Defining Platelet-Derived Components Regulating the Rp Othrombinase Enzyme Complex Francis Ayombil University of Vermont

University of Vermont ScholarWorks @ UVM

Graduate College Dissertations and Theses Dissertations and Theses

2016 Defining Platelet-Derived Components Regulating The rP othrombinase Enzyme Complex Francis Ayombil University of Vermont

Follow this and additional works at: https://scholarworks.uvm.edu/graddis Part of the Biochemistry Commons

Recommended Citation Ayombil, Francis, "Defining Platelet-Derived Components Regulating The rP othrombinase Enzyme Complex" (2016). Graduate College Dissertations and Theses. 533. https://scholarworks.uvm.edu/graddis/533

This Dissertation is brought to you for free and open access by the Dissertations and Theses at ScholarWorks @ UVM. It has been accepted for inclusion in Graduate College Dissertations and Theses by an authorized administrator of ScholarWorks @ UVM. For more information, please contact [email protected].

DEFINING PLATELET-DERIVED COMPONENTS REGULATING THE PROTHROMBINASE ENZYME COMPLEX

A Dissertation Presented

Francis Ayombil

The Faculty of the Graduate College

The University of Vermont

In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Specializing in Biochemistry

May, 2016

Defense Date: March 15, 2016 Dissertation Examination Committee:

Paula B. Tracy, Ph.D., Advisor Bryan A. Ballif, Ph.D., Chairperson Christopher L. Berger, Ph.D. Stephen J. Everse, Ph.D. Scott W. Morrical, Ph.D. Cynthia J. Forehand, Ph.D., Dean of the Graduate College

ABSTRACT

At sites of vascular injury, the critical blood clotting enzyme thrombin is generated from prothrombin via Prothrombinase, a macromolecular, Ca2+-dependent enzymatic complex consisting of the serine protease factor Xa and the non-enzymatic cofactor factor Va, assembled on the membranes of activated platelets. Platelets regulate thrombin formation by providing specific binding sites for the components of Prothrombinase and by releasing a platelet-derived factor V/Va molecule that is more procoagulant than its plasma counterpart and partially resistant to proteolytic inactivation. This dissertation identifies and characterizes the subpopulation of platelet-derived factor V/Va that is responsible for the observed protease resistance, and the mechanism by which Prothrombinase bound to platelets differs from a model system using vesicles composed of 75% phosphatidylcholine (PC) and 25% phosphatidylserine (PS), PCPS vesicles. Previous studies have demonstrated that activated platelets release a dissociable pool of factor V/Va and a non-dissociable, membrane-bound pool, which is covalently attached to the platelet membrane through a glycosylphosphatidyl inositol (GPI) anchor. Data described herein demonstrate unequivocally that the pool of platelet-derived factor V/Va that is resistant to proteolytic inactivation by activated protein C is provided exclusively by the non-dissociable GPI-anchored pool. Further, although this factor Va pool is susceptible to proteolysis by plasmin, the fragments formed are associated with sustained and increased cofactor activity. These observations indicate that tethering of factor Va to the membrane surface via a GPI anchor imparts resistance to proteolytic inactivation and sustained thrombin generation at sites of vascular injury. For several years it has been known that Prothrombinase assembled on PCPS vesicles does not mimic that bound to platelets. While both enzymes cleave prothrombin at Arg271 and Arg320 to form thrombin, prothrombin activation proceeds via the prethrombin-2 pathway (initial cleavage at Arg271) on the platelet surface, in contrast to the meizothrombin pathway (initial cleavage at Arg320) on PCPS vesicles. Using thrombin active site inhibitors, we demonstrate that the preference for either pathway is dictated by the conformation in which prothrombin is bound by the membrane-bound enzyme. The prethrombin-2 pathway of prothrombin activation catalyzed by platelet-bound Prothrombinase is a direct consequence of configuring prothrombin in a proteinase-like state resulting in the exposure of a pseudo-active site that can be stabilized by active site thrombin inhibitors. Conversely, prothrombin is preferentially configured in the zymogen- like state on PCPS vesicles where the meizothrombin pathway is preferred. Additional support for the differential assembly of Prothrombinase on the platelet surface is provided by observations made using prethrombin-1, an intermediate formed by cleavage of prothrombin at Arg155 by the formed thrombin. Prethrombin-1 is converted into fragment-2 and thrombin by platelet-bound Prothrombinase at a substantially higher rate than vesicle-bound Prothrombinase. The decreased rate of prethrombin-1 activation in the model system is due, in part, to inhibition of the vesicle-bound enzyme by the fragment-2 generated in the reaction. Taken together, these data not only provide important molecular insights into the mechanisms by which Prothrombinase bound to activated platelets at sites of vascular injury regulates the procoagulant response to effectively support robust thrombin generation, but also provides potential mechanistic sites that could be targeted therapeutically.

DEDICATION

I would like to dedicate this dissertation to my mom and dad, Mary Ayam and James

Ayombil, whose love and support throughout my studies has been unconditional. Despite being the first generation of my nuclear family to reach college, my parents and siblings did not hesitate to urge me on to reach the highest level of academic education and I am blessed and grateful for their patience, prayers and inspiration. This dissertation is also dedicated to my fiancée, Katherine Bowen for all the support and companionship you have provided me throughout my studies and particularly during the final few months when the future looked uncertain. You stood by me and encourage me along the way and I am forever grateful and looking forward to calling you my wife soon.

ACKNOWLEDGMENTS

Firstly, I would like to thank my mentor and advisor, Dr. Paula Tracy, for the opportunity to work and pursue my graduate studies in her laboratory. My passion for research and the field of coagulation increased significantly under her guidance and I have become a better scientist now. I am also grateful for the informal life conversations and advice she offered me during my stay here and I will forever be grateful. Secondly, thanks go to Dr. Bryan

Ballif, Dr. Christopher Berger, Dr. Stephen Everse, and Dr. Scott Morrical, who as members of my thesis committee provided all the useful feedbacks and contributed directly to the shaping of my dissertation work presented here. I enjoyed our annual committee meetings and thanks for accepting to nurture my graduate career.

Also, I would also like to thank all the past members of the Tracy laboratory for the friendship and favorable working environment which helped me grow as a person and for all the useful suggestions during lab meetings. Special thanks goes to Dr. Beth Bouchard,

Dr. Jay Silveira and Dr. Laura Haynes for the constructive critiques of both my experimental designs and written work over all these years. Many thanks to all my colleagues and friends in the Biochemistry program for the nights out and chalk talks especially to Jamie Abbott and Sarah Abdalla.

Finally, I would like to thank the Biochemistry Dept. and the CMB program for their financial support. Thanks to the Claey’s family, the Bowen family, the Swamp

Donkey’s, my soccer buddies and all the awesome folks I met here in Burlington. You have been instrumental in helping me feel ‘home away from home’ during my studies.

iii

TABLE OF CONTENTS

DEDICATION ...... ii

ACKNOWLEDGMENTS ...... iii

LIST OF TABLES ...... xii

LIST OF FIGURES ...... xiii

CHAPTER ONE: COMPREHENSIVE LITERATURE REVIEW ...... 1

1.1. Hemostasis Overview ...... 1

1.1.1 Introduction to Hemostasis ...... 1

1.1.2 The Coagulant Response – Formation of a Stable Blood Clot ...... 7

1.2. The Generation of Thrombin by Prothrombinase ...... 10

1.2.1. Coagulation Factor V ...... 14 1.2.1.1. Structural Features of Plasma-derived Factor V and Factor Va ...... 14 1.2.1.2. Post-translational Modifications on Factor V ...... 17 1.2.1.3. Functional Properties of Factor Va ...... 22 1.2.1.4. Inactivation of Factors V and Va ...... 28 1.2.1.5. Origins of Plasma-derived and Platelet-derived Factor V/Va ...... 34 1.2.1.6. Functional and Phenotypic Differences between Platelet- derived and Plasma-derived factor V ...... 37 1.2.1.7. Physiological Significance of Platelet-derived vs. Plasma-derived Factor Va ...... 39

1.2.1.8. Megakaryocyte-Endocytosis of Plasma-derived Factor V: Formation of the Functionally and Phenotypically Unique Platelet-derived Factor V/Va pool ...... 41 1.2.1.9. Characterization of the Non-dissociable Platelet-Derived Factor Va Pool .. ……………………………………………………………………………44

1.2.2. Structural Features of Prothrombin ...... 45 1.2.2.1. The Activation of Prothrombin to Thrombin ...... 47

1.2.3. The Assembly of Prothrombinase on the Activated Platelet Surface versus Anionic Phospholipid Vesicles ...... 53

1.3. Summary of Introduction ...... 59

CHAPTER TWO: Platelet-bound prothrombinase preferentially binds its substrate in a ‘proteinase-like state’ to effect prothrombin activation by the prethrombin-2 pathway...... 60

ABSTRACT ...... 61

INTRODUCTION ...... 63

MATERIALS AND METHODS ...... 66

Materials ...... 66

Proteins ...... 66

Iodination of prothrombin...... 67

Isolation of platelets from whole human blood ...... 67

Incorporation of b-FPRcK into prothrombin ...... 68

Prothrombin activation pathway in the presence of DAPA...... 69

Kinetics of prothrombin activation in the presence of DAPA ...... 70

Spectrozyme-TH hydrolysis and estimation of thrombin concentration ...... 70

SDS-PAGE, Western blotting and autoradiography ...... 71

Kinetic data analyses ...... 71

RESULTS ...... 72

Maturation of prothrombin’s pro-active site upon binding to platelet-bound Prothrombinase ...... 72

Stabilization of the proteinase-like state of prothrombin by DAPA dictates its activation by both platelet- and PCPS-bound Prothrombinase ...... 72

DISCUSSION ...... 75

ACKNOWLEDGEMENT ...... 78

AUTHORSHIP ...... 79

REFERENCES ...... 80

FIGURES ...... 85

CHAPTER THREE: Platelet-bound prothrombinase has a unique enzyme-substrate interface that evades product inhibition by prothrombin fragment-2 ...... 92

ABSTRACT ...... 93

INTRODUCTION ...... 95 vi

MATERIALS AND METHODS ...... 98

Materials ...... 98

Coagulation proteins ...... 98

Isolation of platelets from blood ...... 99

Time course of prothrombin activation by platelet-bound Prothrombinase ...... 100

Michaelis-Menten kinetic constants of prethrombin-1 activation by human Prothrombinase ...... 100

Progress curves of complete prethrombin-1 activation by human Prothrombinase . 101

The effect of fragment-2 on prethrombin-1 activation by human prothrombinase .. 102

Spectrozyme TH hydrolysis for determination of thrombin concentrations ...... 102

Electrophoretic analyses of protein samples...... 102

Data analyses ...... 103

RESULTS ...... 105

Activation of prethrombin-1 formed in situ by platelet-bound Prothrombinase ...... 105

Activation of exogenous prethrombin-1 by PCPS-bound Prothrombinase ...... 105

Kinetics constants of prethrombin-1 activation by human Prothrombinase ...... 106

Exogenous fragment-2 inhibition of thrombin generation from prethrombin-1 by PCPS- and platelet-bound prothrombinase ...... 107

Progress curves of prethrombin-1 activation by human prothrombinase ...... 107

vii

DISCUSSION ...... 110

ACKNOWLEDGEMENTS ...... 115

REFERENCES ...... 116

FIGURES ...... 123

CHAPTER FOUR: Functional characterization of the human platelet-derived, glycophosphatidylinosital (GPI)-anchored pool of factor Va ...... 131

ABSTRACT ...... 132

INTRODUCTION ...... 134

MATERIALS AND METHODS ...... 137

Materials ...... 137

Proteins ...... 138

Isolation of human platelets followed by isolation of the dissociable pool of platelet- derived factor Va and the factor Va that is GPI-anchored to thrombin-activated platelets ...... 138

Use of analyses of kinetics of prothrombin activation to define the role of the GPI- anchored factor Va cofactor pool in Prothrombinase assembly and function...... 139

Determination of the effect of APC and plasmin on the cofactor activity associated with dissociable factor Va pool and the GPI-anchored platelet-derived factor Va pool ...... 141

SDS-PAGE and Western blotting analyses ...... 142

viii

RESULTS ...... 144

Characterization of the contribution of GPI-anchored platelet-derived factor Va to Prothrombinase assembly and function on thrombin-activated platelets ...... 144

DISCUSSION ...... 149

ACKNOWLEDGEMENTS ...... 152

AUTHORSHIP ...... 152

REFERENCES ...... 153

TABLE ...... 160

FIGURES ...... 161

CHAPTER FIVE: Characterization of the functional properties of demannosylated plasma-derived factors V and Va ...... 168

ABSTRACT ...... 169

INTRODUCTION ...... 170

MATERIALS AND METHODS ...... 173

Materials ...... 173

Proteins ...... 174

Removal of high mannose glycans on plasma-derived factor V by Endoglycosidase H ...... 174

Endocytosis of plasma-derived factor V by megakaryocyte-like cell-lines (CMKs) ...... 175

The effect of endoglycosidase H treatment on plasma-derived factor Va cofactor activity ...... 175

APC-catalyzed proteolytic inactivation of endoglycosidase H-treated factor Va .... 176

SDS-PAGE and Western blotting analyses ...... 177

RESULTS ...... 178

Removal of high mannose glycans from plasma-derived factor V ...... 178

Effect of endoglycosidase H treatment on factor V and factor Va endocytosis by megakaryocyte-like cell-lines ...... 178

Effect of demannosylation on factor Va cofactor activity ...... 179

Effect of endoglycosidase H treatment on APC-mediated inactivation of factor Va ...... 180

DISCUSSION ...... 182

ACKNOWLEDGEMENTS ...... 184

AUTHORSHIP ...... 185

REFERENCES ...... 186

FIGURES ...... 192

CHAPTER SIX: PERSPECTIVES AND FUTURE DIRECTIONS ...... 198

COMPREHENSIVE BIBLIOGRAPHY ...... 209

LIST OF TABLES

Table Page

Chapter Three: Platelet-bound prothrombinase has a unique enzyme-substrate interface that evades product inhibition by prothrombin fragment-2

Table I. Kinetic parameters for prethrombin-1 activation on membrane surfaces ...... 120 Table II. Normalized rates of thrombin generation from prethrombin-1 or prothrombin activations catalyzed by Prothrombinase bound to PCPS vesicles or thrombin-activated platelets in the presence or absence of Frag 2 and/or thrombin ...... 121 Table III. Parameters of non-linear enzyme kinetic time courses for prethrombin-1 activation by PCPS-bound Prothrombinase...... 122 Chapter Four: Functional characterization of the human platelet-derived, glycophosphatidylinosital (GPI)-anchored pool of factor Va

Table I. Contribution of the two platelet-derived, membrane-bound pools of factor Va to the expression of Prothrombinase activity: dissociably-bound factor Va vs the GPI- anchored factor Va ...... 160

xii

LIST OF FIGURES

Figure Page

Chapter One: Comprehensive Literature Review Figure 1. Components of the human blood and blood vessel...... 2 Figure 2. The Coagulation Cascade...... 9 Figure 3. Contributions of the Components of Prothrombinase to Thrombin Generation...... 13 Figure 4. Domain Structures of Human Factor V and Factor Va...... 15 Figure 5. Schematic of the Post-translational Modifications Identified on Factor V...... 20 Figure 6. Thrombin-mediated Activation of Factor V to Factor Va...... 23 Figure 7. Proteolysis of Factor V by Factor Xa...... 25 Figure 8. Schematic of Factor Va and Factor V Proteolytic Inactivation by Activated Protein C...... 31 Figure 9. Endocytic trafficking of plasma-derived factor V in megakaryocytes to form the unique platelet-derived factor V/Va molecules...... 43 Figure 10. Structural Domains of Human Prothrombin...... 46 Figure 11. The Mechanism of Prothrombin Activation to Thrombin...... 49 Figure 12. The Formation of Prethrombin-1 Intermediate During Prothrombin Activation and Further Conversion to Thrombin...... 52

Chapter Two: Platelet-bound prothrombinase preferentially binds its substrate in a ‘proteinase-like state’ to effect prothrombin activation by the prethrombin-2 pathway

Figure 1. Prothrombinase-mediated pathway of prothrombin activation in the presence of thrombin inhibitors...... 85 Figure 2. Biotinylated FPRck is incorporated into PCPS-bound and platelet-bound prothrombinase...... 86 Figure 3. The effect of increasing concentrations of DAPA on prothrombin activation by platelet-bound Prothrombinase...... 87

xiii

Figure 4. The effect of DAPA on the appearance of prothrombin activation species in reactions by platelet-bound Prothrombinase...... 89 Figure 5. The effect of DAPA on prothrombin activation by PCPS-bound Prothrombinase...... 90 Figure 6. Substrate depletion curves for prothrombin at variable DAPA concentrations in reactions by PCPS-bound Prothrombinase...... 91

Chapter Three: Platelet-bound prothrombinase has a unique enzyme-substrate interface that evades product inhibition by prothrombin fragment-2

Figure 1. Schematic of prethrombin-1 formation and activation...... 123 Figure 2. Prothrombin activation and in situ prethrombin-1 turnover by platelet- Prothrombinase...... 124 Figure 3. Prethrombin-1 activation by Prothrombinase assembled on PCPS vesicles. .. 125 Figure 4. Kinetics of prethrombin-1 activation on PCPS vesicles and platelets...... 126 Figure 5. Relative rates of thrombin generation from prethrombin-1 or prothrombin activation mixtures containing exogenous fragment-2...... 127 Figure 6. Progress curves of prethrombin-1 activation by PCPS- or platelet-bound Prothrombinase: empirical vs theoretical ...... 129 Figure 7. Progress curves of prethrombin-1 activation by PCPS-bound Prothrombinase...... 130

Chapter Four: Functional characterization of the human platelet-derived, glycophosphatidylinosital (GPI)-anchored pool of factor Va

Figure 1. The binding interactions of factor Xa and super washed platelets as measured by prothrombin activation kinetics...... 161 Figure 2. Determination of the kinetic constants of prothrombin activation by Prothrombinase bound to super washed platelets...... 162 Figure 3. Comparison of APC-catalyzed inactivation patterns of GPI-anchored, dissociable pool of platelet-derived, and plasma-derived factor Va molecules...... 163

xiv

Figure 4. Plasmin-mediated proteolysis of plasma-derived factor Va vs non-dissociably bound platelet-derived factor Va...... 165 Figure 5. Immunoblotting analyses of plasmin-catalyzed proteolysis of the dissociable pool of platelet-derived factor Va...... 167

Chapter Five: Characterization of the functional properties of demannosylated plasma- derived factors V and Va

Figure 1. Sites of N-linked glycosylation on the factor Va heavy and light chains...... 192 Figure 2. Electrophoretic mobility shifts following treatment of factor V and factor Va with endoglycosidase H...... 193 Figure 3. Endocytosis and proteolytic processing of endoglycosidase H-treated factor V and factor Va by a megakaryocyte-like cell line...... 194 Figure 4. Kinetic analyses of prothrombin activation by PCPS-bound Prothrombinase assembled in the presence of demannosylated factor Va...... 195 Figure 5. Schematic of activated protein C-catalyzed inactivation of factor Va heavy. 196 Figure 6. APC-mediated proteolysis of thrombin-activated, endoglycosidase H-treated factor V...... 197

CHAPTER ONE: COMPREHENSIVE LITERATURE REVIEW

1.1. Hemostasis Overview

1.1.1 Introduction to Hemostasis

According to the most recent National Heart, Lung, and Blood Institute report (2015), cardiovascular diseases including coronary heart disease, atherosclerosis, cerebrovascular disease, and peripheral vascular disease account for more than 30% of all deaths in the

United States as of 2010. In particular, coronary heart disease remains the most common heart disease resulting in slightly less than half a million deaths annually (~400,000 people representing ~33% of the total cardiovascular disease deaths per year). Recent statistics show that while cardiovascular diseases in the United States have continued to decline in the last 20 years, they remain comparatively higher relative to other countries in the world especially among the female population (National Heart, Lung, and Blood Institute,

Chapter 4. Disease Statistics. Retrieved 8th January 2016; http://www.nhlbi.nih.gov/).

Taken together, the occurrence of these diseases highlights the importance of sustaining cardiovascular research to improve our understanding of the biological and biochemical processes regulating the various functions of the vascular system to effect better disease prevention and treatment.

Higher vertebrates including humans have evolved an ultra-secure and effective distribution system for the delivery of nutrients, defense molecules, and gases to various parts of the body. Our closed vascular system is comprised of a fluid medium (blood), fluid-filled spaces (such as the heart, and blood vessels), surrounded by continuous connective tissues which permeate all tissues in the body (Ruppert, 1983). The human

blood is comprised of a pale yellow liquid known as plasma. The plasma is abundant in soluble proteins including clotting factors, defense particles, and a suspension of specialized cells such as platelets, white blood cells and red blood cells which circulate in the blood vessel network around the entire body (Figure 1).

Figure 1. Components of the human blood and blood vessel. Most of the human blood is comprised of fluid, called the plasma, in which several soluble proteins and cellular components float and circulate around the body. Red blood cells, white blood cells, and platelets are abundant in the plasma and play key roles in both the immune and hemostatic responses. Separating the blood from the subendothelial matrix are the endothelials cells. A breach of vascular integrity allows plasma-based proteins and cells to interact directly with key components in the subendothelial matrix to mount a robust hemostatic response. Evident from the color of effusing blood, red blood cells form the majority of cells in the blood (35-50% of a total volume of ~5 L) and are primarily responsible for the delivery of oxygen (via hemoglobin) to various parts of the body. The white blood cells, on the other hand, play crucial roles during the immune response to invasion by foreign pathogens, and also during the inflammatory response (Williams, 1990). Likewise, human platelets have been implicated in both immune and inflammatory responses. However,

platelets function mainly as the guardians that maintain normal blood circulation within the blood vessels of the vascular system.

Complex, yet ingenious interrelated systems exist to maintain the fluidity of the blood in the vascular system while allowing for the rapid formation of a solid blood clot to prevent excessive blood loss or hemorrhage subsequent to blood vessel injury. These interrelated systems are collectively referred to as hemostasis when they are invoked as part of the body’s normal defense mechanisms to prevent blood loss. Alternatively, these same interrelated systems are invoked during thrombosis which refers to unwanted, pathological, and in some instances life-threatening clot formation. Thrombosis is indeed a pathologic extension of the normal hemostatic response. The hemostatic response in either phenomenon is fundamentally the same, whereas the initiating events are considerably different.

Since platelets are essential for hemostasis, they will play a similar role in thrombosis, which can cause significant mortality and morbidity due to myocardial infarction and stroke. The role that platelets play in angiogenesis, likewise can be beneficial, such as in development of collateral vessels, or detrimental such as in cancer progression and metastasis (Nurden, 2011). Likewise, the role that platelets play in wound healing (Nurden et al., 2008) is similar to their role in inflammation (Gawaz et al., 2005).

One mechanism by which they fulfill these various roles is through their ability to support the assembly and function of the macromolecular enzyme complex, Prothrombinase, which effects the proteolytic conversion of the zymogen prothrombin to the important bioeffector and bioregulatory molecule thrombin, the central enzyme of the coagulation cascade

(Bouchard et al., 2007). As such, platelets and platelet-derived thrombin play key roles in several (patho)physiological processes.

When the vascular integrity is breached, platelets and coagulation factors, circulating in blood in inert forms, interact to effect a “cascade” of events referred to as blood coagulation. The first phase of this response, primary hemostasis, requires the exposure of subendothelial components to which circulating platelets can bind. Platelet exposure to ligands such as von Willebrand factor and collagen triggers their adhesion, activation and degranulation, and aggregation to form a platelet plug (Ruggeri, 2009), in order to localize the subsequent reactions to the site of injury. Activated platelets release constituents from α– and dense granules that promote the hemostatic process and express functional receptors which reinforce the activation and aggregation processes. Likewise, in what is considered a crucial event for the success of secondary hemostasis, activated platelets undergo significant shape change resulting in a 200-fold increase in membrane surface area and express anionic phospholipid binding sites and protein receptors that provide essential catalytic sites for the assembly and function of the coagulation complexes

(Walsh, 1974; Robert et al., 1997; Tracy, 2001). The exposure of subendothelial tissue factor, which complexes with plasma factor VIIa, initiates secondary hemostasis where the coagulation cascade is triggered leading to a series of zymogen to protease conversions and further assembly of two enzyme complexes, intrinsic Tenase and Prothrombinase. The activated platelet membrane provides binding sites for the assembly of both enzyme complexes, which convert factor X to Xa and prothrombin to thrombin, respectively.

Thrombin further converts fibrinogen into fibrin forming the meshwork which strengthens

the unstable platelet plug. The formation of catalytic amounts of thrombin will significantly amplify the coagulation response by its ability to activate platelets and several of the other factors required for hemostatic competence.

The ability of platelets to regulate thrombin generation is central to their roles in hemostasis and thrombosis. While all platelets may become activated following agonist- mediated stimulation, it is now accepted that only a subpopulation of platelets appear competent to function in supporting Prothrombinase assembly and function. This same platelet subpopulation is unique in that it expresses significantly higher membrane concentrations of adhesive receptors on its membrane, suggesting that this population would be the first one recruited to the site of injury (Fager et al., 2010).

Prothrombinase is defined as a macromolecular, enzymatic complex consisting of a Ca2+-dependent, membrane-bound complex of the cofactor factor Va and the serine protease factor Xa (Bouchard et al., 2007). Factor Va can be derived from plasma, following the thrombin-catalyzed proteolytic activation of the procofactor, factor V

(Esmon, 1979; Monkovic & Tracy, 1990a), or subsequent to agonist-induced platelet activation resulting in both factor Va release and it membrane expression (Walsh, 1974;

Harrison & Cramer, 1993; Bouchard et al., 1997b). Thus, in the initial stages of the hemostatic response, activated platelets provide the essential cofactor for Prothrombinase assembly and function, resulting in thrombin generation. Following thrombin generation, additional platelet activation ensues and plasma-derived factor V can be activated to factor

Va. Consequently, subsequent to thrombin-catalyzed platelet activation, platelet-derived and plasma-derived factor Va bind to specific, functional and saturable sites on the

activated platelet membrane surface and in so doing form at least part of the receptor for factor Xa ensuring that prothrombin is activated to thrombin via platelet-bound

Prothrombinase (Bouchard et al., 1997b). The absolute requirement of platelet-derived factor Va for the initiation and maintenance of an effective hemostatic response

(Borchgrevink & Owren, 1961; Janeway et al., 1996; Ang et al., 2009; Bouchard et al.,

2015) is discussed in greater detail in Section 1.2.1.7. Remarkably, the procoagulant platelet subpopulation discussed above, is also defined by those platelets expressing the

GPI-anchored factor Va (Wood et al., 2008). Thus, by adhering, activating and aggregating at a vascular site of injury, platelets are not only able to form a platelet plug, but are also able to provide an essential cofactor in its active form that will be retained at the site of injury protected from inactivation.

The mechanisms by which activated platelets facilitate the appropriate assembly and function of Prothrombinase also guarantee that a procoagulant state is sustained. Since prothrombin’s proteolytic activation by Prothrombinase requires two cleavages at distant sites, two different intermediates can be generated depending on which site is cleaved first.

Thus, prior to thrombin formation platelet-bound Prothrombinase has the option of generating either a prethrombin-2 intermediate which is catalytically inactive, or a meizothrombin intermediate, which is catalytically active and functions as an anticoagulant. Since platelet-bound Prothrombinase only releases the procoagulant enzyme thrombin as its product, robust procoagulant activity is sustained. Consequently, the focus of these studies was to define the mechanism by which platelet-bound Prothrombinase recognizes and activates prothrombin, as well as defining the unique functional properties

of the acquired cofactor, platelet-derived factor Va, which is stored within the platelet’s α- granules.

1.1.2 The Coagulant Response – Formation of a Stable Blood Clot

Normal hemostasis requires that the coagulant response be initiated coincident with the formation of a platelet plug at a site of vascular injury. This will ensure the generation of thrombin and the deposition of fibrin to reinforce the unstable platelet plug formed during the primary phase of the hemostatic response. Insoluble fibrin deposition, and the ultimate cessation of a bleeding episode, are the consequence of various highly-regulated proteolytic activation reactions that involve both plasma- and platelet-derived proteins of the coagulation system. These proteolytic activation reactions are referred to as the coagulation cascade and are crucial to this process. The proteins participating in the coagulation cascade are designated by Roman numerals that are assigned in the order in which they were discovered, e.g. Factor I, now commonly referred to as fibrinogen, and factor V, which retained its original designation. The proteolytically activated form of a coagulant zymogen or procofactor is indicated by a lower case ‘a’ such as factor Va. The physiologically-relevant coagulation cascade mediates a series of zymogen to enzyme conversions (activations) which eventually produces physiological concentrations of thrombin required to stop a bleeding episode (Figure 2). Concurrent with the primary hemostatic response, the extrinsic pathway is initiated when the exposed subendothelial matrix presents reactive extravascular tissue factor (TF) molecules to circulating plasma factor VIIa (FVIIa). The TF-FVIIa binary association is referred to as the extrinsic Tenase enzyme complex, and initiates the extrinsic pathway of the coagulation response

(Nemerson, 1988). The extrinsic pathway is designated as such because the initiator of this pathway, tissue factor, does not circulate in blood, and hence is extrinsic to blood. Extrinsic

Tenase (TF-FVIIa) contributes to the initial generation of thrombin by channeling its proteolytic products to other enzymes such as Prothrombinase, and intrinsic Tenase.

Accordingly, factors IX and X are activated by the extrinsic Tenase complex via limited proteolysis to form factors IXa and Xa, respectively (Osterud & Rapaport, 1977) (Figure

2). Additionally, factor XI can be activated to XIa by thrombin under physiological conditions (Gailani & Renne, 2007). Factor XIa activates factor IX to factor IXa to ensure that the intrinsic Tenase complex is sustained during the coagulant response as the extrinsic

Tenase is rapidly inhibited (Kurachi & Davie, 1977) (Figure 2).

The formation of the serine protease factor IXa results in its association with its non-enzymatic cofactor, factor VIIIa, on the activated platelet membrane forming intrinsic

Tenase. This factors VIIIa-IXa-membrane ternary complex generates significant amounts of factor Xa from X as compared to extrinsic Tenase (Silverberg et al., 1977; van Dieijen et al., 1981). Consistent with this observation, the severe bleeding diatheses accompanying hemophilias A,and B are associated with deficiencies in plasma-derived factors VIII and

IX, respectively (Hemker & Kahn, 1967; Loeliger et al., 1967; Meyer et al., 1967). The final enzyme in the coagulation cascade, Prothrombinase, is formed when the serine protease factor Xa associates with its non-enzymatic cofactor, factor Va, on the activated platelet membrane in the presence of Ca2+. Prothrombinase activates prothrombin to thrombin, the central enzyme in the coagulation response (Barton et al., 1967; Hemker et al., 1967).

Figure 2. The Coagulation Cascade. Following an injury, the highly thrombogenic subendothelial extracellular matrix becomes exposed to plasma constituents allowing for platelet adhesion and activation and initiation of the hemostatic response. Subsequently, the extrinsic Tenase is formed comprised of plasma factor VIIa (FVIIa) and its subendothelial cofactor, tissue factor (TF) in the presence of Ca2+ (extrinsic pathway). Factors IX (FIX) and X (FX) are proteolytically activated to factors IXa (FIXa) and Xa (FXa), respectively by the TF-FVIIa complex. This TF-FVIIa complex is immediately ‘shut down’ by tissue factor pathway inhibitor (TFPI). Meanwhile, on the activated platelet surface, the FVIIIa-FIXa complex forms intrinsic Tenase which generates significant amounts of FXa for the platelet-bound Prothrombinase comprised of the FXa-FVa complex in the presence of Ca2+ to ensure thrombin generation from prothrombin. Thrombin sustains its own generation by further activating FXI to FXIa which then supplies the intrinsic Tenase with FIXa from FIX activation. Thrombin also converts FVIII to FVIIIa. Reactions of the Tenases, and Prothrombinase are significantly enhanced in the presence of cofactors: TF, FVIIIa, and FVa and membrane surfaces.

The thrombin formed, further activates soluble fibrinogen to insoluble fibrin allowing for the formation of the mesh-like network that strengthens the platelet plug (Ferry et al.,

1952). Final consolidation of the fibrin network is achieved through the actions of factor

XIIIa, the transglutaminase, which is formed subsequent to the thrombin-mediated cleavage of the zymogen, factor XIII. Factor XIIIa covalently cross-links fibrin molecules together to form a stable fibrin clot (Ferry, 1952; Ferry et al., 1952; Doolittle et al., 1971).

The initial amounts of thrombin generated when the coagulant response is initiated, amplifies the hemostatic response allowing for sustained thrombin generation required to stop a bleeding episode. Thrombin activates circulating platelets through their protease- activated receptors 1 and 4 (PARs 1 and 4) allowing for the recruitment of additional platelets to the site of injury (Kahn et al., 1998; Kahn et al., 1999). Furthermore, activated platelets express the catalytic membrane surfaces (Walsh, 1974) required for the assembly of the macromolecular enzyme complexes, intrinsic Tenase and Prothrombinase.

Thrombin also activates factors V and VIII to factors Va and VIIIa, respectively (Esmon,

1979; Nesheim & Mann, 1979; Vehar & Davie, 1980). Likewise, the zymogens factors

VII, and XI are activated to VIIa and XIa by thrombin, respectively (Radcliffe &

Nemerson, 1975; Gailani & Broze, 1991). All of these reactions are required for the coagulant response as shown previously in figure 2.

1.2. The Generation of Thrombin by Prothrombinase

Thrombin is often described as the principal enzyme (or protease) in coagulation due to its pivotal role in ensuring clot formation (procoagulant functions), cessation of the procoagulant response (anticoagulant functions such as activation of protein C), and

ultimately clot dissolution. The most important procoagulant functions of thrombin includes the proteolytic cleavage of the procofactors, factors V and VIII, fibrinogen, and the activation of platelets. Additionally, thrombin has been observed to influence both the pro- and anti-inflammatory cell signaling processes as well as promote wound healing. It is thought that thrombin has a limited role during the genesis of the coagulation response; however, the crucial role for this protease in subsequent hemostatic processes cannot be overemphasized as detailed above. Therefore, the generation of significant amounts of thrombin under physiological conditions following an injury must await the expression of procoagulant surfaces such as the activated platelet membrane.

Prothrombinase, a macromolecular enzyme complex comprised of a 1:1 stoichiometric, Ca2+-dependent association of the serine protease, factor Xa, and its non- enzymatic cofactor, factor Va, assembled on a procoagulant membrane surface (such as activated platelets or anionic phospholipids) mediates thrombin generation in vivo (Barton et al., 1967; Mann et al., 1990). Through limited proteolysis, platelet-bound

Prothrombinase catalyzes the activation of the plasma-derived prothrombin zymogen - it’s only known physiological substrate.

Previous studies have shown that the contribution of each non-protease component of the Prothrombinase enzyme is saturable and ensures maximum thrombin generation.

Removal of any one of the components lowers enzymatic activity to a physiologically insignificant amount (as shown in Figure 3). In the absence of Ca2+ ions, factor Xa alone generates negligible amounts of thrombin (0.004 mol/thrombin/min/mol factor Xa) in an extremely slow proteolytic reaction (Nesheim et al., 1979a). The relative rate of thrombin

generation by factor Xa is enhanced by two-fold (0.010 mol/thrombin/min/mol factor Xa) in the presence of Ca2+ ions. The addition of an appropriate membrane surface composed of anionic phospholipids to this reaction enhances the rate of thrombin generation by 10- fold (0.094 mol/thrombin/min/mol factor Xa) as shown in Figure 3. A further addition of the cofactor, factor Va, to complete optimal Prothrombinase enzymatic activity increases the relative rate of thrombin generation by ~300,000-fold (~1200 mol/thrombin/min/mol factor Xa) consistent with desired amounts of thrombin capable of an efficient hemostatic response compared to factor Xa alone (Nesheim et al., 1979a).

The substantial contribution of an appropriate membrane surface, physiologically provided by the activated platelet, to Prothrombinase function is demonstrated by approximately 800-fold decrease in thrombin generation (1.55 mol/thrombin/min/mol factor Xa) in reactions catalyzed by factor Xa in the presence of Ca2+ and factor Va.

Likewise, removal of the cofactor, factor Va from the Prothrombinase enzyme results in a

~10,000-fold reduction in thrombin generation. Thus, both the non-enzymatic cofactor, factor Va, and an appropriate anionic membrane surface are absolutely critical to the assembly and function of Prothrombinase expressing optimal catalytic efficiency (Nesheim et al., 1979a).

In summary, the amount of thrombin generated from prothrombin activation by all or some of the components of Prothrombinase is variable. Similarly, the pathway of prothrombin activation (as discussed in detail in section 1.2.3.1) and the intermediates generated during thrombin formation appear distinct and depend on the membrane surface and (or) components of Prothrombinase that are present. Such observations suggest a

differential membrane-dependent enzyme assembly and function which requires further investigation.

Figure 3. Contributions of the Components of Prothrombinase to Thrombin Generation. The relative rates of thrombin generation are indicated as a function of the constituents of the macromolecular prothrombinase enzyme: the serine protease, factor Xa, green color; the non-enzymatic cofactor, factor Va (showing its heavy chain, HC, and light chain, LC, grey color); assembled on a membrane surface, purple color, in the presence of Ca 2+, orange color. Adapted from the publication of Nesheim and colleagues (Nesheim et al., 1979a). 13

1.2.1. Coagulation Factor V

1.2.1.1. Structural Features of Plasma-derived Factor V and Factor Va

Coagulation factor Va is a key component of Prothrombinase and absolutely essential for normal hemostasis. The earliest observation of its importance in normal hemostasis was made by Dr. Paul Owren, who identified this protein in a patient with a variable, but significant, bleeding diathesis, which could be corrected by administering prothrombin deficient plasma (Owren, 1947). Factor V is also known as proaccelerin or labile factor due to its enhancement of factor Xa’s catalytic activity, and its instability during purification, respectively. The involvement of factor V/Va in the hemostatic response only occurs after initiation of the coagulation cascade suggesting that the participation of this molecule in Prothrombinase is highly regulated.

Plasma-derived factor V is synthesized by hepatocytes. It circulates as a 330 kDa single-chain, inactive, procofactor molecule (Esmon, 1979; Nesheim et al., 1979a). The earliest studies of the structure and function of coagulation factor V were done with the bovine protein and subsequently confirmed using human procofactors. Overall, bovine factor V is approximately 70% homologous to human factor V. However, sequence alignments of the two proteins indicate that the functional ends of both molecules, the N- and C-terminal halves, show as much as 84% and 86% identity, respectively (Guinto et al.,

1992). Several electron microscopy studies have reported that factor V is irregularly shaped displaying globular to oblong structures, and containing distinct appendages or domains

(Lampe et al., 1984; Dahlback, 1985; Mosesson et al., 1990a; Mosesson et al., 1990b).

Interestingly, blood coagulation proteins show a remarkable conservation of domain

structure across species, and homology amongst genes is common. Factor V is composed of the A1-A2-B-A3-C1-C2 domain structural arrangements from the N- to the C-terminus

(Figure 4) (Vehar & Davie, 1980; Gitschier et al., 1984; Toole et al., 1984; Vehar et al.,

1984).

Figure 4. Domain Structures of Human Factor V and Factor Va. Factor V is a single chain inactive glycoprotein with three main domain structures: three A domains, a single B domain, and two C domains (A1-A2-B-A3-C1-C2). When activated by thrombin (IIa, cleavage sites indicated by arrow), the inhibitory B domain is removed. The resulting the N-terminal heavy chain (A1-A2) and the C-terminal light chain (A3-C1-C2) reassociate non-covalently via a calcium ion to form the functional cofactor, factor Va. When factor V is proteolytically activated to factor Va (Figure 4) and described in detail in section 1.2.1.3), the N-terminal heavy chain (containing the A1 & A2 domains) and C-terminal light chain (containing the A3, C1, & C2 domains) remain non-covalently associated via a calcium ion. Crystal structures of full length factor V have been difficult to solve making the interpretation of functional studies challenging. Factor V share similar

domain organization with other proteins such as factor VIII. Factor V has little homology

(14% sequence identity) to factor VIII in the central B domain however, approximately

40% homology is observed in each A or C domain from both proteins (Jenny et al., 1987).

Nonetheless, X-ray crystals of factor Va fragments including proteolytically inactive fragments (factor Vai) (Adams et al., 2004), factor Va domains or homologous proteins have proven useful to understanding functional studies and designing experiments to elucidate structure-function relationships. The X-ray crystal structure of ceruloplasmin has provided some insights into the structure and function of the A domains of factors V and

VIII (Zaitseva et al., 1996). Additional structural evidence from the Fuentes-Prior and

Stoddard laboratories implicate a significant role for factor Va (Macedo-Ribeiro et al.,

1999), and factor VIIIa (Pratt et al., 1999) light chains in anionic phospholipid binding, respectively. Several amino acids in the C2 region of human factors Va and VIIIa appear to make direct contacts with phospholipid membranes in both of their models. Together, these studies support a role for hydrophilic and hydrophobic C2-mediated interactions with the membrane interface allowing for the binding and assembly of these cofactors as part of their respective macromolecular enzyme complexes (Macedo-Ribeiro et al., 1999; Pratt et al., 1999; Adams et al., 2004). Similarly, the crystal structure of factor Vai (activated protein C-inactivated factor Va) from the Everse laboratory (Adams et al., 2004) supported the involvement of hydrophobic amino acids (Trp26, Trp27, and Leu79) and electrostatic interactions (Arg43) in the binding of the C2 domain to membranes.

Several earlier functional studies suggest a role for the factor Va light chain in anionic membrane binding and Prothrombinase assembly (Higgins & Mann, 1983; Tracy

& Mann, 1983), which may be supported by the structural evidence reported some studies above. For example, studies with phosphatidylserine (PS)-containing membranes have implicated a single site around Cys2113 of the C2 domain of factor Va in membrane binding

(Srivastava et al., 2001). Further alanine substitution analyses of previously identified tryptophan and leucine residues on either C domain (C1 and C2) yielded a significant decrease in PS-membrane association and a 4-fold reduction in cofactor function in

Prothrombinase assays implicating an important role for the C2 domain in membrane binding (Majumder et al., 2005; Peng et al., 2005). A crucial role for the C1 domain of factor Va in Prothrombinase function was confirmed by the observation of a significant decrease in the rate of thrombin generation (14,000-fold) when recombinant human factor

Va Ala variants at Tyr1956 and Leu1957 were assayed for cofactor function (Majumder et al.,

2008). Taken together, the domain organization of factor V provides key regulatory regions that dictate the function of the cofactor, factor Va, in Prothrombinase assembly. In summary, factor V’s association with the membrane may involve a combination of factors including 1) penetration of the cofactor into the hydrophobic core (Krieg et al., 1987;

Lecompte et al., 1987; Adams et al., 2004); 2) direct binding through electrostatic interactions (Pusey et al., 1982; Pusey & Nelsestuen, 1984); and/or 3) hydrophobic interactions mediated by key residues within the C domains (Bloom et al., 1979;

Krishnaswamy & Mann, 1988).

1.2.1.2. Post-translational Modifications on Factor V

Since factor V is a very large glycoprotein, it is not surprising that post-translational modifications associated with factor V provide additional ways to regulate factor Va

function. Carbohydrate moieties or glycans are large, hydrophilic post-translational modifications which are regarded as maturation and quality-control determinants in the early secretory pathway of most proteins. However, when retained in a mature protein, they often perform important regulatory and functional roles post-synthesis. While human factor V is inherently a large protein (249 kDa, based on amino acids), glycans contribute significantly (13-25%) to its molecular weight (330, kDa including glycans) (Kane &

Davie, 1986; Kane et al., 1987) as shown in Figure 5A. The glycans on factor V are distributed across all the domains of the protein. In total 37 N-linked glycosylation sites have been observed on factor V including nine in the heavy chain (A1-A2 domains), 25 within the B domain, and three in the light chain (A3-C1-C2 domains) (Jenny et al., 1987).

While some of these glycans on factor V have been confirmed using mass spectrometry and lectin-binding techniques, their functional properties remain to be elucidated. Of the three potential (Asn1675, Asp1982 & Asn2181) light chain sites, Asn1675, and Asn1982 were found to be consistently glycosylated when expressed in a eukaryotic cell system (Kim et al., 1999). However, glycosylation at Asn2181 is variable and exists in two forms: factor

Va1 and factor Va2. The presence or absence of this glycan accounts for the doublet light chains formed (74 kDa, factor Va1 or 71 kDa, factor Va2, respectively) following thrombin activation of factor V (Kim et al., 1999; Nicolaes et al., 1999). Likewise, this differential glycosylation at Asn2181 reportedly decreases the anionic phospholipid membrane affinity of factor Va1 by three to 45-fold compared to factor Va2 (Rosing et al., 1993; Koppaka et al., 1997; Kim et al., 1999). Accordingly, decreased cofactor function of factor Va1 has been reported in various functional assays (Varadi et al., 1996; Hoekema et al., 1997;

Nicolaes et al., 1999). It is thought that the observations described above are the consequence of the inhibitory effect of the bulky glycan at position 2181. Therefore, by limiting access to crucial membrane binding sites on the C2 domain of factor Va, cofactor properties are affected. Studies have shown that Asn1982 (located on the light chain) contains a high mannose N-linked glycan. Earlier work using a combination of immunoaffinity techniques and carbohydrate binding proteins such as lectins predicted the presence of another high mannose N-linked glycan at Asn269 on the heavy chain of factor

Va (Liu et al., 2005). Additional mass spectrometric analyses confirmed the presence of these high mannose residues at the two sites above (Lewandrowski et al., 2006). Further work evaluating the functional properties of these high mannose N-linked glycans (Asn269 and Asn1982) particularly the high mannose residue at Asn1982, will be described in detail in

Chapter Five of this dissertation.

While structural interference posed by glycans may explain some functional consequences, global removal of human factor V glycans has produced conflicting data from various studies. Bruin and colleagues reported that deglycosylation of factor V impaired activation by thrombin, and increased clotting time in thrombin generation assays

(Gumprecht & Colman, 1975; Bruin et al., 1987). Removal of N-linked glycans with endoglycosidase has been reported to increase susceptibility to APC-mediated inactivation of deglycosylated cofactor, factor Va (Fernandez et al., 1997). Conversely, little change with respect to thrombin activation and APC-mediated inactivation of the deglycosylated factor procofactor, V was found in other studies (Fernandez et al., 1997; Silveira et al.,

2002).

Interestingly, unlike the N-linked glycans that are found in all regions of the human factor V molecule O-linked glycans appear limited in their distribution on the cofactor molecule. Treatment with O-glycanases (which removes O-linked glycans) produced electrophoresis mobility shifts only within the B domain suggesting an abundance of this glycan in this region of factor V (Fernandez et al., 1997; Silveira, 1997).

Figure 5. Schematic of the Post-translational Modifications Identified on Factor V. Panel A shows the glycosylation sites and their locations on human factor V. Algorithms have identified 37 N-linked glycosylation sites across the single chain molecule (orange + black + gray). The heavy chain, the B domain, and the light chain contains nine, 25, and three potential N-linked glycans, respectively. Only five of these have been confirmed as partially glycosylated (gray) and glycosylated (black). Panel B indicates the seven potential tyrosine sulfation sites (purple) and the three confirmed phosphorylation sites (blue) on factor V. Phosphorylation is a major post-translational modification which plays important roles in the regulation and function of various proteins in vivo. As shown in Figure 5B, three main sites have been identified as phosphorylation targets on plasma-derived factor 20

V. Using both purified casein kinase II, and a soluble platelet releasate, following thrombin-catalyzed platelet activation, Kalafatis et al. observed the incorporation of radioactive phosphates into one site on the light chain and two sites on the heavy chain of factor V (Kalafatis et al., 1993). In a subsequent study, Rand and colleagues reported limited phosphorylation of the platelet-derived factor Va heavy chain (Rand et al., 1994).

Additional studies indicated that phosphorylation at Ser692 appears unique to the plasma- derived factor Va heavy chain and might be mediated by a casein kinase II-like enzyme in vivo (Gould et al., 2004). The phosphorylation of Ser692 in plasma-derived factor Va and the inability to phosphorylate Ser692 in platelet-derived factor Va have significant functional implications and are discussed in detail in sections 1.2.1.6 and 1.2.1.9 respectively.

Another post-translational modification observed on human factor V is the modification of the hydroxyl of tyrosine residues with sulfates. Using computer algorithms,

Hortin and colleagues predicted seven tryrosine sulfation sites on recombinant factor V grown in the presence of radioactive sulfate (Hortin et al., 1986). Consistent with an overlap with O-glycans, the B domain rich tyrosine sulfates bound the O-linked binding lectin, jacalin, in their affinity purification system (Hortin, 1990). Two groups of tyrosine sulfation sites, Tyr1494, Tyr1510, and Tyr1515 (located in the B domain) and Tyr665, Tyr696, and Tyr698 (located in the heavy chain), are close to thrombin cleavage sites at Arg1545 and

Arg709, respectively. Functional characterization of these modifications by Pittman and colleagues demonstrated that sulfation on factor V correlates with efficient thrombin activation to factor Va and increased cofactor function (Pittman et al., 1994). Interestingly,

studies by Gould and colleagues reported no obvious sulfate modifications of the previously identified tyrosine residues on the heavy chain of purified human plasma- derived factor Va (Gould et al., 2004).

1.2.1.3. Functional Properties of Factor Va

Coagulation factor V circulates in plasma as a single-chain, inactive, protein procofactor expressing negligible if any cofactor activity (approximately 0.3% that of the active cofactor, factor Va) (Nesheim et al., 1979a). This non-enzymatic procofactor must be proteolytically activated prior to its association and function as a key component of

Prothrombinase during hemostasis. Under physiological conditions, factor V is a substrate for a variety of human proteases. Cleavage of factor V by these proteases produces several fragments which show little to full cofactor activity when compared to factor Va. These include elastase (Oates & Salem, 1987; Turkington, 1993; Allen & Tracy, 1995; Samis et al., 1997; Camire et al., 1998a), cathepsin G (Turkington, 1993; Allen & Tracy, 1995;

Camire et al., 1998a), calpain (Rodgers et al., 1987; Bradford et al., 1988), and plasmin

(Lee & Mann, 1989). Other proteases such as factor XIa (Whelihan et al., 2010), factor Xa

(Smith & Hanahan, 1976; Foster et al., 1983; Tracy et al., 1983; Monkovic & Tracy, 1990a, b), and thrombin (Esmon, 1979; Nesheim & Mann, 1979; Nesheim et al., 1979b; Kane &

Majerus, 1981; Suzuki et al., 1982; Foster et al., 1983; Monkovic & Tracy, 1990a, b) which are associated with the coagulation response, cleave factor V to produce significant factor

Va cofactor activity. Principal amongst these proteases is thrombin, which is the main physiological activator of coagulation factor V.

Factor Va is formed following sequential cleavages at three Arg residues (Arg709, Arg1018 and Arg1545) of the single chain factor V molecule. As shown in Figure 6, initial cleavage at Arg709 produces the amino-terminal derived heavy chain (105 kDa) and a 280 kDa fragment (containing the entire B domain and the light chain). Subsequent cleavage at

Arg1018 of the 280 kDa fragment produces a 71 kDa B domain fragment and a 220 kDa intermediate fragment (comprised of a portion of the B domain and the C-terminal half of factor V). A final cleavage at Arg1545 of the 220 kDa fragment releases the carboxyl- terminal derived light chain (71/74 kDa) and a 150 kDa B domain fragment.

Figure 6. Thrombin-mediated Activation of Factor V to Factor Va. Thrombin activates plasma-derived single chain factor V to factor Va via three sequential cleavages. Cleavage at Arg709 occurs first with the generation of the heavy chain fragment (105 kDa). A second cleavage at Arg1018 within the B domain, and a third cleavage at Arg1545 releases the light chain fragment. A calcium-dependent reassociation of the heavy and light chains forms the functional cofactor factor Va.

The heavy chain (105 kDa) and the light chain (74/71 kDa) remain associated non- covalently in the presence of calcium ions to form the functional cofactor, factor Va

(Esmon, 1979; Nesheim & Mann, 1979; Guinto & Esmon, 1982; Suzuki et al., 1982;

Krishnaswamy et al., 1989; Monkovic & Tracy, 1990a).

Interestingly, factor Xa cleaves factor V at the same sites as thrombin (Arg709,

Arg1018, & Arg1545) (Figure 6). However, the order of peptide bond cleavages are different between the two proteases. Cleavage at Arg1018 within the B domain occurs first generating a 150 kDa amino-terminal fragment (comprised of the heavy chain and a portion of the B domain) and a 220 kDa carboxyl-terminal fragment (comprised of the light chain and the remaining half of the B domain) (Figure 7). The generation of these fragments is associated with significant cofactor activity suggesting that cleavage at Arg1018 is crucial for cofactor function (Monkovic & Tracy, 1990a). A second cleavage of the 150 kDa fragment at or near Arg709 generates the heavy chain (105 kDa). While a third cleavage at Arg1545 occurs to release the putative light chain doublet (74/71 kDa) of factor Va, the rate of this cleavage is very slow, and does not contribute to additional cofactor activity compared to the initial cleavage at Arg1018 (Monkovic & Tracy, 1990a) (Figure 7). Other studies from our laboratory showed that cleavage at Arg1545 and the release of the B domain promotes factor

Xa’s association with factor Va (Camire et al., 1998a).

Despite the differential pattern of peptide bond cleavages between thrombin and factor Xa during factor V activation, both proteases show similar catalytic efficiencies: 3.3 x106 M-1s-1 and 4.1 x 106 M-1s-1, respectively. The affinity of factor Xa for factor V is higher with a Km ~7-fold lower than thrombin (10.4 nM vs 71.1 nM). However, consistent

with its higher potency, substrate turnover (kcat) by thrombin appears ~5-fold faster than factor Xa (14.0 min-1 vs 2.6 min-1, respectively) (Monkovic & Tracy, 1990a). Interestingly, prolonged incubation of factor V with either thrombin or factor Xa is associated with a fourth cleavage site near or at Arg643 (Hockin et al., 1997; Erdogan et al., 2007), or Arg1765

(Tracy et al., 1983; Odegaard & Mann, 1987; Thorelli et al., 1997). Cleavage by the former protease is associated with a significant decrease in clotting potential (60%) whereas cleavage by factor Xa has no effect on clotting activity (Tracy et al., 1983; Hockin et al.,

1997).

Figure 7. Proteolysis of Factor V by Factor Xa. Factor Xa activates factor V following cleavages at Arg residues 709, 1018, and 1545. However, the order of bond cleavages are different from those of thrombin. Initial cleavage occurs at Arg1018 within the B domain to generate 150 and 220 kDa fragments. Next, the 150 kDa fragment is cleaved releasing the heavy chain of factor Va (105 kDa), and a 71 kDa B domain-derived fragment. A subsequent slow cleavage at Arg1545 of the 220 kDa fragment releases the light chain of factor Va (71/74 kDa). Together, the heavy (105 kDa) and light (71/74 kDa) chains associate in the presence of calcium to form factor Va. 25

From the studies detailed above, it appears that the B domain of factor V plays an important role in preserving the procofactor state of this coagulation protein molecule.

Pioneering work investigating the role of this region of factor V was done by Kane and colleagues. In their study, they expressed a partial truncated B domain-less factor V species in COS cells and monitored cofactor function in vitro. Factor Vdes811-1491 (FV-DT, lacking a significant portion of the B domain) showed substantial cofactor activity, ~ 30-

38% when compared to factor Va produced by thrombin- or snake venom protease- mediated activation of factor V, respectively (Kane et al., 1990; Keller et al., 1995). A more recent study by Toso and Camire (Toso & Camire, 2004), using similar constructs revealed that the cofactor functions of FV-DT and factor Va were identical in

Prothrombinase assays. Thus, full cofactor activity is expressed in the absence of the B domain region implicating a key role of the central B domain in preserving and maintaining the inactive procofactor state of factor V. Together, these studies suggest that removal of the B domain eliminates conformational constraints within factor Va allowing for productive binding to membranes and interaction with factor Xa during Prothrombinase assembly and function (Toso & Camire, 2004). Additional mutational studies of the amino acid sequences within the highly conserved basic region in the B domain (via deletions or substitutions) suggest that residues 963 through 1008 are essential for maintaining the procofactor state of factor V (Zhu et al., 2007). More recently, it has been shown that an autoinhibitory motif in the B domain of factor V, contributed by the previously identified basic (963-1008) region and a newly identified acidic (1493-1537) region ensures that factor V remains inactive until appropriately activated (Mettine & Camire, 2012).

Once formed, factor Va contributes to substantial thrombin generation as part of

Prothrombinase. This function is possible following the loss of the inhibitory B domain, and exposure of previously buried key amino acid residues (in factor V) for binding to both divalent metal ions, anionic membranes (discussed in section 1.2.1.3), and factor Xa.

Particularly, the exposure of the carboxyl-terminal region of the factor Va heavy chain and access to residues within the A2 domain appears crucial to factor Va function (Kojima et al., 1998; Kalafatis & Beck, 2002; Singh et al., 2003; Beck et al., 2004; Barhoover et al.,

2008b; Barhoover et al., 2008a). The earliest work mapping residues of factor Va involved in Prothrombinase function were performed by Kojima and colleagues. They observed that peptide sequences from Ile311 to Tyr335 noncompetitively inhibited thrombin generation by

50% (Kojima et al., 1998). Similarly, factor V residues 323 to 331 were found to bind factor Xa, and significantly inhibited Prothrombinase in peptide titration studies (Kalafatis

& Beck, 2002). Singh and others identified four residues in factor V including Glu323,

Tyr324, Val331, and Glu330 as mediating the inhibition reported earlier (Singh et al., 2003).

Likewise, an alanine-substituted quadruple mutant of the residues above showed ~400-fold lower Prothrombinase activity compared to wild type factor Va (Barhoover et al., 2008b).

In additional studies, only 25% cofactor activity was observed in clotting assays when

Asp334 and Tyr335 were substituted with alanine residues (Barhoover et al., 2008a). Several other studies have identified amino acids which mediate the binding and association of factor Va with factor Xa during Prothrombinase assembly. These include Glu467, Arg501,

Arg510, Ala511, Asp513, Asp577, Asp578, Arg652 and His1683and with the exception of His1683 all reside within the heavy chain of factor Va (Steen et al., 2002; Steen et al., 2008).

In summary, the majority of the studies detailed above defining the functional regions of plasma-derived factor V/Va were done using anionic membranes. Using key amino acids identified in factor V above, including the structures of homologous proteins such as factor Xa and ceruloplasmin (Brandstetter et al., 1996; Macedo-Ribeiro et al.,

1999), several models of Prothrombinase have been developed. These models have contributed to our present understanding of important structure-function relationships of the cofactor, and the enzyme in general (Adams et al., 2004; Autin et al., 2006; Lee et al.,

2008; Stoilova-McPhie et al., 2008; Lechtenberg et al., 2013). Recently, a crystal structure of Prothrombinase (pseutarin C) from the venom of the Australian Eastern Brown Snake

(Pseudonaja textilis) was solved by the Huntington Lab. While this pre-formed snake

Prothrombinase can function in the presence or absence of an appropriate membrane, the structure of factor Va supports previous observations where the C2 domains form the legs on the membrane surface and allow for tight interactions of the A domains with factor Xa for enzyme formation and Prothrombinase function (Lechtenberg et al., 2013).

1.2.1.4. Inactivation of Factors V and Va

While the importance of thrombin generation cannot be overemphasized during the procoagulant response, a timely regulation of excessive thrombin generation is equally vital to preventing unwanted thrombus growth following an injury. Thrombin down- regulates its own generation via the thrombin-thrombomodulin complex in a mechanism where thrombin’s binding to endothelial derived-thrombomodulin shifts the balance in favor of the anticoagulant response. Therefore, the thrombin-thrombomodulin complex has high affinity for protein C resulting in its conversion to activated protein C (APC). The

major substrates for APC are the cofactors of the intrinsic Tenase (factor VIIIa) and

Prothrombinase (factor Va). APC can also inactivate factor V.

APC-catalyzed inactivation of factor Va requires the ordered proteolysis of the heavy chain at three key residues: Arg506, Arg306 and Arg679 (Salem et al., 1983; Suzuki et al., 1983; Kalafatis et al., 1994). Cleavage at Arg506 produces a 75 kDa fragment with the subsequent cleavage of this intermediate at Arg306 producing a 30 kDa fragment. The final cleavage at Arg679 produces a 30 amino acid peptide (Figure 8A). Nicolaes and colleagues showed that greater than 50% of factor Va’s cofactor activity is lost following initial

506 cleavage at Arg . Consistent with a reduced affinity for factor Xa, the Kd of factor Xa for factor Va, following its cleavage at Arg506 decreased by ~97-fold in Prothrombinase assays

(FVai, 3.9 nM vs FVa, 0.1 nM). A complete loss of factor Va cofactor activity is associated

306 with further cleavage at Arg (Kd of factor Xa for FVai, 196 nM, vs FVa, 20 nM)

(Nicolaes et al., 1995). Interestingly, the functional importance of cleavage at Arg679 appears inconsequential despite contributing to destroying the overall structural integrity of factor Va during the proteolytic process. A recombinant factor V lacking this cleavage site was inactivated by APC at the same rate as the wild type cofactor (Egan et al., 1997;

Heeb et al., 1999a).

The procofactor, factor V, is also inactivated by APC; however, an additional cleavage site at Lys994 (Kalafatis et al., 1994) within the B domain is observed (Figure 8B).

Proteolytic fragments from the APC-catalyzed inactivation of factor V show no cofactor activity when further treated with thrombin. Proteolytic inactivation of factor V first occurs at Arg306 followed by Arg506 within the heavy chain. A final cleavage is observed at both

Arg679 and Lys994 (Kalafatis et al., 1994; Kalafatis et al., 1995). Cleavage at the 306 site by APC is associated with complete inactivation of the factor V molecule, similar to that observed with the cofactor, factor Va.

Like most vitamin K-dependent coagulation proteases, membrane binding, and the presence or absence of cofactors, are vital to optimizing the functional properties of APC and to ultimately achieving catalytic efficiency comparable to other coagulation enzymes such as Prothrombinase. Several studies have demonstrated that cleavage at Arg306 on both factors Va and V depends on the presence of an anionic phospholipid membrane surface

(Suzuki et al., 1983; Kalafatis et al., 1994; Heeb et al., 1995; Nicolaes et al., 1995; Hockin et al., 1997; Heeb et al., 1999a), which can also be provided by the activated platelet

(Solymoss et al., 1988). While, cleavages at Arg506 and Arg679 are only slightly enhanced by membrane binding, the loss of an appropriate membrane surface significantly decreases cleavage at Arg306 by ~1000-fold (Nicolaes et al., 1995; Hockin et al., 1997).

Other studies have shown that factor Xa protects factor Va from APC-catalyzed inactivation at Arg506. Thus, at both sub-saturating and saturating concentrations of factor

Xa, cofactor activity is decreased by approximately 60% (Egan et al., 1997) and 20%

(Kalafatis et al., 1994; Rosing et al., 1995; Egan et al., 1997; Hockin et al., 1997; Heeb et al., 1999b), respectively. The ability of factor Xa to protect factor Va from inactivation was found to be further enhanced (~100-fold) in the presence of prothrombin (Rosing et al., 1995).

Protein S has been proposed as a non-enzymatic cofactor of APC. Impaired function or lower levels of protein S has been reported in some individuals with inherited

thrombophilia (Makris et al., 1997; Esmon, 2000). Interestingly, several studies have shown only a marginal increase (two- to ten-fold) in the catalytic activity of APC in the presence of protein S (Dahlback, 1986; Solymoss et al., 1988; Tans et al., 1991b).

Figure 8. Schematic of Factor Va and Factor V Proteolytic Inactivation by Activated Protein C. The activated protein C-mediated proteolysis of factor Va (A) and factor V (B) are indicated. The heavy and a light chains of factor V are shown in blue with the B domain in grey. The order of bond cleavages are indicated by arrows along with the cleavage sites: Arg305, Arg506, Arg679 and Lys994. The addition of protein S to APC reactions monitoring the proteolysis of factor Va increases cleavage at Arg306 by ~ 20-fold (Rosing et al., 1995). Interestingly, the protection by factor Xa at Arg506 within the factor Va heavy chain is eliminated in the presence of protein S. It has been suggested that in its role as a cofactor for APC, protein S competes for binding to anionic membranes to which Prothrombinase binds thereby weakening the protective effect of the bound factor Xa (van Wijnen et al., 1996).

The effect of glycans on the APC-catalyzed, membrane-dependent inactivation of factor Va was investigated by Fernandez and colleagues. They observed that partial removal of the glycans following treatment with neuraminidase and N-glycanase enhanced

APC-catalyzed inactivation by ~ three-fold (Fernandez et al., 1997). Similar studies from our laboratory revealed that glycosidase treatment of factor V increased its susceptibility to APC-mediated cleavage at Arg306 by ten-fold, but was without effect on factor Va

(Silveira, 1997; Silveira et al., 2002).

Other post-translational modifications including phosphorylation appear to affect inactivation of factor Va by APC. Phosphorylation at Ser692 in plasma-derived factor Va increased its APC catalyzed cleavage by ~ 3-fold with negligible effect on the cleavages at

Arg506 and Arg67 (Kalafatis, 1998). Together, these studies support a role for post- translational modifications in regulating both factor V and factor Va.

The APC-catalyzed inactivation of factor Va as well as factor VIIIa (Kalafatis et al., 1993; Kalafatis et al., 1994) represents an important physiological anticoagulant system poised to prevent unwanted thrombosis. Thus, naturally occurring mutations in coagulation proteins regulated by this system may predispose individuals to thrombosis. APC resistance, identified by Dahlback and colleagues in the early 1990s, is hereditary and associated with venous thrombosis (Dahlback et al., 1993). A missense mutation in the factor V gene resulting in an Arg to Gln substitution at position 506 (factor V R506Q or factor VLeiden) is quite common, and confers resistance to APC-mediated inactivation of the cofactor molecule in individuals carrying the mutation (Bertina et al., 1994; Dahlback,

1994). Homozygosity or heterozygosity of this mutation is associated with a 30-140 fold or 38-fold propensity for venous thrombosis, respectively (Dahlback, 1997).

Studies of the APC-catalyzed inactivation of factor VaLeiden have indicated that, while access to cleavage at Arg506 is eliminated, inactivation will occur via cleavage at

Arg306; however, the rate of inactivation is ~10-fold lower compared to normal factor Va

(Heeb et al., 1995; Kalafatis et al., 1995; Kalafatis & Mann, 1995; Nicolaes et al., 1995;

Dahlback, 1997; Egan et al., 1997). A reduced risk of thrombosis is also associated with variable factor V mutations at Arg306 including factor V R306T (factor VCambridge)

(Williamson et al., 1998) and factor V R306G (factor VHong Kong) (Chan et al., 1998). These observations are consistent with earlier reports that suggest that cleavage at Arg506 and

Arg306 are not mutually exclusive, yet are essential for the ultimate inactivation of factor

Va.

Plasmin, a key fibrinolytic protease derived from plasminogen activation also cleaves factor Va to produce factor Vai. However, compared to APC, plasmin cleaves factor Va at multiple lysine residues in both the heavy and the light chains to produce a mixture of fragmented polypeptides with no cofactor activity. Five (Lys93, Lys109, Arg348,

Lys364, and Lys456) heavy chain and three (Lys1656, Lys1765, and Lys1827) light chain cleavage sites for plasmin have been identified (Zeibdawi & Pryzdial, 2001). Omar and

Mann demonstrated that, similar to APC, plasmin-mediated inactivation of factor Va is significantly increased in the presence of an anionic phospholipid membrane (Omar &

Mann, 1987). Additionally, factor Xa appears to protect the cofactor from inactivation by plasmin. While both the cofactor and the procofactor are inactivated equally by plasmin,

Lee and Mann observed that the initial proteolysis of factor V by plasmin produced fragments with short-lived procoagulant activity (Lee & Mann, 1989).

1.2.1.5. Origins of Plasma-derived and Platelet-derived Factor V/Va

Factor V exists in two pools in human blood. Tracy and colleagues demonstrated that approximately 75-80% of factor V circulates in the plasma (plasma-derived factor V), with the remaining 20-25% residing in platelets (platelet-derived factor V/Va) (Tracy et al., 1982). The average concentration of plasma-derived factor V is ~7-10 µg/mL (20-30 nM). Platelet-derived factor V is stored within α-granules (Vicic et al., 1980) and averages approximately 4,600-14,100 molecules of factor V/platelet (Tracy et al., 1982).

Plasma-derived factor V is synthesized and secreted by hepatocytes (Mattii et al.,

1964; Owen & Bowie, 1977; Mazzorana et al., 1991). Historically, however, the source of human platelet-derived factor V was more difficult to establish unequivocally. Even though there is a single factor V gene, many studies from our laboratory have demonstrated that platelet-derived factor V is phenotypically and functionally different from plasma-derived factor V (Monkovic & Tracy, 1990a, b; Camire et al., 1995; Conlon, 1997; Silveira, 1997;

Camire et al., 1998c; Silveira et al., 2002; Gould et al., 2004; Wood et al., 2008).

Consequently, we hypothesized that human megakaryocytes, platelet precursor cells, synthesized factor V and “re-tailored” it prior to its packaging into platelet α-granules. This hypothesis seemed tenable since human megakaryocytes were shown to contain factor V

(Breederveld et al., 1975; Osterud et al., 1977; Vicic et al., 1980; Tracy et al., 1982) , as were other bone marrow-derived cells, e.g. lymphocytes and monocytes (Tracy et al.,

1985). In addition, studies from other laboratories had demonstrated that megakaryocytes,

derived from guinea pigs (Chiu et al., 1985) and mice (Sun et al., 2003), synthesize factor

V. However, despite collaborations with outstanding megakaryocyte laboratories, we were unable to demonstrate factor V synthesis by human megakaryocytes (unpublished observations). Consequently, we decided to take a genetic approach.

Two individuals heterozygous for the factor VLeiden variant were identified who received either an allogeneic bone marrow transplant or allogeneic liver transplant from individuals expressing wild-type factor V. By assessing the APC-catalyzed proteolytic inactivation of factor Va released by thrombin-activated platelets from these recipients,

Camire, et al. (Camire et al., 1998b) demonstrated that the platelets obtained from the bone marrow recipient released both wild-type factor Va and factor VaLeiden, whereas platelets obtained from the liver recipient released only wild-type factor Va while retaining the factor VaLeiden allele in their megakaryocyte genome. The factor Va released by these two individuals mirrored the factor V/Va that was synthesized in and secreted from their hepatocytes. Therefore, it appeared that the platelet-derived factor V pool originates from megakaryocyte endocytosis of plasma factor V (Camire et al., 1998b).

Several other studies have shown that the majority of proteins found in platelets are acquired from progenitor cells during megakaryocyte maturation and proplatelet formation due to the limited synthetic capabilities of mature platelets. Plasma proteins such as fibrinogen, IgG, and albumin are endocytosed and packaged into α-granules by megakaryocytes (Handagama & Bainton, 1989; Harrison et al., 1989; Handagama et al.,

1990). Using similar approaches, Bouchard and colleagues demonstrated that the platelet pool of factor V may be derived entirely from the factor V plasma pool via megakaryocyte

endocytosis (Bouchard et al., 2005). They demonstrated that factor V was endocytosed in a time-dependent manner by megakaryocytes derived from CD34+ progenitor cells or megakaryocyte-like cell lines, but not by platelets. CD41+ ex vivo-derived megakaryocytes endocytosed factor V, as did subpopulations of the megakaryocyte-like cells, MEG-01 and

CMK. In similar studies, Ayombil and colleagues demonstrated that thrombin-activated lysates from CMK cells, which had endocytosed factor V, express factor Va cofactor activity in clotting assays. Additionally, the electrophoretic phenotype of the endocytosed factor V was similar to that observed in normal platelet lysates (Ayombil et al., 2013).

Bouchard and colleagues further demonstrated that plasma-derived factor V endocytosis by megakaryocytes is mediated by low-density lipoprotein receptor related protein-1 (LRP-1) through a Ca2+ - and clathrin-dependent process (Bouchard et al., 2008).

Based on additional studies (Bouchard et al., 2005; Bouchard et al., 2008), they proposed a two receptor model to explain the mechanism of factor V endocytosis. Plasma-derived factor V binding to a specific receptor facilitates binding to and endocytosis of, a second factor V molecule by LRP-1, or a related family member. Subsequent to its endocytosis, factor V is trafficked through several compartments including the early and late endosomes

(Seifert, 2005), and the endoplasmic reticulum before final packaging into α-granules

(Bouchard et al., 2008) as described in detail in Section 1.2.1.5. Furthermore, we hypothesize that through endocytosis, plasma-derived factor V is phenotypically and proteolytically modified to form the functionally and physically unique pool of protease- resistant, platelet-derived factor V/Va.

1.2.1.6. Functional and Phenotypic Differences between Platelet- derived and

Plasma-derived factor V

In marked contrast to plasma-derived factor V, secretable platelet-derived factor V/Va molecules express significant cofactor activity upon their release from α-granules. Platelet agonists including collagen, epinephrine, arachidonic acid, and ADP mediate platelet- derived factor V/Va release from secretory α-granules (Osterud et al., 1977; Kane et al.,

1980; Vicic et al., 1980; Chesney et al., 1981; Tracy et al., 1985; Monkovic & Tracy,

1990b). When compared to the plasma-derived factor V molecule, platelet-derived factor

V/Va is composed of a mixture of polypeptide fragments with varying molecular weights

(46-330 kDa) (Viskup et al., 1987; Monkovic & Tracy, 1990b; Ayombil et al., 2013).

Between 5-50% of the cofactor activity observed with thrombin-activated plasma-derived factor Va is expressed by platelet-derived factor V/Va upon release. Consistent with these observations polypeptide fragments similar in molecular weights to factor Va heavy and light chains are associated with agonist-mediated platelet activation and secretion (Kane et al., 1980; Vicic et al., 1980; Viskup et al., 1987; Monkovic & Tracy, 1990b; Ayombil et al., 2013). According to Monkovic and Tracy, further activation of platelet-derived factor

V/Va molecules by thrombin or factor Xa results in only a two- to three-fold increase in cofactor activity (Monkovic & Tracy, 1990a). These studies suggest that platelet-derived factor V/Va is pre-packaged and stored in a partially-activated state in α-granules making it capable of mounting a rapid procoagulant response during the initial response to vascular injury when compared to the plasma-derived factor V that requires proteolytic activation prior to expressing cofactor activity.

Several studies have reported additional variable physical and functional characteristics of the two pools of factor V. According to Monkovic and Tracy, there is no obvious electrophoretic mobility differences between the heavy and light chains of factor

Va derived from either the plasma or platelet pools (Monkovic & Tracy, 1990b). However,

Gould and colleagues observed, following N-terminal sequencing, that the light chain of platelet-derived factor Va is cleaved at Tyr1543 and not Arg1545, as is seen with thrombin- or factor Xa-catalyzed factor V activation (Gould et al., 2004). Likewise, Thr402 on the heavy chain of platelet-derived factor Va is O-glycosylated with an N-acetylglucosamine or N-acetylgalactosamine whereas plasma-derived factor Va is not. Additionally, differential phosphorylation at Ser692 of the heavy chain has been reported. While plasma- derived factor V can be phosphorylated by casein kinase II or a casein kinase II-like protease, the platelet-derived molecule is completely resistant to phosphorylation at this site (Kalafatis et al., 1994; Kalafatis, 1998).

Previous studies have reported that plasma- and platelet-derived factor V are differentially activated by various proteases of the coagulation system. Platelet-derived factor V/Va is a better substrate for factor Xa than thrombin. Monkovic and Tracy demonstrated that factor Xa activates platelet-derived factor V/Va ~50-100 times more effectively than thrombin (Monkovic & Tracy, 1990b). In contrast, plasma-derived factor

V is activated equally by factor Xa and thrombin (Monkovic & Tracy, 1990a). Plasmin rapidly inactivates plasma-derived factor Va in the presence of anionic phospholipids (Lee

& Mann, 1989) while platelet-derived factor Va exhibits both increased and sustained activity (up to 4 h) in the presence of plasmin and the activated platelet surface. This

observation suggests that the platelet membrane or some component of the platelet- releasate offers a protective effect since inactivation is significantly delayed in contrast to inactivation reactions on anionic phospholipid membranes (Conlon, 1997). Additionally, studies from our laboratory have shown that platelet-derived factor Va is resistant to inactivation by APC compared to plasma-derived factor Va (Camire et al., 1995; Camire et al., 1998a). Following prolonged incubation with APC (120 min), the cofactor activities of platelet- vs plasma-derived factor Va were decreased by ~50% and 100% respectively, in the presence of activated platelets. The resistance of platelet-derived factor Va to inactivation by APC correlated with reduced cleavage at Arg506. Interestingly, normal platelet-derived factor Va was as resistant to APC-catalyzed inactivation as was the factor

Va from individuals with factor VLeiden (Camire et al., 1995; Heeb et al., 1995; Kalafatis et al., 1995; Nicolaes et al., 1995; Camire et al., 1998a). These observations are consistent with the physiological importance of platelet-secreted factor V/Va molecules as key constituents that regulate the procoagulant response following an injury.

1.2.1.7. Physiological Significance of Platelet-derived vs. Plasma-derived Factor Va

The importance of α-granule stored platelet-derived factor Va to the maintenance of hemostasis is profound and confirmed by several clinical observations that indicate it is the physiologically relevant cofactor pool (Borchgrevink & Owren, 1961; Janeway et al.,

1996; Ang et al., 2009; Bouchard et al., 2015). In an individual completely devoid of plasma- and platelet-derived factor V, appropriate hemostasis was maintained by the prolonged life span of his platelet-derived factor V/Va pool acquired by endocytosis of plasma-derived factor V by megakaryocytes subsequent to his receipt of fresh frozen

plasma (Bouchard et al., 2015). While the patient’s plasma-derived factor V was undetectable by four days following transfusion, 10 days post-transfusion, he was still able to mount robust thrombin generation due to his residual platelet-derived factor Va (~ 7% of normal).

In much earlier studies, Borchgrevink and Owren demonstrated that bleeding associated with congenital factor V deficiency in a patient could be corrected temporarily by the administration of platelets from a normal donor (Borchgrevink & Owren, 1961).

Similarly, as reviewed by Ang and colleagues, several studies demonstrate that platelet transfusions, rather than plasma transfusions, are more effective at correcting bleeding tendencies associated with individuals with non-congenital, acquired factor V inhibitors

(Ang et al., 2009). Furthermore, Nesheim and colleagues observed that a patient with negligible plasma-derived factor V levels (< 1%) due to an acquired factor V inhibitory antibody withstood surgical challenge without a significant bleed. They suggested that the patient’s normal response to bleeding was not compromised due to the presence of intact platelet-derived factor V/Va which was not accessible to the plasma antibody (Nesheim et al., 1986). Furthermore, in Quebec platelet disorder in which affected individuals have normal levels of plasma-derived factor V, they still present life-threatening bleeding conditions due to deficiencies of their α-granule proteins, most notably platelet-derived factor Va (Tracy et al., 1984; Janeway et al., 1996). Finally, studies by Weiss and colleagues reported impaired platelet-derived Prothrombinase activity in a patient expressing a factor V variant, Factor VNew York. They demonstrated that despite having near normal, plasma-derived activity (~80%), the patient showed severe bleeding abnormalities

due to extremely low platelet-derived factor V/Va concentrations (Weiss et al., 2001).

These combined studies offer strong support for the notion that platelet-derived factor Va is the physiologically-relevant cofactor pool.

1.2.1.8. Megakaryocyte-Endocytosis of Plasma-derived Factor V: Formation of the

Functionally and Phenotypically Unique Platelet-derived Factor V/Va pool

In confocal microscopy studies completed in our laboratory, megakaryocyte- endocytosed factor V was shown to initially co-distribute with endocytosed transferrin

(Harding et al., 1983; Bouchard et al., 2005). Consistent with these data, additional studies demonstrated that following its endocytosis, factor V transits through early and late endosomes in a time-dependent manner as demonstrated by its colocalization with antibodies to Rab 5 and Rab 7, respectively (Figure 9). Using brefeldin A, a compound known to disrupt the trans-Golgi network, thus preventing endosomal to Golgi trafficking, we have also shown that endocytosed factor V undergoes partial proteolysis, by an unknown megakaryocyte protease(s), to form a mixture of factor V/Va molecules

(unpublished data). Consistent with this observation, platelet-derived factor V/Va molecules are released from activated platelets in a partially proteolytically-activated state expressing substantial cofactor activity as evidenced by the presence of a bonafide heavy chain and a light chain containing two additional N-terminal amino acids formed by cleavage at Tyr1543 rather than Arg1545 (Gould et al., 2004). Ayombil et al. further demonstrated that subsequent to endocytosis of plasma-derived factor V, the endocytosed protein was described by proteolytic fragments similar in molecular weight to the

heterodimeric subunits of factor Va and expressed cofactor activity in in vitro clotting assays (Ayombil et al., 2013).

In additional trafficking studies, endocytosed factor V co-localized with GM130, a cis-Golgi element consistent with the unique O-glycosylation at Thr402 observed in platelet- derived factor V/Va (Figure 9) (Gould et al., 2004). Furthermore, factor V trafficked to the endoplasmic reticulum where it colocalized with antibodies to PIGK, a member of the cysteine proteases involved in glycosylphosphatidylinositol (GPI)-anchor biosynthesis. In support of this observation, Ser692 in the factor V heavy chain is contained within a consensus sequence for GPI-anchor addition (Wood et al., 2008). Consistent with those combined observations, studies in our laboratory have demonstrated that a pool of platelet- derived factor Va is GPI-anchored to the activated platelet membrane, which results in its resistance to APC-catalyzed inactivation (Chapter Four) (Amthaeur et al., 1993; Camire et al., 1995; Camire et al., 1998c). Anchoring of the platelet-derived factor Va heavy chain to the activated platelet membrane (Wood et al., 2008), may explain the inability of platelet-derived factor Va to be phosphorylated on Ser692, which is marked contrast to what is observed with plasma-derived factor Va (Rand et al., 1994). Likewise, selective trimming of N-linked oligosaccharides occurs in the endoplasmic reticulum and may explain, in part, the different number of high mannose residues on the N-linked glycan at

Asn1982 on the light chains of platelet-derived factor Va, when compared to the plasma- derived factor Va light chains isolated from the same donor platelets (Chapter Five).

Figure 9. Endocytic trafficking of plasma-derived factor V in megakaryocytes to form the unique platelet-derived factor V/Va molecules. Two factor V molecules bind their receptors at the cell membrane of megakaryocytes and are subsequently internalized in a clathrin-dependent manner in the early endosomes and late endosomes as shown by their colocalization with antibodies to Rab 5 and Rab 7, respectively. In the endosomes factor V is partially proteolyzed by unknown megakaryocyte protease(s) before transport in a retrograde faction along compartments of the secretory route including the endoplasmic reticulum (as shown by colocalization with antibody to phosphatidylinositol, GPI, glycan anchor biosynthesis class K, PIGK) where N-linked glycan trimming occurs and a GPI-anchor is added. In the Golgi network (shown by colocalization with antibodies to GM130 and von Willebrand factor), O-linked glycosylation is added to factor V/Va to form the platelet-derived protease-resistant α- granule stored molecule (as shown by colocalization with antibodies to P-selectin, fibrinogen, von Willibrand factor which are stored in the α-granule). Diagram adapted from Bouchard et al., Factor V trafficking, unpublished). Consistent with its eventual packaging and transport to the α-granules, endocytosed factor V colocalized with fibrinogen (Bouchard et al., 2005), von willebrand factor and P- selectin, proteins known to be stored in platelet α-granules (Figure 9). These combined studies are consistent with the retrograde transport of megakaryocyte-endocytosed, plasma-derived factor V as shown in figure 9. Therefore, subsequent to its endocytosis by 43

megakaryocytes, plasma-derived factor V transits through the early and late endosomal compartments before trafficking to the endoplasmic reticulum and trans-Golgi network where it is posttranslationally retailored to become the unique, protease-resistant α-granule pool of platelet-derived factor V/Va.

1.2.1.9.Characterization of the Non-dissociable Platelet-Derived Factor Va Pool

Present evidence suggests that activated platelets contain a non-dissociable pool of factor Va. Early studies by Alberio and colleagues showed that a subpopulation of platelets

(~31% to the total platelet population) express high levels of surface-bound factor V following platelet activation by a combination of convulxin and thrombin (Alberio et al.,

2000). Known as COAT-factor V (COAT-FV), they showed that this pool of factor Va bound factor Xa and supported significant Prothrombinase activity compared to platelets activated independently by either agonist. In additional studies, they proposed, that COAT- platelets, those expressing COAT-FV, are covalently modified by attaching several α- granule proteins to the activated platelet membrane in a reaction wherein a transglutaminase cross-links the proteins to serotonin (Dale et al., 2002). Interestingly, transglutaminases such as factor XIIIa have been reported to interact with factor V.

However, transamination sites were found only in the inhibitory B domain of the procofactor, factor V, which is absent in factor Va (Francis et al., 1986). Contrasting results were reported in recent studies by Jobe and colleagues using murine knockout models of the collagen receptor, FcRγ, and transglutaminase, factor XIIIa. They showed that the FcRγ receptor but not factor XIIIa was important in the formation of the COAT- platelets (Jobe et al., 2005).

In marked contrast, recent studies from our laboratory demonstrate the existence of a functional, yet non-dissociable platelet-derived pool of factor V/Va on thrombin- activated platelets. Using flow cytometric analyses, Fager and colleagues demonstrated that subsequent to thrombin-catalyzed platelet activation, a subpopulation of membrane- bound, platelet-derived factor Va expressed between 35-50% of the cofactor activity associated with the entire membrane-bound factor Va pool. (Fager et al., 2010). Thus, following thrombin activation, two factor Va pools are associated with the activated platelet membrane, a non-dissociable pool and a dissociable pool. Subsequent studies by

Wood and colleagues have shown that this non-dissociable pool of platelet-derived factor

Va is glycosylphosphatidylinositol (GPI)-anchored to the platelet membrane via Ser692 at the C-terminus of the factor Va heavy chain, and within a GPI-consensus sequence (Wood et al., 2008). The functional characterization of this GPI-anchored pool of platelet-derived factor Va is detailed in Chapter Four of this dissertation.

1.2.2. Structural Features of Prothrombin

The prothrombin zymogen (also known as Factor II, Figure 10) is one of the most abundant

(0.1 mg/mL) coagulation proteins found in plasma. Prothrombin is synthesized by the liver and released into the circulation as a 579 amino acid single chain polypeptide with a molecular weight of ~ 72 kDa (Herbert, 1940; Kisiel & Hanahan, 1973; Lundblad et al.,

1976; McDuffie et al., 1979). As shown in Figure 10, human prothrombin is composed of an N-terminal fragment 1 (residues 1-155), followed by fragment 2 (residues 156-271), and a serine protease domain at the C-terminal half (residues 272-579). The membrane- binding γ-carboxyglutamic acid (Gla) domain and kringle 1 domain (residues 1-40 & 65-

143, respectively) make up fragment 1 while kringle 2 domain (residues 170-248) forms part of fragment 2. The serine protease domain is comprised of an A-chain (residues 272-

320), and a catalytic B-chain (residues 321-579) (Heldebrant et al., 1973a; Butkowski et al., 1977; Degen et al., 1983).

Figure 10. Structural Domains of Human Prothrombin. Prothrombin is composed of 579 amino acids. The amino-terminus contains about 40 γ- carboxyglutamic acids forming the Gla domain. This is followed by kringle domains 1 and 2. The Gla domain and kringle 1 form the fragment 1 region of prothrombin (red). The fragment 2 regions contains kringle 2 (blue). The carboxyl-terminal half contains the serine protease domains which consist of the A- and B-chains of the thrombin (gray). The two chains are linked by a disulfide bond between Cys293 and Cys439 as shown above. Several mutational studies have identified key regions of prothrombin that interact with Prothrombinase at the enzyme-substrate interface and may play a role in the activation of the zymogen. Residues within the kringle 1 and 2 domains have been reported to interact with the cofactor, factor Va (Kotkow et al., 1995; Deguchi et al., 1997). Similarly, residues between the autolysis loop (residues 473-487) (Subramanian et al., 2004) and proexosite 1

(Chen et al., 2003) all within the serine protease domain of prothrombin contribute to factor

Va binding in Prothrombinase. Additionally, factor Xa appears to interact with residues

557-571 in the protease domain of prothrombin (Yegneswaran et al., 2003). Studies show that the kringle 2 domain within the fragment 2 region promotes prothrombin activation by

Prothrombinase. According to studies by Krishnaswamy and Walker, this fragment 2 region may mediate conformational changes which expose cleavage sites within the protease domain (Krishnaswamy & Walker, 1997). Recent studies by Pozzi and colleagues showed that a truncated prothrombin mutant, Prothrombin Δ146-167 (with deletion of residues connecting kringle 1 and 2 domains) was defective in thrombin generation due to impaired interaction with factor Va (Pozzi et al., 2014). While the complete crystal structure of prothrombin remains elusive, a recently solved 3.3 Å resolution structure of a

Gla-domainless prothrombin has provided key insights into conformational transitions which may mediate prothrombin activation by Prothrombinase (Pozzi et al., 2013). Their structure shows that residue Arg271 (a cleavage site for Prothrombinase) is located within a highly anionic flexible region of fragment 2, whereas Arg320 (another cleavage site for

Prothrombinase) is buried and inaccessible within the serine protease domain of prothrombin. These recent structural details may offer some mechanistic insight into the presentation of the prothrombin zymogen to Prothrombinase.

1.2.2.1 The Activation of Prothrombin to Thrombin

The prothrombin zymogen is converted into active thrombin following proteolytic cleavages at two key residues: Arg271 and Arg320 (Figure 11). Several studies have shown that the rates and sequence of peptide bond cleavages at the Arg271 site or the Arg320 site are dependent on the reaction conditions: i.e. whether catalysis is being mediated by a complete Prothrombinase enzyme or some components of the enzyme in the presence of

the prothrombin substrate. While the rate of thrombin generation is very slow (0.0044 mol thrombin/min/factor Xa)(Nesheim et al., 1979a), the prethrombin 2 (Pre-2) pathway is predominant when prothrombin is activated by only the serine protease, factor Xa, in solution. In this pathway (Figure 11, arrows on the left), initial cleavage occurs at Arg271 producing a non-enzymatic, yet proteinase-like, Pre-2 intermediate, and fragment 1.2. A second cleavage of the Pre-2 intermediate at Arg320 generates thrombin. When the progress of this reaction is visualized following SDS-PAGE under reducing conditions, the appearance and migration of fragment 1.2 is indicative of the formation of the Pre-2 intermediate (Heldebrant et al., 1973b; Esmon & Jackson, 1974; Rosing et al., 1980). The rate of this reaction is further increased by ~30-fold when anionic phospholipid membrane of defined content (75% phosphatidylcholine & 25% phosphatidylserine; PCPS) is added to this reaction (Nesheim et al., 1979a). Similarly, the Pre-2 pathway of prothrombin activation still persist in the presence of the membrane-bound factor Xa (Esmon et al.,

1974a; Krishnaswamy et al., 1987).

Interestingly, although thrombin generation is enhanced by ~10,000-fold, a shift in the order of peptide bond cleavages of prothrombin becomes apparent when factor Va is included in the reaction mixtures (Nesheim et al., 1979a). In this case, the meizothrombin

(mzIIa) pathway predominates when a complete Prothrombinase is assembled (factors Xa

& Va + Ca2+ ion + PCPS vesicles). Thus, mzIIa, an intermediate with enzymatic properties is generated following initial cleavage at Arg320. Subsequent cleavage at Arg271 of the mzIIa intermediate produces thrombin (Figure 11, arrows on the right). Under reducing conditions, the appearance of fragment 1.2A on a gel is indicative of mzIIa generation and

the utilization of the mzIIa pathway (Rosing et al., 1980; Krishnaswamy et al., 1986;

Rosing et al., 1986; Krishnaswamy et al., 1987).

Figure 11. The Mechanism of Prothrombin Activation to Thrombin. Thrombin is generated from prothrombin by two main pathways. The prethrombin 2 (Pre- 2) pathway occurs when there is an initial cleavage at Arg271 to generate a non-enzymatic Pre-2 intermediate, and fragment 1.2 (Frag 1.2). Subsequent cleavage at Arg320 within the Pre-2 intermediate produces thrombin (arrows on the left). Alternatively, in the meizothrombin (mzIIa) pathway, cleavage at Arg320 occurs first generating an enzymatic mzIIa intermediate. A second cleavage of the mzIIa intermediate at Arg271 produces thrombin (arrows on the right). In both pathways, thrombin, the final enzymatic product of prothrombin activation is composed of the A-chain, and the catalytic B-chain linked together by a disulfide bond (Cys293-Cys439). The enzymatic properties of mzIIa are well-documented and several studies show that it possesses significant anticoagulant properties. Meizothrombin is reportedly more

effective at protein C activation to APC compared to thrombin (Doyle & Mann, 1990;

Hackeng et al., 1996; Cote et al., 1997). Doyle and Mann showed that mzIIa converts protein C to APC at more than 100% the efficiency of thrombin in the absence of thrombomodulin (Doyle & Mann, 1990). Similarly, using recombinant prothrombin, Cote and others showed that in the presence of membrane-bound thrombomodulin, mzIIa is 6- fold more effective than thrombin in protein C activation (Hackeng et al., 1996; Cote et al., 1997). In what may be described as a weak contribution to the procoagulant response, mzIIa marginally activates factor V (Tans et al., 1994), platelets, and converts fibrinogen to fibrin (Doyle & Mann, 1990) in the absence of anionic phospholipids. As described above, the presence of a membrane is crucial, and enhances the catalytic activity of mzIIa consistent with the importance of the Gla-domain in mediating membrane binding interactions which support various function of coagulation proteins.

Studies evaluating the mechanism of prothrombin activation by Prothrombinase in purified systems are done in the presence of active site inhibitors of thrombin such as dansylarginine

N-(3-ethyl-1,5-pentanediyl)amine (DAPA). However, during vascular injury in vivo, the localization of coagulation proteins at the activated platelet surface protects them from inactivating proteins in the plasma and allows for an optimized thrombin generation.

Therefore, the thrombin formed feedback to cleave the prothrombin substrate at Arg155 to generate the Gla-domainless prethrombin-1 (Pre-1) and fragment 1 (which contains the

Gla-domain and kringle 1) (Figure 12). This truncated prothrombin molecule is widely considered a poor substrate for PCPS-bound Prothrombinase due to the absence of the Gla- domain which is believed to anchor the protein to the membrane surface (Stenn & Blout,

1972; Esmon et al., 1974b; Nelsestuen et al., 1974; Bajaj et al., 1975). The conversion of

Pre-1 to thrombin by Prothrombinase, although slow compared to prothrombin, proceeds exclusively through the prethrombin-2 pathway. Initial cleavage at Arg271 produces the

Pre-2 intermediate and fragment 2 followed by cleavage at Arg320 of the Pre-2 intermediate to generate thrombin. Although extremely slow, the Pre-2 intermediate may also be cleaved at Arg284 by the thrombin generated in situ to form the Pre-2 des 1-13 (also known as Pre-

2 prime, Pre-2’), and a thirteen amino acid peptide (F 1-13) (Figure 12) (Downing et al.,

1975). Recently, Wood and colleagues monitored prothrombin activation by mimicking conditions prevalent in vivo following an injury. Thus, collagen and factor Xa were simultaneous added to platelets in the absence of thrombin inhibitors and the reaction products monitored over time. They observed that the Pre-1 formed in situ from thrombin feedback cleavage was effectively converted to thrombin by platelet-bound

Prothrombinase in contrast to its accumulation and slow turnover in reactions by PCPS- bound Prothrombinase (Wood et al., 2011). Their study show a distinction in the utilization of Pre-1 by platelet-bound Prothrombinase vs. PCPS-bound Prothrombinase and support a differential assembly of the enzyme on either membrane surfaces.

Further studies investigating the mechanism by which platelet-bound Prothrombinase effectively utilizes Pre-1 in thrombin generation is detailed in this dissertation (Chapter

Three). Several studies have reported that Prothrombinase bound to model membranes do not mimic that bound to activated platelets. These differences will be highlighted below to raise awareness of the need to conduct research using the more physiologically relevant

surfaces such as activated platelets to delineate the molecular mechanisms regulating thrombin generation.

Figure 12. The Formation of Prethrombin-1 Intermediate During Prothrombin Activation and Further Conversion to Thrombin. Prothrombin activation by prothrombinase assembled on the activated platelet surface occurs devoid of thrombin inhibitors. The thrombin formed can feedback and cleave prothrombin at Arg155 (red) to generate a prethrombin-1 (Pre-1) intermediate. The Pre-1 intermediate is further activated by prothrombinase via cleavage at Arg271 to generated fragment 2 (Frag 2) and the inactive prethrombin-2 (Pre-2) intermediate (Prethrombin-2 pathway). A second cleavage of the Pre-2 intermediate generates thrombin. Thrombin can also cleave the Pre-2 intermediate albeit very slowly at Arg248 (red) to produce a Pre-2 prime (Pre-2’) lacking 13 amino acids of the A-chain (F R1-13).

1.2.3. The Assembly of Prothrombinase on the Activated Platelet Surface versus

Anionic Phospholipid Vesicles

Cholesterol, glycolipids, and phospholipids constitute approximately 70% of the plasma membrane of platelets. Quiescent platelets like most cells have asymmetric distribution of key lipids including phosphatidylethanolamine (PE), phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylinositol (PI), and sphingomyelins. Dormann and colleagues observed that when platelets are activated by various agonists, their surface concentration of PS increases from between 2 to 12% via a flip flop mechanism where previous inner leaflet-based anionic phospholipids (PS) move to the outer leaflet providing key catalytic surfaces for the assembly of coagulation factors during the procoagulant response (Dormann et al., 1998). Similarly, Shi and colleagues reported that the percentage of PS exposed following platelet activation was less than 8% (Shi et al., 2008).

Nonetheless, it is undisputed that activated platelets provide the surface for Prothrombinase assembly during hemostasis (Walsh, 1974). While platelets are the main physiological surface, the majority of published studies evaluating the structure/function relationships of coagulation enzyme complexes in purified systems have been done using phospholipid vesicles of defined content containing 75% phosphatidylcholine and 25% phosphatidylserine (PCPS vesicles). Early studies by Barton and colleagues demonstrated that the surface charge density associated with PS is absolutely essential to factor Xa binding. They observed that optimal procoagulant activity for factor Xa, is associated with lipid surfaces comprised of both PS and PC (Barton et al., 1970).

As described earlier in section 1.2.2.1, thrombin is generated via two major pathways from prothrombin: the Pre-2 or meizothrombin pathway. Interestingly, there is substantial evidence that the meizothrombin pathway (Figure 11, arrows on the right) is utilized exclusively by Prothrombinase bound to PCPS vesicles in purified systems

(Rosing et al., 1980; Krishnaswamy et al., 1986; Rosing et al., 1986; Krishnaswamy et al.,

1987; Tans et al., 1991a). Other studies by Tans and colleagues monitoring thrombin generation in a more complicated system, such as plasma, observed confounding results.

They observed that when the coagulation system was initiated with a Tissue factor-factor

VIIa complex the meizothrombin intermediate was undetectable using immunoblotting techniques (Tans et al., 1991a). Similarly, studies of thrombin generation in whole blood following initiation by recombinant tissue factor revealed the accumulation of both Pre-2 and thrombin after 60 min (Rand et al., 1996). Likewise, using quantitative immunoassays,

Undas and colleagues detected the accumulation and utilization of the Pre-2 intermediate in tissue factor-initiated coagulation in whole blood (Undas et al., 2001). Conversely, using active site labeling and streptavidin binding techniques, Bovill and colleagues observed meizothrombin generation in ex vivo whole blood clotting studies. Collectively, these studies suggest that the Pre-2 pathway (Figure 11, arrows on the left) of thrombin generation is paramount in vivo following an injury albeit the meizothrombin pathway may be utilized to some extent.

A recent seminal publication by Wood and colleagues from our laboratory suggest that when Prothrombinase is bound to the activated platelet surface, thrombin generation proceeds exclusively via the Pre-2 pathway (Wood et al., 2011). Using radiolabeled

recombinant prothrombin species with either of the two cleavage sites at Arg271 or Arg320 mutated to Ala, they showed that when Prothrombinase is bound to collagen- or thrombin- activated platelets, cleavage at the 271 site is preferred followed by cleavage at the 320 site. They therefore, proposed that the generation of thrombin via the Pre-2 pathway by platelet-bound Prothrombinase avoids the formation of the anticoagulant meizothrombin intermediate during injury where the procoagulant response is more desired (Wood et al.,

2011). This mechanism is further supported in recent studies by Haynes and colleagues examining prothrombin activation by platelet-bound Prothrombinase under in vitro flow conditions similar to venous shear rates. They observed that the Pre-2 intermediate remain associated with the platelet-bound enzyme following its formation consistent with its absence in the effluent. On the other hand, in reactions of PCPS-bound Prothrombinase, the effluent contained the meizothrombin intermediate. This study indicated that prothrombin activation by platelet-bound Prothrombinase proceeds via a concerted mechanism in which cleavage at Arg271 is followed immediately by Arg320 and without release of the intermediate from the enzyme (Haynes et al., 2012).

Several other findings support the differential assembly and functions of PCPS- vs platelet-bound Prothrombinase. Prothrombinase components including factor Xa (Kd =

-6 -9 2.7 x 10 M) (Nesheim et al., 1981) and factor Va (Kd = 2.7 x 10 M) (Krishnaswamy &

Mann, 1988) independently bind and associate with phospholipid vesicles (e.g. PCPS).

However, factors Xa and Va cannot bind to quiescent platelets (Bouchard et al., 1997a).

Both factors would associate with activated platelets yet, binding of the former (factor Xa) is absolutely dependent on the binding of the latter (factor Va) to the activated platelet

surface (Tracy et al., 1985). Additionally, while prothrombin does not associate with activated platelets independently of Prothrombinase, binding interactions of both factors

Xa, and Va to phospholipid vesicles can be displaced with excess prothrombin(Nesheim et al., 1993). On either surface, however, factors Xa and Va associate via 1:1 stoichiometric complex in the presence of calcium ions to activate prothrombin to thrombin. This

-10 interaction is defined by a Kd of ~ 1 x 10 M on both the activated platelet surface and phospholipid vesicles, respectively. (Tracy et al., 1985; Krishnaswamy et al., 1988;

Krishnaswamy, 1990).

Several studies suggest that the binding of factor Va appears to provide part of the receptor for factor Xa on activated platelets (Tracy et al., 1992; Nesheim et al., 1993).

Additionally, other studies have shown that the degree of platelet activation correlates with the exposed binding sites for either protein (Bouchard et al., 1997b) and factor Va binds

-10 activated platelets with high affinity (Kd = 3 x 10 M) (Tracy et al., 1979). Early studies by Miletich and colleagues reported that factor V deficiency correlated with reduced factor

Xa binding. They showed that adding human factor V antibodies to thrombin-activated platelets prevents factor Xa binding, and thrombin generation catalyzed by Prothrombinase

(Miletich et al., 1978a). Additionally, they demonstrated that activated platelets express high affinity (0.4 x 10-9 M) binding sites for factor Xa which cannot be competed off by the zymogens: factor X, and prothrombin or the protease of the intrinsic tenase, factor IXaβ

(Miletich et al., 1978b).

When used as a platelet agonist, approximately 50 times more thrombin is required for the expression of maximal binding sites for factor Xa (50 nM thrombin, ~3000

molecules/platelet) compared to factor Va (1 nM thrombin, ~6000 molecules/platelet) on platelets, respectively (Tracy et al., 1979). Recent studies by Fager from our group demonstrated that distinct subpopulation of activated platelets differentially bind factors

Xa and Va. They showed that two subpopulations independently bind either factor Va or both factors Xa, and Va. A third subpopulation, however, binds neither protein in their study (Fager et al., 2010). Together, these studies suggest that not all platelets are the same, and that agonist-simulated platelets perhaps express distinct receptors/binding sites which support the assembly and function of Prothrombinase, in contrast to the indiscriminate binding of the enzyme components on phospholipid vesicles. While the factor Va receptor remains elusive, some studies suggest that the membrane-associated protein, effector cell protease receptor-1 (EPR-1) interacts with, and binds factor Xa (Altieri & Edgington, 1990;

Nicholson et al., 1996). Bouchard and colleagues observed that agonist-simulated platelets express EPR-1 in a concentration-dependent manner and the activity of factor Xa in

Prothrombinase reactions incubated with monoclonal antibodies against EPR-1 was significantly reduced suggesting the antibodies were inhibiting factor Xa binding to the activated platelet surface (Bouchard et al., 1997b).

Studies with phospholipid vesicles appear to implicate a key role for the Gla domain region of factor Xa in membrane binding. Larson and colleagues showed that the activity of a factor Xa construct with Asp substitutions at Gla16, Gla29, and Gla32 was impaired in

Prothrombinase assays on PCPS vesicles. In contrast, significant enzyme activity was reported these constructs were monitored in reactions by platelet-bound Prothrombinase

(Larson et al., 1998). Conversely, characterization of the natural factor X variant (factor

XSt Louis, Gly substitution at Gla7) revealed that, thrombin generation catalyzed by

Prothrombinase was 6-fold higher on phospholipid vesicles compared to thrombin- activated platelets (Rudolph et al., 1996). These studies suggest that, other regions of factor

Xa, besides the proposed Gla-dependent membrane association, are required for the protease to function in Prothrombinase assembled on the activated platelet surface.

Binding interactions of factor Va on phospholipid vesicles (section 1.2.1) have been well characterized compared to the activated platelet surface. However, on either surface, the light chain region including the C1 and C2 domains of factor Va appears important in membrane association (Kane et al., 1982; Tracy & Mann, 1983; Tracy et al.,

1992). Tracy and others reported that the light chain region of factor Va additionally mediates interaction with factor Xa on the activated platelet surface (Tracy & Mann, 1983;

Tracy et al., 1992). Other studies have reported a limited role of the heavy chain of factor

Va in mediating binding interactions with factor Xa (Tracy et al., 1992). The existence of additional unknown binding sites for factors Xa, Va, and, perhaps prothrombin on the activated platelet surface have been postulated and much more work is required (Tracy et al., 1992; Nesheim et al., 1993; Scandura et al., 1996).

While coagulation complex assembly and binding to the membrane surface

(PCPS and platelets) protects from circulating and inactivating inhibitors/proteins, the activated platelet surface appears to offer additional protection from complete factor Va inactivation by APC (Bertina et al., 1994; Camire et al., 1995; Camire et al., 1998a; Oliver et al., 1999). The recent identification of a GPI-anchored pool of platelet-derived factor Va on the activated platelet surface raises an interesting conundrum of whether there is

competition for the remaining binding sites by the dissociable platelet-derived factor V/Va

(in the releasate) and plasma-derived factor Va. Nonetheless, there is evidence for a differential assembly of Prothrombinase on the activated platelet surface compared to phospholipid vesicles of defined content. This dissertation investigates the differential properties of these surfaces in thrombin generation, as well as the effective utilization of prothrombin activation intermediates and the procoagulant contributions of the GPI- anchored pool of platelet-derived factor Va to Prothrombinase function.

1.3. Summary of Introduction

The importance of thrombin generation during the hemostatic response cannot be overemphasized. Congenital deficiencies affecting either cellular components or coagulation proteins involved in both the initiation and propagation of the coagulant response has debilitating consequences for the affected individuals. Thus, insufficient or excess thrombin generation produces hemorrhage, and thrombosis, respectively. Both or either of these imbalances in thrombin generation can be life threatening, if not managed properly. It is therefore imperative that the mechanism and function of the final enzyme complex, Prothrombinase, which regulates the generation of thrombin is adequately understood. Under physiological conditions, activated platelets provide both the catalytic surface and essential cofactor that regulate the function of this enzyme. This dissertation investigates further, the platelet-derived components regulating Prothrombinase assembly and function compared to phospholipid model systems.

CHAPTER TWO: Platelet-bound prothrombinase preferentially binds its substrate in a ‘proteinase-like state’ to effect prothrombin activation by the prethrombin-2 pathway

Francis Ayombil1, Laura M. Haynes1, Brittany L. Todd1,† and Paula B. Tracy1,†,*

1Department of Biochemistry, University of Vermont College of Medicine, Burlington, VT *All correspondences should be addressed to: Paula B. Tracy, PhD, 89 Beaumont Ave, Given Medical Research Building C409, Burlington, VT 05405 Telephone: 802-656-1995, Fax: 802-656-8220, and Email: [email protected] †A portion of this work was presented in an abstract format at the International Society of Thrombosis and Haemostasis XXIV Congress, Amsterdam, Netherlands, July 01, 2013

Short title: Platelet prothrombinase locks prothrombin in a proteinase-like state Key words: Platelets, prothrombinase, proteinase-like, zymogen-like, DAPA

Abbreviations used in this manuscript include: 125I-prothrombin, iodine-125 labeled prothrombin; Factor Xa, Russell’s viper venom activator-activated factor X; factor Va, thrombin-activated factor Va; PC, phosphatidylcholine; PS, phosphatidylserine; PCPS vesicle, 75% PC: 25% PS; FPRck, D-Phe-Pro-Arg-chloromethylketone; b-FPRck, biotinylated D-Phe-Pro-Arg-chloromethylketon; DAPA, dansyl-L-arginine N-(3-ethyl- 1,5-pentanediyl)amide.

ABSTRACT

Platelet-bound prothrombinase effects the generation of hemostatically relevant amounts of thrombin from prothrombin via ordered, and concerted cleavages at Arg271 and Arg320, referred to as the prethrombin-2 pathway of activation. The molecular mechanism by which these distant cleavage sites are presented to the active site of Prothrombinase appears to reside in both the conformational flexibility of prothrombin and the quaternary structure of the membrane-bound enzyme complex, such that prothrombin can be stabilized in one of two states, zymogen-like or proteinase-like, through expression of a pseudo-active site.

Initial cleavage at Arg271 occurs in a proteinase-like state and at Arg320 in a zymogen-like state. Inhibitor binding studies and product formation profiles were used to monitor the changes within the catalytic domain of prothrombin subsequent to its binding to platelet- bound Prothrombinase, or Prothrombinase bound to phospholipid vesicles of defined content [(25% phosphatidylserine:75% phosphatidylcholine), PCPS]. Active site blocked, platelet-bound Prothrombinase was ~ four times more effective in supporting the incorporation of biotinylated D-Phe-Pro-Arg-chloromethylketone (FPRck) into the exposed pseudo-active site of prothrombin when compared to the active site-blocked

PCPS-bound enzyme at equivalent concentrations. Increasing concentrations of the reversible, active site inhibitor dansyl-L-arginine N-(3-ethyl-1,5-pentanediyl)amide

(DAPA) shifted prothrombin equilibrium toward the proteinase-like state, which correlated with an increased rate of cleavage at Arg271 confirming that the platelet-bound enzyme preferred this conformation. In marked contrast, an inverse relationship was observed with the PCPS-bound enzyme. As the equilibrium to the proteinase-like state increased, the

initial cleavage at Arg320 became undetectable, and minimal cleavage at Arg271 appeared until little thrombin was generated. These combined data suggest that platelet-bound

Prothrombin effects thrombin generation solely through the prethrombin-2 pathway by binding its substrate in a proteinase-like conformation.

INTRODUCTION

Prothrombinase, a 1:1 stoichiometric complex of factors Xa and Va assembles in the presence of Ca2+ ions on an appropriate membrane and is the final enzyme complex of the coagulation cascade (1,2). Physiologically, prothrombin is the exclusive substrate for

Prothrombinase and is converted to thrombin (2). The catalytic membrane surface is absolutely essential for accelerating the rate of thrombin generation by approximately

1,000-fold (1). Furthermore, the nature of this membrane surface can dictate the order of peptide bond cleavages. Prothrombinase bound to phospholipid vesicles containing 25% phosphatidylserine and 75% phosphatidylcholine (PCPS vesicles) results in thrombin generation via the anticoagulant enzyme, meizothrombin, following an initial cleavage at

Arg320 (3,4) (Figure 1, right arrow). Conversely, the zymogen prethrombin-2 intermediate predominates when platelet-bound Prothrombinase cleaves prothrombin through an initial cleavage at Arg271 (5,6) suggesting the differential assembly of Prothrombinase bound to either surface (Figure1, left arrow).

Prothrombin and its derivatives have been reported (7-10) to undergo a reversible equilibrium transition between proteinase-like and zymogen-like states that favor initial cleavages at Arg271 and Arg320, respectively. Binding and critical ligand-mediated exosite interactions are thought to influence substrate affinity and the stabilization of prothrombin in either the zymogen- or proteinase-like state (8,9,11). Using prothrombin constructs,

Bianchini et al. demonstrated unequivocally that the equilibrium between zymogen- and proteinase-like states can be shifted in favor of the latter in the presence of active site inhibitors such as dansyl-L-arginine N-(3-ethyl-1,5-pentanediyl)amide (DAPA) and D-

Phe-Pro-Arg-chloromethylketone (FPRck) (12). DAPA binding to prothrombin stabilizes the proteinase-like state and promotes enhanced cleavage at Arg271 rather than Arg320 in reactions by PCPS-bound Prothrombinase.

Steady state kinetic measurements that demonstrate an ~30-fold decrease in the catalytic efficiency of cleavage at Arg271 vs Arg320 when Prothrombinase is bound to PCPS vesicles offer strong support for the preferred cleavage at Arg320 and the formation of meizothrombin in reactions catalyzed by PCPS-bound Prothrombinase (11), whereas prethrombin-2 is transiently generated by Prothrombinase bound to the activated platelet membrane as demonstrate by Wood, et al. (5). The mechanisms of prothrombin activation by platelet- and PCPS-bound Prothrombinase are further differentiated as the anticoagulant meizothrombin intermediate is released from the PCPS-bound enzyme complex, while the prethrombin-2 intermediate is not released from platelet-bound Prothrombinase, revealing a concerted mechanism that supports the procoagulant response required for hemostasis

(6).

Platelet-bound Prothrombinase only releases thrombin, not meizothrombin and we hypothesize that this occurs due to the platelet-bound enzyme complex promoting a shift in the zymogen- to proteinase-like state of prothrombin or a preference for prothrombin already in a proteinase-like state. We demonstrate in the present study that shifting the equilibrium of prothrombin towards the proteinase-like state increases both the rate of prothrombin activation and confirms the utilization of the prethrombin-2 activation pathway during thrombin generation by Prothrombinase assembled on activated platelets.

Furthermore, the transition to the proteinase-like state makes prothrombin a poorer

substrate for the PCPS-bound complex by shifting the proteolytic mechanism to the prethrombin-2 pathway (Fig. 1, left arrow) with decreased utilization of the meizothrombin pathway (Fig. 1, right arrow).

MATERIALS AND METHODS

Materials

Dansylarginine N-(3-ethyl-1,5-pentanediyl)amine (DAPA), Phe-Pro-Arg- chloromethylketone (FPRck), and biotinylated-FPRck (b-FPRck) were purchased from

Haematologic Technologies (Essex Junction, VT). Hirudin was purchased from

Calbiochem (Gibbstown, NJ). Re-crystallized bovine albumin was purchased from MP

Biomedicals (Solon, OH). Arg-Gly-Asp-Ser peptide (RGDS) with greater than 99% purity was purchased from ABI Scientific Inc. (Sterling, VA). The chromogenic thrombin- specific substrate Spectrozyme TH was from Sekisui Diagnostics (San Diego, CA).

Phospholipid vesicles comprised of 75% phosphatidylcholine (wt/wt) and 25% phosphatidylserine (wt/wt) (PCPS) were a generous gift from Dr. Kenneth Mann

(University of Vermont) and prepared according to the procedure of Higgins and Mann

(13) with reagents purchased from Avanti Polar Lipids (Alabaster, AL). IODO-GEN tubes and proprietary protein-free T20 (TBS) blocking buffer were obtained from Thermo Fisher

Scientific (Rockford, IL). Precision plus protein molecular weight standards were purchased from Bio-Rad (Hercules, CA). X-ray autoradiography film, and Western blotting chemiluminescence reagent were purchased from Krackeler Scientific (Albany,

NY), and PerkinElmer Inc. (Waltham, MA), respectively. Streptavidin-conjugated horseradish peroxidase (HRP) was purchased from Sigma (St. Louis, MO).

Proteins

Human thrombin, factor Xa, and EGRck-factor Xa (active site blocked with Glu-Gly-Arg- chloromethylketone) were obtained from Haematologic Technologies (Essex Junction,

VT). Human prothrombin, and factor V were purified as previously described (2,14) from freshly frozen plasma obtained from consenting healthy adults, according to the guidelines of the University of Vermont’s Institutional Review Board and Committee on Human

o Subjects. Proteins were stored in 50% glycerol containing 5 mM CaCl2 at -20 C. Before use, glycerol was removed by dialysis of sufficient dilution. Factor Va was prepared by activating plasma-derived factor V (1.0 µM) with thrombin (20 nM, 10 min, 37oC). The activity of thrombin was quenched by the addition of hirudin (30 nM). Protein concentrations were determined by UV absorbance at 280 nm with the following extinction

1% -1 -1 coefficients (E 280nm, M cm ) and molecular weights (Da): human prothrombin 13.8,

72,000 (15); human thrombin 18.3, 36,700 (16); human factor Xa 11.6, 46,000 (17) and human factor V, 9.6, 330,000 (18).

Iodination of prothrombin

Human prothrombin (0.5 mg) was radiolabeled with iodine-125 (1.0 mCi) according to the procedure previously described for labeling factor V (19). The integrity of the 125I- prothrombin was monitored by SDS-PAGE as described by Laemmli (20) followed by

Coomassie brilliant blue R-250 staining and phosphorimaging. The ability of the 125I- prothrombin to be activated by PCPS-bound prothrombinase was confirmed as reported previously (5).

Isolation of platelets from whole human blood

according to the method of Mustard et al. with some minor modifications as previously reported (21,22). Following isolation, platelets were resuspended in HEPES-Tyrode (HT) buffer (137 mM NaCl, 2.68 mM KCl, 11.9 mM NaHCO3, 0.42 mM NaH2PO4, 1.0 mM

MgCl2, 5 mM CaCl2, and 0.2% dextrose, pH 7.4). Platelet concentrations were measured using a Coulter Z1 particle counter (Beckman Coulter, Brea, CA) or an ADVIA 120

Hematology Analyzer (Siemens Healthcare Diagnostics, Austin, TX) both of which produced identical cell counts. Washed platelets (3 x 108 /mL) were maximally activated

(23) with thrombin (50 nM) in HT in the presence of 1 mM RGDS (5 min, 25oC). Thrombin was inactivated by the addition of hirudin (75 nM).

Incorporation of b-FPRcK into prothrombin

The incorporation of biotinylated FPRcK (b-FPRck) into the active site of prothrombin bound by either PCPS- or activated platelet-associated prothrombinase was monitored in solution. Prothrombin (1.4 µM) was incubated with PCPS vesicles (20 µM), factor Va (5 nM) and active site-blocked EGR-factor Xa (0.5 nM) in 50 mM HEPES, 110 mM NaCl, 5 mM CaCl2 pH 7.4. Equivalent enzyme was assembled on thrombin-activated platelets (1 x

108 /mL) in a reaction mixture containing factor Va (5 nM), RGDS (1 mM), and EGR- factor Xa (5 nM) in HT buffer pH 7.4. Each reaction mixture was preincubated (5 min,

37oC) before the addition of b-FPRck (14 µM) to initiate the reaction. The reaction was incubated for an additional 60 min to allow for stoichiometric incorporation of b-FPRck into the exposed pseudo-active site of prothrombin associated with prothrombinase. Timed aliquots (0 - 60 min) were withdrawn and quenched three-fold into in either quench buffer

(20 mM HEPES, 150 mM NaCl, 50 mM EDTA, 0.1% PEG 8000, pH 7.4) or sample

preparation buffer (62.5 mM Tris, 5 mM ethylenediaminetetraacetic acid, 2% SDS, 10% glycerol, 0.01% bromophenol blue, pH 6.8; SPB). Alternatively, after an hour of incubation, excess b-FPRck was removed by dialysis into 20 mM HEPES, 150 mM NaCl,

2+ 5 mM CaCl2, pH 7.4 (HBS/Ca ). The activity of b-FPR-thrombin derived from PCPS- associated prothrombinase-catalyzed activation of b-FPR-prothrombin was assessed by

Spectrozyme TH hydrolysis (24) as described elsewhere in the Materials and Methods section. For platelet reactions, the following additional steps were taken to remove confounding biotinylated proteins associated with the activated platelets. At predefined time points (0 – 60 min), the reaction was sampled (150 µL) and diluted to 250 µL with

HBS/Ca2+. The diluted sample was layered over 500 µL Apiezon A oil and n-butylpthalate

(9:1) and the platelets were pelleted by centrifugation at 10,000 xg for 2 min at room temperature. About 150 µL of the supernatant was removed and the reaction quenched as described above into SPB containing 1% βME. The incorporation of b-FPRcK into prothrombin was detected by immunoblotting with streptavidin-conjugated horseradish peroxidase.

Prothrombin activation pathway in the presence of DAPA

Prothrombin (1.4 µM) was incubated at 25oC for 15 min with factor Va (20 nM), activated platelets (1 x 108 /mL), RGDS (1 mM), trace 125I-prothrombin (2,000 cpm/µL), and DAPA

(30, 60, 100, or 150 µM) in HT. In parallel experiments, prothrombin (1.4 µM) was incubated with PCPS vesicles (20 µM), factor Va (20 nM), and DAPA (60, 100, 200, or

500 µM) in HBS/Ca2+ with 0.1% PEG 8000, pH 7.4; HBS/Ca2+/PEG). A 50 µL aliquot was removed for the 0 sec time point and each reaction was subsequently initiated with

factor Xa (5 nM or 0.5 nM for platelet or PCPS reactions, respectively). At timed intervals

(0 – 60 min), aliquots (50 µL) were withdrawn and quenched in SPB containing 1% βME.

Samples were subsequently boiled at 90oC for 2 min. The proteins were resolved by SDS-

PAGE (25,000 cpm/lane) and stained with Coomassie brilliant blue R-250, dried (3 h,

80oC) and visualized by autoradiography.

Kinetics of prothrombin activation in the presence of DAPA

Initial velocities of thrombin generation from prothrombin was monitored in two-stage assays. Prothrombin (0 - 10 µM) activation was monitored via Prothrombinase (20 nM factor Va and 0.5 nM factor Xa) assembled on thrombin-activated platelets (1 x 108/mL) in the presence of DAPA (30, 60, 100, 150 µM) and RGDS (1.0 mM) in HT. The reaction was preincubated (15 min, 25oC) before the addition of factor Xa to initiate the reaction.

Aliquots were removed at timed intervals (10, 20, 30 and 40 sec) and diluted three-fold in quench buffer. The concentration of thrombin in each aliquote was determined using

Spectrozyme TH hydrolysis.

Spectrozyme-TH hydrolysis and estimation of thrombin concentration

The concentration of thrombin produced at each sampled time point in Prothrombinase assays was determined by the rate of hydrolysis of the thrombin-specific substrate,

Spectrozyme TH (Sekisui) compared to a standard curve containing the appropriate final concentration of DAPA. Briefly, 80 µL of Spectrozyme TH was added to 20 µL of sample and the rate of hydrolysis of the substrate followed at 405 nm (10 min, 30oC) using a

SpectroMax250 microtiter plate reader (Molecular Devices, Sunnyvale, CA).

SDS-PAGE, Western blotting and autoradiography

Samples quenched in SPB were disulfide bond reduced with 1% βME and boiled for 2 min at 90oC. An equivalent amount of each sample was loaded onto 10% polyacrylamide Tris-

Glycine gels (Invitrogen, Carlsbad, CA) by protein (µg) or radioactivity (CPM) and the proteolytic products resolved by the method of Laemmli (20). Protein bands were transferred onto nitrocellulose membrane as described by Towbin et al. (25), and visualized by immunoblotting using streptavidin-conjugated horseradish peroxidase (0.2 µg/mL).

Alternatively, protein bands were stained with Coomassie brilliant blue R-250, dried at

80oC for 3 h and then exposed to a phosphorimaging screen overnight. The imaging screen was scanned using Molecular Imager FX running Quantity One software v.4.6.3 (Bio-Rad).

Kinetic data analyses

Data obtained from kinetic studies were fitted to linear regression plots to obtain initial rates and subsequently to nonlinear Michaelis-Menten equation using GraphPad Prism v6.0 software (GraphPad) to estimate the kinetic parameters. Alternatively, bands were scanned and quantitative densitometry analyses performed using the Bio-Rad Quantity One v.4.6.3 provided by the NIH.

RESULTS

Maturation of prothrombin’s pro-active site upon binding to platelet-bound

Prothrombinase

The ability of platelet-bound Prothrombinase to bind prothrombin and expose the active site resulting in a proteinase-like state was confirmed by the abundant incorporation of b-

FPRck (Fig. 2). In contrast, minimal b-FPRck was incorporated into the pro-active site of prothrombin. Quantitative densitometry analyses showed approximately four-fold more incorporation of b-FPRck into prothrombin in the presence of the platelet-bound

Prothrombinase compared to the PCPS-bound enzyme (Fig. 2). Furthermore, we confirmed that the b-FPR-thrombin derived from b-FPR-prothrombin following activation had little to no amidolytic activity (< 1%) towards the thrombin-specific chromogenic substrate

Spectrozyme TH, thus demonstrating the active-site specific incorporation of bFPRck.

Stabilization of the proteinase-like state of prothrombin by DAPA dictates its activation by both platelet- and PCPS-bound Prothrombinase

Cleavage of 125I-prothrombin by platelet-bound Prothrombinase produced the nonenzymatic prethrombin-2 intermediate established as a consequence of ordered and concerted cleavages at Arg271 and Arg320 to produce thrombin (5,6). In reactions of platelet- bound Prothrombinase, the appearance of the fragment 1.2 band is followed by the transient appearance of prethrombin-2 similar to reactions of prothrombin activation by factor Xa alone in solution or factor Xa bound PCPS (26). As shown in Fig. 3, the prethrombin-2 pathway of prothrombin activation remained unchanged as the concentration of the reversible active site inhibitor, DAPA, was increased from 30 µM (Fig. 3A.i) to 150 µM

(Fig. 3A.iv). There were no detectable levels of the enzymatic, meizothrombin intermediate. However, the rate of prothrombin consumption and product formation was positively correlated with the concentration of DAPA (Figs. 3A and 3B) as determined by quantitative densitometry. When the fraction of prothrombin remaining during the time course over 60 min was fit to a single exponential phase decay equation, the pseudo-first order rate constant for prothrombin activation (Fig 4A) increased by ~ four-fold at 150 µM

(0.24 min-1) compared to 30 µM DAPA (0.06 min-1) consistent with the increased consumption of prothrombin observed in Fig. 3.

Platelet-bound Prothrombinase preferentially activates prothrombin via the prethrombin-2 intermediate pathway (5,6). We therefore reasoned that shifting the equilibrium of prothrombin towards a proteinase-like state would accelerate the rate of prothrombin activation to thrombin by Prothrombinase assembled on the activated platelet membrane. Fig. 4B shows quantitative densitometry of the full time course of fragment 1.2 generation by platelet-bound Prothrombinase in the presence of increasing concentrations of DAPA. The rate of appearance of fragment 1.2 is consistent with increased consumption of prothrombin and the generation of thrombin as a function of DAPA concentration (Figs.

4A and 4C). Densitometric analyses show minimal prethrombin-2 accumulation (Fig. 4D); however, when the bands were overexposed via autoradiography, a pattern is evident whereby the formation of the transient prethrombin-2 intermediate on the activated platelets surface appears and is turned over more rapidly faster with increasing DAPA concentrations (Fig. 3B).

Likewise, the effect of DAPA on prothrombin activation was monitored on PCPS vesicles, albeit at significantly higher DAPA concentrations (60-500 µM), where it can be seen that increasing the concentration of DAPA alters the product profiles and the rate of prothrombin consumption (Fig. 5). In contrast to Prothrombinase assembled on the activated-platelet membrane, prothrombin is activated by an initial cleavage at Arg320 via the alternative meizothrombin pathway by PCPS-bound Prothrombinase with little to no detection of the prethrombin-2 intermediate (4). As expected, the initial and lingering appearance of the meizothrombin intermediate (indicated by Frag 1.2A) was detected by

SDS-PAGE under reducing conditions consistent with an initial cleavage at Arg320 of prothrombin (Fig. 5A-D). In the presence of 200 µM DAPA (Fig. 5C), the meizothrombin intermediate unexpectedly persisted at relatively high concentrations over the time course of the reaction; however, when 500 µM DAPA (Fig. 5D) was present, virtually no prothrombin activation was observed. When the data obtained by quantitative densitometry were analyzed by a one-phase decay equation as before, we observed that at approximately eight times the initial concentration of DAPA (60 vs 500 µM), the pseudo-first order rate constant for prothrombin consumption was reduced by approximately eight-fold (0.25 vs

0.03 min-1, respectively) on PCPS vesicles (Fig. 6). Our data show an inverse correlation between the concentration of DAPA added to PCPS-Prothrombinase reactions and the rate of prothrombin consumption.

DISCUSSION

Prothrombin is irreversibly converted into thrombin following sequential proteolytic cleavages by Prothrombinase at Arg271 and Arg320 (Fig. 1). The biochemical mechanism underlying this transition from the zymogen- to the proteinase-like state by the

Prothrombinase complex is not completely understood. Several models of Prothrombinase have been proposed (27-30) yet no definitive structure exists for the complex assembled on an appropriate phospholipid membrane. Likewise, a full-length structure of the prothrombin molecule has also remained elusive in part due to the presence of the - carboxyglutamic acid containing Gla-domain; however, recently a structure of a Gla- domainless prothrombin molecule has been solved (31) providing some insight and direction in our interpretations of data from functional studies. Prothrombin exhibits considerable length (~80 Å) with reports of significant conformational flexibility that allows the zymogen to exist in either an extended or partially collapsed state at a 3 to 2 ratio, respectively (31).

We hypothesize that the extended conformation of prothrombin corresponds to the proteinase-like state, while the partially collapsed conformation corresponds to its zymogen-like state (31,32) and that this plasticity defines the preferred conformation of prothrombin for both platelet- and PCPS-bound Prothrombinase. As platelet-bound

Prothrombinase prefers to activate prothrombin via an initial cleavage at Arg271, we postulate that it selectively binds prothrombin in its extended conformation/proteinase-like state; however, thrombin generation remains efficient as an equilibrium concentration of the proteinase-like state is continually reestablished according to Le Chatelier’s principle.

Likewise, we would expect the inverse process to occur when Prothrombinase is assembled on PCPS vesicles.

An alternative hypothesis is that platelet-bound Prothrombinase actively reorients prothrombin to a proteinase-like state using a non-proteolytic mechanism prior to its activation. Non-proteolytic activations have been reported for both the zymogens prothrombin and plasminogen by the bacterial-derived activators staphylocoagulase (33) and streptokinase (34), respectively. In drawing some comparison to these findings, it is reasonable to speculate that the binding of prothrombin by platelet-bound Prothrombinase induces a conformational change which may not expose a functional catalytic site but makes accessible the pseudo-active site and allows for the binding and incorporation of b-

FPRck observed in the present study. Conversely, prothrombin interacting with PCPS- bound Prothrombinase does not transition to a proteinase-like state prior to proteolytic activation and therefore maintains an inaccessible pseudo-active site preventing the incorporation of b-FPRck.

The results of the current study do not definitively demonstrate that either of the above hypotheses is correct, but rather they suggest that a combination of both mechanisms is being utilized by platelet-bound Prothrombinase to promote the activation of prothrombin along a transition of proteinase-like states: 1) Platelets selectively bind the population of circulating prothrombin that exists in a proteinase-like. The rate of prothrombin activation by platelet-bound Prothrombinase can be increased by the addition of DAPA, which shifts the equilibrium of prothrombin towards the proteinase-like state. 2)

Upon binding to platelet-bound Prothrombinase, prothrombin is reoriented such that it is

stabilized in a proteinase-like state with an accessible pseudo-active site; thus, b-FPRck may be incorporated into this pseudo-active site more so than prothrombin in solution or bound to Prothrombinase assembled on PCPS vesicles. 3) Initial cleavage of prothrombin at Arg271 to form prethrombin-2 promotes a further transition to a proteinase-like state (35-

37). As prethrombin-2 is already in the proteinase-like substrate orientation preferred by

Prothrombinase, it remains bound to the enzyme complex prior to and during cleavage at

Arg320 via a concerted mechanism (6). This series of transitions eloquently describes yet another mechanism by which platelets promote a procoagulant response.

In contrast, Prothrombinase bound to PCPS vesicles lacks the mechanisms described above. It has been suggested that prothrombin adopts a zymogen-like character during its activation by Prothrombinase assembled on PCPS vesicles due to the established binding interaction between the Gla-domain containing fragment-1 of prothrombin and the anionic phospholipids of the PCPS membrane restricting the possible orientation the substrate may adapt when in complex with Prothrombinase (10,38,39). Interestingly, decreasing the phosphatidylserine content of PCPS vesicles from 25% to 2.5% promotes initial cleavage of prothrombin at Arg271 and the transient appearance of prethrombin-2 in a manner similar to that observed when Prothrombinase is assembled on the activated platelet membrane with one notable exception: the rate of the reaction is dramatically reduced (11,40). Therefore, it may be possible that while PCPS-bound Prothrombinase would prefer prothrombin in a proteinase-like state, the thermodynamic advantage gained by membrane binding sufficiently offsets this preferred binding conformation. Again, this observation supports the uniqueness of the activated platelet membrane in promoting

effective Prothrombinase assembly and thrombin generation (23,24,41,42) without the requirement for high phosphatidylserine percent composition of their outer membrane leaflet, presumably due to the presence of specific receptors for the components of

Prothrombinase.

Finally, we speculate that while platelets must promote a procoagulant response to maintain hemostasis in response to vascular injury, other physiologic surfaces exist (e.g. microparticles) upon which Prothrombinase may assemble that does not utilize a continuum of proteinase-like states in an effort to thwart undesirable thrombin generation via the release of the anticoagulant intermediate meizothrombin, an active species that has less than 10% the activity of thrombin with regard to platelet activation and fibrin formation (43-45). Therefore, while the secondary meizothrombin activation pathway has served as a red herring in the coagulation field for over 40 years (4,10,46,47), it is not a predominately procoagulant pathway but rather an attempt of the body to avoid unwanted thrombosis.

ACKNOWLEDGEMENT

This work was supported by National Institute of Health [grant P01HL46703, Project 3 to

P.B.T]. F.A. was supported by the Department of Biochemistry, UVM and the Cellular,

Molecular and Biomedical Science Program, College of Medicine, UVM. B.LT was supported by NIH grant T32 HL007594, Hemostasis and Thrombosis Program for

Academic Trainees (to K.G. Mann). The authors would like to acknowledge Dr. Beth

Bouchard and Dr. Jay Silveira for their critical review and helpful suggestions in preparing this manuscript. 78

AUTHORSHIP

Contribution: F.A. designed and performed experiments, analyzed results, and wrote the manuscript; B.L.T designed and performed experiments, and analyzed results. P.B.T. designed the research and wrote the manuscript.

Conflict of interest disclosure: The authors declare no competing financial interest.

Correspondence: Paula B. Tracy, 89 Beaumont Ave, Given Medical Research Building

C409, Burlington, VT 05405; Telephone: 802-656-1995; Fax: 802-656-8220; and Email: [email protected].

REFERENCES

1. Nesheim, M. E., Taswell, J. B., and Mann, K. G. (1979) The contribution of bovine Factor V and Factor Va to the activity of prothrombinase. J Biol Chem 254, 10952- 10962

2. Mann, K. G. (1976) Prothrombin. Methods Enzymol 45, 123-156

3. Rosing, J., Zwaal, R. F., and Tans, G. (1986) Formation of meizothrombin as intermediate in factor Xa-catalyzed prothrombin activation. J Biol Chem 261, 4224- 4228

4. Krishnaswamy, S., Mann, K. G., and Nesheim, M. E. (1986) The prothrombinase- catalyzed activation of prothrombin proceeds through the intermediate meizothrombin in an ordered, sequential reaction. J Biol Chem 261, 8977-8984

5. Wood, J. P., Silveira, J. R., Maille, N. M., Haynes, L. M., and Tracy, P. B. (2011) Prothrombin activation on the activated platelet surface optimizes expression of procoagulant activity. Blood 117, 1710-1718

6. Haynes, L. M., Bouchard, B. A., Tracy, P. B., and Mann, K. G. (2012) Prothrombin Activation by Platelet-associated Prothrombinase Proceeds through the Prethrombin-2 Pathway via a Concerted Mechanism. Journal of Biological Chemistry 287, 38647-38655

7. Pozzi, N., Vogt, A. D., Gohara, D. W., and Di Cera, E. (2012) Conformational selection in trypsin-like proteases. Curr Opin Struct Biol 22, 421-431

8. Kamath, P., Huntington, J. A., and Krishnaswamy, S. (2010) Ligand binding shuttles thrombin along a continuum of zymogen- and proteinase-like states. J Biol Chem 285, 28651-28658

9. Lechtenberg, B. C., Johnson, D. J., Freund, S. M., and Huntington, J. A. (2010) NMR resonance assignments of thrombin reveal the conformational and dynamic effects of ligation. Proc Natl Acad Sci U S A 107, 14087-14092

10. Bradford, H. N., and Krishnaswamy, S. (2012) Meizothrombin is an unexpectedly zymogen-like variant of thrombin. J Biol Chem 287, 30414-30425

11. Orcutt, S. J., and Krishnaswamy, S. (2004) Binding of substrate in two conformations to human prothrombinase drives consecutive cleavage at two sites in prothrombin. J Biol Chem 279, 54927-54936

12. Bianchini, E. P., Orcutt, S. J., Panizzi, P., Bock, P. E., and Krishnaswamy, S. (2005) Ratcheting of the substrate from the zymogen to proteinase conformations directs the sequential cleavage of prothrombin by prothrombinase. Proc Natl Acad Sci U S A 102, 10099-10104

13. Higgins, D. L., and Mann, K. G. (1983) The interaction of bovine factor V and factor V-derived peptides with phospholipid vesicles. J Biol Chem 258, 6503-6508

14. Nesheim, M. E., Katzmann, J. A., Tracy, P. B., and Mann, K. G. (1981) Factor V. Methods Enzymol 80 Pt C, 249-274

15. Kisiel, W., and Hanahan, D. J. (1973) Purification and characterization of human Factor II. Biochimica et biophysica acta 304, 103-113

16. Fenton, J. W., Fasco, M. J., and Stackrow, A. B. (1977) Human thrombins. Production, evaluation, and properties of alpha-thrombin. The Journal of biological chemistry 252, 3587-3598

17. Di Scipio, R. G., Hermodson, M. A., Yates, S. G., and Davie, E. W. (1977) A comparison of human prothrombin, factor IX (Christmas factor), factor X (Stuart factor), and protein S. Biochemistry 16, 698-706

18. Nesheim, M. E., Myrmel, K. H., Hibbard, L., and Mann, K. G. (1979) Isolation and characterization of single chain bovine factor V. J Biol Chem 254, 508-517

19. Monkovic, D. D., and Tracy, P. B. (1990) Activation of human factor V by factor Xa and thrombin. Biochemistry 29, 1118-1128

20. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685

21. Mustard, J. F., Perry, D. W., Ardlie, N. G., and Packham, M. A. (1972) Preparation of suspensions of washed platelets from humans. Br J Haematol 22, 193-204

22. Camire, R. M., Kalafatis, M., Simioni, P., Girolami, A., and Tracy, P. B. (1998) Platelet-derived factor Va/Va Leiden cofactor activities are sustained on the surface of activated platelets despite the presence of activated protein C. Blood 91, 2818- 2829

23. Bouchard, B. A., Catcher, C. S., Thrash, B. R., Adida, C., and Tracy, P. B. (1997) Effector cell protease receptor-1, a platelet activation-dependent membrane protein, regulates prothrombinase-catalyzed thrombin generation. J Biol Chem 272, 9244- 9251

24. Fager, A. M., Wood, J. P., Bouchard, B. A., Feng, P., and Tracy, P. B. (2010) Properties of procoagulant platelets: defining and characterizing the subpopulation binding a functional prothrombinase. Arterioscler Thromb Vasc Biol 30, 2400-2407

25. Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci U S A 76, 4350-4354

26. Esmon, C. T., and Jackson, C. M. (1974) The conversion of prothrombin to thrombin. III. The factor Xa-catalyzed activation of prothrombin. J Biol Chem 249, 7782-7790

27. Lechtenberg, B. C., Murray-Rust, T. A., Johnson, D. J., Adams, T. E., Krishnaswamy, S., Camire, R. M., and Huntington, J. A. (2013) Crystal structure of the prothrombinase complex from the venom of Pseudonaja textilis. Blood 122, 2777-2783

28. Lee, C. J., Wu, S., and Pedersen, L. G. (2011) A proposed ternary complex model of prothrombinase with prothrombin: protein-protein docking and molecular dynamics simulations. Journal of thrombosis and haemostasis : JTH 9, 2123-2126

29. Lee, C. J., Lin, P., Chandrasekaran, V., Duke, R. E., Everse, S. J., Perera, L., and Pedersen, L. G. (2008) Proposed structural models of human factor Va and prothrombinase. Journal of thrombosis and haemostasis : JTH 6, 83-89

30. Adams, T. E., Hockin, M. F., Mann, K. G., and Everse, S. J. (2004) The crystal structure of activated protein C-inactivated bovine factor Va: Implications for cofactor function. Proc Natl Acad Sci U S A 101, 8918-8923

31. Pozzi, N., Chen, Z., Gohara, D. W., Niu, W., Heyduk, T., and Di Cera, E. (2013) Crystal structure of prothrombin reveals conformational flexibility and mechanism of activation. The Journal of biological chemistry 288, 22734-22744

32. Pozzi, N., Chen, Z., and Di Cera, E. (2016) How the Linker Connecting the Two Kringles Influences Activation and Conformational Plasticity of Prothrombin. J Biol Chem

33. Hemker, H. C., Bas, B. M., and Muller, A. D. (1975) Activation of a pro-enzyme by a stoichiometric reaction with another protein. The reaction between prothrombin and staphylocoagulase. Biochim Biophys Acta 379, 180-188

34. Lahteenmaki, K., Kuusela, P., and Korhonen, T. K. (2001) Bacterial plasminogen activators and receptors. FEMS microbiology reviews 25, 531-552

35. Krishnaswamy, S. (2004) Exosite-driven substrate specificity and function in coagulation. J. Thomb Haemost. 3, 54-67

36. Krishnaswamy, S., and Betz, A. (1997) Exosites determine macromolecular substrate recognition by prothrombinase. Biochemistry 36, 12080-12086

37. Hibbard, L. S., Nesheim, M. E., and Mann, K. G. (1982) Progressive development of a thrombin inhibitor binding site. Biochemistry 21, 2285-2292

38. Esmon, C. T., Owen, W. G., and Jackson, C. M. (1974) The conversion of prothrombin to thrombin. II. Differentiation between thrombin- and factor Xa- catalyzed proteolyses. J Biol Chem 249, 606-611

39. Bajaj, S. P., Butkowski, R. J., and Mann, K. G. (1975) Prothrombin fragments. Ca2+ binding and activation kinetics. J Biol Chem 250, 2150-2156

40. Kamath, P., and Krishnaswamy, S. (2008) Fate of membrane-bound reactants and products during the activation of human prothrombin by prothrombinase. J Biol Chem 283, 30164-30173

41. Walsh, P. N. (1974) Platelet coagulant activities and hemostasis: a hypothesis. Blood 43, 597-605

42. Tracy, P. B. (2001) Role of platelets and leukocytes in coagulation. Hemostasis and Thrombosis. Basic Principles and Clinical Practice (R.W. Colman, et al). Lippincott Williams & Wilkins, Philadelphia, 575-596

43. Whelihan, M. F., Zachary, V., Orfeo, T., and Mann, K. G. (2012) Prothrombin activation in blood coagulation: the erythrocyte contribution to thrombin generation. Blood 120, 3837-3845

44. Tans, G., Rosing, J., Thomassen, M. C., Heeb, M. J., Zwaal, R. F., and Griffin, J. H. (1991) Comparison of anticoagulant and procoagulant activities of stimulated platelets and platelet-derived microparticles. Blood 77, 2641-2648

45. Fuster, V., Stein, B., Ambrose, J. A., Badimon, L., Badimon, J. J., and Chesebro, J. H. (1990) Atherosclerotic plaque rupture and thrombosis. Evolving concepts. Circulation 82, II47-59

46. Cote, H. C., Bajzar, L., Stevens, W. K., Samis, J. A., Morser, J., MacGillivray, R. T., and Nesheim, M. E. (1997) Functional characterization of recombinant human meizothrombin and Meizothrombin(desF1). Thrombomodulin-dependent activation of protein C and thrombin-activatable fibrinolysis inhibitor (TAFI), platelet aggregation, antithrombin-III inhibition. J Biol Chem 272, 6194-6200

47. Bovill, E. G., Tracy, R. P., Hayes, T. E., Jenny, R. J., Bhushan, F. H., and Mann, K. G. (1995) Evidence that meizothrombin is an intermediate product in the clotting of whole blood. Arterioscler Thromb Vasc Biol 15, 754-758

FIGURES

Figure 1. Prothrombinase-mediated pathway of prothrombin activation in the presence of thrombin inhibitors. Prothrombin is composed of N-terminal fragments 1 (Frag 1) and 2 (Frag 2) and a C- terminal protease domain containing A- and B chains. Thrombin is formed following two proteolytic cleavages at Arg271 (R271) and Arg320 (R320). The prethrombin-2 (Pre-2) pathway that predominates on the activated platelet surface occurs via an initial cleavage at Arg271 resulting in the formation of fragment-1.2 (Frag 1.2), and a non-enzymatic prethrombin-2 (Pre-2) intermediate. A second cleavage at Arg320 of the Pre-2 intermediate produces α-thrombin (left pathway). When assembled on a phospholipid membrane, such as PCPS vesicles, the meizothrombin pathway predominates with cleavage at Arg320 occurring first producing a proteinase intermediate, meizothrombin (mzIIa). Further cleavage at Arg320 of the mzIIa intermediate produces fragment-1.2 and the active protease α-thrombin (right pathway).

Figure 2. Biotinylated FPRck is incorporated into PCPS-bound and platelet-bound prothrombinase. The incorporation of biotinylated FPRck (b-FPRck) into prothrombin was monitored on both PCPS vesicles and thrombin-activated platelets. As detailed in the methods section, equivalent amounts of Prothrombinase were assembled on either PCPS vesicles (20 µM) or thrombin-activated platelets (1 x 108/mL) in reaction mixtures containing active site blocked EGR-factor Xa (0.5 or 5 nM for PCPS or activated platelets, respectively), and factor Va (20 nM). The reaction was initiated with b-FPRck (14 µM) and prothrombin (1.4 µM) and then incubated for an hour. Aliquots of the reaction were quenched, resolved by SDS-PAGE and the incorporation of biotinylated inhibitor assessed by Western blotting analyses.

Figure 3. The effect of increasing concentrations of DAPA on prothrombin activation by platelet-bound Prothrombinase. (A) Prothrombin (1.4 µM) was activated by human Prothrombinase assembled on thrombin-activated platelets (1 x 108/mL) in the presence of trace 125I-prothrombin, factor Va (20 nM), factor Xa (5 nM) and variable concentrations of DAPA: (i) 30 µM, (ii) 60 µM, (iii) 100 µM, and (iv) 150 µM. Timed aliquots were taken (0 – 60 min), quenched and resolved using 10% polyacrylamide gels via SDS-PAGE. Gels were stained, dried, and visualized by autoradiography. Indicated on the left and right sides of the panels are the molecular weight standards (x103 Da), and the cleavage products (II, Prothrombin; Frag 1.2, Fragment 1.2; Pre-2, Prethrombin-2; & B-chain of thrombin), respectively. (B) Prolonged autoradiographic exposure of protein bands shows the profile of transient prethrombin-2 intermediate during the course of the progress curves at the indicated concentrations of DAPA.

Figure 4. The effect of DAPA on the appearance of prothrombin activation species in reactions by platelet-bound Prothrombinase. The reaction conditions are the same as those described in Fig. 2. The activation of prothrombin by platelet-bound Prothrombinase was assessed over 60 min in the presence of 30 µM (●), 60 µM (■), 100 µM (▲), and 150 µM (▼) DAPA. Shown in the panel are reaction profiles of (A) the fraction of prothrombin remaining; (B) the fraction of Frag 1.2 formed; (C) fraction of B-chain of thrombin formed; and (D) the appearance of the transient Pre-2 intermediate estimated by quantitative densitometric analyses and normalized to the density of the prothrombin band at zero seconds.

Figure 5. The effect of DAPA on prothrombin activation by PCPS-bound Prothrombinase. Prothrombinase assays contained prothrombin (1.4 µM) in the presence of 20 nM factor Va, 0.5 nM factor Xa at increasing concentrations of DAPA (A ) 60 µM, (B) 100 µM, (C) 200 µM and (D) 500 µM. Reaction aliquots were removed and quenched at following timed intervals: 0, 1, 2, 3, 4, 5, 8, 12, 16, 20, 30, 45, and 60 min. Samples were resolved using 10% polyacrylamide gels via SDS-PAGE. Protein bands were detected by Coomassie brilliant blue staining. The molecular weights (x103 Da) are shown on the left. Cleavage products are shown on the right as follows: II, Prothrombin; Frag 1.2A, meizothrombin, Fragment 1.2; Pre-2, Prethrombin-2; and the B-chain of thrombin.

Figure 6. Substrate depletion curves for prothrombin at variable DAPA concentrations in reactions by PCPS-bound Prothrombinase. Reaction conditions are the same as those described in Fig. 5. The rate of prothrombin activation by platelet-bound Prothrombinase was assessed over 60 min. The prothrombin remaining at each time as a fraction of the total at zero seconds is indicated on the y-axis following densitometry analyses of bands obtained from autoradiography. Prothrombin depletion in the presence of 60 µM (■), 100 µM (□), 200 µM (●), and 500 µM (○) DAPA are shown above. The data were fitted to a one-phase decay equation using the GraphPad 6 software.

CHAPTER THREE: Platelet-bound prothrombinase has a unique enzyme-substrate interface that evades product inhibition by prothrombin fragment-2

Francis Ayombil1,†, Laura M. Haynes1,† and Paula B. Tracy1,†,*

1Department of Biochemistry, University of Vermont College of Medicine, Burlington, VT *All correspondence should be addressed to: Paula B. Tracy, PhD, 89 Beaumont Ave, Given Medical Research Building C409, Burlington, VT 05405 Telephone: 802-656-1995, Fax: 802-656-8220, and Email: [email protected] †Portions of this work were presented in abstract formats at the International Society of Thrombosis and Haemostasis XXIII & XXIV Congresses in Tokyo, Japan (2011) and Toronto, Canada (2015)

Short title: Platelet-bound Prothrombinase evades Frag 2 inhibition Key words: Prothrombinase, platelets, PCPS vesicles, fragment-2

Abbreviations used in this manuscript include: γ-carboxyglutamic acid, Gla; Prethrombin- 1, Prethrombin-1; Prethrombin-2, Prethrombin-2; Fragment-2, Frag 2; meizothrombin, mzIIa; Factor Xa, Russell’s viper venom activator-activated factor X; factor Va, thrombin- activated factor Va; PC, phosphatidylcholine; PS, phosphatidylserine; PCPS vesicle, 75% PC: 25% PS; DAPA, dansyl-L-arginine N-(3-ethyl-1,5-pentanediyl)amide.

ABSTRACT

Prothrombin is composed of fragment-1 (in part consisting of the -carboxy-glutamic acid containing Gla-domain), fragment-2, and a serine protease domain. During the hemostatic response, thrombin generated following the proteolytic activation of prothrombin by

Prothrombinase can participate in a feedback reaction whereby fragment-1 is removed from prothrombin generating prethrombin-1. Prethrombin-1 has historically been considered a “dead end substrate”, as loss of the membrane-binding Gla-domain renders its rate of activation negligible by Prothrombinase bound to phospholipid vesicles composed of 75% phosphatidylcholine and 25% phosphatidylserine (PCPS). Contradicting this long-held dogma, Prothrombinase bound to the activated-platelet membrane can effectively activate prethrombin-1 formed in situ to thrombin; however, the pathway of prethrombin-1 activation by platelet-bound Prothrombinase is the same as that observed when Prothrombinase is assembled on PCPS vesicles: Initial cleavage at Arg271 generates fragment 2 and the zymogen intermediate prethrombin-2, while subsequent cleavage at

Arg320 results in the formation of thrombin. While platelet-bound Prothrombinase effectively catalyzes the turnover of both prethrombin-1 and prethrombin-2, PCPS-bound

Prothrombinase is an ineffective activator of prethrombin-1 and prethrombin-2 due to an inability to effectively catalyze cleavage at Arg320. Although the catalytic efficiency of

Prothrombinase assembled on either surface is similar, super physiologic concentrations of prethrombin-1 (>20 µM) can reduce its initial rate of activation three-fold the PCPS-bound

Prothrombinase. Furthermore, exogenously added prethrombin-2 inhibits prothrombin activation by PCPS- but not platelet-bound Prothrombinase. Collectively, these

observations suggest that fragment 2 generated in situ inhibits the activation of prethrombin-1 by PCPS-bound Prothrombinase, while its inability to inhibit the analogous reaction catalyzed by platelet-bound Prothrombinase provides further evidence that platelet-bound Prothrombinase possesses unique structural and functional features that support robust thrombin generation.

INTRODUCTION

The activation of the zymogen prothrombin to the critical serine protease thrombin by Prothrombinase, consisting of the protein cofactor factor Va and the serine protease factor Xa assembled in a 1:1 stoichiometric complex on an appropriate membrane surface, is central to the blood coagulation process, and as such is highly regulated. Prothrombin is composed of an N-terminal fragment-1 (comprised of a -carboxy-glutamic acid containing

Gla-domain; residues 1-155), fragment-2 (Frag 2; residues 156-271), and a C-terminal serine protease domain (comprised of disulfide linked A- and B-chains; residues 272-579)

(Fig. 1) (1). Prothrombin is activated to thrombin following the sequential proteolytic cleavage of two peptide bonds, Arg271-Thr272 and Arg320-Ile321 (Fig. 1). Prothrombin lacking fragment 1 (prethrombin-1) is activated to thrombin by an initial cleavage at Arg271 independent of the surface on which Prothrombinase is assembled, resulting in the formation of the zymogen intermediate prethrombin-2 (2,3). Conversely, initial cleavage of prothrombin at Arg320 would result in the formation of the enzymatically active intermediate meizothrombin des fragment-1 (4-6). Prethrombin-1 formation results from proteolytic cleavage of prothrombin by thrombin at Arg155 (7-9). As loss of the membrane binding Gla-domain dramatically reduces efficiency by which thrombin is generated, this reaction has long been considered a mechanism of product feedback inhibition of extraneous thrombin generation (9-11). Furthermore, the ineffective activation of prethrombin-1 to thrombin is demonstrated by the observation that its activation by

Prothrombinase in the absence of an appropriate membrane surface results in an incomplete

conversion to thrombin and formation of a mixture of approximately equal amounts of prethrombin-2 and thrombin (9).

Our laboratory recently demonstrated that prethrombin-2, resulting from initial cleavage at Arg271, is the only major intermediate generated during prothrombin activation by Prothrombinase assembled on the physiologically relevant surface of the activated platelet membrane (3). We have also demonstrated that prethrombin-1 formed in situ during the activation of prothrombin by Prothrombinase assembled on collagen-activated platelets (3) or thrombin-activated platelets under flow is more efficiently converted to thrombin than PCPS-bound Prothrombinase where both Frag 2 and prethrombin-2, resulting from prothrombin cleavage at Arg271, accumulate during their slow activation to thrombin (2). The mechanism for these apparent differences in prethrombin-1 activation is not known, although a growing body of evidence highlights features of activated platelets to promote efficient thrombin generation (12-15).

In the present study we hypothesize that Frag 2 generated during prothrombin activation inhibits the activation of prethrombin-1 by Prothrombinase assembled on PCPS vesicles, but not activated platelets. Interestingly, Frag 2 has been reported to inhibit prothrombin activation by factor Xa (16), and that it makes key contacts with factor Va

(9,17), factor Xa (18), and (pro)exosite II of prethrombin-2 (4,19) during Prothrombinase assembly. Therefore, we will examine the effect Frag 2 has on the inhibition of prethrombin-1 activation by Prothrombinase assembled on activated platelets and PCPS vesicles in order to understand why a previously considered “dead-end” substrate,

prethrombin-1, may contribute to relevant thrombin generation in vivo and to describe yet another pro-hemostatic feature of platelet-bound Prothrombinase.

MATERIALS AND METHODS

Materials

Phe-Pro-Arg-chloromethylketone (FPRck) and dansylarginine N-(3-ethyl-1,5- pentanediyl)amine (DAPA) were purchased from Haematologic Technologies (Essex

Junction, VT). The peptides Gly-Pro-Arg-Pro (GPRP) and Arg-Gly-Asp-Ser (RGDS) were synthesized and characterized by Think Peptides-Proimmune (Sarasota, FL). Hirudin was obtained from Calbiochem (Gibbstown, NJ) and bovine serum albumin was from MP

Biomedicals (Solon, OH). Phospholipid vesicles composed of 75% phosphatidylcholine

(wt/wt) and 25% phosphatidylserine (wt/wt) (PCPS) were a generous gift from Dr. Kenneth

Mann (University of Vermont) and prepared according to the procedure of Higgins and

Mann (20) with reagents purchased from Avanti Polar Lipids (Alabaster, AL). The chromogenic, thrombin-specific substrate Spectrozyme TH was purchased from Sekisui

Diagnostics (San Diego, CA). Horse anti-mouse IgG-HRP was obtained from Vectors

Laboratories (Burlingame, CA). Burro anti-prethrombin-1 antibody (21) was provided by the University of Vermont Antibody Core Facility.

Coagulation proteins

Human factor Xa and thrombin were purchased from Haematologic Technologies. Human factor V, prothrombin and prothrombin-1 were purified as previously described (22,23) from freshly frozen plasma obtained from consenting, healthy adults, according to the guidelines of the University of Vermont’s Institutional Review Board and Committee on

Human Subjects. Purified proteins were stored in 2 mM CaCl2, 50% glycerol and stored at

-20oC. Excess glycerol was removed by dialyses before experiments. Human factor V (1.0

µM) was activated to factor Va by α-thrombin (20 nM, 10 min, 37oC). The activity of thrombin was quenched by the addition of hirudin (30 nM). Molar protein concentrations were measured by UV absorbance at 280 nm with the following extinction coefficients

1% (E 280nm) and molecular weights (Da): human prothrombin 13.8, 72,000 (24); human prethrombin-1 17.8, 49,900 (22); human α-thrombin 18.3, 36,700 (25); human factor Xa

11.6, 46,000 (26); and human factor V 9.6, 330,000 (27).

Isolation of platelets from blood

Blood from healthy, consenting adults was collected by venipuncture into acid citrate dextrose (71 mM citric acid, 85 mM trisodium, 2% dextrose, pH 4.5) in a ratio of 6 parts of blood to 1 part of acid citrate dextrose. Platelets were subsequently washed and isolated according to the method of Mustard et al. with some minor modifications as previously reported (28,29). Following isolation, platelets were resuspended in HEPES-Tyrode (HT) buffer (137 mM NaCl, 2.68 mM KCl, 11.9 mM NaHCO3, 0.42 mM NaH2PO4, 1.0 mM

MgCl2, 5 mM CaCl2, and 0.2% dextrose, pH 7.4). Platelet concentration was measured using a Coulter Z1 particle counter (Beckman Coulter, Brea, CA) or an ADVIA 120

Hematology Analyzer (Siemens Healthcare Diagnostics, Austin, TX) both of which produced identical cell counts. In some experiments, washed platelets (3 x 108/mL) were maximally activated (12) with thrombin (50 nM) in the presence of 1 mM RGDS in HT buffer pH 7.4 (5 min, 25oC) and activity of thrombin quenched by the addition of hirudin

(75 nM).

Time course of prothrombin activation by platelet-bound Prothrombinase

In these time course reactions, cofactor activity was provided solely by platelet-derived factor Va molecules from α-granule releasates following platelet stimulation. Washed platelets (1 x 108 /mL) were stimulated with thrombin protease-activated receptor (PAR) agonist peptides; PAR1 (100 µM) and PAR4 (500 µM), or thrombin (2 U/mL) and incubated in reaction mixtures containing prothrombin (1.4 µM) and RGDS (1.0 mM) in

HT. The activity of thrombin was inactivated with hirudin (30 nM). The reactions were initiated with 5 or 10 nM factor Xa for thrombin- or PAR-activated platelets following a five min preincubation. Aliquots were withdrawn every 10 sec for two min, and at select intervals for the next 30 min, and subsequently quenched in sample preparation buffer

(312.5 mM Tris, 25 mM EDTA, 10% SDS, 50% glycerol, 0.05% bromophenol blue, pH

6.8; 5X SPB) to make a final of 1X SPB. The proteins were then resolved by SDS-PAGE followed by Western blotting.

Michaelis-Menten kinetic constants of prethrombin-1 activation by human

Prothrombinase

Initial velocities of thrombin generation from prethrombin-1 were monitored in 2-stage assays. Briefly, prethrombin-1 activation (0.5 - 20 µM) was monitored in a reaction mixture containing factor Va (20 nM), DAPA (30 µM), PCPS vesicles (20 µM) in 20 mM HEPES,

150 mM NaCl, pH 7.4 (HBS) containing 5 mM Ca2+ and 0.1% PEG 8000 buffer pH 7.4

(HBS/Ca2+/PEG). Following a five min incubation at 25oC, the reaction was initiated by the addition of factor Xa (0.5 nM). In platelet experiments, reaction conditions were established ensuring equivalent amounts of Prothrombinase (0.5 nM) was assembled on

100

thrombin-activated platelets (1 x 108 /mL) (3) in a reaction mixture containing factor Va

(20 nM) and RGDS (1 mM) in HT. The reaction was initiated by factor Xa (5 nM). At timed intervals, aliquots were removed and diluted three-fold in quench buffer (HBS containing 50 mM EDTA and 0.1% PEG-8000). The Michaelis-Menten kinetic constants

(Km, Vmax, kcat) were estimated from the average of two individual experiments by nonlinear, least squares fit to Eq. 1.

푉푚푎푥⁡푆0 푣0 = ………………………………………………………………Eq. 1 퐾푚⁡+⁡푆0

where vo is the initial rate, So is the initial substrate concentration, and the Km is the substrate concentration at half maximum velocity, Vmax. Vmax is the maximum rate of the reaction and can also be expressed as Eq. 2.

푉푚푎푥 = 푘푐푎푡[퐸]……………………………………………………………. Eq. 2 where kcat is the turnover rate of the enzyme and [E] is the concentration of enzyme.

Progress curves of complete prethrombin-1 activation by human Prothrombinase

The effect of in situ generated Frag 2 during prethrombin-1 activation on product

(thrombin) formation was monitored in vitro. Reaction mixtures contained: prethrombin-1

(1.4 – 40 µM), DAPA (30 or 50 µM), PCPS vesicles (20 µM), and factor Va (20 nM) in

HBS/Ca2+/PEG. Parallel platelet reaction mixtures contained prethrombin-1 (1.4 µM), thrombin-activated platelets (1 x 108/mL) and RGDS (1 mM) in HT. Each reaction was preincubated (5 min, 25oC) and initiated with factor Xa: 0.5 nM or 5 nM for PCPS and platelet experiments, respectively. Timed (0 - 7 h) aliquots were removed, and diluted in

101

quench buffer or SPB. Reaction products were monitored by chromogenic assays or SDS-

PAGE as described below.

The effect of fragment-2 on prethrombin-1 activation by human prothrombinase

Prethrombin-1 (10 µM) was activated to completion (forming Frag 2 and thrombin) by equivalent amounts of PCPS-bound or platelet-bound prothrombinase. Increasing concentration of Frag 2 (containing thrombin) (0-2 µM) were then titrated into prothrombinase assays containing 1.4 µM prethrombin-1 or prothrombin on PCPS or activated platelet surfaces as described above. To determine the initial rates of thrombin generation, aliquots were removed at timed intervals (0 – 5 min) and the reaction stopped by dilution in quench buffer. Experiments to control for the effect of thrombin in the reactions were run in parallel.

Spectrozyme TH hydrolysis for determination of thrombin concentrations

Thrombin activity in the quenched reaction samples was quantified by monitoring the hydrolysis of the chromogenic substrate, Spectrozyme TH. The rate of change in absorbance at 405 nm was monitored (10 min, 25oC) following the addition of 80 µL

Spectrozyme-TH (400 mM) to 20 µL reaction aliquots using a SpectroMax250 microtiter plate reader (Molecular Devices, Sunnyvale, CA). The concentration and rates of thrombin generation were determined from thrombin standard curves assayed concurrently.

Electrophoretic analyses of protein samples

Samples from prothrombin or Prethrombin-1 reactions in SPB containing 1% βME were boiled (2 min) and resolved by SDS-PAGE using 4-12% gradient, 10% or 12% polyacrylamide gels (Invitrogen) as described by Laemmli (30). Resolved proteins were 102

stained with Coomassie brilliant blue R-250. Alternatively, proteins were transferred onto nitrocellulose membranes, probed initially with burro anti-prethrombin-1 (7.5 µg/mL) and developed with HRP-conjugated goat anti-horse IgG (0.2 µg/mL) for detection of the bands by addition of Western Lightning chemiluminescence reagent (Perkin-Elmer, Waltham,

MA) and subsequent X-ray film development.

Data analyses

Data obtained from kinetic studies were fit to linear regression plots and the Michaelis-

Menten equation using GraphPad Prism v6.0 software to determine initial rates, and kinetic constants (Km and Vmax), respectively. Experimentally observed kinetic constants were fit to the integrated Michaelis-Menten derivatization to obtain simulated substrate depletion or progress curves using Wolfram Mathematica 10 software (Champaign, IL).

THEORY

The Michaelis-Menten equation (Eq. 1) was integrated in order to express instantaneous substrate concentration (S) as a function of time (t) as shown below:

푆 −푉 푡 푆 0 푚푎푥 0 퐾 S = 퐾푚W [ 푒 푚 ]…………………………………………………...Eq. 3 Ref (31) 퐾푚 where W is the Lambert function (product log) as previously described (32), So is the initial substrate concentration and t is the time. Experimentally determined Km and Vmax from prothrombinase-mediated prethrombin-1 activation were used with Eq. 3 to obtain a simulated/theoretical substrate depletion or product formation curves.

103

In order to determine the presence of product inhibition in our study, the following equation was utilized:

푣 [푃] = 0 (1 − 푒−휂푡)…………………………………………………………Eq. 4 Ref (33) 휂 where [P] is the concentration of product as a function of the initial rate (vo), time (t), and a function of initial substrate concentration (So) that represents the deviation from linearity in the steady-state enzyme activity resulting from product inhibition or substrate depletion

(η), whereby a positive correlation between η and So is indicative of an enzyme reaction where product inhibition dominates (33).

104

RESULTS

Activation of prethrombin-1 formed in situ by platelet-bound Prothrombinase

Previously, our lab demonstrated that in the absence of the thrombin inhibitor DAPA,

Prothrombinase assembled on collagen-activated platelets effectively converted prethrombin-1 formed during prothrombin activation to thrombin (3). With cofactor activity provided solely by platelet-derived factor Va and under similar conditions to those previously described by Wood et al. (3) platelets were activated by either thrombin or protease-activated receptor peptides (PAR1 and PAR4) as shown in Fig. 2. Our data demonstrate that prothrombin activation proceeds through the prethrombin-2 intermediate pathway consistent with an initial cleavage at Arg271 and independent of the platelet agonist used. Furthermore, our results show that the in situ generated prethrombin-1 by thrombin cleavage of prothrombin at Arg155 was converted into thrombin after 30 min.

Activation of exogenous prethrombin-1 by PCPS-bound Prothrombinase

The cleavage pattern of prethrombin-1 activation and the intermediate species by

Prothrombinase bound to PCPS vesicles were monitored as shown in Fig. 3. As expected, initial cleavage of prethrombin-1 occurs at Arg271 (Fig. 1) generating Frag 2 and the prethrombin-2 intermediate via the prethrombin-2 pathway. The appearance of prethrombin-2 is closely followed by the B-chain of thrombin consistent with a second cleavage of this species at Arg320 by Prothrombinase. No meizothrombin des Frag 1 (initial cleavage at Arg320) was detected as inferred by the lack of Frag 2.A on the gel (Fig. 3).

While thrombin was generated during the time course, a significant amount of the prethrombin-2 intermediate persisted after 120 min of incubation with PCPS-bound

105

Prothrombinase implying a slow or inhibited cleavage at Arg320. Our results demonstrate that Frag 2 generated in situ may be interfering with prethrombin-2 turnover on PCPS vesicles.

Kinetics constants of prethrombin-1 activation by human Prothrombinase

While we initially hypothesized that platelet-bound Prothrombinase possessed a higher affinity for prethrombin-1 compared to the PCPS-bound enzyme where prethrombin-1 is considered a poorer substrate, a Michaelis-Menten kinetic analysis suggested that an alternative mechanism was responsible for this phenomenon. The kinetic constants obtained from prethrombin-1 activation by Prothrombinase assembled on either surface by initial velocity studies are shown in Fig. 4 and Table I. The affinity (Km) of prethrombin-1 for Prothrombinase bound to platelets from two unrelated donors were similar (Km = 3.4 ±

0.49, donor 1 vs Km = 3.0 ± 0.23, donor 2). Surprisingly, the Km of prethrombin-1 for

Prothrombinase assembled on PCPS vesicles (Km = 3.7 ± 0.14, n = 2) was equivalent to that observed for the enzyme complex assembled on the activated platelet membrane

(average Km = 3.2 ± 0.28, n = 2). Likewise, the maximum rate of thrombin generation

(Vmax) is not affected by the surface on which Prothrombinase is assembled. These data are inconsistent with the effective turnover and preferred activation of the prethrombin-1 substrate by platelet-bound Prothrombinase (3). However, we observed that at a high concentration of prethrombin-1 (20 µM), the rate of thrombin generation was 3-fold slower on PCPS-bound vs. platelet-bound Prothrombinase suggesting that Frag 2, Prethrombin-2 or both generated in situ may inhibit Prothrombinase bound to PCPS vesicles (Fig. 4) (34).

106

Exogenous fragment-2 inhibition of thrombin generation from prethrombin-1 by

PCPS- and platelet-bound prothrombinase

To verify the effect of Frag 2 on cleavage and activation of prethrombin-1, we first activated prethrombin-1 to completion generating Frag 2 and thrombin as the final reaction products. As shown in Fig. 5A, the addition of Frag 2/thrombin (2 µM) to prethrombin-1

(1.4 µM) activation mixtures completely abolished the ability of PCPS-bound prothrombinase to convert prethrombin-1 to thrombin (Table II). In contrast, under similar conditions platelet-bound prothrombinase was only inhibited by ~20%. The data in Fig. 5B show that the apparent inhibition by Frag 2 is unique to prethrombin-1 as an alternate substrate for PCPS-bound prothrombinase. Our results show that in either case, the inhibition observed is not due to the thrombin added coincident with prethrombin-2 formation in reactions by both PCPS-bound and platelet-bound prothrombinase (Table II).

Progress curves of prethrombin-1 activation by human prothrombinase

Experimentally observed prethrombin-1 activation progress curves were compared to theoretical curves (from Eq. 3) corresponding to the best fit kinetic parameters (Km = 3.7

-1 vs. 3.2 μM and Vmax = 4.5 vs. 5.5 nMs ; PCPS vs platelets, respectively). We observed that within the first 5 min, thrombin generation from prethrombin-1 (1.4 µM) proceeded with similar initial rates between the theoretical and the experimental conditions with both

PCPS-bound (71.3 ± 0.5 vs 74.3 ± 2.1 nM/min; n = 3) and platelet-bound prothrombinase

(98.4 ± 0.8 vs 112.0 ± 7.0 nM/min; n = 3) respectively (Fig. 6). However, beyond 5 min, there was marked deviation in the amount of thrombin generated in the experimental vs. theoretical progress curves in reactions for prethrombin-1 activation by PCPS-bound

107

prothrombinase (Fig. 6A). In contrast, the experimentally-observed amounts of thrombin formed from prethrombin-1 in reactions by platelet-bound prothrombinase were recapitulated by the theoretical progress curves (Fig. 6B). Following 60 mins of incubation with prethrombin-1 (1.4 µM) approximately 0.3 μM vs. to 0.06 μM substrate remained in the PCPS vesicle vs platelet activation mixtures, respectively.

The data obtained from PCPS-bound Prothrombinase activation of prethrombin-1 at high substrate concentrations are shown in Fig. 7 and Table III. The results show an approximate seven-fold increase (105.1 vs. 726.6 nM/s) in the rate of thrombin generation when substrate concentration was increased eight-fold from 5 µM to 40 µM (Table III). As previously described in Eq. 4, values of η > τ-1, where τ is the time range of data acquisition

(τ = 420 min, τ-1 = 0.002 min-1), and η = deviation from linearity in enzyme kinetics suggest that product inhibition or substrate depletion is responsible for the non-linearity in product formation time courses. Theoretically, a secondary plot of η as a function of So

[Prethrombin-1] where values of η increase with increasing substrate concentration indicates an enzyme reaction where product inhibition dominates. Conversely, when values of η decrease as increased substrate concentration then substrate depletion dominates the enzyme reaction (33). Our data show a deviation from linearity and incomplete activation to thrombin across all the prethrombin-1 concentrations (5, 10, 20 and 40 µM) assayed in reactions by PCPS-bound Prothrombinase (Fig. 7). While almost 75% of the prethrombin-

1 substrate at 5 and 10 µM was converted to thrombin, only 50% of the possible thrombin was formed at higher substrate concentrations (20 and 40 µM prethrombin-1, Fig. 7).

Therefore, at 5 and 10 µM prethrombin-1, the values of η were indistinguishable (0.029 vs.

108

0.028) when the data was fit to Eq. 4. However, a slight increase in the η-value was observed at higher concentrations (20 vs 40 µM, 0.035 vs. 0.034) respectively; Table III).

The data show that as η increases with substrate concentration, it may be surmised that product inhibition is contributing to the non-linearity in prethrombin-1 time courses catalyzed by PCPS-bound Prothrombinase most likely due to the in situ generation of Frag

2 (Fig. 3 and Fig. 5).

109

DISCUSSION

In an effort to understand Prothrombinase function and the molecular mechanism of thrombin generation, several determinants of prothrombin activation have been identified or postulated. The majority of these were determined in studies using model system with phospholipid vesicles of defined content (including PCPS) (6,35-39). Membrane binding is a critical component for Prothrombinase assembly and function in view of the 1,000- fold reduction in rates of thrombin generation in the absence of an appropriate membrane; however, PCPS vesicles do not fully recapitulate the physiologically relevant surface of the activated platelet membrane (12-15,39). Similarly, studies have shown that the pathway of prothrombin activation is dictated by whether Prothrombinase is bound to the activated platelet membrane (2,3) or PCPS vesicles (5,6). To date the biochemical significance of prothrombin’s interaction with the membrane is largely derived from evidence that the Gla- domain within fragment-1 anchors the substrate onto the membrane surface allowing for cleavages at the two distant peptide bonds: Arg271-Thr272 and Arg320-Ile321 on the C- terminal half of the prothrombin molecule to form the active protease, thrombin, and notably this strong membrane interaction is thought to drive initial cleavage at Arg320 via the alternative meizothrombin activation pathway (6). Loss of fragment-1 following cleavage of prothrombin at Arg155 by thrombin produces prethrombin-1, a Gla-domainless intermediate which has a ~10-fold reduced rate in thrombin generation compared to prothrombin (9-11).

With the Gla-domainless prethrombin-1 species as an alternative substrate for

Prothrombinase, we demonstrate in the present study that despite being considered a poorer

110

substrate for PCPS-bound Prothrombinase, it is a viable source for thrombin generation in reactions catalyzed by platelet-bound Prothrombinase. Furthermore, we have deduced from our kinetic analyses of prethrombin-1 activation that: 1) initial cleavage of prethrombin-1 at Arg271 by human Prothrombinase is unaltered by the membrane surface,

PCPS vesicles vs. activated platelets; 2) at higher prethrombin-1 concentrations, the rate of thrombin generation is significantly reduced in reactions by PCPS-bound compared to platelet-bound Prothrombinase; and 3) cleavage at Arg320 of prethrombin-2 formed in situ during prethrombin-1 activation on PCPS vesicles is inhibited by the accumulation of Frag

2, a product of the reaction. These findings are the first to demonstrate that defective membrane binding is not the only reason but rather only partly contributes to the reduced utilization of prethrombin-1 as an excellent source of thrombin generation in reactions by

PCPS-bound Prothrombinase thereby adding to the body of evidence that activated platelets are unique in regards to maximizing the generation of thrombin and promoting an appropriate hemostatic response.

Several lines of evidence from our laboratory and others suggest that human

Prothrombinase is differentially assembled on model phospholipid systems (5,6) compared to the activated platelet membrane (2,3). Wood et al. first made the observation that prethrombin-1 formed in situ during prothrombin activation in the absence of thrombin inhibitors is effectively converted into thrombin by Prothrombinase bound to collagen- activated platelets but not PCPS vesicles (3). Additional data from the present study corroborate those results, and we further demonstrated that the agonist (i.e. thrombin,

PAR1, PAR4 or collagen) used in stimulating platelets does not alter Prothrombinase

111

function during prethrombin-1 activation suggesting that activated platelets maximize their procoagulant potential following an injury. Based on the pattern of effective consumption of the in situ generated prethrombin-1 we anticipated a higher affinity of prethrombin-1 for platelet-bound Prothrombinase compared to the PCPS-bound enzyme. While the affinity

(apparent Km = ~4 µM) of prethrombin-1 activation on PCPS vesicles was consistent with previous studies (34,40), a similar enzyme affinity was obtained on the activated platelet membrane. These findings are inconsistent with the rapid turnover of in situ generated prethrombin-1 during prothrombin activation on the activated platelet (3) and support the conclusion that other features beyond enzyme affinity may account for the differential activation of prethrombin-1.

In contrast to the affinity data, initial rates of thrombin generation from prethrombin-1 were 3-fold lower at high prethrombin-1 concentrations catalyzed by

Prothrombinase assembled on PCPS vesicles compared to when the enzyme complex is assembled on the activated platelet membrane. These results demonstrate a preference for prethrombin-1 by the platelet-bound enzyme and suggest that some form of product inhibition may account for the reduced utilization of this substrate on PCPS vesicles. To understand the possible form of inhibition, it is important to reiterate that during prethrombin-1 activation in situ the prethrombin-2 and Fragment 2 intermediates are formed. In the present study, we have unequivocally demonstrated that prothrombin

Fragment 2 can abolish Prothrombinase activity when the enzyme complex is assembled on PCPS vesicles but not the activated platelet membrane.

112

Furthermore, results from prethrombin-1 progress curves show a significant deviation of the experimental from the predicted when Prothrombinase was assembled on

PCPS vesicles compared to its assembly on the activated platelet surface. In the context of our other finding, we suggest that this deviation is due to the reduced activation of the zymogens prethrombin-1 or prethrombin-2 intermediate to thrombin. Indeed, densitometric analyses indicated the presence of equivalent amounts of both thrombin and prethrombin-2 in prethrombin-1 activation reactions catalyzed by PCPS-bound

Prothrombinase consistent with previous reports (9,41). The results indicate that the fragment-2 formed in situ during the reaction likely inhibits PCPS-bound Prothrombinase to a greater extent compared to the platelet-bound enzyme. Furthermore, analyses of the η- values from increased prethrombin-1 concentration in nonlinear enzyme kinetic time courses support a trend whereby η increases with increased substrate concentration consistent with product inhibition of PCPS-bound Prothrombinase by fragment-2. In conclusion, we speculate that reduced thrombin generation by PCPS-bound

Prothrombinase activation of prethrombin-1 is the result of impaired cleavage at Arg320 of the prethrombin-2 intermediate due to inhibition by fragment-2.

From our present studies, we speculate that fragment-2 may interfere with efficient docking of prethrombin-1 to the PCPS-bound Prothrombinase complex. We hypothesize that at a high local concentration, the anionic fragment-2 may bind to PCPS-bound

Prothrombinase and effectively inhibit the binding of prethrombin-2 to the enzyme complex. Platelet-bound Prothrombinase effectively avoids such feedback interactions due to its unique structural properties (12-15,39). Alternatively, the recent structure of

113

prethrombin-1 also gives potential structural insight as to the underlying mechanism for the observations made in the present study. It shows that the Arg271 cleavage site is ~ 27

Å away from the electronegative region of Fragment 2 (Glu254) and ~35 Å away from the second cleavage site (Arg320), which is fully exposed to solvent for proteolytic attack (42).

Therefore, significant conformational changes are required following prethrombin-1 binding by Prothrombinase to effect cleavages at these two peptide bonds. Likewise, prethrombin-1 exhibits conformational plasticity similar to the full length prothrombin by transitioning between zymogen-like and proteinase-like states (42). As reported here, initial cleavage of prethrombin-1 proceeds via Arg271, which is not preferred by PCPS- bound Prothrombinase (7,9). Therefore, it is possible that the release of fragment-2 may further constrain this undesirable orientation of the prethrombin-2 intermediate on PCPS vesicles further inhibiting its turnover and supporting the notion that conformational plasticity may dictate the fate of each enzymatic step (42). Collectively, our data allow for a model whereby fragment-2 interferes to a greater extent with the binding or recognition of the prethrombin-2 intermediate during prethrombin-1 activation by PCPS-bound

Prothrombinase to a significant extent compared to the platelet-bound Prothrombinase implicating distinct binding interfaces between enzymes assembled on either membrane surface. Therefore, Prothrombinase associated with the physiologically relevant surface of the activated platelet membrane, maximizes prothrombin activation by effectively utilizing intermediate species such as prethrombin-1 and prethrombin-2 in thrombin generation at the site of injury during the hemostatic response.

114

ACKNOWLEDGEMENTS

This work was supported by National Institute of Health (grant P01HL46703, Project 3 to

P.B.T) and University of Vermont College of Medicine. F.A. was supported by the

Department of Biochemistry, UVM and the Cellular, Molecular and Biomedical Science

Program, College of Medicine, UVM. L.M.H was supported by NIH grant T32 HL007594,

Hemostasis and Thrombosis Program for Academic Trainees (to K.G. Mann). We thank

Dr. Christopher L. Berger of UMV for his intellectual contributions to designing kinetic experiments and for his assistance in the analyses of data. The authors would like to acknowledge Dr. Beth Bouchard for critical review and Dr. Jay Silveira for helpful suggestions on experimental design.

AUTHORSHIP

Contribution: F.A. designed and performed experiments, analyzed results, and wrote the manuscript; L.M.H designed and performed experiments, analyzed results and wrote the manuscript. P.B.T. designed the research and wrote the manuscript.

Conflict of interest disclosure: The authors declare no competing financial interest.

Correspondence: Paula B. Tracy, 89 Beaumont Ave, Given Medical Research Building

C409, Burlington, VT 05405; Telephone: 802-656-1995; Fax: 802-656-8220; and Email: [email protected].

115

REFERENCES

1. Mann, K. G., Nesheim, M. E., Church, W. R., Haley, P., and Krishnaswamy, S. (1990) Surface-dependent reactions of the vitamin K-dependent enzyme complexes. Blood 76, 1-16

2. Haynes, L. M., Bouchard, B. A., Tracy, P. B., and Mann, K. G. (2012) Prothrombin Activation by Platelet-associated Prothrombinase Proceeds through the Prethrombin-2 Pathway via a Concerted Mechanism. Journal of Biological Chemistry 287, 38647-38655

3. Wood, J. P., Silveira, J. R., Maille, N. M., Haynes, L. M., and Tracy, P. B. (2011) Prothrombin activation on the activated platelet surface optimizes expression of procoagulant activity. Blood 117, 1710-1718

4. Krishnaswamy, S., and Betz, A. (1997) Exosites determine macromolecular substrate recognition by prothrombinase. Biochemistry 36, 12080-12086

5. Krishnaswamy, S., Church, W. R., Nesheim, M. E., and Mann, K. G. (1987) Activation of human prothrombin by human prothrombinase. Influence of factor Va on the reaction mechanism. J Biol Chem 262, 3291-3299

6. Krishnaswamy, S., Mann, K. G., and Nesheim, M. E. (1986) The prothrombinase- catalyzed activation of prothrombin proceeds through the intermediate meizothrombin in an ordered, sequential reaction. J Biol Chem 261, 8977-8984

7. Heldebrant, C. M., Butkowski, R. J., Bajaj, S. P., and Mann, K. G. (1973) The activation of prothrombin. II. Partial reactions, physical and chemical characterization of the intermediates of activation. J Biol Chem 248, 7149-7163

8. Heldebrant, C. M., and Mann, K. G. (1973) The activation of prothrombin. I. Isolation and preliminary characterization of intermediates. J Biol Chem 248, 3642- 3652

9. Esmon, C. T., and Jackson, C. M. (1974) The conversion of prothrombin to thrombin. III. The factor Xa-catalyzed activation of prothrombin. J Biol Chem 249, 7782-7790

10. Downing, M. R., Butkowski, R. J., Clark, M. M., and Mann, K. G. (1975) Human prothrombin activation. J Biol Chem 250, 8897-8906

11. Malhotra, O. P., Nesheim, M. E., and Mann, K. G. (1985) The kinetics of activation of normal and gamma-carboxyglutamic acid-deficient prothrombins. J Biol Chem 260, 279-287 116

12. Bouchard, B. A., Catcher, C. S., Thrash, B. R., Adida, C., and Tracy, P. B. (1997) Effector cell protease receptor-1, a platelet activation-dependent membrane protein, regulates prothrombinase-catalyzed thrombin generation. J Biol Chem 272, 9244- 9251

13. Fager, A. M., Wood, J. P., Bouchard, B. A., Feng, P., and Tracy, P. B. (2010) Properties of procoagulant platelets: defining and characterizing the subpopulation binding a functional prothrombinase. Arterioscler Thromb Vasc Biol 30, 2400-2407

14. Tracy, P. B., Eide, L. L., and Mann, K. G. (1985) Human prothrombinase complex assembly and function on isolated peripheral blood cell populations. J Biol Chem 260, 2119-2124

15. Krishnaswamy, S., Jones, K. C., and Mann, K. G. (1988) Prothrombinase complex assembly. Kinetic mechanism of enzyme assembly on phospholipid vesicles. J Biol Chem 263, 3823-3834

16. Deguchi, H., Takeya, H., Gabazza, E. C., Nishioka, J., and Suzuki, K. (1997) Prothrombin kringle 1 domain interacts with factor Va during the assembly of prothrombinase complex. The Biochemical journal 321 ( Pt 3), 729-735

17. Church, W. R., Ouellette, L. A., and Messier, T. L. (1991) Modulation of human prothrombin activation on phospholipid vesicles and platelets using monoclonal antibodies to prothrombin fragment 2. J Biol Chem 266, 8384-8391

18. Taneda, H., Andoh, K., Nishioka, J., Takeya, H., and Suzuki, K. (1994) Blood coagulation factor Xa interacts with a linear sequence of the kringle 2 domain of prothrombin. Journal of biochemistry 116, 589-597

19. Anderson, P. J., Nesset, A., and Bock, P. E. (2003) Effects of activation peptide bond cleavage and fragment 2 interactions on the pathway of exosite I expression during activation of human prethrombin 1 to thrombin. J Biol Chem 278, 44482- 44488

20. Higgins, D. L., and Mann, K. G. (1983) The interaction of bovine factor V and factor V-derived peptides with phospholipid vesicles. J Biol Chem 258, 6503-6508

21. McDuffie, F. C., Giffin, C., Niedringhaus, R., Mann, K. G., Owen, C. A., Jr., Bowie, E. J., Peterson, J., Clark, G., and Hunder, G. G. (1979) Prothrombin, thrombin and prothrombin fragments in plasma of normal individuals and of patients with laboratory evidence of disseminated intravascular coagulation. Thromb Res 16, 759-773

22. Mann, K. G. (1976) Prothrombin. Methods Enzymol 45, 123-156

117

23. Nesheim, M. E., Katzmann, J. A., Tracy, P. B., and Mann, K. G. (1981) Factor V. Methods Enzymol 80 Pt C, 249-274

24. Kisiel, W., and Hanahan, D. J. (1973) Purification and characterization of human Factor II. Biochimica et biophysica acta 304, 103-113

25. Fenton, J. W., Fasco, M. J., and Stackrow, A. B. (1977) Human thrombins. Production, evaluation, and properties of alpha-thrombin. The Journal of biological chemistry 252, 3587-3598

26. Di Scipio, R. G., Hermodson, M. A., Yates, S. G., and Davie, E. W. (1977) A comparison of human prothrombin, factor IX (Christmas factor), factor X (Stuart factor), and protein S. Biochemistry 16, 698-706

27. Nesheim, M. E., Myrmel, K. H., Hibbard, L., and Mann, K. G. (1979) Isolation and characterization of single chain bovine factor V. J Biol Chem 254, 508-517

28. Mustard, J. F., Perry, D. W., Ardlie, N. G., and Packham, M. A. (1972) Preparation of suspensions of washed platelets from humans. Br J Haematol 22, 193-204

29. Camire, R. M., Kalafatis, M., Simioni, P., Girolami, A., and Tracy, P. B. (1998) Platelet-derived factor Va/Va Leiden cofactor activities are sustained on the surface of activated platelets despite the presence of activated protein C. Blood 91, 2818- 2829

30. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685

31. Goudar, C. T., Harris, S. K., McInerney, M. J., and Suflita, J. M. (2004) Progress curve analysis for enzyme and microbial kinetic reactions using explicit solutions based on the Lambert W function. Journal of microbiological methods 59, 317-326

32. Corless, R. M., Gonnet, G. H., Hare, D. E. G., Jeffrey, D. J., and Knuth, D. E. (1996) On the Lambert W function. Adv Comput Math 5, 329-359

33. Cao, W., and De La Cruz, E. M. (2013) Quantitative full time course analysis of nonlinear enzyme cycling kinetics. Scientific reports 3, 2658

34. Ayombil, F., Silveira, J. R., Wood, J. P., and Tracy, P. B. (2011) Prethrombin-1, the gla-domainless prothrombin intermediate, is activated efficiently to thrombin by prothrombinase assembled on the activated platelet surface. Journal of thrombosis and haemostasis : JTH 9, P-TU-139

118

35. Rosing, J., Tans, G., Govers-Riemslag, J. W., Zwaal, R. F., and Hemker, H. C. (1980) The role of phospholipids and factor Va in the prothrombinase complex. J Biol Chem 255, 274-283

36. Weinreb, G. E., Mukhopadhyay, K., Majumder, R., and Lentz, B. R. (2003) Cooperative roles of factor V(a) and phosphatidylserine-containing membranes as cofactors in prothrombin activation. J Biol Chem 278, 5679-5684

37. Wu, J. R., and Lentz, B. R. (1994) Phospholipid-specific conformational changes in human prothrombin upon binding to procoagulant acidic lipid membranes. Thromb Haemost 71, 596-604

38. Wu, J. R., Zhou, C., Majumder, R., Powers, D. D., Weinreb, G., and Lentz, B. R. (2002) Role of procoagulant lipids in human prothrombin activation. 1. Prothrombin activation by factor X(a) in the absence of factor V(a) and in the absence and presence of membranes. Biochemistry 41, 935-949

39. Nesheim, M. E., Taswell, J. B., and Mann, K. G. (1979) The contribution of bovine Factor V and Factor Va to the activity of prothrombinase. J Biol Chem 254, 10952- 10962

40. Kim, P. Y., Manuel, R., and Nesheim, M. E. (2009) Differences in prethrombin-1 activation with human or bovine factor Va can be attributed to the heavy chain. Thromb Haemost 102, 623-633

41. Owen, W. G., Esmon, C. T., and Jackson, C. M. (1974) The conversion of prothrombin to thrombin. I. Characterization of the reaction products formed during the activation of bovine prothrombin. J Biol Chem 249, 594-605

42. Chen, Z., Pelc, L. A., and Di Cera, E. (2010) Crystal structure of prethrombin-1. Proc Natl Acad Sci U S A 107, 19278-19283

119

TABLES

Table I. Kinetic parameters for prethrombin-1 activation on membrane surfaces

PCPS vesicles Activated platelets -1 -1 Trial Km, µM Vmax, nMs Donor Km, µM Vmax, nMs 1 3.8 4.2 1 3.4 5.4 2 3.6 4.7 2 3.0 5.6 Mean ± SD 3.7 ± 0.14 4.5 ± 0.35 Mean ± SD 3.2 ± 0.28 5.5 ± 0.14

The kinetic constants (Km and Vmax) were measured from experimental procedures described in Materials and Method by fitting data to the Michaelis-Menten equation using GraphPad software v6. On either membrane surface (PCPS vesicles or activated platelets), initial velocities were limited to within 10% substrate consumption. The kinetic parameters are reported as the mean ± SD, n = 2.

120

Table II. Normalized rates of thrombin generation from prethrombin-1 or prothrombin activations catalyzed by Prothrombinase bound to PCPS vesicles or thrombin-activated platelets in the presence or absence of Frag 2 and/or thrombin

Membrane-bound Relative rate of thrombin generation Prothrombinase 1.4 µM Pre-1 1.4 µM Pre-1 + 2 µM Frag 2 1.4 µM Pre-1 + 2 µM IIa PCPS 1.0 ± 0.02 0.00 ± 0.00 1.00 ± 0.16 Platelet 1.0 ± 0.06 0.77 ± 0.05 1.08 ± 0.01

1.4 µM II 1.4 µM II + 2 µM Frag 2 1.4 µM II + 2 µM IIa PCPS 1.0 ± 0.01 1.11 ± 0.13 1.10 ± 0.09 Platelet 1.0 ± 0.02 1.20 ± 0.10 0.96 ± 0.19

*The rates of thrombin generation were estimated from experimental procedures described in Materials and Methods. The Prothrombinase reactions were saturated with factor Va (20 nM). factor Xa was the limiting reagent. In these reactions the rates obtained were normalized to equivalent concentrations of Prothrombinase/factor Xa (0.5 nM) assembled on each membrane surface. The values shown (mean ± SD) are representative of two or more independent experimental measurements. Prothrombin, prethrombin-1, fragment 2 and thrombin are abbreviated as II, Pre-1, Frag 2, and IIa, respectively.

121

Table III. Parameters of non-linear enzyme kinetic time courses for prethrombin-1 activation by PCPS-bound Prothrombinase.

Prethrombin-1, -1 vo (nM/s) η (min ) [S]o (µM) 5 105.1 0.029 10 207.4 0.028 20 477.9 0.035 40 726.6 0.034 The amount of thrombin formed was estimated from experimental procedures described in Materials and Methods. The initial rate (vo) and η were estimated by fitting data to Eq. 4 using the Mathematica software as previously described (33).

122

FIGURES

Figure 1. Schematic of prethrombin-1 formation and activation. α-Thrombin is generated from prothrombin by ordered cleavages at Arg271 (initial cleavage at R271 - prethrombin-2 pathway) and Arg320 (initial cleavage at R320 - meizothrombin pathway) via membrane-bound Prothrombinase. Feedback proteolysis of prothrombin at Arg155 (R155) by the formed α-thrombin generates fragment-1 (Frag 1) which contains the membrane binding Gla-domain, and a Gla-domainless prethrombin-1 (Pre-1) intermediate which contains the serine protease domain (A- and B-chains). Cleavage of prethrombin-1 at Arg271 (R271) by Prothrombinase generates fragment-2 (Frag 2) and prethrombin-2 (Pre- 2) intermediate via the prethrombin-2 pathway. Subsequent cleavage at Arg320 (R320) of the prethrombin-2 intermediate generates α-thrombin.

123

Figure 2. Prothrombin activation and in situ prethrombin-1 turnover by platelet- Prothrombinase. Washed human platelets (3 x 108 /mL) were activated with A) thrombin (2 U/mL), B) PAR1 (100 µM) or C) PAR4 (500 µM) for 10 min. In reactions with thrombin-activated platelets, 5 nM factor Xa was used to initiate the reactions. For reactions using PAR-activated platelets, 10 nM factor Xa was used. Aliquots of the reaction mixture were removed over time, quenched in sample preparation buffer and the proteins resolved via SDS-PAGE. Proteins were visualized by Western blotting using mAbs against prethrombin-1. Time points are shown on the top of each panel. Molecular weight markers (x103 Da) are shown on the left side of each panel. Prothrombin (II), prethrombin-1 (Pre-1), prethrombin-2 (Pre- 2), fragment 1.2 (Frag 1.2) and the B-chain of thrombin are shown on the right.

124

Figure 3. Prethrombin-1 activation by Prothrombinase assembled on PCPS vesicles. Prethrombin-1 (1.4 µM) was activated by PCPS-bound Prothrombinase (20 µM PCPS, 20 nM factor Va, 0.5 nM factor Xa) in HBS/Ca2+/PEG containing 30 µM DAPA (120 min, 25 oC). Aliquots of the reaction mixture were removed at timed intervals, quenched in sample preparation buffer and the proteins resolved via SDS-PAGE. Proteins were visualized by staining with Coomassie brilliant blue R-250. Molecular weight markers (x103 Da) are shown on the left side of each panel. The substrate, prethrombin-1 (Pre-1), reaction intermediate, prethrombin-2 (Pre-2) and products, B-chain of thrombin (B) and fragment- 2 (Frag 2) are shown on the right. The Frag 2.A intermediate is not observed indicating that initial cleavage at Arg320 of Pre-1 was avoided.

125

Figure 4. Kinetics of prethrombin-1 activation on PCPS vesicles and platelets. Initial rates of α-thrombin generation from prothrombin-1 (0.5-20 µM) by human Prothrombinase assembled on either A) PCPS vesicles (●) or B) thrombin-activated platelets (○) were measured by a two-stage chromogenic substrate assay. Kinetic constants (Km and Vmax) were determined by fitting each data set to the Michaelis-Menten equation.

126

Figure 5. Relative rates of thrombin generation from prethrombin-1 or prothrombin activation mixtures containing exogenous fragment-2. Prethrombin-1 (Pre-1, 10 µM) was activated to completion by membrane-bound Prothrombinase producing equivalent amounts of the products fragment-2 (Frag 2) and thrombin (IIa). The initial rate of thrombin generation from 1.4 µM (A) Prethrombin-1 (Pre-1) or (B) prothrombin (II) by either PCPS-bound Prothrombinase (■) or platelet- bound Prothrombinase (■) were monitored in reaction mixtures containing 2 µM of the in situ generated Frag 2 plus IIa or only IIa. The data are presented as a bar graph with the relative rates normalized to the appropriate initial rate of thrombin generation in the absence of either Frag 2 and/or IIa. The asterisk symbol (*) indicates that no thrombin was formed from Pre-1 on the PCPS vesicles in the presence of exogenously added Frag 2.

127

128

1 4 0 0

1 2 0 0

) M

n 1 0 0 0

(

i 8 0 0

b m

o 6 0 0

r h

T 4 0 0

2 0 0

0 0 1 0 2 0 3 0 4 0 5 0 6 0 T im e (m in )

1 4 0 0

1 2 0 0

) M

n 1 0 0 0

(

i 8 0 0

b m

o 6 0 0

r h

T 4 0 0

2 0 0

0 0 1 0 2 0 3 0 4 0 5 0 6 0 T im e (m in )

Figure 6. Progress curves of prethrombin-1 activation by PCPS- or platelet-bound Prothrombinase: empirical vs theoretical Reaction conditions were as described in Materials and Methods. Timed aliquots of 1.4µM prethrombin-1 activation by 0.5nM Prothrombinase assembled on either (A) PCPS vesicles (20µM) or (B) thrombin-activated platelets (1x108/mL) were removed and quenched in quench buffer. The amount of thrombin generated was followed continuously in a progress curve for PCPS-bound Prothrombinase (▲) vs the predicted curve (∆) based on pre- determined kinetic constants (fit to Eq. 3 in the Materials and Methods). Similar plots are shown for platelet-bound Prothrombinase of experimentally observed (■) vs predicted (□) progress curves. The data are represented as mean ± (SD or SEM) for three separate experiments, which required three different platelet donors.

129

Figure 7. Progress curves of prethrombin-1 activation by PCPS-bound Prothrombinase. Reaction conditions are similar to those described in Fig. 6. The thrombin formed via PCPS-bound Prothrombinase was monitored over 420 min (τ) at the following prethrombin-1 concentrations 5 µM (∆), 10 µM (▲), 20 µM (○) and 40 µM (●).

130

CHAPTER FOUR: Functional characterization of the human platelet-derived, glycophosphatidylinosital (GPI)-anchored pool of factor Va

Francis Ayombil1 and Paula B. Tracy1,*

Short title: Characterization of the GPI-anchored pool of platelet-derived factor Va Key words: factor V/Va, platelets, prothrombinase, activated protein C and plasmin Scientific category: Coagulation proteins

Abbreviations used in this manuscript include: activated protein C, APC; Factor Va, FVa; glycosylphosphatidylinositol, GPI; PC, phosphatidylcholine; PS, phosphatidylserine; PCPS vesicle, 75% PC: 25% PS; DAPA, dansyl-L-arginine N-(3-ethyl-1,5- pentanediyl)amine.

131

ABSTRACT

Downregulation of the procoagulant response requires proteolytic inactivation of factor

Va, the essential cofactor for the Prothrombinase-catalyzed activation of prothrombin to thrombin. The anticoagulant serine protease activated protein C (APC) and the fibrinolytic enzyme plasmin are its most effective physiological inactivators. Several studies, however, demonstrate that a fraction of the platelet-derived factor Va pool is resistant to APC- catalyzed inactivation when compared to its plasma counterpart. The mechanism responsible for this protease resistance is unknown. Subsequent to thrombin-catalyzed platelet activation, platelet-bound factor Va is defined by two subpopulations: a dissociable pool, and a non-dissociable pool that is tethered to the platelet membrane surface via a glycophosphatidylinositol (GPI) anchor on the factor Va heavy chain. Analyses of kinetics of prothrombin activation indicate that the GPI-anchored factor Va pool accounts for between 25-90% of the total platelet-derived factor Va that supports Prothrombinase assembly and function. Furthermore, GPI-anchored factor Va directs Prothrombinase assembly and function with kinetic constants comparable to those measured using the plasma-derived molecule at concentrations that saturate all remaining factor Va binding sites. Remarkably, while the GPI-anchored factor Va was completely resistant to inactivation by APC, the dissociably-bound factor Va pool was inactivated similarly to the plasma-derived molecule. Likewise, the dissociably-bound factor Va was susceptible to plasmin-catalyzed inactivation, whereas the GPI-anchored pool was completely resistant.

Thus, GPI-anchored platelet-derived factor Va plays a critical role in promoting and

132

sustaining thrombin generation at sites of vascular injury due to its covalent retention and resistance to proteolytic inactivation at the activated platelet surface.

133

INTRODUCTION

Activated platelets play several critical roles in normal hemostasis following vascular injury. Their adherence and aggregation at the injury site results in 1) the formation of a platelet plug (1,2), 2) their activation and release of α- and dense-granule constituents, many of which are involved in the procoagulant response (3-5), and 3) their provision of the catalytic surface required for the function and assembly of coagulation complexes (6,7) that will sustain the procoagulant response. Upon platelet activation, factor

Va stored within their α-granules is made immediately available to perform its essential role in the assembly and function of the final enzyme complex, Prothrombinase.

Prothrombinase is a calcium-dependent, 1:1 stoichiometric complex of the non-enzymatic cofactor, factor Va, and the serine protease, factor Xa, bound to the surface of activated platelets adhered to sites of vascular injury that converts the zymogen prothrombin to thrombin (3,8). Factor Va plays a critical role in the formation of Prothrombinase since it is absolutely required for factor Xa binding to activated platelets (9-11), and therefore profoundly influences the amount of thrombin generated. Clinically, this is underscored by observations that deficiencies in the procofactor, factor V, are associated with severe bleeding abnormalities and possibly even death (12,13), whereas inefficient inactivation of factor V and Va by activated protein C (APC), as is seen in factor VLeiden, can lead to venous thrombosis (14) (15,16).

The inactive procofactor factor V is synthesized by, and released from, the liver as a single-chain, 330 kDa protein (17-19) that is distributed between two pools. Of the total factor V found in whole blood, 75-80% circulates in plasma (20-30 nM) and 20-25% is

134

found within the α-granules of platelets (~4600-14,000 molecules/platelet) (4,20). The conversion of plasma-derived factor V to its active state (21,22), requires limited proteolysis by thrombin to yield a 105 kDa N-terminal-derived heavy chain and 75 kDa C- terminal-derived light chain (21,23) associated by Ca2+ whereas, the platelet-derived pool is stored in platelet α-granules in a partial, proteolytically-activated state (factor V/Va)

(4,20) as a result of the endocytosis of plasma-derived factor V by megakaryocytes (24-

26). Its subsequent proteolytic processing and trafficking to α-granules to produce a fraction of the cofactor pool that does need to be activated by thrombin (26-28).

Considering that at a site of vascular injury the concentration of platelet-derived factor

V/Va has been estimated to be as much as 600 times that of plasma factor V (7), any alterations in this cofactor pool will have a profound impact on hemostasis. Indeed, several studies have reported that the platelet-derived factor V/Va pool possesses physical (29-35) and functional (23,27,36,37) characteristics that impart a more procoagulant phenotype than that of its plasma counterpart, and allow for sustained thrombin generation at sites of vascular injury.

Plasma- and platelet-derived factor Va exhibit differential susceptibilities to proteolysis by various proteases including inactivation by activated protein C (APC)

(36,38), and plasmin (39,40). Earlier studies indicated that a substantial fraction of the platelet-derived factor Va pool is totally resistant to proteolytic inactivation by the anticoagulant protease APC (36,38), whereas the plasma-derived factor Va pool is rapidly and completely inactivated by APC. Likewise, the fibrinolytic enzyme, plasmin, rapidly and completely inactivates plasma-derived factor Va, whereas plasmin-catalyzed cleavage

135

of platelet-derived factor Va increases its cofactor activity approximately 3-6 fold, an activity which can be sustained for several hours despite the continued presence of plasmin

(39). These observations are in marked contrast to the rapid and complete inactivation of plasma-derived factor Va catalyzed by plasmin. What physical features make these two cofactor pools different substrates for these various proteases have yet to be determined.

Wood et al. recently identified a non-dissociable, membrane-bound pool of platelet-derived factor Va following extensive washing of thrombin-activated platelets

(10,41), Further studies revealed that the heavy chain of platelet-derived factor Va is covalently associated with the activated platelet membrane via a glycosylphosphatidyl inositol (GPI) anchor at Ser692 that resides within the heavy chain GPI anchor consensus sequence (R678KMHDRLEPEDEES692DADYDYQNRLAAALGIR108). The addition of a

GPI anchor at this site may explain the inability of platelet-derived factor Va, but not plasma-derived factor Va to be phosphorylated at that site (31,41), and which renders the plasma-derived factor Va more susceptible to APC-catalyzed inactivation. Other studies demonstrated that only a subpopulation of platelets expresses this GPI-anchored cofactor pool, which also defines the subpopulation of platelets capable of assembling a functional

Prothrombinase (10). The goal of the current study is define the functional characteristics of the GPI-anchored platelet-derived factor Va pool. Our results demonstrate that GPI- anchored platelet-derived factor Va can account for 25-90% of released/expressed cofactor that supports thrombin generation via Prothrombinase, expresses the same binding affinity for factor Xa, and is resistant to proteolytic inactivation by APC and plasmin.

136

MATERIALS AND METHODS

Materials

Bovine serum albumin was bought from MP Biomedicals (Solon, OH). The peptides Gly-

Pro-Arg-Pro (GPRP), and Arg-Gly-Asp-Ser (RGDS) were synthesized by or purchased from the Protein Core Facility, University of Vermont or ABI Scientific Inc. (Sterling,

VA), respectively. Hirudin was obtained from Calbiochem (Gibbstown, NJ)..

Dansylarginine N-(3-ethyl-1, 5-pentanediyl) amine (DAPA) was obtained from

Haematologic Technologies (Essex Junction, VT). Phospholipid vesicles comprised of

75% phosphatidylcholine (wt/wt) and 25% phosphatidylserine (wt/wt) (PCPS) were a generous gift from Dr. Kenneth Mann (University of Vermont) and prepared according to the procedure of Higgins and Mann (42) with reagents purchased from Avanti Polar Lipids

(Alabaster, AL). The chromogenic thrombin-specific substrate Spectrozyme TH was purchased from Sekisui Diagnostics (San Diego, CA). Nitrocellulose membrane (0.45 µm) was bought from Bio-Rad (Hercules, CA). The Western Lightning chemiluminescence substrate, and iodine-125 were obtained from PerkinElmer (Waltham, MA). IODO-GEN tubes were bought from ThermoScientific (Rockford, IL). Antibodies: Horseradish peroxidase conjugated to horse anti-mouse IgG was purchased from Vectors Labs

(Burlingame, CA). Monoclonal antibodies against human factor Va heavy chain (αhFVa

LC#9)(43) and light chain (αhFVa HC#17)(44) were provided by the University of

Vermont Antibody Core Facility.

137

Proteins

Human factor Xa, α-thrombin, activated protein C (APC), and plasmin were purchased from Haematologic Technologies (Essex Junction). Human prothrombin, and factor V were purified as previously described (17,45) from freshly frozen plasma obtained from consenting healthy adults, according to the guidelines of the University of Vermont’s

Institutional Review Board and Committee on Human Subjects. . Proteins were stored in

o 50% glycerol containing 5 mM CaCl2 at -20 C. Plasma-derived factor V (1.0 µM) was activated to factor Va by α-thrombin (20 nM, 37 oC). The activity of thrombin was quenched by the addition of hirudin (30 nM). Protein concentrations were determined by

1% UV absorbance at 280 nm with the following extinction coefficients (E 280nm) and molecular weights (Da): human APC, 14.5, 56200 (46); human plasmin, 17.0, 83000 (47); human prothrombin, 1.38, 72000 (48); human thrombin, 1.83, 36700 (49); human factor

Xa, 11.6, 46000 (50); and human factor V, 9.6, 330000 (51).

Isolation of human platelets followed by isolation of the dissociable pool of platelet- derived factor Va and the factor Va that is GPI-anchored to thrombin-activated platelets

Blood from nonsmoking, healthy, and consenting adults was collected by venipuncture into acid citrate dextrose (71 mM citric acid, 85 mM trisodium, 2% dextrose, pH 4.5) in a ratio of 6 parts of blood to 1 part of acid citrate dextrose. Platelets were subsequently washed and isolated according to the method of Mustard et al. (36) with some minor modifications as previously reported (52). Following isolation, platelets were resuspended in HEPES-

Tyrode buffer (137 mM NaCl, 2.68 mM KCl, 11.9 mM NaHCO3, 0.42 mM NaH2PO4, 1.0

138

mM MgCl2, 5 mM CaCl2, and 0.2% dextrose, pH 7.4; HT buffer). The platelet concentration was determined using the ADVIA 120 Hematology Analyzer (Siemens

Healthcare Diagnostics). Washed platelets were activated at 3 x 108/mL with thrombin (50 nM) in HT buffer pH 7.4 in the presence of 1 mM RGDS (5 min, 25 oC) and the thrombin inhibited by the addition of 75 nM hirudin.

Isolation of platelet-derived factor Va that is GPI-anchored to the thrombin- activated platelet required the following: Platelets were activated and immediately washed repeatedly as previously described by Wood et al. (10,53). Isolated platelets (1 x 109/mL) in HT buffer containing 5 mM CaCl2, 3.5 mg/mL BSA pH 7.4 (HTA), 3 mM RGDS and 2 mM GPRP were activated with thrombin (50 nM, 2 min, 25oC). Hirudin (75 nM) was then added to inactivate the thrombin. The activated platelet mixture was subjected to centrifugation at 660 xg for 3 min. and the supernatant containing the dissociable pool of platelet-derived factor Va was removed. The remaining platelets were resuspended gently in 1.0 mL HTA buffer containing RGDS (5 mM) and washed repeatedly to isolate the factor Va that is GPI-anchored to the thrombin-activated platelet membrane. Subsequent to their final centrifugation, the platelets were resuspended in HTA and the platelet number determined. These platelets are referred to as “super-washed” platelets.

Use of analyses of kinetics of prothrombin activation to define the role of the GPI- anchored factor Va cofactor pool in Prothrombinase assembly and function.

Initial velocities of thrombin generation from prothrombin were monitored in 2-stage assays. Briefly, prothrombin (0.05 - 15 µM) activation was monitored in a reaction mixture containing 1 x 108/mL super-washed platelets, 30 µM DAPA, and 1 mM RGDS in HT

139

buffer. Subsequent to a 5 min incubation at 25oC, the reaction was initiated with factor Xa

(12 nM). At timed intervals (0, 10, 20, 30 and 40 sec), aliquots were withdrawn and diluted

3-fold into 20 mM HEPES, 150 mM NaCl, 5 mM CaCl2, 0.1% PEG 8000, pH 7.4 containing 50 mM EDTA (quench buffer) to stop the reaction. The thrombin concentration in each sample was determined using the chromogenic substrate Spectrozyme TH (0.4 mM, final). The change in absorbance at 405 nm was monitored (10 min, 30oC) using a

SpectroMax250 microtiter plate reader (Molecular devices, CA) and compared to a standard curve (0 - 100 nM thrombin) prepared daily. The concentrations of thrombin determined were fit to linear regression plots to obtain rates followed by non-linear fitting to the Michaelis-Menten equation using GraphPad Prism v6.0 software (GraphPad Prism,

CA) to determine the kinetic constants (Km and Vmax).

The binding affinity of factor Xa for GPI-anchored platelet-derived factor Va was also determined using a 2-stage assay to measure initial rates of thrombin generation as described above. Platelets isolated from three different donors were used as sources of GPI- anchored platelet-derived factor Va. Prothrombin activation mixtures contained 1.4 µM prothrombin, 1 x 108/mL super-washed platelets, 30 µM DAPA, and 1 mM RGDS in HT buffer. Subsequent to a 5 min incubation at 25oC, the reaction was initiated with variable concentrations of factor Xa (0.25 - 30 nM). At time intervals, aliquots were removed and the thrombin present determined as described above. Data from the initial rates of thrombin generation as a function of factor Xa was fit to one-site binding equation using GraphPad

Prism v6.0 software (GraphPad Prism, CA) to obtain the dissociation constant (Kd) for factor Xa.

140

Near identical assay conditions were used to quantify the amount of platelet-derived

GPI-anchored factor Va bound to thrombin-activated platelets obtained from four different donors. Prothrombin activation mixtures contained 1.4 µM prothrombin, 1 x 108/mL super washed platelets, 30 µM DAPA, and 1 mM RGDS in HT buffer. Subsequent to a 5 min incubation at 25oC, the reaction was initiated with factor Xa (12 nM). At time intervals, aliquots were removed and the thrombin present determined as described above.

Determination of the effect of APC and plasmin on the cofactor activity associated with dissociable factor Va pool and the GPI-anchored platelet-derived factor Va pool

Proteolytic inactivation of the dissociable pool of platelet-derived factor Va and the GPI- anchored platelet-derived factor Va pool were monitored in the presence of APC or plasmin. The supernatant containing dissociable platelet-derived factor Va was prepared by centrifugation of thrombin-activated platelets (1 x 109/mL, 1000 xg, 3 min) (36). The supernatant is also rich in platelet-derived microparticles that provide the membrane surface for Prothrombinase assembly and function. Super-washed platelets retaining GPI- anchored platelet-derived factor Va were prepared as described. Reactions contained 1 x

109/mL super-washed platelets or 500 µL of the mixture containing dissociable platelet- derived factor Va, 1 mM RGDS, and 30 nM hirudin in HT buffer. Following a 5 min incubation (37oC), catalytic amounts of APC (0.25 nM) or plasmin (25 nM) were added.

Timed aliquots (0, 5, 10, 15, 20, 30, 45, 60, 90, 120 min) were removed and quenched in sample preparation buffer (62.5 mM Tris, 5 mM EDTA, 2% SDS, 10% glycerol, 0.01% bromophenol blue, pH 6.8; SPB). Simultaneously, samples were removed, and assayed for residual cofactor activity in a Prothrombinase assay containing 1 x 108/mL protease-treated

141

super-washed platelets, 1 mM RGDS, 30 µM DAPA in the presence of 12 nM factor Xa at

25oC in HT buffer pH 7.4.

The APC- and plasmin-catalyzed inactivation of plasma-derived factor Va used

PCPS vesicles as the required membrane surface. To assess APC-catalyzed inactivation, reaction mixtures contained plasma-derived factor Va (20 nM), hirudin (30 nM), and PCPS vesicles (10 µM) in 20 mM HEPES, 150 mM NaCl, 5 mM CaCl2 containing 0.1% PEG

8000, pH 7.4. (HBS/Ca2+/PEG). For plasmin-catalyzed inactivation experiments, the concentrations of plasma-derived factor Va and PCPS vesicles were increased to 500 nM and 30 µM, respectively. Otherwise reaction conditions were identical to those described above for APC. Each reaction mixture was preincubated for 5 min at 37oC followed by the addition of 0.25 nM APC or 25 nM plasmin, respectively. Aliquots of the reaction mixtures were removed at timed intervals (0, 5, 10, 15, 20, 30, 45, 60, 90, 120 min) and assayed for residual cofactor activity as described. Concurrently, samples were quenched in SPB for immunoblotting analyses.

SDS-PAGE and Western blotting analyses

All samples diluted in SPB were boiled (45 sec, 90oC) under reducing conditions (5% β- mercaptoethanol). Protein samples were analyzed based on their mass (50 ng/lane) using

10% or 4-12% Tris-Gly gradient gels (Invitrogen), according to the method Laemmli et al.

(54) of proteins and peptides resolved via SDS-PAGE were transferred to nitrocellulose as described by Towbin and colleagues. (55) The factor Va antigen was probed with anti- human factor V heavy chain monoclonal antibody (αhFV #17), detected with a horse anti-

142

mouse coupled to horseradish peroxidase (HRP) by enhanced chemiluminescence and exposure and development of imaging film.

143

RESULTS

Characterization of the contribution of GPI-anchored platelet-derived factor Va to

Prothrombinase assembly and function on thrombin-activated platelets

Previous studies from our laboratory demonstrated that when GPI-anchored platelet-derived factor Va is the sole source of the cofactor, it supports Prothrombinase assembly and efficient prothrombin activation. More importantly, the prothrombin cleavage products observed during the reaction indicate that prothrombin was initially cleaved at Arg271 followed by cleavage at Arg320 (the prethrombin-2 pathway of activation) through a concerted mechanism (56). In this study, the Kd for factor Xa binding to GPI- anchored platelet-derived factor Va, expressed by super-washed, thrombin-activated platelets, was determined by monitoring prothrombin activation at varying factor Xa concentrations (Fig. 1). The data shown are representative of experiments performed with platelets isolated from different donors (n = 3). The initial rate of thrombin generation increased as a function of the factor Xa concentration and was maximal at 10 - 12 nM

-9 factor Xa. From these data, an apparent Kd = 1.6 x 10 M was determined to govern factor

Xa binding to the GPI-anchored cofactor, consistent with the values for factor Xa assembly into platelet-bound Prothrombinase reported in earlier studies (9).

Studies were done subsequently to determine the kinetic and binding parameters governing prothrombin activation when the GPI-anchored platelet-derived cofactor was the sole source of factor Va in Prothrombinase assays. Data obtained with two different platelet donors (Fig. 2) yielded Km’s for prothrombin activation that were identical (0.50 vs 0.48 µM) and similar to previous studies (9,57). The different Vmax values (4.01 vs 5.58

144

nMs-1) observed result from the difference of the number of GPI-anchored cofactor molecules expressed on the surface of thrombin-activated platelets between donors.

Subsequently, the amount of GPI-anchored platelet-derived factor Va bound to thrombin-activated, super-washed platelets isolated from four different donors was determined by monitoring the initial rate of thrombin generation catalyzed by platelet- bound Prothrombinase under two conditions. By determining the rate of thrombin generation supported by GPI-anchored platelet-derived factor Va vs. the total platelet- derived factor Va associated with the same donor’s unwashed, thrombin-activated platelets, the percentage of the GPI-anchored pool could be calculated (Table I). The data indicate that between 25% to 90% of the total pool of platelet-derived factor Va, capable of supporting Prothrombinase assembly and thrombin production, is GPI-anchored to the activated platelet membrane similar to earlier observations from our laboratory (10,41).

Effects of APC and plasmin on the cofactor activity of GPI-anchored platelet-derived factor Va

Earlier studies have reported a differential protease susceptibility between plasma and platelet-derived factor Va (36,39). Experiments were performed to compare APC- and plasmin-mediated proteolysis of plasma-derived factor Va to that of the GPI-anchored platelet-derived factor Va pool. APC effects factor Va inactivation and loss of cofactor activity by targeted proteolysis at Arg306, Arg506 and Arg679 within the factor Va heavy chain (58,59). Initial cleavage of the heavy chain at Arg506 results in the generation of 75 kDa intermediate fragment which can be subsequently cleaved at Arg306 to form a 45/42 kDa fragments. Final cleavage at Arg679 of the 42 kDa fragment produces a 30 kDa 145

associated with complete inactivation. The APC-catalyzed inactivation of plasma-derived factor Va bound to PCPS vesicles and GPI-anchored platelet-derived factor V, expressed on the surface of thrombin-activated platelets are shown in Fig. 3. Consistent with several studies, approximately 50% of the plasma-derived factor Va was inactivated within 10 min of the addition of catalytic amounts of APC (Fig 3A, major graph). By two hours, approximately 80% of the cofactor activity of the plasma factor Va was lost, consistent with observations made by others (60,61). Loss of cofactor activity of plasma-derived factor Va correlated with the generation of low molecular weight proteolytic products from immunoblotting analyses (data not shown). In marked contrast, the GPI-anchored platelet- derived factor Va pool was completely resistant to inactivation by APC (Fig. 3A, graph inset). Thus, greater than 95% of the cofactor activity remained after 2 h of incubation of

GPI-anchored factor Va with APC.

In support of the cofactor activity measurements (Fig 3A, graph inset), no proteolytic fragments were observed after prolonged (60 min) incubation of GPI-anchored platelet-derived factor Va with APC (Fig. 3A, immunoblot inset). A faint band representing the 75 kDa species was only identified after two hours of incubation with APC, and did not result in a significant loss of cofactor activity. In contrast, the APC-catalyzed inactivation of the dissociably-bound pool of platelet-derived factor Va was apparent based on the proteolytic profile obtained (Fig. 3B). APC initially cleaved this pool of platelet-derived factor Va at Arg506 to generate a 75 kDa fragment (Fig. 3B), which was further cleaved at

Arg306 to generate the 30 kDa fragment that is associated with complete loss of cofactor function. Despite both these factor Va pools being derived from the platelet, the dissociable

146

pool of platelet-derived factor Va (Fig. 3B) is more susceptible to APC-catalyzed inactivation compared to the GPI-anchored pool (Fig. 3A, inset), which retains close proximity to the activated platelet surface.

Plasmin cleaves the heavy chain of membrane-bound factor Va at multiple sites including Lys309, Lys310, Arg313, Arg348 and Arg506 resulting in the generation of several low molecular weight fragments and loss of cofactor activity (58,59). As shown in Fig 4A, incubation of PCPS-bound plasma-derived factor Va with plasmin resulted in the time- dependent loss of 85% cofactor activity after 30 min of incubation, consistent with observations from other laboratories. (40). In marked contrast, GPI-anchored platelet- derived factor Va retained 100% of its initial cofactor activity following a 40 min incubation with plasmin (Fig. 4B). As the reaction progressed, there was an increase in its cofactor activity, which peaked at ~140% after 2 h, consistent with previous observations

(39).

The plasmin-mediated cleavage profiles were consistent with the different cofactor activities of plasma-derived factor Va and GPI-anchored platelet-derived factor Va. After a 5 min incubation with plasmin, proteolysis of the plasma-derived factor Va heavy chain

(105 kDa) was accompanied by the appearance of fragments of ~75, 50, 45, 30, and 20 kDa (Fig. 4A, inset) with significant proteolysis and inactivation observed by 30 min. A similar pattern of proteolysis and proteolytic fragments were observed in control experiments with or without exogenously added PCPS vesicles in the presence of dissociable platelet-derived factor Va (Fig. 5). In contrast, the heavy chain of GPI-anchored platelet-derived factor Va was > 95% intact and only the ~50 and 45 kDa fragments were

147

detected after a prolonged (2 h) incubation with plasmin (Fig 4B, inset). These fragments are associated with a minimal loss in cofactor activity and may partly explain the negligible loss of activity in the GPI-anchored platelet-derived factor Va pool.

148

DISCUSSION

Platelets adhered and activated at sites of vascular injury promote hemostasis by the release of the contents of α- and dense-granules including von Willebrand factor

(62,63), fibrinogen (3,62,63), plasminogen activator inhibitor-1 (62), and factor V (5,26),

Platelet-derived factor V originates following endocytosis of plasma-derived factor V by platelet precursor cells, megakaryocytes via a clathrin-dependent, receptor-mediated event

(24-26,28). Following its endocytosis and prior to its trafficking to α-granules, factor V undergoes extensive retailoring intracellularly to yield a physically (29-35) and functionally (23,27,36,37) unique cofactor molecule including but not limited to differences in protease susceptibility (23,36,39), differences in N- and O-linked glycosylation (29,35), and resistance to phosphorylation catalyzed by a platelet-associated kinase (31,41). Most recently, Wood et al. identified a subpopulation of platelet-derived factor Va associated with thrombin-activated platelets, which is non-dissociably associated with the activated platelet membrane (10,41). Thus, when activated platelets are extensively washed in the presence or absence of calcium ions, this pool of platelet-derived factor Va remains bound. Additional evidence suggests that this pool of factor Va is GPI- anchored to the activated platelet membrane via the factor Va heavy chain (41). The data shown and described in this study have begun to characterize the functional properties of

GPI-anchored platelet-derived factor Va. This distinct pool of platelet-derived factor Va accounts for 40% - 90%, of the total platelet-derived factor Va that participates in

Prothrombinase assembly and function. When incorporated into Prothrombinase, GPI- anchored platelet-derived factor Va supports thrombin generation with kinetic parameters

149

that are similar to those observed when the plasma-derived molecule is used as the cofactor

(17,21,23). As shown previously it supports the cleavage of prothrombin through the prethrombin-2 pathway of prothrombin activation (56,57) and not the meizothrombin pathway (64-66) indicating that the GPI-anchored platelet-derived factor Va pool will sustain only thrombin generation at the activated platelet surface.

The importance of platelet-derived factor Va to sustaining the procoagulant response when required cannot be overstated, yet processes that down-regulate the activity of Prothrombinase are equally important to maintaining the balance between hemostasis and thrombosis. In contrast to the plasma-derived molecule, which is rapidly and proteolytically inactivated by APC during down-regulation of the procoagulant response

(38,61,67), the GPI-anchored platelet-derived factor Va is totally resistant to proteolysis and inactivation by APC. Similar observations were made by Camire et al. who reported that the total platelet-derived factor Va pool (dissociable plus GPI-anchored) retained

~40% cofactor activity despite prolonged incubation with APC (36). They further demonstrated that when plasma-derived factor Va was added to the activated platelet membrane, it was completely inactivated by APC and the baseline cofactor activity was restored to 40%. Although this GPI-anchored pool was cleaved by the fibrinolytic enzyme, plasmin, to some extent, its cofactor activity was sustained and increased over time consistent with previous studies using the total platelet-derived cofactor pool as the factor

Va source (39). It can be hypothesized that plasmin may have proteolytically inactivated one or more unknown inhibitory determinants on the activated platelet surface resulting in the protease resistance of GPI-anchored platelet-derived factor Va. Alternatively, the

150

delayed release of factor V/Va cofactor molecules from the open canalicular system (68) may partly contribute to the increased cofactor activity and decreased proteolytic inactivation.

Previous observations demonstrated that the non-dissociable, GPI-anchored platelet-derived factor Va pool defines the subpopulation of platelets capable of assembling a functional Prothrombinase, and represents ~50% of the total platelet-derived factor Va pool (10). This subpopulation of platelets capable of binding factor Va and factor Xa has a significantly (4 to 6-fold) increased expression of several adhesion molecules important for normal hemostasis, including glycoprotein Ibα and integrins αIIb and β3, thereby placing these platelets in a unique position to contribute to both hemostatic and thrombotic events. This subpopulation is hypothesized to play a significant role in regulating both normal hemostasis and pathological thrombus formation because the adherent properties of platelets and their ability to mount and sustain a procoagulant response are crucial steps in both of these processes. Increased adhesive receptor density guarantees the adhesion of these procoagulant platelets to von Willebrand factor. Following shear-induced platelet activation, GPI-anchored platelet-derived factor Va is expressed on the activated membrane surface. Factor Xa, formed via the tissue factor/factor VIIa complex, assembles with GPI-anchored platelet-derived factor Va to form Prothrombinase and generate small amounts of thrombin, which subsequently activate more platelets. Recruitment and incorporation of procoagulant platelet subpopulations into the growing thrombus via their increased expression of αIIbβ3 will follow. As these activated platelets also express a pool of platelet-derived factor Va that is resistant to proteolytic inactivation, not only will these

151

platelets promote thrombin generation, but they will support a sustained and prolonged procoagulant response. It can also be hypothesized that these events will be as likely to promote thrombosis as to prevent hemorrhage.

ACKNOWLEDGEMENTS

This work was supported by the National Institutes of Health [P01HL46703-Project 3 to

P.B.T]. F.A. was supported by the Department of Biochemistry, the University of Vermont, and the Cellular, Molecular and Biomedical Science Graduate Program housed within the

College of Medicine. The authors would like to acknowledge Drs. Beth Bouchard and Jay

Silveira for their insightful observations concerning and our work, and Dr. Bouchard for her critical review of this manuscript.

AUTHORSHIP

Contribution: F.A. designed and performed experiments, analyzed results, and wrote the manuscript. P.B.T. designed the research and wrote the manuscript.

Conflict of interest disclosure: The authors declare no competing financial interest.

Correspondence: Paula B. Tracy, 89 Beaumont Ave, Given Medical Research Building

C409, Burlington, VT 05405; Telephone: 802-656-1995; Fax: 802-656-8220; and Email: [email protected].

152

REFERENCES

1. Hellums, J. D., Edmunds, L. H., Grinnell, F., Sutera, S. P., Weiss, H. J., Mann, K. G., and Cooper, S. L. (1982) Platelets revisited. Transactions - American Society for Artificial Internal Organs 28, 627-630

2. Robert, K. A., José, A. L., and Michael, C. B. (1997) Molecular mechanisms of platelet adhesion and activation. The International Journal of Biochemistry & Cell Biology

3. Holmsen, H., and Weiss, H. J. (1979) Secretable storage pools in platelets. Annual review of medicine

4. Osterud, B., Rapaport, S. I., and Lavine, K. K. (1977) Factor V activity of platelets: evidence for an activated factor V molecule and for a platelet activator. Blood 49, 819-834

5. Viskup, R. W., Tracy, P. B., and Mann, K. G. (1987) The isolation of human platelet factor V. Blood 69, 1188-1195

6. Tracy, P. B., Peterson, J. M., Nesheim, M. E., McDuffie, F. C., and Mann, K. G. (1979) Interaction of coagulation factor V and factor Va with platelets. J Biol Chem 254, 10354-10361

7. Tracy, P. B. (2001) Role of platelets and leukocytes in coagulation. Hemostasis and Thrombosis. Basic Principles and Clinical Practice (R.W. Colman, et al). Lippincott Williams & Wilkins, Philadelphia, 575-596

8. Mann, K. G., Nesheim, M. E., Church, W. R., Haley, P., and Krishnaswamy, S. (1990) Surface-dependent reactions of the vitamin K-dependent enzyme complexes. Blood 76, 1-16

9. Tracy, P. B., Eide, L. L., and Mann, K. G. (1985) Human prothrombinase complex assembly and function on isolated peripheral blood cell populations. J Biol Chem 260, 2119-2124

10. Fager, A. M., Wood, J. P., Bouchard, B. A., Feng, P., and Tracy, P. B. (2010) Properties of procoagulant platelets: defining and characterizing the subpopulation binding a functional prothrombinase. Arterioscler Thromb Vasc Biol 30, 2400-2407

11. Bouchard, B. A., Catcher, C. S., Thrash, B. R., Adida, C., and Tracy, P. B. (1997) Effector cell protease receptor-1, a platelet activation-dependent membrane protein, regulates prothrombinase-catalyzed thrombin generation. J Biol Chem 272, 9244- 9251

153

12. Tracy, P. B., Giles, A. R., Mann, K. G., Eide, L. L., Hoogendoorn, H., and Rivard, G. E. (1984) Factor V (Quebec): a bleeding diathesis associated with a qualitative platelet Factor V deficiency. J Clin Invest 74, 1221-1228

13. Chiu, H. C., Rao, A. K., Beckett, C., and Colman, R. W. (1985) Immune complexes containing factor V in a patient with an acquired neutralizing antibody. Blood 65, 810-818

14. Dahlback, B. (1994) Inherited resistance to activated protein C, a major cause of venous thrombosis, is due to a mutation in the factor V gene. Haemostasis 24, 139- 151

15. Kalafatis, M., and Mann, K. G. (1995) Factor V Leiden and thrombophilia. N Engl J Med 332, 1382-1383; author reply 1384

16. Dahlback, B., Carlsson, M., and Svensson, P. J. (1993) Familial thrombophilia due to a previously unrecognized mechanism characterized by poor anticoagulant response to activated protein C: prediction of a cofactor to activated protein C. Proc Natl Acad Sci U S A 90, 1004-1008

17. Nesheim, M. E., Katzmann, J. A., Tracy, P. B., and Mann, K. G. (1981) Factor V. Methods Enzymol 80 Pt C, 249-274

18. Esmon, C. T. (1979) The subunit structure of thrombin-activated factor V. Isolation of activated factor V, separation of subunits, and reconstitution of biological activity. J Biol Chem 254, 964-973

19. Kane, W. H., and Majerus, P. W. (1981) Purification and characterization of human coagulation factor V. J Biol Chem 256, 1002-1007

20. Tracy, P. B., Eide, L. L., Bowie, E. J., and Mann, K. G. (1982) Radioimmunoassay of factor V in human plasma and platelets. Blood 60, 59-63

21. Nesheim, M. E., and Mann, K. G. (1979) Thrombin-catalyzed activation of single chain bovine factor V. J Biol Chem 254, 1326-1334

22. Nesheim, M. E., Taswell, J. B., and Mann, K. G. (1979) The contribution of bovine Factor V and Factor Va to the activity of prothrombinase. J Biol Chem 254, 10952- 10962

23. Monkovic, D. D., and Tracy, P. B. (1990) Activation of human factor V by factor Xa and thrombin. Biochemistry 29, 1118-1128

154

24. Bouchard, B. A., Meisler, N. T., Nesheim, M. E., Liu, C. X., Strickland, D. K., and Tracy, P. B. (2008) A unique function for LRP-1: a component of a two-receptor system mediating specific endocytosis of plasma-derived factor V by megakaryocytes. Journal of thrombosis and haemostasis : JTH 6, 638-644

25. Bouchard, B. A., Williams, J. L., Meisler, N. T., Long, M. W., and Tracy, P. B. (2005) Endocytosis of plasma-derived factor V by megakaryocytes occurs via a clathrin-dependent, specific membrane binding event. Journal of thrombosis and haemostasis : JTH 3, 541-551

26. Camire, R. M., Pollak, E. S., Kaushansky, K., and Tracy, P. B. (1998) Secretable human platelet-derived factor V originates from the plasma pool. Blood 92, 3035- 3041

27. Monkovic, D. D., and Tracy, P. B. (1990) Functional characterization of human platelet-released factor V and its activation by factor Xa and thrombin. J Biol Chem 265, 17132-17140

28. Ayombil, F., Abdalla, S., Tracy, P. B., and Bouchard, B. A. (2013) Proteolysis of plasma-derived factor V following its endocytosis by megakaryocytes forms the platelet-derived factor V/Va pool. Journal of thrombosis and haemostasis : JTH 11, 1532-1539

29. Fernandez, J. A., Hackeng, T. M., Kojima, K., and Griffin, J. H. (1997) The carbohydrate moiety of factor V modulates inactivation by activated protein C. Blood 89, 4348-4354

30. Silveira, J. R., Kalafatis, M., Tracy, P.B. (1997) The influence of the carbohydrate moieties in the inactivation of factor V and factor Va by APC. Blood 90, (Abstract 3166)

31. Kalafatis, M., Rand, M. D., Jenny, R. J., Ehrlich, Y. H., and Mann, K. G. (1993) Phosphorylation of factor Va and factor VIIIa by activated platelets. Blood 81, 704- 719

32. Kim, S. W., Ortel, T. L., Quinn-Allen, M. A., Yoo, L., Worfolk, L., Zhai, X., Lentz, B. R., and Kane, W. H. (1999) Partial glycosylation at asparagine-2181 of the second C-type domain of human factor V modulates assembly of the prothrombinase complex. Biochemistry 38, 11448-11454

33. Nicolaes, G. A., Villoutreix, B. O., and Dahlback, B. (1999) Partial glycosylation of Asn2181 in human factor V as a cause of molecular and functional heterogeneity. Modulation of glycosylation efficiency by mutagenesis of the consensus sequence for N-linked glycosylation. Biochemistry 38, 13584-13591

155

34. Hortin, G. L. (1990) Sulfation of tyrosine residues in coagulation factor V. Blood 76, 946-952

35. Gould, W. R., Silveira, J. R., and Tracy, P. B. (2004) Unique in vivo modifications of coagulation factor V produce a physically and functionally distinct platelet- derived cofactor: characterization of purified platelet-derived factor V/Va. J Biol Chem 279, 2383-2393

36. Camire, R. M., Kalafatis, M., Simioni, P., Girolami, A., and Tracy, P. B. (1998) Platelet-derived factor Va/Va Leiden cofactor activities are sustained on the surface of activated platelets despite the presence of activated protein C. Blood 91, 2818- 2829

37. Pittman, D. D., Tomkinson, K. N., Michnick, D., Selighsohn, U., and Kaufman, R. J. (1994) Posttranslational sulfation of factor V is required for efficient thrombin cleavage and activation and for full procoagulant activity. Biochemistry 33, 6952- 6959

38. Camire, R. M., Kalafatis, M., Cushman, M., Tracy, R. P., Mann, K. G., and Tracy, P. B. (1995) The mechanism of inactivation of human platelet factor Va from normal and activated protein C-resistant individuals. J Biol Chem 270, 20794- 20800

39. Conlon, S. J., Camire, R. M., Kalafatis, M., and Tracy, P. B. (1997) Cleavage of platelet-derived factor Va by plasmin results in increased and sustained cofactor activity on the thrombin-activated platelet surface. Thromb Haemost 77, 2507a

40. Omar, M. N., and Mann, K. G. (1987) Inactivation of factor Va by plasmin. J Biol Chem 262, 9750-9755

41. Wood, J. P., Fager, A. M., and Tracy, P. B. (2008) Platelet-derived factor Va expressed on the surface of the activated platelet is GPI-anchored. Blood 112, 219a

42. Higgins, D. L., and Mann, K. G. (1983) The interaction of bovine factor V and factor V-derived peptides with phospholipid vesicles. J Biol Chem 258, 6503-6508

43. Foster, W. B., Tucker, M. M., Katzmann, J. A., Miller, R. S., Nesheim, M. E., and Mann, K. G. (1983) Monoclonal antibodies to human coagulation factor V and factor Va. Blood 61, 1060-1067

44. van 't Veer, C., Golden, N. J., Kalafatis, M., and Mann, K. G. (1997) Inhibitory mechanism of the protein C pathway on tissue factor-induced thrombin generation.

156

Synergistic effect in combination with tissue factor pathway inhibitor. J Biol Chem 272, 7983-7994

45. Mann, K. G. (1976) Prothrombin. Methods Enzymol 45, 123-156

46. Kisiel, W., Canfield, W. M., Ericsson, L. H., and Davie, E. W. (1977) Anticoagulant properties of bovine plasma protein C following activation by thrombin. Biochemistry 16, 5824-5831

47. Barlow, G. H., Summaria, L., and Robbins, K. C. (1984) Hydrodynamic studies on the streptokinase complexes of human plasminogen, Val442-plasminogen, plasmin, and the plasmin-derived light (B) chain. Biochemistry 23, 2384-2387

48. Kisiel, W., and Hanahan, D. J. (1973) Purification and characterization of human Factor II. Biochimica et biophysica acta 304, 103-113

49. Fenton, J. W., Fasco, M. J., and Stackrow, A. B. (1977) Human thrombins. Production, evaluation, and properties of alpha-thrombin. The Journal of biological chemistry 252, 3587-3598

50. Di Scipio, R. G., Hermodson, M. A., Yates, S. G., and Davie, E. W. (1977) A comparison of human prothrombin, factor IX (Christmas factor), factor X (Stuart factor), and protein S. Biochemistry 16, 698-706

51. Nesheim, M. E., Myrmel, K. H., Hibbard, L., and Mann, K. G. (1979) Isolation and characterization of single chain bovine factor V. J Biol Chem 254, 508-517

52. Mustard, J. F., Perry, D. W., Ardlie, N. G., and Packham, M. A. (1972) Preparation of suspensions of washed platelets from humans. Br J Haematol 22, 193-204

53. Wood, J. P., Fager, A., Silveira, J. R., and Tracy, P. B. (2008) Platelet-derived factor Va expressed on the surface of the activated platelet is GPI-anchored. Blood 112, 219a

54. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685

55. Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proceedings of the National Academy of Sciences of the United States of America 76, 4350-4354

56. Haynes, L. M., Bouchard, B. A., Tracy, P. B., and Mann, K. G. (2012) Prothrombin Activation by Platelet-associated Prothrombinase Proceeds through the

157

Prethrombin-2 Pathway via a Concerted Mechanism. Journal of Biological Chemistry 287, 38647-38655

57. Wood, J. P., Silveira, J. R., Maille, N. M., Haynes, L. M., and Tracy, P. B. (2011) Prothrombin activation on the activated platelet surface optimizes expression of procoagulant activity. Blood 117, 1710-1718

58. Kalafatis, M., and Mann, K. G. (2001) Factor V: Dr. Jeckyll and Mr. Hyde. Adv Exp Med Biol 489, 31-43

59. Kalafatis, M., and Mann, K. G. (2001) The role of the membrane in the inactivation of factor va by plasmin. Amino acid region 307-348 of factor V plays a critical role in factor Va cofactor function. J Biol Chem 276, 18614-18623

60. Kalafatis, M., Bertina, R. M., Rand, M. D., and Mann, K. G. (1995) Characterization of the molecular defect in factor VR506Q. J Biol Chem 270, 4053- 4057

61. Heeb, M. J., Kojima, Y., Greengard, J. S., and Griffin, J. H. (1995) Activated protein C resistance: molecular mechanisms based on studies using purified Gln506-factor V. Blood 85, 3405-3411

62. Harrison, P., and Cramer, E. M. (1993) Platelet alpha-granules. Blood Rev 7, 52-62

63. Sehgal, S., and Storrie, B. (2007) Evidence that differential packaging of the major platelet granule proteins von Willebrand factor and fibrinogen can support their differential release. Journal of thrombosis and haemostasis : JTH 5, 2009-2016

64. Krishnaswamy, S. (2004) Exosite-driven substrate specificity and function in coagulation. J. Thomb Haemost. 3, 54-67

65. Krishnaswamy, S., and Betz, A. (1997) Exosites determine macromolecular substrate recognition by prothrombinase. Biochemistry 36, 12080-12086

66. Krishnaswamy, S., Mann, K. G., and Nesheim, M. E. (1986) The prothrombinase- catalyzed activation of prothrombin proceeds through the intermediate meizothrombin in an ordered, sequential reaction. J Biol Chem 261, 8977-8984

67. Kalafatis, M., Rand, M. D., and Mann, K. G. (1994) The mechanism of inactivation of human factor V and human factor Va by activated protein C. J Biol Chem 269, 31869-31880

158

68. White, J. G., and Clawson, C. C. (1980) The surface-connected canalicular system of blood platelets--a fenestrated membrane system. The American journal of pathology 101, 353-364

159

TABLE

Source of cofactor activity 1 2 3 4 Mean ± SD

Total platelet-derived FVa* 6.73 2.50 4.37 6.45 5.01 ± 1.98

GPI-anchored platelet-derived FVa 2.70 1.11 3.77 1.68 2.32 ± 1.17

% GPI-anchored platelet-derived FVa 40.12 44.40 86.27 26.05 49.21 ± 25.92

*Total platelet-derived factor Va = the sum of dissociably-bound factor Va and GPI-anchored factor Va

160

FIGURES

Figure 1. The binding interactions of factor Xa and super washed platelets as measured by prothrombin activation kinetics. Thrombin-activated platelets were super washed as described in the Materials and Methods. Prothrombinase reactions contained prothrombin (1.4 µM), super washed platelets (1 x 108 m/L), and DAPA (30 µM) in HT buffer pH 7.4. Thrombin generation was initiated by the addition of factor Xa (0.25 – 30 nM) and the initial rates of thrombin generation determined by two-stage assay. The data were fit to a one-site binding equation using GraphPad software.

161

Figure 2. Determination of the kinetic constants of prothrombin activation by Prothrombinase bound to super washed platelets. Thrombin-activated platelets were super washed as described in the Materials and Methods and diluted to 1 x 108/mL. Prothrombinase reactions mixtures contained prothrombin (0.05 - 15 µM) and DAPA (30 µM) in HT buffer. Thrombin generation was initiated by the addition of factor Xa (12 nM), and the initial rates of thrombin generation determined by two stage assay. The data from two donors were fit to the Michaelis-Menten equation using GraphPad Prism software to obtain the Km and Vmax values shown in the inset.

162

Figure 3. Comparison of APC-catalyzed inactivation patterns of GPI-anchored, dissociable pool of platelet-derived, and plasma-derived factor Va molecules. A) APC-inactivation reactions contained thrombin-activated plasma-derived factor Va (20 nM), 10 µM PCPS vesicles, 30 nM hirudin in HBS/Ca2+/PEG buffer, pH 7.4 or Inset: GPI- anchored platelet-derived factor Va (500 µL). Each reaction was initiated with APC (0.25 nM) and aliquots removed at select time points (1, 5, 10, 15, 20, 30, 45, 60, 90 and 120 min) and then immediately assayed for residual cofactor activity in prothrombinase assays at limiting factor Va as described in detailed the Materials and Methods section. Alternatively, at the designated time points, samples were withdrawn and quenched in sample preparation buffer containing 1% βME for immunoblotting analyses using anti- human factor Va (αhFV#17) monoclonal antibody against the heavy chain region. In panel (B), the dissociable platelet-derived factor Va (500 µL) from activated platelets (1 x 109 m/L) was treated with APC (0.25 nM) in HT buffer, pH 7.4 and the proteolytic products monitored by immunoblotting.

163

164

Figure 4. Plasmin-mediated proteolysis of plasma-derived factor Va vs non- dissociably bound platelet-derived factor Va. (A) Thrombin-activated plasma-derived factor Va (500 nM) was treated with plasmin (25 nM) in a reaction mixture containing 30 µM PCPS vesicles and 30 nM hirudin in the presence of HBS/Ca2+/PEG buffer, pH 7.4. Hirudin was added to inhibit the residual thrombin present in the reaction mixture. At selected time points, aliquots of the reaction were removed and immediately assayed for residual cofactor activity in prothrombinase assays at limiting factor Va as described in detailed the Materials and Methods section. Inset: alternatively, at the same designated time intervals, samples were withdrawn and added to SPB containing 5% βME to quench the reaction. Samples were subsequently boiled, resolved by electrophoresis and probed using monoclonal antibodies against human factor Va heavy chain (αhFV#17). Similar experiments were performed in panel (B). The plasmin concentration was unchanged, however the substrate this time round was super- washed platelets (1 x 109 m/L), containing the GPI-anchored pool of platelet-derived factor Va in HT buffer, pH 7.4. Likewise, the graph shows the residual cofactor activity remaining and the inset is a representative of an immunoblot of the factor Va heavy chain proteolytic fragments.

165

166

Figure 5. Immunoblotting analyses of plasmin-catalyzed proteolysis of the dissociable pool of platelet-derived factor Va. Following thrombin (2 U/mL) activation of platelets (1 x 109 /mL), the platelets were gently centrifuged (660 xg, 3 min) and the supernatant pooled to form microparticle rich platelet- releasate containing the dissociable pool of platelet-derived factor Va. Plasmin (25 nM) was added to 500 µL of the dissociable platelet-derived factor Va in the presence of hirudin (30 nM) and RGDS (1 mM) at 37 oC. Aliquots were removed at timed intervals and quenched in SPB. Products of the reactions were resolved by SDS-PAGE followed by Western blotting analyses using monoclonal antibodies against human factor Va heavy chain (αhFV#17).

167

CHAPTER FIVE: Characterization of the functional properties of demannosylated plasma-derived factors V and Va

Francis Ayombil1, Mark E. Jennings2, Paula B. Tracy1,*

1Department of Biochemistry, University of Vermont College of Medicine, 89 Beaumont Ave. Given Medical Research Building C409, Burlington, VT, 05405 2Department of Chemistry, University of Vermont, Cook Physical Sciences Building, 82 University Pl, Burlington, VT 05405

*All correspondence should be addressed to Paula B. Tracy, Ph.D., Email: [email protected], Telephone: 802-656-1995, Fax: 802-656-8220

Short title: Characterization of de-mannosylated plasma-derived factor V Key words: Plasma-derived factor V, prothrombinase, deglycosylation, endoglycosidase H, megakaryocytes

Abbreviations used in this manuscript include: Endoglycosidase H, endo H; factor Va, thrombin-activated factor Va; megakaryocyte-like cell-lines, CMKs; PC, phosphatidylcholine; PS, phosphatidylserine; PCPS vesicle, 75% PC: 25% PS; DAPA, dansyl-L-arginine N-(3-ethyl-1,5-pentanediyl)amine.

168

ABSTRACT

Coagulation factor V exists as two major procoagulant pools in human blood. The platelet- derived pool is acquired by megakaryocyte uptake and subsequent proteolytic modifications of the plasma-derived factor V pool to form a protease-resistant cofactor stored within the platelet α-granule. The functional properties of the platelet-derived molecule have been attributed to both its close association with the platelet membrane, and the post-translational modifications occurring subsequent to endocytic trafficking. One such modification is the presence of high mannose N-linked glycans at Asn269 and Asn1982 on the heavy and light chain regions, respectively. To investigate the importance of these glycans, plasma-derived factor V was incubated with endoglycosidase H to selectively remove these N-linked glycans. Treatment resulted in the removal of approximately 80% of the high mannose glycans on either chain. The proteolytic cleavage profiles obtained by incubation with catalytic amounts of APC were indistinguishable between deglycosylated factor V and the untreated control. When cofactor activity was assessed, deglycosylated factor V expressed 80% of the cofactor activity of the control when assayed in prothrombin time clotting assays. In kinetic studies, the catalytic efficiency (8.2 x 107 vs. 5.0 x 107 M-

1s-1) of Prothrombinase was decreased almost two-fold in the presence of deglycosylated factor Va compared to the control. There was no observed effect on enzyme affinity (Km ~

0.20 µM). Following deglycosylation, factor Va endocytosis was enhanced compared to the untreated samples. These combined data suggest that targeted removal of high mannose residues on plasma-derived factor V/Va may impact its function.

169

INTRODUCTION

The coagulation protein factor Va is absolutely essential for physiological thrombin generation via Prothrombinase assembled on the surface of activated platelets adhered at sites of vascular injury. The procofactor factor V (Mr, 330, 000) is found in two distinct pools in whole blood. Approximately 75-80% circulates in plasma as inactive single chain molecule (1) while the remaining 20-25% is stored in platelet α-granules following its endocytosis by megakaryocytes (1,2). While both plasma- and platelet-derived factor Va can support thrombin generation via Prothrombinase, the platelet-derived factor V/Va pool appears more physiologically relevant (3-5). In addition, platelet-derived factor V expresses several physical differences (6-12), which would enhance its ability to promote and sustain coagulation at sites of vascular injury. It is hypothesized that these modifications occur within the megakaryocyte subsequent to factor V endocytosis. In the megakaryocyte, factor V undergoes substantial proteolysis to form a partially, proteolytically-activated platelet-derived factor V/Va pool (13,14), including cleavage by an unknown protease(s) at Tyr1543 to form the platelet-derived factor Va light chain (12).

In contrast, the plasma-derived factor Va light chain is formed following cleavage by thrombin at Arg1545 (15,16). Using a mass spectrometry-based approach, we also determined that Thr402 in platelet-derived factor V contains an N-acetylgalactosamine or

N-acetylglucosamine not observed in plasma-derived factor V (12). The heavy chain of platelet-derived factor Va is also attached to the platelet membrane via a glycophosphatidylinosital (GPI) anchor (17,18), which confers total resistance of platelet- derived factor Va to inactivation by APC and plasmin (as described in Chapter Four).

170

Mass spectrometry techniques, developed to characterize potential N-linked glycosylation sites in tissue factor (19,20) were applied to the study of human factor Va to determine the sites, composition, and microheterogeneity of the glycan chains. Of the nine potential N-linked glycosylation sites within the heavy chain, glycans were identified at five sites (Asn23, Asn27, Asn269, Asn432 and Asn439), three sites were not glycosylated

(Asn354, Asn526, and Asn639), and one other site (Asn211) still requires identification (Fig.

1). Two of the five glycosylation sites (Asn269 and Asn432) were identified previously (21).

All three sites in the light chain (Asn1675, Asn1982, and Asn2181) were glycosylated. Asn2181 in the light chain had been identified and characterized previously (9,10,21). Four glycans were identified as sialic acid-containing complex carbohydrates (Asn432, Asn439, Asn1675, and Asn2181) and two still require characterization (Asn23 and Asn27). The remaining two sites, Asn269 on the heavy chain and Asn1982 were identified as high-mannose glycans.

Whereas plasma-derived factor Va contains high mannose glycans with either 6, 7, 8, or 9 mannose residues at Asn1982, platelet-derived factor Va glycans at the same site have only

4, 5, or 6 mannose residues, suggesting extensive trimming of this glycan by megakaryocytes following endocytosis (2,12). The high mannose glycans of plasma- and platelet-derived factor Va on Asn269 are similar.

Glycan moieties have been shown to modulate multiple facets of factor V/Va function (15,22-24), including membrane binding (10), cofactor activity (7,12,25), and

APC-catalyzed inactivation (15,22-24). These functional effects can be substantial, e.g. partial removal of carbohydrate from factor V promotes a ten-fold increase its rate of cleavage by APC (6). Thus, the goal of this study was to assess the functional significance

171

of the mannose residues at Asn1982. The high mannose glycans on plasma-derived factor V were removed by treatment with endoglycosidase H, which removes Asn-linked mannose oligosaccharides after the first N-acetylglucosamine residue (26).

172

MATERIALS AND METHODS

Materials

Coomassie Brilliant Blue R-250, and bromophenol blue were purchased from Sigma (St.

Louis, MO). Calcium chloride and polyethylene glycol (PEG) 8000 were obtained from J.

T. Baker (Philipsburg, NJ). Precision plus protein molecular weight standards were purchased from Bio-Rad (Hercules, CA). X-ray autoradiography film, and Western blotting chemiluminescence reagent were obtained from Krackeler Scientific (Albany,

NY), and PerkinElmer Inc. (Waltham, MA), respectively. Tris base and Pierce proprietary protein-Free T20 (TBS) blocking buffer were obtained from Thermo Fisher Scientific

(Rockford, IL). Thrombin-specific inhibitor, hirudin, was purchased from Calbiochem

(Gibbstown, NJ). Dansylarginine N-(3-ethyl-1,5-pentanediyl)amine (DAPA) was bought from Haematologic Technologies (Essex Junction, VT). TriniClot PT Simplastin Excel thromboplastin reagent was purchased from Tcoag (Wicklow, Ireland). Bovine serum albumin was obtained from MP Biomedicals (Solon, OH). Recombinant endoglycosidase

H from Streptomyces plicatus was purchased from Sigma Chemical Co (St. Louis, MO).

Recombinant exoglycosidase α1-2,3 mannosidase cloned from Xanthomonas manihotis was obtained from New England Biolabs Inc. (Ipswich, MA). Phospholipid vesicles comprised of 75% phosphatidylcholine (wt/wt) and 25% phosphatidylserine (wt/wt)

(PCPS) were a generous gift from Dr. Kenneth Mann (University of Vermont) and prepared according to the procedure of Higgins and Mann (27) with reagents purchased from Avanti Polar Lipids (Alabaster, AL). The chromogenic α-thrombin substrate

Spectrozyme TH was purchased from Sekisui Diagnostics (San Diego, CA). Antibodies:

173

Horseradish peroxidase conjugated to horse anti-mouse IgG was purchased from Vectors

Labs (Burlingame, CA). The lectin from Galanthus nivalis (Snowdrop bulb) conjugated to horseradish peroxidase was obtained from US Biologicals (Salam, MA). Monoclonal antibodies against human factor Va heavy chain (αHFVa LC#9) and light chain (αHFVa

HC#17) were provided by the University of Vermont Antibody Core Facility (28,29).

Proteins

Human activated protein C (APC), plasmin, factor Xa, and thrombin, were all obtained from Haematologic Technologies (Essex Junction). Human factor V and prothrombin were purified as previously described (30,31) from freshly frozen plasma obtained from consenting healthy adults, according to the guidelines of the University of Vermont’s

Institutional Review Board and Committee on Human Subjects. Proteins were stored in

o 50% glycerol containing 5 mM CaCl2 at -20 C. Before use, glycerol was removed by dialysis of sufficient dilution and protein concentrations of proteins determined by UV

1% -1 -1 absorbance at 280nm with the following extinction coefficients (E 280nm, M cm ) and molecular weights (Da): human APC, 14.5, 56200 (32), human plasmin, 17.0, 83000 (33), human prothrombin, 13.8, 72000 (34); human thrombin, 18.3, 36700 (35); human factor

Xa, 11.6, 46000 (36); and human factor V, 9.6, 330000 (37).

Removal of high mannose glycans on plasma-derived factor V by Endoglycosidase H

Human plasma-derived factor V (4 µM) was incubated (2 h, 37oC) with 0.23 µM endoglycosidase H under non-protein denaturing conditions in 20 mM HEPES, 150 mM

2+ NaCl, pH 7.4 (HBS) containing 5 mM CaCl2 and 0.1% PEG 8000 (HBS/Ca /PEG). Factor

V was incubated under similar conditions in the absence of endoglycosidase H in parallel,

174

control experiments. Following deglycosylation, factor V (1.0 µM) was activated to factor

Va by α-thrombin (20 nM, 10 min, 37oC). Aliquots of factor Va were either prepared for

SDS-PAGE by dilution in 5X sample preparation buffer (62.5 mM Tris, 5 mM EDTA, 2%

SDS, 10% glycerol, 0.01% bromophenol blue, pH 6.8; SPB), or by diluted in 20 mM Tris-

HCl, 150 mM NaCl, pH 7.4 (TBS) or HBS to assess cofactor activity in various in functional assays.

Endocytosis of plasma-derived factor V by megakaryocyte-like cell-lines (CMKs)

CMK cells were cultured as previously described (2,38). All factor V endocytosis experiments were performed under serum-free conditions as previously described (2,38).

CMK cells (1 x 106/mL) were incubated in the presence of endoglycosidase H-treated plasma-derived factor V or untreated plasma-derived factor V (30 nM, 4 h, 37oC). The cells were extensively washed by centrifugation (250 xg, 7 min) followed by resuspension of the cells in HBS to remove excess and cell surface-bound factor V. The final cell pellets were adjusted to 1 x 106/mL and solubilized in protease inhibitor cocktail to prevent additional proteolysis of the endocytosed protein (13). A portion of the cell lysate was treated with thrombin (2 U/mL, 10 min, 37oC) to convert all of the endocytosed factor V and its proteolytic products (15) to factor Va prior to dilution in SPB for SDS-PAGE and

Western blotting analyses.

The effect of endoglycosidase H treatment on plasma-derived factor Va cofactor activity

The cofactor activity of demannosylated plasma-derived factor Va was monitored in a prothrombin time-based clotting assay using factor V immunodepleted plasma as

175

previously described (31). In this assay, a standard curve is generated by making dilutions

(1:2 through 1:128) of a normal human plasma pool as the factor V source in TBS, pH 7.4.

The clot times of native plasma-derived factor Va and endoglycosidase-treated factor Va were measured and compared to the standard curve.

The cofactor activity of demannosylated factor Va was also assessed via

Prothrombinase assembled on PCPS vesicles. Endoglycosidase H-treated, or untreated factor V were thrombin-activated to factor Va (20 nM) before incubation (5 min, 25oC) with PCPS vesicles (20 µM), DAPA (30 µM) and prothrombin (0 - 2 µM) in

HBS/Ca2+/PEG. The reaction was initiated by the addition of factor Xa (0.5 nM). Timed aliquots (0, 10, 20, 30 and 40 sec) were removed and diluted in quench buffer (20 mM

HEPES, 150 mM NaCl, 50 mM EDTA, 0.1% PEG 8000, pH 7.4). The concentration of thrombin in samples was monitored by chromogenic assay (18). The initial rates of thrombin generation and kinetic constants (Km, Vmax and kcat) governing Prothrombinase were determined using GraphPad Prism v6.0 software.

APC-catalyzed proteolytic inactivation of endoglycosidase H-treated factor Va

Factor Va derived from thrombin-activated endoglycosidase H-treated factor V or untreated factor V were assayed for their susceptibility to APC-catalyzed proteolytic inactivation in vitro. Factor Va (20 nM), and PCPS vesicles (200 µM) in HBS/Ca2+/PEG buffer, pH 7.4 were pre-incubated for 5 min at 37oC. The reaction was initiated by the addition of APC (0.2 nM). At timed intervals (0 - 2 h) aliquots of the reaction were removed and quenched in SPB containing 1% βME. Samples were subsequently boiled and stored for further SDS-PAGE and Western blotting analyses.

176

SDS-PAGE and Western blotting analyses

Factor V/Va fragments were detected by Western blot analyses using monoclonal antibodies against human factor V heavy chain (αhFV HC#17) and light chain (αhFV

LC#9) (28,29). Alternatively, demannosylation was assessed by probing factor V/Va fragments with an HRP-conjugated lectin derived from Galanthus nivalis. To evaluate electrophoretic mobility differences proteins were stained with Coomassie blue following

SDS-PAGE.

177

RESULTS

Removal of high mannose glycans from plasma-derived factor V

We targeted global removal of N-linked high mannose carbohydrates with endoglycosidase H under non-denaturing conditions. No apparent difference in the electrophoretic mobility of endoglycosidase H-treated factor V (lane 2) and the untreated factor V (lane 1) were detected (Fig. 2A). However, upon thrombin activation of the factor

V to factor Va, there was an obvious electrophoretic shift in both the heavy and light chains of demannosylated factor Va (lane 4) as compared to the untreated control (lane 3) indicative of the loss of sugars at Asn269 and Asn1982 (21). Removal of the high mannose glycans was confirmed using Galanthus nivalis (Snowdrop bulb) lectin (Fig. 2B). This lectin is selective, and has high affinity for mannose residues (39). A substantial loss of high mannose glycans was observed following treatment with endoglycosidase H as demonstrated by not only an increase in electrophoretic mobility, but also a decrease in the lectin signal (Fig. 2B). Further ESI LC MS/MS analyses (40) of the treated factor Va samples confirmed that 70-80% of the high mannose glycans were removed at positions

Asn269 and Asn1982 (data not shown). In control experiments, RNase B, which contains a single high mannose residue at Asn34 (41), was fully demannosylated with endoglycosidase

H under the conditions used in this study (data not shown).

Effect of endoglycosidase H treatment on factor V and factor Va endocytosis by megakaryocyte-like cell-lines

The effect of removal of high mannose N-linked glycans from factor V on its endocytosis and subsequent proteolysis by CMK cells was assessed by Western blotting using heavy

178

chain, αhFVa HC#17 (Fig. 3A), and light chain, αhFVa LC#9 (Fig. 3B) monoclonal antibodies. Factor V endocytosis by megakaryocytes is receptor-dependent (42) and mediated by its light chain (43,44). To ensure that any differences that were observed were due to actual differences in factor V endocytosis or proteolysis, the cell lysates from equivalent numbers of cells were subjected to electrophoresis and immunoblotting.

Comparison of the relative amounts and proteolytic pattern of endocytosed factor V demonstrated no substantial differences between untreated and endoglycosidase H-treated factor V although shifts in electrophoretic mobility consistent with glycan removal were observed (Fig. 3A & B, left panels). Following endocytosis, the cells were lysed and treated with thrombin to activate the factor V and its partial activation products (45) to factor Va.

Interestingly, comparison of the relative densities of the heavy chain (105 kDa) and light chain (74 kDa) following conversion of factor V to factor Va in cell lysates demonstrated increased endocytosis of demannosylated factor V (Fig. 3A & B, middle panels).

Interestingly, in both instances, it was observed that endocytosis of the active cofactor, factor Va, was greater than that observed with single chain factor V (Fig. 3A & B, right panels).

Effect of demannosylation on factor Va cofactor activity

Using a prothrombin time-based clotting assay, endoglycosidase H-treated factor Va expressed ~80% the activity of the untreated control. Since subtle differences in cofactor activity may be more pronounced in clotting assays as compared to other measures of cofactor activity, prothrombin activation via Prothrombinase assembled on PCPS vesicles in the presence of untreated and endoglycosidase H-treated factor Va were also compared.

179

The kinetics of prothrombin activation performed in the presence of untreated factor Va

-1 (Km = 0.22 µM, kcat = 18 s , Fig. 4A) were similar to those determined using

-1 endoglycosidase H-treated factor Va (Km = 0.20 µM, kcat = 10 s , Fig. 4B). The catalytic efficiency of Prothrombinase assembled with demannosylated factor Va (kcat/Km = 5.0 x

7 -1 -1 7 -1 -1 10 M s ) was ~60% that of the untreated control (kcat/Km = 8.2 x 10 M s ), consistent with the 20% loss in activity observed in clotting assays. While treatment with endoglycosidase H had a pronounced effect on the Vmax of the reaction as compared to untreated factor Va (5.1 nMs-1 and 8.8 nMs-1, respectively), these studies collectively suggest that removal of high mannose N-linked glycan on factor Va may result in altering its cofactor activity.

Effect of endoglycosidase H treatment on APC-mediated inactivation of factor Va

When sufficient amounts of thrombin are generated following the hemostatic response,

APC promotes factor Va inactivation by cleavage at three major sites in the heavy chain

(Arg306, Arg506 and Arg679, Fig. 5) (46,47). Proteolysis of the heavy chain of factor Va is accompanied by the appearance of a high molecular weight 75 kD fragment consistent with cleavage at Arg506. Subsequent cleavage of the 75 kD fragment at Arg306 produces a 30 kD fragment. Cleavage at Arg306 is membrane-dependent and promotes significant loss of cofactor function (8). The rates and patterns of APC-mediated proteolysis of untreated factor Va (Fig. 6A) and demannosylated factor Va (Fig. 6B) were similar, suggesting that selective removal of the high mannose glycans had no effect on the susceptibility of factor

Va to inactivation by APC. For both cofactors, the 75 kD fragment, consistent with cleavage at Arg506, was observed after one minute of incubation, whereas the 30 kD

180

fragment, consistent with cleavage of the 75 kD fragment at Arg306, was detected within ten min (Fig. 6A & B).

181

DISCUSSION

Previous studies have demonstrated that glycosylation modulates the cofactor function of the plasma-derived protein, factor Va (6,9,10,25,48-50). Various effects on factor Va cofactor function and inactivation by APC were observed; however, in those experiments the consequences of global removal of glycans, and not specific glycan moieties, were monitored. The studies detailed here are the first to characterize the effects of selective removal of high mannose N-linked glycans from plasma-derived factor Va by endoglycosidase H. Treatment with endoglycosidase H removed the majority of the high- mannose glycans as demonstrated by a shift in protein electrophoretic mobility and less reactivity to a mannose binding lectin, which was confirmed by mass spectrometry (80% glycan removal). Endoglycosidase-H treatment, however, appears to moderately decrease the cofactor activity of plasma-derived factor Va by approximately 20% in clotting assays.

Loss of the high mannose glycan slightly increased factor V endocytosis but did not affect its proteolysis by the megakaryocyte-like cell line (CMK). Finally, there was no effect of endoglycosidase-H treatment on APC-mediated, inactivation of factor Va. In unpublished studies by the Bouchard laboratory, selective removal of complex N-linked glycans from factor V was also without effect on factor Va cofactor activity, but a similar increase in factor V endocytosis by megakaryocytes was observed (unpublished data).

N-linked glycans (6,41,49,50) and specifically high mannose glycans (21) play prominent roles in the folding, quality control, and intracellular trafficking of proteins.

Using recombinant factor V and factor VIII molecules, Kaufmann and colleagues determined that high mannose glycans in the B-domains of these proteins mediated their

182

intracellular interactions with Lectin Mannose-binding 1, LMAN1 or Endoplasmic

Reticulum-Golgi Intermediate Compartment 53 kDa, ERGIC-53 (51,52). LMAN1, in conjunction with Multiple Coagulation Factor Deficiency protein 2, MCFD2, functions in the transport of factor V and factor VIII from the endoplasmic reticulum to the Golgi apparatus (53-55). Zheng and others subsequently demonstrated using purified glycans that the carbohydrate recognition domain of LMAN1 binds to Man-α-1,2-Man, the terminal carbohydrate moiety of high mannose glycans (51,56). Using confocal microscopy, it was demonstrated previously that subsequent to its endocytosis by megakaryocytes, factor V is taken up into early and late endosomes and undergoes retrograde trafficking through the

Golgi apparatus and endoplasmic reticulum prior to its storage in α-granules. Its trafficking through the Golgi apparatus is consistent with the presence of a nonconventional modification, O-linked glycosylation, on the platelet-derived cofactor that is not found in its plasma counterpart (12). In addition, as glycoproteins are trafficked through the Golgi apparatus, high mannose glycans undergo substantial trimming by Golgi mannosidases

(51,55,56). Subsequent to factor V endocytosis, high mannose glycans at Asn1982, which contain either 6, 7, 8, or 9 mannose residues undergo extensive trimming as the platelet- derived factor V molecule glycans at the same site have only 4, 5, or 6 mannose residues.

Finally, its trafficking through the endoplasmic reticulum is consistent with the addition of a GPI anchor (57,58). Only a small subset of cellular proteins including SNAREs (soluble

N-ethylmaleimide-sensitive factor attachment protein), acid hydrolase receptors, and processing peptidases undergo retrograde transport (59). Exogenous bacterial (Shiga toxin and cholera toxin) and plant (ricin and abrin) toxins exploit this pathway following their

183

cellular uptake as they require proteolytic activation by furin in the trans-Golgi network.

The active toxins are eventually translocated into the cytosol (59,60). Thus, the retrograde trafficking and functional modification of factor V following its endocytosis represents a unique cellular process.

We have shown that high mannose glycans on factor V have significant effect on factor Va cofactor activity in Prothrombinase assembly yet marginal effect on susceptibility to APC-catalyzed inactivation. The loss of the glycans appears to promote enhanced endocytosis by megakaryocytes. In view of the important roles of mannose residues in the intracellular trafficking of secretory proteins, it is possible that these carbohydrate moieties may play an essential role during factor V trafficking subsequent to endocytosis by megakaryocytes. The results presented here provide a foundation for future studies to evaluate the intracellular route of demannosylated factor V upon uptake by megakaryocytes to gain a better understanding of the importance of selective trimming of mannose residues observed on platelet-derived factor V at Asn1982.

ACKNOWLEDGEMENTS

The authors are grateful to Dylan Souder for his technical assistance and Dr. Kenneth G.

Mann for generously providing us with PCPS vesicles. F.A. was supported by the

University of Vermont Cellular, Molecular, and Biomedical Program, and the Department of Biochemistry. This work was supported by NIH grants P01 HL46703, Project 3 to P.B.T.

The authors would like to thank Jay R. Silveira, PhD and Beth A. Bouchard, PhD for their review and suggestions for this paper. Part of the work in this manuscript was presented as

184

an abstract (Dylan Souder et al., 2014) at the University of Vermont Annual Students

Conference.

AUTHORSHIP

Contribution: F.A. designed, performed experiments, analyzed data, and wrote the manuscript; P.B.T. designed the research, and wrote the manuscript; M.E.J. performed experiments and analyzed data.

Conflict of interest disclosure: The authors have no competing financial interest in this manuscript.

185

REFERENCES

1. Tracy, P. B., Eide, L. L., Bowie, E. J., and Mann, K. G. (1982) Radioimmunoassay of factor V in human plasma and platelets. Blood 60, 59-63

2. Bouchard, B. A., Williams, J. L., Meisler, N. T., Long, M. W., and Tracy, P. B. (2005) Endocytosis of plasma-derived factor V by megakaryocytes occurs via a clathrin-dependent, specific membrane binding event. Journal of thrombosis and haemostasis : JTH 3, 541-551

3. Borchgrevink, C. F., and Owren, P. A. (1961) The hemostatic effect of normal platelets in hemophilia and factor V deficiency. The importance of clotting factors adsorbed on platelets for normal hemostasis. Acta Med Scand 170, 375-383

4. Ang, A. L., Kuperan, P., Ng, C. H., and Ng, H. J. (2009) Acquired factor V inhibitor. A problem-based systematic review. Thromb Haemost 101, 852-859

5. Nesheim, M. E., Nichols, W. L., Cole, T. L., Houston, J. G., Schenk, R. B., Mann, K. G., and Bowie, E. J. (1986) Isolation and study of an acquired inhibitor of human coagulation factor V. J Clin Invest 77, 405-415

6. Fernandez, J. A., Hackeng, T. M., Kojima, K., and Griffin, J. H. (1997) The carbohydrate moiety of factor V modulates inactivation by activated protein C. Blood 89, 4348-4354

7. Silveira, J. R., Kalafatis, M., Tracy, P.B. (1997) The influence of the carbohydrate moieties in the inactivation of factor V and factor Va by APC. Blood 90, (Abstract 3166)

8. Kalafatis, M., Rand, M. D., Jenny, R. J., Ehrlich, Y. H., and Mann, K. G. (1993) Phosphorylation of factor Va and factor VIIIa by activated platelets. Blood 81, 704- 719

9. Kim, S. W., Ortel, T. L., Quinn-Allen, M. A., Yoo, L., Worfolk, L., Zhai, X., Lentz, B. R., and Kane, W. H. (1999) Partial glycosylation at asparagine-2181 of the second C-type domain of human factor V modulates assembly of the prothrombinase complex. Biochemistry 38, 11448-11454

10. Nicolaes, G. A., Villoutreix, B. O., and Dahlback, B. (1999) Partial glycosylation of Asn2181 in human factor V as a cause of molecular and functional heterogeneity. Modulation of glycosylation efficiency by mutagenesis of the consensus sequence for N-linked glycosylation. Biochemistry 38, 13584-13591

186

11. Hortin, G. L. (1990) Sulfation of tyrosine residues in coagulation factor V. Blood 76, 946-952

12. Gould, W. R., Silveira, J. R., and Tracy, P. B. (2004) Unique in vivo modifications of coagulation factor V produce a physically and functionally distinct platelet- derived cofactor: characterization of purified platelet-derived factor V/Va. J Biol Chem 279, 2383-2393

13. Ayombil, F., Abdalla, S., Tracy, P. B., and Bouchard, B. A. (2013) Proteolysis of plasma-derived factor V following its endocytosis by megakaryocytes forms the platelet-derived factor V/Va pool. Journal of thrombosis and haemostasis : JTH 11, 1532-1539

14. Camire, R. M., Pollak, E. S., Kaushansky, K., and Tracy, P. B. (1998) Secretable human platelet-derived factor V originates from the plasma pool. Blood 92, 3035- 3041

15. Monkovic, D. D., and Tracy, P. B. (1990) Activation of human factor V by factor Xa and thrombin. Biochemistry 29, 1118-1128

16. Nesheim, M. E., and Mann, K. G. (1979) Thrombin-catalyzed activation of single chain bovine factor V. J Biol Chem 254, 1326-1334

17. Wood, J. P., Fager, A. M., and Tracy, P. B. (2008) Platelet-derived factor Va expressed on the surface of the activated platelet is GPI-anchored. Blood 112, 219a

18. Fager, A. M., Wood, J. P., Bouchard, B. A., Feng, P., and Tracy, P. B. (2010) Properties of procoagulant platelets: defining and characterizing the subpopulation binding a functional prothrombinase. Arterioscler Thromb Vasc Biol 30, 2400-2407

19. Krudysz-Amblo, J., Jennings, M. E., 2nd, Mann, K. G., and Butenas, S. (2010) Carbohydrates and activity of natural and recombinant tissue factor. J Biol Chem 285, 3371-3382

20. Krudysz-Amblo, J., Jennings, M. E., 2nd, Matthews, D. E., Mann, K. G., and Butenas, S. (2011) Differences in the fractional abundances of carbohydrates of natural and recombinant human tissue factor. Biochim Biophys Acta 1810, 398-405

21. Liu , T., Qian, W. J., Gritsenko, M. A., Camp, D. G., Monroe, M. E., Moore R.J., and al., e. (2005) Human plasma N-glycoproteome analysis by immunoaffinity subtraction, hydrazide chemistry, and mass spectrometry. J Proteome Res. 4, 2070- 2080

187

22. Monkovic, D. D., and Tracy, P. B. (1990) Functional characterization of human platelet-released factor V and its activation by factor Xa and thrombin. J Biol Chem 265, 17132-17140

23. Camire, R. M., Kalafatis, M., Simioni, P., Girolami, A., and Tracy, P. B. (1998) Platelet-derived factor Va/Va Leiden cofactor activities are sustained on the surface of activated platelets despite the presence of activated protein C. Blood 91, 2818- 2829

24. Pittman, D. D., Tomkinson, K. N., Michnick, D., Selighsohn, U., and Kaufman, R. J. (1994) Posttranslational sulfation of factor V is required for efficient thrombin cleavage and activation and for full procoagulant activity. Biochemistry 33, 6952- 6959

25. Bruin, T., Sturk, A., ten Cate, J. W., and Cath, M. (1987) The function of the human factor V carbohydrate moiety in blood coagulation. Eur J Biochem 170, 305-310

26. Trimble, R. B., and Maley, F. (1984) Optimizing hydrolysis of N-linked high- mannose oligosaccharides by endo-beta-N-acetylglucosaminidase H. Anal Biochem 141, 515-522

27. Higgins, D. L., and Mann, K. G. (1983) The interaction of bovine factor V and factor V-derived peptides with phospholipid vesicles. J Biol Chem 258, 6503-6508

28. van 't Veer, C., Golden, N. J., Kalafatis, M., and Mann, K. G. (1997) Inhibitory mechanism of the protein C pathway on tissue factor-induced thrombin generation. Synergistic effect in combination with tissue factor pathway inhibitor. J Biol Chem 272, 7983-7994

29. Foster, W. B., Tucker, M. M., Katzmann, J. A., Miller, R. S., Nesheim, M. E., and Mann, K. G. (1983) Monoclonal antibodies to human coagulation factor V and factor Va. Blood 61, 1060-1067

30. Mann, K. G. (1976) Prothrombin. Methods Enzymol 45, 123-156

31. Nesheim, M. E., Katzmann, J. A., Tracy, P. B., and Mann, K. G. (1981) Factor V. Methods Enzymol 80 Pt C, 249-274

32. Kisiel, W., Canfield, W. M., Ericsson, L. H., and Davie, E. W. (1977) Anticoagulant properties of bovine plasma protein C following activation by thrombin. Biochemistry 16, 5824-5831

188

33. Barlow, G. H., Summaria, L., and Robbins, K. C. (1984) Hydrodynamic studies on the streptokinase complexes of human plasminogen, Val442-plasminogen, plasmin, and the plasmin-derived light (B) chain. Biochemistry 23, 2384-2387

34. Kisiel, W., and Hanahan, D. J. (1973) Purification and characterization of human Factor II. Biochimica et biophysica acta 304, 103-113

35. Fenton, J. W., Fasco, M. J., and Stackrow, A. B. (1977) Human thrombins. Production, evaluation, and properties of alpha-thrombin. The Journal of biological chemistry 252, 3587-3598

36. Di Scipio, R. G., Hermodson, M. A., Yates, S. G., and Davie, E. W. (1977) A comparison of human prothrombin, factor IX (Christmas factor), factor X (Stuart factor), and protein S. Biochemistry 16, 698-706

37. Nesheim, M. E., Myrmel, K. H., Hibbard, L., and Mann, K. G. (1979) Isolation and characterization of single chain bovine factor V. J Biol Chem 254, 508-517

38. Taniguchi, Y., London, R., Schinkmann, K., Jiang, S., and Avraham, H. (1999) The receptor protein tyrosine phosphatase, PTP-RO, is upregulated during megakaryocyte differentiation and Is associated with the c-Kit receptor. Blood 94, 539-549

39. Shibuya, N., Goldstein, I. J., Van Damme, E. J., and Peumans, W. J. (1988) Binding properties of a mannose-specific lectin from the snowdrop (Galanthus nivalis) bulb. J Biol Chem 263, 728-734

40. Chin, E. T., and Papac, D. I. (1999) The use of a porous graphitic carbon column for desalting hydrophilic peptides prior to matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Anal Biochem 273, 179- 185

41. Prien, J. M., Ashline, D. J., Lapadula, A. J., Zhang, H., and Reinhold, V. N. (2009) The high mannose glycans from bovine ribonuclease B isomer characterization by ion trap MS. Journal of the American Society for Mass Spectrometry 20, 539-556

42. Bouchard, B. A., Catcher, C. S., Thrash, B. R., Adida, C., and Tracy, P. B. (1997) Effector cell protease receptor-1, a platelet activation-dependent membrane protein, regulates prothrombinase-catalyzed thrombin generation. J Biol Chem 272, 9244- 9251

43. Bouchard, B. A., Meisler, N. T., Nesheim, M. E., Liu, C. X., Strickland, D. K., and Tracy, P. B. (2008) A unique function for LRP-1: a component of a two-receptor

189

system mediating specific endocytosis of plasma-derived factor V by megakaryocytes. Journal of thrombosis and haemostasis : JTH 6, 638-644

44. Bouchard, B. A., Abdalla, S., and Tracy, P. B. (2013) The factor V light chain mediates the binding and endocytosis of plasma-derived factor V by megakaryocytes. Journal of thrombosis and haemostasis : JTH 11, 2181-2183

45. Ayombil, F., Abdalla, S., Tracy, P. B., and Bouchard, B. A. (2013) Proteolysis Of Plasma-Derived Factor V Following Its Endocytosis By Megakaryocytes Forms The Platelet-Derived Factor V/Va Pool. Journal of thrombosis and haemostasis : JTH

46. Krishnaswamy, S., Williams, E. B., and Mann, K. G. (1986) The binding of activated protein C to factors V and Va. J Biol Chem 261, 9684-9693

47. Egan, J. O., Kalafatis, M., and Mann, K. G. (1997) The effect of Arg306-->Ala and Arg506-->Gln substitutions in the inactivation of recombinant human factor Va by activated protein C and protein S. Protein science : a publication of the Protein Society 6, 2016-2027

48. Barhoover, M. A., Orban, T., Beck, D. O., Bukys, M. A., and Kalafatis, M. (2008) Contribution of amino acid region 334-335 from factor Va heavy chain to the catalytic efficiency of prothrombinase. Biochemistry 47, 6840-6850

49. Rosing, J., Bakker, H. M., Thomassen, M. C., Hemker, H. C., and Tans, G. (1993) Characterization of two forms of human factor Va with different cofactor activities. J Biol Chem 268, 21130-21136

50. Silveira, J. R., Kalafatis, M., and Tracy, P. B. (2002) Carbohydrate moieties on the procofactor factor V, but not the derived cofactor factor Va, regulate its inactivation by activated protein C. Biochemistry 41, 1672-1680

51. Zhang, B., Kaufman, R. J., and Ginsburg, D. (2005) LMAN1 and MCFD2 form a cargo receptor complex and interact with coagulation factor VIII in the early secretory pathway. J Biol Chem 280, 25881-25886

52. Nichols, W. C., Seligsohn, U., Zivelin, A., Terry, V. H., Arnold, N. D., Siemieniak, D. R., Kaufman, R. J., and Ginsburg, D. (1997) Linkage of combined factors V and VIII deficiency to chromosome 18q by homozygosity mapping. J Clin Invest 99, 596-601

53. Hauri, H. P., and Schweizer, A. (1992) The endoplasmic reticulum-Golgi intermediate compartment. Current opinion in cell biology 4, 600-608

190

54. Hong, W. (1998) Protein transport from the endoplasmic reticulum to the Golgi apparatus. J Cell Sci 111 ( Pt 19), 2831-2839

55. Nichols, W. C., Seligsohn, U., Zivelin, A., Terry, V. H., Hertel, C. E., Wheatley, M. A., Moussalli, M. J., Hauri, H. P., Ciavarella, N., Kaufman, R. J., and Ginsburg, D. (1998) Mutations in the ER-Golgi intermediate compartment protein ERGIC-53 cause combined deficiency of coagulation factors V and VIII. Cell 93, 61-70

56. Zheng, C., Page, R. C., Das, V., Nix, J. C., Wigren, E., Misra, S., and Zhang, B. (2013) Structural characterization of carbohydrate binding by LMAN1 protein provides new insight into the endoplasmic reticulum export of factors V (FV) and VIII (FVIII). J Biol Chem 288, 20499-20509

57. Galian, C., Bjorkholm, P., Bulleid, N., and von Heijne, G. (2012) Efficient glycosylphosphatidylinositol (GPI) modification of membrane proteins requires a C-terminal anchoring signal of marginal hydrophobicity. J Biol Chem 287, 16399- 16409

58. Orlean, P., and Menon, A. K. (2007) Thematic review series: lipid posttranslational modifications. GPI anchoring of protein in yeast and mammalian cells, or: how we learned to stop worrying and love glycophospholipids. Journal of lipid research 48, 993-1011

59. Bonifacino, J. S., and Rojas, R. (2006) Retrograde transport from endosomes to the trans-Golgi network. Nature reviews 7, 568-579

60. Wernick, N. L., Chinnapen, D. J., Cho, J. A., and Lencer, W. I. (2010) Cholera toxin: an intracellular journey into the cytosol by way of the endoplasmic reticulum. Toxins 2, 310-325

191

FIGURES

Figure 1. Sites of N-linked glycosylation on the factor Va heavy and light chains. There are approximately 12 potential N-linked glycosylation sites on factor Va: nine on the heavy chain and three on the light chain. Of these, four complex N-linked glycans (●) have been identified on the heavy chain, and two have been identified on the light chain. High mannose N-linked glycans (▲) at Asn269 and Asn1982 in the factor Va heavy and light chains, respectively, have also been identified. Three sites are not glycosylated (Asn354, Asn526, and Asn639), and one other site (Asn211) still requires identification.

192

Figure 2. Electrophoretic mobility shifts following treatment of factor V and factor Va with endoglycosidase H. Factor V and thrombin-activated factor Va were treated with endoglycosidase H as described in Materials and Methods. Samples were resolved by SDS-PAGE and visualized by (A) Coomassie blue staining or (B) Western blotting analyses using mannose binding lectin Galanthus nivalis conjugated to HRP as the probe. Lanes 1 & 5 contain untreated factor V while lanes 2 & 6 contain endoglycosidase H-treated factor V. Lanes 3 & 7 and 4 & 8 contain untreated and treated factor Va, respectively. The electrophoretic mobilities of the molecular weight standards (kDa) are indicated on the left of each panel. The locations of single chain factor V (330 kDa; FV), the factor Va heavy chain (105 kDa; HC) and the factor Va light chain (74/72 kDa; LC) are indicated on the right.

193

Figure 3. Endocytosis and proteolytic processing of endoglycosidase H-treated factor V and factor Va by a megakaryocyte-like cell line. Endoglycosidase H-treated (T) and untreated (C) factor V or factor Va (30 nM) were incubated with CMK cells (1 x 106/mL, 2h, 37oC). Endocytosed factor V and factor Va present in cell lysates were analyzed by Western blotting using monoclonal antibodies against human factor Va (A) heavy chain (αhFV#17) or (B) light chain (αhFV#9). In some instances endocytosed factor V and its proteolytic products were converted to factor Va with thrombin prior to SDS-PAGE and blotting (middle panels). The electrophoretic mobilities of the molecular weight standards (kDa) are indicated on the left of each panel. The locations of single chain factor V (330 kDa; FV), the factor Va heavy chain (105 kDa; HC) and the factor Va light chain (74 kDa; LC) are indicated on the right.

194

Figure 4. Kinetic analyses of prothrombin activation by PCPS-bound Prothrombinase assembled in the presence of demannosylated factor Va. Prothrombinase reaction mixtures contained prothrombin (0.05 -2 µM), PCPS vesicles (20 μM), DAPA (30 µM) in the presence of (A) untreated or (B) endoglycosidase H-treated factor Va (20 nM). Thrombin generation was initiated by the addition of factor Xa (0.5 nM). Initial rates of thrombin generation were determined by two-stage chromogenic assays. The data were fit to the Michaelis-Menten equation to estimate the kinetic parameters (Km, Vmax, and kcat) using GraphPad software v6.

195

Figure 5. Schematic of activated protein C-catalyzed inactivation of factor Va heavy. Factor Va is proteolytically inactivated by activated protein C (APC) via three cleavages at Arg306, Arg506, and Arg679. A 75 kDa fragment is formed following cleavage at Arg506. Further cleavage of this fragment at Arg306 generates a 30 kDa fragment and results in a significant loss of cofactor activity and inactivation. Cleavage at this site (Arg306) is membrane-dependent (left hand pathway). In the alternate pathway, cleavage at Arg306 precedes cleavage at Arg506 resulting in the generation of a 60/62 kDa fragment. This doublet is predominant in reactions of APC-mediated inactivation of factor Va derived from individuals carrying the natural Arg to Glu mutation at position 506 (Factor VLeiden) (right hand pathway). Cleavage at Arg679 produces additional lower molecular weight proteolytic fragments (26/28 kDa and 20/22 kDa doublets), which are associated with only a slight loss in cofactor activity. Fragments shaded light blue are detected by monoclonal antibodies against human factor Va heavy chain (αhFVaHC#17).

196

Figure 6. APC-mediated proteolysis of thrombin-activated, endoglycosidase H- treated factor V. Factor V was demannosylated following treatment with endoglycosidase H and further activated to factor Va as described in Materials & Methods. (A) Untreated or (B) treated factor Va (20 nM) were incubated with APC (0.2 nM) in the presence of PCPS vesicles (200 µM) at 37oC. Aliquots of the reaction were removed at the times indicated in the figure (0 - 120 min) and quenched in SPB. The factor Va proteolytic products were resolved by SDS-PAGE and immunoblotted with monoclonal antibodies against human factor Va heavy chain (αhFVaHC#17). The electrophoretic mobilities of the molecular weight standards (kDa) are indicated on the left of each panel. The locations of heavy chain (105 kDa; HC), the 75 kDa fragment (representing cleavage at Arg506), the 60/62 kDa fragment (representing initial cleavage at Arg306) and the 30 kDa fragment (representing residues 306-506) are indicated on the right.

197

CHAPTER SIX: PERSPECTIVES AND FUTURE DIRECTIONS

When the vascular integrity is breached, the most crucial physiological function of the procoagulant response is the generation of the blood clotting enzyme, thrombin.

Subsequent to its generation, thrombin is essential to the anticoagulant and fibrinolytic responses, and thus the regulation of thrombin generation is central to maintaining the balance between normal hemostasis and pathologic thrombosis. Thrombin is generated from prothrombin by the coordinated efforts of the constituents of Prothrombinase, which includes an appropriate cellular membrane surface, factor Xa and its cofactor factor Va, and calcium ions (1,2). Previous studies of the biochemical functions of the

Prothrombinase complex under static conditions using purified proteins have demonstrated that removal of the membrane surface or factor Va from the enzyme complex results in approximately 1,000-fold, and 10,000-fold reduction, in the rate of thrombin generation, respectively (3). These reports highlight the critical roles of both the membrane binding sites and factor Va to thrombin generation.

Beyond the physiological environment, the required membrane can also be provided by PCPS vesicles, which significantly lessens the complexity of the system.

While these model membranes have been useful in obtaining our present understanding of the assembly and function of Prothrombinase, they do not mimic the physiological surface and constituents provided by the activated platelet membrane. Notably, activated platelets secrete factor Va molecules from their α-granules, which consist of two populations: a dissociable pool and non-dissociable, GPI-anchored pool (4,5). While both pools contribute to Prothrombinase function, the activated platelet membrane additionally

198

provides the catalytic surface for the binding and assembly of Prothrombinase at sites of vascular injury. Additional, significant functional differences between vesicle- and platelet-bound Prothrombinase have been highlighted earlier (Chapter One)

This dissertation examined how the platelet-derived components, particularly the non-dissociable, GPI-anchored pool of platelet-derived factor Va and platelet-bound

Prothrombinase function to maximize the activation of prothrombin to thrombin. Our investigations led to the observations that: 1) the non-dissociable, GPI-anchored pool of platelet-derived factor Va is totally resistant to proteolytic inactivation by APC and only partially susceptible to plasmin-catalyzed proteolysis, such that cofactor activity is sustained in the presence of both proteases and actually enhanced in the presence of plasmin; 2) thrombin generated by Prothrombinase, via the prethrombin-2 pathway, on the activated platelet surface is mechanistically different from the meizothrombin pathway utilized by vesicle-bound Prothrombinase such that prothrombin, configured in a proteinase-like state, is the preferred substrate for the platelet-bound enzyme and sustains the procoagulant response by avoiding the generation of the anticoagulant intermediate, meizothrombin produced when the substrate is configured in a zymogen-like state; and, 3) the Gla-domainless prethrombin-1 intermediate derived from thrombin feedback cleavage of prothrombin at Arg155 is effectively utilized to generate thrombin on the activated platelet surface in contrast to the PCPS vesicle surface, on which its turnover is inhibited by the accumulation of formed fragment-2. Described below are additional future studies and experimental approaches which can be utilized to better understand the observations summarized above.

199

Studies detailed in Chapter Two of this dissertation sought to gain additional insight into the molecular mechanism by which Prothrombinase bound to the activated platelet surface utilizes prethrombin-2, rather than meizothrombin, a pathway of thrombin generation, which is observed with vesicle-bound Prothrombinase. Utilization of the prethrombin-2 pathway maximizes the coagulation response by preventing the formation of the meizothrombin intermediate, an intermediate that expresses significant anticoagulant activity. By exploiting the ability of active site inhibitors to stabilize the conformationally- flexible prothrombin molecule in a proteinase-like state, we were able to show definitively that the platelet-bound enzyme prefers a substrate configured in that state, a configuration that is not recognized by the vesicle-bound enzyme. The present observations adds to a long list of evidence suggesting a differential assembly of Prothrombinase bound to the activated platelet surface vs. PCPS vesicles. The activated platelet surface is complex: as demonstrated by the expression of glycoprotein receptors, cytoskeletal rearrangements and the merger of the granular membrane with that of the platelet membrane, and the exposure of the intricate open canalicular system. Additionally, the activated platelets express distinct and saturable binding sites for components of Prothrombinase compared to the significantly less complex system of PCPS vesicles where binding by Prothrombinase’s components is not unique. We further predict that some unidentified distinct membrane features (such as receptors for factor Xa or factor factor Va) on the platelet surface regulate how the enzyme is assembled and ultimately how the substrate is configured for catalysis.

Therefore, the absence of these unidentified receptors on non-platelet surfaces including

PCPS vesicles may affect how the enzyme is assembled and how the substrate is

200

recognized in comparison to the activated platelet surface. One approach to identify some of these receptors may involve the use of cross-linking agents (such as Sulfo-SBED) to crosslink bound factor Xa, factor Va, and Prothrombinase to the activated platelet- membrane and then utilize mass spectrometric analyses to profile and identify proteins in the cross-linked complex. Crosslinking studies have been previously done in our laboratory therefore this technique will help us to identify potential receptors which can be then be targeted for selective removal with proteases to evaluate their roles in Prothrombinase assembly and function.

We further hypothesize that Prothrombinase bound to non-platelet surfaces prefers its substrate in a zymogen-like state. Previous studies have reported that human lymphocytes, monocytes, and endothelial cells (7), red blood cells (6) and other cellular- derived microparticles support Prothrombinase assembly and thrombin generation via the meizothrombin pathway. Therefore, using DAPA inhibitor binding studies employed in this dissertation, we can test the above hypothesis on these non-platelet surfaces. If indeed

Prothrombinase bound to non-platelet surfaces prefers to bind prothrombin in a zymogen- like configuration, we can then assess the distinct features (via Sulfo-SBED cross-linking studies) of these surfaces e.g. monocytes in comparison to platelets and then eliminate attributes which are similar on either surfaces (including glycoprotein receptors) while documenting distinct cellular features which may affect enzyme assembly. As described above, any cellular feature documented on these non-platelet surfaces could potentially be removed with proteases as well to verify their role in Prothrombinase function.

201

The studies described in Chapter Three continue to investigate the unique functional attributes of Prothrombinase bound to the activated platelet membrane vs. PCPS vesicles. Using prethrombin-1 as an alternate substrate for membrane-bound

Prothrombinase, we were able to conclude that platelet-bound Prothrombinase, binds, or stabilizes, prethrombin-1 in a configuration such that the fragment-2 formed by the initial cleavage of the substrate at Arg271 does not impede the subsequent cleavage at Arg320. In fact, even though both membrane-bound enzymes (platelet vs PCPS vesicle) effect the initial cleavage of prethrombin-1 at Arg271 to generate prethrombin-2 and fragment-2, the formed fragment-2 only inhibits the PCPS-bound enzyme. This inhibition of PCPS-bound

Prothrombinase appears to be mediated by some region within fragment-2 of prothrombin.

One approach to delineating this mechanism of inhibition is to express recombinant prethrombin-1 molecules lacking residues within the kringle 2 and linker 3 (residues 170 -

269) regions within fragment-2 and then monitor thrombin generation via initial rates and progress curves in reactions by PCPS-bound Prothrombinase as described in Chapter Three of this dissertation. Likewise, the effect of fragment-2 on prethrombin-1 activation by

Prothrombinase may be better assessed by preparing and purifying fragment-2 from factor

Xa-catalyzed prethrombin-1 activation reactions (7) and then monitoring the effect of fragment-2 in Prothrombinase titration assays. Data from this studies would also confirm the type of product inhibition occurring on PCPS vesicles. In contrast, we know that when activated, platelets express filopodia and open up their extensive network of canalicular system which may provide an ideal microenvironment for the binding of components of the Prothrombinase complex. With this unique environment absent on PCPS vesicles, it is

202

reasonable to further speculate that during prothrombin activation by the platelet-bound

Prothrombinase the intermediates formed are localized within the microenvironment of the enzyme where their local concentrations are higher allowing for their effective conversion into thrombin. Interestingly, prethrombin-1 activation occurs through cleavage at Arg271 suggesting that the substrate is bound in the proteinase-like state and expresses the pseudo- active site. We know from the Chapter Two of this dissertation that the substrate bound in the proteinase-like configuration is preferred by the platelet-bound Prothrombinase but not the vesicle-bound enzyme. Therefore, by investigating the effect of DAPA-induced prethrombin-1 stabilization on substrate activation by the platelet-bound enzyme, we can appreciate further the mechanism by which all the prothrombin activation species are utilized in thrombin generation at the site of vascular injury. Alternatively, on the PCPS vesicles we anticipate that, prethrombin-1 activation would be further inhibited by a combination of increase presentation of the undesirable proteinase-like substrate to the

PCPS-bound enzyme and the inhibition being mediated by fragment-2.

Studies detailed in Chapter Four allowed the functional characterization of a non- dissociable, GPI-anchored pool of platelet-derived factor Va. This unique factor Va pool is tethered to the platelet membrane via a GPI anchor on the heavy chain at Ser692 and distinguishes the subpopulation of activated platelets that bind a functional Prothrombinase

(4,8). This GPI-anchored factor Va pool supports Prothrombinase assembly and function identically to the non-tethered dissociably-bound cofactor (9,10). Using clotting assays,

Wood et al. estimated that ~40% of the platelet-derived factor Va, capable of supporting thrombin generation, is GPI-anchored to the platelet membrane. Using analyses of kinetics

203

of prothrombin activation, we were able to demonstrate that the GPI-anchored factor Va pool contributed between 25-90% of the total platelet-derived Prothrombinase activity expressed by four different platelet donors. While this approach directly assessed cofactor function, future studies using immunoassays (5) will be more appropriate to quantify the both the expression and variability of GPI-anchored, platelet-derived factor Va among a large group of donors. By generating antibodies against the GPI-anchored factor Va pool via targeting the GPI anchor, we can directly assess the contribution of this pool of platelet- derived factor Va to thrombin generation in Prothrombinase assays.

Earlier studies from our laboratory reported the presence of a subpopulation of platelet-derived factor Va which retained ~50% of its residual cofactor activity or showed a sustained and increased cofactor activity when treated with catalytic amounts of APC

(11) or plasmin (12), respectively. We have demonstrated that the GPI-anchored, platelet- derived factor Va represents this unique pool of protease-resistant factor Va.

Consequently, further study of the susceptibility of this factor Va pool to cleavage and perhaps, inactivation or activation by other proteases known to cleave factor Va, e.g. cathepsin G, elastase (13) and platelet calpains (14) is warranted. Future studies showing resistance of this pool to a known battery of other proteases will provide strong support for a concept of resistance mediated by the GPI-anchor modification on factor Va. Platelet- derived factor Va is critical to the coagulant response via Prothrombinase function; therefore, additional mechanistic insights into the protease resistance phenotype observed here and in the proposed future studies could be a useful therapeutic target. As described in Chapter One of this dissertation, platelet-derived factor Va is enough to sustain

204

hemostatic competence as shown in individuals with antibodies against their plasma factor

V (15) and in congenital factor V deficiency (16). Therefore, for individuals predisposed with thrombophilia including arterial thrombosis, the development and administration of antibodies to the GPI-anchored platelet-derived factor Va may be key to reducing their risks to platelet triggered cardiovascular diseases.

Work described in Chapter Five describes initial attempts to characterize functional properties provided by the high mannose glycans (Asn269 and Asn1982) present on both the heavy and light chains of factor Va. Our initial approach targeted global removal of the high mannose glycans at Asn269 (heavy chain) and Asn1982 (light chain) on plasma-derived plasma factor V using mannose-specific endoglycosidase H. While a marginal difference in cofactor activity was observed in functional assays, an enhanced uptake of both deglycosylated plasma-derived factor V and factor Va by the CMK megakaryocyte-like cell line was observed when compared to the untreated control. Studies have implicated mannose residues in the intracellular trafficking of both secretory and endocytosed proteins

(17-20). Therefore, future experiments should monitor the fate of deglycosylated plasma- derived factor V following its endocytosis by megakaryocytes. Our laboratory has detailed how endocytosed, plasma-derived factor V is trafficked within the megakaryocyte and therefore has the expertise to perform such experiments. It would be interesting to know if the increased uptake of deglycosylated factor Va reported here corresponds to its normal trafficking, yet increased packaging into α-granules, or if the increased uptake is simply due to an enhanced cycling in endosomal compartments with no clear path of trafficking to other organelles such as the endoplasmic reticulum and trans-golgi network.

205

Procedures for the isolation of platelet-derived factor V from donor platelets has been previously published by our laboratory (21); therefore future studies should evaluate the effect of treatment with endoglycosidase H on the cofactor activity of platelet-derived factor Va vs. plasma-derived factor Va.

In conclusion, studies detailed in this dissertation clearly and significantly underscore the functional importance of platelet-derived factor Va particularly the GPI- anchored pool, as well as the activated platelet membrane in the assembly and function of

Prothrombinase, and hence the procoagulant response. However more information is required if the activated platelet and its constituents are to be considered as potential therapeutic targets to inhibit thrombin formation catalyzed by platelet-bound

Prothrombinase, as detailed previously.

References 1. Barton, P. G., Jackson, C. M., and Hanahan, D. J. (1967) Relationship between factor V and activated factor X in the generation of prothrombinase. Nature 214, 923-924

2. Hemker, H. C., and Kahn, M. J. (1967) Reaction sequence of blood coagulation. Nature 215, 1201-1202

3. Nesheim, M. E., Taswell, J. B., and Mann, K. G. (1979) The contribution of bovine Factor V and Factor Va to the activity of prothrombinase. J Biol Chem 254, 10952- 10962

4. Fager, A. M., Wood, J. P., Bouchard, B. A., Feng, P., and Tracy, P. B. (2010) Properties of procoagulant platelets: defining and characterizing the subpopulation binding a functional prothrombinase. Arterioscler Thromb Vasc Biol 30, 2400-2407

5. Tracy, P. B., Eide, L. L., Bowie, E. J., and Mann, K. G. (1982) Radioimmunoassay of factor V in human plasma and platelets. Blood 60, 59-63

206

6. Whelihan, M. F., Zachary, V., Orfeo, T., and Mann, K. G. (2012) Prothrombin activation in blood coagulation: the erythrocyte contribution to thrombin generation. Blood 120, 3837-3845

7. Mann, K. G. (1976) Prothrombin. Methods Enzymol 45, 123-156

8. Wood, J. P., Fager, A. M., and Tracy, P. B. (2008) Platelet-derived factor Va expressed on the surface of the activated platelet is GPI-anchored. Blood 112, 219a

9. Tracy, P. B., Eide, L. L., and Mann, K. G. (1985) Human prothrombinase complex assembly and function on isolated peripheral blood cell populations. J Biol Chem 260, 2119-2124

10. Wood, J. P., Fager, A., Silveira, J. R., and Tracy, P. B. (2008) Platelet-derived factor Va expressed on the surface of the activated platelet is GPI-anchored. Blood 112, 219a

11. Camire, R. M., Kalafatis, M., Simioni, P., Girolami, A., and Tracy, P. B. (1998) Platelet-derived factor Va/Va Leiden cofactor activities are sustained on the surface of activated platelets despite the presence of activated protein C. Blood 91, 2818- 2829

12. Conlon, S. J., Camire, R. M., Kalafatis, M., and Tracy, P. B. (1997) Cleavage of platelet-derived factor Va by plasmin results in increased and sustained cofactor activity on the thrombin-activated platelet surface. Thromb Haemost 77, 2507a

13. Camire, R. M., Kalafatis, M., and Tracy, P. B. (1998) Proteolysis of factor V by cathepsin G and elastase indicates that cleavage at Arg1545 optimizes cofactor function by facilitating factor Xa binding. Biochemistry 37, 11896-11906

14. Bradford, H. N., Annamalai, A., Doshi, K., and Colman, R. W. (1988) Factor V is activated and cleaved by platelet calpain: comparison with thrombin proteolysis. Blood 71, 388-394

15. Ang, A. L., Kuperan, P., Ng, C. H., and Ng, H. J. (2009) Acquired factor V inhibitor. A problem-based systematic review. Thromb Haemost 101, 852-859

16. Bouchard, B. A., Chapin, J., Brummel-Ziedins, K. E., Durda, P., Key, N. S., and Tracy, P. B. (2015) Platelets and platelet-derived factor Va confer hemostatic competence in complete factor V deficiency. Blood 125, 3647-3650

17. Bonifacino, J. S., and Rojas, R. (2006) Retrograde transport from endosomes to the trans-Golgi network. Nature reviews 7, 568-579

207

18. Wernick, N. L., Chinnapen, D. J., Cho, J. A., and Lencer, W. I. (2010) Cholera toxin: an intracellular journey into the cytosol by way of the endoplasmic reticulum. Toxins 2, 310-325

19. Zhang, B., Kaufman, R. J., and Ginsburg, D. (2005) LMAN1 and MCFD2 form a cargo receptor complex and interact with coagulation factor VIII in the early secretory pathway. J Biol Chem 280, 25881-25886

20. Zheng, C., Page, R. C., Das, V., Nix, J. C., Wigren, E., Misra, S., and Zhang, B. (2013) Structural characterization of carbohydrate binding by LMAN1 protein provides new insight into the endoplasmic reticulum export of factors V (FV) and VIII (FVIII). J Biol Chem 288, 20499-20509

21. Gould, W. R., Silveira, J. R., and Tracy, P. B. (2004) Unique in vivo modifications of coagulation factor V produce a physically and functionally distinct platelet- derived cofactor: characterization of purified platelet-derived factor V/Va. J Biol Chem 279, 2383-2393

208

COMPREHENSIVE BIBLIOGRAPHY

Adams, T.E., Hockin, M.F., Mann, K.G. & Everse, S.J. (2004) The crystal structure of activated protein C-inactivated bovine factor Va: Implications for cofactor function. Proc Natl Acad Sci U S A, 101, 8918-23.

Alberio, L., Safa, O., Clemetson, K.J., Esmon, C.T. & Dale, G.L. (2000) Surface expression and functional characterization of alpha-granule factor V in human platelets: effects of ionophore A23187, thrombin, collagen, and convulxin. Blood, 95, 1694-702.

Allen, D.H. & Tracy, P.B. (1995) Human coagulation factor V is activated to the functional cofactor by elastase and cathepsin G expressed at the monocyte surface. J Biol Chem, 270, 1408-15.

Altieri, D.C. & Edgington, T.S. (1990) Identification of effector cell protease receptor-1. A leukocyte-distributed receptor for the serine protease factor Xa. J Immunol, 145, 246-53.

Amthaeur, R., Kodukula, K., Gerber, L. & Udenfriend, S. (1993) Evidence that the putative COOH-terminal signal transamidase involved in glycosylphosphatidylinositol protein synthesis is present in the endoplasmic reticulum. Proc Natl Acad Sci U S A, 90, 3973-3977.

Ang, A.L., Kuperan, P., Ng, C.H. & Ng, H.J. (2009) Acquired factor V inhibitor. A problem-based systematic review. Thromb Haemost, 101, 852-9.

Autin, L., Steen, M., Dahlback, B. & Villoutreix, B.O. (2006) Proposed structural models of the prothrombinase (FXa-FVa) complex. Proteins, 63, 440-50.

Ayombil, F., Abdalla, S., Tracy, P.B. & Bouchard, B.A. (2013) Proteolysis of plasma- derived factor V following its endocytosis by megakaryocytes forms the platelet- derived factor V/Va pool. J Thromb Haemost, 11, 1532-9.

Bajaj, S.P., Butkowski, R.J. & Mann, K.G. (1975) Prothrombin fragments. Ca2+ binding and activation kinetics. J Biol Chem, 250, 2150-6.

Barhoover, M.A., Orban, T., Bukys, M.A. & Kalafatis, M. (2008a) Cooperative regulation of the activity of factor Xa within prothrombinase by discrete amino acid regions from factor Va heavy chain. Biochemistry, 47, 12835-43.

Barhoover, M.A., Orban, T., Beck, D.O., Bukys, M.A. & Kalafatis, M. (2008b) Contribution of amino acid region 334-335 from factor Va heavy chain to the catalytic efficiency of prothrombinase. Biochemistry, 47, 6840-50.

209

Barton, P.G., Jackson, C.M. & Hanahan, D.J. (1967) Relationship between factor V and activated factor X in the generation of prothrombinase. Nature, 214, 923-4.

Barton, P.G., Yin, E.T. & Wessler, S. (1970) Reactions of activated factor X-phosphatide mixtures in vitro and in vivo. J Lipid Res, 11, 87-95.

Beck, D.O., Bukys, M.A., Singh, L.S., Szabo, K.A. & Kalafatis, M. (2004) The contribution of amino acid region ASP695-TYR698 of factor V to procofactor activation and factor Va function. J Biol Chem, 279, 3084-95.

Bertina, R.M., Koeleman, B.P., Koster, T., Rosendaal, F.R., Dirven, R.J., de Ronde, H., van der Velden, P.A. & Reitsma, P.H. (1994) Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature, 369, 64-7.

Bloom, J.W., Nesheim, M.E. & Mann, K.G. (1979) Phospholipid-binding properties of bovine factor V and factor Va. Biochemistry, 18, 4419-25.

Borchgrevink, C.F. & Owren, P.A. (1961) The hemostatic effect of normal platelets in hemophilia and factor V deficiency. The importance of clotting factors adsorbed on platelets for normal hemostasis. Acta Med Scand, 170, 375-83.

Bouchard, B.A., Shatos, M.A. & Tracy, P.B. (1997a) Human brain pericytes differentially regulate expression of procoagulant enzyme complexes comprising the extrinsic pathway of blood coagulation. Arterioscler Thromb Vasc Biol, 17, 1-9.

Bouchard, B.A., Butenas, S., Mann, K.G. & Tracy, P.B. (ed.^eds) (2007) Interactions between platelets and the coagulation system., 2nd edn. Academic Press, San Diego, CA.

Bouchard, B.A., Catcher, C.S., Thrash, B.R., Adida, C. & Tracy, P.B. (1997b) Effector cell protease receptor-1, a platelet activation-dependent membrane protein, regulates prothrombinase-catalyzed thrombin generation. J Biol Chem, 272, 9244-51.

Bouchard, B.A., Williams, J.L., Meisler, N.T., Long, M.W. & Tracy, P.B. (2005) Endocytosis of plasma-derived factor V by megakaryocytes occurs via a clathrin- dependent, specific membrane binding event. J Thromb Haemost, 3, 541-51.

Bouchard, B.A., Meisler, N.T., Nesheim, M.E., Liu, C.X., Strickland, D.K. & Tracy, P.B. (2008) A unique function for LRP-1: a component of a two-receptor system mediating specific endocytosis of plasma-derived factor V by megakaryocytes. J Thromb Haemost, 6, 638-44.

210

Bouchard, B.A., Chapin, J., Brummel-Ziedins, K.E., Durda, P., Key, N.S. & Tracy, P.B. (2015) Platelets and platelet-derived factor Va confer hemostatic competence in complete factor V deficiency. Blood, 125, 3647-50.

Bradford, H.N., Annamalai, A., Doshi, K. & Colman, R.W. (1988) Factor V is activated and cleaved by platelet calpain: comparison with thrombin proteolysis. Blood, 71, 388-94.

Brandstetter, H., Kuhne, A., Bode, W., Huber, R., von der Saal, W., Wirthensohn, K. & Engh, R.A. (1996) X-ray structure of active site-inhibited clotting factor Xa. Implications for drug design and substrate recognition. J Biol Chem, 271, 29988- 92.

Breederveld, K., Giddings, J.C., ten Cate, J.W. & Bloom, A.L. (1975) The localization of factor V within normal human platelets and the demonstration of a platelet-factor V antigen in congenital factor V deficiency. Br J Haematol, 29, 405-12.

Bruin, T., Sturk, A., ten Cate, J.W. & Cath, M. (1987) The function of the human factor V carbohydrate moiety in blood coagulation. Eur J Biochem, 170, 305-10.

Butkowski, R.J., Elion, J., Downing, M.R. & Mann, K.G. (1977) Primary structure of human prethrombin 2 and alpha-thrombin. J Biol Chem, 252, 4942-57.

Camire, R.M., Kalafatis, M. & Tracy, P.B. (1998a) Proteolysis of factor V by cathepsin G and elastase indicates that cleavage at Arg1545 optimizes cofactor function by facilitating factor Xa binding. Biochemistry, 37, 11896-906.

Camire, R.M., Pollak, E.S., Kaushansky, K. & Tracy, P.B. (1998b) Secretable human platelet-derived factor V originates from the plasma pool. Blood, 92, 3035-41.

Camire, R.M., Kalafatis, M., Simioni, P., Girolami, A. & Tracy, P.B. (1998c) Platelet- derived factor Va/Va Leiden cofactor activities are sustained on the surface of activated platelets despite the presence of activated protein C. Blood, 91, 2818-29.

Camire, R.M., Kalafatis, M., Cushman, M., Tracy, R.P., Mann, K.G. & Tracy, P.B. (1995) The mechanism of inactivation of human platelet factor Va from normal and activated protein C-resistant individuals. J Biol Chem, 270, 20794-800.

Chan, W.P., Lee, C.K., Kwong, Y.L., Lam, C.K. & Liang, R. (1998) A novel mutation of Arg306 of factor V gene in Hong Kong Chinese. Blood, 91, 1135-9.

Chen, L., Yang, L. & Rezaie, A.R. (2003) Proexosite-1 on prothrombin is a factor Va- dependent recognition site for the prothrombinase complex. Journal of Biological Chemistry, 278, 27564-9.

211

Chesney, C.M., Pifer, D. & Colman, R.W. (1981) Subcellular localization and secretion of factor V from human platelets. Proc Natl Acad Sci U S A, 78, 5180-4.

Chiu, H.C., Schick, P.K. & Colman, R.W. (1985) Biosynthesis of factor V in isolated guinea pig megakaryocytes. J Clin Invest, 75, 339-46.

Conlon, S.J., Camire, R. M. et al. (1997) Cleavage of platelet-derived factor Va by plasmin results in increased and sustained cofactor activity on the thrombin-activated platelet surface. Thromb Haemost, 77, 2507a.

Cote, H.C., Bajzar, L., Stevens, W.K., Samis, J.A., Morser, J., MacGillivray, R.T. & Nesheim, M.E. (1997) Functional characterization of recombinant human meizothrombin and Meizothrombin(desF1). Thrombomodulin-dependent activation of protein C and thrombin-activatable fibrinolysis inhibitor (TAFI), platelet aggregation, antithrombin-III inhibition. J Biol Chem, 272, 6194-200.

Dahlback, B. (1985) Ultrastructure of human coagulation factor V. J Biol Chem, 260, 1347- 9.

Dahlback, B. (1986) Inhibition of protein Ca cofactor function of human and bovine protein S by C4b-binding protein. J Biol Chem, 261, 12022-7.

Dahlback, B. (1994) Inherited resistance to activated protein C, a major cause of venous thrombosis, is due to a mutation in the factor V gene. Haemostasis, 24, 139-51.

Dahlback, B. (1997) Resistance to activated protein C as risk factor for thrombosis: molecular mechanisms, laboratory investigation, and clinical management. Semin Hematol, 34, 217-34.

Dahlback, B., Carlsson, M. & Svensson, P.J. (1993) Familial thrombophilia due to a previously unrecognized mechanism characterized by poor anticoagulant response to activated protein C: prediction of a cofactor to activated protein C. Proc Natl Acad Sci U S A, 90, 1004-8.

Dale, G.L., Friese, P., Batar, P., Hamilton, S.F., Reed, G.L., Jackson, K.W., Clemetson, K.J. & Alberio, L. (2002) Stimulated platelets use serotonin to enhance their retention of procoagulant proteins on the cell surface. Nature, 415, 175-9.

Degen, S.J., MacGillivray, R.T. & Davie, E.W. (1983) Characterization of the complementary deoxyribonucleic acid and gene coding for human prothrombin. Biochemistry, 22, 2087-97.

212

Deguchi, H., Takeya, H., Gabazza, E.C., Nishioka, J. & Suzuki, K. (1997) Prothrombin kringle 1 domain interacts with factor Va during the assembly of prothrombinase complex. The Biochemical journal, 321 ( Pt 3), 729-735.

Doolittle, R.F., Chen, R. & Lau, F. (1971) Hybrid fibrin: proof of the intermolecular nature of - crosslinking units. Biochem Biophys Res Commun, 44, 94-100.

Dormann, D., Kardoeus, J., Zimmermann, R.E. & Kehrel, B. (1998) Flow cytometric analysis of agonist-induced annexin V, factor Va and factor Xa binding to human platelets. Platelets, 9, 171-7.

Downing, M.R., Butkowski, R.J., Clark, M.M. & Mann, K.G. (1975) Human prothrombin activation. J Biol Chem, 250, 8897-906.

Doyle, M.F. & Mann, K.G. (1990) Multiple active forms of thrombin. IV. Relative activities of meizothrombins. J Biol Chem, 265, 10693-701.

Egan, J.O., Kalafatis, M. & Mann, K.G. (1997) The effect of Arg306-->Ala and Arg506-- >Gln substitutions in the inactivation of recombinant human factor Va by activated protein C and protein S. Protein Sci, 6, 2016-27.

Erdogan, E., Bukys, M.A., Orfeo, T., Mann, K.G. & Kalafatis, M. (2007) Identification of an inactivating cleavage site for alpha-thrombin on the heavy chain of factor Va. Thromb Haemost, 98, 998-1006.

Esmon, C. (2000) The protein C pathway. Crit Care Med, 28, S44-8.

Esmon, C.T. (1979) The subunit structure of thrombin-activated factor V. Isolation of activated factor V, separation of subunits, and reconstitution of biological activity. J Biol Chem, 254, 964-73.

Esmon, C.T. & Jackson, C.M. (1974) The conversion of prothrombin to thrombin. III. The factor Xa-catalyzed activation of prothrombin. J Biol Chem, 249, 7782-90.

Esmon, C.T., Owen, W.G. & Jackson, C.M. (1974a) The conversion of prothrombin to thrombin. V. The activation of prothrombin by factor Xa in the presence of phospholipid. J Biol Chem, 249, 7798-807.

Esmon, C.T., Owen, W.G. & Jackson, C.M. (1974b) The conversion of prothrombin to thrombin. II. Differentiation between thrombin- and factor Xa-catalyzed proteolyses. J Biol Chem, 249, 606-11.

213

Fager, A.M., Wood, J.P., Bouchard, B.A., Feng, P. & Tracy, P.B. (2010) Properties of procoagulant platelets: defining and characterizing the subpopulation binding a functional prothrombinase. Arterioscler Thromb Vasc Biol, 30, 2400-7.

Fernandez, J.A., Hackeng, T.M., Kojima, K. & Griffin, J.H. (1997) The carbohydrate moiety of factor V modulates inactivation by activated protein C. Blood, 89, 4348- 54.

Ferry, J.D. (1952) The Mechanism of Polymerization of Fibrinogen. Proc Natl Acad Sci U S A, 38, 566-9.

Ferry, J.D., Shulman, S. & Foster, J.F. (1952) The conversion of fibrinogen to fibrin. IX. Further flow birefringence studies on inhibited clotting systems. Arch Biochem Biophys, 39, 387-94.

Foster, W.B., Nesheim, M.E. & Mann, K.G. (1983) The factor Xa-catalyzed activation of factor V. J Biol Chem, 258, 13970-7.

Francis, R.T., McDonagh, J. & Mann, K.G. (1986) Factor V is a substrate for the transamidase factor XIIIa. J Biol Chem, 261, 9787-92.

Gailani, D. & Broze, G.J., Jr. (1991) Factor XI activation in a revised model of blood coagulation. Science, 253, 909-12.

Gailani, D. & Renne, T. (2007) Intrinsic pathway of coagulation and arterial thrombosis. Arterioscler Thromb Vasc Biol, 27, 2507-13.

Gawaz, M., Langer, H. & May, A.E. (2005) Platelets in inflammation and atherogenesis. J Clin Invest, 115, 3378-84.

Gitschier, J., Wood, W.I., Goralka, T.M., Wion, K.L., Chen, E.Y., Eaton, D.H., Vehar, G.A., Capon, D.J. & Lawn, R.M. (1984) Characterization of the human factor VIII gene. Nature, 312, 326-30.

Gould, W.R., Silveira, J.R. & Tracy, P.B. (2004) Unique in vivo modifications of coagulation factor V produce a physically and functionally distinct platelet-derived cofactor: characterization of purified platelet-derived factor V/Va. J Biol Chem, 279, 2383-93.

Guinto, E.R. & Esmon, C.T. (1982) Formation of a calcium-binding site on bovine activated factor V following recombination of the isolated subunits. J Biol Chem, 257, 10038-43.

214

Guinto, E.R., Esmon, C.T., Mann, K.G. & MacGillivray, R.T. (1992) The complete cDNA sequence of bovine coagulation factor V. J Biol Chem, 267, 2971-8.

Gumprecht, J.G. & Colman, R.W. (1975) Roles of sialic acid in the function of bovine factor V. Arch Biochem Biophys, 169, 278-86.

Hackeng, T.M., Tans, G., Koppelman, S.J., de Groot, P.G., Rosing, J. & Bouma, B.N. (1996) Protein C activation on endothelial cells by prothrombin activation products generated in situ: meizothrombin is a better protein C activator than alpha- thrombin. Biochem J, 319 ( Pt 2), 399-405.

Handagama, P., Rappolee, D.A., Werb, Z., Levin, J. & Bainton, D.F. (1990) Platelet alpha- granule fibrinogen, albumin, and immunoglobulin G are not synthesized by rat and mouse megakaryocytes. J Clin Invest, 86, 1364-8.

Handagama, P.J. & Bainton, D.F. (1989) Incorporation of a circulating protein into alpha granules of megakaryocytes. Blood Cells, 15, 59-72.

Harding, C., Heuser, J. & Stahl, P. (1983) Receptor-mediated endocytosis of transferrin and recycling of the transferrin receptor in rat reticulocytes. J Cell Biol, 97, 329- 39.

Harrison, P. & Cramer, E.M. (1993) Platelet alpha-granules. Blood Rev, 7, 52-62.

Harrison, P., Wilbourn, B., Debili, N., Vainchenker, W., Breton-Gorius, J., Lawrie, A.S., Masse, J.M., Savidge, G.F. & Cramer, E.M. (1989) Uptake of plasma fibrinogen into the alpha granules of human megakaryocytes and platelets. J Clin Invest, 84, 1320-4.

Haynes, M.L., Beth, A.B., Paula, B.T. & Kenneth, G.M. (2012) Prothrombin Activation by Platelet-associated Prothrombinase Proceeds through the Prethrombin-2 Pathway via a Concerted Mechanism. Journal of Biological Chemistry, 287, 38647-38655.

Heeb, M.J., Kojima, Y., Greengard, J.S. & Griffin, J.H. (1995) Activated protein C resistance: molecular mechanisms based on studies using purified Gln506-factor V. Blood, 85, 3405-11.

Heeb, M.J., Rehemtulla, A., Moussalli, M., Kojima, Y. & Kaufman, R.J. (1999a) Importance of individual activated protein C cleavage site regions in coagulation factor V for factor Va inactivation and for factor Xa activation. Eur J Biochem, 260, 64-75.

215

Heeb, M.J., Kojima, Y., Rosing, J., Tans, G. & Griffin, J.H. (1999b) C-terminal residues 621-635 of protein S are essential for binding to factor Va. J Biol Chem, 274, 36187-92.

Heldebrant, C.M., Noyes, C., Kingdon, H.S. & Mann, K.G. (1973a) The activation of prothrombin. 3. The partial amino acid sequences at the amino terminal of prothrombin and the intermediates of activation. Biochem Biophys Res Commun, 54, 155-60.

Heldebrant, C.M., Butkowski, R.J., Bajaj, S.P. & Mann, K.G. (1973b) The activation of prothrombin. II. Partial reactions, physical and chemical characterization of the intermediates of activation. J Biol Chem, 248, 7149-63.

Hemker, H.C. & Kahn, M.J. (1967) Reaction sequence of blood coagulation. Nature, 215, 1201-2.

Hemker, H.C., Esnouf, M.P., Hemker, P.W., Swart, A.C. & Macfarlane, R.G. (1967) Formation of prothrombin converting activity. Nature, 215, 248-51.

Herbert, F.K. (1940) The estimation of prothrombin in human plasma. Biochem J, 34, 1554-68.

Higgins, D.L. & Mann, K.G. (1983) The interaction of bovine factor V and factor V- derived peptides with phospholipid vesicles. J Biol Chem, 258, 6503-8.

Hockin, M.F., Kalafatis, M., Shatos, M. & Mann, K.G. (1997) Protein C activation and factor Va inactivation on human umbilical vein endothelial cells. Arterioscler Thromb Vasc Biol, 17, 2765-75.

Hoekema, L., Nicolaes, G.A., Hemker, H.C., Tans, G. & Rosing, J. (1997) Human factor Va1 and factor Va2: properties in the procoagulant and anticoagulant pathways. Biochemistry, 36, 3331-5.

Hortin, G., Folz, R., Gordon, J.I. & Strauss, A.W. (1986) Characterization of sites of tyrosine sulfation in proteins and criteria for predicting their occurrence. Biochem Biophys Res Commun, 141, 326-33.

Hortin, G.L. (1990) Sulfation of tyrosine residues in coagulation factor V. Blood, 76, 946- 52.

Janeway, C.M., Rivard, G.E., Tracy, P.B. & Mann, K.G. (1996) Factor V Quebec revisited. Blood, 87, 3571-8.

216

Jenny, R.J., Pittman, D.D., Toole, J.J., Kriz, R.W., Aldape, R.A., Hewick, R.M., Kaufman, R.J. & Mann, K.G. (1987) Complete cDNA and derived amino acid sequence of human factor V. Proc Natl Acad Sci U S A, 84, 4846-50.

Jobe, S.M., Leo, L., Eastvold, J.S., Dickneite, G., Ratliff, T.L., Lentz, S.R. & Di Paola, J. (2005) Role of FcRgamma and factor XIIIA in coated platelet formation. Blood, 106, 4146-51.

Kahn, M.L., Nakanishi-Matsui, M., Shapiro, M.J., Ishihara, H. & Coughlin, S.R. (1999) Protease-activated receptors 1 and 4 mediate activation of human platelets by thrombin. J Clin Invest, 103, 879-87.

Kahn, M.L., Zheng, Y.W., Huang, W., Bigornia, V., Zeng, D., Moff, S., Farese, R.V., Jr., Tam, C. & Coughlin, S.R. (1998) A dual thrombin receptor system for platelet activation. Nature, 394, 690-4.

Kalafatis, M. (1998) Identification and partial characterization of factor Va heavy chain kinase from human platelets. J Biol Chem, 273, 8459-66.

Kalafatis, M. & Mann, K.G. (1995) Factor V Leiden and thrombophilia. N Engl J Med, 332, 1382-3; author reply 1384.

Kalafatis, M. & Beck, D.O. (2002) Identification of a binding site for blood coagulation factor Xa on the heavy chain of factor Va. Amino acid residues 323-331 of factor V represent an interactive site for activated factor X. Biochemistry, 41, 12715-28.

Kalafatis, M., Rand, M.D. & Mann, K.G. (1994) The mechanism of inactivation of human factor V and human factor Va by activated protein C. J Biol Chem, 269, 31869-80.

Kalafatis, M., Bertina, R.M., Rand, M.D. & Mann, K.G. (1995) Characterization of the molecular defect in factor VR506Q. J Biol Chem, 270, 4053-7.

Kalafatis, M., Rand, M.D., Jenny, R.J., Ehrlich, Y.H. & Mann, K.G. (1993) Phosphorylation of factor Va and factor VIIIa by activated platelets. Blood, 81, 704-19.

Kane, W.H. & Majerus, P.W. (1981) Purification and characterization of human coagulation factor V. J Biol Chem, 256, 1002-7.

Kane, W.H. & Davie, E.W. (1986) Cloning of a cDNA coding for human factor V, a blood coagulation factor homologous to factor VIII and ceruloplasmin. Proc Natl Acad Sci U S A, 83, 6800-4.

217

Kane, W.H., Mruk, J.S. & Majerus, P.W. (1982) Activation of coagulation factor V by a platelet protease. J Clin Invest, 70, 1092-100.

Kane, W.H., Devore-Carter, D. & Ortel, T.L. (1990) Expression and characterization of recombinant human factor V and a mutant lacking a major portion of the connecting region. Biochemistry, 29, 6762-8.

Kane, W.H., Lindhout, M.J., Jackson, C.M. & Majerus, P.W. (1980) Factor Va-dependent binding of factor Xa to human platelets. J Biol Chem, 255, 1170-4.

Kane, W.H., Ichinose, A., Hagen, F.S. & Davie, E.W. (1987) Cloning of cDNAs coding for the heavy chain region and connecting region of human factor V, a blood coagulation factor with four types of internal repeats. Biochemistry, 26, 6508-14.

Keller, F.G., Ortel, T.L., Quinn-Allen, M.A. & Kane, W.H. (1995) Thrombin-catalyzed activation of recombinant human factor V. Biochemistry, 34, 4118-24.

Kim, S.W., Ortel, T.L., Quinn-Allen, M.A., Yoo, L., Worfolk, L., Zhai, X., Lentz, B.R. & Kane, W.H. (1999) Partial glycosylation at asparagine-2181 of the second C-type domain of human factor V modulates assembly of the prothrombinase complex. Biochemistry, 38, 11448-54.

Kisiel, W. & Hanahan, D.J. (1973) Purification and characterization of human Factor II. Biochim Biophys Acta, 304, 103-13.

Kojima, Y., Heeb, M.J., Gale, A.J., Hackeng, T.M. & Griffin, J.H. (1998) Binding site for blood coagulation factor Xa involving residues 311-325 in factor Va. J Biol Chem, 273, 14900-5.

Koppaka, V., Talbot, W.F., Zhai, X. & Lentz, B.R. (1997) Roles of factor Va heavy and light chains in protein and lipid rearrangements associated with the formation of a bovine factor Va-membrane complex. Biophys J, 73, 2638-52.

Kotkow, K.J., Deitcher, S.R., Furie, B. & Furie, B.C. (1995) The second kringle domain of prothrombin promotes factor Va-mediated prothrombin activation by prothrombinase. Journal of Biological Chemistry, 270, 4551-4557.

Krieg, U.C., Isaacs, B.S., Yemul, S.S., Esmon, C.T., Bayley, H. & Johnson, A.E. (1987) Interaction of blood coagulation factor Va with phospholipid vesicles examined by using lipophilic photoreagents. Biochemistry, 26, 103-9.

Krishnaswamy, S. (1990) Prothrombinase complex assembly. Contributions of protein- protein and protein-membrane interactions toward complex formation. J Biol Chem, 265, 3708-18.

218

Krishnaswamy, S. & Mann, K.G. (1988) The binding of factor Va to phospholipid vesicles. J Biol Chem, 263, 5714-23.

Krishnaswamy, S. & Walker, R.K. (1997) Contribution of the prothrombin fragment 2 domain to the function of factor Va in the prothrombinase complex. Biochemistry, 36, 3319-30.

Krishnaswamy, S., Mann, K.G. & Nesheim, M.E. (1986) The prothrombinase-catalyzed activation of prothrombin proceeds through the intermediate meizothrombin in an ordered, sequential reaction. J Biol Chem, 261, 8977-84.

Krishnaswamy, S., Jones, K.C. & Mann, K.G. (1988) Prothrombinase complex assembly. Kinetic mechanism of enzyme assembly on phospholipid vesicles. J Biol Chem, 263, 3823-34.

Krishnaswamy, S., Russell, G.D. & Mann, K.G. (1989) The reassociation of factor Va from its isolated subunits. J Biol Chem, 264, 3160-8.

Krishnaswamy, S., Church, W.R., Nesheim, M.E. & Mann, K.G. (1987) Activation of human prothrombin by human prothrombinase. Influence of factor Va on the reaction mechanism. J Biol Chem, 262, 3291-9.

Kurachi, K. & Davie, E.W. (1977) Activation of human factor XI (plasma thromboplastin antecedent) by factor XIIa (activated Hageman factor). Biochemistry, 16, 5831-9.

Lampe, P.D., Pusey, M.L., Wei, G.J. & Nelsestuen, G.L. (1984) Electron microscopy and hydrodynamic properties of blood clotting factor V and activation fragments of factor V with phospholipid vesicles. J Biol Chem, 259, 9959-64.

Larson, P.J., Camire, R.M., Wong, D., Fasano, N.C., Monroe, D.M., Tracy, P.B. & High, K.A. (1998) Structure/function analyses of recombinant variants of human factor Xa: factor Xa incorporation into prothrombinase on the thrombin-activated platelet surface is not mimicked by synthetic phospholipid vesicles. Biochemistry, 37, 5029-38.

Lechtenberg, B.C., Murray-Rust, T.A., Johnson, D.J., Adams, T.E., Krishnaswamy, S., Camire, R.M. & Huntington, J.A. (2013) Crystal structure of the prothrombinase complex from the venom of Pseudonaja textilis. Blood, 122, 2777-83.

Lecompte, M.F., Krishnaswamy, S., Mann, K.G., Nesheim, M.E. & Gitler, C. (1987) Membrane penetration of bovine factor V and Va detected by labeling with 5- iodonaphthalene-1-azide. J Biol Chem, 262, 1935-7.

219

Lee, C.D. & Mann, K.G. (1989) Activation/inactivation of human factor V by plasmin. Blood, 73, 185-90.

Lee, C.J., Lin, P., Chandrasekaran, V., Duke, R.E., Everse, S.J., Perera, L. & Pedersen, L.G. (2008) Proposed structural models of human factor Va and prothrombinase. J Thromb Haemost, 6, 83-9.

Lewandrowski, U., Moebius, J., Walter, U. & Sickmann, A. (2006) Elucidation of N- glycosylation sites on human platelet proteins: a glycoproteomic approach. Mol Cell Proteomics, 5, 226-33.

Liu, T., Qian, W.J., Gritsenko, M.A. & Camp, D.G. (2005) Human plasma N- glycoproteome analysis by immunoaffinity subtraction, hydrazide chemistry, and mass spectrometry. Journal of Proteome Res.,6, 2070-80.

Loeliger, E.A., Hensen, A., Mattern, M.J., Veltkamp, J.J., Bruning, P.F. & Hemker, H.C. (1967) Treatment of haemophilia B with purified Factor IX (PPSB). Folia Med Neerl, 10, 112-25.

Lundblad, R.L., Kingdon, H.S. & Mann, K.G. (1976) Thrombin. Methods Enzymol, 45, 156-76.

Macedo-Ribeiro, S., Bode, W., Huber, R., Quinn-Allen, M.A., Kim, S.W., Ortel, T.L., Bourenkov, G.P., Bartunik, H.D., Stubbs, M.T., Kane, W.H. & Fuentes-Prior, P. (1999) Crystal structures of the membrane-binding C2 domain of human coagulation factor V. Nature, 402, 434-9.

Majumder, R., Quinn-Allen, M.A., Kane, W.H. & Lentz, B.R. (2005) The phosphatidylserine binding site of the factor Va C2 domain accounts for membrane binding but does not contribute to the assembly or activity of a human factor Xa- factor Va complex. Biochemistry, 44, 711-8.

Majumder, R., Quinn-Allen, M.A., Kane, W.H. & Lentz, B.R. (2008) A phosphatidylserine binding site in factor Va C1 domain regulates both assembly and activity of the prothrombinase complex. Blood, 112, 2795-802.

Makris, M., Rosendaal, F.R. & Preston, F.E. (1997) Familial thrombophilia: genetic risk factors and management. J Intern Med Suppl, 740, 9-15.

Mann, K.G., Nesheim, M.E., Church, W.R., Haley, P. & Krishnaswamy, S. (1990) Surface- dependent reactions of the vitamin K-dependent enzyme complexes. Blood, 76, 1- 16.

220

Mattii, R., Ambrus, J.L., Sokal, J.E. & Mink, I. (1964) Production of Members of the Blood Coagulation and Fibrinolysin Systems by the Isolated Perfused Liver. Proc Soc Exp Biol Med, 116, 69-72.

Mazzorana, M., Baffet, G., Kneip, B., Launois, B. & Guguen-Guillouzo, C. (1991) Expression of coagulation factor V gene by normal adult human hepatocytes in primary culture. Br J Haematol, 78, 229-35.

McDuffie, F.C., Giffin, C., Niedringhaus, R., Mann, K.G., Owen, C.A., Jr., Bowie, E.J., Peterson, J., Clark, G. & Hunder, G.G. (1979) Prothrombin, thrombin and prothrombin fragments in plasma of normal individuals and of patients with laboratory evidence of disseminated intravascular coagulation. Thromb Res, 16, 759-73.

Mettine, H.A.B. & Camire, M.R. (2012) A Bipartite Autoinhibitory Region within the B- domain Suppresses Function in Factor V. Journal of Biological Chemistry, 287, 26342-26351.

Meyer, K., Eernisse, J.G., Veltkamp, J.J., Hemker, H.C. & Loeliger, E.A. (1967) Treatment of haemophilia A with purified factor 8 obtained from human plasma by cryoprecipitation. Folia Med Neerl, 10, 49-60.

Miletich, J.P., Majerus, D.W. & Majerus, P.W. (1978a) Patients with congenital factor V deficiency have decreased factor Xa binding sites on their platelets. J Clin Invest, 62, 824-31.

Miletich, J.P., Jackson, C.M. & Majerus, P.W. (1978b) Properties of the factor Xa binding site on human platelets. J Biol Chem, 253, 6908-16.

Monkovic, D.D. & Tracy, P.B. (1990a) Activation of human factor V by factor Xa and thrombin. Biochemistry, 29, 1118-28.

Monkovic, D.D. & Tracy, P.B. (1990b) Functional characterization of human platelet- released factor V and its activation by factor Xa and thrombin. J Biol Chem, 265, 17132-40.

Mosesson, M.W., Church, W.R., DiOrio, J.P., Krishnaswamy, S., Mann, K.G., Hainfeld, J.F. & Wall, J.S. (1990a) Structural model of factors V and Va based on scanning transmission electron microscope images and mass analysis. J Biol Chem, 265, 8863-8.

Mosesson, M.W., Fass, D.N., Lollar, P., DiOrio, J.P., Parker, C.G., Knutson, G.J., Hainfeld, J.F. & Wall, J.S. (1990b) Structural model of porcine factor VIII and

221

factor VIIIa molecules based on scanning transmission electron microscope (STEM) images and STEM mass analysis. J Clin Invest, 85, 1983-90.

Nelsestuen, G.L., Zytkovicz, T.H. & Howard, J.B. (1974) The mode of action of vitamin K. Identification of gamma-carboxyglutamic acid as a component of prothrombin. J Biol Chem, 249, 6347-50.

Nemerson, Y. (1988) Tissue factor and hemostasis. Blood, 71, 1-8.

Nesheim, M.E. & Mann, K.G. (1979) Thrombin-catalyzed activation of single chain bovine factor V. J Biol Chem, 254, 1326-34.

Nesheim, M.E., Taswell, J.B. & Mann, K.G. (1979a) The contribution of bovine Factor V and Factor Va to the activity of prothrombinase. J Biol Chem, 254, 10952-62.

Nesheim, M.E., Myrmel, K.H., Hibbard, L. & Mann, K.G. (1979b) Isolation and characterization of single chain bovine factor V. J Biol Chem, 254, 508-17.

Nesheim, M.E., Kettner, C., Shaw, E. & Mann, K.G. (1981) Cofactor dependence of factor Xa incorporation into the prothrombinase complex. J Biol Chem, 256, 6537-40.

Nesheim, M.E., Furmaniak-Kazmierczak, E., Henin, C. & Cote, G. (1993) On the existence of platelet receptors for factor V(a) and factor VIII(a). Thromb Haemost, 70, 80-6.

Nesheim, M.E., Nichols, W.L., Cole, T.L., Houston, J.G., Schenk, R.B., Mann, K.G. & Bowie, E.J. (1986) Isolation and study of an acquired inhibitor of human coagulation factor V. J Clin Invest, 77, 405-15.

Nicholson, A.C., Nachman, R.L., Altieri, D.C., Summers, B.D., Ruf, W., Edgington, T.S. & Hajjar, D.P. (1996) Effector cell protease receptor-1 is a vascular receptor for coagulation factor Xa. J Biol Chem, 271, 28407-13.

Nicolaes, G.A., Villoutreix, B.O. & Dahlback, B. (1999) Partial glycosylation of Asn2181 in human factor V as a cause of molecular and functional heterogeneity. Modulation of glycosylation efficiency by mutagenesis of the consensus sequence for N-linked glycosylation. Biochemistry, 38, 13584-91.

Nicolaes, G.A., Tans, G., Thomassen, M.C., Hemker, H.C., Pabinger, I., Varadi, K., Schwarz, H.P. & Rosing, J. (1995) Peptide bond cleavages and loss of functional activity during inactivation of factor Va and factor VaR506Q by activated protein C. J Biol Chem, 270, 21158-66.

Nurden, A.T. (2011) Platelets, inflammation and tissue regeneration. Thromb Haemost, 105 Suppl 1, S13-33.

222

Nurden, A.T., Nurden, P., Sanchez, M., Andia, I. & Anitua, E. (2008) Platelets and wound healing. Front Biosci, 13, 3532-48.

Oates, A.M. & Salem, H.H. (1987) The regulation of human factor V by a neutrophil protease. Blood, 70, 846-51.

Odegaard, B. & Mann, K. (1987) Proteolysis of factor Va by factor Xa and activated protein C. J Biol Chem, 262, 11233-8.

Oliver, J.A., Monroe, D.M., Roberts, H.R. & Hoffman, M. (1999) Thrombin activates factor XI on activated platelets in the absence of factor XII. Arterioscler Thromb Vasc Biol, 19, 170-7.

Omar, M.N. & Mann, K.G. (1987) Inactivation of factor Va by plasmin. J Biol Chem, 262, 9750-5.

Osterud, B. & Rapaport, S.I. (1977) Activation of factor IX by the reaction product of tissue factor and factor VII: additional pathway for initiating blood coagulation. Proc Natl Acad Sci U S A, 74, 5260-4.

Osterud, B., Rapaport, S.I. & Lavine, K.K. (1977) Factor V activity of platelets: evidence for an activated factor V molecule and for a platelet activator. Blood, 49, 819-34.

Owen, C.A., Jr. & Bowie, E.J. (1977) Generation of coagulation factors V, XI, and XII by the isolated rat liver. Haemostasis, 6, 205-12.

Owren, P.A. (1947) Parahaemophilia; haemorrhagic diathesis due to absence of a previously unknown clotting factor. Lancet, 1, 446-8.

Peng, W., Quinn-Allen, M.A. & Kane, W.H. (2005) Mutation of hydrophobic residues in the factor Va C1 and C2 domains blocks membrane-dependent prothrombin activation. J Thromb Haemost, 3, 351-4.

Pittman, D.D., Tomkinson, K.N., Michnick, D., Selighsohn, U. & Kaufman, R.J. (1994) Posttranslational sulfation of factor V is required for efficient thrombin cleavage and activation and for full procoagulant activity. Biochemistry, 33, 6952-9.

Pozzi, N., Chen, Z., Pelc, L.A., Shropshire, D.B. & Di Cera, E. (2014) The linker connecting the two kringles plays a key role in prothrombin activation. Proceedings of the National Academy of Sciences of the United States of America, 111, 7630- 7635.

223

Pozzi, N., Chen, Z., Gohara, D.W., Niu, W., Heyduk, T. & Di Cera, E. (2013) Crystal structure of prothrombin reveals conformational flexibility and mechanism of activation. J Biol Chem, 288, 22734-44.

Pratt, K.P., Shen, B.W., Takeshima, K., Davie, E.W., Fujikawa, K. & Stoddard, B.L. (1999) Structure of the C2 domain of human factor VIII at 1.5 A resolution. Nature, 402, 439-42.

Pusey, M.L. & Nelsestuen, G.L. (1984) Membrane binding properties of blood coagulation Factor V and derived peptides. Biochemistry, 23, 6202-10.

Pusey, M.L., Mayer, L.D., Wei, G.J., Bloomfield, V.A. & Nelsestuen, G.L. (1982) Kinetic and hydrodynamic analysis of blood clotting factor V-membrane binding. Biochemistry, 21, 5262-9.

Radcliffe, R. & Nemerson, Y. (1975) Activation and control of factor VII by activated factor X and thrombin. Isolation and characterization of a single chain form of factor VII. J Biol Chem, 250, 388-95.

Rand, M.D., Kalafatis, M. & Mann, K.G. (1994) Platelet coagulation factor Va: the major secretory platelet phosphoprotein. Blood, 83, 2180-90.

Rand, M.D., Lock, J.B., van't Veer, C., Gaffney, D.P. & Mann, K.G. (1996) Blood clotting in minimally altered whole blood. Blood, 88, 3432-45.

Robert, K.A., José, A.L. & Michael, C.B. (1997) Molecular mechanisms of platelet adhesion and activation. The International Journal of Biochemistry & Cell Biology, 1, 91-105.

Rodgers, G.M., Cong, J.Y., Goll, D.E. & Kane, W.H. (1987) Activation of coagulation factor V by calcium-dependent proteinase. Biochim Biophys Acta, 929, 263-70.

Rosing, J., Zwaal, R.F. & Tans, G. (1986) Formation of meizothrombin as intermediate in factor Xa-catalyzed prothrombin activation. J Biol Chem, 261, 4224-8.

Rosing, J., Tans, G., Govers-Riemslag, J.W., Zwaal, R.F. & Hemker, H.C. (1980) The role of phospholipids and factor Va in the prothrombinase complex. J Biol Chem, 255, 274-83.

Rosing, J., Bakker, H.M., Thomassen, M.C., Hemker, H.C. & Tans, G. (1993) Characterization of two forms of human factor Va with different cofactor activities. J Biol Chem, 268, 21130-6.

224

Rosing, J., Hoekema, L., Nicolaes, G.A., Thomassen, M.C., Hemker, H.C., Varadi, K., Schwarz, H.P. & Tans, G. (1995) Effects of protein S and factor Xa on peptide bond cleavages during inactivation of factor Va and factor VaR506Q by activated protein C. J Biol Chem, 270, 27852-8.

Rudolph, A.E., Mullane, M.P., Porche-Sorbet, R., Tsuda, S. & Miletich, J.P. (1996) Factor XSt. Louis II. Identification of a glycine substitution at residue 7 and characterization of the recombinant protein. J Biol Chem, 271, 28601-6.

Ruggeri, Z.M. (2009) Platelet adhesion under flow. Microcirculation, 16, 58-83.

Ruppert, E.E., and Carle, K.J. (1983) Morphology of metazoan circulatory systems. Zoomorphology, 103, 193-208.

Salem, H.H., Broze, G.J., Miletich, J.P. & Majerus, P.W. (1983) Human coagulation factor Va is a cofactor for the activation of protein C. Proc Natl Acad Sci U S A, 80, 1584- 8.

Samis, J.A., Garrett, M., Manuel, R.P., Nesheim, M.E. & Giles, A.R. (1997) Human neutrophil elastase activates human factor V but inactivates thrombin-activated human factor V. Blood, 90, 1065-74.

Scandura, J.M., Ahmad, S.S. & Walsh, P.N. (1996) A binding site expressed on the surface of activated human platelets is shared by factor X and prothrombin. Biochemistry, 35, 8890-902.

Seifert, L.W., Wahle-Ritchie, A., et al (2005) Intracellular trafficking of factor V sebsequent to its endocytosis by megakaryocytes. Blood, 106 (11), 1031.

Shi, J., Pipe, S.W., Rasmussen, J.T., Heegaard, C.W. & Gilbert, G.E. (2008) Lactadherin blocks thrombosis and hemostasis in vivo: correlation with platelet phosphatidylserine exposure. J Thromb Haemost, 6, 1167-74.

Silveira, J.R., Kalafatis, M. & Tracy, P.B. (2002) Carbohydrate moieties on the procofactor factor V, but not the derived cofactor factor Va, regulate its inactivation by activated protein C. Biochemistry, 41, 1672-80.

Silveira, J.R., Kalafatis, M., Tracy, P.B. (1997) The influence of the carbohydrate moieties in the inactivation of factor V and factor Va by APC. Blood, 90, (Abstract 3166).

Silverberg, S.A., Nemerson, Y. & Zur, M. (1977) Kinetics of the activation of bovine coagulation factor X by components of the extrinsic pathway. Kinetic behavior of two-chain factor VII in the presence and absence of tissue factor. J Biol Chem, 252, 8481-8.

225

Singh, L.S., Bukys, M.A., Beck, D.O. & Kalafatis, M. (2003) Amino acids Glu323, Tyr324, Glu330, and Val331 of factor Va heavy chain are essential for expression of cofactor activity. J Biol Chem, 278, 28335-45.

Smith, C.M. & Hanahan, D.J. (1976) The activation of factor V by factor Xa or alpha- chymotrypsin and comparison with thrombin and RVV-V action. An improved factor V isolation procedure. Biochemistry, 15, 1830-8.

Solymoss, S., Tucker, M.M. & Tracy, P.B. (1988) Kinetics of inactivation of membrane- bound factor Va by activated protein C. Protein S modulates factor Xa protection. J Biol Chem, 263, 14884-90.

Srivastava, A., Quinn-Allen, M.A., Kim, S.W., Kane, W.H. & Lentz, B.R. (2001) Soluble phosphatidylserine binds to a single identified site in the C2 domain of human factor Va. Biochemistry, 40, 8246-55.

Steen, M., Villoutreix, B.O., Norstrom, E.A., Yamazaki, T. & Dahlback, B. (2002) Defining the factor Xa-binding site on factor Va by site-directed glycosylation. J Biol Chem, 277, 50022-9.

Steen, M., Tran, S., Autin, L., Villoutreix, B.O., Tholander, A.L. & Dahlback, B. (2008) Mapping of the factor Xa binding site on factor Va by site-directed mutagenesis. J Biol Chem, 283, 20805-12.

Stenn, K.S. & Blout, E.R. (1972) Mechanism of bovine prothrombin activation by an insoluble preparation of bovine factor X a (thrombokinase). Biochemistry, 11, 4502-15.

Stoilova-McPhie, S., Parmenter, C.D., Segers, K., Villoutreix, B.O. & Nicolaes, G.A. (2008) Defining the structure of membrane-bound human blood coagulation factor Va. J Thromb Haemost, 6, 76-82.

Subramanian, Y., Rolf, M.M., José, A.F. & John, H.G. (2004) Prothrombin residues 473- 487 contribute to factor Va binding in the prothrombinase complex. The Journal of biological chemistry, 279, 49019-49025.

Sun, H., Yang, T.L., Yang, A., Wang, X. & Ginsburg, D. (2003) The murine platelet and plasma factor V pools are biosynthetically distinct and sufficient for minimal hemostasis. Blood, 102, 2856-61.

Suzuki, K., Dahlback, B. & Stenflo, J. (1982) Thrombin-catalyzed activation of human coagulation factor V. J Biol Chem, 257, 6556-64.

226

Suzuki, K., Stenflo, J., Dahlback, B. & Teodorsson, B. (1983) Inactivation of human coagulation factor V by activated protein C. J Biol Chem, 258, 1914-20.

Tans, G., Janssen-Claessen, T., Hemker, H.C., Zwaal, R.F. & Rosing, J. (1991a) Meizothrombin formation during factor Xa-catalyzed prothrombin activation. Formation in a purified system and in plasma. J Biol Chem, 266, 21864-73.

Tans, G., Rosing, J., Thomassen, M.C., Heeb, M.J., Zwaal, R.F. & Griffin, J.H. (1991b) Comparison of anticoagulant and procoagulant activities of stimulated platelets and platelet-derived microparticles. Blood, 77, 2641-8.

Tans, G., Nicolaes, G.A., Thomassen, M.C., Hemker, H.C., van Zonneveld, A.J., Pannekoek, H. & Rosing, J. (1994) Activation of human factor V by meizothrombin. J Biol Chem, 269, 15969-72.

Thorelli, E., Kaufman, R.J. & Dahlback, B. (1997) Cleavage requirements for activation of factor V by factor Xa. Eur J Biochem, 247, 12-20.

Toole, J.J., Knopf, J.L., Wozney, J.M., Sultzman, L.A., Buecker, J.L., Pittman, D.D., Kaufman, R.J., Brown, E., Shoemaker, C., Orr, E.C. & et al. (1984) Molecular cloning of a cDNA encoding human antihaemophilic factor. Nature, 312, 342-7.

Toso, R. & Camire, R.M. (2004) Removal of B-domain sequences from factor V rather than specific proteolysis underlies the mechanism by which cofactor function is realized. J Biol Chem, 279, 21643-50.

Tracy, P.B. (2001) Role of platelets and leukocytes in coagulation. Hemostasis and Thrombosis. Basic Principles and Clinical Practice (R.W. Colman, et al). Lippincott Williams & Wilkins, Philadelphia, 575-596.

Tracy, P.B. & Mann, K.G. (1983) Prothrombinase complex assembly on the platelet surface is mediated through the 74,000-dalton component of factor Va. Proc Natl Acad Sci U S A, 80, 2380-4.

Tracy, P.B., Nesheim, M.E. & Mann, K.G. (1983) Proteolytic alterations of factor Va bound to platelets. J Biol Chem, 258, 662-9.

Tracy, P.B., Eide, L.L. & Mann, K.G. (1985) Human prothrombinase complex assembly and function on isolated peripheral blood cell populations. J Biol Chem, 260, 2119- 24.

Tracy, P.B., Nesheim, M.E. & Mann, K.G. (1992) Platelet factor Xa receptor. Methods Enzymol, 215, 329-60.

227

Tracy, P.B., Eide, L.L., Bowie, E.J. & Mann, K.G. (1982) Radioimmunoassay of factor V in human plasma and platelets. Blood, 60, 59-63.

Tracy, P.B., Peterson, J.M., Nesheim, M.E., McDuffie, F.C. & Mann, K.G. (1979) Interaction of coagulation factor V and factor Va with platelets. J Biol Chem, 254, 10354-61.

Tracy, P.B., Giles, A.R., Mann, K.G., Eide, L.L., Hoogendoorn, H. & Rivard, G.E. (1984) Factor V (Quebec): a bleeding diathesis associated with a qualitative platelet Factor V deficiency. J Clin Invest, 74, 1221-8.

Turkington, P.T. (1993) Cathepsin G activates human factor V in vitro. Thromb Res, 72, 333-7.

Undas, A., Brummel, K., Musial, J., Mann, K.G. & Szczeklik, A. (2001) Blood coagulation at the site of microvascular injury: effects of low-dose aspirin. Blood, 98, 2423-31. van Dieijen, G., Tans, G., Rosing, J. & Hemker, H.C. (1981) The role of phospholipid and factor VIIIa in the activation of bovine factor X. J Biol Chem, 256, 3433-42. van Wijnen, M., Stam, J.G., van't Veer, C., Meijers, J.C., Reitsma, P.H., Bertina, R.M. & Bouma, B.N. (1996) The interaction of protein S with the phospholipid surface is essential for the activated protein C-independent activity of protein S. Thromb Haemost, 76, 397-403.

Varadi, K., Rosing, J., Tans, G., Pabinger, I., Keil, B. & Schwarz, H.P. (1996) Factor V enhances the cofactor function of protein S in the APC-mediated inactivation of factor VIII: influence of the factor VR506Q mutation. Thromb Haemost, 76, 208- 14.

Vehar, G.A. & Davie, E.W. (1980) Preparation and properties of bovine factor VIII (antihemophilic factor). Biochemistry, 19, 401-10.

Vehar, G.A., Keyt, B., Eaton, D., Rodriguez, H., O'Brien, D.P., Rotblat, F., Oppermann, H., Keck, R., Wood, W.I., Harkins, R.N., Tuddenham, E.G., Lawn, R.M. & Capon, D.J. (1984) Structure of human factor VIII. Nature, 312, 337-42.

Vicic, W.J., Lages, B. & Weiss, H.J. (1980) Release of human platelet factor V activity is induced by both collagen and ADP and is inhibited by aspirin. Blood, 56, 448-55.

Viskup, R.W., Tracy, P.B. & Mann, K.G. (1987) The isolation of human platelet factor V. Blood, 69, 1188-95.

228

Walsh, P.N. (1974) Platelet coagulant activities and hemostasis: a hypothesis. Blood, 43, 597-605.

Weiss, H.J., Lages, B., Zheng, S. & Hayward, C.P. (2001) Platelet factor V New York: a defect in factor V distinct from that in factor V Quebec resulting in impaired prothrombinase generation. Am J Hematol, 66, 130-9.

Whelihan, M.F., Orfeo, T., Gissel, M.T. & Mann, K.G. (2010) Coagulation procofactor activation by factor XIa. J Thromb Haemost, 8, 1532-9.

Williams, W.J., Nelson, D.A. (1990) Examination of the Blood. Hematology, W.J. Williams, E. Beutler, A.J. Erslev and M.A. Litcman, McGraw Hill, 9-23.

Williamson, D., Brown, K., Luddington, R., Baglin, C. & Baglin, T. (1998) Factor V Cambridge: a new mutation (Arg306-->Thr) associated with resistance to activated protein C. Blood, 91, 1140-4.

Wood, J.P., Fager, A.M. & Tracy, P.B. (2008) Platelet-derived factor Va expressed on the surface of the activated platelet is GPI-anchored. Blood, 112, 219a.

Wood, J.P., Silveira, J.R., Maille, N.M., Haynes, L.M. & Tracy, P.B. (2011) Prothrombin activation on the activated platelet surface optimizes expression of procoagulant activity. Blood, 117, 1710-8.

Yegneswaran, S., Mesters, R.M. & Griffin, J.H. (2003) Identification of distinct sequences in human blood coagulation factor Xa and prothrombin essential for substrate and cofactor recognition in the prothrombinase complex. J Biol Chem, 278, 33312-8.

Zaitseva, I., Zaitsev, V.N., Moshkov, K., Bax, B., Ralph, A. & Lindley, P.F. (1996) The X-ray structure of human serum ceruloplasmin at 3.1 A: nature of the copper centres. J Biol Inorg Chem, 4, 15-23.

Zeibdawi, A.R. & Pryzdial, E.L. (2001) Mechanism of factor Va inactivation by plasmin. Loss of A2 and A3 domains from a Ca2+-dependent complex of fragments bound to phospholipid. J Biol Chem, 276, 19929-36.

Zhu, H., Toso, R. & Camire, R.M. (2007) Inhibitory sequences within the B-domain stabilize circulating factor V in an inactive state. J Biol Chem, 282, 15033-9.

229