Figure 1.
Analysis of FAK-paxillin interaction in wild-type FAK and FAKI936/I998 cells.
(A) Structural view of the 4 helix bundle (α1-α4) of the FAT domain. Paxillin binds to 2 hydrophobic patches located at the interface of α-helix 1/4 and the interface of α-helix 2/3. The I936E and I998E mutations are indicated. (B) Representative blots showing wild-type FAK and FAKI936/I998 immunoprecipitated using anti-FAK Ab and blotted for paxillin and FAK. The expression level of proteins in the corresponding cell lysate is shown. (C) Confocal images from fixed cells expressing wild-type or mutant FAK (fused to GFP) and immunostained for a marker of FAs, paxillin (red). Note the absence of FAK at FAs in mutant cells. Scale bar, 20 μm.
Figure 2.
Phosphorylation status of FAK and FAK substrates in FAK-/- cells re-expressing wild-type FAK and FAKI936/I998.
(A) Graphs show decreased FAK phosphorylation at Tyr397, Tyr576 and Tyr925 in FAKI936/I998 cells compared to wild-type FAK cells (*p<0.013, **p<0.0002; n = 4 to 6 independent experiments). (B) Representative Western blots showing the phosphorylation states of paxillin and p130Cas in FAK-/-, FAK and FAKI936/I998-expressing cells. Graphs show decreased phosphorylation of paxillin in both FAKI936/I998 cells and FAK-/- cells (*p<0.03) compared to wild-type FAK cells (P = 0.0097, F = 6.165) and decrease phosphorylation of p130CAS in FAKI936/I998 cells (***p<0.001) and FAK-/- cells (*p<0.05) compared to wild-type FAK cells (p = 0.0004, F = 12.68; n = 3 to 8 independent experiments).
Figure 3.
Quantification of focal adhesions in FAK, FAKI936/I998 and FAK-/- cells.
(A) Confocal images from fixed FAK-/- cells expressing or not wild-type or mutant FAK-GFP and immunostained for paxillin. Evaluation of paxillin-containing FAs reveals no significant decrease in the number of FAs in FAKI936/I998 cells as compared to FAK-/- and FAK cells (P = 0.0602, F = 2.958; n = 3 independent experiments with 15 to 28 cells and more than 600 FAs counted per condition). Scale bar, 20 μm. (B) Size-distribution of FAs shows a clear deficit of sub-micron sized FAs in FAKI936/I998 (0<FA<0.5 μm2 ***p<0.001; 0.5<FA<1 μm2 **p<0.01) and FAK-/- cells (0<FA<0.5 μm2 ***p<0.001) compared to FAK cells.
Figure 4.
Effects of FAKI936/I998 expression on FA dynamics.
(A) Representative TIRF images of FAK, FAKI936/I998 and FAK-/- cells. Formation of nascent adhesions at the front and dissociation of FAs at the rear (arrows) in FAK, FAKI936/I998 and FAK-/- cells transfected with CFP-paxillin. Arrows point to disassembling FAs at the trailing edge. Insets are magnified views of nascent adhesions over time in the yellow ellipses. Images were acquired at 1 min intervals for 1 h (scale bar, 5 μm) and representative images from the time-lapse sequence are shown at 2 min intervals. Colour images represent the superposition of images taken at t0 (red) and t60 (green). (B) Percentages of stable FAs were calculated from images in (A) as the number of FAs in FAK, FAKI936/I998 and FAK-/- cells at t = 60 min compared to t = 0. (P = 0.0055, F = 9.145; *p<0.03, n = 3 independent experiments).
Figure 5.
Effect of FAKI936/I998 expression on adhesion and migration.
(A) For adhesion assay, cells were seeded in fibronectin-coated wells and allowed to adhere for 1 h before quantification. (P<0.0001, F = 28.96; 3 independent experiments done in sextuplicates). ***p<0.0001 versus FAK cells, †p<0.02 versus FAK-/- cells. (B) For migration assays, confluent cell monolayers were wounded and cells were allowed to migrate for 8 h. Images were taken at t = 0 and t = 8 h, and the distance covered by migrated cells was evaluated. (P<0.0001, F = 21.57; 3 independent experiments done in triplicates). ***p<0.0001 versus FAK cells, *p<0.02 versus FAK cells, ††p<0.005 versus FAK-/- cells.
Figure 6.
Cortactin and paxillin distributions in FAK-/-, wild-type FAK and FAKI936/I998 cells co-transfected with active Src.
(A) Fluorescence images taken at the base of fixed cells expressing wild-type or mutant GFP-FAK and mCherry-SrcY530F and immunostained for cortactin (upper panel) or paxillin (lower panel). Note the presence of cortactin clusters organized in rosette-like (arrowhead) or dot-like (arrow) structures characteristic of Src transformation in FAK cells and the absence of such structures in FAKI936/I998/SrcY530F cells. Note also the membrane localization of SrcY530F in FAKI936/I998 and FAK-/- cells (arrows). Scale bars, 5 μm. (B) Quantification of the percentage of cells having cortactin clusters at their base in the different conditions above. Note the reduced amount of cells with cortactin clusters in FAK-/- and FAKI936/I998 expressing cells. (P = 0.0003, F = 40.92; **p<0.003, ***p<0.002 versus FAK cells, 3 independent experiments with approximately 100 cells counted for each condition).
Figure 7.
Invasion of FAK-/-, wild-type FAK and FAKI936/I998 cells co-transfected with active Src.
Transfected cells were seeded in Matrigel-coated Boyden chambers and allowed to invade for 24 h. Invading cell nuclei were stained with crystal violet and ODs were measured at 595 nm. Bright field images were taken at 4× magnification. Scale bar, 100 μm. (P = 0.0030, F = 7.795; *p<0.03, **p<0.002 versus wild-type FAK cells from 3 independent experiments done in triplicates).
Figure 8.
P-Src distribution and interaction in FAK-/-, wild-type FAK and FAKI936/I998 cells.
(A) Confocal images of fixed FAK-/-cells expressing or not (NT, non-transfected cells) wild-type or mutant FAK-GFP (left panel) and immunostained for paxillin (middle panel) or Src phopsphorylated at tyr 416 (right panel). Note the high level of P-Src at FAs (arrows), identified by paxillin staining, in GFP-FAK cells (upper panel) and the lack of P-Src staining at FAs in FAKI936/I998-GFP cells (lower panel). Note also the low level of P-Src at FAs in non-transfected (NT) FAK-/- cells (arrowheads). Scale bar, 10 μm, insert scale bar, 5 μm. (B) Analysis of the localisation pattern of P-Src and paxillin in the different conditions above. FAs were first segmented and then subjected to co-localisation analysis. Co-localised pixels between paxillin staining (left) and P-Src staining (middle) appear in white on the colocalised image (right) Scale bar, 10 μm.Quantification using Pearson's coefficient (which describes the extent of overlap between image pairs) reveals a significant reduction of correlation in paxillin and P-Tyr416-Src images for both FAKI936/I998 and FAK-/- cells compared to FAK cells (P<0.0001, F = 39.25; ***p<0.0001, n = 3 independent experiments with 15 to 27 cells analyzed) and for FAKI936/I998 cells compared to FAK-/- cells (††p<0.005, n = 3 independent experiments with 21 to 27 cells analyzed). (C) Representative blots showing wild-type FAK and FAKI936/I998 immunoprecipitated using anti-FAK Ab and blotted for FAK, Src and phospho-Src. The expression level of proteins in the corresponding cell lysate is shown.
Figure 9.
p130Cas distribution in FAK-/-, wild-type FAK and FAKI936/I998 cells.
(A) Confocal images of fixed FAK-/-cells expressing or not (NT, non-transfected cells) wild-type or mutant FAK-GFP (left panel) and immunostained for p130Cas (right panel). Note the localization of p130Cas at FAs in both FAK-/-, FAKI936/I998 and FAK cells. Scale bar, 20 μm. (B) Confocal images of fixed FAK-/-cells expressing or not (NT, non-transfected cells) wild-type or mutant FAK-GFP (left panel) and immunostained for P-p130Cas (right panel). Note the high level of P-p130Cas at FAs in GFP-FAK cells (upper panel) and the lack of P-p130Cas staining at FAs in FAKI936/I998-GFP cells (lower panel). Scale bar, 20 μm. Quantification of the mean fluorescence intensity at FAs shows a significant reduction of P-p130Cas staining for both FAKI936/I998 compared to FAK and FAK-/- cells and for FAK-/- cells compared to FAK cells (P = 0.0005, F = 9.434; ***p<0.0005, *p<0.05 and †p<0.05 respectively, n = 3 independent experiments with 12 to 15 cells analyzed and more than 300 FAs analyzed for each condition).