Adaptive genomic evolution of opsins reveals that early mammals flourished in nocturnal environments

Abstract

Background: Based on evolutionary patterns of the vertebrate eye, Walls (1942) hypothesized that early placental mammals evolved primarily in nocturnal habitats. However, not only Eutheria, but all mammals show photic characteristics (i.e. dichromatic vision, rod-dominated retina) suggestive of a scotopic eye design.

Results: Here, we used integrative comparative genomic and phylogenetic methodologies employing the photoreceptive opsin gene family in 154 mammals to test the likelihood of a nocturnal period in the emergence of all mammals. We showed that mammals possess genomic patterns concordant with a nocturnal ancestry. The loss of the RH2, VA, PARA, PARIE and OPN4x opsins in all mammals led us to advance a probable and most-parsimonious hypothesis of a global nocturnal bottleneck that explains the loss of these genes in the emerging lineage (> > 215.5 million years ago). In addition, ancestral character reconstruction analyses provided strong evidence that ancestral mammals possessed a nocturnal lifestyle, ultra-violet-sensitive vision, low visual acuity and low orbit convergence (i.e. panoramic vision).

Conclusions: Overall, this study provides insight into the evolutionary history of the mammalian eye while discussing important ecological aspects of the photic paleo-environments ancestral mammals have occupied.

Keywords: Mammals; Nocturnal bottleneck; Nocturnal lifestyle; Opsins; Panoramic vision; Ultra-violet sensitive vision; Visual acuity.

Fig. 1

Opsin syntenic patterns in Tetrapoda. Assembled genomes were inspected for the opsin genes (yellow boxes) and their neighbouring genes (grey shaded boxes) on both sides. Segmented lines indicate genomic regions with no homologous representative between the analysed genomes. Shaded rectangles express the expected location of the analysed genes in the tetrapod genomes, pointing out the potential and confirmed losses of opsins (i.e. lack of a homologous gene in a homologous region). Missing opsin genes are identified by a red x. Species index: 1. Human (Homo sapiens), 2. Opossum (Monodelphis domestica), 3. Platypus (Ornitorhynchus anatinus), 4. Zebrafinch (Taeniopygia guttata), 5. Carolina anole (Anolis carolinensis), 6. Xenopus (Xenopus laevis), 7. Chicken (Gallus gallus), 8. Chinese turtle (Pelodiscus sinensis), 9. Tasmanian devil (Sarcophilus harrisii), 10. Chimpanzee (Pan troglodytes), 11. Flycatcher (Ficedula albicollis) and 12. Orangutan (Pongo abelii)

Fig. 2

Selective signatures in the mammalian OPN1sw1 opsin. Schematic view of the bovine rhodopsin highlighting the visual opsins spectral tuning sites ([19, 20]): OPN1lw (red), RH2 (green), OPN1sw2 (blue), OPN1sw1 (violet) and RH1 (black). Shared spectral tuning sites are indicated by shared colours. Residues highlighted in red experienced positive selection in mammals. The bar plot depicts the amino acid composition of the OPN1sw1 93 spectral tuning site accounting for the nocturnal and diurnal species and the eutherian orders (or infraclass in the case of marsupials) in which these amino acids were found

Fig. 3

Evolution of opsins and photic-related characters in mammals. Schematic view of the global nocturnal bottleneck hypothesis, linking both the results from the synteny and the ancestral reconstruction analyses: a nocturnal period (represented in grey) is assumed to affect both the mammalian ancestor and the emerging lineages. Global and lineage-specific losses of opsins are indicated with a red cross; opsin losses that were not supported by a consistent synteny are marked with an asterisk (*). Ancestral reconstructions of the activity pattern are represented in pie charts, each slice representing the probability of each state (nocturnal, diurnal and cathemeral). Ancestral reconstructions of the violet and ultra-violet sensitive (VS and UVS) vision are represented by a violet or a black circle, respectively, near the OPN1sw1 opsin (sw1 was used to perform the inferences). Uncertainties about the opsin presence/absence are marked with a double asterisk (**): UVS vision in the mammalian ancestor node is only hypothetical (it could not be determined because the sw1 is not present in monotremes); the therapsid opsin profile is not known since genetic data is not available for therapsids. The potential vision of ancestral mammals was inferred considering both the conopsin content of each node and their spectral sensitivities [blue for OPN1sw2 (sw2), red for OPN1lw (lw) and saturated pink for UV sw1]. Ancestral inferences of the orbit convergence (OC, degrees) and visual acuity (VA, cycles per degree) are indicated in the corresponding nodes. A reconstruction of a skull with an orbital convergence angle of 46° (which correspond to the inferred angle in the node of mammals) is included. The Shennongjia virgin forest (https://commons.wikimedia.org/wiki/File:Shennongjia_virgin_forest.jpg) and the Pristerognathus (https://commons.wikimedia.org/wiki/File:Pristeroognathus_DB.jpg) are licensed under the Attribution-ShareAlike 3.0 Unported license (the license terms can be found on the following link: https://creativecommons.org/licenses/by-sa/3.0/)