Evolution of Tandemly-Replicated Opsin Genes: Molecular Models That Predict Spectral Shifts

A snapshot from an atomistic molecular simulation of a vertebrate cone visual pigment (green – opsin, red – retinal, magenta – palmitoyl moiety) embedded in an explicit lipid bilayer (yellow & purple – lipid head group, grey – lipid tail) and surrounded by a layer of water molecules (cyan).

Group members: Jagdish Patel, Marty Ytreberg

Originated: Summer 2016

Description:

This working group on an as-needed basis, to discuss progress and approaches for modeling a set of tandemly-replicated cone opsins from the zebrafish, medaka, cichlid, and guppy. They selected these RH2-type cone opsins based upon the availability of a reasonable (reasonably homologous) model from an X-ray crystallography structure, and reliable information from the literature regarding sequence data and spectral sensitivity data.

Docking models attempted on these opsins did not predict spectral shift.

Standard all-atom molecular dynamics simulations were then performed to find out if these predict spectral shift. Challenges that were overcome included modeling of the covalently attached chromophore, and the influence of the surrounding lipid bilayer. In addition, we included three “ancestral” RH2 opsins with known spectral sensitivities to increase the scope of the study to seven RH2 opsins.

The molecular dynamics simulations were carried out in December 2016. Visualization of the distinct fluctuations and analysis of internal parameters associated with the chromophore’s heavy atoms suggested that some of these features may predict spectral shift, and quantitative/statistical approaches to determine if this is the case are underway.

RH2 cone opsins of medaka, guppy, and cichlid were added to the analysis to increase the robustness of our approach, providing a total of 14 distinct RH2 sequences. Two of the output metrics (RMSF and the C7-C6-C5-C18 Torsion) were identified as most highly predictive of “blue” vs. “green” peak spectral sensitivity. A mathematical model was generated, using these two parameters, which predicts spectral sensitivity with a correlative R² of 0.94 (!!!!).

The manuscript was submitted in fall 2017 – to Cell, then PNAS, then PLOS Biology, and then eLife. The editorial staff at each journal considered it more appropriate for a specialty journal. In December 2017 we submitted it to PLOS Computational Biology. It was favorably reviewed, and after minor revisions, accepted. It is now published, with an epub date of January 24, 2018.

Publications:

Predicting peak spectral sensitivities of vertebrate cone visual pigments using atomistic molecular simulations

Patel JS, Brown CJ, Ytreberg FM, Stenkamp DL (2018) Predicting peak spectral sensitivities of vertebrate cone visual pigments using atomistic molecular simulations. PLoS Comput Biol 14(1): e1005974. https://doi.org/10.1371/journal.pcbi.1005974