Regulation by protein degradation has become a recurring theme in cancer biology: several oncoproteins, including β-catenin, are controlled by being targeted for degradation, and proteasome inhibitors have even made it into clinical trials as anticancer drugs. But biological control is seldom as simple as it seems. In the January issue of EMBO Reports, Frank Staal and colleagues reveal that phosphorylation, not degradation, might be the key to regulating β-catenin's activity.

β-catenin leads a double life: it's a component of epithelial junctions, but also a co-activator of the TCF family of transcription factors. The β-catenin-mediated induction of TCF's targets stimulates cell proliferation, and the deregulation of this process is an important early step in colorectal tumorigenesis. β-catenin levels are normally kept in check by a 'destruction complex' containing the APC tumour-suppressor protein and glycogen synthase kinase 3β (GSK3β). The accepted view is that phosphorylation of β-catenin by GSK3β allows it to be recognized by a second complex that conjugates β-catenin with ubiquitin — the molecular address tag that sends proteins to their destruction in the proteasome. The physiological switch that blocks β-catenin's destabilization is the extracellar signalling molecule Wnt, which indirectly prevents β-catenin from being phosphorylated by GSK3β.

But is simply increasing the level of β-catenin sufficient to drive transcription of TCF target genes? To find out, the authors blocked β-catenin degradation using a proteasome inhibitor and measured the transcription of TCF-target genes using a reporter construct. To their surprise, stabilizing β-catenin didn't increase transcription of the reporter gene. Experiments in a temperature-sensitive cell line that cannot add ubiquitin to proteins confirmed these results.

All the known oncogenic mutants of β-catenin have mutations in sites that are thought to be phosphorylated by GSK3β, so the authors reasoned that dephosphorylation of β-catenin might be needed not just to protect it from destruction, but also to activate its function as a transcriptional co-activator. To investigate this, the authors measured levels of dephosphorylated β-catenin using an antibody that recognizes only dephospho-β-catenin, under four different conditions. Wnt, transfection with a dominant-positive β-catenin mutant, and the GSK3β inhibitor LiCl increased both the levels of dephospho-β-catenin and TCF-reporter transcription, whereas treatment with a proteasome inhibitor had no effect on either, indicating that dephosphorylation of β-catenin is required for activation of β-catenin as a transcription factor. Immunofluorescence revealed that most of the dephospho-β-catenin was nuclear.

These findings are also supported by previously published clinical data: patients whose tumours have high levels of phosphorylated β-catenin have a better prognosis than those with low levels. This extra level of β-catenin regulation might also turn out to be therapeutically useful because it offers hope of blocking β-catenin's transcriptional function even in the absence of active APC.