Figure 2. The talin head is essential for integrin function in Drosophila.

(a–c) Maternal-zygotic talin null embryos (shown in a) rescued with either full-length WT talinGFP transgene (b) or headless talinGFP transgene (c) and stained for F-actin (green) and integrin (magenta). (d–f) Integrin-dependent phenotypes germband retraction (d), dorsal closure (e) and muscle attachment (f) were assayed in talin-null embryos, WT-talin-rescued embryos, and headless-talin-rescued embryos. The talin head was required for all three processess assayed. Scale bar = 100 µm. (g–h) Recruitment of ubi-promoter driven, GFP-tagged full-length WT talin and headless talin to sites of adhesion was assayed in wild-type embryos (g) and in a talin null background (h). In a WT background, headless-talinGFP competed less well with endogenous talin and was only weakly recruited to sites of adhesion compared to WT (***p<0.001); in the absence of endogenous talin, headless-talin was well recruited to sites of adhesion. (i) FRAP experiments on talinGFP-WT and headless-talinGFP reveal that headless talin is much less stable at sites of adhesion than talinGFP-WT. (j–m) Confocal z-stacks of stage 17 maternal/zygotic-mutant embryos rescued with either full length WT talin (j,l) or headless-talin (k,m). (j–k) adhesions stained for talin (green in j–k; grey in j′–k′) and integrin (red in j–k). In the absence of the talin head, talin was still well recruited. (l–m) Muscles stained for talin (magenta in l–m) and paxillin (green in l–m; grey in l′–m′). Paxillin was not well recruited to adhesions. Scale bars: a–e = 100 µm; j–m = 20 µm.