acad. Sci. USA 85: 8227-8230, 1988. Visser, M.R., G.M. Vercellotti, J.B. McCarthy, J.L. Goodman, T.J. Herbst, et al. Herpes simplex virus inhibits endothelial cell attachment and migration to extracellular matrix proteins. Am.J.Pathol.l34(l):223-230,1989. Visseren, F.L., K.P. Bouter, M.J. Pon, J.B. Hoekstra, D.W. Erkelens, et al. Patients with diabetes mellitus and atherosclerosis; a role for cytomegalovirus? Diabetes Res. Clin. Pract. 36(1): 49-55, 1997. 195 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287. Wakamatsu, S., T. Banno, M. Tohnai, and T. Oguri. Changes of the plasma levels oft-PA and PAI-l in cerebral infarction induced by the venous occlusion. Japanese Journal of Geriatrics 32(8-9): 566-570, 1995. Wakefield, D.L., A. Lloyd, J. Dwyer, S.Z. Salahuddin, and D.V. Ablashi. HHV-6 and myalgic encephalitis. Lancet 1: 1059, 1988. Waldman, W.J., P.W. Adams, C.G. Orosz, and D. Sedmak. T lymphocyte activation by cytomegalovirus-infected allogeneic cultured human cells. Transplantation 54: 887-896, 1992. Weiss, H.J., and B. Lages. Evidence for tissue factor-dependent activation ofthe classic extrinsic coagulation mechanism in blood obtained from bleeding time wounds. Blood 71: 629-635, 1988. Wiman, B., A. Almquist, O. Sigurdardottir, and T. Lindahl. Plasminogen activator inhibitor 1 (PAI) is bound to vitronectin in plasma. Febs. Letters 242: 125-128, 1988. Wu, K.K., and P. Thiagarajan. Role of endothelium in thrombosis and hemostasis. Annu. Rev. Me. 47: 315-31, 1996. Wu, K.K. Inducible cyclooxygenase and nitric oxide synthase. Adv. Pharmacol. 33:179-207, 1995. Wu, C.A., and JD Shanley. Chronic infection of human umbilical vein endothelial cells by human herpesvirus-6. J. Gen. Virology 79: 1247-1256, 1998. Xu, X.M., K. Ohashi, S.K. Sanduja, K.H. Ruan, L.H. Wang, and K.K. Wu. Enhanced prostacyclin synthesis in endothelial cells by retrovirus-mediated transfer of prostaglandin H synthase cDNA. J. Clin. Invest. 91: 1843-1849, 1993. Yamashiroya, H.M., L. Ghosh, R. Yang, and AL. Robertson Jr. Herpesviridae in the coronary arteries and aorta of young trauma victims. Am. J. Pathol. 130(1): 71-79, 1988. Yoshikawa, T., T. Nakashima, S. Suga, Y. Asano, T. Yazaki, et al. Human herpesvirus-6 DNA in cerebrospinal fluid ofa child with exanthem subitum and meningoencephalitis. Pediatrics. 89(1): 888-890, 1992. 196 288. 289. 290. 291. Zhou, H., M. Leon, M.A. Waclawiw, J.J. Popma, Z.X. Yu, et al. Association between prior CMV infection and the risk of restenosis after coronary atherectomy. N. Engl. J. Med. 335: 624- 630, 1996. Zur, M., R.D. Radcliffe, J. Oberdick, and Y. Nemerson. The dual role of factor VII in blood coagulation: initiation and inhibition of a proteolytic system by a zymogen. J. Biol. Chem. 257: 5623- 5631, 1982. Zwaal, R.F., P. Comfurius, and LL. van Deenen. Membrane asymmetry and blood coagulation. Nature 268(5618): 358-360, 1977. Zwaginga, Jaap J., J.J. Sixma, and PG. de Groot. Activation of endothelial cells induces platelet thrombus formation on their matrix. Arteriosclerosis 10(1): 49-61, 1989. 197 AN NV "11111111111111111111111111“ 293017721519