Kindlin-2 (Mig-2): a co-activator of beta3 integrins

Abstract

Integrin activation is essential for dynamically linking the extracellular environment and cytoskeletal/signaling networks. Activation is controlled by integrins' short cytoplasmic tails (CTs). It is widely accepted that the head domain of talin (talin-H) can mediate integrin activation by binding to two sites in integrin beta's CT; in integrin beta(3) this is an NPLY(747) motif and the membrane-proximal region. Here, we show that the C-terminal region of integrin beta(3) CT, composed of a conserved TS(752)T region and NITY(759) motif, supports integrin activation by binding to a cytosolic binding partner, kindlin-2, a widely distributed PTB domain protein. Co-transfection of kindlin-2 with talin-H results in a synergistic enhancement of integrin alpha(IIb)beta(3) activation. Furthermore, siRNA knockdown of endogenous kindlin-2 impairs talin-induced alpha(IIb)beta(3) activation in transfected CHO cells and blunts alpha(v)beta(3)-mediated adhesion and migration of endothelial cells. Our results thus identify kindlin-2 as a novel regulator of integrin activation; it functions as a coactivator.

Figure 1.

Sequences of the membrane-distal region of β₃ CT have essential roles in integrin α_IIbβ₃ activation. (A) Alignment of integrin β CT sequences, highlighting (red) the conserved regions, the two NXXY/F motifs and one T/S cluster. (B) Suppression of integrin α_IIbβ₃-mediated cell spreading by expressed β₃ CT depends on conserved sequences in its membrane-distal region. After transient transfection with plasmids encoding the indicated β₃ CT-containing chimera (β₃ CT/PSGL-1), adhesion of the α_IIbβ₃-CHO cells to fibrinogen was examined. The adherent cells were fixed and stained with the anti-PSGL-1 mAb, KPL-1, for visualization by fluorescence microscopy (10× objective). Bar, 20 μm. (C) Conserved residues in the membrane-distal region support α_IIbβ₃ activation. Plasmids encoding α_IIb and β₃ or its mutants were transiently transfected to CHO cells. The transfected cells were stained with 2G12 to assess α_IIbβ₃ expression or PAC1 to assess α_IIbβ₃ activation. FACS was used to measure the mean fluorescence intensity (MFI) of 2G12 or PAC1 binding, and relative MFI of PAC1 binding was normalized to integrin expression levels based on 2G12 staining (Ma et al., 2006). The error bars represent means ± SD of three independent experiments.

Figure 2.

Kindlin-2 enhances talin-induced integrin α_IIbβ₃ activation. EGFP-fused β₃ CT binding proteins were transiently transfected into α_IIbβ₃-CHO cells. Their effects on α_IIbβ₃ activation were evaluated by PAC1 binding (A and D) and expression levels were measured by Western blotting with anti-GFP antibody (B and E). Representative FACS histograms of PAC1 binding to talin-H and/or kindin-2 (kind-2) positive cells (C). Error bars (A and D) represent means ± SD (n = 3). *, P < 0.05; **, P < 0.01 (versus vector).

Figure 3.

Distinct binding sites for kindlin-2 and talin in β₃ CT. (A) Lysates of CHO cells, HUVECs, or out-dated platelets were incubated with GST or GST-fused β₃ CT bearing the indicated mutations in the presence of glutathione-Sepharose. After washing, the precipitates were analyzed by SDS-PAGE. The loading of the GST proteins was assessed by Coomassie blue staining. The associated kindlin-2 or talin-H was detected in Western blots with anti-kindlin-2 or anti-talin-H. (B) Amino acid sequences of β₃ CT C-terminal peptide corresponding to Y⁷⁴⁷-T⁷⁶² and a mutant peptide with two loss-of-function mutations, S⁷⁵²P and Y⁷⁵⁹A. (C) The pull-down assay was performed in the presence of indicated peptides. The influence of these peptides on kindlin-2 or talin-H binding to β₃ CT was evaluated by SDS-PAGE and Western blotting.

Figure 4.

Both the N and C terminus of kindlin-2 are required for β₃ CT association and support of talin-induced integrin activation. (A) Organization of predicated domains of kindlin-2 protein. The FERM domain is shown in yellow, in which the F₂ subdomain is split by the PH domain. Deletion mutations from N terminus (ΔN) or C terminus (ΔC) are indicated. (B) The lysates of CHO cells transfected with EGFP-kindlin-2 with indicated mutations were used for pull-down assays. After incubating with GST fusion β₃ CT (wild-type) and glutathione-Sepharose, kindlin-2 protein bound to the β₃ CT was evaluated by SDS-PAGE and Western blotting using anti-GFP antibody. Kindlin-2 expression levels in lysates are also shown. (C) CHO cells expressing α_IIbβ₃ were transiently transfected with empty EGFP vector or cDNA encoding the indicated proteins. Binding of PAC1 to the different transfectants was assessed by FACS and relative MFI of PAC1 binding were calculated as described in Materials and methods. Error bars represent means ± SD (n = 3). **, P < 0.01 (versus talin-H).

Figure 5.

Endogenous kindlin-2 supports β₃ integrin function in cells. (A) Subcellular localizations of kindlin-2 and β₃ integrin. HUVECs spread on fibrinogen for 30 or 60 min were stained with the anti-kindlin-2 mAb and anti-β₃ subunit polyclonal antibody followed by AlexaFluor 568 anti–mouse IgG and AlexaFluor 488 anti–rabbit IgG. Bar, 10 μm. (B) RNAi suppression of kindlin-2 expression in CHO cells. Expression of kindlin-2 in parental CHO cells (non-T), kindlin-2 siRNA (SiKind-2), or control RNA (SiControl) transfectants was analyzed by Western blotting with kindlin-2 or actin antibodies. (C) CHO cells expressing α_IIbβ₃ were transiently transfected with vector or talin-H, together with control RNAs (SiControl) or siRNAs targeting kindlin-2 (SiKind-2). The binding of PAC1 to the different transfectants was assessed by FACS and MFI of PAC1 binding was calculated. The error bars are means ± SD (n = 3). (D) RNAi suppression of kindlin-2 expression in HUVECs. (E and F) Non-transfected (Non-T) or HUVECs transfected with control RNAs (SiControl) or targeted siRNAs for Kindlin-2 (SiKind-2) were used in adhesion assays (E) or migration assays (F). The adherent or migrated cells were fixed, stained, and counted (10× objective). The error bars are means ± SD of three independent experiments. (G) HUVECs transfected with control RNAs (SiControl) or targeted siRNAs for kindlin-2 (SiKind-2) were stimulated with PMA, and adhesion to fibrinogen was measured. (H) Kindlin-2 as an integrin coactivator. Integrin activation depends on interaction of talin-H with the NPLY⁷⁴⁷ motif and the membrane-proximal clasping region. Kindlin-2 facilitates activation by associating with the C-terminal regions of β₃ CT, involving the TS⁷⁵²T and NITY⁷⁵⁹ motifs.