Deletion of platelet CLEC-2 decreases GPIbα-mediated integrin αIIbβ3 activation and decreases thrombosis in TTP

Abstract

Microvascular thrombosis in patients with thrombotic thrombocytopenic purpura (TTP) is initiated by GPIbα-mediated platelet binding to von Willebrand factor (VWF). Binding of VWF to GPIbα causes activation of the platelet surface integrin αIIbβ3. However, the mechanism of GPIbα-initiated activation of αIIbβ3 and its clinical importance for microvascular thrombosis remain elusive. Deletion of platelet C-type lectin-like receptor 2 (CLEC-2) did not prevent VWF binding to platelets but specifically inhibited platelet aggregation induced by VWF binding in mice. Deletion of platelet CLEC-2 also inhibited αIIbβ3 activation induced by the binding of VWF to GPIbα. Using a mouse model of TTP, which was created by infusion of anti-mouse ADAMTS13 monoclonal antibodies followed by infusion of VWF, we found that deletion of platelet CLEC-2 decreased pulmonary arterial thrombosis and the severity of thrombocytopenia. Importantly, prophylactic oral administration of aspirin, an inhibitor of platelet activation, and therapeutic treatment of the TTP mice with eptifibatide, an integrin αIIbβ3 antagonist, reduced pulmonary arterial thrombosis in the TTP mouse model. Our observations demonstrate that GPIbα-mediated activation of integrin αIIbβ3 plays an important role in the formation of thrombosis in TTP. These observations suggest that prevention of platelet activation with aspirin may reduce the risk for thrombosis in patients with TTP.

© 2022 by The American Society of Hematology.

Figures

Graphical abstract

Figure 1.

CLEC-2 deficiency impairs GPIbα-mediated integrin αIIbβ3 activation. Aggregation of washed WT and CLEC-2–deficient platelets in response to 10 μg/mL of VWF and 2 μg/mL of botrocetin, which facilitates VWF binding to GPIbα, in the absence (A) or presence (C) of GR144053 trihydrochloride (GR). (B and D) Quantification of platelet aggregation shown in (A) and (C), respectively. (E) Flow cytometry analysis of binding of mAb Jon/A, which specifically binds to activated integrin αIIbβ3, to platelets. Washed platelets were stimulated with 2 μg/mL of botrocetin (Botro) in the presence or absence of 10 μg/mL of VWF from human plasma. After that, platelets were immediately incubated with mAb Jon/A, and mAb Jon/A binding was analyzed with flow cytometry. The data are representative of 5 independent experiments, and data represent mean ± standard deviation (SD). *P < .05.

Figure 2.

CLEC-2 deficiency impairs platelet aggregation on immobilized VWF under flow. Washed platelets treated with albumin or 20 μg/mL of RGD, an integrin αIIbβ3 antagonist, were perfused over immobilized plasma VWF (150 μg/mL) at 10 dyne/cm2 for 10 minutes. After removing nonadherent platelets, adhered platelets were fixed with 2% paraformaldehyde (PFA), stained with anti-CD41 antibodies, and observed with microscopy. (A) Representative images of platelet aggregation on immobilized VWF under flow. (B) Quantification of platelet aggregation on immobilized VWF under flow. Adherent platelets described in (A) were quantified by analyzing the pixels of CD41-positive areas in each field. The data are representative of 5 independent experiments, and data represent mean ± SD. *P < .05.

Figure 3.

CLEC-2 deficiency impairs GPIbα signaling. (A) Akt, (B) p38 MAPK, and (C) Lyn activation in platelets stimulated with human plasma VWF and botrocetin, which facilitates VWF binding to GPIbα. Washed platelets were preincubated with dimethyl sulfoxide (DMSO) or 10 μM of PP2, an inhibitor of Src family kinases, at room temperature for 30 minutes and then stimulated with 2 μg/mL of botrocetin with or without 10 μg/mL of VWF for 5 minutes. Then platelets were lysed, and activation of Akt and p38MAPK was analyzed by western blotting. P-Akt, P-p38, and P-Lyn are the activated kinases. (D) Aggregation of PP2-treated platelets in response to botrocetin and VWF. Washed platelets were pretreated with DMSO or 10 μM of PP2 at room temperature for 30 minutes, and then aggregation of WT and CLEC-2–deficient platelets in response to botrocetin and VWF was observed. (E) Quantification of platelet aggregation shown in (D). (F) Flotation assay of lipid raft localization of GPIbα. Washed platelets were lysed with 1% triton X-100, and the lysate was applied to the top of 5% to 40% gradient of Optiprep. After centrifugation, the distribution of GPIbα, Lyn, and β actin in fractions from top to bottom of the gradient was analyzed by western blot. (G) Quantification of GPIbα that located in lipid rafts shown in (F). The data are representative of 5 independent experiments.

Figure 4.

The extracellular domain of CLEC-2 is required for GPIbα-mediated platelet activation. (A) Schematic for the construction of a chimeric CLEC-2 that fused human IgG Fc with the CLEC-2 extracellular domain (CLEC-2/FC). hFC, human IgG Fc control; N, N-terminus; C, C-terminus. (B) Recombinant CLEC-2 inhibited VWF-induced platelet aggregation. Washed WT platelets were pretreated with Fc or CLEC-2/Fc at 37°C for 1 hour, and aggregation of platelets in response of human plasma VWF was measured. (C) Quantification of platelet aggregation shown in (B). (D) Activation of integrin αIIbβ3 on CLEC-2/Fc-treated platelets. Washed WT platelets were pretreated with Fc or CLEC-2/Fc at room temperature for 1 hour and then treated with VWF in the presence of botrocetin for 5 minutes at 37°C. Then, integrin αIIbβ3 activation was measured with antibody Jon/A by flow cytometry. (E) Recombinant CLEC-2 inhibited platelet aggregation on immobilized VWF under flow. WT platelets were preincubated with Fc or CLEC-2/Fc (30 μg/mL) for 1 hour at 37°C and were then perfused over plates coated with VWF from human plasma (100 μg/mL) at 10 dyne/cm2 for 10 minutes. Adherent platelets were fixed with 2% PFA and stained with anti-mouse CD41, followed by Alexa 488-conjugated secondary antibodies. Representative images were presented. (F) Quantification of aggregated platelets shown in (E). The data are representative of 3 independent experiments and represent mean ± SD. *P < .05.

Figure 5.

Platelet CLEC-2 deficiency protects mice from thrombotic injury in a mouse model of TTP. Mice were injected with antibodies to ADAMTS13 and human plasma VWF to induce TTP. At 24 hours after TTP induction, blood was collected to analyze VWF levels and platelet count, and lungs were collected to analyze thrombus formation. (A) Representative immunofluorescence images of thrombi in lungs of mice with TTP. (B) Quantification of thrombus formation in lung shown in (A). Quantification of the pixels of CD41-positive areas in each field of images. (C) Platelet counts in mice with TTP. Platelet counts in mice at 7 days before experiments (basal level) and at 24 hours after induction of TTP (by injecting antibodies to ADAMTS13 and VWF) were quantified. Each dot represented a datum from 1 mouse. Comparisons marked by asterisks are significantly different (P < .05). (D) Blood LDH activity of mice at basal level and at 24 hours after TTP induction. (E) Analysis of VWF in blood of mice with TTP by electrophoresis on agarose gel and by densitometric scanning. The data are representative of 5 independent experiments, and data represent mean ± SD. *P < .05.

Figure 6.

Inhibition of integrin αIIbβ3 activation reduces thrombus formation in TTP. WT mice were injected with antibodies to ADAMTS13 and human plasma VWF to induce TTP. At 24 hours after TTP induction, blood was collected to analyze platelet count, and lungs were collected to analyze thrombus formation. Saline control or eptifibatide (15 mg/kg, body weight) was injected when VWF was administrated and then was given every 6 hours after the first administration. (A) Representative immunofluorescence images of thrombi in lungs of saline- and eptifibatide-treated mice with TTP. (B) Quantification of thrombus formation in lung shown in (A). Quantification of the pixels of CD41-positive areas in each field of images. (C) Platelet counts in mice with TTP. Platelet counts in mice at 7 days before experiments (basal level) and at 24 hours after induction of TTP were quantified. Each dot represented a datum from 1 mouse. The data are representative of 5 independent experiments, and data represent mean ± SD. *P < .05.

Figure 7.

Inhibition of platelet activation by aspirin reduces thrombus formation in a mouse model of TTP. (A) Enzyme-linked immunosorbent assay analysis of TxA2 release from platelets. Washed platelets were treated with saline or botrocetin in the presence or absence of VWF for 7 minutes at 37°C, and released TxA2 was measured by detecting TxB2, a stable metabolite of TxA2, in the supernatant. TTP in WT mice was induced by injection of antibodies to ADAMTS13 and plasma VWF. Aspirin (5 mg/kg, body weight) was given 3 days before VWF administration through oral gavage, twice a day, until the end of the experiment. In mice without TTP, VWF was not administrated after antibody injection. At 24 hours after TTP induction with VWF, mice were euthanized and platelet count in blood and thrombus formation in lung were analyzed. (B) Aggregation of platelets from mice fed with the vehicle or aspirin. Platelet-rich plasma from mice treated with the vehicle control or aspirin was stimulated with arachidonic acids (500 μg/mL), and aggregation was monitored. Quantification of aggregation was inserted inside the figure. (C) Flow cytometry analysis of surface expression of CD61 on platelets from mice treated with the vehicle or aspirin. (D) Representative immunofluorescence images of thrombi in lungs of saline- and aspirin-treated mice with TTP. (E) Quantification of thrombus formation in lung shown in (D). Quantification of the pixels of CD41-positive areas in each field of images. (F) Platelet counts in mice with TTP. Platelet counts in mice at 7 days before experiments (basal level) and at 24 hours after induction of TTP were quantified. Each dot represented a datum from 1 mouse. The data are representative of 5 independent experiments, and data represent mean ± SD. *P < .05.