Most vertebrate retinas are composed of two types of photoreceptors: rods, for dim light vision, and cones, for color vision under brighter conditions. However, several interesting exceptions to this duplex organization are seen in the squamate reptiles, a scaly group that contains lizards and snakes. In 1934, Gordon Walls, a post-doc studying visual physiology at the University of Michigan, put forward an inventive theory of visual evolution that sought to explain these exceptions.
Walls observed that some lizards and snakes lacked rods entirely – they possessed a pure-cone retina. These reptiles lived a bright, diurnal lifestyle, and may have had less need for low-light rod vision. However, many evolutionary relatives of the pure-cone reptiles possessed the exact opposite – pure-rod retinas. These reptiles had adapted to a dim, nocturnal lifestyle. Finally, some squamates had intermediate retinas with eclectic mixes of rods, cones and, occasionally, photoreceptors that resembled both rods and cones.
Based on this diversity, and the evolutionary relationships of pure-rod and pure-cone reptiles, Walls developed a ‘transmutation’ theory of photoreceptors. This theory proposed that rods and cones were not fully independent, but could swap classes through evolution. Unfortunately, Walls had limited means to test his theory, because he lacked a reliable technique to identify the visual pigments, which serve as molecular signatures of the different photoreceptor types.
Two recent papers have revisited Walls’ transmutation theory and found evidence that he was basically on the right track. In the first, Simões and colleagues, writing in the Proceedings of the Royal Society B, used microspectrophotometry and genetic sequencing to identify the visual pigments expressed by a diverse group of snakes – some with all-cone retinas, some with all-rod retinas and others with intermediate photoreceptor types. They found that all-cone retinas still expressed visual pigment genes characteristic of rods. The opposite was also true: all-rod retinas expressed cone visual pigments.
The second study, by Schott and colleagues, writing in PNAS, investigated the garter snake, which possesses an all-cone retina. They looked closely at the ultrastructure and gene expression of these cones, and found that they exhibit many rod-like features, and are thus likely to have evolved from rods. Together, these two studies support Walls’ original theory that rods can transmutate into cones, and cones can transmutate into rods.
But things may be even more complicated than Walls originally predicted. By combining gene expression studies with phylogenetics, Simões and colleagues suggest that both cone-like rods and rod-like cones have evolved independently at least twice. Indeed, this work, combined with previous studies of the gecko eye, indicates that some reptilian retinas have undergone repeated, radical shifts in photoreceptor composition. This is in contrast to the retinas of fishes, birds and mammals, where the rod/cone dichotomy is thought to have remained relatively static for millions of years.
Why are photoreceptors so evolutionarily fluid in reptiles compared with other vertebrates? One possibility is that squamate visual pigments have somehow been decoupled from photoreceptor morphology, allowing rod pigments to be expressed in cones, and vice versa. Future studies could use transcriptional profiling of intermediate photoreceptors to investigate the genetic mechanisms of transmutation. It remains to be seen whether there is just one or multiple different ways for rods and cones to swap identity. Although many people shiver at the thought, a deep gaze into the eyes of serpents could help us understand how our own retinas ended up with their characteristic duplex structure.
