Genetic Variation in the APC gene associates with moderate APC allele-specific expression and is correlated with increased susceptibility to common forms of colorectal cancer

Although germline variation in allele-specific expression (ASE) has been shown to associate with highly penetrant familial cancers, its role in sporadic common cancers is less well-defined. ASE of the adenomatous polyposis coli (APC) gene plays a role in familial adenomatous polyposis (FAP), a familial colon cancer. We hypothesized that moderate ASE variation in the APC gene contributes to susceptibility to common forms of colorectal cancer (CRC). There were two main aims of this study: First, we investigated whether APC ASE was associated with sporadic cancer by performing germline analysis in CRC cases (n=53) and controls (n=68) using denaturing high performance liquid chromatography (DHPLC). We compared the means, medians, and variances of ASE between cases and controls. Our results showed that while there was no significant difference in average ASE values between cases and controls (mean p=0.57, median p=0.45), there were significant differences in the distribution of ASE values between cases and controls; cases had a significantly larger variance (p=0.0004). Further analyses confirmed that cases had a significantly higher proportion of individuals more than 1.0 (p=0.001, OR=3.97) or 1.645 (p=0.005, OR=13.46) standard deviations from the mean than controls, confirming that deviations of ASE values from the overall mean were associated with increased risk of CRC. Second, we investigated whether genetic variants in the APC gene were associated with its ASE. Twenty-three tag SNPs were genotyped in the APC gene, and eight passed quality control and were polymorphic. Four markers were significantly associated with APC ASE in the case/control analyses, rs41115 (additive model p=0.0091), rs41115 (dominant model p=0.0030), rs467033 (dominant model p=0.0427), rs971517 (additive model p=0.0010), and rs2431507 (p=0.0247). Three markers were significantly associated with APC ASE in case-only analyses, rs41115 (p=0.0058), rs41115 (p=0.0013), and rs971517 (p=0.0103). None of the markers showed significant association with APC ASE in controls. In conclusion, our study detected a wider distribution in APC ASE among CRC cases compared to CRC controls, found that CRC risk seemed to increase with the degree of ASE, and discovered genetic variants in APC that were associated with ASE. Our combined results support the conclusion that APC, a gene that predisposes to FAP, also plays an important role in sporadic cancer via ASE.