Neuronal cell fate in the retina is determined by a stepwise process to generate seven main cell types — among them rod and cone photoreceptors — from common pools of progenitor cells. Swaroop and colleagues elegantly demonstrate that the transcription factor neural retina leucine zipper (NRL) is not only essential but sufficient to drive differentiation of precursors to rod photoreceptors.

NRL is a transcriptional regulator of rod-specific genes, and mice lacking NRL (Nrl−/−) develop rod-free, cone-only retinas. The questions remained of whether or not rods and cones are established from a common pool of competent cells, and whether Nrl expression can drive the transformation of precursors into rod-only retinas. Swaroop and colleagues generated transgenic mice expressing Nrl under the control of cone–rod homeobox (Crxp-Nrl) on a wild-type (Crxp-Nrl/WT) or Nrl−/− (Crxp-Nrl/Nrl−/−) background. Crx is expressed in postmitotic cells of the retina that can develop into either cone or rod photoreceptors.

Retinal sections of Crxp-Nrl mice of both backgrounds contained only rods and no cones. Electroretinograms (ERGs) obtained during darkness and light showed that rod-mediated dark adaptation was normal in both Crxp-Nrl mouse lines, but impaired in Nrl−/− mice. Cone-derived ERGs showed an impaired response in both Crxp-Nrl mouse lines, but normal and supernormal responses in wild-type and Nrl−/− mice respectively, as expected. These results indicate that Nrl expression converts all cells of the common cone/rod precursor pool into rods.

Furthermore, the researchers tested the influence of Nrl on cells already determined to become cone photoreceptors by generating transgenic mice expressing Nrl under the control of the promotor for the S-opsin gene (BPp-Nrl/WT), which drives expression in a subset of cone photoreceptors. Histological and immunohistochemical analyses as well as ERG recordings indicate a significant decrease in S-opsin-positive cells, suggesting a switch in fate from cone to rod photoreceptors. This finding indicated that Nrl can also modulate cone fate. Furthermore, the researchers found that NRL can bind to the promotor regions of two cone-specific genes, and consequently might directly suppress cone development.

Nrl is therefore essential and sufficient for the determination of rod photoreceptor cell fate. This report clearly demonstrates post-mitotic plasticity in the mammalian retina in vivo, and makes way for the exploitation of engineered rod (or even cone) photoreceptors from retinal stem cells as a therapeutic tool for the treatment of diseases with a degenerating retinal phenotype.