Direct Binding of Rap1 to Talin1 and to MRL Proteins Promotes Integrin Activation in CD4+ T Cells

Abstract

Agonist-induced Rap1 GTP loading results in integrin activation involved in T cell trafficking and functions. MRL proteins Rap1-interacting adapter molecule (RIAM) and lamellipodin (LPD) are Rap1 effectors that can recruit talin1 to integrins, resulting in integrin activation. Recent work also implicates direct Rap1-talin1 interaction in integrin activation. Here, we analyze in mice the connections between Rap1 and talin1 that support integrin activation in conventional CD4⁺ T (Tconv) and CD25^HiFoxp3⁺CD4⁺ regulatory T (Treg) cells. Talin1(R35E, R118E) mutation that disrupts both Rap1 binding sites results in a partial defect in α_Lβ₂, α₄β₁, and α₄β₇ integrin activation in both Tconv and Treg cells with resulting defects in T cell homing. Talin1(R35E,R118E) Tconv manifested reduced capacity to induce colitis in an adoptive transfer mouse model. Loss of RIAM exacerbates the defects in Treg cell function caused by the talin1(R35E,R118E) mutation, and deleting both MRL proteins in combination with talin1(R35E,R118E) phenocopy the complete lack of integrin activation observed in Rap1a/b-null Treg cells. In sum, these data reveal the functionally significant connections between Rap1 and talin1 that enable α_Lβ₂, α₄β₁, and α₄β₇ integrin activation in CD4⁺ T cells.

FIGURE 1.

Rap1 binding to talin1 F0 or F1 subdomain contributes to integrin activation in CD4⁺ T cells and Treg cells. (A) Diagram depicting the role of talin1 in integrin activation. Mutations R35E and R118E in talin1 block Rap1 binding to F0 and F1 subdomains, respectively. (B–D) Binding of control and Tln1-cR35E or Tln1-cR118E mutant CD4⁺ T cells to soluble integrin ligands. Cellular stimulation was achieved with 100 nM PMA. Data represent mean ± SEM (n = 4 mice), normalized to the unstimulated control condition. Statistical significance in (B–D) was measured against the WT control T cells. One-way ANOVA with Bonferroni post hoc test. ***p < 0.001.

FIGURE 2.

Blockade of Rap1 binding to both F0 and F1 subdomains in talin1 prevents integrin activation in CD4⁺ T cells. (A) Both R35E and R118E mutations in talin1 F0 and F1 subdomains, respectively, prevent Rap1–talin1 interaction. (B) Peripheral blood cell counts of Tln1-cR35E,R118E mice. Values are mean ± SEM (n = 6 mice). Two-tailed t test; no significant differences were observed. PMN, polymorphonuclear neutrophils; Ly, lymphocytes; Mono, monocytes. (C) Binding of control or Tln1-cR35E,R118E mutant CD4⁺ T cells to soluble integrin ligands. Cells were stimulated with 100 nM PMA. Data represent mean ± SEM (n = 10 mice), normalized to the unstimulated control condition. Two-way ANOVA with Bonferroni post hoc test.***p < 0.001.(D) Static adhesion of control or Tln1-cR35E,R118E mutant CD4⁺ T cells to integrin ligands. Cells were stimulated with 1μg/ml SDF1α or 100nM PMA. The number of arrested adherent cells was plotted. Data are mean ± SEM (n = 6mice). Two-way ANOVA with Bonferroni post hoc test.*p < 0.05,***p < 0.001.(E and F) In vivo competitive homing of Tln1-cR35E,R118E mutant CD4⁺ T cells to different lymphoid tissues. CD4⁺ T cells were isolated from either control or Tln1-cR35E,R118E mice, differentially labeled, and mixed before injection into C57BL/6 mice. The ratio of Tln1-cR35E,R118E mutant to control CD4⁺ T cells was determined by flow cytometry from PLNs and MLNs or spleens 3h after injection. Data represent mean ± SEM (n = 9 mice). One-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (G) In vitro cell proliferation assay. CD4⁺ T cells were isolated from spleens and labeled with CFSE before stimulation with plate-bound anti-CD3/CD28 mAb for 4 d at 37°C. Proliferation index was plotted. Data represent mean ± SEM (n = 4 mice). Two-way ANOVA with Bonferroni post hoc test. **p < 0.01. (B–E and G) Statistical significance was measured against the WT control T cells or mice.

FIGURE 3.

Loss of interaction between Rap1 and talin1 protects mice from colitis. Conventional T cells (10⁶) from WT or Tln1-cR35E,R118E mice were adoptively transferred into Rag1^−/− mice. (A) Changes in body weight are plotted. Values are normalized as a percentage of the original weight. Two-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (B) Percentage survival. Data represent mean ± SEM (n = 12 mice). Gehan-Breslow-Wilcoxon test. (C) mRNA expression of IL-1β, TNF-α, IL-6, IFN-γ, and IL-17A in distal colon tissues from individual groups of mice relative to PBS-injected control mice at Day 100. Results are normalized to expression of GAPDH. Mean ± SEM are plotted. Two-tailed t test; *p < 0.05, **p < 0.01. (A–C) Statistical significance was measured against the WT control T cells or mice.

FIGURE 4.

Rap1–talin1 direct interaction synergizes with Rap1–RIAM–talin1 to mediate integrin activation in CD4⁺ T cells. (A) RIAM functions as a Rap1 effector that recruits talin1 to the plasma membrane, in parallel to Rap1 binding to talin1, to enable its interaction with the integrin β cytoplasmic tail. (B) In vivo competitive homing of Tln1-cR35E,R118E; RIAM-cKO CD4⁺ T cells to different lymphoid tissues. CD4⁺ T cells were isolated from either control or Tln1-cR35E,R118E; RIAM-cKO mice, differentially labeled, and mixed before injection into C57BL/6 mice. The ratio of Tln1-cR35E,R118E; RIAM-cKO mutant to control CD4⁺ T cells was determined by flow cytometry from PLNs and MLNs or spleens (SP) 3 h after injection. Data represent mean ± SEM (n = 6 mice). One-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (C) Binding of control, Tln1-cR35E,R118E, or Tln1-cR35E,R118E; RIAM-cKO mutant CD4⁺ T cells to soluble integrin ligands. Cells were stimulated with 100 nM PMA. Data represent mean ± SEM (n = 4 mice), normalized to the unstimulated control condition. Two-way ANOVA with Bonferroni post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001. (D and E) CD4⁺ T cells isolated from control or Tln1-cR35E,R118E mice were transduced with retroviruses encoding either EGFP-tagged RIAM or EGFP alone. Integrin activation in EGFP⁺ cells was analyzed by flow cytometry. (D) The level of EGFP expression was determined at 3 d after transduction. (E) Binding of transduced CD4⁺ T cells to soluble integrin ligands. Cells were stimulated with 100 nM PMA. Data represent mean ± SEM (n = 4 mice). Two-way ANOVA with Bonferroni post hoc test. **p < 0.01, ***p < 0.001. (B–D) Statistical significance was measured against the WT control T cells.

FIGURE 5.

Rap1 binding to talin1 participates to activate integrins in Treg cells. (A) Mouse breeding strategy to selectively restrict the expression of talin1(R35E,R118E) in Treg cells. (B) Peripheral blood cell counts of Tln1-rR35E,R118E mice. Values are mean ± SEM (n = 6 mice). Two-tailed t test; *p < 0.05, **p < 0.01. PMN, polymorphonuclear neutrophils; Ly, lymphocytes; Mono, monocytes. (C) Binding of control or Tln1-rR35E,R118E mutant Treg cells to soluble integrin ligands. Cells were stimulated with 100 nM PMA. Data represent mean ± SEM (n = 8 mice), normalized to the unstimulated control condition. Two-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (D) In vivo competitive homing of Tln1-rR35E,R118E mutant Treg cells to different lymphoid tissues. Treg cells were isolated from either control or Tln1-rR35E,R118E mice, differentially labeled, and mixed before injection into C57BL/6 mice. The ratio of Tln1-rR35E,R118E mutant to control Treg cells was determined by flow cytometry from PLNs and MLNs or spleens (SP) 3 h after injection. Data represent mean ± SEM (n = 9 mice). One-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (E and F) Treg cell suppressive activity. Treg cells isolated from CD45.2 congenic control or Tln1-rR35E,R118E mice were mixed with responder T cells at the indicated Treg/responder cell ratios. Responder cells were CFSE-labeled CD4⁺CD25⁻ naïve T cells isolated from CD45.1 congenic C57BL/6 mice and activated by anti-CD3 (5 μg/ml), anti-CD28 (5 μg/ml), and IL-2. CFSE⁺ populations gated on CD45.1⁺ cells were analyzed by flow cytometry at Day 4 to determine the proliferation index using FlowJo software (n = 3 mice). Two-way ANOVA with Bonferroni post hoc test. **p < 0.01. (B–D and F) Statistical significance was measured against the WT control T cells or mice.

FIGURE 6.

Both LPD and Rap1 binding to talin1 contribute to activate integrins in Treg cells. (A and B) Representative morphology (A) and lymphoid tissues (B) of Tln1-rR35E,R118E; RIAM-rKO mice. ALN, aortic lymph nodes; BLN, brachial lymph nodes; ILN, inguinal lymph nodes; SP, spleen; THY, thymus. (C) Binding of control, RIAM-rKO, Tln1-rR35E,R118E, or Tln1-rR35E,R118E; RIAM-rKO mutant Treg cells to soluble integrin ligands. Cells were stimulated with 100 nM PMA. Data represent mean ± SEM (n = 4 mice), normalized to the unstimulated control condition. Two-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (D) Treg cell suppressive activity. Treg cells isolated from CD45.2 congenic control or Tln1-rR35E,R118E; RIAM-rKO mice were mixed with responder T cells at the indicated Treg/responder cell ratios. Responder cells were CFSE-labeled CD4⁺CD25⁻ naive T cells isolated from CD45.1 congenic C57BL/6 mice and activated by anti-CD3 (5 μg/ml), anti-CD28 (5 μg/ml), and IL-2. CFSE⁺ populations gated on CD45.1⁺ cells were analyzed by flow cytometry at Day 4 to determine the proliferation index using FlowJo software (n = 4 mice). Two-way ANOVA with Bonferroni post hoc test. **p < 0.01, ***p < 0.001. (E) In vivo competitive homing of Tln1-rR35E,R118E; RIAM-rKO Treg cells to different lymphoid tissues. Treg cells were isolated from either control or Tln1-rR35E,R118E; RIAM-rKO mice, differentially labeled, and mixed before injection into C57BL/6 mice. The ratio of Tln1-rR35E,R118E; RIAM-rKO mutant to control Treg cells was determined by flow cytometry from PLNs and MLNs or spleens 3 h after injection. Data represent mean ± SEM (n = 8 mice). One-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (F) Binding of control, Tln1-rL325R, or Tln1-rR35E,R118E; RIAM-rKO; LPD-rKO mutant Treg cells to soluble integrin ligands. Cells were stimulated with 100 nM PMA. Data represent mean ± SEM (n = 4 mice). Two-way ANOVA with Bonferroni post hoc test. ***p < 0.001. (C–F) Statistical significance was measured against the WT control T cells.