The current work may help build a foundation for future therapies in which cells of the inner retina--and potentially other cells, including photoreceptors--are regenerated in situ, in the living human eye, says Reh. In theory, such treatments might allow physicians to replace retinal neurons "precisely at the spot where they're needed, without disruptions or discontinuities," he says.

In lower vertebrates like fish and chickens, retinal cells are known to generate new neurons in response to damage, often restoring sight. While mammals do not have the same self-healing capacity, some previous research has suggested that under particular circumstances, mammals' retinas might be able to generate new neurons. Reh's current work offers more definitive evidence that immature cells, derived from muller glia, can differentiate again into mature neurons, says Michael Young of the Schepens Eye Research Institute.

More research is needed before retinal regeneration can be attempted in humans. "We need much more control over the basic cellular processes"--trying to regenerate different types of neurons and making sure that they function properly in vivo--"before we can treat real people with blinding disease," says Anand Swaroop of the National Eye Institute.

For example, scientists need to show that regenerated neurons behave normally in the eye, integrating into circuits with other cells and contributing to vision. "It is hard enough to grow different cell types," says Rose. "But will they function? Will they do what the cells they are replacing normally would do? That's really tough."

Reh says that growing new cells in the eye could be preferable to transplanting cells, an approach that his team is also working on. "Transplantation involves a tricky surgery; the cells may not go exactly where you want them to go," says Reh, and some cells could cause an immune reaction. "Developing methods to stimulate regeneration may prove to be the best option in the long run."