In a study published in the journal Nature, researchers at Dana-Farber Cancer Institute and an international team of associates used a technique involving viral proteins to prioritize which of the many mutated genes in tumor cells are likely involved in cancer and which are mere bystanders. The technique may help researchers decide which mutated genes are most deserving of further study, and help pharmaceutical firms develop drugs with better chances of success.
"Technology for sequencing the entire genome of cancer cells often turns up dozens or hundreds of mutations in those cells," says study leader Marc Vidal, PhD, director of the Center for Cancer Systems Biology (CCSB) at Dana-Farber. "The challenge is to distinguish between mutations that cause or sustain the cancer and those that are just clutter. Viral proteins could be a tool for making such distinctions."
The approach, developed by an interdisciplinary team, takes its cue from two decades of discoveries into the role of some viruses in cancer. After being infected by a virus, a cell starts producing proteins encoded by the virus. These viral proteins alter the normal functions of the cell, interacting with cellular proteins and disrupting the function of cellular genes. If these alterations spoil the regular, orderly pace of cell division, the cell may become cancerous.
In many cases, the disruptions produced by cancer viruses are similar to those triggered by gene mutations. This raised an intriguing hypothesis for the members of the investigative team. "What if we examine the interactions between viral and cell proteins systematically – if we look for interactions across all the proteins from cancer-associated viruses with all the proteins in normal cells?" remarks study contributor Michael Cusick PhD, of the CCSB. "That viral proteins and gene mutations cause cancer by similar mechanisms led us to ask, could we use the viral proteins as stand-ins for identifying proteins – and the underlying gene mutations – associated with cancer?"
The study brought together researchers from the United States – at Dana-Farber, Brigham and Women's Hospital, the University of California, San Francisco – and Canada – at the University of Toronto and Mt. Sinai Hospital in Toronto. Researchers used 144 proteins from four viruses linked to cancer. They screened the proteins one by one to see if they interacted with or changed the expression of the roughly 20,000 different proteins in human cells. The result was the identification of roughly 950 cell proteins potentially involved in cancer.
An analysis of the results showed the technique to be as successful in identifying cancer-related mutations as the two most common methods currently in use. By combining all three methods, scientists can attain even more precision in determining which genes might be involved in cancer.
The results confirm the usefulness of viruses as tools for prying open the genetic secrets of cancer. "Viruses have spent millions of years perfecting the art of interacting with cells – of taking advantage of gene pathways that are essential for the cell and may therefore be essential to cancer," Cusick remarks. "We're now using them to guide us to the mutations that are the most important in cancer." The lead authors of the study are Orit Rozenblatt-Rosen, PhD, Megha Padi, PhD, Guillaume Adelmant, PhD, Michael Calderwood, PhD, and Thomas Rolland, PhD, of Dana-Farber, and Rahul Deo, of UCSF.
The research was funded primarily by a Center of Excellence in Genomic Science grant (P50HG004233) from the National Human Genome Research Institute of the National Institutes of Health (NIH). It also received additional NIH grant support, (RO1CA093804; RO1CA063113; PO1CA050661; RO1CA081135; RO1CA066980; UO1CA141583; RO1CA131354; RO1CA047006; RO1CA085180; T32HL007208; KO8HL098361; KO8CA122833; F32GM095284; and K25HG006031).