Dr Shaun P. Jackson, Australian Center for Blood Diseases, Department of Medicine, Monash Medical School, Box Hill Hospital, Box Hill, Victoria, Australia 3128. Tel.: +613 98950350; fax: +613 98950332; e-mail: email@example.com
Summary. Recent in vivo studies have highlighted the dynamic and complex nature of platelet thrombus growth and the requirement for multiple adhesive receptor–ligand interactions in this process. In particular, the importance of von Willebrand factor (VWF) in promoting both primary adhesion and aggregation under high shear conditions is now well established. In general, the efficiency with which platelets adhere and aggregate at sites of vessel wall injury is dependent on the synergistic action of various adhesive and soluble agonist receptors, with the contribution of each of the individual receptors dependent on the prevailing blood flow conditions. In this review, we will discuss the major platelet adhesive interactions regulating platelet thrombus formation under high shear, with specific focus on the VWF (GPIb and integrin αIIbβ3) and collagen receptors (GPVI and integrin α2β1). We will also discuss the signaling mechanisms utilized by these receptors to induce platelet activation with specific emphasis on the role of cytosolic calcium flux in regulating platelet adhesion dynamics. The role of soluble agonists in promoting thrombus growth will be highlighted and a model to explain the synergistic requirement for adhesive and soluble stimuli for efficient platelet aggregation will be discussed.
Over the past 10 years a great deal of progress has been made in understanding the mechanisms regulating platelet thrombus growth. In particular, a much clearer picture has emerged regarding the specific roles of various adhesive proteins and receptors in thrombogenesis and the influence of rheological factors on this process. Much has also been learned about the mechanisms used by adhesive proteins and soluble agonists to induce platelet activation and the signaling pathways involved. An area that is less clearly defined is the signaling relationships operating between soluble agonist and adhesion receptors to regulate the various stages of thrombus development. Moreover, we continue to have limited insight into how these signals are integrated to modulate the dynamics of thrombus growth.
In this review we will discuss the role and contribution of platelet adhesive receptors and ligands in the initiation and propagation of thrombus growth and highlight the signaling mechanisms used by these receptors. We will also discuss the contribution of soluble agonists in potentiating platelet activation and aggregate formation and highlight potential synergies with adhesive receptors in this process. We will primarily focus on thrombus development under high shear conditions, as the specific receptor–ligand interactions mediating this process have been most clearly delineated, and because clinically this is of relevance to the pathogenesis of the acute coronary syndromes and stroke. It should be emphasized that the contribution of specific adhesive interactions and soluble stimuli to thrombus development is dependent on the prevailing blood flow conditions, therefore caution should be exercised when attempting to extrapolate from conclusions based on high shear experiments to those occurring in low shear environments. Towards the latter part of this review, we will discuss a potential working model to explain the dynamics of thrombus growth in the context of calcium signaling. In particular, we will highlight the co-operative signaling relationship operating between GPIb and integrins αIIbβ3 and the importance of adhesive and soluble costimuli in regulating the efficiency of platelet adhesion and aggregation in a shear field.
Platelet thrombus dynamics
Platelet adhesion to exposed subendothelium is a complex multistep process involving a diverse array of adhesive ligands (von Willebrand factor [VWF], collagen, fibronectin, thrombospondin and potentially laminin) and receptors on the platelet surface (GPIb/V/IX, GPVI, integrins αIIbβ3, α2β1, α5β1 and α6β1) [1–13]. The nature of the adhesive interactions operating during platelet adhesion and thrombus growth is largely influenced by the prevailing rheological conditions. For example, under low shear conditions, such as those experienced in larger arteries and veins, platelet adhesion to the vessel wall is thought to primarily involve one or more fibrillar collagens, fibronectin and laminin [1–3,11–13]. However, under high shear conditions, the initial tethering of platelets to the damaged subendothelium is critically dependent on the binding of platelet GPIb/V/IX to subendothelial bound VWF [14–21]. While sufficient to promote binding of platelets to the injured vessel wall, this adhesive interaction is characterized by a rapid dissociation rate, resulting in platelet translocation at the site of injury. This rapid deceleration in platelet velocity allows adhesive interactions with slower intrinsic binding kinetics (i.e. GPVI and/or integrins) to mediate firm platelet adhesion.
In contrast to primary adhesion, subsequent platelet– platelet adhesion (aggregation or cohesion) is exclusively mediated by two receptors, GPIb/V/IX and integrin αIIbβ3[5,16,22–26], with the contribution of GPIb becoming progressively more important with increasing blood flow. The relative roles of VWF and fibrinogen in supporting platelet aggregation and thrombus growth have been defined using in vitro and in vivo models [15,25,27–29]. Under high shear, VWF is the major ligand promoting platelet aggregation with fibrinogen and/or fibrin playing a stabilizing role, whereas at low shear fibrinogen is thought to be the primary ligand supporting thrombus growth. The finding that thrombus formation can persist in the absence of VWF and fibrinogen has raised the interesting possibility that a third integrin αIIbβ3 ligand, perhaps fibronectin, facilitates platelet aggregation in vivo[27,30].
Recent advances in intravital microscopy have highlighted the dynamic nature of platelet thrombus growth in vivo[26,27,31,32]. In particular, they have established that the majority of platelets tethering to the injured vessel wall or to the surface of thrombi translocate for a variable period prior to forming firm adhesion contacts or detaching back into bulk flow. This phenomenon appears to be a general feature of thrombus formation, occurring in the arterial and venous circulation . A critical determinant regulating thrombus growth is the proportion of translocating platelets forming stable adhesion contacts, primarily a reflection of the activation status of integrin αIIbβ3. In general, the affinity status of integrin αIIbβ3 reflects the cumulative action of multiple platelet-activating stimuli (Fig. 1). In the following sections, we will examine the specific contribution of individual adhesion receptors to the generation of intracellular signals (outside-in signaling) and discuss potential mechanisms by which these signals may cooperate to regulate integrin αIIbβ3 activation and platelet adhesion dynamics under flow.
GPIb/V/IX signaling during platelet adhesion to VWF
The dynamics of platelet adhesion on a purified VWF matrix are strikingly similar to those observed during thrombus development in vivo, in that platelets tethering to the VWF surface translocate prior to forming firm adhesion contacts [4,5]. Recent studies have provided insight into the mechanisms controlling adhesion dynamics on VWF [23,33–36] that may be relevant to the understanding of thrombus growth on more complex biological surfaces. While the importance of the VWF–GPIb interaction in supporting platelet–subendothelial and platelet–platelet interactions has been clearly delineated, the contribution of this adhesion event to the initiation of platelet activation remains less clearly defined. There are three important issues with respect to GPIb signaling: (i) does GPIb signal; (ii) how does it transduce signals; (iii) what is the physiological significance of GPIb signaling?
Considerable controversy continues to exist with respect to whether GPIb signals. Some studies have failed to detect signaling downstream of GPIb , others have detected weak activating signals [23,36–42], while a few reports have demonstrated strong platelet activation by VWF [43–45]. Much of the discrepancy between different studies is most likely due to the large variations in experimental strategies used to examine GPIb signaling, including the use of different ligands (venom peptides, human or bovine VWF, recombinant VWF fragments, soluble or immobilized VWF), artificial modulators (ristocetin, botrocetin), cell types (human and mouse platelets, transfected CHO and K562 cell lines) and functional assays (suspension aggregation studies vs. flow-based adhesion assays, static vs. shear conditions). Factors that modulate the affinity and number of bonds between the VWF A1 domain and GPIb are likely to have a significant influence on signaling and the effects of shear forces on GPIb signaling are only just beginning to be defined [23,36]. Based on the weight of current evidence in the literature, it is reasonable to conclude that GPIb probably does generate intracellular signals; however, the absolute importance of these signals for normal thrombus development remains unclear.
The mechanisms by which GPIb generates signals remains controversial , with some studies suggesting direct signaling through the cytoplasmic tails of GPIb/V/IX , and others suggesting signaling through physically associated surface molecules, such as FcR γ-chain [45,47] and FcRγIIa [43,44,48], while yet others suggesting indirect mechanisms linked to ADP release [23,28,49–51] and possibly thromboxane A2 generation [37,43]. The cytoplasmic tails of the individual subunits of the GPIb/V/IX receptor complex do not have intrinsic kinase activity, nor do they bind GTP-binding proteins or become phosphorylated by tyrosine kinases. It has been proposed that GPIbα signals directly as a consequence of its association with the cytoskeletal structural protein, actin-binding protein (ABP-280)  and/or signaling molecules such as 14–3-3ζ[53–57], calmodulin , Src kinases  and possibly phosphoinositide 3-kinase (PI 3-kinase) . Alternatively, GPIb has been proposed to transduce signals indirectly through its physical association with the ITAM-bearing receptors, FcR γ-chain [45,47,60] or FcRγIIa [43,44,48]. Studies examining shear-induced platelet aggregation have suggested that VWF primarily stimulates platelet activation through an indirect pathway linked to ADP release [49–51]. According to these studies, the VWF–GPIb interaction initiates platelet activation by stimulating transmembrane calcium influx through channels in the platelet plasma membrane. This increase in cytosolic calcium promotes dense granule secretion of ADP, which induces integrin αIIbβ3 activation through engagement of the purinergic receptors, P2Y1 and P2Y12. Of course, these distinct models of GPIb signaling are not necessarily mutually exclusive and may contribute to platelet activation to varying degrees depending on the experimental conditions.
Co-operative signaling between GPIb and integrin αIIbβ3 regulating platelet adhesion on VWF
Recent in vitro flow studies have demonstrated an important role for GPIb-induced calcium signals in initiating integrin αIIbβ3 activation, necessary for stable platelet adhesion on VWF [23,36]. These transient calcium spikes occur independently of extracellular calcium, ADP and TXA2, indicating that signals operating downstream of GPIb may directly regulate calcium release from internal stores. Sustained calcium oscillations appear to be dependent on a distinct calcium signal operating downstream of integrin αIIbβ3, involving both intracellular calcium mobilization and transmembrane influx (Fig. 2). The mechanism by which integrin αIIbβ3 promotes intracellular calcium release has not been elucidated, although one proposed mechanism involves signaling processes akin to that utilized by ITAM-bearing receptors (e.g. GPVI/FcR γ-chain) (Fig. 3). This is primarily based on the observation that many of the signaling molecules implicated in integrin αIIbβ3 signaling are also involved in ITAM receptor signaling. For example, the Src kinase Fyn, known to phosphorylate ITAM tyrosine motifs, has also been shown to phosphorylate the β3 tail of integrin αIIbβ3. Similarly, the non-receptor tyrosine kinase Syk, which binds to the cytoplasmic tails of ITAM-bearing receptors [62,63], has also been demonstrated to bind integrin αIIbβ3. Other similarities include the recruitment of adaptor molecules to ligated receptors; LAT (Linker for Activation of T cells) in the case of ITAM-receptors [62,63] and Shc in the case of integrin αIIbβ3. These adaptor molecules may then facilitate the binding and activation of additional signaling molecules including PLC and PI 3-kinase, eventually leading to PI turnover, calcium flux and PKC activation. Inhibition of PI 3-Kinase or PKC is known to abrogate integrin activation and associated sustained calcium flux under flow [23,33,36]. These enzymes are likely to be important contributors to the positive feedback loop linking integrin αIIbβ3 activation and sustained calcium flux under high shear.
The demonstration that GPIb and integrin αIIbβ3 elicit distinct calcium signals during platelet adhesion to VWF provides insight into the mechanism by which dynamic changes in cytosolic calcium directly translate into changes in platelet translocation behavior under flow. For example, single calcium spikes elicited by GPIb are sufficient to induce transient platelet arrest through the reversible activation of integrin αIIbβ3, contributing to the stop–start nature of platelet translocation [23,36]. Subsequent sustained calcium oscillations are involved in promoting firm adhesion. On a VWF substrate, the efficiency of this latter process is low, as evidenced by the small proportion of translocating platelets exhibiting sustained calcium oscillations. This aspect of VWF function is likely to be important for regulating the fidelity of hemostatic plug formation, as excessive activation of platelets through the VWF–GPIb–integrin αIIbβ3 axis per se, would represent a significant risk factor for the development of pathological thrombi. In the following sections we will discuss the important role of adhesive and soluble costimuli in potentiating GPIb and integrin αIIbβ3-derived calcium signals, necessary for the efficient conversion of translocating platelets to stably adherent cells.
Signaling pathways promoting platelet adhesion to collagen
In contrast to VWF, collagen is a highly efficient substrate at supporting stable platelet adhesion and thrombus growth. More than 90% of platelets tethering to this matrix form stationary adhesion contacts and once adherent these platelets provide highly reactive surfaces for the recruitment of additional platelets. Considering that initial platelet tethering to collagen under high shear conditions requires VWF binding to GPIb (Fig. 4), these findings reinforce the important role of the two major platelet collagen receptors, GPVI and integrin α2β1, in promoting efficient platelet adhesion and activation under flow. Recent studies on mice platelets lacking GPVI have suggested a dominant role for this receptor in promoting platelet adhesion to collagen . Integrin α2β1 also appears to contribute to platelet adhesion through amplification of signals initiated by GPVI [67,68]. The precise details underlying GPVI signaling have been described in detail elsewhere [67,69] and will only be briefly summarized here (Fig. 4). GPVI binding to collagen results in the clustering or capping of associated FcR γ-chains. ITAM domains within the cytoplasmic tails of the clustered FcR γ-chains are phosphorylated by the Src kinases, Fyn and Lyn, providing binding sites for Syk. Subsequent recruitment of the adaptor proteins SLP-76 and LAT promotes accumulation of numerous other proteins [70,71], including PI 3-kinase, Grb2, Vav, WASP and PKC. Ultimately, this signaling cascade leads to the phosphorylation and activation of PLCγ2 generating a robust calcium signal that promotes efficient platelet activation. Despite the importance of GPVI in promoting platelet adhesion and activation on collagen in vitro, its deficiency does not appear to significantly undermine the hemostatic function of platelets in vivo, at least in mice . How important this receptor is in promoting pathological thrombus formation, particularly in diseased vessels enriched in types I and III fibrillar collagens, remains unclear.
Insight into the mechanisms regulating the thrombogenic potential of distinct adhesive substrates has been obtained from flow studies examining cytosolic calcium flux during platelet adhesion to VWF or type I fibrillar collagen . These studies have identified three important factors regulating the efficiency of platelet adhesion and aggregation: (i) the proportion of primary adherent cells eliciting a cytosolic calcium response; (ii) the magnitude and pattern of the calcium signal; and (iii) the duration of the calcium signal. For example, on a VWF matrix, only a small percentage of tethered platelets (<10%) exhibit an oscillatory calcium response and the mean Δ[Ca2+]c in these cells is relatively low (<250 nm), resulting in sustained cohesive contacts in ∼40% of tethered platelets (<5% of the total tethered platelet population). In contrast, >90% of platelets that tether to collagen display high oscillatory Δ[Ca2+]c responses (>900 nm) and the calcium levels are sustained in all cells, resulting in the rapid formation of irreversible platelet adhesion contacts. On a collagen matrix these primary adherent cells provide extremely efficient nuclei for the subsequent recruitment of platelets from flowing blood. The elevated and sustained calcium response elicited by type I collagen is transmitted to subsequent layers within the developing thrombus. These findings suggest that the efficient communication of calcium signals between aggregating platelets plays an important role in dictating the rate and extent of thrombus growth.
An important outstanding issue relates to the mechanism by which an adhesive substrate that only interacts with the primary adherent layer of platelets propagates activating signals that drive thrombus growth. As will be discussed in the next section, the concept of costimuli potentiating the initial GPIb and integrin αIIbβ3 calcium signals relevant to efficient platelet adhesion on a collagen substrate is also likely to be applicable to platelet aggregation. However, in this latter situation, costimulation primarily occurs through the activation of one or more soluble agonist G-protein coupled receptors, rather than through tyrosine-kinase linked receptors.
Role of soluble agonists in promoting thrombus growth under high shear conditions
In this section of the review we will briefly discuss the contribution of soluble agonists in promoting platelet thrombus growth. Due to space limitations, there will only be a very brief discussion on the mechanisms used by ADP, TXA2 and thrombin to promote platelet activation, as more detailed reviews on this subject can be found elsewhere [74–77]. In the final part of the review, we will discuss potential spatial and temporal signaling relationships operating between soluble agonist and adhesion receptors relevant to the dynamic regulation of platelet adhesion and thrombus growth.
The role of ADP in promoting platelet aggregation and thrombus growth is well established . This agonist binds to two major platelet purinergic receptors (P2Y1 and P2Y12), which play an important role in potentiating platelet activation induced by other stimuli. A good deal of recent progress has been made in defining the signaling mechanisms utilized by these receptors to induce platelet activation. P2Y1 is coupled to Gq and ligation of this receptor in platelets is associated with PLCβ-mediated intracellular calcium elevation and integrin αIIbβ3 activation [78–80]. P2Y12, on the other hand, is coupled to the inhibition of adenylyl cyclase, potentiating platelet activation by suppression of cAMP .
A considerable body of evidence supports an important role for ADP in promoting platelet activation under high shear conditions. For example, studies examining shear effects on platelets using a cone-and-plate viscometer have demonstrated an important requirement for ADP for shear-induced platelet activation [28,49–51,82]. Recent studies in our laboratory have demonstrated a key role for P2Y12 activation in driving the progressive accrual of platelets within developing aggregates under conditions of blood flow . Similarly, analysis of thrombus formation on immobilized fibrillar collagen has demonstrated an important role for ADP in promoting three-dimensional thrombus growth under high shear [83–86]. The P2Y1 and P2Y12 receptors may have distinct, complementary roles in thrombus development, with the P2Y1 receptor promoting initial thrombus growth and the P2Y12 receptor helping to sustain platelet activation required for stable thrombus formation. With the recent generation of mice with specific deletions in the P2Y1 and P2Y12 genes, considerable new insight into the relative roles of these two receptors in promoting thrombus formation under high shear conditions is likely to be gained over the next few years.
TXA2 activates platelets by binding to a specific G-protein receptor coupled to the Gq and G12/13 family members [78,87,88]. Platelet activation through the Gq pathway activates PLCβ while G12/13 regulates MLC phosphorylation through activation of Rho-kinase. The role of TXA2 in supporting the hemostatic function of platelets is well established; however, its role in promoting thrombus growth, particularly under pathological shear conditions, is less clearly defined. For example, studies examining platelet aggregation in a cone-and-plate viscometer have demonstrated that inhibition of TXA2 generation has no effect on shear-induced platelet aggregation [50,89,90]. Similarly, in laminar flow studies, pretreating platelets with aspirin has no inhibitory effect on thrombus formation at pathological shear rates (10 500 s−1), although at lower shear rates thrombus growth is reduced [15,91–93]. These in vitro findings have been verified in animal thrombosis models in which increasing the degree of stenosis in injured dog coronary arteries, thereby increasing shear forces, overcomes the antithrombotic effects of aspirin [94–96]. These findings may partly explain clinical observations that aspirin fails to prevent occlusive thrombosis in patients exhibiting high degrees of arterial stenosis .
Thrombin is one of the most potent activators of platelets and has a well-established role in promoting thrombus formation under all shear conditions. Thrombin activates platelets through multiple cell surface receptors including the GPIb/V/IX complex and the protease-activated receptors (PARs). In human platelets, thrombin cleaves two PARs, PAR1 [98,99] and PAR4 [98,100], with the former receptor coupled to Gq, G12/13 and Gz/i and the latter receptor coupled to Gq and G12/13 only [98,100–102]. Thrombin cleavage of PAR1 is associated with the activation of numerous signaling pathways involving the activation of PLCβ, PI 3-kinase and RhoA/Rho kinases. Thrombin cleavage of PAR4 is also sufficient to induce platelet activation , although the rate of platelet activation through this receptor is significantly slower and more sustained than through PAR1 [74,103,104]. Recent evidence suggests that thrombin can also induce platelet activation through non-catalytic means following removal of GPV from the GPIb/V/IX complex . The mechanism for platelet activation under these conditions has not been defined but requires thrombin binding to GPIb.
Early studies examining the role of thrombin in promoting platelet thrombus formation in ex vivo perfusion systems suggested an important role for this agonist at low (100 s−1) to intermediate (650 s−1) shear rates, but little contribution at higher shear rates (>2600 s−1) [106,107]. However, more recent studies on PAR3 and PAR4 knockout mice have demonstrated defective thrombus formation under high shear conditions in vivo[108,109]. Similarly, blockade of tissue factor has also been demonstrated to inhibit thrombus formation at high shear rates using in vitro thrombosis models . Whether thrombin is primarily involved in promoting initial thrombus growth or is more important for stabilizing formed thrombi remains unclear, although ex vivo perfusion studies have suggested that the latter function may be more important .
Co-operative signaling between soluble agonists and adhesion receptors enabling efficient platelet adhesion under flow
A key outstanding issue relates to the mechanism by which soluble agonists co-operate with adhesive stimuli to regulate the spatial and temporal sequences of thrombus growth. This is a complex issue, due in part to the dynamic nature of thrombus formation and also because of the constraints imposed by blood flow on the platelet adhesion/activation process. For example, high shear forces can have complex effects on platelet adhesion. On the one hand, it can promote platelet–matrix interactions by enhancing VWF binding to GPIb, and on the other hand, it can undermine the formation of stable platelet–matrix interactions by applying high tensile forces to forming bonds. Furthermore, rapid blood flow has complex and poorly characterized effects on soluble agonist generation and clearance at the site of vascular injury. Thus by promoting platelet activation and release of soluble agonists, high shear may indirectly potentiate platelet reactivity, a finding that has been most clearly demonstrated in closed experimental systems, such as the cone-and-plate viscometer, in which shear-dependent build-up of ADP plays an important role in inducing platelet activation [28,49–51,82]. However, in flowing blood, ADP released at the luminal surface of developing thrombi may be quickly removed from the site of injury, limiting the platelet-activating effects of such stimuli. Analogous to the coagulation cascade, in which efficient thrombin generation requires localization of enzyme/substrate reactions on immobilized surfaces, efficient platelet activation by soluble stimuli is also likely to require platelet localization to the site of injury. Thus transient adhesion of platelets to the developing thrombus may promote release of ADP and TXA2 in the immediate vicinity of platelet–platelet adhesion contacts, co-ordinating localized platelet activation. This type of spatial signaling provides a mechanism of achieving high local concentrations of excitatory signals at sites of cell–cell contact, which may counteract, at least in part, the rapid clearance of soluble agonists by flowing blood.
The schematic diagram in Fig. 5 provides a simple conceptual model to help explain the synergistic requirement for platelet adhesive events and soluble agonist stimulation for efficient thrombus growth on a VWF matrix under high shear. According to this model, the initial interaction of platelets with the vessel wall or with the surface of developing thrombi through the ‘VWF–GPIb–integrin αIIbβ3 axis’ leads to weak platelet activation, resulting in the majority of platelets undergoing a variable period of surface translocation prior to forming firm adhesion contacts. Central to this model is the concept that costimulation of platelets through adhesion or soluble agonist receptors is required to potentiate and sustain activation signals initiated by the VWF–GPIb–integrin αIIbβ3 axis, leading to platelet deceleration and arrest in the shear field. Based on currently available evidence, cytosolic calcium appears to be the dominant second messenger regulating platelet adhesion dynamics, as there is a strong inverse correlation between calcium flux and platelet translocation behavior .
This model also proposes an important role for integrin αIIbβ3 in co-ordinating calcium signals during platelet aggregate formation under flow. This is based on recent evidence that the formation of stable platelet–platelet adhesion contacts requires the efficient communication of calcium signals between platelets. According to this model, ADP release at the site of platelet–platelet contact serves to sustain integrin αIIbβ3-derived calcium signals by a P2Y12-linked signaling mechanism . An important element of the current model that requires further investigation is the relationship between integrin αIIbβ3 ligation and ADP release. While purely speculative, an attractive hypothesis is that transient integrin αIIbβ3 engagement of VWF promotes release of ADP at the point of platelet–platelet contact, thereby co-ordinating localized platelet activation.
Progress in the understanding of the roles of specific adhesion receptors in promoting primary platelet adhesion and aggregation, coupled with an improved understanding of the signaling mechanisms used by adhesion and G-protein coupled receptors, has provided a framework to begin understanding the signaling mechanisms regulating thrombus formation. However, there is still much to be learned. For example, while there is abundant evidence from in vitro studies that GPIb and integrin αIIbβ3 outside-in signaling plays a potentially important role in promoting both primary adhesion and aggregation, the importance of these signals for effective thrombus growth in vivo, particularly under high shear conditions, remains ill-defined. We also do not have a clear understanding of the importance of the individual collagen receptors in promoting thrombus growth in vivo and whether targeted inhibition of one or more of these receptors is likely to lead to a safe and effective antithrombotic approach. With the increasing availability of mice with specific deletions of one or more hemostatic components, combined with confocal imaging techniques that allow real-time analysis of thrombus growth in vivo, a clearer picture as to the role of individual adhesive and soluble agonist receptors in this process should soon emerge.
We would like to thank Dr Yuping Yuan, Prof Hatem H. Salem and Sascha Hughan for helpful input and discussions. Authors' work cited in this review was supported by funding from the NH & MRC, NHF and The Wellcome Trust.