Dimerization deficiency of enigmatic retinitis pigmentosa-linked rhodopsin mutants

Abstract

Retinitis pigmentosa (RP) is a blinding disease often associated with mutations in rhodopsin, a light-sensing G protein-coupled receptor and phospholipid scramblase. Most RP-associated mutations affect rhodopsin's activity or transport to disc membranes. Intriguingly, some mutations produce apparently normal rhodopsins that nevertheless cause disease. Here we show that three such enigmatic mutations-F45L, V209M and F220C-yield fully functional visual pigments that bind the 11-cis retinal chromophore, activate the G protein transducin, traffic to the light-sensitive photoreceptor compartment and scramble phospholipids. However, tests of scramblase activity show that unlike wild-type rhodopsin that functionally reconstitutes into liposomes as dimers or multimers, F45L, V209M and F220C rhodopsins behave as monomers. This result was confirmed in pull-down experiments. Our data suggest that the photoreceptor pathology associated with expression of these enigmatic RP-associated pigments arises from their unexpected inability to dimerize via transmembrane helices 1 and 5.

Figure 1. Rhodopsin structure showing sites of RP mutations.

Three views of rhodopsin (PDB accession 4J4Q) are shown in the context of a lipid bilayer (light grey slab); TM helices 1 (TM1, light blue) and 5 (TM5, dark blue) are highlighted. The amino acids that are affected in the RP mutants under study are shown as space-filling CPK representations (F45 (red), V209 (green) and F220 (gold)). Ballesteros–Weinstein numbers corresponding to the affected residues are as follows: 1.40 (F45), 5.44 (V209) and 5.55 (F220).

Figure 2. Expression of RP mutants.

(a) SDS–PAGE analysis of purified WT opsin and RP-associated mutants. The gel was visualized by Coomassie staining. BSA (1 μg–200 ng) was run alongside and used for quantification purposes. (b) Fluorescence micrographs of non-permeabilized COS-7 cells expressing WT and mutant opsin as indicated. Cell surface opsin (green) was detected with the Ret-P1 antibody that recognizes an N-terminal (extracellular) epitope; nuclei (red) were stained with 4,6-diamidino-2-phenylindole (DAPI). Scale bar, 20 μm. (c) Rhodopsin immunostaining (green) in cross-sections of rhodopsin knockout mouse retinas transfected with WT, F45L, V209M and F220C rhodopsin constructs, respectively. Nuclei are stained with Hoechst (blue). Scale bar, 10 μm.

Figure 3. RP mutants are functional visual pigments.

(a) Ultraviolet-visible spectra of purified WT rhodopsin and RP-linked rhodopsin mutants. The solid traces correspond to spectra measured in the dark, whereas the dashed traces represent spectra obtained after illuminating the samples for 90 s with >515 nm light. (b) G protein activation assay. The traces show the time course of intrinsic fluorescence of the G protein transducin. Fluorescence increases on GDP–GTP exchange in the Gα subunit, corresponding to its activation. The sample containing transducin and a catalytic amount of rhodopsin was illuminated for 60 s, starting at t=0 s. Nucleotide exchange was initiated by addition of GTPγS (arrrow). At t=400 s, an excess amount of light-activated WT rhodopsin was added to test for the completeness of transducin activation. (c) Transducin fluorescence measurement as in b, conducted before and after light activation of rhodopsin (arrow hν). After adding GTPγS at t=60 s, the intrinsic fluorescence of transducin did not change due to lack of active rhodopsin. At t=200 s, the sample was illuminated (arrow) to activate rhodopsin and initiate GDP–GTP exchange. (d) Retinal release. The traces show the intrinsic fluorescence increase on retinal Schiff base hydrolysis and retinal release from light-activated rhodopsin. The reaction was initiated by illumination of rhodopsin at t=90 s. At t=3,100 s, NH₂OH was added to hydrolyse the remaining retinal Schiff base, thereby facilitating the release of any remaining retinal. For b–d, the fluorescence was normalized to the level at the start of the recording and the traces in each panel are vertically displaced for clarity (the displacements are 0.02 units in b, 0.05 units in c and 0.2 units in d).

Figure 4. Scramblase activity of RP mutants.

(a) Schematic representation of the scramblase activity assay. Details may be found in the text and Supplementary Note 1. (b) Fluorescence traces obtained on adding dithionite to NBD-PC-containing protein-free liposomes (grey trace) and proteoliposomes reconstituted with different opsins (coloured traces) at a high PPR (∼1 g mol⁻¹). Each trace represents the mean of three independent experiments; the variation between experiments was negligible and is contained within the thickness of the lines. (c) Scramblase activity as a function of the amount of WT opsin reconstituted. Scramblase assays were performed for 400 s with vesicles reconstituted at different PPRs (PPR*, in units of grams of protein per mole of phospholipid, obtained by scaling the measured PPR as described in ‘Methods'). The plot represents the dependency of p(≥1) scramblase (the probability of a vesicle having at least one scramblase) on the PPR*. The line represents the data fit calculated as described in ‘Methods' and in Supplementary Notes 2 and 3. The data are from five independent protein preparations. (d) As in c, for the F45L mutant. The trace for the WT sample is from c. (e) As in d, for the V209M mutant. (f) As in d, for the F220C mutant. Data shown in d–f are from three independent protein preparations in each case. (g) Molar mass (M) of the functionally reconstituted scramblase (mean±s.e., corresponding to Supplementary Table 2) derived from fits of the p(≥1) scramblase versus PPR* data for WT opsin, Ops* and the RP-associated mutants. Dashed lines indicate the molecular weights of opsin monomer and dimer. (h) Schematic illustration of reconstitution modes of WT and RP mutant opsins. All opsins are purified in DDM as monomers. Opsin-DDM micelles are combined with DDM-destabilized vesicles and detergent withdrawal is initiated by adding SM2 BioBeads. When detergent is withdrawn WT opsin dimerizes (or multimerizes), while still within DDM micelles. As DDM continues to be removed from the system, the dimers insert into the available vesicles. For opsin RP mutants, dimerization does not occur and the proteins enter vesicles directly, as monomers.

Figure 5. Pull-down experiments.

(a) Schematic representation of the fusion proteins. (b) Protocol for the pull-down experiment. (c) Purified Ops-FG-SNAP and Ops-FG proteins are shown in the SDS–PAGE/Coomassie-stained gel panels labelled ‘Purified proteins' (5 μl samples were loaded from ∼100 ng μl⁻¹ stocks; it is noteworthy that the Ops-FG-SNAP constructs stain more intensely than the Ops-FG proteins). The pull-down experiment was carried out as described in ‘Methods' and as outlined in b. Ops-FG-SNAP protein (2 μg) corresponding to either WT or RP-associated mutants was covalently pre-bound to SNAP capture resin in 0.1% (w/v) DDM and the resin was washed before adding cognate Ops-FG protein (2 μg in 0.1% (w/v) DDM). The mixture was treated with Biobeads as described in ‘Methods' before removing the supernatant, washing the resin and analysing the resulting pulldowns by SDS–PAGE/Coomassie staining after elution from the resin with Laemmli buffer (panel labelled ‘Pull-down'). The pull-down gel is representative of four independent experiments. Entire gels corresponding to the panels ‘Purified proteins' and ‘Pull-down' are shown in Supplementary Fig. 6C,D.

Figure 6. Rhodopsin dimer models.

(a,d) Cross-sectional view of two distinct dimer interfaces (DI-1 and DI-2) in rhodopsin. The numbered circles correspond to individual helices in the TM helical bundle of each uniquely coloured monomer. (b,c) Top and side views of the interaction interface DI-1 formed by TM1-helix 8 in the opsin crystal structure PDB accession 3CAP. F45 is shown as a CPK model in red. (e,f) Top and side views of the interaction interface DI-2 in which TM5 plays a role. This structural model, obtained by docking the opsin structure (PDB accession 4J4Q) using the HADDOCK webserver, corresponds to cluster-6 in Supplementary Fig. 9. V209 (green) and F220 (gold) are shown as CPK models.