Involvement of Oxidative and Endoplasmic Reticulum Stress in RDH12-Related Retinopathies

Retinol dehydrogenase 12 (RDH12) is expressed in photoreceptor inner segments and catalyses the reduction of all-trans retinal (atRAL) to all-trans retinol (atROL), as part of the visual cycle. Mutations in RDH12 are primarily associated with autosomal recessive Leber congenital amaurosis. To further our understanding of the disease mechanisms, HEK-293 cell lines expressing wildtype (WT) and mutant RDH12 were created. The WT cells afforded protection from atRAL-induced toxicity and oxidative stress. Mutant RDH12 cells displayed reduced protein expression and activity, with an inability to protect cells from atRAL toxicity, inducing oxidative and endoplasmic reticulum (ER) stress, with upregulation of sXBP1, CHOP, and ATF4. Pregabalin, a retinal scavenger, attenuated atRAL-induced ER stress in the mutant RDH12 cell lines. A zebrafish rdh12 mutant model (rdh12u533 c.17_23del; p.(Val6AlafsTer5)) was generated through CRISPR-Cas9 gene editing. Mutant fish showed disrupted phagocytosis through transmission electron microscopy, with increased phagosome size at 12 months post-fertilisation. Rhodopsin mislocalisation and reduced expression of atg12 and sod2 indicated early signs of a rod-predominant degeneration. A lack of functional RDH12 results in ER and oxidative stress representing key pathways to be targeted for potential therapeutics.


Introduction
Leber congenital amaurosis (LCA) is a severe early onset autosomal recessive retinal dystrophy, and mutations in the retinol dehydrogenase 12 (RDH12) gene (OMIM: 608830) account for approximately 10% of all cases [1]. RDH12 is located on chromosome 14, consisting of seven exons and encoding a protein that is 35 kDa, with the majority of mutations being missense. RDH12-LCA is a progressive condition that typically presents with night blindness, nystagmus, and central loss of vision in early childhood, leading to complete blindness in adulthood [2]. Autosomal dominant mutations in RDH12 have also been linked to a later onset milder retinitis pigmentosa (RP) phenotype [3,4].
RDH12 is expressed in photoreceptor inner segments and is responsible for the reduction of all-trans retinal (atRAL) to all-trans retinol (atROL), as part of the visual cycle [5]. atRAL is primarily reduced to atROL in photoreceptor outer segments (POSs) by RDH8; however, in periods of intense illumination, atRAL leaks to the inner segments where it is reduced by RDH12 [6]. RDH12 can also act on medium chain aldehydes, suggesting a possible role in the detoxification of lipid peroxidation products [7]. In COS-7 cells transfected with various RDH12 missense variants, mutant protein was expressed at lower levels and displayed significantly reduced enzyme activity, compared with wildtype (WT) RDH12 [8]. Nuclei were counterstained with DAPI (4′,6-diamidino-2-phenylindole) (blue). Scale bar = 10 µM. (C) RDH12 activity assay using HPLC showed no atROL was detected in untransfected cells, but it was found in WT cells, confirming active functional RDH12. (D) Enzyme activity of all mutant proteins was significantly reduced compared with WT RDH12. (E) RDH12 protects against atRAL-induced toxicity. Cells were dosed with increasing concentrations of atRAL for 24 h and cell viability was assessed by MTT assay. atRAL is toxic to untransfected cells, whereas cells expressing WT RDH12 were protected from atRAL-induced cell death. p.Y226C and p.S13* mutant RDH12 did not protect cells from atRAL toxicity, whereas p.A109P mutant protein offered significantly higher protection than untransfected cells at 100 µM atRAL concentration. Three independent experiments were performed. Data are expressed as mean ± SEM, and analysed using two-way ANOVA, followed by Dunnetts multiple comparison test. * p ≤ 0.05, *** p ≤ 0.001.
In order to study the disease mechanisms and screen potential therapeutics for RDH12-retinopathies, three mutant RDH12 stable HEK-293 cell lines were created; two p.Y226C and p.S13* cell lines show no RDH12 protein expression, while p.A109P shows reduced expression. (B) Co-staining of calnexin (red) and GFP (green) confirmed localisation of RDH12 to the endoplasmic reticulum. Nuclei were counterstained with DAPI (4 ,6-diamidino-2-phenylindole) (blue). Scale bar = 10 µM. (C) RDH12 activity assay using HPLC showed no atROL was detected in untransfected cells, but it was found in WT cells, confirming active functional RDH12. (D) Enzyme activity of all mutant proteins was significantly reduced compared with WT RDH12. (E) RDH12 protects against atRAL-induced toxicity. Cells were dosed with increasing concentrations of atRAL for 24 h and cell viability was assessed by MTT assay. atRAL is toxic to untransfected cells, whereas cells expressing WT RDH12 were protected from atRAL-induced cell death. p.Y226C and p.S13* mutant RDH12 did not protect cells from atRAL toxicity, whereas p.A109P mutant protein offered significantly higher protection than untransfected cells at 100 µM atRAL concentration. Three independent experiments were performed. Data are expressed as mean ± SEM, and analysed using two-way ANOVA, followed by Dunnetts multiple comparison test. * p ≤ 0.05, *** p ≤ 0.001.
In order to study the disease mechanisms and screen potential therapeutics for RDH12retinopathies, three mutant RDH12 stable HEK-293 cell lines were created; two expressing missense mutations (p.Y226C and p.A109P) and one nonsense mutation (p.S13*). Site-directed mutagenesis was used to introduce mutations into the WT RDH12 plasmid, and mutagenesis was confirmed by Sanger sequencing. Western blot probing for RDH12 was used to analyse levels of protein expression. The p.Y226C mutant showed no protein expression, consistent with previous publications [21]. Cells expressing the p.A109P mutant showed reduced expression of RDH12, whereas the p.S13* mutant showed no protein expression, consistent with it being a null mutant, thus no full length protein is produced ( Figure 1A). The HPLC activity assay showed significantly reduced RDH12 activity in the p.Y226C (3.5 ± 1.6%) and p.A109P lines (2.7 ± 0.3%) (p < 0.0001) and no activity in the p.S13* line ( Figure 1D).

RDH12 Protects Cells from All-Trans Retinal-Induced Toxicity
High concentrations of atRAL are toxic to cells. Hence, in order to determine the ability of RDH12 to protect cells against this stressor, and whether mutant cells retain this ability, increasing concentrations of atRAL were added to cells for 24 h and cell viability was assessed by MTT assay. In untransfected cells, increasing concentrations of atRAL resulted in a dose-dependent decrease in cell viability. However, in cells expressing WT RDH12, no significant decrease in cell viability was observed with atRAL up to 100 µM ( Figure 1E). At 100 µM atRAL, WT cells offered significant protection with 84% cell viability compared to untransfected cells, which displayed only 26% viability (p < 0.001). In all mutant lines, decreased cell viability was observed compared to WT at concentrations of 50 µM atRAL and above, consistent with the reduced expression and activity of mutant RDH12. In cells expressing the p.Y226C and p.S13* mutants, no increased protection against atRAL toxicity was observed compared to the untransfected cells. However, the p.A109P mutant cells did offer increased protection (64% cell viability) against 100 µM atRAL toxicity compared to the untransfected cells (p = 0.023), consistent with reduced expression levels.

atRAL Induces Oxidative Stress in Mutant Cell Lines
Excess atRAL has been shown to induce oxidative stress in ARPE-19 cells, resulting in an increase in ROS [13]. SOD is the first line of defence antioxidant enzyme that dismutates superoxide anions to hydrogen peroxide. SOD activity was measured following dosing with atRAL for 24 h. In untransfected cells, this significantly reduced SOD activity (p < 0.0001), indicating an increase in ROS beyond the capacity of SOD enzymes, resulting in oxidative stress. In contrast, SOD activity was not affected by addition of atRAL in WT RDH12 cells, demonstrating that protection was conferred by the WT protein. SOD activity was reduced in all three mutant lines, indicating an increase in oxidative stress levels and impairment of the compensatory defence mechanisms (Figure 2A). mRNA expression of oxidative stress markers was also analysed by RT-qPCR in mutant cell lines treated with atRAL, which resulted in a significant increase in NRF2 in the p.Y226C line (p = 0.0132) and similarly in HO-1 in the p.Y226C and p.S13* mutant lines (p = 0.004 and p = 0.0006, respectively). CAT mRNA expression was significantly reduced in all three mutant cell lines ( Figure 2B). N-acetylcysteine amide (NACA), an antioxidant, was tested for its ability to reduce atRAL-induced oxidative stress. NACA was used at a concentration of 750 µM based on previous publications [22,23]. Treatment with 750 µM NACA for 24 h did not restore atRAL induced reduction in SOD activity, nor did it significantly attenuate atRALinduced changes in NRF2 and CAT expression. However, NACA did show a general trend with reducing atRAL-induced upregulation of HO-1 expression in all mutant lines and significantly in the p.S13* line (p = 0.0173) (Supplementary Materials Figure S1). Dosing with atRAL causes a significant reduction in SOD activity in mutant cells. Statistical significance was analysed using two-way ANOVA and Sidak's multiple comparison test. ** p ≤ 0.01, *** p ≤ 0.001. (B) Expression of oxidative stress markers NRF2, SOD2, CAT, GPX1, and HO-1 was analysed by RT-qPCR following treatment with 50 µM atRAL for 24 h. Statistical significance was analysed with paired t-test. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

Pregablain Attenuates atRAL-Induced ER Stress in RDH12 Mutant Cell Lines
Expression of sXBPI through RT-qPCR was slightly increased in all mutant lines compared to WT RDH12 cells, with a 2.3-, 2-, and 2.2-fold increase in p.Y226C, p.A109P, and p.S13* cells, respectively. This could be a result of RDH12 misfolding caused by the mutations, triggering the UPR pathway. Incubation with 50 µM atRAL for 24 h resulted in a significant increase in expression of sXBP1 by 14.6-, 6.6-, and 9.6-fold in the p.Y226C, p.A109P, and p.S13* cells, respectively, compared to undosed cells. Treatment with atRAL also significantly increased expression of CHOP by 6.3, 5.2 and 7.6-fold and ATF4 by 3.4-, 3.5-, and 4-fold in p.Y226C, p.A109P, and p.S13* lines, respectively ( Figure 3).
Pregabalin is a primary amine containing drug, commonly used to treat epilepsy, nerve pain, and anxiety [24], and was shown to protect Rdh12 -/-knockout mice from lightinduced retinal degeneration [25]. As shown in Figure 3, co-treatment of 1 mM pregabalin with atRAL for 24 h significantly reduced expression of sXBP1 in the p.Y226C (p < 0.0001) and p.S13* (p = 0.0178) lines, and reduced expression of ATF4 in all mutant cell lines. These data show that pregabalin can reduce ER stress. NACA also significantly reduced the expression of sXBP1, CHOP, and ATF4 in p.S13* line (Supplementary Materials Figure S2). Tauroursodeoxycholic acid (TUDCA), a drug that has ER stress inhibiting properties, was Dosing with atRAL causes a significant reduction in SOD activity in mutant cells. Statistical significance was analysed using two-way ANOVA and Sidak's multiple comparison test. ** p ≤ 0.01, *** p ≤ 0.001. (B) Expression of oxidative stress markers NRF2, SOD2, CAT, GPX1, and HO-1 was analysed by RT-qPCR following treatment with 50 µM atRAL for 24 h. Statistical significance was analysed with paired t-test. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

Pregablain Attenuates atRAL-Induced ER Stress in RDH12 Mutant Cell Lines
Expression of sXBPI through RT-qPCR was slightly increased in all mutant lines compared to WT RDH12 cells, with a 2.3-, 2-, and 2.2-fold increase in p.Y226C, p.A109P, and p.S13* cells, respectively. This could be a result of RDH12 misfolding caused by the mutations, triggering the UPR pathway. Incubation with 50 µM atRAL for 24 h resulted in a significant increase in expression of sXBP1 by 14.6-, 6.6-, and 9.6-fold in the p.Y226C, p.A109P, and p.S13* cells, respectively, compared to undosed cells. Treatment with atRAL also significantly increased expression of CHOP by 6.3, 5.2 and 7.6-fold and ATF4 by 3.4-, 3.5-, and 4-fold in p.Y226C, p.A109P, and p.S13* lines, respectively ( Figure 3). h and cell viability was assessed to determine the optimal dose. A concentration of 100 µM TUDCA was chosen, as it was well tolerated by the cells (95% cell viability), with a minimal final concentration of 0.5% dimethyl sulfoxide (DMSO). TUDCA did not reduce atRAL-induced ER stress in any of the mutant cell lines (Supplementary Materials Figure  S3). Figure 3. atRAL-induced ER stress was attenuated by pregabalin. Cells were dosed with 50 µM atRAL, 1 mM pregabalin (PGB), or both for 24 h. RT-qPCR was performed analysing the expression of ER stress markers. Dosing with atRAL significantly increased expression of sXBP1, ATF4, and CHOP in p.Y226C (A), p.A109P (B), and p.S13* (C) cell lines. Dosing with pregabalin reduced expression of ER stress markers. Statistical significance was analysed using one-way ANOVA and Sidak's multiple comparison test. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

Generation and Characterisation of rdh12 u533 Zebrafish
An in vivo rdh12 zebrafish mutant model was generated using CRISPR-Cas9 mutagenesis. The rdh12 u533 mutant line carried a 7 base pair (bp) deletion c.17_23del; p.(Val6AlafsTer5), which leads to a premature termination codon in exon 1 ( Figure 4A). Expression of rdh12 mRNA transcript was significantly reduced in the mutant fish at 5 days post-fertilisation (5 dpf) (p = 0.0087) ( Figure 4B). Embryos were characterised through retinal histology, TUNEL assay, and rhodopsin immunostaining, but no evidence of a disease phenotype was observed up to 5 dpf, with no difference in SOD activity or atRAL levels between wildtype (wt) and rdh12 u533 fish (Supplementary Materials Figure  S4). At 12 months post-fertilisation (12 mpf), retinal histology and TUNEL assay in the rdh12 u533 fish were akin to wt controls. However, rhodopsin was mislocalised to the photoreceptor inner segments and outer nuclear layer in the rdh12 u533 retina ( Figure 4H). Blue and red/green opsin expression was similar between the wt and rdh12 u533 fish ( Figure 4I- Dosing with pregabalin reduced expression of ER stress markers. Statistical significance was analysed using one-way ANOVA and Sidak's multiple comparison test. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001. Pregabalin is a primary amine containing drug, commonly used to treat epilepsy, nerve pain, and anxiety [24], and was shown to protect Rdh12 -/knockout mice from lightinduced retinal degeneration [25]. As shown in Figure 3, co-treatment of 1 mM pregabalin with atRAL for 24 h significantly reduced expression of sXBP1 in the p.Y226C (p < 0.0001) and p.S13* (p = 0.0178) lines, and reduced expression of ATF4 in all mutant cell lines. These data show that pregabalin can reduce ER stress. NACA also significantly reduced the expression of sXBP1, CHOP, and ATF4 in p.S13* line (Supplementary Materials Figure S2). Tauroursodeoxycholic acid (TUDCA), a drug that has ER stress inhibiting properties, was also tested. Cells were treated with TUDCA at various concentrations (25-300 µM) for 24 h and cell viability was assessed to determine the optimal dose. A concentration of 100 µM TUDCA was chosen, as it was well tolerated by the cells (95% cell viability), with a minimal final concentration of 0.5% dimethyl sulfoxide (DMSO). TUDCA did not reduce atRAL-induced ER stress in any of the mutant cell lines (Supplementary Materials Figure S3).

Generation and Characterisation of rdh12 u533 Zebrafish
An in vivo rdh12 zebrafish mutant model was generated using CRISPR-Cas9 mutagenesis. The rdh12 u533 mutant line carried a 7 base pair (bp) deletion c.17_23del; p.(Val6AlafsTer5), which leads to a premature termination codon in exon 1 ( Figure 4A). Expression of rdh12 mRNA transcript was significantly reduced in the mutant fish at 5 days post-fertilisation (5 dpf) (p = 0.0087) ( Figure 4B). Embryos were characterised through retinal histology, TUNEL assay, and rhodopsin immunostaining, but no evidence of a disease phenotype was observed up to 5 dpf, with no difference in SOD activity or atRAL levels between wildtype (wt) and rdh12 u533 fish (Supplementary Materials Figure S4). At 12 months post-fertilisation (12 mpf), retinal histology and TUNEL assay in the rdh12 u533 fish were akin to wt controls. However, rhodopsin was mislocalised to the photoreceptor inner segments and outer nuclear layer in the rdh12 u533 retina ( Figure 4H). Blue and red/green opsin expression was similar between the wt and rdh12 u533 fish ( Figure 4I-L). Retinal ultrastructure was examined using transmission electron microscopy (TEM), which revealed a significant increase in the size of phagosomes in the mutant RPE (p = 0.0001) ( Figure 5F). Expression of autophagy markers was reduced in the rdh12 u533 fish, with atg12 significantly reduced by 1.9-fold compared with the wt fish (p = 0.0038) ( Figure 5G). No significant difference in atRAL levels was observed in the 16 mpf retinas between the wt and rdh12 u533 fish ( Figure 6A). There was no significant difference in SOD and CAT activity at 16 mpf, although mRNA expression of sod2 was significantly reduced (p = 0.0085) at 12 mpf in the rdh12 u533 fish ( Figure 6E). No significant difference in expression of ER stress markers at 12 mpf was observed between the wt and rdh12 u533 fish (Supplementary Materials Figure S5). L). Retinal ultrastructure was examined using transmission electron microscopy (TEM), which revealed a significant increase in the size of phagosomes in the mutant RPE (p = 0.0001) ( Figure 5F). Expression of autophagy markers was reduced in the rdh12 u533 fish, with atg12 significantly reduced by 1.9-fold compared with the wt fish (p = 0.0038) ( Figure  5G). No significant difference in atRAL levels was observed in the 16 mpf retinas between the wt and rdh12 u533 fish ( Figure 6A). There was no significant difference in SOD and CAT activity at 16 mpf, although mRNA expression of sod2 was significantly reduced (p = 0.0085) at 12 mpf in the rdh12 u533 fish ( Figure 6E). No significant difference in expression of ER stress markers at 12 mpf was observed between the wt and rdh12 u533 fish (Supplementary Materials Figure S5).

Discussion
Mutations in RDH12 are associated with severe early onset retinal degeneration. However, owing to the lack of phenotype observed in the Rdh12 -/-mouse models, the exact disease mechanism of RDH12-retinopathies is not known. Lack of functional RDH12 is thought to result in a build-up of atRAL, which is toxic to cells and induces a number of apoptotic pathways. In this study, HEK-293 cell lines expressing WT and mutant RDH12 were created to study the effects of RDH12 dysfunction. WT RDH12 protected cells against atRAL-induced cell death and oxidative stress, and mutant RDH12 lost this protective capability, resulting in a decrease in cell viability and an increase in atRAL-induced oxidative and ER stress. We generated a CRISPR-Cas9 rdh12 mutant zebrafish model that had a late-onset disease phenotype showing evidence of mild/early rod photoreceptor degeneration with associated RPE changes.
Several studies have linked a build-up of atRAL to an accumulation of ROS in ARPE-19 cells, leading to oxidative stress [12][13][14]17]. Exposure of Abca4 -/-Rdh8 -/-mice to bright light induced atRAL build-up, which also led to an increase in ROS in the photoreceptors and CAT (C) was analysed in 16 mpf retina, but no significant differences were found between wt and rdh12 u533 mutant fish. (D-I) Expression of oxidative stress markers was analysed by RT-qPCR in retinas from wt and rdh12 u533 fish at 12 mpf. mRNA expression of sod2 was significantly reduced in rdh12 u533 fish retinas. Data are expressed as mean ± SEM; * p < 0.05.

Discussion
Mutations in RDH12 are associated with severe early onset retinal degeneration. However, owing to the lack of phenotype observed in the Rdh12 -/mouse models, the exact disease mechanism of RDH12-retinopathies is not known. Lack of functional RDH12 is thought to result in a build-up of atRAL, which is toxic to cells and induces a number of apoptotic pathways. In this study, HEK-293 cell lines expressing WT and mutant RDH12 were created to study the effects of RDH12 dysfunction. WT RDH12 protected cells against atRAL-induced cell death and oxidative stress, and mutant RDH12 lost this protective capability, resulting in a decrease in cell viability and an increase in atRAL-induced oxidative and ER stress. We generated a CRISPR-Cas9 rdh12 mutant zebrafish model that had a late-onset disease phenotype showing evidence of mild/early rod photoreceptor degeneration with associated RPE changes.
Several studies have linked a build-up of atRAL to an accumulation of ROS in ARPE-19 cells, leading to oxidative stress [12][13][14]17]. Exposure of Abca4 -/-Rdh8 -/mice to bright light induced atRAL build-up, which also led to an increase in ROS in the photoreceptors [14]. Exposure of ARPE-19 cells to atRAL resulted in upregulation of NRF2 and antioxidant genes HO-1 and y-GCSh [13]. Similarly, atRAL treatment increased the expression of NRF2 in the p.Y226C cell line, and increased HO-1 expression in all mutant lines. The redox state of the cell is controlled by a fine balance between the level of ROS and antioxidants. Oxidative stress is a result of an imbalance between ROS and antioxidant levels. Under mild stress, antioxidant enzymes scavenge ROS to maintain homeostasis in the cell. However, if ROS levels become too high, cellular homeostasis is disrupted and antioxidant enzymes become overwhelmed as they cannot cope with the high ROS load [26]. The three main classes of antioxidant enzymes are SOD, CAT, and GPX. A significant reduction in CAT mRNA expression and SOD enzyme activity was observed in atRAL-treated mutant cells, indicating that the antioxidant defence mechanisms of the cell are impaired and cannot protect the cells against the high free radical load. In the rdh12 u533 zebrafish retina at 12 mpf, sod2 expression was significantly reduced by 2.2-fold compared with wt fish. Selective knockout of Sod2 in the RPE of mice resulted in reduced ERG c-wave peak and increased RPE cell area in the central retina. EM also revealed disorganised cristae and enlarged mitochondria in the RPE. Increased oxidative stress was seen in the photoreceptors and increased Nrf2 expression in the RPE/choroid complex of Sod2 mice [27]. In another Sod2 RPE knockdown mouse model, suppression of Sod2 also resulted in increased oxidative stress, loss of ERG response, shortened inner and outer segments, and a thinner outer nuclear layer [28]. Overall, SOD activity was not affected in the rdh12 u533 fish up to 16 mpf, and expression of other antioxidant genes (sod1, cat, gpx1, and hmox-1) did not differ significantly between wt and rdh12 u533 retinas, suggesting that sod1 and other antioxidant enzymes can compensate for the reduced sod2 expression at this timepoint, thereby preventing oxidative stress and cellular damage. Possible later timepoints may reveal further disturbances as the disease progresses.
ROS have been shown to co-localise with mitochondria and the ER in atRAL-treated ARPE-19 cells and induce ER stress via activation of the PERK-eIF2α-ATF4 pathway with increased expression of ER stress markers BiP, ATF4, and CHOP [13]. Upregulation of CHOP via activation of ATF4 triggers apoptosis [29]. Similarly, in our study, incubation with atRAL resulted in an increase in mRNA expression of sXBP1, ATF4, and CHOP in all mutant lines, indicating that atRAL activated ER stress induced apoptosis. Pregabalin lowered expression of ER stress markers in mutant HEK-293 cells. ER stress and differences in atRAL levels were not seen in the rdh12 u533 fish at 12 mpf, so pregabalin was not tested in the zebrafish model. Pregabalin is a primary amine containing drug, which can bind free atRAL via a Schiff base and lower its concentration in the retina, thereby preventing the formation of A2E. Pregabalin reduced atRAL levels in Rdh12 -/mice [25], and reduced ROS production in light-exposed Abca4 -/-Rdh8 -/mice [14]. Pregabalin also had a neuroprotective effect in a rat model of diabetic retinopathy, by reducing retinal glutamate, nitric oxide, and malondialdehyde and increasing glutathione levels, and restored retinal histology, by increasing retinal thickness and ganglion cell layer count [30]. Pregabalin is commonly used to treat epilepsy, nerve pain, and anxiety [24]. The most common adverse effects reported with pregabalin use are dizziness, sleeplessness, blurred vision, impaired concentration, dry mouth, oedema, and weight gain [24]. However, recently, the Medicines and Healthcare Products Regulatory Agency (MHRC) have reported severe respiratory depression linked to pregabalin [31].
In the rdh12 u533 zebrafish retina at 12 mpf, we found mislocalised rhodopsin, larger phagosomes, and reduced expression of autophagy markers with a significant decrease in atg12 mRNA expression. Atg5 was found to co-localise with phagosomes containing POS in control mouse RPE. RPE-specific deletion of Atg5 in mice resulted in a reduced ERG response at 16 weeks of age. Atg5-RPE knockout mice also showed accumulation of phagosomes with undigested POS and the POS were not able to penetrate the RPE beyond the apical surface. These mice also had a higher level of 11-cis, all-trans, and 13-cis retinal [32]. In a zebrafish model of USH2A-related RP, the ush2a rmc1 mutant, disruption in rhodopsin localisation and autophagy was also noted, with increased autophagosomes and increased gene expression of atg5 and atg12 and LC3 protein expression [33]. Similar to RDH12-related RP, the retinal degeneration in USH2A patients is later onset, developing from adolescence, compared to those with LCA. Kim et al. [32] suggested that phagocytosis of the POS by the RPE and the visual cycle converge via a non-canonical autophagy pathway, both of which are thought to be disrupted in RDH12-related retinopathies. In this pathway, termed LC3-associated phagocytosis (LAP), following phagocytosis of the POS, the phagosome associates with the Atg12-Atg5-Atg16L complex, triggering lipidation of microtubule-associated protein 1 light chain 3 (LC3) to LC3-II and its recruitment to the phagosome. This signals for the lysosome to fuse with the phagosome, forming a phagolysosome, resulting in degradation of ingested POS and its recycling back to the photoreceptors. This pathway converges with the visual cycle, as POS phagocytosis aids in recycling of the visual pigments back to the photoreceptors. TEM imaging revealed enlarged phagosomes in the RPE of rdh12 u533 fish, indicating a disruption of RPE phagocytosis, with inefficient digestion of POS. Disruption of phagocytosis along with a build-up of retinoids leads to the formation of lysosomal bodies called lipofuscin. The retinoid pigments in lipofuscin are autofluorescent, and are seen on fundus imaging [34]. RDH12-LCA patients have widespread RPE atrophy and fundus imaging typically shows central hypoautofluorescence corresponding to RPE cell death, surrounded by hyperautofluorescent lesions, indicating regions of high toxicity and diseased RPE [35].
In conclusion, through the generation of two models of RDH12-disease, an in vitro human cell line and an in vivo zebrafish mutant model, we have seen that mutations in RDH12 result in an inability of the enzyme to reduce atRAL, leading to toxic levels of this retinoid in cells. This triggers apoptosis, through induction of the oxidative stress and ER stress pathways. Although no gross abnormalities were noted in retinal histology in the rdh12 u533 fish and atRAL levels were comparable to wt fish, early indicators of disrupted pathways are observed in the adult fish including rhodopsin mislocalisation and increased phagosomal size with changes in autophagy related expression, suggesting that stress in the retina is effectively controlled so as not to lead to apoptosis or severe disruption at least up till 12-16 months of age. This corresponds more with the autosomal dominant RDH12-associated RP phenotype, which is a later-onset and relatively mild phenotype in comparison to the severe early-onset LCA phenotype. A similar difference between the human and zebrafish phenotypes has been observed previously in the modelling of KCNJ13-LCA, where the homozygous zebrafish kcnj13 td15 showed an adult-onset retinal degeneration, contrasting with the early severe phenotype seen in patients [36]. It is possible that compensatory mechanisms exist in the zebrafish, accounting for these differences. Although no other rdh12 isoforms exist in the zebrafish, several retinol dehydrogenases are present in the retina that may compensate for lack of rdh12 in zebrafish. In addition, zebrafish are a cone-dominant species, and the cone-specific visual cycle, which is independent of the RPE, may play a more prominent role in the regeneration of the visual chromophore in zebrafish. As the mouse models of RDH12 disease also do not show a severe phenotype, a species specific role may exist for RDH12 in humans that is either redundant or compensated for in other species. Further monitoring of the fish at later time points will provide more insight into the rdh12 phenotype. We have identified a number of key pathways to be targeted for potential therapeutics, including pregabalin. Drugs with ER stress lowering properties, in addition to antioxidants and retinal scavengers, represent a new class of potential drugs that can be targeted for RDH12-related retinopathies.

Stable HEK-293 Cell Line Generation
The ORF expression clone was obtained from GeneCopoeia (pEZ_M98), which encoded RDH12 (NM_152443) with a C-terminal GFP tag. Three known patient mutations (c.677A > G, p.Y226C; c.325G > C, p.A109P; and c.38C > A, p.S13*) were introduced into plasmids using the Quickchange II Site-Directed Mutagenesis kit (Agilent, Santa Clara, CA, USA). Primers were designed, according to kit instructions, with approximately 10 bases of correct sequence either side of the mutation (Supplementary Materials Table S1). Colonies with the correct mutation were verified by Sanger sequencing.
Cells were plated in six-well plates at a density of 600,000 cells per well and, 24 h later, they were transfected using Lipofectamine 2000 Reagent (Thermo Fisher Scientific). Twenty-four hours after transfection, cells were passaged into 96-well plates in selection media (DMEM high glucose, 10% FBS, 1 mg/mL G418) using serial dilutions to achieve low cell density. After 10 days, all non-transfected cells died off and, after a further 2 weeks, transfected cells formed distinct colonies. Individual colonies were picked into 24-well plates and expanded. This was defined as passage 0. For continued culture, G418 concentration was reduced to 0.5 mg/mL. All further experiments were carried out on cells from passage 5 onwards.

Western Blot
Cells were lysed in RIPA buffer (Thermo Fisher Scientific)(city, state abbreviation if USA or Canada, country)), and the protein concentration was quantified using BCA Protein Assay kit (Thermo Fisher Scientific). Samples were analysed by Western blot as previously described [37], using anti-RDH12 (1:1000) overnight at 4 • C. Anti-RDH12 and RDH12 transfected HEK293T cell lysate for use as a positive control were kindly gifted by Professor Debra Thompson (University of Michigan). Membrane was stripped and re-probed with anti-β-actin (A2228; 1:5000; Sigma-Aldrich, St. Louis, MO, USA), as a loading control.

RDH12 Activity Assay
Cell pellets were resuspended in 1× PBS, lysed by sonication, and centrifuged at 13,000 rpm for 10 min. Supernatant was collected and the protein concentration was determined using BCA Protein Assay kit. Reactions were carried out in 500 µL PBS, using 50 µg protein, 1 mM NADPH (Cayman Chemical, Ann Arbor, MI, USA), and 2.5 mM atRAL (Sigma R2500). Reaction was initiated with the addition of NADPH and incubated at 37 • C for 15 min, and stopped by the addition of 500 µL cold methanol. Retinoids were extracted according to the protocol by Chetyrkin et al. [38] on Waters Sep-Pak C18 column. HPLC analysis was performed using a Waters ACQUITY ® UPLC ® (Waters, Wilmslow, UK). Elution was monitored at 325 nm for atROL and 360 nm for atRAL with a Waters ACQUITY 2996 PDA. The stationary phase was Waters ACQUITY UPLC BEH C18 1.7 µm, 2.1 × 50 mm column and the mobile phase consisted of acetonitrile/water 0.1% formic acid (87.5:12.5). The analysis was 8 min, with a flow rate of 0.25 mL/min and a 10 µL injection.

Cell Viability MTT Assay
Cells were plated at a density of 40,000 cells per well in a 96-well plate. Cells were incubated with various concentrations of atRAL for 24 h. Cell viability was determined using MTT (M6494; Thermo Fisher Scientific). Following dosing, treatment media was replaced with 100 µL fresh media. MTT was prepared at a concentration of 5 mg/mL in PBS, and 10 µL was added to each well. Cells were incubated for 4 h at 37 • C. All but 25 µL media was removed, and MTT was solubilised with 50 µL DMSO. Absorbance was read at 540 nm on a Safire2 plate reader (TECAN, Männdorf, Switzerland).

Drug Dosing
For all dosing experiments, atRAL was used at a concentration of 50 µM. Pregabalin (Sigma) was used at a concentration of 1 mM, NACA (Tocris Bioscience, Bristol, UK) at 750 µM, and TUDCA (Cayman Chemical) at 100 µM. Cells were plated in six-well plates at a density of 700,000 cells per well. After 24 h, drugs were added to the cells in culture media. Twenty-four hours later, cells were pelleted and stored at −80 • C for further analysis.

RT-qPCR
Total RNA was extracted from cells using the RNeasy mini kit and from whole embryos and enucleated adult eyes using the RNeasy FFPE kit (Qiagen, Hilden, Germany). cDNA was synthesised from 1 µg of RNA using the Superscript II First Strand cDNA synthesis kit (Thermo Fisher Scientific), according to the manufacturer's instructions. Transcript levels were analysed using SYBR Green MasterMix (Thermo Fisher Scientific) on a StepOne Real-Time PCR system (Applied Biosystems, Thermo Fisher Scientific), under standard cycling conditions. All samples were assayed in triplicate. Primer sequences are shown in Table S2.

SOD and CAT Activity Assay
SOD activity was determined using the SOD assay kit (Sigma; 19160) according to the manufacturer's instructions. CAT activity was determined using the catalase activity assay kit (Abcam; ab83464), according to the manufacturer's instructions.

Zebrafish Husbandry
wt, AB strain (wt), and rdh12 u533 zebrafish were bred and maintained according to local UCL and U.K. Home Office regulations for the care and use of laboratory animals under the Animals Scientific Procedures Act at the UCL Bloomsbury campus zebrafish facility. Zebrafish were raised at 28.5 • C on a 14 h light/10 h dark cycle. UCL Animal Welfare and Ethical Review Body approved all procedures for experimental protocols, in addition to the U.K. Home Office (License no. PPL PC916FDE7). All approved standard protocols followed the guidelines of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research Ethics. Zebrafish were terminally anaesthetised in 0.2 mg/mL Tricaine (MS−222) for sample collection.

CRISPR-Cas9 Generation of rdh12 Fish
CRISPR guides were designed using Benchlings CRISPR Guide Design Software, targeting exon 1 of rdh12, and the following guide was used: 5 -TGGCGTTCGCGGCGGGTTTAGGG. Mutagenesis was carried out as previously described [33]. Mutations were analysed by Sanger sequencing using the forward primer 5 -GGAGGCTGCTGAACACATTC and reverse primer 5 -CGATTTCTGGAGCAGATCATGTC. Injected zebrafish were raised to adulthood (F0) and outcrossed with wild-type fish to produce the F1 generation. F1 finclip DNA was assessed for heterozygous mutation carriers. These fish were then outcrossed with wild-type fish to produce an F2 generation. Heterozygous fish carrying the same mutation from the F2 generation were then in-crossed to produce the F3 generation. Homozygous fish and their wt siblings were used for all further characterisations.

Immunohistochemistry
Cryosections were prepared from adult enucleated eyes or whole embryos and immunostained as previously described [36]. Primary antibodies used were 4D2 (1:200; Abcam), 1D4 (1:200; Abcam), and anti-blue opsin (1:200; gifted by Professor David Hyde, University of Notre Dame). Slides were incubated with primary antibodies overnight at 4 • C, washed, and then incubated with appropriate Alex Fluor secondary antibodies (Thermo Fisher Scientific) for 2 h at RT. Slides were imaged on Leica LSM 710 confocal microscope.

TUNEL Assay
TUNEL assay was performed on cryosections using the ApopTag Plus Fluorescein In Situ Apoptosis Detection Kit (Merck-Millipore, Burlington, MA, USA), according to the manufacturer's instructions. Slides were imaged on a Leica LSM 710 confocal microscope.

Retinal Histology
Whole embryos were fixed in 4% PFA/PBS overnight at 4 • C, then embedded using JB-4 embedding kit (Polysciences Inc., Warrington, PA, USA). Sections were cut at a thickness of 7 µm. For adult fish, enucleated eyes embedded for electron microscopy, as described below, were sectioned at a thickness of 1 µm. Sections were stained with 1% toluidine blue and imaged on an a Axioplan 2 microscope (ZEISS Microscopy, Jena, Germany).

HPLC Analysis of atRAL Levels in Zebrafish
Retinoids were extracted according to the protocol by Costaridis et al. [39]. Approximately 250 embryos or adult enucleated eyes were collected and sonicated in 1 mL stabilising buffer (PBS, 0.5% ascorbic acid, 0.5% EDTA, 0.3% sodium sulphate, pH 7.3). Sample was extracted twice in 2 mL extraction solvent (8:1 ethyl acetate:methyl acetate plus 0.5% butylated hydroxytoluene), with vigorous shaking for 30 min. Extracts were combined and dried under nitrogen gas and redissolved in acetonitrile. HPLC analysis was performed using a Waters ACQUITY ® UPLC ® . Elution was monitored at 360 nm for atRAL with a Waters ACQUITY 2996 PDA. The stationary phase was Waters ACQUITY UPLC BEH C18 1.7 µm, 2.1 × 50 mm column and the mobile phase consisted of acetonitrile/water 0.1% formic acid (87.5:12.5). The analysis lasted 8 min, with a flow rate of 0.25 mL/min and a 10 µL injection.

Transmission Electron Microscopy
Enucleated eyes from 12 mpf zebrafish were fixed and embedded as previously described [33]. Blocks were sectioned using a UC7 ultramicrotome (Leica Microsystems) and 70 nm sections were picked up on Formvar-coated 2 mm slot copper grids (Gilder Grids Ltd., Grantham, UK) and post-stained with Reynolds lead citrate for 5 min. Sections were viewed using a 120 kV Tecnai G2 Spirit transmission electron microscope (FEI Company, Eindhoven, The Netherlands) and images were captured using a OneView UltraScan ® 4000 camera (Gatan Inc., Pleasanton, CA, USA) and Serial EM software [40]. The images were stitched together using IMOD [41]. Visualisation was done in 3dmod. Images were analysed using ImageJ.

Statistical Analysis
All statistical analysis was performed using GraphPad Prism 8. All data are expressed as mean ± SEM of at least three independent experiments. For all analyses, Shapiro-Wilk normality test was initially carried out to determine if data are normally distributed, and the appropriate statistical test was chosen. For comparison between two groups, data were analysed either using paired or unpaired t-test, depending on experimental design or Mann-Whitney test. For more than two groups, statistical significance was analysed using one-way ANOVA, followed by Sidak's multiple comparison test. For grouped analyses, two-way ANOVA with Dunnett's multiple comparison test was used. A p-value of <0.05 was considered significant.