Wnt-1 was found as an oncogene activated by the Mouse Mammary Tumor Virus in murine breast cancer (Nusse and Varmus, 1982). APC was first isolated as a tumor suppressor gene in human colon cancer (reviewed in Polakis 1997). After establishing that APC and b-catenin bind to each other (Su et al, 1993; Rubinfeld 1993) activating mutations in the human b-catenin gene were found in human colon cancer and melanomas (Morin et al, 1997) These mutations alter specific b-catenin residues important for GSK3 phosphorylation and stability (see table). The role for Frat/GBP in cancer is illustrated by its activation by proviral insertion in mouse lymphomas (Jonkers, 1997). A recent example of the link between cancer and Wnt signaling comes from the identification of the Drosophila gene legless as a homolog of Bcl-9, a gene implicated in B cell malignancies (Kramps, 2002)

Interestingly, mutations in the human AXIN1 gene were reported in human hepatocellular carcinomas (Satoh et al, 2000; reviewed by Clevers, 2000). These findings underscore the important predictive value of these pathways: by establishing how genes control signaling (e.g. whether they act as repressors or activators) one can make educated guesses how these gene might contribute to human cancer. A repressor can be a tumor suppressor gene and an activator a dominant oncogene. The same principle applies to other pathways, such as the Hedgehog-Patched pathway and the TGF beta pathway.

TCF1 can also act as a tumor suppressor gene (Roose, 1999), as Tcf1 mutant mice develop adenomas in the gut and mammary glands. Introduction of a mutant APC allele into these mice increases the number of tumors, suggesting that Tcf1 acts as a feedback repressor of target genes of the pathway.

Wnt5A, which can act to counteract signaling by other Wnts, appears to be able to act as a tumor suppressor, as mice that are mutant for Wnt5A develop lymphoid malignancies (Liang et al, 2003). There is also evidence that members of the SFRP family are epigenetically inactivated in colon cancer (Suzuki et al, 2004). Because the SFRPs may sequester and inactivate Wnt, loss of SFRP expression may contribute to Wnt activation in cancer.

There are several excellent reviews on Wnt signaling and cancer: Polakis, 2000 and Bienz and Clevers,