NEW YORK (GenomeWeb) – In a study published online today in Nature Communications, researchers from the Institute of Cancer Research, London and elsewhere describe rare, disruptive colorectal cancer (CRC)-related risk mutations in new and known risk genes, which were found by sequencing the exomes of more than 1,000 individuals with familial forms of the disease.
The team believes the analysis was sufficiently powered to pick up all of the major CRC risk variants. Still, since the rare variants found so far accounted for between 15 and 31 percent of the familial cases considered, the group cautioned that future studies involving at least 10,000 affected individuals will be needed to unearth variants with smaller effects on CRC predisposition.
"Each cancer gene that has been discovered, or common genetic variant that we will continue to uncover, provides us with new insights into the underlying biology of the disease, and increases our ability to assess people for their risk," senior author Richard Houlston, a molecular and population genetics researcher at ICR London, said in a statement.
Houlston and his colleagues sifted through tens of thousands of CRC cases in an ICR repository, focusing on 1,028 individuals with CRC who had early onset disease and came from families that were especially prone to the disease.
After capturing protein-coding portions of each genome with Illumina TruSeq exon capture kits, the team used the HiSeq 2000 to sequence these exomes. The same approach was used to do exome sequencing on 1,644 unaffected controls. Sequences for 57 individuals were excluded from the team's subsequent analysis owing to low data quality or because individuals had non-European ancestry that might make such comparisons difficult.
When they compared mutations patterns in the remaining 1,006 cases and 1,609 controls — adding in Illumina HumanExome array data for another 5,552 individuals with CRC and nearly 6,800 without — the researchers did not find that any one moderate frequency variant showed statistically significant ties to CRC risk.
On the other hand, their gene-based search for rarer variants revealed enrichment of nonsense and frame-shift mutations in known CRC risk genes such as MSH2, MLH1, and APC. Fewer 'predicted damaging' mutations turned up in those genes and non-synonymous changes were not significantly enriched in other CRC contributors such as MSH6 and PMS2.
In an effort to narrow in on previously undetected susceptibility genes, the team focused on exome sequences from the 863 cases missing high-penetrance mutations in CRC risk genes. For that analysis, it added in whole-genome sequence data on another 188 familial CRC cases and nearly 3,600 controls from the UK10K control dataset.
The search led to recurrent variants in three suspicious new genes: IL12RB1, LIMK2, and POLE2. Disruptive POLE2 also turned up in the researchers' subsequent gene-set enrichment analyses, as did disruptive changes to the telomere length-regulating genes POT1 and MRE11A.
Mutations in latter gene, which is also known for its role in double-strand DNA break repair, was detected in three of the CRC cases, the team noted. And one of the alterations overlapped with a mutation reported in ovarian cancer. Likewise, the POT1 gene has been implicated in both familial glioma familial melanoma, and other cancer susceptibility conditions.
In their search for non-dominant mutations linked to CRC risk, the researchers identified one case marked by recessive, compound heterozygous mutations in NTHL1, a gene involved in base-excision DNA repair, while their common variant analysis revealed apparent associations with the ATF1 gene.
"Our study clarifies the genetic architecture of CRC and probably discounts the existence of further major high-penetrance susceptibility genes, which individually account for [less than 1percent] of the familial risk," the authors wrote.