Platelet Tissue Factor: Expression of Pro-Coagulant Activity depends on Gpibα Activation and Signaling through Lyn-Mediated Phosphorylation. A Platelet-Based Model of Hemostasis

Human platelets synthesize and store functionally silent tissue factor (TF) which has procoagulant activity (PCA) after platelet activation. We now explored the location of inactive and active forms of TF and mechanisms of activation. We reported that resting, non-permeabilized human platelets express scant surface TF, which was strikingly enhanced and co-localized with GPIbα after stimulation (Blood2007;109:5242). The externalization of TF was confirmed by immunoprecipitation (Ip) from biotinylated membranes before and after platelet activation. Moreover, TF and GPIbα co-precipitated in Ip assays and both glycoproteins were prominently found in lipid rafts (Lr) domains. TF-dependent PCA, assessed by FXa generation, was negligible or absent in resting, leukocyte-free platelet preparations. Interestingly, FVII was found in platelet membranes (western blot) and no exogenous FVIIa was needed to trigger TF-dependent FXa generation from washed platelets. Platelet responses to activation (TF-dependent PCA, platelet aggregation and secretion) depended on the agonist used. With 1IU/mL VWF + 1.2 mg/mL Ristocetin (Ris) PCA increased 5 to 10-fold 2 min after stimulation and platelets formed large aggregates with null 14C-serotonin secretion. In contrast, both platelet aggregation and secretion were normal 2 min after activation with 1IU thrombin, 10μM TRAP or 2μg/mL collagen; and PCA induced by thrombin was only ≈1/2 of that elicited by VWF+Ris, whereas 2 min after TRAP or collagen stimulation no PCA induction was detected. Removal of membrane cholesterol (by methyl-β-cyclo-dextrin) or disruption of Lr (with filipin III) abolishes the VWF-Ris-induced PCA. Lr’s from resting platelets contain TF of Mr ≈60kDa. Five min after activation an increase in Lr’s TF was observed, but its Mr depended on the agonist: species of ≈47kDa and ≈60kDa were found after VWF-Ris and TRAP activation, respectively. Given that GPIbα signals through Lyn, member of the Src family, inhibition of signaling with PP2 resulted in 80% fall of VWF-Ris-induced PCA. Ip assays revealed that Lyn co-precipitated with both GPIbα and TF in VWF-Ris activated, but not resting platelets. The phosphokinase activity of Lyn on TF was tested. A polyclonal antibody raised against the phosphorylated (Ser253/Ser258) cytoplasmic domain of TF (Dr. W. Ruf, La Jolla) recognized membrane TF only in activated, not in resting platelets. These findings indicate that VWF-induced activation of GPIbα, subsequent signaling with Lyn-induced serine phosphorylation, along with a change of TF to a ≈47kDa species, triggers human platelets PCA. The previously described role of phospho-disulphide isomerases (PDI) in TF activation through SH groups oxidation was also explored. TF and PDI co-precipitated in resting and activated platelet membranes and antagonizing PDI with bacitracin inhibited TF-dependent PCA. After platelet activation and labeling with MPB, a sulfhydryl-specific probe, TF protein (with PCA) was detected on the plasma membrane, denoting presence of reduced thiols. Furthermore, platelet incubation with phenylarsine oxide, a blocking reagent of vicinal SH groups, or HgCl2, a potent oxidant of thiol groups, had no effect on platelet PCA. Thus, it seems unlikely that TF activation depends on SH oxidation. Finally, we found that platelet TF was sufficient to speed up the clotting of plasma. In fact, clotting time of PRP (2 × 108 platelets mL−1) incubated with ionophore A23187 (2 min, 37°C) and then re-calcified, was 59 ± 6 sec, whereas clotting time in re-calcified PRP without previous activation was 137 ± 19 sec (n=19, p<0.0001). Taken together, our results highlight the crucial role of platelets, not only in assembling clotting complexes and reactions on their surface, but also providing enough TF to trigger the whole process. This novel, comprehensive understanding of hemostasis, (“platelet-based hemostasis model”), unifies primary and secondary hemostasis around the platelets, which would be able to synchronize and modulate the times of both processes ensuring a confined thrombin generation and adequate deposit of fibrin when and where it is needed. It also emphasizes the self-sufficiency of intravascular components to carry out both normal hemostasis and thrombus formation. In this context, platelet PCA may become a central pharmacological target for preventing or managing bleeding and thrombotic disorders.