Splice Variant of Retinal G-Protein-Coupled Receptor Deletion-Mediated Dysregulation of Autophagy Increases the Susceptibility to Age-Related Macular Degeneration-Like Defects

Age-related macular degeneration (AMD) is one of the most common causes of severe, irreversible vision loss. Dysregulation of the inflammatory system, complement pathway, lipid signaling, angiogenesis, and extracellular matrix are involved in the development of AMD [1]. It has been estimated that 288 million people worldwide will have AMD by 2040 [2]. Yellowish deposits, which are called drusen, develop in the macula, and abnormalities occur in the retinal pigment epithelium (RPE) in the early stage of AMD. Late-stage AMD progresses to geographic atrophy (GA; dry AMD) or choroidal neovascularization (wet AMD) [3]. Long-term administration of vascular endothelial growth factor inhibitors can help manage choroidal neovascularization [4]. However, there is no effective treatment for drusen in the early stages of AMD and GA in dry AMD. Understanding the pathogenesis of dry AMD would help in the development of treatments.

Dry AMD is pathologically characterized by RPE dysfunction and a consequent decrease in number of rod photoreceptors [1]. Although GA affects most cell types in the retina [5], the mechanism of retinal degeneration is poorly understood. Autophagic dysfunction is associated with general aging [6], and impaired autophagy has been found in several age-related neurodegenerative disorders, including Alzheimer’s and Parkinson’s diseases [7, 8]. Autophagy is a catabolic process essential for the maintenance of homeostasis through active “self-consumption,” in which toxic waste is sequestered and then degraded by lysosomes [9]. Autophagy is also critical for the removal of damaged organelles and repairing cell functions. The identification of autophagy-related proteins (ATG) and autophagic lysosome reformation have been fundamental advances in understanding autophagy. Recently, it has been shown that impaired autophagy is an important factor in the pathogenesis of dry AMD [10].

In our previous research, we found that the human gene encoding retinal G-protein-coupled receptor (RGR) can generate an exon-skipping splice variant, termed RGR-opsin (RGR-d), which is present in small and large drusen as a persistent component [11]. In particular, older mice with mutant RGR-d showed a gradual ocular pathology with degenerative changes in retinal structures (photoreceptors, RPE, and choriocapillaris) that presented as AMD-like defects [12]. However, it remains unclear how RGR-d caused the changes in RPE function. Based on these studies, we hypothesized that RGR-d may impair autophagy and cause RPE distortion, and this dysfunction may cause the subsequent effects on the photoreceptors, leading to retinal dysfunction and degeneration. Accordingly, we investigated the involvement of RGR-d in the dysfunction and dysregulation of autophagy in RPE cells using a combination of in vivo and in vitro experiments. We explored the changes in the signaling pathways early in the pathogenesis of dry AMD. We aimed to identify treatment targets before severe progression occurs, such as in cases of GA.

The study was approved by the Ethics Committee of Peking University People’s Hospital. All experiments were conducted in accordance with the Association for Research in Vision and Ophthalmology’s Statement on the Use of Animals in Research. The RGR-d mutant mice were donated by Dr. Henry K.W. Fong, Department of Ophthalmology, Keck School of Medicine of University of Southern California, University of Southern California, Los Angeles, CA, USA. The RGR-d mice were generated as we previously described [12]. Mice were housed in a colony room on a 12-h light/12-h dark cycle, and they were provided with food and water ad libitum. All experiments were performed when the mice were 12 months old. For fundus observation, after anesthetizing the mice with an intraperitoneal injection of 1.25% tribromoethyl alcohol, 1% atropine eye drops (Santen Pharmaceuticals Co., Ltd., Osaka, Japan) were administered to dilate their pupils. carbomer ophthalmic gel (Bausch & Lomb, Rochester, NY, USA) was used to coat the cornea. The microscope lens of the small animal fundus imaging system (OPTO-RIS, Optoprobe Science LTD, UK) was adjusted to touch the cornea and obtain a clear fundus photograph. Fundus fluoroscopy was performed by intraperitoneally injecting the mice with 1.7 mL/kg of 2% fluorescein sodium and repeating the photography procedure. Animals were euthanized by cervical dislocation following anesthesia with tribromoethanol. All efforts were made to reduce the rats’ suffering and to minimize the number of animals used. In this study, a total of 42 mice were used. Mice were euthanized, and the eyes were enucleated and processed for further analysis.

The eyes were prepared and removed the anterior segment of the eye. The eyecups were prepared, washed in PBS three times, fixed in 2.5% glutaraldehyde at 4°C in the dark for overnight. After being washed with PBS (3 times for 10 min), the samples were postfixed with 1% osmium for 3 h. The eyecups were then rinsed with PBS and dehydrated in acetone (50%, 70%, and 100% for 3 × 10 min). After embedding and hardening in resin, sagittal sections were cut from the tissues by using an ultramicrotome. The samples were stained with 3% uranyl acetate combined with lead citrate. Copper grids with 50 mesh grids were selected to support the samples. Transmission electron microscopic (TEM) images were obtained using a JEM-1400PLUS (JEOL, Tokyo, Japan).

Mice were euthanized and the eyes were enucleated. After fixing the eyes in 4% paraformaldehyde (Beyotime, Shanghai, China), they were dehydrated and embedded in paraffin. Hematoxylin and eosin (H&E) staining was performed on 5-μm-thick sagittal sections of tissues crossing the optic nerves. Images were taken using an Olympus microscope (Tokyo, Japan).

Retinal function was evaluated in RGR-d mice and wild-type (WT) mice (12 months old) with flash electroretinography (fERG; RETIport, Roland Consult, Brandenburg an der Havel, Germany). The mice were allowed to adapt to the dark and anesthetized with an intraperitoneal injection of 1.25% tribromoethyl alcohol (0.4 mL/20 g). Under illumination with a red light (>650 nm), the pupil was dilated using tropicamide phenylephrine eye drops (Santen Pharmaceuticals Co., Ltd.). We placed platinum ring electrodes onto the corneal surface. Subdermal grounding electrodes were also used. The stimulation intensity was initially set to 0.001 cd.s/m² and increased as described previously [13]. The amplitudes of the a- and b-waves were recorded at all stimulation intensities. The amplitude of the a-waves was measured from baseline to the troughs of the a-wave. Previous studies have pointed the electroretinography (ERG) a-wave originates from photoreceptor cells [14, 15]. The b-wave, which represents the activity of the inner retina, was measured from the base of the negative a-wave to the peak of the b-wave. The quantitative results were derived as the means of six independent experiments.

Human retinal pigment epithelial (ARPE-19) cells were cultured in Dulbecco’s modified eagle medium (DMEM F12, lot 2323161, Gibco, Carlsbad, CA, USA) containing 10% fetal bovine serum (lot 2168090 RP, Gibco) and 1% streptomycin-penicillin, in an atmosphere containing 5% CO₂ at a temperature of 37°C. When the cells were >90% confluent, they were dissociated using 0.25% trypsin-EDTA (Ref 25200-056, Gibco). There were 5 × 10⁴ ARPE-19 cells that were seeded in six-well plates to carry out lentivirus transduction. When the cells were approximately 30% confluent, they were transfected with lentivirus-control, lentivirus-CMV-MCS-RGR-3FLAG-SV40-puromycin, and lentivirus-CMV-MCS-RGR-d-3FLAG-SV40-puromycin (Shanghai Genechem Co., Ltd., Shanghai, China). Once the cells reached 70% confluence, puromycin solutions of appropriate concentration were used to select stable expression cell lines (multiplicity of infection, MOI = 5). MG-132, a proteasome inhibitor (ab141003, Abcam, Cambridge, UK), was used to enrich the overexpressed protein. In this study, ARPE-19 cells were treated with 2 μm MG-132 for 24 h. Stable transformants of ARPE-19 expressing RGR or RGR-d, and normal control (NC) cells were cultured in plates or the upper chambers of Transwell filter inserts (Costar, Corning Inc., Corning, NY), as the RGR group, RGR-d group, and control group. Palmitic acid (PA; catalog No. P5585, Sigma-Aldrich, St. Louis, MO, USA) was initially dissolved in dimethyl sulfoxide and added to complete cell culture growth medium containing 5% bovine serum albumin (catalog No. A8806, Sigma-Aldrich; final concentration: 0.2 mm) to simulate the increased oxidative stress in early AMD. Twenty-four hours later, the cells were ready for use in subsequent experiments.

Cell viability was assessed using the cell counting kit-8 (CCK-8, 40203ES76, Yeasen Biotechnology Shanghai Co., Ltd.). Cells were seeded in 96-well plates at a density of 5 × 10³ cells per well and incubated for 24 h. Stable transformants of ARPE-19 expressing RGR or RGR-d, along with NC cells, were treated with 2 μm MG-132 for 24 h, with or without 0.2 mm PA. After removing the medium, 100 µL of 10% CCK-8 reagent in serum-free DMEM F12 medium was added to each well, and the cells were then incubated in a 5% CO₂ incubator at 37°C for 1 h. Absorbance was measured at 450 nm using a multifunctional microplate reader (Tecan, Männedorf, Switzerland). Each experiment included 3 independent wells and was performed in triplicate. Results are presented as mean ± SEM.

A transepithelial electrical resistance (TER) assay and tracer flux assay were performed to assess the barrier function of ARPE-19 cell monolayers. A volt-ohm meter with an electrode (EVOM2, Beijing Jingong Hongtai Technology Co., Ltd., China) was used to monitor the TER. To allow complete formation of epithelial cell junctions, the cells were cultured on inserts. Cells (1 × 10⁴) were seeded to the upper chambers (6.4 mm diameter, 0.4 mm pore size; Corning) and incubated for 12 h, and 800 mL DMEM F12 culture medium was added in the lower chamber. TER was measured at baseline (unstimulated) and after treating with 2 μm MG-132 for 24 h, with or without 0.2 mm PA, as described in the section “Cell lentivirus transfection and stimulation,” and the media in all chambers of the Transwell were replaced. The cells were cultured in an incubator (DMEM F12 without FBS) 2 h before measuring the resistance values of the individual monolayers. Before each test, the background resistance was measured using culture medium (DMEM F12 without FBS) and a blank filter. Four measurements were taken for each insert and averaged.

After stimulation, the medium was removed. Fluorescein isothiocyanate (FITC)-dextran (catalog No. 46945, Sigma-Aldrich) was added to the upper chamber (concentration: 1 mg/mL), and 800 mL of the culture medium was added to the lower chamber. Samples of the medium were collected from the bottom chamber at 15, 45, 60, and 90 min. The fluorescence intensity of each sample was determined using a multifunctional microplate reader (Tecan, Männedorf, Switzerland) with excitation/emission peaks of 488/530 nm, respectively.

The cells were seeded in a 24-well plate growing on glass coverslips. After the stimulations described in the section “Cell lentivirus transfection and stimulation,” carboxylate-modified microspheres (FluoSpheres, catalog No. F8803, Invitrogen, Waltham, MA, USA) were used to evaluate the change in phagocytosis function. The microspheres were added to the complete cell culture growth medium (final concentration: 2 μL/mL). After incubation for 1.5 h, the cells were washed three times using phosphate-buffered saline (PBS; HyClone, Logan, UT, USA), fixed for 15 min in 4% paraformaldehyde at room temperature, and then immersed in 0.3% Triton X-100 for permeabilization. After washing with PBS, Alexa Fluor phalloidin (catalog No. A12381, Invitrogen) was used to label the cytoskeleton. Cells were counterstained with DAPI dihydrochloride. A laser confocal microscope (TCS SP8 WLL, Leica, Wetzlar, Germany) was used to image the cells.

After stimulation, the cells were fixed for 15 min in 4% paraformaldehyde, and nonspecific binding was blocked by immersing them in 5% goat serum for 1 h at 37°C. The cells were first incubated with a ZO-1 polyclonal antibody (1:200, 40–2200, Invitrogen) at 4°C overnight, rinsed with PBS, and incubated for 60 min at room temperature with Alexa Fluor 488 AffiniPure Goat Anti-Rabbit IgG (H+L; 1:1,000, 33106ES60, Yisheng, Beijing, China) as the secondary antibody. Counterstaining was performed using DAPI (ab104139, Abcam), and images were obtained using the laser confocal microscope (TCS SP8 WLL, Leica).

The anterior segment of each eye was removed, and the retina was harvested after the eye had been enucleated. Samples were collected and suspended in the lysis buffer (catalog No. 89900, Thermo Fisher Scientific, Waltham, MA, USA) containing Pierce Protease and Phosphatase Inhibitor (catalog No. A32959, Thermo Fisher Scientific). After lysis by ultrasonication and centrifugation at 10,000 g for 15 min, the supernatant was retrieved and mixed with loading buffer (catalog No. REF NP0007, Thermo Fisher Scientific) to denature the proteins. Protein concentrations were determined using the BCA Protein Assay Kit (Beyotime). Protein separation was achieved by SDS-PAGE (REF NP0301, Thermo Fisher Scientific). The proteins were then transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA), which were subsequently blocked for 1 h with Intercept Blocking Buffer (LI-COR, Lincoln, NE, USA) and incubated with the primary antibody overnight. The antibodies used for this assay were anti-LC3 rabbit mAb (#4108, Cell Signaling Technology, Danvers, MA, USA), anti-Beclin-1 (D40C5) rabbit mAb (#3495, Cell Signaling Technology), anti-ATG5 rabbit mAb (#12994, Cell Signaling Technology), anti-SQSTM1/P62 rabbit mAb (#39749, Cell Signaling Technology), and anti-β-actin antibody (#4967, Cell Signaling Technology). After incubation with the indicated antibodies, the membranes were washed and incubated for 1 h at room temperature with the secondary antibody. An imaging system (ODYSSEY CLx, LI-COR) was used to visualize and quantify the immunoblots.

All of the statistical analyses were performed using SPSS Statistics 20 (IBM, Armonk, NY, USA) and GraphPad Prism Inc. (Dotmatics, Boston, MA, USA). Representative images are shown, and the quantitative results represent the mean values from three or six independent experiments. The Mann-Whitney U test was used to evaluate differences between two groups. To detect differences among three or more groups, we used the one-way ANOVA test. We considered p values of ≤0.05 to indicate statistical significance.

Our previous histological study [12] observed a drusen-like basal deposit in the peripheral retina in 10-month-old RGR-d mice, and multiple areas of depigmentation and lesions in 31-month-old RGR-d mice. In the present study, we found that the fundus imaging was normal in 6-month-old WT and RGR-d mice. In 12-month-old WT mice, there was mild vitreous syneresis. However, some retinal atrophy of the periphery fundus was seen in 12-month-old RGR-d mice. Fluorescein fundus angiography showed some atrophy and patchy penetrating hyperfluorescent foci on the peripheral fundus (Fig. 1).

In RGR-d mice, the morphological assay of H&E-stained retinal tissues showed thinning of each layer of the retinal structure in the center and periphery (online suppl. Fig. 1A and B; for all online suppl. material, see https://doi.org/10.1159/000541991) and abnormal pigmentation of the RPE (Fig. 1a3, a4). To observe the changes in the RPE-Bruch’s membrane-choriocapillaris complex of RGR-d mice, we performed embedded and sectioned eyeballs of WT and RGR-d mice by TEM. In RGR-d mice, the base ultrastructure of the RPE was disordered compared with the WT mice. In 12-month-old RGR-d mice (Fig. 2), TEM analysis showed extension of some diffuse granular deposits toward the RPE (Fig. 2b1). The morphology of choroidal microvascular endothelial cells was distorted comping by narrowing of choriocapillaris lumen, here were fewer pigment granules in the RPE cells, and the granules were malformed (Fig. 2b1, b2). The morphology of the photoreceptor outer segments was structurally damaged and disordered in RGR-d mice (Fig. 2b3), and a thickened Bruch’s membrane was observed along the wall of the choriocapillaris endothelial cells compared with WT mice (Fig. 2c).

Previous studies did not explored changes in retinal function. We used fERG to evaluate the retinal function of the 12-month-old RGR-d mice and representative images were recorded sequentially (Fig. 3). The a-wave is a large negative wave that originates in the photoreceptor cells and is followed by b-wave, a large positive wave, which originates from bipolar cells and Müller cells. The a- and b-waves thus represents the function of the outer layer and the inner nuclear layer of the retina, respectively. In RGR-d mice, the a- and b-wave amplitudes for scotopic fERG were lower at a stimulation intensity of 3.0 cd.s/m² compared with WT mice, although they were not significantly different. The a-wave amplitudes during photopic fERG in the WT and RGR-d mice were −31.35 ± 6.36 versus −16.55 ± 6.75 µV at 1.0 cd.s/m², −20.72 ± 4.43 versus −14.93 ± 3.07 µV at 3.0 cd.s/m², and −16.63 ± 4.37 versus −13.68 ± 3.48 µV at 10.0 cd.s/m². The amplitudes of the a-waves in photopic fERG at 1.0 cd.s/m² were significantly lower in the RGR-d mice than in the WT eyes (Fig. 3b). There were no significant differences in the b-wave amplitudes measured during photopic fERG. These results indicated that the photoreceptor outer segments of RGR-d mice were dysfunctional.

Autophagy is critical for maintaining cellular homeostasis by removing protein aggregates and injured organelles [9]. Impaired autophagy has been implicated in AMD and other age-related neurodegenerative diseases [10]. We assessed the alterations in autophagy-related protein expression levels in the RPE and retinal neurepithelium of WT and RGR-d mice to assess the status of the autophagy pathway. LC3II/I protein expression in the RPE and retinal neurepithelium was lower in RGR-d mice than in WT mice, and Beclin-1 expression was also downregulated (Fig. 4). ATG5 protein, which is upstream of the autophagy regulatory pathway, also showed decreased expression. The autophagic substrate P62 was increased in the RPE and retinal neurepithelium of RGR-d mice (Fig. 4a, b1). These protein expression profiles indicated that autophagy regulation was decreased in the RPE and retinal neurepithelium of RGR-d mice in the early stage of AMD-like defect formation.

Oxidative stress increases with aging in retina [16]. To determine the changes in RPE function in cells overexpressing RGR-d, PA was used to simulate oxidative stress conditions. In online supplementary Figure 2, cell viability assays demonstrate that MG-132, and the combination of the two compounds (MG-132 + PA) do not induce cell death. The cellular barrier function was estimated by measuring the permeability of an epithelial cell monolayer. The TER assay of ARPE-19 cells expressing RGR or RGR-d, and NC cells showed that the resistance of the epithelial monolayer was decreased in cells overexpressing RGR-d protein under oxidative stress conditions. Before PA stimulation, the TER did not differ significantly among the three groups (Fig. 5a). Under PA stimulation, the TER was decreased in the three groups. The TER was significantly decreased in the RGR-d group compared with the NC group or RGR group (58.00 ± 10.92 vs. 43.13 ± 7.78 ohms/cm², 59.50 ± 6.60 vs. 43.13 ± 7.78 ohms/cm², respectively), and there was no difference between the NC and RGR groups after PA stimulation. In the tracer flux assay, stimulation with PA caused an increase in FITC-dextran passage across the ARPE-19 cell monolayer in the three groups (Fig. 5b1–b5), without significant differences between the NC and RGR groups. At 45, 60, and 90 min (Fig. 5b3–b5), the fluorescence intensity was significantly greater in the RGR-d group than in the other groups.

We also evaluated the role of RGR-d in regulating the intercellular junction by performing immunofluorescent labeling of ZO-1. Figure 6a showed that treatment with PA disrupted the tight junctions in three groups. The fluorescent signal of ZO-1 was decreased, and the tight junction was obviously disorganized in the PA-treated RGR-d group. We measured the phagocytotic activity in each groups by using FluoSpheres. Representative images of these cell groups are shown in Figure 6b and c. After PA stimulation, the number of FluoSpheres detected was significantly decreased in the RGR-d group than in the other groups. We observed accumulation of FluoSpheres between cells that were not engulfed by the ARPE-19 cells. Our results showed that RGR-d overexpression disrupted the function of ARPE-19 cells under oxidative stress conditions.

Cells transfected with lentivirus overexpressed protein in the presence of MG-132 (online suppl. Fig. 3). Exposure of three groups of cells to 0.2 mm PA for 24 h led to a significant increase in the LC3II/I ratio in the NC and RGR groups, but not in the RGR-d group (Fig. 7). We then measured the protein expression levels of ATG5, Beclin-1, and P62 and that their expression levels did not differ significantly among the three groups before PA treatment. However, exposure to oxidative stress downregulated ATG5 and Beclin-1 and upregulated P62 in the RGR-d group. However, these changes were not significant in the PA-treated RGR-d group. These results demonstrated that oxidative stress exposure leads to dysfunctional autophagy in cells overexpressing RGR-d.

In this study, using mice subjected to multiplex gene editing using the clustered regularly interspaced short palindromic repeats (CRISPR) method [12], we found mild changes in the fundus of early dry AMD-like defects at 12 months old; however, the retinal function had already been altered by this time. The autophagy marker LC3II/I was downregulated. Dysregulated autophagy has been found in retina and RPE-choriocapillaris complex. To explore the effect that altered RGR to RGR-d on the function of cells, we constructed a lentivirus vector that induced RGR and RGR-d overexpression in ARPE-19 cells and used PA treatment to simulate the oxidative stress that occurs in early dry AMD. Our study suggested that cells overexpressing RGR-d were more prone to dysregulation and dysfunction of autophagy under oxidative stress.

AMD is a complex process involving oxidative stress, hypoxia, inflammation, lipid metabolism, immune dysregulation, and environmental risk factors [17, 18]. There are several rodent models to help reveal the pathology of dry AMD. Oxidative stress was shown to cause RPE damage in animal models, including nuclear factor erythroid 2-related factor 2 deficient (Nrf2^−/−) mice [19], superoxide dismutase deletion (Sod2^−/−) mice [20], and high-fat diet-fed peroxisome proliferator-activated receptor-γ coactivator 1α (Pgc-1α^+/−) mice [21]. Some models of inflammation can also simulate the early stage of AMD changes, including membrane cofactor protein knockout (CD46^−/−) mice [22] and complement factor H transgenic mice [23]. Although these models have some limitations, they have contributed to advancing our understanding of AMD itself and elucidate potential new therapeutic targets. In the present study, using RGR-d mice, we found that RGR-d disrupted autophagy in the retina and resulted in the dysfunction of RPE and the development of an early AMD-like pathophysiology. RGR is a membrane protein that is localized in the cytoplasm of RPE and Müller cells [24, 25]. Furthermore, it is an RPE opsin that is produced in during retinal development and in primary cell cultures [26]. It is important that cells synthesize 11-cis-retinal at a normal rate in the light and dark [27‒29]. RGR-d, an exon-skipping splice variant of RGR, was identified in donor retina and RPE of AMD patients [30]. The pathogenesis of aged RGR-d mutant mice also shows AMD-like defects [12]. In our previous study, we only performed ultrastructural observations using TEM but did not perform fundus fluorescein angiography or retinal function evaluation, in mice at about 10 months old. Severe choriocapillaris degeneration was observed in 21-month-old RGR-d mice and multiple depigmentation was found in 31-month-old RGR-d mice. Here, we therefore investigated the changes in retinal function and signal pathways at an early age in RGR-d mice showing lamellar retinal atrophy and a thickened Bruch’s membrane. We found that the outer retinal layer of 12-month-old RGR-d mice may have retinal functional deficits before multiple depigmentation and lesions appear. The autophagy pathways were dysregulated in the RPE and retinal neurepithelium of 12-month-old RGR-d mice, which showed the early stage of AMD-like defects. Furthermore, under oxidative stress, the cells overexpressing RGR-d showed dysfunction in the epithelial monolayer permeability, tight junctions, and phagocytosis. Dysregulated autophagy was also observed in the cells overexpressing RGR-d under oxidative stress.

Vitreous turbidity was observed in WT mice at 12 months old and the fundus image and fluorescein fundus angiography were normal. However, the fundus phenotype of RGR-d mice revealed distinct hypopigmentation of regions of the RPE in the retina and penetrating hyperfluorescent foci. In dry AMD, the degeneration of RPE results in early hyperfluorescent foci at a time when the loss of the RPE enhances the visibility of the underlying choroidal vasculature perfused with the intravascular fluorescein dye. The surrounding choroidal stromal tissue becomes blurred in the later phases of fluorescein angiography [31]. Electron microscopy images showed thickening of the Bruch’s membrane and abnormal morphology of the RPE in in vivo models of AMD [20]. Here, we also observed distortion of the photoreceptors. The interconnection between the RPE and photoreceptors plays an important role in maintaining the photoreceptor metabolism [20]. ERG is a standard method for evaluating the impairment of retinal function and each ERG component represents the activity of different groups of retinal cells [32]. In photopic fERG at an intensity of 1.0 cd.s/m², the amplitudes of the a-waves were decreased in RGR-d mice, indicating reduced functions of the photoreceptor cells.

Autophagy is a catabolic process that has been explored in some neurodegenerative diseases [7]. The process is often upregulated in response to environmental stresses to maintain normal cellular renewal [6]. The overall autophagic capacity is increased in aging, but there is increasing evidence indicating impaired autophagy in the latter stages of mouse models for AMD [10]. In the present work, we found decreased LC3-II/I expression in the RPE and retinal neurepithelium in RGR-d mice. LC3 is involved in the biogenesis of autophagosomes and is present in all types of autophagic membranes. LC3I is converted to LC3II by lysosomal enzymes, and the rate of conversion reflects the progression of autophagy [33, 34].More than 30 ATGs have been identified [35]. We observed decreased ATG5 expression in the retina of RGR-d mice, consistent with the findings reported in other animal models of AMD [36]. A previous study found that deletion of Atg5 impaired autophagy in the retina, and the capacity for removing the RPE was downregulated, leading to AMD-like RPE defects and partially penetrant retinal degeneration [36]. Beclin-1, another component of the autophagy pathway, possesses a BH3 domain and belongs to the BH3-only protein family [37]. In RGR-d mice, Beclin-1 expression was decreased. P62 is key to the degradation of autophagy cargoes. In the present study, the expression of P62 increased as the autophagy function decreased in RGR-d mice. Furthermore, we found the same changes in these autophagy pathway proteins in ARPE-19 cells overexpressing RGR-d under oxidative stress.

AMD is a multifactorial disorder involving most components of the retina [5], including the RPE, which provides the retina with nutrients and performs phagocytosis of the photoreceptor outer segments. Normal RPE function ensures the survival and metabolism of photoreceptors and the maintenance of vision. Under oxidative stress, the autophagic system was upregulated to cope with pathological stimulation in ARPE-19 cells in the NC and RGR overexpression groups. However, in the RGR-d overexpression group, the autophagic system was disordered and affected the function of the cells. In the RGR-d group, there was significantly greater loss of barrier function and phagocytosis. Our previous work has shown that RGR-d mice developed certain AMD-like pathophysiologies and the phenotypic effects progressed gradually in this mouse model [12]. In addition, in the current study, based on the mouse models and transfected cells, we found that RGR-d is a dysfunctional protein linked to AMD-like defects and it may be involved in the dysfunction of autophagy in the early stage of AMD-like defects.

There are several limitations to our study. The ARPE-19 cell line in culture loses cell polarity and cannot fully replicate RPE function. A previous study noted that the RGR protein is also abundant in Müller cells, but we did not assess changes in Müller cells when RGR mutated to RGR-d [25]. We plan to investigate changes in cellular function in primary mouse RPE cells and Müller cells when RGR mutates to RGR-d. Our findings indicate that autophagy plays a role in the dysregulation of cellular function leading to AMD, which is consistent with previous research [38, 39]. Based on these findings, enhancing autophagy through regulators may offer a potential therapeutic approach for AMD. Some studies suggest that increasing autophagy can mitigate AMD-like phenotypes in mice [40‒42]. Therefore, further research is needed to determine whether boosting autophagy can improve AMD-like defects in RGR-d mice and to explore the relationship between RGR-d and the autophagy signaling pathway.

In summary, the RGR-d mouse model showed fundus changes and retinal dysfunction at 12 months old and dysregulated autophagy was already apparent in the early stage of AMD-like defects. Cells overexpressing RGR-d were more prone to dysregulated and dysfunctional autophagy under oxidative stress. We will explore the relationship between RGR-d and dysregulated autophagy further and identify the treatment targets before severe progression occurs.

The authors would like to thank Henry K.W. Fong, Department of Ophthalmology, Keck School of Medicine of University of Southern California, University of Southern California, Los Angeles, CA, USA.

The study has been performed in accordance with the Declaration of Helsinki. This study protocol was reviewed and approved by the Ethics Committee of Peking University People’s Hospital, Approval No. (2019PHC027). And the experiments complied with the Association for Research in Vision and Ophthalmology’s Statement on the Use of Animals in Research. All presentations of the study have consent for publication.

The following authors have no financial disclosures: Y.G., N.X., H.Y., J.L., L.H., L.Z., W.D., Z.L., and M.Z. The authors have no conflicts of interest to declare.

This work was supported by the National Natural Science Foundation of China (Grant No. 82171060) and the National Key Research and Development Program of China (No. 2020YFC2008200; No. 2020YFC2008203). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Y.G. performed the study, performed the data analysis, and wrote the manuscript; N.X., H.Y., and J.L. performed the data analysis; L.H., L.Z., W.D., and Z.L. revised the manuscript; and M.Z. designed and performed the study.

All data generated or analyzed during this study are included in this article and its online supplementary material files. Further inquiries can be directed to the corresponding author.