Antithrombotic effects of targeting alphaIIbbeta3 signaling in platelets

Abstract

alphaIIbbeta3 interaction with fibrinogen promotes Src-dependent platelet spreading in vitro. To determine the consequences of this outside-in signaling pathway in vivo, a "beta3(Delta760-762)" knockin mouse was generated that lacked the 3 C-terminal beta3 residues (arginine-glycine-threonine [RGT]) necessary for alphaIIbbeta3 interaction with c-Src, but retained beta3 residues necessary for talin-dependent fibrinogen binding. beta3(Delta760-762) mice were compared with wild-type beta3(+/+) littermates, beta3(+/-) heterozygotes, and knockin mice where beta3 RGT was replaced by beta1 C-terminal cysteine-glycine-lysine (EGK) to potentially enable signaling by Src kinases other than c-Src. Whereas beta3(+/+), beta3(+/-) and beta3/beta1(EGK) platelets spread and underwent tyrosine phosphorylation normally on fibrinogen, beta3(Delta760-762) platelets spread poorly and exhibited reduced tyrosine phosphorylation of c-Src substrates, including beta3 (Tyr(747)). Unlike control mice, beta3(Delta760-762) mice were protected from carotid artery thrombosis after vessel injury with FeCl(3). Some beta3(Delta760-762) mice exhibited prolonged tail bleeding times; however, none demonstrated spontaneous bleeding, excess bleeding after surgery, fecal blood loss, or anemia. Fibrinogen binding to beta3(Delta760-762) platelets was normal in response to saturating concentrations of protease-activated receptor 4 or glycoprotein VI agonists, but responses to adenosine diphosphate were impaired. Thus, deletion of beta3 RGT disrupts c-Src-mediated alphaIIbbeta3 signaling and confers protection from arterial thrombosis. Consequently, targeting alphaIIbbeta3 signaling may represent a feasible antithrombotic strategy.

Figure 1

Generation of β3 knockin mice. (A) Wild-type β3^+/+ (WT) and mutant β3 cytoplasmic domain sequences. β3(Δ760-762) lacks the 3 C-terminal residues of β3 (arginine-glycine-threonine [RGT]), while in β3/β1(glutamic acid–glycine-lysine [EGK]) those residues have been replaced with the respective 3 C-terminal residues of β1. (B) Gene-targeting strategy. A 7.2-kb targeting vector for the mouse β3 gene contained a Neo cassette flanked by 2 lox P sequences between β3 exons 14 and 15. Either of the β3 mutations was introduced into exon 15. B, BamHI; E, EcoRI; X, XhoI. (C) Southern blot analysis with a 3′ probe of R1 embryonic stem (ES) cell genomic DNA transfected with the β3(Δ760-762) targeting vector and digested with EcoRI. (D) PCR genotyping of final floxed β3 cytoplasmic domain mutants. (E) Mouse platelet lysates blotted with antibodies recognizing the extracellular domain of β3 (anti-β3) or the β3 C-terminus (anti-β3 C-terminus). (F) Platelet lysate (10 μg) from β3^+/+, β3^+/−, and β3(Δ760-762) mice were resolved on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotted with antibody to β3 or c-Src. (G) Surface expression of αIIbβ3 was determined by flow cytometry using an antibody to αIIb. Data are expressed as mean fluorescence intensity in arbitrary units plus or minus standard error of the mean [SEM] (n = 5-13).

Figure 2

β3(Δ760-762) mice are resistant to carotid artery thrombosis. (A) Carotid artery blood flow profiles from 1 representative WT and 2 β3(Δ760-762) mice. Arteries were exposed to FeCl₃ for 3 minutes beginning at t = 0, and blood flow was measured for at least 30 minutes. (B) Times to thrombotic occlusion for each mouse studied: WT, n = 13; β3(Δ760-762), n = 11; β3/β1(EGK), n = 7; β3^+/−, n = 5. Two β3(Δ760-762) mice (arrows) showed either an approximately 90% reduction in initial flow rate, or restoration of flow within 6 minutes after complete occlusion. (C) Tail bleeding times for initial cessation of bleeding. Each circle indicates 1 animal: WT, n = 56; β3(Δ760-762), n = 46; β3/β1(EGK), n = 31; β3^+/−, n = 18. Box graphs indicate median plus or minus SEM, and the total distribution for each genotype. (D) Frequency of rebleeding from tail wounds: WT, n = 56; β3(Δ760-762), n = 34; β3/β1(EGK), n = 27; β3^+/−, n = 18.

Figure 3

Platelet spreading on immobilized fibrinogen. (A) Platelets were incubated for 45 minutes on fibrinogen-coated coverslips in the absence or presence of 0.5 mM MnCl₂. Platelets were fixed and stained with rhodamine-phalloidin (red) to label F-actin or with an antibody to phosphotyrosine (green). Wild-type platelets showed filopodia (arrowheads) and minimal lamellipodial extension, both of which were accentuated by extrinsic activation of αIIbβ3 with MnCl₂. Scale bar = 10 μm. (B) Platelets were plated on fibrinogen as in panel A, in the absence or presence of MnCl₂, ADP, or PAR4 peptide. After fixation and staining, platelet spreading (mean surface area) was quantified by image analysis (see “Methods” for details). *P < .01. Data represent means plus or minus SEM of 349 to 859 platelets analyzed on 6 to 10 separate coverslips for each experimental condition.

Figure 4

Platelet tyrosine phosphorylation, fibrin clot retraction, and binding properties of β3 cytoplasmic domain mutants. (A) Platelets were incubated in suspension in the presence of 250 μg/mL fibrinogen plus or minus 0.5 mM MnCl₂ for 20 minutes at room temperature and then lysed. Protein tyrosine phosphorylation was assessed by immunoblotting with antiphosphotyrosine antibodies (pTyr). Blots were stripped and reprobed with an antibody to c-Src. (B) Phosphorylation of β3 Tyr⁷⁴⁷ was detected with a phospho-specific antibody. Blots were reprobed with an antibody to total β3 and phosphorylation data were normalized for β3 expression. The bar graph depicts the fold-increase in β3 Tyr⁷⁴⁷ phosphorylation induced by MnCl₂ + fibrinogen. Data are means (± range) for 2 independent experiments. (C) Pull-down of c-Src and talin from mouse platelet lysate (PltLys) by recombinant wild-type and mutant β3 cytoplasmic domain model proteins. An αIIb cytoplasmic domain protein was used as a negative control. Bound c-Src and talin were detected by immunoblotting. Loading of pull-down beads with recombinant cytoplasmic domains was monitored by staining with Coomassie brilliant blue. (D) Fibrin clot retraction using platelet-rich plasma from wild-type or mutant mice. Clotting was initiated with 9.4 U/mL thrombin and 1.9 mM CaCl₂. Data represent mean (± SEM) for least 3 separate experiments performed in duplicate on at least 4 animals with each genotype. Experiments with wild-type platelets in the presence of 10 μM cytochalasin D (cyto D) or vehicle (dimethyl sulfoxide [DMSO]) were performed in duplicate on platelets from 4 animals to confirm that clot retraction was dependent on actin polymerization in this assay.

Figure 5

Platelet aggregation and fibrinogen binding. (A) Platelet aggregation in platelet-rich plasma was stimulated with ADP or PAR4 receptor-activating peptide (AYPGKF). Results are representative of at least 3 experiments, performed in duplicate on at least 4 animals from each genotype. (B) Specific FITC-fibrinogen binding to platelets was assessed by flow cytometry. Fibrinogen binding is reported as mean fluorescence intensity in arbitrary fluorescence units and was normalized for αIIbβ3 surface expression measured with antibody to αIIb. *P ≤ .05 compared with wild-type platelets. (C) Specific fibrinogen binding in the absence (□) and presence (■) of 0.5 mM MnCl₂. Results in B and C are representative of at least 6 independent experiments for each genotype.