G protein subunit Gα₁₃ binds to integrin α_IIbβ₃ and mediates integrin “outside-in” signaling

Abstract

Integrins mediate cell adhesion to the extracellular matrix and transmit signals within the cell that stimulate cell spreading, retraction, migration, and proliferation. The mechanism of integrin outside-in signaling has been unclear. We found that the heterotrimeric guanine nucleotide-binding protein (G protein), Gα₁₃, directly bound to the integrin β₃ cytoplasmic domain, and that Gα₁₃-integrin interaction was promoted by ligand binding to the integrin α_IIbβ₃ and by guanosine triphosphate (GTP)-loading of Gα₁₃. Interference of Gα₁₃ expression or a myristoylated fragment of Gα₁₃ that inhibited interaction of α_IIbβ₃ with Gα₁₃ diminished activation of protein kinase c-Src and stimulated the small GTPase RhoA, consequently inhibiting cell spreading and accelerating cell retraction. We conclude that integrins are non-canonical Gα₁₃-coupled receptors that provide a mechanism for dynamic regulation of RhoA.

Integrins mediate cell adhesion and transmit signals within the cell that lead to cell spreading, retraction, migration, and proliferation (1). Thus, integrins have pivotal roles in biological processes such as development, immunity, cancer, wound healing, hemostasis and thrombosis. The platelet integrin, α_IIbβ₃, typically displays bidirectional signaling function (2, 3). Signals from within the cell activate binding of α_IIbβ₃ to extracellular ligands, which in turn triggers signaling within the cell initiated by the occupied receptor (so-called “outside-in” signaling). A major early consequence of integrin “outside-in” signaling is cell spreading, which requires activation of the protein kinase c-Src and c-Src-mediated inhibition of the small guanosine triphosphatase (GTPase) RhoA (4-7). Subsequent cleavage of the c-Src binding site in β₃ by calpain allows activation of RhoA, which stimulates cell retraction (7, 8). The molecular mechanism coupling ligand-bound α_IIbβ₃ to these signaling events has been unclear.

Heterotrimeric guanine nucleotide-binding proteins (G proteins) consist of Gα, Gβ and Gγ subunits (9). G proteins bind to the intracellular side of G-protein coupled receptors (GPCR) and transmit signals that are important in many intracellular events (9-11). Gα₁₃, when activated by GPCRs, interacts with Rho guanine-nucleotide exchange factors (RhoGEF) and thus activates RhoA (11-14), facilitating contractility and rounding of discoid platelets (shape change). To determine whether Gα₁₃ functions in signaling from ligand-occupied integrin, we investigated whether inhibition of Gα₁₃ expression with small interfering RNA (siRNA) affected α_IIbβ₃-dependent spreading of platelets on fibrinogen, which is an integrin ligand. We isolated mouse bone marrow stem cells and transfected them with lentivirus encoding Gα₁₃ siRNA. The transfected stem cells were transplanted into irradiated C57/BL6 mice (15). Four to six weeks after transplantation, nearly all platelets isolated from recipient mice were derived from transplanted stem cells as indicated by the enhanced green fluorescent protein (EGFP) encoded in lentivirus vector (Fig. S1, Fig. 1A). Platelets from Gα₁₃ siRNA-transfected stem cell recipient mice showed >80% decrease in Gα₁₃ expression (Fig. 1B). When platelets were allowed to adhere to immobilized fibrinogen [α_IIbβ₃ binding to immobilized fibrinogen does not require prior “inside-out” signaling activation (16)], platelets depleted of Gα₁₃ spread poorly as compared with control platelets (Fig. 1A, Fig. S2). The inhibitory effect of Gα₁₃ deficiency is unlikely to be caused by its effect on GPCR-stimulated Gα₁₃ signaling because (i) washed resting platelets were used and no GPCR agonists were added, and (ii) prior treatment with 1 mM aspirin [which abolishes thromboxane A₂ (TXA₂) generation (17)] did not affect platelet spreading on fibrinogen (Fig. S2), making it unlikely the endogenous TXA₂-mediated stimulation of Gα₁₃. Furthermore, Gα₁₃ siRNA inhibited spreading of Chinese hamster ovary (CHO) cells expressing human α_IIbβ₃ (123 cells) (18), which was rescued by an siRNA-resistant Gα₁₃ (Fig. S3). Thus, Gα₁₃ appears to be important in integrin “outside-in” signaling leading to cell spreading.

Fig. 1. The role of Gα₁₃ in integrin outside-in signaling.

(A) Confocal microscopy images of spreading scrambled siRNA control platelets or Gα₁₃-depleted platelets (Gα₁₃-siRNA) on fibrinogen, without or with Y27632. Merged EGFP (green) fluorescence and Alex Fluor 546-conjugated phalloidin (Red) fluorescence. (B) Western blot comparison of Gα₁₃ abundance in platelets from mice innoculated with control siRNA- or Gα₁₃-siRNA-transfected bone marrow stem cells. (C, D, E) Mouse platelets from scrambled siRNA- or Gα₁₃ siRNA-transfected stem cells were allowed to adhere to immobilized fibrinogen, solubilized and analyzed for c-Src Tyr⁴¹⁶ phosphorylation and RhoA activation.

To determine whether Gα₁₃ serves as an early signaling mechanism that mediates integrin-induced activation of c-Src, we measured phosphorylation of c-Src at Tyr⁴¹⁶ (which indicates activation of c-Src) in control and fibrinogen-bound cells. Depletion of Gα₁₃ in mouse platelets or 123 cells abolished phosphorylation of c-Src Tyr⁴¹⁶ (Fig. 1C, Fig. S3), indicating that Gα₁₃ may link integrin α_IIbβ₃ and c-Src activation. Because c-Src inhibits RhoA (7, 19), we also tested the role of Gα₁₃ in regulating activation of RhoA. RhoA activity was suppressed to baseline 15 minutes after platelet adhesion, and became activated at 30 minutes (Fig. 1C), which is consistent with transient inhibition of RhoA by c-Src (7). The integrin-dependent delayed activation of RhoA was not inhibited by depletion of Gα₁₃, indicating its independence of the GPCR-Gα₁₃-RhoGEF pathway (Fig. 1C). In contrast, depletion of Gα₁₃ accelerated RhoA activation (Fig. 1C). Thus, Gα₁₃ appears to mediate inhibition of RhoA. The inhibitory effect of Gα₁₃ depletion on platelet spreading was reversed by Rho-kinase inhibitor Y27632 (Fig. 1A), suggesting that Gα₁₃-mediated inhibition of RhoA is important in stimulating platelet spreading. These data are consistent with Gα₁₃ mediating integrin “outside-in signaling” inducing c-Src activation, RhoA inhibition, and cell spreading.

The integrin α_IIbβ₃ was co-immunoprecipitated by anti-Gα₁₃ antibody, but not control IgG, from platelet lysates (Fig. 2A). Conversely, an antibody to β₃ immunoprecipitated Gα₁₃ with β₃ (Fig. 2B). Co-immunoprecipitation of β₃ with Gα₁₃ was enhanced by GTP-γS or AlF₄^- (Fig. 2A, Fig. S4). Thus, β₃ is present in a complex with Gα₁₃, preferably the active GTP-bound Gα₁₃. To determine whether Gα₁₃ directly binds to the integrin cytoplasmic domain, we incubated purified recombinant Gα₁₃ (20) with agarose beads conjugated with glutathione S-transferase (GST), or a GST-β₃ cytoplasmic domain fusion protein (GST-β₃CD). Purified Gα₁₃ bound to GST-β₃CD, but not to GST (Fig. 2C). Purified Gα₁₃ also bound to the β₁ integrin cytoplasmic domain fused with GST (GST-β₁CD) (Fig. 2D). The binding of Gα₁₃ to GST-β₃CD and GST-β₁CD was detected with GDP-loaded Gα₁₃, but enhanced by GTP-γS and AlF₄^- (Fig. 2C, 2D), indicating that the cytoplasmic domains of β₃ and β₁ can directly interact with Gα₁₃, and GTP enhances the interaction. The Gα₁₃-β₃ interaction was enhanced in platelets adherent to fibrinogen, and by thrombin, which stimulates GTP binding to Gα₁₃ via GPCR (Fig. 2E). Hence, the interaction is regulated by both integrin occupancy and GPCR signaling.

Fig. 2. Binding of Gα₁₃ to β₃ and the inhibitory effect of mSRI peptide.

(A) Proteins from platelet lysates were immunoprecipitated with control IgG or antibody to Gα₁₃ with or without 1 μM GDP, 1 μM GTP or 30 μM AlF₄^-. Immunoprecipitates were immunoblotted with anti-Gα₁₃ or anti-β₃ (MAb15). See Fig. S4 for quantitation. (B) Proteins from platelet lysates were immunoprecipitated with control mouse IgG, anti-α_IIbβ₃ (D57 (24)) or an antibody to the glycoprotein Ibα (GPIb). Immunoprecipitates were immunoblotted with anti-Gα₁₃, anti-β₃, or anti-GPIb antibodies. (C, D) Purified GST-β3CD (C) or GST-β1CD (D) bound to glutathione beads was mixed with purified Gα₁₃ with or without 1 μM GDP, 1 μM GTPγS or 30 μM AlF₄^-. Bound proteins were immunoblotted with anti-Gα₁₃. Quantitative data are shown as mean±SD and p value (t-test). (E) Lysates of control platelets or platelets adherent to fibrinogen in the absence or presence of 0.025 U/ml thrombin were immunoprecipitated with anti-Gα₁₃, and then immunoblotted with MAb15. Quantitative data are shown as mean±SD and p value (t-test). (F) Lysates from 293FT cells transfected with Flag-tagged wild type Gα₁₃ or indicated truncation mutants (see Fig. S5) were precipitated with GST-β₃CD- or GST-bound glutathione beads. Bead-bound proteins were immunoblotted with anti-Flag (Bound). Flag-tagged protein amounts in lysates are shown by anti-Flag immunoblot (Input). (G) Protein from platelet lysates treated with 0.1% DMSO, 250 μM scrambled control peptide (Ctrl) or mSRI were immunoprecipitated with anti-Gα₁₃. Immunoprecipitates were immunoblotted with anti-Gα₁₃ or anti-β₃. See Fig. S4 for quantitation.

To map the β₃ binding site in Gα₁₃, we incubated cell lysates containing Flag-tagged wild type or truncation mutants of Gα₁₃ (Fig. S5) with GST-β₃CD beads. GST-β₃CD associated with wild type Gα₁₃ and the Gα₁₃ 1-212 fragment containing α helical region and switch region I (SRI), but not with the Gα₁₃ fragment containing residues 1-196 lacking SRI (Fig. 2F). Thus, SRI appears to be critical for β₃ binding. To further determine the importance of SRI, Gα₁₃-β₃ binding was assessed in the presence of a myristoylated synthetic peptide, Myr-LLARRPTKGIHEY (mSRI), corresponding to the SRI sequence of Gα₁₃ (197-209) (21). The mSRI peptide, but not a myristoylated scrambled peptide, inhibited Gα₁₃ binding to β₃ (Fig. 2G), indicating that mSRI is an effective inhibitor of β₃-Gα₁₃ interaction. Therefore, we further examined whether mSRI might inhibit integrin signaling. Treatment of platelets with mSRI inhibited integrin-dependent phosphorylation of c-Src Tyr⁴¹⁶ and accelerated RhoA activation (Fig. 3A). The effect of mSRI is unlikely to result from its inhibitory effect on the binding of RhoGEFs to Gα₁₃ SRI because Gα₁₃ binding to RhoGEFs stimulates RhoA activation, which should be inhibited rather than promoted by mSRI (21). Thus, these data suggest that β₃-Gα₁₃ interaction mediates activation of c-Src and inhibition of RhoA. Furthermore, mSRI inhibited integrin-mediated platelet spreading (Fig. 3B), and this inhibitory effect was reversed by C3 toxin (which catalyzes the ADP ribosylation of RhoA) or Y27632, confirming the importance of Gα₁₃-dependent inhibition of RhoA in platelet spreading. Thrombin promotes platelet spreading, which requires cdc42/Rac pathways (22). However, thrombin-promoted platelet spreading was also abolished by mSRI (Fig. 3B), indicating the importance of Gα₁₃-β₃ interaction. Thus, Gα₁₃-integrin interaction appears to be a mechanism that mediates integrin signaling to c-Src and RhoA, thus regulating cell spreading.

Fig. 3. Effects of mSRI on integrin-induced c-Src phosphorylation, RhoA activity and platelet spreading.

(A) Washed human platelets pre-treated with DMSO, mSRI, or scrambled control peptide were allowed to adhere to fibrinogen and then solubilized at indicated time points. Proteins from lysates were immunoblotted with antibodies to c-Src phosphorylated at Tyr⁴¹⁶, c-Src, or RhoA. GTP-bound RhoA was measured by association with GST-RBD beads (25). See Fig. S4 for quantitative data. (B) Spreading of platelets treated with 0.1% DMSO, scrambled control peptide, or mSRI, in the absence or presence of C3 toxin, Y27632, or 0.01 U/ml thrombin. Platelets were stained with Alexa Fluor 546–conjugated phalloidin.

To further determine whether Gα₁₃ mediates inhibition of integrin-induced RhoA-dependent contractile signaling, we investigated the effects of mSRI and depletion of Gα₁₃ on platelet-dependent clot retraction (shrinking and consolidation of a blood clot requires integrin-dependent retraction of platelets from within) (7, 8). Clot retraction was accelerated by mSRI and depletion of Gα₁₃ (Fig. 4, A and B, Fig. S6), indicating that Gα₁₃ negatively regulates RhoA-dependent platelet retraction and coordinates cell spreading and retraction. The coordinated cell spreading-retraction process is also important in wound healing, cell migration and proliferation (23).

Fig. 4. The role of Gα₁₃ in clot retraction and dynamic RhoA regulation.

(A) Effect of 250 μM mSRI peptide on clot retraction of human platelet-rich plasma compared with DMSO and scrambled peptide. Clot sizes were quantified using Image J (mean±SD, n=3, t-test). (B) Comparison of clot retraction (mean±SD, n=3, t-test) mediated by control siRNA platelets and Gα₁₃-depleted platelets. (C, D, E, F) Platelets were stimulated with thrombin with or without 2 mM RGDS, and monitored for turbidity changes of platelet suspension caused by shape change and aggregation (C). The platelets were then solubilized at indicated time points and analyzed for amount of β₃ coimmunoprecipitated with Gα₁₃ (D) and amount of GTP-RhoA bound to GSTRBD beads (E) by immunoblot. (F) quantitative data (mean±SD) from 3 experiments. (G) A schematic illustrating Gα₁₃-dependent dynamic regulation of RhoA and crosstalk between GPCR and integrin signaling.

The function of Gα₁₃ in mediating the integrin-dependent inhibition of RhoA contrasts with the traditional role of Gα₁₃, which is to mediate GPCR-induced activation of RhoA. However, GPCR-mediated activation of RhoA is transient, peaking at 1 minute after exposure of platelets to thrombin, indicating the presence of a negative regulatory signal (Fig. 4, D and F). Furthermore, thrombin-stimulated activation of RhoA occurs during platelet shape change before substantial ligand binding to integrins (Fig. 4, C, D and F). In contrast, following thrombin stimulation, β₃ binding to Gα₁₃ was diminished at 1 minute when Gα₁₃-dependent activation of RhoA occurs, but increased after the occurrence of integrin-dependent platelet aggregation (Fig. 4, E and F). Thrombin-stimulated binding of Gα₁₃ to α_IIbβ₃ and simultaneous RhoA inhibition both require ligand occupancy of α_IIbβ₃, and are inhibited by the integrin inhibitor RGDS (Fig. 4, D-F). Thus, our study demonstrates not only a function of integrin α_IIbβ₃ as a non-canonical Gα₁₃-coupled receptor but also a new concept of Gα₁₃-dependent dynamic regulation of RhoA, in which Gα₁₃ mediates initial GPCR-induced RhoA activation and subsequent integrin-dependent RhoA inhibition (Fig. 4G). These findings are important for our understanding on how cells spread, retract, migrate, and proliferate, which is fundamental to development, cancer, immunity, wound healing, hemostasis and thrombosis.