Kroon et al., Nature Biotechnology, Advance Online Publication (20 February 2008)

The full picture.
Human ES cells can eventually give rise to cells that resemble pancreatic beta cells (labeled β).

A Sweet Success for Embryonic Stem Cells

Scientists have for the first time prodded human embryonic stem cells to become functional pancreatic cells. The work, published online today in Nature Biotechnology, marks an important step toward using embryonic stem cells to treat diabetes.

One of the most sought-after prizes in embryonic stem (ES) cell research is a method to turn the cells into pancreatic beta cells. These cells produce insulin in response to sugar in the blood, and they are damaged or missing in type 1 diabetes. If scientists could find a reliable way to make beta cells from human ES cells, they might be able to replenish patients' supply. But so far, no one has managed to produce functional beta cells in the laboratory.

Now, developmental biologist Emmanuel Baetge and his colleagues at the biotech company Novocell in San Diego, California, report that they have managed to coax human ES cells to behave like beta cells in mice. For several years, the group has been attempting to use the molecular signals that prompt pancreatic development in a fetus to direct human ES cells to become beta cells in a dish. In previous work, the researchers described a technique that seemed to produce insulin-producing cells, but the cells did not respond to glucose, a key characteristic of working beta cells.

So the researchers backed up a step. Instead of trying to get mature beta cells in a dish, they treated ES cells with signals that coaxed them to form pancreatic endoderm, a type of cell similar to those present in a human fetus 6 to 9 weeks old. They implanted these still-immature cells into mice, hoping that the final signals that prompt beta-cell formation could be provided by the animals' bodies. Thirty days after the implant, the researchers could detect the human version of C-peptide, a byproduct of insulin production, in the animals' blood. (The researchers looked at C-peptide instead of insulin because it is easier to tell the difference between the human and mouse versions.) After 2 months, the levels of human C-peptide went up when the mice received a dose of glucose, indicating that the implanted cells were responding to blood sugar. Finally, the researchers selectively killed the animals' own beta cells with a toxin, which usually causes mice to become diabetic. But rodents that had received implanted human cells did not develop diabetes, showing that the implanted cells could take the place of the animals' beta cells.

The work is "highly significant for the field," says diabetes expert Teresa Ku of the Beckman Research Institute in Duarte, California, whose group has also been working on ways to turn ES cells into beta cells. Baetge says that Novocell is already meeting with the U.S. Food and Drug Administration to discuss what additional safety tests will be necessary before they can consider starting human trials. But Ku notes that it is not clear whether the pancreatic endoderm cells are self-renewing. If not, she says, they might give out after a few years, and patients would have to undergo multiple transplants. A fully differentiated beta cell in the culture dish is still the surer bet, she says.

Related sites

Posted in Health, Biology