Scientists have identified three new stretches of DNA that appear to boost the chance of getting diabetes. The findings, confirmed in roughly 32,000 people with and without the disease, suggest that so-called genome-wide association (GWA) studies are paying off. Scientists caution that they have a long way to go before the new genetics can be applied to patients, but they're hopeful that after years of trying, they're finally beginning to decipher the genetics of common diseases.
Many common diseases are extremely complex at the genetic level. Unlike rare disorders such as cystic fibrosis--which has been traced back to a single gene--cancer, heart disease, and type 2 diabetes likely involve many genes, each playing a mild role in boosting susceptibility. Hunting these genes down is a chore, but new technology allows scientists to "scan" entire genomes for specific disease markers. One of the first GWA studies, in age-related macular degeneration, found a gene that boosted disease risk significantly (ScienceNOW, 11 March 2005), and a handful found since then, like the first, at least double the chance of suffering from this retinal deterioration.
Now type 2 diabetes is in the limelight. Reporting online today in Science, three groups in the U.S. and the U.K. describe their collective discovery of three genetic regions that each boost the risk of the disease by roughly 10% to 40%. One of the three, a gene called CDKAL1, was also found by scientists at deCODE genetics in Reykjavik in their study of 1399 diabetes patients and 5275 controls in Iceland. DeCODE's findings are described in a paper also published today online in Nature Genetics.
The power of this work, say scientists, comes in part from the enormous number of DNA samples surveyed, more than 32,000, with one of the three variants independently replicated by the deCODE team. The three new associations found were vanishingly unlikely to have occurred by chance. However, the role each individual gene plays in diabetes appears to be quite small.
Separately, the Science papers confirmed six other genetic regions that others had previously identified as having a connection to type 2 diabetes. Kári Stefánsson, the chief executive officer of deCODE, notes that a smaller sample size may help explain why his team didn't report two of the three new variants found by the three groups that pooled their data. The Icelandic group found associations for those two, he said, but they weren't strong enough to pursue.
Getting teams to cooperate in such a cutthroat field may be tough, but it's "absolutely required" in order to find genes like these, says Daniel E. Weeks, a statistical geneticist at the University of Pittsburgh in Pennsylvania. The findings are just the beginning of what GWA studies will accomplish, notes David Altshuler, the director of the program in medical and population genetics at the Broad Institute in Cambridge, Massachusetts, who helped lead one of the teams reporting results today. The next step is to sequence these regions and confirm the relevant genes. Figuring out how they work "is going to take great creativity and insight," says Altshuler, as will determining how and when to apply the results to patients.