NEW YORK (GenomeWeb) – A study published online this week in Nature Communications suggests variants at four previously undetected sites in the genome can heighten susceptibility to basal cell carcinoma.
Using variant data gleaned from thousands of recently sequenced Icelandic genomes, researchers from Amgen subsidiary Decode Genetics, the University of Iceland, and elsewhere imputed patterns for nearly 25 million variants in the genomes of more than 270,000 individuals with or without basal cell carcinoma from the same population.
"The Icelandic Cancer Registry has recorded histopathologically confirmed diagnoses of [basal cell carcinoma] since 1981, making studies using population-based case ascertainment possible," senior author Kari Stefansson, Decode's founder and CEO and a researcher at the University of Iceland, and his colleagues wrote.
The genome-wide association study uncovered variants with ties to basal cell carcinoma at four new loci on chromosomes 2, 8, and 10 — associations that the team verified through additional testing on thousands of cases and controls from non-Icelandic populations.
Although sun exposure is a known risk factor for basal cell carcinoma and other forms of skin cancer, the researchers noted, susceptibility to the disease appears to have a heritable component as well.
Past association and candidate gene studies have unearthed risk variants in some of the same genes that influence pigmentation in individuals of European ancestry, for example, though genes with no known ties to pigmentation have also been implicated in basal cell carcinoma.
To search for still more genetic factors behind basal cell carcinoma risk, Stefansson and his colleagues turned to population genetic information in the whole-genome sequences of 2,636 Icelandic individuals, focusing on more than 19.5 million SNPs and nearly 5.5 million small insertions and deletions.
These variants were then imputed in the genomes of basal cell carcinoma cases and controls who had either been tested using arrays to pick up key genotyping and long-range phasing patterns, or who were closely related to these chip-typed individuals.
Using this approach, the team ultimately imputed variant patterns in 4,752 individuals with basal cell carcinoma from the Icelandic Cancer Registry and another 266,358 unaffected controls from the same population.
A comparison of SNP and indel profiles in those with the disease and those without led the researchers to four loci: one chromosome 2 site near the oncogene MYCN, and another near the CASP8-ALS2CR12 locus, along with variants in or around the chromosome 8 gene ZFHX4 and GATA3 on chromosome 10.
The researchers went on to verify these associations through testing on another 1,475 individuals with basal cell carcinoma from Spain, Eastern Europe, or Denmark and 4,733 controls from these populations.
Their follow-up fine mapping analyses — again helped along by the available Icelandic genome sequences — suggested a basal cell carcinoma-linked variant on chromosome 10 is in linkage disequilibrium with two other variants that turn up at a conserved GATA3 transcription factor binding site.
When the team used array data to look for expression shifts coinciding with the skin cancer susceptibility variants, meanwhile, it found signs of altered CASP8 splicing in the presence of basal cell carcinoma-linked SNPs — results that were further fleshed out through RNA sequencing on blood samples from 261 genotyped individuals and analyses of skin expression data generated for the GTEx project.
"While the mechanisms at two of the loci remain cryptic," the study's authors concluded, "we provide testable hypotheses for roles of CASP8 and GATA3 variants in epidermal carcinogenesis."