NEW YORK (GenomeWeb News) – Using a mutagenesis screen in combination with high-throughput DNA analysis, an international team of researchers has identified hundreds of potential cancer genes in mice.
In a paper published online today in Cell, researchers from the Netherlands Cancer Institute and the Wellcome Trust Sanger Institute triggered cancer in wild type mice and mice lacking the tumor suppressors p53 or p19ARF using so-called retroviral insertion mutagenesis. Then they analyzed 500 resulting tumors. In so doing, they found more than 250 new genes that may be involved in cancer — and they started unraveling networks containing new and previously identified cancer-related genes.
The Moloney murine leukemia virus, or MuLV, causes white blood cell cancer in mice by integrating into host DNA and influencing gene activity and/or disrupting certain genes. In order to understand where these insertions are — and how they contribute to cancer — the researchers infected both wild-type newborn mice and newborn mice lacking p53 or p19ARF with the virus and monitored each for tumor development.
“Human cancers are generally thought to be formed by the stepwise accumulation of mutations that disrupt genes within a cell, and the virus mimics this process as it inserts itself into the mouse genome,” co-senior author David Adams, an experimental cancer geneticist at the Wellcome Trust Sanger Institute, explained in a statement. “The virus then acts as a ‘tag,’ allowing us to identify where it has integrated and which gene or genes have been disrupted.”
They eventually isolated genomic DNA from 510 tumors — mainly of the spleen, thymus, and lymph nodes — and identified viral insertion sites using linker-mediated PCR and shotgun subcloning. When the researchers subsequently mapped the insertions onto the mouse genome, they were able to pinpoint the genes affected by the virus in each tumor.
Using this approach, the team identified 10,806 independent mutations in more than 300 regions of the mouse genome that may contribute to tumorigenesis. Among them were known and suspected tumor suppressors as well as characterized or suspected oncogenes. The researchers were also able to begin constructing networks of genes that were mutated in the same tumor as well as mutations that were mutually exclusive.
Intriguingly, when they then compared their results in mice with genes in the Cancer Gene Consensus, they found that dozens of the common insertion sites identified in mice were in or near mouse orthologs of human cancer genes — including genes linked to non-hematopoietic tumors. Similarly, some of the same signaling pathways that are known to be disrupted in human cancers were also affected in the mouse screen.
And when the team looked at copy number variation by hybridizing 713 cell lines collected through the Sanger Institute’s Cancer Genome Project to 10,000 SNP arrays, they found that more candidate genes from the mouse screen were amplified in these samples than one would expect at random.
“This scale of analysis allows identification of a high number of new candidate oncogenes and tumor suppressors and detects highly significant combinations of co-occurring or mutually exclusive genes,” the authors wrote.
The researchers are reportedly doing similar experiments using other viruses and mutagenesis techniques in an effort to identify a similar set of suspect genes for bowel, lymphoma, and breast cancers.
Data from the study is incorporated into an online resource called MutaPedia, intended to help others identify cancer genes and delve into interactions between oncogenic mutations.