SFRP4 as an epigenetic biomarker of colorectal cancer risk

Abstract

There is a lack of robust biomarkers of CRC risk. Epigenetic changes in the WNT-related SFRP4, a gene whose expression is down-regulated early in CRC development, may be a potential CRC risk biomarker. If SFRP4 promoter methylation proved to be a useful biomarker of CRC risk, it would have potential implications for CRC screening. In addition, it could be used as a surrogate endpoint for investigations of CRC risk modifying interventions. SFRP4 methylation at several CpG sites was quantified in macroscopically normal rectal mucosal biopsies from volunteers at a relatively lower and higher CRC risk in two studies viz. the BORICC Study and the DISC Study. In the BORICC Study, the mean SFRP4 methylation of the 5 CpG sites investigated was significantly (p=0.036) higher in those in the higher risk group than in healthy controls. In the DISC Study, SFRP4 methylation was also higher at all CpG sites in the higher risk groups than in healthy controls but the differences were not statistically significant. In the BORICC Study SFRP4 methylation was also quantified in buccal cells matched to the rectal biopsies for the volunteers at a relatively lower and higher CRC risk. In contrast with the findings from the rectal mucosa, SFRP4 methylation was significantly (p<0.001) lower at all CpG sites in those in the higher risk group than in healthy controls. At CpG sites 1 and 4 only, SFRP4 methylation in the rectal biopsies and buccal cells was correlated significantly (p=0.001 and p=0.041 respectively). The healthy controls in the DISC Study were entered into a 50 day dietary intervention study and randomised to two potential chemoprevention agents; resistant starch and polydextrose in a 2  2 factorial design. SFRP4 methylation levels were quantified before and after the dietary intervention. Individually, resistant starch and polydextrose had no detectable effect on SFRP4 methylation levels. However, there was evidence of an interaction between the two intervention agents which was qualitatively similar at all CpG sites investigated. This interaction was statistically significant (p=0.008) at CpG site 2. The biological interpretation of this interaction cannot be determined until the study is unblinded. This study has provided preliminary evidence that SFRP4 methylation may be a novel epigenetic biomarker of CRC risk and that measurement in DNA from buccal cells may be a useful surrogate for invasive measurements on rectal mucosa.