Properdin: a tightly regulated critical inflammatory modulator

Summary The complement alternative pathway is a powerful arm of the innate immune system that enhances diverse inflammatory responses in the human host. Key to the effects of the alternative pathway is properdin, a serum glycoprotein that can both initiate and positively regulate alternative pathway activity. Properdin is produced by many different leukocyte subsets and circulates as cyclic oligomers of monomeric subunits. While the formation of non‐physiological aggregates in purified properdin preparations and the presence of potential properdin inhibitors in serum have complicated studies of its function, properdin has, regardless, emerged as a key player in various inflammatory disease models. Here, we review basic properdin biology, emphasizing the major hurdles that have complicated the interpretation of results from properdin‐centered studies. In addition, we elaborate on an emerging role for properdin in thromboinflammation and discuss the potential utility of properdin inhibitors as long‐term therapeutic options to treat diseases marked by increased formation of platelet/granulocyte aggregates. Finally, we describe the interplay between properdin and the alternative pathway negative regulator, Factor H, and how aiming to understand these interactions can provide scientists with the most effective ways to manipulate alternative pathway activation in complex systems.


I N V I T E D R E V I E W
Properdin: a tightly regulated critical inflammatory modulator Adam Z. Blatt | Sabina Pathan | Viviana P. Ferreira

| INTRODUCTION
While the basic principles of complement activation are wellcharacterized, complex interactions between complement and other cells and proteins, especially in the context of disease pathogenesis, have precluded, with few exceptions, successful use of anti-complement therapeutics in the clinic for common diseases.
Highlighting a difficulty in understanding the complete scope of complement activity is the intricate interplay between regulatory mechanisms that govern complement activation. While most human cells and tissues express membrane-bound regulatory proteins to protect themselves from complement attack, soluble regulatory proteins are recruited to inflammatory sites through interactions with both complement proteins and other ligands defined by the unique composition of the local microenvironment. The soluble complement protein properdin is both a regulator and an initiator of the complement alternative pathway that has potent effects on the level of complement activation. Understanding the complex biology of properdin, however, has proven difficult due to the different sources of properdin used in biochemical studies and its intricate self-associations. Despite these difficulties, experimental data have proven properdin to be both a powerful inflammatory modulator that can greatly enhance tissuedamaging inflammation and a critical promoter of microbial clearance in multiple disease models. Here, we discuss basic principles of, and recent advances in understanding, properdin biology, highlighting areas that require future studies. We progress to describe the interplay between properdin and the negative regulator Factor H, the potential This article is part of a series of reviews covering Preformed Mediators of Defense appearing in Volume 274 of Immunological Reviews.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. of properdin inhibitors in treating human disease with an emphasis on the role of properdin in thromboinflammation, and mention potential consequences of properdin inhibition as evidenced in animal models.
Finally, we briefly elaborate on the utility of using novel strategies to promote properdin activity on the surface of microorganisms. To facilitate discussion, we begin with a condensed review of the complement system, which has been extensively reviewed elsewhere. 1-3

| THE COMPLEMENT SYSTEM
The complement system is a group of soluble blood proteins that rapidly activate in a cascade-like manner to orchestrate inflammatory and immune responses in the human host. Complement activity can be initiated by one of three pathways termed the classical, lectin, and alternative pathways, which all converge at the cleavage of the central protein C3 by an enzymatic complex called the C3 convertase.
C3 cleavage releases the small fragment C3a, and the resulting larger fragment, C3b, can bind covalently to nearby hydroxyl-and aminogroups via an exposed thioester bond. C3b bound to, or in close proximity to, the C3 convertase changes the specificity of the complex, allowing for the cleavage of C5 to C5a and C5b. The terminal pathway activates through a series of conformational changes in complement proteins, initiated by the binding of C5b to C6, followed by sequential binding of C7, C8, and C9 that results in the formation of the membrane attack complex (MAC; C5b-9). 1,3 The classical and lectin pathways utilize distinct initiation mechanisms that lead to the generation of the same C3 convertase, denoted C4b2b. Classical pathway activation is primarily a result of antibodyantigen interactions that promote the activation of the C1 complex, while the lectin pathway activates in response to recognition of specific carbohydrate patterns by various lectin molecules. The alternative pathway, whose activation mechanism will be discussed below, is unique among the three complement pathways because it can activate without the requirement for specific antigen recognition. 2 Each complement pathway generates the same set of effector molecules to carry out its functions. C3b and its breakdown products iC3b and C3dg tag surfaces for recognition by complement receptors expressed on human cells (CR1, CR2, CR3, CR4, and CRIg). 1 The anaphylatoxins, C3a and C5a, bind to the G-protein-coupled receptors C3a receptor (C3aR) and C5a receptor 1 (C5aR1; CD88), respectively, to promote inflammation and other diverse functions, including cerebellar development, homing of stem cells to the bone marrow, and tissue fibrosis. 4 The MAC is capable of directly forming pores in membranes, leading to direct microbial killing, and sublytic levels of MAC stimulate a variety of proinflammatory responses in human cells, including cell and inflammasome activation. 5

| The alternative pathway
The alternative pathway continuously activates at a low level in the fluid phase of blood due to spontaneous hydrolysis of the labile thioester bond in C3, in a process known as 'tick-over'. An additional source of C3(H 2 O) may be formed by contact activation on certain cells, such as platelets and artificial surfaces. 6,7 Upon hydrolysis of its thioester bond, C3 undergoes a conformational change to form C3(H 2 O) and gains the ability to bind Factor B. [8][9][10] Factor B in association with C3(H 2 O) is cleaved by Factor D to Ba and Bb, to generate the fluid-phase alternative pathway C3 convertase, C3(H 2 O)Bb.
C3(H 2 O)Bb cleaves additional C3 molecules to C3a and C3b, exposing the labile thioester bond on C3b. 8 The thioester bond is rapidly hydrolyzed by water, and only a small percentage of C3b is able to bind covalently to nearby hydroxyl-and amino-groups on surfaces before hydrolysis. 11 Covalently bound C3b tags surfaces for additional alternative pathway activation. Surface-bound C3b, which is structurally similar to C3(H 2 O), recruits Factor B molecules, which are cleaved by Factor D in an identical manner to that which occurs in the fluid phase. 12 The surface-bound alternative pathway C3 convertase, C3bBb, cleaves more C3 molecules to C3a and C3b, which, if bound covalently to the surface, act as focal points for the assembly of more alternative pathway convertases, effectively amplifying its own F I G U R E 1 Activation of the alternative pathway. Described from upper right to left: The thioester bond in C3 is spontaneously hydrolyzed by water, leading to the formation of C3(H 2 O), which can recruit Factor B (FB). Once bound to C3(H 2 O), FB is cleaved by Factor D (FD) to Bb to form the alternative pathway fluid phase C3 convertase, C3(H 2 O)Bb. The C3 convertase cleaves C3 to C3a and C3b, which can bind covalently to nearby amino-and hydroxyl-groups via its thioester group. C3b covalently bound to a surface recruits FB, which is subsequently cleaved by FD to form the alternative pathway cell-surface C3 convertase, C3bBb. While C3bBb has a half-life of only approximately 90 seconds, properdin (P) stabilizes the convertase to increase its activity 5-to 10-fold activity. 13 Figure 1 summarizes alternative pathway activation. C3b initially deposited by the classical or lectin pathways can also serve as sites for alternative pathway activation; therefore, the alternative pathway is an amplification loop for all complement activity. The alternative pathway may account for approximately 80% of terminal pathway activity even when complement is initially activated by the classical 14 or lectin pathways 15 (Fig. 2).

| Alternative pathway regulation
Its spontaneous nature and its ability to amplify all complement activity make regulation of the alternative pathway a necessity in the human host in order to prevent excessive inflammation and tissue damage. Human cells and tissues are protected from complement attack by various membrane-bound complement regulatory proteins, including CD55, CD59, CD46, and CR1. 1 However, the serum glycoprotein, Factor H (Fig. 3A), which has been reviewed elsewhere, 16 has also proven to be critical to limiting alternative pathway activation on the surface of several cell types, even in the presence of membrane-bound regulators. Factor H is also the primary regulator of the alternative pathway in the fluid phase, preventing complement consumption via uncontrolled alternative pathway activation.
Factor H functions by accelerating the decay of the convertases by promoting the dissociation of Bb from C3(H 2 O) and from C3b, 17 known as dense deposit disease (DDD), characterized by insufficient fluid-phase regulation that causes consumption of C3. 35,36 Contrarily, mutations primarily located in domains 19-20 that limit Factor H-mediated cell-surface protection, but retain fluid-phase regulation, F I G U R E 2 Alternative pathway amplifies all complement activity. C3b originally deposited on a surface by the classical pathway can act as a site for formation of the alternative pathway cell-surface C3 convertase. The alternative pathway deposits more C3b on the surface, which can act as additional sites for alternative pathway C3 convertase formation. Therefore, even minor complement activity initiated by the classical (or lectin) pathway can be quickly and efficiently amplified by the alternative pathway can be used in vitro and in vivo as an efficient tool to (i) evaluate the contribution of Factor H cell-surface protection in inflammatory disease models 31,[45][46][47][48] and (ii) enhance the understanding of the pathogenic mechanisms that lead to aHUS. 24,38,[49][50][51] Alternative pathway convertases are inherently unstable, with a half-life of only approximately 90 seconds. 52 Properdin, the focus of this review, has a well-defined role in which it acts as a positive regulator of complement activity by stabilizing the alternative pathway convertase, thus increasing its activity 5-to 10-fold. 53 Properdin stabilizes the convertase by primarily binding C3b, although it makes additional contacts with Factor B and Bb. [54][55][56] Properdin binds to C3bB and C3bBb with higher affinity than to C3b alone. 55 Interestingly, properdin has been shown to directly limit Factor H cofactor activity 55,57 and is predicted to affect Factor H decay accelerating activities via structural data, 54 suggesting that defective Factor H regulation may enhance the positive regulatory functions of properdin.

| MOLECULAR STRUCTURE AND CHARACTERISTICS OF PROPERDIN
Properdin is a 53-kDa monomer composed of six complete thrombospondin type 1 repeat (TSR) domains labeled TSR 1-6 58,59 and a truncated N-terminal TSR domain containing key conserved residues denoted TSR 0. 59  Deletion of TSR 3 has no effect on binding to C3b or sulfatides nor oligomer formation. Domain 4 and 5 deletion mutants fail to form trimers and tetramers but retain the ability to dimerize, while deletion of TSR6 completely inhibits oligomerization. 62 Collectively, the results from the study by Higgins et al. 62 suggest that multiple TSR domains act in concert to mediate properdin oligomerization and function, with key roles implicated for domains 5 and 6. The critical nature of domains 5 and 6 for properdin functionality were supported by subsequent studies in which antibodies raised against human TSR5 66 and mouse TSR5-6 67 produced in Escherichia coli effectively inhibited properdin function in vitro and in vivo, respectively. Furthermore, the Y387D point mutation in TSR6 abrogated the ability of properdin to bind C3b and regulate the alternative pathway. This mutation is associated with type III properdin deficiency, characterized by severe functional impairment despite normal plasma levels of properdin. 68 Two other mutations, R73W (TSR1) and Q316R (TSR5), are related to type II properdin deficiency, defined by very low levels of circulating properdin. 69 Properdin recovered from these affected individuals had oligomerization defects and primarily consisted of dimers. 70 Other previously identified mutations in the properdin gene primarily lead to type I properdin deficiency defined by a complete lack of circulating properdin, 69 and thus these mutations have not provided useful insights into the domains/residues mediating properdin function.
Structural studies by Sun et al. 59 and Alcorlo et al. 54 proposed models for oligomerization. These studies illustrated roles for TSRs 0-1 and 5-6 in mediating contacts at the vertices of properdin oligomers, Ziegler et al. 78 showed that higher levels of properdin were found in the urine of patients with renal disease vs healthy individuals, and Corvillo et al. 79 proposed using serum properdin concentrations as a diagnostic marker for a subset of C3 glomerulopathy patients. Studies are lacking, however, that shed light on the dynamics of properdin metabolism in the context of diseases that do not affect the kidney or that have excessive leukocyte activation or leukocyte dysfunction.
Another key issue that remains unresolved is the molecular composition of properdin produced by different leukocyte sources. T-cell-derived properdin was shown to be approximately 100× more active than serumderived properdin in an alternative pathway hemolysis assay 72 ; however, the reason for this enhanced functionality was never determined. Saggu

| PROPERDIN AS AN INITIATOR OF ALTERNATIVE PATHWAY ACTIVITY: EVIDENCE AND SETBACKS
When Pillemer first described properdin in 1954, he proposed it acted as an initiator of alternative pathway activity. 80 This theory was later discounted and replaced with the well-established role of properdin Upon repeated freeze-thaw cycles, purified properdin forms nonphysiological aggregates (P n ) that have distinctly different properties from physiological oligomers. First, non-physiological aggregates have 'activated properdin activity', meaning they induce alternative pathway activation in solution leading to complement consumption. 65,92 Second, non-physiological aggregates bind non-specifically to surfaces, including live Jurkat and Raji cells, 83 Neisseria species, 93 and nonactivated platelets. 31 The physiological oligomers, however, can act as pattern recognition molecules binding selectively to specific surfaces, including zymosan, necrotic Jurkat and Raji cells, 83 Chlamydia pneumoniae, 82 and activated platelets. 31 The arrangement of properdin monomers in non-physiological aggregates (described by Farries as 'large amorphous aggregates') remains unknown, but given the highly positively charged nature of properdin monomers, non-specific ionic interactions with negatively charged anions could account for their robust binding abilities.
Unfractionated properdin isolated from plasma, which contains properdin polymers (P 2 , P 3 , P 4 , and non-physiologicial aggregated P n ) is visualized as approximately 53-kDa monomers in SDS-PAGE under both reduced and non-reduced conditions. Although some studies detect purified properdin at high molecular weights (approximately 220 kDa) in SDS-PAGE, and others have used SDS-PAGE to rule out the presence of aggregates in their properdin preparations, this is inadequate because the polymeric forms (including the non-physiological aggregates) are not visible on an SDS-PAGE 83 due to the non-covalent nature of the polymer associations. The multimers can easily be detected, however, when the same unfractionated properdin preparations are separated by size exclusion or ion exchange chromatography. 83 Interestingly, in a coimmunoprecipitation assay, Pauly et al. 94 discovered C3(H 2 O), C3 fragments, clusterin, and immunoglobulins associated with properdin in human serum. Also, C3b 2 -natural IgG complexes have been shown to stimulate complement amplification in a properdin-dependent manner. 95 These associations between properdin and other molecules may account for the properdin that migrates at a high molecular weight in SDS-PAGE in some purified preparations. The nature of the interactions between properdin and these other molecules in human serum, which resist SDS, denaturing, and/or reducing agents, remains to be determined.
Differences have been noted in the binding capabilities of purified properdin vs properdin contained in serum. For instance, purified physiological properdin oligomers bound directly to zymosan in solution, 83 but properdin required C3 activity to bind zymosan in the presence of serum in subsequent studies. 96,97 Similarly, binding of properdin trimers to thrombin-activated platelets was inhibited in the presence of normal human serum. 31 These results have led to the proposition that properdin binding is regulated in serum by unknown inhibitors. As mentioned above, C3(H 2 O), C3 fragments, clusterin, and immunoglobulins were associated with properdin in human serum. 94 While the physiological effects of the interactions between these molecules and properdin remain unknown, it is possible that they could account for the qualitative differences in binding observed for purified properdin vs properdin contained in serum. Recent evidence also indicates that monomeric C-reactive protein (mCRP) inhibits the binding of unfractionated properdin to proximal tubular epithelial cells. 88 Because mCRP is generated and present mainly on cell surfaces or local microenvironments, 98 it is likely that the inhibitory capacity of mCRP may be limited to cell surfaces and not affect properdin activity in normal human serum. The ability of serum to inhibit direct properdin binding can also be influenced by the specific ligands, as Xu et al. 91 showed that unfractionated properdin and properdin in either normal human serum or C3-depleted serum bound directly to necrotic cells to approximately the same extent. This adds another layer of complexity to properdin studies, indicating that direct properdin binding is influenced not only by the type of properdin used for the study but also by the specific ligands tested. Inhibitors in serum may fine-tune properdin-initiated functions by limiting direct properdin binding to some, but not all, surfaces.
Because of the ability of neutrophils to secrete properdin from their secondary granules upon stimulation, 76  Conversely, Harboe et al. 97 showed that neutrophil-derived properdin was incapable of binding directly to solid-phase zymosan, a surface previously shown to directly bind physiological properdin oligomers. 83 The discrepancies between the latter two studies may be due to the different methodologies used to determine properdin binding F I G U R E 6 Properdin functions. Properdin (P) enhances alternative pathway activity in one of two ways: (1) acting as a positive regulator of pre-existing alternative pathway activity or (2) initiating alternative pathway activity. Properdin acts as a positive regulator by stabilizing the alternative pathway C3 and C5 convertases, increasing their activity 5-to 10-fold. As a pattern recognition molecule, properdin binds selectively to specific surfaces upon which it recruits C3b or C3(H 2 O) to initiate alternative pathway activity (solid-phase ELISA 97 vs flow cytometry), 83

| PROPERDIN AND THROMBOINFLAMMATION
Within the past decade, there has been a renewed interest in evaluating the role of properdin in disease pathogenesis. This has led to the development of properdin-deficient mice, which have been utilized in murine models of asthma, 103,104 arthritis, 77,105,106 AAA, 99,103 non-septic 107 and septic 74 shock, and C3 glomerulopathy. 36,49 These studies have provided key information about the breadth of diseases exacerbated by properdin function and many have been reviewed elsewhere. 71,108,109 Here, we will focus on the interactions of human properdin with platelets and neutrophils and its potential impact on thromboinflammation in the vasculature. To facilitate discussion, we begin with a brief introduction about platelet/granulocyte aggregates.

| Platelet/granulocyte aggregates
Platelets not only are key mediators of vascular hemostasis as well as pathologic thrombosis, but they secrete multiple proinflammatory mediators, many of which recruit and activate neutrophils. 110 While neutrophils are key players in the innate immune response, they also express tissue factor to aid in thrombi formation 111,112  (ii) upregulation of adhesion molecules on neutrophils to promote infiltration into the vasculature; (iii) increased production of reactive oxygen species; and (iv) increased tissue factor expression to enhance thrombi formation. 117-119 Therefore, excess PGA formation has significant pathologic potential, which is reflected in the beneficial effects of limiting PGA formation in various animal vascular injury models. [131][132][133][134] Furthermore, excess PGA formation is a phenomenon observed in patients suffering from diverse chronic inflammatory diseases, including acute coronary syndromes, 135 inflammatory bowel disease, 136 inflammatory lung disease, 137 and diabetes. 138

| Platelets and complement
Platelet activation is a prerequisite for PGA formation, and activat-  Fig. 8.
F I G U R E 7 Main physical contacts governing platelet/granulocyte aggregate interactions. P-selectin initially tethers activated platelets to granulocytes via binding to PSGL-1. This interaction leads to an intracellular signaling cascade that activates and increases the expression of CR3 on granulocytes. CR3 binds multiple ligands on platelets including GPIIb/IIIa through a fibrinogen bridge, GP1bα directly and via a HMWK bridge, ICAM-2, and JAM-3. CD40L and TREM-1L, expressed on activated platelets, enhance granulocyte functions via binding to their granulocyte counter-receptors, CD40 and TREM-1, respectively F I G U R E 8 Complement activation on platelets. Platelets activate both the alternative and classical pathways on their surface. Expression of the surface receptor, gC1qR, is increased after stimulation with shear stress and binds C1q to initiate classical pathway activation. Chondroitin sulfate released from platelets upon activation also binds C1q to initiate classical pathway activity in the fluid phase. Physiological properdin (P) binds directly to activated platelets and recruits C3(H 2 O) to initiate alternative pathway activity. C3(H 2 O) also binds directly to activated platelets and leads to alternative pathway activation in the presence of properdin. Properdin-mediated alternative pathway activation on platelets is controlled by Factor H

| Properdin and neutrophils
While C5a generation was not directly measured in the study by Saggu et al., 31 complete activation of the terminal pathway on platelets implies that platelet-mediated alternative pathway activation leads to the generation of C5a, a potent neutrophil chemoattractant and stimulator. 148

| Properdin and PGA formation
In line with this theory, researchers have elucidated roles for complement in enhancing PGA formation in human whole blood using extracorporeal circuits that simulate coronary artery bypass [151][152][153][154][155] and ex vivo flow cytometry assays. 145,156-159 Rinder et al. 152,155 demonstrated that PGA formation in an extracorporeal circuit was dependent on C5a-mediated effects, a result that was also seen by Lappegard et al. 151 using a heparin-coated circuit. The latter study also established a role for the alternative pathway in enhancing C5a, by demonstrating the ability of an anti-Factor D antibody to limit PGA formation and upregulation of CR3 on neutrophils. 151  Ruef et al. 158 showed that unfractionated properdin could induce platelet/leukocyte aggregate (PLA) formation in human whole blood and significantly enhanced PLA formation in the presence of the weak platelet agonist, ADP in an ex vivo flow cytometry assay. Using a similar system, Blatt et al. 156 recently addressed the role of physiological properdin oligomers in their ability to induce PGA, determining that they enhance PGA formation in the presence of thrombin receptoractivating peptide (TRAP). Each properdin form (dimers, trimers, and tetramers) also sensitized whole blood to PGA formation when added to blood stimulated with a dose of TRAP that induced little to no PGA formation by itself. Properdin tetramers were the most potent oligomer with regard to enhancing PGA formation and alternative pathway activity. These data were in agreement with the high degree of 'native properdin activity' originally ascribed to properdin tetramers by Pangburn 65 and also suggest that high levels of properdin (especially properdin tetramers) in the local microenvironment have the potential to exacerbate thromboinflammation in the presence of relatively weak platelet stimulation. In addition, it was determined that inhibition of endogenous properdin in the blood, using inhibitory antihuman properdin monoclonal antibodies, reduced TRAP-induced PGA formation by approximately 50% and to the same level as inhibiting all complement activity with compstatin, an inhibitor that prevents C3 cleavage. 156 The overall effect of properdin was attributed to its ability to enhance the generation of C5a leading to the upregulation of CR3 on granulocytes, 156

| POTENTIAL FOR PROPERDIN INHIBITORS AS LONG-TERM PROPHYLACTICS FOR PREVENTION OF THROMBOINFLAMMATION
Because increased levels of PGA are associated with many different human diseases, [117][118][119][135][136][137][138] targeting platelet/neutrophil interactions could be a widely applicable, beneficial, and long-term therapeutic option to limit thromboinflammation in diverse clinical settings. Effective inhibition of PGA formation with minimal side effects, however, does not come without its challenges. Despite references made to the pathologic effects of PGA formation, 117-119 platelet/ neutrophil interactions are a normal process during vascular homeostasis, as neutrophils help limit platelet activation by degrading ADP, phagocytosing platelets, and releasing pentraxin 3 that limits platelet adhesion to fibrinogen. 162 In turn, platelet-derived inflammatory mediators help activate and recruit neutrophils in response to various pathogens. 163 Platelet/neutrophil interactions in the vasculature are thus quite dynamic and completely inhibiting their binding longterm by targeting P-selectin or CR3 (key receptors on platelets and neutrophils, respectively, for mediating stable aggregate formation) could have detrimental side effects. P-selectin deficient mice have an increased susceptibility to skin infections and show prolonged bleeding times, 164 while CR3 has both key proinflammatory and protective immunomodulatory roles, and thus its dysfunction is associated with diverse clinical phenotypes ranging from increased risk for infection to an increased susceptibility to the development of systemic lupus erythrematosus. 165 In addition, therapeutics that inhibit platelet activation (limiting all platelet/neutrophil interactions, 166 such as clopidogrel), are associated with rare, but potentially life-threatening bleeding phenotypes. 167 The studies by Blatt et al. and Hamad et al., discussed in the previous section, showed that while inhibiting endogenous properdin, all complement activity, or specifically C5a-C5aR1 interactions impairs PGA formation, the extent of inhibition approximates only 50% of the total TRAP-induced PGA formation in their systems. 6,145,156 This indicates that there is a significant complement-independent portion of PGA formation that is likely mediated by direct contacts or soluble non-complement mediators. By binding to PSGL-1, P-selectin induces the activation of CR3 to an intermediate-affinity state capable of binding platelet receptors. Subsequent stimulation by chemokines induces a high-affinity conformation for CR3. 168 Based on these previous studies, 6,145,156 it is likely that properdin-mediated C5a aids in the transition of CR3 from an intermediate-to high-affinity conformation, increasing the chance that neutrophils stably bind to platelets.
This may help to explain why inhibition of properdin or C5a only limits, rather than completely abrogates, TRAP-mediated PGA formation.
The partial effect of complement inhibition on TRAP-mediated PGA formation leads to the possibility that complement activity could be the force that increases granulocyte accumulation at thromboinflammatory sites to a level that becomes pathologic, rather than homeostatic. This theory is supported by epidemiologic and experimental evidence that has independently suggested roles for both granulocytes 169 and complement 170 in cardiovascular disease, an example of a set of diseases associated with increased circulating PGAs. 135 Fig. 9.
The consequence of inhibiting properdin in the presence of impaired Factor H cell-surface protection was discussed above for the PGA model, 156 and was also demonstrated in a different model by Lesher et al. 49 , who showed that inhibition of properdin in the presence of rH19-20 prevented lysis of human and murine erythrocytes by sera.
We have compared the inhibition of properdin to inhibition of C5 on sheep erythrocytes using the same rH19-20-mediated hemolysis assay as originally described, 24 and found properdin inhibitors to be approximately fourfold more effective than C5 inhibitors in preventing hemo-  156 Given that the mechanism for the effect of properdin on TRAP-mediated PGA formation was related to its ability to increase C5a generation, these results were in agreement with findings from the studies by Harboe et al. 14 that showed the alternative pathway accounted for approximately 80% of terminal pathway activation when complement activity was initiated by either the classical or lectin pathway. 15 A clear example of a beneficial effect of inhibiting properdin in an in vivo disease model in which complement activity is not initiated by the alternative pathway comes from investigating the role of complement in a murine model of AAA. 99,177 Not only did these studies show that properdin-deficient mice were resistant to AAA development, 99 a vascular inflammatory disease characterized by neutrophil infiltration, but it also showed that complement activity was initiated by lectin pathway-mediated activation following the deposition of anti-fibrinogen natural antibodies. 177 The fact that properdin inhibition was equally effective as lectin pathway inhibition in this model gives credence to the possibility that properdin inhibitors can have potent long-term effects in inflammatory vascular disorders, regardless of the mechanism by which complement activity is initiated.
While the ex vivo data support a role for properdin-enhanced generation of C5a, properdin inhibition would limit the generation of all complement effector molecules, and the potential in vivo effects of this should be considered. Decreased C3b and iC3b generation may limit neutrophil and monocyte infiltration into inflamed vasculature, 178,179 while decreased C3a levels would limit vasodilatory and proinflammatory effects on the vasculature. 4 A potentially attractive therapeutic strategy may be the use of properdin inhibitors in combination with C5a inhibitors (C5 cleavage inhibitors or C5a receptor antagonists). Properdin inhibitors would F I G U R E 9 Model for the mechanism of the effects of properdin on platelet/granulocyte aggregate (PGA) formation and complement activation on each cell type. (A) Activated platelets initially tether to granulocytes via P-selectin/PSGL-1 interactions, where they activate the classical pathway(CP) on their surface and/or secrete chondroitin sulfate, a known activator of the CP. (B) Properdin-enhanced alternative pathway(AP) activity amplifies CP activity initiated by platelets, leading to the deposition of C3b on the granulocyte surface. The AP can then use deposited C3b to amplify its activity directly on the granulocyte surface. The AP also activates spontaneously on (C) neutrophils and activated platelets, and (D) AP activity is enhanced by high levels of properdin oligomers (P 2 , P 3 , and especially P 4 ), secreted from neutrophils. (E) Properdin-enhanced AP activity ultimately leads to increased levels of C5a that binds to C5aR1 on neutrophils to enhance CR3 expression. (F) Factor H regulates AP/properdin-mediated PGA formation. Figure originally printed in: Blatt et al 156 limit the generation of both C5a and MAC, while C5 inhibitors or C5a receptor antagonists could account for potential C5 cleavage by thrombin or other coagulation enzymes. 180,181 To our knowledge, a therapeutic regimen such as this has never been tested in clinical trials.
With the rate of advances in understanding the roles of properdin and C5a in various disease models, these two types of inhibitors could be employed simultaneously in a way that potentially lowers the effective dose of each inhibitor, thus limiting potential side effects.
To date, no properdin inhibitors have been described that can specifically target different properdin oligomers. Given the enhanced functionality of properdin tetramers compared with other oligomers, 65,156 inhibitors that specifically target properdin tetramers may be especial-  shown to be dependent on properdin-mediated generation of C5a. 185 While these studies both showed worsened outcomes with properdin deficiency, the detrimental effect of the absence of properdin was due to very different mechanisms in these complex diseases. These studies, thus, highlight the context-specific effects of properdin inhibition, which would not be universally applicable to all patients, in particular those that are normally properdin-sufficient, and could potentially be alleviated by careful clinical monitoring.

| Potential consequences of properdin inhibition
Steiner et al. 186 found a protective role for properdin in the development of early atherosclerotic lesions in male mice fed a low-fat diet, but this protective function was eliminated upon feeding mice a highfat diet. While these results are intriguing, due to the strict criteria for the protective effect of properdin (only male mice fed a low-fat diet), it is uncertain how properdin inhibition would affect the development of atherosclerosis, a complex inflammatory disease associated with multiple risk factors, in humans.
Properdin deficiency improved outcomes in a murine model of zymosan-induced non-septic shock, but exacerbated disease in LPSinduced non-septic shock. 107 This study serves as another example of context-specific roles for properdin that would need to be clinically monitored. It should be noted, however, that LPS is a potent neutrophil activator that induces full neutrophil degranulation. 187 Thus, transiently increased levels of properdin due to massive LPS-mediated neutrophil activation during non-septic shock may compensate for decreased properdin function due to properdin inhibitors in patients that are normally properdin sufficient.

| UNDERSTANDING THE INTERPLAY BETWEEN PROPERDIN AND FACTOR H IN THE MICROENVIRONMENT
Studies by Saggu et al. 31 and Blatt et al. 156 highlight the opposing roles of properdin and Factor H in the microenvironment. Aside from binding to key complement components in order to carry out their regulatory functions, both properdin and Factor H bind to a diverse set of ligands other than complement proteins. In addition to sulfatides and the polyanionic structures already mentioned, unfractionated properdin has recently been shown to bind to neutrophil myeloperoxidase, 87 Factor H-related protein-5 (CFHR-5) 81 ), and monomeric C-reactive protein (mCRP). 88 Not only does Factor H bind mCRP, 188 but myeloperoxidase has been shown to interfere with the interaction between mCRP and Factor H, 189  Factor H also binds to many of the same polyanions associated with properdin binding, including heparin, 50,[191][192][193][194][195][196][197][198] heparan sulfate, 102,193,194,196 and chondroitin sulfate C. 26,199 Association with polyanions increases Factor H regulation on surfaces by directing it to specific cells 50,193 and inducing oligomerization via the C-terminal domains. 200  Interesting studies by Ruseva et al. 36  domains of Factor H to another protein that will direct the molecule to inflammatory sites where complement is activating. In TT30, Factor H N-terminal domains are linked to CR2 to specifically target the molecule to sites of C3d accumulation. 203 Mini-Factor H uses the intrinsic cell-localizing capabilities of the Factor H C-terminus, by linking the N-terminal regulatory domains directly to domains 19-20. 204,205 An exciting possibility is the potential use of properdin inhibitors in conjunction with these novel therapeutics. Extra control exerted over alternative pathway activation by TT30 or mini-FH may augment the effects of properdin inhibition, although studies need to be conducted to determine potential synergistic effects. Combined use of properdin inhibitors and Factor H promoters may be particularly attractive for the treatment of patients containing aHUS-related mutations in Factor H. Factor H promoters could directly compensate for the intrinsic defect in Factor H cell-surface protection, while properdin inhibitors, which have proven to be effective at limiting TRAP-mediated PGA formation 156 and complement activation on human erythrocytes 49

when
Factor H regulation is impaired, could limit or prevent disease flareups resulting from alternative pathway activation that overwhelms the level of regulation provided by Factor H promoters.

| PROMOTING PROPERDIN ACTIVITY ON PATHOGENS
Much of this review has been dedicated to the potential utility of inhibiting properdin in thromboinflammatory diseases and the associated consequences; however, properdin undoubtedly is a key weapon in the fight against invading microorgansims. A murine model of polymicrobial sepsis showed increased morbidity and mortality in the presence of genetic properdin deficiency, 74 and properdin enhances complement activation on Chlamydia pneumoniae, 82 Neisseria meningitidis and gonorrhoeae, 93,206 fungal glycans, 96 and Escherichia coli. 74,89,97 Therefore, finding ways to direct properdin activity to the microbial surface could provide novel ways to resolve infections.
Recombinantly produced properdin, predominantly containing non-physiological aggregates, enabled opsonization of Streptococcus pneumoniae and Neisseria meningitidis by human serum. In the case of Neisseria meningitidis, the recombinant properdin also mediated in vitro killing in serum. The injection of recombinant properdin was subsequently shown to have beneficial effects in murine infection models with both pathogens. 207 Spitzer et al. 89 showed that a single chain antibody (scFv) targeted to murine or human erythrocytes and linked to properdin could induce complement activation on the erythrocytes. Utilizing this targeting strategy against microorganisms could prove to be effective, especially if the scFv is directed against bacterial proteins known to bind complement regulatory proteins. For instance, Neisseria meningitidis expresses several proteins that recruit Factor H to its surface, [208][209][210] and thus scFv's directed against Factor H-binding proteins could promote alternative pathway activity on the bacteria by simultaneously increasing properdin activity and decreasing Factor H regulation. This strategy would also likely bypass any need for antibody-dependent classical pathway initiation due to the potent complement initiation properties of scFv-properdin. 89 Many bacteria other than Neisseria species including Borrelia burgdorferi, 211 Yersinia enterocolitica, 212 and Streptococcus pyogenes 213 bind Factor H (reviewed in 16) to protect themselves from the alternative pathway, therefore scFv-properdin targeted against Factor H-binding proteins could be an efficient strategy for treating infections resulting from a diverse array of pathogens.

| CONCLUDING REMARKS
Properdin is a powerful inflammatory modulator that is tightly regulated by Factor H and has the ability to both initiate and positively regulate the alternative pathway amplification loop for all complement activity. While current evidence is lacking for an initiating role for properdin in vivo or ex vivo, there remains a possibility that properdin initiation enables activation of the alternative pathway on the surface, but quickly becomes irrelevant upon deposition of multiple C3b molecules, allowing for its more prominent role as a positive regulator. With the development of novel reagents that can differentiate between initiating and regulatory roles of properdin, as well as discern properdin function resulting from different properdin oligomers or sources, critical advances will be made in understanding the complex biology of properdin in the context of diverse inflammatory and thromboinflammatory environments. A better understanding of properdin biology will open the door for the generation of novel properdin inhibitors that can be employed alone or in combination with other complement inhibitors to prevent or limit disease pathogenesis. By studying the molecular basis for the interplay between properdin and Factor H, we will better be able to predict the potential effects of either inhibiting or promoting properdin activity and could develop strategies to maximize effects on the alternative pathway by employing methods that simultaneously inhibit properdin and promote Factor H activity, or vice versa. These types of strategies may find particular usefulness in the clinic to limit inflammation in diseases marked by excessive alternative pathway activation or to treat microbial infections. With these thoughts in mind, we can collectively contribute to the knowledge required to make significant advances in the treatment or prevention of multiple disease etiologies.