RhoA Sustains Integrin αIIbβ3Adhesion Contacts under High Shear*

The small GTPase RhoA modulates the adhesive nature of many cell types; however, despite high levels of expression in platelets, there is currently limited evidence for an important role for this small GTPase in regulating platelet adhesion processes. In this study, we have examined the role of RhoA in regulating the adhesive function of the major platelet integrin, αIIbβ3. Our studies demonstrate that activation of RhoA occurs as a general feature of platelet activation in response to soluble agonists (thrombin, ADP, U46619, collagen), immobilized matrices (von Willebrand factor (vWf), fibrinogen) and high shear stress. Blocking the ligand binding function of integrin αIIbβ3, by pretreating platelets with c7E3 Fab, demonstrated the existence of integrin αIIbβ3-dependent and -independent mechanisms regulating RhoA activation. Inhibition of RhoA (C3 exoenzyme) or its downstream effector Rho kinase (Y27632) had no effect on integrin αIIbβ3 activation induced by soluble agonists or adhesive substrates, however, both inhibitors reduced shear-dependent platelet adhesion on immobilized vWf and shear-induced platelet aggregation in suspension. Detailed analysis of the sequential adhesive steps required for stable platelet adhesion on a vWf matrix under shear conditions revealed that RhoA did not regulate platelet tethering to vWf or the initial formation of integrin αIIbβ3 adhesion contacts but played a major role in sustaining stable platelet-matrix interactions. These studies define a critical role for RhoA in regulating the stability of integrin αIIbβ3adhesion contacts under conditions of high shear stress.

Platelet adhesion and aggregation at sites of vascular injury is essential for the arrest of bleeding and for subsequent vessel wall repair. The ability of platelets to adhere under conditions of rapid blood flow requires the synergistic contribution of multiple receptor-ligand interactions, foremost of which involves the interaction between von Willebrand factor (vWf) 1 and the two major platelet adhesion receptors, glycoprotein (GP) Ib/V/IX and integrin ␣ IIb ␤ 3 . The vWf⅐GP Ib/V/IX interaction is characterized by a rapid association rate that enables efficient platelet tethering to the injured vessel wall, whereas subsequent integrin ␣ IIb ␤ 3 engagement of vWf is important for promoting platelet arrest (1). The vWf⅐GP Ib/V/IX interaction is indispensable for normal platelet function, because it slows platelet movement at the vessel wall thereby enabling receptors with slower intrinsic binding kinetics, i.e. integrins, to engage adhesive ligands. A similar dual-step adhesion mechanism, involving selectins and ␤ 2 integrins, is employed by leukocytes to adhere to post-capillary venules at sites of inflammation (2)(3)(4).
A key requirement of the adhesion contacts formed by platelets and leukocytes is their ability to resist the detaching effects of rapidly flowing blood. In the case of platelets, the shear forces operating on adhesive bonds can be extremely high, particularly at sites of arterial narrowing where shear forces can exceed 380 dyne/cm 2 (10,000 s Ϫ1 ). The ability of the GP Ib/V/IX complex to maintain platelet adhesive interactions under such high shear conditions is due to the large number of bonds formed between vWf and GP Ib, the inherent biomechanical stability of the vWf⅐GP Ib interaction, and anchorage of the receptor complex to the membrane skeleton (1). The factors that regulate the stability of integrin ␣ IIb ␤ 3 adhesion contacts in a shear field remain less clearly defined. It is well established that integrin engagement of adhesive ligands leads to receptor clustering and the formation of firm anchorage points with the actin-based cytoskeleton (5), however, the importance of such post-ligand binding events for sustaining platelet adhesion in a shear field remains unclear.
The adhesive function of many integrins is regulated in part by RhoA, a member of the Ras homologous (Rho) family of small GTPases (6). RhoA activation has been extensively studied in a range of cultured cell and is critical for the formation of actin stress fibers and focal adhesions (7). Several RhoA effector proteins have been implicated in the formation of these structures, including p140mDia and Rho kinase (p160 ROCK , ROK␣). The effect of Rho kinase on focal adhesion formation appears to involve phosphorylation and inactivation of myosin light chain (MLC) phosphatase leading to increased MLC phosphorylation and enhanced myosin II contractility, resulting in tension on actin filaments and their subsequent alignment. Furthermore, RhoA-mediated activation of phosphatidylinositol-4-phosphate 5-kinase induces the formation of phosphatidylinositol (4,5)-bisphosphate, a widely acting phospholipid that promotes actin polymerization and cytoskeletal-membrane attachment (8,9). The culmination of these and other RhoA-mediated signaling events leads to bundling of actin filaments into actin cables or thick stress fibers and the clustering of their associated integrin receptors into focal adhesions (7). These cytoskeletal changes induced by RhoA leads to strong cell attachment points with the extracellular matrix (8).
The platelet contains relatively high levels of RhoA, along with other RhoA effectors (10). However, despite this, very little is known about the mechanisms of RhoA activation or its contribution to integrin-mediated adhesive responses in platelets. Recent studies have demonstrated RhoA activation in response to thrombin (11) and the thromboxane mimetic, U46619 (12), with activation of RhoA by the latter agonist reported to occur independent of integrin ␣ IIb ␤ 3 (12). The role of RhoA in regulating the adhesive function of integrin ␣ IIb ␤ 3 remains controversial (10,(13)(14)(15)(16). Although initial reports suggested a potentially important role in regulating integrin ␣ IIb ␤ 3 activation (10), more recent studies have not supported these conclusions (13,14,17). In general, these latter findings are in keeping with studies on cultured cells in which changes in the activation status of RhoA do not correlate with alterations in integrin affinity (18).
In the current study, we have investigated the role of RhoA in regulating the adhesive function of integrin ␣ IIb ␤ 3 . Our studies have demonstrated the existence of integrin ␣ IIb ␤ 3 -dependent and -independent mechanisms regulating RhoA activation. Although RhoA activation appears to represent a general feature of platelet activation, our studies do not support an important role for RhoA in regulating the affinity status of integrin ␣ IIb ␤ 3 . Rather, we demonstrate an important role for RhoA in regulating the stability of integrin ␣ IIb ␤ 3 adhesion contacts. The ability of RhoA to sustain integrin ␣ IIb ␤ 3 ⅐matrix interactions appears critical for efficient platelet adhesion in a shear field.
Preparation of Native Human von Willebrand Factor-Human von Willebrand factor (HvWf) was purified to homogeneity from plasma cryoprecipitate according to the method of Montgomery and Zimmerman (24).
Preparation of Asialo Human von Willebrand Factor-Asialo vWf was prepared from vWf that had been purified from a human dried Factor VIII Fraction (CSL Ltd., Victoria, Australia). Briefly, reconstituted Factor VIII fraction was applied to a Sepharose CL-6B sizeexclusion column, fractions were collected, and vWf was detected by SDS-PAGE and ristocetin cofactor activity. Fractions containing purified vWf were pooled and concentrated by centrifugation. Purified vWf (100 g/ml) was dialyzed in 0.05 M sodium acetate, 150 mM NaCl, 8 mM CaCl 2 , pH 6.0, at 4°C, followed by incubation with 0.1 unit of ␣2-3,6,8neuraminidase, Vibrio cholerae (Calbiochem-Novabiochem Corp.) for 3 h at 37 C. De-sialylated vWf was dialyzed in 10 mM Tris-HCl, 150 mM NaCl, pH 7.4, and platelet aggregating activity was confirmed by addition of 20 g/ml purified protein to citrated platelet-rich plasma.
Purification of Fibrinogen-Human fibrinogen was purified from fresh frozen plasma, according to the method of Jakobsen and Koerulf (25).
Preparation of Washed Platelets and Red Blood Cells-Whole blood (anticoagulated with acid-citrate-dextrose) was collected from healthy volunteers who had not received any anti-platelet medication in the preceding 2 weeks. Washed platelets were prepared as previously described (21) and resuspended in Tyrode's buffer (10 mM Hepes, 12 mM NaHCO 3 , pH 7.4, 137 mM NaCl, 2.7 mM KCl, 5 mM glucose) containing 1 mM calcium, where indicated. Autologous red blood cells (RBCs) were obtained as described previously (23).
Preparation and Treatment of Platelets with C3 Exoenzyme-A construct encoding recombinant C3 exoenzyme (a gift from Prof. Keith Burridge, University of North Carolina, Chapel Hill, NC) was expressed as a glutathione S-transferase (GST) fusion protein in Escherichia coli, followed by thrombin-mediated cleavage of the GST tag, as described by Dillon and Feig (26). Purified C3 exoenzyme was dialyzed against Tyrode's buffer overnight at 4°C. Platelets in platelet washing buffer (pH 6.5) were incubated with purified C3 exoenzyme (100 g/ml) in the presence of 200 units of hirudin for 4 h at room temperature. Using these conditions, we routinely achieved ϳ80% inhibition of RhoA (Fig.  3A), without significant loss of platelet reactivity (data not shown).
Preparation and Treatment of Platelets with Y27632-The Rho kinase inhibitor Y27632 was prepared as described previously by Uehata and colleagues (63). Washed platelets resuspended in Tyrode's buffer were incubated with vehicle alone (Me 2 SO) or Y27632 (20 M) for 10 min at room temperature.
ADP-ribosylation of RhoA by C3 Exoenzyme-ADP-ribosylation of RhoA in platelet extracts was analyzed using a modified method as described by Leng and colleagues, et al. (13). Briefly, platelets incubated with vehicle alone or the indicated concentrations of C3 exoenzyme, were lysed on ice by the addition of an equal volume of lysis buffer (20 mM Hepes, pH 7.4, 145 mM NaCl, 1.5% Triton X-100, 0.8% deoxycholate, 0.2% SDS, 3 mM EGTA, 2 mM MgCl 2 , 50 g/ml phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 10 g/ml aprotinin). Immediately prior to lysis, an equivalent amount of C3 exoenzyme was added to control platelets. In vitro ADP-ribosylation was performed immediately at 30°C for 60 min, by addition of 200 M GTP, 10 M NAD ϩ , 10 g/ml C3 exoenzyme, and 05-2.5 Ci/assay [ 32 P]NAD ϩ . Reactions were terminated by the addition of sample buffer, boiled, and analyzed on SDS-PAGE, followed by autoradiography.
Assay for GTP-bound RhoA-The level of GTP-bound RhoA in platelet lysates was measured as described previously (27,28). Western blots were quantified by densitometry (GelPro software), and RhoA activation expressed as -fold increase over the level of active RhoA in resting platelets, after correcting for protein loading (histograms).
Aggregation Studies-Washed platelets were stimulated with the indicated concentrations of agonist, in the presence of calcium (1 mM) and fibrinogen (500 g/ml). Aggregation initiated by thrombin was performed in the absence of fibrinogen. All aggregations were initiated by stirring the suspensions at 950 rpm for 10 min at 37°C in a fourchannel automated platelet analyzer (Kyoto Daiichi, Japan). The extent of platelet aggregation was defined as the percentage change in optical density as measured by the automated platelet analyzer.
Static Adhesion and Spreading Assays-Static adhesion assays were performed using a modified method of Yap et al. (23). Briefly, glass coverslips were coated with 10 g/ml HvWf for 2 h at room temperature, followed by blocking with 10% heat-inactivated human serum for 60 min at room temperature. Platelets were incubated with immobilized matrices for between 30 -60 min, unless otherwise indicated, and visualized using phase contrast microscopy. Images were captured, and the surface area of adherent platelets was determined using MCID software (Imaging Research Inc.). The number of C3 exoenzyme-treated adherent platelets and surface area of C3 exoenzyme-treated adherent platelets was expressed as a percentage of vehicle-treated control platelets.
In Vitro Flow-based Adhesion Assays-Flow assays were performed using glass microcapillary tubes (Microslides, Vitro Dynamics Inc.) coated with 100 g/ml HvWf, according to the method described by Yap et al. (23). Microcapillary tubes were blocked with 10% heat-inactivated human serum prior to use. Washed platelets reconstituted with RBCs (50% hematocrit) were perfused across coated microcapillary tubes at shear rates of 600 and 1800 s Ϫ1 . Platelet tethering and stationary adhesion was visualized in real time using differential interference contrast microscopy and the first 2 min of flow video recorded for off-line analysis. In all studies, off-line analyses of platelet tethering and stationary adhesion were performed as described previously (23). Briefly, any cell forming an adhesion contact with immobilized HvWf for greater than 40 ms was scored as a tethered cell. Stationary adhesion was arbitrarily defined as cells not moving more than a single cell diameter over a 10-s period. The duration of stationary platelet adhesion contacts was determined on individual platelets that had formed stationary interactions with immobilized vWf for a period of Ն1 s. Individual cells were analyzed every second for a period up to 30 s, and stationary adhesion was arbitrarily defined as cells not moving more than a single cell diameter over a 1-s period. For analysis of platelet spreading, adherent platelets were washed for 15 min with Tyrode's buffer (1800 s Ϫ1 ), visualized using phase contrast microscopy, and video recorded for off-line analysis. Platelet spreading was analyzed as described for the static adhesion assays.
Shear-induced Platelet Aggregation Studies-Control or Y27632-or C3 exoenzyme-treated platelets, resuspended in Tyrode's buffer, pH 7.5, and containing 1 mM CaCl 2 , were subjected to shear (5000 s Ϫ1 ) in the presence of 20 g/ml HvWf for 5 min, using a cone-and-plate viscometer, as described previously (29). Platelets were visualized using inverse phase contrast microscopy (ϫ5 objective), and images were captured using a charge-coupled device camera. Five random fields were captured from each experimental sample, and quantification of single platelets was performed using MCID software.
Immunofluorescence Microscopy-PAC1 immunofluorescence was performed under static and flow conditions, as described previously (23). Platelets were imaged via confocal microscopy, using ϫ100 oil immersion, and quantitation was performed using Leica software.
Statistical Analysis-Significant differences were detected using an unpaired t test with one-way analysis of variance, using Prism software (GraphPad Software for Science, San Diego, CA).

RESULTS
Integrin ␣ IIb ␤ 3 -dependent and Independent Pathways Regulating RhoA Activation-Platelets respond to a multitude of activating stimuli in vivo, including soluble agonists generated at the site of vascular injury, adhesive substrates present within the damaged vessel wall, and hemodynamic forces generated by the flow of blood. In an attempt to gain further insight into the functional role of RhoA in platelets, we initially examined whether RhoA activation was restricted to a subset of platelet-activating stimuli or occurred as a general feature of platelet activation. Initially, platelets were stimulated in suspension with a diverse range of soluble agonists, including potent agonists such as thrombin, collagen, and the thromboxane A 2 mimetic, U46619, or weaker agonists such as ADP. To monitor RhoA activation directly, we utilized an assay that selectively isolates GTP-bound active RhoA from cell lysates (27) by precipitation with the RhoA binding domain of the downstream effector Rhotekin (GST-RBD), as described under "Experimental Procedures." Consistent with previous reports (11,12), RhoA activation was induced by thrombin (1 unit/ml) and U46619 (500 nM) (Fig. 1, left). In addition, RhoA activation was also induced by collagen (10 g/ml) and ADP (25 M), but with the latter agonist the rate and extent of RhoA activation was substantially less than with the other agonists (Fig. 1A, and data not shown).
In cultured cells, the activation status of RhoA is regulated by soluble growth factors and integrin adhesion receptors (18,30). However, to date, there is no evidence that platelet integrins are involved in regulating RhoA. To examine the potential involvement of the major platelet integrin ␣ IIb ␤ 3 in this process, platelets were pretreated with the integrin ␣ IIb ␤ 3blocking antibody c7E3 Fab, prior to the performance of platelet aggregation studies. In thrombin-stimulated platelets, c7E3 Fab-pretreatment resulted in a reduction of ϳ70% in RhoA activation (Fig. 1B), whereas with U46619, collagen, and ADP, integrin ␣ IIb ␤ 3 inhibition resulted in Ͼ90% inhibition of RhoA activation. Our findings with respect to U46619 are in direct contrast to those recently reported by Gratacap and colleagues (12) who demonstrated that integrin ␣ IIb ␤ 3 blockade with RGDS peptides had no inhibitory effect on RhoA activation. To investigate potential explanations for this difference, we compared the effects of RGDS peptide and c7E3 Fab on the activation of RhoA induced by U46619 under identical experimental conditions employed by Gratacap et al. (12). Consistent with previous findings (12), we did not detect any inhibitory effect of RGDS peptide on RhoA activation induced by U46619, however, c7E3 Fab dramatically inhibited RhoA activation under identical experimental conditions (Fig. 1C). Findings similar to c7E3 Fab were observed with another potent integrin ␣ IIb ␤ 3 antagonist, aggrastat (data not shown). These findings suggest that RGDS peptide may not completely block ligand binding to integrin ␣ IIb ␤ 3 , a finding consistent with previous reports (31).
To examine whether RhoA activation represents a general feature of platelet activation, changes in RhoA activity were assessed following platelet adhesion to immobilized substrates and in platelets exposed to high shear stress. As demonstrated in Fig. 2, adhesion and spreading of platelets on either a fibrinogen (100 g/ml) or human vWf (10 g/ml) matrix was associated with significant RhoA activation, similar to the levels observed with soluble agonist stimulation. Pretreating platelets with c7E3 Fab prior to adhesion to vWf inhibited RhoA activation by Ͼ95% (Fig. 2B). To examine the effects of high fluid shear stress on RhoA activation, shear-induced platelet activation (SIPA) studies were performed using a cone-andplate viscometer, as described under "Experimental Procedures." In control studies, we demonstrated that high shear rates (typically Ն3000 s Ϫ1 ) were required to induce aggregation of washed platelets independent of the addition of exogenous stimuli. Furthermore, aggregation under these experimental conditions was dependent on vWf binding to GPIb and integrin ␣ IIb ␤ 3 , because it was completely inhibited by pretreating platelets with blocking antibodies against either receptor (data not shown). As demonstrated in Fig. 2A, exposure of platelets in suspension to high shear (5000 s Ϫ1 ) in the presence of purified soluble human vWf, induced platelet aggregation and robust RhoA activation. In contrast, exposure of platelets to relatively low levels of shear (Ͻ1000 s Ϫ1 ) failed to induce RhoA activation (data not shown), consistent with the requirement for a threshold level of shear to induce platelet activation independent of other exogenous stimuli. Similar to the findings on immobilized vWf, blocking ligand binding to integrin ␣ IIb ␤ 3 with c7E3 Fab abolished RhoA activation (Fig. 2B). Taken together, these studies suggest that RhoA activation represents a general feature of platelet activation. Furthermore, they demonstrate a major role for integrin ␣ IIb ␤ 3 in regulating the activation state of RhoA.
Effect of Inhibiting RhoA on Platelet Adhesion and Aggregation-The role of RhoA in regulating platelet aggregation remains controversial. Several studies have demonstrated inhibition of platelet aggregation in the presence of the specific RhoA inhibitor, C3 exoenzyme (10,16), or the Rho kinase inhibitor Y27632 (15), whereas others have failed to show any inhibitory effect (13). To investigate the role of RhoA in regulating platelet aggregation, platelets were pretreated with C3 exoenzyme. Incubation of platelets with increasing concentrations of C3 exoenzyme (0 -100 g/ml) demonstrated a dose-dependent ADP-ribosylation of RhoA, with ϳ80% inhibition of this small GTPase following incubation with 100 g/ml C3 exoenzyme for 4 h at room temperature (Fig. 3A). In all subsequent experiments, high concentrations (100 g/ml) of C3 exoenzyme were used to inhibit RhoA. As demonstrated in Fig.  3B, pretreating platelets with C3 exoenzyme had no inhibitory effect on platelet aggregation induced by thrombin or the PAR1 specific agonist, TRAP. Furthermore, the same concentration of C3 exoenzyme had no effect on the ability of these agonists to induce binding of the activation-specific integrin ␣ IIb ␤ 3 antibody, PAC1 (Fig. 3C). Similar findings were observed using the Rho kinase inhibitor Y27632 (20 M) (Fig. 3B and data not  shown). Moreover, neither C3 exoenzyme nor Y27632 had any inhibitory effect on platelet aggregation induced by collagen, U46619, or ADP (Fig. 3B), demonstrating that RhoA is unlikely to play a major role in regulating integrin ␣ IIb ␤ 3 activation induced by soluble agonists.
We next examined the consequences of inhibiting RhoA activity on platelet adhesion, shape change, and spreading on immobilized fibrinogen or HvWf. Previous studies have demonstrated a potentially important role for RhoA in regulating platelet shape change in response to soluble agonist stimulation (11,(32)(33)(34). As demonstrated in Fig. 4 (A and B), inhibiting RhoA or RhoA kinase had no effect on platelet adhesion, shape change, and spreading on immobilized vWf or on the ability of these cells to bind PAC-1 (Fig. 4C). Similar findings were apparent on a fibrinogen matrix (data not shown). Platelet adhesion and spreading on these matrices is associated with dramatic reorganization of the cytoskeleton, leading to the formation of two distinct actin-based structures, filopodial bundles and broad lamellipodial networks (35,36). Although Cdc42 and Rac are well known to regulate filopodial and lamellipodial formation, respectively, some degree of RhoA activity is also required for these cytoskeletal changes in a number of cell types (37). To investigate the effect of inhibiting RhoA on actin cytoskeletal remodeling, C3 exoenzyme-or Y27632-treated platelets were fixed following adhesion to immobilized vWf and F-actin stained with tetramethyl rhodamine isothiocyanateconjugated phalloidin. In contrast to the well-documented inhibitory effects of RhoA/Rho kinase inhibitors on cytoskeletal reorganization of cultured motile cells, both C3 exoenzyme-and Y27632-treated platelets exhibited a similar pattern of F-actin staining to control platelets (data not shown). Collectively, these studies do not support a major role for the RhoA signaling pathway in regulating integrin ␣ IIb ␤ 3 activation and cytoskeletal reorganization in spreading platelets.
Effect of RhoA Inhibitors on Shear-induced Platelet Aggregation-To investigate the effects of RhoA signaling inhibitors on shear-induced platelet aggregation (SIPA), Y27632 or C3 exoenzyme-treated platelets were sheared for 5 min in a coneand-plate viscometer. Inhibiting RhoA or Rho kinase had a major effect on SIPA. As demonstrated in Fig. 5, both Y27632 and C3 exoenzyme prevented the formation of large platelet aggregates with the majority of cells remaining as single platelets or small aggregates (Ͻ10 cells). SIPA is a complex process initiated by shear-induced binding of soluble vWf to the GPIb/ V/IX, leading to ADP release and the activation of integrin ␣ IIb ␤ 3 (38). In an attempt to gain further insight into the mechanism by which RhoA regulates SIPA, we investigated the effects of the RhoA inhibitors on platelet aggregation induced by asialo-vWf. This latter platelet agonist utilizes a similar multistep adhesion process as SIPA, except that it does not require high shear or artificial modulators to initiate platelet activation, because removal of the sialic acid residues from vWf results in spontaneous vWf binding to GP Ib. Similar to shearinduced platelet aggregation, As-vWf binding to GP Ib induces platelet aggregation in an ADP-and integrin ␣ IIb ␤ 3 -dependent manner (39,40). To investigate the role of RhoA in regulating As-vWf-induced platelet aggregation, C3 exoenzyme-treated platelets were resuspended in platelet-poor plasma, and aggregation induced by addition of As-vWf, as described under "Experimental Procedures." As demonstrated in Fig. 5B, C3treated platelets were able to aggregate at a similar rate and extent as control platelets stimulated with As-vWf, with no  4. Examination of a role for RhoA and Rho kinase in platelet adhesion, spreading, and integrin ␣ IIb ␤ 3 activation under static conditions. Washed platelets pretreated with vehicle alone (Control) or C3 exoenzyme (100 g/ml, C3), were applied to coverslips coated with vWf (10 g/ml), as described under "Experimental Procedures." C, in some experiments, adhesion was carried out in the presence of PAC-1 antibody (1 g/ml). Adhesion (A) and spreading (B) were examined using differential interference contrast microscopy (A, ϫ100 oil objective; B, ϫ100 oil objective with 4ϫ zoom). C, PAC-1 binding was quantified using confocal microscopy (ϫ100 oil objective, with 2ϫ zoom). Images are taken from one experiment representative of three. Histograms represent the means Ϯ S.E. from three separate experiments (n ϭ 3).

FIG. 5. Examination of a role for RhoA and Rho kinase in shear-induced platelet aggregation (SIPA). A, washed platelets
were preincubated with C3 exoenzyme (100 g/ml) or Y27632 (20 M). Pretreated platelets were sheared at 5000 s Ϫ1 for 5 min using a coneand-plate viscometer. Single platelets and aggregates were imaged using phase-contrast microscopy (ϫ5 objective). Images are taken from one experiment representative of three. SIPA was quantified by capturing five random fields from each experiment. The number of single platelets in each field was quantified offline using computer-assisted analysis (MCID software). Results represent the means Ϯ S.E. of three independent experiments. Statistical analysis was performed using a t test comparing control versus C3 exoenzyme-or Y27632-treated platelets (*, p Ͻ 0.05; **, p Ͻ 0.01). B, effect of RhoA on asialo-vWf-mediated aggregation. Washed platelets preincubated with either vehicle alone or C3 exoenzyme (100 g/ml) were stirred in the presence of asialo-vWf (20 g/ml), and aggregation was monitored in a four-channel automated platelet analyzer. The aggregation tracing is taken from one experiment representative of five. Tabulated results represent the means Ϯ S.E. of five separate experiments (n ϭ 5). tions, we examined the effect of C3 exoenzyme and Y27632 on platelet adhesion to immobilized vWf using an in vitro flowbased platelet adhesion assay (23). A major advantage of this assay system is that it allows detailed real-time analysis of each step of the platelet adhesion process, providing information on changes in dynamics of the vWf⅐GPIb and vWf⅐integrin ␣ IIb ␤ 3 adhesion contacts. In these studies, washed platelets were preincubated with either vehicle, C3 exoenzyme or Y27632, then reconstituted with RBCs prior to perfusion through vWf-coated microcapillary tubes at 600 or 1800 s Ϫ1 , as described under "Experimental Procedures." Analysis of the dynamics of individual platelet⅐matrix contacts demonstrated that the majority of platelets were able to tether normally to HvWf, regardless of pretreatment by C3 exoenzyme (Fig. 6A, line graph) or Y27632 (Fig. 6B, line graph). In contrast, the number of platelets forming sustained stationary adhesion contacts with the vWf matrix was significantly reduced by C3 exoenzyme at 600 s Ϫ1 (Fig. 6A) and 1800 s Ϫ1 (data not shown). Similar findings were observed with Y27632 (Fig. 6B, and data not shown).
To investigate whether the decrease in the ability of C3 exoenzyme-treated platelets to form stable adhesion contacts was due to a decrease in integrin ␣ IIb ␤ 3 activation, we compared the binding of PAC-1 to control or C3 exoenzyme-or Y27632-treated platelets following adhesion to HvWf at 1800 s Ϫ1 . These studies demonstrated no difference in PAC-1 binding to platelets treated with either inhibitor (data not shown). However, a potential limitation of these studies was the possibility that we were biasing our results by only selecting a subset of cells for analysis, i.e. those cells that had activated sufficient levels of integrin ␣ IIb ␤ 3 to form stable adhesion contacts. To overcome this technical limitation, flow-based adhesion studies were performed on a human vWf matrix in the presence of ristocetin, or alternatively, on a purified bovine vWf matrix. Ristocetin promotes stationary platelet adhesion in a shear field independent of integrin ␣ IIb ␤ 3 , due to the ability of ristocetin to increase the affinity of the GPIb⅐vWf interaction. Similarly, bovine vWf has a high affinity for human GPIb, allowing sustained stationary platelet adhesion independent of integrin ␣ IIb ␤ 3 (data not shown). In contrast to human vWf alone, inhibiting RhoA or Rho kinase had no effect on the number of platelets forming stationary adhesion contacts on either bovine vWf or human vWf in the presence of ristocetin (data not shown), consistent with the lack of effect of RhoA signaling on the adhesive function of GPIb. Furthermore, neither C3 exoenzyme nor Y27632 had any inhibitory effect on the level of PAC-1 binding to these cells (Fig. 7). These studies do not support an important role for RhoA in regulating the affinity status of integrin ␣ IIb ␤ 3 under shear conditions.
To investigate the possibility that RhoA modulates firm platelet adhesion by influencing the stability of the integrin ␣ IIb ␤ 3 ⅐vWf interaction, we examined the duration of stationary platelet adhesion contacts with the human vWf matrix in control or C3 exoenzyme-or Y27632-treated platelets. In initial control studies, we demonstrated that blocking the ligand binding function of integrin ␣ IIb ␤ 3 by pretreating platelets with the integrin ␣ IIb ␤ 3 antagonist, c7E3 Fab, reduced stationary adhesion contact formation beyond 1 s down to 10 -20% of cells. The ability of a small percentage of platelet to remain stationary longer than 1 s under these experimental conditions was due to the formation of thin membrane tethers (41), rather than incomplete blockade of integrin ␣ IIb ␤ 3 . In contrast, ϳ60 -65% of control, C3 exoenzyme-, and Y27632-treated platelets tethering to the vWf surface formed stationary adhesion contacts for at least 1 s. The lack of inhibitory effect of C3 exoenzyme or Y27632 on initial adhesion contact formation is consistent with the ability of integrin ␣ IIb ␤ 3 to become activated and engage vWf under these experimental conditions. Analysis of the proportion of platelets maintaining stationary adhesion contacts over time revealed a significant difference between control, C3 exoenzyme-, and Y27632-treated platelets. As demonstrated in Fig. 8, in control platelets there was a time-dependent reduction in the percentage of platelets remaining stationary from 1 to 10 s with ϳ25% sustaining firm adhesion contacts for at least 10 s. In contrast, with both C3 exoenzyme-and Y27632treated platelets, Ͻ10% of cells sustained stable adhesion con- Pretreated cells were reconstituted with RBCs prior to perfusion through vWf-coated microcapillary tubes (100 g/ml) (1800 s Ϫ1 ). Platelets were incubated with PAC-1 antibody prior to fixation and staining with a fluorescein isothiocyanate-conjugated secondary antibody. PAC-1 binding was imaged using a confocal microscope (ϫ100 oil objective 2ϫ zoom). Images were taken from one experiment representative of three. The histogram represents the means Ϯ S.E. from three separate experiments.
tacts with the vWf matrix for up to 10 s. These results demonstrate an important role for RhoA in maintaining the strength of integrin ␣ IIb ␤ 3 adhesion contacts with vWf under shear conditions. DISCUSSION Platelet thrombus formation is critically dependent upon the adhesive and signaling function of integrin ␣ IIb ␤ 3 . Despite considerable progress in defining the mechanisms initiating inte- grin ␣ IIb ␤ 3 activation and adhesion (14,23,(42)(43)(44)(45)(46)(47)(48), there remains limited information on the factors involved in sustaining integrin ␣ IIb ␤ 3 adhesion contacts, especially under pathophysiologically relevant shear conditions. The studies in this manuscript demonstrate an important role for the small GTPase RhoA in this process. Our studies demonstrate that RhoA activation occurs as a general feature of platelet activation, and, with the majority of platelet stimuli examined, integrin ␣ IIb ␤ 3 outside-in signaling appears to play a major role in regulating RhoA activity. Analysis of the effects of C3 exoenzyme or Y27632 on platelet function, over a broad range of experimental conditions, does not support an important role for RhoA or Rho kinase in regulating the affinity status of integrin ␣ IIb ␤ 3 . Rather, our studies suggest a potentially important role for RhoA signaling in stabilizing integrin ␣ IIb ␤ 3 adhesion contacts, thereby promoting efficient platelet adhesion and aggregation in a shear field.
Our studies have demonstrated the existence of integrin ␣ IIb ␤ 3 -dependent and -independent signaling pathways linked to RhoA activation. The relative importance of these two pathways for RhoA activation appears to be agonist-specific. For example, in thrombin-stimulated platelets, a substantial proportion of RhoA activation occurs independent of ligand binding to integrin ␣ IIb ␤ 3 , whereas, with ADP, collagen, vWf, and U46619, integrin ␣ IIb ␤ 3 signaling appears to play the predominant role in promoting RhoA activation, at least under the experimental conditions employed in this study. Our findings with regards to U46619 appear to contradict a recent study by Gratacap et al. (12), who demonstrated that RhoA activation by U46619 occurred independent of ligand binding to integrin ␣ IIb ␤ 3 . However, as demonstrated in the current study, and from other previous reports (31), the exclusion of integrin ␣ IIb ␤ 3 from platelet signaling events must be interpreted with caution when based on studies utilizing RGDS peptides in isolation. This technical issue aside, it is quite possible, and likely, that U46619 and other agonists utilize integrin ␣ IIb ␤ 3 -independent mechanisms to promote RhoA activation. For example, several previous studies have demonstrated a potentially important role for RhoA in the regulation of integrin ␣ IIb ␤ 3 -independent functional responses, such as platelet shape change (11,32,34,49). Consistent with this, thrombin induces rapid RhoA/Rho kinase activation (11), 2 the timing of which mirrors MLC phosphorylation and platelet shape change (11). Furthermore, there is evidence supporting a direct role for G-protein-coupled receptors in the regulation of RhoA activity independent of integrins (50 -53). Although the precise mechanisms regulating RhoA activity in platelets remain to be fully elucidated, the existence of integrin ␣ IIb ␤ 3 -dependent and -independent pathways in this process provides for distinct spatial and temporal roles for RhoA in the modulation of platelet activation and adhesion responses.
Our findings that RhoA activity in platelets is regulated by both soluble agonists and integrin ␣ IIb ␤ 3 is consistent with a large volume of evidence in cultured cells that RhoA activation is modulated through a complex interplay between soluble stimuli and integrins (27, 54 -58). The upstream signaling elements involved in RhoA activation remain a complex and relatively undefined area. RhoA, like all Rho family members, is activated by guanine-nucleotide exchange factors (GEFs) (59). Exchange factor activity can be regulated by a variety of signaling events, depending upon the nature of the stimulus. Many of these signaling events remain to be determined, however, some progress has been made into the mechanism of RhoA GEF activation mediated by G-protein-coupled receptors.
Studies in mammalian cells have demonstrated that G 12/13 is capable of RhoA activation, via the direct binding of G␣ 12/13 to p115 Rho GEF (50 -52), leading to an increase in its exchange factor activity (51). Whether this same signaling pathway exists in platelets remains to be determined; however, studies by Klages et al. (32) have demonstrated that selective G 12/13 activation is sufficient to induce Rho/Rho kinase-mediated phosphorylation of MLC. Activation of RhoA by integrins remains poorly understood (60), although there is evidence for a potentially important role for tyrosine kinases such as Src and Syk in this process (61,62).
Our findings do not suggest a major role for RhoA in regulating cytoskeletal changes necessary for platelet shape change and spreading on adhesive substrates, a finding consistent with previous studies by Leng et al. (13). This may not be surprising, given that the cytoskeletal changes associated with these events, such as formation of actin bundles in filopodia and actin networks in lamellipodia, have been attributed to the activation of other Rho family members, such as Cdc42 and Rac, respectively. However, in other cells types abolishing RhoA signaling has been demonstrated to inhibit cell spreading (37). For example, under experimental conditions leading to complete inhibition of RhoA, cell adhesion contact formation with the matrix is significantly weakened, thereby undermining the ability of cells to efficiently spread. In addition, changes in RhoA activity have been demonstrated to indirectly influence the activation status of other Rho family members (8). In platelets, the lack of effect of C3 exoenzyme on cytoskeletal reorganization and platelet spreading may be due to incomplete inhibition of RhoA, or alternatively, inhibiting RhoA in platelets may not significantly influence the activation state of other Rho family members. Future studies will be required to address this issue.
A major finding from the present study is the requirement for RhoA signaling for maintaining integrin ␣ IIb ␤ 3 adhesion contacts in a shear field. Our studies do not support an important role for RhoA in regulating integrin ␣ IIb ␤ 3 activation, a finding consistent with observations in many other cell types. The precise mechanism by which RhoA sustains integrin ␣ IIb ␤ 3 adhesion contacts remains to be established, however, based on findings in other cells, it is likely that RhoA plays an important role in regulating integrin ␣ IIb ␤ 3 avidity. In cultured cells, activation of RhoA results in the development of large integrin clusters that are anchored to actin stress fibers, forming focal adhesion contacts (FAs). The ability of RhoA to induce focal contact assembly has been extensively investigated and involves downstream targets, including Rho kinase. In its activated form, this serine/threonine kinase catalyzes the phosphorylation of myosin phosphatase, down-regulating its activity and elevating MLC phosphorylation, thereby promoting contractility, integrin clustering, and FA formation (7). Although platelets do not contain typical FAs such as those observed in fibroblasts, they do contain "focal adhesion-like" complexes, which share many of the components of true FAs. In support of a role for RhoA in regulating these structures, Leng and colleagues (13) have previously reported that C3 exoenzyme treatment of platelets reduces the number of vinculin-containing adhesion complexes in spread platelets.
Finally, our studies provide new insight into the importance of integrin ␣ IIb ␤ 3 outside-in signaling for the process of shearinduced platelet aggregation. More specifically, we have demonstrated that integrin ␣ IIb ␤ 3 -dependent RhoA activation following platelet adhesion to vWf is essential for stabilizing integrin ␣ IIb ␤ 3 adhesion contacts. These findings have potentially important implications for the regulation of platelet adhesive behavior in vivo. For example, pathological thrombi typically form at sites of arterial narrowing where shear forces may be as high as 380 dyne/cm 2 (10,000 s Ϫ1 ). The demonstration that Rho signaling has a relatively selective role in regulating integrin ␣ IIb ␤ 3 adhesiveness under shear conditions raises the interesting possibility that Rho inhibitors may reduce platelet thrombus formation in vivo. The ability to inhibit integrin ␣ IIb ␤ 3 adhesiveness, without potentially undermining the normal signaling mechanisms regulating integrin ␣ IIb ␤ 3 activation, would represent a novel approach for regulating platelet adhesiveness in vivo.