Fig. 3. The methyl-CpG binding function of kaiso is required to rescue the xKaiso knockdown phenotype. The percentages of presented phenotypes are indicated. (A) Embryos injected with 10 ng of CMO develop normally by tadpole stage. (B) Over-expression of 750 pg wild-type human kaiso RNA does not affect normal development of Xenopus embryos. (C) Injection of 750 pg of wild-type human kaiso RNA together with 5 ng KMO leads to complete rescue of 26% of the embryos. (D) Apoptotic phenotype produced by injection of 5 ng KMO. (E) Co-injection of 750 pg of C522R human kaiso mutant RNA with 5 ng KMO cannot rescue the phenotype of xKaiso depletion. (F) Location of the kaiso mutant C522R amino acid substitution (red) in the third zinc finger, leading to loss of the ability to bind methylated DNA. (G) Protein gel of recombinant wild-type human kaiso and C522R (C>R) mutant proteins (arrow). (H) Pull-down experiment showing p120^ctn (arrow) binds both wild-type human kaiso and C522R (C>R) mutant proteins in vitro, but not human kaiso protein lacking the ZF domain (ZF). (I) EMSA experiment using recombinant C522R (C>R) mutant and wild-type human kaiso proteins with methylated (lanes 1, 4), non-methylated Sm oligos (lanes 2, 5) and human matrilysin (Hmat) oligo (lanes 3, 6) as probes. The kaiso-specific band shift is arrowed. The C522R mutant shows no DNA-binding activity.