Integrin signaling through Arg activates p190RhoGAP by promoting its binding to p120RasGAP and recruitment to the membrane

Abstract

The Rho family GTPases RhoA (Rho), Rac1, and Cdc42 are essential effectors of integrin-mediated cell attachment and spreading. Rho activity, which promotes formation of focal adhesions and actin stress fibers, is inhibited upon initial cell attachment to allow sampling of the new adhesive environment. The Abl-related gene (Arg) tyrosine kinase mediates adhesion-dependent inhibition of Rho through phosphorylation and activation of the Rho inhibitor p190RhoGAP-A (p190). p190 phosphorylation promotes its binding to p120RasGAP (p120). Here, we elucidate the mechanism by which p120 binding regulates p190 activation after adhesion. We show that p190 requires its p120-binding domain to undergo Arg-dependent activation in vivo. However, p120 binding does not activate p190RhoGAP activity in vitro. Instead, activation of p190 requires recruitment to the cell periphery. Integrin-mediated adhesion promotes relocalization of p190 and p120 to the cell periphery in wild-type fibroblasts, but not in arg(-/-) fibroblasts. A dominant-negative p120 fragment blocks p190:p120 complex formation, prevents activation of p190 by adhesion, and disrupts the adhesion-dependent recruitment of p190 to the cell periphery. Our results demonstrate that integrin signaling through Arg activates p190 by promoting its association with p120, resulting in recruitment of p190 to the cell periphery where it inhibits Rho.

Figure 1.

Arg is required for adhesion-dependent Rho inhibition. (A) Top, relative Rho activity plotted as a function of time. Active Rho levels were assessed in wild-type (WT, diamonds), arg^−/− (squares), and arg^−/− fibroblasts expressing an Arg-yellow fluorescent protein (arg^−/− + Arg-YFP) (triangles) held in suspension (0 min) or plated on FN for 10, 20, 30, or 60 min by using an ELISA assay. Relative Rho activity was determined by dividing the absorbance reading at each time point by the absorbance of the 0 time point for each cell type. Mean ± SE, n ≥ 3. Analysis of variance between all cell types: 10-min time point, p = 0.0003; 20-min time point, p = 0.0001; 30-min time point, p = 0.0026; and 60-min time point, p = 0.0486. Post hoc Student-Newman-Keuls test for each time point (*p < 0.05). Bottom, total Rho was determined for each cell type at each time point to ensure equal loading for the ELISA assay. One hundred micrograms of total protein extract was immunoblotted for RhoA. (B) One hundred micrograms of protein extract from WT, arg^−/−, and arg^−/− + Arg-YFP were immunoblotted for Arg, p190, p120, and Rho. Our current lot of anti-Arg antisera cross-reacts slightly with endogenous Abl, and these cross-reacting bands in arg^−/− and arg^−/− + Arg-YFP lysate are indicated with an asterisk (*).

Figure 2.

Stimulation of p190 by Arg requires the p120-binding domain. (A) Domain structure of p190RhoGAP (p190). Domains indicated are as follows: GTP-B, GTP-binding domain; RBR, p120RasGAP-binding region; GAP, GTPase-activating protein domain. Truncation mutants lacking the GTP-binding domain (381-C); both the GTP-binding domain and the middle sequence (760-C); and a truncation mutant expressing only the GAP domain (1135-C) were generated. (B) Rhotekin pull-down assays were performed from lysates of HEK293 cells transfected with expression vectors for HA-Rho alone (Ctrl) or together with p190 +/− Arg, or the various p190 truncation mutants +/− Arg as indicated. Active Rho pull-down (top) and total cell lysate immunoblots (middle) were probed as indicated. After quantifying bands by densitometry, active Rho levels in each sample were normalized to total Rho in the sample and this was normalized to the control ratio. Mean ± SE, n ≥ 3 (bottom). (C) HA-tagged p190 was immunoprecipitated with anti-HA antibody from lysates of HEK293 cells that had been untransfected (Ctrl) or transfected with expression vectors for full-length p190 or the various truncation mutants as indicated. Immunoprecipitates on the same blot were probed for p120 (top) and p190 (middle) as indicated. Total protein lysates (25 μg of protein) were probed for p120 (bottom).

Figure 3.

Binding of p120 does not activate p190RhoGAP in vitro. (A) Coomassie stain of purified p190 (lane 2), in vitro phosphorylated p190 (lane 3), and purified p190:p120 complex (lane 4). Standard protein molecular masses are shown for 150 and 100 kDa (lane 1). (B) GTP hydrolysis by RhoA was monitored over time in the absence of p190 (triangles), or in the presence of p190 (diamonds) or in vitro phosphorylated p190 (squares). (C) GTP hydrolysis by RhoA was measured over time in the absence of p190 (diamonds), or in the presence of p190 (triangles) or p190:p120 complex (1:1 M ratio; squares). Mean ± SE, n ≥ 3.

Figure 4.

p190 activation in vivo requires p120 binding. (A) p120 binds to p190 in a phosphotyrosine-dependent manner. The model depicts binding of the p120-2-3-2 fragment to p190, which inhibits the binding of p120 to p190. (B) p190 was immunoprecipitated from lysates of HEK293 cells that had been transfected with expression vectors for HA-p190, Arg, and/or the HA-tagged 2-3-2 fragment as indicated. Immunoprecipitates on the same blot were probed for p120 and HA-p190 (top two panels) as indicated. Total protein lysates (100 μg of protein) were probed for p120, Arg, or HA-2-3-2 fragment (bottom three panels). (C) Rhotekin pull-down assays were performed from lysates of HEK293 cells transfected with expression vectors for HA-Rho alone (lane 1), or together with HA-p190, Arg, and/or the 2-3-2 fragment as indicated (lanes 2–8). Active Rho pull-down (top) and total cell lysate immunoblots (total protein panels) were probed and quantitated as described in Figure 2. Mean ± SE, n ≥ 3 (bottom).

Figure 5.

Dominant-negative p120 blocks adhesion-dependent p190 activation. (A) Rhotekin pull-down assays were performed from lysates of nonstimulated (−FN; lanes 1–4) or FN-stimulated (+FN; lanes 5–8) HEK293 cells. Cells were transfected with expression vectors for HA-Rho alone (lanes 1 and 5) or together with HA-p190 and/or the 2-3-2 fragment as indicated (lanes 2–4 and 6–8). Active Rho was quantified as in Figure 2. Mean ± SE, n ≥ 3 (bottom). (B) p190 was immunoprecipitated from lysates of nonstimulated (−FN; lanes 1–4) or FN-stimulated (+FN; lanes 5–8) HEK293 cells. Cells were transfected with expression vectors for GFP alone (control; lanes 1 and 5) or HA-p190 and/or 2-3-2 fragment as indicated (lanes 2–4 and 6–8). Pull-downs (top three panels) were probed as indicated. Total protein lysates (100 μg of protein) were probed for p120RasGAP (bottom).

Figure 6.

Arg is required for p190:p120 colocalization at the cell periphery. WT (A–E, K–O), arg^−/− (F–J, P–T), or arg^−/− + Arg-YFP cells (U–AD) were plated on coverslips coated with either 0.1% PLL (A–J) or 10 μg/ml FN (K–AD), and stained for p190 (A, F, K, P, and V), p120 (B, G, L, Q, and AA), and F-actin (C, H, M, R, W, and AB). Merged images (D, I, N, and S) and enlargements (E, J, O, and T) of each channel in top grid show p190 in red, p120 in green, and F-actin in blue. Merged images and enlargements in bottom grids show either p190 (X–Y) or p120 (AC and AD) in red and F-actin (X and Y, AC and AD) in blue. Enlarged regions are indicated by white dotted boxes in merge panels. White arrowheads highlight regions of p190:p120 colocalization (N and O), or individual p190 (X and Y) or p120 localization (AC and AD). Bar in A (applies to A–D, F–I) and K (applies to K–N, P–S, U–X, Z–AC), 10 μm.

Figure 7.

The 2-3-2 fragment inhibits p190 recruitment to the cell periphery. Wild-type cells were transfected with either RFP (A and K) or 2-3-2-RFP (F and P), plated on FN-coated coverslips, and in some cases immunostained for p190 (L and Q). p120 localization in RFP or 2-3-2-RFP transfected cells was assessed by cotransfection with p120-GFP (B and G). p190FF-YFP was transfected into wild-type cells to asses its localization (W). F-actin was stained in all cell types (C, H, M, R, and X). Merged images and enlargements show either p120 (D, E, I, and J), p190 (N and O, S and T), or p190FF (Y and Z) in green and F-actin in blue. Enlargement regions are indicated by white dotted boxes in merge panels. White arrowheads highlight regions of p120 (D and E, I and J) or p190 localization (N and O). p120, p190, and p190FF localization at the cell periphery was quantified. (U and V) Black bars indicate RFP-transfected cells for p190 and p120 localization, and gray bars indicate 2-3-2-RFP transfected cells, with number of cells (n ≥ 50) indicated in each instance. Student's t test between p190 localization in RFP versus 2-3-2-RFP–expressing cells, *p < 0.0001. Bar in A, 10 μm.