Fig 2: Specificity of the Effect of TAT-Cx43266-283 on FAK-Dependent GSC MotilityPrimary G13 (A and B) or G16 (C and D) GSCs were incubated with 50 µM TAT, TAT-Cx43274-291, TAT-Cx43266-283, 5 µM FAK inhibitor 14 (FAK inh 14), or 5 µM FAK inh 14 + TAT-Cx43266-283, and their random movements were recorded by time-lapse microscopy for 14 hr. Track-plots showing the trajectories described by at least seven representative G13 (A) or G16 (C) GSCs for 14 hr. The origin of each cell’s trajectory is at the intersection of the x and y axes. Scale bars, 50 µm. The migration length was extracted from the track-plots using the total path length described by each individual cell in at least ten movies from three independent experiments. The line bisecting the plot corresponds to the mean (ANOVA; ***p < 0.001, *p < 0.05; versus control; +++p < 0.001; versus TAT; ###p < 0.001, ##p < 0.01; versus TAT-Cx43274-291; @p < 0.05; FAK inh 14 versus FAK inh 14 + TAT-Cx43266-283; ns, not significant). Non-significant differences were found when the control, 50 µM TAT and 50 µM TAT-Cx43274-291 were compared. The trajectories of cells that died during the time lapse were not considered.

Fig 3: The FERM domain leads to activation of endogenous FAK in a tyrosine 397 dependent manner.HA-FERM and HA-FERM Y397F injected embryos in one blastomere at the animal pole of two cell stage embryos were processed for whole mount immunostaining using an HA antibody (green) to reveal expressing cells and the respective phospho-specific antibodies (red) as indicated. In each case individual signals for each secondary are shown in addition to a merged image and finally an intensity color coded image of the respective phospho-specific antibody signal. HA-FERM and HA-FERM Y397F injected cells are indicated with red stars and un-injected cells with white stars. (A–D) Levels of phosphorylated tyrosine 576 are elevated in HA-FERM overexpressing cells compared to controls. (E–H) Overexpression of HA-FERM Y397F has no effect on the endogenous levels of phosphorylated tyrosine 576. HA-FERM Y397F expressing cells have the same levels of phosphorylated endogenous FAK on tyrosine 576 with neighboring control cells. (I–L) Levels of phosphorylated tyrosine 861 are elevated in HA-FERM expressing cells compared to controls. (M–P) Overexpression of HA-FERM Y397F has no effect on the endogenous levels of phosphorylated tyrosine 861. HA-FERM Y397F expressing cells have the same levels of phosphorylated endogenous FAK on tyrosine 861 with neighboring control cells. (Q) Total lysates from HA-FERM injected gastrula stage embryos contain comparable levels of endogenous FAK as un-injected controls but elevated levels of phosphorylated FAK on tyrosines 397, 576 and 861. Blotting using the anti-P-Y397 antibody shows that the exogenously expressed FERM is heavily in trans phosphorylated on tyrosine 397 (2nd row). The intensity values from the densitometry analysis were normalized against total FAK and present the average increase in phosphorylation from three independent experiments (R–U) Confocal images of A6 Xenopus cells transfected with HA-FERM. Cells were stained with anti-HA (R) and anti-P-Y576 (S). T is the merged image and U an intensity color coded image of the anti-P-Y576 signal. Transfected cells are shown with red stars and controls with white stars. HA-FERM transfected cells show reduced levels of tyrosine 576 phosphorylation suggesting that FERM expression blocks FAK activation in these cells. Scale bars: (A) 40 µm, (E) 30 µm, (I) 20 µm, (M) 50 µm, (R) 20 µm.

Fig 4: FAK is heavily phosphorylated in mesodermal tissues and integrin-free regions of cells.(A) Intensity color coded confocal section of the dorsal lip region from a whole mount immunostained gastrula stage embryo using a P-Y576 FAK antibody. Mesodermal cells (white arrow) display much higher levels of phospho-FAK than endodermal cells lining the forming archenteron (white arrowheads) and the endodermal cells of the blastopore (red arrow). (B) Same as A but showing the anterior mesendoderm and the animal cap from a whole mount immunostained gastrula stage embryo. The superficial cells of the animal cap (white arrowhead) show lower levels of phospho-FAK signal compared to deep cells (white arrow) and mesendodermal cells (red arrow). (C) High magnification color coded narrow optical section of superficial cells of the animal cap reveals that the apical surface of these cells display similar levels of phospho-FAK compared to the basolateral region while (D) the apical region of the deep cells of the animal cap facing the fibronectin ECM display significantly elevated levels of phospho-FAK compared to the basolateral region. In addition, in the deep cells of the ectoderm labeling of phospho-FAK in the basolateral region is relatively uniform but the apical region displays distinct foci of higher signal intensity (E–G) Confocal optical sections from whole mount immunostained embryos using integrin-ß1 (green) and P-Y397 FAK antibodies (red). Integrin-ß1 and P-Y397 FAK colocalize on the plasma membrane at cell–cell boundaries (white arrowhead). However phosphorylated FAK is also present on the apical region of the outermost cells of the embryo where integrin-ß1 is absent (red arrow). (H–J) Same embryo as above but the cells facing the blastocoel cavity are imaged. Integrin-ß1 and P-Y397 FAK colocalize on the apical surface of the cells facing the blastocoel (white arrows). Scale bars: (A) 100 µm, (B) 50 µm, (C, D) 10 µm, (E) 30 µm, (H) 20 µm.

Fig 5: Protein expression analysis of NISCH interacting partners and downstream pathways in middle ear by immunohistochemistry.Middle ear sections of Nisch+/+, Nischedsn/edsn and Itga5tm1Hyn/+; Nischedsn/edsn mice at 3 wk, stained with (A) NISCH, (B) ITGA5, (C) p-PAK1, (D) p-LIMK1/2, (E) RAC1, (F) NF-?B p65, (G) FAK and (H) p-SMAD2 antibodies. To quantify the results middle ear epithelial cells in wild-types and mutants were counted in four middle ears from each genotype. Scale bar = 100 µm. ns P > 0.05; * P < 0.05; ** P < 0.01; *** P < 0.001. Error bars indicate standard error of mean. The data was analysed by one-way ANOVAs and Holm-Sidak’s multiple comparison procedures for post-hoc testing.