For years, researchers have investigated how the body loses the ability to produce enough insulin, a hallmark of diabetes. Now an intriguing theory is emerging, and it suggests a potential treatment that few scientists had considered.
The hormone insulin helps shuttle glucose, or blood sugar, from the bloodstream into individual cells to be used as energy. But the body can become resistant to insulin, and the beta cells of the pancreas, which produce the hormone, must work harder to compensate. Eventually, the thinking goes, they lose the ability to keep up. “We used to say that the beta cells poop out,” said Alan Saltiel, director of the Life Sciences Institute at the University of Michigan. In reality, he added, this shorthand meant “we have no idea what’s going on.”
Some evidence suggested that large numbers of these cells died through a process of programmed cell death called apoptosis. But that was at best a partial explanation. Now, researchers at Columbia University have put forth a surprising alternative.
In mice with Type 2 diabetes, the researchers showed that beta cells that had lost function were not dead at all. Most remained alive, but in a changed form. They reverted to an earlier developmental, “progenitor,” state.
It’s as if these cells are “stepping back in time to a point where they look like they might have looked during their development,” said Dr. Domenico Accili, director of the Columbia University Diabetes and Endocrinology Research Center, who led the new work.
If researchers could find a way to reverse the process, coaxing them to become beta cells again, these cells might regain the ability to produce insulin. “It’s a pretty new idea,” Dr. Saltiel said, and one that “offers a lot of hope.”
In earlier work, Dr. Accili sought to understand what happened to beta cells at the molecular level as diabetes progressed. He investigated the role played by a protein called FOXO1, which seems to disappear as beta cells stop producing insulin.
In the new work, published in September in the journal Cell, Dr. Accili, Chutima Talchai, then a postdoctoral fellow in his laboratory, and their colleagues genetically engineered mice that lacked FOXO1 in beta cells. At first the animals appeared normal. But as they were subjected to stress — pregnancy for the females, aging for the males — the mice developed high blood sugar, decreased insulin secretion and other signs of Type 2 diabetes.
The mice also began to produce proteins normally found only during fetal development. Some of their beta cells, Dr. Accili found, had come to resemble progenitor cells. Similar tostem cells, these are destined to become, or “differentiate” into, hormone-producing cells as the animal matures. Their appearance in adult mice was a surprise. Dr. Accili also showed that this process — in which beta cells lost their identities, or de-differentiated — accounted for virtually all of the animals’ decrease in insulin-producing capacity. It wasn’t happening just on the margins.
Other laboratories are scrutinizing the odd phenomenon. Matthias Hebrok, director of the diabetes center at the University of California, San Francisco, said that he and his colleagues also had evidence that stress could cause insulin-producing cells to stop working properly and revert to a less mature state.
A range of physiological stresses, including obesity, pregnancy and aging, all tend to increase demand on beta cells to produce more insulin, Dr. Accili said. It may be that they are “taking a little rest,” he said, in returning to a less active state. Although it’s not yet clear why this might happen, the finding may lend support to the view that doctors should focus on relieving stress on the beta cells rather than pushing them to produce more insulin, which may speed the progression of diabetes, Dr. Accili said.
If something similar occurs in humans with Type 2 diabetes, scientists may find ways to nudge de-differentiated beta cells to return to their mature, insulin-producing form. In Dr. Accili’s research, some of the progenitor cells went on to produce another hormone,glucagon, which acts to raise blood sugar.
This suggests that these cells still have the ability to change into other types of endocrine cells. “Why not become beta cells again?” Dr. Accili asked.
Indeed, research in mice has shown that when large numbers of beta cells are destroyed,some hormone-producing alpha cells morph into beta cells. Another study has shown that through the addition of proteins called transcription factors, another pancreatic cell type, called acinar cells, can become beta cells.
“The pancreas is not an entirely static collection of cells. It’s much more dynamic than we previously appreciated,” said Dr. Mark A. Magnuson, professor of molecular physiology and biophysics at Vanderbilt University Medical Center. “This is the concept of cellular plasticity, and it’s very much in play.”
Still, Dr. Magnuson added, turning progenitor cells efficiently into functioning beta cells that would respond normally to glucose remains a challenge. “There are a lot of people thinking about it, but it’s a tough nut to crack.”

I found this article novel in the fact that we believed for so long that insulin producing B cells in the pancreas reached programmed cell death due to stress that pushed for them to produce more insulin (diabetes, pregnancy etc.) but instead they revert back to a developmental progenitor state in mice, which could then differentiate into other new hormone producing cells as the mouse matures. It is interesting then that the cells seem to recognize the stress and rather than pushing a disease like diabetes forward more aggressively, that they revert to an immature state. What is causing this, what mechanisms are involved? If this also occurs in humans, the thought of being able to force these now progenitor cells into becoming Beta cells again could have substantial influence. Additionally, other cells in the pancreas as the article mentions, have the ability to morph into Beta cells. This suggests that the pancreas is very diverse and contains a sense of plasticity.