Fig. 4

Abnormal distribution and glycosylation of CMAC components in MZtmem2 mutants. (A-G) Immunofluorescence shows localization of paxillin (green; A,B), pFAK (pY³⁹⁷) (red; C,D), βDG (blue; E-G), and glycosylated αDG (green; E-G) relative to muscle fibers, marked with Phalloidin (red; A,B,E-G) or the antibody F310 (green; C,D); lateral views, dorsal up, at 26hpf. (A-D) Both paxillin (B′) and pFAK (D′) are recruited to the MZtmem2 MTJ. However, compared with the concentrated and robust localization in Mtmem2 siblings (A′,C′), paxillin appears disorganized (B′), and pFAK levels are diminished (D′). (E-G) The antibody IIH6, which recognizes a glycosylated epitope of αDG within its laminin-binding site (Ervasti and Campbell, 1993), detects αDG at the MTJ in Mtmem2 siblings (E′), but detects only trace amounts of glycosylated αDG at the MZtmem2 MTJ (F′, Table S3). In contrast, βDG is readily detectable at the MZtmem2 MTJ (F′′). αDG glycosylation can be rescued in MZtmem2 mutants by injection of tmem2 mRNA (G′, Table S3). (H) Bar graph compares immunostaining intensity for glycosylated αDG, relative to levels of βDG, at the MTJ at 26hpf; error bars indicate s.e.m. For each condition, we measured the mean pixel intensity (MPI) of immunostaining at five different MTJs in each of three representative embryos; see also Fig. S5. Introduction of full-length Tmem2, but not the Tmem2 ectodomain, caused improvement in αDG glycosylation in MZtmem2 mutants. Asterisk indicates significant difference from MZtmem2 (Student′s t-test; P<0.005).