Abstract
There have been conflicting reports on the requirement of GSK-3beta-mediated phosphorylation of the tumor suppressor adenomatous polyposis coli (APC) vis-à-vis its ability to bind and degrade beta-catenin. Using a unique combination of loss of function for Shaggy/GSK-3beta and a gain of function for human APC in Drosophila, we show that misexpressed human APC (hAPC) can still sequester Armadillo/beta-catenin. In addition, human APC could suppress gain of Wnt/Wingless phenotypes associated with loss of Shaggy/GSK-3beta activity, suggesting that sequestered Armadillo/beta-catenin is non-functional. Based on these studies, we propose that binding per se of beta-catenin by APC does not require phosphorylation by GSK-3beta.
