The Identification of Colorectal Cancer Susceptibility Genes Using a Cross-Species, Systems Genetics Approach.

摘要

Colorectal cancer (CRC) is the third most commonly diagnosed and third leading cause of cancer-related death in the United States. As much as 35% of risk for developing this disease is due to genetic risk factors. Genome-wide association studies (GWAS) have identified ~40 independent common risk variants that contribute to genetic predisposition. However, these variants fail to explain the majority of the genetic component for risk. Identification of additional susceptibility alleles for CRC is critical for developing genetic screening tools capable of predicting individuals at heightened genetic risk. This dissertation describes three projects aimed at uncovering CRC susceptibility genes using a cross-species systems genetics approach.;The first project (Chapter 2) involves assessing single nucleotide polymorphisms (SNPs) at loci that were linked to CRC risk by GWAS for allele-specific somatic copy number gains or losses in human colorectal tumor genomes. We hypothesized that GWAS-identified SNPs exhibit allele-specific copy number changes (termed 'allele-specific imbalance,' or ASI) in the tumor genome of patient specimens in much the same fashion as oncogenes are amplified and tumor suppressors are lost. We tested this hypothesis using quantitative genotyping to detect relative gains or losses of GWAS-identified SNP alleles in a cohort of human paired colorectal tumor/normal DNA samples. Testing of 17 SNPs revealed statistically significant allele-specific copy number changes at one SNP, rs6983267 at 8q24, which suggests that ASI occurs at CRC risk loci but perhaps at low frequency.;The second project discussed in this document (Chapter 3) describes a large ASI screen of SNPs tagging for candidate genes at the human orthologs of three murine CRC susceptibility quantitative trait loci (QTL). We prioritized genes for inclusion in this study based on RNA-seq data generated from the normal colons of the CRC-resistant and CRC-susceptible mouse parental strains that were used to linkage-map these QTLs. Because no protein-damaging coding SNPs were identified by RNA-seq, we focused largely on genes showing differential expression between the strains and/or genes with a documented role in cancer-relevant pathways or processes. Two SNPs in SNX10 emerged from our screen and warrant further investigation.;The third project (Chapter 4) reports our preliminary functional investigations of the hypoxia-inducible factor Epas1, whose gene maps to the Scc4 susceptibility QTL. This gene is differentially expressed in the colons of the CRC-resistant and CRC-susceptible mouse strains used to map Scc4. Furthermore, five tagging SNPs within EPAS1 show suggestive evidence of ASI in human colorectal tumor/normal DNA pairs, but were not statistically significant after correction for multiple comparisons. These data prompted us to investigate EPAS1/Epas1 for functional effects in human and mouse colon cell lines. We tested this gene for effects on beta-catenin/TCF-mediated transcriptional activity under conditions of normoxia and hypoxia in SW480 and HCT116 human colon adenocarcinoma cell lines. Under hypoxic culture conditions, we observed activation of our TOPflash beta-catenin/TCF reporter construct when Epas1 expression was enforced, suggesting a plausible role for EPAS1 in modifying CRC susceptibility. In summary, our data validate our cross-species approach as an innovative strategy for uncovering novel candidate CRC susceptibility genes.