Through a series of painstaking experiments involving mice, the Harvard biologists pinpointed three crucial molecular switches that, when flipped, completely convert a common cell in the pancreas into the more precious insulin-producing ones that diabetics need to survive.

The feat, published online yesterday by the journal Nature, raises the tantalizing prospect that patients suffering from not only diabetes but also heart disease, strokes, and many other ailments could eventually have some of their cells reprogrammed to cure their afflictions without the need for drugs, transplants or other therapies.

"It's kind of an extreme makeover of a cell," said Douglas Melton, codirector of the Harvard Stem Cell Institute, who led the research. "The goal is to create cells that are missing or defective in people. It's very exciting."

The findings left other researchers in a field that has become accustomed to rapid advances reaching for new superlatives to describe the potential implications.

"I'm stunned," said Robert Lanza, chief scientific officer of Advanced Cell Technology in Worcester, Mass., a developer of stem-cell therapies. "It introduces a whole new paradigm for treating disease."

"I think it's hugely significant," said George Daley, a stem-cell researcher at Children's Hospital in Boston.

Even the harshest critics of embryonic stem-cell research hailed the discovery as a major, welcome development.

"I see no moral problem in this basic technique," said Richard Doerflinger of the U.S. Conference of Catholic Bishops, a leading opponent of the study of embryonic stem cells because it involves destroying human embryos. "This is a 'win-win' situation for medicine and ethics."

Melton and others cautioned that years of research lay ahead on whether the development would translate into cures.

"It's an important proof of concept," said Lawrence Goldstein, a stem-cell researcher at the University of California, San Diego. "But these things always look easier on the blackboard than when you have to do them in actual patients."

Although the experiment involved mice, there was optimism the approach would work in people.

"You never know for sure - mice aren't humans," Daley said. "But the biology of pancreatic development is very closely related."

Melton has already started experimenting with human cells in the laboratory and hopes to start planning the first studies involving people with diabetes within a year. "I would say within five years we could be ready to start human trials," Melton said.

Others have already started trying the approach on other cells, including those that could be used to treat spinal cord injuries and neurogenerative disorders such as Lou Gehrig's disease.

"The idea, to be able to reprogram one adult neuron type into another for repair in the nervous system, is very exciting," said Paola Arlotta, who is working in the Center for Regenerative Medicine at Massachusetts General Hospital-Harvard Medical School.

The research is the latest development in the explosive field of "regenerative medicine," which is trying to create replacement tissues and body parts tailored to patients. That dream appeared within reach after scientists discovered human embryonic stem cells, which can develop into any type of body cell.

But stem-cell research has been plagued by political and ethical debates because the cells can only be obtained by destroying embryos, which has been opposed by President Bush and others who believe that even the earliest stages of human life have moral standing.

Scientists last year shocked the field when they announced they had discovered how to manipulate the genes of adult cells to turn them back into the equivalent of embryonic cells - entities dubbed "induced pluripotent stem," or "iPS," cells - which could then be coaxed into any type of cell in the body.

The new work takes further advantage of the prowess scientists have developed in harnessing the once-mysterious inner workings of cells - this time to skip the intermediary step of iPS cells and directly transform adult cells.