The researchers report they were able to transform about one in 5,000 cells—enough to get several iPS cells from a single culture dish—and then coax them to become nerve cells or heart tissue on the benchtop. Genetic scans indicated that the cells were more similar to embryonic tissue than to the original fibroblasts.
"What I find remarkable," Kriegstein says, "is essentially the same steps that worked in the mouse were able to work with human cells…. Everybody assumed that it'd be a different story in reprogramming the human cells."
Thomson's team reports producing a similar cellular alchemy using two of the same genes—Oct4 and Sox2—and two different ones—Nanog and Lin28,—making it less likely that the Japanese finding was a fluke.
Thomson, who in 1998 became the first scientist to extract human stem cells from embryos, says his group began seeking these factors four years ago, but chose to work with human cells. As of last spring, he says, his group, led by lab member Junying Yu, had pared an initial list of 100-plus genes to 14.
Then came Yamanaka, who Thomson says beat him to the punch because mouse cells grow much faster than human cells do, allowing more rapid experimentation. "We thought, 'oh no, it's already been done; we've been beaten,'" he recalls.
As many as a dozen major labs, he says, have since tried but failed to make reprogramming work in human cells. His team plugged along, testing gene combinations in four cell types in varying degrees of differentiation, hopeful that this would eventually lead to the correct genetic recipe.
In the online edition of Science, he and his colleagues report that Oct4 and Sox2 were capable of converting neonatal foreskin fibroblasts into cells similar to Yamanaka's, whereas Nanog significantly boosted the frequency of reprogramming and Lin28 upped it by a moderate amount.
Although Oct4 and Sox2 were well-known players in embryonic cells, "we thought this would be such a complicated problem, we never tested those genes up front," Thomson says. "It's kind of remarkably lucky that three or four [genes] are sufficient. I wasn't optimistic it would work."
The new results still leave researchers with the task of double-checking that reprogrammed cells are safe and truly have the same potential as the embryonic variety. They also have to figure out how to circumvent problems with the viral delivery system, which may disrupt important genes, resulting in cancer. Kriegstein notes that the hunt will likely commence for small molecules capable of activating the key genes.
Thomson predicts that companies such as Madison-based Cellular Dynamics International, which he co-founded, that test drug candidates for dangerous heart toxicity, could begin using cells derived from reprogramming in their assays within a year.
Whatever the source of pluripotent cells, Thomson says, researchers face the same scientific challenges—namely, understanding how to convert them into key tissues such as the beta islet cells that are impaired in diabetics, and then how to introduce them safely and effectively into humans.
Opponents of research on human embryos might contend that reprogramming happened because of the federal restrictions on embryonic research, but Thomson believes the stigma on the field made researchers wary and delayed the discovery of reprogramming by several years. "I'm cautiously optimistic" about reprogrammed cells, he says. "The worry is the politics will get involved again and squash caution."