"It's a proof of principle," said Dr. Lawrence Chan, professor of medicine and molecular and cellular biology as well as chief of the division of diabetes, endocrinology and metabolism at the College. "The exciting part of it is that mice with diabetes are 'cured.' "
In the research, which is described in a report in Nature Medicine's online edition today, Chan and his colleagues used the NeuroD gene, a transcription factor that induces the liver to produce cells that make insulin and the three hormones associated with the pancreas' endocrine system.
The gene was attached to a so-called "gutless" adenovirus from which all toxic genes had been removed. This viral vector is a very efficient way to introduce genes into liver cells. Alone, NeuroD partially corrected the disease in the diabetic mice. Combined with a beta cell growth factor called Btc, the gene therapy complete cured the mice's diabetes for at least four months.
An added benefit is that the cells in the liver also produce glucagon, somostatin and pancreatic polypeptide, which may play a role in controlling insulin production and release.
"Until now it has not been possible to induce the formation of islets by any gene therapy approach," said Chan.
It does not mean that the treatment can be used in people immediately.
"It's farther from people than I would like," he said. He knows of no stumbling blocks to its effectiveness in people.
The main stumbling block is the vector or virus used to take the gene into the cells. Chan and his colleagues used the safest viral vector available today, but he expects even safer ones to be available within the decade.
"We want to use the safest vector possible," he said.
The treatment has advantages over transplant of islet cells, the insulin producers in the pancreas, because it avoids the lifelong use of powerful immunosuppressive drugs and eliminates the need to find a compatible donor.
Chan credits one of his postdoctoral students, Dr. Hideto Kojima, with much of the work in developing this protocol.