|Home | About | Journals | Submit | Contact Us | Français|
The platelet receptor for von Willebrand factor (VWF), the glycoprotein Ib-IX (GPIb-IX) complex, mediates platelet adhesion at sites of vascular injury and transmits signals leading to platelet activation. VWF/GPIb-IX interaction sequentially activates the Src Family Kinase (SFK) Lyn, phosphoinositide 3-kinase (PI3K) and Akt, leading to activation of integrin αIIbβ3, and integrin-dependent stable platelet adhesion and aggregation. It remains unclear how Lyn activates the PI3K/Akt pathway following ligand binding to GPIb-IX.
Using platelet-specific Rac1−/− mice and the Rac1 inhibitor NSC23766, we examined the role of Rac1 in GPIb-IX-dependent platelet activation. Rac1−/− mouse platelets and NSC23766-treated human platelets were defective in GPIb-dependent stable adhesion to VWF under shear stress, integrin activation, thromboxane A2 (TXA2) synthesis and platelet aggregation. Interestingly, GPIb-induced activation of Rac1 and the guanine nucleotide exchange factor (GEF) for Rac1,Vav, was abolished in both Lyn−/− and PP2-treated platelets but was unaffected by the PI3K inhibitor LY-294002, indicating that Lyn mediates activation of Vav and Rac1 independently of PI3K. Furthermore, GPIb-induced activation of Akt was abolished in Rac1-deficient platelets, suggesting that Rac1 is upstream of the PI3K/Akt pathway.
A Lyn/Vav/Rac1/PI3K/Akt pathway mediates VWF-induced activation of integrin αIIbβ3 to promote GPIb-IX-dependent platelet activation.
Under the high shear rate flow conditions present in arteries and capillaries, platelet adhesion to the site of vascular injury is mediated by the interaction between subendothelial-bound VWF and its platelet receptor, the GPIb-IX complex.1–3 The interaction between VWF and GPIb-IX not only mediates transient platelet adhesion to the injured vessel wall but also initiates a signal transduction cascade culminating in the activation of integrin αIIbβ3, leading to stable platelet adhesion, spreading, and aggregation.2, 4–6 This process is essential for thrombosis under high shear conditions. GPIb-IX-mediated platelet activation signaling is known to involve several intracellular signaling molecules and pathways, including the SFK Lyn, the PI3K/Akt pathway, and the cGMP and mitogen-activated protein kinase (MAPK) pathways.2, 7–16 GPIb-IX signaling is also greatly amplified by the activation of the immunoreceptor tyrosine-based activation motif (ITAM), TXA2, and adenosine diphosphate (ADP) signaling pathways 2, 8, 17, 18. Although the importance of GPIb-IX-mediated platelet activation to arterial thrombosis is well understood, the signaling pathway that regulates GPIb-IX-dependent platelet activation is poorly characterized. Thus far, the identified most proximal step to GPIb-IX that propagates platelet activation signals is the activation of Lyn and Lyn-dependent activation of the PI3K/Akt pathway. Interestingly, although the PI3K/Akt pathway is activated downstream of Lyn and is critical for promoting GPIb-IX-mediated platelet activation, the exact molecular mechanisms governing this process are unclear.
The Rho family small GTPase (guanosine triphosphatase), Rac1, has been shown to be important for arteriole thrombosis in vivo.19, 20 Rac1 is activated by binding to guanosine triphosphate (GTP) and plays roles in multiple cellular processes, including actin polymerization21, lamellipodia formation19, 22–24 and cell retraction25. However, the role of Rac1 in GPIb-IX-mediated signal transduction has not been explored.
Here we show that Rac1 is a critical mediator of GIb-IX-dependent early signaling leading to integrin activation, stable platelet adhesion under shear stress, TXA2 production, and platelet aggregation. Most importantly, we demonstrate that Rac1 mediates Lyn-dependent activation of the PI3K/Akt signaling pathway during GPIb-IX-dependent platelet activation. Thus, our study reveals an important new link in the GPIb-IX signaling pathway and a novel mechanism of Rac1-dependent PI3K/Akt activation.
Mice containing the Rac1 conditional allele (Rac1loxP/loxP) 26 were crossed with mice carrying the Pf4-Cre transgene (Pf4-Cre+).27 Pf4-Cre+/Rac1loxP/loxP mice are notated as Rac1−/− and Pf4-Cre−/Rac1loxP/loxP mice as wild type (WT). Mice were maintained on a mixed SV/129/C57/Bl-6 background and littermates used as control. Animal usage and protocol were approved by the institutional animal care committee of the University of Illinois at Chicago.
Human and murine platelets were prepared as previously described and used at 3 × 108/mL. 28, 29 Analysis of platelet adhesion under flow was performed as described previously. 11, 12 Data are representative of ≥ 3 experiments and statistical significance was determined via ANOVA and post-test.
Buffers contained 1% BSA and 2 mM MgCl2. Washed human (1 × 108/mL) and mouse (2 × 108/mL) platelets were stimulated with ristocetin/botrocetin ± VWF in the presence of 5 or 2 μg/mL Oregon Green-labeled fibrinogen (Oregon Green-Fg: Molecular Probes) for 20 or 10 min., respectively. Integrilin and RGDS were used to determine the level of non-specific binding of Oregon Green-Fg to the platelet surface. Platelets were diluted 1:20 in Modified-Tyrode’s buffer and analyzed with a BD Accuri C6 flow cytometer. Specific fibrinogen binding was determined by subtracting the geometric means of fluorescence intensity of the nonspecific binding (integrin antagonist present) from the total binding (integrin antagonist absent).
2 mM MgCl2, 1% BSA, and 10 mM EDTA were added to all buffers and 1 × 108 platelets/mL were used. Washed human and mouse platelets were incubated 5 or 10 min. with VWF ± ristocetin/botrocetin, respectively. Platelets were fixed with a final concentration of 1% paraformaldehyde (PFA) in HEPES-saline buffer for 20 min., centrifuged at 100 g for 1 min., and the pellet stained for 30 min. in 0.1 mL Modified-Tyrode’s buffer containing 4 μg/mL SZ-29-FITC. Samples were diluted 1:20 with Modified-Tyrode’s buffer and analyzed using a BD Accuri C6 flow cytometer. Specific VWF binding was determined by subtracting the geometric means of fluorescence intensity of VWF binding in the presence of botrocetin/ristocetin alone from the VWF binding in the absence of botrocetin/ristocetin.
Platelet aggregation 30 and measurement of the TXA2 metabolite, TXB2 15, 17 was performed as described previously. Briefly, following platelet aggregation for 8 minutes, the reaction was stopped with EDTA and aspirin, the platelets briefly spun down in a microfuge, and the resultant supernatant used in a TXB2 ELISA (Assay Designs). TXB2 is used to indicate levels of TXA2. Data are from ≥ 3 experiments and expressed as mean ± standard error of the mean (SEM). Statistical significance was determined using the Student’s t-test.
SDS-PAGE analysis of phosphorylated signaling proteins was performed as described previously.12 Anti-pan phospho SFK (#2101: Tyr416), which reacts with phosphorylated Lyn, anti-phospho P38 (#9211: Thr180/Tyr182), anti-phospho Akt (#4058: Ser473), and anti-pan Akt (#2920) antibodies are from Cell Signaling Technology. Anti-pan phospho Vav (sc-16408-R), anti-pan Vav (sc-55482), anti-Lyn (sc-7274) and anti-P38 (sc-535) antibodies are from Santa Cruz Biotechnology.
In order to study the role of Rac1 in GPIb-IX-induced platelet activation, we generated mice that lack expression of Rac1 exclusively in megakaryocytes and platelets and used a small molecule inhibitor of Rac1, NSC23766. To obtain megakaryocyte- and platelet-specific conditional Rac1 knockout mice, Pf4-Cre transgenic mice27 were cross-bred with mice containing the Rac1loxP/loxP allele.26 Deletion of Rac1 in Rac1−/− mouse platelets was verified via Western Blot analysis of platelet lysates (Figure 1A). Rac1−/− mouse platelets did not show any noticeable difference in routine hematological parameters including morphology and counts compared to WT platelets. This is similar to previous reports, where Rac1−/− mouse platelets were generated using the Mx1-Cre transgene.19, 20
We initially evaluated the effect of Rac1 deficiency on GPIb-IX-induced, integrin-dependent stable platelet adhesion to VWF under shear stress, which was created using a cone-plate rheometer. As shown in Figure 1, stable adhesion of human and mouse platelets to VWF was completely inhibited by the integrin antagonist RGDS (Figure 1B–D). This is consistent with our previous results that stable platelet adhesion to VWF under shear flow conditions requires not only GPIb-IX but also GPIb-IX-dependent activation of integrin αIIbβ3.11, 12. Stable platelet adhesion to VWF under flow is partially inhibited by the cyclooxygenase (COX) inhibitor aspirin, which is also consistent with our previous data that the COX pathway plays an amplifying role in promoting platelet adhesion under shear stress.11, 12 Interestingly, stable platelet adhesion to VWF was dramatically impaired in Rac1−/− mouse platelets or NSC23766-treated human platelets (Figure 1B–D). These data indicate that Rac1 plays a critical role in GPIb-IX-mediated, integrin-dependent platelet adhesion to VWF under shear stress. Importantly, deficiency of Rac1 led to significantly greater inhibition of stable platelet adhesion than aspirin (Figure 1B–D), indicating that the role of Rac1 in GPIb-IX-mediated platelet activation exceeds the COX-dependent amplification signaling pathway.
We next investigated whether Rac1 is important for GPIb-IX-induced integrin activation. WT and Rac1−/− mouse platelets were stimulated with VWF in the presence of botrocetin, and analyzed for the binding of Oregon Green-Fg using flow cytometry. VWF/botrocetin-induced fibrinogen binding was diminished in Rac1−/− mouse platelets compared to WT platelets, indicating that Rac1−/− mouse platelets were defective in GPIb-IX-induced activation of αIIbβ3 (Figure 2A, B). These results were recapitulated in human platelets, where VWF/ristocetin-induced fibrinogen binding was abolished by Rac1 inhibitor-treatment, as compared to the control solvent DMSO (Figure 2C, D). These results demonstrate that Rac1 plays an important role in the signaling pathway of GPIb-IX-dependent integrin activation.
To exclude the possibility that the functional defects observed in Rac1-deficient platelets derive from a defect in the VWF-binding function of GPIb-IX, washed WT and Rac1−/− mouse platelets or washed human platelets treated with DMSO or NSC23766 were incubated with human VWF in the presence of botrocetin or ristocetin, respectively, to induce the binding of VWF to GPIb-IX. The platelets were then fixed, stained with a FITC-labeled anti-VWF antibody, and analyzed via flow cytometry. As expected, botrocetin or ristocetin induced the binding of soluble VWF to platelets. There was no difference in botrocetin-induced binding of VWF to WT and Rac1−/−mouse platelets (Figure 3A, B) or ristocetin-induced binding of VWF to DMSO- and NSC23766-treated human platelets (Figure 3C, D). Therefore, Rac1 is not involved in regulating the VWF binding function of GPIb-IX, but rather is important in GPIb-IX-dependent signal transduction.
In order to further confirm the role of Rac1 in GPIb-IX-mediated platelet activation, we also evaluated the effect of Rac1 deficiency on VWF-induced platelet aggregation. VWF-induced platelet aggregation is characterized by two distinct waves.2, 8 The first wave mainly comprises VWF/GPIb-IX-dependent platelet agglutination and an often minor component of GPIb-IX-induced integrin-dependent platelet aggregation.9, 32 The second wave represents TXA2-, secretion- and integrin-dependent platelet aggregation. It requires GPIb-IX-induced activation of integrin αIIbβ3 and subsequent TXA2 synthesis, which initiates integrin- and TXA2-dependent granule secretion and secretion-dependent amplification of platelet activation. As expected, stimulation of WT mouse platelets with VWF/botrocetin or stimulation of human platelets with VWF/ristocetin lead to two waves of platelet aggregation (Figure 4A, B). However, Rac1−/− platelets and WT mouse platelets treated with NSC23766 both had a defect in the second-wave of VWF-induced aggregation (Figure 4A). Similarly, NSC23766-treated human platelets were also defective in the second wave of VWF-induced platelet aggregation (Figure 4B). Thus, these results further verify that Rac1 is required for VWF/GPIb-IX-mediated platelet activation in both human and mouse platelets. In addition, Rac1−/− mouse platelets (Figure 4C) or NSC23766-treated human platelets (Figure 4D) were defective in GPIb-IX-mediated TXA2 synthesis, indicating that Rac1 is important in the early GPIb-IX-mediated signaling pathway that is upstream of TXA2 synthesis.
Previous studies from our laboratory and from others have shown that GPIb-IX-mediated early signaling leading to integrin activation involves GPIb-IX-dependent activation of the SFK Lyn, and Lyn-dependent activation of the PI3K/Akt pathway, which subsequently stimulates the cGMP and MAPK pathways.9–12, 32, 33 To determine how Rac1 is involved in GPIb-IX-mediated platelet activation, we determined whether and how loss of function of Rac1 affected this signaling pathway by evaluating VWF/GPIb-IX-mediated phosphorylation of SFK, Akt and p38 MAPK during the early agglutination phase of VWF-induced platelet aggregation (2 minute time point) prior to the second wave of platelet aggregation.2, 8 VWF/GPIb-IX-induced phosphorylation of Akt was abolished in Rac1−/− mouse platelets (Figure 5A) or NSC23766-treated human platelets (Figure 5B), indicating that Rac1 is upstream of Akt in the GPIb-IX signaling pathway. Furthermore, platelet deficiency in Rac1 also abolished GPIb-IX-mediated phosphorylation of P38 MAPK (Figure 5A, B). These data suggest that Rac1 is required for activation of the PI3K/Akt pathway as well as the downstream p38 MAPK pathway. Consistent with the observation that Rac1 functions upstream of the TXA2 synthesis pathway (Fig 4), aspirin had no effect on GPIb-IX-mediated phosphorylation of Akt and P38, which was fully inhibited by NSC23766 in the presence of aspirin. In contrast, VWF/GPIb-IX-induced phosphorylation of SFKs was not negatively affected by deficiency of Rac1 (Figure 5A, B). Thus, Rac1 is not required for GPIb-IX-dependent activation of SFKs.
To assess whether Lyn and/or PI3K were involved in GPIb-IX-mediated activation of Rac1, the levels of GTP-bound Rac1 in human platelets treated with the SFK inhibitor PP2 or PI3K inhibitor LY294002, or in Lyn−/− mouse platelets, were assayed following ligation of GPIb-IX. Rac1 was activated within 1 minute and remained active throughout VWF-induced platelet aggregation in WT platelets (Figure 6A). However, VWF-induced GTP loading of Rac1 was abolished in Lyn−/− (Figure 6A) and PP2-treated platelets (Figure 6B), indicating that Rac1 is activated downstream of Lyn in the GPIb-IX signaling pathway. Interestingly, VWF-induced GTP-loading of Rac1 remained unaffected by LY294002 (Figure 6B, C), indicating that Rac1 activation occurs independent of PI3K. In fact, under the same conditions, LY294002- and NSC23766-treated platelets were both defective in GPIb-IX-mediated activation of Akt and P38 MAPK but not SFKs (Figure 5B). Taken together, these results indicate not only that Rac1 becomes activated during GPIb-IX-dependent platelet activation, but that Rac1 is activated downstream of Lyn and functions upstream of the PI3K/Akt/MAPK pathway.
Rac1 can be activated by multiple GEFs, including Vav. Recent studies suggest that Vav can be activated by Lyn-dependent phosphorylation at Tyr174.34 To evaluate whether Lyn may mediate GPIb-IX-dependent activation of Rac1 via phosphorylation of Vav, we determined whether ligation of GPIb-IX induces Vav phosphorylation and whether GPIb-IX-dependent Vav phosphorylation is affected in Lyn−/− mouse platelets or PP2-treated platelets during the early phase of VWF-induced platelet agglutination/aggregation prior to the second wave of platelet aggregation. Indeed, ligation of GPIb-IX induced Vav activation, which was abolished in Lyn−/− mouse platelets (Figure 6D) or PP2-treated human platelets (Figure 6E). In contrast, activation of Vav was unaffected by either LY294002 or NSC23766 (Figure 6E). Thus, Lyn mediates GPIb-IX-induced activation of Vav and Rac1 independent of the PI3K pathway. Collectively, we have demonstrated that a novel Lyn-Vav-Rac1-PI3K-Akt signaling pathway is important in the early phase of GPIb-IX-mediated signal transduction leading to platelet activation.
In this study, we demonstrate that Rac1 plays a critical role in stimulating GPIb-IX dependent platelet activation. We further show that Rac1 is important in the early GPIb-IX signaling pathway leading to activation of integrin αIIbβ3 and stable platelet adhesion under flow. Importantly, we have discovered that Rac1 and its GEF, Vav, are activated downstream from Lyn, and that Rac1 is required for activating the PI3K/Akt pathway. Thus, our study not only reveals a novel role for Rac1 in platelet activation but also as an important mediator of GPIb-IX-induced, Lyn-dependent activation of the PI3K/Akt signaling pathway leading to integrin activation.
We conclude that Rac1 is required for GPIb-IX-mediated platelet activation. This conclusion is supported by the data that Rac1−/− mouse platelets and NSC23766-treated human platelets were defective in GPIb-IX-dependent integrin activation and stable platelet adhesion to VWF under flow. Moreover, the TXA2-, -secretion-, and integrin-dependent second wave of platelet aggregation was abolished in Rac1-deficient platelets without affecting VWF binding, further supporting an important role for Rac1 in GPIb-IX-mediated platelet activation signaling. The platelet activation process induced by GPIb-IX can be divided into early phase GPIb-IX-specific signaling events and late phase amplification signaling pathways shared by all other platelet agonists.2, 8 Several signaling molecules and events have been shown to be important in the early GPIb-IX signaling pathway leading to integrin activation and stable platelet adhesion, including the SFKs c-Src and Lyn, the PI3K/Akt pathway, intracellular calcium oscillation, cGMP-dependent protein kinase (still controversial), and p38/ERK MAPKs.2, 8, 10, 13, 14, 18, 35–37 Some other signaling molecules, such as components of the ITAM signaling pathway, have been shown to be important in the late signal amplification phase of GPIb-IX-induced platelet activation, which facilitates aggregation and the recruitment of additional platelets to the growing thrombus.2 However, the distinct role of many other signaling molecules in either the early or late phase of GPIb-IX-induced platelet activation remains poorly characterized. This is because the aggregation response of platelets to GPIb-IX-specific early phase receptor signaling is often masked by VWF-mediated platelet agglutination and the second phase of VWF-induced platelet aggregation is complicated by the platelet response to amplification signals induced via the integrin, ITAM, TXA2, and ADP signaling pathways. It is thus difficult to use the routine platelet aggregation assay to effectively specify the role of a particular signaling molecule in GPIb-IX-specific platelet responses. We thus determined the specific role of Rac1 in early GPIb-IX signaling by analyzing the effect of Rac1 deficiency in platelet adhesion to VWF under flow conditions. Under shear stress, stable platelet adhesion to immobilized VWF requires the “early” GPIb-IX mediated signaling pathway that stimulates the inside-out activation of integrin, leading to integrin-dependent stable platelet adhesion. Stable platelet adhesion under shear stress does not require molecules important in the secondary signaling pathways such as Syk and ITAM 11, 12, although it is amplified by TXA2. Our data demonstrate that stable platelet adhesion to VWF under shear stress was dramatically impaired in Rac1−/− mouse platelets and NSC23766-treated human platelets. Furthermore, the impairment in Rac1−/− mouse platelets and NSC23766-treated platelets is significantly greater than the inhibitory effects of saturable concentrations of aspirin. These data indicate that the role of Rac1 in stable platelet adhesion is not limited to its role in TXA2 synthesis and signal amplification, but rather that Rac1 plays an important role in the early, TXA2-independent GPIb-IX-dependent signaling pathway that activates αIIbβ This conclusion is further supported by the data showing that Rac1 stimulates activation of Akt and P38 MAPK independent of the TXA2 pathway (Figure 5B) and that Rac1 is also upstream of the TXA2 synthesis pathway (Figure 4C, D). Importantly, we have provided direct evidence that Rac1 is critical for GPIb-IX-dependent integrin activation (Figure 2). These results, for the first time, indicate a new small GTPase-dependent signaling mechanism in the early GPIb-IX signaling pathway, leading to integrin activation and stable platelet adhesion under shear stress. These results, however, do not exclude an important role for Rac1 also in the secondary platelet amplification pathways such as granule secretion and integrin outside-in signaling, which have been demonstrated previously.26, 38
It is established that the SFK Lyn and the PI3K/Akt pathway are important in early GPIb-IX signaling.11–13, 15, 16 It remains unclear, however, as to how GPIb-IX activates these molecules and how these molecules are linked into a signaling pathway or network that mediates GPIb-IX signaling leading to integrin activation. Previous studies have shown the co-immunoprecipitation of GPIb-IX with PI3K39 and Lyn40, and it was postulated that the GPIb-associated p85 subunit of PI3K functions as a scaffold that recruits SFKs to GPIb, thereby facilitating the activation of SFKs.40 Here we show that GPIb-IX-mediated activation of the PI3K/Akt signaling pathway requires Rac1. We also show that GPIb-IX-mediated activation of Rac1 requires Lyn, and likely involves Lyn-dependent phosphorylation of Vav. Furthermore, the catalytic function of PI3K is not required for GPIb-IX-mediated activation of Rac1, as LY294002 had no effect on GPIb-IX-dependent Rac1 activation. Thus, although we do not exclude the role of PI3K as a scaffold, our study reveals a novel signaling mechanism in which ligand binding to GPIb-IX induces the sequential activation of the SFK Lyn, the GEF Vav, and Rac1. Rac1 then mediates activation of the PI3K/Akt signaling pathway. Thus, this study represents a significant advance by identifying an important molecule in the early GPIb-IX signaling pathway leading to integrin activation.
The identification of Rac1 as a downstream effector of Lyn that stimulates the PI3K/Akt pathway is a novel finding not only to the GPIb-IX signaling pathway in platelets, but also may have general implications to the mechanisms regulating PI3K activation. The SFKs Lyn/Fyn and Rac1 have both been implicated in stimulating the activity of PI3K by binding to the p85 subunit of PI3K.41–43 Our data, however, indicate that Lyn activates the Rac1 GEF Vav and thus Rac1, and requires Rac1 to mediate activation of PI3K. It will be interesting to further investigate whether this new model of Lyn- and Rac1-dependent PI3K activation is a common mechanism in other cell types.
We thank Dr. Radek Skoda, University Hospital Basel, and Dr. Mark Ginsberg, University of California, San Diego for providing Pf4-Cre mice. We also thank Dr. Lowell, University of California, San Francisco for providing the Lyn knockout mice, and Dr. Changgeng Ruan, Soochow University for providing monoclonal antibodies.
SOURCES OF FUNDING
This work is supported in part by NHLBI/NIH grants HL062350, HL068819 and HL080264 (X.D.). M.K.D. is a recipient of the American Heart Association Midwest Affiliate Predoctoral Fellowship Award #11PRE7590050.