This research encompassed both newly diagnosed wAMD patients and control patients with simple cataracts, selected by two experts based on fundus fluorescein angiography (FFA) and indocyanine green angiography (ICGA) assessments. Aqueous humor was collected from wAMD patients during their first intravitreal injection, while controls underwent cataract surgery. Following centrifugation at 3000 rpm for 5 minutes, samples were stored at -80°C before analysis. All participants provided written, informed consent, and the study received approval from the institute’s ethics committee at Chongqing University Jiangjin Hospital (approval reference number: KY2023006).

Adult C57BL/6 mice (4-8 weeks old) were housed at Chongqing University of Arts and Sciences, National & Local Joint Engineering Research Center of Targeted and Innovative Therapeutics, Chongqing, China. Ethical approval for all animal experiments was obtained from the Experimental Animal Welfare Ethics Committee of Chongqing University of Arts and Sciences (approval reference number: CQWLDF202304), adhering to the declaration on the use of animals in ophthalmic and vision research by the Association for Research in Vision and Ophthalmology (ARVO).

After dilating pupils with compound tropicamide eye drops (Shenyang Sinqi Pharmaceutical Co., Ltd.), C57BL/6 mice underwent laser photocoagulation. Laser spots were induced in each of the four quadrants using an OculightSlx Krypton Red Laser system (power 200 mW, duration 75 ms, spot size 75 μm). One day after laser treatment, 2 μl of Metrnl (100ng/ml) or PBS was injected into the vitreous. Mice were euthanized one week later for immunohistochemistry assays and qRT‐PCR.

Primary human umbilical vein endothelial cells (HUVECs) were cultured in endothelial cell growth medium (ECM, Promocell). Human embryonic kidney 293T (HEK293T) and mouse macrophages (RAW264.7 cells) cultures were performed in DMEM (purchased from Procell Company, Wuhan, China) containing 10% FBS. Human macrophages (THP-1 cells) were cultured in RPMI-1640 (Life Technologies) with 10% FBS (Biological Industries). These cells were placed in 37°C, 5% CO₂ incubator.

All of cell lines present in this study were obtained from Zhejiang Meisen Cell Technology Co., Ltd.

Cell viability was determined using the CCK-8 assay (Beyotime). In brief, before detecting the optical density (OD) at 450 nm using a microplate reader, each culture well was added with 10μl of CCK-8 solution and was incubated for 2h at 37°C with 5% CO₂.

EdU cell proliferation staining was performed using an EdU kit (Beyotime). Briefly, cells (8 × 10³ cells/well) were cultured with/without Metrnl (100nM) on round coverslips in 24-well plates for 24 h. After fixed with 4% paraformaldehyde and permeated with 0.3% Triton X-100 for 15 min, cells were incubated with EdU for 2h. They were treated with the Click Reaction Mixture for 30 min at room temperature in a dark place before being counterstained with Hoechst 33342 for 10 minutes.

The levels of Metrnl, IL-1β, IL-6, and TNF-α in clinical aqueous humor, choroid of mice, and the media of RAW264.7 cell cultures were measured using ELISA kits (Abclonal Technology, Wuhan, China/MeiKe, Jiangsu, China) following the provided protocols.

Cells were lysed using a nuclear and cytoplasmic protein extraction kit (Beyotime, Shanghai, China). Protein concentration was determined with a BCA kit (Beyotime, Shanghai, China). Subsequently, the protein samples underwent separation through SDS-PAGE gel electrophoresis and were then transferred to PVDF membranes. After blocking with 5% skimmed milk for 1 hour, the membranes were incubated overnight at 4°C with primary antibodies: Anti-Metrnl (ABclonal, USA; 1:1000), anti-IκBα (ABclonal, USA; 1:1000), anti-NF‐κB (ABclonal, USA; 1:1000), PCNA (ABclonal, USA; 1:1000), and β-tubulin (Proteintech, USA; 1:1000). They were then treated for 1 hour at room temperature with HRP-conjugated secondary antibodies. Visualization of the bands was achieved using ECL reagents (ThermoFisher, MA, USA).

Cells were lysed with NP40 solution for 30 minutes on ice. Following centrifugation, the supernatant was collected in the EP tube. In vitro CO-IP assay, magnetic beads were incubated with the rest of supernatants for overnight at 4°C. In vivo CO-IP assay, after incubation with antibodies for 6 hours, the magnetic beads were incubated overnight with the collected supernatant. Then, these magnetic beads were separated from the mixture and resuspended in 80μl NP-40 lysis buffer, which were mixed with 20μl of SDS-loading buffer to boil for 10 minutes. Finally, samples were collected for Western blot.

The total RNA was extracted from RAW264.7and THP-1cells with RNAfast200 (Fastagen, Shanghai, China). The purity and amounts of RNA were estimated by spectrophotometer. Then, cDNA was gained though reverse transcription process with HiScript IV RT SuperMix for qPCR (Vazyme, Nanjing, China). SYBR Green Real-time PCR Master Mix (Vazyme, Nanjing, China) was used for qRT-PCR. The primer sequences were designed from the PrimerBank database ( Supplementary Table 1 ).

Initially, three kinds of conditioned medium were collected from RAW264.7 cells. The first was normal conditioned medium (N-CM), the second was conditioned medium stimulated by LPS (L-CM), and the last was conditioned medium from the cells overexpressing Metrnl, meanwhile stimulated by LPS (M-CM). In 96-well plates, 50ul of Matrigel (BD Biosciences) was applied per well and placed at 37°C for 30 min to solidify. Then P5 HUVECs (1.5× 10⁴) cells were seeded on top and categorized into three groups. Each group received 50 μl of ECM and 50 μl of the previously mentioned conditioned medium. Photographs were taken at 4h, 12h, and 24h intervals.

HUVECs were seeded in 6-well plates and incubated until reaching full confluence. After overnight incubation in serum-free medium, cells were scratched to create an artificial wound using a 200 μL pipette tip. The scratched cells were washed off with PBS, and each well received 1ml of DMEM and 1ml of the aforementioned conditioned medium, incubated at 37°C, 5% CO₂. Scratch areas were photographed at 0, 12, and 24 hours. The migration rate at 24 hours was calculated as follows: Migration rate at 24 hours = (width at 0 hours - width at 24-hour time point)/width at 0 hours × 100%.

All data were quantified using ImageJ 1.46r (National Institutes of Health, USA).

Transwell chambers were placed into 24-well plates. A total of 4 × 10⁴ HUVECs were resuspended in 200 μl of serum-free DMEM, evenly distributed in the upper chamber. The lower chambers were separately supplemented with 400 μl of DMEM and 400 μl of the aforementioned conditioned medium. After 12 hours, the membranes in the upper chambers were fixed for 30 minutes in 4% paraformaldehyde and stained for 20 minutes in Crystal Violet Staining Solution at room temperature. Using a x20 microscope, five fields of view were photographed for each membrane.

As described in previous experiments (13), we extracted the choroid-sclera complex from 6-8 week old mice. The complex was subsequently sliced into 1-mm pieces and embedded into Matrigel (BD Biosciences). Once solidified, the pieces were treated with 200 μl of various conditioned medium and 200 μl of DMEM. After 10 days, photographs of choroidal explants were captured, and the sprouting areas were quantified using ImageJ (National Institutes of Health).

Sterile coverslips were placed into a 24-well plate and incubated with poly-L-Lysine for 20 minutes. Following the protocols from Millipore, USA, coverslips were sequentially treated with glutaraldehyde, biotinylated-gelatin, and growth media. HUVECs were seeded on the coverslips and cultured until a confluent monolayer was achieved in a 37°C, 5% CO₂ incubator. After an overnight incubation in serum-free medium, various supernatants were added to the plates as a stimulant for 12 hours. Before fixing cells with 3.7% formaldehyde, the culture media was removed, and cells were incubated with fluorescein-streptavidin for 5 minutes. Subsequently, each well was supplemented with anti-VE cadherin antibody overnight at 4°C and anti-Mouse IgG, Cy3 conjugate at room temperature for 1 hour. Following a brief washing, the coverslips were inverted onto a glass slide with slide mounting media. Finally, images were captured using fluorescence microscopy.

Serum-free DMEM was prepared and supplemented with HighGene plus Transfection reagent (ABclonal). Plasmids expressing genes such as Metrnl, UCHL-1, and others were added to the mixture. After thorough mixing, the mixture was transfected into HEK293T or THP-1 cells for 24 hours. When transfecting plasmids into RAW264.7 cells, the mixture was prepared in a similar manner and incubated for 24 hours. SiRNA targeting Metrnl and UCHL-1 were transfected using a comparable approach. All steps were carried out in accordance with the provided protocols.

GraphPad Prism 9 was used for statistical analysis. The unpaired Student’s t test was applied to compare two separate experimental groups. The Analysis of Variance (ANOVA) test was utilized for nonparametric analysis of multiple comparisons. A P-value<0.05 was statistically significant.

The expression levels of inflammatory cytokines (IL-1β, IL-6, and TNF-α) and Metrnl were initially assessed using ELISA. In the samples of aqueous humor, IL-1β and IL-6 exhibited a significant increase compared to the control group, while there was no significant difference in TNF-α expression ( Figures 1A–C ). The concentration of Metrnl showed a decreasing trend in wAMD patients’ aqueous humor, although this was not statistically significant ( Figure 1D ). A similar trend was observed in the expression of IL-1β, IL-6, and TNF-α in LPS-induced macrophages (RAW 264.7 cells) and the choroid-sclera complex of CNV mice ( Figures 1E–G, I–K ). Consistent with wAMD patients’ aqueous humor, the expression of Metrnl was suppressed ( Figures 1H, L ). Furthermore, Metrnl was evaluated by western blot, revealing a decrease in the protein level of Metrnl in LPS-treated macrophages ( Figures 1M, N ). These findings suggest that both inflammatory cytokines and Metrnl play crucial roles in the occurrence and development of CNV.

To investigate the impact of Metrnl on regulating macrophage activation, we initially examined the growth and proliferative ability of macrophages after the exogenous administration of Metrnl. The EdU assay revealed that increasing the expression of Metrnl did not affect the viability and proliferative ability of macrophages ( Supplementary Figures 1D, E ). We further investigated the function of Metrnl in the production of pro-inflammatory cytokines. Firstly, we validated the efficiency of overexpression of Metrnl in macrophages ( Figures 2A, B, F, G ). Subsequently, qRT-PCR was applied to assess the effects in different treated groups. RAW264.7 cells and THP-1 cells were treated with LPS to elevate mRNA levels of IL-1β, TNF-α, and IL-6, while overexpression of Metrnl significantly reduced them ( Figures 2C–E, H–J ). These results indicate that Metrnl can modulate pro-inflammatory cytokines in macrophages.

It is well known that activated macrophages play an essential role in regulating the process of angiogenesis. In the next step, we explored the role of Metrnl in regulating angiogenesis in activated macrophages. CCK-8 and EdU assays were employed to assess the influence of exogenous Metrnl on HUVECs, and the data illustrated that Metrnl did not impact the growth and proliferative ability of HUVECs ( Supplementary Figures 1A–C ). We then used conditioned medium from macrophages to assess the migratory ability of endothelial cells through migration assays ( Figures 3F, H ) and transwell experiments ( Figures 3A, B ). The results demonstrated that Metrnl significantly inhibited inflammation-induced endothelial cell migration. Furthermore, we explored whether Metrnl could influence the angiogenesis of HUVECs ( Figures 3C–E ). As expected, the L-CM group promoted tube formation, and the M-CM group could attenuate this effect. Additionally, it is well-established that the integrity and function of blood vessels are crucial in the process of CNV, where inflammatory cytokines significantly affect vascular permeability. We further employed an in vitro vascular permeability image model to assess the cell–cell adhesion of HUVECs. The permeability among HUVECs stimulated by L-CM was significantly worse than that in N-CM, and M-CM reversed the L-CM effect ( Figures 3G, I ). Therefore, the results suggest that Metrnl could reduce the permeability of neovascularization.

Considering that Metrnl affects endothelial cell function by attenuating LPS-mediated macrophage inflammatory responses, we sought to investigate its effect on angiogenesis in isolated choroidal explants. The results showed that the sprouting area of choroidal explants was increased in the L-CM group, which was mitigated in the M-CM group ( Figures 4A, B ). Inhibiting the inflammatory response is indeed beneficial for suppressing the process of CNV. Exogenous Metrnl was given in CNV model mice by intravitreal injection to assess the effect of Metrnl in vivo. Compared with the control group, Metrnl significantly suppressed the area of CNV ( Figures 4C, D ). Moreover, Metrnl reduced the levels of IL-1β, IL-6, and TNF-α in choroid tissue from CNV mice ( Figures 4E–G ). The results imply that Metrnl suppresses the process of CNV through anti-inflammatory effects.

To assess the mechanism behind the regulatory effect of Metrnl on the secretion of pro-inflammatory factors in macrophages, mass spectrometry was applied to identify potential Metrnl-binding proteins. A total of 194 proteins were found to specifically bind to Metrnl ( Figure 5A ). After analyzing the abundance of these proteins, we identified highly enriched proteins potentially interacting with Metrnl and involved in inflammatory pathways. Next, these potential proteins were confirmed to interact with Metrnl through Co-IP assays. We constructed overexpression plasmids Myc-UCHL-1, Myc-CPT1B, and Myc-OTUD6B. Then, these plasmids were co-transfected with Flag-Metrnl in HEK293T cell lines, respectively. The Co-IP assay validated that Metrnl interacts with UCHL-1 both in vitro and in vivo ( Figures 5B–E ; Supplementary Figures 2A, B ). Additionally, we predicted two potential interacting residues between Metrnl and UCHL-1 using molecular docking techniques. The molecular docking results showed that UCHL-1 and Metrnl could interact with each other, with putative binding sites being serine 244, glutamine 247, or arginine 153 ( Figure 5F ). Research evidence has shown that UCHL-1 is involved in the regulation of inflammation through activating/inactivating the NF-κB signaling pathway. We assessed the impact of Metrnl on the regulation of the UCHL-1/NF-κB pathway in macrophages ( Figure 5G ). The results showed that Metrnl did not regulate UCHL-1 expression but affected the downstream NF-κB signaling pathway of UCHL-1. NF-κB nuclear translocation was increased in LPS-induced macrophages; nevertheless, NF-κB nuclear translocation was inhibited with Metrnl overexpression. These results suggest that Metrnl may be involved in the NF-κB signaling pathway by binding and regulating UCHL-1 activity.

To validate that Metrnl relies on UCHL-1 to regulate the activation of inflammatory signaling pathways, we conducted western blot assays. The results revealed that overexpressing the Flag-Metrnl plasmid in RAW264.7 cells inhibited NF-κB translocation from the cytoplasm to the nucleus and increased the expression of IκBα in the cytoplasm ( Figure 6A ). This phenomenon could be reversed by the overexpression of UCHL-1. Similarly, when Metrnl was inhibited in RAW264.7 cells, cytoplasmic IκBα and NF-κB were down-regulated, while nuclear NF-κB was up-regulated. Simultaneous knockdown of UCHL-1 could reverse this condition ( Figure 6B ). Moreover, the above results were verified in THP-1 cells ( Figures 6C, D ). To further examine the biological process, supernatants were collected from macrophages with knocked down Metrnl or concurrently knocked down Metrnl and UCHL-1. We treated HUVECs with these supernatants and performed tube formation assays ( Figures 6E, F ). The results illustrated that the deficiency of Metrnl affected the biological function of endothelial cells induced by the inflammation of macrophages. Taken together, our findings suggest that Metrnl inhibited the angiogenic capabilities of HUVECs by interacting with UCHL-1 in macrophages, which was linked to the NF-κB signaling pathway.