Abstract
The origin of dim-light vision in vertebrates has been attributed to an early duplication event in an ancestral form of pigment called RH, which led to the formation of the RH1 (rhodopsin) and RH2 (rhodopsin-like) opsins. Two sites (122 and 189) are known to critically underlie rhodopsin's sensitivity in low light; however, the evolutionary trajectory leading to dim-light vision remains unclear. We built a large comparative dataset of vertebrate RH1 and RH2 sequences and inferred their respective ancestral protein sequences alongside their common ancestor. Sequence comparisons revealed that known critical sites underpinning RH1 kinetics did not evolve at the origin of this molecule, and instead first evolved in RH (M122E) as well as later in the RH1 of ancestor of jawed vertebrates (P189I). To assess the phenotypic impacts of these changes, we conducted in vitro protein expression and found that the retinal release rate of RH1 became significantly slower following its divergence from RH2, with further deceleration at the origin of jawed vertebrates. We also constructed single RH1 mutants and found that the replacement S295A accounted for the observed difference in retinal release between RH and RH1. Finally, to validate these phenotypic consequences in vivo, we compared zebrafish overexpressing RH, RH1, or the mutant RH-S295A, and recorded increased locomotor rebound following light-intensity reduction in the latter two treatments. We conclude that highly sensitive dim-light vision in vertebrates evolved via three main stages of molecular adaptation, at the origins of RH and RH1, and subsequently in the ancestral RH1 of jawed vertebrates.