The Activation Pathway of Human Rhodopsin in Comparison to Bovine Rhodopsin

Abstract

Rhodopsin, the photoreceptor of rod cells, absorbs light to mediate the first step of vision by activating the G protein transducin (Gt). Several human diseases, such as retinitis pigmentosa or congenital night blindness, are linked to rhodopsin malfunctions. Most of the corresponding in vivo studies and structure-function analyses (e.g. based on protein x-ray crystallography or spectroscopy) have been carried out on murine or bovine rhodopsin. Because these rhodopsins differ at several amino acid positions from human rhodopsin, we conducted a comprehensive spectroscopic characterization of human rhodopsin in combination with molecular dynamics simulations. We show by FTIR and UV-visible difference spectroscopy that the light-induced transformations of the early photointermediates are very similar. Significant differences between the pigments appear with formation of the still inactive Meta I state and the transition to active Meta II. However, the conformation of Meta II and its activity toward the G protein are essentially the same, presumably reflecting the evolutionary pressure under which the active state has developed. Altogether, our results show that although the basic activation pathways of human and bovine rhodopsin are similar, structural deviations exist in the inactive conformation and during receptor activation, even between closely related rhodopsins. These differences between the well studied bovine or murine rhodopsins and human rhodopsin have to be taken into account when the influence of point mutations on the activation pathway of human rhodopsin are investigated using the bovine or murine rhodopsin template sequences.

Keywords: G protein-coupled receptor (GPCR); human; protein dynamic; receptor; rhodopsin; structure-function.

FIGURE 1.

Sequence alignment of human and bovine rhodopsin. Green, differences between both sequences; orange, conserved D(E)RY motif; blue, TM3-TM5 microdomain (Glu-122, Trp-126, and His-211) near the retinal β-ionone ring; magenta, conserved NPXXY(X)_5,6F motif; red, Schiff base lysine (Lys-296); the bold lettering denotes the known retinitis pigmentosa positions.

FIGURE 2.

Homology model and areas of interest. a, homology model of dark state human rhodopsin based on bovine rhodopsin structure (PDB entry 1U19) containing 11-cis-retinal bound via a protonated Schiff base to Lys-296 (red). Black boxes highlight the regions (amino acids 297–300) and (amino acids 194–198) that are in the focus of this study. Differences between human (green) and bovine (gray) amino acids are indicated. b and d, enlarged side view of the two regions shows the details of the structural differences. c and e, sequence alignment segments of bovine and human rhodopsin for residues 294–304 and 190–200 as well WebLogo diagram of all available mammalian rhodopsin sequences. Dashed arrows below e indicate the known secondary structure elements, namely EL2, TM5, and the T₅E₂C region (residues 194–198; TEC).

FIGURE 3.

Comparison of the region around residue 298 (spacefill depiction) of inactive bovine (a) and human (b) rhodopsin. Shown are interior pockets that fit a probe of radius 0.8 Å (orange). The change of Ser-298 in bovine rhodopsin into Ala-298 in human rhodopsin presumably leads to a continuous pore accessible to water molecules (small red spheres) near residue 298 and the retinal SB.

FIGURE 4.

Conformational flexibility of T₅E₂C and adjacent residues analyzed by MD simulations of bovine (a and c) and human (b and d) rhodopsin holoproteins in their dark (a and b) and active (c and d) state. For each system three representative conformations (red, orange, yellow) obtained by cluster analysis are shown. The tube thickness relates to fluctuations of each residue (root mean square fluctuations) within a given cluster. In the simulations of bovine rhodopsin a subset of the EL2/TM5 region shows a flapping motion (solid arrows), whereas in human rhodopsin the whole region shows a sliding (dashed arrows) and, in the active state additionally, a flapping motion (solid arrow) effecting more residues. e and f, hydrogen bonds in the TEC region of bovine and human rhodopsin. e, in bovine rhodopsin a backbone-backbone hydrogen bond is prevalent between Glu-196 and Thr-198, whereas in human rhodopsin a hydrogen bond is observed between Glu-5 and Lys-195 (f).

FIGURE 5.

FTIR difference and double difference spectra of the early intermediates of human (green lines) and bovine (gray lines) rhodopsin. a, Batho spectra were recorded at pH 7.4, 80 K. The double difference (human minus bovine, blue line) indicates marginal deviations at this early stage of the photoreaction. b, Lumi spectra obtained at pH 7.4, 173K; the double difference (red line) reveals more distinct shifts of characteristics bands in the amide-I region.

FIGURE 6.

Coupled FTIR and UV-visible (inset) spectroscopy of human (green lines) and bovine (gray lines) rhodopsin under conditions favoring Meta I (pH 8, 0 °C). a, difference spectra in H₂O. Remarkable are the large deviations between both pigments, indicated by the intense dd-spectrum (blue line). Especially, the most prominent band of bovine Meta I is absent in human spectrum (shown by the strong band at 1663 cm⁻¹ in the dd-spectrum). b, the same measurement in ²H₂O. The large deviations between both pigments remain as demonstrated by the dd-spectrum (orange line). c, FTIR difference spectra of human rhodopsin and TEC chimera (blue line). The T₅E₂C region (amino acids 194–196) of human rhodopsin (LKP) was replaced with the bovine sequence (PHE). Note the reappearance of the Meta I marker band at 1660 cm⁻¹ in the spectrum of the chimera. d, FTIR difference spectra of human rhodopsin and SB chimera (dark red line). The sequence between the Schiff base and the NPXXY(X)_5,6F motif (amino acids 297–300) of human rhodopsin (SAAI) was replaced with that of bovine (TSAV). Only small deviations in the amide-I and amide-II regions are observed in the dd-spectrum.

FIGURE 7.

Coupled FTIR and UV-visible (insets) spectroscopy of the late human (green lines) and bovine (gray lines) intermediates. a, conditions favoring a mixture of Meta states (pH 8, 30 °C). In the double difference spectrum (purple line) the bands are less intense as compared with the Meta I conditions. b, spectra recorded at pH 5, 30 °C indicate a transition from the dark state toward the fully active state Meta II_bH⁺. The weak bands in the double difference (blue line) demonstrate that the active states of human and bovine rhodopsin are largely similar. c, spectra of the complex formation of active Meta II_bH⁺ and high affinity G_tαCT peptide (Meta II·peptide). The double difference (red line) exhibits more pronounced bands than in the case of Meta II_bH⁺. d, in contrast, the peptide binding spectra (pbs, difference spectrum of Meta II·peptide-minus-difference spectrum of Meta II_bH⁺ formation) of both pigments are closely similar. Note that, due to difference formation, the influence of the dark state on these spectra is negligible.

FIGURE 8.

Flash photolysis, titration behavior, and G protein activation assay of human (green lines) and bovine (gray lines) rhodopsin. a, formation of Meta II as monitored by the absorption increase at 380 nm after a 532-nm laser flash under Meta II conditions (pH 5, 20 °C). All signals show a biphasic time course. Meta II formation in human rhodopsin predominantly follows the fast kinetics, whereas in bovine rhodopsin both components are almost equally expressed. b, pH titration curves of the absorbance change at 1744 cm⁻¹ (protonation state of Asp-83 as monitor for the retinal Schiff base) normalized to the acidic and alkaline end point levels derived from a regular Henderson-Hasselbalch fit. c, G protein activation (pH 7, 20 °C) as monitored by change of the intrinsic tryptophan fluorescence. Both proteins show similar G_t activation.