NEW YORK (GenomeWeb News) – Retrotransposon movement in somatic cells may introduce insertions in the genome that contribute to tumorigenesis in certain cancer types, according to a study in the early, online edition of Science today by members of the Cancer Genome Atlas Research Network and their collaborators.
"Our analysis suggests that some [transposable element] insertions provide a selective advantage during tumorigenesis, rather than being merely passenger events that precede clonal expansion," co-corresponding authors Peter Park and Peter Kharchenko, both based at Harvard Medical School and Brigham and Women's Hospital, and their colleagues wrote. "We observe differential deregulation of TE activity across and within different cancer types."
Using their computational method, dubbed TE analyzer, or Tea, the investigators profiled transposable element insertion sites in whole-genome sequence data representing tumor and matched normal samples from 43 individuals with glioblastoma, multiple myeloma, colorectal cancer, prostate cancer, and ovarian cancer.
When they compared the jumping gene insertion patterns in tumor and normal samples and across the five cancer types, the researchers found that the epithelial cancers —colorectal, prostate, and ovarian cancer — contain somatic insertions that were often associated with highly mutated genes, genes with altered expression patterns, or parts of the genome where DNA methylation is reduced in cancer.
On the other hand, they did not find evidence of somatic retrotransposition in the blood cancer, multiple myeloma, or in the brain cancer, glioblastoma.
Although transposable elements are common in the genomes of humans and other mammals, their ability to hop around the genome in somatic cells is generally limited by epigenetic and other mechanisms, the study authors explained. Nevertheless, they noted that there are situations in which jumping genes can move around the genome, introducing insertion polymorphisms or, in some instances, disease-related mutations.
Consequently, researchers suspected that retrotransposons movement might be more pronounced than usual in cancers due to alterations affecting the cellular safeguards that normally prevent rampant TE activity in the genome.
To explore that possibility, researchers pulled in paired-end genome sequence data generated for matched tumor-normal samples from 43 individuals, including five with colorectal cancer, seven with prostate cancer, eight with ovarian cancer, seven with multiple myeloma, and 18 with glioblastoma.
They then characterized TE insertion positions and the mechanisms associated with them in the genomes using the newly developed Tea computational strategy, which involves aligning paired-end genome sequencing reads not only to the human reference genome, but also to a custom assembly representing canonical and divergent TE sequences.
Using this approach, the team tracked down 194 somatic insertions. Of these, 183 involved were so-called L1 retrotransposons, 10 were Alu element insertions, and one was an ERV insertion.
PCR-based validation testing on a subset of the insertions indicated that the Tea software identified authentic TE insertions with 97 percent accuracy, though the study's authors noted that there are likely insertions missed by the computational method, which does not pick up insertion events that occur in repeat-rich regions.
Researchers detected TE insertions in all of the colorectal, ovarian, and prostate cancer samples tested, though the incidence of these insertions varied from one tumor to the next within each cancer type. For instance, one colorectal cancer sample had an inordinate number of L1 insertion sites at 102, while the average number of these insertions across the other four colon cancer samples tested was nine.
On the other hand, the team did not find somatic L1 or Alu insertions in any of the blood or brain tumor samples. A multiple myeloma tumor did harbor an ERV insertion, but the pattern of this insertion suggests that it was introduced through the microhomology-mediated break-induced repair mechanism rather than traditional retrotransposition.
In the cancers containing somatic L1 or Alu insertions, insertions tended to fall in the introns or untranslated portions of genes with high overall mutation rates. They also tended to coincide with shifts in gene expression, researchers explained, noting that somatic TE insertion in a gene most often corresponded with reduced expression.
Data for the five colon cancer genomes and on another 228 colon cancer exomes indicated that somatic L1 insertions are also enriched in parts of the genome that are known to have lower-than-usual methylation in these cancers, researchers reported.
Meanwhile, researchers' analyses of insertion polymorphisms in 44 normal genomes indicated that these germline TE insertions — which stem from retrotransposition events in gametes or during early embryonic development — differ from insertions in cancer genomes, both in their distribution and in their epigenetic context.
"Although a more extensive panel of matched genomic and epigenetic data is needed to investigate the functional impact of retrotransposition events and the pathways involved," the study's authors concluded, "our analysis reveals the extent of TE insertions in human tumors and lays the foundation for determining the role of these events in human neoplasia."
The Tea software that researchers used for analyzing transposable element insertion patterns in the current study is available through a website developed by Park's computational genomics lab.