NEW YORK (GenomeWeb) – An international team led by investigators at St. Jude Children's Research Hospital has uncovered inherited variants at a locus near the tumor suppressor coding gene CDKN2A that contribute to the risk of developing acute lymphoblastic leukemia during childhood.
As they reported yesterday in Nature Communications, the researchers did array-based genotyping on thousands of children with or without B-cell ALL, focusing on coding portions of the genome. Their search led to a missense variant in CDKN2A, which appears to curb activity of a tumor suppressor protein called p16-INK4A.
The team's subsequent screening of more than 2,400 children with ALL led to still more functionally relevant variants in and around the same locus, while mouse cell line experiments supported the notion that alterations in p16-INK4A can dial up the risk of leukemic transformation.
"These results provide direct functional evidence for the influence of inherited genetic variation on ALL risk, highlighting the important and complex roles of CDKN2A-CDKN2B tumor suppressors in leukemogenesis," corresponding author Jun Yang, a pharmaceutical sciences and hematological malignancies researcher at St. Jude, and his colleagues wrote.
The CDKN2A-CDKN2B locus on chromosome 9 encodes no less than three tumor suppressor proteins: p16-INK4A, p14-ARF, and p15-INK4B. It is also tends to get deleted in tumors from children with B-ALL or T-cell ALL and is one of half a dozen loci that have been implicated in ALL risk through past GWAS, the team noted.
Because most ALL risk variants described previously have turned up in sequences between genes or in intronic sequences in genes that don't code for proteins, the researchers decided to focus their new analysis on protein-coding portions of the genomes.
Using the Illumina Infinium HumanExome array, the team genotyped 1,773 children of European descent with B-ALL at almost 250,000 SNPs.
When they compared the genotyping patterns to those in 10,448 children from the same population without ALL, the researchers found three sites with significant ties to ALL risk: non-coding variants in ARID5B and IKZF1 — genes implicated in ALL risk in the past — a coding SNP at the CDKN2A-CDKN2B locus.
They verified the association with the latter SNP, known as rs3731249, through testing on a validation cohort comprised of 409 Danish children with ALL and nearly 1,600 without.
The rs3731249 SNP falls in both an untranslated region of p14-ARF and a coding sequence for p16-INK4A, the study's authors noted. But they suspected the latter change — a missense mutation — was likely behind the newly detected association.
Indeed, the team's mouse progenitor cell line experiments indicated that elevated expression of p16-INK4A proteins produced from CDKN2A sequences containing the rs3731249 variant were more apt to transition to leukemia when the classic BCR-ABL1 gene fusion was present. In contrast, higher-than-usual levels of wild type p16-INK4A appeared to slow this leukemic transformation.
The team track down 19 more variants in the CDKN2A-CDKN2B region that are suspected of having functional effects when it did targeted Illumina HiSeq 2000 sequencing of the locus in germline DNA samples for 2,407 children with ALL, including 1,450 cases from the first phase of the GWAS.
"These results provided functional evidence for the influence of inherited genetic variants on ALL leukemogenesis," Yang and co-authors concluded, "further indicating that a continuum of genetic variations in both host and tumor genomes contribute to malignant transformation and cancer risk."