Like the brain, the retina is not good at healing itself. Though human embryonic stem (ES) cells have been coaxed into becoming retinal pigment epithelial cells capable of rescuing vision in a mouse model1 another type of cells, light-detecting photoreceptors, have proved much harder to produce in culture. Now, researchers led by Masayo Takahashi at the RIKEN Center for Stem Cell Biology in Kobe, Japan, have derived putative photoreceptors from mouse, monkey and human ES cells2. Notably, the human cells were differentiated under fully defined culture conditions that did not require animal-derived ingredients or coculture.

Working first with mouse cells, Takahashi and colleagues found that they could obtain retinal progenitors by exposing clumps of ES cells to antagonists of two well-known signaling pathways (Wnt and NODAL) and then sorting for cells expressing a retinal transcription factor, Rx. To move the cells further toward photoreceptors, Takahashi and colleagues blocked a pathway (Notch) necessary for self-renewal. This caused cells to differentiate into precursors and spontaneously into cone photoreceptors. By adding more ingredients, including fibroblast growth factor and retinoic acid, the researchers obtained putative rod photoreceptors.

Takahashi and colleagues then adapted their protocol to primate cells, taking care to avoid using serum or any undefined components. After a protocol lasting over 300 days, they were ultimately able to obtain both rod and cone photoreceptors from human cells for a total efficiency of around 30%.

Most untreatable blindness is due to problems not with the structures of the eye, but rather with the neural cells of the retina. If sufficient quantities of these cells could be found, they might make cell therapies possible. The field still has many obstacles to surmount: culture must become faster and more efficient, better techniques to purify photoreceptors from other cells must be developed, and the cells must be able to survive and function in the eye. So far, these cells have been identified by cell markers and morphology rather than function; thus, this work is a significant step toward obtaining cells with therapeutic potential.