nAMD patients (n = 8) and age-related cataracts as controls (n = 8) were recruited from the Ophthalmology Department of the Second Hospital of Hebei Medical University. Informed written consent was obtained from each participant. The study was approved by the Ethics Committee of the Second Hospital of Hebei Medical University. The detailed clinical characteristics of the participants were shown in Supplementary Table S1. No significant differences between the two groups in terms of age, sex, BMI, smoking, or alcohol consumption were observed.

The inclusion criteria were as follows: (1) age over 60 years; (2) a first-time diagnosis of nAMD based on the Age-Related Macular Degeneration Preferred Practice Pattern (Flaxel et al., 2020), or a diagnosis of cataract requiring surgical treatment; and (3) no history of ocular treatments such as intravitreal anti-VEGF injections, photodynamic therapy, or photocoagulation.

Exclusion criteria were as follows: (1) the existence of ocular diseases other than cataracts or nAMD; (2) a history of intraocular disease or prior ocular surgery; and (3) the presence of serious systemic conditions such as cancer, autoimmune diseases, congestive heart failure, hypertension, or diabetes mellitus.

Aqueous humor samples were collected from nAMD patients prior to their first intravitreal injection of anti-VEGF drugs, and from control subjects during cataract surgery. Using a 1 cc syringe and a cannula, 50–150 μL of aqueous humor was obtained from each participant and immediately transferred into Eppendorf tubes, then stored at −80 °C. For proteomic sequencing, every four individual samples in each group were randomly pooled and mixed, resulting in two paired samples.

Shanghai Applied Protein Technology Co., Ltd. conducted the TMT-based quantitative proteomic analysis, following the procedures detailed in previous research (Liu et al., 2021). Mascot 2.2 and Proteome Discoverer 1.4 were utilized for protein identification and quantification.

Human umbilical vein endothelial cells (HUVECs) were cultured in endothelial cell medium, supplemented with 10% fetal bovine serum (FBS, Gibco) and ECGS (ECM, ScienCell), at 37 °C in a humidified 5% CO₂ atmosphere. ARPE-19 cells (ATCC) were grown in DMEM/F12 medium containing 10% FBS.

Transfection was performed once the cells attained 70%–80% confluence. RARRES1 small interfering RNA (si-RARRES1) and si-NC (negative control siRNA), as well as pcDNA3.1 control vector and the pcDNA3.1-RARRES1 overexpression plasmid, were designed and provided by GenePharma. Lipofectamine 2000 (Invitrogen) was employed for transfection, following the manufacturer’s instructions. The siRNA or plasmid complexes were premixed and added to cells in 6-well plates. The final concentration of siRNA was 50 nM, and plasmids were transfected at a concentration of 2 µg per well. The medium was exchanged for fresh culture medium with 10% FBS and no antibiotics, 6 h post-transfection.

HUVECs were collected for subsequent assays 48 h after transfection. When ARPE-19 cells reached 90%–95% confluence, they were starved in medium containing 1% FBS for 24 h, followed by H₂O₂ treatment for 48 h. H₂O₂ was freshly prepared at a concentration of 200 µM and added directly to the culture medium. Samples were then collected for subsequent analysis.

Cell viability was measured with the Cell Counting Kit-8 (CCK-8; Beyotime). Briefly, cells were seeded in 96-well plates at a density of 5 × 10³ cells per well in 100 μL of complete medium. After 48 h of transfection, 10 μL of CCK-8 solution was added to each well. The cells were then incubated at 37 °C for 1 h. Cell viability was determined by measuring the optical density at 450 nm using a microplate reader (BioTek). Blank wells containing medium and CCK-8 without cells were used to subtract background absorbance.

For the Transwell assay, serum-starved transfected HUVECs (5 × 10⁴ cells/well) were placed in the upper chamber (8 μm pore size, 24-well plate; Corning), with the lower chamber filled with medium containing 1% FBS. Cells were suspended in 200 μL of serum-free medium before seeding into the upper chamber. After 24 h of incubation at 37 °C in a humidified atmosphere with 5% CO₂, the cells that had migrated to the lower surface of the membrane were fixed with 4% paraformaldehyde for 15 min and stained with 0.1% crystal violet for 20 min. Non-migrated cells on the upper surface were gently removed using a cotton swab. the number of migrated cells were visualized and counted under a light microscope (Olympus).

Matrigel (Cat# 356230, Corning) was thawed overnight on ice, aliquoted into precooled 96-well plates (60 μL per well), and incubated at 37 °C for 30 min. HUVECs were transfected with siRNA or plasmid cDNA for 48 h, after which they were seeded on Matrigel at a density of 5 × 10³ cells per well in 100 μL of complete endothelial basal growth medium (EBM, supplemented with EGM SingleQuots, Lonza), and the cells were cultured at 37 °C in a 5% CO₂ atmosphere. After 18 h of incubation, tube-like structures were observed, and phase-contrast images were captured using a microscope (Olympus). Tube formation was evaluated with ImageJ software.

Apoptosis in ARPE-19 cells was assessed using an Annexin V-FITC Apoptosis Detection Kit (Cat#: E-CK-A217, Elabscience, Wuhan, China), following the manufacturer’s protocol. After reaching approximately 75% confluence, cells were exposed to 400 μmol/L H₂O₂ for 48 h to induce oxidative stress. Following treatment, cells were gently washed twice with cold PBS and harvested using trypsinization without EDTA to preserve membrane integrity. Cells were then resuspended in 1× binding buffer at a concentration of 1 × 10⁶ cells/mL. A total of 100 µL of the cell suspension was incubated with 5 µL of Annexin V-FITC and 5 µL of propidium iodide (PI) in the dark at room temperature for 15 min. Apoptotic cells were analyzed using a FACS LSRFortessa flow cytometer (BD Biosciences) within 1 h of staining.

Intracellular ROS levels were measured using DCFH-DA dye (Cat#: E-BC-K138-F, Elabscience). Cells were incubated with DMEM containing 10 µM DCFH-DA for 30 min at 37 °C in the dark to prevent photobleaching. After incubation, cells were washed three times with serum-free DMEM to remove excess dye. The labeled cells were subsequently viewed using a fluorescence microscope (Nikon) using an excitation wavelength of 488 nm and an emission wavelength of 525 nm.

C57BL/6 male mice (6–8 weeks old) (Beijing HFK Bioscience) were housed under SPF conditions. They were provided with ad libitum access to food and water and maintained on a 12-h light/dark cycle. All animal experiments were approved by the Ethics Committee of the Second Hospital of Hebei Medical University.

Laser-induced CNV was induced in the mice following previously described protocols (Wang et al., 2020). The mice were anesthetized with chloral hydrate, and 0.5% tropicamide (Henry Schein) was applied to dilate the pupils. Laser photocoagulation was performed using an ophthalmic laser system (Nd, 532 nm; Coherent Novus Omni) with a 50 μm spot diameter, 100 ms duration, and 300 mW power. In each eye, five laser spots were applied around the optic disc, and the successful disruption of Bruch’s membrane was confirmed by the formation of a bubble.

Using a Harvard Apparatus microinjector and pulled glass micropipettes, intravitreal injections were performed as outlined in the previous description (Mori et al., 2001). Using a dissecting microscope, the sharpened tip of the micropipette was inserted through the sclera just behind the limbus into the vitreous cavity. Each micropipette was calibrated to deliver 2 μL of the vehicle, which contained around 1 × 10¹³ viral genome copies (vgc) of AAV2-Rarres1 and AAV2-ZsGreen, when the foot switch was activated.

The eyeballs were enucleated 7 days after laser injury and fixed in eyeball fixative solution for 24 h at 4 °C. After dehydration through graded ethanol (70%, 80%, 95%, and 100%) and embedding in paraffin, the eye tissues were sectioned to a thickness of 4 μm using a microtome. The sections were stained with hematoxylin and eosin (HE) following standard protocols and mounted with neutral gum for histological evaluation under a light microscope.

Seven days post-photocoagulation, the CNV lesion size was quantified using immunofluorescence staining of choroidal flat mounts with IB4 (FITC-IB4; Sigma-Aldrich). The flat mounts were incubated to specifically label endothelial cells within CNV lesions. After staining, samples were washed and mounted with antifade mounting medium.

TRIzol reagent (Takara) was used to extract total RNA from treated ARPE-19 cells or retinal tissues. The purity and concentration of RNA were assessed by measuring absorbance at 260/280 nm using a NanoDrop spectrophotometer (Thermo Fisher Scientific). RNA integrity was confirmed by agarose gel electrophoresis. cDNA synthesis was performed with the PrimeScript 1st strand cDNA synthesis kit (Monad Biotech). mRNA levels were quantified with a SYBR qRT-PCR kit (Monad Biotech) on a CFX Connect Real-Time PCR System (Bio-Rad). Gene expression was determined using the 2^−ΔΔCT method, normalizing to GAPDH mRNA. Primer sequences for qRT-PCR are provided in Supplementary Table S2.

The cells or triturated RPE, choroid, or scleral tissues were lysed in RIPA buffer on ice and then boiled in SDS loading buffer. The resulting proteins were resolved by 6%–12% SDS-PAGE and transferred onto nitrocellulose membranes. The membranes were blocked with 5% skim milk, cut horizontally when needed, and incubated overnight at 4 °C with the selected primary antibodies. The primary antibodies include RARRES1 (sc-390461, Santa Cruz), phospho-JNK (HY-P80831, MCE), JNK (HY-P80197, MCE), phospho-ERK (HY-P81156, MCE), ERK (HY-P80393, MCE), Nrf-2 (80593-1-RR, Proteintech), SIRT-1 (HY-P80319, MCE), Occludin (502601, ZEN), β-actin (HRP-66009, Proteintech). The membranes were washed and then incubated with the corresponding secondary antibodies for 1 h at room temperature. Protein bands were visualized with an enhanced chemiluminescence (ECL) reagent and analyzed using a Bio-Rad GelDoc imaging system. If necessary, the membranes were stripped with stripping buffer (Solarbio) for 30 min at room temperature, and additional proteins were detected on the same membranes.

Seven days after laser photocoagulation, CNV model mice were anesthetized with an intraperitoneal injection of a ketamine (80 mg/kg) and xylazine (10 mg/kg) mixture. Pupils were dilated using topical application of 1% tropicamide and 2.5% phenylephrine eye drops. Fluorescein sodium (Cat# HJ20130649; Lisheng Pharmaceutical) was intraperitoneally injected at a dose of 5 μg/g body weight. Fluorescein angiography (FFA) images were captured using a retinal angiography system (Heidelberg Engineering).

Using Prism 9 (GraphPad Software) and SPSS software, statistical analyses were carried out. Data are expressed as means ± standard deviations (SD) from at least three independent experiments. Brown-Forsythe ANOVA with Dunnett’s T3 test was used to compare the multiple groups, and t-test was used to compare two groups. A p-value of less than 0.05 was deemed statistically significant.

Proteomic analysis of aqueous humor samples from nAMD patients identified 1,241 expressed proteins, with 134 of them being differentially expressed-68 upregulated and 66 downregulated-based on a threshold of P < 0.05 and a log2 fold change >1.2 or <0.83 (Student’s t-test). Notably, RARRES1 was one of the significantly downregulated proteins in the aqueous humor of nAMD patients (Figure 1). This finding suggests that RARRES1 may play a critical role in the pathophysiology of nAMD, warranting further functional investigation.

To explore the expression profile of RARRES1 in ARPE-19 cells under both normal and oxidative stress conditions, we conducted comparative analyses using Western blotting, qPCR, and immunofluorescence staining. ARPE-19 cells were treated with different concentrations of H₂O₂ (0, 200, 400, and 600 μM) for 48 h for both Western blot and qPCR analyses. Western blot results revealed a significant downregulation of RARRES1 protein expression in H₂O₂-treated ARPE-19 cells compared to the control group (0 μM) (Figure 2A). This was further confirmed by qRT-PCR, which showed a marked reduction in RARRES1 mRNA levels following H₂O₂ exposure (Figure 2B). Immunofluorescence analysis of cells treated with 400 μM H₂O₂ revealed significantly weaker RARRES1 fluorescence intensity compared to untreated controls (Figure 2C). Notably, this expression pattern aligns with our previous findings of reduced RARRES1 levels in aqueous humor samples from patients with nAMD. These results demonstrate that oxidative stress directly suppresses RARRES1 expression in retinal pigment epithelial cells, paralleling clinical observations in nAMD.

To determine whether RARRES1 regulates H₂O₂-induced oxidative stress in ARPE-19 cells, we knocked down RARRES1 using siRNA and overexpressed RARRES1 using the pcDNA3.1 plasmid. Cells transfected with negative control siRNA (si-NC) and pcDNA3.1 served as the respective controls. RARRES1 expression was notably enhanced in comparison to the negative control group (Supplementary Figure S1). To assess whether RARRES1 protects ARPE-19 cells from H₂O₂-induced damage, we performed flow cytometry to analyze apoptosis and conducted CCK-8 assays. The results showed that knockdown of RARRES1 significantly increased H₂O₂-induced apoptosis, with fewer cells surviving H₂O₂ exposure in the RARRES1 knockdown group compared to the si-NC group. Conversely, overexpression of RARRES1 significantly suppressed H₂O₂-induced apoptosis and improved cell viability compared to the pcDNA3.1 group (Figure 3A). The DCFH-DA fluorescent probe was used to measure ROS production. As shown in Figure 3B, knockdown of RARRES1 increased ROS generation in ARPE-19 cells, whereas overexpression of RARRES1 reduced ROS production. These data indicate that RARRES1 plays a protective role against oxidative stress by reducing ROS levels and apoptosis in RPE cells.

To investigate whether RARRES1 is involved in inflammation and angiogenesis, we used qPCR to examine the expression of proangiogenic and inflammatory factors. It was found that knockdown of RARRES1 significantly upregulated the mRNA expression of VEGFA, VEGFR2, IL-1β, MCP-1, IL-8, and IL-6 in the presence of H₂O₂ (Figure 4A). In contrast, overexpression of RARRES1 significantly downregulated the mRNA levels of these genes in H₂O₂-insulted cells (Figure 4B).

Next, we assessed the protein expression of the tight junction protein Occludin via Western blot analysis. Notably, RARRES1 knockdown significantly reduced Occludin expression (Figure 4C), whereas RARRES1 overexpression restored Occludin levels in H₂O₂-insulted cells (Figure 4D).

The conditioned medium from RARRES1 knockdown cells further enhanced the angiogenic potential of HUVECs, as evidenced by increased cell numbers and junction formation (Supplementary Figure S2).

Together, these findings suggest that RARRES1 not only suppresses inflammation and angiogenic signaling but also helps maintain RPE barrier integrity under oxidative stress conditions.

Western blot analysis was conducted to explore the molecular mechanisms behind RARRES1’s antioxidative action, focusing on the JNK signaling pathway and proteins associated with oxidative stress. The results revealed that si-RARRES1 transfection significantly increased the expression of p-JNK, while notably suppressing the levels of both Nrf2 and SIRT1 in ARPE-19 cells under oxidative stress, compared to the si-NC control group (Figure 5A). In contrast, overexpression of RARRES1 in H₂O₂-treated cells significantly inhibited p-JNK activation and enhanced the expression of Nrf2 and SIRT1 relative to the pcDNA3.1 controls (Figure 5B). This suggests that RARRES1 exerts its antioxidative effects by modulating key stress-response pathways, suppressing pro-apoptotic JNK signaling while promoting protective Nrf2 and SIRT1 activity.

To investigate the effects of RARRES1 on angiogenesis, HUVECs were transfected with siRNA targeting RARRES1 and the pcDNA3.1 RARRES1 plasmid. Immunofluorescence confirmed the expression of RARRES1 in HUVECs (Supplementary Figure S3). Transfection efficiency was evaluated via Western blotting and qPCR. As shown in Supplementary Figure S4, RARRES1 expression was lower in the si-RARRES1 groups compared to the si-NC groups, and higher in the RARRES1 groups compared to the pcDNA3.1 groups, confirming successful transfection.

Cell viability was assessed using CCK-8 assays, revealing that knockdown of RARRES1 promoted the viability of HUVECs compared to the si-NC group, while overexpression of RARRES1 inhibited HUVEC viability compared to the pcDNA3.1 group (Figure 6A). The si-RARRES1 group exhibited a markedly greater number of migrating cells compared to the si-NC group, while RARRES1 overexpression markedly suppressed HUVEC migration (Figure 6B). In the tube formation assay, knockdown of RARRES1 increased the number of vessel junctions, whereas overexpression of RARRES1 reduced the number of vessel junctions (Figure 6C). Knockdown of RARRES1 resulted in upregulation of VEGFA and VEGFR2 expression, whereas RARRES1 overexpression downregulated their mRNA levels (Figure 6D). Collectively, these results demonstrate that RARRES1 negatively regulates key processes involved in angiogenesis, highlighting its potential role as an anti-angiogenic factor.

We first assessed the expression pattern of RARRES1 in the retina-choroid tissues of mice. RARRES1 was predominantly localized to the RPE layer and within CNV lesions, with its expression markedly reduced in the CNV group compared to the control (Supplementary Figure S5). To explore how RARRES1 expression changes in vivo, we established a laser-induced CNV mouse model. Western blotting and qPCR analyses revealed a significant downregulation of RARRES1 at both the protein (Figure 7A) and mRNA (Figure 7B) levels in the CNV group compared to normal mice, suggesting that suppression of RARRES1 might be involved in the pathogenesis of CNV.

Next, to investigate the role of RARRES1 in CNV formation, we used adeno-associated virus type 2 (AAV2) to construct a RARRES1 overexpression vector and a negative control. C57BL/6J mice were intravitreally injected with AAV2, and 4 weeks later, Western blotting and qPCR confirmed successful overexpression of RARRES1 in the retina-RPE-choroid tissues, as evidenced by significantly increased protein and mRNA levels compared to the control group (Supplementary Figure S6).

After confirming robust transgene expression, laser-induced CNV was induced. HE staining showed that both the length and area of CNV were substantially reduced in the AAV2-RARRES1 group compared to the control group (Figure 7C) 7 days after the model was established. CNV activity and lesion size were compared between the AAV2-RARRES1 and AAV2-ZsGreen control groups. AAV2-RARRES1 group had a smaller CNV area and volume compared to the AAV2-ZsGreen group, as shown by immunostaining with IB4 (Figure 7D). Reduced vascular leakage was found in the AAV2-RARRES1 group compared to the AAV2-ZsGreen group in the late phase of FFA (Figure 7E). Interestingly, it was found that RARRES1 knockdown reduced Occludin expression in HUVECs, whereas its overexpression increased Occludin expression (Supplementary Figure S7). These in vivo findings firmly establish that RARRES1 suppresses CNV progression, validating its potential therapeutic value in nAMD.