Dr. Fred Levine describes a way for tweaking ordinary adult cells in the pancreas so they become insulin-producing beta cells. About 21 million Americans have the disease, which occurs due to the loss or malfunction of beta cells or because the body resists its own insulin. Beta cells form and live in pancreatic islet cells, an area that makes up 2 percent of the pancreas. The theory that pancreatic cells other than islet cells could produce insulin had been discounted in a widely regarded paper published in the journal Nature in 2004.

But working on experiments with mice, Levine and his colleagues crafted what they say is “a new way of thinking.” Their theory, though, remains many years away from being tested on humans.

They discovered that when mixed with cells from pancreatic fetal tissue, non-islet pancreatic cells could be coaxed into becoming beta cells.

“What we found was that 10 percent of the cells implanted in the mouse had turned into beta cells,” said Levine, who works for the Burnham Institute and the University of California San Diego's Rebecca and John Moores Cancer Center, both in La Jolla.

Initially, the strategy described by Levine and his colleagues might be most applicable for people with so-called juvenile, or Type I, diabetes. This version of diabetes, which affects about 200,000 children and adolescents in the United States, occurs when a person is born with faulty or missing pancreatic islet beta cells so not enough insulin is produced.

It also shows up in people whose islet cells are mistakenly destroyed by their body's immune system.

Theoretically, Levine said, his experimental method might be adapted to benefit the much larger group of people with Type II diabetes. With this form of the disease, the body develops resistance to insulin. Type II patients would gain greatly from a reactivation of beta cells, Levine said.