Biologists have hunted for weak spots in cancer cells for years, hoping to find clues to the disease that can be exploited. That should get easier thanks to a mass-screening technique reported in the 1 February issue of Science that may provide a cost-effective and powerful way to pick out new drug targets against cancer.
As genetic technology has grown more sophisticated and cheaper, scientists have begun dissecting a cancer cell's arsenal on a massive scale. In 2005, the National Institutes of Health in Bethesda, Maryland, launched The Cancer Genome Atlas (TCGA), a $1.5 billion search for genes that are mutated in a host of cancers (Science, 16 December 2005, p. 1751). Some scientists have criticized TCGA for focusing on gene sequencing while diverting funds from functional studies that can determine which of the hundreds of mutations are most important. One person with such concerns is geneticist Stephen Elledge of Harvard Medical School in Boston.
With molecular biologist Gregory Hannon of Cold Spring Harbor Laboratory in New York state, Elledge developed genetic tools that examine how genes function in human cancer cells. As they report in two new studies, the pair and their colleagues constructed viral vectors, each one containing an RNA molecule designed to shut down a gene with a complementary sequence. The vectors also contained DNA bar codes, sequences that the researchers could look for later to determine which small RNAs were having a big effect on a cell's behavior. The researchers inserted between 10,000 and 40,000 of these small RNAs at once into breast cancer, colon cancer, and normal human cells in the lab. The main analysis, of 10,000 short hairpin RNAs, targeted about 3000 different genes, says Elledge. Then they waited to see which small RNAs would blunt survival or growth of cancer cells without affecting normal ones. The theory is that those genes hit by RNAi are acting in concert with abnormalities in the cancer cell to cause out-of-control proliferation.
On this first pass, roughly two dozen genes fit the bill, says Elledge. "It looks like you can get a lethality signature," a pattern of genes that affect how tumor cells proliferate. Elledge cautions that this is just a start in determining which proteins might make good drug targets and that the technique won't pick up every one. The work cost about $30,000 to conduct once the tools were in place.
Other experts are hopeful that this approach will pay dividends. "I'm a very big fan," says cancer geneticist Ronald DePinho of the Dana-Farber Cancer Institute in Boston, who also chairs TCGA’s advisory board. He believes that given cancer's complexity, it is necessary to both survey the genomes of tumor cell and examine how the cells respond when genes are silenced. If the two methods independently come up with genes that are interconnected, "it would be a much more powerful way of prioritizing" what to study next, says computational biologist Michael Bittner of the Translational Genomics Research Institute in Phoenix, Arizona.