Embryonic axis induction by the armadillo repeat domain of beta-catenin: evidence for intracellular signaling.

beta-catenin was identified as a cytoplasmic cadherin-associated protein required for cadherin adhesive function (Nagafuchi, A., and M. Takeichi. 1989. Cell Regul. 1:37-44; Ozawa, M., H. Baribault, and R. Kemler. 1989. EMBO [Eur. Mol. Biol. Organ.] J. 8:1711-1717). Subsequently, it was found to be the vertebrate homologue of the Drosophila segment polarity gene product Armadillo (McCrea, P. D., C. W. Turck, and B. Gumbiner. 1991. Science [Wash. DC]. 254:1359-1361; Peifer, M., and E. Wieschaus. 1990. Cell. 63:1167-1178). Also, antibody perturbation experiments implicated beta-catenin in axial patterning of the early Xenopus embryo (McCrea, P. D., W. M. Brieher, and B. M. Gumbiner. 1993. J. Cell Biol. 123:477-484). Here we report that overexpression of beta-catenin in the ventral side of the early Xenopus embryo, by injection of synthetic beta-catenin mRNA, induces the formation of a complete secondary body axis. Furthermore, an analysis of beta-catenin deletion constructs demonstrates that the internal armadillo repeat region is both necessary and sufficient to induce axis duplication. This region interacts with C-cadherin and with the APC tumor suppressor protein, but not with alpha-catenin, that requires the amino-terminal region of beta-catenin to bind to the complex. Since alpha-catenin is required for cadherin-mediated adhesion, the armadillo repeat region alone probably cannot promote cell adhesion, making it unlikely that beta-catenin induces axis duplication by increasing cell adhesion. We propose, rather, that beta-catenin acts in this circumstance as an intracellular signaling molecule. Subcellular fractionation demonstrated that all of the beta-catenin constructs that contain the armadillo repeat domain were present in both the soluble cytosolic and the membrane fraction. Immunofluorescence staining confirmed the plasma membrane and cytoplasmic localization of the constructs containing the armadillo repeat region, but revealed that they also accumulate in the nucleus, especially the construct containing only the armadillo repeat domain. These findings and the beta-catenin protein interaction data offer several intriguing possibilities for the site of action or the protein targets of beta-catenin signaling activity.