NEW YORK (GenomeWeb News) – A Japanese team has published an integrated genomic and transcriptomic analysis focused on a form of kidney cancer called clear-cell renal cell carcinoma.
As they reported online in Nature Genetics last night, researchers from the University of Tokyo, Kyoto University, and other centers in Japan did whole-genome, whole-exome, and/or RNA sequencing on samples from more than 100 individuals with ccRCC. To that, they added information on gene expression patterns, copy number profiles, and methylation status in the tumors, assessed using arrays.
Together, the available data helped the team find new and known genetic and epigenetic glitches that are recurrent in ccRCC tumors. These included collections of coincident alterations in subsets of the tumors that seemed to coincide with better or worse patient outcomes.
"This integrated molecular analysis unmasked new correlations between DNA methylation, gene mutation, and/or gene expression and copy number profiles," corresponding author Seishi Ogawa, a cancer researcher affiliated with the University of Tokyo and Kyoto University, and his colleagues noted, "enabling the stratification of clinical risks for patients with ccRCC."
Past studies of ccRCC — itself a subtype of the renal cell carcinoma form of kidney cancer — suggest that these tumors are prone to problems involving a proteolytic pathway centered on the VHL gene.
Within the past year or two, teams based in the US and China have used whole-genome, whole-exome, and targeted sequencing to unearth recurrent mutations affecting chromatin-related genes in ccRCC and other renal cell carcinoma subtypes as well.
In an effort to characterize the full complement of mutations that can contribute to ccRCC, authors of the study set out to do an integrated analysis focused on a fairly large set of tumors from that RCC subtype.
To that end, they used Illumina's HiSeq 2000 instrument to sequence the whole genomes of matched tumor-normal samples from 14 ccRCC cases. Those samples — as well as tumor-normal pairs from 92 more individuals with ccRCC — were also subjected to whole-exome sequencing and RNA sequencing.
To complement that data, the group did deeper targeted sequencing and array-based methylation, copy number, and gene expression assessments on 240 matched tumor-normal samples — experiments performed with the Illumina MiSeq platform and microarrays from Illumina, Affymetrix, and Agilent, respectively.
As expected from past work on ccRCC, the team's analysis of the tumors turned up multiple mutations and genetic glitches that inactivate the VHL gene.
Moreover, the VHL gene and three other genes from the same chromosome 3 locus — SETD2, BAP1, and PBRM1 — made up four of the top five most recurrently mutated genes in ccRCC, the researchers reported.
By design, though, the cases selected for the discovery stage of the study included individuals whose tumors did not harbor any obvious VHL alterations. In several of those tumors, the team saw recurrent mutations to the elongin gene TCEB1, which tended to correspond with loss of heterozygosity involving a region on chromosome 8.
As it turned out, the sorts of changes detected in the TCEB1 gene were changes expected to interfere with interactions with VHL, altering the function of the VHL-containing pathway even in the absence of mutations to complex components themselves.
The researchers tallied up mutations to several other pathways that were quite common, too, including one signaling pathway that contains PI3-kinase, AKT, and mTOR gene products and another based around p53.
The available data also made it possible to look at epigenetic shifts, mutations, and gene expression patterns that frequently coincide with one another in the kidney cancer.
For instance, the team tracked down a subset of ccRCC tumors with higher-than-usual methylation, a preponderance of somatic mutations, and a tendency toward loss of heterozygosity in regions on chromosomes 9 and 14. Such methylation-related tumor subsets seemed to coincide with better or worse patient outcomes, researchers reported, as did mutations to certain genes such as BAP1.
"By integrating multiple layers of different comprehensive analyses, we unmasked unique correlations between somatic mutations, DNA methylation, gene expression, and copy number alterations," the study's authors concluded, "which were also linked to the clinical behaviors of tumors."