Roles and interactions among protease-activated receptors and P2ry12 in hemostasis and thrombosis

Abstract

Toward understanding their redundancies and interactions in hemostasis and thrombosis, we examined the roles of thrombin receptors (protease-activated receptors, PARs) and the ADP receptor P2RY12 (purinergic receptor P2Y G protein-coupled 12) in human and mouse platelets ex vivo and in mouse models. Par3(-/-) and Par4(+/-) mouse platelets showed partially decreased responses to thrombin, resembling those in PAR1 antagonist-treated human platelets. P2ry12(+/-) mouse platelets showed partially decreased responses to ADP, resembling those in clopidogrel-treated human platelets. Par3(-/-) mice showed nearly complete protection against carotid artery thrombosis caused by low FeCl(3) injury. Par4(+/-) and P2ry12(+/-) mice showed partial protection. Increasing FeCl(3) injury abolished such protection; combining partial attenuation of thrombin and ADP signaling, as in Par3(-/-):P2ry12(+/-) mice, restored it. Par4(-/-) mice, which lack platelet thrombin responses, showed still better protection. Our data suggest that (i) the level of thrombin driving platelet activation and carotid thrombosis was low at low levels of arterial injury and increased along with the contribution of thrombin-independent pathways of platelet activation with increasing levels of injury; (ii) although P2ry12 acts downstream of PARs to amplify platelet responses to thrombin ex vivo, P2ry12 functioned in thrombin/PAR-independent pathways in our in vivo models; and (iii) P2ry12 signaling was more important than PAR signaling in hemostasis models; the converse was noted for arterial thrombosis models. These results make predictions being tested by ongoing human trials and suggest hypotheses for new antithrombotic strategies.

Fig. 1.

Effect of Par3, Par4, and P2ry12 heterozygous and homozygous null mutations on mouse platelet responses, and PAR1 antagonism on human platelet responses to thrombin and ADP. Washed platelets were incubated with the indicated concentrations of thrombin or ADP; binding of antibodies specific for activated α_IIbβ₃ was measured by flow cytometry. Antibodies JON/A and PAC-1 recognize activated mouse and human α_IIbβ₃, respectively. (A) Thrombin responses in mouse platelets with the indicated Par genotypes. (B) Thrombin responses in human platelets incubated with SCH530348 (100 nM), vehicle, or nothing for 1 h before stimulation. (C) Thrombin and (D) ADP responses in mouse platelets with the indicated P2ry12 genotypes. Results were expressed as percent of maximal response to thrombin in each experiment. Mean ± SEM is shown. n = 7–9 in A; 3 in B; and 10–17 in C and D.

Fig. 2.

Effect of mutations that cause partial or complete loss of thrombin or ADP signaling in platelets on carotid thrombosis at different levels of FeCl₃ injury. Littermates from intercrosses of (A–C) Par4^+/−, (D and F) Par3^+/−, (E) Par3^+/−:P2ry12^+/−, and (G–I) P2ry12^+/− mice were studied. Carotid thrombosis was triggered by 4 (A, D, and G), 8 (B, E, and H), and 20% (C, F, and I) FeCl₃ injury. The percent of arteries without occlusion as a function of time after injury is shown. Flow as a function of time after injury is shown in Fig. S4. The number for each group is indicated. The cross in E was used to generate the Par3^+/+ and Par3^−/− data shown here and the more complex comparison in Fig. 3C.

Fig. 3.

Effect of combined partial or complete loss of thrombin and ADP signaling on carotid thrombosis. Littermates from intercrosses of (A) Par4^+/−:P2ry12^−/− or Par4^−/−:P2ry12^+/−, (B) Par4^+/−:P2ry12^+/−, and (C) Par3^+/−:P2ry12^+/− mice were studied as in Fig. 2. In A, wild-type mice from the same colony were studied in parallel. Carotids were injured with (A) 20, (B) 4, and (C) 8% FeCl₃. Flow as a function of time after injury is shown in Fig. S5.

Fig. 4.

Effect of combined partial or complete loss of thrombin and ADP signaling on hemostasis. (A–C) Littermates from Par4^+/−:P2ry12^+/−, Par4^+/−:P2ry12^−/−, and Par4^−/−:P2ry12^+/− intercrosses, and (D–F) Par3^+/−:P2ry12^+/− intercrosses were studied. (A and D) Tail-bleeding time in seconds, (B and E) tail blood loss (AU based on A575 nm), and (C and F) small vessel bleeding during surgery were measured. Mean ± SEM is shown. n =16–65 in A; 16–60 in B; 10–39 in C; 13–27 in D; 13–25 in E; and 10–18 in F.

Fig. 5.

Comparison of the effects of Par3, Par4, and P2yr12 heterozygous or homozygous mutations, alone and in combination, on thrombosis vs. hemostasis. Carotid thrombosis was induced by (A) 4, (B) 8, and (C) 20% FeCl₃. The percent of carotids open at the end of the protocol vs. mean tail blood loss measured as in Fig. 4 are shown for the genotypes indicated. Sample size ranged from 23 to 96 for hemostasis indices and 8 to 50 for thrombosis indices. Similar results were obtained using plots of other indices of protection against thrombosis and impaired hemostasis (Figs. S8 and S9).