Integrins have a critical role in thrombosis and haemostasis1. Antagonists of the platelet integrin αIIbβ3 are potent anti-thrombotic drugs, but also have the life-threatening adverse effect of causing bleeding2,3. It is therefore desirable to develop new antagonists that do not cause bleeding. Integrins transmit signals bidirectionally4,5. Inside-out signalling activates integrins through a talin-dependent mechanism6,7. Integrin ligation mediates thrombus formation and outside-in signalling8,9, which requires Gα13 and greatly expands thrombi. Here we show that Gα13 and talin bind to mutually exclusive but distinct sites within the integrin β3 cytoplasmic domain in opposing waves. The first talin-binding wave mediates inside-out signalling and also ligand-induced integrin activation, but is not required for outside-in signalling. Integrin ligation induces transient talin dissociation and Gα13 binding to an EXE motif (in which X denotes any residue), which selectively mediates outside-in signalling and platelet spreading. The second talin-binding wave is associated with clot retraction. An EXE-motif-based inhibitor of Gα13–integrin interaction selectively abolishes outside-in signalling without affecting integrin ligation, and suppresses occlusive arterial thrombosis without affecting bleeding time. Thus, we have discovered a new mechanism for the directional switch of integrin signalling and, on the basis of this mechanism, designed a potent new anti-thrombotic drug that does not cause bleeding.
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Shattil, S. J. & Newman, P. J. Integrins: dynamic scaffolds for adhesion and signaling in platelets. Blood 104, 1606–1615 (2004)
Coller, B. S. Anti-GPIIb/IIIa drugs: current strategies and future directions. Thromb. Haemost. 86, 427–443 (2001)
Serebruany, V. L., Malinin, A. I., Eisert, R. M. & Sane, D. C. Risk of bleeding complications with antiplatelet agents: meta-analysis of 338,191 patients enrolled in 50 randomized controlled trials. Am. J. Hematol. 75, 40–47 (2004)
Hynes, R. O. Integrins: bidirectional, allosteric signaling machines. Cell 110, 673–687 (2002)
Moissoglu, K. & Schwartz, M. A. Integrin signalling in directed cell migration. Biol. Cell 98, 547–555 (2006)
Tadokoro, S. et al. Talin binding to integrin beta tails: a final common step in integrin activation. Science 302, 103–106 (2003)
Ye, F., Kim, C. & Ginsberg, M. H. Molecular mechanism of inside-out integrin regulation. J. Thromb. Haemost. 9 (Suppl. 1). 20–25 (2011)
Gong, H. et al. G protein subunit Gα13 binds to integrin αIIbβ3 and mediates integrin “outside-in” signaling. Science 327, 340–343 (2010)
Shen, B., Delaney, M. K. & Du, X. Inside-out, outside-in, and inside-outside-in: G protein signaling in integrin-mediated cell adhesion, spreading, and retraction. Curr. Opin. Cell Biol. 24, 600–606 (2012)
Moser, M., Nieswandt, B., Ussar, S., Pozgajova, M. & Fassler, R. Kindlin-3 is essential for integrin activation and platelet aggregation. Nature Med. 14, 325–330 (2008)
Ma, Y. Q., Qin, J., Wu, C. & Plow, E. F. Kindlin-2 (Mig-2): a co-activator of β3 integrins. J. Cell Biol. 181, 439–446 (2008)
Obergfell, A. et al. Coordinate interactions of Csk, Src, and Syk kinases with αIIbβ3 initiate integrin signaling to the cytoskeleton. J. Cell Biol. 157, 265–275 (2002)
Flevaris, P. et al. A molecular switch that controls cell spreading and retraction. J. Cell Biol. 179, 553–565 (2007)
Patil, S. et al. Identification of a talin-binding site in the integrin β3 subunit distinct from the NPLY regulatory motif of post-ligand binding functions. The talin N-terminal head domain interacts with the membrane-proximal region of the β3 cytoplasmic tail. J. Biol. Chem. 274, 28575–28583 (1999)
Wegener, K. L. et al. Structural basis of integrin activation by talin. Cell 128, 171–182 (2007)
Haling, J. R., Monkley, S. J., Critchley, D. R. & Petrich, B. G. Talin-dependent integrin activation is required for fibrin clot retraction by platelets. Blood 117, 1719–1722 (2011)
Petrich, B. G. et al. Talin is required for integrin-mediated platelet function in hemostasis and thrombosis. J. Exp. Med. 204, 3103–3111 (2007)
Coller, B. S. Interaction of normal, thrombasthenic, and Bernard-Soulier platelets with immobilized fibrinogen: defective platelet-fibrinogen interaction in thrombasthenia. Blood 55, 169–178 (1980)
Ugarova, T. P. et al. Conformational changes in fibrinogen elicited by its interaction with platelet membrane glycoprotein GPIIb-IIIa. J. Biol. Chem. 268, 21080–21087 (1993)
Du, X. et al. Ligands “activate” integrin αIIbβ3 (platelet GPIIb-IIIa). Cell 65, 409–416 (1991)
Arias-Salgado, E. G., Lizano, S., Shattil, S. J. & Ginsberg, M. H. Specification of the direction of adhesive signaling by the integrin β cytoplasmic domain. J. Biol. Chem. 280, 29699–29707 (2005)
Goksoy, E. et al. Structural basis for the autoinhibition of talin in regulating integrin activation. Mol. Cell 31, 124–133 (2008)
Xi, X., Bodnar, R. J., Li, Z. Y., Lam, S. C. T. & Du, X. P. Critical roles for the COOH-terminal NITY and RGT sequences of the integrin β3 cytoplasmic domain in inside-out and outside-in signaling. J. Cell Biol. 162, 329–339 (2003)
Krishnadas, A., Rubinstein, I. & Onyuksel, H. Sterically stabilized phospholipid mixed micelles: in vitro evaluation as a novel carrier for water-insoluble drugs. Pharm. Res. 20, 297–302 (2003)
O’Brien, K. A., Gartner, T. K., Hay, N. & Du, X. ADP-stimulated activation of Akt during integrin outside-in signaling promotes platelet spreading by inhibiting glycogen synthase kinase-3β. Arterioscler. Thromb. Vasc. Biol. 32, 2232–2240 (2012)
Delaney, M. K., Liu, J., Zheng, Y., Berndt, M. C. & Du, X. The role of Rac1 in glycoprotein Ib-IX-mediated signal transduction and integrin activation. Arterioscler. Thromb. Vasc. Biol. 32, 2761–2768 (2012)
Cho, J. et al. Protein disulfide isomerase capture during thrombus formation in vivo depends on the presence of β3 integrins. Blood 120, 647–655 (2012)
O’Brien, K. A., Stojanovic-Terpo, A., Hay, N. & Du, X. An important role for Akt3 in platelet activation and thrombosis. Blood 118, 4215–4223 (2011)
Marjanovic, J. A., Li, Z., Stojanovic, A. & Du, X. Stimulatory roles of nitric-oxide synthase 3 and guanylyl cyclase in platelet activation. J. Biol. Chem. 280, 37430–37438 (2005)
Nimura, N., Kinoshita, T., Yoshida, T., Uetake, A. & Nakai, C. 1-Pyrenyldiazomethane as a fluorescent labeling reagent for liquid chromatographic determination of carboxylic acids. Anal. Chem. 60, 2067–2070 (1988)
We thank T. Kozasa, B. Kreutz and C. Chow for providing purified recombinant Gα13 protein; and B. Petrich and D. Critchley for providing talin−/− mice. We acknowledge that H. Gong performed experiments for this project. This work is supported by grants from National Heart, Lung, and Blood Institute (HL080264, HL062350 (X.D.) and HL109439 (J.C.)).
X.D. holds pending patents on the EXE-motif-containing peptide inhibitors.
Extended data figures and tables
Extended Data Figure 1 A schematic showing how selective inhibitors of integrin outside-in signalling work as anti-thrombotics.
Blue arrows indicate steps that are inhibited.
a, b, Lysates from CHO cells expressing similar levels of wild-type (WT) αIIbβ3, β3 C-terminal truncation mutants (Δ759, Δ741, Δ728 and Δ715) complexed with wild-type αIIb (a), and β3 EXE motif mutants (E731A, E732A, E733A and EEE to AAA) complexed with wild-type αIIb (b) were immunoprecipitated with anti-Gα13 antibody or equal amount of control rabbit IgG. Immunoprecipitates and lysates (equivalent of 10% used for immunoprecipitation) were immunoblotted with anti-Gα13 and anti-β3 antibodies. c, d, GST–β3CD (WT) or GST–β3(AAA)CD (AAA mutant) proteins immobilized in glutathione-coated microtitre wells were incubated with increasing concentrations of Gα13 (c), or increasing concentrations of THD (d). After washing, bound Gα13 and THD were respectively detected using anti-Gα13 or anti-talin (mouse IgG was used as a specificity control) followed by secondary horseradish peroxidase-labelled anti-IgG antibody. e, In addition to the AAA mutation, conserved mutations of EEE to DED and EEE to QSE (as found in β5) were introduced to the β3 cytoplasmic domain. These mutants were co-transfected with wild-type αIIb into CHO cells, which were sorted to achieve comparable expression levels with wild-type-αIIbβ3-expressing cells (as shown in Extended Data Fig. 4e). Lysates from these cells were immunoprecipitated with anti-β3 or equal amount of pre-immune rabbit serum. Lysates (10%) and immunoprecipitates were immunoblotted with anti-Gα13 or anti-β3. f, Lysates from human platelets (with or without stimulation with 0.025 U ml−1thrombin) were immunoprecipitated with anti-Gα13 antibody or equal amount of control rabbit IgG. Immunoprecipitates were immunoblotted with anti-Gα13 and anti-β1 antibodies. Gα13 is associated with β1, which is increased after thrombin stimulation.
Extended Data Figure 3 Ligand occupancy induces switch of integrin αIIbβ3 from the talin-bound to the Gα13-bound state.
a, To determine the effect of integrin activation and ligand occupancy on Gα13–β3 association, human platelets were incubated with or without 1 mM MnCl2 and 30 μg ml−1 fibrinogen for 5 min at 22 °C. Platelet lysates were then immunoprecipitated with anti-β3 or pre-immune rabbit serum. Lysates (10%) and immunoprecipitates were immunoblotted with anti-β3 or anti-Gα13. b, c, Washed human platelets were stimulated with 0.025 U ml−1 α-thrombin with or without adding 2 mM EDTA (an inhibitor of the ligand binding function of integrins), stirred (1,000 r.p.m.) at 37 °C, solubilized at various time points, and immunoprecipitated with anti-β3 or equal amounts of pre-immune rabbit serum. Lysates (10%) and immunoprecipitates were immunoblotted with anti-Gα13, anti-talin or anti-β3 antibodies. b, Western blot results. c, Turbidity changes in platelet suspension indicating integrin-dependent platelet aggregation. Note the inhibitory effect of EDTA on talin dissociation and Gα13 binding to β3. d, As additional controls for Fig. 2a to exclude the possibility of loss of talin and β3 in platelet lysates to insoluble fraction during integrin signalling, washed human platelets were stimulated with 0.025 U ml−1 α-thrombin in the absence or presence of 2 mM integrin inhibitor RGDS, stirred (1,000 r.p.m.) at 37 °C, and then solubilized at various time points as in Fig. 2a. Solubilized platelets were centrifuged at 14,000g for 10 min to separate lysates from insoluble pellets. Pellets were dissolved in SDS sample buffer to the same volume as the lysates after diluting them 1:1 with 2× SDS sample buffer, and both were immunoblotted with anti-β3 and anti-talin antibodies. Note that the levels of talin and β3 in platelet lysates kept essentially constant during the course of platelet aggregation and, with low concentrations of thrombin used to stimulate platelets, very little insoluble β3 and talin were present in the pellet, which were detectable only after prolonged exposure (5-min exposure compared to 10 s of normal exposure time) and with no obvious variation during the course of platelet aggregation.
Extended Data Figure 4 Effects of shRNA-induced talin knockdown and talin knockout on integrin signalling.
a, Western blot comparison of talin 1 expression levels in mouse platelets derived from control shRNA- or talin-shRNA-transfected bone marrow stem cells. Western blots of Gα13, and integrin β1 and β3 are also shown. b, Adhesion of unstimulated mouse platelets to immobilized fibrinogen for 1 h. Adherent platelets were quantified as percentage of total platelets loaded (mean ± s.d., n = 4). c, Turbidity changes in mouse platelet suspension stimulated with 5 μM ADP in the presence of 20 μg ml−1 fibrinogen, with or without 1 mM MnCl2, as detected using an aggregometer. d, Fluorescence microscopy images of phalloidin-stained mouse platelet spreading on fibrinogen for 1 h, with or without 1 mM MnCl2. e, Quantification of surface areas of individual adherent platelets as shown in Fig. 2g (mean ± s.e.m.).
a, Flow cytometric analysis of integrin αIIbβ3 expression levels in β3−/− mouse platelets transfected with wild-type or AAA mutant β3 using bone marrow stem cell transplantation technology in comparison with C57BL/6 mouse platelets. β3−/− platelets were used as a negative control. αIIbβ3 complex was detected using an anti-mouse αIIb antibody. b, Mouse platelets expressing recombinant wild-type or AAA mutant β3 as in a were lysed and immunoprecipitated with anti-β3 or equal amounts of pre-immune rabbit serum. Lysates (10%) and immunoprecipitates were immunoblotted with anti-SRC or anti-β3 antibodies. c, d, Spreading of phalloidin-stained wild-type platelets (EEE), AAA mutant platelets and AAA mutant platelets incubated with mP6Scr or mP6 on immobilized fibrinogen for 1 h. c, Typical fluorescence microscopy images. d, Quantification of surface areas of individual platelets (mean ± s.e.m.).
Extended Data Figure 6 Effects of mutational disruption of the EXE motif on integrin outside-in signalling.
a, Expression levels of wild-type or the EXE motif (QSE, DED or AAA) mutants of β3 in complex with wild-type αIIb in CHO cells, as determined by flow cytometry. Mouse IgG was used as a negative control. b, c, Spreading of CHO-1b9 cells expressing wild-type αIIbβ3, and QSE, DED or AAA mutant αIIbβ3 on fibrinogen for 1 h. b, Quantification of surface areas of individual cells (mean ± s.e.m.). c, Typical microscopy images. d, Flow cytometric analysis of wild-type αIIbβ3, AAA or Y747A mutant αIIbβ3 expression in CHO cells. Mouse IgG was used as a control. e, CHO cells expressing wild-type, AAA or Y747A β3 without (top panels) or with (bottom panels) co-expression of recombinant THD were solubilized and immunoprecipitated with anti-β3 or pre-immune serum. 10% lysates and immunoprecipitates were immunoblotted with anti-talin, anti-Gα13 or anti-β3 antibodies. f, Typical western blots for Fig. 3g. Wild-type or AAA-mutant-αIIbβ3-expressing CHO-1b9 cells were allowed to adhere to immobilized fibrinogen, solubilized at various time points, and analysed for RHOA activation and SRC Tyr 416 phosphorylation.
Extended Data Figure 7 mP6 selectively inhibits integrin outside-in signalling without affecting inside-out signalling.
a–d, Washed human platelets were stimulated with 0.025 U ml−1 α-thrombin in the absence or presence of 250 μM myristoylated peptides, mP13 (a, b) and mP6 (c, d) with stirring (1,000 r.p.m.) at 37 °C, and then solubilized at various time points. Lysates were immunoprecipitated with anti-β3 rabbit serum or equal amounts of pre-immune serum. Lysates (10%) and immunoprecipitates were immunoblotted with anti-Gα13, anti-talin or anti-β3 antibodies. a, c, Typical western blot results. b, d, Typical turbidity changes in platelet suspension indicating integrin-dependent platelet aggregation. e, Quantification of human platelet spreading on immobilized fibrinogen for 1 h, without or with treatment with DMSO, mP6Scr, or mP6 as shown in Fig. 4b (mean surface area ± s.e.m.). f, Flow cytometric analysis of PAR4-AP-induced Oregon Green-labelled soluble fibrinogen binding to human platelets pre-treated with 100 μM mP6Scr or 100 μM mP6 stimulated with increasing concentrations of PAR4-AP. Integrilin-treated platelets were used as a negative control. g, Flow cytometric analysis of 100 μM PAR4-AP-induced PAC1 binding to human platelets pre-treated with 100 μM mP6Scr or mP6. Integrilin-treated platelets were used as negative control. h, Flow cytometric analysis of PAR4-AP-induced Oregon Green-labelled soluble fibrinogen binding to human platelets pre-treated with solvent DMSO, mP13Scr or mP13. Resting platelets were used as a negative control.
Extended Data Figure 8 The in vivo effect of mP6: selective inhibition of thrombosis but not haemostasis.
a, Representative images of laser-induced mouse cremaster arteriolar thrombosis (red) in the context of the bright-field microvascular histology, visualized by infusion of nonblocking rat anti-mouse GPIbβ antibody conjugated to DyLight 649. The C57BL/6 mice were injected with 5 μmol kg−1micellar formulated mP6 or mP6Src (negative control), 12 μmol kg−1Integrillin or buffer, 3 min before laser-induced arteriolar wall injury. White arrows indicate the directions of the blood flow. b, The mean platelet fluorescence intensity for 30 thrombi (performed in three mice) for each treatment at selected time points (mean ± s.e.m., n = 30, t-test). Fluorescence in mP6- and Integrilin-treated mice is minimal. c, Comparison of mP6 (5 μmol kg−1) with the same dose of Integrilin and their respective controls in occlusion time of FeCl3-induced carotid artery thrombosis in mice. Typical arterial blood flow charts of FeCl3-induced occlusive thrombosis are shown. d, Comparison of mP6 (5 μmol kg−1) with the same dose of Integrilin and controls in mouse tail bleeding analysis. Released haemoglobin levels were used as a parameter to assess blood loss (mean ± s.d., n = 10).
a, Estimation of intracellular levels of 1-pyrenyldiazomethane (PDAM)-conjugated mP6 and mP6Scr following incubation with platelets for 5 min. Platelets were pelleted by centrifugation, and the amounts of PDAM-conjugated peptides in platelet lysates were estimated (mean ± s.d., n = 3). b, Haemogram of mouse whole blood before or 1 h after injection of mP6 or mP6Scr (5 μmol kg−1), showing no significant differences.
This file shows Intravital videomicroscopy of the cremaster muscle arteriolar circulation and platelet thrombus formation (red) after laser injury, with HEPES buffer injection. (MOV 8136 kb)
This file shows Intravital videomicroscopy of the cremaster muscle arteriolar circulation and thrombus formation (red) after laser injury, with Integrilin (12 µM per kg) injection. (MOV 9312 kb)
This file shows Intravital videomicroscopy of the cremaster muscle arteriolar circulation and thrombus formation (red) after laser injury, with mP6Scr micelle (5 µM per kg) injection. (MOV 8138 kb)
This file shows Intravital videomicroscopy of the cremaster muscle arteriolar circulation and thrombus formation (red) after laser injury, with mP6 micelle (5 µM per kg) injection. (MOV 7981 kb)
About this article
Cite this article
Shen, B., Zhao, X., O’Brien, K. et al. A directional switch of integrin signalling and a new anti-thrombotic strategy. Nature 503, 131–135 (2013). https://doi.org/10.1038/nature12613
Journal of Translational Medicine (2021)
Nature Communications (2021)
GNA13 regulates BCL2 expression and the sensitivity of GCB-DLBCL cells to BCL2 inhibitors in a palmitoylation-dependent manner
Cell Death & Disease (2021)
Nature Reviews Drug Discovery (2020)
Structure-guided design of pure orthosteric inhibitors of αIIbβ3 that prevent thrombosis but preserve hemostasis
Nature Communications (2020)