Gα13 Switch Region 2 Binds to the Talin Head Domain and Activates αIIbβ3 Integrin in Human Platelets*

Background: Switch Regions of G protein Gα subunits activate downstream cell signaling events. Results: Gα13 Switch Region 2 forms a calcium-dependent bi-molecular complex with the head domain of talin. Conclusion: Binding of the Gα13 to the talin head domain promotes αIIbβ3 integrin activation. Significance: These results provide a new paradigm for inside-out signaling and αIIbβ3 integrin activation in platelets. Even though GPCR signaling in human platelets is directly involved in hemostasis and thrombus formation, the sequence of events by which G protein activation leads to αIIbβ3 integrin activation (inside-out signaling) is not clearly defined. We previously demonstrated that a conformationally sensitive domain of one G protein, i.e. Gα13 switch region 1 (Gα13SR1), can directly participate in the platelet inside-out signaling process. Interestingly however, the dependence on Gα13SR1 signaling was limited to PAR1 receptors, and did not involve signaling through other important platelet GPCRs. Based on the limited scope of this involvement, and the known importance of G13 in hemostasis and thrombosis, the present study examined whether signaling through another switch region of G13, i.e. Gα13 switch region 2 (Gα13SR2) may represent a more global mechanism of platelet activation. Using multiple experimental approaches, our results demonstrate that Gα13SR2 forms a bi-molecular complex with the head domain of talin and thereby promotes β3 integrin activation. Moreover, additional studies provided evidence that Gα13SR2 is not constitutively associated with talin in unactivated platelets, but becomes bound to talin in response to elevated intraplatelet calcium levels. Collectively, these findings provide evidence for a novel paradigm of inside-out signaling in platelets, whereby β3 integrin activation involves the direct binding of the talin head domain to the switch region 2 sequence of the Gα13 subunit.

Even though GPCR signaling in human platelets is directly involved in hemostasis and thrombus formation, the sequence of events by which G protein activation leads to ␣IIb␤3 integrin activation (inside-out signaling) is not clearly defined. We previously demonstrated that a conformationally sensitive domain of one G protein, i.e. G␣ 13 switch region 1 (G␣ 13 SR1), can directly participate in the platelet inside-out signaling process. Interestingly however, the dependence on G␣ 13 SR1 signaling was limited to PAR1 receptors, and did not involve signaling through other important platelet GPCRs. Based on the limited scope of this involvement, and the known importance of G 13 in hemostasis and thrombosis, the present study examined whether signaling through another switch region of G 13 , i.e. G␣ 13 switch region 2 (G␣ 13 SR2) may represent a more global mechanism of platelet activation. Using multiple experimental approaches, our results demonstrate that G␣ 13 SR2 forms a bimolecular complex with the head domain of talin and thereby promotes ␤3 integrin activation. Moreover, additional studies provided evidence that G␣ 13 SR2 is not constitutively associated with talin in unactivated platelets, but becomes bound to talin in response to elevated intraplatelet calcium levels. Collectively, these findings provide evidence for a novel paradigm of insideout signaling in platelets, whereby ␤3 integrin activation involves the direct binding of the talin head domain to the switch region 2 sequence of the G␣ 13 subunit.
Cells possess multiple G protein signaling pathways that contribute to the different cellular responses involved in development, function, survival, and disease (1,2). Consequently, G protein signaling remains an area of intense investigation. It is well established that classic heterotrimeric G protein signaling first requires ligand binding to a G-protein-coupled receptor. The G-protein subunits (G␣, G␤, and G␥) then become activated and interact with downstream effectors to initiate signal transduction events (3,4). In this regard, the conformationally sensitive switch regions 1 and 2 of the G␣ subunit are known to be directly involved in G protein activation and participate in its downstream signaling mechanisms (5)(6)(7). G 13 is one such G protein that is ubiquitously expressed and abundant in all cell types. It is known to be essential for numerous in vivo processes including embryogenesis, angiogenesis, chemokinesis, hemostasis, and thrombosis (2)(3)(4). In this connection, we previously demonstrated that human platelet shape change, aggregation, and secretion can be dependent on G␣ 13 switch region 1 (G␣ 13 SR1) 3 signaling (8). However, these studies also provided evidence that the critical importance of this G␣ 13 SR1 signaling pathway is limited to PAR1-mediated platelet activation. Based on this consideration, the present study examined whether a separate G␣ 13 switch region signaling mechanism, i.e. G␣ 13 SR2 may explain the global importance of G 13 for in vivo platelet function. Using peptide affinity chromatography of native platelet proteins and immunoaffinity purification of native platelet G␣ 13 -protein interactions, our results demonstrate that the amino acid sequence of G␣ 13 SR2 (but not G␣ 13 SR1) directly binds to the talin-␣IIb␤3 integrin-kindlin-3 complex in human platelets. Furthermore, dissociation of this complex revealed that the binding partner for G␣ 13 SR2 is the head domain (also designated as FERM domain) of talin (and not ␣IIb, ␤3 integrin or kindlin-3. Importantly, this G␣ 13 SR2-talin binding interaction was promoted by increased intraplatelet calcium levels and prevented by calcium chelation. The ability of talin to form a specific complex with G␣ 13 SR2 was further confirmed by bi-molecular binding measurements using recombinant talin head domain, G␣ 13 SR2 peptides and GST-G 13 SR2 fusion proteins. Lastly, studies measuring fibronectin adhesion of NIH3T3 fibroblasts suggest that the binding interaction between talin and G␣ 13 SR2 is not limited to platelet signaling, but may represent a more universal mechanism of integrin activation.
Human Platelet Functional Studies-The platelet count in the freshly drawn PRP was adjusted to 3 ϫ 10 8 platelets/ml with calcium-free Tyrode's buffer (pH 7.4). Platelets were pre-incubated with peptides or vehicle for 1 min prior to incubation with U46619, TRAP1, TRAP4, ADP, or A23187. Aggregation was measured using the turbidimetric method (9), with a model 400 Chrono-Log aggregometer.
Solubilized Platelet Membrane Preparation-Solubilized platelet membranes were prepared as previously described (10). Briefly, platelets were sonicated and the membranes were sedimented by ultracentrifugation (100,000 ϫ g) for 45 min at 4°C. The membrane pellet was then solubilized in cold buffer (25 mM Tris-HCl, 5 mM MgCl 2 , pH 7.4, plus 10 mM CHAPS and 5 mM DTT).
Ca 2ϩ Mobilization Assay-Human PRP was diluted to 2.5 ϫ 10 6 platelets/ml in Tyrode's buffer containing 0.1% BSA. Cells were loaded for 1 h at 37°C with FLIPR calcium-sensitive dye, according to the manufacturer's protocol. The samples were measured using a FlexStation plate reader (Molecular Devices). Cells were excited at 485 nm, and Ca 2ϩ fluorescence was detected at an emission wavelength of 525 nm (11).
BAPTA Loading of Platelets-Platelets in PRP were loaded with diethel-BAPTA-AM (15 M) for 30 min at 37°C.
RhoA G-LISA Activation-RhoA activation was measured with a G-LISA™ assay. The platelets were treated with agonists alone or were pre-incubated with 250 M peptide prior to stimulation with the agonists. After 5 min, platelet lysates were prepared, and the bound active Rho-family protein was detected at 490 nm.
Src ELISA Activation-Src activation was measured with a Src ELISA activation assay. The platelets were treated with agonists alone or were pre-incubated with 250 M of peptide prior to stimulation with the agonists. After 15 min, platelet lysates were prepared, and the bound phosphorylated Src was detected at 450 nm.
SulfoLink Peptide Affinity Columns-All the peptides used for affinity chromatography were synthesized with a N-terminal cystine and coupled to the SulfoLink agarose beads according to the manufacturer's protocol.
Antibody Affinity Columns-Antibody against G␣ 13 was raised in rabbits immunized with N-terminal peptide (40 -48 amino acid sequence) of G␣ 13 (12). This antibody was used to generate an immunoaffinity protein-A Sepharose column to purify G␣ 13 and its associated proteins from solubilized human platelets. The specific column-bound proteins were eluted using a 2.5 pH glycine buffer into neutralization buffer (pH 8.0).
Flow Cytometry Analysis of Platelets using FITC-labeled PAC1 Antibody-The platelet count in human PRP was adjusted to 1 ϫ 10 6 platelets/ml using Tyrode's buffer (pH 7.4). The platelets were stimulated with 0.5 M U46619 or 25 M TRAP1 for 5 min and then labeled for 15 min at room temperature with FITC-conjugated PAC1 antibody against activated ␣llb␤3. The samples were then analyzed by flow cytometry in a Beckman Coulter CyAn II Flow Cytometer.
Dot Blot Assay-Recombinant proteins were pipetted directly onto nitrocellulose blotting membranes, rinsed briefly in PBS (pH 7.4) and 0.1% Tween 20 (PBST), and incubated with 5% Milk-PBST for 1 h at room temperature. The membrane was rinsed briefly and incubated with 5% BSA-PBST with 0.5 M biotinylated peptide for 1 h at room temperature. Subsequently, each blot was incubated with 5% BSA-PBST with streptavidin-HRP for 30 min at room temperature. Each blot was developed using ECL chemiluminescent substrates.
GST Pull-down Assay-Recombinant GST-G 13 SR2 and GST-G 13 SR2 random proteins (10 g each) were incubated with glutathione-Sepharose 4B resin in Binding Buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 2 mM EDTA, 1 mM MgCl 2 , 0.1% Nonidet P-40, 5% BSA) for 1 h at room temperature. Beads were washed in the binding buffer twice and incubated with 40 g of His-TalinH for 1 h at room temperature, followed by two washes using the wash buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 2 mM EDTA, 1 mM MgCl 2 , 0.1% Nonidet P-40). Samples were eluted off beads by boiling in 3ϫ SDS sample buffer. Samples were resolved using 10% SDS-PAGE Gels, transferred to nitrocellulose membranes, and probed with an anti-talin head domain antibody.
Construction of Mouse G␣ 13 Plasmids-A 2.8 kb cDNA encoding wild-type mouse G␣ 13 (GeneBank TM ID 057665) subcloned into the pCMV-SPORT6 vector (clone ID 4918717; Open Biosystems). Three new plasmids were then created based on this vector by selectively mutating either the G␣ 13 SR2 Arg 227 to Ala or the G␣ 13 SR2 Arg 232 to Ala using the Stratagene II site-directed mutagenesis kit (Agilent Technologies) following the manufacturer's instructions. The primers used for these point mutation reactions (R227A, 5Ј-GTTGATGTAGGTGG-CCAGGCATCAGAACGGAAACGCTG-3Ј and R232A, 5Ј-GCCAGAGATCAGAACGGAAAGCCTGGTTTGAATGCT-TTGACA-3Ј were made by Integrated DNA Technologies. All plasmids used in experiments were amplified from ␣5-competent E. coli cultures (NEB) and then purified using endotoxinfree maxiprep kits (Qiagen) and sequenced by the DNA Services Facility at the UIC Research Resources Center.
Cell Culture and Plasmid Transfection-Mouse embryonic fibroblasts (NIH3T3) were purchased from ATCC and grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% calf bovine serum (CBS), 100 units/ml of penicillin and 0.1 mg/ml streptomycin, under 10% CO 2 at 37°C. When the cells reached 30 -40% confluence, the plasmid transient transfection was carried out using Fugene HD (Promega) according to manufacturer's instructions. Forty-eight hours after the transfection, the cells were either trypsinized with 0.25% Trypsin EDTA for the cell adhesion assay or harvested for Western blot using cell lysis buffer).
Cell Adhesion Assay-Non-treated, Corning 96 Well Black Flat Bottom Polystyrene Microplates were coated with 200 l of mouse fibronectin (10 g/ml), at 37°C for 1 h. The plates were rinsed with 200 l of PBS twice, and the nonspecific binding sites were blocked with 200 l heat-treated 1% BSA/PBS solution 37°C for 1 h. Cultured NIH3T3 cells were harvested with 0.25% trypsin EDTA, and the cell suspension was centrifuged (80 ϫ g) for 2 min. The pelleted cells were resuspended in colorless DMEM (Mediatech, Inc) at a concentration of 5 ϫ 10 6 cells/ml. 5 l/ml of calcein AM (Life Technologies) was then added, and the cell suspension was incubated at 37°C for 30 min. The cells were then separated from the suspension medium by centrifugation (80 ϫ g) for 2 min and resuspended (5 ϫ 10 6 cells/ml) in adhesion buffer (Hank's balanced salt solution plus 1% BSA). Following resuspension, 100 l of the cells were added to each well, and the plates were incubated for 2 h at 37°C. Finally, the wells were washed 4ϫ with PBS buffer read for fluorescence at 517 nm using a fluorescein filter.
Quantification of Western Blots/Dot Blots and Statistical Analysis-NIH software, ImageJ, was used for densitometric analysis of blots. Unless otherwise specified, significance was determined between samples using unpaired Student's t test (p Ͻ 0.05). All results are representative of data obtained using at least three separate experiments/and or 3 separate units of human PRP.
Rap1 Activity Assay-Rap1 activity was determined as previously described (16). Human PRP was adjusted to 2 ϫ 10 8 platelets/ml in Tyrode's buffer. Platelets were pre-treated with either 150 M Myr-G 13 SR2 pep , 150 M Myr-G13SR2 Random pep , DMSO, or 10 ng/ml PGI 2 for 2 min prior to stimulation with 1.0 M A23187 for 5 min on ice with gentle agitation. Platelets were immediately lysed using 2ϫ RIPA buffer with Complete Protease Inhibitor Mixture (Roche) for 15 min with gentle agitation. Clarified lysate was obtained through centrifugation at 14,000 ϫ g at 4°C for 10 min. The supernatant was incubated with 5 g of RalGDS bound to nickel bead resin (GoldBio) for 90 min at 4°C with gentle agitation. Resin was washed with 1ϫ RIPA buffer and eluted by boiling in SDS loading buffer. Bound Rap1 was detected via Western blotting using anti-Rap1 antibody.
TalinH ELISA-Recombinant 6x-His tagged Talin head domain (talinH) was added to Immulon2 Removawells (Dynatech Laboratories, Inc.) in 50 mM sodium carbonate/bicarbonate buffer (pH 9.4) and incubated overnight at 4°C to generate a standard curve. Samples were blocked with 5% BSA-PBST for 1 h at room temperature, washed thoroughly, and incubated with an anti-His antibody (Santa Cruz Biotechnology). TalinH concentration was determined using TMB substrate at 450 nm. Subsequently, GST-G 13 SR2, GST-G13SR2 Random pep , and GST were added to Immulon 2 wells at concentrations ranging from 1.0 g/ml to 1.0 ng/ml and probed with an anti-GST antibody (Santa Cruz Biotechnology). Equal loading concentrations were normalized by absorbance values. To determine talinH binding, GST fusion proteins were added to Immulon2 wells in equal concentrations overnight at 4°C. Each well was blocked in 5% BSA-PBST, incubated with talinH, and washed thoroughly. TalinH binding was measured by probing with an anti-His antibody and developed using TMB substrate at 450 nm. Each value was converted to picogram of talinH using the linear range of the standard curve. Each analysis was performed in triplicate and statistical significance was determined using unpaired Student's t test (p Ͻ 0.05).
PAC1 Integrin Activation Assay-CHO cells stably expressing ␣IIb␤3 integrins (CHOA5 cells) were a generous gift from Dr. Mark Ginsberg. CHOA5 cells were transfected with either G␣13 wild type (WT), or G␣13 R227A or R232A mutants. After 24 h integrin activation was determined using a PAC1 monoclonal antibody (Biolegend) and an Alexa555 conjugated secondary antibody (Life Technologies). Cells were fixed in 0.5% PFA for 15 min at room temperature and washed thoroughly in PBS. Cells were blocked and permeabilized for 30 min on ice using a PBS buffer containing 4% BSA and 0.1% Saponin. After thorough washing in PBS, G␣13 was probed using an anti-G␣13 antibody (Santa Cruz) and an Alexa 405 conjugated secondary antibody (Life Technologies). Cells were analyzed using a BD LSRii flow cytometer and PAC1 activation was measured for all cells and G␣13-positive cells. To evaluate the effect of Talin and G␣13 together, CHOA5 cells were co transfected with G␣13 WT, R227A, or R232A and either GFP (Control) or GFP Talin (Experimental). Integrin activation was determined by gating for G␣13-, GFP-, and PAC1-positive cells. The overall change in percent integrin activation was determined as follows: ⌬Activation ϭ (% Experimental/% Control). Each analysis was performed in triplicate and statistical significance was determined using unpaired Student's t test (p Ͻ 0.05).

Results
Agonist Activation of ␣IIb␤3 Is Dependent on G␣ 13 Switch Region 2 Signaling-To investigate G␣ 13 SR2 signaling in human platelets, a myristolyated peptide comprising the complete amino acid sequence of G␣ 13 SR2 was synthesized. This peptide was employed as both a competitive inhibitor of G␣ 13 SR2 signaling in intact platelets and as an immobilized ligand to enable the purification of G␣ 13 SR2-associated proteins. In the intact platelet studies, it was necessary to demonstrate that the myristolyated G␣ 13 SR2 peptide gained access to the platelet cytosolic compartment. To this end, human PRP was treated with myristoylated G 13 SR2 pep , non-myristoylated G 13 SR2 pep , or Vehicle (DMSO). After incubation, the PRP samples were centrifuged and the supernatant was removed to obtain the unbound fraction as shown in Fig. 1a (left panel). The pellet was then lysed and centrifuged to separate the membrane and cytosolic fractions as shown in the middle and right panel of Fig. 1a, respectively. Western blotting data clearly show that the non-myristoylated peptide is primarily located in the unbound fraction, while the myristoylated peptide is primarily taken up into the cytosolic fraction. These results demonstrate that myristolyated G 13 SR2 pep permeates the platelet plasma membrane and has access to intraplatelet signaling processes. The next series of experiments examined whether this G␣ 13 SR2 peptide (G 13 SR2 pep ) interfered with human platelet aggregation in PRP. It was found (Fig. 1b) that G 13 SR2 pep blocked platelet aggregation (trace c) stimulated by all the agonists examined, including U46619, TRAP1, ADP, TRAP4, and A23187. Clearly, this inhibitory profile is quite different from that associated with a peptide representing G␣ 13 SR1 (G 13 SR1 pep ), which only blocked aggregation mediated by TRAP1 (trace b). Furthermore, a single change in the amino acid sequence of G 13 SR2 pep at amino acid 227 (from Arg to Ala; G 13 SR2 227 mutant pep ) significantly reduced its inhibitory effects (trace d), demonstrating the high degree of specificity of G 13 SR2 pep . To investigate the mechanism by which G 13 SR2 pep blocks platelet aggregation, we first examined its role in primary platelet aggregation stimulated by ADP (3 M) in human PRP. Specifically, aspirin (ASA) treatment was used to block the secondary aggregation process that derives from the secretion of platelet dense granules. The results (aggregation upper panel and secretion lower panel) are presented in Fig. 1c. Trace e and bar e illustrate the ADP-mediated aggregation and secretion response in PRP pretreated with 75 M G 13 SR2 Random pep alone. It can be seen that pretreatment of the PRP with 1.0 mM ASA resulted in partial inhibition of this aggregation (upper panel, trace f) and a complete inhibition of dense granule secretion (lower panel, bar f), relative to no ASA treatment (trace e and bar e). Thus, in the presence of 1.0 mM ASA, the residual aggregation (trace f) represents the primary aggregation response. Furthermore, the results illustrate that G 13 SR2 pep (75 M) caused almost complete inhibition of the residual, ADP-induced, primary aggregation response (upper panel, trace g), even though it only produced partial inhibition of secretion (lower panel, bar g). Finally, when platelets were treated with both 1.0 mM ASA and 75 M G 13 SR2 pep both aggregation (trace h) and secretion (bar h) were completely blocked. Similar results were obtained using A23187 under the same experimental conditions (data not shown). These data demonstrate that G 13 SR2 pep peptide is capable of blocking primary human platelet aggregation.
The next experiments investigated the fundamental differences between G 13 SR1 and G 13 SR2 signaling pathways by comparing the abilities of G 13 SR1 pep and G 13 SR2 pep to modulate downstream platelet effectors. It was found that G 13 SR1 pep substantially blocked RhoA, calcium, and Src activation induced by TRAP1 but not by U46619. Surprisingly however, G 13 SR2 pep did not block these signaling pathways induced by either TRAP1 or U46619 (Fig. 1d). These findings indicate that G 13 SR1 and G 13 SR2 have completely different signaling profiles in human platelets, and that unlike G 13 SR1 signaling, the signaling events mediated by G␣ 13 SR2 lie further downstream from RhoA, calcium mobilization, or Src.
Based on these findings, experiments were undertaken to investigate whether G 13 SR2 pep interferes with a separate and well-known mediator of platelet aggregation, i.e. the Rap1/ RIAM signaling pathway. To examine this possibility, human PRP was activated with 1.0 M A23187 to mobilize calcium, and treated with either PGI 2 to inhibit Rap1 activation, or with G 13 SR2 pep or G 13 SR2 random pep . After treatment, the platelets were lysed and incubated with RalGDS to harvest active GTPbound form of Rap1 that was in turn detected using immunoblotting with a Rap1 antibody (Fig. 1, panel e) The lower panel of blots show total Rap1 in PRP, while the upper panel of blots show the active form of Rap1 upon each treatment (Fig. 1). It is clear that A23187 increases the amount of active Rap1 as compared with platelets treated with vehicle (DMSO), while PGI 2 treatment abrogates Rap1 activation in agreement with previous findings (16). Finally, when pre-treated with G 13 SR2 pep or G 13 SR2 random pep , the activity of Rap1 remained unchanged. Together, these data indicate that G 13 SR2 pep acts downstream of the calcium mobilization and Rap1 pathways.
On this basis, we next examined the possibility that G␣ 13 SR2 modulates a later step in the platelet activation process, i.e. ␣IIb␤3 integrin activation (17). These experiments employed FITC-PAC1 antibody and flow cytometry analysis to measure agonist-mediated activation of platelet ␣IIb␤3 (18). It was found that G 13 SR2 pep did indeed inhibit ␣IIb␤3 activation in response to both TRAP1 (Fig. 1f) and U46619 (Fig. 1g). Thus, while G␣ 13 SR1 signaling induces calcium mobilization, RhoA activation and Src activation specifically through PAR1 (19) receptor stimulation, G␣ 13 SR2 signaling is recruited by multiple platelet receptors and acts at a further downstream target involving ␣IIb␤3 activation.

G␣ 13 Switch Region 2 Interacts with the Talin-␣llb␤3-kindlin-3 Complex by Binding to the Talin Head
Domain-To identify the mechanism by which G␣ 13 SR2 modulates ␣llb␤3 function, we performed experiments to determine whether the amino acid sequence of G␣ 13 SR2 interacts with ␣llb␤3 integrin or an ␣llb␤3-associated protein such as talin (20). To this end, peptide-affinity chromatography was employed. Briefly, G 13 SR2 pep was coupled to a SulfoLink agarose column to purify G␣ 13 SR2-associated proteins from solubilized human platelets. Western blot analysis of the affinity eluates from the G 13 SR2 pep column demonstrated the purification of intact talin, ␣llb␤3, and kindlin-3 (Fig. 2a). The specificity of this affinity purification was demonstrated by the finding that an affinity column using G 13 SR2 random pep did not purify any of these platelet proteins (Fig. 2a). These findings provide evidence that the amino acid sequence comprising G␣ 13 SR2 directly binds to a protein complex (13,(21)(22)(23)(24)(25)(26) containing talin, ␣llb␤3, and kindlin-3. This notion was further confirmed by measuring the association of endogenous platelet G␣ 13 with FIGURE 1. Effects of G 13 SR1 pep and G 13 SR2 pep on human platelet function. a, myristoylated or non-myristoylated peptide were incubated with platelet PRP, which was then separated into unbound, membrane bound, or cytosolic fractions. An antibody raised against G 13 SR2 sequence was used for immunoblotting. ImageJ densiometric analysis indicates that roughly 5% of the myristoylated peptide remains unbound in the supernatant and that 95% partitions to the cytosolic fraction. b, traces denoted as a are representative of PRP pretreated with either 500 M G 13 SR1 random pep or G 13 SR2 random pep . Unless otherwise specified in the figure;  this protein complex. Specifically, an antibody generated against the N-terminal peptide sequence of G␣ 13 (12) was coupled to an immunoaffinity column (G␣ 13 -Ab) and used to purify G␣ 13 -associated proteins from solubilized platelets. Western blot analysis of the eluates from the G␣ 13 -Ab immunoaffinity column again demonstrated the purification of talin, ␣llb␤3, and kindlin-3 (Fig. 2b), whereas an immunoaffinity column using the pre-immune IgG did not result in the purification of any of these proteins (Fig. 2b).
The next series of experiments were designed to identify the binding partner of G␣ 13 within the protein complex. Our data demonstrate that G␣ 13 binds to talin, but not to ␣llb, ␤3 integrin or kindlin-3. Specifically, when the talin-␣llb␤3kindlin-3 complex was dissociated by heat treatment of solubilized platelets, both the G 13 SR2 pep affinity column and the G␣ 13 -Ab immunoaffinity column purified only talin and not ␣llb, ␤3 integrin, or kindlin-3 (Fig. 2, c and d). In addition, since the talin 8D4 antibody is known to detect both whole talin as FIGURE 2. Binding of G 13 SR2 pep or native platelet G␣ 13 protein to talin in solubilized human platelet membranes. Because the preparation of platelet membranes involves physical manipulation, the results from these studies reflect partially activated platelets. In panel a the solubilized platelet membrane eluates from the G 13 SR2 pep or the G 13 SR2 random pep affinity columns were immunoblotted for whole talin (220 kDa) ␣llb␤3 (130 kDa, 115 kDa) and kindlin-3 (72 kDa). In panel b the solubilized platelet membrane eluates from the G␣ 13 -Ab immunoaffinity column or the pre-immune IgG column were immunoblotted for G␣ 13 (43 kDa), talin, ␣llb␤3, and kindlin-3. In panels c and d the components of the talin-␣llb␤3-kindlin-3 complex were dissociated by heating the solubilized platelet membrane samples at 100°C for 45 min prior to affinity chromatography. The resulting eluates from the G 13 SR2 pep (or G 13 SR2 random pep ) columns (panel c) and the G␣ 13 Ab immunoaffinity (or control pre-immune IgG) columns (panel d) were immunoblotted for talin, ␣llb, ␤3, and kindlin-3. In panel e the solubilized platelet membrane eluates from the G␣ 13 -Ab immunoaffinity column or the control pre-immune IgG column were immunoblotted for the talin head domain. In panel f competition binding was performed by incubating solubilized platelet membranes with 250 M of G 13 SR2 pep (or G 13 SR2 random pep ) as the competing agent prior to G␣ 13 -Ab immunoaffinity column chromatography. In panel g the solubilized platelet membrane eluates from an affinity columns using G 13 SR2 pep or a mutant peptide with a single amino acid substitution (R227A; G 13 SR2 227 mutant pep ) were immunoblotted for the talin head domain. In panel h the solubilized platelet membrane eluate from a peptide affinity columns using G 13 SR1 pep was immunoblotted for whole talin, ␣llb, ␤3, kindlin-3, and p115-RhoGEF. In cases where the gel was probed with more than one antibody, the panel represents a composite of separate blots where each lane was run using its own molecular weight standards. All results are representative of data obtained using at least three separate units of human PRP, and representative data were chosen for each panel.
well as the talin rod domain (21), the absence of the talin rod domain (Fig. 2, a and b) suggests that the head domain of talin serves as the binding region for G␣ 13 . This finding was confirmed by blotting the G␣ 13 -Ab immunoaffinity column eluate with the TA205 antibody that specifically recognizes the talin head domain (13) (Fig. 2e). Evidence that the binding of G␣ 13 to the talin head domain is mediated through the G␣ 13 SR2 sequence was provided by binding competition studies. In these experiments, solubilized platelet membranes were pre-incubated with G 13 SR2 pep before protein purification by the G␣ 13 -Ab immunoaffinity column. The results demonstrated that competition with G 13 SR2 pep inhibited native G␣ 13 binding to the talin head domain (Fig. 2f). Taken together, these findings suggest that G␣ 13 specifically binds to the talin head domain through the G 13 SR2 sequence, and does not directly associate with either ␣llb, ␤3 integrin, or kindlin-3. Furthermore, the aggregation data shown in Fig. 1a suggest that Arg-227 within the G 13 SR2 sequence may be important for talin binding. This notion was confirmed by the finding that an affinity column using the G 13 SR2 227 mutant pep (R227A) was not effective in purifying the talin head domain (Fig. 2g).
Our previous results demonstrated that G 13 switch region 1 (G 13 SR1) binds to and signals through p115-RhoGEF in human platelets. In order to examine whether G 13 SR1 also interacts with the talin-␣llb␤3-kindlin-3 complex, we again employed peptide-affinity chromatography as described above. In this case, the column was prepared using G 13 SR1 pep coupled to Sul-foLink agarose beads. However, Western blot analysis of the affinity eluates from the G 13 SR1 pep column demonstrated only the purification p115-RhoGEF and not the purification of talin, ␣llb, ␤3 integrin, or kindlin-3 (Fig. 2h).
Additional experiments provided independent confirmation of the direct binding interaction between G␣ 13 SR2 and the talin head domain. To this end, we employed a bi-molecular in vitro binding reaction outside of the platelet milieu. Using a dot blot assay, it was found that recombinant talin head domain directly binds to G 13 SR2 pep but not to the G 13 SR2 random pep (Fig. 3, a and  b). Moreover, the specificity of G 13 SR2 pep -talin binding complex was demonstrated by using two different G 13 SR2 mutant peptides, each with a single R to A substitution at positions 227 or 232 (G 13 SR2 227 mutant pep and G 13 SR2 232 mutant pep ). The results demonstrated (Fig. 3, a and b) that each mutant exhibited a significantly reduced binding affinity for talin head domain. To further demonstrate the specificity of this binding interaction between G 13 SR2 pep and talin head domain, we utilized an alternate experimental approach. Specifically, recombinant GST-G 13 SR2 and GST-G 13 SR2 random fusion proteins were generated and incubated with recombinant talin head domain to quantify their binary interactions. The results show that while GST-G 13 SR2 bound to talin, GST-G 13 SR2 random did not (Fig. 3c). Taken together, these results add additional support to the notion that G 13 SR2 can form a bi-molecular complex with the talin head domain.
G␣ 13 Switch Region 2 Directly Binds to Talin but Not to ␤3 Integrin-Our previous experiments (Fig. 2d) indicated that G␣ 13 is capable of forming a ternary complex with talin but not with ␤3 integrin. This notion was further investigated by additional dot blot binding assays measuring the ability of biotiny-lated G 13 SR2 pep , (or G 13 SR2 random pep ) to bind with recombinant GST-talin head domain, GST-␤3 cytoplasmic tail, or GST control. It was found that G 13 SR2 pep bound only to GST-talin head domain, but not to the GST-␤3 cytoplasmic tail (Fig. 3, d  and e). Furthermore, the specificity of the binding interaction between G␣ 13 SR2 and talin head domain was again demonstrated by the finding that biotinylated G 13 SR1 pep bound to neither GST-talin head domain nor GST-␤3 cytoplasmic tail (Fig. 3, d and e), but did effectively bind to recombinant fulllength p115-RhoGEF (Fig. 3f). This preference of G␣ 13 SR1 for binding to p115-RhoGEF is consistent with our findings in solubilized platelet membranes (Fig. 2h) and our previously published results (8). To demonstrate that the recombinant GST-␤3 integrin cytoplasmic tail employed in the previous experiment (Fig. 3, d and e) was indeed functional, we tested the binding activity of GST-␤3 integrin cytoplasmic tail with talin head domain (22). As expected, talin head domain significantly bound GST-␤3 integrin (Fig. 3g) under the same experimental conditions as employed in Fig. 3, d and e. Finally, an ELISA was used to examine G␣ 13 SR2 interaction with talin. Using a talinH standard curve (Fig. 3h), the picograms of talinH bound to GST, G 13 SR2 pep , or GST-G 13 SR2 random pep is illustrated in Fig. 3i. Collectively, the above findings provide strong evidence that G␣ 13 SR2 mediates integrin activation by directly binding to the talin head domain and not to the ␤3 integrin cytoplasmic tail.
Intraplatelet Calcium Stimulates the Binding of G␣ 13 Switch Region 2 to Talin-We next examined a possible mechanistic basis for the binding interaction between G␣ 13 SR2 and talin head domain. Since intraplatelet calcium is known to be a common mediator of ␤3 integrin activation, we first tested the effects of the calcium ionophore A23187 on G␣ 13 -talin binding. Specifically, intact human platelets (PRP) were treated with 10 M A23187, solubilized, and then subjected to G␣ 13 -Ab immunoaffinity chromatography as described above. The results demonstrated (Fig. 4a) that elevation of intraplatelet calcium by A23187 caused a notable increase in G␣ 13 -talin head domain binding. This observation was further confirmed by chelating available calcium with EGTA and BAPTA. Under these conditions, the binding of G␣ 13 to talin was completely abolished, even in the presence of A23187 (Fig. 4a). These results indicate that G␣ 13 is not constitutively associated with talin in truly un-activated platelets, but only becomes bound to talin in the presence of elevated calcium.
G␣ 13 Switch Region 2 Signaling Is Essential for Adhesion of NIH3T3 Cells-To determine whether G␣ 13 SR2-talin signaling is limited to human platelets or whether it may serve as a more global mechanism of integrin activation, cell adhesion experiments were performed using fibroblasts. Specifically, we expressed both wild-type and mutant forms of the G␣ 13 protein in cultured NIH3T3 cells and measured the effects of these G␣ 13 SR2 mutations on cell adhesion to fibronectin. The rationale for these experiments derives from our previous observations indicating that arginine 227 and arginine 232 contained within G␣ 13 SR2 each play a critical role in G␣ 13 SR2-talin binding and integrin activation. Thus, if G␣ 13 SR2 signaling is in fact essential for the process of cell adhesion, overexpression of the mutant forms of G␣ 13 should interfere with this process, due to the loss of important arginine residues within the G␣ 13 SR2 sequence. The results from these experiments are consistent with this notion. Specifically, Western blot analysis (Fig. 4b) of transfected NIH3T3 cells demonstrated that the expression levels of the wild-type G␣ 13 were comparable to the expression levels of both the G␣ 13 R227A and the G␣ 13 R232A mutant proteins. However, despite these equal expression levels, the G␣ 13 SR2 mutant (R227A and R232A) cells were significantly less capable of adhering to fibronectin compared with the wild- . Bi-molecular binding between G 13 SR2 and talin head domain (talinH). Panel a depicts a dot blot of talin (2 g, 1 g, 0.5 g) probed with biotinylated G 13 SR2 peptides (G 13 SR2 pep , G 13 SR2 227 mutant pep , G 13 SR2 232 mutant pep , and G 13 SR2 random pep ). Panel b depicts quantification of the bottom row of panel a. The symbol (*) represent significance (p Ͻ 0.05) relative to G 13 SR2 pep binding to His-talinH. Panel c depicts a GST pull-down assay where His-talinH is bound to either GST-G 13 SR2 or GST-G 13 SR2 random fusion proteins. GST-G 13 SR2 and GST-G 13 SR2 random inputs are shown by Ponceau stain (bottom panel), His-talinH input is shown using Coomassie staining (middle panel), and bound His-talinH (top panel) is shown using an anti-talin antibody. Panel d depicts a dot blot where GST-talinH (top row), GST-␤3 cytoplasmic tail (middle row), and GST (bottom row) are probed with G 13 SR1 pep (first column), G 13 SR1 random pep (second column), G 13 SR2 pep (third column), or G 13 SR2 random pep (fourth column) biotinylated peptides. Panel e depicts quantification of panel d. The symbol (*) represents significance (p Ͻ 0.05) between GST-talinH binding to G 13 SR2 pep and G 13 SR2 pep random . Panel f depicts a dot blot where recombinant His-p115-RhoGEF full length is probed with G 13 SR1 pep or G 13 SR1 random pep biotinylated peptides. To the left is a quantification of this dot blot. The symbol (*) represents significance (p Ͻ 0.05) between G 13 SR1 pep and G 13 SR1 random pep . Panel g depicts a dot blot where GST-␤3 cytoplasmic tail and GST are overlaid with recombinant His-talinH, and probed using anti-talin antibody. The bar illustrates quantification of this dot blot. The symbol (*) represents significance (p Ͻ 0.05) between GST-␤3 cytoplasmic tail and GST. Panel h depicts a talin standard curve ELISA. Panel i depicts the binding of talin to GST fusion proteins in an ELISA. Absorbance values (450 nm) were recorded for talin binding to GST-G 13 SR2, GST-G 13 SR2 Random , and GST. Absorbance values were converted to picograms using the talin standard curve in panel h, and significance was established using the Student's unpaired t test. Each experiment was repeated three times (n ϭ 3) and representative data were chosen for each figure.
type G␣ 13 SR2 cells (Fig. 4b). Collectively, these findings suggest that in addition to human platelet inside-out signaling, G␣ 13 SR2 may play a fundamental role in the process of integrin-mediated cell adhesion in other cell types.
To further examine the effect of G␣ 13 in integrin activation, we transfected wild type G␣ 13 and G␣ 13 SR2 mutants into CHO cells stably expressing ␣IIb␤3, (CHOA5 cells). After a 24 h transfection, each sample was gated for G␣ 13 expression and integrin activation using a PAC1 antibody. The results demonstrate that G␣ 13 increases the total percent of cells with active integrins when compared with vector alone (Fig. 4c). Similarly, the G␣ 13 or G␣ 13 SR2 mutants were transfected into CHOA5 cells (Fig. 4d). After gating for G␣ 13 expression and PAC1 activation, the data show that SR2 mutations significantly reduced the percent of cells with active integrins. To determine if G␣ 13 SR2 is mediating integrin activation through talin, we co-transfected the G␣ 13 constructs with either GFP or GFP tagged full-length talin. CHOA5 cells were gated for G␣ 13 , GFP, or GFP-talin, and PAC1 signal, and the change in integrin activation between GFP and GFPtalin cells was calculated. The results demonstrate that WT G␣ 13 causes substantially greater integrin activation than either of the G␣ 13 SR2 mutants (Fig. 4e).

Discussion
In the present study, the underlying basis for platelet insideout signaling has been further investigated by examining a direct mechanistic link between G␣ 13 signaling and ␣llb␤3 integrin activation. The results demonstrate evidence for such a link involving switch region 2 of the G 13 ␣ subunit. Our past (8) and present findings suggest that G␣ 13 actually plays two important and distinct roles in the platelet activation process, depending on whether signaling occurs through its Switch Region 1 or through its Switch Region 2 (see model in Fig. 5). In this regard, we found that signaling through G␣ 13 SR1 is a classical GPCR-linked event that derives from agonist interaction with a G 13 -coupled receptor. The activated G␣ 13 SR1 then causes stimulation of multiple downstream effectors, including RhoA, calcium, and Src, among others. The specificity of this signaling pathway is consistent with the notion that inhibition of G␣ 13 SR1 signaling has very selective and limited effects on platelet function. Thus, G␣ 13 SR1 inhibitors only block signaling by GPCRs that are primarily coupled to G␣ 13 , i.e. PAR1 (Fig.  1b, trace b) (8), and do not block platelet aggregation in response to other platelet agonists, e.g. ADP, TRAP4, U46619, A23187, etc., because these agonists do not heavily rely upon  Fig. 4 panel a shows that increased intraplatelet calcium levels promote native G␣ 13 -talin binding. Platelets from untreated human PRP or A23187 (10 M)-treated PRP were solubilized and subjected to G␣ 13 -Ab immunoaffinity column chromatography. In separate experiments, the PRP was supplemented with EGTA (3 mM) and the platelets were loaded with BAPTA prior to A23187 (10 M) stimulation. The platelets were then solubilized and subjected to G␣ 13 -Ab immunoaffinity column chromatography. G␣ 13 blotting, shown in the left two lanes, was used as a loading control. All results are representative of data obtained using at least 3 separate units of human PRP, and representative data were chosen. In panel b cultured mouse embryonic fibroblasts (NIH3T3) were transfected with plasmids containing G␣ 13 wild-type, G␣ 13 R227A, or G␣ 13 R232A. Forty-eight hours after transfection, the cells were harvested for Western blot analysis of G␣ 13 or allowed to adhere to fibronectin-coated plates. The Western blots illustrated at the top of the figure illustrate comparable levels of G␣ 13 expression in each case. In panel c a G␣ 13 plasmid or vector alone was transfected into CHO cells stably expressing platelet ␣IIb␤3 integrins (CHOA5). Percent integrin activation was determined using a monoclonal PAC1 antibody. In panel d the G␣ 13 wild-type, G␣ 13 R227A, or G␣ 13 R232A plasmids were transfected into CHOA5 cells, gated for G ␣13 expression and the percent of cells with active integrins was quantified. In panel e each G␣ 13 construct was co-transfected with either GFP or GFP tagged full-length talin and the change in integrin activation between G ␣13 co-expression with GFP and GFP-talin. Significance (*) (p Ͻ 0.05) was determined by ANOVA using GraphPad PRISM 5.04 statistical software (San Diego, CA). All results are representative of at least three separate experiments (n ϭ 3). G 13 signal transduction (8). Our present and previous demonstration that G␣ 13 SR1 signaling is specific for PAR1-mediated activation is inconsistent with recent work (27,28) describing a binding interaction between G␣ 13 SR1 and platelet ␤3 integrin, as a basis for ␣llb␤3 integrin activation. If this binding interaction were indeed the mechanism of G␣ 13 -mediated integrin activation in platelets, one would expect inhibition of G 13 SR1 signaling to not only block platelet stimulation caused by PAR1, but platelet stimulation caused by all activating agents. However, this was not found to be the case (8). Moreover, the present studies have been unable to demonstrate the direct binding of either native platelet G␣ 13 or G 13 SR1 pep to ␤3 integrin in a bi-molecular reaction. The basis for our inability to demonstrate this direct binding interaction is presently unclear.
On the other hand, it is apparent that the signaling mechanism of G␣ 13 SR2 is quite different from that of G␣ 13 SR1 in the sense that it is not directly coupled to GPCR activation. Rather, it appears to occur as a consequence of GPCR-mediated increases in intraplatelet calcium, and to be directly linked to talin activation of ␤3 integrin. Furthermore, since integrin activation serves as the final common pathway for platelet-platelet adhesion, inhibition of G␣ 13 SR2 signaling produces global inhibition of platelet aggregation in response to all agonists (Fig. 1b,  trace c).
Because of these two divergent signaling pathways, the sequence of events that ultimately leads to integrin activation may be quite different depending on the GPCR that is stimulated. Thus, for GPCRs that are extensively coupled to G 13 , such as PAR1, G␣ 13 SR1, and G␣ 13 SR2 signaling events are sequentially induced, where PAR1-mediated activation of G␣ 13 SR1 leads to intraplatelet calcium mobilization (8). This increase in intraplatelet calcium levels then promotes multiple platelet signaling events (29 -31) including binding of G␣ 13 SR2 to whole talin and subsequent cleavage of the talin rod domain, presumably by calpain (32)(33)(34). Once formed, the G␣ 13 SR2-talin head domain complex binds to the ␤3 cytoplasmic tail leading to integrin activation.
In contrast, other platelet receptors that do not extensively couple to G 13 (e.g. P2Y 1 , TPR, PAR4, etc.) are capable of increasing intraplatelet calcium levels by separate signaling mechanisms, such as G q -mediated PLC activation and consequent IP3 formation. Therefore, in such instances, elevated calcium is itself sufficient to initiate G␣ 13 SR2-mediated integrin activation, and GPCR-mediated G␣ 13 SR1 signaling is not required. Taken together, our findings therefore provide a new paradigm for inside-out platelet signaling by identifying G␣ 13 SR2 as an essential link between membrane receptor stimulation, increased intracellular calcium levels, and ␤3 integrin activation.
Author Contributions-S. S. participated in writing the manuscript as well as in the design of experiments and analysis of the data in Fig.  1 (b, d, f, and g), Fig. 2 (ah), Fig. 4a, and preparation of the model in Fig. 5. J. S. participated in writing the manuscript as well as in the design of experiments and analysis of the data in Fig. 1 (a and e), Fig.  3 (a-i), Fig. 4 (ce), and preparation of the model in Fig. 5. X. Z. participated in writing the manuscript as well as in the design of experiments and analysis of the data in Fig. 4b. A. H. C. participated in writing the manuscript, as well as overseeing, coordinating, and conceiving the experiments performed at Tufts University. G. C. L. conceived the study, participated in writing the manuscript, as well as overseeing, coordinating and conceiving the experiments performed at the University of Illinois at Chicago.  13 Switch Region signaling pathways. G 13 SR1 is a classical receptor-coupled signal transduction pathway where agonist interaction with a surface membrane GPCR leads to G 13 SR1 activation and subsequent downstream signaling events, e.g. calcium mobilization, RhoA and Src activation, etc. In contrast, signaling through G 13 SR2 is not directly GPCR-dependent, but rather calcium-dependent. In this case, elevation of intraplatelet calcium promotes the recruitment of talin by the CalDAG-GEF/Rap1/RIAM pathway, binding of G 13 SR2 to whole talin, cleavage of the talin rod domain, binding of the G␣ 13 -talin head complex to the cytoplasmic tail of ␤3 integrin, and dissociation of G␣ 13 from the talin head complex. Because of these significant differences in signaling mechanisms, selective inhibition of G 13 SR1 and G 13 SR2 signaling produces markedly different effects on human platelet function. Thus, inhibition of G 13 SR1 signaling only blocks PAR1-mediated platelet aggregation, whereas inhibition of G 13 SR2 signaling blocks aggregation in response to all platelet agonists.