A Dissertation Entitled Role of Complement Regulatory Protein

A Dissertation

entitled

Role of Complement Regulatory Protein Properdin in Complement Activation on

Platelets and in the Formation of Platelet-Leukocyte Aggregates

Gurpanna Saggu

Submitted to the Graduate Faculty in partial fulfillment of the requirements for the

Doctor of Philosophy Degree in Biomedical Sciences

______Viviana P. Ferreira, D.V.M., Ph.D., Committee Chair

______Stanislaw Stepkowski, D.V.M., Ph.D., Committee Member

______R. Mark Wooten, Ph.D., Committee Member

______Z. Kevin Pan, M.D., Ph.D., Committee Member

______Guillermo Vazquez, Ph.D., Committee Member

______Patricia R. Komuniecki, Ph.D., Dean College of Graduate Studies

The University of Toledo May, 2014

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author. An Abstract of

Role of Complement Regulatory Protein Properdin in Complement Activation on Platelets and in the Formation of Platelet-Leukocyte Aggregates

Gurpanna Saggu

Submitted to the Graduate Faculty in partial fulfillment of the requirements for the Doctor of Philosophy Degree in Biomedical Sciences

University of Toledo,

May 2014

Patients with inflammatory cardiovascular disease have an increased number of circulating activated platelets and platelet-leukocyte aggregates (PLAs), both of which play a central role in the initiation and progression of disease. Activated platelets can activate the complement system on their surface, with potential consequences in vascular inflammation and thrombosis. Properdin, a positive regulator of the alternative pathway (AP) of complement, is produced mainly by stimulated leukocytes. The mechanisms by which properdin participates in complement activation on platelets and in PLA formation remain unknown. We have shown that the physiological forms of human properdin bind directly to activated, but not resting, platelets. The binding of properdin promotes AP complement activation on activated platelets, as measured by C3b and C9 deposition on their surface, by forming novel C3 convertases on the platelet [C3(H2O),Bb]. Removal of surface proteins by treating platelets with a low dose of proteinase K, leads to reduced properdin binding to activated platelets. On the other hand, chondroitin sulfate-A (a glycosaminoglycan that is released by platelets upon activation)

iii increases the binding of properdin to activated platelets by ~4 fold. These results suggest that interaction of properdin with activated platelets may depend on a protein receptor as well as glycosaminoglycans (i.e. proteoglycans). We have also determined that properdin released by PMA-stimulated neutrophils binds to activated platelets. Since activated neutrophils and platelets directly interact with one another in pro-inflammatory microenvironments, the fresh properdin produced by neutrophils would be available to platelets at high concentrations. Using ex- vivo whole blood assays, we show that properdin leads to an increase in PLA formation in TRAP (thrombin receptor activating peptide)-stimulated whole blood and inhibition of properdin leads to decrease in PLA formation. Our data also show that properdin-mediated PLA formation is controlled by complement regulatory protein factor H. Altogether, the results support a role for properdin in the cellular microenvironment, contributing to complement activation on activated platelets and PLA formation, with potential consequences in inflammation pathophysiology.

To my father who encouraged me to take this path, and my mother who assured me

that I was capable of it. To my brother who has been my strength throughout.

Acknowledgements

First and foremost I want to thank my major advisor Dr. Viviana P. Ferreira. It has been an honor to be her first Ph.D. student. I appreciate all her contributions to make my Ph.D. experience productive and stimulating. The dedication she has for her work was contagious and motivational for me, even in the tough times during the pursuit of my Ph.D.

I would like to thank to my committee members, Dr. Mark Wooten, Dr. Stanislaw

Stepkowski, Dr. Kevin Pan and Dr. Guillermo Vazquez, for all their time, constructive criticism and advice they have offered.

I would like to thank Dr. Claudio Cortes who spent long hours training me during the initial phase. I would also like to thank Heather Emch, Dr. Galia Ramirez, Laci

Bloomfield and Adam Blatt for all their support.

I would also like to acknowledge all the faculty, staff and students in the Dept. of

Medical Microbiology and Immunology. They have created a very professional and friendly work environment.

Lastly, I want to thank my parents and brother for supporting me in all my pursuits; and I want to thank my loving, supportive, encouraging, and (not so) patient husband Mithun, whose unconditional support during this Ph.D. is so appreciated.

Contents

Abstract………………………………………………………………………………………………………. iii

Acknowledgements……………………………………………………………………………...... vi

Contents……………………………………………………………………………………………………... vii

List of Tables…………………………………………………………………………………………….... xvi

List of Figures…………………………………………………………………………………………….. xvii

List of Abbreviations………………………………………………………………………………….. xxi

1 Introduction…………………………………………………………………………………...... 1

1.1 The Complement System……………………………………………………………….... 1

1.1.1 Background…………………………………………………………………………….. 1

1.1.2 Activation of the complement system……………………………………. 4

1.1.2.1 Classical Pathway………….……………………………………………... 4

1.1.2.2 Lectin pathway………………………………………………………...... 6

1.1.2.3 Alternative pathway…………………………………………………….. 8

1.1.2.3.1 Complement component C3…………………..………….... 8

1.1.2.3.2 Initiation of the alternative pathway: Formation

of the initial fluid phase C3 convertase……………………………... 9

1.1.2.3.3 Initial C3b deposition and formation of C3b,Bb

convertase………………………………………………………………………. 10

vii

1.1.2.3.4 Role of properdin in amplification of the

alternative pathway………………………………………………………... 11

1.1.2.4 C5 convertase and the terminal complement pathway………………………………………………………………………………….... 12

1.1.2.5 Regulatory proteins that allow the alternative pathway to distinguish between host/self cells (non- activating surfaces) and activating/pathogenic surfaces……….. 14

1.1.2.5.1 Soluble negative regulators of the alternative

pathway…………………………………………………………………………. 14

(a) Factor I……………………..…………………………………. 14

(b) Factor H………………………………………………………. 14

1.1.2.5.2 Membrane-bound negative regulators of the

alternative pathway………………………………………………………… 17

(a) Decay accelerating factor……………………………… 17

(b) CR1……………………………………………………………... 17

(d) CD59……………………………………………………………. 17

1.1.2.6 Properdin: a positive regulatory protein with functions in complement initiation……………………………………...... 18

1.1.2.6.1 Properdin sources…………………………………………...... 18

1.1.2.6.2 Properdin structure…………………………………………... 19

1.1.2.6.3 Role of properdin as an initiator versus stabilizer

of the alternative pathway convertase……………………………... 20

viii

1.1.2.6.4 Binding of properdin to a variety of cell surfaces

and potential binding ligands for properdin……………………... 21

1.1.2.6.5 Importance of separating physiological forms of

properdin (P2-P4) from aggregated (“activated”) properdin

to study the specificity of properdin-target

interactions.……………………………………………………………………. 22

1.1.2.6.6 The role of properdin in the local

microenvironment………………………………………………………….. 24

1.2 Platelets…………………………………………………………………………………………... 27

1.2.1 Background…………………………………………………………………………….. 27

1.2.1.1 Platelet origin………………………………………………………………. 27

1.2.1.2 Platelet structure and function……………………………………. 27

1.2.2 Platelet activation…………………………………………………………………... 30

1.2.2.1 Platelet agonists…………………………………………………………... 31

1.2.2.1.1 Thrombin and thrombin receptor-activating

peptide (TRAP)………..…………………………………………………...... 31

1.2.2.1.2 Arachidonic acid………………………………………………... 32

1.2.2.1.3 Adenosine Diphosphate…………………………………….. 33

1.2.2.2 Platelet granules………………………………………………………….. 34

1.2.3 Interactions between platelets and the complement system… 36

1.3 Platelet-leukocyte aggregates………………………………………………………… 39

1.3.1 Background…………………………………………………………………………….. 39

1.3.2 Mechanisms of platelet-leukocyte aggregate formation……….. 40

1.3.3 Activation of complement on platelets and on leukocytes…….. 41

2 Hypothesis…………………………………………………………………………………………….. 44

3 Specific aims………………………………………………………………………………………….. 45

4 Materials and Methods…………………………………………………………………………. 47

4.1 General Methods……………………………………………………………………………... 47

4.1.1 Buffers…………………………………………………………………………………….. 47

4.1.2 Antibodies………………………………………………………………………………. 48

4.1.3 Other reagents………………………………………………………………………... 50

4.1.4 Serum and additional complement proteins…………..……………... 51

4.1.5 Separation of physiological forms of properdin…………………..... 53

4.1.6 Platelet isolation and activation…………………………………………….. 54

4.2 Specific Methods (separated by the Aim involved)………………………… 56

4.2.1 Methods Aim 1: To determine the molecular mechanisms

involved in alternative pathway activation on activated platelets.... 56

4.2.1.1 Measurement of properdin binding to platelets………………... 56

4.2.1.2 Measurement of platelet activation by properdin……………... 57

4.2.1.3 Measurement of presence of properdin or C3 components

on platelet surface……………………………………………………………………… 57

4.2.1.4 Measurement of properdin binding in presence of

chondroitin sulfate-A or proteinase K……………………………………….... 58

4.2.1.5 Measurement of factor H binding to platelets…………………… 58

4.2.1.6 Measurement of CD46, CD55 and CD59 on platelet surface 59

4.2.1.7 Measurement of alternative pathway complement

activation on platelets (C3b/iC3b and C5b-9 deposition)…………….. 59

4.2.1.8 Separation of C3 and C3(H2O)………………………………………….. 60

4.2.1.9 Measurement of recruitment to the platelet surface of C3

components by properdin and of properdin by C3(H2O)……………… 61

4.2.1.10 Measurement of C3 convertase [C3(H2O),Bb and C3b,Bb]

formation on platelets………………………………………………………………... 62

4.2.1.11 Measurement of properdin binding to platelets in NHS…... 62

4.2.1.12 PMN isolation, activation, and assessment of the ability of

properdin and C3 components from PMN supernatants to bind to

stimulated platelets…………………………………………………………………… 63

4.2.2 Methods Aim 2: To determine the role of properdin in platelet-leukocyte aggregate formation…………………………………………. 64

4.2.2.1 Measurement of the formation of platelet-leukocyte

aggregates in the presence of TRAP and/or properdin………………… 64

4.2.2.2 Measurement of alternative pathway-mediated hemolysis

of rabbit erythrocytes by compstatin (control experiment)…………. 65

4.2.2.3 Measurement of the ability of anti-C3 antibodies to inhibit

binding of properdin to platelets………………………………………………… 65

4.2.2.4 Measurement of formation of platelet-leukocyte

aggregates in the presence of complement inhibitors and/or

properdin………………………………………………………………………………….. 66

4.2.2.5 Assessment of inhibition of C5a-mediated neutrophil

activation by C5aRa………………………………………………………………….... 67

4.2.2.6 Measurement of the formation of platelet-leukocyte

aggregates in the presence of rH19-20 (a competitive inhibitor of

factor H)…………………………………………………………………………………..... 68

4.3 Statistics………………………………………………………………………………………….. 68

5 Results………………………………………………………………………………………………….. 70

5.1 Specific Aim 1: To determine the molecular mechanisms involved

in alternative pathway activation on activated platelets…………………….. 70

5.1.1 Does properdin bend to platelets?...... 70

5.1.1.1 Separation of physiological forms of properdin from

purified human properdin…………………………………………………………. 72

5.1.1.2 Do purified physiological forms of properdin bind to

platelets?...... 72

5.1.1.3 Does properdin binding depend upon the agonist of

platelet activation?...... 74

5.1.1.4 Is the level of properdin binding proportional to the level

of P-selectin (CD62P) on the platelet surface?...... 76

5.1.1.5 Does native properdin directly activate platelets?...... 78

5.1.1.6 Does properdin binding to activated platelets require

previous C3 fragment deposition on the platelets?...... 78

5.1.1.7 Does properdin bind to proteins and/or

glycosaminoglycans on the surface of activated platelets…………….. 82

5.1.2 Does properdin bound to platelets initiate complement

activation?......

xii

5.1.2.1 Does properdin bound to normal human platelets lead to

C3b deposition on the platelets?...... 84

5.1.2.2 Does properdin bound to platelets lead to C9 deposition

on the platelets?...... 89

5.1.3 Is factor H regulation needed for controlling properdin-

mediated complement activation on platelets?...... 92

5.1.3.1 Does factor H control properdin-mediated C3b deposition

on activated platelets?...... 92

5.1.3.2 Do aHUS-related mutations in factor H affect its ability to

control properdin mediated C3b deposition on activated

platelets?...... 94

5.1.4 What is the mechanism by which the platelet-bound

properdin activates complement?...... 97

5.1.4.1 Separation of C3(H2O) from C3...... 97

5.1.4.2 Does properdin bound to the surface of the platelet recruit

C3 components to the platelet surface?...... 98

5.1.4.3 Do the recruited C3 components on the platelet surface

lead to convertase formation?...... 101

5.1.5 Does native properdin released by neutrophils bind to

activated platelets?...... 104

5.2 Specific Aim 2: To determine the role of properdin in platelet- leukocyte aggregate formation…………………………………………………………….. 107

xiii

5.2.1 Does properdin increase platelet-leukocyte aggregate formation and is this effect dependent on complement activation? 107

5.2.1.1 Standardization of an assay that allows the measurement

of platelet-leukocyte aggregate formation…………………………………... 107

5.2.1.2 Does properdin increase platelet-leukocyte aggregate

formation?...... 109

5.2.1.3 Does inhibition of endogenous properdin in blood

(without adding properdin) reduce the formation of TRAP-

induced PLA?...... 113

5.2.1.4 Is the effect of properdin on platelet-leukocyte aggregate

formation dependent on complement activation?...... 115

5.2.2 Is the effect of properdin on platelet-leukocyte aggregate formation mediated by C5a receptors?...... 116

5.2.2.1 Analysis of the available C5aR-antagonists in their ability

to inhibit C5a-mediated activation of neutrophils………………………... 118

5.2.2.2 Is the effect of properdin on platelet-leukocyte aggregate

formation mediated by C5aR?...... 118

5.2.3 Is the properdin-mediated increase in platelet-leukocyte aggregate formation controlled by factor H?...... 119

5.2.3.1 Does factor H control the properdin-mediated increase in

PLA formation?...... 120

5.2.3.2 Are aHUS-related mutants of factor H limited in their

ability to control PLA formation?...... 122

xiv

6 Discussion……………………………………………………………………………………………... 125

Reference List…………………………………………………………………………………………….. 144

A Identification of a novel mode of complement activation on stimulated platelets mediated by properdin and C3(H2O). 2013. J.

Immunol. 190: 6457-6467……………………………………...... 165

B Local release of properdin in the cellular microenvironment: role in pattern recognition and amplification of the alternative pathway of complement. 2013. Front Immunol. 3: 412……………………………………………….. 181

List of Tables

Table 1: List of sources of properdin………………………………………...... 19

Table 2: Platelets agonists, their receptors and their intracellular effects…………………. 34

Table 3: Major contents of platelet granules……………………………………………………………. 35

xvi

List of Figures

Figure 1: Complement system overview………………………………………………………………… 2

Figure 2: Structure of C1 complex…………………………………………………………………………. 5

Figure 3: Classical pathway……………………………………………………………………………………. 6

Figure 4: Lectin pathway……………………………………………………………………………………….. 7

Figure 5: Schematic representation of molecular reactions in the formation of C3b

and C3(H2O) from C3………………………………………………………………………………. 9

Figure 6: Alternative pathway………………………………………………………………………………... 11 11

Figure 7: Terminal complement pathway……………………………………………………………….. 13

Figure 8: Domains of factor H………………………………………………………………………………… 16

Figure 9: Properdin structure…………………………………………………………...... 20

Figure 10: Release of properdin in the local microenvironment………………………………. 26

Figure 11: Non-activated and activated platelet morphology……………...... 29

Figure 12: Mechanism of PLA formation………………………………………………………………… 40

Figure 13: Purification of human properdin from plasma……………………………………….. 53

Figure 14: Unfractionated properdin binds to both non-activated and activated

platelets………………………………………………………………………………………………... 71 Figure 15: Separation of physiological properdin forms from purified properdin……. 72

Figure 16: Analysis of binding of physiological forms of properdin to activated

platelets stimulated by different agonists…………………………………………...... 73 Figure 17: Dose and time dependent binding of properdin to platelets……………………. 74

xvii

Figure 18: The level of properdin binding to platelets depends on the agonist…………. 75

Figure 19: The binding of properdin to activated platelets is not proportional to

CD62P levels on the platelet surface………………………………………………………. 77 Figure 20: Properdin does not activate platelets…………………………………………………….. 79

Figure 21: The binding of properdin to platelets is not mediated by C3…………………… 80

Figure 22: C3 components and properdin are not present on the surface of washed

activated platelets…………………………………………………………………………………. 81 Figure 23: Properdin binding is increased in the presence of chondroitin sulfate-A..... 82

Figure 24: Properdin binding is reduced upon removal of surface proteins……...... 83

Figure 25: Presence of CD46, CD55 and CD59 on platelets………………………………………. 84

Figure 26: Schematic for properdin-mediated C3b deposition in properdin-depleted

serum……………………………………………………………………………………………...... 85

Figure 27: Properdin-mediated C3b deposition in properdin-depleted serum…………. 87

Figure 28: Schematic for properdin-mediated C3b deposition in normal human

serum…………………………………………………………………………………………………… 88

Figure 29: Properdin-mediated C3b deposition in normal human serum…………………. 89

Figure 30: Schematic for properdin-mediated C5b-9 deposition in properdin-

depleted serum and normal human serum……………………………………………... 90 Figure 31: Properdin promotes formation of C5b-9 complexes on the surface of

activated platelets…………………………………………………………………………………. 91 Figure 32: Factor H binds to washed, activated platelets…………………………………………. 93

Figure 33: Properdin-mediated complement activation is exacerbated when cell

surface protection by factor H is inhibited…………………………………………….. 94

Figure 34: aHUS mutants are impaired in their ability to control properdin-

xviii

mediated C3b deposition………………………………………………………………………. 96

Figure 35: Separation of C3(H2O) from C3. …………………………………………………………….. 97

Figure 36: Analysis of binding of C3, C3b and C3(H2O) to thrombin-activated

platelets with or without properdin on their surface………………………………. 99

Figure 37: Analysis of binding of C3, C3b and C3(H2O) to arachidonic acid-activated

platelets with or without properdin on their surface………………………………. 100

Figure 38: Ability of thrombin and arachidonic acid-activated platelets to form a

functional convertase………………………………………...... 103

Figure 39: Normal human serum inhibits binding of properdin to activated platelets. 104

Figure 40: Properdin released by activated neutrophils binds to activated platelets,

but C3 released by activated neutrophils does not………………………………….. 106

Figure 41: General schematic of PLA experiment setup……………………...... 108

Figure 42: Properdin increases formation of platelet-granulocyte aggregates…………. 110

Figure 43: Properdin increases formation of platelet-monocyte aggregates…………….. 112

Figure 44: Effect of properdin on platelet-leukocyte aggregates is dose dependent….. 113

Figure 45: Inhibition of endogenous properdin leads to decrease in PLA formation…. 114

Figure 46: Inhibition of complement in serum by compstatin………………………………….. 116

Figure 47: Analysis of inhibition of complement activation on properdin-mediated

platelet-granulocyte and platelet-monocyte aggregates………………………….. 117

Figure 48: Complement-mediated effect of properdin on platelet-granulocyte

aggregation is not mediated by C5a-C5aR interaction……………………………... 119

Figure 49: Properdin-mediated platelet-granulocyte aggregate formation is

controlled by factor H…………………………………………………………………………… 121

Figure 50: Properdin-mediated platelet-monocyte aggregate formation is controlled

by factor H……………………………………………………………………………………………. 122

xix

Figure 51: Increase in platelet-monocyte aggregation caused by rH19-20 is

dependent on complement…………………………………………………………………….. 123

Figure 52: aHUS-related rH19-20 mutants of factor H are not affected in their ability

to compete with full-length factor H and thus increase the formation of

platelet-leukocyte aggregates………………………………………………………………… 124

Figure 53: Model 1: Properdin-mediated complement activation on activated

platelets………………………………………………………………………………………………... 142

Figure 54: Model 2: Role of properdin in PLA formation…………………………………………. 143

List of Abbreviations

A414 nm………….. Absorbance at 414nm AA……………….. Arachidonic acid ACD……………... Acid citrate dextrose ADP……………... Adenosine diphosphate aHUS……………. Atypical hemolytic uremic syndrome ARMD…………... Age-related macular degeneration

C5aR……………. Receptor for C5a C5aRa…………... Antagonist for C5aR CD62P………….. P-selectin COX……………... Cyclooxygenase CR1……………... Complement receptor 1 CR2……………... Complement receptor 2 CR3……………... Complement receptor 3

DAF……………... Decay accelerating factor

EDTA…………… Ethylene diamine tetraacetic acid EGTA…………… Ethylene glycol tetraacetic acid ERs……………..... Rabbit erythrocytes ESC3…………….. C3b-coated sheep erythrocytes fD……………...... Factor D fB……………...... Factor B

GAG……………... Glycosaminoglycan GVB……………... Gelatin veronal buffer GVBE…………… Gelatin veronal buffer with EDTA

HBSS……………. Hanks balanced salt solution HETE…………… 12-(S)-hydroxyeicosatetraenoic acid

IgG………………. Immunoglobulin

LOX…………….... 12-lipoxygenase LPS……………..... Lipopolysaccharide

xxi

MAC……………... Membrane attack complex MASP-1………... MBL associated protease-1 MASP-2………... MBL associated protease-2 MBL……………... Mannose binding lectin MCP……………... Membrane cofactor protein MgEGTA……….. Magnesium-Ethylene glycol tetraacetic acid

NHS …………….. Normal human serum

OCS……………… Open canalicular system

P2………………… Properdin dimers P3………………… Properdin trimers P4………………… Properdin tetramers Pn………………… Properdin oligomers PARs……………. Protease-activated receptors PBS……………… Phosphate buffered saline PGE1……………. Prostaglandin E1 PGE2 ……………. Prostaglandin E2 PGH2……………. Prostaglandin H2 PLA……………… Platelet-leukocyte aggregates PLA2…………….. Phospholipase A2 PLC……………… Phospholipase-C PMA…………….. Phorbol 12 myristate 13 acetate PMNs…………… Polymorphonuclear cells PMPs…………… Platelet micro-particles PNH……………... Paroxysmal nocturnal hemoglobinuria PRP……………… Platelet rich plasma PSGL-1…………. P-selectin GP ligand-1

RBC……………… Red blood cell rH19-20……….. Recombinant domains 19-20 of factor H RT………………… Room temperature sC5b-9…………. Soluble C5b-9 SCRs…………….. Short consensus repeats SD………………... Standard deviation

TRAP…………… Thrombin receptor activating peptide TREM-1……….. Triggering receptor expressed on myeloid cells-1 TSR……………… Thrombospodin repeat TXA2……………. Thromboxane A2

WT………………. Wild type

xxii

Chapter 1

Introduction

1.1 The Complement System

1.1.1 Background

The complement system, discovered by Jules Bordet in 1896, is a major part of the innate immune response. It was first described as the heat labile component of the serum that ‘complemented’ antibody mediated killing of bacteria. It is now known that the complement system comprises a complex network of more than 30 proteins in plasma and on cell surfaces. All complement proteins combined are at a concentration of about 3g/L of plasma, and they make up more than 15% of the globular fraction of plasma (1). These proteins can undergo activation in a cascade of steps, serve as regulatory proteins or act as surface receptors (2). This system can allow for an immune response against infectious organisms, damaged tissue, and

‘non-self’ surfaces. The complement system consists of three distinct pathways, which depend on different molecules for their activation: the classical pathway, the

CLASSICAL LECTIN ALTERNATIVE PATHWAY PATHWAY PATHWAY

Antigen:antibody complexes Recognition of PAMPs Spontaneous hydrolysis

Modified host proteins by lectins or ficolins of C3 to C3(H2O) C-reactive protein Amyloid P protein MBL MASPs C1q Ficolins C2 C3 fB pentraxins C4 fD C1r fP C1s fH C2 C4 fI C3b

Adaptive & Opsonization Inflammation Membrane Humoral Phagocytosis Chemotaxis Disruption Immunity C3b deposition C3a, C5a release C5b-C9 activation Antigen processing

Figure 1: Complement system overview. Complement activation occurs through one or more of three pathways. Each pathway uses a unique set of recognition molecules in order to initiate. All activation pathways lead to cleavage of a central C3 molecule and attachment of C3b to target sites. Finally, targets of complement activation undergo numerous processes all directed either at killing or at developing lasting immunity, as well as release of pro- inflammatory mediators such as C3a and C5a which can recruit immune-system cells and promote inflammation. (Adapted from Pangburn et al, Vaccine, 2008, 26 Suppl 8:15-21 )

lectin pathway and the alternative pathway (Fig 1). Although these pathways activate differently, they converge at the formation of the C3 convertase, which leads to deposition of C3b fragments on pathogenic surfaces. The complement system protects the host in various ways including: (a) the generation of many 2

opsonizing fragments such as C3b, iC3b and C3d that bind covalently to pathogens and help in their engulfment by phagocytes including polymorphonuclear cells

(PMNs), macrophages, B cells, CD4+T cells, and CD8+ T cells. (Reviewed in (3)); (b) the production of small sub products of activation (C3a and C5a) that are released into the fluid phase of blood and activate certain cells including PMNs, dendritic cells, mast cells, eosinophils and basophils or act as chemotactic agents for other cells such as PMNs and macrophages (Reviewed in (4)); (c) the formation of the membrane attack complex (MAC) in the final stages of complement activation, which acts by creating pores in the membranes of target surfaces, including pathogens.

Besides the role of complement in innate immunity, it also plays an important role in generating adaptive immunity (Reviewed in (5)). Opsonization of antigens allows for enhanced uptake of complement-coated antigens by antigen presenting cells. Also, B-cells express complement receptor 2 (CR2 or CD21), which is a receptor for C3b fragments (C3d, C3dg). This receptor, along with CD19 and

CD81, forms the B-cell co-receptor. When the B-cell receptor (immunoglobulin) comes in contact with an antigen, the CR2 part of the co-receptor can recognize the complement fragments on the antigen, resulting in significant lowering of the antigen threshold needed for efficient B cell activation (6,7).

The activation of complement plays an essential role in the generation of immune responses against infectious organisms. It is also known to play an important role in clearance of dying and dead cells through opsonization and 3

promotion of phagocytosis (8-10) as well as in clearance of immune-complexes via

C3b deposition on the complex (Reviewed in (11)). Although complement is beneficial to the body’s defense system, complement activation is also a major contributor to inflammatory disorders such as ischemia/reperfusion injury (12), autoimmune disorders (Reviewed in (13)) and atherosclerosis (14-16). When complement activation becomes deregulated or excessive, it can participate in contributing to or initiating several pathologies.

1.1.2 Activation of the complement system

The complement system can be activated by three major pathways: the classical pathway, lectin pathway and alternative pathway.

1.1.2.1 Classical Pathway

The classical pathway is initiated when C1q recognizes and binds to the Fc region of an antibody (usually IgG and IgM) bound to an activating surface. Aside from binding antibodies, the C1q molecule can also bind to cell wall components and membrane proteins of various microorganisms, modified host proteins and phospholipids, C-reactive protein and amyloid P protein. The C1q molecule is a hexamer, with each unit consisting of a globular head and a collagenous tail, and the six globular heads are connected by their collagenous tails. C1q forms a part of the

C1 complex, which consists of a single C1q molecule bound to two molecules each of

the serine proteases C1r and C1s (Fig 2). Binding of C1q to the surface leads to a conformational change in the collagenous region of C1q, which causes autoactivation of C1r, leading to the activation of C1s. The activated C1s molecule then cleaves complement component C4 into C4a and C4b. The cleaved C4b fragment binds covalently to –NH2 groups (proteins) or –OH groups (carbohydrates) present on the activating surface. C1s also cleaves complement component C2 into C2a and

C2b. The C2b fragment binds to the surface-bound C4b, and forms C4b,2b; the C3 convertase of the classical pathway. This convertase can now cleave C3 molecules into C3a and C3b. C3b molecules can bind covalently to the target surface in a mode identical to C4b and lead to opsonization (Fig 3). The C3b molecule also associates

Figure 2: Structure of C1 complex. The C1 complex consists of one molecule of C1q and two molecules each of C1r and C1s. C1q is composed of six identical subunits with globular heads and long collagen-like tails, resembling a “bunch of tulips”. Two molecules each of C1r and C1s bind to the tail region.

with C4b,2b to form C4b,2b,3b; the C5 convertase of the classical pathway. This convertase cleaves C5, initiating the terminal activation cascade and leading to formation of C5b-9 (membrane attack complex), which perforates the cell membrane and leads to cell damage via osmotic imbalance.

Figure 3: Classical pathway. The classical pathway is initiated through the interaction of the C1 complex to the pathogen surface or with antibodies bound to the surface. The binding of the globular heads of C1q to the Fc regions of immunoglobulins leads to autoactivation of C1r which then cleaves and activates C1s. Activated C1s can digest C4 and C2 in order to form the C4b,2b complex – the C3 convertase – which digests C3 molecules generating C3b that covalently binds to the pathogen surface.

1.1.2.2 Lectin pathway

The activation of the lectin pathway is very similar to the classical pathway, but the lectin pathway uses different initiation molecules for activation. These activation molecules (mannose binding lectin (MBL), L-ficolins or H-ficolins) recognize specific carbohydrates and carbohydrate patterns on the microbial

surface. MBL and ficolins are very similar to C1q in structure and function. However,

MBL is a collectin made up of 2-6 units, each of which has a collagen like N-terminal region, a neck region, and a globular C-terminal lectin domain. The MBL molecule forms a complex with two serine proteases: MBL associated protease-1 (MASP-1) and MBL associated protease-2 (MASP-2). When the MBL complex is bound to the activating surface, MASP-2 can cleave C4 and C2 and lead to the formation of a C3 convertase in a mode similar to that of the classical pathway. The rest of the cascade is identical to that of the classical pathway (Fig 4).

MBL/Ficolins (Collectins) C4

C2 C3 C3a

C4b C2b C3b C3b C3b Carbohydrates

C3 convertase

Figure 4: Lectin pathway. The lectin pathway initiates through the binding of the mannan binding lectin (MBL) or ficolins to sugar residues on the microbial surface. MBL or ficolins are bound to MBL-associated serine proteases (MASPs) 1 and 2. Similar to C1s, MASP2 activates C4 and C2 which then leads to formation of the C3 convertase and deposition of C3b.

1.1.2.3 Alternative pathway

The mechanism of activation of the alternative pathway is very distinct from that of the classical and lectin pathways. This pathway does not require specific protein-protein or protein-carbohydrate interactions and is activated spontaneously and everywhere in an organism by the constant mechanism of C3 ‘tickover’

(Reviewed in (17)). This pathway is able to discriminate between the self- and non- self surfaces due to strict regulation of the pathway by regulatory proteins present both in the plasma as well as on the cell membrane of the host. The proteins factor

B, factor D and properdin are exclusive to the alternative pathway of complement.

Activation of this pathway leads to generation and deposition of large amounts of

C3b, which then initiates the terminal activation cascade. The activation and regulation of the alternative pathway are described below in detail.

1.1.2.3.1 Complement component C3

Complement component C3 is the central and most abundant protein of the complement system. It is present in the plasma at a concentration of approximately

1g/L. C3 consists of a larger α-chain of 115kDa and a smaller β-chain of 75kDa. The two chains are linked to one another by a disulfide bridge (Fig. 5A). When C3 undergoes proteolysis by the C3 convertase, a small C3a fragment of 9kDa is released from the α-chain of C3, and the remaining molecule is known as C3b (Fig

5C).

1.1.2.3.2 Initiation of the alternative pathway: Formation of the initial fluid phase C3 convertase

The alternative pathway is initiated in the fluid phase by the spontaneous hydrolysis of the thioester bond in the C3 molecule leading to formation of C3(H2O)

(Fig. 5B). C3(H2O) is structurally different from C3b as it has a complete α-chain

β β A B H OH +H2O SS S C O SS CS O α α C3 C3(H O) Proteolytic 2 activation β C S S α S C O C3a Metastable C3b

+R-OH group +H2O on target surface β β D E S S S S α α S C O S C O H OH H OR Target Fluid phase Surface C3b Surface bound C3b

Figure 5: Schematic representation of molecular reactions in the formation of C3b and C3(H2O) from C3. Initiation of the alternative pathway occurs by spontaneous hydrolysis of the putative thioester in native C3, generating a C3 molecule with C3b like properties known as C3(H2O) (A, B). Proteolytic cleaveage of a C3 molecule by a C3 convertase leads to release of the C3a fragment for the α- chain of C3 leaving behind a metastable C3b (A, C). The intramolecular thioester bond in the α-chain of C3 becomes the reactive group of proteolytically produced metastable C3b (C). If the C3b molecule is not bound to a surface immediately, it is inactivated by hydrolysis of the thioester bond and is lost in the fluid phase (D). If close enough to a surface, the highly reactive thioester reacts with the –NH2 or – OH groups on the surface and allows the C3b to bind to a surface (E).

whereas in C3b, the C3a domain has been proteolytically cleaved from the α-chain.

On an SDS-PAGE, C3(H2O) is indistinguishable from C3. C3(H2O) molecules can bind to factor B in the presence of Mg++ ions (C3(H2O),B). Factor D can then proteolytically cleave factor B and release a 33kDa fragment known as Ba. The remaining factor B, now known as Bb, remains bound to the C3(H2O) leading to formation of C3(H2O),Bb, the fluid phase C3 convertase (Fig. 6A).

1.1.2.3.3 Initial C3b deposition and formation of C3b,Bb convertase

Once the C3(H2O),Bb convertase is formed, it can proteolytically cleave C3 molecules to form C3a and a metastable form of C3b. This metastable C3b exposes a highly reactive thioester bond that allows it to bind to any surface that has an available –NH2 group (proteins) or –OH group (carbohydrates), if it is close enough

(Fig. 5E). While the thioester bond in intact C3 has a half-life of 231h at 37°C (18), the metastable C3b has a half-life of 60µs (19). If the C3b molecule is not bound to a surface immediately, it is inactivated by hydrolysis of the thioester bond and is lost in the fluid phase (Fig 5D). The surface-bound C3b can now bind factor B in the presence of Mg++ ions. Factor B can then be cleaved by the serine protease factor D, leading to the formation of C3b,Bb, the surface bound C3 convertase. This C3b,Bb can now cleave many C3 molecules and generate C3b. Each new C3b fragment has the capacity to form a new C3 convertase, thus leading to efficient amplification of the alternative pathway (Fig 6B).

1.1.2.3.4 Role of properdin in amplification of the alternative pathway

Properdin, a high molecular weight polymer, is the only positive regulator of the alternative pathway of complement. It functions by binding to and stabilizing the

C3 convertase (Fig. 6B). The C3b,Bb protease is susceptible to dissociation (20) and has a half-life of less than 90 seconds (21). Properdin binds to this complex and increases its half-life by 4- to 10- fold (22). It binds to C3b via a region near the C-

A Ba fB fB fD Bb +H2O C3 C3(H2O) C3(H2O) C3(H2O) Fluid phase C3 convertase

Bb C3 C3(H O) 2 Ba fB P Fluid phase fB fD Bb P Bb C3 convertase C3b C3b C3b C3b C3b C3b

Figure 6: Alternative pathway. The alternative pathway of complement represents a true safegaurd system of the human host, which is triggered spontanously and everywhere in an organism. (A) Spontaneous hydrolysis of a C3 molecule in the fluid phase leads to generation of C3(H2O) which can bind factor B (fB). fB then gets cleaved by factor D (fD) yielding the fluid phase C3 convertase C3(H2O),Bb. (B) The spontaneous fluid phase C3 convertase digests C3, generating C3b fragments which have the ability to bind covalently to nearby membranes. Bound C3b can now interact with fB, which is then cleaved by fD, generating the membrane-bound C3 convertase which can cleave C3. Properdin (P), a positive regulatory protein, binds to the C3 convertase and stabilizes it, allowing the convertase to cleave many C3 molecules for efficient amplification.

terminus of its α-chain. The structure and function of properdin is discussed in detail in the review paper attached in appendix B and will be further discussed ahead (section 1.1.2.6).

1.1.2.4 C5 convertase and the terminal complement pathway

To this point we have described how the C3 convertases of the classical, lectin as well as the alternative pathway are generated. The next step is the generation of the

C5 convertase. For all pathways, this is brought about by the attachment of a second

C3b molecule to the corresponding C3 convertase. For the classical and lectin pathway, the C5 convertase is generated by binding of C3b to C4b,2b to form

C4b,2b,3b. For the alternative pathway C3b binds to C3b,Bb to form C3b2,Bb. The C5 convertase can now cleave C5. C5 is a 192kDa (23) two-chain glycoprotein (24) that binds to the C5 convertase through C3b (25) and thus can be cleaved by C2b or Bb to form C5a and C5b. C5b can then initiate the non-enzymatic assembly of the components of the terminal complement pathway. C5b binds C6 to generate a C5b,6 complex, which then recruits a C7 molecule causing a conformational change in C7 that exposes its hydrophobic site and allows it to insert into the lipid bilayer (26).

Binding of the next component, C8, to this complex exposes a hydrophobic site on

C8 allowing C8 to insert into the membrane. This complex then promotes polymerization of 10-26 molecules of C9, leading to the formation of a membrane pore, which has an external hydrophobic face and an internal hydrophilic channel of about 100Å (27). This pore leads to the loss of cellular homeostasis by allowing the free passage of water and solutes across the lipid bilayer (Fig 7). 12

C3 C6 C8 C9 C5 C7

C5a C5b-C9 Bb P (MAC) C3b C3b C5b

C5 convertase Figure 7: Terminal complement pathway. Binding of (Alternative Pathway) C3b to a C3 convertase generates a C5 convertase (C4b,2b,3b for the classical and lectin pathways OR

C3b(n),Bb for the alternative pathway) . The C3b in the C5 C3 convertase can non-covalently bind a C5 molecule which then gets digested into C5a and C5b by C2b (classical or OR lectin pathway) or Bb (alternative pathway). Digestion of C5 is the last enzymatic step of the pathway. C5b can then bind C6 to form the C5b,6 complex which then recruits a C7 molecule causing a conformational change C4b C2b C3b in C7. C5b-C7 can strongly bind to the membrane without disrupting the lipid bilayer. Binding of C8 to this complex exposes a hydrophobic site on C8 allowing it to insert C5 convertase into the membrane. Next, a C9 molecule binds to the C5b- (Classical/Lectin C8 complex and undergoes a conformational change, Pathway) which allows it to cross the membrane and also allows

binding of additional C9 molecules, upto a maximum of 18, leading to the formation of a membrane pore or the membrane attack complex (MAC).

1.1.2.5 Regulatory proteins that allow the alternative pathway to distinguish between host/self cells (non-activating surfaces) and activating/pathogenic surfaces.

1.1.2.5.1 Soluble negative regulators of the alternative pathway

There are two major soluble negative regulatory proteins of the alternative pathway of complement present in the plasma – factor H and factor I.

(a) Factor I: Factor I is a highly specific serine protease that can cleave the α- chain of C3b as well as C4b, which inactivates the fragments. Factor I requires the presence of co-factors to carry out its enzymatic function. Plasma protein factor H and soluble complement receptor 1 (CR1), as well as the surface proteins membrane cofactor protein (MCP) and CR2 serve as co-factors for the cleavage of C3b. The presence of factor H increases the affinity of factor I for C3b by 15 fold (28).

(b) Factor H: Factor H (formerly known as β1H) is a 150kDa serum glycoprotein, present in the plasma at a concentration of 116 – 562 µg/mL (29). It is expressed constitutively in the liver (30), but can also be expressed locally by various cells such as endothelial cells (31), epithelial cells, retinal pigment epithelial cells (32), and mesenchymal cells (33). Factor H is a complement regulatory protein that functions in two ways: (a) it accelerates the decay of the alternative pathway C3 convertase (C3b,Bb) and (b) it acts as a co-factor for factor I to allow factor-I mediated cleavage and inactivation of C3b (34-37). If factor H is absent in the fluid phase, it leads to spontaneous activation of complement and consumption of C3 and 14

factor B in the fluid phase (38). Factor H is also essential for discrimination of self- cells (host cells) from non-self cells and thus helps control complement activation on these cells by the described mechanisms. Efficient recognition of cell surfaces by factor H requires the surface to have specific markers such as polyanions (i.e. sulfated glycosaminoglycans) along with initial C3b fragments (39). Surfaces that effectively activate the alternative pathway (i.e. various pathogens) lack polyanion markers and thus cannot efficiently bind factor H, leaving the surface unprotected from the alternative pathway.

Factor H is composed of 20 highly conserved short consensus repeats (SCRs)

(40) of about 60 amino acids in length linked to one another with about 3-8 amino acid spacers (Reviewed in (41)). The four N-terminal domains contain a C3b binding site that is responsible for the decay acceleration and co-factor activities of factor H

(35,42-44). The remaining sixteen domains contain various binding sites, which allow functional effectiveness of domains 1-4. Additional C3b binding sites have been mapped in regions 7-15 and 19-20 (45-48). Factor H also has sites for heparin and other polyanions in SCR 7, SCR 9-15 and SCR 20 (46,48-53). The SCR 19-20 sites for C3b and polyanion binding have been shown to be essential for interaction of factor H with host surfaces (39,46,54-60) (Fig 8). In agreement with these findings, our laboratory has previously shown that recombinant domains 19-20 (rH19-20) can efficiently compete with full length factor H for binding to C3b and polyanions on cells (39,54,55).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Complement Polyanion C3b and regulatory binding Polyanion domains domains binding domains C3b binding domains aHUS-linked domains

Figure 8: Domains of factor H. Factor H consists of 20 short consensus repeats (SCRs) connected by 3-8 amino acid spacers. The figure shows the regions of factor H in which the critical functions of factor H such as complement regulation, C3b binding, recognition of host cells are located and also shows the domains that can contain mutations most commonly associated with aHUS.

Mutations in the domains of factor H that allow recognition of host cells result in diseases such as atypical hemolytic uremic syndrome (aHUS) (61,62) and age-related macular degeneration (ARMD) (62). aHUS is a disease characterized by thrombocytopenia, hemolytic anemia and renal failure (63). Mutational studies have shown that most of the aHUS related mutations lie on or indirectly affect the C- terminal domains, mainly domain 20, of factor H (64-69); Reviewed in (41)). These mutations in domain 20 lead to defective binding of factor H to C3b, to polyanions or to both (39). Thus, aHUS related mutations lead to a defect in binding of factor H to cell surfaces, and disrupt factor H-mediated cell surface protection.

1.1.2.5.2 Membrane-bound negative regulators of the alternative pathway

There are four membrane proteins that are present on host cells and have regulatory activity for the alternative pathway (as well as classical and lectin pathways): decay accelerating factor (DAF, CD55), complement receptor 1 (CR1), membrane cofactor protein (MCP; CD46), and CD59.

(a) Decay accelerating factor: It accelerates the decay of the C3/C5 convertases (70, 71), in a mode similar to factor H.

(b) CR1: It is a transmembrane protein that induces decay of C3/C5 convertases and acts as a cofactor for factor I, allowing cleavage and inactivation of

C3b.

(c) Membrane cofactor protein: It is a transmembrane protein that has one main known function: it allows the factor I-mediated cleavage and inactivation of

C3b (72).

(d) CD59: It is a GPI-anchored protein that binds to C8 in the C5b-8 complex and prevents the binding of C9, thus preventing the formation of the lytic MAC.

1.1.2.6 Properdin: a positive regulatory protein with functions in complement initiation

(Structure and function of properdin is also discussed in detail in a review paper in appendix B)

Properdin is the only known positive regulator of the alternative pathway of complement and functions by stabilizing the C3 convertase of the pathway. It was first described in 1954 by Louis Pillemer (73). In 1971, Götze and Müller-Eberhard discovered the ‘C3 activator system’ for complement activation (74) and proposed that properdin was essential for stabilization of the C3 convertase required for amplification of the pathway (75).

1.1.2.6.1 Properdin sources

While most other complement proteins are produced mainly in the liver, properdin is not synthesized in the liver and is made by various cell types including peripheral blood monocytes (76), neutrophils (77), primary T cells (78), shear- stressed endothelial cells (79), dendritic cells (80), bone marrow progenitor cell lines (reviewed in (81)) and adipocytes (82) (Table 1) resulting in properdin serum levels of 4–25 µg/ml (81,83-85). Multiple inflammatory agonists, such as TNF-α, C5a or fMLP, can stimulate the release of properdin (Table 1) from the granules of neutrophils into the pro-inflammatory microenvironment to induce local alternative pathway activation.

Table 1: List of sources of properdin

(Reproduced from Cortes et al, Front Immunol, 2013, 3:412)

1.1.2.6.2 Properdin structure

Properdin, a highly positively charged protein (isoelectric point > 9.5), exists as cyclic dimers (P2), trimers (P3), and tetramers (P4) of head-to-tail associations of monomeric subunits (86,87) (Fig 9). Each monomer weighs ~53 kDa (88), and is composed of seven thrombospondin repeat (TSR) type I domains (TSR0-TSR6)

(89) (Fig 9). Domain deletion studies (90) have shown that TSR3 is not required for

C3b,Bb stabilization nor sulfatide binding. TSR4 is important in stabilizing the

C3b,Bb convertase complex, and TSR5 is required for binding to C3b. TSR5 is also

involved in sulfatide binding, but the region of TSR5 involved for binding sulfatides differs from the region involved in binding to C3b (91).

The native P2, P3, P4 forms of properdin have greater affinity for cell-bound

C3b,Bb or C3b,B than for cell-bound C3b. Binding of properdin to all three molecules

C3b,Bb C3b,B or C3b, is more favorable when these molecules are bound to a cell membrane versus their respective fluid phase counterparts (92). Properdin is also known to interact with C3(H2O) (a C3b-like protein; Fig. 5B) (93).

Figure 9: Properdin structure. Properdin consists of identical monomeric units connected to one another in a head to tail manner. Each monomer consists of seven thrombospondin repeat (TSR) type I domains.

1.1.2.6.3 Role of properdin as an initiator versus stabilizer of the alternative pathway convertase

Upon discovery by Pillemer (73), properdin was originally thought to be an initiator of the alternative pathway of complement (73). This controversial view was later replaced by the widely accepted notion that properdin serves as a positive regulator that amplifies the alternative pathway by extending the half-life of the C3 and C5 convertases (20,22) (as described in section 1.1.2.3.4). Recent reports 20

propose that properdin acts as a pattern recognition molecule (as discussed ahead).

This view is consistent with the complement initiation function proposed over 50 years ago (73) and has re-opened the controversy regarding the functions of properdin.

1.1.2.6.4 Binding of properdin to a variety of cell surfaces and potential binding ligands for properdin

Recent studies have reported properdin binding directly to various non-self surfaces: zymosan (94,95), rabbit erythrocytes, Neisseria gonorrhoeae (95), certain

Escherichia coli strains (95,96), early (97) or late apoptotic cells (85), necrotic cells

(85,94), live human leukemia T cell lines (97), normal human proximal tubular epithelial cells (98), Chinese hamster ovary cells (97), neutrophils (77,99), and cartilage oligomeric matrix protein (100). Properdin, covalently bound to a biosensor surface, can subsequently recruit C3b and factor B to form C3b,Bb,P

(101). Importantly, this study conducted by Hourcade also shows that even when properdin binds to surface-bound C3b, properdin can still recruit C3b and factor B to form a new convertase. This goes beyond the “convertase stabilizer” function, in which properdin binds only once the convertase is already formed and suggests that properdin may serve as a pattern recognition molecule for alternative pathway initiation on targets. Additional evidence supporting the ability of properdin to initiate complement activation (by forming de novo C3 convertases on cell surfaces) comes from studies where human embryonic kidney cells (102) or E. coli (95) were

transfected with a vector expressing a transmembrane form of properdin on the cell surface, turning the cell surface into an activator of the alternative pathway.

Properdin may interact directly with surfaces by recognizing specific surface ligands including DNA on late apoptotic and necrotic cells (85), glycosaminoglycan

(GAG) chains of surface proteogycans on proximal tubular epithelial cells (98) and T cells (97) (including heparin (103), heparan sulphate (97,104), dextran sulfate

(105), fucoidan (105), and chondroitin sulphate (97), and bacterial lipopolysaccharide (LPS) and lipooligosaccharide (106), all of which are negatively charged molecules. Additional studies are needed for identifying the receptors for properdin on other surfaces on which properdin has been found to bind, as discussed above.

1.1.2.6.5 Importance of separating physiological forms of properdin (P2-P4) from aggregated (“activated”) properdin to study the specificity of properdin-target interactions.

Biochemical studies of serum-derived pure properdin indicate that non- physiological high molecular weight, highly positively charged polymers (known as

Pn or “activated” properdin) form during long term storage and freezing/thawing

(86,107). Although “activated” properdin (or Pn) retains the ability to stabilize the alternative pathway convertases, it possesses the abnormal capacity to activate complement in solution (consumption of complement) (86) and bind non- specifically to surfaces such as live T cells and Nesseriae (94,108). The studies

mentioned in section 1.1.2.6.4 (except (94,108,109) and parts of (85,97,99)) were carried out with unfractionated pure properdin potentially containing aggregates.

Studies using physiological forms of properdin (P2-P4) separated from non- physiological aggregates, by ion exchange and/or size exclusion chromatography, found native properdin does not bind to some previously described surfaces, such as rabbit erythrocytes, live Jurkat cells and Neisseria sp. (94,108). However, native properdin forms do bind to necrotic cells, yeast cell wall components (94),

Chlamydia pneumoniae (109), and activated platelets ((110); Appendix A), suggesting it is a highly selective recognition molecule. In addition, neutrophil- derived native/physiological properdin can bind to apoptotic T cells (97) and neutrophils (77,99), while properdin, in the context of C3-deficient serum can bind to dying cells (85). Interestingly, T cell-derived properdin is ~100 times more active than serum properdin (78) when tested in a traditional AP hemolytic assay, but the molecular mechanisms involved in the increased activity remain unknown.

Although it has been speculated that serum-derived, unfractionated properdin (that can contain aggregated “activated” properdin) may be similar to native neutrophil- or T cell-derived properdin, biochemical experimental evidence is lacking.

Moreover, activated properdin forms (Pn), are not normally in circulation (or are tightly controlled) since their presence leads to systemic complement activation and consumption (86). Based upon available experimental evidence, future studies should be carried out only with native properdin forms (separated from “activated” properdin), or with fresh leukocyte-derived properdin, in order to effectively

determine specific interactions between properdin and surfaces and not over- estimate the role of properdin (due to aggregates) in complement activation.

1.1.2.6.6 The role of properdin in the local microenvironment

Transient increases in local properdin concentrations due to cell production

(i.e. T cells, monocytes, and neutrophils) would likely lead to stabilization

(Reviewed in (81)) and generation/initiation of the alternative pathway convertases, thus greatly amplifying the complement response to a local stimulus, in particular because these cells also synthesize the other complement proteins necessary for complement activation.

Properdin released by phagocytes binds to apoptotic and necrotic cells

(85,97), and this may aid in their removal directly or through properdin-mediated complement activation (Fig. 10). Properdin-mediated complement activation may also be important for further recruitment of proinflammatory cells to infection sites.

At sites of inflammation where many different properdin-producing cells are in close proximity and cytokine release and complement activation occurs, neutrophils rapidly secrete properdin upon degranulation stimuli (Table 1). Endogenous native properdin has been detected on the surface of isolated, non-stimulated neutrophils

(77) and TNF/fMLP-stimulated neutrophils (99), independently from C3 (99).

Unfractionated properdin (known to contain non-physiological complement- activating aggregates, as described above), when incubated with isolated resting neutrophils, promotes complement activation on neutrophil membranes (99), and

when added to whole blood, induces the formation of platelet-leukocyte aggregates

(111). The exact mechanism of complement activation on neutrophils remains to be determined and properdin-mediated initiation is possible. Complement activation on neutrophils results in increased release of complement products such as C5a fragments and MAC that could further activate neutrophils, endothelial cells or other cells in close contact with adherent neutrophils and contribute to a pro- inflammatory microenvironment.

Figure 10: Release of properdin in the local microenvironment. Properdin (P) released by immune cells may directly bind to surfaces and promote alternative pathway complement activation. P may recruit C3b or C3(H2O) to form a C3 convertase and further promote C3b deposition on surfaces (P-mediated complement activation). Sources of C3b may be derived from C3 convertases of the alternative, lectin or classical pathways. C3(H2O) derived from ‘tick over’ C3 hydrolysis may also bind to cell-bound properdin, forming a C3(H2O),Bb convertase on the cell. In addition, properdin can bind to C3b on surfaces and recruit additional C3b and factor B, generating new convertases. P-mediated complement activation may participate in opsonization, MAC deposition and C3a and C5a release which are important processes in inflammatory immune responses. Finally, locally released P may carry out functions that are independent from complement activation/amplification. (Reproduced from Cortes et al, Front Immunol, 2013, 3:412; Appendix B)

1.2 Platelets

1.2.1 Background

Platelets are small, anucleate circulating blood particles which have the primary role of accumulating at sites of vascular injury and initiating a blood clot to arrest blood loss. Giulio Bizzozero, the ‘father of the platelet’, described platelets as a novel morphological element with important roles in hemorrhage and thrombosis.

More recently, platelets have been found to also play important roles in inflammation.

1.2.1.1 Platelet origin

Platelets arise from large, nucleated bone marrow cells known as megakaryocytes. These megakaryocytes rearrange their cytoplasm into proplatelets

(long extensions resembling beads on a string), which extend through the endothelial lining of blood vessels and finally result in release of platelets into the blood stream (112). Human blood contains approximately 150,000 to 400,000 platelets/µl, with an average lifespan of about 10 days (113), before being degraded in the liver or spleen.

1.2.1.2 Platelet structure and function

Platelets circulate in the blood as oval shaped discs that lack a nucleus and genomic DNA and have a diameter of about 2-3 µm. They have a bilamellar plasma

membrane composed of phospholipids. Their plasma membrane is mostly smooth with invaginations that form the open canalicular system (OCS), which is a series of plasma membrane indentations that run through the interior of the platelet (112).

Although the role of the OCS remains somewhat unclear, it is thought that the series of folded membranes provide the platelets with a large surface area to take up molecules for storage and release them upon activation (Reviewed in (114)).

Platelets also have a well-defined cytoskeletal system, composed primarily of actin filaments, that allows them to maintain their shape. The cytoplasm also contains various organelles including mitochondria, lysosomes, peroxisomes, α granules and dense granules.

Platelets lack a nucleus and thus cannot transcribe nuclear genes or regulate gene expression by transcription, however, they are rich in mitochondria that contain genetic material that can be actively transcribed in platelets (115). Platelets can carry out protein synthesis (116) as they contain small amounts of mRNA, an endoplasmic reticulum, and ribosomes. Thus, platelets have a dynamic proteome. In the non-activated state, platelets have lower translational levels, but higher levels of protein synthesis are attained upon thrombin activation. Assessing the differences in the platelet proteome will give more insight into platelet physiology and their function in physiological versus pathological conditions.

Platelets in the resting state have a discoid shape. Upon activation, such as when they come in contact with sites of vascular damage, they change their shape from normal discs to spheres with finger-like dendritic protrusions that considerably increase their size (Fig 11). They do so by initiating rapid calcium

Platelet Discoid granules Platelets with Platelet shaped finger-like granules being platelets dendritic released protrusions

Non-Activated Activated platelets platelets

Figure 11: Non-activated and activated platelet morphology. Resting or non- activated platelets are discoid structures that lack a nucleus but contain various organelles and granules. Upon activation by different stimuli, platelets change their shape to spiny spheres with finger like filopodia and pseudopods and release their granule contents. influx and reorganization of their actin cytoskeleton (117,118).

The most characterized function of platelets is to participate in hemostasis.

When circulating platelets are exposed to subendothelial structures at sites of

vascular damage, they rapidly get activated and undergo shape change. At this time, the receptors for adhesive and clotting proteins also increase on the surface of the platelets. These platelets attract more platelets to form clumps and plug or seal leaks in the vasculature. Platelets also participate in inflammatory processes when activated, since they release various cytokines and chemokines, that can recruit leukocytes to the injured endothelium (Reviewed in (114)). Recently there has also been a growing body of evidence that shows activated platelets can directly activate complement (119,120), suggesting that platelets can focus complement activation to sites of vascular injury.

1.2.2 Platelet activation

Platelet activation is a rapid process that occurs when platelets come in contact with (i) agonists such as thrombin, ADP, collagen, platelet activating factor, arachidonic acid (Reviewed in (112)), (ii) thrombogenic surfaces such as an injured endothelium (Reviewed in (112)), and (iii) artificial surfaces such as hemodialysis equipment, vascular grafts and stents (121). Platelets can also get activated when they are subject to physical stimuli such as high shear stress (122). Activation of platelets leads to changes in platelet shape due to rapid rearrangement of the cytoskeleton, release of platelet micro-particles (PMPs), and release of platelet granules that causes activation of surrounding platelets and thus leads to platelet aggregation.

1.2.2.1 Platelet agonists

Various agonists differ in their ability to activate platelets. Some agonists, such as thrombin, arachidonic acid and collagen, are classified as strong agonists

(which cause degranulation of the platelet), whereas others such as ADP and serotonin are classified as weak agonists (only result in shape change). At sites of endothelial damage, various agonists are produced.

1.2.2.1.1 Thrombin and thrombin receptor-activating peptide (TRAP)

Thrombin is a serine protease that, besides activating platelets, is also involved in various other functions such as converting fibrinogen to fibrin, regulating vessel tone, and smooth muscle cell proliferation. It is produced by cleavage of its precursor prothrombin in the coagulation cascade. Thrombin carries out its signaling through a class of G-protein coupled receptors called protease- activated receptors (PARs) (123). Of the four known PAR receptors (PAR 1-4), two are present on human platelets (PAR-1 and PAR-4) and mediate the signaling action of thrombin (123). Thrombin activates PAR receptors by cleaving the N-terminus of the receptor, exposing a tethered ligand that can interact with specific sites on the extracellular loops of the receptor. This causes a conformational change in the receptor, promoting exchange of GTP for GDP initiating signaling, which leads to activation of PI3-kinase and release of cytosolic calcium (Reviewed in (124)).

Thrombin receptor activating peptide (TRAP) has the same amino acid sequence as

the ‘tethered ligand’ (SFLLRN) and can thus bind to the tethered ligand binding sites on the PAR receptor and mimic thrombin receptor activation (125).

1.2.2.1.2 Arachidonic acid

Arachidonic acid is a major polyenoic fatty acid. It is a precursor for an important group of biologically active compounds called the eicosanoids, which include molecules such as thromboxane A2 (TXA2). Arachidonic acid is stored in platelets and released upon activation (126). It has also been found in abundant quantities in psoriasis inflamed skin tissue (127) and is also contained in resting leukocytes (128). Endothelial cells release arachidonic acid in conditions of hypoxia, in response to lysophospholipids and in response to agents that elevate calcium levels (Reviewed in (129)). Exogenous arachidonic acid can penetrate into platelets and be acted upon by cyclooxygenase (COX) enzymes (130). In platelets, arachidonic acid is released from the phospholipids of the membrane by the action of the lipid- cleaving enzyme phospholipase A2 (PLA2). Arachidonic acid is then converted to prostaglandin H2 (PGH2) by PGH2 synthase -1 or -2, also known as COX -1 or -2. It is also converted to 12-(S)-hydroxyeicosatetraenoic acid (HETE) by 12-lipoxygenase

(LOX). In platelets, PGH2 can be converted to TXA2 by thromboxane synthase or to prostaglandin E2 (PGE2) by cyclic prostaglandin E synthase. TXA2 induces irreversible platelet activation by binding to its specific receptors - known as TP receptors. Human platelets have two subtypes of the TP receptors – TPα and TPβ.

TPα couples Gq and inhibits adenyl cyclase activity leading to intra platelet activation signals. TPβ couples to Gi. Both isoforms of TP receptor lead to activation 32

of phospholipase-C (PLC) causing an increase in cytosolic calcium (131,132). The arachidonic acid pathway also leads to the production of PGE2, which at low concentrations (nM), can increase the platelet aggregation response to suboptimal concentrations of different platelet agonists (133,134). Besides being enzymatically cleaved, arachidonic acid can also be non-enzymatically converted to isoeicosanoids such as 8-iso-PGF2α. This compound activates platelets and can also potentiate the response of platelets to suboptimal concentrations of other platelet agonists (135).

1.2.2.1.3 Adenosine Diphosphate

Adenosine diphosphate (ADP), which is released by damaged endothelial cells and red blood cells as wells as from the dense granules of platelets, mediates platelet activation via P2Y1 and P2Y12 G-protein linked nucleotide receptors present on human platelets (136). Stimulation of P2Y1, which is coupled to Gq, causes platelet shape change and mobilization of cytosolic calcium via activation of PLC.

ADP can also stimulate platelets via the P2X1 receptor. This is a ligand-gated ion channel, which allows transmembrane calcium flux (137).

Table 2: Platelets agonists, their receptors and their intracellular effects

1.2.2.2 Platelet granules

It has become evident that platelets are also active players in immunity and inflammation. They have a number of intracellular granules in which they can store biologically active molecules. Upon activation, platelets either expose their granule contents on the platelet surface or release them into circulation. There are three major types of platelet granules: α-granules, dense granules and lysosomes.

α-granules are the most abundant type of granules in the platelet ranging from about 40-100 per platelet. α-granules acquire their content by both biosynthesis as well as endocytosis. Adhesion molecules, such as P-selectin (CD62P), are expressed on the platelet surface upon activation, and chemokines, such as

RANTES, are released into circulation upon α-granule release. It has also been

shown that platelet α-granules contain growth, angiogenesis, and coagulation factors. It has been suggested that platelet α-granules are not homogeneously packed and the release of a certain kind of α-granules may depend on the kind of stimulus received by the platelet (138).

Dense granules are present in much fewer numbers as compared to the α- granules. They contain calcium and magnesium ions, serotonin, ADP and ATP.

Lysosomes have similar contents as most other cell types in which they are present, including cathepsins and hydrolases.

Table 3: Major contents of platelet granules

1.2.3 Interactions between platelets and the complement system

Besides their well-known role as mediators of hemostasis, platelets are also known to play important roles in modulating thrombosis and inflammation.

Excessive or inadvertent platelet activation is common at sites of endothelial damage, underlying many cardiovascular disorders, such as myocardial infarction, unstable angina, and stroke (139). Patients with chronic inflammatory conditions such as unstable atherosclerosis, hypercholesterolemia, and coronary disease have a higher number of activated platelets (140,141) and platelet-leukocyte aggregates

(142) circulating in the blood.

Stimulated platelets activate the complement system on or near their surface

(119,120,143-146). In general, normal inflammatory processes require complement activation for an effective immune response, as well as for efficient removal of spent cells from circulation. In pathological acute and chronic inflammatory diseases (i.e. systemic lupus erythematosus, cancer, atherosclerosis, ischemia/reperfusion injury, neuroinflammation, among others), excessive complement activation contributes to tissue damage and leads to elevated release of pro-inflammatory by-products (i.e.

C5a, C3a) and C5b-9 end product (MAC) (147), which in turn participate in leukocyte recruitment, vascular inflammation, platelet activation, and thrombosis

(148,149). Thus, understanding the molecular mechanisms by which complement activates on stimulated platelets becomes essential for understanding its role in platelet-mediated physiology and disease pathogenesis.

Both the classical (120) and the alternative pathways (119) have been shown to activate on the platelet surface. Activation of the classical pathway on platelets has been associated with the expression of gC1qR on the activated platelet surface

(120). Alternative pathway activation on platelets has been shown to be associated with the expression of P-selectin on the activated platelet surface (119). Recent studies also indicate that complement activates in the microenvironment surrounding the stimulated platelet upon release of chondroitin sulfate-A (145).

Interestingly, complement activation on platelets occurs despite the presence of complement regulatory proteins (150,151). In addition, individuals with paroxysmal nocturnal hemoglobinuria (PNH) or aHUS, diseases in which the activity of one or more complement regulatory proteins (i.e. factor H, CD59, CD55, CD46) is impaired (55,152), have exacerbated complement activation on their platelets

(153,154).

Studies by Del Conde et al. (119) show that C3b binds directly to activated platelets by using P-selectin (CD62P) as a receptor, and they proposed this as a mechanism by which alternative pathway activation occurs on the surface of activated platelets. In contrast, Hamad et al. (146) detected binding of only C3(H2O) to stimulated platelets, which was not the result of proteolytic cleavage of C3 or alternative pathway complement activation. Therefore, the mechanisms by which the alternative pathway of complement activates on stimulated platelets remain controversial. Our study intends to contribute to the understanding of the molecular mechanisms involved in this phenomenon, specifically with regard to the

contribution of complement regulatory protein properdin. We hypothesize that properdin may act as a selective initiator of complement on activated platelets.

1.3 Platelet-leukocyte aggregates

1.3.1 Background

When platelets are activated in circulation at sites of inflammation and vascular injury, they not only adhere to one another and the damaged vascular site, but they also adhere to leukocytes. Higher numbers of circulating platelet-leukocyte aggregates (PLAs) are found in the blood of patients with inflammatory cardiovascular conditions such as unstable atherosclerosis (114,140,141,155-158), hypercholesterolemia (159,160), and coronary disease (140,141). PLAs are considered to be reflective of plaque instability and vascular thrombosis (161), and when PLAs are prevented in an animal model of vascular injury, there is reduced myocardial reperfusion injury and preserved endothelial function (162). In addition, binding of platelets to monocytes leads to increased expression and adhesive capacity of integrins, causing increased monocyte transmigration and atherogenic capacity of monocytes (163). Thus, PLAs are not only indicative of cardiovascular disease, but may also play an important role in the initiation and progression of disease. Platelet-leukocyte binding may also be an initial step before clearance of spent platelets by phagocytosis (164).

1.3.2 Mechanisms of platelet-leukocyte aggregate formation

Platelets communicate with leukocytes through multiple surface receptors as well as soluble molecules. The initial interaction of leukocytes with platelets occurs via binding of P-selectin to its ligand P-selectin GP ligand-1 (PSGL-1) (165). The P- selectin/PSGL-1 interaction induces tyrosine kinase-dependent activation of Mac-

1/CR3 (complement receptor 3; CD11b/CD18), a β2 integrin, by causing conformational changes. This leads to induction of activated Mac-1/CR3 which interacts with GPIb and GPIIbIIIa on the platelet surface, resulting in stable adhesion

(166,167) (Fig 12). PSGL-1 engagement on monocytes leads to enhanced tissue factor activity in monocytes (168,169) along with mobilization of NFκB that leads to

Figure 12: Mechanism of PLA formation. The initial interaction of platelets with leukocytes occurs through engagement of P-selectin by its ligand PSGL-1. This interaction induces activation of Mac-1/CR3 (complement receptor 3; CD11b/CD18), a β2 integrin, by causing conformational changes, leading to induction of activated Mac-1/CR3, which then interacts with GPIb and GPIIbIIIa on the platelet surface, resulting in stable adhesion.

cytokine expression (170). PSGL-1 engagement on neutrophils is required for oxidative burst (171). Activation of Mac-1/CR3 also induces β1 integrin expression, which is important for neutrophil adhesion and migration (172). Subsequent interactions between platelets and leukocytes include: a) CD40-CD40L interaction, causing upregulation of cytokine and chemokine release by monocytes (173); and b)

TREM-1 (triggering receptor expressed on myeloid cells-1) – TREM-1 ligand interaction, causing enhanced neutrophil activation (174).

1.3.3 Activation of complement on platelets and on leukocytes

Platelets can activate complement, on or near their surface, by both the classical (120,145) and the alternative (119) pathways. PLA formation is dependent on complement activation as inhibition of all complement activation leads to a decrease in activation of leukocytes (as measured by CD11b) as well as a decrease in

PLA formation in the presence of TRAP stimulation (145). In the same study the effect of complement was shown to depend on interaction of C5a with its receptor on leukocytes. It has also been shown that patients with E. coli-induced hemolytic uremic syndrome have an increased number of PLAs associated with C3 and C9 when their blood is stimulated with shiga toxin or LPS (175). This study also showed that the dominant pathway that plays a role in shiga toxin-mediated PLA formation is the alternative pathway. This study was conducted with blood in the presence of ethylene glycol tetraacetic acid (EGTA), a strong chelating agent for Ca++ ions which allows activation of only the alternative pathway, or with ethylene

diamine tetraacetic acid (EDTA), a strong chelating agent for metal ions that inhibits all complement pathways by chelating Ca++ and Mg++ ions. Chelating agents like

EGTA and EDTA alter the ion concentrations of blood and thus, interfere with PLA formation (176). As mentioned previously, properdin is mainly produced by leukocytes (76-78). When unfractionated properdin (that contains non- physiological, complement-activating aggregates) is added to citrated whole blood it leads to increased PLA (111). This study, as well as the previously mentioned study

(175), was carried out with citrate-anticoagulated blood. Anticoagulants such as citrate and EDTA significantly inhibit complement activation (177) due to their calcium and magnesium binding properties and may interfere in the formation of

PLA due to the same reason. Thus, the role of alternative pathway activation and physiological properdin in PLA formation, as well as the molecular mechanisms involved, remain unknown.

Endogenous properdin has been detected on the surface of isolated, non- stimulated neutrophils (77) as well as TNF/fMLP-stimulated neutrophils (99), independently from C3 (99). Interestingly, proinflammatory and coagulation- induced stimuli allow neutrophils to activate the alternative pathway of complement in an autocrine and paracrine fashion, despite the presence of membrane-bound complement regulatory proteins, on neutrophil surfaces (99).

Complement activation on neutrophils and platelets results in deposition of C3 fragments such as C3b, iC3b and C3d on these cells. These fragments can be recognized by complement receptors such as CR1, CR2, CR3, and/or CR4 present on the cells (178). Mac-1/CR3, which plays an essential role in PLA formation as 42

described above, is a receptor for iC3b (166,167). Thus, it is possible that interaction of Mac-1 on leukocytes with iC3b on the platelets may contribute to PLA formation.

In addition, complement activation also leads to release of complement products such as C5a fragments and MAC, which could further activate neutrophils, endothelial cells, or platelets in close contact with the neutrophils (149,179-181), leading to an increase in expression of complement receptors (182) and further contributing to a pro-inflammatory microenvironment. Production of properdin by various cells at pro-inflammatory sites may lead to transient high concentrations of properdin in the microenvironment, making it available to bind to activated platelets and neutrophils. Our study aims to contribute to elucidating the specific roles of properdin in the formation of PLAs.

Chapter 2

Hypothesis

Properdin promotes complement activation on platelets and the formation of platelet-leukocyte aggregates.

Sub-Hypothesis 1: Properdin interacts directly with activated platelets and

initiates complement activation.

Sub-Hypothesis 2: Properdin increases the formation of platelet-leukocyte

aggregates by promoting complement activation on the platelet/leukocyte

interface.

Chapter 3

Specific Aims

1. To determine the molecular mechanisms involved in alternative pathway activation on activated platelets.

1.1. Does properdin bind to platelets?

1.2. Does properdin bound to platelets initiate complement activation?

1.3. Is factor H regulation needed for controlling properdin-mediated

complement activation on platelets?

1.4. What is the molecular mechanism by which the platelet-bound properdin

activates complement?

1.5. Does native properdin released by neutrophils bind to activated

platelets?

2. To determine the role of properdin in platelet-leukocyte aggregate formation.

2.1. Does properdin increase platelet-leukocyte aggregate formation and is

this effect dependent on complement activation?

2.2. Is the effect of properdin on platelet-leukocyte aggregate formation mediated by C5a receptors?

2.3. Is the properdin-mediated increase in platelet-leukocyte aggregate formation controlled by factor H?

Chapter 4

Materials and Methods

4.1 General Methods

4.1.1 Buffers

The buffers used were:

A.1. Citrate buffer: 9.35 mM Na3Citrate, 4.75 mM Citric acid, 17.35 mM Dextrose,

145 mM NaCl, pH 6.5

A.2. Tyrode’s buffer: 136.9 mM NaCl, 2.7 mM KCl, 983.8 μM MgCl2 6H2O, 3.2 mM

Na2HPO4, 3.5 mM HEPES, 0.35% BSA, 5.5 mM Dextrose, 2 mM CaCl2; pH 7.4.

A.3. Tyrode/PGE/Hep buffer: Tyrode’s buffer containing 1 µM Prostaglandin E1

(PGE1) and 2 IU/mL Heparin

A.4. Modified HEPES Tyrode buffer: 119 mM NaCl, 5 mM KCl, 25 mM HEPES

buffer, 2 mM CaCl2, 2 mM MgCl2, 0.35% BSA, 6 g/liter glucose

A.5. GVB=: 5 mM veronal, 145 mM NaCl, 0.004% NaN3, 0.1% Gelatin.

A.6. GVBE: GVB= + 10 mM Ethylene diamine tetraacetic acid (EDTA)

A.7. Phosphate buffered saline (PBS): 10 mM sodium phosphate, 140 mM NaCl,

0.02% NaN3, pH 7.4

A.8. 0.1M Magnesium-Ethylene glycol tetraacetic acid (MgEGTA): 0.1 M EGTA

+ 0.1 M Mg++. Adjust pH to 7.3 (EGTA will dissolve at this pH)

A.9. Mono S buffer A: 50 mM sodium phosphate, pH 6.0

A.10. Mono S buffer B: 50 mM sodium phosphate, 0.5 M NaCl, pH 6.0

A.11. Red blood cell (RBC) lysis/fixing buffer (Biolegend)

A.12. Hanks balanced salt solution with calcium and magnesium (HBSS+2)

(Gibco)

4.1.2 Antibodies

The following murine monoclonal antibodies were used in this study:

B.1. IgG1 anti-human properdin (#1; Quidel)

B.2. IgG1 isotype control (eBioscience)

B.3. IgG1 anti-human C3/C3b (Cedarlane) 48

B.4. IgG1 anti-human iC3b (Cedarlane)

B.5. PE-conjugated IgG1 anti-human C3/C3b (Cedarlane)

B.6. APC-conjugated IgG1 anti-human CD42b (Biolegend)

B.7. PE/Cy5-conjugated IgG1 anti-human CD62P (Biologend)

B.8. FITC-conjugated IgMκ PAC-1 monoclonal antibody (BDBiosciences)

B.9. IgG2a anti-human C5b-9 neo-epitope (Dako)

B.10. IgG2a anti-human factor Bb neoantigen (Abd Serotec)

B.11. AF488-conjugated IgG1 anti-human CD11b (Biolegend)

B.12. PE-conjugated IgG2a anti-human CD16b (Biolegend)

B.13. PE-conjugated IgG1 anti-human CD45 (Biolegend)

B.14. IgG1 anti-human factor H (#2; Quidel)

B.15. FITC-conjugated IgG1 anti-human CD46 (Serotec)

B.16. FITC-conjugated IgG1 anti-human CD55 (Biolegend)

B.17. FITC-conjugated IgG1 anti-human CD59 (Chemicon)

B.18. FITC-conjugated mouse IgG1 isotype control (Chemicon)

The following polyclonal antibodies were used:

B.19. AF488-conjugated goat anti-mouse polyclonal IgG (Invitrogen)

B.20. F(ab’)2 polyclonal goat anti-C3b IgG (LifeSpan BioSciences)

B.21. Affinity purified goat anti-properdin polyclonal IgG

4.1.3 Other reagents

C.1. Lepirudin (Refludan) (Bayer)

C.2. Compstatin (Tocris Bioscience)

C.3. Compstatin control peptide (Tocris Bioscience)

C.4. C5aR antagonist W-54011 (Calbiochem)

C.5. Normal human serum (NHS) (CompTech)

C.6. Properdin depleted serum (CompTech)

C.7. Factor B (fB) (CompTech)

C.8. Factor D (fD) (CompTech)

C.9. C5a (CompTech)

C.10. Thrombin (Sigma)

C.11. Arachidonic acid (AA) (Chronolog)

C.12. Adenosine di-phosphate (ADP) (Sigma)

C.13. FITC-labeled Annexin V (Molecular Probes)

C.14. Chondroitin sulfate-A (Sigma)

C.15. Proteinase K (New England Biolabs)

C.16. Phorbol 12 myristate 13 acetate (PMA) (Enzo Life Sciences)

C.17. HALT protease inhibitor (Thermo Scientific)

C.18. Thrombin receptor activating peptide (TRAP) (Bachem)

4.1.4 Serum and additional complement proteins

Purification of properdin was carried out as previously described (94).

Briefly, freshly collected, never frozen, pooled normal human plasma (obtained from

Innovative Research) was passed over an anti-properdin affinity column using an

ÄKTApurifier FPLC system (GE). The bound properdin was eluted using glycine-HCl.

In order to eliminate of remaining impurities, the properdin was first passed over a

QAE column, which is a strong anion exchanger, and subsequently passed over an anti-IgG column, an anti-albumin column, and an anti-C1q column. The purified

protein was stored at -80°C. The purity of the final product was assessed by running a 10% SDS-PAGE (183) (Fig. 13A) and Ouchterlony test (184). The function of the purified properdin was tested using an AP50 hemolytic assay. Briefly, lysis of rabbit erythrocytes (ERs) was measured by mixing 5x107cells on ice with GVB=, 0.1M

MgEGTA (10 mM final concentration), and properdin-depleted serum (20% final concentration) in the presence of increasing volumes of a 1/10 dilution of NHS to achieve a final concentration of 0.1 - 1.5% of NHS in 100 μl or in the presence of increasing volumes of a 1/1000 dilution of purified properdin to achieve a final concentration of 0.01 – 0.15 μg/ml of purified properdin in 100 μl. The mix was immediately transferred to 37°C water bath and incubated for 20 min. To determine the extent of hemolysis induced by freshly purified properdin versus the previous batch of properdin or NHS, 500µl cold GVBE= was added and the samples were centrifuged for 3 min at 2000g. Supernatant (200µl) was transferred to a microtiter plate and the OD of the supernatant was determined at 414 nm. The percent lysis was determined by subtracting the absorbance at 414 nm (A414nm) of the background lysis control and dividing by the maximum lysis in water. The newly purified properdin conserved 100% of the activity of the NHS and 136% activity of the previous batch of properdin that was used as a control (Fig. 13B).

Purification of C3 (185) and the generation of C3b (186) were carried out as previously described in the cited references, by other laboratory members.

A Reduced Non-Reduced B kDa 100 188 98 80 62 49 38 60 28 1/10 NHS # 1

%Lysis 1/10 NHS # 2 17 40 1/1000 P #23 14 1/1000 P #24 20 6 EDTA control 3 0 0 5 10 15 Volume (ul)

Figure 13: Purification of human properdin from plasma. (A) Purity of the isolated properdin was checked by running an SDS-PAGE and comparing with the previous batch of properdin. Properdin appeared as a single band free of any visible impurities at the expected molecular weight of ~53kDa. (B) Functional analysis of

purified properdin by AP50 assay was carried out as follows: Rabbit erythrocytes

(ERs) (5x107cells) were mixed on ice with GVB=, 0.1M MgEGTA (10 mM final concentration) and properdin depleted serum (20% final concentration) in the presence of increasing concentrations of two different batches of NHS (0.1 – 1.5%) or purified properdin (present batch # 24 and a previous batch; 0.01 – 0.15 μg/ml). The mix was immediately transferred to 37°C water bath and incubated for 20 min. Hemolysis was assessed by adding 500 µl cold GVBE= and centrifuging for 3 min at 2000g. Supernatant (200 µl) was transferred to a microtiter plate and the OD of the supernatant was determined at 414 nm. The percent lysis was determined by

subtracting the A414nm of the background lysis control and dividing by the maximum lysis in water. X axis: 0, 1, 3, 6, 9, 12, 15 μl on the X axis represents 0, 0.1, 0.3, 0.6, 0.9, 1.2 and 1.5 % serum (for NHS#1 and #1) and 0, 0.01, 0.03, 0.06, 0.09, 0.12, 0.15 μg/ml properdin (for P # 23 and # 24).

4.1.5 Separation of physiological forms of properdin

Physiological polymeric forms of properdin (P2-P4) were separated from non-physiological aggregated forms (Pn) by gel filtration chromatography. The Pn

forms are known to accumulate after prolonged storage and freeze/thaw cycles and to induce non-specific complement activation in solution (86) and on certain surfaces (94,187)). Briefly, pure properdin (5 mg) was loaded onto a Phenomenex

Bio Sep-Sec-S4000 column (600 x 7.8 mm) with a guard column (75 x 7.8 mm), and eluted at a flow rate of 0.5 ml/min in PBS. The concentration was measured by spectrophotometry using an extinction coefficient of 1.87. Purified, physiological forms of properdin were stored at 4°C and used within two weeks of separation, as previously described (86).

4.1.6 Platelet isolation and activation

Human platelets were isolated from the blood of healthy donors. Blood was collected via venipuncture using protocols approved by The Institutional Review

Board from the University of Toledo College of Medicine and Life Sciences. Written informed consent was obtained from all donors, in accordance with the Declaration of Helsinki. Blood was drawn into ACD tubes (BD Vacutainer) and mixed by gentle inversion of the vacutainer tubes to allow uniform mixing of the anti-coagulant with the blood, while avoiding excessive agitation to prevent activation. Platelet-rich plasma was separated by centrifugation at 200g for 15 min at room temperature

(RT) with no brake to avoid mixing of the layers that separated during centrifugation. The blood separates into (i) a bottom red blood cell layer which occupies 50-80% of the total volume, (ii) a middle very thin band of white blood cells, and (iii) a top straw-colored layer of platelet rich plasma (PRP). The PRP was

collected carefully and slowly using a glass Pasteur pipette without disturbing the lower layers and transferred to a new 15ml conical tube. The platelets were washed twice using citrate buffer (which helps prevent platelet activation) at 440g for 10 min at RT. The platelets were counted using the center RBC chamber of a hemocytometer slide and were then resuspended to a final concentration of 1x108 cells/ml in Tyrode’s buffer, which is an iso-osmotic phosphate buffer with physiological pH containing glucose as a source of metabolic energy and physiological concentrations of divalent cations.

Agonists used to activate the platelets included thrombin at 1 IU/ml (or varying doses as specified in the figure legends), arachidonic acid at 1 mM (or varying doses as specified in the figure legends), or ADP at 20 µM for 30 min at 37°C.

The platelets were then washed once with Tyrode/PGE/Hep buffer (to prevent further platelet activation and aggregation) by centrifuging at 2000g for 10 min at

RT. To assess platelet activation, expression of CD62P (using PE/Cy5-mouse IgG anti-human CD62P) was detected on CD42b+ (APC-mouse IgG anti-human CD42b) platelets. Finally, platelets were washed and fixed with 1% paraformaldehyde for 30 min at 4°C, prior to acquisition using a BD FACSCalibur flow cytometer (BD

Biosciences). A minimum of 10,000 events per sample were acquired and the data were analyzed using FlowJo software version 7.6.5.

4.2 Specific Methods (separated by Aim involved)

4.2.1 Methods Aim 1: To determine the molecular mechanisms involved in alternative pathway activation on activated platelets.

4.2.1.1 Measurement of properdin binding to platelets

Non-activated, thrombin-activated, arachidonic acid-activated, or ADP- activated platelets (2x106 platelets/100 µl) were incubated with properdin P2-P4 forms (0-25 μg/ml) in Tyrode/PGE/Hep for 1 hour at RT. Platelets were washed twice with Tyrode/PGE/Hep by centrifuging at 2000g for 10 min at RT. Binding of properdin was assessed by flow cytometry using an anti-properdin monoclonal antibody (used at 10 µg/ml in 100µl), followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody (used at 10 µg/ml in 100µl). After washing, platelets were stained with APC-mouse anti-human CD42b and with PE/Cy5-mouse anti-human

CD62P (both used at 10 µg/ml in 100µl). Finally, platelets were washed, fixed, and the data were acquired and analyzed as described in section 4.1.6. In some experiments, properdin was added to the activated platelets or to C3b-coated sheep erythrocytes (ESC3b; control) in the presence of F(ab’)2 polyclonal goat anti-C3b antibodies.

4.2.1.2 Measurement of platelet activation by properdin

To assess whether properdin can activate platelets, washed human platelets

(1 x 108/ml) were incubated without agonist or with thrombin, arachidonic acid, or properdin (25µg/ml) for 30 min at 37°C. The platelets were then washed and activation was assessed as a measure of CD62P expression using a PE/Cy5-labeled anti-CD62P monoclonal antibody (used at 10 µg/ml in 100µl), gpIIbIIIa expression using a FITC-labeled PAC-1 monoclonal antibody (used at 0.01 µg/ml in 100µl), or

Annexin V binding using FITC-labeled Annexin V. After washing, an APC-labeled

CD42b monoclonal antibody was used to gate on the platelets. Platelets were then washed, fixed, acquired and analyzed as described in section 4.1.6.

4.2.1.3 Measurement of presence of properdin or C3 components on platelet surface

Platelets (1x108 platelets/ml) were activated using thrombin (1UI/ml) or arachidonic acid (5 mM). Platelets were washed and the presence of C3 components or properdin on the surface of platelets was determined by FACS using an anti-

C3/C3b monoclonal antibody (5 µg/ml in 100µl), anti-properdin monoclonal antibody (10 µg/ml in 100µl), or an IgG1 isotype control antibody (either 5 or 10

µg/ml in 100µl) followed by Alexa Fluor 488-conjugated anti-mouse IgG (10 µg/ml in 100µl). After washing an APC-labeled CD42b monoclonal antibody (10 µg/ml in

100µl) was used to gate on the platelets. Finally, platelets were washed, fixed, acquired and analyzed as described in section 4.1.6.

4.2.1.4 Measurement of properdin binding in presence of chondroitin sulfate-A or proteinase K

For some experiments, properdin (10 μg/ml) was incubated with 0-5 mg/ml chondroitin sulfate-A for 30 min at RT. NA or thrombin-activated platelets (2x106 platelets/100 µl) were then added to the mixture and incubated for 60 min at RT.

Properdin binding was assessed as described in section 4.2.1.1.

To assess if properdin binding to platelets was through a protein receptor, thrombin-activated platelets were incubated with 0-100 µg/ml proteinase K at RT for 30 min. Platelets were washed and incubated with properdin for 60 min at RT.

Properdin binding was assessed as described in section 4.2.1.1. As a control for proteinase K treatment, the level of CD62P expression was assessed by FACS using a

PE/Cy5-labelled anti-CD62P monoclonal antibody (10 µg/ml in 100 µl).

4.2.1.5 Measurement of factor H binding to platelets

Non-activated, thrombin-activated or arachidonic acid-activated platelets

(2x106 platelets/100µl) were incubated with factor H (100 μg/ml) in

Tyrode/PGE/Hep for 1 hour at 37°C. Platelets were washed with Tyrode/PGE/Hep by centrifuging at 2000g for 10 min at RT. Binding of factor H was assessed by flow cytometry using an anti-factor H monoclonal antibody (10 µg/ml in 100µl), followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody (10 µg/ml in 100µl).

After washing, platelets were stained with APC-mouse anti-human CD42b monoclonal antibody and with PE/Cy5-mouse anti-human CD62P monoclonal

antibody (both at 10 µg/ml in 100µl). Finally, platelets were washed, fixed, and the data were acquired and analyzed as described in section 4.1.6.

4.2.1.6 Measurement of CD46, CD55 and CD59 on platelet surface

Non-activated or thrombin (1UI/ml) -activated platelets (1x108 platelets/ml) were washed and the presence of CD46, CD55, CD59 on the surface of platelets was determined by FACS using an anti-CD46 (2.5 µg/ml in 100µl), anti-CD55 (2.5 µg/ml in 100µl), or anti-CD59 (concentration unavailable, used at 1:40 dilution) monoclonal antibodies (or an IgG1 isotype control). An APC-labeled CD42b monoclonal antibody (10 µg/ml in 100µl) was used to gate on the platelets. Finally, platelets were washed, fixed, acquired and analyzed as described in section 4.1.6.

4.2.1.7 Measurement of alternative pathway complement activation on platelets

(C3b/iC3b and C5b-9 deposition)

Non-activated, thrombin-activated, or arachidonic acid-activated platelets

(2x106 platelets/100 µl) were incubated in the absence or presence of properdin.

Washed platelets were then incubated with properdin-depleted serum (60%) for various time points at 37°C. In separate experiments, platelets (2x106 platelets/100

µl) were incubated with the properdin-depleted serum as described above, but in the presence of 25 µM rH19-20, a competitive inhibitor of factor H-mediated cell surface regulation (54,55,188-190). For some experiments, platelets were incubated with properdin-depleted serum in the presence of 25 µM of aHUS-associated mutant forms of rH19-20: D1119G (mutation in domain 19), L1189F or R1215G (mutations 59

in domain 20). In all experiments, platelets were incubated with serum in the presence of 5 mM Mg-EGTA, which strongly chelates Ca++ ions needed for classical and lectin pathway activation, thus allowing to measure complement activation solely by the alternative pathway. Negative controls were incubated with serum in presence of 10 mM EDTA, which chelates both Ca++ and Mg++ ions to inhibit complement activation. Complement activation was stopped by washing samples with cold Tyrode’s buffer containing 10 mM EDTA. Deposition of C3b was detected using PE-anti C3/C3b or an unlabeled anti-C3b monoclonal antibody (5 µg/ml in

100 µl), followed by AF488-goat anti-mouse IgG (10 µg/ml in 100 µl). Similarly,

C5b-9 was detected using anti-C5b-9 neo-epitope antibody (0.95 µg/ml in 100µl) followed by AF488-goat anti-mouse IgG (10 µg/ml in 100µl). The platelets were stained with APC-mouse anti-human CD42b and PE/Cy5-mouse anti-human CD62P antibodies (both at 10 µg/ml in 100µl) and analyzed as described in section 4.1.6.

For measurement of iC3b deposition, the only differences in the method were that the platelets were incubated with pooled properdin (2 µg/ml for arachidonic acid- activated platelets and 10 µg/ml for thrombin-activated platelets) and stained with an unlabeled anti-iC3b monoclonal antibody (5 µg/ml in 100 µl). The rest of the method was identical to the measurement of C3b deposition.

4.2.1.8 Separation of C3 and C3(H2O)

C3(H2O) was separated from C3 by cation exchange chromatography as described previously (185). Briefly, C3 was incubated at 37°C for 2 hours to convert intermediate/inactive forms of C3 to C3(H2O). The sample was diluted with Mono S 60

buffer A and loaded onto a 1 ml Mono S column. The column was washed with buffer A and eluted using a 20 ml salt gradient (0-100%) of Mono S buffer B at a flow rate of 1 ml/min. The peaks were collected using a fraction collector, and purified C3 and C3(H2O) were dialyzed against PBS. The concentration was determined by spectrophotometry (using and extinction coefficient of 1.1), and both

C3 and C3(H2O) were stored at 4°C, and used within two weeks of separation.

4.2.1.9 Measurement of recruitment to the platelet surface of C3 components by properdin and of properdin by C3(H2O)

Non-activated, thrombin-activated, or arachidonic acid-activated platelets

(2x106 platelets/100 µl) were incubated with or without properdin (25 μg/ml) in

Tyrode/PGE/Hep buffer for 1 hour at RT. Platelets were washed twice with

Tyrode/PGE/Hep and then incubated with C3, C3b, or C3(H2O) (100 μg/ml), or, in the case of thrombin-activated platelets, with varying concentrations of C3(H2O) (0-

100 µg/ml). Binding of C3 components was assessed by flow cytometry using an anti-C3 monoclonal antibody (5 µg/ml in 100µl), followed by an Alexa Fluor 488- conjugated anti-mouse IgG antibody (10 µg/ml in 100µl). Alternatively, the activated platelets were incubated with C3(H2O) (100 µg/ml) first, followed by washing and subsequent incubation with P3 (0-25 µg/ml). Binding of properdin was assessed by flow cytometry as described in section 4.2.1.1.

4.2.1.10 Measurement of C3 convertase [C3(H2O),Bb and C3b,Bb] formation on platelets

Thrombin-activated or arachidonic acid-activated platelets (2x106 platelets/100 µl) were incubated with one of the following for 1 hour at RT: (a)

Tyrode’s buffer alone, (b) properdin (25 µg/ml), (c) C3(H2O) (50 or 100 µg/ml), (d)

C3b (50 or 100 µg/ml), or (e) first with properdin, followed by washing, and then

C3(H2O) or C3b (1 hour each). Alternatively, C3(H2O) or C3b was added first, followed by washing and then properdin (1 hour each). After washing, the ability to form Bb was assessed by resuspending the pellet in 100µl of factor D (2 µg/ml) along with factor B (80 µg/ml) for 30 min at RT. The formation of Bb was assessed by flow cytometry using anti-human complement factor Bb neo-epitope monoclonal antibody (10 µg/ml in 100µl) followed by an Alexa Fluor 488-conjugated anti- mouse IgG antibody (5 or 10 µg/ml in 100µl). The platelets were then stained with

APC-mouse anti-human CD42b and PE/Cy5-mouse anti-human CD62P monoclonal antibodies (both at 10 µg/ml in 100µl) and analyzed as described in section 4.1.6.

4.2.1.11 Measurement of properdin binding to platelets in the presence of NHS

Non-activated, thrombin-activated or arachidonic acid-activated platelets

(2x106 platelets/100µl) were incubated with or without properdin (25µg/ml) in

Tyrode/PGE/Hep, in the presence of NHS (0-60%; in 10 mM EDTA) for 30 min at

RT. Platelets were washed with Tyrode/PGE/Hep and binding of properdin was assessed by flow cytometry as described in section 4.2.1.1.

4.2.1.12 PMN isolation, activation, and assessment of the ability of properdin and C3 components from PMN supernatants to bind to stimulated platelets

Fresh blood was drawn from healthy volunteers into EDTA tubes (BD

Vacutainer), and used immediately (within 2 hrs of drawing). Polymorphonuclear

(PMN) cells were isolated by using a PolymorphprepTM gradient (Axis-Shield PoC

AS) following manufacturer’s instructions (>90% pure PMN). Briefly, 20ml of whole blood was carefully layered over 20ml of polymorphprep at RT in a 50ml conical tube and spun at 500g for 35 min without brakes to avoid disruption of layers that separated during centrifugation. The blood will separate into (i) a top band of plasma, (ii) a second band of mononuclear cells (iii) a band of PMN cells and (iv) a bottom RBC pellet. The plasma layer as well as the first buffy band of cells was removed using a Pasteur pipette and then the buffy coat containing the PMNs was harvested using a Pasteur pipette and transferred to a fresh 15ml conical tube. The collected cell suspension was diluted with an equal volume of half concentration

HBSS2+ (Gibco) and PMNs were spun down by centrifuging at 400g for 10 min. This allowed for the removal of any residual contaminating RBCs. An additional RBC lysing step was included by resuspending the pellet in RBC lysis buffer, incubating 7 min at 37°C, and then spinning the cells down. The cells were finally washed with

HBSS+2 + 0.2% BSA and resuspended in the same.

PMN cells (2.5x107 cells/ml), in HBSS+2 + 0.2% BSA, were activated using

PMA (10 ng/ml) for 30 min at 37°C, as previously described (77). The supernatant was collected after centrifuging the cells at 600g for 10 min at 4°C and centrifuged

again at 13,000g for 10 min to remove cell debris. HALT protease inhibitor (1:100) was added to the supernatant. Activation of PMN was verified by flow cytometry by double-gating on forward scatter (FSC) and side scatter (SSC) for PMN cells and subsequently on CD16b positive cells. CD11b level was then measured on the gated population. Supernatant (50 μl) of PMA-activated neutrophils was incubated with non-activated or thrombin-activated (1 U/ml) platelets (2x106) in a final reaction volume of 100 μl containing 10 mM EDTA. Binding of properdin was assessed as described in section 4.2.1.1. Binding of C3 components was assessed as described in section 4.2.1.3.

4.2.2 Methods Aim 2: To determine the role of properdin in platelet-leukocyte aggregate formation.

4.2.2.1 Measurement of the formation of platelet-leukocyte aggregates in the presence of TRAP and/or properdin

Human whole blood was collected via venipuncture from healthy donors.

Blood was drawn into vacutainer tubes (BD Vacutainer) containing 50 µg/ml lepirudin (Refludan). Blood (20 µl) was stimulated with 0-20 µM TRAP in the presence or absence of purified properdin (0-100µg/ml) and modified HEPES

Tyrode’s buffer (total reaction volume of 80 µl) at 37°C for 15 min. The reaction was stopped using an RBC lysis/fixation solution, which lyses the red blood cells while fixing all other cells at the same time. The samples were then washed and stained using PE-labeled anti-CD45 (leukocyte marker) (0.375 µg/ml in 80 µl) and APC- labeled anti-CD42b (platelet marker) monoclonal antibody (2.5 µg/ml in 80 µl), 64

along with appropriate isotype controls (APC Mouse 1gG1 isotype control). A total of 20,000 events of the granulocyte + monocyte population were acquired using a flow cytometer.

4.2.2.2 Measurement of alternative pathway-mediated hemolysis of rabbit erythrocytes in the presence of compstatin (control experiment)

Compstatin is a cyclic peptide inhibitor of complement, which selectively binds C3 and inhibits proteolytic cleavage by the C3 convertase, thus inhibiting complement activation ((191); reviewed in (192)). Washed ERs (2x107/ml in 100µl) were incubated with 8% pooled NHS along with increasing concentrations of compstatin or compstatin control peptide in the presence of 50 mM Mg-EGTA in

GVB= for 20 min at 37°C (with mixing every 5 min). Cold GVBE (400µl) was then added to the samples to stop complement activity (due to the presence of EDTA in the buffer as well as the cold temperature of the buffer) and the samples were centrifuged at 1000g for 2 min at 4°C to pellet the erythrocytes. Supernatants (200

µl) were collected and the A414nm was measured using a spectrophotometer.

4.2.2.3 Measurement of the ability of anti-properdin antibody to inhibit the binding of properdin to platelets

Purified forms of properdin (10 µg/ml) were incubated with anti- properdin#1 (0 – 100 µg/ml) or IgG1 isotype control (100 µg/ml) in the presence of

Tyrode/Hep/PGE for 20 min at RT. The properdin-antibody mix was incubated with

arachidonic acid-activated platelets (2 x 106/100 µl) for 1 hr at RT, washed and stained with an affinity purified polyclonal anti-properdin antibody (1:200 dilution in 100 µl) followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody (5 or

10 µg/ml in 100 µl). The platelets were then stained with APC-mouse anti-human

CD42b and PE/Cy5-mouse anti-human CD62P monoclonal antibodies (both at 10

µg/ml in 100µl) and analyzed as described in section 4.1.6. As a positive control experiment for the antibody we measured inhibition of properdin binding to known ligand C3b on sheep erythrocytes (EsC3bs). For this, the properdin-antibody mix described above was incubated instead with EsC3bs (5 x 106/100 µl) for 1 hr at 4°C, washed and stained with an affinity purified polyclonal anti-properdin antibody

(1:200 dilution in 100 µl) followed by an Alexa Fluor 488-conjugated anti-mouse

IgG antibody (5 or 10 µg/ml in 100 µl).

4.2.2.4 Measurement of formation of platelet-leukocyte aggregates in the presence of complement inhibitors and/or properdin

Lepirudin anti-coagulated blood (20 µl) was stimulated with 0-20 µM TRAP in the presence or absence of purified properdin (0-100 µg/ml) and modified HEPES

Tyrode’s buffer (total reaction volume of 80 µl) at 37°C for 15 min. For some experiments, this stimulation was carried out either in the presence of 100 µg/ml of anti-properdin#1 monoclonal antibody (that inhibits properdin convertase- stabilizing function) or IgG1 isotype control. For experiments designed to test the role of all complement activity, blood was incubated with 50 µM compstatin for 5

min at RT prior to the stimulation step. For experiments designed to test the role of

C5aR in PLA formation we used a C5aR antagonist (C5aRa), W-54011, to inhibit the

C5a-C5aR interaction. This antagonist is a tetrahydronaphthalenyl carboxamide compound that has been shown to inhibit the binding of C5a to human neutrophils, as well as C5a-induced intracellular Ca2+-mobilization, chemotaxis, and reactive oxygen species (ROS) generation. The blood was incubated with 50 µM C5aRa (W-

54011) or DMSO (C5aRa dilutent) control for 5 min at RT prior to the stimulation step. Finally, the reaction was stopped using an RBC lysis/fixation solution, which lyses the red blood cells while fixing all other cells at the same time. The samples were then washed, stained and analyzed as stated in section 4.2.2.1.

4.2.2.5 Assessment of inhibition of C5a-mediated neutrophil activation by C5aRa

Neutrophils were isolated as described in section 4.2.1.10. Isolated, washed neutrophils were incubated with 50 µM C5aRa (W-54011), or DMSO control for 5 min at RT prior to activation with 0.2 µM C5a at 37°C for 20 min. Neutrophils were then washed using HBSS+2 (Gibco) + 0.2% BSA and stained with AF488-mouse anti human CD11b and PE-mouse anti-human CD16b monoclonal antibodies, along with appropriate isotype controls. The neutrophil population was selected by double- gating on FSC and SSC for PMN cells and subsequently on CD16b positive cells.

Neutrophil activation was assessed by measuring levels of CD11b (as a measure of

CR3/Mac-1) on the gated population.

4.2.2.6 Measurement of the formation of platelet-leukocyte aggregates in the presence of rH19-20 (a competitive inhibitor of factor H)

Lepirudin anti-coagulated blood (20 µl) was stimulated with 0-20 µM TRAP in the presence or absence of rH19-20 (0-22.5 µM) and modified HEPES Tyrode’s buffer (total reaction volume of 80 µl) at 37°C for 15 min. To test the effect of aHUS- associated mutations on the ability of rH19-20 to compete with factor H binding to

PLAs, TRAP stimulation was carried out in the presence of the following aHUS- related mutants of rH19-20: D1119G (mutation in domain 19), L1189F or R1215G

(mutations in domain 20). For some experiments, this stimulation was carried out in the presence of 100 µg/ml of anti-properdin#1 monoclonal antibody or IgG1 isotype control in order to inhibit endogenous properdin in the blood or properdin being actively released by the leukocytes. For experiments designed to test the role of all complement activity, the blood was incubated with 50 µM compstatin, an inhibitor of all complement activation, for 5 min at RT prior to the stimulation step.

The reaction was stopped using an RBC lysis/fixation solution (Biolegend). The samples were then washed, stained and analyzed as described in section 4.2.2.1.

4.3 Statistics

The data were analyzed with GraphPad Prism 4.0 software. Unpaired

Student’s t tests were used to determine the statistical significance of the difference between the groups assessed for properdin binding, CD62P expression, C3b or C5-9 deposition and PLA formation. One way analysis of variance Dunnett’s Multiple

comparison test was used when comparing all the conditions of resting and thrombin-stimulated platelets incubated with different C3 components (C3,

C3(H2O), C3b) in the presence or absence of properdin. Unpaired Student’s t tests were used to determine significance of differences in binding of C3 components to arachidonic acid-activated platelets (with or without properdin). Two way analysis of variance-Bonferroni’s posttest was applied to determine statistical significance of whether properdin recruits C3(H2O) to the surface of activated platelets, or vice versa, at different C3(H2O) or properdin concentrations, respectively.

Chapter 5

Results

5.1 Specific Aim 1: To determine the molecular mechanisms involved in alternative pathway activation on activated platelets.

A significant part of the results presented here was recently published in The

Journal of Immunology (Appendix A). Authors retain the permission to reuse original figures or tables in the author's own work.

5.1.1 Does properdin bind to platelets?

Binding of pure human properdin to activated human platelets was first assessed by flow cytometry. Figure 14 shows that pure properdin bound to both washed non-activated and activated platelets. Properdin subjected to prolonged storage and/or freeze/thaw cycles is known to accumulate non-physiological, high molecular weight aggregates (Pn) (86,107). These Pn forms, also known as “activated properdin”, have the ability to spontaneously activate complement in serum

(86,107) and bind non-specifically to surfaces ((94,108); reviewed in (187)). Since we know that unseparated properdin has non-physiological Pn forms, it was

essential to purify physiological forms of properdin and separate out the non- physiological aggregate forms of properdin.

A NA B Act 250 250 No P No 200 200 P 150 150

100 100 Pure Properdin Pure Properdin 50 50

Counts 0 Counts 0 100 101 102 103 104 100 101 102 103 104 Properdin binding

Figure 14: Unfractionated properdin binds to both non-activated and activated platelets. Non-activated (NA) (A) or thrombin-activated (1 U/ml; Act) (B) platelets (2x106 platelets/100 µl) were incubated in the presence or absence of 25 μg/ml of unfractionated properdin for 30 min at RT in Tyrode/PGE/Hep for 1 hour at RT. Platelets were washed twice with Tyrode/PGE/Hep (each time centrifuged at 2000g for 10 min at RT) and binding of properdin was assessed by FACS using an anti-properdin monoclonal antibody, followed by an Alexa Fluour 488-conjugated anti-mouse IgG antibody. After washing, platelets were stained with APC-mouse anti-human CD42b to gate on the platelet population. Finally, platelets were washed and fixed with 1% paraformaldehyde for 30 min at 4°C, prior to acquisition using BD FACS Calibur flow cytometer. A minimum of 10,000 events per sample were acquired and the data was analyzed using FlowJo software version 7.6.5. (Supplemental figure S1 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

5.1.1.1 Separation of physiological forms of properdin from purified human properdin

The native physiological forms of properdin (P2, P3 and P4) were separated from the non-physiological aggregate forms (Pn) by size exclusion chromatography, as previously described (94) and summarized in the Materials and Methods section.

Based on the size of the aggregates, the properdin forms resolved as separate peaks and eluted in the order of Pn, P4, P3 and P2 forms as shown in figure 15.

P 1000 3 Figure 15: Separation of physiological properdin forms 800 from purified properdin. The pure, frozen properdin sample was thawed P2 600 and separated by size exclusion P4 chromatography. The thawed 400 properdin sample, in PBS, was loaded onto a Phenomenex BioSep-SEC-

Pn Absorbance(280 nm) 200 S4000 gel filtration column (7.8 × 600 mm) and was eluted at a flow 0 rate of 0.5 ml/min. 0 10 20 30 40 50 60 Time (min)

5.1.1.2 Do purified physiological forms of properdin bind to platelets?

We purified the physiological forms of properdin as mentioned above, and tested the binding of the physiological forms of properdin to washed non-activated

(Fig. 16A) and thrombin-activated platelets (Fig. 16B). The data indicate that P2, P3 and P4 bind only to activated platelets (Fig. 16A-C), but not to non-activated platelets. The data also indicate that binding of properdin to activated platelets was dose-dependent (Fig. 17A) as well as time-dependent (Fig. 17B).

A B C 250 250 120 No NA NA Act Thr Act Thr P No 100 200 200 P 80

150 P2 150 P2 60 P3 P3 100 P4 100 P4 40 P2-P4 P2-P4 50 50 Pbinding (GMFI) 20 ns

Counts 0 Counts 0 0 0 1 2 3 4 0 1 2 3 4 10 10 10 10 10 10 10 10 10 10 No P P2 P3 P4 P2-P4 Properdin binding

Figure 16: Analysis of binding of physiological forms of properdin to activated platelets stimulated by different agonists. Non-activated (NA) (A) or thrombin-activated (Act Thr) (B) platelets (2x106 platelets/100 µl) were

incubated in the presence or absence of 25 μg/ml of P2, P3, P4 or P2-P4 pool (in a ratio of 1:2:1) for 30 min at RT in Tyrode/PGE/Hep buffer. Platelets were washed and binding of properdin was assessed by FACS using an anti- properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. (C) Graphical representation of properdin binding to thrombin-activated platelets, as determined in A and B, and expressed as means and standard deviations (SD) of triplicate geoMean fluorescence intensity (GMFI) values. An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.01(**) and p<0.001(***)). In (C), the differences between the ability of properdin to bind to activated platelets versus non activated platelets were all significant (p<0.05). Results are representative of 3 separate experiments shown as means and standard deviations of triplicate observations. (Figure 1 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

A B 35 NA 60 NA Act Thr Act Thr 30 25 40 20 15 20 10 Pbinding(GMFI) Pbinding(GMFI) 5 0 0 0 10 20 30 0 20 40 60 Properdin (µg/ml) Time (min)

Figure 17: Dose and time dependent binding of properdin to platelets. Non-activated (NA) or thrombin-activated (Act Thr) platelets (2x106

platelets/100 µl) were incubated in the presence of 0-30 μg/ml of P2-P4 pool (in

a ratio of 1:2:1) for 30 min in (A) or 25 μg/ml P2-P4 pool (in a ratio of 1:2:1) for 0 – 60 min in (B) at RT in Tyrode/PGE/Hep buffer. Platelets were washed and binding of properdin was assessed by FACS using an anti-properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. Results are shown as means and standard deviations of triplicate observations. (Panel (A) is from figure 1 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

5.1.1.3 Does properdin binding depend upon the agonist of platelet activation?

The extent to which stimulated platelets activate complement on or near their surface, when exposed to plasma or serum, is higher on platelets stimulated with strong agonists versus weak agonists (120,144). Therefore, we sought to determine whether platelets activated by different agonists would preserve the ability to bind properdin. Platelets were activated with thrombin, arachidonic acid

(strong agonists), or ADP (weak agonist), and compared in their ability to bind properdin. The data in figure 18 show that the binding of properdin to platelets 74

stimulated with arachidonic acid is ~4-fold higher than thrombin-stimulated platelets, while no binding of properdin was detected to ADP-activated platelets.

Thus, in figures 16 and 17, the use of thrombin-activated platelets allowed more stringent conditions for assessing significance of properdin binding, because these platelets bind notably less properdin than arachidonic acid-activated platelets. This data suggests that the binding of properdin to activated platelets depends on the mode of stimulation of the platelets.

600 no P with P **

400

200 ***

P binding (GMFI) ns

0 NA Act Act Act ADP Thr AA

Figure 18: The level of properdin binding to platelets depends on the agonist. Non-activated (NA), ADP-activated (Act ADP), thrombin-activated (Act Thr), arachidonic acid-activated (Act AA) platelets (2x106 platelets/100 µl) were

incubated in the presence or absence of 25 μg/ml of P2-P4 pool (in a ratio of 1:2:1) for 60 min at RT in Tyrode/PGE/Hep buffer. Platelets were washed and the binding of properdin was assessed by FACS using an anti-properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.01(**) and p<0.001(***)). Results are representative of 3 separate experiments shown as means and standard deviations of triplicate observations. (Figure 1 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

5.1.1.4 Is the level of properdin binding proportional to the level of P-selectin (CD62P) on the platelet surface?

It has been reported that alternative pathway complement activation on platelets requires CD62P (119). The authors indicate that C3b binds to CD62P, forms a C3 convertase, and activates the complement system leading to the deposition of C3b and C5b-9. They show that C3b bound to CD62P can lead to the formation of a C3 convertase in the presence of factor B. Since platelets activated by different agonists expose varying levels of CD62P on their surface (193), we determined if the level of properdin binding was proportional to the level of CD62P exposed on the platelet surface. The maximum CD62P levels on thrombin- and arachidonic acid- activated platelets were achieved with <1 U/ml or <1 mM, respectively (Fig. 19A-B, indicated by arrow). As expected, platelet activation, as measured by exposure of CD62P, varied between the different stimuli (thrombin > arachidonic acid; Fig. 19A-C). Interestingly, although the maximum level of CD62P expression is significantly higher (~1.5-2 fold) on platelets stimulated with thrombin (Fig. 19A,C) versus arachidonic acid (Fig. 19B-C), the binding of properdin to the arachidonic acid-activated platelets (Fig. 19E,F) is significantly higher (~3 fold) than to the thrombin-activated platelets (Fig. 19D,F). Thus, properdin binding is not directly proportional to the level of CD62P exposed on activated platelets.

A B C

1200 1200 800 *** 1000 1000 600 800 800 600 600 400 400 400 200 200 200 0 0 0 CD62P expression (GMFI) CD62P expression CD62P expression (GMFI) CD62P expression

0.0 0.5 1.0 0 2 4 6 8 10 (GMFI) CD62P Expression NA Act Act Thrombin (IU/ml) Arachidonic Acid (mM) Thr AA D BE F 250 *** 400 400 200 300 300 150 200 200 100 100 100 50 P binding (GMFI) P binding P binding (GMFI) P binding P binding (GMFI) P binding 0 0 0 0.0 0.5 1.0 0 2 4 6 8 10 NA Act Act Thrombin (IU/ml) Arachidonic Acid (mM) Thr AA

Figure 19: The binding of properdin to activated platelets is not proportional to CD62P levels on the platelet surface. Platelets were activated using thrombin (Thr) (0-1 IU/ml in (A) and (D); 1 IU/ml in (C) and (F)) or arachidonic acid (AA) (0-10 mM in (B) and (E); 1 mM in (C) and (F)). Non-activated (NA) or activated (Act) platelets (2x106 platelets/100 µl) were incubated in the presence of 25 μg/ml of P3 for 60 min at RT. (D)-(F), Binding of properdin was assessed by FACS using an anti-properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. (A)-(C), Platelet activation was assessed by using a PE/Cy5 labeled anti-CD62P monoclonal antibody. The arrows in (A) and (B) represent the dose of activator, used in (C) and (F), where maximal CD62P expression was achieved. An APC-labeled anti-CD42b monoclonal antibody was used to gate on platelets. The data are representative of two independent experiments shown as means and standard deviations. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.001(***)). (Figure 2 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

5.1.1.5 Does native properdin directly activate platelets?

We determined if native properdin can directly activate platelets, thus potentially contributing to the increased availability of activated platelets on which complement can activate. Incubating the resting platelets with properdin during the activation step does not result in an increase in expression of CD62P or gpIIbIIIa, or annexin V binding on the platelet surface (Fig. 20).

5.1.1.6 Does properdin binding to activated platelets require previous C3 fragment deposition on the platelets?

Complement C3 is present in platelets (119) and its spontaneously hydrolyzed form (C3(H2O)), as well as its C3b fragment, have been shown to bind to platelets (119,146). C3(H2O) and C3b also bind to properdin when the alternative pathway C3 and C5 convertases are formed (C3b,Bb,P or C3(H2O),Bb,P) (93,194). In order to determine if properdin was binding through C3 components on the platelets (which would indicate that properdin is binding indirectly to the platelet surface), we assessed whether the binding of properdin could be inhibited by polyclonal F(ab’)2 anti-C3 antibodies. F(ab’)2 antibodies were used to block C3 components and thus inhibit binding of properdin to C3 components. A simple diagram of the method used is shown in figure 21 A and B. Figure 21C shows that 10

µg/ml of the antibody completely inhibited the binding of properdin to C3b- opsonized sheep erythrocytes (EsC3b). These sheep erythrocytes have a very large amount of C3b on their surface (>125,500 C3b/cell) due to the fact that they are 78

A B C CD62P gpIIbIIIa Annexin V binding 600 25 250

20 200 400 15 150

10 100 200 5 50 CD62P Expression (GMFI) CD62P Expression 0 0 (GMFI) Binding V Annexin gpIIbIIIa Expression (GMFI) gpIIbIIIa Expression 0 No Thr AA P No Thr AA P No Thr AA P Agonist Agonist Agonist

Figure 20: Properdin does not activate platelets. Washed human platelets were incubated without agonist or with thrombin, arachidonic acid, or properdin (25 µg/ml) for 30 min at 37°C. The platelets were then washed and activation was assessed as a measure of (A) CD62P expression using a PE/Cy5- labeled anti-CD62P monoclonal antibody, (B) gpIIbIIIa expression using a FITC- labeled PAC-1 monoclonal antibody (IgMκ; BDBiosciences), or (C) Annexin V binding using FITC-labeled Annexin V (Molecular Probes). After washing, an APC-labeled CD42b monoclonal was used to gate on the platelets. Graphs are representative of two separate experiments. (Supplemental figure S2 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

subjected to multiple rounds of amplification of C3b deposition in vitro. On the contrary, not even 10-fold more (100 µg/ml) of the same antibody was able to inhibit the binding of P2-4 to thrombin- or arachidonic acid-activated platelets (Fig.

21D-E). This occurs even though the platelets have significantly less C3b molecules/cell than the EsC3b since they have been subjected to only one round of complement amplification (as a result of their incubation with serum). In addition, neither C3 components nor properdin were detected on the surface of washed thrombin- or arachidonic acid-activated platelets by flow cytometry (Fig. 22). These data indicate that the physiological forms of properdin bind selectively to activated platelets, independently from C3. 79

A B C3b coated sheep erythrocytes (ESC3b) Activated platelets

F(ab’)2 F(ab’)2 anti-C3 anti-C3

Incubate with properdin Incubate with properdin

Determine properdin Determine properdin binding by flow cytometry binding by flow cytometry

C D E 100 200 200 No No Isotype ESC3b Act Thr Act AA 80 P P 150 150 P + anti C3 60 P +/- anti C3 100 P +/- anti C3 100 40 P 50 50 20 Counts Counts Counts 0 0 0 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 Properdin binding

Figure 21: The binding of properdin to platelets is not mediated by C3. (A) and (B) General schematic of experimental steps. (C) C3b coated sheep erythrocytes (ESC3b) (5x106 ESC3bs/100 µl) in GVB= were incubated with or without 10 µg/ml anti-C3 polyclonal antibody for 15 min at 4°C. Without washing,

P2-P4 (10 µg/ml) were added to the cells and incubated further for 1 h at 4°C. Cells were washed and properdin binding was detected using an anti-properdin monoclonal antibody or an IgG1 isotype control, followed by an Alexa Fluor 488- conjugated anti-mouse IgG antibody. Thrombin activated (Thr-Act) (D) or arachidonic acid activated (AA-Act) (E) platelets (2x106 platelets/100 µl) were incubated with or without 100 µg/ml anti-C3 polyclonal antibody for 30 min at RT.

Without washing, P2-P4 (10 µg/ml) were added to the cells and incubated further for 30 min at RT. Platelets incubated without anti-C3 and without P2-P4 were used as a negative control. Platelets were washed and the binding of properdin was assessed using an anti-properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. (Figure 3 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

A B 50 Isotype 50 Isotype Anti C3b 40 40 Anti C3b

30 30

20 20 C3b C3b (GMFI) C3b (GMFI) C3b 10 10

0 0 NA Act Thr NA Act AA C D 50 50 Isotype Isotype 40 Anti-P 40 Anti-P

30 30

20 20

10 10 Properdin (GMFI) Properdin Properdin (GMFI) Properdin 0 0 NA Act Thr NA Act AA Figure 22: C3 components and properdin are not present on the surface of washed activated platelets. Platelets (1x108 platelets/ml) were activated using thrombin (1UI/ml) (A and C) or arachidonic acid (5 mM) (B and D). Platelets were washed and presence of C3 components on the surface of platelets was determined by FACS using an anti-C3/C3b monoclonal (A and B), anti-properdin monoclonal (C and D), or an IgG1 isotype control antibody followed by Alexa Fluor 488-conjugated anti-mouse IgG. After washing an APC-labeled CD42b monoclonal antibody was used to gate on the platelets. Results are shown as means and standard deviations (SD) of triplicate geoMean fluorescence intensity (GMFI) values. Differences between the isotype control and specific antibody groups are non-significant (p>0.05). (Supplemental figure S3 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

5.1.1.7 Does properdin bind to proteins and/or glycosaminoglycans on the surface of

activated platelets?

Properdin is known to interact with various ligands on cell surfaces such as

glycosaminoglycans (heparin (103), heparin sulfate (97,104), dextran sulfate (105),

fucoidan (105), chondroitin sulfate (97)), as well as DNA (85), and bacterial LPS

(106). Since platelets are known to release chondroitin sulfate A upon activation, we

wanted to determine if chondroitin sulfate A acts as a ligand for properdin binding

to platelets. We found that at low doses of chondroitin sulfate A, properdin binding

was increased approximately four-fold (Fig. 23), while higher doses had no effect on

binding of properdin to the activated platelets. We also determined that the removal

A B Figure 23: Properdin binding is increased in the presence of 250 NA 80 400 chondroitin sulfate-A. Chondroitin NA +P 200 200 Act sulfate-A (CS-A) was incubated with P3

60 MFI) Geo 0 (

CD62P level CD62P Act + P (10 μg/ml) for 30 min at RT. Non- 0 50 100 150 activated (NA) or thrombin-activated (Geo MFI) (Geo 40 Proteinase K (µg/ml) 100 (Act) platelets (2x106 platelets/100 µl) were then added to the mixture and 20 50 incubated for 60 min at RT. Platelets were Binding P Binding Binding MFI)(Geo P

P 0 0 washed and binding of properdin was 0 50 100 0 1 2 3 4 5 assessed using an anti-properdin Proteinase K (ug/ml) CS-A (mg/ml) monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population.

of surface proteins from the platelet surface, by treating the platelets with

increasing doses of proteinase K, led to a dose-dependent decrease in properdin

binding to platelets (Fig. 24). The inset showing decrease in surface CD62P 82

expression represents a control for removal of surface proteins with increasing doses of proteinase K. These results indicate that both proteins, as well as glycosaminoglycans (or proteoglycans) play in role in the interaction of properdin with platelets.

A B 250 NA 80 400 NA +P 200 200 Act

60 MFI) Geo 0 (

CD62P level CD62P Act + P 0 50 100 150

(Geo MFI) (Geo 40 Proteinase K (µg/ml) 100

20 50 Binding P Binding Binding MFI)(Geo P

P 0 0 0 50 100 0 1 2 3 4 5 Proteinase K (ug/ml) CS-A (mg/ml)

Figure 24: Properdin binding is reduced upon removal of surface proteins. Non-activated (NA) or thrombin-activated (Act) platelets (2x106 platelets/100 µl) were incubated with proteinase K (0 – 100 µg/ml) at RT for 30 min. Platelets were

washed and incubated with P2-P4 for 60 min at RT. Binding of properdin was assessed using an anti-properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. Inset: As a control for proteinase K treatment, the level of CD62P expression was assessed by FACS using a PE/Cy5- labelled anti-CD62P antibody. FSC vs SSC profile was used to gate on the platelet.

5.1.2 Does properdin bound to platelets initiate complement activation?

5.1.2.1 Does properdin bound to normal human platelets lead to C3b deposition on the platelets?

Host cell surfaces are protected by membrane-bound and fluid-phase complement regulatory proteins. Platelets have surface-bound CD55, CD46, and

CD59 (Fig. 25) (151,195) and are able to bind factor H (Fig. 32; addressed in section

5.1.3) (196). Thus, the binding of properdin to the surface of platelets will not necessarily lead to deposition of C3b on normal platelets. In order to determine functional consequences of the interaction between properdin and stimulated platelets, we investigated whether physiological

A B NA Act Anti-CD46 Anti-CD46 Anti-CD55 Anti-CD55 Anti-CD59 Anti-CD59 Isotype Isotype Counts Counts

Figure 25: Presence of CD46, 55, 59 on platelets. Platelets (1x108 platelets/ml) were activated using thrombin (1UI/ml). Non-activated (NA) (A) or thrombin- activated (Act) (B) platelets (2x106 platelets/100 µl) were washed and presence of CD46, CD55, CD59 was determined by FACS using a FITC labeled anti-CD46, anti- CD55 or anti-CD59 or an IgG1 isotype control. After washing, an APC-labeled CD42b monoclonal antibody was used to gate on the platelet population.

forms of properdin that are bound Activated platelets

to activated platelets have the ability to promote complement P activation. As shown in the schematic in figure 26, washed, non-activated and thrombin- or Incubate with properdin- arachidonic acid-activated depleted serum at 37°C platelets were incubated in the presence or absence of purified Determine C3b deposition properdin. Platelets were then by flow cytometry

washed and exposed to properdin- Figure 26: Schematic for properdin- mediated C3b deposition in depleted serum in order to study properdin-depleted serum. alternative pathway activation mediated only by properdin bound to the activated platelets. Figure 27A,D shows that rapid C3b deposition occurs on activated platelets pre-incubated with properdin in as early as 5 min compared to activated platelets alone, non-activated platelets, or non-activated platelets pre-incubated with properdin. Maximum C3b deposition on properdin-bound thrombin-activated platelets was observed after 20 min (Fig. 27A), whereas for properdin-bound arachidonic acid-activated platelets maximum C3b deposition was observed in as early as 5 min (Fig. 27D). Thrombin- activated platelets with properdin on their surface induced ~2.2-fold increase in

C3b deposition versus thrombin-activated platelets alone, at 30 minutes (Fig. 27B).

Only the group of thrombin-activated platelets with properdin on their surface 85

induced significant C3b deposition when exposed to P-depleted serum for 30 minutes versus zero minutes (Fig. 27B). Moreover, we observed that C3b deposition was ~10-fold higher on arachidonic acid-stimulated platelets with properdin on their surface (Fig. 27E) when compared to the thrombin-activated platelets with surface-bound properdin (Fig. 27B; right bars). Arachidonic acid-activated platelets without properdin on their surface induced C3b deposition, although ~4-fold less than platelets with properdin on their surface (Fig. 27E). Both thrombin and arachidonic acid-activated platelets induced iC3b deposition on their surface only when properdin was present on the surface (Fig. 27C and F).

A B C ** 150 20 60 0 min 0 min 30 min ** 30 min 15 40 100 10 n.s. 20 50 5 iC3b deposition (GMFI) depositioniC3b C3b deposition (GMFI) deposition C3b C3b deposition (GMFI) deposition C3b 0 0 0 Thr-Act Thr-Act Thr-Act Thr-Act Thr-Act Thr-Act 0 10 20 30 40 +P +P +P +P Time (min) Compstatin EDTA NA NA+P Act Act+P D E F 1000 600 3000 Act + P 0 min 0 min 30 min *** 30 min 800 400 2000 600

400 Act 200 1000 200 NA + P C3b deposition (GMFI) deposition C3b C3b deposition (GMFI) iC3b deposition (GMFI) deposition iC3b NA 0 0 0 0 20 40 60 AA-Act AA-Act AA-Act AA-Act AA-Act AA-Act +P +P Time (mins) +P +P EDTA Compstatin NA NA+P Act Act+P

Figure 27: Properdin-mediated C3b/iC3b deposition in properdin- depleted serum. Non-activated (NA), thrombin activated (Thr-Act) or arachidonic acid-activated (AA-Act) platelets (2x106 platelets/100 µl) were incubated in the presence or absence of P3 or P2-4 (25 µg/ml) for 60 min at RT. Platelets were then washed and incubated with 60% properdin-depleted serum in the presence of Mg-EGTA at 37°C for various time points in (A) and (C) or 30 min in (B), (C), (E) and (F). Deposition of C3b on the surface of platelets was assessed using a PE-labeled anti-C3 monoclonal antibody in (A), unlabeled anti-C3/C3b monoclonal antibody followed by AF488- conjugated anti-mouse IgG in (B), (D) and (E). Deposition of iC3b on the surface of platelets was assessed using an unlabeled anti-iC3b monoclonal antibody followed by AF488-conjugated anti-mouse IgG in (C) and (F). An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. As controls, platelets (in the presence or absence of properdin) were incubated with 60% properdin-depleted serum in EDTA (10mM). In (A), the EDTA controls gave an average GMFI of 5.3±0.69 at 40 min and in (C), the EDTA controls gave an average GMFI of 30.02±25.39 at 60 min (not shown in graph). The graphs show one representative experiment of three independent experiments. Each bar in (B) and (D) represents the mean and standard deviation of triplicate observations. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.01(**); p<0.001 (***)). (Panels A, B, D, E from Figure 4 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

Similarly, we also measured Activated platelets

C3b deposition on thrombin-activated platelets P in NHS (Fig. 28) because in a normal host, the properdin- bound platelets will come in contact with the NHS that Incubate with normal human serum at 37°C contains properdin. The data showed similar kinetics Determine C3b deposition of C3b deposition to that by flow cytometry observed with properdin- Figure 28: Schematic for properdin- depleted serum, with rapid mediated C3b deposition in normal human serum. C3b deposition on platelets incubated with properdin as early as 5 min and maximum C3b deposition reached at

15-20 min (Fig. 29A). Minor, but significant C3b deposition was seen on activated platelets without properdin in NHS. C3b deposition was significantly enhanced when properdin was present on the surface of the activated platelets before exposure to the NHS (Fig. 29B). These data indicate that alternative pathway activation on stimulated platelets is greatly enhanced when properdin is bound to the platelet surface, even when complement regulatory functions on the platelets are normal.

A B *** 30 *** 100 0 min 30 min 75 20

50 *** 10 25 C3b deposition (GMFI) deposition C3b C3b deposition (GMFI) deposition C3b 0 0 0 20 40 60 Thr-Act Thr-Act Thr-Act Time (min) +P +P NA NA+P Act Act+P EDTA

Figure 29: Properdin-mediated C3b deposition in normal human serum. Non-activated (NA) or thrombin activated (Thr-Act) platelets (2x106

platelets/100 µl) were incubated in the presence or absence of P3 or P2-4 (25 µg/ml) for 60 min at RT. Platelets were then washed and incubated with 60% NHS in the presence of Mg-EGTA at 37°C for various time points in (A) or 30 min in (B). Deposition of C3b on the surface of platelets was assessed using a PE-labeled anti-C3 monoclonal antibody in (A) or unlabeled anti-C3/C3b monoclonal antibody followed by AF488-conjugated anti-mouse IgG in (B). An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. As controls, platelets (in the presence or absence of properdin) were incubated with 60% properdin-depleted serum in EDTA (10mM). In (A), the EDTA controls gave an average GMFI of 6.9±0.24 at 60 min (not shown in graph). The graphs show one representative experiment of three independent experiments. Each bar in (B) represents the mean and standard deviation of triplicate observations. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.001 (***)).

5.1.2.2 Does properdin bound to platelets lead to C9 deposition on the platelets?

Deposition of C3b on host cell surfaces may not necessarily cause complement to progress to its terminal stage (formation of membrane attack complex; C5b-9; MAC) due to complement regulation, as explained in the preceding 89

section. Thus, we Activated platelets

investigated whether properdin-mediated P complement activation on the surface of platelets leads to formation of the Incubate with properdin- MAC, in spite of depleted serum or normal human serum at 37°C complement regulation, using the method shown in Determine C5b-9 deposition figure 30. Washed non- by flow cytometry activated, thrombin- Figure 30: Schematic for properdin- mediated C5b-9 deposition in properdin- activated or arachidonic depleted serum and normal human serum. acid-activated platelets were incubated in the presence or absence of properdin and exposed to properdin- depleted serum or NHS for 60 min. C5b-9 deposition on the platelets was measured using an antibody specific for a neo epitope on C9 that is exposed only when C9 is incorporated into the MAC. Only the thrombin-activated platelets that had been preincubated with properdin showed an increase in C5b-9 deposition as compared to activated platelets alone, non-activated platelets or non-activated platelets preincubated with properdin (Fig. 31 A, C). In the case of arachidonic acid-activated platelets, MAC deposition was observed on activated platelets without properdin on their surface, and a significant increase was observed on platelets with properdin on their surface (Fig. 31 B, D). 90

A B ** 80 300 0min 0min ** 60min 60min 60 60min EDTA 60min EDTA 200

100 20 C5b-9 Deposition (GMFI) Deposition C5b-9 C5b-9 Deposition (GMFI) Deposition C5b-9 0 0 NA NA+P Act-ThrAct-Thr+P NA NA+P Act-AA Act-AA+P C D *** 30 ** 500 0min 0min 60min 60min 400 60min EDTA 60min EDTA 20 300

200 10 100 C5b-9 Deposition (GMFI) Deposition C5b-9 C5b-9 Deposition (GMFI) Deposition C5b-9 0 0 NA NA+P Act-Thr Act-Thr+P NA NA+P Act-AA Act-AA+P

Fig 31: Properdin promotes formation of C5b-9 complexes on the surface of activated platelets. Non-activated (NA), thrombin-activated (Act- Thr) (A) and (C) or arachidonic acid-activated (Act-AA) (B) and (D) platelets

(2x106 platelets/100 µl) were incubated in the presence or absence of P2-4 or

P3 (25 µg/ml) for 60 min at RT. Platelets were then washed and incubated with 60% properdin-depleted serum (A) and (B) or 60% normal human serum (C) and (D) in the presence of Mg-EGTA at 37°C for 0 or 60 min (or with EDTA at 37°C for 60 minutes only as a control). Deposition of C5b-9 complexes was assessed using an anti-C5b-9 neo-epitope monoclonal antibody followed by AF488 conjugated anti-mouse IgG. An APC-labeled anti- CD42b monoclonal antibody was used to gate on platelets. The results are representative of two separate experiments shown as means and SDs of triplicate observations. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.005 (**)). (Panels A, B from figure 8 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

5.1.3 Is factor H regulation needed for controlling properdin-mediated complement activation on platelets?

The regulatory mechanisms that participate in complement regulation on cell surfaces (mentioned in the preceding sections) often have redundant functions.

Thus, the loss of one regulator does not imply that regulation of complement on that cell will be measurably affected.

5.1.3.1 Does factor H control properdin-mediated C3b deposition on activated platelets?

We confirmed that factor H can bind to activated platelets (Fig. 32)

(196,197). Mutations in the C-terminus of factor H impair the efficient binding of factor H to cell surfaces and have been associated with deposition of complement on the platelet surface in patients with aHUS, which may result in thrombosis and thrombocytopenia (153). Thus, we examined the effect of inhibiting factor H regulation on the platelet surface during properdin-mediated complement activation on platelets. In order to impair factor H binding to platelets, we used a competitive inhibitor of factor H cell surface complement regulation known as rH19-20 (54,55,188-190). This inhibitor is a recombinant protein consisting of domains 19-20 of factor H that competes with full-length factor H for binding to cell surface C3b and polyanions, inhibiting the ability of factor H to protect cell surfaces, without affecting fluid phase complement regulation (54).

A NA B Thr C AA Counts Counts Counts

fH binding

Isotype control Anti fH#2 (Quidel)

Figure 32: Factor H binds to washed, activated platelets. Non-Activated (NA) (A), thrombin-activated (Thr) (B) or arachidonic acid-activated (AA) (C) platelets (2x106 platelets/100 µl) were incubated in the presence of factor H (100 µg/ml) for 30 min at 37°C. Platelets were then washed and binding of factor H was detected using an unlabeled anti-factor H monoclonal antybody (anti fH#2;Quidel) or an IgG1 isotype control, followed by AF488-conjugated anti-mouse IgG. An APC- labeled anti-CD42b monoclonal antibody was used to gate on platelets.

The data in figure 33 indicate that inhibiting factor H complement regulation using rH19-20 on thrombin- or arachidonic acid-activated platelets with properdin increases C3b deposition by ~4-fold (thrombin-activated; Fig. 33A) and ~8-fold

(arachidonic acid-activated; Fig. 33B) when compared to activated platelets with properdin, but without rH19-20. Inhibiting factor H complement regulation in the absence of properdin induces significant C3b deposition as compared to the EDTA control. This increase is similar to the C3b deposition that occurs when the platelets have properdin on their surface and factor H regulation has not been inhibited.

A B Thr-Act AA-Act 300 *** 2500 *** *** 2000 200 *** *** 1500 *** *** 1000 100 *** *** 500 *** C3b deposition (GMFI) deposition C3b C3b deposition (GMFI) deposition C3b 0 0 EDTA + - - - - EDTA + - - - - P + - - + + P + - - + + rH19-20 - - + - + rH19-20 - - + - +

Figure 33: Properdin-mediated complement activation is exacerbated when cell surface protection by factor H is inhibited. Thrombin-activated (Thr-Act; (A)) or arachidonic acid-activated (AA-Act; (B)) (2x106 platelets/100

µl) were incubated in the presence or absence of P2-P4 (25 µg/ml) for 60 min at RT. Platelets were then washed and incubated with 60% properdin-depleted serum in presence of Mg-EGTA with or without rH19-20 (25 µM) at 37°C for 30 min. Deposition of C3b on the surface of platelets was assessed using an unlabeled anti-C3/C3b monoclonal antibody followed by AF488-conjugated anti- mouse IgG. An APC-labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. As controls, platelets (in the presence of properdin) were incubated with 60% properdin depleted serum in EDTA (10mM). The graphs show one representative experiment of two independent experiments. Each bar represents the mean and standard deviation of triplicate observations. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.001 (***)).. (Figure 5 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

5.1.3.2 Do aHUS-related mutations in factor H affect its ability to control properdin mediated C3b deposition on activated platelets?

Patients with aHUS are known to be associated with an increased risk of thrombosis ((198); Reviewed in (199)). aHUS-related mutations of factor H have

been associated with deposition of complement on the platelet surface in patients with aHUS (153). This study shows that platelets taken from patients with aHUS mutations in SCR 15, as well as in the C-terminal domain-SCR 20, have greater C3 and C9 deposits as compared to platelets from normal donors. The study also shows that incubation of normal platelets with aHUS patient serum leads to C3 and C9 deposition on platelets as well as induces platelet aggregation.

We examined whether aHUS associated mutants of factor H are impaired in their ability to control properdin-mediated C3b deposition on activated platelets.

We used three aHUS-related mutants of factor H: D1119G, L1189F and R1215G (39).

These mutations are located in either domain 19 (D1119G) or domain 20 (L1189F and R1215G) of factor H. D1119G is greatly impaired in its ability to bind C3b whereas R1215G is partially impaired as compared to the wild type (WT) protein.

On the other hand, L1189F has slightly enhanced C3b binding capacity as compared to WT rH19-20. The heparin binding capacity of D1119G and L1189F is slightly enhanced whereas that of R1215G is slightly reduced as compared to WT. Ferreira et al (39) has shown that the D1119G and the R1215G mutants of rH19-20 are almost completely impaired in their ability to bind to C3b-coated host-like erythrocyte surfaces, whereas L1189F is only partially impaired. As mentioned in the previous section, WT rH19-20 competes with full length factor H for binding to the platelet surface and thus inhibits the ability of factor H to control C3b deposition on the platelet surface. We found that all the mutants tested were not able to compete with factor H and inhibit its ability to control C3b deposition on the platelet surface. This data suggests that factor H with any of the three aHUS-related 95

mutations tested is significantly impaired in its ability to control C3b deposition on activated platelets as compared to WT factor H (Fig. 34). Thus, patients with aHUS may be more susceptible to properdin-mediated complement activation.

*** ** 500 no P *** with P *** 400

300

200 *** 100 C3b deposition (GMFI) deposition C3b

Figure 34: aHUS mutants are impaired in their ability to control properdin- mediated C3b deposition. Non-Activated (NA), thrombin-activated (Thr) 2x106

platelets/100 µl) were incubated in the presence or absence of P2-P4 (25 µg/ml) for 60 min at RT. Platelets were then washed and incubated with 60% properdin- depleted serum in presence of Mg-EGTA with or without WT rH19-20 or mutants D1119G, L1189F or R1215G (25 µM each) at 37°C for 30 min. Deposition of C3b on the surface of platelets was assessed using an unlabeled anti-C3/C3b monoclonal antibody followed by AF488-conjugated anti-mouse IgG. An APC- labeled anti-CD42b monoclonal antibody was used to gate on the platelet population. The results are representative of two separate experiments. Statistical significance was assessed by determining the p-value using an unpaired t-test (p< 0.01 (**); p<0.001 (***)).

5.1.4 What is the mechanism by which the platelet-bound properdin activates complement?

We next sought to define the molecular mechanisms involved in properdin- mediated complement activation on stimulated platelets. Properdin can bind to C3b

(22) and C3(H2O) (93,194), which are structurally similar. We therefore analyzed the ability of the properdin that is bound to the activated platelet to recruit C3b and

C3(H2O) to the platelet surface as a potential first step in convertase formation.

5.1.4.1 Separation of C3(H2O) from C3.

The first step to determine C3 100 whether platelet bound properdin can 1.5

80 BufferB % recruit C3b and C3(H2O) was to separate C3(H2O) from C3. Briefly, C3 1.0 60 was incubated at 37°C for 2 hours to 40 0.5 C3(H2O) convert intermediate/inactive forms Absorbance (280 nm) 20 of C3 to C3(H2O). The sample was 0 0 0 5 10 15 20 25 30 35 diluted with Mono S buffer A and Elution vol. (ml) loaded onto a 1 ml Mono S column. Figure 35: Separation of C3(H2O) from C3. C3 was separated from The column was washed with buffer A C3(H2O) by cation exchange chromatography as described in the and eluted using a 20 ml salt gradient Methods section. (**)). (Figure 6 from Saggu et al, JImmunol, 2013, (0-100%) of buffer B at a flow rate of 190(12):6457-67; Appendix 1)

1 ml/min. Purified C3 and C3(H2O) were dialyzed against PBS, stored at 4°C and used within two weeks of separation. 97

Figure 35 shows the separation of C3(H2O) from C3 by ion-exchange chromatography.

5.1.4.2 Does properdin bound to the surface of the platelet recruit C3 components to the platelet surface?

We analyzed the ability of the properdin that is bound to the activated platelet to recruit C3b and C3(H2O) to the platelet surface, as a potential first step in convertase formation (Figs. 36 and 37). It has been proposed that C3b binds directly to activated platelets via CD62P (119). Nevertheless, Hamad et al., by using antibodies against a neo epitope found on C3(H2O), detected binding of C3(H2O), but not of C3b or non-activated C3, to activated platelets (146). The data in figure 36A, using thrombin-activated platelets, confirm that C3 and C3b do not significantly bind directly to resting or thrombin-activated platelets, and that C3 and C3b do not bind even when properdin is present on the platelet surface. On the other hand,

C3(H2O) binds significantly to activated platelets without (p<0.05) and with

(p<0.001) properdin on their surface (Fig. 36A). Moreover, properdin that is bound directly to activated platelets recruits ~1.5-fold more C3(H2O) to the platelet surface when compared to activated platelets alone (Fig. 36B). On the contrary, the C3(H2O) that is bound directly to activated platelets is not able to recruit more properdin than the activated platelets without C3(H2O) (Fig. 36C).

A * B DC 80 6 C3 *** 45 Act Act C3b *** Act + C3(H O) Act + P3 2 5 C3(H2O) *** * ** 60 4 30 * 3 40

2 15 O) Binding (GMFI) Binding O)

2 20

1 (GMFI) P Binding C3(H Binding of C3of components Binding 0 0 0 0 25 50 75 100 0 5 10 15 20 25

(fold increase compared to no C3) no to compared increase (fold NANA NA-PNA ThrAct-Act ThrAct-P-Act +P +P C3(H O) (μg/ml) Properdin (μg/ml) 2

Figure 36: Analysis of binding of C3, C3b and C3(H2O) to thrombin-activated platelets with or without properdin on their surface. (A) Non activated (NA) or thrombin-activated (Thr-Act) platelets (2x106 platelets/100 µl) were incubated in the

presence or absence of P3 (25 µg/ml) for 30 min at RT, washed and then incubated

without or with C3, C3(H2O) or C3b (100 µg/ml) at RT for 1 h. Binding of C3 forms was assessed by FACS using an anti-C3/C3b monoclonal antibody followed by AF488- conjugated anti-mouse IgG. APC-labeled anti CD42b monoclonal antibody was used to

gate on platelets. Binding of C3, C3b, and C3(H2O) to each group of platelets (NA, NA+P, Thr-Act, Thr-Act+P) was normalized to the GMFI of each respective group incubated without any C3 components. The results represent the mean and SDs of seven independent experiments. Statistical analysis was carried out comparing all columns of NA conditions to the NA + C3 group, and all columns of Act conditions to the Act + C3 group by one way analysis of variance-Dunnett’s Multiple comparison test. Statistical significance is indicated as follows: p<0.05(*), p<0.001(***). An unpaired Student’s t-test was used to evaluate the statistical significance between the binding of

C3(H2O) to the Act versus the Act+P group (p<0.05(*)). (B) Properdin recruits C3(H2O) to the surface of activated platelets. Thrombin-activated (Act) platelets (2x106

platelets/100 µl) were incubated in the presence or absence of P3 (25 µg/ml) for 30

min at RT. Platelets were then washed and incubated with increasing doses of C3(H2O)

(0-100 µg/ml) at RT for 1 h. Binding of C3(H2O) was assessed by FACS as described in

(A). (C) C3(H2O) does not recruit properdin to the surface of activated platelets. Thrombin activated (Act) platelets (2x106 platelets/100 µl) were incubated in the

presence or absence of C3(H2O) (100 µg/ml) at RT for 60 min. Platelets were then washed and incubated with increasing doses of properdin (0-25 µg/ml) at RT for 60 min. Properdin binding was assessed using an anti-properdin monoclonal antibody followed by AF488-conjugated anti-mouse IgG. An APC-labeled anti CD42b antibody was used to gate on the platelet population. Each graph (B) and (C) is representative of two independent experiments shown as means and SDs. Statistical analysis (p<0.05(*), p<0.01(**), p<0.001(***)) was performed by using two way analysis of variance-Bonferroni’s posttest. All data without a p-value were non-significant. (**)). (Figure 6 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

When analyzing arachidonic acid-activated platelets (Fig. 37), both C3(H2O) and C3b bind significantly to activated platelets without and with properdin on their surface (Fig. 37). Properdin bound to activated platelets recruits ~3-fold more C3b

(p<0.01) and ~2-fold more C3(H2O) (p<0.05) to the platelet surface when compared to activated platelets alone (Fig. 37).

* 1 2 0 C3 C3b

1 0 0 C3(H2O) 8 0 ** 6 0

4 0

2 0

Binding of C3 components NA NA AA AA +P Act Act+P (fold increase compared to no C3) (fold compared increase

Figure 37: Analysis of binding of C3, C3b and C3(H2O) to arachidonic acid- activated platelets with or without properdin on their surface. Non activated (NA) or arachidonic acid-activated (AA-Act) platelets (2x106 platelets/100 µl)

were incubated in the presence or absence of P2-4 (25 µg/ml) for 30 min at RT,

washed and then incubated with C3, C3(H2O) or C3b (50 µg/ml) at RT for 1 h. Binding of C3 forms was assessed by FACS using an anti-C3/C3b monoclonal antibody followed by AF488-conjugated anti-mouse IgG. APC-labeled anti CD42b monoclonal antibody was used to gate on platelets. Binding of C3, C3b, and

C3(H2O) to platelets was normalized to the GMFI observed for each group incubated without any C3 components. The results show one representative experiment of two independent experiments. An unpaired student’s t test was

used to evaluate the statistical significance between the binding of C3(H2O) to the Act versus the Act-P group (p<0.05(*)).(Figure 7 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

100

5.1.4.3 Do the recruited C3 components on the platelet surface lead to convertase formation?

We next assessed the ability of platelet-bound properdin, C3b, and C3(H2O) to generate C3 convertases. In order to do this, thrombin or arachidonic-activated platelets were incubated with either (a) buffer, (b) properdin, (c) C3(H2O) or C3b, or

(d) properdin followed by washing and then C3(H2O) or C3b. A group with C3(H2O) first, followed by properdin, was not included for the thrombin activated platelets, as C3(H2O) does not recruit properdin to the platelet surface (Fig. 36C). After washing the cells, factors B and D were added to the groups of platelets mentioned above and the formation of Bb, due to proteolytic cleavage by factor D, was assessed by flow cytometry using an anti-Bb neoepitope antibody (Fig. 38A). Although thrombin-activated platelets bind C3(H2O) directly (Fig. 36A), formation of Bb was detected only on the surface of thrombin-activated platelets where C3(H2O) was additionally recruited to the surface by properdin (Fig. 38B).

For arachidonic acid-activated platelets, formation of Bb was detected on the surface of platelets where C3(H2O) (Fig. 38C, solid line) or C3b (Fig. 38D, solid line) was recruited to the surface by properdin. When platelets first received C3(H2O) or

C3b, convertase formation occurred only after properdin had been recruited (Fig.

38C-D, dotted lines). Altogether, these data indicate that on thrombin-activated platelets, only the recruitment of C3(H2O) by properdin leads to de novo formation of novel cell-bound [C3(H2O),Bb] convertases, while in the case of arachidonic acid- activated platelets, recruitment of both C3(H2O) and C3b by properdin leads to convertase formation. Moreover, C3(H2O) that is bound directly to arachidonic acid- 101

activated platelets can form, albeit to a lesser extent, novel convertases that require the recruitment of properdin for convertase stabilization, facilitating convertase detection (Fig. 38C, dotted line).

102

A Act platelet Act platelet Act platelet Act platelet C3b C3(H O) C3b C3(H2O) 2 P P

C3b C3(H O) P 2

Incubate with fB and fD

Determine Bb formation by flow cytometry B C D

150 Thr-Act alone AA-Act alone AA-Act alone +P +P +P +C3b 150 +C3(H2O) 150 +C3(H2O) +C3b P +C3(H2O) P 100 100 100 +PC3(H2O) +PC3b +PC3(H2O)

50 50 50

0 Counts Counts 0 Counts 0 100 101 102 103 104 100 101 102 103 104 Bb formation Bb formation Bb formation

Figure 38: Ability of thrombin and arachidonic acid-activated platelets to form a functional convertase: (A) Schematic representation of general experimental setup. (B) Thrombin-activated (Thr-Act) platelets were incubated in presence of P3 (25 μg/ml) or C3(H2O) (100 μg/ml), or sequentially with both

(P3C3(H2O)) or with neither for 60 min each at RT in Tyrode’s buffer. Platelets were then washed and incubated in presence of fB (80 μg/ml) and fD (2 μg/ml) for 30 min at RT in Tyrode’s buffer. (C) Arachidonic acid-activated platelets were incubated in the presence of P2-P4 (25 μg/ml), C3(H2O) (100 μg/ml) or sequentially with both (PC3(H2O) or C3(H2O)P) or neither for 60 min each at RT in Tyrode’s buffer. (D) Arachidonic acid-activated platelets were incubated in the presence of P2-P4 (25 μg/ml), C3b (100 μg/ml) or sequentially with both (PC3b or C3bP) or neither for 60 min each at RT in Tyrode’s buffer. Platelets were then washed and incubated in the presence of fB (80 μg/ml) and fD (2 μg/ml) for 30 min at RT. Bb formation on activated platelets was assessed by using an anti-Bb neo-epitope monoclonal antibody, followed by an Alexa Fluor 488

F(ab’)2 goat conjugated anti-mouse IgG antibody. APC-labeled anti-CD42b monoclonal antibody was used to gate on platelets. One representative experiment of two separately performed experiments is shown. (Figure 6and 7 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

103

5.1.5 Does native properdin released by neutrophils bind to activated platelets?

It has been previously shown that serum inhibits binding of properdin to surfaces (94,97). In agreement with this, properdin binding to platelets was inhibited in a dose-dependent manner by NHS (Fig. 39). This suggests that properdin binding to activated platelets may occur only when properdin is readily available to activated platelets in the microenvironment and that properdin- mediated complement activation is tightly controlled.

A B 100 100 NA NA 80 NA + P 80 NA + P Thr-Act AA-Act 60 Thr-Act+P 60 AA-Act+P

40 40

20 20 % Properdin binding % Properdin % Properdin binding % Properdin 0 0 0 20 40 60 0 20 40 60 NHS (%) NHS (%)

Figure 39: Normal human serum inhibits binding of properdin to activated platelets. Non-Activated (NA), thrombin-activated (Thr-Act) or arachidonic acid- activated (AA-Act) platelets (2x106 platelets/100 µl) were incubated with or

without P3 (25 µg/ml) in Tyrode/PGE/Hep, in the presence of NHS (0-60%; in 10 mM EDTA to avoid complement activation) for 30 min at RT. Platelets were washed with Tyrode/PGE/Hep and binding of properdin was assessed by flow cytometry using an anti-properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. After washing, platelets were stained with an APC-mouse anti-human CD42b monoclonal antibody to gate on the platelet population. One representative experiment of two separately performed experiments is shown. (Supplemental Figure S4 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

104

Neutrophils release properdin upon activation by various inflammatory stimuli such as PMA, fMLP, C5a and TNF-α (77), which may increase the local concentration of properdin in a pro-inflammatory microenvironment. In order to determine whether properdin derived from the activated neutrophil supernatants binds to platelets, supernatants from PMA-activated neutrophils were incubated with non-activated and thrombin-activated platelets (that bind significantly less pure properdin than arachidonic-acid-activated platelets). The activated platelets themselves do not have properdin on their surface (Fig. 22). As shown in figure 40A-

B, the properdin in the neutrophil supernatant bound only to activated platelets. C3 from activated neutrophil supernatants did not bind to the platelets (Fig. 40C-D), indicating that the binding of neutrophil-derived properdin to the platelets occurred independently from C3.

105

A B 200 200 NA Act 150 150

100 100

50 50

Counts 0 Counts 0 100 101 102 103 104 100 101 102 103 104 P binding C D 250 250 NA Act 200 200

150 150

100 100

50 50

Counts 0 Counts 0 100 101 102 103 104 100 101 102 103 104 C3 components binding

Figure 40: Properdin released by activated neutrophils binds to activated platelets, but C3 released by activated neutrophils does not. Non-activated (NA) (in (A) and (C)) or thrombin-activated (Act) (in (B) and (D)) platelets were incubated with supernatant from PMA-activated neutrophils for 60 min at RT in Tyrode/PGE/Hep with 10 mM EDTA. Platelets were washed and the binding of properdin (in (A) and (B)) and C3 components (in (C) and (D)) was assessed by FACS using anti- properdin antibody (dark line) (in (A) and (B)), or anti-C3 antibody (dark line) (in (C) and (D)) followed by an Alexa Fluor 488 conjugated anti- mouse IgG antibody. IgG1 monoclonal antibody was used as isotype control (grey filled). APC-labeled anti-CD42b monoclonal antibody was used to gate on platelets. One representative experiment of two separately performed experiments is shown. (Panel (A) and (B) from figure 9 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

106

5.2 Specific Aim 2: To determine the role of properdin in platelet- leukocyte aggregate formation.

5.2.1 Does properdin increase platelet-leukocyte aggregate formation and is this effect dependent on complement activation?

5.2.1.1 Standardization of an assay that allows the measurement of platelet-leukocyte aggregate formation.

In order to study the role of properdin in PLA formation we used a well- characterized ex-vivo PLA formation assay (200). Briefly, as shown in figure 41, whole blood from healthy donors was collected in lepirudin (Refludan) (50 µg/ml) anti-coagulant. Lepirudin is a highly specific thrombin inhibitor (201) with no adverse effects on complement activation or PLA formation and is the recommended anticoagulant to be used in models where complement activation

(202) or PLA formation (176) needs to be studied. Although lepirudin inhibits thrombin, it still allows stimulation of platelets through the PAR receptors PAR-1 and PAR-4. The blood was stimulated with 0-100 µg/ml TRAP at 37°C for 15 min in the presence or absence of physiological forms of properdin (P2-P4). The reaction was stopped using a RBC Lyse/Fix solution and the samples were then washed and stained using specific monoclonal antibodies for a leukocyte marker (anti-CD45) and a platelet marker (anti-CD42b), along with appropriate isotype controls, and acquired by flow-cytometry. The anti-CD45 antibody stains the different leukocyte populations differentially (Fig. 41A; the highest fluorescence corresponds to lymphocytes, lower fluorescence to monocytes and the lowest to granulocytes). We

107

gated on the granulocyte and monocyte populations separately based on their CD45 staining and granularity, and determined the associated CD42b fluorescence

(platelet staining) as an indication of PLA formation. Figures 41B, C (square inset) indicate the population of cells that are positive for both platelets and leukocytes, in either unstimulated blood (Fig. 41B) or TRAP-stimulated blood (Fig. 41C), respectively. Thus, an increase in the CD42b signal (platelets) associated with each leukocyte population, is an indication of PLA formation (Fig. 41).

Collect blood in anticoagulant tubes

Incubate blood with agonist (and properdin)

Lyse RBCs and fix the other cells. B (Unstimulated)

Stain with leukocyte and platelet specific antibody Leukocyte marker - CD45

C Acquire by flow (TRAP cytometry -stimulated)

Platelet marker -marker Platelet CD42b Leukocyte marker - CD45

Fig. 41: General schematic of PLA experiment

108

5.2.1.2 Does properdin increase platelet-leukocyte aggregate formation?

It has been previously shown that when unfractionated properdin is added to non-activated whole blood it leads to increased PLA formation (111).

Unfractionated properdin contains non-physiological aggregate forms of properdin, and thus it remains to be determined if the physiological forms of properdin have an effect on PLA formation. It is also not known whether properdin leads to an increase in PLA formation when the PLAs are induced in whole blood using TRAP as an agonist. This is important because, in a proinflammatory environment, there is activation of platelets as well as leukocytes and thus formation of aggregates. Also, the above mentioned study was carried out in citrate anti-coagulated blood. Anti- coagulants such as citrate significantly inhibit complement activation (177) and thus should not be used if the effects of complement are to be evaluated. Therefore, it still remains to be determined if physiological forms of properdin play a role in PLA formation and if this role is dependent on complement.

Using the previously mentioned ex-vivo method for PLA formation in human whole blood, we evaluated the effect of adding of physiological forms of properdin to TRAP-stimulated, lepirudin-anticoagulated whole blood. Figure 42A shows the gating strategy based on CD45 versus side scatter for gating on granulocyte and monocyte populations. Figures 42B, C are representative dot plots showing an increase in percent of granulocytes with associated CD42b fluorescence (i.e. platelets) upon addition of properdin to TRAP-stimulated blood. Figure 42D is a bar graph of the same data showing that the addition of properdin leads to ~2.6 fold

109

increase in platelet-granulocyte aggregate formation in TRAP-stimulated whole blood compared to TRAP-stimulated whole blood without added properdin.

Monocytes are known to form aggregates with platelets significantly faster than granulocytes in stimulated whole blood (165). They are also known to bind more platelets per cell than the granulocytes (165). The presence of circulating platelet-monocyte aggregates can serve as markers of platelet activation, plaque

no P with P B C D 80 no P *** with P 60

A Granulocytes 40 ve 20

0 CD42b CD42b + CD45 NA Act E F G 100 no P * 80 with P 60 Monocytes

ve 40 20

CD42b 0 CD45 CD42b + NA Act Figure 42: Properdin increases formation of platelet-granulocyte aggregates. Lepirudin (50 µg/ml) anti-coagulated whole blood (20 µl) was incubated either

alone, with P3 (100 µg/ml), with TRAP (20 µM) or simultaneously with P3 and TRAP in a total reaction volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b monoclonal antibody. Leukocytes were gated based on their CD45 staining and granularity. (A) Gating strategy to select granulocyte and monocyte population. (B) and (C) TRAP-activated granulocyte population with and without properdin; (D) Representative bar graph for granulocytes; (E) and (F) TRAP-activated monocyte population with and without properdin; (G) Representative bar graph for monocytes. One representative experiment of three separately performed experiments is shown here. Statistical significance was assessed by determining the p-value using an unpaired Student’s t-test (p< 0.05 (*); ,p<0.001 (***)).

110

activation and acute myocardial infarction (Reviewed in (164)). Upon analyzing the monocyte population in the TRAP-stimulated whole blood, we detected a minor properdin-mediated increase in platelet-monocyte aggregate formation (Fig. 42E-

G). Since the level of platelet-monocyte aggregate formation even in the absence of properdin was almost maximum (90%), it may not be possible to detect highly significant increase in platelet-monocyte aggregates, if any, in these conditions.

Thus, we titrated the dose of the agonist (TRAP) to find the dose at which sub- optimum (~50%) platelet-monocyte aggregate formation was observed (Fig. 43A).

Using the sub-max activation dose we determined that properdin leads to a significantly more platelet-monocyte aggregate formation than TRAP alone (Fig.

43B). Our data also show that the effect of properdin on platelet-leukocyte aggregation is dose dependent (Fig. 44A-B). Properdin causes a significant, ~2.5 fold increase in platelet-granulocyte aggregates starting at the lowest dose of properdin tested (Fig. 44A), whereas a significant increase in platelet-monocyte aggregate formation due to properdin was seen only with doses of properdin greater than 25

µg/ml (Fig. 44B).

111

A B 100 80 ** no P 70 80 with P 60 60 50 40 40 30 20 20 +veCD42b Monocytes 10 % +veCD42b Monocytes 0 0 0 5 10 15 20 NA Act TRAP (µM)

Figure 43: Properdin increases formation of platelet-monocyte aggregates. (A) Lepirudin (50 µg/ml) anti-coagulated whole blood (20 µl) was incubated with increasing doses of TRAP (0 – 20 µM) in a total reaction volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b antibody. Leukocytes were gated based on their CD45 staining and granularity. (B) Lepirudin (50 µg/ml) anti-coagulated whole blood (20 µl) was incubated either alone, with P3 (100

µg/ml), with TRAP (6.5 µM) or simultaneously with P3 and TRAP in a total reaction volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed and stained as in (A). One representative experiment of three separately performed experiments is shown here. Statistical significance was assessed by determining the p-value using an unpaired Students t-test (p< 0.01 (**)

112

A B 100 NA 100 NA Act Act 80 80

60 60

40 40

20 20 %CD42b +ve Monocytes +ve %CD42b %CD42b +ve Granulocytes +ve %CD42b 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Properdin (g/ml) Properdin (g/ml)

Figure 44: Effect of properdin on platelet-leukocyte aggregates is dose dependent. Lepirudin (50 µg/ml) anti-coagulated whole blood (20 µl), in the presence of TRAP (20 µM in (A), 6.5 µM in (B)), was incubated with increasing doses of properdin (0 – 100 µg/ml) in a total reaction volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b monoclonal antibody. Leukocytes were gated based on their CD45 staining and granularity. (A) shows dose dependent formation of platelet-granulocyte aggregates. (B) shows dose dependent formation of platelet-monocyte aggregates. One representative experiment of two separately performed experiments is shown here.

5.2.1.3 Does inhibition of endogenous properdin in blood (without adding properdin) reduce the formation of TRAP-induced PLA?

In order to determine the role of endogenous properdin in whole blood

(without adding exogenous purified properdin), we used a monoclonal anti- properdin antibody (Quidel anti-P#1) to inhibit endogenous properdin. This antibody inhibits complement-dependent functions of properdin, as determined by inhibition of properdin binding to C3b coated sheep erythrocytes (Fig. 45A). It also inhibits the binding of properdin to activated platelets (Fig. 45B). Upon inhibiting endogenous properdin with this antibody, the data show a significant decrease in

113

A B EsC3bs Thr-Act 300 P+aP#1-100µg/ml 200 P+aP#1-50µg/ml P+aP#1-50µg/ml No P P+aP#1-10µg/ml 150 P+aP#1-10µg/ml 200 P+IgG1 isotype P+IgG1 isotype 100 No P P 100 P 50 Counts Counts 0 0 100 101 102 103 104 100 101 102 103 104 Properdin binding Properdin binding C D 60 * 100 *** 50 80 40 60 30 40 20

10 20

0 0 % % CD42bMonocytes +ve for % % CD42b Granulocytes +ve for

Figure 45: Inhibition of endogenous properdin leads to decrease in PLA formation. (A) Properdin (2 µg/ml) was incubated with an anti-properdin monoclonal (anti-P#1) for 15 min at 4°C. C3b-coated sheep erythrocytes (EsC3bs) (5 x 106/100 µl) were then added to the mixture and incubated for 60 min at 4°C.

EsC3bs were washed and stained with an affinity purified polyclonal anti-properdin antibody followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody (B) Properdin (10 µg/ml) was incubated with anti-P#1 for 10 min at RT. Thrombin- activated (Thr-Act) platelets (2x106 platelets/100 µl) were then added to the mixture and incubated for 60 min at RT. Platelets were washed and binding of properdin was assessed using an anti-properdin monoclonal antibody, followed by an Alexa Fluor 488-conjugated anti-mouse IgG antibody. An APC-labeled anti- CD42b monoclonal antibody was used to gate on the platelet population. (C) and (D) Lepirudin (50 µg/ml) anti-coagulated whole blood (20 µl), in the presence of TRAP (20 µM in (C), 7.5 µM in (D)), was incubated with either anti-P#1 (100 µg/ml), IgG1 isotype control (100 µg/ml) or neither in a total reaction volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b antibody. Leukocytes were gated based on their CD45 staining. One representative experiment of two separately performed experiments is shown here. An unpaired Student’s t-test was used to evaluate the statistical significance between groups (p<0.05(*), p<0.001 (***)).

114

platelet-granulocyte aggregates (Fig. 45C) as well as platelet-monocyte aggregates

(Fig. 45D), unlike the isotype control for anti-P#1. This data shows that endogenous properdin plays a role in formation of PLAs.

5.2.1.4 Is the effect of properdin on platelet-leukocyte aggregate formation dependent on complement activation?

Since the antibody used for inhibition of endogenous properdin can inhibit both complement-dependent functions (Fig. 45A) of properdin as well complement- independent functions of properdin (by inhibiting properdin binding to platelets)

(Fig. 45B), we sought to determine if the role of properdin in PLA is dependent on complement activation. We used a cyclic peptide inhibitor of complement known as compstatin, which selectively inhibits C3 ((191); Reviewed in (192)). To confirm the function of this inhibitor and determine the appropriate dose to use, we tested it in a standard AP50 assay using rabbit erythrocytes. Figure 46 shows that alternative pathway mediated lysis of the erythrocytes was reduced to ~1.5% using 20 µM compstatin and almost reached 0% by 40 µM compstatin (Fig. 46). Using this inhibitor at 50 µM we observed a significant decrease in platelet-granulocyte aggregate formation as compared to agonist alone (Fig. 47A). This decrease caused by compstatin was of the same magnitude as the decrease caused by the anti- properdin inhibitory antibody (Fig. 47A). The compstatin control peptide had no effect on platelet-leukocyte aggregates (data not shown). This result indicates that most (or all) of the observed properdin function that leads to increased platelet- granulocyte aggregate formation is dependent on complement activity (i.e. C3 and

115

100 Figure 46: Inhibition of complement in Compstatin serum by compstatin. Rabbit erythrocytes 80 (2x107/100ul) were incubated with 8% Compstatin Control Peptide normal human serum in GVB= (with 2.5 mM 60 MgEGTA) in the presence of increasing concentrations of compstatin (0 – 40 µM) (A414) 40 for 20 min at 37°C (mixing every 5 min). The reaction was stopped by adding 400 μl Lysis cold GVBE. The cells were spun down and

% 20 absorbance of the supernatant was measured at 414 nm and % hemolysis was 0 calculated. One representative experiment 0 10 20 30 40 Concentration (µM) of two separately performed experiments is shown here.

C5 convertase stabilization by properdin or properdin-mediated complement activation). On the other hand, we found that inhibition of complement using compstatin did not cause a decrease in platelet-monocyte aggregate formation.

However, the anti-properdin inhibitory antibody caused a significant decrease in aggregate formation to similar levels as those in non-activated blood (Fig. 47B). This result indicates that the role of properdin in platelet-monocyte aggregate formation is critical, is not mediated by complement-dependent functions of properdin, and may be a direct (complement-independent) effect of properdin.

5.2.2 Is the effect of properdin on platelet-leukocyte aggregate formation mediated by C5a receptors?

Leukocytes have receptors for C5a (C5aR) (Reviewed in (203)). C5a, upon receptor engagement, induces increases in both β1 and β2 integrin expression on neutrophils (204), enhances adhesive interactions of neutrophils to endothelial cells 116

Figure 47: Analysis of inhibition of A * B complement activation 35 * 100 n.s. ** on properdin- 30 80 * mediated platelet- 25 granulocyte and 20 60 n.s. platelet-monocyte 15 40 aggregates. Lepirudin 10 (50 µg/ml) anti- 20 5 coagulated whole blood %CD42b+ve Monocytes %CD42b+ve %CD42b+ve Granulocytes %CD42b+ve 0 0 was preincubated with or without compstatin (50 µM) for 5 min at RT. Subsequently, blood

without compstatin (20 µl) was incubated with TRAP alonewithout (20 µM in compstatin (A), 7.5 µM (20 in (B)) alone or with TRAP and anti-P#1 in a total reaction volumeµl) of 80 µl for 15 min at 37°C. Similarly, whole blood preincubated with compstatin (20 µl) was also

incubated with TRAP (20 µM in (A), 7.5 µM in (B)) in a total reaction volume of 80 µl for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC- labeled anti-CD42b antibody. Leukocytes were gated based on their CD45 staining and granularity. One representative experiment of two separately performed experiments is shown here. An unpaired Student’s t-test was used to evaluate the statistical significance between groups (p<0.05(*)).

(205) and induces respiratory burst. It has been previously shown that inhibition of

C5a-C5aR in the presence of TRAP stimulation leads to a significant reduction in both platelet-granulocyte, as well as platelet-monocyte aggregate formation (145).

Thus, we sought to determine if the effect of properdin on platelet-leukocyte aggregates was also dependent on the C5a-C5aR interaction as has been previously described for TRAP-stimulated blood.

117

5.2.2.1 Analysis of the available C5aR-antagonists in their ability to inhibit C5a- mediated activation of neutrophils.

We first analyzed the ability of the three commercially available C5aR antagonists (C5aRa) to inhibit C5a-mediated activation of neutrophils by flow cytometry. Only one (W-54011) was able to effectively inhibit C5a-mediated activation of the neutrophils by 75 to 100% (Fig. 48A). The results shown here are represented as % CD11b expression relative to the maximum expression observed with C5a using non-activated neutrophils as background.

5.2.2.2 Is the effect of properdin on platelet-leukocyte aggregate formation mediated by C5aR?

In our study, since we found only platelet-granulocyte aggregate formation

(and not platelet-monocyte aggregate formation) to be dependent on complement, we assessed the effect of inhibiting C5a-C5aR interaction on platelet-granulocyte aggregate formation. We used the above mentioned C5aRa, W54011, to determine if the effect of properdin in PLA formation was mediated by C5aR. The data shows that the DMSO control had significantly reduced PLA formation as compared to the group without DMSO. We found that there was no further decrease in PLA formation upon inhibition with the C5aRa as compared to the DMSO control in both the groups to which properdin was added, as well as in the groups without added properdin

(Fig. 48B). Thus, our data indicates that C5aR-mediated activation is not critical for

PLA formation, whether in the presence or absence of properdin.

118

A B 100 no P with P n.s. 90 80 150 70 n.s. 60 100 * 50 * 50 40 30 0 20

(CD11b expression) (CD11b 10 Relative % activation % Relative -50 %Granulocytes CD42b+ve for 0

NA+DMSOC5a+DMSOC5a+C5aRa

Figure 48: Complement-mediated effect of properdin on platelet- granulocyte aggregation is not mediated by C5a-C5aR interaction. (A) Polymorphonuclear cells (5 x 105/100 µl) were preincubated with 50 µM C5aRa for 5 min at RT. PMNs were then stimulated with C5a (0.2 µM) at 37°C for 20 min. Cells were washed and stained with AF488-conjugated anti-CD11b monoclonal antibody. A PE-labeled anti-CD16b monoclonal antibody was used to gate on the PMN population. (B) Lepirudin (50 µg/ml) anti-coagulated whole blood was preincubated with compstatin (50 µM) or C5aRa (50 µM) (or DMSO as a control) for 5 min at RT. The preincubated blood (20 µl) was incubated with TRAP (20 µM) in the presence or absence of properdin (100 µg/ml) in a total volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b monoclonal antibody. Leukocytes were gated based on their CD45 staining and granularity. One representative experiment of three separately performed experiments is shown. Unpaired Student’s t-test was used to evaluate the statistical significance between groups (p<0.05(*)).

5.2.3 Is the properdin-mediated increase in platelet-leukocyte aggregate formation controlled by factor H?

Patients with aHUS, who lack factor H regulation, are known to be at an increased risk for thrombosis (153,198). The role that complement regulatory proteins properdin and factor H play in controlling the formation of platelet- 119

granulocyte aggregates remains unknown. Thus, we evaluated if the properdin- mediated increase in PLA formation is controlled by factor H.

5.2.3.1 Does factor H control the properdin-mediated increase in PLA formation?

To address this question, we used rH19-20 as a competitive inhibitor of factor H binding to cell surfaces, as described in section 1.3.1. The data in figure 49A show a significant (approximate 2-fold) increase in platelet-granulocyte aggregate formation when rH19-20 is added to TRAP-stimulated blood. Upon inhibition of endogenous properdin function using the inhibitory monoclonal antibody (anti-

P#1), the formation of platelet-granulocyte aggregates was significantly reduced

(Fig. 49A) by ~5-fold (when rH19-20 was added) or by ~2.5-fold (when rH19-20 was not added), bringing the level of platelet-granulocyte aggregates down to similar levels when comparing these two groups that received anti-P#1. Similar reduction in platelet-granulocyte aggregate formation was seen upon inhibiting complement activation using compstatin (Fig. 49A), indicating that factor H controls properdin-mediated platelet-granulocyte aggregation in a complement-dependent manner. The data in Fig. 49B shows that increasing doses of rH19-20 lead to a dose- dependent increase in platelet-granulocyte aggregation.

In the case of platelet-monocyte aggregates, inhibition of factor H regulation using rH19-20 also leads to a significant increase in platelet-monocyte aggregate formation (Fig. 50). The magnitude of difference between the groups with and without rH19-20 varied depending on the dose of TRAP used. Thus, lower doses of

TRAP allowed the detection of a larger window between the +/- rH19-20 groups 120

A no rH19-20 B 70 with rH19-20 *** 100 ** *** NA Act 60 80 50 40 60 Granulocytes

ve 30 40 20 20 10 %CD42b +ve Granulocytes +ve %CD42b %CD42b + %CD42b 0 0 0 5 10 15 20 25 rH 19-20 (μM)

Figure 49: Properdin-mediated platelet-granulocyte aggregate formation is controlled by factor H. (A) Lepirudin (50 µg/ml) anti-coagulated whole blood was incubated with either TRAP (20 µM) alone, with TRAP + rH19-20 (13.5 µM), with TRAP + anti-P#1 (100 µg/ml) or with TRAP + rH19-20 + anti-P#1 in a total volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b antibody. Leukocytes were gated based on their CD45 staining and granularity. One representative experiment of three separately performed experiments is shown. (B) Lepirudin anti-coagulated whole blood incubated with increasing doses of rH19-20 and processed as stated above. An unpaired Student’s t-test was used to evaluate the statistical significance between groups (p<0.01(**); p<0.001(***)).

(Fig. 50). We next tested the effect of complement inhibition in the presence of rH19-20. The data in figure 51 shows that all the increase in platelet-monocyte aggregate formation caused by rH19-20 was inhibited upon inhibition of complement activation (Fig. 51). Thus, the increase in platelet-monocyte aggregate formation mediated by inhibition of factor H is complement-dependent. Altogether, these data indicate, for the first time, that TRAP-induced PLA formation is promoted

121

by properdin and is significantly regulated by factor H.

100 no rH19-20 with rH19-20 ** 80 * ** n.s.

Monocytes 60 ve 40

20 %CD42b %CD42b + 0 NA Act Act Act Act 6.5 7.5 8.5 20 TRAP (μM)

Figure 50: Properdin-mediated platelet-monocyte aggregate formation is controlled by factor H. Lepirudin (50 µg/ml) anti-coagulated whole blood was incubated with increasing doses of TRAP (6.5 – 20 μM) in the presence or absence of rH19-20 (13.5 µM) in a total volume of 80 µl with modified HT buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC- labeled anti-CD42b monoclonal antibody. Leukocytes were gated based on their CD45 staining and granularity. One representative experiment of three separately performed experiments is shown here. An unpaired Student’s t-test was used to evaluate the statistical significance between groups (p<0.05(*);p<0.01(**)).

5.2.3.2 Are aHUS-related mutants of factor H limited in their ability to control PLA formation?

We have already determined that aHUS-related mutants of factor H are impaired in their ability to control properdin-mediated C3b deposition on activated platelets by using mutant forms of rH19-20 and comparing it to WT rH19-20 (Fig.

122

34). We have also shown that factor H regulates PLA formation because it is

70 * * Figure 51: Increase in platelet- monocyte aggregation caused by 60 rH19-20 is dependent on 50 complement. Lepirudin (50 µg/ml) anti-coagulated whole blood, in the 40 presence or absence of compstatin (50 30 μM) was incubated with TRAP (6.5 μM) in the presence or absence of rH19-20 20 (13.5 µM) in a total volume of 80 µl with 10 modified HT buffer for 15 min at 37°C.

%Monocytes +ve for CD42b +ve for %Monocytes The samples were fixed using RBC 0 lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b antibody. Leukocytes were gated based on their CD45 staining. One representative expt of two separately

performed experiments is shown. Statistical significance was assessed by determining the p-value using an unpaired t-test (p<0.05 (*)).

significantly increased when cell surface regulation by factor H is inhibited by WT rH19-20 (Fig. 49 and 50). Thus, we hypothesized that aHUS-related mutants of factor H would be affected in their ability to control PLA formation. To test this, we used three aHUS-related mutants of rH19-20: D1119G, L1189F and R1215G

(described in section 5.1.3.2) as competitors in a PLA assay. The data show that all the mutants caused as much of an increase in PLA formation as did WT rH19-20

(Fig. 52). This was surprising given the previous data indicating that these mutant forms of rH19-20 could not compete with full length factor H and thus were unable to cause an increase in properdin-mediated C3b deposition on platelets (Fig. 34).

123

The data indicate that these aHUS-related mutations do not affect the ability of factor H to control PLA formation.

100 Figure 52: aHUS-related rH19-20 mutants of factor H are not affected in their ability 80 to compete with full-length factor H and thus increase the formation of platelet- 60 leukocyte aggregates. Lepirudin (50 µg/ml) anti-coagulated whole blood was 40 incubated in the presence of TRAP (20 µM) with either WT rH19-20 or rH19-20 mutants 20 (D1119G, L1189F, R1215G) (13.5 µM) in a total volume of 80 µl with modified HT %Granulocytes +ve for CD42b for +ve %Granulocytes 0 buffer for 15 min at 37°C. The samples were fixed using RBC lyse/fix buffer for 10 min at RT, washed and stained using PE labeled anti-CD45 monoclonal antibody and APC-labeled anti-CD42b monoclonal antibody. Leukocytes were gated based on

their CD45 staining and granularity. One representative experiment of two separately performed experiments is shown here.

124

Chapter 6

Discussion

In this thesis, it is hypothesized that properdin promotes complement activation on platelets and the formation of platelet-leukocyte aggregates. Properdin

(i) interacts directly with activated platelets and initiates complement activation on their surface and (ii) increases the formation of platelet-leukocyte aggregates by promoting complement activation on the platelet/leukocyte interface. The data presented in this thesis mostly support this hypothesis.

Our studies reveal that the physiological forms of properdin bind to stimulated platelets, but not to resting platelets, in a manner that is not proportional to P-selectin exposure, leading to the formation of a novel C3 convertase (P-

C3(H2O),Bb) on its surface and allowing the activation of the alternative pathway of the complement system. In addition, C3(H2O) can also initiate activation of complement, as long as properdin is present to stabilize the convertases. The data also show that properdin released by neutrophils binds activated platelets. In addition, properdin-initiated complement activation is regulated by factor H, and aHUS-related mutants affect factor H regulation in this context. Our studies also go 125

on to show that properdin increases PLA formation and inhibition of endogenous properdin leads to a decrease in PLA formation. A decrease in platelet-granulocyte aggregate formation by inhibition of complement activation shows that most of the observed properdin-mediated platelet-granulocyte aggregate formation is dependent on complement activity, but does not depend on C5a-C5aR interaction.

The absence of an effect of complement inhibition on platelet-monocyte aggregate formation indicates that properdin-mediated platelet-monocyte aggregation may be independent of complement activity. The data also show that the properdin- mediated increase in PLA is controlled by factor H, but aHUS related mutants of factor H are not affected in their ability to control PLA. Our studies collectively suggest that properdin is essential for alternative pathway activation to proceed on activated platelets, as well as for increasing PLA formation by both complement- dependent (i.e. promotion of complement activation at the platelet/leukocyte interface) and/or complement-independent mechanisms.

Our results show that pure human properdin, that contains non- physiological aggregates, binds to both non-activated and activated platelets (Fig.

14). On the other hand, physiological forms of properdin bind only to platelets that have been activated by strong agonists (Fig. 18), but not to non-activated platelets

(Fig. 16 A-C), in a dose- and time-dependent manner (Fig. 17). Studies using physiological properdin forms separated from non-physiological aggregates have shown that native properdin does not bind to some of the surfaces previously described in literature, such as rabbit erythrocytes, live Jurkat cells and Neisseria sp

126

(94,108). Thus it was essential to carry out our study with native properdin forms separated from non-physiological aggregates.

Thrombin activates platelets via PAR receptors by pathways dependent on phospholipase C and/or phospholipase A2, the latter of which includes the arachidonic acid transformation pathway. The arachidonic acid pathway bypasses the need for agonist receptors and activates platelets via the cyclooxygenase (COX) pathway by using thromboxane synthase to produce the platelet agonist thromboxane A2 (206). Aside from platelets, neighboring activated cells such as leukocytes and endothelial cells also express COX isoenzymes (207), which could further contribute to platelet activation. Inhibitors of COX lead to the inhibition of complement-enhanced platelet aggregation and release of serotonin (159-161,208-

210). Our data show that properdin, by itself, does not activate the platelets as measured by CD62P expression, gpIIbIIIa expression and Annexin V binding (Fig.

20). Platelets activated by weak agonists, such as ADP and epinephrine, support less complement activation than platelets activated by thrombin or arachidonic acid

(144). The capacity of platelets to activate complement on their surface when exposed to plasma or serum is proportional to the extent of platelet activation (144) and alternative pathway activation has been shown to occur due to the binding of

C3b to activated platelets via P-selectin (CD62P) (119). Herein, the data show that arachidonic acid-stimulated platelets induced significantly lower overall exposure of

P-selectin compared with thrombin-activated platelets (Fig. 19A-C). Nevertheless, properdin binding was ~3-4 fold higher (Fig. 19D-F) and C3b deposition was ~10-

127

fold higher (Fig. 27) on the arachidonic acid-stimulated platelets, when compared to thrombin-activated platelets at maximal platelet P-selectin expression for both agonists. As previously mentioned, it has been reported that activated platelets can bind C3b via P-selectin on their surface (119). The bound C3b forms a C3 convertase and leads to complement activation (119). Our data show that, unlike the

C3b/CD62P mechanism for alternative pathway complement activation described by del Conde et al. (119), properdin binding to activated platelets and C3b deposition are not proportional to the expression of CD62P. Thus, properdin binding may depend on agonist-specific varying exposure of cell surface marker(s) on the platelets. In agreement with this notion, proteomic expression on platelets has been shown to significantly vary depending on whether the platelets were activated using thrombin, arachidonic acid, or collagen (211).

Native forms of properdin are known to interact with cell bound C3b (with greater affinity than to fluid phase C3b) (92) and also with C3(H2O) (93). Platelets can also bind C3b as well as C3(H2O) (119). Thus, it was necessary to assess whether properdin was binding to platelets through C3 products. The data show that properdin does not rely on C3 fragment deposition on the platelet for binding

(Fig. 21). Also, the presence of C3 components cannot be detected on the surface of activated platelets by flow cytometry (Fig. 22). Properdin is a highly positively charged protein (isoelectricpoint >9.5) that can interact with certain surfaces directly by recognizing glycosaminoglycan (GAG) chains of surface proteoglycans

(97,104). Candidate sulfated GAGs shown to interact with properdin include heparin

128

(103), heparan sulfate (97,104), dextran sulfate (105), fucoidan (105), and chondroitin sulfate (97). Interestingly, chondroitin sulfate-A, which is released by platelets and found on the platelet surface upon activation (212,213), enhances the binding of properdin to the activated platelets (Fig. 23). In addition, removal of surface proteins from activated platelets leads to a dose-dependent reduction in the binding of properdin to platelets (Fig. 24). These results suggest that both proteins as well as glycosaminoglycans (i.e. glycosaminoglycan side chains of proteoglycans) may play a role in the binding of properdin to the platelet surface. Other ligands for properdin on cells include DNA on late apoptotic and necrotic cells (85), and bacterial LPS and lipooligosaccharide (106). Thus, all cell surface molecules

(identified to date) shown to interact directly with properdin on cells are negatively charged.

When properdin is covalently bound to a biosensor surface, it can recruit C3b and factor B to form C3b,Bb,P (Hourcade, 2006). Additional studies demonstrating the ability of properdin to initiate complement activation (by forming de novo C3 convertases on cell surfaces) include studies where human embryonic kidney cells

(102) or E. coli (95) were transfected with a vector expressing a transmembrane form of properdin on the cell surface, leading to alternative pathway activation on the cell. Our data shows that alternative pathway activation proceeds when properdin is bound to the surface of activated platelets, as measured by C3b/iC3b

(Fig. 27, 29) and C5b-9 deposition (Fig. 31), after exposing the platelets to properdin-depleted serum or NHS. NHS was used as a serum control that has

129

undergone less post-extraction processing than the depleted sera, and the results were similar. Interestingly, this properdin-initiated complement activation occurs even when complement regulation on the platelets is normal (i.e. normal membrane-bound and fluid phase complement regulatory proteins). Thus, in physiological conditions, when complement regulation on the platelets is normal, complement activation may aid in the clearance of spent platelets from circulation and help to keep the prothrombotic effects of activated platelets in check. At sites of tissue injury where there is activation of platelets and neutrophils leading to release of neutrophil-derived properdin, properdin-mediated complement activation may additionally contribute to local inflammation by the release of pro-inflammatory mediators such as C3a and C5a and/or directly to tissue damage.

The fluid phase complement regulatory protein factor H can bind to activated platelets (196,197) (Fig. 32). Patients with aHUS have mutations in the C-terminus of factor H (which impair its ability to efficiently bind to cell surfaces) and are associated with the presence of complement deposits on their platelets (153). We tested complement activation on platelets when surface regulation by factor H was inhibited. We used recombinant protein rH19-20 to assess complement activation in the absence of cell surface factor H regulation. rH19-20, which consists of the two C- terminal domains 19 and 20 of factor H, is a recombinant protein that competes with full length factor H for binding to C3b and polyanions on cell surfaces (54,55).

We found that C3b deposition is significantly enhanced when cell surface protection by factor H is blocked, especially on activated platelets that have properdin on their

130

surface (Fig. 33). The presence of aHUS-related mutations (D1119G, L1189F,

R1215G) in domains 19-20 of factor H significantly impaired the ability of rH19-20 to enhance properdin-mediated C3b deposition on the surface of activated platelets

(Fig. 34). Clinical data suggests that enhanced platelet-associated complement activation correlates with increased thrombotic events in patients with aHUS

(153,198) due to mutations in the C-terminus of factor H that impair cell surface protection. Properdin-mediated complement activation may contribute to these phenomena (214).

Recently, Hamad et al. (146) showed, using specific antibodies, that the C3 that binds to platelets consists mainly of C3(H2O), and not C3b. Because C3(H2O) is generated in the fluid phase of blood by the spontaneous hydrolysis of the thioester bond in C3, the C3(H2O) that was bound to the platelet was non-proteolytically activated (146). In agreement with these results, our data with thrombin-activated platelets show that purified C3(H2O) indeed binds to activated platelets while C3b binding cannot be detected (Fig. 37A). Platelet-bound C3(H2O) does not lead to the formation of a C3 convertase on the thrombin-activated platelet surface when exposed to factors B and D (Fig. 38B). On the other hand, properdin, by recruiting

C3(H2O) (Fig. 36B) to the surface of activated platelets, allows the formation of a novel cell bound C3(H2O),Bb convertase in the presence of factors B and D (Fig.

38B), which can lead to the activation of the alternative pathway as measured by

C3b and C9 deposition (Figs. 27, 29 and 31). It is likely that platelet-bound properdin, aside from acting as an initiating point for alternative pathway activation

131

on the platelets, is also stabilizing the newly formed convertase, facilitating detection of the convertase in our experimental system by making it more resistant to decay versus the platelets without properdin on their surface.

Interestingly, the results with arachidonic-acid activated platelets, which were not assessed in the previous study (146), indicate that the activated platelets bind detectable levels of C3b in addition to C3(H2O) (Fig. 37). Aside from the

C3(H2O) that is available due to C3 tickover (Reviewed in (215)), Nilsson and

Nilsson-Ekdahl have hypothesized (216) that the rate of hydrolysis of C3 to C3(H2O) may be accelerated by the interaction of C3 with certain biological surfaces, such as platelets and that this C3(H2O) may serve as an initiating molecule of the alternative pathway. As mentioned, C3(H2O) no longer has a reactive thioester for interacting covalently with cell surfaces, and is normally found forming part of fluid phase C3 convertases for initiating the alternative pathway. However, figure 38C shows that

C3(H2O) bound to arachidonic acid-activated platelets can in fact lead to the formation of a novel cell-bound convertase, which can only be detected if properdin is present to stabilize the convertase (Fig. 38C, dotted line). Although we did not detect direct binding of whole C3 to thrombin-activated platelets, and only minimal binding of C3 to arachidonic acid-activated platelets, increased availability of

C3(H2O) may be triggered by contact activation of C3 on gas bubbles, such as those that occur in cardiopulmonary devices and in decompression sickness (reviewed in

(216)), potentially contributing to complement activation on platelets.

132

It is not known why C3(H2O) binds more than C3b to the activated platelets and to properdin on the platelets (Figs. 36-37), despite being components that are structurally and functionally similar (18). CR2 (CD21; C3d/iC3b receptor) may be a receptor for C3(H2O) on B lymphocytes (194), and CR2 has been identified on the surface of platelets (217). It is also possible that the C3a region that is present in

C3(H2O), but not in C3b, may contain a site important for interacting with activated platelets. The interacting region between C3 (1402-1435aa) and properdin (TSR-5) has been previously determined (90). Our data indicate for the first time that additional recruitment of C3(H2O) by properdin, on activated platelets leads to de novo formation of cell bound [C3(H2O),Bb] convertases. These results suggest that properdin on the platelet surface may act as a second contact point for C3(H2O) increasing its avidity for activated platelets and allowing C3(H2O) to form a novel functional C3 convertase [C3(H2O),Bb].

Neutrophils are known to store properdin in their secondary granules and release it upon activation by stimuli such as PMA, fMLP, C5a, and TNF-α (77).

Properdin derived from PMN cells binds to activated platelets (Fig. 40A-B).

Neutrophils are also known to store C3 components in their granules and release them upon activation (218). We did not detect binding of C3 components from PMN supernatants to activated platelets as measured by flow cytometry (Fig. 40C-D), indicating that binding of properdin released by neutrophils does not require presence of C3 components on the platelet surface. Under physiological inflammatory conditions and normal complement regulation, properdin may direct

133

low level complement activation on activated platelets, and contribute to opsonizing spent platelets for removal. Serum inhibits the ability of properdin to bind to activated platelets in a dose-dependent manner (Fig. 39). Serum has also been shown to inhibit binding of purified physiological properdin to zymosan, necrotic cells, and C. pneumonia (94,109). Inhibitors of properdin binding to surfaces in serum are yet to be identified. It is possible that properdin that is freshly secreted by different cell types (76,77,79) and does not quickly encounter a nearby platelet surface to bind, will eventually lose its ability to bind to surfaces directly once it comes in contact with serum. This regulation would prevent unwanted properdin- mediated complement damage at more distant/bystander cell surfaces while keeping the conventional function of stabilizing the C3 and C5 convertases of the alternative pathway intact. As mentioned, properdin binds DNA and sulfated glucoconjugates. Thus, fluid phase forms of DNA (219) or glycoproteins could potentially serve as regulators of properdin/surface interactions once properdin has left the microenvironment of the cells producing it (i.e. neutrophils).

Properdin-mediated complement activation (by inducing formation of MAC and release of C3a and C5a) may stimulate degranulation and activation of other resting platelets (148,149). Moreover, the properdin-induced complement activation on platelets significantly increases when cell surface protection by factor

H is inhibited using a competitive inhibitor (Fig. 33). Thus, this properdin-mediated mechanism may be exacerbated in diseases where complement regulation is compromised (i.e. PNH, aHUS) (153,154).

134

Complement also plays a role in tissue damage in many inflammatory diseases that are associated with increased platelet activation and coagulopathies that are not directly associated with defects in complement regulation (i.e. cardiovascular disease, certain cancers, sepsis, among others) (147). Thus, it is possible that in vascular injury, the local inflammatory leukocytes may readily produce properdin (76-79,81,99). This properdin would be available to platelets at high concentrations in the local microenvironment as they become activated by the damaged endothelium and interact with leukocytes (forming PLAs (200)), potentially contributing to complement-mediated disease pathogenesis. In agreement with this notion, Ruef et al. (111) showed that pure, unfractionated properdin (with non-physiological polymers), when added to whole blood, increases the formation of PLAs. This study, however, had various limitations including: (a) using citrate as an anticoagulant, which significantly inhibits complement activation due to its calcium and magnesium binding properties (177);

(b) assessing the effect of unfractionated properdin that contains non-physiological aggregates, which are known to bind to surfaces non-specifically (94), including non-activated platelets (Fig. 14; (110)). Our study aimed to overcome these limitations by using only physiological forms of properdin, and by using an appropriate anticoagulant (lepirudin) that is a specific thrombin inhibitor (201) and does not interfere with complement activation (202).

The data show that physiological forms of properdin lead to significantly more platelet-granulocyte aggregate formation in TRAP-stimulated, lepirudin-

135

anticoagulated whole blood as compared to TRAP stimulation alone (Fig. 42D).

Upon titrating the dose of TRAP to achieve sub-optimal platelet-monocyte aggregate formation (Fig. 43A), we observed that properdin also significantly increases the formation of platelet-monocyte aggregates (Fig. 43B). The effect of properdin on

PLA formation was found to be dose-dependent. While properdin had a significant effect on platelet-granulocyte aggregate formation even at the lowest dose tested

(6.25 µg/ml) (Fig. 44A), the effect of adding properdin on platelet-monocyte aggregate formation was seen only at doses higher than (25 µg/ml) (Fig. 44B). This indicates that greater amounts of properdin are required to have a significant effect on platelet-monocyte aggregates as compared to platelet-granulocyte aggregates, or that the endogenous properdin in blood could be sufficient for effective induction of platelet-monocyte aggregates.

Upon inhibition of endogenous properdin in the blood using a monoclonal antibody (anti-P#1) that inhibits complement dependent functions of properdin

(Fig. 45A) as well as the binding of properdin to platelets (Fig. 45B), the data showed a significant decrease in platelet-granulocyte (Fig. 45C) and platelet- monocyte aggregate formation (Fig. 45D). To determine if the role of properdin in

PLA formation is mediated by complement, we inhibited complement activation using compstatin. The dose of compstatin used in our studies (50 µM) is greater than the dose required to see 100% inhibition of complement in NHS (Fig. 46). In the case of platelet-granulocyte aggregates, inhibition of complement using compstatin led to a significant decrease in PLA formation (Fig. 47A), to a level

136

similar to that observed when endogenous properdin was inhibited. These results suggest that all the effect of properdin on the formation of platelet-granulocyte aggregates is likely due to the complement-mediated functions of properdin (i.e. convertase stabilization). Interestingly, in the case of platelet-monocyte aggregates, although inhibiting properdin dramatically reduced platelet-monocyte aggregates to levels similar to non-activated blood, the inhibition of complement with compstatin did not lead to significant reduction of aggregates (Fig. 47B). This suggests that the role of properdin in the formation of platelet-monocyte aggregates may be direct and include functions beyond those mediated by complement activation, such as potentially serving as a bridge between both cell types. Future studies aimed at understanding the molecular mechanisms involved in potential complement- independent function(s) of properdin in these phenomena are warranted.

We next wanted to understand the mechanism by which complement plays a role in platelet-granulocyte aggregate formation. Leukocytes have receptors for C5a on their surface (Reviewed in (203)). Hamad et al (145) has shown that TRAP- mediated PLA formation is dependent, at least in part, on C5a-C5aR interaction and can be inhibited using a C5aR antagonist. The data show that although the antagonist that we tested (W-54011) was capable of inhibiting C5aR-mediated activation of granulocytes (Fig. 48A), it did not inhibit platelet-granulocyte aggregate formation in whole blood as compared to its DMSO control (Fig. 48B). The

DMSO control by itself led to a significant reduction in PLA formation as compared to when no DMSO was added to the sample (Fig. 48B). Our result differs from the

137

previous report (145), however, it is not clear in that study what the C5aRa was dissolved in and if the effect of this vehicle on PLA formation was tested (as we have tested for our DMSO control). Thus, although the effect of properdin on platelet- granulocyte aggregates is complement-dependent, it is not mediated by C5aR stimulation in our hands. Given the relevance of these findings, confirmation could be carried out by using a direct inhibitor of C5a (i.e. inhibiting anti-C5a antibody) in future studies.

Our data shows that properdin leads to deposition of C3b/iC3b (Fig. 27, 29) on the surface of activated platelets. It is possible that the effect of properdin may be mediated by the interaction of C3b/iC3b on platelets with its receptors on leukocytes (CR1 and CR3). Soluble C5b-9 (sC5b-9) is also formed during complement activation on platelets (145). sC5b-9 is capable of activating platelets

(Reviewed in (220)) and blockade of C5b-9 generation can inhibit activation of platelets as well as leukocytes (180). Thus, it is possible that sC5b-9 may be playing a role in the interaction of platelets with leukocytes. Granulocytes also have receptors for C3a, although the expression of the C3a receptor on granulocytes is 20 times lower than that of the C5a receptor (221). It is also known that C3a is 50-100 fold less potent than C5a in causing neutrophil transient shape change and respiratory burst (222). However, the effect of inhibiting C3a-C3aR interactions on platelet-granulocyte aggregate formation remains to be tested.

It is known that patients with aHUS, who lack factor H regulation due to mutations in the protein, are at a higher risk for thrombosis ((198); Reviewed in 138

(199)). Our data show that inhibition of factor H regulation at cell surfaces using a competitive inhibitor of factor H binding (rH19-20) leads to a significant increase in platelet-granulocyte aggregate formation (Fig. 49) as well as platelet-monocyte aggregate formation (Fig. 50A). This is the first study to show that factor H plays a critical role in controlling PLA formation. Our data also show that this exacerbation of PLA formation upon inhibition of factor H is mediated by properdin, because inhibition of endogenous properdin leads to almost complete reduction in platelet- granulocyte aggregate formation that was caused by rH19-20 (Fig. 49A). We also found that for platelet-granulocyte aggregates, inhibition of complement activation in the presence or absence of rH19-20 reduced platelet-granulocyte aggregate formation to the same level as when properdin was inhibited, which was significantly and consistently lower than basal TRAP induced platelet-granulocyte aggregate formation (Fig. 49A). In the case of platelet-monocyte aggregates, inhibition of factor H increased platelet-monocyte aggregate formation (Fig. 50), while compstatin inhibited all the rH19-20-mediated increase (Fig. 51). Altogether, these data indicate that factor H controls PLA formation in a complement- dependent manner.

Unexpectedly, we found that three aHUS-related mutants of rH19-20

(D1119G, L1189F and R1215G) were not affected in their ability to compete with full-length factor H and thus promoted PLA formation in a mode similar to the WT rH19-20 (Fig. 52). This suggests that mutations at sites within the C-terminus other than the ones tested may perturb the interaction of factor H with platelets, and thus

139

warrant further investigation. It is also possible that factor H binds to sites other than C3b and polyanions in the context of PLA; for example factor H can bind neutrophils via integrin CR3 (223). Thus, aHUS-related mutants of rH19-20 used in this study, which are affected in their ability to inhibit binding of factor H to C3b and polyanions (39), may not be impaired in their ability to inhibit binding of factor H to other ligands/receptors for factor H on cells. Factor H also binds platelets directly, in the absence of complement activation, via domains other than 19-20 (i.e. 1-7)

(196). Thus, the effect of mutations in domain 19-20 could be tested in combination with other domains that may play a role in the binding of factor H to platelets in the context of PLA. Alternatively, certain aHUS patients have nonsense mutations in factor H that result in truncated factor H without the C-terminus (224), which can no longer bind efficiently to cell surfaces. Thus, based on our results indicating a central role for the factor H C-terminus in controlling PLA formation, these individuals would likely be susceptible to increased PLA formation and thrombosis.

The data collectively indicate a new mechanism of alternative pathway activation on stimulated platelets that is not proportional to the expression of

CD62P, is initiated by properdin, and requires the recruitment of C3(H2O) or C3b for the formation of a novel initiating cell-bound C3 convertase [P-C3(H2O),Bb] or of [P-

C3b,Bb] (Fig. 53; model). C3(H2O) can also initiate convertase formation, but requires properdin for stabilization. Thus, in a pro-inflammatory microenvironment, where both activated leukocytes as well as activated platelets are present, properdin produced by the activated leukocytes may be available to

140

activated platelets in high concentrations. The data also demonstrate that properdin plays an important role in PLA formation and while the effect of properdin in granulocytes is mostly complement-dependent, for monocytes, properdin acts in ways independent of complement activation. We also show that the formation of platelet-granulocyte aggregates is independent of C5a-C5aR interaction and that

PLA formation is controlled by factor H (Fig. 54; model). Overall, these mechanisms may contribute to complement activation on platelets and increased PLA formation in physiological inflammation as well as contributing to complement-mediated tissue damage in inflammatory diseases.

141

P P P P C3 C5a C3(H2O) C5b-C9 (MAC) Bb C3b C3b C3b P Bb P C3(H2O),Bb Activated Platelet and C3b,Bb convertases C3(H O) 2 Bb P

Figure 53: Model: Properdin-mediated complement activation on activated platelets. Properdin (orange triangles) released by PMNs binds to activated platelets and can recruit C3(H2O) (thrombin-activated platelets) or both C3(H2O) and C3b (arachidonic acid–activated platelets) to stimulated platelets, allowing the formation of a novel cell-bound C3(H2O),Bb or a C3b,Bb convertase. C3(H2O) can also bind to stimulated platelets and in the presence of properdin can promote C3(H2O),Bb convertase formation. These events can then lead to alternative pathway-complement activation (as shown by C3b and MAC deposition) on the platelet surface. Properdin that does not encounter a nearby cell surface to bind may lose the ability to bind to surfaces directly soon after it is in contact with blood, therefore preventing unwanted properdin-mediated complement damage in surrounding areas while keeping the conventional function of stabilizing the C3 and C5 convertases of the alternative pathway. (Figure 9 from Saggu et al, JImmunol, 2013, 190(12):6457-67; Appendix A)

142

Local pro-inflammatory microenvironment Granulocytes Increased Inflammation Soluble Leukocyte P C3a, P P MAC P P P C5a P P P P P P P P P P P Convertase formation P P P CR3/Mac-1 Convertase Complement C3 ? Blood vessel stabilization initiation Interaction of C3 opsonization products with leukocyte C3b Bb Bb receptors Bb C3(H2O) C3b C3b iC3b C3(H2O) Activated Platelet ? fH Complement activation controlled by fH Monocytes

Figure 54: Model: Role of properdin in PLA formation. In a pro-inflammatory microenvironment, properdin produced by the activated leukocytes may be available to activated platelets in high concentrations. The effect of properdin in granulocytes is mostly complement-dependent. For monocytes, properdin acts in ways independent of complement activation (i.e. potentially serving as a bridge). Formation of platelet-granulocyte aggregates is independent of C5a-C5aR interaction. PLA formation is controlled by factor H. Overall, these mechanisms may contribute to complement activation on platelets and increased PLA formation in physiological inflammation as well as to complement-mediated tissue damage in inflammatory diseases.

143

Reference List

1. Walport, M. J. 2001. Complement. First of two parts. N. Engl. J. Med. 344: 1058-1066.

2. Pangburn, M. K., V. P. Ferreira, and C. Cortes. 2008. Discrimination between host and pathogens by the complement system. Vaccine 26 Suppl 8: I15-I21.

3. Carroll, M. C. 1998. The role of complement and complement receptors in induction and regulation of immunity. Annu. Rev. Immunol. 16: 545-568.

4. Monk, P. N., A. M. Scola, P. Madala, and D. P. Fairlie. 2007. Function, structure and therapeutic potential of complement C5a receptors. Br. J. Pharmacol. 152: 429-448.

5. Carroll, M. C. 2004. The complement system in regulation of adaptive immunity. Nat. Immunol. 5: 981-986.

6. Molina, H., V. M. Holers, B. Li, Y. Fung, S. Mariathasan, J. Goellner, J. Strauss- Schoenberger, R. W. Karr, and D. D. Chaplin. 1996. Markedly impaired humoral immune response in mice deficient in complement receptors 1 and 2. Proc. Natl. Acad. Sci. U. S. A 93: 3357-3361.

7. Marchbank, K. J., C. C. Watson, D. F. Ritsema, and V. M. Holers. 2000. Expression of human complement receptor 2 (CR2, CD21) in Cr2-/- mice restores humoral immune function. J. Immunol. 165: 2354-2361.

8. Taylor, P. R., A. Carugati, V. A. Fadok, H. T. Cook, M. Andrews, M. C. Carroll, J. S. Savill, P. M. Henson, M. Botto, and M. J. Walport. 2000. A hierarchical role for classical pathway complement proteins in the clearance of apoptotic cells in vivo. J. Exp. Med. 192: 359-366.

9. Flierman, R., and M. R. Daha. 2007. The clearance of apoptotic cells by complement. Immunobiology 212: 363-370.

10. Trouw, L. A., A. M. Blom, and P. Gasque. 2008. Role of complement and complement regulators in the removal of apoptotic cells. Mol. Immunol. 45: 1199-1207.

144

11. Tausk, F., and I. Gigli. 1990. The human C3b receptor: function and role in human diseases. J. Invest Dermatol. 94: 141S-145S.

12. Arumugam, T. V., I. A. Shiels, T. M. Woodruff, D. N. Granger, and S. M. Taylor. 2004. The role of the complement system in ischemia-reperfusion injury. Shock 21: 401-409.

13. Holers, V. M. 2003. The complement system as a therapeutic target in autoimmunity. Clin. Immunol. 107: 140-151.

14. Yasojima, K., C. Schwab, E. G. McGeer, and P. L. McGeer. 2001. Complement components, but not complement inhibitors, are upregulated in atherosclerotic plaques. Arterioscler. Thromb. Vasc. Biol. 21: 1214-1219.

15. Niculescu, F., and H. Rus. 2004. The role of complement activation in atherosclerosis. Immunol. Res. 30: 73-80.

16. Niculescu, F., and H. Rus. 1999. Complement activation and atherosclerosis. Mol. Immunol. 36: 949-955.

17. Muller-Eberhard, H. J. 1988. Molecular organization and function of the complement system. Ann. Rev. Biochem. 57: 321-347.

18. Pangburn, M. K., R. D. Schreiber, and H. J. Muller-Eberhard. 1981. Formation of the initial C3 convertase of the alternative complement pathway. Acquisition of C3b-like activities by spontaneous hydrolysis of the putative thioester. J. Exp. Med. 154: 856-867.

19. Sim, R. B., T. M. Twose, D. S. Paterson, and E. Sim. 1981. The covalent-binding reaction of complement component C3. Biochem. J. 193: 115-127.

20. Medicus, R. G., O. Gotze, and H. J. Muller-Eberhard. 1976. Alternative pathway of complement: Recruitment of precursor properdin by the labile C3/C5 convertase and the potentiation of the pathway. J. Exp. Med. 144: 1076-1093.

21. Pangburn, M. K., and H. J. Muller-Eberhard. 1986. The C3 convertase of the alternative pathway of human complement. Enzymatic properties of the biomolecular proteinase. Biochem. J. 235: 723-730.

22. Fearon, D. T., and K. F. Austen. 1975. Properdin: Binding to C3b and stabilization of the C3b-dependent C3 convertase. J. Exp. Med. 142: 856-863.

23. Discipio, R. G., C. A. Smith, H. J. Muller-Eberhard, and T. E. Hugli. 1983. The activation of human complement component C5 by a fluid phase C5 convertase. J. Biol. Chem. 258: 10629-10636.

145

24. Nilsson, U. R., R. J. Mandle, and J. A. McConnell-Mapes. 1975. Human C3 and C5: subunit structure and modifications by trypsin and C42-C423. J. Immunol. 114: 815-822.

25. Vogt, W., G. Schmidt, B. Von Buttlar, and L. Dieminger. 1978. A new function of the activated third component of complement: Binding to C5, an essential step for C5 activation. Immunology 34: 29-40.

26. Muller-Eberhard, H. J. 1984. The membrane attack complex. Springer Semin. Immunopathol. 7: 93-141.

27. Pangburn, M. K., and H. J. Muller-Eberhard. 1984. The alternative pathway of complement. Springer Seminars in Immunopathology 7: 163-192.

28. Discipio, R. G. 1992. Ultrastructures and interactions of complement factors H and I. J. Immunol. 149: 2592-2599.

29. Esparza-Gordillo, J., J. M. Soria, A. Buil, L. Almasy, J. Blangero, J. Fontcuberta, and C. S. Rodriguez de. 2004. Genetic and environmental factors influencing the human factor H plasma levels. Immunogenetics 56: 77-82.

30. Schwaeble, W., J. Zwirner, T. F. Schulz, R. P. Linke, M. P. Dierich, and E. H. Weiss. 1987. Human complement factor H: expression of an additional truncated gene product of 43 kDa in human liver. Eur. J. Immunol. 17: 1485- 1489.

31. Brooimans, R. A., A. A. Van der Ark, W. A. Buurman, L. A. Van Es, and M. R. Daha. 1990. Differential regulation of complement factor H and C3 production in human umbilical vein endothelial cells by IFN-gamma and IL-1. J. Immunol. 144: 3835-3840.

32. Chen, M., J. V. Forrester, and H. Xu. 2007. Synthesis of complement factor H by retinal pigment epithelial cells is down-regulated by oxidized photoreceptor outer segments. Exp. Eye Res. 84: 635-645.

33. Tu, Z., Q. Li, H. Bu, and F. Lin. 2010. Mesenchymal stem cells inhibit complement activation by secreting factor H. Stem Cells Dev.

34. Harrison, R. A., and P. J. Lachmann. 1980. The physiological breakdown of the third component of human complement. Mol. Immunol. 17: 9-20.

35. Pangburn, M. K., R. D. Schreiber, and H. J. Muller-Eberhard. 1977. Human complement C3b inactivator: Isolation, characterization, and demonstration of an absolute requirement for the serum protein BIH for cleavage of C3b and C4b in solution. J. Exp. Med. 146: 257-270.

146

36. Weiler, J. M., M. R. Daha, K. F. Austen, and D. T. Fearon. 1976. Control of the amplification convertase of complement by the plasma protein BIH. Proc. Natl. Acad. Sci. USA 73: 3268-3272.

37. Whaley, K., and S. Ruddy. 1976. Modulation of the alternative complement pathway by BIH globulin. J. Exp. Med. 144: 1147-1163.

38. Schreiber, R. D., M. K. Pangburn, P. Lesavre, and H. J. Muller-Eberhard. 1978. Initiation of the alternative pathway of complement: recognition of activators by bound C3b and assembly of the entire pathway from six isolated proteins. Proc. Natl. Acad. Sci. USA 75: 3948-3952.

39. Ferreira, V. P., A. P. Herbert, C. Cortes, K. A. McKee, B. S. Blaum, S. T. Esswein, D. Uhrin, P. N. Barlow, M. K. Pangburn, and D. Kavanagh. 2009. The binding of factor H to a complex of physiological polyanions and C3b on cells is impaired in atypical hemolytic uremic syndrome. J. Immunol. 182: 7009- 7018.

40. Kristensen, T., R. A. Wetsel, and B. F. Tack. 1985. Structure of the human complement control protein factor H., 44 ed. 1531.

41. Ferreira, V. P., M. K. Pangburn, and C. Cortes. 2010. Complement control protein factor H: the good, the bad, and the inadequate. Mol. Immunol. 47: 2187-2197.

42. Alsenz, J., J. D. Lambris, T. F. Schulz, and M. P. Dierich. 1984. Localization of the complement component C3b binding site and the cofactor activity for factor I in the 38 kDa tryptic fragment of factor H. Biochem. J. 224: 389-398.

43. Gordon, D. L., R. M. Kaufman, T. K. Blackmore, J. Kwong, and D. M. Lublin. 1995. Identification of complement regulatory domains in human factor H. J. Immunol. 155: 348-356.

44. Kuhn, S., and P. F. Zipfel. 1996. Mapping of the domains required for decay acceleration activity of the human factor H-like protein 1 and factor H. Eur. J. Immunol. 26: 2383-2387.

45. Jokiranta, T. S., P. F. Zipfel, J. Hakulinen, S. Kuhn, M. K. Pangburn, J. D. Tamerius, and S. Meri. 1996. Analysis of the recognition mechanism of the alternative pathway of complement by monoclonal anti-factor H antibodies: evidence for multiple interactions between H and surface bound C3b. FEBS Lett. 393: 297-302.

46. Jokiranta, T. S., J. Hellwage, V. Koistinen, P. F. Zipfel, and S. Meri. 2000. Each of the three binding sites on factor H interacts with a distinct site on C3b. J. Biol. Chem. 275: 27657-27662. 147

47. Schmidt, C. Q., A. P. Herbert, D. Kavanagh, C. Gandy, C. J. Fenton, B. S. Blaum, M. Lyon, D. Uhrin, and P. N. Barlow. 2008. A new map of glycosaminoglycan and C3b binding sites on factor H. J. Immunol. 181: 2610-2619.

48. Sharma, A. K., and M. K. Pangburn. 1996. Identification of three physically and functionally distinct binding sites for C3b in human complement factor H by deletion mutagenesis. Proc. Natl. Acad. Sci. USA 93: 10996-11001.

49. Blackmore, T. K., and D. L. Gordon. 1996. SCR 7 is a major heparin/sialic acid binding site of complement factor H., 33 ed. 15.

50. Blackmore, T. K., J. Hellwage, T. A. Sadlon, N. Higgs, P. F. Zipfel, H. M. Ward, and D. L. Gordon. 1998. Identification of the second heparin-binding domain in human complement factor H. J. Immunol. 160: 3342-3348.

51. Ormsby, R. J., T. S. Jokiranta, T. G. Duthy, K. M. Griggs, T. A. Sadlon, E. Giannakis, and D. L. Gordon. 2006. Localization of the third heparin-binding site in the human complement regulator factor H. Mol. Immunol. 43: 1624- 1632.

52. Pangburn, M. K., M. A. L. Atkinson, and S. Meri. 1991. Localization of the heparin-binding site on complement factor H. J. Biol. Chem. 266: 16847- 16853.

53. Prodinger, W. M., J. Hellwage, M. Spruth, M. P. Dierich, and P. F. Zipfel. 1998. The C-terminus of factor H: monoclonal antibodies inhibit heparin binding and identify epitopes common to factor H and factor H-related proteins. Biochem. J. 331 ( Pt 1): 41-47.

54. Ferreira, V. P., A. P. Herbert, H. G. Hocking, P. N. Barlow, and M. K. Pangburn. 2006. Critical role of the C-terminal domains of factor H in regulating complement activation at cell surfaces. J. Immunol. 177: 6308-6316.

55. Ferreira, V. P., and M. K. Pangburn. 2007. Factor H mediated cell surface protection from complement is critical for the survival of PNH erythrocytes. Blood 110: 2190-2192.

56. Jokiranta, T. S., Z. Z. Cheng, H. Seeberger, M. Jozsi, S. Heinen, M. Noris, G. Remuzzi, R. Ormsby, D. L. Gordon, S. Meri, J. Hellwage, and P. F. Zipfel. 2005. Binding of complement factor H to endothelial cells is mediated by the carboxy-terminal glycosaminoglycan binding site. Am. J. Pathol. 167: 1173- 1181.

57. Jozsi, M., M. Oppermann, J. D. Lambris, and P. F. Zipfel. 2007. The C-terminus of complement factor H is essential for host cell protection. Mol. Immunol. 44: 2697-2706. 148

58. Pangburn, M. K. 2002. Cutting edge: Localization of the host recognition functions of complement factor H at the carboxyl-terminal: implications for hemolytic uremic syndrome. J. Immunol. 169: 4702-4706.

59. Pickering, M. C., E. G. de Jorge, R. Martinez-Barricarte, S. Recalde, A. Garcia- Layana, K. L. Rose, J. Moss, M. J. Walport, H. T. Cook, C. de, Sr., and M. Botto. 2007. Spontaneous hemolytic uremic syndrome triggered by complement factor H lacking surface recognition domains. J. Exp. Med. 204: 1249-1256.

60. Ram, S., A. K. Sharma, S. D. Simpson, S. Gulati, D. P. McQuillen, M. K. Pangburn, and P. A. Rice. 1998. A novel sialic acid binding site on factor H mediates serum resistance of sialylated Neisseria gonorrhoeae. J. Exp. Med. 187: 743- 752.

61. de Cordoba, SR., and E. G. de Jorge. 2008. Translational mini-review series on complement factor H: genetics and disease associations of human complement factor H. Clin. Exp. Immunol. 151: 1-13.

62. Holers, V. M. 2008. The spectrum of complement alternative pathway- mediated diseases. Immunol. Rev. 223: 300-316.

63. Kavanagh, D., A. Richards, and J. Atkinson. 2008. Complement regulatory genes and hemolytic uremic syndromes. Annu. Rev. Med. 59: 293-309.

64. Bettinaglio, C. J., P. F. Zipfel, B. Amadei, E. Daina, S. Gamba, C. Skerka, N. Marziliano, G. Remuzzi, and M. Noris. 2001. The molecular basis of familial hemolytic uremic syndrome: mutation analysis of factor H gene reveals a hot spot in short consensus repeat 20. J. Am. Soc. Nephrol. 12: 297-307.

65. Herbert, A. P., D. Uhrin, M. Lyon, M. K. Pangburn, and P. N. Barlow. 2006. Disease-associated sequence variations congregate in a polyanion- recognition patch on human factor H revealed in 3D structure. J. Biol. Chem. 281: 16512-16520.

66. Jokiranta, T. S., V. P. Jaakola, M. J. Lehtinen, M. Parepalo, S. Meri, and A. Goldman. 2006. Structure of complement factor H carboxyl-terminus reveals molecular basis of atypical haemolytic uremic syndrome. EMBO J. 25: 1784- 1794.

67. Perez-Caballero, D., C. Gonzalez-Rubio, M. E. Gallardo, M. Vera, M. Lopez- Trascasa, S. Rodriguez de Cordoba, and P. Sanchez-Corral. 2001. Clustering of missense mutations in the C-terminal region of factor H in atypical hemolytic uremic syndrome. Am. J. Hum. Genet. 68: 478-484.

68. Perkins, S. J., and T. H. Goodship. 2002. Molecular modelling of the C-terminal domains of factor H of human complement: a correlation between haemolytic 149

uraemic syndrome and a predicted heparin binding site. J. Mol. Biol. 316: 217-224.

69. Richards, A., M. R. H. Buddles, R. L. Donne, B. S. Kaplan, E. Kirk, M. C. Venning, C. L. Tielemans, J. A. Goodship, and T. H. Goodship. 2001. Factor H mutations in hemolytic uremic syndrome cluster in exons 18-20, a domain important for host cell recognition. Am. J. Hum. Genet. 68: 485-490.

70. Nicholson-Weller, A., J. P. March, S. I. Rosenfeld, and K. F. Austen. 1983. Affected erythrocytes of patients with paroxysmal nocturnal hemoglobinuria are deficient in the complement regulatory protein, decay accelerating factor. Proc. Natl. Acad. Sci. USA 80: 5066-5070.

71. Pangburn, M. K., R. D. Schreiber, and H. J. Muller-Eberhard. 1983. Deficiency of an erythrocyte membrane protein with complement regulatory activity in paroxysmal nocturnal hemoglobinuria. Proc. Natl. Acad. Sci. USA 80: 5430- 5434.

72. Seya, T., M. Okada, M. Matsumoto, K. S. Hong, T. Kinoshita, and J. P. Atkinson. 1991. Preferential inactivation of the C5 convertase of the alternative complement pathway by factor I and membrane cofactor protein (MCP). Mol. Immunol. 28: 1137-1147.

73. Pillemer, L., L. Blum, I. H. Lepow, O. A. Ross, E. W. Todd, and A. C. Wardlaw. 1954. The properdin system and immunity. I. Demonstration and isolation of a new serum protein, properdin, and its role in immune phenomena. Science 120: 279-285.

74. Gotze, O., and H. J. Muller-Eberhard. 1971. The C3-activator system: an alternative pathway of complement activation. J. Exp. Med. 134: 90s-108s.

75. Gotze, O., and H. J. Muller-Eberhard. 1974. The role of properdin in the alternative pathway of complement activation. J. Exp. Med. 139: 44-57.

76. Whaley, K. 1980. Biosynthesis of the complement components and the regulatory proteins of the alternative complement pathway by human peripheral blood monocytes. J. Exp. Med. 151: 501-516.

77. Wirthmueller, U., B. Dewald, M. Thelen, M. K. Schafer, C. Stover, K. Whaley, J. North, P. Eggleton, K. B. Reid, and W. J. Schwaeble. 1997. Properdin, a positive regulator of complement activation, is released from secondary granules of stimulated peripheral blood neutrophils. J. Immunol. 158: 4444-4451.

78. Schwaeble, W., W. G. Dippold, M. K. Schafer, H. Pohla, D. Jonas, B. Luttig, E. Weihe, H. P. Huemer, M. P. Dierich, and K. B. Reid. 1993. Properdin, a positive

150

regulator of complement activation, is expressed in human T cell lines and peripheral blood T cells. J. Immunol. 151: 2521-2528.

79. Bongrazio, M., A. R. Pries, and A. Zakrzewicz. 2003. The endothelium as physiological source of properdin: role of wall shear stress. Mol. Immunol. 39: 669-675.

80. Reis, E. S., J. A. Barbuto, and L. Isaac. 2006. Human monocyte-derived dendritic cells are a source of several complement proteins. Inflamm. Res. 55: 179-184.

81. Schwaeble, W. J., and K. B. Reid. 1999. Does properdin crosslink the cellular and the humoral immune response? Immunol. Today 20: 17-21.

82. Pattrick, M., J. Luckett, L. Yue, and C. Stover. 2009. Dual role of complement in adipose tissue. Mol. Immunol. 46: 755-760.

83. Fijen, C. A., B. R. van den, M. Schipper, M. Mannens, M. Schlesinger, F. G. Nordin, J. Dankert, M. R. Daha, A. G. Sjoholm, L. Truedsson, and E. J. Kuijper. 1999. Properdin deficiency: molecular basis and disease association. Mol. Immunol. 36: 863-867.

84. Nolan, K. F., and K. B. Reid. 1993. Properdin. Methods Enzymol. 223: 35-46.

85. Xu, W., S. P. Berger, L. A. Trouw, H. C. de Boer, N. Schlagwein, C. Mutsaers, M. R. Daha, and K. C. van. 2008. Properdin binds to late apoptotic and necrotic cells independently of c3b and regulates alternative pathway complement activation. J. Immunol. 180: 7613-7621.

86. Pangburn, M. K. 1989. Analysis of the natural polymeric forms of human properdin and their functions in complement activation. J. Immunol. 142: 202-207.

87. Smith, C. A., M. K. Pangburn, C.-W. Vogel, and H. J. Muller-Eberhard. 1984. Molecular architecture of human properdin, a positive regulator of the alternative pathway of complement. J. Biol. Chem. 259: 4582-4588.

88. Nolan, K. F., and K. B. Reid. 1990. Complete primary structure of human properdin: a positive regulator of the alternative pathway of the serum complement system. Biochem. Soc. Trans. 18: 1161-1162.

89. Sun, Z., K. B. Reid, and S. J. Perkins. 2004. The dimeric and trimeric solution structures of the multidomain complement protein properdin by X-ray scattering, analytical ultracentrifugation and constrained modelling. J. Mol. Biol. 343: 1327-1343.

151

90. Higgins, J. M., H. Wiedemann, R. Timpl, and K. B. Reid. 1995. Characterization of mutant forms of recombinant human properdin lacking single thrombospondin type I repeats. Identification of modules important for function. J. Immunol. 155: 5777-5785.

91. Perdikoulis, M. V., U. Kishore, and K. B. Reid. 2001. Expression and characterisation of the thrombospondin type I repeats of human properdin. Biochim. Biophys. Acta 1548: 265-277.

92. Farries, T. C., P. J. Lachmann, and R. A. Harrison. 1988. Analysis of the interactions between properdin, the third component of complement (C3), and its physiological activation products. Biochem. J. 252: 47-54.

93. Pangburn, M. K., and H. J. Muller-Eberhard. 1980. Relation of a putative thioester bond in C3 to activation of the alternative pathway and the binding of C3b to biological targets of complement. J. Exp. Med. 152: 1102-1114.

94. Ferreira, V. P., C. Cortes, and M. K. Pangburn. 2010. Native polymeric forms of properdin selectively bind to targets and promote activation of the alternative pathway of complement. Immunobiology 215: 932-940.

95. Spitzer, D., L. M. Mitchell, J. P. Atkinson, and D. E. Hourcade. 2007. Properdin can initiate complement activation by binding specific target surfaces and providing a platform for de novo convertase assembly. J. Immunol. 179: 2600-2608.

96. Stover, C. M., J. C. Luckett, B. Echtenacher, A. Dupont, S. E. Figgitt, J. Brown, D. N. Mannel, and W. J. Schwaeble. 2008. Properdin plays a protective role in polymicrobial septic peritonitis. J. Immunol. 180: 3313-3318.

97. Kemper, C., L. M. Mitchell, L. Zhang, and D. E. Hourcade. 2008. The complement protein properdin binds apoptotic T cells and promotes complement activation and phagocytosis. Proc. Natl. Acad. Sci. U. S. A 105: 9023-9028.

98. Gaarkeuken, H., M. A. Siezenga, K. Zuidwijk, K. C. van, T. J. Rabelink, M. R. Daha, and S. P. Berger. 2008. Complement activation by tubular cells is mediated by properdin binding. Am. J. Physiol Renal Physiol 295: F1397- F1403.

99. Camous, L., L. Roumenina, S. Bigot, S. Brachemi, V. Fremeaux-Bacchi, P. Lesavre, and L. Halbwachs-Mecarelli. 2011. Complement alternative pathway acts as a positive feedback amplification of neutrophil activation. Blood 117: 1340-1349.

152

100. Happonen, K. E., T. Saxne, A. Aspberg, M. Morgelin, D. Heinegard, and A. M. Blom. 2010. Regulation of complement by cartilage oligomeric matrix protein allows for a novel molecular diagnostic principle in rheumatoid arthritis. Arthritis Rheum. 62: 3574-3583.

101. Hourcade, D. E. 2006. The role of properdin in the assembly of the alternative pathway C3 convertases of complement. J. Biol. Chem. 281: 2128-2132.

102. Vuagnat, B. B., J. Mach, and J. M. Le Doussal. 2000. Activation of the alternative pathway of human complement by autologous cells expressing transmembrane recombinant properdin. Mol. Immunol. 37: 467-478.

103. Yu, H., E. M. Munoz, R. E. Edens, and R. J. Linhardt. 2005. Kinetic studies on the interactions of heparin and complement proteins using surface plasmon resonance. Biochim. Biophys. Acta 1726: 168-176.

104. Zaferani, A., R. R. Vives, P. P. van der, J. J. Hakvoort, G. J. Navis, G. H. van, M. R. Daha, H. Lortat-Jacob, M. A. Seelen, and B. J. van den. 2011. Identification of tubular heparan sulfate as a docking platform for the alternative complement component properdin in proteinuric renal disease. J. Biol. Chem. 286: 5359- 5367.

105. Holt, G. D., M. K. Pangburn, and V. Ginsgurg. 1990. Properdin binds to sulfatide [Gal(3-SO4)beta1-1Cer] and has a sequence homology with other proteins that bind sulfated glycoconjugates. J. Biol. Chem. 265: 2852-2855.

106. Kimura, Y., T. Miwa, L. Zhou, and W. C. Song. 2008. Activator-specific requirement of properdin in the initiation and amplification of the alternative pathway complement. Blood 111: 732-740.

107. Farries, T. C., J. T. Finch, P. J. Lachmann, and R. A. Harrison. 1987. Resolution and analysis of 'native' and 'activated' properdin. Biochem. J. 243: 507-517.

108. Agarwal, S., V. P. Ferreira, C. Cortes, M. K. Pangburn, P. A. Rice, and S. Ram. 2010. An evaluation of the role of properdin in alternative pathway activation on Neisseria meningitidis and Neisseria gonorrhoeae. J. Immunol. 185: 507-516.

109. Cortes, C., V. P. Ferreira, and M. K. Pangburn. 2011. Native properdin binds to Chlamydia pneumoniae and promotes complement activation. Infect. Immun. 79: 724-731.

110. Saggu, G., C. Cortes, H. N. Emch, G. Ramirez, R. G. Worth, and V. P. Ferreira. 2013. Identification of a novel mode of complement activation on stimulated platelets mediated by properdin and C3(H2O). J. Immunol. 190: 6457-6467.

153

111. Ruef, J., P. Kuehnl, T. Meinertz, and M. Merten. 2008. The complement factor properdin induces formation of platelet-leukocyte aggregates via leukocyte activation. Platelets. 19: 359-364.

112. Thon, J. N., and J. E. Italiano. 2012. Platelets: production, morphology and ultrastructure. Handb. Exp. Pharmacol. 3-22.

113. George, J. N. 2000. Platelets. Lancet 355: 1531-1539.

114. Semple, J. W., J. E. Italiano, Jr., and J. Freedman. 2011. Platelets and the immune continuum. Nat. Rev. Immunol. 11: 264-274.

115. Gnatenko, D. V., J. J. Dunn, S. R. McCorkle, D. Weissmann, P. L. Perrotta, and W. F. Bahou. 2003. Transcript profiling of human platelets using microarray and serial analysis of gene expression. Blood 101: 2285-2293.

116. Booyse, F. M., T. P. Hoveke, and M. E. Rafelson, Jr. 1968. Studies on human platelets. II. Protein synthetic activity of various platelet populations. Biochim. Biophys. Acta 157: 660-663.

117. Nachmias, V. T. 1980. Cytoskeleton of human platelets at rest and after spreading. J. Cell Biol. 86: 795-802.

118. Hartwig, J. H. 1992. Mechanisms of actin rearrangements mediating platelet activation. J. Cell Biol. 118: 1421-1442.

119. Del Conde, I., M. A. Cruz, H. Zhang, J. A. Lopez, and V. Afshar-Kharghan. 2005. Platelet activation leads to activation and propagation of the complement system. J. Exp. Med. 201: 871-879.

120. Peerschke, E. I., W. Yin, S. E. Grigg, and B. Ghebrehiwet. 2006. Blood platelets activate the classical pathway of human complement. J. Thromb. Haemost. 4: 2035-2042.

121. Gemmell, C. H., S. M. Ramirez, E. L. Yeo, and M. V. Sefton. 1995. Platelet activation in whole blood by artificial surfaces: identification of platelet- derived microparticles and activated platelet binding to leukocytes as material-induced activation events. J. Lab Clin. Med. 125: 276-287.

122. Kroll, M. H., J. D. Hellums, L. V. McIntire, A. I. Schafer, and J. L. Moake. 1996. Platelets and shear stress. Blood 88: 1525-1541.

123. Kahn, M. L., M. Nakanishi-Matsui, M. J. Shapiro, H. Ishihara, and S. R. Coughlin. 1999. Protease-activated receptors 1 and 4 mediate activation of human platelets by thrombin. J. Clin. Invest 103: 879-887.

154

124. Brass, L. F. 2003. Thrombin and platelet activation. Chest 124: 18S-25S.

125. Shankar, R., C. A. de la Motte, E. J. Poptic, and P. E. DiCorleto. 1994. Thrombin receptor-activating peptides differentially stimulate platelet-derived growth factor production, monocytic cell adhesion, and E-selectin expression in human umbilical vein endothelial cells. J. Biol. Chem. 269: 13936-13941.

126. Neufeld, E. J., and P. W. Majerus. 1983. Arachidonate release and phosphatidic acid turnover in stimulated human platelets. J. Biol. Chem. 258: 2461-2467.

127. Hammarstrom, S., M. Hamberg, B. Samuelsson, E. A. Duell, M. Stawiski, and J. J. Voorhees. 1975. Increased concentrations of nonesterified arachidonic acid, 12L-hydroxy-5,8,10,14-eicosatetraenoic acid, prostaglandin E2, and prostaglandin F2alpha in epidermis of psoriasis. Proc. Natl. Acad. Sci. U. S. A 72: 5130-5134.

128. Chilton, F. H., J. S. Hadley, and R. C. Murphy. 1987. Incorporation of arachidonic acid into 1-acyl-2-lyso-sn-glycero-3-phosphocholine of the human neutrophil. Biochim. Biophys. Acta 917: 48-56.

129. Bogatcheva, N. V., M. G. Sergeeva, S. M. Dudek, and A. D. Verin. 2005. Arachidonic acid cascade in endothelial pathobiology. Microvasc. Res. 69: 107-127.

130. Sautebin, L., D. Caruso, G. Galli, and R. Paoletti. 1983. Preferential utilization of endogenous arachidonate by cyclo-oxygenase in incubations of human platelets. FEBS Lett. 157: 173-178.

131. Brass, L. F., C. C. Shaller, and E. J. Belmonte. 1987. Inositol 1,4,5-triphosphate- induced granule secretion in platelets. Evidence that the activation of phospholipase C mediated by platelet thromboxane receptors involves a guanine nucleotide binding protein-dependent mechanism distinct from that of thrombin. J. Clin. Invest 79: 1269-1275.

132. Shenker, A., P. Goldsmith, C. G. Unson, and A. M. Spiegel. 1991. The G protein coupled to the thromboxane A2 receptor in human platelets is a member of the novel Gq family. J. Biol. Chem. 266: 9309-9313.

133. MacIntyre, D. E., and J. L. Gordon. 1975. Calcium-dependent stimulation of platelet aggregation by PGE. Nature 258: 337-339.

134. Vezza, R., R. Roberti, G. G. Nenci, and P. Gresele. 1993. Prostaglandin E2 potentiates platelet aggregation by priming protein kinase C. Blood 82: 2704- 2713.

155

135. Patrono, C., and G. A. FitzGerald. 1997. Isoprostanes: potential markers of oxidant stress in atherothrombotic disease. Arterioscler. Thromb. Vasc. Biol. 17: 2309-2315.

136. Hollopeter, G., H. M. Jantzen, D. Vincent, G. Li, L. England, V. Ramakrishnan, R. B. Yang, P. Nurden, A. Nurden, D. Julius, and P. B. Conley. 2001. Identification of the platelet ADP receptor targeted by antithrombotic drugs. Nature 409: 202-207.

137. Oury, C., E. Toth-Zsamboki, J. Vermylen, and M. F. Hoylaerts. 2002. P2X(1)- mediated activation of extracellular signal-regulated kinase 2 contributes to platelet secretion and aggregation induced by collagen. Blood 100: 2499- 2505.

138. White, G. C., and R. Rompietti. 2007. Platelet secretion: indiscriminately spewed forth or highly orchestrated? J. Thromb. Haemost. 5: 2006-2008.

139. Quinn, M. 2010. Platelet Physiology. In Platelet Function: Assessment, Diagnosis, and Treatment. M. Quinn and D. Fitzgerald, eds. Humana Press, Totowa, NJ. 3-20.

140. Trip, M. D., V. M. Cats, F. J. van Capelle, and J. Vreeken. 1990. Platelet hyperreactivity and prognosis in survivors of myocardial infarction. N. Engl. J. Med. 322: 1549-1554.

141. van Zanten, G. H., G. S. de, P. J. Slootweg, H. F. Heijnen, T. M. Connolly, P. G. de Groot, and J. J. Sixma. 1994. Increased platelet deposition on atherosclerotic coronary arteries. J. Clin. Invest 93: 615-632.

142. Furman, M. I., S. E. Benoit, M. R. Barnard, C. R. Valeri, M. L. Borbone, R. C. Becker, H. B. Hechtman, and A. D. Michelson. 1998. Increased platelet reactivity and circulating monocyte-platelet aggregates in patients with stable coronary artery disease. J. Am. Coll. Cardiol. 31: 352-358.

143. Peerschke, E. I., W. Yin, D. R. Alpert, R. A. Roubey, J. E. Salmon, and B. Ghebrehiwet. 2009. Serum complement activation on heterologous platelets is associated with arterial thrombosis in patients with systemic lupus erythematosus and antiphospholipid antibodies. Lupus 18: 530-538.

144. Peerschke, E. I., W. Yin, and B. Ghebrehiwet. 2010. Complement activation on platelets: implications for vascular inflammation and thrombosis. Mol. Immunol. 47: 2170-2175.

145. Hamad, O. A., K. N. Ekdahl, P. H. Nilsson, J. Andersson, P. Magotti, J. D. Lambris, and B. Nilsson. 2008. Complement activation triggered by

156

chondroitin sulfate released by thrombin receptor-activated platelets. J. Thromb. Haemost. 6: 1413-1421.

146. Hamad, O. A., P. H. Nilsson, D. Wouters, J. D. Lambris, K. N. Ekdahl, and B. Nilsson. 2010. Complement component C3 binds to activated normal platelets without preceding proteolytic activation and promotes binding to complement receptor 1. J. Immunol. 184: 2686-2692.

147. Ricklin, D., G. Hajishengallis, K. Yang, and J. D. Lambris. 2010. Complement: a key system for immune surveillance and homeostasis. Nat. Immunol. 11: 785- 797.

148. Wiedmer, T., C. T. Esmon, and P. J. Sims. 1986. Complement proteins C5b-9 stimulate procoagulant activity through platelet prothrombinase. Blood 68: 875-880.

149. Polley, M. J., and R. L. Nachman. 1983. Human platelet activation by C3a and C3a des-arg. J. Exp. Med. 158: 603-615.

150. Nicholson-Weller, A., J. P. March, C. E. Rosen, D. B. Spicer, and K. F. Austen. 1985. Surface membrane expression by human leukocytes and platelets of decay-accelerating factor, a regulatory protein of the complement system. Blood 65: 1237-1244.

151. Morgan, B. P. 1992. Isolation and characterization of the complement- inhibiting protein CD59 antigen from platelet membranes. Biochem. J. 282 ( Pt 2): 409-413.

152. Kerr, H., and A. Richards. 2012. Complement-mediated injury and protection of endothelium: lessons from atypical haemolytic uraemic syndrome. Immunobiology 217: 195-203.

153. Stahl, A. L., F. Vaziri-Sani, S. Heinen, A. C. Kristoffersson, K. H. Gydell, R. Raafat, A. Gutierrez, O. Beringer, P. F. Zipfel, and D. Karpman. 2008. Factor H dysfunction in patients with atypical hemolytic uremic syndrome contributes to complement deposition on platelets and their activation. Blood 111: 5307- 5315.

154. Devine, D. V., R. S. Siegel, and W. F. Rosse. 1987. Interactions of the platelets in paroxysmal nocturnal hemoglobinuria with complement. Relationship to defects in the regulation of complement and to platelet survival in vivo. J. Clin. Invest 79: 131-137.

155. Becker, R. C., R. P. Tracy, E. G. Bovill, K. G. Mann, and K. Ault. 1994. The clinical use of flow cytometry for assessing platelet activation in acute

157

coronary syndromes. TIMI-III Thrombosis and Anticoagulation Group. Coron. Artery Dis. 5: 339-345.

156. Fitzgerald, D. J., L. Roy, F. Catella, and G. A. FitzGerald. 1986. Platelet activation in unstable coronary disease. N. Engl. J. Med. 315: 983-989.

157. Wagner, D. D. 2005. New links between inflammation and thrombosis. Arterioscler. Thromb. Vasc. Biol. 25: 1321-1324.

158. Kirchhofer, D., M. A. Riederer, and H. R. Baumgartner. 1997. Specific accumulation of circulating monocytes and polymorphonuclear leukocytes on platelet thrombi in a vascular injury model. Blood 89: 1270-1278.

159. Broijersen, A., A. Hamsten, M. Eriksson, B. Angelin, and P. Hjemdahl. 1998. Platelet activity in vivo in hyperlipoproteinemia--importance of combined hyperlipidemia. Thromb. Haemost. 79: 268-275.

160. Broijersen, A., F. Karpe, A. Hamsten, A. H. Goodall, and P. Hjemdahl. 1998. Alimentary lipemia enhances the membrane expression of platelet P-selectin without affecting other markers of platelet activation. Atherosclerosis 137: 107-113.

161. Furman, M. I., M. R. Barnard, L. A. Krueger, M. L. Fox, E. A. Shilale, D. M. Lessard, P. Marchese, A. L. Frelinger, III, R. J. Goldberg, and A. D. Michelson. 2001. Circulating monocyte-platelet aggregates are an early marker of acute myocardial infarction. J. Am. Coll. Cardiol. 38: 1002-1006.

162. Hayward, R., B. Campbell, Y. K. Shin, R. Scalia, and A. M. Lefer. 1999. Recombinant soluble P-selectin glycoprotein ligand-1 protects against myocardial ischemic reperfusion injury in cats. Cardiovasc. Res. 41: 65-76.

163. da Costa Martins, P. A., J. M. van Gils, A. Mol, P. L. Hordijk, and J. J. Zwaginga. 2006. Platelet binding to monocytes increases the adhesive properties of monocytes by up-regulating the expression and functionality of beta1 and beta2 integrins. J. Leukoc. Biol. 79: 499-507.

164. Freedman, J. E., and J. Loscalzo. 2002. Platelet-monocyte aggregates: bridging thrombosis and inflammation. Circulation 105: 2130-2132.

165. Rinder, H. M., J. L. Bonan, C. S. Rinder, K. A. Ault, and B. R. Smith. 1991. Dynamics of leukocyte-platelet adhesion in whole blood. Blood 78: 1730- 1737.

166. Diacovo, T. G., S. J. Roth, J. M. Buccola, D. F. Bainton, and T. A. Springer. 1996. Neutrophil rolling, arrest, and transmigration across activated, surface-

158

adherent platelets via sequential action of P-selectin and the beta 2-integrin CD11b/CD18. Blood 88: 146-157.

167. Evangelista, V., S. Manarini, S. Rotondo, N. Martelli, R. Polischuk, J. L. McGregor, G. G. de, and C. Cerletti. 1996. Platelet/polymorphonuclear leukocyte interaction in dynamic conditions: evidence of adhesion cascade and cross talk between P-selectin and the beta 2 integrin CD11b/CD18. Blood 88: 4183-4194.

168. Osterud, B. 1995. Cellular interactions in tissue factor expression by blood monocytes. Blood Coagul. Fibrinolysis 6 Suppl 1: S20-S25.

169. Celi, A., G. Pellegrini, R. Lorenzet, B. A. De, N. Ready, B. C. Furie, and B. Furie. 1994. P-selectin induces the expression of tissue factor on monocytes. Proc. Natl. Acad. Sci. U. S. A 91: 8767-8771.

170. Weyrich, A. S., M. R. Elstad, R. P. McEver, T. M. McIntyre, K. L. Moore, J. H. Morrissey, S. M. Prescott, and G. A. Zimmerman. 1996. Activated platelets signal chemokine synthesis by human monocytes. J. Clin. Invest 97: 1525- 1534.

171. Ruf, A., and H. Patscheke. 1995. Platelet-induced neutrophil activation: platelet-expressed fibrinogen induces the oxidative burst in neutrophils by an interaction with CD11C/CD18. Br. J. Haematol. 90: 791-796.

172. Werr, J., E. E. Eriksson, P. Hedqvist, and L. Lindbom. 2000. Engagement of beta2 integrins induces surface expression of beta1 integrin receptors in human neutrophils. J. Leukoc. Biol. 68: 553-560.

173. Cerletti, C., C. Tamburrelli, B. Izzi, F. Gianfagna, and G. G. de. 2012. Platelet- leukocyte interactions in thrombosis. Thromb. Res. 129: 263-266.

174. Haselmayer, P., L. Grosse-Hovest, L. P. von, H. Schild, and M. P. Radsak. 2007. TREM-1 ligand expression on platelets enhances neutrophil activation. Blood 110: 1029-1035.

175. Stahl, A. L., L. Sartz, and D. Karpman. 2011. Complement activation on platelet-leukocyte complexes and microparticles in enterohemorrhagic Escherichia coli-induced hemolytic uremic syndrome. Blood 117: 5503-5513.

176. Bournazos, S., J. Rennie, S. P. Hart, and I. Dransfield. 2008. Choice of anticoagulant critically affects measurement of circulating platelet-leukocyte complexes. Arterioscler. Thromb. Vasc. Biol. 28: e2-e3.

177. Mollnes, T. E., P. Garred, and G. Bergseth. 1988. Effect of time, temperature and anticoagulants on in vitro complement activation: consequences for 159

collection and preservation of samples to be examined for complement activation. Clin. Exp. Immunol. 73: 484-488.

178. Fearon, D. T. 1984. Cellular receptors for fragments of the third component of complement. Immunol. Today 5: 105-110.

179. Ando, B., T. Wiedmer, and P. J. Sims. 1989. The secretory release reaction initiated by complement proteins C5b-9 occurs without platelet aggregation through glycoprotein IIb-IIIa. Blood 73: 462-467.

180. Rinder, C. S., H. M. Rinder, B. R. Smith, J. C. Fitch, M. J. Smith, J. B. Tracey, L. A. Matis, S. P. Squinto, and S. A. Rollins. 1995. Blockade of C5a and C5b-9 generation inhibits leukocyte and platelet activation during extracorporeal circulation. J. Clin. Invest 96: 1564-1572.

181. Wiedmer, T., and P. J. Sims. 1985. Effect of complement proteins C5b-9 on blood platelets. Evidence for reversible depolarization of membrane potential. J. Biol. Chem. 260: 8014-8019.

182. Fearon, D. T. 1983. The human C3b receptor. Springer Seminars in Immunopathology 6: 159-172.

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

184. OUCHTERLONY, O. 1958. Diffusion-in-gel methods for immunological analysis. Prog. Allergy 5: 1-78.

185. Pangburn, M. K. 1987. A fluorimetric assay for native C3. The hemolytically active form of the third component of human complement. J. Immunol. Methods 102: 7-14.

186. Rawal, N., and M. K. Pangburn. 1998. C5 convertase of the alternative pathway of complement. Kinetic analysis of the free and surface-bound forms of the enzyme. J. Biol. Chem. 273: 16828-16835.

187. Cortes, C., J. A. Ohtola, G. Saggu, and V. P. Ferreira. 2012. Local release of properdin in the cellular microenvironment: role in pattern recognition and amplification of the alternative pathway of complement. Front Immunol. 3: 412.

188. Renner, B., V. P. Ferreira, C. Cortes, R. Goldberg, D. Ljubanovic, M. K. Pangburn, M. C. Pickering, S. Tomlinson, A. Holland-Neidermyer, D. Strassheim, V. M. Holers, and J. M. Thurman. 2011. Binding of factor H to tubular epithelial cells limits interstitial complement activation in ischemic injury. Kidney Int. 80: 165-173. 160

189. Takeda, K., J. M. Thurman, S. Tomlinson, M. Okamoto, Y. Shiraishi, V. P. Ferreira, C. Cortes, M. K. Pangburn, V. M. Holers, and E. W. Gelfand. 2012. The critical role of complement alternative pathway regulator factor H in allergen-induced airway hyperresponsiveness and inflammation. J. Immunol. 188: 661-667.

190. Banda, N. K., G. Mehta, V. P. Ferreira, C. Cortes, M. C. Pickering, M. K. Pangburn, W. P. Arend, and V. M. Holers. 2013. Essential role of surface- bound complement factor H in controlling immune complex-induced arthritis. J. Immunol. 190: 3560-3569.

191. Sahu, A., B. K. Kay, and J. D. Lambris. 1996. Inhibition of human complement by a C3-binding peptide isolated from a phage-displayed random peptide library. J. Immunol. 157: 884-891.

192. Ricklin, D., and J. D. Lambris. 2008. Compstatin: a complement inhibitor on its way to clinical application. Adv. Exp. Med. Biol. 632: 273-292.

193. Taylor, M. L., N. L. Misso, G. A. Stewart, and P. J. Thompson. 1995. Differential Expression of Platelet Activation Markers CD62P and CD63 Following Stimulation with PAF, Arachidonic Acid and Collagen. Platelets. 6: 394-401.

194. Schwendinger, M. G., M. Spruth, J. Schoch, M. P. Dierich, and W. M. Prodinger. 1997. A novel mechanism of alternative pathway complement activation accounts for the deposition of C3 fragments on CR2-expressing homologous cells. J. Immunol. 158: 5455-5463.

195. Nicholson-Weller, A., D. B. Spicer, and K. F. Austen. 1985. Deficiency of the complement regulatory protein, "decay-accelerating factor", on membranes of granulocytes, monocytes and platelets in paroxysmal nocturnal hemoglobinuria. N. Engl. J. Med. 312: 1091-1096.

196. Vaziri-Sani, F., J. Hellwage, P. F. Zipfel, A. G. Sjoholm, R. Iancu, and D. Karpman. 2005. Factor H binds to washed human platelets. J. Thromb. Haemost. 3: 154-162.

197. Mnjoyan, Z., J. Li, and V. Afshar-Kharghan. 2008. Factor H binds to platelet integrin alphaIIbbeta3. Platelets. 19: 512-519.

198. Licht, C., F. G. Pluthero, L. Li, H. Christensen, S. Habbig, B. Hoppe, D. F. Geary, P. F. Zipfel, and W. H. Kahr. 2009. Platelet-associated complement factor H in healthy persons and patients with atypical HUS. Blood 114: 4538-4545.

199. Kavanagh, D., T. H. Goodship, and A. Richards. 2006. Atypical haemolytic uraemic syndrome. Br. Med. Bull. 77-78: 5-22.

161

200. Barnard, M. R., L. A. Kreuger, A. L. Frelinger, III, M. I. Furman, and A. D. Michelson. 2003. Whole blood analysis of leukocyte-platelet aggregates. In Current Protocols in Cytometry.

201. Chang, J. Y. 1983. The functional domain of hirudin, a thrombin-specific inhibitor. FEBS Lett. 164: 307-313.

202. Mollnes, T. E., O. L. Brekke, M. Fung, H. Fure, D. Christiansen, G. Bergseth, V. Videm, K. T. Lappegard, J. Kohl, and J. D. Lambris. 2002. Essential role of the C5a receptor in E coli-induced oxidative burst and phagocytosis revealed by a novel lepirudin-based human whole blood model of inflammation. Blood 100: 1869-1877.

203. Wetsel, R. A. 1995. Structure, function and cellular expression of complement anaphylatoxin receptors. Curr. Opin. Immunol. 7: 48-53.

204. Molad, Y., K. A. Haines, D. C. Anderson, J. P. Buyon, and B. N. Cronstein. 1994. Immunocomplexes stimulate different signalling events to chemoattractants in the neutrophil and regulate L-selectin and beta 2-integrin expression differently. Biochem. J. 299 ( Pt 3): 881-887.

205. Jagels, M. A., P. J. Daffern, and T. e. Hugli. 2000. C3a and C5a enhance granulocyte adhesion to endothelial and epithelial cell monolayers: epithelial and endothelial priming is required for C3a-induced eosinophil adhesion. Immunopharmacology 46: 209-222.

206. Michelson, A. D., M. R. Barnard, L. A. Krueger, A. L. Frelinger, III, and M. I. Furman. 2002. Flow Cytometry. In Platelets. A. D. Michelson, ed. Academic Press/Elsevier Science, New York. 297-315.

207. Rocca, B., and C. Patrono. 2010. Prostanoid generation in platelet function: assessment and clinical relevance. In Platelet Function: Assessment, Diagnosis, and Treatment. M. Quinn and D. Fitzgerald, eds. Humana Press, Totowa, NJ. 267-282.

208. Maclouf, J., G. Folco, and C. Patrono. 1998. Eicosanoids and iso-eicosanoids: constitutive, inducible and transcellular biosynthesis in vascular disease. Thromb. Haemost. 79: 691-705.

209. Polley, M. J., and R. L. Nachman. 1979. Human complement in thrombin- mediated platelet function: uptake of the C5b-9 complex. J. Exp. Med. 150: 633-645.

210. Polley, M. J., R. L. Nachman, and B. B. Weksler. 1981. Human complement in the arachidonic acid transformation pathway in platelets. J. Exp. Med. 153: 257-268. 162

211. Majek, P., Z. Reicheltova, J. Stikarova, J. Suttnar, A. Sobotkova, and J. E. Dyr. 2010. Proteome changes in platelets activated by arachidonic acid, collagen, and thrombin. Proteome. Sci. 8: 1-13.

212. Okayama, M., K. Oguri, Y. Fujiwara, H. Nakanishi, H. Yonekura, T. Kondo, and N. Ui. 1986. Purification and characterization of human platelet proteoglycan. Biochem. J. 233: 73-81.

213. Ward, J. V., and M. A. Packham. 1979. Characterization of the sulfated glycosaminoglycan on the surface and in the storage granules of rabbit platelets. Biochim. Biophys. Acta 583: 196-207.

214. Heinen, S., F. G. Pluthero, V. F. van Eimeren, S. E. Quaggin, and C. Licht. 2013. Monitoring and modeling treatment of atypical hemolytic uremic syndrome. Mol. Immunol. 54: 84-88.

215. Lachmann, P. J. 2009. The amplification loop of the complement pathways. Adv. Immunol. 104: 115-149.

216. Nilsson, B., and E. K. Nilsson. 2012. The tick-over theory revisited: is C3 a contact-activated protein? Immunobiology 217: 1106-1110.

217. Nunez, D., C. Charriaut-Marlangue, M. Barel, J. Benveniste, and R. Frade. 1987. Activation of human platelets through gp140, the C3d/EBV receptor (CR2). Eur. J. Immunol. 17: 515-520.

218. Botto, M., D. Lissandrini, C. Sorio, and M. J. Walport. 1992. Biosynthesis and secretion of complement component (C3) by activated human polymorphonuclear leukocytes. J. Immunol. 149: 1348-1355.

219. Rhodes, A., S. J. Wort, H. Thomas, P. Collinson, and E. D. Bennett. 2006. Plasma DNA concentration as a predictor of mortality and sepsis in critically ill patients. Crit. Care doi:10. 1186/cc4894.

220. Sims, P. J., and T. Wiedmer. 1991. The response of human platelets to activated components of the complement system. Immunol. Today 12: 338- 342.

221. Zwirner, J., O. Gotze, G. Begemann, A. Kapp, K. Kirchhoff, and T. Werfel. 1999. Evaluation of C3a receptor expression on human leucocytes by the use of novel monoclonal antibodies. Immunology 97: 166-172.

222. Ehrengruber, M. U., T. Geiser, and D. A. Deranleau. 1994. Activation of human neutrophils by C3a and C5A. Comparison of the effects on shape changes, chemotaxis, secretion, and respiratory burst. FEBS Lett. 346: 181-184.

163

223. Losse, J., P. F. Zipfel, and M. Jozsi. 2010. Factor H and factor H-related protein 1 bind to human neutrophils via complement receptor 3, mediate attachment to Candida albicans, and enhance neutrophil antimicrobial activity. J. Immunol. 184: 912-921.

224. Smith, R. J., J. Alexander, P. N. Barlow, M. Botto, T. L. Cassavant, H. T. Cook, C. de, Sr., G. S. Hageman, T. S. Jokiranta, W. J. Kimberling, J. D. Lambris, L. D. Lanning, V. Levidiotis, C. Licht, H. U. Lutz, S. Meri, M. C. Pickering, R. J. Quigg, A. L. Rops, D. J. Salant, S. Sethi, J. M. Thurman, H. F. Tully, S. P. Tully, d. van, V, P. D. Walker, R. Wurzner, and P. F. Zipfel. 2007. New approaches to the treatment of dense deposit disease. J. Am. Soc. Nephrol. 18: 2447-2456.

164

Appendix A

Identification of a novel mode of complement activation on stimulated platelets mediated by properdin and C3(H2O).

This article was originally published in The Journal of Immunology. Gurpanna

Saggu, Claudio Cortes, Heather N. Emch, Galia Ramirez, Randall G.Worth, and

Viviana P. Ferreira. 2013. Identification of a novel mode of complement activation on stimulated platelets mediated by properdin and C3(H2O). J. Immunol. 190: 6457-

Link to article on Journal of Immunology website: http://www.jimmunol.org/content/190/12/6457.full.pdf+html

Authors retain the permissions to include the final, published version of the article in a thesis and/or dissertation in print. If required by the degree-conferring institution, an electronic version of the final, published version may be deposited into a thesis repository as long as a link to the article on The Journal of Immunology website is included.

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

Appendix B

Local release of properdin in the cellular microenvironment: role in pattern recognition and amplification of the alternative pathway of complement.

This article was originally published in Frontiers of Immunology., Claudio Cortes, Jennifer Ohtola, Gurpanna Saggu, and Viviana P. Ferreira. 2013. Local release of properdin in the cellular microenvironment: role in pattern recognition and amplification of the alternative pathway of complement. F. Immunol. 3: 412. Copyright Cortes, Ohtola, Saggu, and Ferreira.

Reproduced with permission under the terms of Creative Commons Attribution License, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics, etc.

181

182

183

184

185

186

187

188