Integrin αII b tail distal of GFFKR participates in inside-out αII b β3 activation

Abstract

Background: Increases in ligand binding to integrins (activation) play critical roles in platelet and leukocyte function. Integrin activation requires talin and kindlin binding to integrin β cytoplasmic tails. Research has focused on the conserved GFFKR motif in integrin αII b tails, integrin β cytoplasmic tails and the binding partners of β tails. However, the roles of αII b tail distal of GFFKR motif are unexplored.

Objective: To investigate the role of αII b tail distal of GFFKR in talin-mediated inside-out integrin signaling.

Methods: We used model cell systems to examine the role of αII b tail distal of GFFKR in bidirectional αII b β3 signaling and αII b β3 -talin interactions.

Results: Deletion of amino acid residues after the GFFKR motif in αII b tail moderately decreased β3 (D723R)-induced activation, abolished talin-induced αII b β3 activation in model cells, and inhibited agonist-induced αII b β3 activation in megakaryocytic cells. Furthermore, residues in αII b tail distal of GFFKR did not affect outside-in αII b β3 signaling or αII b β3 -talin interaction. Addition of non-homologous or non-specific amino acids to the GFFKR motif restored αII b β3 activation in model cells and in megakaryocytic cells. Molecular modeling indicates that β3 -bound talin sterically clashes with the αII b tail in the αII b β3 complexes, potentially disfavoring the α-β interactions that keep αII b β3 inactive.

Conclusion: The αII b tail sequences distal of GFFKR participate in talin-mediated inside-out αII b β3 activation through its steric clashes with β3 -bound talin.

Keywords: Integrin αIIbβ3; cell adhesion; kindlin-2 protein, human; signal transduction; talin.

Figure 1

The presence, not the sequence, of αIIb tail distal of GFFKR is required for inside-out αIIbβ3 activation. (A) Truncation of αIIb tail after GFFKR significantly inhibited β3(D723R)-induced αIIbβ3 activation, while presence of irrelevant residues after GFFKR restores such activation. CHO cells were transiently transfected with αIIb and β3 construct as indicated. Integrin activation was expressed as normalized PAC1 binding and calculated as (MFI-MFI₀)/MFI_D57 where MFI was the mean fluorescence intensity of bound activation-specific antibody PAC-1 and MFI₀ was that in the presence of an αIIbβ3-specific antagonist, eptifibatide, and MFI_D57 was the mean fluorescence intensity of anti-αIIbβ3 antibody, D57. (B) Truncation of the αIIb tail after GFFKR completely abolished talin- and kindlin-induced αIIbβ3 activation, while presence of irrelevant residues after GFFKR restores such activation. CHO cells stably expressing αIIbβ3 or αIIbβ3 mutant were transfected with a transfection marker (tomato) and pcDNA, THD, THD+kindlin-1 or THD+kindlin-2. Integrin activation in transfected cells was expressed as normalized PAC1 binding as described in (A). In (A) and (B), asterisks indicate statistical significance at 95% confidence level with two-tail, unpaired t-test. (C) Protein expression of transfected cells from (B) were analyzed by western blotting with anti-Flag antibody for kindlins, anti-HA for THD, anti-αIIb tail (Rb8276), anti-β3 tail (Rb8275), and anti-GAPDH. (D) CHO cells stably expressing αIIbβ3 or αIIbΔ996β3 were transfected with increasing amount of cDNA encoding THD and a transfection marker (tomato). Integrin activation in transfected cells was expressed as normalized PAC1 binding as described in (A). Right panel of (D) shows the expected THD expression in transfected cells. Error bars in (A), (B) and (C) are standard errors of at least 3 independent experiments.

Figure 2

The presence, not the sequence, of αIIb tail distal of GFFKR is required for thrombin-induced αIIbβ3 activation in human acute megakaryocytic leukemia cells (CMK). (A) CMK cells activate αIIbβ3 in response to thrombin stimulation. Unstimulated, EDTA-treated, and thrombin-stimulated CMK cells were double stained with GPIb, a marker for megakaryocytes and platelets, and activation specific antibody PAC1 as described in methods. As previously reported [–24], PAC1 binding was increased in GPIb-high CMK cells upon thrombin stimulation. Histogram shows the overlay of the gated regions (box) from the FACS plots. (B) CMK cells were transduced with αIIbβ3 or mutant as indicated, doubled stained with GPIb and PAC1, and analyzed as in (A). The thrombin-induced PAC1 binding was used to measure the activation of each integrins upon thrombin stimulation and was calculated as (MFI_thrombin−MFI_unstimulated), where MFI_thrombin was the mean fluorescence intensity of bound PAC-1 in the presence of 1U thrombin and MFI_unstimulated was that before stimulation. CMK cells express endogenous αIIbβ3 and thus can respond to thrombin (bar of the uninfected). Thrombin-induced αIIbβ3 activation was increased in CMK cells transduced with wt αIIbβ3, αIIbΔ1001β3 or αIIb_5Aβ3 but not with αIIbΔ996β3. (C) CMK cells from (B) were double stained with GPIb and a rabbit polyclonal anti-αIIbβ3 antibody to measure the expression level of αIIbβ3 in GPIb-high cells. CMK cells transduced with αIIbβ3, αIIbΔ996β3, αIIbΔ1001β3 or αIIb_5Aβ3 had similar high level of αIIbβ3 expression over the uninfected, indicating that the defects of αIIbΔ996β3 seen in (B) is a direct result of lost responsiveness of αIIbΔ996β3 to thrombin stimulation. Error bars in B and C indicate standard errors of 3 experiments.

Figure 3

The first few residues distal of GFFKR are required for optimal αIIb and β3 TMD-tail interactions. (A) Schematic representation of the αIIb and β3 TMD-tail interaction assays as previously described [26]. Flag tagged αIIb-TMD-tail-TAP and Tac-β3-TMD-tail fusion proteins were co-expressed in CHO cells. αIIb fusion protein was captured by calmodulin beads and amount of β3 TMD tail bound to the αIIb was analyzed by western blotting. (B) Western blots of the input and captured αIIb or β3 were shown. αIIb TMD-tail was blotted with anti-Flag and β3 TMD-tail was blotted with anti-Tac. αIIbΔ996 and αIIb_5A bind less to β3 than wt αIIb and αIIbΔ1001 do.

Figure 4

Deletion of αIIb tail after GFFKR does not affect αIIbβ3-talin interactions or outside-in integrin signaling. (A) CHO cells stably expressing αIIbβ3 wt or αIIbΔ996β3 were transfected with HA-talin and lysed after culturing for 24 hours. αIIbβ3 was immunoprecipitated with anti-αIIbβ3 (Rb8053) and precipitants were eluted with SDS-PAGE loading buffer, and detected by western blotting with anti-HA for talin or anti-β3. αIIbΔ996 truncation did not affect talin binding. (B) Recombinant αIIbβ3 or αIIbΔ996β3 was captured onto D57-conjugated agarose beads. The immobilized integrin beads were then incubated with various concentrations of purified THD, washed, eluted and analyzed by SDS-PAGE and western blotting. Bound THD was detected by blotting with anti-His6 tag and immobilized integrins were detected by coomassie staining (bottom two panels). There is no statistically significant difference between αIIbβ3 and αIIbΔ996β3 in their affinity for THD. Error bars indicate standard deviation of 3 independent experiments. (C) CHO cells stably expressing αIIbβ3 wt or αIIbΔ996β3 were serum starved overnight and plated on fibrinogen-coated plates for the specified periods of time, lysed and probed for signaling and control proteins as indicated in the western blots. αIIbΔ996 truncation did not affect signaling as assessed by phosphorylation of Akt or FAK. (D) CHO cells stably expressing αIIbβ3 or αIIbΔ996β3 were plated on fibrinogen-coated coverslips for 60 minutes, fixed with formaldehyde and imaged using phase-contrast microscope. There is no significant difference of cell spreading between cells expressing αIIbβ3 and αIIbΔ996β3. Scale bar is 50µm.

Figure 5

Binding of talin to β3 tail results in steric clashes between talin and αIIb. (A) Shown in red and blue are the structure of integrin αIIb and β3 TMD complex in cellular membranes [13]. The F3 subdomain of THD is shown in green. The structure of the entire complex was assembled by building a homology model of β3 tail after the β1D in the β1D-Talin F2F3 complex structure [18] and aligning the resulting β3-Talin F2F3 model to the β3 of αIIβ3 TMD complex structure determined in cellular membrane [13]. Talin head overlaps with αIIb tail in the model, indicating potential steric hindrance. (B) Sequence alignment of α tails. The α tails display no significant sequence similarity after GFFKR motif, suggesting no requirement for sequence specificity despite a number of them are regulated by talin and kindlins. (C) The capacity of THD to activate αIIbβ3 was not affected when αIIb tail was swapped with non-homologous sequences from α5 or αL. Integrin activation was expressed as normalized PAC1 binding and calculated as (MFI−MFI₀)/ MFI_D57 where MFI was the mean fluorescence intensity of bound PAC-1 and MFI₀ was that in the presence of an αIIbβ3-specific antagonist, eptifibatide, and MFI_D57 was the mean fluorescence intensity of D57. Asterisks indicate statistical significance at 95% confidence level with two-tail, unpaired t-test. Error bars are standard errors of at least 3 experiments. (D) Protein expressions of the transfected cells from (C).

Figure 6

Models of αIIb tail participating in talin-mediated αIIβ3 activation. (A) CHO cells stably expressing αIIβ3 (A5 cells) were transfected with a transfection marker (GFP), and pcDNA, or THD as indicated. Cells were treated with latrunculin A (2 µM) to inhibit actin polymerization or blebbistatin (50 µM) to inhibit myosin II ATPase activity. αIIβ3 activation was expressed as normalized PAC1 binding and calculated as (MFI−MFI₀)/ MFI_D57 where MFI was the mean fluorescence intensity of bound PAC-1 and MFI₀ was that in the presence of an αIIbβ3-specific antagonist, eptifibatide, and MFI_D57 was the mean fluorescence intensity of D57. The lower panel showed the protein expressions of the transfected cells. Asterisks indicate statistical significance at 95% confidence level with two-tail, unpaired t-test. Error bars are standard errors of 4 experiments. (B) Cells from (A) were fixed with 3.7% formaldehyde, permeabilized with 0.2% triton-100, stained with alexa488-phalloidin for actin, and imaged at 60× with a fluorescence microscope as described in the methods. At the concentration used in (A), latrunculin A demonstrably disrupted the cellular F-actin and blebbistatin markedly changed cell morphology, presumably as a result of blocked myosin II activity and loss of contractility. Scale Bar is 10 µm. (C) Model of talin-induced integrin activation. Besides causing topological changes in the β3 TMD, talin also clashes with the residues distal of GFFKR in αIIb tail, potentially displacing the αII tail. Combined effects from both mechanisms result in the loss of α-β interactions and integrin activation.