Researchers use skin cells from type 1 diabetes patients to produce cells that make insulin

WORCHESTER, Mass.—A group of researchers have been able touse skin cells from people with type 1 diabetes to produce cells that madeinsulin in response to changing blood sugar levels, though not as efficiently asnormal insulin-producing cells do.

A study published recently in the Proceedings of theNational Academy of Sciences describes away to create induced pluripotent stem (iPS) cells from ordinary adult cellstaken from patients with type 1 diabetes. These stem cells then can bereprogrammed to produce all of the cell types relevant to the disease.

"What you get is the ability to watch, for the first time,type 1 diabetes develop," says senior author Douglas Melton, a professor ofnatural sciences at Harvard University and co-director of the Harvard Stem CellInstitute. "Until you watch a disease develop, you will not understand themechanism, and you therefore cannot devise any kind of sensible treatment orcure."

The major immediate implication from this experiment is thatscientists now have a preliminary lab model of human type 1 diabetes cells, andthe hope is that an animal model of the disease could be developed from thisresearch.

"These cells, which capture all the genes involved in T1diabetes (because the cells come from a patient), are going to become apowerful model to study how T1 D develops," Melton says. "It is not ourintention to use these cells for transplantation into patients, but rather toreconstruct or recapitulate the disease so that we can understand its cause."

To test whether their lab-made cells could function likenormal beta cells, Melton's group exposed them to glucose in a dish. When sugarlevels were high, the cells produced more of a protein that beta cells releasewhen they break down sugar; when glucose levels were low, the protein levelswere low as well.

As a result, by capturing flawed cells and studying theirimperfections, Melton notes that it is possible to gain an understanding of thegenetic cause of the disease. The science could also be applied to otherdiseases, he says.

"Work at the Harvard Stem Cell Institute onneurodegeneration (ALS and SMA) is another example," adds Melton.

Melton and his colleagues were able to show that thereprogrammed iPS cells can be spurred to differentiate into tissue resemblingthe insulin-producing pancreatic beta cells that are destroyed by the immunesystem in type 1 diabetes.

"Directing the differentiation of the pluripotent stem cellswas accomplished in a multi-step process, adding signals such as growth factorsand inducing molecules, at each stage," Melton says. "This has been done,albeit inefficiently, by many labs and the problem is how to make each stepmore efficient so that at the end one gets many beta or beta-like cells. Withpresent methods, one gets less than 1 percent of the cells becoming beta-likecells."

Embryonic stem (ES) cells have long been the gold standardfor deriving pluripotent cell lines. But ES cells can only be used to createdisease models for disorders such as cystic fibrosis, where the geneticunderpinnings are straightforward. Because the genetics underlying type1 diabetes are complex and poorly understood, researchers have no way toidentify diabetes-specific ES cells.

Therefore, iPS cells derived from patients known to bediabetic offer the best hope for modeling the disease by allowing researchersto generate diabetes-specific versions of all the relevant cell types: Thepancreatic beta cells, the immune cells that destroy them and the thymus cellsthat orchestrate their destruction.

Ultimately, Melton plans to construct a "living test tube"for probing the interplay between the beta cells and the immune system indiabetes. He hopes to use the diabetic iPS cells to generate all three relevantcell types and then to put those cells into a so-called humanized mouse thatcan accept human cells to see how they interact.

"We have to make beta cells more efficiently, then makecells of the immune system (by making hematopoietic stem cells), then makethymic epithelium (to educate the immune cells) and then, finally, put all 3components into a living test tube ( an immunodeficient mouse) to reconstructthe disease," he says. "It's an ambitious, some might say overly ambitious,experiment, but if it works, we will for the first time be able to watch humanT1D develop."

Measuring continued success will start with the results ofthat experiment. If it works, Melton says "then we'll have a celebration."

According to Melton, researchers could use such a model toaddress specific questions about how type 1 diabetes develops and progresses.It would also be possible to use this model to test potential treatments. Buthe cautions that all these applications are still a long way off. A humanizedmouse carrying differentiated diabetic iPS cells doesn't yet exist, and itcould be years until it does.

Moving forward, as these methods of making beta cells becomemore established, Dr. Rohit Kulkarni, a diabetes expert at Joslin DiabetesCenter in Boston, told Time Magazine thatthe strategy could be expanded to help patients with either type 1 or 2diabetes.

"It might even be more relevant for other types of diabeteswhere there is no immune-system attack," he says. In those cases, simplyreplacing nonfunctioning beta cells might go a long way toward treating or evencuring the disease.

The research also is good news to Susan Solomon, CEO of theNew York Stem Cell Foundation, which provided some of the funding for thestudy.

"This is a big deal," she says in a statement. "Tackling thebasic biology of type 1 diabetes, which is a very complex disease, is acritical step. With these cells, we can see in a dish what's happening to theimmune system, and if you don't understand the immune response, you get nowherewith type 1 diabetes."