Today our manuscript "A system-level, molecular evolutionary analysis of mammalian phototransduction" was published in BMC Evolutionary Biology!
In this article we investigated whether the structure of protein-protein interactions in the visual phototransduction system accounts for some of the variation in the evolutionary rates seen in the underlying genes. Visual phototransduction is the first step in vision. It is the biochemical process that converts a visual stimulus into a neuronal response. Two phototransduction pathways exist in vertebrates, corresponding to the rod (low-light vision) and cone (bright, color vision) photoreceptor cells.
In order to investigate the effects of system structure on evolutionary rates, we hand-constructed a network containing all of the known protein interactions involved in phototransduction in both the rod and cone photoreceptor cells. Evolutionary rates were estimated on a phylogenetic tree with 12 mammalian species. Using this, we found that proteins that are more topographically central, using the graph-theoretic closeness centrality measurement, are subject to stronger evolutionary constraint, manifested as lower ratios of non-synonymous to synonymous substitution rates. We also found that different protein classes tend to be more strongly conserved than others; in particular, G-proteins have undergonoe stronger constraint while ion channels and pumps have experienced relative relaxation. This indicates that selection on complex phenotypes is manifested in a non-random manner on the underlying genes.
Additionally, we describe two detected events of positive selection. One is in the gene OPN1SW, encoding the blue/short-wave cone opsin in the phylogenetic branch leading to the Old World apes. By homology, we predict the specific locus under positive selection to have a putative role in spectral tuning of the receptor, however this is called into doubt by experimental studies. The other gene to be found to have been under positive selection is SLC24A1, the rod-specific sodium-calcium-potassium ion exchanger in the branch leading to the rodents. We were unable to make predictions about the functional relevance of this event.