Whole-Exome Sequencing
Research Models
Regina Kelder

Whole-Exome Sequencing

Mouse-specific bioinformatics tools identify somatic mutations in mice. Our ongoing coverage of AACR 2017

Cancer immunotherapy shows us that we have the biological systems and tools to fight back. So why does the cancer win so much of the time? For every person who responds well to a checkpoint inhibitor, four others do not.

Success might be found by deciphering the tumor’s playbook, a task made easier by high throughput sequencing technologies like whole-exome sequencing (WES), a technique for sequencing all of the exonic regions in a genome. Exome sequencing has already revealed novel disease associations in humans, and there are even a handful of companies offering otherwise healthy people the chance to have their exomes sequenced. You can also analyze the entire exome of cancer patients’ tumors to pick up variations in potentially important genes.

This same approach is being used in modeling cancer in mice. A poster presented on Monday at the American Association of Cancer Research meeting in Washington, D.C. used WES to analyze the mutational profiles of 14 different mouse cancer models—ranging from non-small cell lung cancer to melanoma, colon and breast cancer. Four were genetically engineered mouse (GEM) models including the KP model, the other 10 were syngeneic mouse cell lines, which retain intact immune systems and can be particularly relevant for studies of immunologically-based targeted therapies (see Feb. 11, 2015 Eureka post.)..  

In clinical settings, WES is seen as increasingly important in making treatments more precise. Not every tumor is alike, but not every tumor is unique. Analyze 1,000 patients’ tumors and you might identify one or two variants that a majority have in common, which could be a good target for a new drug.

In the Charles River study led by Bruno Zeitouni, Senior Scientist in Bioinformatics at Charles Discovery Research Services GmbH and presented by Julia Schueler, Head of Tumor Biology Modeling, the same kind of question was asked of animals. The models were sequenced using WES along with DNA from the original mouse strains. Raw sequencing data were analyzed by mouse-specific bioinformatics tools they developed in-house to identify somatic mutations in the tumor models.

Different levels of mutation rates among the various models were observed—the highest was a renal cancer model and the lowest were the four GEM models—and common/specific mutations among the KP and KP-derived cell lines models. They also checked mutations in 23 selected targetable genes—such as DNA repair genes BRCA1 and BRCA2 associated with breast cancer—and learned that the KP models harbored the KRAS G12D mutation and the TP53 whole gene deletion known to be associated with human colon cancers while a melanoma syngenic model, with 11 mutations in these cancer genes, had one of the heaviest mutation loads.

Ultimately the study suggested that the models with highest mutation rates tended to be more sensitive to checkpoint inhibitors alone or in combination with another checkpoint inhibitor, though it wasn’t always the case. While the two melanoma models both had similar mutation rates, one responded to the drugs, the other did not.

Which perhaps suggests that the tumor models, despite all this interrogation, still have more secrets to tell.