Diabetic mice cast suspicion on protein.

When scientists arrive in the laboratory on a Monday morning and discover that some of their prized mice died over the weekend, the misfortune tends to get their attention. One such incident has now thrown a spotlight on a protein that may lie at the heart of type 2 diabetes, which strikes as much as 6 percent of the U.S. population.

Unlike type 1 diabetes, an autoimmune disorder in which the body destroys cells that make the hormone insulin, type 2 diabetes involves both decreased production of insulin and a diminished ability on the part of many tissues to respond to the hormone. Scientists have long found this combination difficult to explain with a single gene mutation, and it is one of many factors pointing to the involvement of several genes in the development of type 2 diabetes.

Yet investigators who have engineered mice to lack a single gene--the one that encodes a protein called insulin receptor substrate 2 (IRS-2)--have found to their surprise that the animals display both of the main characteristics of type 2 diabetes. Many of the mice ultimately fall into a coma and die if untreated.

This discovery "brings together the two arms of the disease in a common molecular pathway," says Morris F. White of the Howard Hughes Medical Institute (HHMI) at Harvard Medical School in Boston.

White and his colleagues study how cells respond to insulin, which regulates the concentration of the sugar glucose in blood. Years ago, they discovered IRS-1 and realized that it helps transfer insulin's signal into cells.

When an insulin molecule arrives at a cell, it latches onto a surface protein known as the insulin receptor. That interaction causes a chemical modification of proteins in the cell, such as IRS-1, which then disperse and activate other proteins, explains White.

Mice made to lack IRS-1 don't develop diabetes, however. They are smaller than normal and some tissues don't respond efficiently to insulin, but the mice compensate with insulin-making beta cells that enlarge and proliferate.

Although discouraged by that finding, White went on to create mice without IRS2, a closely related protein his group had also discovered. When some of those mice died of dehydration at around 8 to 10 weeks, investigators found that the animals' urine was flooded with glucose,

Further studies revealed that the animals' skeletal muscle and liver had developed a significant resistance to insulin. The animals were also unable to boost insulin production. Mice without IRS-2 start life with fewer beta cells than normal mice, the researchers found. Moreover, instead of making more beta cells as their resistance to insulin developed, these mice continued to lose the insulin producers over time.

It's not clear what role IRS-2 plays in beta cells or even if it acts there only in response to insulin. Whatever the trigger, the protein may help the cells enlarge, proliferate when needed, or avoid death from various stresses, such as high concentrations of sugar in the blood.

Other diabetes researchers remain skeptical that IRS-2 plays a major role in the human illness, noting that White and his colleagues have so far failed to unearth people with diabetes who have flaws in the gene for IRS-2.

"Time will tell how relevant IRS-2 is to the human condition," says Graeme I. Bell, an HHMI investigator at the University of Chicago. "Mutations in it don't seem very important in man."

Preliminary investigations have hinted that some people with type 2 diabetes have altered IRS-2 activity, White counters. In those people, mutations in other genes or environmental factors such as body weight may regulate the protein and any role it may play in diabetes.

White suggests that stimulating IRS-2 activity could offer a treatment for either type of diabetes. "if we can figure out what's triggering IRS-2 in beta cells, we have a potential drug target to keep beta cells alive," he says.