Researchers at the University of California, Los Angeles (UCLA) have created the first network of mammalian genetic interactions by cross-referencing previously published genome maps from four different mammalian species. This network could help our understanding of genetic diseases rooted in abnormal cell division and unrestrained cell growth, such as cancer.
The researchers, led by Desmond Smith, professor in the department of molecular and medical pharmacology at UCLA, found extensive overlap with regard to gene interaction networks while comparing the genome maps of humans, dogs, cats, and mice (1). They then combined the networks by rearranging the non-human species onto the scaffold of the human genome. The group’s current work stems from their 2008 discovery that radiation hybrid panels from the 1980s could be used to map gene expression levels and the loci that regulate those expression levels (2).
When radiation hybrid mapping was conceived by cell biologist Henry Harris in the 1970s, it was a solution without a problem; nobody knew how to exploit this technique. In 1990s, David Cox—then a professor of genetics and pediatrics at Stanford University School of Medicine—proposed using radiation hybrid panels in the Human Genome Project. These panels were employed to produce long-range maps of mammalian chromosomes, providing the coordinates of genes. But once the project was completed, these radiation hybrid maps, as well as the technique itself, quickly faded from the research community’s memory.
“We realized we could take this old data that was gathering dust in databases and, instead of using it to make high resolution maps, we could ask a different question,” Smith told BioTechniques.
Although the original purpose of radiation hybrid mapping was to determine if a gene was present in each radiation hybrid plane, Smith wondered if these maps could provide information about inheritance patterns between particular genes: did genes tend to be inherited together or separately in the radiation hybrid cells? If gene A is present, does that mean that gene Z is always absent? Or if Z is present, is A always absent?
“I realized that in principle we could use this data to make genetic interaction maps, depending on whether genes tend to appear together or not appear together in different radiation hybrid clones,” said Smith.
Relying on statistics to verify trends of co-inheritance, a conjecture might then be made regarding interactive relationships. “If you see statistically a co-inheritance or lack of co-inheritance, you can then deduce that A and Z must somehow biologically interact to help cells divide or kill cells off,” said Smith.
For example, if researchers look up a gene that is associated with unrestrained growth in a cancer cell on the radiation hybrid map, the team could identify all the genes that interact with it and determine if indirectly tampering one of these has an effect on outcomes.
Smith expects that about a hundred of these interactions will be relevant to cancer. His team plans to test these interactions in living cells to find out if they reoccur in an in vivo setting. The group will begin by hunting the most notorious genes, such as MYC, a usual suspect in breast and brain cancer; HRAS, which plays a lead role in bladder cancer, and EGFR, mutations of which are linked to lung and breast cancer.
For now, the team has a lot of data to sort through: computational analysis revealed 7 million interactions with statistical significance in the human genome.
“Now we’re going to cherry-pick the ones that might be medically relevant, and see if our discovery stands up to further scrutiny when we actually do an experiment,” said Smith. “I’ve always been seduced by the beauty of this method. It never left my mind.”
1. Lin, A., R.T. Wang, S. Ahn, C.C. Park, and D.J. Smith. 2010. A genome-wide map of human genetic interactions inferred from radiation hybrid genotypes. Genome Res. 8:1122-32.
2. Park, C.C., S. Ahn, J.S. Bloom, A. Lin, R.T. Wang, T. Wu, A. Sekar, A.H. Khan, et al. 2008. Fine mapping of regulatory loci for mammalian gene expression using radiation hybrids. Nat Genet. 4:421-9.