In most colorectal cancers, Wnt/b-catenin signaling is activated by loss-of-function mutations in the adenomatous polyposis coli (APC) gene and plays a critical role in tumorigenesis. Tankyrases poly(ADP-ribosyl)ate and destabilize Axins, a negative regulator of b-catenin, and upregulate b-catenin signaling. Tankyrase inhibitors downregulate b-catenin and are expected to be promising therapeutics for colorectal cancer. However, colorectal cancer cells are not always sensitive to tankyrase inhibitors, and predictive biomarkers for the drug sensitivity remain elusive. Here we demonstrate that the short-form APC mutations predict the sensitivity of colorectal cancer cells to tankyrase inhibitors. By using well-established colorectal cancer cell lines, we found that tankyrase inhibitors downregulated b-catenin in the drug-sensitive, but not resistant, colorectal cancer cells. The drug-sensitive cells showed higher Tcf/LEF transcriptional activity than the resistant cells and possessed "short" truncated APCs lacking all seven b-catenin-binding 20-amino acid repeats (20-AARs). In contrast, the drug-resistant cells possessed "long"APCretaining two or more 20-AARs. Knockdown of the long APCs with two 20-AARs increased b-catenin, Tcf/LEF transcriptional activity and its target gene AXIN2 expression. Under these conditions, tankyrase inhibitors were able to downregulate b-catenin in the resistant cells. These results indicate that the long APCs are hypomorphic mutants, whereas they exert a dominant-negative effect on Axin-dependent b-catenin degradation caused by tankyrase inhibitors. Finally, we established 16 patient-derived colorectal cancer cells and confirmed that the tankyrase inhibitor-responsive cells harbor the short-form APC mutations. These observations exemplify the predictive importance of APC mutations, the most common genetic alteration in colorectal cancers, for molecular targeted therapeutics.
ASJC Scopus subject areas