Applied stretch stimulates vinculin binding at focal adhesions, p130Cas phosphorylation and Rap activation. (A–D) B16F1 cells were seeded onto fibronectin-coated (5 µg/cm²) silicone rubber plates overnight before applying 10% equibiaxial stretch to the substrata for indicated times. (A) Cells were stained using vinculin antibodies and Rhodamine–phalloidin to visualize F-actin (left panels). Total vinculin intensity at adhesion complexes was quantified for unstretched control cells as well as stretched cells (left graph) and normalized to β1 integrin intensity (right graph). n>30 cells from three independent experiments. Scale bar: 10 µm. (B) Phosphorylation of Y165 in p130Cas was probed using phospho-specific antibodies. Total p130Cas is shown below. (C) The amount of active Rap1 (Rap1-GTP) was determined by performing a Ral-GDS-GST pulldown and probing with anti-Rap1 antibodies. Total Rap1 from a fraction of the lysates is shown below. (D) The levels of active GTP-bound Rac1 and RhoA were determined using G-LISA assays. Rap1, Rac1, and RhoA activities were normalized to time 0 (pre-stretch) samples. The mean±s.e.m. from three experiments is graphed. *P<0.05, **P<0.005, ***P<0.001 (unpaired two-tailed Student's t-tests).

Tension-induced focal adhesion formation and tumor cell invasion are dependent on Rap activation. (A,B) B16F1 cells stably expressing empty vector, Rap1V12 or Rap1GAPII were seeded within 3D collagen I+fibronectin (FN) gels and stretched for 24 h. Vinculin fluorescence is shown (A) and quantified (mean±s.e.m., n>30 cells; B). Scale bar: 10 μm. White arrows indicate the direction of applied stretch. (C) The elongation factor for >40 cells from three experiments was determined. (D) B16F1 cells stably expressing vector, Rap1V12 or Rap1GAPII were cultured in collagen gels and released to determine the percentage contraction compared to gels without cells added. The mean±s.e.m. from three experiments is graphed. (E,F) B16F1 cells transfected with either scrambled control siRNA or Rap1a plus Rap1b siRNAs were assessed for the expression of Rap1, Rap2 and β-actin by immunoblotting (E), and for elongation in 3D collagen gels (F). Scale bar: 50 μm. *P<0.05, **P<0.005, ***P<0.001 (paired two-tailed Student's t-tests).

Stretch-induced recruitment of vinculin to focal adhesions near the leading edge requires Rap activation and actin polymerization. (A) B16F1 cells were cultured on fibronectin-coated (2.5 µg cm⁻²) silicone rubber plates overnight before applying 10% equibiaxial stretch to the substrata for 10 min. Cells were treated with or without 1 µM latrunculin A for the last 3 min of applied stretch, fixed and stained for vinculin and β1 integrin. The ratio of vinculin intensity to β1 integrin intensity per adhesion was quantified for >20 cells from three experiments and graphed. Only adhesions at the leading edge were quantified. (B,C). B16F1 cells transiently expressing the empty vector or Rap1GAPII–FLAG were plated on fibronectin-coated silicone rubber plates overnight before applying 10% equibiaxial stretch to the substrata for 10 min where indicated. Cells were fixed and stained for vinculin and β1 integrin (B). The ratio of vinculin to β1 integrin intensities per adhesion was quantified for >15 cells from three experiments and graphed according to the distance from the leading edge of the cell (C). See also Fig. S1. Scale bar: 2 µm. *P<0.05; **P<0.01; ***P<0.001; NS, not significant (unpaired two-tailed Student's t-tests).

Rap activation is required for actin polymerization as well as PIP₃ accumulation and Rac1 activation at the leading edge of tumor cells. (A) B16F1 cells were transfected with actin–GFP plus either Rap1GAPII or an empty vector and then seeded onto fibronectin-coated (2.5 µg cm⁻²) chamber slides for 2 h. Peripheral actin structures were photobleached and the recovery of fluorescence was determined over at least 30 s. Examples are shown in left panels. The mean±s.e.m. of the percentage recovery of actin-GFP for >30 cells from three experiments is graphed. PB, pre-bleach. Scale bars: 10 µm. ***P<0.001 (paired two-tailed Student's t-tests). (B,C) B16F1 cells were transiently transfected with vector, Rap1GAPII–FLAG or Rap1N17-Myc and either Akt (PH)–GFP or Raichu-Rac1 before being plated on fibronectin-coated (2.5 µg/cm²) coverslips for the indicated times. (B) Cells were fixed and stained for FLAG or Myc to identify transfected cells. F-actin was visualized using Rhodamine–phalloidin and the amount of co-localized Akt(PH)–GFP and F-actin was quantified. Cells were divided into two equal regions (as indicated by the dotted lines), corresponding to leading and trailing edges when these were present, and the amount of colocalized Akt(PH)–GFP and F-actin in each half of the cell was determined. In each graph, the bars on the left represent the trailing part of the cell and the bars on the right represent the leading part of the cell. The mean±s.e.m. for 30 cells from three experiments is graphed and the numbers above each bar is the mean value. (C) Cells were fixed and stained for FLAG or Myc and the amount of activated Rac1 in each half of the cell, as indicated by the FRET intensity of the Raichu-Rac1 fusion protein, was determined by measuring the CFP fluorescence before and after bleaching YFP. FRET intensity in the trailing (left bars) and leading (right bars) parts of the cell are graphed (mean±s.e.m.). n=30 cells from 3 experiments. Scale bars: 10 µm.

Acute shear flow activates Rap1 to stabilize integrins at established focal adhesions. (A) B16F1 cells were resuspended in DMEM and either lysed or passed through a 28 G needle for 2 min. The amount of Rap1-GTP in cell lysates was assessed by western blotting. (B) For FRAP analysis, COS-7 cells were transfected with α5-integrin–GFP and seeded onto fibronectin-coated flow chamber slides. A syringe was used to draw medium from the end of the chamber, creating a 20-dyne shear force that could be rapidly turned on or terminated. Adhesions within cells were photobleached and allowed to recover for 90 s, either under shear or not. Shear was then applied (blue lines) or terminated (red lines) for 1 min and proximal adhesions were subsequently bleached and allowed to recover for a further 90 s. The percentage recovered GFP signal (i.e. the mobile fraction) for α5-integrin–GFP after each 90-s period is graphed. (C,D) B16F1 cells were transfected with α5-integrin–GFP, β1-integrin–GFP, or the G429N extended β1-integrin–GFP. For FRAP analysis, focal adhesions were photobleached under 20 dynes of shear or not. The percentage recovery of the GFP signal in the subsequent 90-s period is plotted. Each dot represents an individual ROI. k_on and k_off values that reflect the rates at which α5-integrin–GFP associated with and dissociated from focal adhesions were estimated from FRAP curves that were generated under no shear conditions. *P<0.01; ***P<0.001; NS, not significant (paired two-tailed Student's t-tests).