Researchers in the lab of Trey Ideker, Ph.D., at University of California, San Diego recently published results of a genome-wide analysis comparing paired normal and tumor tissues that led to the identification of mutations in noncoding regions linked to changes in target gene expression. Although previous whole-genome sequencing (WGS) studies of cancer tissue have uncovered recurrent somatic mutations at noncoding loci, there exists little evidence of functional associations between these mutations and regulatory control of particular genes. One clear exception are mutations that occur in the promoter of the telomerase reverse transcriptase (TERT) gene.
The current study was able to make important associations between noncoding mutations and gene expression because the method Ideker’s team used to identify somatic mutations in noncoding regions was fundamentally different than the approach previously taken. As Wei Zhang, Ph.D., lead author of the article in Nature Genetics and a postdoc in Ideker’s lab, explained, “they focused on mutations that are highly recurrent, beyond the background mutation rate.”
In contrast, “our analysis focused on mutations associated with changes in gene expression.” The new findings provide fascinating insights into the potential implications of functional somatic mutations in noncoding regions of the cancer genome. Using specific genes as examples, Zhang, et al. began to examine the link and possible causal relationship between a specific mutated noncoding locus, a change in target gene expression, and a cancer phenotype. As researchers begin to explore the significance of somatic mutations in noncoding regions of the cancer genome and learn if and at what point in the cancer life cycle they may exert an effect—from tumorigenesis to metastasis—then at least some of these recurrently mutated loci may one day have a role in advancing cancer diagnostics or therapeutics.
Identifying Somatic eQTLs
The study by Zhang, et al. involved WGS analysis with matched mRNA expression profiles of 930 tumor-normal tissue pairs representing 22 cancer types. These samples were acquired from The Cancer Genome Atlas (TCGA). The researchers searched for single nucleotide variations (SNVs) in noncoding regions. “Instead of looking at the entire genome, we focused on regions that are known to have a regulatory impact and where there are recurrent mutations,” said coauthor Jason Kreisberg, Ph.D. Recurrent SNVs that occurred within 50 bp of each other were grouped into clusters.
Click here to access the rest of this article.