Bell’s group’s work (#9) is a tour de force that can scarcely be given adequate attention in a summary as brief as this. The problem, in a nutshell, is that type 2 diabetes is not a simple disease caused by a single gene. If it were, it would have been pinned down long ago, such is the interest in the disease. Rather, it is multifactorial, with both environmental and genetic components, and even then the genetics of diabetes is not simple. Bell and Craig Hanis had previously shown that in a population of Mexican Americans, who are very susceptible to diabetes and who had been studied for more than 20 years, there is a statistical linkage between the presence of the disease and a region on chromosome 2. But the "gene," which they called NIDDM1, was an extremely long stretch of DNA, with no precise location for the defect that causes diabetes.

"It was extraordinarily difficult," said Bell at the time, "because the gene location is not precisely defined.…Rather, it is only a probability that the gene will be in a particular region, so the region that you have to search is much larger than for a single-gene disorder."

Gradually, using a variety of sequencing and statistical techniques, the group reduced the search area from 1.7 Mb to 66 Kb and eventually homed in on a previously unknown gene that made a protein called calpain-10. The protein is a protease. That is, it cleaves other proteins. "This protease was not on anyone’s list of favorite genes for affecting either insulin secretion or insulin action or hepatic glucose production," Bell explained. "Having gotten this far we now end up with many more challenges ahead." Challenges such as how exactly calpain-10 affects insulin’s activity and the basis of its involvement in diabetes are currently being investigated.

At present there does not seem to be any connection between calpain-10 and the subject of the #10 paper, resistin. This is an entirely new hormone which is secreted by fat cells and seems to make other cells resistant to insulin. Crucially, resistin offers a plausible causal link between type 2 diabetes and obesity, one of the biggest risk factors for the disease. Mitchell Lazar and his team at the University of Pennsylvania started from a new class of anti-diabetic drugs, thiazolidinediones (TZDs). Lazar has described the use of insulin to treat diabetes as "talking louder to deal with a poor telephone connection, rather than fixing the line." TZDs seem be an amplifier, boosting a cell’s sensitivity to insulin.

Lazar’s group looked for genes that are active during the formation of fat cells, but which are downregulated in mature fat cells that are exposed to TZD. Of the few that fit the bill, one coded for a small protein that would be exported from the cell. It was produced in mouse white fat cells, but not in other tissues. The protein was also found circulating in the serum, decreased after a fast, and increased when the mice were fed again. Levels were higher in mice that were genetically predisposed to obesity and diabetes. Giving resistin to normal mice impaired insulin metabolism. And there is a human version of resistin.