Mice used in this study were maintained on a 12-h light/12-h dark cycle. M-opsin knockout (Opn1mw^−/−) mice have been previously described (Zhang et al., 2017). We have demonstrated previously that there are no differences in cone morphology, retinal structure, or scotopic/photopic ERG responses between Opn1mw^−/− hemizygous males and homozygous females, so both were used in this study (Zhang et al., 2017; Deng et al., 2018). All experimental procedures involving animals in this study were approved and conducted in strict accordance with relevant guidelines and regulations by the Institutional Animal Care and Use Committee at University of Florida and West Virginia University, the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the National Institutes of Health.

Human OPN1LW missense mutations were generated using a Q5^® site-directed mutagenesis kit (New England Biolabs, Ipswitch, MA, USA) following a manufacturer provided protocol and a previously characterized plasmid used by the lab that contains OPN1LW cDNA with an HA tag fused in-frame at the C-terminus. After mutagenesis, each mutant was then cloned into an AAV vector containing a cone-specific PR2.1 promoter (Mancuso et al., 2009) and sequences were confirmed by Sanger sequencing. All mutant vectors were packaged in AAV serotype 8 Y733F by transfection of HEK293 cells and purified according to previously published methods (Zolotukhin et al., 2002).

Before injection, eyes of 1-month-old Opn1mw^−/− mice were dilated using Tropi-Phen drops (Pine Pharmaceuticals, Tonawanda, NY) approximately 15–30 min prior to administration of anesthesia by intramuscular (IM) injection of ketamine (80 mg/kg) and xylazine (10 mg/kg) in sterile phosphate buffered saline (PBS). Trans-corneal subretinal injections were then carried out using a 33-gauge blunt end needle attached to a 5 mL Hamilton syringe after an entering puncture was placed with a 25-gauge needle at the edge of the cornea. Viral vector mixed with fluorescein dye (0.1% final concentration, 1 μL volume of mixture) was delivered into the subretinal space as described previously (Pang et al., 2006, 2008). Eyes were treated post-injection with Neomycin/Polymixin B Sulfates/Bacitracin Zinc ophthalmic ointment (Bausch & Lomb, Inc., Tampa, FL). An intraperitoneal (IP) injection of antisedan (Orion Corporation, Espoo, Finland) was then given to reverse anesthesia.

Following humane euthanasia of mice, retinas from uninjected and mutant opsin injected Opn1mw^−/− mice, as well as from wild-type (WT) controls (three of each), were carefully dissected and flash frozen on dry ice. Each retinal sample was then homogenized by sonication in 300 μL of protein extraction buffer [0.23 M sucrose, 2 mM EDTA, 5 mM Tris–HCl (pH 7.5)] containing 1X cOmplete^™ protease inhibitor cocktail (Millipore Sigma, Burlington, MA, USA). Supernatant was collected after centrifugation and a Pierce^™ BCA Protein Assay Kit was used to determine protein concentration. Retinal lysate was then divided into aliquots containing equal amounts of total protein (40 μg) and subjected to SDS-PAGE gel electrophoresis on a 4–15% Mini-PROTEAN^® TGX^™ Precast Gels (Bio-Rad, Hercules, CA, USA). After SDS-PAGE gel electrophoresis, proteins were transferred to an Immobilon-FL membrane (Millipore Sigma, Burlington, MA, USA) and probed with anti-red/green opsin (Millipore, AB5405, 1:1000 dilution) antibody. Anti-alpha tubulin (cat # T5168, 1:5000; Millipore Sigma, Burlington, MA, USA) served as a loading control. Immunoblot was visualized with the Odyssey Infrared Fluorescence Imaging System (LICOR Biosciences, Lincoln, NE, USA).

Preceding ERG testing, animals were light-adapted to inhibit rod responses and pupils were dilated with Tropi-Phen drops (Pine Pharmaceuticals, Tonawanda, NY). Photopic ERGs were carried out 4-weeks post-injection using a UTAS Visual Diagnostic System with a Big Shot Ganzfeld dome (LKC Technologies, Gaithersburg, MD, USA). For ERG testing, all mice were anesthetized with ketamine (80 mg/kg) and xylazine (10 mg/kg) by IM injection. Cone ERGs from mutant opsin injected eyes were recorded with green channel middle wavelength light (530 nm) stimulation followed by red channel long wavelength light (630 nm), both at various intensities (−0.6, 0.4, and 1.4 log cd s/m²). Both M-cone (green) and L-cone (red) ERGs produced similar ERG results. Red light ERG recordings taken at 1.4 log cd s/m² were used for statistical analysis and for plotting data. S-cone ERGs (360 nm) were recorded at intensities of −0.6 and 0.4 log cd s/m². ERG data are presented as average ± SD (n = 8 for each mutant injected group). Flat ERGs lacking a measurable a-or b-wave were assigned a b-wave maximum amplitude of 1 μV. Statistical analysis of ERG data comparing mutant injected eyes vs. untreated and WT controls is performed by one-way ANOVA and Dunnett’s multiple comparisons test (GraphPad Prism). Statistical significance is defined as a p value of <0.01.

After euthanasia, eyes were marked at the dorsal position using the Change-a-Tip Deluxe Cautery tool (Braintree Scientific, Braintree, MA). Next, eyes were enucleated and a 16-gauge needle was used to introduce an opening along the edge of the cornea before being incubated in 4% paraformaldehyde (PFA) in 1X PBS for 2 h at room temperature. The cornea was dissected carefully away from the eye and the lens was removed. Next, eyecups were rinsed in 1X PBS and placed in 30% sucrose overnight at 4°C for cryoprotection before being embedded in Tissue TEK O.C.T. compound (Sakura Finetek USA, Inc., Torrance, CA) and frozen at −80°C. Retinas were sectioned from dorsal to ventral at 12 μm thickness. For immunohistochemistry (IHC), retinal cross-sections were rinsed in 1X PBS and then blocked [3% Bovine serum albumin (BSA), 0.3% Triton X-100 in 1X PBS] for 1 h at room temperature before incubating in primary antibody overnight at 4°C. Anti-L/M-opsin (MilliporeSigma, AB5405, 1:1000), anti-S-opsin (MilliporeSigma AB5407, 1:1000 dilution), anti-HA (Roche, rat monoclonal, clone 3F10, 1:200), anti-PDE6α’ (ABgene, AP9728c, 1:200), and anti-cone transducin γ subunit (gift from Dr. Vadim Arshavsky’s lab, Duke University) antibodies were used. Following primary incubation, sections were washed and incubated with Alexa-594 or Alexa-488 IgG secondary antibodies (molecular Probes, Eugene OR, 1:500) at room temperature for 2 h. Sections were washed after secondary and subsequently mounted with Vectashield Mounting Medium for Fluorescence (H-1400, Vector lab, Inc. Burlingame, CA, USA) and cover slipped. A Leica Fluorescence Microscope LAS X Widefield System was used to image the stained sections.

Prior to enucleation, a Change-a-Tip Deluxe Cautery tool (Braintree Scientific, Braintree, MA) was used to mark eyes on the cornea at the dorsal position. A 16-gauge needle was used to poke a hole along the edge of the cornea before the enucleated eye was introduced into 4% PFA in 1X PBS. The cornea and lens were removed following a 2-h incubation at room temperature and a radial cut was done to mark the dorsal position. After dissecting the neural retina from the eyecup, retinal whole mounts were blocked in 3% BSA containing 0.3% Triton-X-100 in 1X PBS overnight at 4°C. On the next day, retinal whole mounts were placed in primary antibody containing biotinylated peanut agglutinin (PNA; Vector Laboratories, Burlingame, CA, 1:500) and anti-S-opsin antibody in 1% BSA in 1X PBS overnight at 4°C. Whole mounts were subsequently washed in 0.05% Tween-20 in 1X PBS before being incubated in secondary antibody containing Fluorescein Avidin D (Vector Laboratories, Burlingame, CA, 1:500) and anti-rabbit IgG Alexa-594 (1:500) in 1X PBS. Finally, whole mount retinas were washed and ventral, temporal, and nasal cuts were made before being coverslipped using ProLong Gold Antifade Mountant (Thermo Fisher, Waltham, MA). Whole mounts were imaged using a Nikon C2 confocal microscope and processed using FIJI software (Schindelin et al., 2012). Four images were captured (two ventral, two dorsal) for each retinal whole mount, and PNA-positive cells from K82E (n = 6 eyes), P187S (n = 8), and M273K (n = 6) images were counted within an area equivalent of 0.04 mm² per image using the particle analysis tool in ImageJ. Counts were averaged for the dorsal and ventral regions and the SD was calculated. Statistical analysis was performed by two-way ANOVA with Tukey’s multiple comparisons test to compare differences between each mutant group for dorsal or ventral regions (*p ≤ 0.05, **p < 0.002, ***p < 0.001).

All data are presented as the mean ± SEM unless otherwise noted. Figure legends contain details about sample size. All analyses were carried out using GraphPad Prism 9 software by ordinary 1-way ANOVA with Dunnett post hoc test (comparing two groups) or by two-way ANOVA with Dunnett post hoc test for multiple comparisons unless noted differently. Significance is indicated as *p ≤ 0.05, **p < 0.002, or ***p < 0.001.

The cDNA of each OPN1LW cone opsin missense mutant (K82E, P187S, and M273K) fused with a C-terminal HA tag was cloned into an AAV vector under a cone-specific PR2.1 promoter and packaged into an AAV8-Y733F capsid. The amino acid position of these mutations in human cone opsin are illustrated in Figures 1A,B (Stenkamp et al., 2002). Each OPN1LW mutant AAV vector was delivered by subretinal injection into one eye of Opn1mw^−/− mice at 1 month of age, with the contralateral eye serving as an uninjected control. Retinas were collected 30 days post-injection and OPN1LW protein levels were analyzed by western blot using an anti-red/green opsin antibody that binds the C-terminus of human L/M-opsin (Jones et al., 2011). We were able to identify OPN1LW-K82E expression in injected eyes. However, expression of OPN1LW-P187S was minimal and we were unable to detect expression of OPN1LW-M273K (Figure 1C).

To determine the capability of these OPN1LW missense mutants in generating responses to light, we performed long- (630 nm) wavelength cone ERGs on treated Opn1mw^−/− mice at 1-month post-injection. Uninjected Opn1mw^−/− eyes do not exhibit an L-or M-cone mediated light response but have a normal S-cone ERG (Zhang et al., 2017). We were unable to detect green-or red-light ERG responses in Opn1mw^−/− eyes expressing mutant P187S or M273K. However, K82E-injected eyes generated an average b-wave amplitude of 54.4 ± 20.2 μV (average ± SD, n = 8), demonstrating a visual response that is significantly lower than WT controls (121.6 ± 8.3 μV, n = 8, p < 0.0001), but significantly higher than the flat ERGs from untreated contralateral control eyes (p < 0.0001) (Figures 2A,B). Together, these results suggest that the only OPN1LW mutant in this study that is capable of generating responses to light is K82E, but its efficiency to do so is reduced in comparison to WT OPN1LW.

We next determined the subcellular localization of these OPN1LW mutants in injected Opn1mw^−/− eyes by immunohistochemistry (IHC). Previously, we have demonstrated dorsal Opn1mw^−/− cones exhibit shortened outer segments, but remain viable, while ventral cones retain normal cone structure due to the presence of S-opsin (Deng et al., 2018). In addition to each mutant, we also examined the expression and localization of two endogenous COS-specific proteins involved in phototransduction, phosphodiesterase α subunit (PDE6α’) and cone transducin γ subunit (GNGT2), in both the dorsal and ventral regions of the injected Opn1mw^−/− retinas.

Localization of K82E was detected mainly within the COS, both in the dorsal and ventral retina (Figure 3). GNGT2 and PDE6α’ were localized normally to the COS within the ventral retinas that express S-opsin. However, in the dorsal retina, both proteins were partially mislocalized to the inner segment and endoplasmic reticulum (ER). Previous studies have demonstrated that binding of 11-cis-retinal is required for proper trafficking of cone opsins to the COS (Zhang et al., 2008). Furthermore, trafficking of downstream phototransduction proteins cone transducin and cone PDE6 co-traffic with 11-cis-retinal bound cone opsin. Therefore, these results suggest that while K82E mutants properly bind chromophore and are correctly transported to the COS, binding and trafficking of GNGT2 and PDE6α’ to the OPN1LW mutant is inhibited as these proteins are trapped in the inner segment. The mislocalization of these two proteins explains the partial cone ERG rescue observed in K82E-injected eyes, as the cones lack the proper levels of phototransduction proteins in the light-sensing outer segments.

We were unable to detect expression of P187S by IHC in both the dorsal and ventral region of injected retinas. In the dorsal region that lacks endogenous M-opsin or AAV-mediated P187S, GNGT2 and PDE6α’ were either undetectable or severely reduced, but appeared to be localized normally in the ventral retina where S-opsin is present (Figure 4). These data suggest that P187S is misfolded and fails to exit the perinuclear ER, preventing co-trafficking of GNGT2 and PDE6α’. Interestingly, while we were unable to detect M273K by western blot analysis, IHC staining showed that M273K was localized to the cone inner segment and ER in both the dorsal and ventral areas of injected Opn1mw^−/− retinas, suggesting that M273K is likely partially misfolded, which may interfere with efficient 11-cis-retinal binding (Figure 5). PDE6α’ and GNGT2 were barely detectible in the dorsal retinas, but were expressed normally and localized correctly to the COS in cones expressing S-opsin within the ventral region of the same retina.

To determine if expression of the OPN1LW mutants has a toxic effect on cone viability, we next examined the retinas of WT eyes injected with each mutant. Cone viability was assessed by peanut agglutinin (PNA) staining by counting the number of PNA⁺ cells in both the dorsal and ventral retina and compared to the contralateral, uninjected eyes at 1-month post-injection (Figures 6A–C). PNA specifically binds to the extracellular glycoprotein matrix of cone outer and inner segment sheaths and is an indicator of cone viability (Blanks and Johnson, 1983). All three mutants significantly reduced the number of viable cones in the dorsal retina compared to untreated WT controls to a various extent, ranging from a 31% to a 74% decrease, but only P187S (31.3% decrease) also showed a significant reduction in viable cones within the ventral retina (Figure 6C). Short- (360 nm) wavelength ERGs were also performed on WT injected eyes to determine the toxicity of these various L-opsin mutants on S-cone function. Measurements of S-cone ERGs in each mutant treated eye showed that expression of K82E, P187S, or M273K did not result in a significant ERG reduction compared to WT eyes (Figure 6D), suggesting these mutant proteins do not alter cone viability but are partially or severely misfolded.