ADAP is required for normal alphaIIbbeta3 activation by VWF/GP Ib-IX-V and other agonists

Abstract

Interaction between von Willebrand factor (VWF) and platelet GP Ib-IX-V is required for hemostasis, in part because intracellular signals from VWF/GP Ib-IX-V activate the ligand-binding function of integrin alphaIIbbeta3. Because they also induce tyrosine phosphorylation of the ADAP adapter, we investigated ADAP's role in GP Ib-IX-V signal transduction. Fibrinogen or ligand-mimetic POW-2 Fab binding to alphaIIbbeta3 was stimulated by adhesion of ADAP+/+ murine platelets to dimeric VWF A1A2 but was significantly reduced in ADAP-/- platelets (P<.01). alphaIIbbeta3 activation by ADP or a Par4 thrombin receptor agonist was also decreased in ADAP-/- platelets. ADAP stabilized the expression of another adapter, SKAP-HOM, via interaction with the latter's SH3 domain. However, no abnormalities in alphaIIbbeta3 activation were observed in SKAP-HOM-/- platelets, which express normal ADAP levels, further implicating ADAP as a modulator of alphaIIbbeta3 function. Under shear flow conditions over a combined surface of VWF A1A2 and fibronectin to test interactions involving GP Ib-IX-V and alphaIIbbeta3, respectively, ADAP-/- platelets displayed reduced alphaIIbbeta3-dependent stable adhesion. Furthermore, ADAP-/- mice demonstrated increased rebleeding from tail wounds. These studies establish ADAP as a component of inside-out signaling pathways that couple GP Ib-IX-V and other platelet agonist receptors to alphaIIbbeta3 activation.

Figure 1

Bleeding times in wild-type or ADAP knockout mice. Tails of ADAP^+/+ or ADAP^−/− mice were warmed and then transected 0.5 mm from the end and immersed in 0.9% isotonic saline solution at 37°C. Blood flow was monitored, and the time when bleeding arrested initially was noted as the bleeding time. Tails were monitored further for a minimum of 60 seconds to determine if bleeding recurred. (A) Bleeding times ± standard error of the mean (SEM). (B) Percentage of mice that showed rebleeding. Results are calculated from measurements on a total of 33 ADAP^+/+ and 37 ADAP^−/− mice.

Figure 2

Characterization of ADAP^−/− platelets. Washed platelets from ADAP^+/+ and ADAP^−/− mice were incubated with the appropriate antibodies, and binding was determined by flow cytometry. (A) αIIbβ3 surface expression. (B) GP Ibα surface expression. (C) α-granule secretion. Platelets were resting, or activated with 250 μM Par4-activating peptide, and the binding of an anti–murine P selectin antibody was determined. There was no statistically significant difference in P selectin expression between ADAP^+/+ and ADAP^−/− platelets in response to Par4-activating peptide despite statistically significant increases in P selectin expression compared with resting platelets. Results shown represent a summary of at least 4 experiments and show the average ligand binding ± standard error of the mean (SEM).

Figure 3

Integrin activation induced by GP Ib-IX-V. Washed platelets from ADAP^+/+ (A-C) or ADAP^−/− (D-F) mice were allowed to settle on dmA1A2 VWF. The activation state of αIIbβ3 was reported by the binding of soluble FITC-fibrinogen (green). A noninhibitory antibody to αIIbβ3 was used to view the cell outline (red). (A and D) Platelets were on dmA1A2 VWF in the presence of a cocktail of inhibitors against ADP and thromboxane A2. (B and E) As in panels A and D, with the addition of EDTA. (C and F) Platelets were on dmA1A2 VWF and activated with PMA without inhibitors. Results shown are representative of at least 3 experiments with similar results. (G-H) Quantification of specific, activation-dependent FITC-fibrinogen (G) or POW-2 Fab (H) binding to αIIbβ3 in response to GP Ib-IX-V ligation, as shown in panels A to F for FITC-fibrinogen. Results for ADAP^−/− (white bars) are depicted as a percentage of the soluble ligand that is specifically bound by ADAP^+/+ platelets (black bars) ± SEM. Data represent a summary of 4 experiments.

Figure 4

GP Ib-IX-V–induced αIIbβ3 activation under shear flow. Whole blood from ADAP^+/+ or ADAP^−/− mice was adjusted to correct for thrombocytopenia of ADAP^−/− mouse blood and aspirated over a coverslip coated with dmA1A2 VWF, with or without fibronectin (FN), and placed in a parallel plate flow chamber. The total number of platelets tethering to the surface was measured, and the percentage of these platelets that remained stationary for 3.5 seconds or 10 seconds was determined at 20 seconds, 1 minute, or 2 minutes after the initiation of flow. Results represent stationary platelets, as a percentage of the total number of platelets contacting the surface within the sampling time, ± SEM. (A) Platelets stationary for 3.5 seconds. (B) Platelets stationary for 10 seconds. Stationary adhesion (dependent on αIIbβ3 activation) is much higher for platelets on the combined dmA1A2 VWF/FN surface than on dmA1A2 VWF alone. Results shown are the summary of at least 3 experiments. *P < .05 in a Student t test for ADAP^−/− platelets compared with ADAP^+/+ platelets.

Figure 5

Cytoskeletal rearrangements in response to adhesion to dmA1A2 VWF and fibronectin under shear flow. Platelets were allowed to interact with dmA1A2 VWF alone or together with FN as described in Figure 4. Cells were fixed after 2 minutes of flow, stained for αIIb, imaged by deconvolution microscopy, and scored for filopodial extensions. (A) ADAP^+/+ or (B) ADAP^−/− platelets adherent on the combined A1A2 VWF/FN surface. Arrowheads illustrate filopodia, which were shorter and less numerous in panel B. (C) Quantitative representation of the percentage of cells with filopodial protrusions on dmA1A2 VWF alone or together with FN ± SEM (*P < .05). Results shown are the summary of 3 separate experiments.

Figure 6

Soluble FITC-fibrinogen binding to platelets activated by ADP, Par4 receptor-activating peptide, or convulxin. Platelets were stimulated with varying concentrations of agonist, and the binding of soluble FITC-fibrinogen to platelets was determined by flow cytometry. ADAP^+/+ or ADAP^−/− platelets stimulated with (A) ADP, (B) Par4, or (C) convulxin. Results are shown as the percentage of ADAP^+/+ FITC-fibrinogen binding to ADAP^+/+ cells for a given agonist concentration ± SEM (*P < .05; **P < .01). Results are a summary of at least 4 separate experiments.

Figure 7

SKAP-HOM protein expression and function in ADAP^+/+ or ADAP^−/− mice. (A) Washed platelets from ADAP^+/+ or ADAP^−/− mice were lysed, subjected to SDS gel electrophoresis, and Western blotted for SKAP-HOM. Blots were then reprobed for ADAP and integrin β3, the latter to monitor gel loading. Two separate anti–SKAP-HOM antibodies were used to confirm results. (B) SKAP-HOM regulation by ADAP in CHO cells. CHO cells were transfected with 1 μg ADAP alone, 3 μg EGFP-SKAP-HOM or EGFP-SKAP-HOM lacking the SH3 domain (dSH3) alone, or a combination of ADAP and EGFP-SKAP-HOM. After 48 hours, cells were subjected to SDS polyacrylamide gel electrophoresis and probed for ADAP and SKAP-HOM. Platelet lysate from wild-type mice, prepared as in Figure 6A, was used as a positive control, and blots were reprobed with a β-actin antibody to monitor gel loading. (C) Platelets from wild-type or SKAP-HOM knockout mice were allowed to adhere to dmA1A2 VWF as described in Figure 2, and integrin activation was assessed by FITC-fibrinogen binding in the presence or absence of a cocktail of inhibitors or PMA. Results in panels A and B are representative of at least 3 experiments, with similar results. Results in panel C are for ADAP^−/− platelets (white bars), depicted as a percentage of soluble ligand that is specifically bound by ADAP^+/+ platelets (black bars) ± SEM, and are a summary of 3 separate experiments.