RhoA and Rac1 GTPases Differentially Regulate Agonist-Receptor Mediated Reactive Oxygen Species Generation in Platelets

Agonist induced generation of reactive oxygen species (ROS) by NADPH oxidases (NOX) enhances platelet aggregation and hence the risk of thrombosis. RhoA and Rac1 GTPases are involved in ROS generation by NOX in a variety of cells, but their roles in platelet ROS production remain unclear. In this study we used platelets from RhoA and Rac1 conditional knockout mice as well as human platelets treated with Rhosin and NSC23767, rationally designed small molecule inhibitors of RhoA and Rac GTPases, respectively, to better define the contributions of RhoA and Rac1 signaling to ROS generation and platelet activation. Treatment of platelets with Rhosin inhibited: (a) U46619 induced activation of RhoA; (b) phosphorylation of p47phox, a critical component of NOX; (c) U46619 or thrombin induced ROS generation; (d) phosphorylation of myosin light chain (MLC); (e) platelet shape change; (f) platelet spreading on immobilized fibrinogen; and (g) release of P-selectin, secretion of ATP and aggregation. Conditional deletion of RhoA or Rac1 gene inhibited thrombin induced ROS generation in platelets. Addition of Y27632, a RhoA inhibitor, NSC23766 or Phox-I, an inhibitor of Rac1-p67phox interaction, to human platelets blocked thrombin induced ROS generation. These data suggest that: (a) RhoA/ROCK/p47phox signaling axis promotes ROS production that, at least in part, contributes to platelet activation in conjunction with or independent of the RhoA/ROCK mediated phosphorylation of MLC; and (b) RhoA and Rac1 differentially regulate ROS generation by inhibiting phosphorylation of p47phox and Rac1-p67phox interaction, respectively.

Agonist induced generation of reactive oxygen species (ROS) including superoxide anion (O -2 ) and hydrogen peroxide (H 2 O 2 ) enhance platelet aggregation and hence the risk of thrombosis [14,15]. Although diverse biochemical reactions contribute to ROS generation, NADPH oxidases (NOX) have emerged as critical sources of agonist induced ROS generation [16]. Two isoforms of NOX, namely NOX1 and NOX2 and their regulatory subunits p22 phox , p47 phox and p67 phox , have been characterized in platelets [14,15,17,18] and recent reports have confirmed that ROS activities play a role in regulation of platelet activation [18][19][20][21]. In particular, Delaney et al. [18] have recently reported that ROS generation by NOX leads to platelet secretion and aggregation via the Syk/phospholipase Cγ2/calcium signaling pathway.
Previous studies have shown that small GTPases RhoA [22] and Rac1 [23,24] are involved in NOX activation. RhoA can trigger ROS generation via the ROCK mediated phosphorylation of p47 phox [22] whereas Rac GTPases activate NOX by interacting with p67 phox to promote its binding to NOX [16,24]. Moreover, it has been shown that ROS generation by Rac1 does not involve phosphorylation of p47 phox [24].
Agonist-receptor induced phosphorylation of myosin light chain (MLC) via Gα 13 /RhoA/ ROCK leads to platelet shape change and secretion [25]. However, RhoA is activated not only by agonist-receptor mediated activation of Gα 13 , but also directly and reversibly by ROS leading to stress fiber formation [26]. The bidirectional positive feedback loops for activation of RhoA by ROS and generation of ROS by RhoA suggest that a RhoA-ROS signaling circuit is involved in regulation of platelet activation. Reports that inhibition of RhoA blocks activation of ROCK as well as phosphorylation of p47 phox and ROS production [22] further support the possibility that a RhoA/ROS signaling contributes to platelet activation. In this study we investigated the effects of gene targeting or pharmacologic inhibition of RhoA on ROS generation by thrombin or U46619, two of the agonists known to activate RhoA [27] and generate ROS [20], to better understand the role of RhoA in ROS generation and platelet activation.
We have shown earlier that deficiency or inhibition of Rac1 GTPase inhibits platelet secretion and aggregation induced by diverse agonists including thrombin and U46619 [8]. Although Rac1 GTPase has been shown to be critical in ROS generation by NOX enzymes [16] so far its role in agonist induced ROS generation in platelets has remained to be determined.
Here we report that RhoA, through ROCK/p47 phox signaling, generates ROS that, at least in part, contributes to platelet activation in conjunction with or independent of the RhoA mediated ROCK phosphorylation of MLC. Our data show that RhoA and Rac1 differentially regulate ROS generation by inhibiting phosphorylation of p47 phox and Rac1-p67 phox interaction, respectively.

Materials
Rhosin was custom synthesized as described [28]. Collagen was obtained from Chrono-Log Corporation (Havertown, PA). The anti-PAK, anti-phospho-PAK, anti-p47phox and anti-GAPDH antibodies were purchased from Cell Signaling Technology, Boston, MA. The antiphosho-p47phox was purchased from MyBioSource, San Diego, CA. HRP-conjugated goat anti-mouse IgG and HRP-conjugated goat anti-rabbit IgG were obtained from Thermo Scientific-Pierce, Rockford, IL. All other chemicals and reagents were purchased either from Sigma-Aldrich or from specifically noted sources. murine platelets containing apyrase (3 U/ml) were layered over cover slips in the presence or absence of Rhosin. After a five minute incubation at 37°C the cover slips were rinsed with PBS to remove free platelets. Platelets on cover slips were then fixed with 4% paraformaldehyde for ten minutes, rinsed with PBS twice and permeabilized with 0.1% Triton X-100 for 60 seconds. After two rinses with PBS platelets were stained with Alexa 594-phalloidin to visualize F-actin [5]. A Carl Zeiss LSM-510 confocal Axioplan 200 microscope and a Plan-Neofluar 100x/1.45 oil objective was used to generate platelet images. Digital images were processed using Zen 2007 software from Carl Zeiss.
Statistical analysis. Data are expressed as means ± SD or SE as described in figure legend. A p value of <0.05 indicates statistically significant difference between the control and test samples.

Rhosin inhibited RhoA GTPase activation in platelets
We have shown earlier that Rhosin specifically inhibits activation of RhoA in the NIH 3T3 cells by binding to RhoA at the site required for its activation by Rho-GEF [28]. In this study we investigated the effects of Rhosin on activation of platelet Rho GTPases to demonstrate that Rhosin specifically inhibits activation of RhoA but not that of Rac1 and Cdc42. Washed human platelets were incubated with Rhosin or DMSO for two minutes and then stimulated withTXA 2 analog U46619, a known inducer of RhoA activation. Blots in Fig 1A and 1C show that Rhosin inhibited U46619 induced RhoA activation in a concentration-dependent manner. The data in bar graph ( Fig 1D) show that Rhosin significantly inhibited RhoA activation with minimal effects on Rac1 or Cdc42 activation.

Gene targeting of RhoA deleted expression of RhoA in platelets
RhoA induces ROS generation via ROCK mediated phosphorylation of p47phox [22]. We investigated the possibility that if RhoA is involved in ROS generation then genetic deficiency of RhoA should inhibit ROS generation. To test this possibility we generated RhoA deficient mice as detailed in the methods section and characterized expression of RhoA and related Rho proteins Rac1 and Cdc42 in washed platelets from RhoA -/and matching RhoA +/+ mice by Western blotting. Blots in Fig 1E and the bar graph ( Fig 1F) show that gene targeting of RhoA completely depleted RhoA with a minimal effect on the expression of Rac1 or Cdc42.

Inhibition of RhoA GTPase blocked phosphorylation of p47 phox
Platelet activation by diverse agonists leads to reactive oxygen species (ROS) generation and ROS have been shown to regulate platelet activation [14]. Based on the reports that RhoA/ ROCK mediated phosphorylation of p47 phox , the organizing subunit of NOX, regulates ROS generation and ROS in turn directly activates RhoA, we investigated the possibility that RhoA also regulates platelet activation by affecting ROS generation. A two minute pre-incubation of washed human platelets with Rhosin inhibited U46619 or thrombin induced phosphorylation of p47 phox (Fig 2A). Addition of Y27632, a known inhibitor of RhoA/ROCK signaling [27], to platelets also blocked phosphorylation of p47 phox (Fig 2A). However, Phox-I, an inhibitor of Rac1-p67phox interaction [33], necessary for NOX mediated ROS generation, did not inhibit phosphorylation of p47phox ( Fig 2C). Quantitative analysis of these blots indicates that inhibition of RhoA GTPase by Rhosin or Y27632 prevents phosphorylation of p47 phox (Fig 2B) whereas inhibition of Rac1-p67phox interaction does not affect phosphorylation of p47phox ( Fig 2D). Washed human platelets were incubated with U46619 (0.01 μM) for one minute. The reactions were terminated by adding ice-cold HEPES-buffered Tyrode's solution containing protease inhibitors cocktail. GTP loading of RhoA, Rac1 and Cdc42 was analyzed as described in the methods section. A two minute pre-incubation of platelets with Rhosin inhibited U46619 induced RhoA-GTP formation in a concentration-dependent manner. Rhosin minimally inhibited Cdc42-GTP and Rac1-GTP formation. (E-F) Conditional RhoA knockout mice were generated as described previously [29]. The Western blots and the bar graph show that gene targeting of RhoA completely deleted RhoA expression and partially decreased Cdc42 or Rac1 expression in platelets. Total RhoA, Rac1, Cdc42 and GADPH are shown as loading controls. The data in bar graphs are mean ± SE from three experiments (*p<0.05).

Inhibition of RhoA GTPase blocked ROS generation
Since activation of p47 phox is required for agonist induced ROS generation by NOX, we next investigated the possibility that inhibition of p47phox phosphorylation should prevent ROS generation. Addition of Rhosin to platelets blocked ROS generation induced by U46619 ( Fig  3A) or thrombin (Fig 3B). Treatment of platelets with Y27632 also inhibited thrombin induced ROS generation (Fig 3C). We further investigated the role of RhoA in ROS generation using the RhoA -/platelets. As shown in Fig 3D, addition of thrombin to platelets from RhoA -/mice generated significantly less ROS than platelets from the matching wild type mice. These data indicate that RhoA regulates NOX mediated ROS generation by ROCK mediated phosphorylation of p47 phox .  Inhibition of RhoA prevented phosphorylation of myosin light chain and platelet shape change Activated RhoA increases phosphorylation of myosin light chain (MLC) via its effector ROCK by inhibiting MLC phosphatase [34]. We investigated the possibility that if Rhosin prevents RhoA activation then it should also inhibit phosphorylation of MLC. Addition of Rhosin (30 μM) to aspirin (1 mM) treated platelets containing apyrase (3 U/ml) two minutes prior to stimulation with U46619 or thrombin inhibited phosphorylation of MLC (Fig 4A and 4B). These findings further confirm that Rhosin is an effective inhibitor of RhoA activation in platelets.
The role of RhoA in inducing phosphorylation of MLC in cytoskeletal reorganization leading to platelet shape change is well known [27]. We investigated the possibility that Rhosin by inhibiting RhoA/ROCK mediated MLC phosphorylation blocks platelet shape change. The effect of Rhosin on platelet shape change was recorded using an aggregometer by monitoring the decrease in light transmittance following addition of an agonist. U46619 or thrombin induced platelet shape change in aspirin (1 mM) treated washed human platelets in the presence of apyrase (3 U/ml). Incubation of platelets with Rhosin (30 μM) inhibited U46619 or thrombin induced platelet shape change (Fig 4C and 4D). These results show that Rhosin blocks U46619 or thrombin induced platelet shape change mediated by RhoA.

Inhibition of RhoA inhibited platelet spreading on immobilized fibrinogen
Binding of ligands such as fibrinogen to integrin αIIbβ 3 induces outside-in signaling leading to cytoskeletal reorganization that results in morphological changes namely platelet spreading, formation of filopodia, lamellipodia and stress fibers. RhoA has been shown to be involved in platelet cytoskeletal reorganization. We therefore investigated the possibility that inhibition of RhoA by Rhosin may prevent outside-in signaling mediated platelet spreading. Aspirin (1mM) treated washed RhoA +/+ platelets with or without Rhosin and RhoA -/platelets were layered over immobilized fibrinogen and platelet morphological changes were visualized by confocal microscopy. Rhosin treated platelets (Fig 5B), and RhoA-deficient platelets (Fig 5C), as compared to matching controls ( Fig 5A) exhibited significantly less spreading on fibrinogen and a decrease in the numbers of filopodia. Platelet spreading was diminished in Rhosin treated or RhoA -/platelets by 64% and 71% respectively (Fig 5D). Eighty percent of the DMSO treated, 44% of Rhosin treated and only 22% of RhoA -/platelets exhibited filopodia. In platelets expressing filopodia Rhosin or RhoA deficiency decreased the number of filopodia by 51% and 63% respectively (Fig 5E). These findings clearly show that RhoA plays a critical role in integrin mediated cytoskeletal reorganization and its deficiency due to gene targeting or inhibition by Rhosin diminishes platelet spreading.

Inhibition of RhoA blocked platelet release of p-selectin, ATP secretion and aggregation
Pharmacologic inhibition or gene targeting of RhoA has been shown to result in defective platelet function [35][36][37]. RhoA deficiency has been reported to result in significantly diminished release of P-selectin in response to thrombin and impaired aggregation induced by thrombin or protease activated receptor peptide-4 [35]. We investigated the possibility that if RhoA is involved in platelet secretion and aggregation than inhibition of RhoA by Rhosin should block secretion from the dense and α-granules as well as platelet aggregation. Addition of Rhosin to platelets two minutes before stimulation with U46619 or thrombin inhibited release of P-selectin (Fig 6A and 6B), secretion of ATP (Fig 6C and 6D) and aggregation ( Fig  6E and 6F) in a concentration-dependent manner. These data indicate that RhoA plays a critical role in platelet activation.

Inhibition or gene targeting of Rac1 GTPase blocked ROS generation
The role of Rac GTPases in ROS generation by NOX1 and NOX2 has been well documented [16]. Rac GTPases activate NOX by enhancing binding of p67 phox to NOX2 [16]. If Rac1 GTPase is essential for ROS generation then gene targeting or inhibition of Rac GTPases should inhibit ROS generation in platelets. To test this possibility we investigated ROS generation in platelets from Rac1 -/mice and platelets treated with NSC23766, a specific inhibitor of Rac GTPases [38]. Addition of NSC23766 to washed human platelets two minutes before stimulation with thrombin blocked ROS generation ( Fig 7A). As shown in Fig 7B, thrombin stimulation led to significantly less ROS production in platelets from Rac1 -/mice as compared with platelets from the matching wild type mice. These data indicate that Rac1 GTPase plays a critical role in agonist induced ROS generation in platelets. Interaction of activated i.e. GTP bound Rac GTPases with p67 phox is essential for activation of NOX leading to ROS generation [16,[39][40][41]. We investigated the effects of Phox-I, a rationally designed small molecule inhibitor of Rac1-P67 phox interaction [33] to determine if inhibition of Rac1-p67 phox interaction prevents ROS generation. Addition of Phox-I to platelets two minutes before stimulation with thrombin inhibited ROS generation in platelets (Fig 7A). These data suggest that the regulation of ROS generation by Rac1 depends on Rac1-p67 phox interaction.

Discussion
A possible role for ROS in platelet activation was reported more than thirty-nine years ago [42,43]. Since then ROS generation in platelets stimulated by diverse agonists has been shown by flow cytometry [20] and a number of reports have shown that agonist induced ROS generation in platelets involves NADPH oxidase (NOX) [20,[44][45][46]. Platelets have been shown to express NOX1 and NOX2 [17,18] and patients with an inherited deficiency of NOX2 also known as gp91 phox have been reported to have impaired platelet function [47]. Rho family GTPases Rac1 and RhoA have been shown to regulate ROS generation by NOX. RhoA has been shown to play a critical role in agonist-receptor mediated platelet activation via the classical RhoA/ ROCK/MLC phosphorylation pathway [35][36][37]48]. However, platelet activation by diverse agonists not only leads to phosphorylation of MLC but also phosphorylation of p47 phox and ROS generation. The ability of ROS to directly and reversibly activate RhoA leading to stress fiber formation [26] suggests that activation of RhoA is yet another mechanism involved in ROS mediated platelet activation. In this study thrombin and U46619, two of the agonists known to activate RhoA GTPase, were used to better understand the role of RhoA in ROS generation and platelet activation.
Our findings that platelets treated with Rhosin, a rationally designed small molecule inhibitor of RhoA [28], and platelets from RhoA -/mice generated less ROS in response to thrombin (Fig 3B and 3D) shows for the first time that RhoA plays a critical role in agonist-induced ROS generation in platelets. Inhibition of thrombin induced ROS generation in platelets treated with Y27632 (Fig 3C), a known inhibitor of RhoA [27], or in platelets from RhoA -/-, as compared to RhoA +/+ , platelets (Fig 3D) provide further evidence that RhoA is involved in agonist induced ROS generation in platelets.
A role of agonist/RhoA/ROCK mediated MLC phosphorylation in platelet shape change and secretion has been known for some time [36,37,48]. Inhibition of RhoA blocks downstream effector ROCK and consequently ROCK mediated phosphorylation of target proteins including MLC. Our findings that Rhosin inhibited thrombin or U46619 induced phosphorylation of MLC (Fig 4A and 4B), and shape change (Fig 4C and 4D) in the presence of aspirin and apyrase, inhibitors of secondary mediators namely TXA 2 and ADP released during platelet activation, suggest that Rhosin inhibits platelet shape change by blocking RhoA/ROCK knockout mice were generated as described previously [30]. Addition of thrombin (0.1 U/ml) induced significantly less ROS generation in Rac1 -/-, as compared to Rac1 +/+ , platelets (mean ± SE, n = 4, *p < 0.001). Generation of reactive oxygen species in dcf-da loaded washed platelets was monitored by flow cytometry as detailed in the Methods section.
doi:10.1371/journal.pone.0163227.g007 mediated phosphorylation of MLC. These findings are in agreement with a recent report that gene targeting of RhoA abolishes thrombin or U46619 induced phosphorylation of MLC and platelet shape change [35].
Our findings that pharmacologic targeting of RhoA by Rhosin or genetic deletion of RhoA inhibited spreading of platelets on immobilized fibrinogen (Fig 5) suggest that RhoA is involved in integrin αIIbβ 3 -dependent spreading of platelets. Others have reported that RhoA is either required or is not needed for platelet spreading to occur [49,50]. Gong et al. [51] have reported that initially integrin αIIbβ 3 -dependent inhibition of RhoA leads to platelet spreading and at the later stage termination of RhoA inhibition leads to RhoA-dependent contraction. Reasons for the discrepancies in the role of RhoA in platelet spreading is not clear at this time. However, different fibrinogen coating densities have been shown to dramatically affect integrin αIIbβ 3 -mediated platelet signaling and spreading [52]. We incubated platelets on fibrinogen (3 μg/ml) for five minutes. Gong et al. [51] incubated platelets for 90 minutes on 100 μg/ml fibrinogen coated coverslips [26]. Pleines et al. used 200 μg/ml fibrinogen to study spreading of RhoA -/platelets for thirty minutes [35]. The differences between the density of fibrinogen and or duration of time platelets were exposed to fibrinogen used in our and other studies may be responsible for the discrepancies in our and their observations.
RhoA has been shown to activate NOX by phosphorylating p47phox [22] and NOX generated ROS leads to platelet secretion and aggregation via the Syk/phospholipase Cγ2/calcium signaling pathway [18]. Our findings that Rhosin inhibited RhoA/ROCK/p47 phox /NOX mediated ROS generation (Fig 3) and U46619 or thrombin induced release of P-selectin, ATP secretion and aggregation (Fig 6) suggests that inhibition of RhoA/ROCK/NOX/ROS, at least in part, prevents platelet activation in conjunction with or independent of the RhoA/ROCK/MLC signaling.
Rac GTPases are integral part of ROS generation by NADPH oxidase isoforms NOX1 and NOX2. Binding of activated Rac GTPases to p67 phox activates NOX2 by facilitating binding of p67 phox to NOX2. Our findings that inhibition of Rac1 GTPase by NSC23766 (Fig 7A) or gene targeting of Rac1GTPase (Fig 7B) diminished ROS generation in platelets clearly demonstrate that Rac1 is essential for agonist induced ROS generation in platelets. Moreover, our data showing that Phox-I, an inhibitor of Rac1-p67 phox interaction, blocks thrombin induced ROS generation ( Fig 7A) without affecting phosphorylation of p47 phox (Fig 2C and 2D) concur with other published reports [16,33] that Rac1-p67 phox /NOX signaling plays a critical role in Rac1 mediated ROS generation. These observations, together with a report by others that Rac1 is not involved in phosphorylation of p47 phox [24] suggest that Rac1 and RhoA utilize distinct signaling to regulate agonist induced ROS generation in platelets.