NEW YORK (GenomeWeb) – An international team led by investigators at the Fred Hutchinson Cancer Research Center has identified a combination of rare and common genetic variants coinciding with colorectal cancer (CRC) susceptibility.
The researchers started by doing whole-genome sequencing on more than 2,100 CRC cases and controls, using variants imputed from these and other data in a two-stage genome-wide association study that included nearly 125,500 individuals with or without CRC. The findings, appearing online today in Nature Genetics, revealed 40 previously undocumented CRC-associated variants and 55 variants implicated in CRC risk in prior studies.
The set included a rare, apparently protective variant in the intron of the CHD1 gene, the team noted, and highlighted roles for low-frequency variants, variants affecting immune function and signaling pathways, and somatic mutation drivers in CRC. Those involved suggested the collection of CRC contributors identified in the study and prior analyses may lead to new targets for preventing or treating cancer.
"To date, the search for new targets for cancer therapy has been limited to focus primarily on the molecular characteristics of cancer cells. We think there is a huge opportunity in using the GWAS approach to inform drug development for colorectal cancer," co-first author Jeroen Huyghe, a public health sciences researcher at the Fred Hutchinson Cancer Center, said in a statement.
Using Illumina HiSeq 2500 paired-end sequencing, the researchers did whole-genome sequencing on 1,439 individuals with CRC and 720 unaffected controls. Using sequence data covering each genome to 3.8- to 8.6-fold coverage, on average, and population reference sequence data from the Haplotype Reference Consortium, they got haplotype phasing information for 31.8 million variants, including rare variants.
With variants imputed from these data, the team did GWAS and GWAS meta-analyses that relied on array-based genotyping profiles for 34,869 CRC cases and 29,051 controls from the discovery phase of the study and a validation group that included 23,262 more individuals with CRC and 38,296 without. A subset of participants were genotyped using a custom Illumina array that included known cancer-related loci as well as suspicious variants found in the first phase of the association study.
From the GWAS meta-analysis and their subsequent conditional analysis, the researchers uncovered known CRC contributors, along with new associations involving common, low-frequency, and rare variants. They went on to delve into 40 new variants implicated in CRC with functional annotation, open chromatin maps, and chromatin accessibility profiling by ATAC-seq.
The team also came up with a polygenic score based on 95 new and previously known risk variants, though it cautioned that the polygenic risk score is expected to be less accurate in non-European populations due to the over-representation of European participants in the current study.
In a statement, co-first author Tabitha Harrison, a genetic epidemiologist at the Fred Hutchinson Cancer Center, noted that the team is "broadening the study population to include people from diverse ethnic backgrounds."
"This will give us a more complete understanding of risk across the entire population," Harrison explained.