• b /38 If! 36 U'IIVCI::>lLy r ree crate 11111111111111111111111111111111111111111111111111111111111111111111111111111111 34300000347058 Universiteit Vrystaat -HIE-RDI-E E-KSE-MPlA-AR-M-AG ONDER' GEEN OMSTANDIGHEDE UIT DIE \. l'::~BUOTEEK VERWYDER WORD NIEo RECOMBINANT PRODUCTION AND EVALUATION OF A MULTIFUNCTiONAL HAEMOSTATIC FUSION PROTEIN by WALDA BRENDA VAN ZYL Submitted in fulfilment of the requirements for the degree of Philosophiae Doctor In the Department of Haematology and Cell Biology, Faculty of Health Sciences, University of the Orange Free State, Bloemfontein, South Africa OCTOBER 1999 Promoter: Prof. G.H.J. Pretorius Co-promoter: Prof. H.F. Kotzé ACKNOWLEDGEMENTS I wish to express my gratitude and appreciation to the following persons and institutions: Prof. G.H.J. Pretorius, for his invaluable guidance, patience and support during this study. Prof. H.F Kotzé, for his helpful advice, contributions and time. Prof. P.N. Badenhorst, Head of the Department of Haematology and Cell Biology, University of the Orange Free State, for his continuous interest in my work. Kim Alexander, Seb Lamprecht, Jan Roodt, Paul Slater and Elmarie Wentzel for their assistance and help. The Medical Research Council and Central Research Fund of the University of the Orange Free State, for financial support. A special thanks to Colleen Schultz and Nerina Visser, for all the coffee and laughs and for their invaluable support. All my thanks to Elize Muller and Trudi van Rensburg for all the talks vial e-mail and for their support. All my friends and colleagues from the Departments of Haematology and Cell Biology, Human Genetics and Botany, for their friendship and support. My parents and family for their love and encouragement. My husband, Ludi for all his love and interest in my work and well-being. Lastly, I would like to thank the lord for giving me the perseverance and courage to complete this study, to Him all the Honour. DECLARATION I, the undersigned, hereby declare that the word contained within this thesis is my original and independent work and has not in its entirety or in part been submitted to any university for a degree. All the sources I have made use of or quoted have been acknowledged by complete references. W.B. van Zyl October 1999 Dedicated to my husband, Ludi CONTENTS PAGE LIST OF FIGURES LIST OF TABLES iv ABBREVIATIONS vi 1. GENERAL INTRODUCTION 1 1.1 INTRODUCTION 1 1.2 AIM AND SCOPE OF THE PROJECT 2 1.3 REFERENCES 4 2. LITERATURE REVIEW 8 2.1 INTRODUCTION 8 2.2 THROMBOSIS AND HAEMOSTASIS 8 2.2.1 Platelet function 9 2.2.2 Coagulation 14 2.2.3 Fibrinolysis 18 2.3 NOVEL ANTICOAGULANT AND ANTITHROMBOTIC 22 AGENTS 2.3.1 Inhibition of platelet aggregation 23 2.3.2 Direct inhibition of thrombin 33 2.3.3 Indirect inhibition of thrombin 43 2.3.4 Activation of fibrinolysis 46 2.3.5 Combination proteins 52 2.4 REFERENCES 56 3. PRODUCTION OF A RECOMBINANT ANTITHROMBOTIC AND 83 FIBRINOLYTIC PROTEIN, PLATSAK, IN Escherichia coli ABSTRACT 83 1 MATERIALS AND METHODS 84 1.1 Host strain 84 1.2 Enzymes and chemicals 84 1.3 Recombinant DNA techniques 84 1.4 Cloning strategy 84 1.5 Media and cultivation 84 1.6 Protein extraction and purification 86 1.7 Measurement of antithrombin activity 86 1.8 Measurement of antiplatelet activity 86 1.9 Measurement of fibrinolytic activity 86 2 RESULTS AND DISCUSSION 87 REFERENCES 89 4. OPTIMIZATION OF ANTI PLATELET ACTIVITY 91 4.1 INTRODUCTION 91 4.2 MATERIALS AND METHODS 91 4.2.1 Host strain 91 4.2.2 Enzymes and chemicals 91 4.2.3 Recombinant DNA techniques 92 4.2.4 Cloning strategy 92 4.2.5 Cultivation, protein purification and measurement of 92 antiplatelet activity 4.3 RESULTS AND DISCUSSION 93 4.4 CONCLUSIONS 97 4.5 REFERENCES 97 5. OPTIMIZATION OF PURIFICATION OF PLATSAK 99 5.1 INTRODUCTION 99 5.2 MATERIALS AND METHODS 99 5.2.1 Conventional metal affinity chromatography using Ni- 99 NTA resin 5.2.2 Metal chromatography using an FPLC-column 101 5.3 RESULTS AND DISCUSSION 102 5.4 CONCLUSIONS 105 5.5 REFERENCES 107 6. PLATSAK, A POTENT ANTITHROMBOTIC AND FIBRINOLYTIC 108 PROTEIN, INHIBITS ARTERIAL AND VENOUS THROMBOSIS IN A BABOON MODEL SUMMARY 108 INTRODUCTION 109 MATERIALS AND METHODS 112 Preparation of PLATSAK 112 Animals studied 112 Study protocol 113 Graft imaging and quantification of platelet deposition 115 Laboratory measurements 115 RESULTS 116 DISCUSSION 118 ACKNOWLEDGEMENTS 123 REFERENCES 123 7. CONCLUSIONS 128 7.1 REFERENCES 131 SUMMARY 134 LIST OF FIGURES PAGE CHAPTER 1 Fig. 1.1 Schematic representation of the composition of 3 PLATSAK and the function of the individual components. CHAPTER 2 Fig. 2.1 An overview of haemostasis. 10 Fig. 2.2 A schematic representation of the important participants 13 in thrombus formation. Fig. 2.3 Coagulation cascade. 15 Fig.2.4 Schematic representation offibrin formation and 17 subsequent lysis by plasmin. Fig. 2.5 The interactions of physiological thrombolysis in which 20 the process is locally favoured but systemically inhibited. Fig. 2.6 A model showing the molecular link between coagulation 21 and fibrinolysis. Fig. 2.7 Schematic summary of the different interaction 34 mechanisms between thrombin and its inhibitors. ii CHAPTER 3 Fig. 1 Reverse translation of the amino acid sequence of the 85 antithrombotic peptide and chemical synthesis of two overlapping oligonucleotides. Fig. 2 A schematic representation of the strategy followed for 86 construction of the PLATSAK gene. Fig. 3 SOS-PAGE analysis of protein purification procedure. 87 Fig.4 Determination of antithrombotic activity using TT and 88 aPTT. Fig. 5 A. Determination of thrombin inhibition using 88 Chromozyme TH as substrate for thrombin. Fig. 5 B. Determination of the inhibition of ADP induced platelet 88 aggregation. CHAPTER 4 Fig.4.1 Cloning strategy followed to produce PLATSAK2 and 94 PLATSAK3. Fig 4.2 SOS-PAGE analysis of intracellular proteins to follow the 96 induction of protein synthesis. CHAPTER 5 Fig. 5.1 Interaction between the His-tag and the Ni-NTA-resin. 100 iii Fig. 5.2 SDS-PAGE of purification using Poros MC column. 104 Proteins were eluted with a decrease in pH. Fig. 5.3 SDS-PAGE of purification using Poros MC column. 106 Proteins were eluted with 0.5 M imidazole. CHAPTER 6 Fig. 1 Schematic representation of the composition of PLATSAK 111 and the function of the individual components. Fig. 2 A schematic representation of the experimental set-up to 114 study platelet deposition. Fig. 3 The effect of PLATSAK on platelet deposition in control 117 animals and after treatment. iv LIST OF TABLES PAGE CHAPTER 2 Table 2.1 Summary of antiplatelet proteins. 26 Table 2.2: Summary of direct antithrombins. 35 Table 2.3: Primary structure of hirunorm Vand its comparison 40 with hirudin variant 2. Table 2.4: Summary of indirect antithrombins. 44 Table 2.5: Summary of fibrinolytic proteins. 47 CHAPTER 3 Table 1 Summary of primers. 85 Table 2 Comparison of L-values obtained from 89 thrombelastography. CHAPTER 4 Table 4.1 Summary of primers. 93 v CHAPTER 6 Table 1: Summary of the changes that were observed in platelet 119 count, aPTT, TAT complexes and FOP levels. vi ABBREVIATIONS U2-PI - u2-plasma inhibitor 1111n- 111-lndium ABE - Anion binding exosite ADP - Adenosine diphosphate APC - Activated protein C aPTT - Activated partial thromboplastin time AT 111-Antithrombin III AV - Arteriovenous cDNA - Complementary DNA CHO - Chinese hamster ovary EDTA - Ethylenedinitrilo tetraacetic acid ELISA - Enzyme-linked immuno-sorbent assay FOP - Fibrinogen degradation products FPA - Fibrinopeptide A FPLC - Fast protein liquid chromatography Gp - Glycoprotein HPI - Hookworm platelet inhibitor HLA - Human leukocyte antigen IPEC - Immortalized porcine endothelial cells IPTG - lsopropyl-ê-Dvthiopalactopyranoslde kDa - Kilodalton KGD - Lysine-Glycine-Aspartic acid KGDS - Lysine-Glycine-Aspartic Acid-Serine LAPP - Leech antiplatelet protein Ni-NTA - Nickel-nitrilotriacetic acid NTA - Nitrilotriacetic acid PAI-1 - Plasminogen activator inhibitor-1 PAI-2 - Plasminogen activator inhibitor-2 PAI-3 - Plasminogen activator inhibitor-3 PCR - Polymerase chain reaction vii PLATSAK - Platelet-Anti!hrombin-St~phylo.!sinase PPACK - O-Phenyl-L-Prolyl-L-Argenyl-chloromethylketone Pr - Primer psi - pressure per square inch rAPC - Recombinant activated protein C RGD - Arginine-Glycine-Aspartic acid RGONP - Arginine-Glycine-Aspartic acid-Asparagine-Proline RGOS - Arginine-Glycine-Aspartic acid-Serine RGOW - Arginine-Glycine-Aspartic acid-Tryptophan rLAPP - Recombinant leech antiplatelet protein rscu-PA - Recombinant single chain urokinase-type plasminogen activator rTAP - Recombinant tick anticoagulant peptide SAK - Staphylokinase scu-PA - Single chain urokinase-type plasminogen activator SOGE - Serine-Aspartic acid-Glycine-Glutamic acid SOS-PAGE - Sodium dodecyl sulphate polyacrylamide gel electrophoresis TAFI - Thrombin activatable fibrinolysis inhibitor TAP - Tick anticoagulant peptide TAT - Thrombin-antithrombin III TF - Tissue factor TFPI - Tissue factor pathway inhibitor t-PA - Tissue-type plasminogen activator TRAP - Thrombin receptor activating peptide TT - Thrombin time TxA2 - Thromboxane A2 u-PA - Urokinase-type plasminogen activator VWF - Von Willebrandt factor 1 CHAPTER 1 GENERAL INTRODUCTION 1.1 INTRODUCTION: Haemostasis is the processes involved in the prevention of blood loss during vascular injury, while thrombosis is a pathological outcome of vascular disease. Haemostasis involves complex interactions between damaged vessel wall surfaces, activated blood platelets and activated coagulation factors. These interactions ultimately lead to the production of thrombin, which is responsible for the conversion of soluble fibrinogen to insoluble fibrin. Thrombin also acitvates platelets and activates coagulation factors Vand VIII. Cross-linking between fibrin strands results in the formation of a haemostatic plug at the site of vascular injury. Following wound repair, the haemostatic plug will be removed by a physiological process called fibrinolysis, which is also a delicately balanced cascade of interactions amongst several proteins. Since thrombin plays a central role in the maintenance of vascular integrity, it has been the focus of research in the field of antithrombotic agents in the past few years. Hirudin, a potent inhibitor of thrombin, is produced by the medicinal leech and has been studied intensively (Markward, 1970; Rydel et aI, 1990; Rydel et aI, 1991). Fragments and derivatives of hirudin have also been studied thoroughly (Krstenansky and Mao, 1987; Mao et al, 1988; Naski et al, 1990; Schmitz et al, 1991). Much research has also been devoted to the prevention of platelet aggregation, in particular blockade of the receptor of fibrinogen on the activated platelet membrane, glycoprotein lib/Ilia (Caller, 1985; Deckmyn et aI, 1994; Foster et aI, 1994). Several disintegrins from snake venoms were studied and their role in the inhibition of platelet activation has been elucidated (Dennis et aI, 1990; Niewiarowski et aI, 1990; Savage et aI, 1990; Scarborough et aI, 1993). 2 Exogenous activators of fibrinolysis, streptokinase and staphylokinase, have been studied for their potential as fibrinolytic agents (Collen et aI, 1993; Collen and Lijnen, 1994; Collen, 1997). Several combination proteins, which simultaneously target haemostasis at different levels, were recently developed. The activities of these novel proteins include combined antithrombin and antiplatelet activity (Knapp et aI, 1992), fibrinolytic and antiplatelet activity (Smith et aI, 1995) as well as fibrinolytic and antithrombin activity (Lijnen et aI, 1995). Another interesting approach in terms of fibrin targeting of an antithrombin was followed by Bode et al (1994). They fused the Fab' of a monoclonal antibody, 59D8, to recombinant hirudin. The resultant antithrombin was substantially more effective than recombinant hirudin. 1.2 AIM AND SCOPE OF THE STUDY: In this study, I have developed a novel chimeric protein that would target haemostasis at three levels. The protein, named PLATSAK, was designed to inhibit the action of thrombin, prevent platelet aggregation and activate fibrinolysis. It consists of staphylokinase (SAK), linked via a factor Xa cleavage site, to an antithrombotic and antiplatelet peptide (Fig. 1.1). The chimera was designed to act as a local drug delivery system. Theoretically, the high fibrin-specificity of SAK should transport PLATSAK to a fibrin-containing thrombus and thus to an environment that contains high concentrations of factor Xa. The peptide can then be released by factor Xa in the proximity of recently activated thrombin and activated platelets. Additional platelet aggregation and fibrin formation can subsequently be prevented by the antiplatelet and antithrombotic peptide. The peptide was designed to contain three inhibitory regions. Firstly, on its N-terminus it has the RGD-sequence for binding to the fibrinogen receptor (Gp Ilb/llla) to prevent fibrinogen binding to platelets and so also platelet aggregation (Ruoslahti and Pierschbacher, 1987). That is followed by a part of fibrinopeptide A (residues 8-16) to block the active site of thrombin (Martin et aI, 1992) and the C-terminus of hirudin (residues 54-65) to block the anion binding site of thrombin (Markwardt, 1970). 3 Hirudin SAK (54-65) Gp lib/ilia Activates - inhibits platelet plasminogen aggregation - promotes fibrinolysis - targets a thrombus Active site - inhibits proteolysis Releases peptide in vivo Anion binding exosite - inhibits fibrin formation Fig. 1.1 Schematic representation of the composition of PLATSAK and the function of the individual components. The staphylokinase part is linked to the antiplatelet and antithrombotic peptide via the recognition sequence of factor Xa. 4 I have decided to use Escherichia coli cells as expression host for the recombinant production of PLATSAK. It is a well-known organism in terms of its genetic and biochemical characteristics and has been widely used as expression host for the production of novel agents. Furthermore, it is a fast growing organism and is easy and relatively inexpensive to cultivate. The aim of the study was to successfully construct the hybrid gene, express it in E.coli cells and produce sufficient amounts of the protein to evaluate its biological activities in vitro. Furthermore, the project was aimed at the estimation of the potential of PLATSAK in an in vivo model. 1.3 REFERENCES: Bode C, Hudelmayer M, Mehwald P, Bauer S, Freitag M, Van Hodenberg E, NewelI JB, Kublet W, Haber E, Runge MS. Fibrin-targeted recombinant hirudin inhibits fibrin deposition on experimental clots more efficiently than recombinant hirudin. Circulation 1994;90: 1956-63. Collen D. Thrombolytic therapy. Thromb Haemast 1997;78:742-6. Collen 0, Lijnen HR. Staphylokinase, a fibrin-specific plasminogen activator with therapeutic potential? Blood 1994;84:680-6. Collen 0, Schlott B, Engelborghs Y, Van Hoef B, Hartmann M, Lijnen HR, Behnke D. On the mechanism of the activation of human plasminogen by recombinant staphylokinase. J Bioi Chem 1993;268:8284-9. Caller BS. A new murine monoclonal antibody reports inactivation-dependent change in the conformation and/or micro-environment of the platelet glycoprotein IIb/llla complex. J Clin Invest 1985;76:101-8. 5 Deckmyn H, Stanssens P, Hoet B, Declerck PJ, Lauwereys M, Gansemans Y, Tornai I, Vermylen J. An echistatin-like Arg-Gly-Asp (RGD)-containing sequence in the heavy chain CDR3 of a murine monoclonal antibody that inhibits human platelet glycoprotein Ilblllla function. Br J Haematol 1994;87:562-71. Dennis MS, Henzei WJ, Pitti RM, Lipari MT, Napier MA, Deisher TA, Bunting S, Lazarus RA. Platelet glycoprotein Ilblllla protein antagonists from snake venoms: Evidence for a family of platelet aggregation inhibitors. Proe Natl Acad Sci USA 1990;87:2471-5. Foster MR, Hornby EJ, Brown S, Hann M, Kitchin J, Pike N, Ward P. Inhibition of human platelet aggregation by GR91669, a prototype fibrinogen receptor antagonist. Thromb Res 1994;75:269-84. Knapp A, Degenhardt T, Dodt J. Hirudisins. J Bioi Chem 1992;267:24230-4. Krstenansky JL, Mao SJT. Antithrombin properties of the C-terminus of hirudin using synthetic unsulfated NCt-acetyl-hirudin45-65. FEBS Lett 1987;211:10- 6. Lijnen HR, Wnendt S, Schneider J, Janocha E, Van Hoef B, Collen 0, Steffens GJ. Functional properties of a recombinant chimeric protein with combined thrombin inhibitory and plasminogen-activating potential. Eur J Biochem 1995;234:350-7. Mao SJT, Yates MJ, Owen TJ, Krstenansky JL. Interaction of hirudin with thrombin: Identification of a minimal binding domain of hirudin that inhibits clotting activity. Biochemistry 1988;27:8170-3. Markwardt F. Hirudin as an inhibitor of thrombin. Methods Enzymol 1970;19:924- 32. 6 Martin PO, Robertson W, Turk 0, Huber R, Bode W, Edwards BFP. The structure of residues 7-16 of the Aa-chain of human fibrinogen bound to bovine thrombin at 2.3-A resolution. J Bioi Chem 1992;267:7911-20. Naski MC, Fenton II JW, Maraganore JM, Olsen ST, Shafer JA. The COOH- terminal domain of hirudin. J Bioi Chem 1990;265: 13484-9. Niewiarowski S, Cook JJ, Stewart GJ, Gould RJ. Structural requirements for expression of antiplatelet activity of disintegrins. Circulation 1990;82 (Supplement 111):370(Abstract). Ruoslahti E, Pierschbacher M.D. New perspectives in cell adhesion: ROG and integrins. Science 1987;238:491-7. Rydel TJ, Ravichandran KG, Tulinsky A, Bode W, Huber R, Roitsch C, Fenton II JW. The structure of a complex of recombinant hirudin and human a- thrombin. Science 1990;249:277-80. Rydel TJ, Tulinsky A, Bode W, Huber R. Refined structure of the hirudin-thrombin complex. J Mol Bioi 1991 ;221 :583-601. Savage B, Marzec UM, Chao BH, Harker LA, Maraganore JM, Ruggeri ZM. Binding of the snake venom-derived proteins applaggin and echistatin to the arginine-glycine-aspartic acid recognition site(s) on platelet glycoprotein IIblllla complex inhibits receptor function. J Bioi Chem 1990;265:11766-72. Scarborough RM, Rose JW, Naughton M, Phillips DR, Nannizzi L, Arfsten A, Campbell AM, Charo IF. Characterization of the integrin specificities of disintegrin isolated from American pit viper venerns. J Bioi Chem 1993;268: 1058-65. 7 Schmitz T, Rothe M, Dodt J. Mechanisms of the inhibition of a-thrombin by hirudin-derived fragments hirudin(1-47) and hirudin(45-65). Eur J Biochem 1991; 195:251-6. Smith JW, Tachias K, Madison EL. Protein loop grafting to construct a variant of tissue-type plasminogen activator that binds platelet integrin allb~3. J Bioi Chem 1995;270:30486-90. 8 CHAPTER 2 LITERATURE REVIEW 2.1 INTRODUCTION: Remarkable strides have been made to unravel the molecular mechanisms that underlie thrombosis and haemostasis, especially with regard to the central role of thrombin and blood platelets. This knowledge and the high incidence of vascular related diseases led to a drastic increase in the development of novel and potent antithrombotic agents. These agents are targeted to either effectively prevent the formation of thrombin or inhibit its activity, or to inhibit platelet-platelet interactions. Furthermore, thrombolytic agents can be utilised to dissolve excessive fibrin clots and can also be used in conjunction with antithrombotic agents to prevent reocclusion of blood vessels following successful reperfusion. This review will briefly focus on the molecular principles of thrombosis and haemostasis, .especially with regard to the role of platelets, coagulation and fibrinolysis. The recent development and mode of action of several novel and potent antiplatelet, antithrombotic and fibrinolytic agents will also be discussed in detail. 2.2 HAEMOSTASIS AND THROMBOSIS: Haemostasis is a defence mechanism that prevents excessive blood loss by maintaining vascular integrity. Thrombosis on the other hand, is a pathological consequence of vascular disease. The haemostatic process involves complex and integrated interactions between damaged vessel wall surfaces, activated blood platelets and activated coagulation factors to form a localized haemostatic plug to prevent blood loss. Haemostasis involves complicated systems of activation and I inhibition, which limit generalized extension of thrombosis beyond the area of 9 damage (Fig. 2.1). Under strictly controlled conditions the haemostatic plug is subsequently dissolved by fibrinolysis. When these contact mechanisms are overcome by excessive platelet activation, platelets and coagulation become major contributors to the development of thrombosis and can lead to morbidity and mortality due to myocardial infarction and stroke (Harker, 1990). 2.2.1 Platelet Function: Blood platelets are produced by bone marrow megakaryocytes. They are about 3 urn in diameter and 1 urn thick and thus the smallest of all blood cells (Blockmans et aI, 1995). Regardless of their small size they are biologically among the most active and play a substantial role in haemostasis, thrombosis and atherosclerosis (Wu, 1996). Resting platelets circulate the vascular system as discoid anuclear cells. Cellular constraints and factors like prostacyclin and nitric oxide prevent adhesion and aggregation of resting platelets and allow close contact to the endothelial cell lining without adhering to it (Wu, 1996). P-selectin, a cell adhesion molecule present in platelets and endothelial cells, is stored in the secretory granules and is rapidly expressed on the plasma membrane upon activation (Hsu- Un et aI, 1984). Both platelet and endothelial P-selectin mediate leukocyte adhesion (McEver et aI, 1995) and are believed to be responsible to position circulating platelets close to the vessel wall. Following vascular injury, platelets form the first line of defence to prevent excessive blood loss. This occurs through a series of well-defined reactions that culminate in the formation of a platelet plug at the area of injury. The reactions include activation of platelets, change in platelet shape, release of their granular content and ultimately platelet aggregation (Harker, 1990; Ruggeri, 1997). It is generally accepted that binding of platelets to collagen in vivo is the trigger that starts the haemostatic process. Intact endothelial cells, adjacent to the injury, produce inhibitors of both coagulation and platelets to control the size of the hemostatic plug and contain it to the area of injury (Blockmans et aI, 1995). 10 ~VESSELINJURV~ VWF ~ I TISSUE FACTOR ~PI ......... HEPARIN NO - - - - - - - - ~ ADP ADPases - - - - - - ~ TX~ FIBRIN PLATELET ./ AGGREGATES Jr RECANALIZATION FIBRINOL VSIS ~_A_N_T_IP_L_A_SM___IN -+» I CELL MIGRATION PAl l AND PROLIFERATION HEALING Fig. 2.1 An overview of haemostasis (Colman et ai, 1994). Solid lines indicate activation and broken lines inhibition. 11 The platelet plasma membrane consists of a phospholipid bilayer containing the glycoproteins (Gp), which serve as receptors for adhesive proteins and seven transmembrane receptors for activating and inhibitory agents to bind to (Blockmans et aI, 1995). The glycoproteins and their ligands play essential roles in platelet plug formation. Of particular interest are Gp la/lla, IbN/IX and lib/Ilia. Under low flow conditions, fibronectin binding to Gp la/lla play the most important role in platelet adhesion (Weiss, 1995), while van Willebrand factor (VWF) is most important under high flow conditions (Ruggeri, 1997). Circulating VWF cannot bind to platelets, but is capable of binding to collagen. When bound to collagen, VWF undergoes certain conformational changes, which enable it to bind to Gp Ib/IXN. Initially, the A 1 domain of VWF rapidly binds to Gp lb« on platelet membranes (Savage et aI, 1996). This bond has a high dissociation rate, resulting in detachment of the tailing end of the platelet and a forward rotational movement due to the torque produced by the blood flow. It does, however, slow the progress of the platelets across the damaged area. There are two schools of thought on the mechanisms that anchor platelets at the site of injury. The first suggests that the rolling movement of the platelet continues until Gp lib/Ilia becomes activated and binds to the RGDS sequence in the C1 domain of VWF (Savage et aI, 1992; Ruggeri, 1997). The second proposes that the anchor is when Gp la/lla binds to collagen in the subendotheluim (Vermylen et aI, 1997). Thrombin, adenosine diphosphate (ADP) and thromboxane A2 (TxA2) are interdependent agonists that activate platelets (Harker 1994; Ware and Caller, 1995). Thrombin is the most potent and primary agonist of platelets and activation is achieved by binding to its receptor on platelets (Vu et aI, 1991). ADP is released from activated platelets and is responsible for recruitment of surrounding platelets, while TxA2 induces receptor activation (Harker, 1994). Activation results in a change in platelet shape to markedly increase the membrane surface area to facilitate platelet-subendothelium and platelet-platelet interactions. Activation also results in the release of the contents of platelet granules. The granules contain platelet specific proteins (platelet factor 4 and p-thromboglobulin), adhesion proteins (fibrinogen, von Willebrandt factor, fibronectin, vitronectin, and thrombospondin), coagulation factors (factors Vand XIII and protein S), mitogenic 12 factors (platelet derived growth factor, endothelial cell growth factor and epidermal growth factor), fibrinolytic proteins (u2-antiplasmin and plasminogen activator inhibitor-1) and plasma proteins (albumin and immunoglobulins). Once activated, platelets can interact to form the hemostatic plug. The 40 000 - 50 000 Gp lib/Ilia receptors and the adhesion proteins fibrinogen, VWF, fibronectin and vitroneetin play essential roles in platelet aggregation. The Gp Ilb/llla receptors on resting platelets bind fibrinogen weakly and are incapable of binding to VWF. Activation changes the Gp Ilb/llla receptor into a high-affinity binding site for fibrinogen and VWF. Gp lib/Ilia recognises two binding domains on fibrinogen: an RGD-sequence in each of the two A-alpha chains and a dodecapeptide in the gamma chain (Ruggeri, 1993; Ginsberg et al, 1995). Fibrinogen molecules bind to Gp lib/Ilia to form molecular bridges between two adjacent platelets to stabilise the thrombus. Activated Gp lib/Ilia on the luminal side of the platelet plug may bind circulating fibrinogen or VWF, which in turn may interact with Gp lib/Ilia of activated platelets in the neighbourhood. This process can be repeated several times to add a new layer of platelets to form the haemostatic plug (Ware and Coiier, 1995). Platelet adhesion and aggregation and the relevant participants in thrombus formation are schematically shown in Fig. 2.2. In concert with the events of platelet adhesion and aggregation, the coagulation cascade is activated and thrombin is responsible for the conversion of fibrinogen to fibrin (see Section 2.2.2). Activated platelets are responsible for providing a negatively charged surface for the coagulation factor complexes to bind to. The conversion of factor X to Xa and prothrombin to thrombin is facilitated to enhance thrombus formation on the activated platelet membrane (Tracy 1988; Mann et aI, 1990). Several factors control the growth of a hemostatic plug. Firstly, aggregating agents are removed from the site of plug formation by the flowing blood and its concentration is diluted in the area of thrombogenesis (Weiss, 1995). Secondly, thrombin stimulates adjacent intact endothelial cells to release platelet inhibitors 13 Gp IbNIX-vWF-collagen +Gp la/lla-collagen o Gp lib/ilia -fibrinogen-Gp lib/ilia Fig. 2.2 A schematic representation of the important participants in thrombus formation. VWF binds to collagen and Gp IbN/IX to promote platelet adhesion to the exposed subendothelium. This process is helped by the binding of collagen to Gp la/lla. Binding of fibrinogen, VWF, fibronectin or vitronectin to Gp Ilb/llla on the membrane of activated platelets results in platelet aggregation. Thrombin, ADP and TxA2 are important receptors for platelet agonists on the platelet membrane. 14 like prostacyclin and nitric oxide (Ware and Caller, 1995). Thirdly, thrombin binds to thrombomodulin on the endothelial cell membrane to activate protein C, which is regarded as the most important in vivo inhibitor of thrombin formation (Esrnon, 1993). 2.2.2 Coagulation: The factors of the coagulation cascade circulate as inert pro-enzymes. They are activated on negatively charged membrane surfaces, provided by activated platelets, by another activated coagulation factor in the presence of a eo-enzyme and Ca-ions (Fig. 2.3). The coagulation factors are serine proteases and activation of a pro-enzyme lower down in the cascade is achieved by removing a peptide from the pro-enzyme at a serine residue in the active site (Kay, 1988). Thrombin is the final product of this series of reactions (Fig. 2.3). Prothrombin is cleaved through a series of steps to thrombin A and B and the biologically active Cf.- thrombin. Activation of thrombin leads to the exposure of the anion binding exosite (ABE), which enables thrombin to bind to negatively charged molecules like heparin (Rosenberg and Damus, 1973), fibrinogen (Fenton et al, 1988) and the C-terminus of hirudin (Grutter et al, 1990; Naski et al, 1990; Rydel et al, 1990). Residues 7-16 of thrombin are essential for the catalytic efficiency of the molecule and represents the anionic exosite domain (Berliner et aI, 1985; Lord et aI, 1990; De Cristofaro and Castagnola, 1991). The active site is responsible for its amidolytic activity and results in hydrolysis of small substrates like tripeptide p-nitroanilide and binding of peptides like fibrinopeptide A (Martin et al, 1992). Additionally, thrombin has an apolar binding site adjacent to its catalytic site in the fibrinopeptide groove, accounting for thrombin binding to compounds like proflavin (Sonder and Fenton, 1984). Thrombin transforms soluble fibrinogen to insoluble fibrin by releasing fibrinopeptides A and B from fibrinogen (Blomback and Blomback, 1972). Fibrin 15 Vessel Injury _ Inhibition 10 VIla/TF Complex IX IX @ XIa X ® Xa Va _-oc::::------~) Thrombin prothromC 0 ~ CD Thrombin FibrinOg~ Fibrin Fig. 2.3 Coagulation cascade hypothesised by Broze (1992). Exposure of tissue factor (TF) to FVlla initiates coagulation (A). The VllalTF complex produces small amounts of factors IXa and Xa (B), until tissue factor pathway inhibitor (TFPI) inhibits this process (C). From this point only the action of factors IXa and Villa (0) can generate additional factor Xa (E) to maintain coagulation. Thrombin activation of factor XI (F) and Xla autoactivation (G) may produce additional factor IXa (H). 16 formation and subsequent lysis is illustrated in Fig. 2.4. Following release of fibrinopeptides A and B, fibrin dimerisation occurs. Factor XlIla crosslinks a dimer with another monomer to enhance resistance of the thrombus towards plasmin degradation. Lengthening of the polymer occurs in a progressive, half-overlap, side-to-side interaction of monomers producing broad sheets of fibrin (Hermans and McDonagh, 1982). Lysis of the thrombus is discussed in Section 2.2.3. Thrombin activates platelets by binding to its receptor on the platelet membrane (Coughlin, 1993) and is also responsible for activation of factor XIII, which cross- links fibrin threads. Furthermore, thrombin promotes its own production by activation of factors V and VIII. Simultaneously, it downregulates its own production by binding to thrombomodulin on the endothelium surface to activate the coagulation inhibitor, protein C (Esmon et aI, 1982). Activated protein C, requiring protein S as a co-factor, inhibits thrombin production by inhibiting factors Va and Villa (Scully, 1992). Thrombin thus interacts with cells as well as with circulating proteins. All these interactions require the anion-binding site, while the catalytic domain of thrombin is responsible for its activational function (Arnaud et al, 1994). Thrombin plays a central role in co-ordinating the molecular and cellular interactions essential for vascular lesion formation. Firstly, thrombin is the principal mediator of thrombus formation at in vivo sites of vascular injury. Secondly, thrombin is a potent growth factor that stimulates proliferation of vascular smooth muscle cells at vascular injury sites in vivo. Thirdly, thrombin controls the effects of other growth factors. Fourthly, thrombin regulates inflammatory processes in blood leukocytes and vascular vessel cells. On this basis, Harker et al (1995) suggested that inhibition of thrombin or its receptor function would ultimately lead to disruption of both thrombus formation and vascular lesion formation. 17 ®-+-® Fibrinogen ~Thrombin @-- Fig. 2.6 A model showing the molecular link between coagulation and fibrinolysis. The thrombin-thrombomodulin complex activates both protein C and TAFI, which down-regulates the production of thrombin and plasmin, respectively (Nesheim et al, 1997). 22 complex with plasminogen. This complex is capable of converting plasminogen to plasmin (Collen et aI, 1993). Streptokinase has a low affinity for fibrin (Wohl et aI, 1978), while staphylokinse is highly fibrin-specific (Collen and Lijnen, 1994). 2.3 NOVEL ANTICOAGULANT AND ANTITHROMBOTIC AGENTS: The principal in vivo inhibitor of thrombin is antithrombin III (AT Ill), which requires hepariniods to enhance its inactivation of thrombin. When heparin binds to AT Ill, it causes a conformational change to enhance its accessibility to bind to thrombin (Rosenburg and Damus, 1973). Heparin is widely used as an anticoagulant, but it has several disadvantages. Heparin does not inhibit thrombin-induced platelet activation and it can be inactivated by platelet factor 4, which is released from activated platelets in the vicinity of the thrombus (Harker, 1994). In addition, the heparin-antithrombin III complex cannot inhibit thrombin bound to a thrombus, most likely because it is too large to diffuse into the thrombus (Hogg and Jackson, 1989; Weitz et aI, 1990). Furthermore, the efficiency of heparin therapy diversifies within and amongst patients and thrombocytopenia develops in 5-10% of patients (Bates, 1997). As a result, we have witnessed a marked increase in the development of new and more potent anticoagulant and antiplatelet agents in the past decade. In this regard, two major avenues have been followed. The first involved coagulation and initially thrombin was targeted. Subsequent developments include antagonists for factor Xa, factor Vila and tissue factor. The other avenue targeted blood platelets, in particular the Gpllb/llla receptor on the platelet membrane. In the case where major pathological abnormalitiés are related to platelet deposition and not so much to fibrin formation, it may be more desirable to rely on antiplatelet agents than on antithrombins (Caller et aI, 1991). As early as the 1950's antithrombotic agents were identified and isolated from blood sucking animals like ticks, assassin bugs, leeches and horse flies (Markwardt, 1994). It was, however, only during the past ten to fifteen years that such anticoagulant agents could be fully characterised using recombinant DNA 23 technology. This literature review will focus on same of the antiplatelet and antithrombotic agents found in nature, as well as on synthetic agents developed from these native haemostatic agents. 2.3.1 Inhibition of platelet aggregation: Some antiplatelet agents prevent the action of a single aggregatory agonist, without affecting the activity of other agonists. For example, aspirin prevents formation of TxA2, but does not effect platelet aggregation stimulated by other agonists such as ADP or thrombin (Gates et aI, 1988). Fibrinogen binding to Gp Ilb/llla is a mutual event during platelet aggregation and is independent of the initiating stimulus. Inhibition of fibrinogen binding to Gp lib/Ilia is thus a more effective pharmacological approach than prevention of the activity of individual agonists (Foster et al, 1994). The murine monoclonal antibody, 7E3, was the first platelet Gp lib/Ilia antagonist to be developed (Caller, 1985). It completely inhibited in vitro platelet aggregation and prevented thrombosis in animal models of arterial thrombosis and thrombolysis (Hanson et aI, 1988; Gold et aI, 1988). The high immunogenicity of the antibody led to the development of a derivative product, a chimaeric monoclonal 7E3 Fab, via recombinant technology. This hybrid molecule consists of mouse-derived variable regions of the original molecule linked to the constant region of human immunoglobulin IgG (Caller et aI, 1989). The primary mechanism of c7E3 Fab is to block the Gp IIb/llla receptor on activated platelet membranes to prevent adhesive ligands such as fibrinogen and VWF to bind to it. This prevents platelet aggregation and ultimately leads to impaired thrombus formation. Christopaulos et al (1993) observed with flow cytometry that c7E3 Fab can move from one platelet to another and can thus be redistributed to new platelets entering the circulation. Additionally, Gp IIb/llla receptors have been shown to be involved in platelet adhesion and platelet spreading (Weiss et aI, 1986) and inhibition can result in a decrease in the release of granular contents due to a lower level of platelet activation. This may in turn result in a decrease in the local concentration of the inhibitors of fibrinolysis, 24 plasminogen activator inhibitor-1 (PAI-1) and u2-plasma inhibitor (u2-PI), which are both released from platelets (CoIIer, 1996). Another profibrinolytic effect of c7E3 Fab may evolve from its prevention of factor Xllla to bind to platelets (Cox and Devine, 1994). In addition, factor Xllla cross-linking of fibrin enhances its resistance to fibrinolysis and factor Xllla also mediates cross-linking of u2-PI to fibrin. Furthermore, reducing the number of platelets in a thrombus, due to inhibition of Gp Ilb/llla, could decrease the local availability of platelet factor XIII. Factor Xllla cross-linking reactions would thus be reduced even further (CoIIer, 1997). Therefore, c7E3 Fab has the potential to serve as both an antithrombotic and a profibrinolytic drug. Another murine monoclonal antibody, MA-16N7C2, inhibits Gp lib/Ilia function and is the first antibody described with an echistatin-like RGD-containing sequence in the CDR3-region of its heavy chain (Deckmyn et aI, 1994). The antibody also recognises the Gp lib/Ilia complex on resting platelets, but platelet aggregation accelerates binding and increases its affinity for the complex. The antiplatelet activity of MA-16N7C2 was confirmed in a baboon model of platelet-dependent arterial thrombosis. The effects were dose-dependent and long lasting, suggesting that MA-16N7C2 is a potent and long-acting Gp Ilb/llla inhibitor (Kotzé et aI, 1995). In order to produce more powerful fibrinogen receptor antagonists, Foster et al (1994) studied the inhibition of platelet aggregation by non-peptidic compounds. They were constructed by replacing the Arg-Gly of RGD by alkyl chains of varying lenghts. The most potent in vitro compound was GR91669 {8- [aminoiminomethyl)thio]-L-aspartyl-L-phenylalanine}. In in vitro studies GR91669 inhibited marmoset platelet aggregation in whole blood, similarly to inhibition of platelet aggregation in human whole blood. Marmosets were thus chosen as the animal model for ex vivo studies. These studies showed that GR91669 inhibited platelet aggregation significantly, but reversibly over a period of a few hours. The short half-life of GR91669 and the lack of activity when taken orally of other peptidic fibrinogen receptor agonists, accentuate the potential of non-peptidic compounds (Foster et aI, 1994). 25 The physical characteristics of novel antiplatelet agents found in nature are summarised in Table 2.1. Disintegrins isolated from snake venoms represent a new class of low molecular weight RGD-containing peptides (Niewiarowski et aI, 1990). A novel platelet aggregation inhibitor, contortrostatin, was isolated from southern copperhead snake venom and it has an apparent molecular weight of 9 kDa (Trikha et al, 1990). Dennis et al (1990a) described the purification, complete amino acid sequence and biological activity of several snake venom proteins, which are potent Gp lib/Ilia antagonists and inhibitors of platelet aggregation. These proteins are kistrin from Agkistrodon rhodostoma, bitan from Bitis arietans, three isoforms of trigramin from Trimeresusus gramineus and an isoform of echistatin from Echis carinatus. All four peptides consist of between 47 and 83 residues and were able to inhibit platelet aggregation significantly (Dennis et aI, 1990a). Kistrin contains six intramolecular disulfide bonds and it binds reversibly to Gp Ilb/llla in nanomolar concentrations (Dennis et al, 1990a). Yasuda and Gold (1991) used kistrin in conjunction with a recombinant tissue-type plasminogen activator in a canine model of coronary artery thrombosis. Kistrin increased both the rate and extent of thrombolysis and prevented reocclusion. Shebuski et al (1990) successfully used echistatin in an animal model of thrombosis. Musial et al (1990) compared the action of RGDS and four disintegrins from viper venoms (echistatin, flavoridin, albolabrin and bitistatin). Their results confirmed that disintegrins are potential candidates for antiplatelet agents (Musial et aI, 1990). Savage et al (1990) studied the platelet-binding characteristics of two snake venom-derived proteins, applaggin and echistatin, from Agkistrodon piscivorus piscivorus and Echis carinatus, respectively. Applaggin is a disulfide-linked homodimer and its RGD-motif is situated at residues 50-52, while echistatin is a single chain, with its RGD-motif at residues 24-26. Both proteins were able to inhibit platelet secretion and aggregation of platelets stimulated by ADP, collagen and human thrombin. A monoclonal antibody LJ-CP3, which inhibits binding of Table 2.1: Summary of antiplatelet proteins. ---- Name Species Size Reference Expression host Reference Accutin Agkistro don 47 residues Yeh et al (1998a) - - acutus 7 Cys 5.2 kDa Albolabrin Trimeres urus 73 residues Musial et al (1990) - - albolabris Applaggin Agkistrodon 72 residues Savage et al - - piscivorus 17.7 kDa (1990) piscivorus Barbourin Sistrurus m. 73 residues Scarborough et al - - barbouri 12 Cys (1991 ) Basiiiein Crotalus basilicus 72 residues Scarborough et al - - (1993a) Bitan Bitis arietans 83 residues Dennis et al - - (1990a) Bitistatin Bitis arietans 84 residues Musial et al (1990) - - 9 kDa Calin Hirudo Munro eta/(1991) - - medicinales 65 kDa Cerastin Crotalus cerastes 73 residues Scarborough et al - - cerastes (1993a) Cereberin Crotalus viridis 72 residues Scarborough et al - - cereberus (1993a) Concortrostatin Southern 70-80 residues Trika et al (1990) - - copperhead snake 9 kDa Cotiarin Bothrops cotiara 72 residues Scarborough et al - - (1993a) N (J) Table 2.1 Continued Name Species Size Reference Expression host Reference Crotatroxin Crotalus atrox 72 residues Scarborough et a/ - - (1993a) Decorsin Macrobdella 39 residues Krezel eta/(1994) - - decora 4.4 kDa Durissin Crotalus durissus 72 residues Scarborough et a/ - - durissus (1993a) Echistatin Echis carinatus 48 residues Dennis et a/ - - 5.4 kDa (1990a) Flavoridin Trimeresurus 70 residues Musial et a/ (1990) - - flavoviridis Flavostatin Trimeresurus 68 residues Maruyama et a/ - - flavoviridis 12 Cys (1997) Hookworm Ancy/ostoma Chadderdon and platelet inhibitor caninum 15 kDa Cappello (1999) Jararacin Bothrops jararaca 73 residues Scarborough et a/ - - (1993a) Kistrin Agkis trodon 68 residues Dennis et a/ - - rhodostoma 12 Cys (1990a) I , Lachesin Lachesis mutus 73 residues Scarborough et a/ - - (1993a) I Leech antiplatelet Haementeria Connolly et a/ Yeast Keiler et a/ (1992) protein officina/is 16 kDa (1992) J Lutosin Crotalus viridis 73 residues Scarborough et a/ - - /utosus (1993a)_ I Molossin Crotalus tnotoesus 73 residues Scarborough et a/ - - I motossus (1993a) N --J Table 2.1 Continued Name Species Size Reference Expression host Reference Moubatin Ornithodoros Waxman and - - moubata 17 kDa Connolly (1993) Pallidipin Triatoma Noeske-Jungblut Baby hamster Noeske-Jungblut pallidipensis 19 kDa et a/ (1994) kidney cells eta/(1994) Trigramin Trimeresusus 72 residues Dennis et a/ - - gramineus 7.5 kDa (1990a) Viridin Crotalus viridis 71 residues Scarborough et a/ - - viridis (1993a) N CX) 29 RGD-containing proteins to Gp lib/Ilia, also prevented applaggin binding to platelets. These findings confirm that the binding of the two venom proteins are mediated by an RGD-motif (Savage et aI, 1990). Scarborough et al (1993a) characterized the disintegrin specificities of eleven different disintegrins isolated from the venoms of the American pit viper genera Bothrops, Crotalus and Lachesis. All disintegrins consisted of 71-73 amino acids, containing twelve highly conserved cysteine residues. These disintegrins all share a high homology of primary amino acid sequence with other peptides in this family of 71-73 residue integrins, such as trigramin (Huang et aI, 1987; Huang et aI, 1989), albolabrin (Musial et al, 1990), kistrin (Dennis et al, 1990a), applaggin (Chao et al, 1989) and flavoridin (Musial et al, 1990). The eleven disintegrins studied by Scarborough et al (1993a) could be organized in two distinct groups in terms of structure and function. The first group consisted of cerastin, lutosin, crotatroxin and durissin and they had only four amino acid differences in their primary amino acid sequence. All members had identical C- terminal sequences from residues 50-73 and all had an RGDW-sequence at residues 51-54. This group was more potent in preventing fibrinogen binding to Gp lib/Ilia than inhibiting the binding of vitronectin to aV~3. The second group included molossin, viridin, cereberin, basilicin, lachesin, jararacin and cotiarin. In contrast to the first group, the disintegrin activity of the second group resided in the RGDNP-sequence at residues 51-55 and it was more effective in inhibition of vitronectin binding to aV~3 than fibrinogen binding to Gp II/lila. These results indicated that the amino acid sequence immediately downstream from the RGD- sequence play a crucial role in determining integrin specificity and affinity (Scarborough et al, 1993a). Yeh et al (1998a) isolated a new disintegrin, accutin, from Agkistrodon acutus venom, which contains an RGD-sequence and seven Cys residues at positions highly homologous to other disintegrins. Although accutin did not affect the change in platelet shape caused by thrombin, ADP, collagen or U46619 activation, it did inhibit platelet aggregation stimulated by these agonists. Furthermore, it 30 prevented binding of the monoclonal antibody 7E3 to activated platelets. Accutin thus belongs to the short chain disintegrin family, acting specifically on the binding epitape of lib/Ilia, overlapping with that of 7E3 and blacks the receptor for fibrinogen binding (Yeh et aI, 1998a). In addition, accutin also inhibited binding of 7E3 to integrin av~3, its receptor on human umbilical vein endothelial cells. Yeh et al (1998b) studied the effect of accutin on the binding of other anti-integrin monoclonal antibodies to several other receptors, such as all~1, alll~1 and aV~1 on human umbilical vein endothelial cells, but no inhibition was detected. Interestingly, accutin had an unexpected in vivo anti-angiogenic effect on ten day old chick embryo cells and it induced apoptotic DNA fragmentation in human umbilical vein endothelial cells. These characteristics give accutin great potential in the field of antimetastatie agents (Yeh et al, 1998). Scarborough et al (1991) isolated a novel disintegrin, barbourin, from the venom of Sistrurus m. barbouri. In contrast to all other disintegrins, barbourin inhibited Gp II/lila binding via a KGD-sequence instead of the RGD-sequence. Scarborough et al (1993b) incorporated the KGD-motif on cyclic peptides and optimized the peptides in terms of cyclic ring size, hydrophobic binding site interactions and Iysyl side chain function. An optimal display of KGD on cyclic peptides resulted in a high affinity and selectivity for Gp Ilb/llla, which were virtually identical to that of barbourin. This study demonstrated that the specificity and potency of disintegrins could successfully be mimicked by small cyclic peptides (Scarborough et aI, 1993b). This synthetic cyclic KGD-heptapeptide, with high affinity and specificity for the Gp lib/Ilia integrin, was named integrelin. Binding of integrelin inhibits platelet aggregation and prevents thrombosis. It is a rapid-acting and highly potent agent with a short half-life (Charo et al, 1992). Tcheng et al (1995) performed the first clinical trial of integrelin during a routine, elective, low- and high-risk coronary intervention study. Profound and sustained inhibition of platelet function was achieved (Tcheng ef al, 1995). Flavostatin is a novel 68 amino acid disintegrin found in the venom of Trimeresurus flavoviridis (Maruyama et al, 1997). It contains an RGD-sequence 31 and twelve conserved Cys residues. Flavostatin effectively inhibited ADP, collagen and thrombin receptor activating peptide (TRAP)-induced platelet aggregation (Maruyama et al, 1997). Decorsin originates from the leech Macrobdella decora. It has a single RGD- sequence, which is situated at the apex of an extended loop (Krezel et aI, 1994). This structural positioning of the RGD-sequence is commonly found amongst disintegrins like kistrin (Adler et al, 1991), echistatin and flavoridin (Dennis et al, 1993). Decorsin is far more effective in preventing platelet aggregation than the pentapeptide GRGDV. Although it contains an RGD-sequence, decorsin shows only approximately 16% amino acid sequence similarity with other snake venom Gp lib/Ilia antagonists. Decorsin was the first Gp lib/Ilia antagonist and inhibitor of platelet aggregation isolated from leeches (Seymour et aI, 1990). Connolly et al (1992) described a protein that inhibited collagen-induced platelet aggregation. The protein was isolated from Haementeria officinalis and the purified protein was designated leech antiplatelet protein (LAPP). According to Keiler et al (1992), LAPP prevented platelet adhesion to collagen, but had no effect on aggregation when platelets were stimulated by ADP, thrombin, arachidonic acid, U46619 or A23187 (Connolly et aI, 1992). The gene encoding LAPP was cloned and expressed in yeast (Keiler et aI, 1992), enabling in vivo studies in an animal model of thrombosis (Schaffer et aI, 1993). rLAPP was able to completely inhibit collagen-mediated platelet aggregation, but had no significant effect on the rate and extent of platelet deposition on a collagen surface. In contrast, a peptidyl fibrinogen receptor antagonist, L-366763 (acetylated-Cys-Asn- Pro-Arq-Gly-Asp-Cys-Nl-l-). completely prevented platelet deposition at the same dosage that inhibited ex vivo aggregation. These results demonstrated that inhibition of collagen-induced platelet aggregation by rLAPP alone was not sufficient to prevent platelet-dependent thrombosis in the animal model studied and that other mechanisms are crucial for the development of thrombosis (Schaffer et al, 1993). 32 The medicinal leech, Hirudo medicinales, became famous for the production of hirudin, the most potent known thrombin inhibitor in nature (Markwardt, 1970). Interestingly, Munro et al (1991) isolated an inhibitor of collagen-mediated platelet adhesion and aggregation from the same leech and the inhibitor was named calin. It inhibits aggregation by rapidly (1-10 min) binding to collagen. However, no cleavage of collagen occurs as in the case of collagenases. Calm's rapid interaction with collagen may explain the prolonged bleeding phenomenon seen after leech bites (Munro et aI, 1991). Waxman and Connolly (1993) purified an antiplatelet protein from yet another blood-sucking organism, the soft tick, Ornithodoros moubata. The protein was called moubatin and interfered in haemostasis by preventing collagen-stimulated platelet aggregation. The blood-sucking bug, Triatoma pallidipensis, also produces an inhibitor of collagen-induced platelet aggregation, called pallidipin (Noeske-Jungblut et aI, 1994). Interestingly, it had no effect on platelet adhesion to collagen, but inhibited the release of ADP from platelets. No inhibition of aggregation in response to ADP, thrombin, TxA2, mimetic U44619 or phorbol ester was observed. Its gene was cloned from a cDNA library and the recombinant product was produced in baby hamster kidney cells. Recombinant pallidipin had antiplatelet effects identical to those of the native inhibitor (Noeske-Jungblut et aI, 1994). Recently, Chadderdon and Cappella (1999) isolated an inhibitor of platelet aggregation and adhesion from adult Ancylostoma caninum hookwarms. The protein of approximately 15 kDa was named hookworm platelet inhibitor (HPI). HPI blocked platelet aggregation in response to epinephrine, thrombin and ADP. Furthermore, it also inhibited the binding of resting platelets to immobilized fibrinogen and collagen, suggesting interactions with Gp lib/Ilia and Gp la/lla. Monoclonal antibodies were used to confirm blockade of cell surface integrins Gp lib/Ilia and Gp la/lla (Chadderdon and Cappella, 1999). It is clear that there are a vast number of products that can inhibit platelet- dependent thrombogenesis by preventing the binding of inhibitors to Gp lib/Ilia. Integrelin and c7E3 Fab have been tested in clinical studies where their efficiency 33 was proven. Results document that Gp lib/Ilia blockade reduces the incidence of clinically significant ischemic events in the entire spectrum of patients undergoing coronary intervention. It is proposed that in the cases where major pathological abnormalities are related to platelet deposition and not so much to fibrin formation, it may be more desirable to rely on antiplatelet agents than on antithrombins (Caller et aI, 1991). 2.3.2 Direct inhibition of thrombin: The development of direct and indirect thrombin inhibitors has led to a new dimension in the management of thrombotic and vascular disorders and had immense implications on drug development research (Fareed et aI, 1999). De Simone et al (1998) schematically compared the binding of fibrinogen to thrombin and the binding of the different classes of direct thrombin inhibitors to thrombin (Fig. 2.7). Table 2.2 summarises the physical characteristics of the direct antithrombins. More detailed descriptions of the design and mode of action of the different direct thrombin inhibitors will follow in the text. Hirudin is produced by the salivary glands of the leech Hirudo medicinales and is the most potent and specific thrombin inhibitor found in nature (Markwardt, 1970). The molecule is a single carbohydrate-free polypeptide and is stabilised by three intramolecular disulphide bridges. It contains a sulphated tyrosine on position 63. Unlike heparin, hirudin does not require any cofactors for its anticoagulant activity. Hirudin is composed of a cysteine-rich amino terminus and an acidic carboxy- terminus. Inhibition of thrombin is accomplished by the formation of a tight 1:1 stoichiometric complex.' The 48-amino acid globular N-terminus, which is stabilized by three disulphide bridges, binds within the active site of thrombin (Rydel et aI, 1990; Rydel et aI, 1991). The C-terminus, which is a highly acidic region and also contains a sulphated tyrosine at position 63, binds through ionic and hydrophobic interactions to the ABE of thrombin (Rydel et aI, 1990). Hirudin binds to thrombin in a non-substrate mode (Fig. 2.7F), with the N-terminal tail of hirudin parallel to thrombin segment Se~14_Glu217(Stone and Hofsteenge, 1986; Grutter et aI, 1990; Rydel et aI, 1990, Markwardt, 1994). This results in inhibition of both the proteolytic and cellular activities of the enzyme (Fenton, 1989). 34 A Fibrinogen c: I~I' I Exo,,,, 1.::=1 =====..l=~r-ï I 1'·5,,214 C·GI~216. Active SIle:Tnrornbin A'p189 B Substrate-lice Inhibuor C Non-substrate-lice InhibitorC N ""-u-o"-,, ---. r- .--_---,N c:_1 Exosue 11 N·5er214 C.GI)·216 N.5er214 C·Gly216 ACtive site: Active sue Tarorntnn A'p189 Thrombin Aspl89 o E Erosue Inhibitor Fibnnogen-Iike Inhibitor C I' r- Cl L-2;Exosue 1 i Erosue 1 Ir- N·5er214 C·Gly216 N.5er214. C·Gly216 Active Site: I Active SHe:Thmmbin Asp 189 Thrombin Asp 189 F I H'.": I G ,c Hirunonn IIW I ~rl U=OSi'=' ~IW)r- Exosue Ir- I'·5er214 C·Gly216 I N·5er214 .' C·Gly216Active site Active: siteTnrombin Asp 189 Thrombin Aspl89 Fig. 2.7 Schematic summary of the different interaction mechanisms between thrombin and its inhibitors (De Simone et aI, 1998). A: Interaction of fibrinogen with ABE and the active site of thrombin. B: Peptide-based inhibitors like PPACK interact with the active site in a substrate mode by aligning its backbone (like fibrinogen) in an antiparallel manner to thrombin segment Se~14_Glu217. C: Some active site-directed peptidic inhibitors also interact in a nonsubstrate mode and align their backbone in parallel fashion to thrombin segment Ser214_Glu217e,g. BMS-183507 (Iwanowich et aI, 1994). D: ABE-directed inhibitors interact merely with the fibrinogen recognition sequence of thrombin. They include hirugen and hirugen-related peptides, hirudin54-65and hirullin50-62. E: Multisite- directed inhibitors bind to both the active site and the ABE in an antiparallel manner and are hirulogs, hirulog derivatives and hirutonins. F: Hirudin interacts with the ABE and with the active site in a parallel fashion. G: Hirunorms bind in a similar manner to thrombin than hirudin. Table 2.2: Summary of direct antithrombins. Name Species Size Reference Expression host Reference Bifrudin Hirudinaria Electricwala et al - - manillensis 7 kDa (1991 ) Bothrojaracin Bothrops jararaca Zingali et al COS cells Arocas et al 27 kDa (1993) (1997) I Haemadin Haemadipsa 57 residues Stube et al (1993) Escherichia coli Stube eta/(1993) I sylvestris 5 kDa Hirudin Hirudo 65 residues Markwardt (1970) Escherichia coli Dodt et al (1986) medicinales 6 Cys 7 kDa Hirullin P18 Hirudinaria 61 residues Krstenansky et al - - manillensis (1990) Rhodnin Rhodnius prolixus 103 residues Friedrich et al Escherichia coli Friedrich et al 11 kDa (1993) (1993) L.... CJ.) (Jl 36 Two unique characteristics distinguish hirudin from conventional protease inhibitors. Firstly, most serine protease inhibitors contain a reactive site for interaction with the active site of the target enzyme. In contrast, none of the three lysine residues of hirudin is involved in such an interaction (Braun et aI, 1988). Secondly, hirudin contains a compact N-terminus, but a disordered C-terminus, in contrast to most other serine protease inhibitors, which are compact molecules (Folkers et al, 1989) Unlike native hirudin, recombinant hirudin lacks a sulphate group on Tyr63 (Oodt et aI, 1984; Oodt et aI, 1986). As a result, the desulphonated compound has a tenfold-reduced affinity for a-thrombin (Stone and Hofsteenge, 1986). The affinity of r-hirudin for thrombin could, however, be restored to equivalent levels to that of wild type hirudin by introducing phosphotyrosine into position 63 (Hofsteenge et aI, 1990). A negatively charged Tyr63 thus plays a substantial role in determining the affinity of hirudin for thrombin. Over the last decade, several groups investigated the activity of different hirudin fragments. Schmitz et al (1991) found that the N-terminal fragment of hirudin (Hir" 47) inhibited all enzymatic functions of thrombin. Krstenansky and Mao (1987) studied a chemically synthesized unsulphated NU_acetyl_hirudin45-6(5Hir45-65). This fragment was able to inhibit blood coagulation and the release of fibrinopeptide A by thrombin, but was unable to inhibit the amidolytic activity of thrombin. This indicates that the C-terminus of hirudin occupies the ABE, which was later confirmed by Oodt et al (1990). Schmitz et al (1991) showed that Hir45-65inhibits the interaction between thrombin and thrombomodulin. This results in inhibition of the activation of protein C, activation of platelets and endothelial cells, and interactions between thrombin and factors Vand VIII. The binding of hirudin to thrombin does not inhibit binding of antithrombin III and thrombin, since antithrombin III does not recognize the anion-binding site (Dennis et aI, 1990b). Mao et al (1988) determined that the shortest C-terminal fragment with inhibitory activity was Hir56-65w, hile maximum activity was obtained with Hir54-65.The binding of thrombin exosite inhibitors, like Hir54-65,to thrombin is presented in Fig. 70 (Banner and Hadvary, 1991; Stubbs et al, 1992; Priestle et al, 1993). 37 Phe56 appeared to be crucial for maintaining activity, since replacement with Glu or Leu lead to complete loss of inhibitory activity of hirudin and its fragments. Even when Phe56 was replaced with D-Phe to determine conformational requirements, inhibition was also completely lost. Circular dishroism spectra showed that binding of hirudin C-terminal peptides to thrombin lead to significant conformational changes and that loss of thrombin activity might be due to a lack of conformational change in the case of D-Phe (Mao et al, 1988). The leech Hirudinaria manillensis belongs to the same family as Hirudo medicina/es and it produces an antithrombin similar to hirudin, called bifrudin (Electricwala et al, 1991). The antithrombotic effect of bufrudin in human plasma had a potency similar to hirudin variant I (Dodt et al, 1984) at equivalent dosage. It did, however, not cross-react with monoclonal antibodies directed against recombinant hirudin variant 1. Amino acid sequence analysis showed four differences up to residue 25 between the two antithrombins. These results indicate that bufridin is a potent thrombin inhibitor with biological activity similar to hirudin, but it differs in terms of structural and immunological properties (Electricwala et al, 1991). Naski et a/ (1990) studied the effect of a synthetic N-acetylated C-terminus of hirudin, called hirugen. They found that hirugen [Ac-Asn-Gly-Asp-Phe-Glu-Glu-Ile- Pro-Glu-Glu-Tyr(S03)Leu] competitively inhibited the action of a-thrombin towards fibrinogen, but with minimal inhibition of the active site. These results again illustrated the interaction between the C-terminus of hirudin and the anion-binding site of thrombin. The binding of hirugen as exosite inhibitor is shown in Fig. 2.70 (Krstenansky and Mao, 1987; Krstenansky et al, 1987; Jakubowski and Maraganore, 1990; Skrzypczak-Junken et al, 1991). /n vitro studies with hirugen revealed that it competitively inhibited fibrinogen cleavage and platelet activation by thrombin (Jakubowski et al, 1990; Cadroy et al, .1991). Hirugen prevented ex vivo platelet deposition in low-shear flow chambers connected to arteriovenous shunts in baboons. However, it failed to affect ex vivo platelet deposition on collagen type I-coated tubing (Cadroy et al, 1991). 38 Another hirudin variant produced by Hirudinaria manillensis was designated hirullin P18. Like hirudin, it has a highly acidic C-terminus. Phe51 is a crucial residue and corresponds to the important Phe56 of hirudin. The hirullin50-62 peptide 54 65 (SDFEEFSLDDIEQ) binds thrombin with similar affinity as unsulfated hirudin - (Krstenansky et ai, 1990). Crystallographic studies showed that the association of residues 51-55 of hirullin with thrombin is similar to that of hirudin and hirugen. The remaining residues also interact with and bind to thrombin, but binding is achieved through a conformational adjustment of the peptide with respect to the conformation of hirudin and hirugen. It causes the side group of IIe60of hirullin to point in the opposite direction of Tyr63 of hirudin and hirugen, but allows the residues to interact with the 310turn of the hydrophobic binding pocket of thrombin. The hydrophobic interaction is thus accomplished through a conformational readjustment (Qiu et aI, 1993). Maraganore et al (1990) designed a range of antithrombotic peptides called hirulogs. They consist of an active-site specific sequence with an Arg-Pro scissile bond, a polymeric linker of glycyl residues (6 to 18 A in length) and a sequence for binding to the ABE of thrombin. In contrast to synthetic hirudin C-terminal peptides, hirulogs inhibited both the fibrinogen clotting activity of thrombin and its proteolytic activity. Fig. 2.7E shows the binding of hirulogs to both the active site and the ABE of thrombin in an antiparallel manner (Maraganore et al, 1990; Skrzypczak-Jankun et al, 1991; Qiu et al, 1992). Hiru log-1 [(D-Phe )-Pro-Arg-Pro-( GIY)4-Asn-GIy-Asp-P he-G Iu-GIu-Ile-P ro-Glu-G Iu- Tyr-Leu] was capable of inhibiting thrombin catalyzed hydrolysis of p-nitroanilide at nanomolar concentrations (DiMaio et al, 1990; Maraganore et al, 1990). The optimal length of the oligoglycyl spaeer forming a molecular bridge between the active-site and the ABE sequence appeared to be at least three to four glycyl residues (Maraganore et al, 1990). Witting et al (1992b) studied the effect of hirulog-B2 on thrombin inhibition. Hirulog-B2 has D-cyclohexylalanine substituted in the first position and is highly specific for binding to the active site and to the ABE of thrombin and could have strong pharmaceutical potential. The efficiency of hirulogs as antithrombotic drugs was clearly demonstrated in animal models of both venous and arterial thrombosis (Maraganore and Adelman, 1996). The slow 39 cleavage of the Pro-Arg bond by thrombin is however a major limitation. The direct thrombin inhibition by hirulogs is therefore only temporary and in time, hirulogs are converted into less potent hirugen-like peptides (Witting et aI, 1992a). A new class of thrombin inhibitors, called hirunorms, was designed to mimic the binding mode of hirudin, without the undesirable substrate-like features of hirulogs. Hirunorm V is a synthetic polypeptide of 26 residues and was constructed to interact with the active site of thrombin through its N-terminus. Its C-terminal domain should bind the ABE of thrombin (Lombardi et aI, 1996). The two binding regions are divided by a spaeer segment of the appropriate length to allow interaction with both the active site and ABE. The interaction between hirunorm and thrombin is shown in Fig. 2.7G (De Simone et aI, 1998). The molecular structure of hirunorms makes them resistant to proteolysis by thrombin. Hirunorm V showed high antithrombotic activity in different rat models and compared well with hirudin in inhibiting arterial thrombosis (Lombardi et aI, 1996). The primary structure of hirunorm V is presented and compared to the amino acid sequence of hirudin variant 2 (Harvey et aI, 1986) in Table 2.3. Additionally, the activities of other proteases such as plasmin, tPA and trypsin are not affected by hirunorms (Cirillo et aI, 1996). The hirunorms are also insensitive to plasma proteases and are readily inactivated by rat liver or kidney extracts (Cirillo et aI, 1996). X-ray crystal structure analysis of hirunorm V showed that its C-terminus interacts with the ABE, similarly to the interaction between hirudin and thrombin (De Simone et al, 1998). These crystallographic studies confirm the accuracy of the hirunorm design and provide convincing proof of its interaction as a synthetic thrombin inhibitor with high potency. In the inhibition of thrombin by hirudin-derivatives, the focus is on the segments inhibiting the active site and ABE. The role of the linker (residues 49-54) was limited to provision of an appropriate spaeer between the two inhibitory segments. None of the side chains of the linker contributed significantly to binding and could be replaced without considerably affecting the activity of hirudin (Yue et aI, 1992). 40 Table 2,3: Primary structure of hirunorm Vand its comparison with hirudin variant 2 (De Simone et al, 1998). Hirudin Hirunorm V lie 1 Chg1 Thr ValL j j Tyr 2-Nal Thr4 Thr4 Asp" Asp" I D-Alao Globular domain Gly' I p-Ala8 Pro"o Pro~ Glu4~ Glu1u Serou Ser11 Hiso1 His 1L Asn'" h-Phe lj AsnOj GI/4 Gly::>'! Gly'::> Asp::>::> Asp" Pheoo Tyrlf Glu::>' Glulo Glu::>o Glul~ lIe::>~ IleLu Proou ProL1 Gluol Aiba Glu bL AibLj Tyro.) Tyr" Leu04 Cha'" Glnb!:l D-Glu'::o 41 In complex with a-thrombin, the linker was mainly exposed to the solvent-outside of a deep groove between the active site and the ABE. The linker had exceptional conformational flexibility, but no interaction with thrombin (Yue et ai, 1992). f\f- Acetyl[D_Phe45,Arg47]hirudin45-65(P53) is a bivalent thrombin inhibitor with a Kj- value of 5.6 nM (DiMaio et ai, 1990). It consists of an active site inhibitor segment, [f\f-Acetyl-(dF)PRP], an ABE inhibitor segment, hirudin55-65(DFEEIPEEYLQ-OH), and a connecting linker, hirudin49-54(QSHNDG). Szewczuk et al (1993) used P53 as a model to study the structure-function relationship of the connecting linker. They used eo-amino acids, which modified the length of the linker as well as the number and location of the peptide bonds. The length of the linker was calculated in terms of the number of atoms participating in the formation of the backbone. No side chains were included in the synthesis of the linker. A minimum of eleven atoms was required to bridge the two inhibitory segments, but the potency of this linker was weak (Kj-value of 26 nM). An inhibitor with a 13-atom linker showed the highest potency with a Kj-value of 0.51 nM. Molecular dynamic simulation of inhibitors with a 13-atom linker suggested that the linker interacted with thrombin through hydrogen bonds. These inhibitors may be reclassified as trivalent rather than bivalent inhibitors (Szewczuk et ai, 1993). A schematic representation of its interaction with thrombin is shown in Fig.7E (Szewczuk et ai, 1993). Hirutonins belong to a family of bifunctional inhibitors containing a noncleavable moiety mimicking the scissile bond (Zdanov et aI, 1993). Hirutonin-2 is an analog of (D)Phe-Pro-Arg-Gly-hirudin49-65. The crystal structure of the thrombin-hirutonin- 2 complex shows that the C- and N-terminal segments are well ordered, while the linker is partially disordered. Hirutonin-6 has the same N-terminus as hirutonin-2, connected to a shortened ABE binding segment by a short non-peptidyl linker. The linker is situated near die bottom of the groove connecting the active site and the ABE, forming a short anti parallel ~-sheet-like organisation with Leu40_Leu41and interacting with Glu39_Leu4o_Leu41via van der Waals farces. The N-terminal fragment of the two hirutonins binds thrombin similarly to other inhibitors with this motif, while binding of the C-terminal fragments to the ABE are similar to that of hirudin and hirulogs. Interaction between thrombin and hirutonins are shown in Fig 2.7E (Zdanov et al, 1993). 42 A slow tight-binding thrombin inhibitor, haemadin, was isolated from the Indian leech, Haemadipsa sylvestris (Strube et aI, 1993). Haemadin is highly thrombin specific, with no inhibition of other proteases like trypsin, chymotrypsin, factor Xa and plasmin. The N-terminus of haemadin shows no sequence similarity with hirudin, but six of the twelve C-terminal residues are identical to those of hirudin. In contrast to hirudin, haemadin does lack a tyrosine residue in its C-terminus. The haemadin gene was cloned and expressed in E. coli. Recombinant haemadin displayed a similar inhibition constant and specific activity than native haemadin, suggesting that posttranslational modifications are not essential for its activity (Strube et al, 1993). Bothrojaracin is a potent thrombin inhibitor that was isolated from the venom of Bothrops jararaca. This unique thrombin inhibitor does not bind to the active site, but interacts with both anion-binding exosites 1 and 2 (Zingali et al, 1993). Bothrojaracin has two disulfide-linked polypeptides, chains A and B with molecular masses of 15 kDa and 13 kDa, respectively. In contrast to other ligands, which recognise exosite 1, bothrojaracin lacks an acidic sequence similar to the C- terminal tail of hirudin. Arocas et al (1997) used COS cells to express functional bothrojaracin, which was able to bind to and inhibit the function of thrombin. A highly specific thrombin inhibitor was isolated from the assassin bug, Rhodnius prolixus, and was named rhodniin (Friedrich et al, 1993). It displays an internal sequence homology of residues 6-48 with residues 57-101, indicating a two- domain structure. Rhodniin forms a 1:1 complex with thrombin. Domain 1 binds to the active site of thrombin, with His 10 pointing into the specificity pocket. The cDNA of rhodnin could be cloned and expression in E. coli yielded a potent recombinant thrombin inhibitor (Friedrich et al, 1993). D-Phenyl-L-Prolyl-L-Arginyl-chloromethylketone (PPACK) is a specific inhibitor of the catalytic site of thrombin. The effect is brought about by irreversible alkylation of the active site histidine (Kettner and Shaw, 1979). Bode et al (1989) did crystallographic studies to investigate the formalion of the stoichiometric complex between human a-thrombin and PPACK. They found that the exceptional 43 specificity of PPACK could be explained by a hydrophobic cage in thrombin, formed by lie174,Trp215,Leu99, His57, Tyr60A and Trp600. PPACK interacts with the active site of thrombin in a substrate mode by aligning its backbone in an antiparallel manner to thrombin segment Se~14_Glu217(Bode et al, 1989) as illustrated in Fig. 2.7B. This interaction is similar to that of fibrinogen, which is also in a substrate mode (Fig. 2.7A). Lumsden et a/ (1993) investigated the effect of PPACK on thrombosis following endarterectomy in baboons. Continuous intravenous infusion of PPACK (100 nmol/kg/min) permanently interrupted thrombosis, by irreversibly inactivating the thrombin that was bound to the thrombus. Infusion was started after surgical haemostasis was established to avoid abnormal surgical bleeding. These safe and lasting characteristics make PPACK an attractive candidate for use in mechanical interventional vascular procedures applied in the management of symptomatical arterial disease in humans (Lumsden et ai, 1993). 2.3.3 Indirect inhibition of thrombin: Thrombin generation is not affected by direct antithrombins and some thrombin molecules may escape inhibition. This can, however, be prevented by inhibition of factor Xa and thus disruption of the thrombin feedback loop. The physical characteristics of the indirect thrombin inhibitors are summarised in Table 2.4. Waxman et a/ (1990) isolated a factor Xa inhibitor form the soft tick, Ornithodoros moubata. The inhibitor has a single chain and was designated tick anticoagulant peptide (TAP). Interaction between TAP and factor Xa involves initial low-affinity binding to an exosite followed by high-affinity binding to the catalytic site of factor Xa (Vlasuk et al, 1991). TAP was superior to heparin in the prevention of venous thrombi in a rabbit model (Vlasuk et al, 1991) and also in preventing heparin- resistant platelet thrombi in a primate model of arteriovenous shunts grafted with collagen and endarterectomised carotids (Schaffer et al, 1991; Kotzé et al, 1997). Biologically active recombinant TAP (rTAP) was produced in the yeast Saccharomyces cerevisiae. Native TAP and rTAP have the same amino acid Table 2.4: Summary of indirect antithrombins. Name Species Size Reference Expression host Reference Annexin V Homo sapiens 319 residues Funakoshi et al Escherichia coli Thiagarajan and placenta 36.5 kDa (1987a) Benedict (1997) Antistasin Haementeria 119 residues Nutt et al (1988) Insect baculovirus Nutt et al (1991) officin alis 15 kDa Protein C Homo sapiens Esmon et al AV12 cell line Ehrlich et al 62 kDa (1982) (Syrian hamster (1989) tumor cells) Tick anticoagulant Omithodoros 60 residues Waxman etal Yeast Neeper et al peptide moubata 6.8 k Da (1990) (1990) ~ ~ 45 composition, primary structure, electrophoretic mobility and inhibition of factor Xa (Neeper et ai, 1990). Due to its small molecular size, TAP is a weak immunogen, which makes it an attractive option for the prevention and treatment of thrombosis (Verstraete and Zoldhelyi, 1995). The Mexican leech Haementeria officinalis produces a tight-binding and highly specific inhibitor of factor Xa, which was purified to homogeneity and called antistasin (Dunwiddie et al, 1993). Antistasin has a high cysteine content (Nutt et al, 1988). Recombinant antistasin was produced in an insect cell baculovirus host vector system (Nutt et aI, 1991). Antistasin has been studied thoroughly in various thrombosis models and found to be superior to heparin in preventing platelet-rich thrombi in dacron-grafted arteriovenous femoral grafts (Schaffer et ai, 1992). Clinical development of antistasin is, however, highly unlikely, because of its strong immunogenicity, (Verstraete and Zoldhelyi, 1995). A similar factor Xa inhibitor was isolated from Haementeria ghilianii (Condra et al, 1989). A factor Xa inhibitor was also found in the hookworm, Ancylostoma duodenalis (Capello et aI, 1993). Annexins are a family of calcium-dependent anionic-phospholipid-binding proteins (Barton et aI, 1991). Annexin V was originally isolated from the placenta and characterised as placental anticoagulant protein I (Funakoshi et aI, 1987a; Funakoshi et aI, 1987b). It binds to anionic phospholipid on the platelet membrane and prevents the binding of factors Xa and Va to platelets (Thiagarajan and Tait, 1990; Thiagarajan and Tait, 1991). Annexin V was used effectively as inhibitor of thrombosis in a venous thrombosis model (Romoisch et aI, 1991; Van Ryn McKenna et al, 1993; Van Heerde et al, 1994). Thiagarajan and Benedict (1997) successfully used recombinant annexin V as an inhibitor of arterial thrombosis in a rabbit carotid artery injury model. Intravenous infusion of annexin V significantly inhibited arterial thrombosis, without impairing the haemostatic response (Thiagarajan and Benedict, 1997). Protein C is a human coagulation factor inhibitor and is involved in the down- regulation of thrombin production. Thrombin, bound to thrombomodulin on the endothelium surface, activates protein C (Esmon et aI, 1982). Activated protein C 46 (APC) requires protein S as a co-factor and inhibits thrombin production by inhibiting coagulation factors Va and Villa (Scully, 1992). Gruber et al (1989) investigated the in vivo antithrombotic properties of APC in a baboon model of thrombus formation on prosthetic vascular grafts. Infusion of human APC led to inhibition of clotting as measured by aPTT and to reduction of vascular graft platelet deposition as determined by the real-time scintillation camera imaging of 111ln-labeled platelet deposition. Template bleeding times showed that haemostatic plug formation remained normal. Gruber et al (1990) proved in a primate model of arterial thrombosis that recombinant activated protein C (rAPC), like human plasma-derived APC, inhibited thrombus formation without impairing primary haemostasis. Hanson et al (1993) presented a novel and effective antithrombotic strategy by the infusion of thrombin to activate endogenous protein C in a baboon model. Gruber et al (1993) observed that APC levels are elevated during thrombolytic therapy and that this may help to prevent reocclusion during or after thrombolysis (Gruber et aI, 1993). In 1994 Gruber et al observed that APC indeed reduced the diameter and relative number of fibrin fibers in plasma clots. Protein C is thus also involved in enhancing the efficacy of thrombolysis (Gruber et al, 1994). In conclusion, there are many direct and indirect thrombin inhibitors that inhibit thrombin with different potency. One must always keep in mind that complete inhibition of thrombin can lead to the development of a bleeding tendency, which may in itself be as dangerous to the patient as thrombosis. It was therefore suggested that a mild to medium strength antithrombotic may be the best choice (Fenton, 1998). 2.3.4 Activation of fibrinolysis: Fibrinolysis is an essential process to dissolve fibrin clots and to prevent excessive fibrin formation that can disrupt normal blood flow. Several profibrinolytic proteins have been studied intensively to assess their potential as therapeutic agents. The physical characteristics of these profibrinolytic proteins are summarized in Table 2.5. Table 2.5: Summary of fibrinolytic proteins. Name Species Size Reference Expression host Reference Bat plasminogen Desmodus GardelI et al Escherichia coli Gardeli et al activator (Bat-PA) rotundus 49 kDa (1989) (1989) Hemetin Haementaria Malinconico et al - - i ghilianii 120 kDa (1984) ! Plasminogen Agkistrodon halys Park et al (1998) Baculovirus Park et al (1998) activator 32 kDa Staphylokinase Staphylococcus 136 residues Collen et al (1993) Escherichia coli Sako (1985) aureus 15.5 kDa Streptokinase Streptococci Cederholm- - - strains 48 kDa Williams et al (1979) Tissue-type Homo sapiens 530 residues Pennica et al Escherichia coli Pennica et al plasminogen 68 kDa (1983) (1983) activator Urokinase-type Homo sapiens 411 residues Husain et al Escherichia coli Zamarron et al plasminogen 54 kDa (1983) (1984) activator - --- - -- ~---~ .j::>. -....,J 48 Two bacterial plasminogen activators, streptokinase and staphylokinase, are currently intensively researched to determine their suitability as thrombolytic agents. Streptokinase, produced by Streptococci, forms a 1:1 stoichiometric complex with plasminogen, which is in turn capable of converting other plasminogen molecules to plasmin. The streptokinase-plasminogen complex is resistant to inhibition by circulating u2-antiplasmin (Cederholm-Williams et aI, 1979). Streptokinase causes temporary hypertension in many patients and significant allergic reactions in some patients. It is also highly antigenic, which makes repeated administration undesirable (Collen, 1997). Staphylokinase is produced by certain strains of Staphylococcus aureus. Like streptokinase, staphylokinase forms a 1:1 stoichiometric complex with plasminogen. Unlike the streptokinase-plasminogen complex, the staphylokinase- plasminogen complex is inactive and requires activation to staphylokinase-plasmin. Activation of the plasminogen part of the molecule results in exposure of the active site and conversion of the complex to an effective plasminogen activator (Collen et aI, 1993). In the absence of fibrin, the complex is rapidly degraded by U2- antiplasmin (Lijnen et al, 1991), which is also in contrast to the resistant streptokinase-plamin(ogen) complex. Staphylokinase molecules released from the complex can be recycled to interact with other plasminogen molecules (Silence et aI, 1993a). However, in the presence of fibrin, the inhibition rate is more than 100- fold reduced, making staphylokinase highly fibrin-specific (Lijnen et aI, 1992; Silence et aI, 1993b). Staphylokinase does not activate plasminogen in the absence of fibrin, probably due to a weak affinity between plasminogen and staphylokinase in plasma (Sakharov et aI, 1996). Furthermore, u2-antiplasmin inhibits the generation of active staphylokinase-plasmin complexes. Trace amounts of plasmin are found at the fibrin surface and it forms an active staphylokinase-plasminogen complex. This complex is bound to fibrin via lysine binding sites on the plasmin molecule and it is thus protected from rapid inhibition by u2-antiplasmin. Following digestion of the fibrin clot, the complex is released and inhibited. This leads to prevention of further plasminogen activation (Collen, 1997). 49 The tertiary structure of staphylokinase has been elucidated by Rabijns et al (1997). It has a flattened structure composed of a mixed five-stranded ~-sheet, packed on a single a-helix of twelve residues. Mutagenesis of staphylokinase by Silence et al (1995) showed that two pairs of charged amino acids (46:50 and 65:69) are crucial for its interaction with plasmin. These four amino acids map to the side of the molecule, comprising the a-helix. Functional staphylokinase:plasmin complex formation also depends on a cluster of charged amino acids at the N-terminus (residues 11-14), which extends from the same side as the two pairs of charged amino acids (Silence et aI, 1995). Furthermore, activation of plasminogen by the staphylokinase:plasmin complex requires processing of staphylokinase by plasmin. Ten N-terminal residues are removed by plasmin leading to the exposure of Lys 11 as the new amino terminus (Schlott et aI, 1997). Mutagenesis of the catalytic domain of plasmin showed that Arg719 mediates complex-formation and the plasminogen activation potential of staphylokinase:plasmin (Jespers et aI, 1998). This critical residue is situated in an extended loop at the western edge of the active-site cleft (Lamba et aI, 1996). Jespers et al (1999) used alanine-scanning mutagenensis and phage display to study the structural and functional basis of plasminogen activation by staphylokinase. For the purpose of this study, the staphylokinase.u-plasrnin complex was investigated to deduce a coherent docking model of the crystal structure of staphylokinase on the homology-based model of u-plasrnln, the catalytic domain of plasmin. Staphylokinase binding is mediated by two surface- exposed loops, 174 and 215, at the edge of the active site cleft of u-plasmin. The binding epitape of staphylokinase involves the N-terminus and the five-stranded mixed ~-sheet. The a-helix and the ~2 strand do not participate in the binding to ~l- plasmin, but are essential to induce plasminogen activation by the staphylokinase.u-plasrnin complex. A topologically distinct activation epitape thus exists. Binding of staphylokinase to the catalytic domain of plasmin allows protrusion of the activation epitape into a broad groove near the catalytic domain. This process generates a competent binding pocket for u-plasrninoqen, which buries approximately 2500 A of the staphylokinase.u-plasmin complex upon 50 binding (Jespers et al, 1999). While the manuscript of Jespers et al (1999) was in preparation, this deduced staphylokinase.u-plasmin.u-plasrnln complex was fully confirmed by X-ray crystallography (Parry et aI, 1998). Staphylokinase has been studied intensively in several animal models and in humans (Collen and Lijnen, 1994). Like streptokinase, staphylokinase is also highly immunogenic, which precludes repeated administration. Staphylokinase contains three non-overlapping immunodominant epitopes. Two of these epitopes could be eliminated by site-directed mutagenesis, where clusters of two to three charged amino acids were replaced with alanine, (Collen et aI, 1996a). The variants SakSTAR.M38 (Lys35, Glu38, Lys74, Glu75 and Arg77 substituted by Ala) and SakSTAR.M89 (Lys74, Glu75, Arg77, Glu8Dand ASp82substituted by Ala) were thrombolytically active and induced significantly less antibody formation than the wild type staphylokinase. This was found in animal models and in patients with peripheral arterial occlusion (Collen et aI, 1996b). These results indicate that staphylokinase can be manipulated to produce a less immunogenic protein which has potential clinical value as a thrombolytic agent. Tissue-type plasminogen activator (t-PA) IS a human fibrinolytic protein and is mainly produced by the endothelial cells. t-PA is synthesized as a single chain molecule of 530 amino acids and has a molecular weight of 68 kDa (Pennica et aI, 1983). It consists of a finger domain, an epidermal growth factor domain, two kringle domains and a catalytic domain. t-PA activates plasminogen by cleavage of the Arg561_Va156b2ond. The single chain form is converted to a two-chain form by plasmic cleavage of the Arg278_lle279bond (Ichinose et aI, 1984), which increases its fibrin-specificity (Husain et aI, 1989). Plasminogen activation is thus localized to sites of vascular injury. The fibrin specificity of t-PA is situated in its kringle domains and finger domain (Pennica et aI, 1983). The binding site of t-PA on fibrin has been localized to residues 149-161 (Bosma et aI, 1988), although other evidence indicates that t-PA binds to both the D and E domains of fibrin (Hasan et al, 1992). 51 A single-chain variant (alteplase) and a double-chain variant (duteplase) of t-PA has been produced (Collen, 1997). A deletion mutant of t-PA, consisting of the kringle 2 and protease domain, was constructed and administered as a bolus for coronary artery thrombolysis in patients with acute myocardial infarction. This variant was found to be equipotent to streptokinase (INJECT Investigators, 1995). Potent mutants were also constructed by substitution or deletion of one or a few selected amino acids. TNK-rt-PA, in which Thr103was replaced with Asn, Asn117 with Gin, and Lys296-His-Arg-Arg with Ala-Ala-Ala-Ala, was found to have an 8-fold slower clearance and a 2DD-fold enhanced resistance to PAI-1. This mutant had an increased potency on platelet-rich clots and was more effective upon bolus administration at a lower dose that rt-PA in in vivo models (Keyt et al, 1994; Collen et al, 1994). Urokinase-type plasminogen activator (u-PA) is found in large amounts in urine. u- PA contains 411 amino acids and has a molecular weight of 55 kDa (Husain et aI, 1983). Like t-PA it also cleaves the Arg561_Va156b2ond of plasminogen. It has structural similarities to t-PA, but contains only one kringle domain and lacks the finger domain. The single chain form can be rapidly converted to the two-chain form by plamin or kallikrein (Zamarron et al, 1984; Ichinose et al, 1986). The two- chain forms have greater enzymatic activity, but are less fibrin-specific than the single-chain form (Zamarron et aI, 1984). Both two-chain forms are inhibited by PAI-1 and PAI-2, while the single chain form is not (Stump et aI, 1986; Kruithof et aI, 1986). u-PA has been successfully used to treat thrombotic disorders, like venous thrombosis, pulmonary embolism, arterial thrombosis, acute myocardial infarction and acute reocclusion following percutaneous transluminal angioplasty (Loscalzo, .19 Park et al (1998) studied a novel plasminogen activator from Agkistrodon halys venom. First, they constructed a cDNA library and used a probe, based on the consensus sequence of serine protease from snake venom, for screening purposes. A positive clone was successfully expressed in a baculovirus system as a 32 kDa protein and was purified to homogeneity. Western analysis indicated that the band was indeed a snake venom component. The recombinant protein enhanced fibrinolysis by activation of plasminogen (Park et aI, 1998). 52 The Amazon leech, Haementaria ghilianii, produces a fibrinolytic enzyme called hemetin. It is a metalloprotease, which does not inhibit other coagulation factors. Hemetin is also not inactivated by plasma protein inhibitors (Malinconico et aI, 1984). A family of closely related plasminogen activators was isolated from the saliva of the vampire bat, Oesmodus rotundus (GardelI et al, 1989; Gardeli et al, 1990). They exist as three major forms with structural similarities to human t-PA. They seem to be more fibrin selective and more potent than human t-PA (GardelI et aI, 1991; Kratschmar et al, 1991). 2.3.5 Combination proteins: In some cases it is preferable to combine different activities in a single molecule to target different levels of haemostasis. Such a multifunctional agent could have enhanced efficacy and so lead to more effective treatment of a patient. Knapp et al (1992) designed a novel recombinant protein with combined antithrombotic and antiplatelet activity. The disintegrin activity was obtained by placing an RGD-sequence at the tip of a finger-like loop of hirudin. Native hirudin contains SDGE at the tip of the protruding finger (residues 32-35). Variants were obtained by replacing SDGE with RGDS to obtain hirudisin and KGDS to obtain hirudisin-1. Inhibition studies indicated that hirudisin is a 2-fold more potent thrombin inhibitor than hirudisin-1 or r-hirudin. Hirudisin also prevented ADP- induced platelet aggregation. Hirudisin is an important example of the success of novel proteins with combined antithrombotic and antiplatelet activities. Smith et al (1995) utilized the strategy of loop grafting, in which the amino acid sequence of a biologically active, flexible loop on one protein is used to replace a surface loop on an unrelated protein. HCDR3 from Fab-9, an antibody selected to bind the ~3-integrins with nanomolar affinity (Smith et aI, 1994), was grafted onto the epidermal growth factor-like module of human t-PA. This variant of t-PA (LG-t- PA) was cloned and expressed in COS cells. LG-t-PA bound to platelet integrin 53 (XllbP3 with nanomolar affinity, maintaining full enzymatic activity and was normally stimulated by its physiological co-factor fibrin. Like the donor antibody, binding of LG-t-PA to platelet integrin (XllbP3 was dependent on divalent cations and was inhibited by an RGO-containing peptide, indicating that LG-t-PA binds specifically to the ligand binding site of the integrin (Smith et aI, 1995). The use of combined profibrinolytic and antiplatelet activity can therefore enhance the fibrin-specificity of the thrombolytic agent. Riesbeck et al (1998) fused a HLA class I leader sequence with hirudin, linked to domains three and four of human C04 and the intracytoplasmic sequence of either C04 or human P-selectin. Mouse fibroblasts, Chinese hamster ovary (CHO)-K1 cells, immortalized porcine endothelial cells (IPECs) and a pituitary secrectory cell line (016/16) were transfected with the hirudin fusion constructs. Hirudin was expressed at the cell surface, where it bound thrombin and prevented the formation of fibrin in an in vitro assay using human plasma. Hirudin-C04-P- selectin fusion proteins accumulated in storage granules of secretory cells. When activated with phorbol ester, the fusion proteins were relocated to the cell surface, where the exposed hirudin was fully active. These fusion proteins can be used proactively in situations where thrombotic complications are anticipated, such as vascular surgery and transplantation. Constitutive hirudin expression would be absent on normal endothelial cells, but could be rapidly expressed when the complement system is activated. This approach is designed to inhibit trace amounts of thrombin generation before clot formation (Riesbeck et aI, 1998). A monoclonal antibody, 5908, recognises an epitope on fibrin that is only exposed after thrombin cleaves the p-chain of fibrinogen to release fibrinopeptide B (Hui et aI, 1983). Bode et al (1994) utilised this exclusive fibrin-binding property and covalently linked recombinant hirudin to the Fab' of 5908. The resultant fibrin- targeting molecule has a molecular weight of approximately 57 kOa,. which corresponds to a 1:1 molar ratio of hirudin and 5908Fab'. Hirudin-5908Fab' was 10 times more effective than hirudin to inhibit fibrin deposition onto experimental clot surfaces in fibrin solution and human plasma. Furthermore, it inhibited 54 peptidolytic activity more effectively than free hirudin. Therefore, fibrin-targeting can thus greatly increase antithrombotic activity of known antithrombotic drugs (Bode et al, 1994). Lijnen et al (1995) fused a C-terminal fragment (Ser47_Leu411) of recombinant single chain urokinase-type plasminogen activator (rscu-PA) to the C-terminus of hirudin (Asn53_Gln65) via a fourteen amino acid linker to achieve combined fibrinolytic and antithrombotic potential. The recombinant chimeric protein was produced in E. coli. The fibrinolytic potency of the chimera in plasma was maintained, when compared to the truncated form of rscu-PA and wild-type scu- PA. The chimera led to prolongation of thrombin time in normal human plasma and inhibited thrombin induced platelet aggregation. The amidolytic activity of thrombin remained unchanged, since the chimera left the active site of thrombin unblocked and available for small synthetic substrates. Their results also indicated that a higher concentration of the chimera was required for inhibition of coagulation than for fibrin clot lysis or for inhibition of thrombin-induced platelet aggregation. Therapeutically effective concentrations for clot lysis and prevention of platelet aggregation may thus be reached without influencing the systemic blood coagulation system (Lijnen et aI, 1995). Szarka et al (1999) analysed the suitability of staphylokinase as a fusion protein by fusing it to hirudin. Both N-and C-terminal fusions were constructed and the recombinants were expressed in Bacillus subtilis. Removal of the first ten amino acids by plasmin caused release and subsequent instability of the N-terminal fusion proteins in the presence of plasmin. Site-directed mutagenesis was performed at Lys10 and Lys11 to produce a plasmin-resistant variant. It could, however, not be done without interference with activation of staphylokinase by plasmin. Two putative plasmin cleavage sites (Lys 135and Lys136)are located at the C-terminus of staphylokinase. However, both these sites are resistant to plasmin cleavage, leading to the production of stable C-terminal fusion. Staphylokinase activity in terms of plasminogen activation were tested using a chromogenic substrate for plasmin. Activity of the C-terminal fusion was 55 indistinguishable from that of wild-type staphylokinase. Staphylokinase is thus a potential protein to use in a fusion for the addition of fibrinolytic activity to antithrombotic agents (Szarka et ai, 1999). An interesting approach to inhibit thrombin generation was followed by Riesbeck et al (1997), where a recombinant tissue factor pathway inhibitor (TFPI) was utilised to inhibit FXa. Full length or a truncated TFPI, lacking the third Kunitz domain, was fused to domains three and four and the carboxy-terminal of human CD4. The fusion proteins were expressed in a mouse fibroblast cell line and both proteins were tethered to the cell surface, where they were able to bind FXa. Inhibition of FXa activity was verified using a chromogenic assay. Genetic manipulation of endothelial cells to express functional TFPI may inhibit the development of coronary artery heart disease following cardiac allotransplantation and may inhibit thrombosis in the context of xenotransplantation (Riesbeck et ai, 1997). Recently, a modified hirutonin was introduced. An Asp-Ser sequence was inserted between the active site blocking moiety and the anion binding site moiety of hirutonin (LeBlond et ai, 1999). When washed platelets were stimulated with thrombin receptor activated peptide, the modified hirutonin displayed disintegrin activity and antithrombin acitivity was maintained. However, when platelets in platelet rich plasma was stimulated with ADP, no disintegrin activity was observed (Van Wyk et ai, unpublished observations). It is evident that the combination of antithrombin and fibrinolytic or antithrombin and disintegrin activity into one peptide can be done successfully. It is too early to assess whether this approach will be a success, since very few of these peptides were tested in vivo. In addition, the modified peptide usually lost some of the activity of the parent molecule and also some of the activity of the molecule that was added. It is therefore still an open question whether the less active combination peptide will be as effective as the more specific parent molecules. 56 2.4 REFERENCES: Adler M, Lazarus RA, Dennis MS, Wagner G. Solution structure of kistrin, a potent platelet aggregation inhibitor and Gp Ilb/llla antagonist. Science1991 ;253:445-8. Arnaud E, Lafay M, Gaussem P, Picard V, Jandrot-Perrus M, Aiach M, Rendu F. An autoantibody directed against human thrombin anion-binding exosite in a patient with arterial thrombosis: effects on platelets, endothelial cells, and protein C activation. Blood 1994;84:1843-50. Arocas V, Castro HC, Zingali RB, Guillin MC, Jandrot-Perrus M, Bon C, Wisner A. Molecular cloning and expression of bothrojaracin, a potent thrombin inhibitor from snake venom. Eur J Biochem 1997;248:550-7. Bajzar L, Manual R, Nesheim ME. Purification and characterization, of TAFI, a thrombin activatable fibrinolysis inhibitor. J Bioi Chem 1995;270:14477-84. Banner OW, Hadvary P. Crystallographic analysis at 30 A resolution of the binding to human thrombin of four active site-directed inhibitors. J Bioi Chem 1991 ;266:20085-93. Barton GJ, Newman RH, Freemont PS, Crumpton MJ. Amino acid sequence analysis of annexin super-gene family of proteins. Eur J Biochem 1991; 198:749-60. Bates, ER. Clinical trail results with hirudin and bivalirudin for acute coronary artery syndromes. Semin Thromb Hemost 1997;23:575-81. Belin D. Biological and facultative secretion of plasminogen activator inhibitor-2. Thromb Haemost 1993;70:144-7. 57 Berliner LJ, Sugawara Y, Fenton II JW. Human thrombin binding to non- polymerized fibrin-Sepharose: evidence for an anionic binding region. Biochemistry 1985;24:7005-9. Blockmans 0, Deckmyn H, Vermylen J. Platelet activation. Blood Rev 1995;9: 143-56. Blomback B, Blomback M. The molecular structure of fibrinogen. Ann NY Acad Sci 1972;202:77-97. Bode C, Hudelmayer M, Mehwald P, Bauer S, Freitag M, Von Hodenberg E, NewelI JB, Kublet W, Haber E, Runge MS. Fibrin-targeted recombinant hirudin inhibits fibrin deposition on experimental clots more efficiently than recombinant hirudin. Circulation 1994;90: 1956-63. Bode W, Mayr I, Baumann U, Huber R, Stone SR, Hofsteenge J. The refined 1.9- A crystal structure of human o-thrornbin: interaction with D-Phe-Pro-Arg chloromehylketone and significance of the Tyr-Pro-Pro-Trp insertion segment. EMBO J 1989;8:3467-75. Bosma PJ, Rijken DC, Niewenhuizen W. Binding of tissue-type plasminogen activator to fibrinogen fragments. Eur J Biochem 1988; 172:399-404. Braun PJ, Dennis S, Hofsteenge J, Stone SR. Use of site-directed mutagenesis to investigate the basis for the specificity of hirudin. Biochemistry 1988;27:6517-22. Broze GJ. Why do hemophiliacs bleed? Hosp Pract 1992;71-86. Cadroy Y, Maraganore JM, Hanson SR, Harker LA. Selective inhibition by a synthetic hirudin peptide of fibrin-dependent thrombosis in baboons. Proc Natl Acad Sci USA 1991;88:1177-81. 58 Capeilo M, Clyne LP, Mc.Phedran P, Hotez PJ. Ancylostoma factor Xa inhibitor: Partial purification and its identification as a major hookworm-derived anticoagulant in vitro. J Infect Dis 1993;167:1474-7. Cederholm-Williams SA, De Cock F, Lijnen HR, Collen D. Kinetics of the reactions between streptokinase, plasmin and alpha 2-antiplasmin. Eur J Biochem 1979; 100: 125-32. Chadderdon RC, Cappello M. The hookworm platelet inhibitor: functional blockade of integrins Gp Ilb/llla (allbP3) and Gp la/lla (a2P1) inhibits platelet aggregation and adhesion in vitro. J Infect Dis 1999; 179: 1235-41. Chao BH, Jakubowski JA, Savage B, Chow EP, Marzec UM, Harker LA, Maraganore JM. Agkistrodon piscivorus piscivorus platelet aggregation inhibitor: a potent inhibitor of platelet activation. Proc Natl Acad Sci USA 1989;87:8050-4. Charo IF, Scarborough RM, Du Mée CP, Wolf 0, Phillips DR, Swift RL. Pharmacodynamics of the Gp lib/Ilia antagonist integrelin: phase I clinical studies in normal healthy volunteers. Circulation 1992;86(Suppl I):1- 260(Abstract) . Christopoulos C, Mackie I, Lahiri A, Machin S. Flow cytometric observations on the in vivo use of Fab fragments of a chimaeric monoclonal antibody to platelet glycoprotein lib/Ilia. Blood Coagul Fibrinolysis 1993;4:729-37. Cirillo M, Lippi A, Subissi A, Agnelli G, Criscuoli M. Experimental pharmacology of hirunorm: a novel synthetic peptide thrombin inhibitor. Thromb Haemost 1996;76:384-92. Collen D. Thrombolytic therapy. Thromb Haemost 1997;78:742-6. 59 Collen 0, Bernaerts R, Declerck PJ, De Cock F, Demarsin E, Jenné S, Laroche Y, Lijnen HR, Silence K, Verstreken M. Recombinant staphylokinase variants with altered immunoreactivity, I: construction and characterization. Circulation 1996a;94: 197-206. Collen 0, Lijnen HR. Staphylokinase, a fibrin-specific plasminogen activator with therapeutic potential? Blood 1994;84:680-6. Collen 0, Moreau H, Stockx L, Vanderschueren S. Recombinant staphylokinase variants with altered immunoreactivity, II: thrombolytic properties and antibody induction. Circulation 1996b;94:207-16. Collen 0, Schlott B, Engelborghs Y, Van Hoef B, Hartmann M, Lijnen HR, Behnke D. On the mechanism of the activation of human plasminogen by recombinant staphylokinase. J Bioi Chem 1993;268:8284-9. Collen 0, Stassen JM, Yasuda T, Refino CJ, Paoni N, Keyt B, Roskams T, Guerrero JL, Lijnen HR, Gold HK, Bennett WF. Comparative thrombolytic properties of tissue-type plasminogen activator and of a plasminogen activator inhibitor-1-resistant glycosylation variant, in a combined arterial and venous thrombosis model in the dog. Thromb Haemost 1994;72:98- 104. Coiier BS. A new murine monoclonal antibody reports inactivation-dependent change in the conformation and/or micro-environment of the platelet glycoprotein lib/Ilia complex. J Clin Invest 1985;76: 101-8. Coller BS. Augmentation of thrombolysis with antiplatelet drugs. Coron Artery Dis 1996;6:911-4. Coller BS. Gp Ilb/llla antagonists: pathophysiologic and therapeutic insights from studies of c7Es Fab. Thromb Haemost 1997;78:730-5. 60 Coller BS, Folts JD, Smith SR, Scudder LE, Jordan R. Abolition of in vivo platelet thrombus formation in primates with monoclonal antibodies to the platelet Gp Ilb/llla receptor. Circulation 1989;80:1766-74. Coiier BS, Scudder LE, Beer J, Gold HK, Folts JD, Cavagnaro J, Jordan R, Wagner C, luliucci J, Knight 0, Ghrayeb J, Smith C, Weisman, HF, Berger H. Monoclonal antibodies to platelet glycoprotein lib/Ilia as antithrombotic agents. Ann NY Acad Sci 1991;614:193-213. Colman RW, Marder VJ, Salzman EW, Hirsh J. Overview of hemostasis. In Hemostasis and Thrombosis, Third Edition. (Eds. Colman RW, Hirsh R, Marder VJ, Salzman EW.) J.B. Lippencott Company, Philadelphia; 1994. Condra C, Nutt E, Petroski CJ, Simpson E, Friedman PA, Jacobs JW. Isolation and structural characterization of a potent inhibitor of coagulation factor Xa from the leech Haementaria ghi/ianii. Thromb Haemost 1989;61 :437-41. Connolly TM, Jacobs JW, Condra C. An inhibitor of collagen-stimulated platelet activation from the salivary glands of the Haementeria officina/is leech. J Bioi Chem 1992;267:6893-8. Coughlin SR. Thrombin receptor structure and function. Thromb Haemost 1993;66:184-7. Cox AD, Devine DV. Factor Xllla binding to activated platelets is mediated through activation of glycoprotein lib/Ilia. Blood 1994;83:1006-16. Davies MC, Englert ME, De Renzo EC. Interaction of streptokinase and human plasminogen. I. Combining of streptokinase and plasminogen observed in the ultracentrifuge under a variety of experimental conditions. J Bioi Chem 1964;239:2651-6. De Cristofaro R, Castagnola M. Kinetic aspects of release of fibrinopeptides AP and AY by human alpha-thrombin. Haemostasis 1991 ;21:85-90. 61 Deckmyn H, Stanssens P, Hoet B, Declerck PJ, Lauwereys M, Gansemans Y, Tornai I, Vermylen J. An echistatin-like Arg-Gly-Asp (RGD)-containing sequence in the heavy chain CDR3 of a murine monoclonal antibody that inhibits human platelet glycoprotein lib/Ilia function. Br J Haematol 1994;87:562-71. De Munk GAW, Groenveld E, Rjken DC. Urokinase and its complex with plasmingen activator-3/protein C inhibitor in urine. Fibrinolysis 1994;8:9- 15. Dennis MS, Carter P, Lazarus RA. Binding interactions of kistrin with glycoprotein Ilb/llla: Analysis by site-directed mutagenesis. Proteins Struct Funct Genet 1993;15:312-21. Dennis MS, Henzei WJ, Pitti RM, Lipari MT, Napier MA, Deisher TA, Bunting S, Lazarus RA. Platelet glycoprotein Ilb/llla protein antagonists from snake venoms: Evidence for a family of platelet aggregation inhibitors. Proe Natl Acad Sci USA 1990a;87:2471-5. Dennis S, Wallace A, Hofsteenge J, Stone SR. Use of fragments of hirudin to investigate thrombin-hirudin interaction. Eur J Biochem 1990b;188:61-6. De Simone G, Lombardi A, Galdiero S, Nastri F, Della Morte R, Staiano N, Pedone C, Bolognesi M, Pavone V. Hirunorms are true hirudin mimeties. The crystal structure of human a-thrombin-hirudin V complex. Protein Sci 1998;7:243-53. DiMaio J, Gibbs B, Munn D, Levebvre J, Ni F, Konishi Y. Bifunctional thrombin inhibitors based on the sequence of hirudin45-65. J Bioi Chem 1990;265:21698-703. Dodt J, Kohier S, Schmitz T, Wilhelm B. Distinct binding sites of Ala48-Hirudin1- 47 and Ala48-Hirudin48-65 on a-thrombin. J Bioi Chem 1990;265:713-8. 62 Dodt J, MOller H-P, SeemOller U, Chang J-Y. The complete amino acid sequence of hirudin, a thrombin specific inhibitor. FEBS Lett 1984; 165: 180-4. Dodt J, Schmitz T, Schafer T, Bergmann C. Expression, secretion and processing of hirudin in E. coli using the alkaline phosphatase signal sequence. FEBS Lett 1986;202:373-7. Dunwiddie CT, Waxman L, Vlasuk GP, Friedman PA. Purification and characterization of inhibitors of blood coagulation factor Xa from hematophageous organisms. Methods EnzymoI1993;223:291-312. Ehrlich HJ, Jaskunas SR, Grinnell BW, Van SB, Bang NU. Direct expression of recombinant activated human protein C. J Bioi Chem 1989;264:14298- 304. Electricwala A, Sawyer RT, Jones CP, Atkinson T. Isolation of a thrombin inhibitor from the leech Hirudinaria manillensis. Blood Coagul Fibrinolysis 1991 ;2:83-9. Esmon CT. Molecular events that control the protein C molecular anticoagulant pathway. Thromb Haemost 1993;70:29-35. Esmon NL, Owen WG, Esmon CT. Isolation of a membrane-bound cofactor for thrombin-catalyzed activation of protein C. J Bioi Chem 1982;257:859-64. Fareed J, Callas 0, Hoppensteadt DA, Lewis BE, Bick RL, Walenga JM. Antithrombin agents as anticoagulants ad antithrombotics: implications in drug development. Semin Hematol 1999;36:42-56. Fenton II JW. Thrombin interactions with hirudin. Semin Thromb Hemost 1989; 15:265-8. Fenton II JW, Ofosu FA, Brezniak MA, Hassouna Hl. Thrombin and Antithrombotics. Semin Thromb Hemost 1998;24:87-91. 63 Fenton II JW, Olsen TA, Zabinski MP, Wilner GD. Anion-binding exosite of human alpha-thrombin and fibrin(ogen) recognition. Biochemistry 1988;27:7106- 12. Folkers PJM, Clore GM, Oriscoll PC, Oodt J, Kohier S, Gronenborn AM. Solution structure of recombinant hirudin and the Lys47-vGlu mutant: A nuclear magnetic resonance and hybrid distance geometry-dynamical simulated annealing study. Biochemistry 1989;28:2601-17. Foster MR, Hornby EJ, Brown S, Hann M, Kitchin J, Pike N, Ward P. Inhibition of human platelet aggregation by GR91669, a prototype fibrinogen receptor antagonist. Thromb Res 1994;75:269-84. Francis CW, Marder VJ. Mechanisms of fibrinolysis. In: Williams Haematology Fifth Edition. (Eds. Beutier E, Lichtman MA, Coiier BS, Kipps TJ.) Mc Graw-Hilllnc, New York; 1995. Francis CW, Marder VJ, Martin SE. Plasmic degradation of crosslinked fibrin. I. Structural analysis of the particulate clot and identification of new macromolecular-soluble complexes. Blood 1980;56:456-64. Friedrich T, Kroger B, Bialojan S, Lemaire HG, Hoffken HW, Reuschenbach P, OUe M. Dodt J. A kazal-type inhibitor with thrombin specificity from Rhodnius prolixus. J Bioi Chem 1993;268:16216-22. Funakoshi T, Heimark R, Hendrickson L, McMullen BA, Fujikawa K. Human placental anticoagulant proteins: isolation and characterization. Biochemistry 1987a;26:5572-8. Funakoshi T, Hendrickson LE, McMullen BA, Fujikawa K. Primary structure of human placental anticoagulant protein. Biochemistry 1987b;26:8087-92. 64 GardelI SJ, Duong LT, Diehl RE, York JD, Hare TR, Register RB, Jacobs JW, Dixon RA, Friedman PA. Isolation, characterisation and cDNA cloning of a vampire bat salivary plasminogen activator. J Bioi Chem 1989;264: 17947- 52. GardelI SJ, Hare TR, Bergum PW, Cuca BC, O'Neill-Palladino L, lavodny SM. Vampire bat salivary plasminogen activator is quiescent in human plasma in the absence of fibrin unlike human tissue plasminogen activator. Blood 1990;76:2560-4. GardelI SJ, Ramjit DR, Stabilito JJ, Fujita T, Lynch JJ, Cuca GC, Jain 0, Wang SP, Tung JS, Mark GE, Shebuski RJ. Effective thrombolysis without marked plasminemia after bolus intravenous administration of vampire bat salivary plasminogen activator in rabbits. Circulation 1991 ;84:244-53. Ginsberg MH, Du X, O'Toole TE, Loftus JC. Platelet integrins. Thromb Haemost 1995;74:352-9. Gold HK, Collen BS, Yasuda T, Saito T, Fallon JT, Guerrero L, Leinbach RC, liskind AA, Collen D. Rapid and sustained coronary artery recanalization with combined bolus injection of recombinant tissue-type plasminogen activator and monoclonal antiplatelet Gp IIb/llla in a dog model. Circulation 1988;77:670-7. -Gruber A, Griffin JH, Harker LA, Hanson SR. Inhibition of platelet-dependent thrombosis formation by human activated protein C in a primate model. Blood 1989;73:639-42. Gruber A, Hanson SR, Kelly AB, Yan BS, Bang N, Griffin JH, Harker LA. Inhibition of thrombus formation by activated recombinant protein C in a primate model of arterial thrombosis. Circulation 1990;82:578-85. Gruber A, Mori E, Del Zoppo GJ, Waxman L, Griffin JH. Alteration of fibrin network by activated protein C. Blood 1994;83:2541-8. 65 Gruber A, Pal A, Kiss RG, Sas G, Griffin JH. Generation of activated protein C during thrombolysis. Lancet 1993;342:1275-6. Grutter MG, Priestle .JP, Rahuel J, Grossenbacher H, Bode W, Hofsteenge J, Stone SR. Crystal structure of the thrombin-hirudin complex: a novel mode of serine protease inhibition. EMBO J 1990;9:2361-5. Hanson SR, Griffin JH, Harker LA, Kelly AB, Esmon CT, Gruber A. Antithrombotic effects of thrombin-induced activation of endogenous protein C in primates. J Clin Invest 1993;92:2003-12. Hanson SR, Pareti FI, Ruggeri ZM, Marzec UM, Kunicki TJ, Montgomery RR, Zimmerman TS, Harker LA. Effects of monoclonal antibodies against the platelet glycoportein Ilblllla complex on thrombosis and haemostasis in the baboons. J Clin Invest 1988;81:149-58. Harker LA. Pathogenesis of thrombosis. In Haematology, Fourth Edition. (Eds. Williams WJ, Beutier E, Erslev AJ, Lichtman MA.) McGraw-Hill Inc., New York; 1990. Harker LA. Strategies for inhibiting the effects of thrombin. Blood Coagul Fibrinolysis 1994;5:S4 7-S85. Harker LA, Hanson SR, Runge MS. Thrombin hypothesis of thrombus generation and vascular lesion formation. Am J Cardiol 1995;75: 12B-7B. Harvey RP, Degryse E, Stefani L, Schamber F, Cazenave J-P, Courtney M, Tolstoshev P, Lecocq L-P. Cloning and expression of a cDNA coding for the anticoagulant hirudin from the bloodsucking leech, Hirudo medicinales. Proc Natl Acad Sci USA 1986;83: 1084-8. Hasan AA, Chang WS, Budzynski AZ. Binding of fibrin fragments to one-chain and two-chain tissue-type plasminogen activator. Blood 1992;79:2313-21 66 Hermans J, McDonagh J. Fibrin: structure and interactions. Semin Thromb Hemost 1982;8:11-24. Hofsteenge J, Stone SR, Donella-Deana A, Pinna L-A. The effect of substituting phosphotyrosine for sulphotyrosine on the activity of hirudin. Eur J Biochem 1990; 188:55-9. Hogg PJ, Jackson CM. Fibrin monomer protects thrombin from inactivation by heparin-antithrombin Ill: implications for heparin efficacy. Proc Natl Acad Sci USA 1989;86:3619-23. Hsu-Lin SC, Berman CL, Furie BC, August 0, Furie B. A platelet membrane protein expressed during platelet activation and secretion. J Bioi Chem 1984;259:9121-6. Huang TF, Holt JC, Kirby EP, Niewiarowski S. Trigramin: primary structure and its inhibition of von Willebrandt factor binding to glycoprotein Ilb/llla complex on human platelets. Biochemistry 1989;28:661-6. Huang TF, Holt JC, Lukasiewicz JL, Niewiarowski S. Trigramin: a low molecular weight peptide inhibiting fibrinogen interaction with platelet receptors expressed on glycoprotein Ilb/llla complex. J Bioi Chem 1987;262:16157- 63. Hui KY, Haber E, Matsueda GR. Monoclonal antibodies to a synthetic fibrin-like peptide bind to human fibrin but not fibrinogen. Science 1983;222: 1129- 32. Husain SS, Gurewich V, Lipinski B. Purification and partial characterization of a single-chain high molecular weight form of urokinase from human urine. Arch Biochem Biophys 1983;220:31-8. 67 Husain SS, Hasan AA, Budzynski AZ. Differences between binding of one-chain and two-chain tissue plasminogen activators to non-cross-linked and cross-linked clots. Blood 1989;74:999-1006. Ichinose A, Fukikawa K, Suyama I. The activation of prourokinase by plasma kallikrein and its inactivation by thrombin. J Bioi Chem 1986;261 :3486-9. Ichinose A, Kisiel W, Fukikawa K. Proteolytic activation of tissue plasminogen activator by plasmin and tissue enzymes. FEBS Lett 1984;175:412-8. INJECT Investigators. Randomised, double-blind comparison of reteplase double- bolus administration with streptokinase in acute myocardial infarction: trial to investigate equivalence. Lancet 1995;346:329-36. Jakubowski JA, Maraganore JM. Inhibition of coagulation and thrombin-induced platelet activation by a synthetic dodecapeptide modeled on the carboxy- terminus of hirudin. Blood 1990;75:399-406. Jespers L, Van Herzeele H, Lijnen HR, Van Hoef B, De Maeyer M, Collen D, Lasters I. Arginine 719 in human plasminogen mediates formation of the staphylokinase:plasmin activator complex. Biochemistry 1998;37:6380-6. Jespers L, Vanwetswinkel S, Lijnen HR, Van Herzeele N, Van Hoef B, Demarsin E, Collen D, De Maeyer M. Structural and functional basis of plasminogen activation by staphylokinase. Thromb Haemost 1999;81 :479-85. Kay LA. Whole blood coagulation in haemostasis. In Essentials of haemostasis and thrombosis. Churchill Livingstone Inc., Edinburgh; 1988. Keiler PM, Schultz LD, Condra C, Karczewski J, Connolly TM. An inhibitor of collagen-stimulated platelet activation from the salivary glands of the Haementeria officina/is leech, II: cloning of the eDNA and expression. J Bioi Chem 1992;267:6899-904. 68 Kettner C, Shaw E. D-Phe-Pro-ArgCH2CI-A selective affinity label for thrombin. Thromb Res 1979;14:969-73. Keyt BA, Paoni NF, Refino CJ, Berleau L, Nguyen H, Chow A, Lai J, Pena L, Pater C, Ogez J. A faster-acting and more potent form of tissue plasminogen activator. Proc. Natl Acad Sci USA 1994;91 :3670-4. Knapp A, Degenhardt T, Dodt J. Hirudisins. J Bioi Chem 1992;267:24230-4. Kotzé HF, Badenhorst PN, Lamprecht S, Meiring M, Van Wyk V, Nuyts K, Stassen JM, Vermylen J, Deckmyn H. Prolonged inhibition of acute arterial thrombosis by high dosing of a monoclonal anti-platelet glycoprotein lib/Ilia antibody in a baboon model. Thromb Haemost 1995;74:751-7. Kotzé HF, Lamprecht S, Badenhorst PN, Roodt JP, Van Wyk V. Transient interruption of arterial thrombosis by inhibition of factor Xa results in long- term antithrombotic effects in baboons. Thromb Haemost 1997;77: 1137- 42. Kratschmar J, Haendler G, Langer G, Boidol W, Bringmann P, Alagon A, Donner P, Schleuning WO. The plasminogen activator family from the salivary gland of the vampire bat Oesmodus rotundus: cloning and expression. Gene 1991;105:229-37. Krezel AM, Wagner G, Seymour-Ulmer J, Lazarus RA. Structure of the RGD protein decorsin: conserved motif and distinct function in leech proteins that affect blood clotting. Science 1994;264:1944-7. Krstenansky JL, Mao SJT. Antithrombin properties of the C-terminus of hirudin using synthetic unsulfated NU-acetyl-hirudin45-65. FEBS Lett 1987;211:10- 6. 69 Krstenansky JL, Owen TJ, Yates MT, Mao SJT. Anticoagulant peptides: nature of the interaction of the C-terminal region of hirudin with a noncatalytic binding site of thrombin. J Med Chem 1987;30: 1688-91. Krstenansky JL, Owen TJ, Yates MT, Mao SJT. The C-terminal binding domain of hirullin P18. Antithrombotin activity and comparison to hirudin peptides. FEBS Lett 1990;269:425-9. Kruithof EKO, Tran-Thang C, Bachmann F. The fast-acting inhibitor of tissue-type plasminogen activator in plasma is also the primary plasma inhibitor of urokinase. Thromb Haemost 1986;55:65-9. Lamba 0, Bauer M, Huber R, Fischer S, Rudolph R, Kohnert U, Bode W. The 2.3 A crystal structure of the catalytic domain of recombinant two-chain human tussue-type plasminogen activator. J Mol Bioi 1996;258: 117-35. LeBlond L, DiMaio J, Winocour PD. Insertion of the Asp-Ser-Phe sequence in the P'position of hirutonin provides molecules having both antithrombin and disintegrin activity. Thromb Res 1999;93: 171-81. Lecander I, Astedt B. Isolation of a new specific plasminogen activator from pregnancy plasma. Br J HaematoI1986;62:221-8. Levin EG, Marzec U, Anderson J, Harker LA. Thrombin stimulates tissue plasminogen activator release from cultured human endothelial cells. J Clin Invest 1984;74:1988-95. Lijnen HR, Collen D. Interaction of plasminogen activators and inhibitors with plasminogen and fibrin. Semin Thromb Hemost 1982;8:2-10. Lijnen, HR, Van Hoef B, De Cock F, Okada K, Ueshima S, Matsuo 0, Collen D. On the mechanism of fibrin-specific plasminogen activation by staphylokinase. J Bioi Chem 1991 ;266: 11826-32. 70 Lijnen, HR, Van Hoef B, Matsuo 0, Collen D. On the molecular interactions between plasminogen-staphylokinase, alpha 2-antiplasmin and fibrin. Biochem Biophys Acta 1992; 1118: 144-8. Lijnen HR, Wnendt S, Schneider J, Janocha E, Van Hoef B, Collen D, Steffens GJ. Functional properties of a recombinant chimeric protein with combined thrombin inhibitory and plasminogen-activating potential. Eur J Biochem 1995;234:350-7. Lombardi A, Nastri F, Della Morte R, Rossi A, De Rosa A, Staiano N, Pedone C, Pavone V. Rational design of true hirudin mimetics: Synthesis and characterization of multisite-directed a-thrombin inhibitors. J Med Chem 1996;39:2008-17. Lord ST, Byrd P, Hede KL, Wei C, Colby TJ. Analysis of fibrinogen Aa-fusion proteins. J Bioi Chem 1990;265:838-43. Loscalzo J. Thrombolytic therapy. In: Williams Haematology Fifth Edition. (Eds. Beutier E, Lichtman MA, Coller BS, Kipps TJ.) Mc Graw-Hill Inc, New York; 1995. Lumsden AB, Kelly AB, Schneider PA, Krupski WC, Dodson T, Hanson SR, Harker LA. Lasting safe interruption of endarterectomy thrombosis by transiently infused antithrombin peptide D-Phe-Pro-Arg-CH2CI in baboons. Blood 1993;81:1762-70. Malinconico SM, Katz JB, Budzynski A. Hemetin: anticoagulant protease from the salivary gland of the leech Haementaria ghilianii. J Lab Clin Med 1984; 103:44-58. Mann KG, Neshim ME, Church WR, Haley P, Krishnaswamy S. Surface dependent reactions of vitamin K dependent enzyme complex. Blood 1990;70: 1-16. 71 Mao SJT, Yates MT, Owen TJ, Krstenansky JL. Interaction of hirudin with thrombin: identification of a minimal binding domain of hirudin that inhibits clotting activity. Biochemistry 1988;27:8170-3. Maraganore JM, Adelman BA. Hirulog: a direct thrombin inhibitor for management of acute coronary syndromes. Coron Artery Dis 1996;7:438-48. Maraganore JM, Bourdon P, Jablonski J, Ramachandran KL, Fenton II JW. Design and characterization of hirulogs: a novel class of bivalent peptides inhibitors of thrombin. Biochemistry 1990;29:7095-101. Markwardt F. Hirudin as an inhibitor of thrombin. Methods Enzymol 1970;19:924- 32. Markwardt F. The development of hirudin as an antithrombotic drug. Thromb Res 1994;74: 1-23. Martin PO, Robertson W, Turk 0, Huber R, Bode W, Edwards BFP. The structure of residues 7-16 of the Aa-chain of human fibrinogen bound to bovine thrombin at 2.3-A resolution. J Bioi Chem 1992;267:7911-20. Maruyama K, Kawasaki T, Sakai Y, Taniuchi Y, Shimizu M, Kawashima H, Takenaka T. Isolation and amino acid sequence of flavostatin, a novel disintegrin from the venom of Trimeresurus flavoviridis. Peptides 1997; 18:73-8. McEver RP, Moore KL, Cummings RD. Leukocyte trafficking mediated by selectin- carbohydrate interactions. J Bioi Chem 1995;270: 11025-8. Munro R, Jones CP, Sawyer RT. Calin - a platelet adhesion inhibitor from the saliva of the medicinal leech. Blood Coagul Fibrinolysis 1991 ;2:179-84. 72 Musial J, Niewiarowski S, Rucinski B, Stewart GJ, Cook JJ, Williams JA, Edmunds Jr LH. Inhibition of platelet adhesion to surfaces of extracorporeal circuits by disintegrins. Circulation 1990;82:261-73. Naski MC, Fenton II JW, Maraganore JM, Olsen ST, Shafer JA. The COOH- terminal domain of hirudin. J Bioi Chem 1990;265: 13484-9. Neeper MP, Waxman L, Smith DE, Schulman CA, Sardana M, Ellis RW, Schaffer LW, Siegl PKS, Vlasuk GP. Characterization of recombinant tick anticoagulant peptide. J Bioi Chem 1990;265: 17746-52. Nesheim M, Wang W, Boffa M, Nagashima M, Morser J, Bajzar L. Thrombin, thrombomodulin and TAFI in the molecular link between coagulation and fibrinolysis. Thromb Haemost 1997;78:386-91. Niewiarowski S, Cook JJ, Stewart GJ, Gould RJ. Structural requirements for expression of antiplatelet activity of disintegrins. Circulation 1990;82 (Supplement 111):370(Abstract). Noeske-Jungblut C, Kratschmar J, Haendler B, Alagon A, Possani L, Verhallen P, Donner P, Schleuning WO. An inhibitor of collagen-induced platelet aggregation from the saliva of Triatoma pallidipennis. J Bioi Chem 1994;269:5050-3. Nutt EM, Gasic T, Rodkey J, Gasic GJ, Jacobs JW, Friedman PA, Simpson E. The amino acid sequence of antistasin. A potent inhibitor of factor Xa reveals a repeated internal structure. J Bioi Chem 1988;263;10162-7. Nutt EM, Jain 0, Lenny AB, Schaffer L, Siegl PK, Dunwiddie CT. Purification and characterization of recominant antistasin: a leech-derived inhibitor of coagulation factor Xa. Arch Biochem Biophys 1991 ;285:37-44. 73 Dates JA, Fitzgerald GA, Branch RA, Jackson EK, Knapp HR, Roberts LJ. Clinical implication of prostaglandin and thromboxane A formation. N Engl J Med 1988;319:689-98. Park 0, Kim H, Chung K, Kim DS, Yun Y. Expression and characterization of a novel plasminogen activator from Agkistrodon ha/ys venom. Toxicon 1998;36:1807-19. Parry MAA, Fernandez-Catalan, C, Bergner A, Huber R, Hopfner KP, Schlott B, Guhrs KH, Bode W. The ternary microplasmin-staphylokinase- microplasmin complex is a proteinase-cofactor-substrate complex in action. Nature Struct Bioi 1998;5:917-23. Pennica 0, Holmes WE, Kohr WJ, Harkins RN, Vehar GA, Ward CA, Bennett WF, Yelvenson E, Seeburg PH, Heyneker HL, Goeddel DV, Collen D. Cloning and expression of human tissue-type plasminogen activator cDNA in E.co/i. Nature 1983;301:214-21. Plow EF, Collen D. The presence and release of alpha-2-antiplasmin from human platelets. Blood 1981 ;58:1069-74. Priestle JP, Rahuel J, Rink H, Tones M, Grutter MG. Changes in interaction in complexes of hirudin derivatives and human a-thrombin due to different crystal forms. Protein Sci 1993;2: 1630-42. Oiu X, Padmanabhan KP, Carperos VE, Tulinsky A, Kline T, Maraganore JM, Fenton II JW. Structures of hirulog 3-thrombin complex and nature of S' subsites of substrates and inhibitors. Biochemistry 1992;31: 11689-97. Oiu X, Yin M, Padmanabhan KP, Krstenansky JL, Tulinsky A. Structures of thrombin complexes with a designed and a natural exosite peptide inhibitor. J Bioi Chem 1993;268:20318-26. 74 Rabijns A, De Bondt G, De Ranter C. Three-dimensional structure of staphylokinase, a plasminogen activator with therapeutic potential. Nature Struct Bioi 1997;4:357-60. Riesbeck K, Chen 0, Kemball-Cook G, McVey JH, George AJT, Tuddenham EGO, Dorling A, Lechler RI. Expression of hirudin fusion proteins in mammalian cells. Circulation 1998;98:2744-52. Riesbeck K, Dorling A, Kemball-Cook G, McVey JH, Jones M, Tuddenham EGO, Lechler RI. Human tissue factor pathway inhibitor fused to C04 binds both FXa and TF/FVlla at the cell surface. Thromb Haemost 1997;78:1488-94 Romoisch J, Seiffge 0, Reiner G, Paques EP, Heimburger N. In vivo antithrombotic effect potency of placental protein 4 (annexin V). Thromb Res 1991;61:93-104. Rosenburg RO, Oamus PS. The purification and mechanism of action of human antithrombin-heparin cofactor. J Bioi Chem 1973;248:6490-505. Ruggeri ZM. Mechanisms of shear-induced platelet adhesion and aggregation. Thromb Haemost 1993;70:119-23. Ruggeri ZM. Mechanisms initiating platelet thrombus formation. Thromb Haemost 1997;78:611-6. Rydel TJ, Ravichandran KG, Tulinsky A, Bode W, Huber R, Roitsch C, Fenton II JW. The structure of a complex of recombinant hirudin and human (J.- thrombin. Science 1990;249:277-80. Rydel TJ, Tulinsky A, Bode W, Huber R. Refined structure of the hirudin-thrombin complex. J Mol Bioi 1991 ;221 :583-601. 75 Sakharov DV, Lijnen HR, Rijken DC. Interactions between staphylokinase, plasminogen and fibrin. Staphylokinase discriminates between free plasminogen and plasminogen bound to partially degraded fibrin. J Bioi Chem 1996;271 :27912-8. Sako T. Overproduction of staphylokinase in Escherichia coli and its characterization. Eur J Biochem 1985;149:557-63. Savage B, Marzec UM, Chao BH, Harker LA, Maraganore JM, Ruggeri ZM. Binding of the snake venom-derived proteins applaggin and echistatin to the arginine-glycine-aspartic acid recognition site(s) on platelet glycoprotein Ilb/llla complex inhibits receptor function. J Bioi Chem 1990;265:11766-72. Savage B, Saldivar E, Ruggeri ZM. Initiation of platelet adhesion by arrest onto fibrinogen or translocation on von Willebrand factor. Cell 1996;84:289-97. Savage B, Shattil SJ, Ruggeri ZM. Modulation of platelet function through adhesion receptors: a dual role for glycoprotein lib-Ilia (integrin Ilb(3) mediated by fibrinogen and glycoprotein Ib-von Willebrand factor. J Bioi Chem 1992;267: 11300-6. Scarborough RM, Naughton M, Teng W, Rose JW, Phillips DR, Nannizzi L, Arfsten A, Campbell AM, Charo IF. Design of potent and specific integrin antagonists. J Bioi Chem'1993b;268:1066-73. Scarborough RM, Rose JW, Hsu MA, Phillips DR, Fried VA, Campbell AM, Nannizzi L, Charo IF. Barbourin. A Gp Ilb/llla-specific integrin antagonist from the venom of Sistrurus m. barbouri. J Bioi Chem 1991 ;266:9359-62. Scarborough RM, Rose JW, Naughton M, Phillips DR, Nannizzi L, Arfsten A, Campbell AM, Charo IF. Characterization of the integrin specificities of disintegrin isolated from American pit viper venoms. J Bioi Chem 1993a;268: 1058-65. 76 Schaffer LW, Davidson JT, Siegl PKS, Gould RJ, Nutt RF, Brady SF, Connolly TM. Recombinant leech antiplatelet protein prevents collagen-mediated platelet aggregation but not collagen graft thrombosis in baboons. ArterioscIer Thromb 1993;13:1593-601. Schaffer LW, Davidson JT, Vlasuk GP, Dunwiddie CT, Siegl PKS. Selective factor Xa inhibition by recombinant antistasin prevents vascular graft thrombosis in baboons. ArterioscIer Thromb 1992; 12:879-85. Schaffer LW, Davidson JT, Vlasuk GP, Siegl PKS. Antithrombotic afficacy of recombinant tick anticoagulant peptide; a potent inhibitor of factor X in a primate model of arterial thrombosis. Circulation 1991 ;84: 1741-8. Schlott B, Guhrs KH, Hartmann M, Rocker A, Collen D. NH2-terminal structural motifs in staphylokinase required for plasminogen activation. J Bioi Chem 1997;272:6067 -72. Schmitz T, Rothe M, Dodt J. Mechanisms of the inhibition of a.-thrombin by hirudin-derived fragments hirudin(1-47) and hirudin(45-65). Eur J Biochem 1991; 195:251-6. Scully MF. The biochemistry of blood clotting: the digestion of a liquid to form a solid. In Essays in biochemstry, vol 27. (Eds. Tipton KF.) Portland Press, London; 1992. Seymour JL, Henzei WJ, Nevins B, Stults JT, Lazarus RA. Decorsin. A potent glycoprotein llb/llla antagonist and platelet aggregation inhibitor from the leech Macrobdella decora. J Bioi Chem 1990;265:10143-7. Shebuski RJ, Ramjit DR, Sitko GR, Lumma PK, Garsky VM. Prevention of canine coronary artery thrombosis with echistatin, a potent inhibitor of platelet aggregation from the venom of the viper, Echis carinafus. Thromb Haemost 1990;64:576-81. 77 Silence K, Collen D, Lijnen HR. Interaction between staphylokinase, plasmin(ogen) and alpha 2-antiplasmin. Recycling of staphylokinase after neutralization of the plasmin-staphylokinase complex by alpha 2- antiplasmin. J Bioi Chem 1993a;268:9811-6. Silence K, Collen D, Lijnen HR. Regulation by alpha 2-antiplasmin and fibrin of the activation of plasminogen with recombinant staphylokinase in plasma. Blood 1993b;82: 1175-83. Silence K, Hartmann M, Guhrs KH, Gase A, Schlott B, Collen 0, Lijnen HR. Structure-function relationships in staphylokinase as revealed by "clustered charge to alanine" mutagenesis. J Bioi Chem 1995;270:27192-8. Skrzypczak-Jankun E, Carperos VE, Ravichandran KG, Tulinsky A, Westbrook M, Maraganore JM. Structure of hirugen and hirulog 1 complexes of a- thrombin. J Mol Bioi 1991 ;221: 1379-93. Smith JW, Hu 0, Satterthwait A, Pinz-Sweeney S, Barbas CF. Building synthetic antibodies as adhesive ligands for integrins. J Bioi Chem 1994;269:32788-95. Smith JW, Tachias K, Madison EL. Protein loop grafting to construct a variant of tissue-type plasminogen activator that binds platelet integrin allbP3. J Bioi Chem 1995;270:30486-90. Sonder SA, Fenton II JW. Proflavin binding within the fibrinopeptide groove adjacent to the catalytic site of human «-thrornbin. Biochemistry 1984;23: 1818-23. Stone SR, Hofsteenge J. Kinetics of the inhibition of thrombin by hirudin. Biochemistry 1986;25:4622-8. 78 Strube K-H, Krëqer B, Bialojan S, Otte M, Dodt J. Isolation, sequence analysis and eloning of haemadin. J Bioi Chem 1993;268:8590-5. Stubbs MT, Oschkinat H, Mayr I, Huber R, Angliker H, Stone SR, Bode W. The interaction of thrombin with fibrinogen. A structural basis for its specificity. Eur J Biochem 1992;206: 187-95. Stump DC, Thienpont M, Collen D. Purification and characterisation of a novel inhibitor of urokinase from human urine. Quantification and preliminary characterisation in plasma. J Bioi Chem 1986;261: 12759-66. Szarka SJ, Sihota EG, Habibi HR, Wong S-L. Staphylokinase as a plasminogen activator component in recombinant fusion proteins. Appl Environm Microbiol 1999;65:506-13. Szewczuk Z, Gibbs BF, Vue SY, Purisima E, Zdanov A, Cygler M, Konishi Y. Design of a linker for trivalent thrombin inhibitors: interaction of the main chain of the linker with thrombin. Biochemistry 1993;32:3396-404. Tcheng JE, Harrington RA, Kottke-Marchant K, Kleiman NS, Ellis SG, Kereiakes DJ, Mick MJ, Navetta FI, Smith JE, Worley SJ, Miller JA, Joseph DM, Sigmon KN, Kitt MM, Du Mée CP, Califf RM, Topoi EJ. Multicenter, randomized, double-blind, placebo-controlled trial of the platelet integrin glycoprotein Ilblllla blocker integrelin in elective coronary intervention. Circulation 1995;91:2151-7. Thiagarajan P, Benedict CR. Inhibition of arterial thrombosis by recombinant annexin V in a rabbit carotid artery injury model. Circulation 1997;96:2339-47. Thiagarajan P, Tait JF. Binding of annexin V/placental anticoagulant protein I to platelets: evidence for phosphatidylserine exposure in the procoagulant response to activated platelets. J Bioi Chem 1990;265: 17420-3. 79 Thiagarajan P, Tait JF. Collagen-induced exposure of procoagulant activity in platelets. J Bioi Chem 1991 ;266:24302-6. Tracy PB. Regulation of thrombin generation at cell surface. Semin Thromb Hemost 1988;14:227-33. Trikha M, Retzios AD, Markland FS. Characterization of a novel platelet aggregation inhibitor (contortrostatin) from the southern copperhead snake venom. Blood 1990;76 (Supplement):479a (Abstract). Van Heerde WL, Sakariassen KS, Hemker HC, Sixma JJ, Reutelingsperger CPM, DeGroot PG. Annexin V inhibits the pro-coagulant activity of matrices of TNF-stimulated endothelium under flow conditions. Arterioscler Thromb 1994;14:824-30. Van Ryn McKenna J, Merk H, Muller TH, Buchanan MR, Eisert WGo The effect of heparin and annexin V on fibrin accretion after injury in the jugular vein of rabbits. Thromb Haemost 1993;69:227-30. Vermylen J, Hoylaerts MF, Arnout J. Antibody-mediated thrombosis. Thromb Haemost 1997;78:420-6. Verstaete M, Zoldhelyi P. Novel antithrombotic drugs in development. Drugs 1995;49:856-84. Vlasuk GP, Ramjit 0, Fujita T, Dunwiddie CT, Nutt EM, Smith DE, Shebuski RJ. Comparison of the in vivo anticoagulant properties of standard heparin and the highly selective factor Xa inhibitors antistasin and tick anticoagulant peptide (TAP) in a rabbit model of venous thrombosis. Thromb Haemost 1991 ;65:257-62. Vu TK, Wheaton VI, Hung OT, Charo I, Coughlin SR. Domains specifying thrombin receptor interaction. Nature 1991 ;353:674-7. 80 Ware JA, Coller BS. Platelet morphology; biochemistry and function. In: Williams Haematology Fifth Edition. (Eds. Beutier E, Lichtman MA, Coller BS, Kipps TJ.) Mc Graw-Hill Inc, New York; 1995. Waxman L, Connolly TM. Isolation of a inhibitor selective for collagen-stimulated platelet aggregation from the soft tick Ornithodoros moubata. J Bioi Chem 1993;268:5445-9. Waxman L, Smith DE, Arcuri KE, Vlasuk GP. Tick anticoagulant peptide (TAP) is a novel inhibitor of blood coagulation factor Xa. Science 1990;248:593-6. Weiss HJ. Flow-related platelet deposition on subendothelium. Thromb Haemost 1995;74:352-9. Weiss HJ, Turitto VT, Baumgartner HR. The role of shear rate and platelets in promoting fibrin formation on rabbit subendothelium: studies utilizing patients with quantitative and qualitative platelet defects. J Clin Invest 1986;78: 1072-82. Weitz Jl, Hudoba M, Massel 0, Maraganore J, Hirsh J. Clot-bound thrombin is protected from inhibition by heparin-antithrombin III but is susceptible to inactivation by antithrombin Ill-independent inhibitors. J Clin Invest 1990;86:385-91. Wiman B, Collen D. On the mechanism of the reaction between human a2- antiplasmin and plasmin. J Bioi Chem 1979;254:9291-7. Witting Jl, Bourdon P, Brezniak DV, Maraganore JM, Fenton II JW. Thrombin- specific inhibition by and slow cleavage of hirulog-1. Biochem J 1992a;283:737 -43. Witting Jl, Bourdon P, Maraganore JM, Fenton II JW. Hirulog-1 and -B2 thrombin specificity. Biochem J 1992b;287:663-4. 81 Wohl RC, Summaria L, Arzadon L, Robbins DC. Steady-state kinetics of activation of human and bovine plasminogens by streptokinase and its equimolar complexes with various activated forms of human plasminogen. J Bioi Chem 1978;253:1402-7. Wu KK. Platelet activation mechanisms and markers in arterial thrombosis. J Intern Med 1996;239:17-34. Yasuda T, Gold HK. Kistrin, a polypeptide platelet Gp Ilblllla receptor antagonist, enhances and sustains coronary arterial thrombolysis with recombinant tissue-type plasminogen activator in a canine preparation. Circulation 1991;83:1038-47. Yeh CH, Peng HC, Huang TF. Accutin, a new disintegrin, inhibits angiogenensis in vitro and in vivo by acting as integrin alphavbeta3 antagonist and inducing apoptosis. Blood 1998b;92:3268-76. Yeh CH, Peng HC, Yih JB, Huang TF. A new short chain RGD-containing disintegrin, accutin, inhibits the common pathway of human platelet aggregation. Biochem Biophys Acta 1998a;1425:493-504. Vue S-Y, DiMaio J, Szewczuk Z, Purisima EO, Ni F, Konishi Y. Characterisation of the interactions of a bifunctional inhibitor with alpha-thrombin by molecular modeling and peptide synthesis. Protein Eng 1992;5:77-85. Zamarron C, Lijnen HR, Van Hoef B, Collen D. Biological and thrombolytic properties of proenzyme and active forms of urokinase. I. Fibrinolytic and fibrinogenolytic properties in human plasma in vitro of urokinase obtained from human urine or by recombinant DNA technology. Thromb Haemost 1984;52: 19-23. 82 Zdanov A, Wu S, Di Maio J, Konishi Y, Li Y, Wu X, Edwards BFP, Martin PO, Cygler M. Crystal structure of the complex of human a-thrombin and nonhydrolyzable bifunctional inhibitors hirutonin-2 and hirutonin-6. Proteins 1993; 17:252-65. Zingali RB, Jandrot-Perrus M, Guillin MC, Bon C. Bothrojaracin, a new thrombin inhibitor isolated from Bothrops jararaca venom: characterization and mechanism of thrombin inhibition. Biochemistry 1993;32: 10794-802. 83 CHAPTER3 THROMBOSIS Pergamon RESEARCH Thrombosis Research 88 (1997) 419-426 REGULAR ARTICLE Production of a Recombinant Antithrombotic and Fibrinolytic Protein, PLATSAK, in Escherichia colli Walda B. van ZyJl, Gert H. J. Pretorius', Manfred Hartmann/ and Harry F. Kotzé' 'Department of Haematology, University of the Orange Free State, Bloemfontein, South Africa; 21nstitut fur Molekulare Biotechnologie, Jena, Germany. (Received 19 August 1997 by Editor P. Badenhorst; revised/accepted 3 October 1997) Abstract amidolytic activity of thrombin. The fibrinolytic ac- tivity was almost equal to that of recombinant sta- The three main components involved in thrombosis phylokinase as measured with a thrombelastograph. and haemostasis are thrombin, platelets, and plas- Platelet aggregation was not markedly inhibited min. Almost all inhibitors of thrombosis are focused by PLATSAK, probably due to the unfavourable either on the inhibition of thrombin or on the inhibi- three dimensional structure, with the Arg-Gly-Asp tion of platelets. We designed a construct using the sequence buried inside. Our results confirm that it fibrinolytic activity of staphylokinase, fused via a is feasible to design and produce a hybrid multifurie- cleavable linker to an antithrombotic peptide of 29 tional protein that targets various components of the amino acids. The peptide was designed to include haemostatic process. © 1998 Elsevier Science Ltd. three inhibitory regions: (1) the Arg-Gly-Asp (RGD) amino acid sequence to prevent fibrinogen binding Key Words: Staphylokinase; Antiplatelet; Antithrombin; to platelets; (2) a part of fibrinopeptide A, an inhibi- Recornbinant tor of thrombin; and (3) the tail of hirudin, a potent direct antithrombin. The amino acid sequence of latelets and thrombin both 'play pivotal roles the 29 amino acid peptide was reverse translated, in thrombogenesis whereas plasmin dissolves and the gene was chemically synthesised and cloned P the thrombus. It is therefore not surprising into an expression vector as a 3' fusion to the staphy- that we have experienced a drastic increase in the lokinase gene. Gene expression was induced in design and development of new anti thrombotic E. coli Top 10 cells and the fusion protein, desig- agents that can either inhibit the action of throrn- nated PLATSAK, was purified using metal affinity bin or prevent platelet aggregation, and fibrinolytic chromatography. The purified fusion protein sig- agents that can enhance fibrinolysis. There are po- nificantly lengthened the activated partial thrombo- tent new inhibitors of platelet function, whereas plastin time and thrombin time and inhibited the powerful inhibitors of thrombin function and pro- duction are currently under study. The platelet inhibitors include synthetic Arg-Gly-Asp (RGD)- Abbreviations: SAK, staphylokinase: Pr, primer; FXa, factor Xa; peptides [1,2], snake venoms [3,4], and the Fïab'), TI, thrombin time; aPTI, activated partial thromboplastin time; fragment of the monoclonal antibody 7E3 [5]. peR, polymerase chain reaction; IPTO, isopropyl-[3-D-thiogalac- topyranoside; NT A, nitrilo-tri-acetic acid; ADP, adenosine di- Amongst the thrombin inhibitors are hirudin [6], phosphate. hirudin fragments [7], the hirulogs [8], hirugen [9], Corresponding author: O.HJ. Pretorius, Department of Haema- and synthetic hirudin C-terminal peptides [10]. In- tology, UOFS, P.O. Box 339, Bloemfontein, 9300, South Africa. hibitors of thrombin production include recornbi- Tel: +2751 4053656; Fax: +2751 4308485; E-mail: (gnhmghjp@ rncd.uovs.ac.za). nant tick anticoagulant peptide [lI] and recornbi- 0049-3848/97 $17.00 + .00 © 1998 Elsevier Science Ltd. Printed in the USA. All rights reserved. PIl S0049-3848(97)00277-6 84 W.B. van Zyl et al.lThrombosis Research 88 (1997) 419-426 nant activated protein C [12]. Collen and Lijnen [13] idues 54-65) were obtained from Sigma. Chemicals suggested that the effectivity of anticoagulant and for media were purchased from Difco. All chemi- antiplatelet agents could be enhanced by conjunctive cals used were of analytical grade. use with thrombolytic agents like staphylokinase (SAK) or streptokinase. SAK is a 136 amino acid 1.3. Recombinant DNA Techniques fibrinolytic protein produced by the bacterium Staphylococcus aureus which forms a 1:1 complex DNA techniques were performed according to the with plasmin(ogen). This complex then activates general criteria currently in use [17]. Fragments other plasminogen molecules [14]. for cloning were purified from agarose gels using For the purpose of this study, we designed an Qiagen purification kits. E. coli transformation antithrombotic peptide of29 amino acids that com- were performed as described [18]. bines antithrombin and anti platelet activity and used SAK to provide fibrinolytic activity. SAK was 1.4. Cloning Strategy linked to the anti thrombotic peptide through a se- quence that can be cleaved by activated coagula- The amino acid sequence encoding the antithrom- tion factor X. The factor Xa target sequence was botic peptide was reverse translated using optimal introduced to enable release of the anti thrombotic universal codons [19]. Two 60-mer oligonucleo- peptide in vitro and in vivo. The peptide was de- tides, designated primers 1 and 2 (Table 1), were signed to include three inhibitory regions: (1) an chemically synthesised. These two primers overlap amino acid sequence (RGD) derived from fibrino- 22 base pairs and the lacking ends were filled in gen, which is essential for binding of fibrinogen by PCR (Figure 1). The purified gene was subse- to its receptor (Glycoprotein lIb/IlIa) on platelet quently cloned into the Sma I restriction site of membranes [15]; (2) a part of fibrinopeptide A pUCBM21 (Figure 2) and the resulting plasmid was (residues 8-16), an inhibitor of thrombin [16]; and designated pWBl. Using primers 2 and 3 (Table 1), (3) the C-terminal tail of hirudin (residues 54-65), the gene was amplified from pWBl and cloned a direct antithrombin [6]. We hypothesised that simultaneously with the SAK gene and the Factor the construct, named PLATSAK (Platelet-Anti- Xa cleavage site into the Eco RI and the Sal I !hrombin-§.t~phylokinase), would prevent platelet restriction sites of pWB6 (Figure 2). The Factor aggregation, inhibit the action of thrombin and en- Xa cleavage site was constructed 'as two comple- hance fibrinolysis. In this paper we report on the mentary oligonucleotides, primers 4 and 5 (Table construction, purification, and in vitro characteri- 1). The newly constructed plasmid was designated sation of this recombinant fusion protein. pWBM3. To aid purification of the recombinant fusion protein, PCR was applied to add six consecu- tive histidine residues to the C-terminus of the 1. Materials and Methods protein, using primers 6 and 7 (Table 1). The newly constructed plasmid was called pWBM3H. A plas- 1.1. Host Strain mid containing only the SAK gene was named pWBM and produced in parallel with pWBM3H. Escherichia coli Top 1OF' [mer A, ~(mrr-hsdRMS- Gene expression was under control of the strongly mcrBC), 120 seconds. Interestingly, the level of FDP in plasma did not increase after administration of PLATSAK. These results demonstrate that PLATSAK effectively inhibited platelet deposition in both arterial- and venous-type thrombosis in an animal model. Introduction Several studies in animal models have emphasized the importance and essence of blood platelets and thrombin in thrombogenesis. The pivotal role of the adhesive protein receptor, glycoprotein (Gp) Ilblllla, in platelet-platelet interactions, has been demonstrated in canines (1-3) and non-human primates (4-6), by using selective antagonists. Of particular importance is the fact that Arg-Gly-Asp (RGD) containing peptides and proteins effectively block platelet- platelet interactions (5,7). The crucial role of thrombin in thrombogenesis has been demonstrated by using direct thrombin inhibitors (8,9). In other studies the formation of thrombin was inhibited with activated protein C (10), tissue factor pathway inhibitor (11) or recombinant tick anticoagulant, which inhibits activated factor Xa (12). In view of the importance of the RGD binding domain on the platelet adhesive proteins and the role of thrombin in thrombogenesis, together with the central role of fibrinolysis, multifunctional peptides have been designed and developed. The approaches were to combine antiplatelet and antithrombin activity, to combine antithrombin and fibrinolytic activity, to combine antiplatelet and fibrinolytic activity and to combine antithrombin, antiplatelet and fibrinolytic activity in the same molecule. Peptides that combine antiplatelet and antithrombin activity include the hirudisins (13), a chimeric antithrombin peptide (14) and a modified hirutonin (15). These peptides markedly inhibited thrombin, but had variable affectivity in inhibiting platelet-platelet interactions. Examples of peptides where antithrombin activity were combined with fibrinolytic activity, include one where r-hirudin were coupled with streptokinase (16) and one where 110 a 40 kDa part of recombinant single chain urokinase-type plasminogen activator were coupled to the tail (Hir53-65) of hirudin (17). Both proteins retained their fibrinolytic and antithrombin activity. A protein where an RGD binding domain was introduced into tissue-type plasminogen activator binds to Gp Ilb/llla and retains its fibrinolytic activity (18). PLATSAK is a protein that combines antiplatelet, antithrombin and fibrinolytic activity (19). The protein consists of staphylokinase linked, via a factor Xa cleavage site, to an antiplatelet and antithrombin peptide (Fig. t). Staphylokinase (SAK) is a highly fibrin-specific fibrinolytic protein (20) and should transport the chimera to a fibrin clot and thus to a factor Xa rich environment. The antiplatelet and antithrombin peptide can then be released by factor Xa to inhibit additional platelet aggregation and fibrin formation. The peptide was designed to contain three inhibitory regions. Firstly, on its N-terminus it has the RGD-sequence for binding to the fibrinogen receptor (Gp lib/Ilia) and so prevents platelet aggregation (21). That is followed by a part of fibrinopeptide A (residues 8-16) to inhibit the active site of thrombin (22) and the C-terminus of hirudin (residues 54-65) to bind to and block the anion binding site of thrombin (23). In vitro the chimera significantly inhibited the action of thrombin and activated plasminogen, but had no marked effect on platelet aggregation (19). In this study we evaluated the antithrombotic effect of PLATSAK in a baboon model of arterial and venous thrombosis (24). The baboon was used because the composition and function of its hemastatic apparatus closely resembles that of humans (25). Thrombogenic devices consisting of Dacron vascular graft material and an expansion chamber were inserted as extension segments into chronic arteriovenous (AV) shunts (26). The effect of a high dosage of PLATSAK on platelet deposition was investigated in real time by scintillation camera imaging of the deposition of 111ln-labeled platelets. 111 Hirudin SAK (54-65) Gp lib/ilia Activates - inhibits platelet plasminogen aggregation - promotes fibrinolysis - targets a thrombus Active site - inhibits proteolysis Releases peptide in vivo Anion binding exosite - inhibits fibrin formation Fig. 1 Schematic representation of the composition of PLATSAK and the function of the individual components. The staphylokinase part is linked to the antiplatelet and antithrombotic peptide via the recognition sequence of factor Xa. 112 Materials and Methods Preparation of PLA TSAK PLATSAK was recombinantly produced in E. coli cells and cultivation was performed as described in detail (19). The cells were harvested by centrifugation (3000xg for ten minutes) and proteins were extracted by repeated cycles of freezing and thawing (27). This procedure led to an enrichment of the recombinant proteins and the proteins could be extracted in any buffer of choice. In this study the proteins were dissolved in saline. The enriched fractions were run on SOS-PAGE gels and the percentage of PLATSAK present was determined by densitometry (Hoefer Scientific Instruments, San Francisco, USA). The extract contained 46% PLATSAK. In order to exclude possible effects of contaminating E. coli proteins on thrombogenesis, cells that were transformed with the empty vector were treated in exactly the same manner as the PLATSAK producing cells. These extracts were then used in the control baboons. Protein concentrations were determined using the BCA Protein Assay system from Pierce (Rockford, Illinois, USA). Animals studied Three normal male baboons (Papio ursinus) were used in this study. The animals weighed approximately ten kg and were disease-free for at least six weeks prior to use. To enable handling, the baboons were sedated with approximately 1 mg/kg ketamine hydrochloride (Premier Pharmaceuticals, Johannesburg, SA). The peripherai blood platelet count was normal (371 ± 112 X 109/1). All procedures were approved by the Ethics Committee for Animal Experimentation of the University of the Orange Free State in accordance with the National Code for Animal Use in Research, Education, Diagnosis and Testing of Drugs and Related Substances in South Africa. The baboons supported 113 permanent Teflon-Silastic AV shunts implanted in the femoral vessels (24). These shunts do not detectably shorten platelet survival or produce measurable platelet activation (25). Study protocol The experimental setup is schematically presented in Fig. 2. Thrombogenic devices were prepared as previously described (24,26). A piece of Dacron vascular graft material (1.26 crrr) served as a generator of platelet-dependent arterial-type thrombosis. An expansion chamber (3.77 cm2) was used to generate coagulation-dependent venous thrombosis. Initial wall shear rates were approximately 318 sec-1 for the graft section and less than 10 sec" for the expansion chamber. Under these conditions the devices cause arterial-type platelet-dependent and venous-type coagulation-dependent thrombosis, respectively (26). The thrombogenic device, prefilled with saline to avoid a blood-air interface, was incorporated as an extension segment in the permanent AV shunt, using Teflon connectors (24). The control baboons were treated with an equimolar bolus of the freeze-thaw extract from E. coli cells that did not produce PLATSAK to show that the proteins did not affect thrombogenesis. The bolus was administrated ten minutes before control devices were placed. The devices were kept in place for 60 minutes or until it occluded, where after they were removed and blood flow through the permanent AV shunt re-established. Sixty minutes later the baboons were treated with a bolus of 3.68 mg/kg PLATSAK to attain a plasma concentration of approximately 0.10 mg/ml plasma. After ten minutes the thrombogenic devices were placed for 120 minutes, whereafter they were removed to re-establish blood flow through the permanent AV shunt. 114 ~ 10 mm . 10 mm ~ 10 mm . ••••• ••• to ••••••••••••••• ~ 10 ••••••• vascular graft Expansion Chamber ~ Detector -..__ --- A3 MOS Computer Fig. 2 A schematic representation of the experimental set-up to study platelet deposition. The enlarged Dacron graft was used to simulate arterial thrombosis, while the enlarged expansion chamber was used to simulate venous thrombosis. 115 Graft imaging and quantification of platelet deposition Autologous platelets were labeled with 111ln-tropolone as previously described (28). Imaging and quantification of 111ln-platelets were done as described (24). In short, image acquisition of the grafts was done with a Searle Pho scintillation camera fitted with a high resolution collimator. The images were stored on and analyzed with a Medical Data Systems A3 computer (Medtronic, Ann Arbor, MI) interfaced with the scintillation camera. Dynamic image acquisition, three minute images (128 X 128 byte mode) for one hour in the control studies and two hours in the PLATSAK studies, was started simultaneously with the placement and start of blood flow through the devices. A three-minute image of a three ml autologous blood sample was also acquired each time the grafts were imaged to determine circulating blood radioactivity (blood standard). This blood was transferred to a thrombogenic device and imaged in the same geometry as the thrombogenic device incorporated into the AV shunt. Regions of interest of the graft and expansion segments were selected to determine the deposited and circulating radioactivity in the dynamic image. Radioactivity in regions of similar sizes was determined for the graft and extension segments of the blood filled device to determine circulating radioactivity. This was subtracted from the radioactivity in the graft and extension regions to calculate deposited radioactivity. The total number of platelets deposited was calculated as described (9,12,24). Laboratory measurements The times at which blood samples were collected can be seen in the results section. The platelet count of blood samples collected in 2 mg/ml disodium EDTA were determined with a Technicon H2 blood cell analyser (Bayer. Diagnostics, Tarratown, NY). The same blood sample was also used to determine whole blood and plasma radioactivity. Blood radioactivity was corrected by subtracting plasma radioactivity in order to calculate platelet 116 radioactivity (24). The radioactivity in the different samples was determined in a well-type scintillation counter. Blood was also collected in 3.8% sodium citrate (nine volumes blood to one volume citrate) to determine activated partial thromboplastin time (aPTT), plasma levels of the thrombin-antithrombin III (TAT) complex and fibrinogen degradation products (FOP). The aPTT was measured on a fibrinometer (Clotex II, Hyland Division, Travenol Laboratories, Costa Mesa, CA) using reagents supplied by the manufacturer. TAT complexes were determined using an ELlSA-method (Enzygnost Combipack, Behring, Marburg, Germany) as recommended by the suppliers. FOP in plasma was determined semi-quantitatively using a latex agglutination method following the manufacturer's instructions (Diagnostica Stago, Asnieres-Sur-Seine, France). Results The results on platelet deposition are summarized in Fig. 3. In the three control studies the thrombogenic devices occluded after 42, 39, 42 minutes. Whereas platelet deposition onto the graft segment started almost immediately, there was a delay of approximately six minutes before they started accumulating in the expansion chamber. In addition, the graft segments accumulated approximately six times more platelets than the expansion chambers at the end of the control studies. Treatment with PLATSAK inhibited platelet deposition onto the Dacron vascular graft by approximately 84% after 39 minutes, i.e. at the time of occlusion of the control devices. Platelet deposition was however not completely interrupted, but deposited at a much slower rate than in the control studies, i.e. 1.43 ± 0.71 x 107 platelets/crnvrriin versus 8.85 ± 1.83 x 107 platelets/cmvrnln. The inhibitory effect of PLATSAK on platelet deposition in the expansion chambers was more pronounced, i.e. 94% less platelets deposited at 39 minutes than in the control studies. In addition, the start of deposition was delayed by approximately twelve to fifteen minutes. The rate of platelet deposition following treatment was also much slower than in the control studies, 0.22 ± 0.12 x 107 117 Dacron vascular graft N"3 E aC;-J o "f"" )( -2 "'0 Q) ;ot: IJ) co.. Q) "'01 J!! ~.s ..!!! n, 50 75 100 125 Time (minutes) Expansion chamber N" 0.6 -ECJeon "f"" )( :; 0.4 Q) ;ot: IJ) co. Q) "J'!0! 0.2 ~.s ..!!! o, 25 50 75 100 125 Time (minutes) Fig. 3 The effect of PLATSAK on platelet deposition in control animals (-) and after treatment (----). Values are given as a mean ± SEM. 118 platelets/crnvmin versus 4.39 ± 1.97 x 107 ptatelets/crrrrmin. None of the thrombogenic devices occluded during the 120 minutes of the study after treatment with PLATSAK. The changes in platelet count, aPTT, TAT complexes and FOP levels are summarized in Table 1. Infusion of the freeze-thaw extract from PLATSAK non- producing E. coli decreased the platelet count slightly to approximately 80%. Placement of the thrombogenic devices caused a further decrease of approximately 21% during the 39 to 42 minutes that the devices remained potent. After treatment, placement of the thrombogenic devices for 120 minutes, decreased the platelet count by 8%. The aPTT was not affected by thrombogenesis in the control studies. It was markedly prolonged by treatment with PLATSAK, especially towards the end of the study. The TAT complexes increased approximately twenty-fold in the control study. After the thrombogenic devices were removed, and before placement of the second thrombogenic device 30 minutes later, it decreased to normal values. After treatment with PLATSAK, placement of the thrombogenic devices increased the TAT complexes two- to three-fold. The changes that we observed in the plasma levels of FOP are of interest. The test only became positive toward the end of the control study. Thereafter, the FOP levels remained constant in spite of treatment with PLATSAK. Discussion Thathrornboqenic devices that were used in this study typically represents conditions under which an arterial-type (graft segments) and a venous-type (expansion chambers) thrombus forms. The wall shear rate in the graft segment, ~ 320 sec", approximates that found in medium to large sized arteries, while the wall shear rate in the expansion chamber, < 10 sec", represents that found in the venous blood vessels (26,29). The finding that platelet deposition onto the graft segment was approximately six-fold more than in the expansion chamber (Fig. 3) 119 Table 1: Summary of the changes that were observed in platelet count, aPTT, TAT complexes and FOP levels. (Mean ± SO; NO = not done) Study Time Platelet count aPTT TAT FDP (min) (x 109/1) (sec) ()lg/I) ()lg/ml) Control Pre-bolus 371 ± 112 33 ± 3 13.1 ±4.4 Neg Pre-graft 340 ± 94 34 ± 3 21.5±15.5 Neg 10 NO 34 ±4 36.6 ± 16.7 Neg 39 - 42 268 ± 55 40 ± 5 251.5 ± 26.9 20 PLATSAK Pre-graft 267 ± 55 65 ± 8 27.6 ± 16.8 20 10 NO 72 ±12 25.5 ± 13.4 20 120 244 ± 62 > 120 61.0 ± 21.0 20 *p<0.05, students t-test. End versus pre-graft. further indicates arterial-type, platelet-dependent thrombogenesis versus venous- type coagulation-dependent thrombogenesis (29). One must also bear in mind that anticoagulants such as heparin and low molecular weight heparin markedly inhibit platelet deposition in the expansion chamber, but has little efffect on platelet deposition onto graft or collagen surfaces (4,5,8,26). On the other hand, blockade of the Gp lib/Ilia receptor markedly inhibits platelet deposition onto the graft surface (4,6), but not in flow expansion chambers (4). We have no indication that the extract from PLATSAK non-producing E. coli cells inhibited platelet deposition in the control studies. Firstly, platelet deposition onto the graft segments and in the expansion chambers was rapid (Fig. 3) and all three devices occluded within 42 minutes following their placement. Secondly, the number of platelets deposited onto the graft surface and in the expansion chamber compares favourably with the results obtained in studies where similar 120 devices were used (26,29). Thirdly, there was no significant lengthening in the aPTT (Table 1). The injected proteins also did not cause thrombocytopenia since the peripheral blood platelet count was not markedly affected by infusion (Table 1). PLATSAK markedly inhibited platelet deposition onto the graft segment and in the expansion chamber when the thrombogenic devices were placed ten minutes after a bolus infusion (Fig. 3). In both cases the onset of platelet deposition was delayed when compared to the control studies. Furthermore, none of the devices occluded during the 120 minutes of the study. The differences that were observed in the inhibition of platelet deposition are of interest. At 39 minutes, when the control devices occluded, platelet deposition onto the graft segment was inhibited by approximately 84% and in the expansion chamber by approximately 94%. At 120 minutes inhibition was approximately 50% and 85%, respectively, when compared to the number of platelets deposited at 39 minutes in the control studies. These results suggest that the extent of inhibition of platelet deposition onto the graft segment was markedly less than in the expansion chamber. However, when inhibition of the rates of platelet deposition are compared, it seems that inhibition onto the graft surface and in the expansion chamber was equivalent. In the case of inhibition of platelet deposition onto the graft surface, the rate of deposition was inhibited by 83 ± 8%. The rate of platelet deposition in the chamber was inhibited by 94 ± 4%, i.e. slightly, but not significantly, more than in the case of the graft surface. This is difficult to explain, especially in view of our in vitro results with PLATSAK (19), where we have shown that the antithrombin activity was much more pronounced than the antiplatelet activity. One must also bear in mind that direct inhibition of thrombin with, for example hirudin, requires a much higher dose to inhibit platelet- dependent thrombosis than coagulation-dependent thrombosis (30). Therefore, one would have expected that the effect of PLATSAK should have been more pronounced in the expansion chamber since the in vitro studies suggest it (19) and in vivo results with direct thrombin inhibition support it (30). A plausible 121 explanation for our finding that platelet-dependent and coagulation-dependent thrombosis was equivalently inhibited by PLATSAK may be that its antiplatelet effect is more pronounced in vivo than in vitro. Alternatively, one may conclude that aggregation studies in vitro is not entirely suited to test antiplatelet effects of the blockade of the Gp Ilb/llla receptor, especially when the antiplatelet effect is not very pronounced. It was clearly shown that not all platelets have to be functional to cause a normal platelet aggregation in response to agonists added in vitro (31). b Ten to twenty minutes after the infusion of PLATSAK the aPTT lengthened from approximately 33 seconds to 65 seconds to 72 seconds. It only lengthened to >120 seconds towards the end of the study (Table 1). In general, when an antithrombin is infused, the aPTT lengthens to near steady state levels within ten to fifteen minutes (9,32). We do not know if the design of PLATSAK can explain the results of treatment on the aPTT. The antithrombin/antiplatelet peptide is coupled to SAK with a factor Xa cleavage site (Fig. 1). It is plausible that when coupled to SAK, the antithrombin activity of PLATSAK is not optimal. Initially, before placement of the thrombogenic device and during the first ten minutes after the device was placed, most of the antithrombin part could still be attached to SAK, because not enough factor Xa is available to free the peptide. At later stages, on the other hand, factor Xa cleavage could provide enough of the peptide to inhibit thrombin and so lengthen the aPTT. Treatment with PLATSAK either decreased the production of thrombin, or PLATSAK formed complexes with thrombin to make it inaccessible for binding to antithrombin Ill, or both. After treatment with PLATSAK, appreciably less TAT complexes formed (Table 1). FOP in plasma could only be measured at the end of the control studies, indicating some activation of fibrinolysis through normal mechanisms. It does not appear that SAK resulted in additional activation of 122 fibrinolysis, since the FOP levels in plasma does not seem to increase in the studies were PLATSAK was used (Table 1). One must however not overinterpret these results since the method to measure FOP is semi-quantitative. We have not measured indices that could assess if PLATSAK increased the risk of developing a bleeding tendency. There were indications that it was the case. When we removed the permanent AV shunts after each study, we had some difficulty to stop bleeding at the sutures where the shunt access sites were closed. We do not know whether this was the result of inhibition of thrombogenesis by the peptide or activation of fibrinolysis by SAK. Indications are that it was activation of fibrinolysis. Firstly, the half-life of small peptides is very short (5,8), while that of SAK is relatively long (20). Secondly, one baboon developed a haematoma at the site of surgery 24 hours after the AV shunt was removed and bleeding effectively stopped. If it was indeed the case, one must redesign the thrombus targeting characteristic of PLATSAK. In conclusion, we have shown that PLATSAK inhibited platelet deposition in both arterial- and venous-type thrombosis. Our results show that it is feasible to combine different antithrombotic strategies with success into the same molecule. We could not distinguish which strategy was responsible for inhibition. The results do, however, indirectly suggest that both the inhibition of thrombin and blockade of Gp IIb/llla played a part. It does not appear that fibrinolysis contributed to the lower number of platelets that were deposited. One must bear in mind that these conclusions are based on the results obtained in three baboons. In view of the difficulties to produce sufficient amounts of PLATSAK, and because the effect of PLATSAK on platelet deposition was so dramatic, we decided not to study more baboons. The results on the three baboons were regarded as sufficient proof that combination of different inhibitory moieties into one molecule is a viable strategy for the development of antithrombotic drugs. 123 Acknowledgements This study was financially supported by the Central Research Fund of the University of the Orange Free State and the South African Medical Research Council. We would like to acknowledge the technical assistance of Elmarie Wentzel and Charmaine Combrink. References 1. Haskei EJ, Adams SP, Feigen LP, Saffitz JE, Gorczynski RJ, Sobel BE, Abendschein DR. Prevention of reoccluding platelet-rich thrombi in canine femoral arteries with a novel peptide antagonist of platelet glycoprotein Ilblllla receptors. Circulation 1989;80: 1775-82. 2. Bates ER, McGillem MJ, Mickelson JK, Pitt B, Mancini GBJ. A monoclonal antibody against the platelet glycoprotein lib/Ilia receptor complex prevents platelet aggregation and thrombosis in a canine model of coronary angioplasty. Circulation 1991;84:2463-9. 3. Willette RN, Sauermelch CF, Rycyna R, Sarkar S, Feuerstein GZ, Nichois AJ, Ohlstein EH. Antithrombotic effects of a platelet fibrinogen receptor antagonist in a canine model of carotid artery thrombosis. Stroke 1992;23:703-11. 4. Hanson SR, Pareti FI, Ruggeri ZM, Marzec UM, Kunicki TJ, Montgomery RR, Zimmerman TS, Harker LA. Effect of monoclonal antibodies against the platelet glycoprotein Ilblllla complex on thrombosis and haemostasis in the baboon. J Clin Invest 1988;81:149-58. 124 5. Cadroy Y, Houghten RA, Hanson SR. RGDV peptide selectively inhibits platelet-dependent thrombus formation in vivo: Studies using a baboon model. J Clin Invest 1989;84:939-44. 6. Kotzé HF, Badenhorst PN, Lamprecht S, Meiring M, Van Wyk V, Nuyts K, Stassen JM, Vermylen J, Deckmyn H. Prolonged inhibition of acute arterial thrombosis by high dosing of a monoclonal anti-platelet glycoprotein Ilb/llla antibody in a baboon model. Thromb Haemost 1995;74:751-7. 7. Ruggeri ZM, Houghton RA, Russel SR, Zimmerman TS. Inhibition of platelet function with synthetic peptides designed to be high affinity antagonists of fibrinogen binding to platelets. Proc Natl Acad Sci USA 1986;83:5708- 12. 8. Hanson SR, Harker LA. Interruption of acute platelet-dependent thrombosis by the synthetic antithrombin D-phenylalanyl-L-propyl-L-arginyl chloromethyl ketone. Proe Natl Acad Sci USA 1988;85:3184-8. 9. Kotzé HF, Lampreeht S, Badenhorst PN. A four hour infusion of recombinant hirudin results in long-term inhibition of arterial-type thrombosis in baboons. Blood 1995;85:3158-63. 10. Gruber A, Griffen JH, Harker LA, Hanson SR. Inhibition of platelet dependent thrombus formation by human activated protein C in a primate model. Blood 1989;73:639-42. 11. Haskei EJ, Torr SR, Day KC, Palmier MO, Wun TC, Sobel BE, Abendschein OR. Prevention of arterial reocclusion after thrombolysis with recombinant lipo-protein-associated coagulation inhibitor. Circulation 1991;84:821-7. 125 12. Kotzé HF, Lamprecht S, Badenhorst PN, Roodt JP, Van Wyk V. Transient interruption of arterial thrombosis by inhibition of factor Xa results in long- term antithrombotic effects in baboons. Thromb Haemost 1997;77:1137- 42. 13. Knapp A, Degenhardt T, Dodt J. Hirudisins. J Bioi Chem 1992;267:24230- 4. 14. Church FC, Phillips JE, Woods JL. Chimeric antithrombin peptide: Characterization of an Arg-Gly-Asp (RGD)- and hirudin carboxyl terminus-containing synthetic peptide. J Bioi Chem 1991;266: 11975-9. 15. Leblond L, DiMaio J, Winocour PD. Insertion of the Asp-Ser-Phe sequence in the P'position of hirutonin provides molecules having both antithrombin and disintegrin activity. Thromb Res 1999;93:171-81. 16. Phaneuf MD, Ozaki CK, Johnstone MT, Loza J-P, Quist WC, LoGerfo FW. Covalent linkage of streptokinase to recombinant hirudin: A novel thrombolytic agent with antithrombotic properties. Thromb Haemost 1994;71 :481-7. 17. Lijnen HR, Wendt S, Schneider J, Janocha E, Van Hoef B, Collen 0, Steffeus GJ. Functional properties of a recombinant chimeric protein with combined thrombin inhibitory and plasminogen-activating potential. Eur J Biochem 1995;234:350-7. 18. Yamada T, Shimada Y, Kikuchi M. Integrin-specific tissue-type plasminogen activator engineered by introduction of the Arg-Gly-Pro sequence. Biochem Biophys Res Commun 1996;228:306-11. 126 19. Van Zyl WB, Pretorius GHJ, Hartmann M, Kotzé HF. Production of a recombinant antithrombotic and fibrinolytic protein, PLATSAK, in Escherichia co/i. Thromb Res 1997;88:419-26. 20. Collen 0, Lijnen HR. Staphylokinase, a fibrin-specific plasminogen activator with therapeutic potential? Blood 1994;84:680-6. 21. Ruoslahti E, Pierschbacher MD. New perspectives in cell adhesion: RGD and integrins. Science 1987;238:491-7. 22. Martin PO, Robertson W, Turk 0, Huber R, Bode W, Edwards BFP. The structure of residues 7-16 of the Aa-chain of human fibrinogen bound to bovine thrombin at 2.3A resolution. J Bioi Chem 1992;267:7911-20. 23. Markwardt F. Hirudin as an inhibitor of thrombin. Methods Enzymol 1970; 19:924-32. 24. Hanson SR, Kotzé HF, Savage B, Harker LA. Platelet interactions with Dacron vascular grafts. A model of acute thrombosis in baboons. Arteriosclerosis 1985;5:595-603. 25. Hanson SR, Harker LA. Baboon models of acute arterial thrombosis. Thromb Haemost 1987;58:801-5. 26. Cadroy Y, Horbett TA, Hanson SR. Discrimination between platelet- mediated and coagulation-mediated mechanisms in a model of complex thrombus formation in vivo. J Lab Clin Med 1989; 113:436-48. 27. Johnson BH, Hecht MH. Recombinant proteins can be isolated from E. co/i cells by repeated cycles of freezing and thawing. Biotechnology 1994; 12:1357 -60. 127 28. Kotzé HF, Lamprecht S, Badenhorst PN, Van Wyk V, Roodt JP, Alexander K. In vivo inhibition of acute platelet-dependent thrombosis in a baboon model by BAY U3405, a thromboxane A2-receptor antagonist. Thromb Haemast 1993;70:672-5. 29. Hanson SR, Griffen JH, Harker LA, Kelly AB, Esmon CT, Gruber A. Antithrombotic effects of thrombin-induced activation of endogenous protein C in primates. J Clin Invest 1993;92:2003-12. 30. Harker LA. Strategies for inhibiting the effects of thrombin. Blood Coagul and Fibrinolysis 1994;5:S47-S58. 31. Cerskus AL, Ali M, Davies BJ, McDonald JWD. Possible significance of small numbers of functional platelets in a population of aspirin-treated platelets in vitro and in vivo. Thromb Res 1980;18:389-97. 32. Meiring SM, Lotter MG, Badenhorst PN, Bucha E, Novak G, Kotzé HF. Sites of elimination and pharmacokinetics of recombinant I-BI-Iepirudin in baboons. J Pharm Sci 1999;88:523-9. 128 CHAPTER 7 CONCLUSIONS My aim was to design and express a synthetic gene encoding an antithrombin and antiplatelet peptide and to evaluate its in vitro activity and in vivo potential. The chemically synthesized gene was fused downstream from the gene encoding staphylokinase (SAK), separated only by the recognition sequence of factor Xa (Fig 1.1). SAK was fused to the peptide to enhance its stability towards proteolytic degradation during production in Escherichia coli and furthermore, to add fibrinolytic activity to the range of activities of the chimera. The resultant chimeric protein was designated PLATSAK (PLatelet- Antilhrombin-§tAphyloKinase) and had fibrinolytic, antithrombotic and antiplatelet activity. Since SAK is highly fibrin-specific, it can transport the chimera to a fibrin clot and thus also to a factor Xa-rich environment. The high fibrin-specificity of SAK enabled the development of PLATSAK as a local drug delivery system. Six consecutive histidine residues were added immediately downstream of the C-terminus of the peptide to use as an affinity tag during metal affinity chromatography to aid purification of PLATSAK. Production of the chimeric protein in E. coli was successful and PLATSAK comprised 46% of the intracellular protein content after one hour of the induction of gene expression. Metal affinity chromatography was used to purify PLATSAK to homogeneity. In vitro tests showed that PLATSAK was able to inhibit thrombin activity markedly. The mechanism of interaction of PLATSAK with thrombin is not clear and it is not certain whether the antithrombin activity is due to the hirudin or to the FPA components. It can be assumed that the tail of hirudin will bind to the anion binding exosite of thrombin (Naski et al, 1990; Rydel et aI, 1990) and the FPA sequence to the active site (Martin et aI, 1992). However, it is unlikely that one molecule of PLATSAK would be able to inhibit both the exosite and the active site of a single thrombin molecule. In vitro, the antiplatelet activity was negligible, most likely due to an unfavourable three- 129 dimensional folding of the protein. Attempts at repositioning the RGD-motif to the C-terminus or the N-terminus of PLATSAK were made, but there was no success in improving the in vitro antiplatelet activity. However, more attempts at improving the in vitro antiplatelet effect need to be performed. More RGD- motifs could be placed at strategic sites on the molecule. The three- dimensional structure of SAK can be utilized to determine protruding loops for placing RGD-motifs at the tips of these loops. This strategy will hopefully result in the positioning of RGD-motifs at optimal sites for interaction with its receptors on activated platelet membranes. Data obtained by thrombelastography showed that the fibrinolytic activity of PLATSAK was slightly lower than that of recombinant SAK. It is possible that the addition of the antithrombotic peptide to the C-terminus of SAK led to a slight distortion of its tertiary structure or may somewhat interfere in the binding of PLATSAK to plasmin(ogen). To test the activity of a drug in an animal model, it is preferable to use a preparation of the highest possible purity. For the in vivo evaluation of PLATSAK, the purification procedure used at that stage had to be optimized for large-scale purification. Although different resins, different metals and different elution procedures were tested, none proved to be sufficient for purifying the large amounts of purified proteins that were required for the in vivo studies. This lack of a suitable purification method left no other choice than to use an enriched intracellular extract of PLATSAK-producing E. coli cells for the animal studies. Three baboons were studied in a well-established model of arterial and venous thrombosis (Cadroy et al, 1989) to evaluate the in vivo effectivity of PLATSAK on the inhibition of platelet deposition. The thrombogenic devices that were used typically represent conditions under which an arterial-type (using graft material) and a venous-type (using expansion chambers) thrombus form. In the control studies an equimolar extract of native E. coli proteins were administered as a bolus and haemostasis was not markedly affected. The same baboons were subsequently treated with a bolus of 3.68 mg/kg PLATSAK. After 120 minutes inhibition was approximately 50% on the graft material and 85% in the expansion chamber, when compared to the number of platelets deposited at 42 130 minutes in the control studies. Although no significant inhibition of platelet aggregation could be observed in the in vitro studies, it appeared that platelet aggregation was markedly inhibited by PLATSAK in the in vivo model. This illustrated that results of in vitro platelet aggregation studies not necessarily predict the in vivo antiplatelet effectivity as is also suggested by Cerskus et al (1980). aPTI was lengthened to more than 120 seconds, indicating significant inhibition of thrombin. Furthermore, appreciably less TAT complexes were formed, indicating less thrombin available for complex formation with antithrombin Ill. FOP appeared in the plasma towards the end of the control studies and did not increase after treatment of PLATSAK. It thus appeared that PLATSAK did not result in additional activation of fibrinolysis during the two hours of the study. However, removal of the permanent AV-shunts at the end of the study was accompanied by some bleeding, suggesting that activation of fibrinolysis by PLATSAK was taking place. The combination of different activities into one molecule is not in concert with the current pharmacological approach. Most agents presently studied for their potential pharmacological use, are focused on only one specific activity, like hirudin for the inhibition of thrombin (Markwardt, 1994), annexin V for inhibition of factor Xa (Thiagarajan and Benedict, 1997), integrelin for the prevention of platelet-platelet interactions (Tcheng et aI, 1995) and staphylokinase for the activation of plasminogen (Collen et ai, 1996). Furthermore, due to its size and composition, PLATSAK is highly immunogenic and can only be administered once. To enhance its ability as a potential antithrombin and fibrinolytic agent the SAK portion need to be reduced to the minimum structure required for activity. The published structure of SAK in complex with u-plasmin can be used to determine the parts of SAK involved in binding (Jespers et aI, 1999). A minimal molecule with sufficient fibrinolytic and a low immune response can thus be constructed. A multifunctional agent like PLATSAK is not likely to be used as a routine antithrombin and will only be used under very specific and strictly controlled circumstances. The strict monitoring of the effect of PLATSAK on haemostasis can only be done in hospital circumstances. Due to its potent 131 activity it will not be used as precaution, but rather as treatment during procedures like endarterectomy and angioplasty. Furthermore, it may also be used for the treatment of thrombotic events like deep venous thrombosis. In order to ascertain the exact in vivo potential of PLATSAK the large scale purification should be optimized, probably by using alternative fusion systems. The evaluation of other fusion, expression and purification systems, however, falls outside the boundaries of the current study. Effective large scale purification of PLATSAK will enable me to study more animals to determine a dose dependent response and to evaluate possible side-effects and bleeding tendencies. Since this study was aimed at the construction, production and determination of the pharmacological value of a novel antithrombotic agent, it did not give new insights into the complex interactions of haemostasis and thrombosis. However, the interesting finding that both arterial and venous thrombosis was equivalently inhibited by PLATSAK might shed some light on the respective thrombotic mechanisms. This needs to be investigated with further in vivo studies. The important contribution of this study is that it proved possible to design a multifunctional recombinant chimeric protein that can modulate haemostasis and thrombosis in animal models. The study verified that it is feasible to combine different haemostatic strategies with success into the same molecule. The results obtained on three baboons were regarded as sufficient proof that combination of different inhibitory moieties into one molecule is a viable strategy for the development of anti thrombotic agents. 7.1 REFERENCES: Cadroy Y, Houghten RA, Hanson SR. RGDV peptide selectively inhibits platelet-dependent thrombus formation in vivo: Studies using a baboon model. J Clin Invest 1989;84:939-44. 132 Cerskus AL, Ali M, Davies BJ, McDonald JWD. Possible significance of small numbers of functional platelets in a population of aspirin-treated platelets in vitro and in vivo. Thromb Res 1980;84:939-44. Collen 0, Moreau H, Stockx L, Vanderschueren S. Recombinant staphylokinase variants with altered immunoreactivity, II: thrombolytic properties and antibody induction. Circulation 1996;94:207-16. Jespers L, Vanwetswinkel S, Lijnen HR, Van Herzeele N, Van Hoef B, Demarsin E, Collen 0, De Maeyer M. Structural and functional basis of plasminogen activation by staphylokinase. Thromb Haemostas 1999;81 :479-85. Markwardt F. The development of hirudin as an antithrombotic drug. Thromb Res 1994;74:1-23. Martin PO, Robertson W, Turk 0, Huber R, Bode W, Edwards BFP. The structure of residues 7-16 of the Aa-chain of human fibrinogen bound to bovine thrombin at 2.3-A resolution. J Bioi Chem 1992;267:7911-20. Naski MC, Fenton II JW, Maraganore JM, Olsen ST, Shafer JA. The COOH- terminal domain of hirudin. J Bioi Chem 1990;265:13484-9. Rydel TJ, Ravichandran KG, Tulinsky A, Bode W, Huber R, Roitsch C, Fenton II JW. The structure of a complex of recombinant hirudin and human a- thrombin. Science 1990;249:277-80. Tcheng JE, Harrington RA, Kottke-Marchant K, Kleinman NS, Ellis SG, Kereiakes DJ, Mick MJ, Navetta FI, Smith JE, Worley SJ, Miller JA, Joseph DM, Sigman KN, Kitt MM, Du Mée CP, Califf RM, Topoi EJ. Multicenter, randomized, double-blind, placebo-controlled trial of the platelet integrin glycoprotein Ilb/llla blocker integrelin in elective coronary intervention. Circulation 1995;91 :2151-7. 133 Thiagarajan P, Benedict CR. Inhibition of arterial thrombosis by recombinant annexin V in a rabbit carotid artery injury model. Circulation 1997;96:2339-47. 134 SUMMARY Platelets and coagulation both play a pivotal role in thrombosis, one of the major life-threatening diseases in our society. We have thus experienced a drastic increase in the development of potent and secure antithrombotic, antiplatelet and fibrinolytic agents during the past decade. Recently, much research has been devoted to the development of chimeric proteins, where haemostasis is simultaneously targeted at different levels. For the purpose of this study, such a chimera, named PLATSAK, was designed. A 29 amino acid antithrombotic and antiplatelet peptide, comprising three inhibitory regions, was linked to staphylokinase via a cleavable factor Xa recognition sequence. The overall activities of PLATSAK should include inhibition of thrombin, prevention of platelet aggregation and activation of fibrinolyis. The gene encoding PLATSAK was expressed in E. coli cells under controlled conditions. PLATSAK was produced as a strongly expressed protein of 18 kDa and was purified form native E. coli proteins using metal affinity chromatography. In vitro analysis of PLATSAK activity revealed strong inhibition of thrombin and potent fibrin degradation. However, no effect on platelet aggregation could be observed. Several attempts at producing more potent antiplatelet variants were unsuccessful. According to its in vitro activity, PLATSAK appeared to be a potent novel haemostatic agent and was prone to be evaluated in an in vivo system. The in vivo activity of PLATSAK was evaluated by assessing its effect on platelet deposition in a baboon model of arterial and venous thrombosis. Dacron vascular graft segments and expansion chambers, inserted as extensions into permanent femoral arteriovenous shunts, were used to simulate arterial and venous thrombosis, respectively. PLATSAK (3.68 mg/kg) was administered as a bolus. Platelet deposition onto the graft surface and in the expansion chamber was imaged in real time with a scintillation camera as the deposition of lllln-labelled platelets. After two hours, platelet deposition in the graft 135 I, segments and expansion chambers was inhibited by 50% and 85% respectively when compared to control studies. The aPTT was lengthened to >120 seconds. Interestingly, the level of FOP in plasma did not increase after administration of PLATSAK. These results demonstrate that PLATSAK effectively inhibited platelet deposition in both arterial- and venous-type thrombosis an animal model. This is in contrast to the lack of the antiplatelet activity of PLATSAK in vitro. This illustrates that in vitro platelet aggregation results can not be directly applied to an in vivo situation. In summary, the recombinant production of a multifunctional haemostatic fusion protein, PLATSAK, was successful. In vitro PLATSAK showed significant antithrombin and fibrinolytic activity, but trivial antiplatelet activity. In vivo studies revealed that PLATSAK is a potent antithrombin and also prevented platelet deposition on thrombogenic material. The strong immuun response of PLATSAK however needs to be investigated and a variant with a weak immunogenic nature needs to be produced.