Beyond spectral tuning: human cone visual pigments adopt different transient conformations for chromophore regeneration

Abstract

Human red and green visual pigments are seven transmembrane receptors of cone photoreceptor cells of the retina that mediate color vision. These pigments share a very high degree of homology and have been assumed to feature analogous structural and functional properties. We report on a different regeneration mechanism among red and green cone opsins with retinal analogs using UV-Vis/fluorescence spectroscopic analyses, molecular modeling and site-directed mutagenesis. We find that photoactivated green cone opsin adopts a transient conformation which regenerates via an unprotonated Schiff base linkage with its natural chromophore, whereas red cone opsin forms a typical protonated Schiff base. The chromophore regeneration kinetics is consistent with a secondary retinal uptake by the cone pigments. Overall, our findings reveal, for the first time, structural differences in the photoactivated conformation between red and green cone pigments that may be linked to their molecular evolution, and support the proposal of secondary retinal binding to visual pigments, in addition to binding to the canonical primary site, which may serve as a regulatory mechanism of dark adaptation in the phototransduction process.

Keywords: Color vision; G-protein coupled receptors; Ligand binding; Retinal; Visual phototransduction.

Fig. 1

Regeneration of photoactivated green and red cone pigments with 11CR and 9CR. a, b Immunopurified human recombinant green cone pigment was analyzed by UV–visible spectrophotometry in the dark (1, solid line) and after addition of exogenous ligand, i.e., 2.5-fold of either 11CR (a) or 9CR (b) over opsin (2, dotted line). The pigment was then illuminated for 30 s (>495 nm) and a spectrum was recorded after 30 min (3, dashed line). c, d Red cone opsin was spectroscopically measured in an analogous way to green cone opsin and spectra were measured in the dark (1, solid line), after 11CR (c) or 9CR (d) addition (2, dotted line), and 30 min after illumination (3, dashed line)

Fig. 2

Photoactivated green cone opsin regenerates with 11CR by means of an unprotonated Schiff base linkage. a The UV–vis spectrum of purified green cone pigment was measured in the dark (1, solid line). After addition of 2.5-fold 11CR, it was illuminated for 30 s (λ > 495 nm) (2, dotted line). Then, the sample was split into two aliquots in which one was acidified with 2 N H₂SO₄ immediately (3, dashed line) and the other 30 min (4, dot-dashed line) later. b A difference spectrum was obtained by subtracting the spectrum of the sample acidified after 30 min and the sample acidified immediately after photoactivation. c Green cone pigment, in a fluorimetric micro-cuvette was illuminated for 30 s (λ > 495 nm) after a stable baseline was obtained. Then, 2.5-fold of 11CR to the concentration of the pigment was added immediately after photoactivation and mixed thoroughly. The dotted line represents the fast regeneration process immediately after 11CR addition

Fig. 3

Phe309 regulates Schiff base protonation during 11CR regeneration of photoactivated green cone opsin. a Sequence alignment between the human red and green opsins and bovine rhodopsin (in gray), excluding the N- and C-terminus. Residues highlighted in blue indicate differences between red and green opsins (numbered according to the general scheme [7]). b Molecular model of the inactive red opsin showing the proposed interaction between Tyr309^7.40 (in TM7) and Thr201 (in ECL2). The cartoon of TM7 and the conformation of Tyr309 in the active state are shown with a transparent representation, and the arrow indicates the movement. c 11CR regeneration experiment was carried out with purified F309Y green cone opsin mutant. After measuring a dark-state spectra (1), 11CR was added prior to illumination (2) and the regenerated spectrum was recorded after 30 min (3) showing a band at 530 nm and the sample was acidified (4). c inset. A closer look at the visible region of dark (1) and (2) regenerated spectra. d The difference spectrum, obtained by subtracting the acidified spectrum and the regenerated spectrum, shows a clear difference band at 530 nm

Fig. 4

Regeneration and retinal release kinetics of green and red cone opsin with 9CR. a The absorbance increase at the λ_max in the visible region of each regenerated red/green pigment was plotted and fit with a saturation curve. Chromophore regeneration with 9CR was measured for red and green cone opsins for 30 min. The dark-state spectral values of red and green cone opsin were normalized to have equivalent dark absorbance and the regenerated absorbances were also normalized by the same factor. b Absorbance changes at 380 nm for the regenerated samples were plotted and fit to a single exponential decay curve

Fig. 5

Regeneration and secondary retinal uptake kinetics of red and green cone opsin with 11CR. a 2.5-fold of 11CR to the concentration of red cone opsin was added in the dark and illuminated for 30 s (λ > 495 nm). UV–visible spectra of the sample were obtained every min for 30 min. The λ_max at the visible region of each regenerated spectrum was plotted and fit with a saturation curve. b In the regeneration experiments with 11CR, the regeneration was followed at 380 nm and the absorbance change at this wavelength was monitored for 30 min and the data fit to a sigmoidal curve. c Molecular model of the active red opsin showing the potential coexistence of two retinal molecules one in the primary site (as observed in the rhodopsin structure with PDB id 2X72 [49] in pale yellow) and another in the proposed secondary sites (showing representative poses obtained from flexible docking; in multiple colors). TM helices are represented as cylinders and side-chains of neighboring residues are shown with sticks

Fig. 6

Photoactivated intermediate conformations and 11CR regeneration of red and green cone pigments. a Cone pigments heterologously expressed in mammalian cells were regenerated with 11CR and immunopurified. b Exogenous 11CR (2.5 molar ratio over opsin) was added in the dark, to the immunopurified regenerated sample, and the sample was subsequently illuminated (λ > 495 nm). This causes the isomerization of opsin-bound 11CR to ATR resulting in a protein conformational change that facilitates fast pre-binding of free 11CR to a secondary, lower affinity, retinal binding site followed by chromophore regeneration. c Fast chromophore regeneration would be achieved by translocation and covalent binding of this pre-bound 11CR to the chromophoric binding pocket by displacing ATR which in turn would move out from the protein. Further, the secondary uptake of 11CR by the visual pigment, with a slower kinetics, may be due to the conformational change of the regenerated protein coupled with the reduced availability of bulk 11CR that would synergistically decelerate 11CR binding to the secondary site. d Under our experimental conditions that would lead, at this stage, to a 11CR-regenerated visual pigment harboring an additional 11CR bound to the postulated secondary binding site