Role of heterotrimeric G proteins in platelet activation and clot formation in platelets treated with integrin αIIbβ3 inhibitor.

Mechanisms of platelet activation are triggered by thrombin, adenosine diphosphate (ADP), epinephrine, thromboxane A2, and other soluble agonists which induce signaling via heterotrimeric Gαq, Gαi, and Gα12/13 proteins. We have undertaken a study addressing the contribution of these G proteins to platelet activation and clot formation in the presence of eptifibatide, thus excluding outside-in signaling provided by integrin αIIbβ3-fibrinogen engagement. Selective and combined activation of the G proteins was achieved by using combinations of platelet agonists and inhibitors. Platelet activation in platelet-rich plasma was evaluated by P-selectin expression using flow cytometry. Contribution of platelets to whole blood clotting was assessed by rotation thromboelastometry (ROTEM). Selective signaling of Gαq or Gαi but not Gα12/13 promoted P-selectin expression. Further enhancement of P-selectin expression was achieved by ADP-induced combined signaling of Gαq and Gαi, and to more extent by U46619 at high concentration (1.5 μM) induced combined signaling of Gαq and Gα12/13 while maximal P-selectin expression was achieved by thrombin receptor-activating peptide (TRAP)-induced combined signaling of Gαq, Gαi, and Gα12/13. In ROTEM, selective activation of Gαq, Gαi, or Gα12/13 failed to affect blood clotting. Combined signaling of Gαq and Gαi or Gαq and Gα12/13 or all three G proteins shortened the clotting time and stimulated clot strength. Pretreatment of platelets with acetylsalicylic acid did not change the effect of ADP but inhibited the effect of TRAP. Signaling of Gαq and Gα12/13 triggered by U46619 also stimulated clot formation. Combined signaling of either Gαq and Gαi or Gαq and Gα12/13 is sufficient to stimulate maximal platelet activation and enhanced clot formation in platelets treated with inhibitor of integrin αIIbβ3. It could be suggested that outside-in signaling is not necessarily required to fulfill these platelet functions.


Introduction
Platelet activation by thrombin (or thrombin receptor-activating peptide, TRAP), adenosine diphosphate (ADP), epinephrine, and thromboxane A 2 (TXA 2 ) is mediated by signaling via heterotrimeric GTP-binding proteins (G proteins) of the Gαq, Gαi, and Gα12/13 families which interact with downstream effectors [1][2][3]. Thrombin activates platelets via protease-activated receptors (PAR)-1 coupled to the members of all three G protein families [4] and PAR-4 coupled to Gαq and Gα12/13 [5,6]. In addition, thrombin deliberates arachidonic acid from membrane phospholipids leading to TXA 2 generation and consequently to Gα12/13 activation not depending on the integrin αIIbβ3-fibrinogen binding [7,8]. ADP binds to P2Y 1 and P2Y 12 receptors coupled to Gαq and Gαi proteins, respectively [9,10]. In addition, ADP induces TXA 2 generation via coordinated outside-in signaling through integrin αIIbβ3-fibrinogen complex [7]. TXA 2 binds to thromboxane-prostanoid (TP) receptors coupled to both Gαq and Gα12/13 proteins [11,12]. G protein-mediated signaling cascades act in synergy leading to inside-out signaling through integrin αIIbβ3 which then provides outside-in signaling following fibrinogen/fibrin binding [3]. Eventually, because all platelet agonists can induce TXA 2 generation and ADP release, their effects involve all G protein signaling pathways. This positive-feedback mechanism has obscured the analysis of the roles of individual G proteins in platelet function. In experimental conditions, the G proteins may be selectively activated or inhibited demonstrating their role in platelet function [3,13,14].
Platelets promote thrombus formation by several mechanisms such as releasing several coagulation factors, providing platform to coagulation process by phosphatidylserine expression, as well as by clot retraction [15][16][17]. Thromboelastometry as a global coagulation test suits well for evaluation of platelet contribution to clot formation [18][19][20][21] and using modification of thromboelastometry, we have also shown that clot firmness greatly depends on platelet activation by specific soluble agonists [22]. The main advantage of this method is that it reproduces all stages of clot formation including clotting time (CT), which reflects the activity of the coagulation cascade, as well as clotting propagation and final clotting strength.
In our previous study, we demonstrated that pretreating of normal blood with eptifibatide, an inhibitor of integrin αIIbβ3, significantly impaired clot formation resembling the clinical situation which occurs in patients with Glanzmann thrombasthenia [20]. Further aggravation of clot formation was observed by dilution of blood due to reduction of platelet count and coagulation factors. These findings prompted us to use this experimental model (platelet pretreatment with eptifibatide plus hemodilution) in order to unmask the proximate effect of G proteins on platelet activation and contribution to clot formation. To our best knowledge, no studies have been performed concerning the G protein downstream signaling without additional outside-in platelet activation and clot formation.

Specific reagents
TRAP and acetylsalicylic acid (ASA; an inhibitor of cyclooxygenase) were obtained from Sigma-Aldrich (St. Louis, MO, USA); ADP was obtained from DiaMed (Cressier, Switzerland); U46619, a stable TXA 2 analogue, AR-C66096, an inhibitor of P2Y 12 ADP receptor coupled to Gαi 2 protein, and MRS2500, an inhibitor of P2Y 1 ADP receptor coupled to Gαq protein, were all obtained from Tocris Bioscience (Bristol, UK); Eptifibatide, a αIIbβ3 integrin inhibitor (Integrilin®), was obtained from Schering-Plough (Kenilworth, NJ, USA). The reagents were used at the following concentrations: TRAP 50 μM, ADP 1.25 μM, ASA 0.5 mM, AR-C66096 and MRS2500 10 μM, and U46619 1.5 and 0.1 μM. ASA was dissolved with DMSO. All other reagents were dissolved in saline. The corresponding amounts of DMSO and Tris-buffered saline (25 mM Tris, 150 mM NaCl, pH 7.4) were introduced in control samples. Approach to the G proteins' signaling is presented in Table I.

Blood collection and conventional analysis
The study was approved by the local ethics committee, and each volunteer signed a written informed consent in accordance with the Declaration of Helsinki. Blood was drawn from adult healthy male and female volunteers (ranging in age from 21 to 54 years) who were free of any medication affecting the hemostasis system for at least 2 weeks prior to blood donation. The first 2 mL of blood was discarded. Blood samples were collected into tubes containing 3.2% sodium citrate or EDTA (Vacuette; Greiner Bio-One GmbH, Kremsmunster, Austria). Whole blood (WB) samples were rested for 30 min at room temperature and processed within 3 h after collection. The samples were considered suitable for the study when the platelet count was within the range of 2-4 × 10 8 mL −1 and hematocrit of 36-48%, which were determined in EDTA-anticoagulated blood and measured in LH-750 device (Beckman Coulter, Oakley, UK). Fibrinogen concentration in plasma of citrated blood was determined by the Clauss method ranging at 2.5-3.6 g/L. Prothrombin time and activated partial thromboplastin time were within normal ranges and measured in ACL TOP 700 (Instrumentation Laboratory, Bedford, MA, USA).

Blood preparation
To reduce the concentration of coagulation factors and, therefore, to unmask the effect of G protein activation by exogenous agonists, citrate-anticoagulated blood was diluted by 40% with Trisbuffered saline. Platelet-rich plasma (PRP) was obtained by centrifugation of blood at 130 g for 12 min and collecting the upper fraction. Platelet-poor plasma (PPP) was obtained by centrifugation of the remaining blood at 2300 g for 15 min. Platelet concentration in PRP was adjusted if needed to 2 × 10 8 mL −1 with autologous PPP. All PRP samples were pretreated for 20 min with 25 μg/mL eptifibatide followed by incubation of PRP for 20 min with ASA, AR-C66096, MRS25000, or corresponding vehicles. Thereafter, platelets were activated by TRAP, ADP, or U46619.

Flow cytometry
PRP samples were pretreated with appropriate antagonists and agonists. Aliquot of the mixture was incubated for 20 min with 90 μL mix of PE-conjugated anti-CD62P (pre-diluted 1:100) and PC7-conjugated anti-CD41 (pre-diluted 1:50) monoclonal antibodies (BioLegend, San Diego, CA, USA). Samples were diluted 5 times with Tyrode buffer and analyzed by flow cytometry. Platelets were gated according to CD41 positive events. Mean fluorescence intensity (MFI) of anti-CD62P was evaluated. Data analysis was performed using the FlowJo software (Tree Star Inc., Ashland, OR, USA).

Rotation thromboelastometry
Reconstitution of WB for processing clot formation was performed by mixing PRP with PPP and packed cells not significantly changing the hematocrit level obtained after the hemodilution. Rotation thromboelastometry (ROTEM) measurements were conducted in ROTEM TM device (Pentapharm, Munich, Germany) using 300 μL of reconstituted WB placed into cups following introduction of 20 μL of NATEM reagent (CaCl 2 at 17 mM) and platelet function modulators. ROTEM tests were performed according the manufacturer's instructions at 37°C and ran for 60 min. The following variables were analyzed: clotting time (CT, s), the time period from introducing the blood into the cup to beginning of blood coagulation; α-angle (αA, degrees), the rate of clot development; and maximum clot firmness (MCF, mm) reflecting clot strength. To ensure uniform methodology, all experiments were performed by the same researcher.

Statistical analysis
The data were presented as mean ± standard deviation (SD) and analyzed using one-way analysis of variance (ANOVA) followed by Holm-Sidak's post hoc test for pairwise comparisons. Differences were considered statistically significant if the p value was less than 0.05.

Effect of G proteins on platelet activation
Combined activation of Gαq, Gαi, and Gα12/13 was achieved by stimulation of platelets with TRAP which was followed by  Figure 1). Pretreatment of platelets with ASA led to a reduction of MFI to 80.3 ± 9.6 compared to non-treated with ASA (p = 0.003). Combined induction of Gαq and Gαi signaling by ADP also activated platelets (MFI 36.9 ± 8.6, p = 0.002 compared to resting platelets). In contrast to TRAP, pretreatment of platelets with ASA did not change the extent of platelet activation induced by ADP (38.2 ± 7.6). When Gαi and Gαq pathways were selectively activated (treating platelets with MRS2500 or AR-C66096, respectively, before ADP), P-selectin expression was slightly reduced compared to ADP but still was significantly higher compared to resting platelets (28.3 ± 7.0 and 28.3 ± 8.3; p = 0.043 for both). Combined signaling of Gαq and Gα12/13 by U46619 at 1.5 μM also activated platelets (56.6 ± 5.7; p < 0.001). In contrast, with selective activation of Gα12/13 by U46619 at 0.1 μM, no difference of P-selectin expression compared to untreated platelets was observed (6.7 ± 1.1 versus 8.2 ± 2.0).

Clotting time
It must be taken into consideration that shortening of CT reflects stimulation of thrombin generation. Triple activation of Gαq, Gαi, and Gα12/13 by TRAP and double activation of Gαq and Gαi by ADP led to similar shortening of CT (421 ± 20 and 458 ± 53 versus 599 ± 50 s, both with p < 0.001, Figure 2A). Pretreatment of TRAP-stimulated platelets with ASA withdrew the effect of TRAP and set the CT level nearer to control (546 ± 62 s, p = 0.209 versus control, p < 0.001 versus TRAP), while the effect of ASA on ADP stimulation was negligible (490 ± 48 s, p = 0.788 versus ADP). Double activation of Gαq and Gα12/13 by 1.5 μM U46619 led to shortening of CT to 477 ± 55 s (p < 0.001). Sole activation of Gαi reduced CT to 535 ± 31 s (p = 0.039), while sole activation of Gαq or Gα12/13 did not change the CT level (558 ± 50 and 591 ± 46 s, respectively).

Maximum clot firmness
Combined activation of all three G proteins by TRAP as well as combined activation of Gαq and Gαi proteins by ADP enlarged the clot strength (30.6 ± 3.2 and 28.8 ± 5.1 versus 22.7 ± 4.9 mm, p < 0.001 and p = 0.005, respectively, Figure 2C). Pretreatment of blood with ASA followed by activation with either TRAP or ADP precluded their stimulatory effect of clot strength (

Discussion
In this study, we aimed to evaluate the involvement of platelet Gαq, Gαi, and Gα12/13 proteins in platelet activation and clot formation. To better understand the proximate contribution of G proteins to these platelet functions, we used an experimental model in which outside-in signaling from the αIIbβ3 integrin was excluded by eptifibatide. Additionally, we reduced clot Mean fluorescence intensity (MFI) of anti-CD62P was evaluated. Data analysis was performed using the FlowJo software. The results are presented as mean ± SD (n = 4) and analyzed by one-way ANOVA followed by Holm-Sidak's post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001 versus resting platelets; ## p < 0.01 with versus without ASA.
formation by hemodilution as it was observed previously using diluted normal blood treated with eptifibatide and diluted blood of patients with Glanzmann thrombasthenia [20]. Selective and combined activation of the G proteins was achieved by using various platelet agonists and inhibitors. We found that combined signaling of Gαq, Gαi, and Gα12/13 proteins induced by TRAP substantially activated eptifibatidepretreated platelets, as evidenced by enhanced P-selectin expression, as well as promoted WB clot formation. Combined signaling of Gαq and Gαi proteins induced by ADP also activated platelets and blood clotting, though the extent of these effects was significantly lower compared to TRAP. Preincubation of platelets with ASA significantly reduced the stimulatory effects of TRAP but did not alter the effects of ADP. The possible explanation for this finding is that stimulation of platelets with TRAP (or thrombin) leads to platelet activation accompanied by TXA 2 generation without the need for formation of the integrin αIIbβ3-fibrinogen complex. In contrast, TXA 2 generation following platelet stimulation with ADP requires coordinated outside-in signaling through the αIIbβ3 integrin [7,8]. In line of this evidence, Malmsten et al. [23] much earlier showed defective thromboxane generation with ADP but not with thrombin in patients with Glanzmann thrombasthenia.
In the present study, selective signaling of Gαq or Gαi triggered lower but still significant platelet activation. In contrast, these G proteins, did not affect blood clotting. Likewise, in another study, when outside-in signaling was not eliminated, platelet aggregation was induced by concomitant but not selective signaling of Gαi and Gαq proteins [24].
U46619, a stable TXA 2 analogue, which at high concentration (1.5 μM) activated both Gα12/13 and Gαq [12,22,25], in our study synergistically stimulated both platelet activation and clot formation. In contrast, U46619 at low concentration (0.1 μM), which triggered selective signaling of Gα12/13 [26], failed to stimulate platelet activation and blood clotting. These results are in accordance with the data that stimulation of Gαq-deficient platelets by U46619 was not accompanied by P-selectin expression [27]. Likewise, when outside-in signaling was not excluded, U46619 at 0.1 μM induced only platelet shape change without further integrin activation and aggregate formation [22]. In line with these observations, clot retraction was shown to be induced by activation of Gαq and Gαi but not Gα12/13 [28]. Though, low doses of U46619 triggered tyrosine phosphorylation of different proteins independently of signals ensuing from integrin αIIbβ3 [26,29].
The results of our study suggest that Gαq and Gαi proteins are able to induce platelet activation and promote clot formation without support by integrin outside-in signaling. Regarding Gα12/13, we cannot assert or deny its role in platelet function in this model, despite some difference of the effects of Gαq, Gαi, Figure 2. Effect of G protein signaling on clot formation. Citrate-anticoagulated blood was diluted by 40% with Tris-buffered saline and then PRP was prepared. The PRP was pretreated with 25 μg/mL eptifibatide for 20 min and additionally with 10 μM MRS2500, 10 μM AR-C66096, or 0.5 mM ASA for another 20 min. Thereafter, reconstitution of WB for processing clot formation was performed by mixing PRP with PPP and packed cells. Reconstituted WB (300 μL) was put into cups and supplemented with 20 μL of NATEM reagent (CaCl 2 at 17 mM) and platelets were stimulated with 50.0 μM TRAP, 1.25 μM ADP, or U46619 at either 0.1 or 1.5 μM. ROTEM tests were performed according the manufacturer's instructions at 37°C and ran for 60 min. Clotting time (A), alpha angle (B), and maximal clot firmness (C) were analyzed. Statistical analysis was performed by one-way ANOVA followed by Holm-Sidak's post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001 versus resting platelets; ## p < 0.01, ### p < 0.001 with versus without ASA. The results of 10 independent experiments are presented. and Gα12/13 proteins triggered by TRAP and the effects of Gαq and Gαi triggered by ADP. This difference may be due to higher potential effect of TRAP as the strongest platelet agonist on platelet function, compared to ADP. Furthermore, the fact that signaling of Gαq and Gα12/13 at high concentration of U46619 is stronger than selective signaling of Gα12/13 at low concentration of U46619 does not deny the involvement of Gα12/13 in platelet function because high amount U46619 may exert more intensive activation of Gα12/13, which by itself may mediate this effect of U46619.
Thus, this is the first study demonstrating the contribution of platelet G protein signaling to platelet activation and clot formation in condition when the additive platelet activation by αIIbβ3 integrin outside-in signaling was precluded. We show that signaling of Gαq or Gαi but not Gα12/13 exerted platelet activation and contributed to WB clot formation. Nevertheless maximal effect on platelet activation was achieved using all three G proteins together. We suggest that outside-in signaling is not necessarily required to fulfill these platelet functions.