With such a model, Melton says, researchers could begin to address specific questions about how type 1 diabetes develops and progresses. For example, if it were possible to "reboot" a diabetic patient's immune system, would he necessarily redevelop the disease? Is the disease process the same in all affected patients? And which of the three cell types is the first to go awry? "It's quite amazing that we don't know the answer to that," Melton says.

It would also be possible to use this model to test potential treatments, says Melton. But he cautions that all these applications are still a long way off. A humanized mouse carrying differentiated diabetic iPS cells doesn't yet exist, and it could be years until it does.

One potential roadblock is that different samples from the same patient don't yield identical cell lines. The adult cells are reprogrammed using viruses that incorporate themselves into the genome and encode proteins known to sometimes revert cells to an earlier developmental state. These viruses don't insert themselves in a predictable way, meaning the proteins they specify aren't manufactured uniformly among different infected cells. "There's no certainty that you can just make one iPS clone from an individual and assume it's going to be representative of that individual," says Loring.

Although the main goal of the project was to create a model of type 1 diabetes, the diabetic iPS cells may have therapeutic potential as well. In theory, iPS cells derived from a particular patient could be turned into betalike cells that could then be implanted to replace the cells that her immune system has destroyed. But there would be nothing to stop the patient's immune system from attacking the implanted cells just as it attacks new beta cells naturally manufactured by the body. "That's the catch-22 with this particular disease, and autoimmune diseases in general," says Loring.

However, Loring says, it may be possible to cloak the implanted cells in order to hide them from the immune system. If so, it would be preferable to use cells derived from the patient to be treated. If the cloaking were to fail, the implanted cells would at least be recognized as "self," avoiding a catastrophic rejection response. This would obviate the need for dangerous antirejection immunosuppressant therapy.

Melton suspects, however, that modifying beta cells enough to hide them from the immune system would so thoroughly alter them that they would no longer serve their normal function. And in any case, he says, the current methods for generating betalike cells from iPS cells are too inefficient to be therapeutically useful.