UBRARY Michigan State " University This is to certify that the thesis entitled The Ultrastructual Characterization of the Antithrombin III Stationary Cofactor Found on Bovine Aortic Endothelitga present 'by Margaret Ellen Hogan has been accepted towards fulfillment of the requirements for M. S . degree in Patholggy .WIW Major professor Date ;' 0-7639 MS U is an Afimmu've Action/Equal Opportunity Institution MSU LIBRARIES .-:—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. THE ULTRASTRUCTUAL CHARACTERIZATION OF THE ANTITHROMBIN III STATIONARY COFACTOR FOUND ON BOVINE AORTIC ENDOTHELIUM. By Margaret Ellen Hogan A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Pathology 1987 ABSTRACT THE ULTRASTRUCTUAL CHARACTERIZATION OF THE ANTITHROMBIN III STATIONARY COFACTOR FOUND ON BOVINE AORTIC ENDOTHELIUM. By Margaret Ellen Hogan A stationary cofactor for Antithrombin III (ATIII) exists on the surface of isolated bovine aortic endothelial cells. This cofactor is involved in the ATIII inactivation of the serine protease, thrombin. Previously, its characterization was limited to kinetic studies of the cell-bound and isolated cofactor, which demonstrated heparin-like acceleration of ATIII inactivation of various serine proteases. This research ultrastructurally characterizes the cofactor to provide morphological correlation with kinetic studies. Endothelial cells grown on microculture beads were examined using 1251 radiolabelled ATIII. The cells showed specific binding of ATIII that was inhibited by: Preincubation of the cells with cold ATIII or a heparin-specific enzyme, and preincubation of labelled ATIII with heparin or thrombin solutions. Results were cosupported by immuno-gold labelling. The cofactor resembled isolated heparin in activity and enzyme specific degradation, however, unlike isolated heparin, it bound and held ATIII to the cell surface. To my parents, for all their love and giving me the chance to explore. To Jill, for lots of support and friendship. To Chris Knight, for the perfect scientific perspective. ii ACKNOWLEDGEMENT I would like to thank the Staff and Friends of the Center for Electron Optics for their help and encouragement. For without their unique blend of talents I would have surely completed this thesis a much less rounded individual. iii TABLE OF CONTENTS List of Tables.................................................. List of Figures................................................. Abbreviations................................................... Introduction.................................................... Literature Review........................................... Materials and Methods........................................... Endothelial Cell Isolation and Culture...................... Microcarrier Bead Preparation............................... Iodination of Antithrombin III.............................. Dowex Column Preparation.................................... Radioassay of GAG-ATIII Binding............................. ATIII Binding to Endothelial Cells.......................... Enzyme Treatment of Endothelial Cells....................... Colloidal Gold Labelling.................................... Electron Microscopy Preparation............................. Transmission Electron Microscopy......................... Scanning Electron Microscopy............................. Autoradiography............................................. Quantitation of ATIII Binding............................... Statistical Analysis........................................ ResaltBOOOOOOOOOOOO0.0..O000......0..OIOIOOOOOOOOOOOOOO0.0.0.... DiscussionO0.0.COO...OO0.0...0..O0.0...OOOOOOOOOOOOO0.0.0.000... S‘lmarYOOOOI.0.0...0.0.0..0...OOOOCOOOOOOOOCOOOO0.00.00.00.00... Appendix A. Endothelial Cell Isolation and Culture............. Appendix B. Formulations....................................... Appendix C. Statistical Analysis of Data....................... Bibliography.................................................... iv Page v vi vii 1 5 14 14 14 15 15 15 16 17 17 18 19 19 20 20 21 22 28 34 35 42 43 51 Table Table Table 1. 2. 3. Table 4. Table LIST OF TABLES Radioassay of ATIII Binding to GAGs..................... 22 ATIII Binding Measured by LM Autoradiography............ 23 Autoradiography of Enzyme Treatment..................... 25 Count and Weight Conversions of ATIII Binding Study..... 43 Count and Weight Conversions of Enzyme Study............ 48 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure LIST OF FIGURES Structure of Heparin.................................... ATIII Interaction With Heparin.......................... Interaction of Heparin, ATIII and Thrombin.............. The Endothelial Cell Role in Maintaining Blood Fluidity. Thrombin Involvement in Hemostasis and Thrombosis....... Enzyme Treatment Flowchart.............................. Common Sample Preparation for TEM and SEM............... ATIII Binding Autoradiography........................... Gold Labelling of ATIII Binding......................... SEM of Gold Labelling of ATIII Binding.................. Microcarrier Cultures - TEM............................. Enzymatic Cleavage Points of Heparinase and Heparatinase Isolated Aorta Preparation.............................. Digestion and Removal of Cells.......................... Collection of Cells..................................... Endothelial Cell Tissue Culture......................... vi 12 17 18 24 24 27 29 31 36 37 38 41 ADP AMP Anti-AT ATIII BAT BEI BEEM CS ddHZO DME DS EDTA EM FBS GAG HA HAT HCl HEPES Hep Sulf HS PA PE PBS PGI PHG RT SEI SEM TEM Th tPA vWF 2 LIST OF ABBREVIATIONS Adenosine diphosphate Adenosine monophosphate Antibodies against ATIII Antithrombin III Bovine ATIII Backseatter electron image Better Equipment for Electron Microscopy (vendor) Chondroitin sulfate Double distilled water Dulbecco's modification of Eagle's medium Dermatan sulfate Ethylenediamine tetraacetic acid Electron microscopy Foetal bovine serum Glycosaminoglycan Hyaluronic acid Human ATIII Hydrochloric acid Hydroxyethylpiperazine ethane sulphonic acid Heparan sulfate Heparan sulfate Plasminogen activator Phosphate buffer Phosphate buffered saline Prostacyclin Phosphate-HEPES-Glucose buffer Room temperature Secondary electron image Scanning electron microscopy Transmission electron microscopy Thrombin tissue Plasminogen activator vonWillebrand Factor vii INTRODUCTION Endothelial cell involvement in maintaining blood fluidity encompasses many actions, some passive, some active. One endothelial cell characteristic, previously described and tested here, is that of providing a membrane surface instrumental in the Antithrombin III (ATIII) inactivation of circulating blood coagulation proteins. Though this compound has been described to be heparin-like in nature, no substantial proof has been given of its true identity, or of its relative concentration in the body. This nonthrombogenic function of the endothelium links the vascular wall to yet another facet in the control of unwanted thrombosis. Heparin is not naturally found in the circulating blood in appreciable amounts. Yet, this endogenous glycosaminoglycan (GAG) (once concentrated) has been a key theraputic anticoagulant for over thirty years. Heparin is a linear, polymetric molecule that has an average molecular weight of 10,000 to 15,000, but ranges from 3,000 to 45,000. It is composed of repeating units of glucosamine and one of two uronic acids (iduronic or glucuronic acid) both in pyranose form (Johnson and Mulloy, 1976) (Figure 1). Its primary mode of action is as a catalyst for Antithrombin III (ATIII), a circulating serine protease inhibitor. Of the coagulation proteins affected, thrombin inactivation by ATIII is quantitatively the most important. Heparin-mediated inactivation of the clotting factors include two 1 2 mechanisms. One involves its specific binding to ATIII through a critical tryptophan residue (Blackburn, et a1. 1984) contained in a lysine block on the ATIII molecule. This binding produces a conformational change in the ATIII that makes its reactive site more accessible to the active center of the activated clotting factors (Lindahl and Hook, 1978) (Figure 2). CH e 600' e 20‘, o chfloo 00- ca 3 on 0 on owe o 0" o mu a: mu at am Figure 1. Structure of Heparin. Two forms of the repeating units of heparin. (A) Glucuronic acid and (B) Iduronic acid, the only difference between the two uronic acids is the location of the carboxyl group. Areas that can be sulfated are indicated by an asterisk (*). (After Johnson and Mulloy, 1976). SERINE ATIII MASS (3+ 8“°"=CE) CB + U‘V'Z HEPARIN a? (‘b + a FAST ; m Figure 2. ATIII Interaction With Heparin. Acceleration of ATIII inhibition of a serine protease via a conformational change in ATIII by heparin binding. The ATIII binding site of heparin as shown by ("'), is specific. (After Lindahl and Hook, 1978) The second mechanism is the_ non-specific binding of heparin to circulating plasma proteins. This binding is due to heparin's high negative charge density. It is this characteristic that is often exploited in the purification of plasma factors IX, XI and thrombin. In the case of thrombin, heparin helps bring the ATIII molecule in closer proximity to the thrombin molecule, therefore facilitating the inactivation of thrombin by ATIII (Figure 3). Tum I .\¢s\\\\\\\\\\\\\ 1 fl 9 MEPAR iN ATIII lion—Specific Iinding Binding sno sn- L, __JI§§§S§§§§!T1""‘ l J "BEAR"! Figure 3. Interaction of Heparin, ATIII and Thrombin. Illustration of heparin-mediated ATIII inactivation of thrombin. (A) ATIII binds to a specific site on heparin, while thrombin binds to heparin non-specificly. The resulting closeness of the two molecules allow binding. (B) After binding, heparin is released from the ATIII-Thrombin complex. The anionic density of heparin is an important factor in its anticoagulant activity. It has been shown that with a decrease in sulphation there is a related decrease in anticoagulant activity (Hurst, et al. 1979; Ayotte, et a1. 1981; Riesenfield, et al. 1981). Also, the molecular weight of heparin has been shown to be related to anticoagulant activity. The larger molecules are found to promote a stronger inhibition of thrombin by antithrombin III (Andersson, et a1. 1979; Laurent, et al. 1978). The ATIII-thrombin interaction is kinetically very slow. In the presence of heparin, this reaction is accelerated 200-fold. Recently a heparin-like cofactor for ATIII was observed to be attached to the surface of isolated endothelial cells. It was shown to have ATIII binding capabilities as well as be involved in the ATIII inactivation of thrombin (Dryjski, et al., 1982, 1983). It was this cofactor that was examined in this current research using electron microscopy via the binding of ATIII. My characterization involved three primary objectives: 1) the 4 binding characteristics of the cofactor, 2) the determination of the most likely GAG to be responsible for the activity, and 3) the localization / quantitation of the cofactor on the endothelial cell. These three characterizations were examined as follows. Hatton, et al. (1978), demonstrated the binding characteristics of various endothelial GAGs bound to a substrate. One of their observations was the selective binding of heparin to ATIII. This selective binding was utilized in my experiments to tag the stationary cofactor. In this research, ATIII binding was quantitated on the endothelial cell, and compared with binding of ATIII complexed with various molecules (GAGs and thrombin). To further characterize the cofactor, I employed treatment of the endothelium with various GAG-specific enzymes. The enzymes heparinase and heparitinase were used. Heparinase is specific for heparin and heparitins C and D, while heparitinase is specific for heparitins A and B (heparan sulfate). Heparitinase does not digest heparin, heparitin C or D. The objective of the research was to distinguish between the three most likely GAGs involved with the cofactor activity (heparin, dermatan sulfate and heparan sulfate). Localization of the cofactor was approached with the use of electron microscopy (EM) combined with immunogold labelling. The previously described ATIII binding experiments were designed to give some insight to the actual location of the cofactor while providing the information for the binding study. The examination of the endothelial cell - coagulation protein interaction through electron microscopic methods has not previously been widely utilized. The use of EM to demonstrate and localize this 5 interaction ultrastructurally made this approach invaluable to correlate previously examined kinetic experiments. The methods used in my study, autoradiography and immunogold cytochemistry, were used to probe the surface as well as the interior of the endothelial cell. In this way a number of characteristics were examined. The most important of these being: 1) the quantitation of binding sites per cell, 2) the condition and maturity of the cells tested, and 3) the binding characteristics of the cofactor itself. Literature Review The vascular endothelium plays a significant role in maintaining the fluidity of circulating blood. This ability is made possible by the varied functions of the endothelial cell which include; 1) formation of a relative, mechanical barrier between the blood components and the sub-endothelial matrix; 2) the synthesis or metabolism of mediators that regulate the interaction between the blood components and the vessel wall; 3) control of vascular repair through cell contraction, cell migration and proliferation, and thrombolysis; and 4) the maintenance of thromboresistance (Figure 4). ADP l L Nail2 AMP + odenosine WIF iibronoctin tPA GAG! Protein C lnacflvo Tiwombin g Protein Ca Manic-i Barrier Figure 4. The Endothelial Cell Role in Maintaining Blood Fluidity. 6 The vascular endothelium is a single layer of cells that line all blood vessels in the body. They provide a physical barrier between the pro-coagulant factors in the blood (platelets, coagulation proteins) and the subcellular matrix. This matrix, when exposed, causes the induction of the hemostatic plug through direct platelet interaction and / or the activation of the coagulation cascade. Therefore, under physiologic conditions, the endothelium inhibits hemostasis and thrombosis, but when injured, promotes these two processes (Gimbrone, 1981). Mediators synthesized or metabolized by the endothelium are primarily involved in platelet interaction, though some are active in mechanical regulation of vessel tone. The most important of these synthesized compounds are von Willebrand factor (vWF), fibronectin, and the glycosaminoglycans (GAGs). The von Willebrand factor is synthesized and released (Jaffe, et al. 1974) both into the circulation and subendothelially, where it is absorbed to exposed collagen (Sakariassen, et al., 1979). Following vascular injury, the role of vWF in hemostasis is to act as a cofactor in the adhesion of platelets to exposed subendothelial collagen, thereby promoting clot formation. Fibronectin synthesis and secretion is towards the subcellular matrix (Yamada and Olden, 1978; Mosher, et a1. 1982) where it acts as a substrate for factor XIIIa (the fibrin stabilization factor) following vessel injury. This interaction causes crosslinking between fibronectin, fibrin or collagen (Masher, et al., 1979), contributing to a stable, hemostatic plug. Endothelial production of GAGs primarily plays a passive role in the cell's non-thrombogenic nature. The proteoglycans found in greatest abundance are, in increasing 7 concentration, hyaluronic acid, chondroitin sulfate, dermatan sulfate (chondroitin sulfate B) and heparan sulfate (Buonassisi, 1973; Gardais, et al., 1973). Of those listed, only the sulphated GAGs dermatan sulfate and heparan sulfate have been shown to have appreciable anticoagulant activity (Hook, et al., 1984). Even this activity, however, requires approximately 70-times the concentration of each to equal that of heparin, a widely used anticoagulant (Thilo, et al. 1983). Their activity is believed to mimic the action of heparin. Heparin is a natural, heterogenous GAG produced primarily in the mast cell, and is found circulating in the blood in only minute quantities. One of the anticoagulant effects of heparin is through its interaction with ATIII, a circulating anticoagulant protein. It catalyzes the rate at which ATIII neutralizes the proteolytic activities of several activated clotting factors in the intrinsic and common coagulation pathways (Bjornsson and Greenberg, 1985). Heparin can also have a direct effect on blood factors through electostatic binding that interfere with enzymes' procoagulant activities. It is believed that some of heparin's activity occurs while absorbed to the surface of the endothelium (Barzu, et al., 1985). It is in this relationship that the anticoagulant characteristics of the secreted GAGs are related to heparin, and, therefore, provide a non-thrombogenic function of the endothelial cell. Upon injury, the endothelium undergoes a number of changes to resolve the damage and limit hemorrhage. These changes include endothelial cell contraction, cell migration and proliferation, and thrombolysis. With vascular injury there is a brief period of vasoconstriction, which in small vessels serves to reduce blood flow. 8 The formation of a hemostatic plug, via platelet interaction and blood coagulation factors then covers the site of injury. Migration and regeneration of the endothelial layer follows fibrin clot contraction, therby renewing the endothelial layer. At the same time, the hemostatic plug is being removed through fibrinolytic substances (tissue plasminogen activator (tPA)) secreted from the endothelial cell. The outcome is a repaired, non-thrombogenic cell layer and the removal of the hemostatic plug (Robbins, et al. 1984; Ogston, 1983). Another function of the endothelium is that of thromboresistance. This characteristic involves both passive and active mechanisms. As stated before, the endothelial GAGs provide a surface that is passively non-thrombogenic. Active thromboresistance by the endothelium is maintained through several mechanisms, including the following: 1) synthesis and release of prostacyclin; 2) secretion of tPA; 3) degradation of ADP by membrane-bound ADPase; 4) uptake and degradation . of vasoactive amines; 5) contribution of a cofactor (thrombomodulin) in the thrombin-dependant activation of Protein C; and 6) uptake, inactivation and clearance of thrombin (Thompson and Harker, 1983). Prostacyclin (PGIZ) is a very potent inhibitor of platelet aggregation, it has also been shown to be a potent stimulator of cAMP accumulation (Hopkins and Gorman, 1981). Prostacyclin is produced through membrane arachadonic acid-ester conversion by first, cyclo-oxygenase then by prostacyclin synthetase. The resulting prostaglandin has at least two actions. One is to dilate vessels locally (therefore increase blood flow), and the other is to increase intra-platelet concentrations of cAMP and so depress platelet aggregation. PGI2 is the most active platelet aggregation inhibitor, 9 and like most highly active substances, is short-lived in the system. A number of different forms of plasminogen activator are synthesized and secreted from the endothelium, a fibrin-dependant (tPA) and fibrin-independent (urokinase-like) PA activity. All which show species, localization and activity differences (Todd, 1959; Lang, 1981; Levin and Loskutoff, 1982). Their function is to cleave plasminogen into the active, fibrinolytic enzyme, plasmin. Plasmin is not normally found in the blood, and requires the conversion of the zymogen plasminogen, via some activating factor, to become the active enzyme. The half-life of plasmin in the body was calculated to be 2.5 minutes (Matsuo, 1982). Plasmin acts on the insoluble fibrin of the hemostatic plug, converting it to a variety of soluble degradation products. Secretion of tPA by the endothelial cells has been linked to thrombin stimulation of the endothelium (Levin, et al., 1984), postulated to be stimulated by catacholamines (Cash, 1978), and can be stimulated by a number of drugs (Davidson, et al., 1972; Nilsson, 1978). ADP regulation through endothelial ecto-ADPase is involved in the control of ADP-induced platelet aggregation (Cooper, et a1. 1979). ADP is one of the secreted constituents of the platelet's dense granules, which is released upon activation. Degrading ADP by endothelial cell membrane-bound ADPases causes the production of AMP and adenosine, both potent inhibitors of platelet aggregation (Pearson, et a1. 1978). This leads to a negative feedback system that helps maintain blood fluidity. The removal of vasoactive amines by the endothelium maintains blood fluidity. The vasoactive amines are released by platelets and include S-hydroxytryptamine (serotonin), histamine and epinephrine (Robbins, et al., 1984). Serotonin and histamine have action on smooth muscle and 11 are believed to be involved in blood vessel constriction. Epinephrine has been shown to increase the effect of ADP on platelet aggregation (Ardlie, et al., 1966). Thrombomodulin is a receptor on the endothelial cell that binds thrombin, a serine protease. This binding results in a 20,000-fold increase in the conversion of Protein C to active Protein Ca (Esmon and Owen, 1981). Thus, this endothelial cell receptor has the ability to accelerate the rate of thrombin-dependant Protein C conversion. This in turn allows the active Protein C to inhibit the conversion of the zymogens, Factors V and VIII to their active forms, both of which are important as cofactors in the coagulation cascade (Rosenberg and Rosenberg, 1984). Another endothelial involvement in Protein C mediation is the synthesis and release of Protein S (Stern, et al., 1986). Protein S is a regulatory plasma protein, which is an essential portion of the Protein C anticoagulant pathway. It acts as a non-enzymatic cofactor which promotes binding of the activated Protein C to membrane surfaces. Once bound to a phospholipid surface, activated Protein C can effectively exert its anticoagulant function (walker, 1984). Thrombin circulates in the blood as the zymogen, prothrombin. Upon cleavage (of a single arginyl-isoleucine bond), the zymogen undergoes a change permitting the expression of enzymatic activity. Thrombin has several functions in thrombosis and hemostasis (Figure 5), and therefore, thrombin clearance from the circulation is essential to maintain blood fluidity (Fenton, et a1. 1977). Thrombin is modulated by the serine protease inhibitor ATIII. Antithrombin III forms a 1:1 stoichiometric complex with thrombin 12 rendering it enzymatically inactive (Abildgaard, 1969). Complex formation is greatly enhanced in the presence of heparin, resulting in a more rapid neutralization of thrombin. [man "39:53 fi§:€fi\\}“!%::fi 5:» m nuns gggéi’dfinonflfin "”'"'“‘“‘“Q 53 N m Q m an- Thu hm plush 2215 RE! ”'7' sffnfibfl 3r1- Figure 5. Thrombin Involement in Hemostasis and Thrombosis. Sites of thrombin activity are shown by the asterisk (*). Taken from Manual of Hemostasis and Thrombosis. Thompson, et al., 1983. The endothelium has been noted for its ability to bind and inactivate thrombin (Dryjski, et al. 1983). It has been shown that the endothelium quickly removes thrombin from the circulation in rabbit (Lollar and Owen, 1980), bovine (Busch, et al., 1982) and human (Awbrey, et al., 1979) endothelial cell cultures. The identity of this thrombin receptor was attributed to two, possibly different, sites. The first, was thrombomodulin thrombin binding (Lollar, et a1. 1982), and the other, was a specific thrombin receptor that upon binding inactivated the thrombin (Lollar and Owen, 1980; Busch, et a1. 1982). The distinction was made when thrombin—ATIII complexes were produced upon contact with the endothelium (Busch, et al., 1982). This work was substantiated with perfusion studies by Marcum and coworkers in 1983 and 1984. What was found was a heparin-like, stationary cofactor for circulating ATIII. When ATIII bound to the surface cofactor, thrombin 13 binding was greatly increased and resulted in the liberation of an inactive thrombin-ATIII complex. Antagonists of the ATIII cofactor activity of heparin significantly reduced the capacity of the preparation to inhibit thrombin (Busch and Owen, 1982). The possible presence of heparin attached to the vascular endothelium poses many questions. The objectives of this research was to distinguish the cofactor from the most likely GAGs involved in the activity, and to determine the nature of the cofactor in regards to its binding characteristics and the endothelial response to the cofactor manipulation. MATERIALS AND METHODS Endothelial Cell Isolation and Culture: Bovine aorta were obtained at a slaughter house, and immersed in ice-cold sterile Dulbecco's Modified Eagle Medium (DME) to transport. The cell isolation was carried out in a laminar flow hood using the following modifications of previously described isolation procedures (Booyse, et al., 1975; Gimbrone, 1976; Huey, 1985). The aortic endothelial cells were released from the intima by digesting with a solution of collagenase (36U/ml) dissolved in sterile DME at 370 C. The resulting cell suspensions were seeded into 100mm tissue culture plates and grown in a humidified, 370 C incubator with 10% C02. Once the primary endothelial cell cultures reached confluency, they were washed briefly in sterile PBS, and removed from the plates using the splitting buffer, 0.02% EDTA, 0.5% BSA in Phosphate-HEPES-Glucose (PHG) buffer (5 minutes at 37 o C). The cells were pelleted at 1800 rpm for 5 minutes. The splitting buffer was removed, and the cells were divided into 100mm petri plates (no tissue culture surface) which contained 1.5ml of prepared microcarrier beads (Cytodex 1, Pharmacia). Media was changed every other day until the cells were confluent on the beads. See Appendix A for detailed protocol. Microcarrier Bead Preparation: The microcarriers (Cytodex 1) were prepared by hydrating lgm of beads in 50mls of 0.22 Diffco gelatin in double distilled water (ddHZO), and then autoclaving the head 14 15 slurry at 1200 C, 15psi for 20 minutes (liquid cycle). The sterile beads were then distributed to petri plates in a concentration of 1.5ml of bead slurry for every 30mls of culture media - providing approximately 200,000 beads/plate and a surface area of 180cm2 (Hogan, et al., 1987). Iodination of Antithrombin III: Iodination of ATIII was by the Chloramine T method (Greenwood, et al., 1980). One mCi 1251 (ICN Biomedicals, Inc. 456 mCi/ml) was added to ATIII (5mgs/1.5mls PBS) and swirled. 250ul of Chloramine T (1mg/ml PBS) was then added dropwise to the protein mixture and allowed to react for 30 minutes on ice. To terminate the reaction, 250ul sodium metabisulfite (1mg/ml PBS) was added and the solution swirled. The calculated specific activities were, 1.3 x 109 cpm/mg for Bovine ATIII and 6.1 x 109 cpm/mg for Human ATIII. To remove unbound 1251, the mixture (2.0mls) was passed through a column packed with Dowex 1, prepared as follows. Dowex Column Preparation: Dowex 1 was suspended in 50:50 ethanol:acetone (45gm resin/100mls) for 1 hour. The suspension was filtered dry and resuspended in ddHZO and mixed with a magnetic stirrer. Solid sodium acetate was added to give a final concentration of 1N. The Dowex was filtered dry, and the cake was placed into ddHZO. Concentrated hydrochloric acid was added to obtain a final concentration of 3M (approximately a 1:4 dilution of the HCl stock). The resin was again filtered and washed with 0.5M HCl, followed with ddHZO until the filtrate was neutral. The Dowex was then in the chloride ion form and ready to use. Radioassay of GAG-ATIII Binding: The GAGs, heparan sulfate (HS, Miles Scientific Inc.), dermatan sulfate (DS, Sigma #C4259), hyaluronic 16 acid (HA, Sigma #H1751) and chondroitin sulfate (CS, Sigma #C8529) were prepared to a final concentration of 0.1mg/ml in Phosphate Buffer (PB). Heparin (H, Upjohn Co.) and thrombin (Th, DiaTech Inc. #303061) were diluted to 0.1U/ml in PB. Rabbit antibodies to ATIII (anti-ATIII) and purified ATIII (bovine (BAT, Sigma #A9141) and human (HAT, USDA isolate)) were diluted to 0.1mg/ml in PB. PB alone was used as a control. 50ul of each of the above solutions were allowed to incubate in a 96 well microtiter assay plate for 1 hour at room temperature. Each well was washed three times with PB (GAG binding was approximated by staining test wells with 32 alcian blue), and was blocked with 3% BSA followed by a final wash in PB. 100,000 cpm/well of 1251 -ATIII (bovine or human) in PB was added in a volume of 50ul, and was incubated 30 minutes at room temperature. Plates were washed three times with PB, dried, the wells removed and counted in a gamma counter. ATIII Binding to Endothelial Cells: Into 1500ul microfuge tubes approximately 0.5mls of bead/cell suspension was placed. Each sample was washed twice with PHG to remove residual media. The buffer was removed, and half of the samples were put on ice and the other half left at room temperature. 300ul of each of the test solutions (see below) were added to the tubes, and all were incubated 30 minutes. The test solutions were removed, the samples washed twice with PHG and fixed in 2% glutaraldehyde in PB. Treatments (incubated at both 4 and 270 C) Volume Labelled ATIII alone (diluted with PB) 150u1 each Labelled ATIII + Heparin (0.1U/m1) 150u1 each Labelled ATIII + Heparan Sulfate (0.1mg/ml) 150ul each Labelled ATIII + Thrombin (0.1U/m1) 150ul each ATIII (pre-incubation, 30 minutes) then labelled ATIII 300ul each PB 300ul 17 Enzyme Treatment of Endothelial Cells: 0.5mls of the cell-covered beads were aliquoted into each of six 1500ul microfuge tubes. The cells were then.washed two times with PHG to remove the media, the supernatants removed, and were incubated with the enzymes (Heparinase, 5U/reaction/500ul; Heparitinase, 5U/reaction/500ul; Miles Scientific, Inc.). After 30 minutes at 37° C, the incubation solution was removed, and the cells were washed two times with PHG. Dual incubations with radio-labelled ATIII were done at both room temperature and on ice. Again, after a 30 minute incubation the reaction mixture was taken off, the cells washed twice with PHG, and then were prepared for microscopy (Figure 6.). 125 Enzyme Treatmeng I—ATIII Treatment (30 minutes, 37 C) (30 minutes) Heparitinase room temperature (270 C) Heparitinase 40 C Heparinase room temperature (270 C) Heparinase 4° C PHG room temperature (270 C) PHG 4° c Endothelial Cells Heparinase Heparitinase RT ice RT ice ATIII ATIII ATIII ATIII tag tag tag tag Figure 6. Enzyme Treatment Flowchart. Colloidal Gold Labelling: Microcarrier beads cultured to confluency, were removed from the growth medium and pooled. The 18 suspension was washed with warm (370 C) PHG buffer, and aliquoted into individual reaction tubes. Supernatants were removed and discarded after the cells settled, and half of the samples were incubated with the ATIII reaction mixtures (as used in the ATIII Binding to Endothelial Cells Study), and the other half prefixed with 0.5% 4 paraformaldehyde in PHG buffer (15 minutes, 37°C.), then reacted with the identical solutions. After the 30 minute incubation, the cells were washed then fixed in 0.5% paraformaldehyde (15 minutes, 270 C), washed again and blocked with 3% BSA. Anti-ATIII prepared in rabbits, was then added to the cell mixtures, and reacted 30 minutes at 370 C. After washing twice with PHG, Protein A-Gold (1:4 dilution) prepared according to Bendayan (1984a) was incubated with the cells (30 minutes, 370 C). The cells were again washed to remove unbound gold, and fixed in paraformaldehydelglutaraldehyde, and prepared for electron microscopy (Bendayan, 1984b). Controls were prepared by leaving out one reagent (ATIII, anti-ATIII or both) and replacing it with PHG. Procedures remained the same. Electron Microscopy Preparation: A common procedure was followed for both TEM and SEM preparation of the cell cultures (Figure 7). Endothelial Cells Fixation wash Dehydration /\ Figure 7. Common Sample Preparation for TEM and SEM. SEM prep TEM prep Cell cultures were fixed in a 0.5% paraformaldehyde/1.0% glutaraldehyde 19 solution prepared in PHG buffer (1 hour, 40 C). Post-fixation was in 1.0% osmium tetroxide in PHG for 2 hours at room temperature. After fixation the samples were washed repeatedly with 50 mM HEPES buffer, then double distilled water to remove any glucose in the samples. Specimens were dehydrated through graded ethanol, then divided into two portions, one for SEM and the other for TEM preparation. Transmission Electron Microscopy: Dehydrated samples were infiltrated with Spurrs-Mollenhauer resin (Klomparens, et al., 1986), after the following schedule: Alcohol : Acetone 3:1 1 hour Alcohol : Acetone 1:3 1 hour Acetone 100% 1 hour Acetone : Resin 3:1 3 hours Acetone : Resin 1:1 3 hours Acetone : Resin 1:3 3 hours Resin 100% overnight Samples in 100% resin were placed under mild vacuum (20 psi) 1 hour to insure complete infiltration/exchange of the bead. The following day the cell-covered beads were added to the top of filled BEEM (Better Equipment for Electron Microscopy) capsules, and allowed to sink to the tip. Again the specimens were placed under vacuum (20 psi, 15 minutes), then polymerized in a 600 C oven for 48 hours. Blocks were sectioned with a diamond knife, stained with uranyl acetate and examined with a JEOL 100 CXII transmission electron microscope operated at 80 kV. Scanning Electon Microscopy: Dehydrated samples were placed in porous baskets and critical point dried. The dried samples were then attached to aluminum stubs with sticky tape, and coated with carbon or gold. Carbon evaporation was done with a Ladd evaporator with a rotary stage (5nm of carbon), and gold was applied with a Film-Vac, Inc. 20 sputter coater (15nm of gold). Secondary electron images (SEI) and backscattered electron images (BEI) were both run at 15kV and were performed on the same JEOL 350E scanning electron microscope. Autoradiography: TEM- Dark gold sections (100nm) of labelled cells were cut and transfered to plastic-coated (0.52 collodion in amyl acetate) glass slides. The slides were then carbon coated (approx. 4nm) using a Ladd vacuum evaporator. Coated slides were dipped in Kodak emulsion NTB 2, diluted to obtain a gold reflection over the sections. Specimens were allowed to dry, and then stored, desiccated at 40 C for exposure. Light microscopy- 2 micrometer sections were collected on clean glass slides and then dipped into undiluted Kodak emulsion NTB 2. When dry, the slides were stored as were the slides with the thin sections. Exposed emulsion development-. After various exposure times, glass slides with the exposed emulsion were allowed to reach room temperature (from 40 C), were developed in half-strength Kodak D19, for 2 minutes, fixed 2-3 minutes in full strength fixer, washed for 10 minutes and dried (Budd, 1971; Klomparens, et al., 1986). For LM examination, slides were coverslipped and examined. For TEM, the collodion was floated off the slides using hydrofluoric acid. Grids were then placed on the area that contained sections, picked up, stained with uranyl acetate and examined. Quantitation of ATIII Binding: Unstained autoradiographic thick sections (1 micrometer) were photographed and enlarged to a final magnification of 1000K. The number of exposed silver grains per cell section was determined, then each cell was cut out, pooled by treatment and weighed (to the 10-4 place) on a Mettler H15 balance. All weights 21 were compared to known standards (pre-determined by micrographing, enlarging (to 1000K), cutting out and weighing a grid of known area). All micrographs were printed on Kodak Polycontrast RCII, medium weight paper. Counts per cell slice was converted to counts per whole cell (based on an average endothelial cell size of 10um3). Statistical Analyisis: Data was analized by using a Random Analysis of Variance, where the variance was tested for heterogeneity with a F-max test. Multiple comparisons were made using Dunnett's test for comparison of all treatments to a control, and Bon Ferroni's test for the comparison of treatments. See Appendix C for complete analysis. RESULTS The ATIII binding to substrate-attached GAGs had approximately the same order of affinity in both bovine and human ATIII, with only a change in order between Hyaluronic Acid (HA) and Dermatan Sulfate (DS) (Table 1). Table 1. Radioassay of ATIII Binding to GAGs Human ATIII Bovine ATIII (mean cpm) (mean cpm) Heparan Sulfate 5601.3 2633.3 Dermatan Sulfate 5834.6 4538.6 Hyaluronic Acid 7971.6 4014.6 Chondroitin Sulfate 9118.6 4605.3 Heparin 9618.6 5060.0 Thrombin 12800.0 6493.3 Human ATIII 5663.0 2264.0 Bovine ATIII 5491.3 2803.6 Load Count per 50ul (cpm) 55895 73542 Note: Mean cpm based on N=6 The rank of the GAGs in regards to their ability to bind ATIII was as follows: Bovine ATIII: H > CS > DS > HA > HS Human ATIII: H > CS > HA > DS > HS With both forms of ATIII, heparin bound to a substrate demonstrated the greatest ability to bind ATIII, where heparan sulfate was closest to a negative control in binding ATIII. 22 23 Attempts to localize radiolabelled ATIII binding to the surface of the endothelium using decreased temperature (4°C) were unsuccessful. When comparing binding experiments done at RT and 4°C, no substantial difference was found. Essentially, all labelling was internalized. Autoradiographs for TEM were sparse in silver grains (Figure 8A), therefore quantitation of ATIII binding was made at the light microscope level (Figure 8B). The EM was then used to identify areas within the cell that tended to show the highest levels of binding. These areas were dispersed with no apparent organelle containment. ATIII binding to endothelial cells measured by autoradiographic means showed a significant difference via Dunnett's test between the treatments and the control. When the treatments were compared to their ability to block ATIII binding to the cells, only ATIII pre-bound to thrombin and ATIII pre-bound to heparin showed any significant reduction in binding (via Bon Ferroni's test) to the cells (Table 2). The amount of thrombin reacted with the ATIII was far below saturation levels, this was to prevent precipitation of the complexed molecules following reaction. Table 2. ATIII Binding Measured by LM Autoradiography Binding at 4°C Treatment mean counts / cell variance Control 16.92 2.793 Heparin 23.45 8.789 Thrombin 24.41 14.933 Heparan Sulfate 34.52 8.544 ATIII 37.23 13.573 Binding at RT Treatment mean counts / cell variance Control 16.86 1.361 24 Figure 8. ATIII Binding Autoradiography. Demonstration of the radiolabelled ATIII bound to the endothelial cell. A) TEM of three labelled cells. Nuclei (Nu) and mitocondria (m) visible, as well as the exposed silver grains (arrows). Stained with uranyl acetate only. Bar ' 1 micrometer. B) LM of the autoradiographs. B1 is a stained section (toluidine blue) to show nuclei (Nu) of the cells, and the bead (B). BZ is unstained autoradiograph section showing bead (B) surrounded by cells. Both bars = 10 micrometers. Figure 9. Gold Labelling of ATIII Binding. TEM demonstration of gold labelling. A) Single cell with nucleus (Nu) present, showing common sparse labelling of gold (arrow). Portion of microculture bead present (asterisk). Bar = 1 micrometer. B) Enlargement of bead (B) surface showing labelled cell fragment (arrows). Bar ' 1 micrometer. 25 Table 2. ATIII Binding Measured by LM Autoradiography (cont.) Binding at RT Treatment mean counts / cell variance Heparin 22.87 9.345 Thrombin 28.17 12.916 Heparan Sulfate 33.28 7.961 ATIII 36.30 14.938 Notes: See Appendix A for statistical analysis. 2 Mean counts per cell is taken to be equivalent to 10um . Mean counts based on N'10 Enzyme treatment of endothelial cells with heparinase and heparitinase followed by binding of labelled ATIII (Table 3) showed that endothelial cells enzymatically treated with heparinase showed a marked decrease in their ability to bind labelled ATIII. Cells treated with heparitinase, specific for heparan sulfate, showed no significant difference with cells having no pretreatment. Table 3. Autoradiography of Enzyme Treatment Treatment mean counts / cell variance Contol 29.14 5.95 Heparitinase 28.26 5.42 Heparinase 23.64 2.61 Note: See Appendix A for statisical analysis. 2 Mean counts per cell is taken to be equivalent to 10um . Mean counts based on N'10 Pre-embedding gold labelling of the cofactor gave little information at the TEM level. The incidence of the cofactor on the cell surface was infrequent enough that examination produced few, if any, of the gold markers (Figure 9A). Pre-fixation of the cells did prevent internalization of the gold label. This was an improvement over the attempts to reduce internalization with a decrease in 26 temperature. All label found was either on the surface of the cells or within convolutions on the exposed microcarrier bead surface in close proximity to the cell (Figure 9B). At the SEM level, the gold labelled cofactor was evident and easily observed using backscatter electron detection (Figure 10B). When compared to the bead surface (Figures 100 and 10D), specific binding was demonstrated. Gold labelling was not quantified, and was used only as a confirmation of the ATIII binding ability of the endothelial cells. 27 Figure 10. SEM of Gold Labelling of ATIII Binding. Visualization of labelling using BEI. A) SEM of untreated microcarrier cell cultures showing one bead confluent with cells. Bar ' 10 micrometers. B) Secondary (B1) and backscatter (BZ) electron images of labelled cells. Nucleus (Nu) and gold particles (arrows) are evident. Bars ' 5 micrometers. C) BEI of sub-confluent microcarrier bead showing cell specific binding. Nucleus (Nu) can be seen on micrograph as well as on line drawing (D). Exposed bead surfces (asterisks) and gold labeling (small stars) can be seen. Bars = 5 micrometers. DISCUSSION The ultrastructural examination of endothelial cell monolayers posed a difficult, technical problem. Monolayers are difficult to embed for TEM examination, especially if a minimum of cell disruption is desired. For this reason, I established method for the use of endothelial cells grown on sectionable cross-linked dextran beads for EM. These microcarrier beads provided an excellent growth matrix, as well as made correlation between SEM, TEM and LM possible. Another advantage was the ability to pool cell cultures, thus allowing proper population sampling, something that would not have been possible had the cells been grown on tissue culture plates. Figure 11 shows the cells grown on the beads (A), demonstrating a retention of cell morphology (Figure 11B). To begin the characterization of the stationary cofactor, it was necessary to establish some idea as to its general nature. It had previously been described as able to bind ATIII, and with ATIII accelerate the inactivation of thrombin (Busch and Owen, 1982). In addition to this, its close association to the endothelial cell prompted its description as heparin-like. Taking this into consideration, it was then necessary to determine what possible GAGs, common to the endothelium, were possibly responsible for this cofactor activity. Hatton, et al. 1978, ran a series of experiments that demonstrated 28 29 Figure 11. Microcarrier Cultures - TEM. Micrographs of cell grown on cross-linked dextran beads. A) TEM of cells on bead (B). Four cells can be seen. Nuclei (Nu) are distinct. Bar - 10 micrometers. B) Enlargement of endothelial cells showing intact morphology. Nuclei (Nu) have retained bilayer membranes. Cell-to-cell contact is allowed (arrows), and cell maintain fine cytoplasmic projections (asterisks). Bar - 1 micrometer. 30 ATIII-heparin specific binding, to the exclusion of the other GAGs tested. I repeated these experiments, with some modifications, and came to the same conclusions as Hatton's group. The ATIII binding experiments clearly showed that, with the exception of heparin, binding of the inhibitor to the GAGs was negligible. These results allowed the use of ATIII to probe the surface of the endothelial cell. The next step was to determine if ATIII binding to the cofactor could be inhibited. It has been described that the most likely GAG responsible for the cofactor activity is heparan sulfate (Buonassisi and Colburn, 1982). They report, using 358 labelling, finding protein - bound carbohydrate chains both at the cell surface and in the supernatant growth medium of rabbit endothelial cultures. They assume the identity of the GAG to be heparan sulfate via resistance to degradation by chondroitinases AC and ABC. It was for this reason heparan sulfate was included in the bulk of my characterization experiments, even though it showed reduced affinity for ATIII. In the inhibition experiments, radiolabelled ATIII was pre-incubated ‘with various test solutions, then incubated with the endothelial cells to assess the degree of ATIII blockage. Pilot experiments used cold ATIII pre-incubations, followed by cell incubation, then probing with radiolabelled antibodies to ATIII. This raised the question whether the decrease in binding observed, was due to inhibition of ATIII binding or if the antibody was hindered in its binding to ATIII. By eliminating the secondary labelling, a direct measure of binding could be determined. Another problem arose due to the endothelial cell's predilection to endocytosis. This prompted the use of 4°C incubations to try to hinder 31 the endocytosis of label into the cell. Even at this temperature, the cells maintained a level of endocytosis that essentually resulted in a surface devoid of label. Therefore, autoradiographic labelling was quantitated in the interior of the cells. The results of the inhibition experiments showed that heparin, thrombin pre-incubation with ATIII could inhibit the binding of ATIII to the endothelial cell surface. Also, incubation of the endothelial cells with unlabbeled ATIII could also inhibit labbelled ATIII binding via cofactor saturatation. This was substantiated with incubations at both 4°C and 27°C (room temperature). Heparan sulfate showed no significant inhibition of ATIII binding. This tends to direct thought to a more heparin-like molecule as opposed to that of heparan sulfate. To further characterize the cofactor and perhaps distinguish between its heparin or heparan sulfate nature, enzymatic digestion of the endothelial cell surface was performed. The enzymes used were specific for heparin and heparan sulfate cleavage. The enzymes, isolated from Flavobacterium heparinum, digest the molecules at different glucosyl linkages (Miles Scientific, Product Profile). This is based on the number of iduronic acids the molecules possess. Heparinase Heparat inane Figure. 12. Enzymatic Cleavage Points of Heparinase and Heparatinase. Arrows indicate sites of cleavage. 32 Heparin is rich in the iduronic acid form, while heparan sulfate is rich in the glucuronic acid form. Therefore, the enzymes are more specific for one or the other. These experiments resulted in a marked decrease in the endothelial cell's ability to bind ATIII after digestion with the heparinase and not heparitinase. This points to a GAG rich in iduronic acid components. The calculation of the number of binding sites per endothelial cell was based on the number of silver grains per one micrometer section in control tissue. This reflects approximately 300 to 350 binding sites, as determined by maximum ATIII binding on a cell 10 um3. This calculation is much lower than that of Marcum, et al., 1986. Their estimation was of 58000 protease inhibitor binding sites per cell for a heparan sulfate-like cofactor on the surface of endothelial cells. This could either be due to the measuring of additional binding sites unmasked by the cofactor isolation, or that the age of the cells in question has a bearing on the material Marcum's group characterized. They used solubilization of the cell surface followed by affinity fractionization to identify a cofactor showing anticoagulant activity, specifically, the acceleration of ATIII inactivation of thrombin. This study employed cloned endothelial cells, that had under gone "less than 70 population doublings” (approximately 2 months old). These cells are essentially different than the endothelial cells used in my studies. Goldsmith, et al., 1984 examined the change in endothelial cell properties and kinetics over time while being grown in tissue culture. They found that there was a gradual decline in prostacyclin release as soon as the first population doubling. Angiotensin converting factor 33 release also dropped off, both signs of a decrease in cellular response. When discussing cultured endothelial cell kinetics, it was found that the cells tended to arrest their growth upon subculturing, to the point that highly subcultured cells could be used for models of in vitro senescence. To parallel Marcum's cloned cell cultures (approximately 2 months old) to the cultures used in my experiments (approximately 2 weeks old) would be innappropriate unless only general comparisons were made. Of interest was Marcum's finding that a high affinity (for ATIII) surface fraction contain the Gch-AMN-3-O-SO3 which represents the ATIII-binding region of heparin, which constitutes a structual marker for heparin. Also, the complexing of ATIII was completely eliminated by pre-treatment of the cells with a purified Flavobacterium heparinase (digests both heparin and heparan sulfate), substantiating the enzyme treatments performed in my experiments. These two observations again suggest a possible sub-population of heparin on the surface of the endothelial cell. Though the cofactor has been shown to be similar to heparin in kinetic studies, one characteristic it does not share with the GAG is its ability to bind ATIII without subsequent release. Heparin only is in contact with ATIII long enough to conformationally change ATIII into a more accessible molecule, this binding does not endure. This is not the case of the surface cofactor. It has been shown to bind ATIII even after repeated washings. These morphological findings, along with kinetic studies from other laboratories, may help in the understanding of the role the endothelial cell plays in the regulation and prevention of unwanted thrombosis. 1) 2) 3) 4) 5) 6) 7) SUMMARY Only light pre-fixation of the cells could prevent endocytosis, as apposed to low temperature incubations Heparin or thrombin incubated with ATIII could inhibit the ATIII's binding to the endothelial cell surface ATIII incubated with heparan sulfate showed no decrease in its ability to bind to the endothelial cell surface Digestion of the endothelial cell surface with heparinase destroyed the ability of the cell to bind ATIII Digestion of the cells with heparitinase showed no significant change in ATIII binding to the cell ATIII binding sites per cell is approximately 300 to 350 at maximun binding (based on one cell - 10 micrometers3) Unlike isolated heparin, this ATIII cofactor binds ATIII and can maintain union even after repeated washings of the cell surface 34 APPENDICES 35 APPENDIX A Endothelial Cell Isolation and Culture This section provides detailed description of the isolation procedure of the bovine aortic endothelial cells and their subsequent tissue culturing. Appendix B provides the formulations of the buffers used in these protocols. A.1. Endothelial Cell Isolation Bovine aorta were obtained at a slaughter house, and immersed in ice-cold, sterile DME for transport. The cell isolation was carried out in a laminar flow hood on a coated, absorbant mat. The aorta was rinsed out with sterile PBS throughout the entire procedure to prevent drying of the endothelium. The external fat and facia was then removed, being careful to leave the intercostal arteries intact (Figure 13A). These arteries were then tied off with suture close to the point of branching (Figures 13B and 13C). The small end of the aorta was then clamped off with a hemostat to produce a bag-like structure. To facilitate hanging the aorta during incubation, either a portion of the brachiocephalic artery can be left attached to the aorta (Figure 14A), or the aorta can be cut near the upper, larger opening providing a means of attaching it to a ring stand. In this instance, a plastic-coated clamp was ethanol sterilized, and the single prong was used to hang the aorta at an appropriate height that allowed free suspension. The suspended aorta was then filled with warm DME to check for leaks. The color of the media provided a means to locate any leaks from sutures or cuts in the tissue. Once the aorta is leak-free, the 36 Figure 13. Isolated Aorta Preparation. A) Excess fat and facia is removed from the vessel. B) Tying off of the intercostal arteries close to the main vessel. C) Trimming of the tied off vellel and the suture cord. 37 Figure 14. Digestion and Removal of Cells. A) Aorta is hung from a ring stand, producing a water-tight, blind tude. B) Warm collagenase mixture is added to the vessel, is covered with foil, and incubated. C) Aorta is drained, refilled with media and shaken to dislodge loosened cells. 38 Figure 15. Collection of Cells. A) After shaking, vessel is cut off at narrow (distal) end. B) and media is decanted into sterile centrifuge tubes. 39 DME was replaced with a collagenase mixture of 36U/ml (CLS, Cooper Biomedical, Lot 46N6783) in 37°C. DME. The aorta was covered with a sterile piece of foil, and was incubated for 20 minutes at 37°C. (Figure 14B). The collagenase mixture was then decanted from the aorta and discarded. The aorta was filled three-quarters full with DME culture media, the large end clamped shut, and was vigorously shaken for about 3 minutes (Figure 14C). The bottom (small end) was cleaned with ethanol, and cut near the clamp to allow decanting of the media into sterile centifugation tubes. Before decantation, the open end was wrapped with a disposable tissue to prevent contamination of the cell suspension from liquid on the surface of the vessel (Figures 15A and 15B). Media was added 4 more times with progressively rigorous agitation and rubbing to dislodge the attached cells. The tubes of media were spun in a swingout rotor at 1800 rpm for 5 minutes. A.2. Endothelial Cell Tissue Culture The resulting pellet was resuspended in fresh culture media and was used to seed 100mm tissue culture plates, that were placed in a humidified incubator at 37°C. and 102 C02. The media was changed at 6 hours and again at 24 hours to remove red blood cells and other debris. After this point, the media was changed every other day until the cells reached confluency (usually 8 to 10 days). Detachment from the primary cell plates was facilitated by a breif wash with sterile, 37°C. PBS, followed by 0.021 EDTA, 0.5% BSA, PHG solution which was allowed to incubate on the plates 5 minutes at 37°C. 40 This incubation should not be exceeded due to possible cell damage by the EDTA. The loosened cells were gently washed off, pelleted as before and split into 100mm bacteriological petri plates in a ratio of 1:3. To each of the plates microculture beads were added, gently swirled, placed in the incubator, and not moved again for at least 18 hours. The bead cultures were handled in the same manner as the primary cultures except extra care was required when changing media as not to draw off the beads with the media (Figure 16). 41 Figure 16. Endothelial Cell Tissue Culture. A) Endothelial cell sheet isolated from the aorta. 0 hours. B) Cell sheet after attachment to culture dish. Notice spreading of cells. 24 hours. C) Sub-confluent culture on tisue culture dish. Some cells still spread, others beginning to round up. 5 days. D) Typical "cobble-stone” appearance of confluent endothelial cultures. 8 days. E) Freashly seeded microcarrier beads. Cells just begining to attach and spread (arrows). 2 days. F) Confluent bead cultures. 8 days. All bars - 50 micrometers. Q ‘ “.O‘ .1: yil"...h\bul . o i n v V! Ho“.(lc",’0 "."ou‘lw m 041,) :5 ..vs ‘0..i0‘. oOfiJi, o n.. [It 5 I .7 i .4.- ‘.. I I” ........ r45; 0. .09!“ .23; Av 9....) no \ a f \ \O'fv- 'aflm .- o .0 . ~ .1 “.4“ . L . .. .. B ”hunk"? .wftr o . . w\.0.ni? .0 4 o . o w...“ .10..;...5. I 42 APPENDIX B Formulations This section provides the formulations of the buffers used in the experimental procedures. DME Culture Media: DME, with 4gm/liter glucose, pH 7.5 10% FBS 20mM HEPES, pH 7.5 2mM Glutamine Penicillin 100000U/liter Streptomycin 100mg/liter PB (Phosphate buffer): 0.1 M monobasic sodium phosphate 0.1 M dibasic sodium phosphate to give a pH of 7.5 PBS (Phosphate buffered saline): 0.8% Sodium chloride 0.1M Phosphate buffer, pH 7.5 PHG (Phosphate-HEPES-Glucose buffer): 0.1 M PBS, pH 7.5 15mM HEPES, pH 7.5 11mM Glucose Splitting Buffer (Endothelial Cell Culture): 0.022 EDTA 0.5% BSA in PHG buffer, approximately 5mls per 100mm plate 43 APPENDIX C Statistical Analysis of Data The purpose of this section is to provide the statistics used in the data supplied in the results section. All of the measurements as well as the formulas are included. C.1. ATIII Binding Study - Autoradiography Table 4. Count and Weight Conversions of ATII Binding Study Treatments Counts Weight Coungs Mean Treatment Vari- (gm) 10um Total ance Heparin, 4°C 131 .5810 25.71 131 .5502 19.65 152 .6278 19.98 114 .5005 18.79 109 .3455 26.04 61 .1867 26.97 140 .4676 24.71 165 .5808 23.45 92 .2953 25.71 23.45 211.01 8.789 Heparan Sulfate, RT 174 .3602 33.87 123 .2706 37.52 74 .1650 37.02 49 .1485 30.23 149 .3844 31.99 222 .4651 36.39 108 .3186 30.98 123 .3370 30.12 176 .5109 31.43 33.28 299.55 7.961 Heparin, RT 187 .6227 24.79 170 .6426 21.84 212 .6614 26.46 84 .3612 19.19 364 1.0829 27.74 205 .7380 22.93 63 .2785 18.67 175 .6755 21.38 22.87 183.00 9.345 Thrombin, RT 136 .5247 25.39 139 .3531 32.49 141 .5288 24.01 135 .4253 26.19 142 .3032 30.66 120 .2832 34.97 141 .3793 30.68 152 .5092 24.64 131 .4703 24.99 119 .3546 27.69 28.17 281.71 12.916 44 Table 4. (cont.) Treatments Counts Weight Count: Mean Treatment Vari- (gm) 10um Total ance Thrombin, 4°C 93 .3581 21.35 87 .2828 25.39 27 .0934 23.86 60 .2298 21.55 123 .2939 34.54 56 .1938 23.85 74 .2834 21.55 82 .3018 22.43 46 .1509 25.16 24.41 219.68 14.933 Heparan Sulfate, 4°C 71 .1462 38.08 182 .4295 34.99 161 .4395 31.24 96 .2457 32.23 107 .2579 34.24 113 .2294 37.66 140 .3654 33.62 118 .3351 31.06 141 .2919 39.87 102 .2611 32.24 34.52 345.23 8.544 cold ATIII, 4°C 89 .4920 14.93 136 .5440 20.63 212 1.0347 16.91 121 .5941 16.81 161 .8726 15.15 139 .7075 16.22 149 .7450 16.51 96 .4329 18.30 64 .2874 18.38 100 .5370 15.38 16.92 169.22 2.793 cold ATIII, RT 98 .3979 21.33 56 .2307 20.03 58 .2337 20.48 101 .4555 18.30 83 .3419 20.04 77 .3211 19.79 96 .4070 19.47 90 .3888 19.11 124 .5332 19.19 129 .5960 16.86 19.46 194.56 1.361 ATIII*, RT 242 .5820 34.32 224 .5031 34.82 275 .7412 30.62 268 .7343 30.12 163 .3660 36.76 78 .1642 39.21 331 .6670 40.96 184 .3735 40.66 211 .4442 39.21 36.30 326.68 14.938 45 Table 4. (cont) Treatments Counts Weight Count: Mean Treatment Vari- (gm) 10um Total ance ATIII*, 4°C 277 .3933 38.04 170 .4097 34.24 121 .1646 30.67 184 .3680 41.27 195 .4407 36.52 166 .2158 43.49 124 .1265 40.45 208 .2517 34.10 104 .1113 38.56 41 .0968 34.96 37.23 372.31 13.573 C0101. F-max test for heterogeneity of variances between treatments: F' largest varience / smallest variance ' 14.938 / 1.361 ' 10.977 ' calculated F-max For F-max tabled: a= total number of groups ' 10 n' smallest number of observations = 8 p' probability level - 0.05 F-max(tabled)' 11.7 No significant heterogeneity of variance within all treatments 0.1.2. ATIII Binding at 4°C. Analysis of Variance: 1. Correction term 2. Sums of Squares (Total) 3. Sums of Squares (Treatment) C'(sum X)2 / total observations -(1317.45)2 / 48 -1735674.5 / 48 ~36159.885 T 88 o'sum(X2) - C '39414.636 - 36159.885 '3254.75 SSTr-sum(T2 / r) - c '38952.781 - 36159.885 '2792.896 46 TB totals of each group r‘ number of replicates ANOVA Table: Source dF SS MS F treatment 4 2792.9 698.2 65.0 error 43 461.9 10.7 total 47 3254.8 -- 4. F- MS(treatment) / MS(error) ' 698.2 / 10.7 - 65.0 = calculated F value 5. F(tabled)' 2.59 at 0.05 probability F(calc) > F(tabled) therefore reject HO' all treatments are the same. Multiple Comparison of Data: Dunnett's test for comparison of all treatments to a control Treatment Mean control 16.92 heparin 23.45 thrombin 24.41 hep sulf 34.52 ATIII 37.23 1. Dunnett's Critical Value Dunnett's t value x square root of (2MS(error) / r) '2.22 x 1.46 .3025 2. Comparison of Means Control:Heparin 6.53 Control:Thrombin 7.49 Control:Heparan Sulfate 17.60 Control:ATIII 20.31 -all show significant difference from control (difference of treatment means > Dunnitt's critical value) Bon Ferroni's test for the comparison of treatments 1. Bon Ferroni's Critical Value Bon Ferroni's t value x square root of (2MS(error) / r) '2.30 x 1.46 .3036 47 2. Comparison of Means ATIII:Heparan Sulfate 2.71 no significant difference ATIII:Heparin 13.78 significant difference C.1.3. ATIII Binding at RT Analysis of Variance: 1. Correction term C'(sum X)2 / rt -(1285.50)2 / 46 -1652510.3 / 4e .35924.136 2. Sums of Squares SSTO'sum(X2) - C (Total) =38160.49 - 35924.136 '2236.35 3. Sums of Squares SSTr'sum(T2 / r) - C (Treatment) B1811.18 ANOVA Table: Source df SS MS F treatment 4 1811.2 452.8 43.66 error 41 425.2 10.3 4. F= MS(treatment) / MS(error) ' 452.8 / 10.4 = 43.66 ' calculated F value 5. F(tabled)' 2.59 at 0.05 probability F(calc) > F(tabled) therefore reject Ho' all the treatments are the same Multiple Comparison of Data: Dunnett's test for comparison of all treatments to a control Treatment Mean control 16.86 heparin 22.87 thrombin 28.17 hep sulf 33.28 ATIII 36.30 1. Dunnett's Critical Value 48 Dunnett's t value x square root of (2MS(error) / r) ”2.22 x 1.46 1“3.25 2. Comparison of Means Control:Heparin 6.01 Control:Thrombin 11.31 Control:Heparan Sulfate 16.42 Control:ATIII 19.44 -all show significant difference from control (difference of treatment means > Dunnett's critical value) Bon Ferroni's test for comparison of treatments 1. Bon Ferroni's Critical Value Bon Ferroni's t value x square root of (2MS(error) / r) '2.30 x 1.46 ll=3.36 2. Comparison of Means ATIII:Heparan Sulfate ATIII:Heparin 3.02 no significant difference 13.43 significant difference C.2. Enzyme Treatment Study - Autoradiography Table 5. Count and Weight Conversions of Enzyme Study Treatments Counts Weight Counts2 Mean Treatment Vari- (gm) 10um Total ance No Enzyme 33 .1069 25.48 89 .2821 26.04 163 .4238 31.75 141 .3835 30.35 137 .3836 29.48 320 .7872 33.53 59 .1758 27.70 213 .5943 29.58 157 .4569 28.36 29.14 262.27 5.95 Heparitinase 316 .8208 31.77 245 .7697 26.27 157 .4286 30.23 228 .6475 29.06 143 .4433 26.62 155 .5058 25.41 158 .4123 31.63 49 Table 5. (cont.) Treatments Counts Weight Count: Mean Treatment Vari- (gm) 10um Total ance Heparitinase 125 .3775 27.32 179 .5674 26.04 28.26 254.35 5.42 Heparinase 104 .3749 22.89 216 .7754 22.99 193 .6678 23.85 213 .7008 26.09 184 .6642 22.86 116 .3723 25.72 113 .3740 24.94 107 .3852 22.93 101 .3858 20.61 23.64 212.88 2.61 C.2.1. F-max test for heterogeneity of variances between treatments. P“ largaest variance / smallest variance ' 5.95 / 2.61 = 2.28 3 calculated F-max For F-max tabled: a' total number of groups ' 3 n' smallest number of observations = 9 p‘ probability level ' 0.05 F-max(tabled)= 5.34 No significant heterogeneity of variance C.2.2. ATIII Binding with Enzyme Treatment. Analysis of Variance: 1. Correction term C'(sum X)2 / total observations -(729.5)2 / 27 -532170.25 / 27 '19710.009 2. Sums of Squares SST I'sum(X2) - C (Total) ° '282.03 3. Sums of Squares SSTr-sum(T2 / r) - C (Treatment) '19866.374 - 19710.009 '156.37 50 T' totals for each group r‘ number of replicates ANOVA Table: Source dF SS MS F treatment 2 156.4 78.2 15.04 error 24 125.6 5.2 total 26 282.0 -- 4. F' MS(treatment) / MS(error) ‘ 78.2 / 5.2 ' 15.04 ' calculated F value 5. F(tabled)‘ 3.40 at 0.05 probability F(calc) > F(tabled) therfore reject Ho = all treatments are the same. Multiple Comparison of Data: Dunnett's test for comparison of all treatments to a control Treatment Mean control 29.14 heparinase 23.64 heparitinase 28.26 1. Dunnett's Critical Value Dunnett's t value x square root of (2MS(error) / r) I'2.01 x 1.08 '2.16 2. Comparison of Means Control:Heparinase 5.50 significant difference Control:Heparitinase 0.88 no significant difference BIBLIOGRAPHY BIBLIOGRAPHY Albildgaard, U. 1969. Binding of Thrombin to Antithrombin III. Scandinavian Journal of Clinical and Laboratory Investigation Andersson, L.-O., Barrowcliffe, T.W., Holmer, E., Johnson, E.A. and i— Soderstrom, G. 1979. Molecular Weight Dependency of the Heparin i1? Potentiated Inhibition of Thrombin and Activated Factor X. Effect of Heparin Neutralization in Plasma. Thrombosis Research 15:531- 541. .q Ardlie, N.G., Glew, G. and Schwartz, C.J. 1966. Influence of Catecholamines on Nucleotide-induced Platelet Aggregation. Nature 212:415-417. Awbrey, B.J., Hoak, T.C. and Owen, W.G. 1979. Binding of Human Thrombin to Cultured Human Endothelial Cells. Journal of Biological Chemistry 254:4092-4095. Ayotte, L.. Lorimeau, J.C. and Perlin, A.S. 1981. Influence of Variations in the Chemical Structure of Heparin on its Anticoagulant and Anti-factor Xa Activities. Thrombosis Research 22:97-102. Barzu, T., Molho, P., Tobelem, G., Petitou, M. and Caen, J. 1985. Binding and Endocytosis of Heparin By Human Endothelial Cells in Culture. Biochimica et Biophysica Acta 845:196-203. Bendayan, M. 1984a. Protein A-Gold Electron Microscopic Immunocytochemistry: Methods, Applications, and Limitations. Journal Electron Microscopy Technique 1:243-270. Bendayan, M. 1984b. Enzyme-Gold Electron Microscopic Cytochemistry: A New Affinity Approach for the Ultrastructural Localization of Macromolecules. Journal of Electron Microscopy Technique 1:349-372 Bjornsson, T.D. and Greenberg, C. 1985. Heparinization in the Surgical Patient. Infections in Surgery Febuary:87-97. Blackburn, M.N., Smith, R.L., Carson, J. and Sibley, C.C. 1984. The Heparin-binding Site of Antithrombin III. The Journal of Biological Chemistry 259:939-941. Booyse, F.M., Sedlak, B.J. and Rafelson, M.E. 1975. Culture of Arterial 51 52 Endothelial Cells: Characterization and Growth of Bovine Aortic Cells. Thrombosis Diathesis and Haemorrhage 34:825-839. Budd, C.C. 1971. Recent Development in Light and Electron Microscope Autoradiography. Annual Review of Cytology 31:21—56. Buonassisi, V. 1973. Sulfated Mucopolysaccharide Synthesis and Secretion in Endothelial Cell Cultures. Experimental Cell Research 76:363-369. Busch, G., Cancilla, P.A., DeBault, L.E., Goldsmith, J.C. and Owen, W.G. 1982. Use of Endothelium Cultured on Microcarriers as a Model for the Microcirculation. Laboratory Investigation 47(5):498-504. Busch, C. and Owen, W.G. 1982. Identification In Vitro of an Endothelial Cell Surface Cofactor for Antithrombin III. Journal of Clinical Investigation 69:726-729. Cash, J.D. 1978. Control Mechanisms of Activator Release. IN Progress in Chemical Fibrinolysis and Thrombolysis. Volume 3. Davison, J.F., Rowan, R.M., Samama, M.M. and Desnoyers, P.C. (eds). New York. Ravin Press. Cooper, D.R., Lewis, G.P., Lieberman, G.E., Webb, H. and Westwick, J. 1979. ADP Metabolism in Vascular Tissue, a Possible Thrombo-regulatory Mechanism. Thrombosis Research 14:901-914. Davidson, J.F., Lochhead, M., McDonald, G.A. and McNicol, G.P. 1972. Fibrinolytic Enhancement by Stanazol: a Double Blind Trial. British Journal of Haematology 22:543-547. Dryjski, M., Olsson, P. and Swedenborg, J. 1982. Uptake and Inhibition of Thrombin by the Vascular Wall. Thrombosis Research 27:467-475. Dryjski, M., Larsson, R., Olsson, P. and Swedenborg, J. 1983. Effect of Glycosaminoglycans and Antithrombin III on Uptake and Inhibition of Thrombin by the Vascular Wall. Thrombosis Research 32:355-363. Esmon, C. and Owen, W.G. 1981. Identification of an Endothelial Cell Cofactor by Thrombin-catalyzed Activation of Protein C. Proceedings of the National Academy of Sciences 78:2249-2252. Fenton, J.W., Fasco, M.J., Stackrow, A.B., Aronson, D.L., Young, A.M. and Finlayson, J.S. 1977. Human Thrombins. Production, Evaluation and Properties of Alpha-thrombin. Journal of Biological Chemistry 252:3587-3598. Gardais, A., Picard, J. and Hermelin, B. 1973. Glycosaminoglycan (GAG) Distribution in Aortic Wall From Five Species. Comparative Biochemistry and Physiology 44B:507-515. Gimbrone, M.A. 1976. Culture of Vascular Endothelium. Progress in Hemostasis and Thrombosis 3:1-28. 53 Gimbrone, M.A. 1981. Vascular Endothelium and Atherosclerosis IN Moore, S. (ed). The Biochemistry of Disease. New York, Marcel Dekker. Goldsmith, J.C., McCormick, J.J. and Yen, A. 1984. Endothelial Cell Kinetics. Laboratory Investigation 51(6):643-647. Gggenwood, P.C., Hunter, W.M. and Glover, J.S. 1980. The Preparation of I-labeled Human Growth Hormone of High Specfic Activity. Biochemical Journal 89:114-123. Hatton, M.W., Berry, L.R. and Regoeczi. 1978. Inhibition of Thrombin by Antithrombin III in the Presence of Certain Glycosaminoglycans Found in the Mammalian Aorta. Thrombosis Research 13: 655-670. Hogan, M.E., Hassouna, H.I. and Klomparens, K.L. 1986. Sectionable Microcarrier Beads: An Improved Method for the Preparation of Cell Monolayers for Electron Microscopy. Journal of Electron Microscopy Technique 5:159-169. Hook, M., Kjellen, L., Johansson, S. and Robinson, J. 1984. Cell - Surface Glycosaminoglycans. Annual Review of Biochemistry 53: 847-869. Hopkins, N.K. and Gorman, R.R. 1981. Regulation of Endothelial Cell Cyclic Nucleotide Metabolism by Prostacyclin. Journal of Clinical Investigation 67:540-546. Huey, S. 1985. Mini-Masters manuscript. Michigan State University, Department of Biochemistry. Hurst, R.E., Menter, J.M., West, 8.8., Settine, J.M. and Coyne, E.H. 1979. Structural Basis for the Anticoagulant Activity of Heparin. I. Relationship to the Number of Charged Groups. Biochemistry 18: 4283-4287. Jaffe, E.A., Hoyer, L.W. and Nachman, R.L. 1974. Synthesis of von Willebrand factor by Cultured Human Endothelial Cells. Proceedings of the National Academy of Sciences 71:1906-1909. Johnson, E.A. and Mulloy, B. 1976. The Molecular Weight Range of Mucosal Heparins. Carbohydrate Research 1951:119-127. Klomparens, K.L., Flegler, S.L. and Hooper, G.R. 1986. IN Procedures for Transmission and Scanning Electron Microscopy for Biological and Medical Science: A Laboratory Manual. Second Edition. Ladd Research Industries. pp 146. Laug, W.E. 1981. Secretion of Plasminogen Activators of Cultured Bovine Endothelial Cells: Partial Purification, Characterization and Evidence for Multiple Forms. Thrombosis and Haemostasis 45: Laurent, T.C., Tengblad, A., Thunberg, L., Hook, M. and Lindahl, U. 54 1978. On the Molecular Weight Dependence of the Anticoagulant Activity of Heparin. Biochemical Journal 175:691-701. Levin, E. G. and Loskutoff, D. J. 1982. Regulation of Plasminogen Activator Production By Cultured Endothelial Cells. IN Endothelium Fishman, A. P., (ed). Annals of the New York Academy of Sciences 401: 184-194. Levin, E.G., Marzec, U., Anderson, J. and Harker, L.A. 1984. Thrombin Stimulates Tissue Plasminogen Activator Release from Cultured Human Endothelial Cells. Journal of Clinical Investigation 74:1988-1995. Lindahl, U. and Hook, M. 1978. Glycosaminoglycans and Their Binding to Biological Macromolecules. Annual Review of Biochemistry 47:385-417 Lollar, P. and Owen, W.G. 1980. Clearance of Thrombin from Circulation In Rabbits by High Affinity Binding Sites On Endothelium. Journal of Clinical Investigation 66:1222-1230. Lollar, P., MacIntosh, S. and Owen, W.G. 1982. Interaction of Thrombin With the Vascular Endothelium in a Recirculating Langendorff Rabbit Heart Preparation. IN Endothelium. Fishman, A. P., (ed). Annals of the New York Academy of Sciences 401: 234-240. Marcum, J.A., Fritze, L., Galli, S.J., Karp, G. and Rosenberg, R.D. 1983. Microvascular Heparinlike Species With Anticoagulant Activity. American Journal of Physiology 245:H725-H733. Marcum, J.A., McKenney, J.B. and Rosenberg, R.D. 1984. Acceleration of Thrombin-Antithrombin Complex Formation in Rat Hindquarters Via Heparinlike Molecules Bound to the Endothelium. Journal of Clinical Investigation 74:341-350. Marcum, J.A., Atha, D.H., Fritze, L.M., Nawroth, P., Stern, D. and Rosenberg, R.D. 1986. Cloned Bovine Aortic Endothelial Cells Synthesize Anticoagulantly Active Heparin Sulfate Proteoglycans. The Journal of Biological Chemistry 261:7507-7517. Matsuo, O. 1982. Turnover of Tissue Plasminogen Activator in Man. Thrombosis and Haemostasis 48:242-246. Mosher, D.R., Schad, P.E. and Kleinman, H.K. 1979. Cross-linking of Fibronectin to Collagen by Blood Coagulation Factor XIIIa. Journal of Clinical Investigation 64:781-787. Mosher, D. P., Doyle, M. J. and Jaffe, E. A. 1982. Synthesis and Secretion of Thrombospondin and Fibronectin by Cultured Endothelial Cells. IN Pathobiology of the Endothelial Cell. Academic Press, New York. Nilsson, I.M. 1978. Effect of Drugs on Activator Synthesis and Release. IN Progress in Chemical Fibrinolysis and Thromolysis. Volume 3. Davison, J.F., Rowan, R.M., Samama, M.M. and Desnoyers, P.C. (eds). 55 New York, Raven Press. Ogston, D. 1983. IN The Physiology of Hemostasis. Harvard University Press. Cambridge. pp.1-10. Pearson, J.D., Carleton, J.S., Hutchings, A. and Gordon, J.L. 1978. Uptake and Metabolism of Adenosine by Pig Aortic Endothelial and Smooth Muscles Cells in Culture. Biochemical Journal 170:265-271. Riesenfeld, J., Thunberg, L., Hook, M. and Lindahl, U. 1981. The Antithrombin-binding Sequence of Heparin. The Journal of Biological Chemistry 256:2389-2394. Robbins, S.L., Cotran, R.S. and Kumar, V. 1984. IN Pathologic Basis of Disease. Third Edition. Philadelphia. W.E. Saunders Co. Rosenberg, R.D. and Rosenberg, J.S. 1984. Natural Anticoagulant Mechanisms. Journal of Clinical Investigation 74:1-6. Sakariassen, K.S., Bolhuis, P.A. and Sixma, J.J. 1979. Human Blood Platelet Adhesion to Artery Subendothelium is Mediated by Factor VIII-Von Willebrand Factor Bound to the Subendothelium. Nature 279:638-648. Stern, D., Brett, J., Harris, K. and Nawroth, P. 1986. Participation of Endothelial Cells in the Protein C-Protein S Anticoagulant Pathway: The Synthesis and Release of Protein S. The Journal of Cell Biology 102:1971-1978. Thilo, D.G.S., Heinrich, D. and Lasch, H.G. 1983. Endothelial Cell Control Functions in Blood Coagulation. IN The Endothelial Cell - A Pluripotent Control Cell of the Vessel Wall. Thilo, D.G.S., Freshney, R.I. (eds). S. Karger. New York. pp.29-54. Thompson, A.R. and Harker, L.A. 1983. Manual of Hemostasis and Thrombosis. Third Edition. Philadelphia, F.A. Davis Company. Todd, A.S. 1959. The Histological Localization of Fibrinolysin Activator. Journal of Pathology and Bacteriology 78:281-288. Walker, F. 1984. Protein 8 and the Regulation of Activated Protein C. Seminars in Thrombosis and Hemostasis 10:131-138. Yamada, K.M. and Olden, K. 1978. Fibronectins - adhesive glycoproteins of Cell Surface and Blood. Nature 275:179-184.