Introduction: The aims of this study were to investigate the molecular alterations of cuproptosis-related genes and to construct the cuproptosis-related circular RNA (circRNA)-microRNA (miRNA)-mRNA networks in neovascular age-related macular degeneration (nAMD). Methods: The transcriptional profiles of laser-induced choroid neovascularization (CNV) mouse models and nAMD patient samples were obtained from sequencing and from the GEO database (GSE146887), respectively. The expression levels of ten cuproptosis-related genes (FDX1, DLAT, LIAS, DLD, PDHB, MTF1, CDKN2A, GLS, LIPT1, and PDHA1) were extracted and verified in both mouse CNV models and patient peripheral blood mononuclear cells (PBMCs) samples. The cuproptosis-related circRNA-miRNA-mRNA network was further constructed based on miRNet database, the dataset GSE131646 of small RNA expression profile, and the dataset GSE140178 of circRNA expression profile in mouse CNV models. Results: The significant upregulation of Cdkn2a and Mtf1 and the downregulation of other 5 cuproptosis-related genes were verified in the mouse CNV model, but only CDKN2A significantly upregulated in PBMCs of patients with nAMD. Four miRNAs were detected in the intersection between miRNet prediction and sequencing data: miR-129-5p, miR-129-2-3p, miR-182-5p, and miR-615-3p. There were 9 circRNAs at the intersection of hsa-miR-182-5p and hsa-miR-615-3p predictions, one circRNA predicted by hsa-miR-129-5p and GSE140178 (hsa-circASH1L), and one circRNA predicted by hsa-miR-182-5p and hsa-miR-615-3p (hsa-circNPEPPS). Conclusion: This study suggested the repression of cuproptosis in nAMD pathologies and constructed a cuproptosis-related network of 8 cuproptosis-related genes, 4 miRNAs, and 11 circRNAs.

Age-related macular degeneration (AMD) is the leading cause of blindness in developed countries [1]. Late-stage AMD can progress rapidly in its choroidal neovascular (CNV) form (neovascular AMD [nAMD]). Although legal blindness and visual impairment caused by nAMD have decreased substantially with the introduction of anti-vascular endothelial growth factor treatment, the cumulative incidences of new blindness in patients receiving as-required anti-vascular endothelial growth factor therapy have been as high as 12% over 3 years [2], and undertreatment is very common in real-world practices. Given the high burden on public health, there is still an urgent need to further investigate nAMD pathologies and develop new nAMD treatments.

Copper is an indispensable trace metal element for various biological processes [3]. Reduced copper levels [4] and downregulated copper influx transporter CTR1 [5] have been observed in the retinal pigment epithelium (RPE)/choroid complex of AMD patients. Furthermore, clinical trials have shown that a higher intake of copper is significantly associated with lower risks of nAMD [6]. These findings suggest the involvement of copper in nAMD pathologies.

Recently, a new copper-mediated cell death pathway, cuproptosis, has been identified. It is characterized by direct copper binding to the lipoylated components of the tricarboxylic acid (TCA) cycle [7]. Considering that TCA-related metabolic substrates have been shown to significantly increase in patients with nAMD [8], the cuproptosis-related pathologies in nAMD need to be studied.

RNAs, such as microRNAs (miRNAs) and circular RNAs (circRNAs), have been found to regulate various forms of cell death [9]. circRNA-miRNA or circRNA-protein interactions can regulate autophagy, pyroptosis, and ferroptosis [10]. Moreover, copper-related circRNA-miRNA-mRNA networks have been constructed in various pathologies [11, 12]. For example, copper-related miRNAs, upregulated by copper deficiency and mediating the downregulation of copper protein, are well recognized in the literature [13]. Therefore, a cuproptosis-related circRNA-miRNA-mRNA network can also be expected in nAMD.

In this study, our aim was to comprehensively investigate the molecular alterations and clinical relevance of cuproptosis-related circRNA-miRNA-mRNA networks in nAMD. It was mainly based on sequencing in both laser-induced CNV models and samples from patients with nAMD. Bioinformatic analyses in several databases, such as miRNet and starBase, were also conducted. Our analysis highlights the importance of cuproptosis in nAMD development and lays a foundation for the therapeutic application of cuproptosis-related noncoding RNA regulators in nAMD.

Animal Models of Laser-Induced CNV

Laser-induced photocoagulation was performed as described before [14]. Briefly, mice (6–8 weeks, male) were anesthetized with 1% sodium pentobarbital (0.1 mL/10 g body weight; Guge Biotech, Wuhan, China), and the pupils were dilated with 1% tropicamide (Santen, Osaka, Japan). Four laser-injured spots were induced by a 532 nm laser (120 mW, 100 ms; Visulas 532S, Carl Zeiss, Jena, Germany) around the optic nerve using a slip lamp delivery system. The successful induction of CNV in mouse models was confirmed by color fundus images and/or fluorescein angiography (2% fluorescein sodium; Fluorescite, Alcon, Tokyo, Japan) 7 days after laser induction. Also at 7 days post-induction, mice were sacrificed and eyes were enucleated for subsequent analyses.

Microarray Analysis of the Mouse CNV Model

Total RNA was extracted from RPE-choroid-sclera complexes with TRIzol (ThermoFisher, Carlsbad, USA). RNA integrity was detected by Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, USA). Only samples with RNA Integrity Number ≥7 were subjected to further analysis. The libraries were constructed using TruSeq Stranded Total RNA with Ribo-Zero Gold according to the manufacturer’s instructions. Then they were sequenced on the Illumina sequencing platform and 150 bp/125 bp paired-end reads were generated. Raw reads generated during high-throughput sequencing were in the fastq format. To get the raw reads filtered, Trimmomatic software was first used for adapter removal. These analyses were provided by the commercial service of OEbiotech (Shanghai, China).

Data Acquisition and Differential Expression Analyses

The transcriptional profiles for nAMD patient samples were obtained from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The dataset GSE146887 [15], deposited in GPL18573 Illumina NextSeq 500 (Homo sapiens), contains 4 CNV samples in the macular regions and 4 healthy controls. It provides a high-resolution RNA-sequencing-based transcriptional signature of CNV membranes in patients with nAMD. Expression levels of ten cuproptosis-related genes were extracted: FDX1, DLAT, LIAS, DLD, PDHB, MTF1, CDKN2A, GLS, LIPT1, and PDHA1. Differential expression was considered statistically significant under the criteria of |Log2 (fold change)| >1 and p < 0.05.

Ophthalmic Examination of the nAMD Patients and the Collection of Their Peripheral Blood Mononuclear Cells

Patients with nAMD were recruited and examined in the Department of Ophthalmology, Shanghai General Hospital. They should be older than 55 years old and had reduced or distorted central vision. The diagnosis of nAMD was made on CNV evidences from advanced diagnostic measures such as spectral domain optical coherence tomography (OCT; Spectralis HRA, Heidelberg, Heidelberg, Germany) with fundus fluorescein angiography (Spectralis HRA) and/or swept-source OCT angiography (SVision, Luoyang, China) [16]. Cases with CNV secondary to other ocular pathologies such as pathological myopia or cases combined with other ocular disorders such as glaucoma were excluded. Patients with age-related cataract were recruited into the control group, with age matched with those in the nAMD group. Comprehensive ocular examinations were applied to exclude any other ophthalmic disorder, including slit lamp examination, tonometry (TX-20, Canon, Tokyo, Japan), spectral domain OCT, swept-source OCT angiography, and fundus photography (CLARUS 500, Carl Zeiss, Aalen, Germany). Both groups excluded patients with severe systematic disorders such as cancer.

Blood samples were drawn and PBMCs were isolated using a standard protocol as described before [17]. Briefly, blood cells were collected by centrifugation at 500 g for 5 min and red blood cells were removed with lysis (C3702, Beyotime, Shanghai, China). Remaining PBMCs were obtained by centrifugation at 500 g at 4°C, washed by phosphate buffered saline, and resuspended in RZ reagent (RK145, TianGen Biotech, Beijing, China) for subsequent RNA extraction.

RNA Extraction and Quantitative Real-Time Reverse Transcription-Polymerase Chain Reaction (RT-qPCR)

Total RNA was extracted from RPE-choroid-sclera complexes with an RNAprep Kit (DP419, TianGen Biotech). cDNA was synthesized with a PrimeScript RT reagent Kit (RR047Q, Takara, Shiga, Japan) and analyzed by RT-qPCR (Applied Biosystems, Foster City, USA) according to the manufacturer’s instructions. ΔΔCt method was used to obtain the fold changes in RNA expression. Primer sequences are listed in Table 1.

Table 1.

Primers for RT-qPCR validation

TargetForward primersReverse primers
Gapdh 5′-AGG​TCG​GTG​TGA​ACG​GAT​TTG-3′ 5′-TGT​AGA​CCA​TGT​AGT​TGA​GGT​CA-3′ 
Fdx1 5′-CAA​GGG​GAA​AAT​TGG​CGA​CTC-3′ 5′-TTG​GTC​AGA​CAA​ACT​TGG​CAG-3′ 
Dlat 5′-CTT​TAG​CCT​CCA​AAG​CGA​GAG-3′ 5′-AGA​TTG​TAA​ATG​TTC​CAC​CCT​GG-3′ 
Lias 5′-CCT​GGG​GTC​CCG​GAT​ATT​TG-3′ 5′-GAA​GGT​CTG​GTC​CAT​TAT​GCA​A-3′ 
Dld 5′-GAG​CTG​GAG​TCG​TGT​GTA​CC-3′ 5′-CCT​ATC​ACT​GTC​ACG​TCA​GCC-3′ 
Pdhb 5′-AGG​AGG​GAA​TTG​AAT​GTG​AGG​T-3′ 5′-ACT​GGC​TTC​TAT​GGC​TTC​GAT-3′ 
Mtf1 5′-ACA​CCT​TCG​TCT​GTA​ATC​AGG​A-3′ 5′-CTG​CAC​GTC​ACA​CTC​AAA​TGG-3′ 
Cdkn2a 5′-CGC​AGG​TTC​TTG​GTC​ACT​GT-3′ 5′-TGT​TCA​CGA​AAG​CCA​GAG​CG-3′ 
Gls 5′-CTA​CAG​GAT​TGC​GAA​CAT​CTG​AT-3′ 5′-ACA​CCA​TCT​GAC​GTT​GTC​TGA-3′ 
Lipt1 5′-TGC​TTC​CGA​TTA​CTT​TGT​CAG​C-3′ 5′-TCC​AGT​CTT​CAA​AAG​CCA​GAT​TT-3′ 
Pdha1 5′-TGT​GAC​CTT​CAT​CGG​CTA​GAA-3′ 5′-TGA​TCC​GCC​TTT​AGC​TCC​ATC-3′ 
GAPDH 5′-CAT​GAG​AAG​TAT​GAC​AAC​AGC​CT-3′ 5′-AGT​CCT​TCC​ACG​ATA​CCA​AAG​T-3′ 
FDX1 5′-TTC​AAC​CTG​TCA​CCT​CAT​CTT​TG-3′ 5′-TGC​CAG​ATC​GAG​CAT​GTC​ATT-3′ 
DLAT 5′-CGG​AAC​TCC​ACG​AGT​GAC​C-3′ 5′-CCC​CGC​CAT​ACC​CTG​TAG​T-3′ 
LIAS 5′-GTA​TGT​GAG​GAA​GCT​CGA​TGT​C-3′ 5′-CAC​CCA​TCA​ACA​TGA​TCG​TGG-3′ 
DLD 5′-CTC​ATG​GCC​TAC​AGG​GAC​TTT-3′ 5′-GCA​TGT​TCC​ACC​AAG​TGT​TTC​AT-3′ 
PDHB 5′-GCA​GCA​GTG​CTA​TCT​AAA​GAA​GG-3′ 5′-CCA​GGA​AAT​TGA​ACG​CAG​GAC-3′ 
MTF1 5′-CAG​TGC​GGA​GAA​CAC​TTG​C-3′ 5′-TGC​ACA​TAA​CCC​TGG​GAC​ATT-3′ 
CDKN2A 5′-GAT​CCA​GGT​GGG​TAG​AAG​GTC-3′ 5′-CCC​CTG​CAA​ACT​TCG​TCC​T-3′ 
GLS 5′-AGG​GTC​TGT​TAC​CTA​GCT​TGG-3′ 5′-ACG​TTC​GCA​ATC​CTG​TAG​ATT​T-3′ 
LIPT1 5′-TTG​CTA​AAG​AGC​CCT​TAC​CAA​G-3′ 5′-TCA​TCC​GTT​GGG​TTT​ATT​AGG​TG-3′ 
PDHA1 5′-TGG​TAG​CAT​CCC​GTA​ATT​TTG​C-3′ 5′-ATT​CGG​CGT​ACA​GTC​TGC​ATC-3′ 
mmu_circZNF644 5′-AGT​CTC​TCC​TCA​CAA​GAC​CCA-3′ 5′-GCA​CCA​GTA​ATG​TCG​GTG​TTT-3′ 
mmu_circSHOC2 5′-ATC​GCC​TTG​GCC​TGA​GAT​AC-3′ 5′-GGA​GTC​ACA​TCC​GTC​CGT​TA-3′ 
hsa_circTKT 5′-GAC​CGC​TTC​ATC​GAG​TGC​TA-3′ 5′-CGG​TGA​AAG​CTT​GTT​TCG​GG-3′ 
hsa_circCREB3L2 5′-AGG​ACA​CTT​GTA​CTT​AGT​TTG​GG-3′ 5′-GTC​TGA​GGC​AGG​CTG​AAG​G-3′ 
hsa_circTCP1 5′-AGG​CCC​AGG​TTA​ACC​CAG​AA-3′ 5′-ATA​ACG​CAC​TGC​TTC​CTT​GC-3′ 
hsa_circATXN1 5′-ACC​AGT​ACA​AGT​TTA​TTG​TTT​CAG​T-3′ 5′-TGC​AGG​CTG​AAA​TCC​ACT​CT-3′ 
hsa_circMAML1 5′-TCC​CCA​TAT​TCT​TCT​ACT​GCC​C-3′ 5′-CCA​TTG​TTT​TCC​GCG​CTA​CC-3′ 
hsa_circZNF644 5′-CCG​TTG​ATG​TAT​CAG​CCA​CA-3′ 5′-TGG​CAA​GCC​CAT​TTA​ACA​CA-3′ 
hsa_circPRKDC 5′-TGG​AGG​CTC​TTC​TGT​GAT​TTT​GA-3′ 5′-CAT​GGG​GAA​GTG​AGC​AAC​GA-3′ 
hsa_circPMAIP1 5′-CAG​CCT​AGA​GGC​AGC​TAT​TT-3′ 5′-GGA​GTC​CCC​TCA​TGC​AAG​TT-3′ 
hsa_circSHOC2 5′-ACA​GAC​TGT​CAG​CAA​TAC​CCA-3′ 5′-TGG​ATC​TAA​AAG​CAC​TAT​GCC​CA-3′ 
hsa_circNPEPPS 5′-TGG​TAG​GAA​AGC​TGC​TTG​GAA-3′ 5′-AAG​GTT​CCC​GTA​CCT​GTT​TGC-3′ 
hsa_circASH1L 5′-AGG​GCA​GTA​CAG​ACT​TTG​GC-3′ 5′-TCT​CCC​TGT​GAA​TGA​AGA​CGG-3′ 
TargetForward primersReverse primers
Gapdh 5′-AGG​TCG​GTG​TGA​ACG​GAT​TTG-3′ 5′-TGT​AGA​CCA​TGT​AGT​TGA​GGT​CA-3′ 
Fdx1 5′-CAA​GGG​GAA​AAT​TGG​CGA​CTC-3′ 5′-TTG​GTC​AGA​CAA​ACT​TGG​CAG-3′ 
Dlat 5′-CTT​TAG​CCT​CCA​AAG​CGA​GAG-3′ 5′-AGA​TTG​TAA​ATG​TTC​CAC​CCT​GG-3′ 
Lias 5′-CCT​GGG​GTC​CCG​GAT​ATT​TG-3′ 5′-GAA​GGT​CTG​GTC​CAT​TAT​GCA​A-3′ 
Dld 5′-GAG​CTG​GAG​TCG​TGT​GTA​CC-3′ 5′-CCT​ATC​ACT​GTC​ACG​TCA​GCC-3′ 
Pdhb 5′-AGG​AGG​GAA​TTG​AAT​GTG​AGG​T-3′ 5′-ACT​GGC​TTC​TAT​GGC​TTC​GAT-3′ 
Mtf1 5′-ACA​CCT​TCG​TCT​GTA​ATC​AGG​A-3′ 5′-CTG​CAC​GTC​ACA​CTC​AAA​TGG-3′ 
Cdkn2a 5′-CGC​AGG​TTC​TTG​GTC​ACT​GT-3′ 5′-TGT​TCA​CGA​AAG​CCA​GAG​CG-3′ 
Gls 5′-CTA​CAG​GAT​TGC​GAA​CAT​CTG​AT-3′ 5′-ACA​CCA​TCT​GAC​GTT​GTC​TGA-3′ 
Lipt1 5′-TGC​TTC​CGA​TTA​CTT​TGT​CAG​C-3′ 5′-TCC​AGT​CTT​CAA​AAG​CCA​GAT​TT-3′ 
Pdha1 5′-TGT​GAC​CTT​CAT​CGG​CTA​GAA-3′ 5′-TGA​TCC​GCC​TTT​AGC​TCC​ATC-3′ 
GAPDH 5′-CAT​GAG​AAG​TAT​GAC​AAC​AGC​CT-3′ 5′-AGT​CCT​TCC​ACG​ATA​CCA​AAG​T-3′ 
FDX1 5′-TTC​AAC​CTG​TCA​CCT​CAT​CTT​TG-3′ 5′-TGC​CAG​ATC​GAG​CAT​GTC​ATT-3′ 
DLAT 5′-CGG​AAC​TCC​ACG​AGT​GAC​C-3′ 5′-CCC​CGC​CAT​ACC​CTG​TAG​T-3′ 
LIAS 5′-GTA​TGT​GAG​GAA​GCT​CGA​TGT​C-3′ 5′-CAC​CCA​TCA​ACA​TGA​TCG​TGG-3′ 
DLD 5′-CTC​ATG​GCC​TAC​AGG​GAC​TTT-3′ 5′-GCA​TGT​TCC​ACC​AAG​TGT​TTC​AT-3′ 
PDHB 5′-GCA​GCA​GTG​CTA​TCT​AAA​GAA​GG-3′ 5′-CCA​GGA​AAT​TGA​ACG​CAG​GAC-3′ 
MTF1 5′-CAG​TGC​GGA​GAA​CAC​TTG​C-3′ 5′-TGC​ACA​TAA​CCC​TGG​GAC​ATT-3′ 
CDKN2A 5′-GAT​CCA​GGT​GGG​TAG​AAG​GTC-3′ 5′-CCC​CTG​CAA​ACT​TCG​TCC​T-3′ 
GLS 5′-AGG​GTC​TGT​TAC​CTA​GCT​TGG-3′ 5′-ACG​TTC​GCA​ATC​CTG​TAG​ATT​T-3′ 
LIPT1 5′-TTG​CTA​AAG​AGC​CCT​TAC​CAA​G-3′ 5′-TCA​TCC​GTT​GGG​TTT​ATT​AGG​TG-3′ 
PDHA1 5′-TGG​TAG​CAT​CCC​GTA​ATT​TTG​C-3′ 5′-ATT​CGG​CGT​ACA​GTC​TGC​ATC-3′ 
mmu_circZNF644 5′-AGT​CTC​TCC​TCA​CAA​GAC​CCA-3′ 5′-GCA​CCA​GTA​ATG​TCG​GTG​TTT-3′ 
mmu_circSHOC2 5′-ATC​GCC​TTG​GCC​TGA​GAT​AC-3′ 5′-GGA​GTC​ACA​TCC​GTC​CGT​TA-3′ 
hsa_circTKT 5′-GAC​CGC​TTC​ATC​GAG​TGC​TA-3′ 5′-CGG​TGA​AAG​CTT​GTT​TCG​GG-3′ 
hsa_circCREB3L2 5′-AGG​ACA​CTT​GTA​CTT​AGT​TTG​GG-3′ 5′-GTC​TGA​GGC​AGG​CTG​AAG​G-3′ 
hsa_circTCP1 5′-AGG​CCC​AGG​TTA​ACC​CAG​AA-3′ 5′-ATA​ACG​CAC​TGC​TTC​CTT​GC-3′ 
hsa_circATXN1 5′-ACC​AGT​ACA​AGT​TTA​TTG​TTT​CAG​T-3′ 5′-TGC​AGG​CTG​AAA​TCC​ACT​CT-3′ 
hsa_circMAML1 5′-TCC​CCA​TAT​TCT​TCT​ACT​GCC​C-3′ 5′-CCA​TTG​TTT​TCC​GCG​CTA​CC-3′ 
hsa_circZNF644 5′-CCG​TTG​ATG​TAT​CAG​CCA​CA-3′ 5′-TGG​CAA​GCC​CAT​TTA​ACA​CA-3′ 
hsa_circPRKDC 5′-TGG​AGG​CTC​TTC​TGT​GAT​TTT​GA-3′ 5′-CAT​GGG​GAA​GTG​AGC​AAC​GA-3′ 
hsa_circPMAIP1 5′-CAG​CCT​AGA​GGC​AGC​TAT​TT-3′ 5′-GGA​GTC​CCC​TCA​TGC​AAG​TT-3′ 
hsa_circSHOC2 5′-ACA​GAC​TGT​CAG​CAA​TAC​CCA-3′ 5′-TGG​ATC​TAA​AAG​CAC​TAT​GCC​CA-3′ 
hsa_circNPEPPS 5′-TGG​TAG​GAA​AGC​TGC​TTG​GAA-3′ 5′-AAG​GTT​CCC​GTA​CCT​GTT​TGC-3′ 
hsa_circASH1L 5′-AGG​GCA​GTA​CAG​ACT​TTG​GC-3′ 5′-TCT​CCC​TGT​GAA​TGA​AGA​CGG-3′ 

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway Enrichment Analyses

GO and KEGG pathway enrichment analyses were performed based on the transcriptional profiles for nAMD patient samples. They were conducted to forecast functional annotations of cuproptosis-related genes with Database for Annotation, Visualization, and Integrated Discovery tool (https://david.ncifcrf.gov/).

Protein-Protein Interaction (PPI) Network Construction and Gene Cluster Identification

The online Search Tool for the Retrieval of Interacting Genes (STRING; http://string-db.org) is used to describe and display the PPI network of the cuproptosis-related genes. All the 10 cuproptosis-related genes were input into the STRING and identified significant interactions with interaction scores ≥0.9 (high confidence). The STRING analysis results were imported into Cytoscape (https://cytoscape.org/) and analyzed by the cytoHubba plug-in unit. Then, the genes were ranked according to the Maximal Clique Centrality algorithm. The genes were also analyzed in Metascape, with the Molecular Complex Detection (MCODE) algorithm to automatically extract protein complexes in it. The MCODE parameter criteria were executed by default.

circRNA-miRNA-mRNA Network Construction

miRNet (https://www.mirnet.ca/) is a miRNA-centric network visual analytics platform. The dataset GSE131646 [18], deposited in GPL19057 Illumina NextSeq 500 (Mus musculus), contains small RNA expression profile in mouse CNV models (n = 3 for CNV models and n = 3 for controls). Based on miRNet and GSE131646, altered miRNAs with |Log2 (fold change)| >1.5 and p < 0.05 were reserved in miRNA prediction. Competing miRNAs-cuproptosis regulatory networks were predicted by Cytoscape.

The dataset GSE140178 [19] contains the circRNA expression profile in mouse CNV models (n = 3 for CNV models and n = 3 for controls). It was deposited in GPL21826 074663 Arraystar Mouse circRNA microarray V2 platform and altered circRNAs with |Log2 (fold change) | >1.5 and p < 0.05 were reserved. The top 20 significantly up- and downregulated circRNAs in CNV models were extracted from GSE140178. They were intersected with circRNAs predicted by starBase (https://starbase.sysu.edu.cn/starbase2/index.php) from selected miRNAs to construct the upstream regulatory networks. Expression levels of the predicted circRNAs were verified in PBMCs of patients with nAMD. The final circRNA-miRNA-mRNA network was also constructed by Cytoscape.

Statistical Analyses

The Student’s t test was used to evaluate statistical differences between the study group and the control group (p < 0.05).

Cuproptosis-Related Genes Were Involved in CNV/nAMD Pathologies

The successful induction of CNV in mouse models is presented in Figure 1a. There were 5 samples in the control group and 6 samples in the laser-induced CNV group used in RT-qPCR verification. A total of 16 PBMC samples were collected from 16 patients with nAMD (10 females and 6 males, 72.06 ± 7.94 years old). Six patients without clinical fundus pathologies were recruited (4 female and 2 male, 74.83 ± 10.32 years old) into the control group.

Fig. 1.

Expression profiles in cuproptosis-related genes in the mouse CNV model. a Confirmation of successful induction of CNV in mouse. The left is a color fundus image and the right is a fluorescein angiography image. b The heatmap of overall differential expression signatures. c The differential expression signatures of cuproptosis-related genes. d RT-qPCR verified the upregulation of Cdkn2a and Mtf1 and the downregulation of 5 cuproptosis-related genes (n = 5 for the control group, n = 6 for the CNV group).

Fig. 1.

Expression profiles in cuproptosis-related genes in the mouse CNV model. a Confirmation of successful induction of CNV in mouse. The left is a color fundus image and the right is a fluorescein angiography image. b The heatmap of overall differential expression signatures. c The differential expression signatures of cuproptosis-related genes. d RT-qPCR verified the upregulation of Cdkn2a and Mtf1 and the downregulation of 5 cuproptosis-related genes (n = 5 for the control group, n = 6 for the CNV group).

Close modal

Differential expression profiles were detected in mouse CNV models (Fig. 1b) and in human CNV samples (Fig. 2a, b). Expression profiles of ten cuproptosis-related genes were both detected in two sets (Fig. 1c, 2c). RT-qPCR verified the upregulation of Cdkn2a and Mtf1 (fold change = 5.756 and 1.311, respectively; p = 0.0485 and 0.0378, respectively) and the downregulation of Fdx1, Dld, Pdhb, Gls, and Lipt1 (fold change = 0.634, 0.748, 0.614, 0.634, and 0.636, respectively; p = 0.006, 0.012, 0.002, 0.010, and 0.005, respectively) in the mouse CNV model (Fig. 1d), but among them, only CDKN2A (fold change = 8.532; p = 0.037) significantly upregulated in PBMCs of patients with nAMD (Fig. 2d). The involvement of cuproptosis-related genes in CNV pathologies can be confirmed from our results.

Fig. 2.

Expression profiles of cuproptosis-related genes in the nAMD. The heatmap (a) of overall differential expression signatures and the volcano plot (b). c The differential expression signatures of cuproptosis-related genes. d RT-qPCR verified the significant upregulation of CDKN2A in PBMCs of patients with nAMD (n = 6 for the control group, n = 16 for the nAMD group). MTF1 was too lowly expressed to be detected by RT-qPCR and its expression levels were not shown.

Fig. 2.

Expression profiles of cuproptosis-related genes in the nAMD. The heatmap (a) of overall differential expression signatures and the volcano plot (b). c The differential expression signatures of cuproptosis-related genes. d RT-qPCR verified the significant upregulation of CDKN2A in PBMCs of patients with nAMD (n = 6 for the control group, n = 16 for the nAMD group). MTF1 was too lowly expressed to be detected by RT-qPCR and its expression levels were not shown.

Close modal

Enrichment Analysis and PPI Network Construction

Figure 3a, b present the 12, 7, and 5 statistically significant processes detected by GO, KEGG, and REACTOME analyses, respectively, in ascending order of strength. Both GO and KEGG analyses detected lipoic acid metabolism with the highest strength >3 (Fig. 3a, b). Regulation of the pyruvate dehydrogenase (PDH) complex was also recognized as the most relevant pathway by REACTOME (Fig. 3c).

Fig. 3.

Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. a GO analysis. b KEGG analysis. Both analyses detected lipoic acid metabolism with the highest strength >3. c Regulation of PDH complex was also recognized as the most relevant pathway by REACTOME.

Fig. 3.

Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. a GO analysis. b KEGG analysis. Both analyses detected lipoic acid metabolism with the highest strength >3. c Regulation of PDH complex was also recognized as the most relevant pathway by REACTOME.

Close modal

The PPI network was predicted by STRING (Fig. 4a) and visualized by Cytoscape (Fig. 4b). In Figure 4a, protein is represented by differently colored nodes with some 3D protein structures known or predicted. This network had 17 edges, and its average node degree was 3.4. The local clustering coefficient was 0.733 (PPI enrichment p < 0.001). The close relationships among LIPT1, PDHB, PDHA1, DLAT, and DLD were well recognized, while GLS and FDX1 were predicted only to interact with DLD and LIAS, respectively (Fig. 4b). The remaining two proteins, CDKN2A and MTF1, were less connected with the others and were not found in this network.

Fig. 4.

The protein-protein interaction (PPI) network was predicted by STRING (a) and visualized by Cytoscape (b). This network had 17 edges, which represented protein-protein associations. In (a), known interactions were in light blue (from curated databases) or purple (experimentally determined), and predicted interactions were in green (gene neighborhood), red (gene fusion), or blue (gene co-occurrence). Other relationships, such as text mining and co-expression, were indicated in grass green and black. Its average node degree was 3.4. The local clustering coefficient was 0.733 (PPI enrichment p < 0.001). The close relationships among LIPT1, PDHB, PDHA1, DLAT, and DLD were well recognized. In (b), ranks of genes were also reflected by the colors of the nodes. More relationships with other analyzed genes, higher the rank and darker the nodes were. c PPI network predicted by Metascape.

Fig. 4.

The protein-protein interaction (PPI) network was predicted by STRING (a) and visualized by Cytoscape (b). This network had 17 edges, which represented protein-protein associations. In (a), known interactions were in light blue (from curated databases) or purple (experimentally determined), and predicted interactions were in green (gene neighborhood), red (gene fusion), or blue (gene co-occurrence). Other relationships, such as text mining and co-expression, were indicated in grass green and black. Its average node degree was 3.4. The local clustering coefficient was 0.733 (PPI enrichment p < 0.001). The close relationships among LIPT1, PDHB, PDHA1, DLAT, and DLD were well recognized. In (b), ranks of genes were also reflected by the colors of the nodes. More relationships with other analyzed genes, higher the rank and darker the nodes were. c PPI network predicted by Metascape.

Close modal

Metascape also showed a PPI network based on the physical interactions in STRING (physical score >0.132) and BioGrid (Fig. 4c). In addition, it retained glyoxylate metabolism and the glycine degradation (Log10 [p value] = −15.1), acetyl-CoA biosynthetic process from pyruvate (Log10 [p value] = −14.3), and the acetyl-CoA biosynthetic process (Log10 [p value] = −12.8) as functional descriptions for these five MCODE components.

miRNA Prediction

There were 395 miRNAs detected and 30 transcription factors predicted to interact with cuproptosis-related genes by miRNet (Fig. 5). The miRNA-mRNA network had a total of 763 edges.

Fig. 5.

miRNA prediction by miRNet. There were 395 miRNAs (blue dots) detected and 30 transcription factors (yellow dots) predicted to interact with cuproptosis-related proteins.

Fig. 5.

miRNA prediction by miRNet. There were 395 miRNAs (blue dots) detected and 30 transcription factors (yellow dots) predicted to interact with cuproptosis-related proteins.

Close modal

To narrow the scope of miRNA prediction, GSE131646 [18] of miRNA sequencing in mouse CNV models was screened (Fig. 6a, b). There were 4 upregulated miRNAs and 38 downregulated miRNAs in the mouse CNV models (Fig. 6c). The intersection between the miRNet prediction and the dataset highlighted four miRNAs: miR-129-5p, miR-129-2-3p, miR-182-5p, and miR-615-3p (Fig. 6d). The integrating miRNA-mRNA network of these selected miRNAs and cuproptosis-related mRNAs was generated using Cytoscape (Fig. 6e). Considering the negative nature of miRNA-mRNA regulation, the focus should be placed on three pairs centered on MTF1.

Fig. 6.

Narrow down the scope of miRNA prediction. a, b GSE131646 of miRNA sequencing in mouse CNV models was screened. c There were four upregulated miRNAs and 38 downregulated miRNAs in the mouse CNV models. d The intersection between the miRNet prediction and the GSE131646 dataset highlighted 4 miRNAs: miR-129-5p, miR-129-2-3p, miR-182-5p, and miR-615-3p. e The integrating miRNA-mRNA network of these selected miRNAs and cuproptosis-related mRNAs were generated by Cytoscape. Arrows besides showed the up-/downregulation of the corresponding elements, indicated by RT-qPCR verifications or sequencing data.

Fig. 6.

Narrow down the scope of miRNA prediction. a, b GSE131646 of miRNA sequencing in mouse CNV models was screened. c There were four upregulated miRNAs and 38 downregulated miRNAs in the mouse CNV models. d The intersection between the miRNet prediction and the GSE131646 dataset highlighted 4 miRNAs: miR-129-5p, miR-129-2-3p, miR-182-5p, and miR-615-3p. e The integrating miRNA-mRNA network of these selected miRNAs and cuproptosis-related mRNAs were generated by Cytoscape. Arrows besides showed the up-/downregulation of the corresponding elements, indicated by RT-qPCR verifications or sequencing data.

Close modal

circRNA Prediction, Validation, and circRNA-miRNA-mRNA Network Construction

The starBase reported 83, 144, and 16 human circRNAs interacting with hsa-miR-129-5p, hsa-miR-182-5p, and hsa-miR-615-3p, respectively. The 100 circRNAs with the highest prediction degrees and their interactions with miRNAs are presented in Figure 7a. The target of miR-129-2-3p cannot be reported by starBase.

Fig. 7.

circRNA prediction. a One hundred circRNAs with the highest prediction degrees among the starBase predicted circRNAs interacting with hsa-miR-129-5p, hsa-miR-182-5p, and hsa-miR-615-3p. Their interactions with miRNAs were also presented. b Interactions between starBase predicted circRNAs and the top 20 up-/downregulated circRNAs from GSE140178. As circRNA IDs from different databases or sequencing platforms were named differently, intersections of predicted circRNAs were made based on their gene symbols. For example, though 40 circRNAs were extracted from GSE140178 in total, both mmu_circRNA_26681 and mmu_circRNA_26686 were labeled as circMsh3, and only 39 gene symbols from GSE140178 were used in the intersection. c Details about the selected circRNAs were listed. There were 9 circRNAs at the intersection of hsa-miR-182-5p and hsa-miR-615-3p predictions, which were listed in red. The only circRNA predicted by hsa-miR-129-5p and GSE140178 was hsa-circRNA8777 (gene symbol: ASH1L, in yellow), and the only circRNA predicted by hsa-miR-182-5p and hsa-miR-615-3p was hsa-circRNA3411 (gene symbol: NPEPPS, in green).

Fig. 7.

circRNA prediction. a One hundred circRNAs with the highest prediction degrees among the starBase predicted circRNAs interacting with hsa-miR-129-5p, hsa-miR-182-5p, and hsa-miR-615-3p. Their interactions with miRNAs were also presented. b Interactions between starBase predicted circRNAs and the top 20 up-/downregulated circRNAs from GSE140178. As circRNA IDs from different databases or sequencing platforms were named differently, intersections of predicted circRNAs were made based on their gene symbols. For example, though 40 circRNAs were extracted from GSE140178 in total, both mmu_circRNA_26681 and mmu_circRNA_26686 were labeled as circMsh3, and only 39 gene symbols from GSE140178 were used in the intersection. c Details about the selected circRNAs were listed. There were 9 circRNAs at the intersection of hsa-miR-182-5p and hsa-miR-615-3p predictions, which were listed in red. The only circRNA predicted by hsa-miR-129-5p and GSE140178 was hsa-circRNA8777 (gene symbol: ASH1L, in yellow), and the only circRNA predicted by hsa-miR-182-5p and hsa-miR-615-3p was hsa-circRNA3411 (gene symbol: NPEPPS, in green).

Close modal

We also included the top 20 up-/downregulated circRNAs reported by GSE140178 [19]. Predicted circRNAs intersected between the starBase report and this sequencing database. There were 9 circRNAs at the intersection of the hsa-miR-182-5p and hsa-miR-615-3p predictions (Fig. 7b), which are listed in Figure 7c. The only circRNA predicted by both hsa-miR-129-5p and GSE140178 was hsa-circRNA8777 (circASH1L), and the only circRNA predicted by both hsa-miR-182-5p and hsa-miR-615-3p was circRNA3411 (circNPEPPS).

We also verified the expression levels of these circRNAs in the human PBMC samples (Fig. 8). Note that each host gene could have several circRNAs, but here we only tested one circRNA from each host gene ranked first in the database. In addition, circCREB3L2, circTCP1, circATXN1, circPRKDC, and circPMAIP1 were too lowly expressed to be detected by RT-qPCR. Therefore, expression levels of only 6 circRNA were presented: circNPEPPS was significantly upregulated in the nAMD samples, while circTKT and circZNF644 were significantly downregulated.

Fig. 8.

circRNAs expression levels in nAMD. RT-qPCR verified the significant upregulation of circNPEPPS (circRNA3411) and downregulation of circMAMIL1 (circRNA6572) and circZNF644 (circRNA8595) in PBMCs of patients with nAMD. Another 5 circRNAs were too lowly expressed to be detected by RT-qPCR and their expression levels were not shown (n = 6 for the control group, n = 16 for the nAMD group).

Fig. 8.

circRNAs expression levels in nAMD. RT-qPCR verified the significant upregulation of circNPEPPS (circRNA3411) and downregulation of circMAMIL1 (circRNA6572) and circZNF644 (circRNA8595) in PBMCs of patients with nAMD. Another 5 circRNAs were too lowly expressed to be detected by RT-qPCR and their expression levels were not shown (n = 6 for the control group, n = 16 for the nAMD group).

Close modal

Finally, the cuproptosis-related circRNA-miRNA-mRNA networks in nAMD were constructed (Fig. 9). Considering the downregulation of miR-129-5p and miR-182-5p, and the negative regulation between circRNA and miRNA, only circNPEPPS was selected as the promising upstream of this regulation network. The schematic diagram of circNPEPPS/miR-129-5p/MTF1 is presented in Figure 10.

Fig. 9.

Construction of cuproptosis-related circRNA-miRNA-mRNA network with 8 cuproptosis-related genes, 4 miRNAs, and 11 circRNAs.

Fig. 9.

Construction of cuproptosis-related circRNA-miRNA-mRNA network with 8 cuproptosis-related genes, 4 miRNAs, and 11 circRNAs.

Close modal
Fig. 10.

Schematic diagram of circNPEPPS/miR-129-5p/MTF1 axis. The circNPEPPS functions as a competing sponge that binds miR-129-5p and thus prevents it from binding and suppressing its target MTF1 mRNA. The circNPEPPS, miR-129-5p, and MTF1 mRNA are presented in brown, green, and gray, respectively. The core binding sites between miR-129-5p and MTF1 mRNA 3’-UTR, and the core binding site between circNPEPPS and miR-129-5p are shown in red.

Fig. 10.

Schematic diagram of circNPEPPS/miR-129-5p/MTF1 axis. The circNPEPPS functions as a competing sponge that binds miR-129-5p and thus prevents it from binding and suppressing its target MTF1 mRNA. The circNPEPPS, miR-129-5p, and MTF1 mRNA are presented in brown, green, and gray, respectively. The core binding sites between miR-129-5p and MTF1 mRNA 3’-UTR, and the core binding site between circNPEPPS and miR-129-5p are shown in red.

Close modal

In this study, we explored the expression signature of ten cuproptosis-related genes in mouse CNV models and verified them in nAMD samples. Functional analyses showed enrichment of lipoic acid metabolism. In addition, we predicted upstream circRNA/miRNA regulators. A cuproptosis-related circRNA-miRNA-mRNA network was constructed to provide promising regulatory targets for nAMD treatment.

Several metal elements have been identified as beneficial supplements for AMD prevention, such as zinc [20], calcium [21], and copper [6]. Among them, calcium is reported to delay AMD progression by enhancing copper metabolism [22]. Iron is known to induce RPE death and further AMD development via ferroptosis [23]. Copper is involved in iron oxidation and is frequently co-localized with iron within mitochondria [24]. Previously, copper has been shown to stimulate blood vessel formation in the avascular cornea of rabbits [25]. Copper-chitosan treatments have been demonstrated to promote corneal epithelial wound healing via nitric oxide metabolism [26]. However, copper nanoparticles and ions induce retinal defects and retinal cell apoptosis via upregulating unfold protein responses and reactive oxygen species [27]. Copper also participates in high glucose-induced inflammation in RPE cells, which can be rescued by copper chelation [28]. The role of copper-induced cell death, namely cuproptosis, has not yet been well studied in nAMD.

Cuproptosis targets lipoylated TCA cycle proteins and causes subsequent toxic cell stress [7]. There are seven positive regulators (FDX1, LIAS, LIPTS, DLD, DLAT, PDHA1, and PDHB) and three negative regulators (MTF1, GLS, and CDKN2A) in cuproptosis [7]. We found that CDKN2A was consistently upregulated in both mouse CNV models and nAMD human tissues. Cyclin dependent kinase inhibitor 2 (CDKN2A/Cdkn2a), also known as P16INK4A/p16Ink4a (p16), is an important tumor-suppressor gene [29] and a senescence maker [30]. Life-long removal of the p16-positive cell can delay the onset of these phenotypes, and late-life clearance can reverse established age-related disorders [31]. The upregulation of CDKN2A in nAMD is in accordance with its role in aging; therefore, downregulation of CDKN2A might be a promising strategy for nAMD treatment.

In addition, in this study, Pdhb, Gls, and Mtf1 were significantly differentially expressed in mouse CNV models. An MTF1-containing small molecule compound, E3330, can reduce CNV [32], indicating the potential treatment effect of MTF1. In addition, the products of all three genes, along with Cdkn2a, are part of the PDH complex rather than the lipoic acid pathway. This is consistent with previous findings regarding PDH kinase 1 phosphorylation and PDH complex inhibition in nAMD [33]. Considering these findings, copper-driven TCA cycle disturbance can be confirmed in nAMD.

Furthermore, we identified upstream miRNA regulators in nAMD pathology. Four of the predicted miRNAs, miR-129-5p, miR-129-2-3p, and miR-182-5p, were downregulated, while miR-615-3p was upregulated. In particular, both miR-129-2-3p and miR-129-5p belong to the miR-129 family. Considering the negative association between miRNAs and mRNAs in sponging relation, along with the upregulation of their common target Mtf1, miR-129-5p, miR-129-2-3p, and miR-182-5p can be regarded as three promising miRNA targets (Fig. 6e). In addition, miR-129-5p [34], miR-129-2-3p [35], and miR-182-5p [36] have all been reported to inhibit angiogenesis in cancers, highlighting the treatment potential of these miRNAs in nAMD.

circRNAs are a special type of noncoding RNA with covalently closed loops, and they act as sponges for downstream miRNAs. In our cuproptosis-related circRNA-miRNA-mRNA network, 11 circRNAs were recruited. Previous studies have shown that knockdown of circATXN1 inhibits the tube formation and angiogenesis of glioma-associated endothelial cells [37], in agreement with the antiangiogenesis effect of upstream miR-129-5p and miR-182-5p. circCREB3L2 was once reported in canine mammary tumors [38], while circPRKDC knockdown has been seen to promote skin wound healing by enhancing keratinocyte migration [39]. Based on the regulation principles of circRNA-miRNA-mRNA, we identified the circNPEPPS/miR-129-5p/MTF1 axis (Fig. 10) as a promising regulator in cuproptosis-related nAMD pathologies. Though no previous study on circNPEPPS is available, it can be predicted to have a similar role in CNV reduction as MTF1 [32]. It may also decrease the area of neovascularization via miR-129-5p inhibition, as in corneal tissues [40]. The biological effects of the remaining circRNAs have not yet been elucidated, calling for further studies in this area.

To the best of our knowledge, this is the first study to explore the correlations between cuproptosis-related genes and the development of AMD. Cell death is fundamental to AMD pathology and is also meaningful for treatment development [41]. The roles of apoptosis [42], necroptosis [43], and ferroptosis [44] have been well recognized in AMD. Cuproptosis is a newly identified programmed cell death that differs from any known mechanism, which may indicate a novel target for AMD therapy [9, 45]. In this study, we further constructed a regulatory network of circRNA-miRNA-mRNA, which provides a powerful tool for exploring the molecular mechanism of cuproptosis in AMD.

There were some limitations in this study. The first was the use of the whole RPE-choroid-sclera complex in RNA sequencing. The RPE-choroid or RPE-choroid-sclera complex at CNV spots may be better choices, but the RPE-choroid-sclera complex is also a well-accepted sample for CNV study [46‒48]. Also, the dataset GSE131646 [18] of small RNA profiles and the dataset GSE140178 [19] of circRNA profiles we used for bioinformatic analysis were both obtained from mouse RPE-choroid-sclera complexes. For the purpose of tissue consistency, we chose RPE-choroid-sclera complexes to explore transcriptional profiles. Second, although the validation of cuproptosis-related genes was made in PBMCs of patients with nAMD, advanced confirmation in nAMD ocular samples is warranted. Also, no dataset with a sufficiently large sample size and clinical prognostic information (something that is urgently needed for future research) was available to test the predictive/prognostic effects of these differentially expressed genes.

This study explored the role of cuproptosis in nAMD pathologies and constructed a cuproptosis-related circRNA-miRNA-mRNA network that originated from microarray sequencing and bioinformatic analyses. The identification of this network provides novel insights into the cuproptosis and noncoding RNA mechanisms of nAMD. It also provides promising targets and biomarkers for nAMD prevention and treatment.

This study was reviewed and approved by the Institutional Review Board at Shanghai General Hospital (2023SQ099) and was performed adhering to the principles of the Declaration of Helsinki. Written informed consents were provided by enrolled patients. All data analyzed were de-identified for the interests of patients. Animal experiments were approved by the Institutional Review Board at Shanghai General Hospital (2019AW055), and the procedures were performed adhering to the principles of the Association for Research in Vision and Ophthalmology.

The authors declare no conflict of interest.

This study was supported by the National Natural Science Foundation of China (81970845), Science and Technology Innovation Action Plan of Shanghai Science and Technology Commission (2014090067000), the Science and Technology Commission of Shanghai Municipality (20Z11900400), the Shanghai Hospital Development Center (SHDC2020CR2040B, SHDC2020CR5014), and the Shanghai Collaborative Innovation Center for Translational Medicine (TM202115PT).

Conceptualization, methodology, investigation, analysis, data curation, writing, and visualization: M.Z. and J.W.; methodology, resources, data curation, writing, and visualization: Y.W. and X.W.; conceptualization, resources, manuscript review and editing, supervision, project administration, and funding acquisition: X.S. and L.H.

Additional Information

Min Zhang and Jiali Wu contributed equally to this work.

Data are not publicly available due to ethical reasons. Further inquiries can be directed to the corresponding author.

1.
Lambert
NG
,
ElShelmani
H
,
Singh
MK
,
Mansergh
FC
,
Wride
MA
,
Padilla
M
.
Risk factors and biomarkers of age-related macular degeneration
.
Prog Retin Eye Res
.
2016
;
54
:
64
102
.
2.
Mitchell
P
,
Liew
G
,
Gopinath
B
,
Wong
TY
.
Age-related macular degeneration
.
Lancet
.
2018
;
392
(
10153
):
1147
59
.
3.
Tapiero
H
,
Townsend
DM
,
Tew
KD
.
Trace elements in human physiology and pathology. Copper
.
Biomed Pharmacother
.
2003
;
57
(
9
):
386
98
.
4.
Erie
JC
,
Good
JA
,
Butz
JA
,
Pulido
JS
.
Reduced zinc and copper in the retinal pigment epithelium and choroid in age-related macular degeneration
.
Am J Ophthalmol
.
2009
;
147
(
2
):
276
82.e1
.
5.
Aloysius Dhivya
M
,
Aberami
S
,
Nikhalashree
S
,
Biswas
J
,
Liu
W
,
Irudayaraj
J
.
Copper mediates mitochondrial biogenesis in retinal pigment epithelial cells
.
Biochim Biophys Acta Mol Basis Dis
.
2020
;
1866
(
10
):
165843
.
6.
Agrón
E
,
Mares
J
,
Clemons
TE
,
Swaroop
A
,
Chew
EY
,
Keenan
TDL
.
Dietary nutrient intake and progression to late age-related macular degeneration in the age-related eye disease studies 1 and 2
.
Ophthalmology
.
2021
;
128
(
3
):
425
42
.
7.
Tsvetkov
P
,
Coy
S
,
Petrova
B
,
Dreishpoon
M
,
Verma
A
,
Abdusamad
M
.
Copper induces cell death by targeting lipoylated TCA cycle proteins
.
Science
.
2022
;
375
(
6586
):
1254
61
.
8.
Han
G
,
Wei
P
,
He
M
,
Teng
H
.
Glucose metabolic characterization of human aqueous humor in relation to wet age-related macular degeneration
.
Invest Ophthalmol Vis Sci
.
2020
;
61
(
3
):
49
.
9.
Cui
L
,
Gouw
AM
,
LaGory
EL
,
Guo
S
,
Attarwala
N
,
Tang
Y
.
Mitochondrial copper depletion suppresses triple-negative breast cancer in mice
.
Nat Biotechnol
.
2021
;
39
(
3
):
357
67
.
10.
Zhou
X
,
Zhan
L
,
Huang
K
,
Wang
X
.
The functions and clinical significance of circRNAs in hematological malignancies
.
J Hematol Oncol
.
2020
;
13
(
1
):
138
.
11.
Fu
XZ
,
Zhang
XY
,
Qiu
JY
,
Zhou
X
,
Yuan
M
,
He
YZ
.
Whole-transcriptome RNA sequencing reveals the global molecular responses and ceRNA regulatory network of mRNAs, lncRNAs, miRNAs and circRNAs in response to copper toxicity in Ziyang Xiangcheng (Citrus junos Sieb. Ex Tanaka)
.
BMC Plant Biol
.
2019
;
19
(
1
):
509
.
12.
Yuan
Y
,
Zhou
X
,
Kang
Y
,
Kuang
H
,
Peng
Q
,
Zhang
B
.
Circ-CCS is identified as a cancer-promoting circRNA in lung cancer partly by regulating the miR-383/E2F7 axis
.
Life Sci
.
2021
;
267
:
118955
.
13.
Pilon
M
.
The copper microRNAs
.
New Phytol
.
2017
;
213
(
3
):
1030
5
.
14.
Zhao
X
,
Gao
M
,
Liang
J
,
Chen
Y
,
Wang
Y
,
Wang
Y
.
SLC7A11 reduces laser-induced choroidal neovascularization by inhibiting RPE ferroptosis and VEGF production
.
Front Cell Dev Biol
.
2021
;
9
:
639851
.
15.
Schlecht
A
,
Boneva
S
,
Gruber
M
,
Zhang
P
,
Horres
R
,
Bucher
F
.
Transcriptomic characterization of human choroidal neovascular membranes identifies calprotectin as a novel biomarker for patients with age-related macular degeneration
.
Am J Pathol
.
2020
;
190
(
8
):
1632
42
.
16.
Schmidt-Erfurth
U
,
Chong
V
,
Loewenstein
A
,
Larsen
M
,
Souied
E
,
Schlingemann
R
.
Guidelines for the management of neovascular age-related macular degeneration by the European Society of Retina Specialists (EURETINA)
.
Br J Ophthalmol
.
2014
;
98
(
9
):
1144
67
.
17.
Oca
AI
,
Perez-Sala
A
,
Pariente
A
,
Ochoa
R
,
Velilla
S
,
Pelaez
R
.
Predictive biomarkers of age-related macular degeneration response to anti-VEGF treatment
.
J Pers Med
.
2021
;
11
(
12
):
1329
.
18.
Zhang
L
,
Liu
S
,
Wang
JH
,
Zou
J
,
Zeng
H
,
Zhao
H
.
Differential expressions of microRNAs and transfer RNA-derived small RNAs: potential targets of choroidal neovascularization
.
Curr Eye Res
.
2019
;
44
(
11
):
1226
35
.
19.
Liu
X
,
Zhang
L
,
Wang
JH
,
Zeng
H
,
Zou
J
,
Tan
W
.
Investigation of circRNA expression profiles and analysis of circRNA-miRNA-mRNA networks in an animal (mouse) model of age-related macular degeneration
.
Curr Eye Res
.
2020
;
45
(
9
):
1173
80
.
20.
Age-Related Eye Disease Study Research Group
.
A randomized, placebo-controlled, clinical trial of high-dose supplementation with vitamins C and E, beta carotene, and zinc for age-related macular degeneration and vision loss: AREDS report no. 8
.
Arch Ophthalmol
.
2001
;
119
(
10
):
1417
36
.
21.
Tisdale
AK
,
Agrón
E
,
Sunshine
SB
,
Clemons
TE
,
Ferris
FL
3rd
,
Chew
EY
.
Association of dietary and supplementary calcium intake with age-related macular degeneration: age-related eye disease study report 39
.
JAMA Ophthalmol
.
2019
;
137
(
5
):
543
50
.
22.
Klevay
LM
.
Calcium can delay age-related macular degeneration via enhanced copper metabolism
.
Med Hypotheses
.
2020
;
135
:
109467
.
23.
Zhao
T
,
Guo
X
,
Sun
Y
.
Iron accumulation and lipid peroxidation in the aging retina: implication of ferroptosis in age-related macular degeneration
.
Aging Dis
.
2021
;
12
(
2
):
529
51
.
24.
Ugarte
M
,
Osborne
NN
,
Brown
LA
,
Bishop
PN
.
Iron, zinc, and copper in retinal physiology and disease
.
Surv Ophthalmol
.
2013 Nov-Dec
58
6
585
609
.
25.
Harris
ED
.
A requirement for copper in angiogenesis
.
Nutr Rev
.
2004
;
62
(
2
):
60
4
.
26.
Tellios
V
,
Liu
H
,
Tellios
N
,
Li
X
,
Hutnik
CML
.
Administration of nitric oxide through a novel copper-chitosan delivery system in human corneal and limbal epithelial cell injury
.
Invest Ophthalmol Vis Sci
.
2018
;
59
(
2
):
967
77
.
27.
Zhao
G
,
Sun
H
,
Zhang
T
,
Liu
JX
.
Copper induce zebrafish retinal developmental defects via triggering stresses and apoptosis
.
Cell Commun Signal
.
2020
;
18
(
1
):
45
.
28.
Aloysius Dhivya
M
,
Sulochana
KN
,
Bharathi Devi
SR
.
High glucose induced inflammation is inhibited by copper chelation via rescuing mitochondrial fusion protein 2 in retinal pigment epithelial cells
.
Cell Signal
.
2022
;
92
:
110244
.
29.
Liggett
WH
Jr
,
Sidransky
D
.
Role of the p16 tumor suppressor gene in cancer
.
J Clin Oncol
.
1998
;
16
(
3
):
1197
206
.
30.
Faust
HJ
,
Zhang
H
,
Han
J
,
Wolf
MT
,
Jeon
OH
,
Sadtler
K
.
IL-17 and immunologically induced senescence regulate response to injury in osteoarthritis
.
J Clin Invest
.
2020
;
130
(
10
):
5493
507
.
31.
Baker
DJ
,
Wijshake
T
,
Tchkonia
T
,
LeBrasseur
NK
,
Childs
BG
,
van de Sluis
B
.
Clearance of p16Ink4a-positive senescent cells delays ageing-associated disorders
.
Nature
.
2011
;
479
(
7372
):
232
6
.
32.
Li
Y
,
Liu
X
,
Zhou
T
,
Kelley
MR
,
Edwards
P
,
Gao
H
.
Inhibition of APE1/Ref-1 redox activity rescues human retinal pigment epithelial cells from oxidative stress and reduces choroidal neovascularization
.
Redox Biol
.
2014
;
2
:
485
94
.
33.
Vallée
A
,
Lecarpentier
Y
,
Guillevin
R
,
Vallée
JN
.
Aerobic glycolysis hypothesis through WNT/Beta-Catenin pathway in exudative age-related macular degeneration
.
J Mol Neurosci
.
2017
62
3–4
368
79
.
34.
Zeng
A
,
Yin
J
,
Li
Y
,
Li
R
,
Wang
Z
,
Zhou
X
.
miR-129-5p targets Wnt5a to block PKC/ERK/NF-κB and JNK pathways in glioblastoma
.
Cell Death Dis
.
2018
;
9
(
3
):
394
.
35.
Zhao
Y
,
Ren
F
,
Wang
X
,
Liu
Y
.
Circ-TFF1 promotes breast cancer progression through the miR-129-2-3p/IRAK1 Axis
.
Biochem Genet
.
2022
;
60
(
6
):
2155
70
.
36.
Yan
S
,
Wang
H
,
Chen
X
,
Liang
C
,
Shang
W
,
Wang
L
.
MiR-182-5p inhibits colon cancer tumorigenesis, angiogenesis, and lymphangiogenesis by directly downregulating VEGF-C
.
Cancer Lett
.
2020
;
488
:
18
26
.
37.
Liu
X
,
Shen
S
,
Zhu
L
,
Su
R
,
Zheng
J
,
Ruan
X
.
SRSF10 inhibits biogenesis of circ-ATXN1 to regulate glioma angiogenesis via miR-526b-3p/MMP2 pathway
.
J Exp Clin Cancer Res
.
2020
;
39
(
1
):
121
.
38.
Zhu
Y
,
Lu
B
,
Wu
J
,
Li
S
,
Jia
K
.
CircRNA expression profiles in canine mammary tumours
.
Vet Sci
.
2022
;
9
(
5
):
205
.
39.
Han
D
,
Liu
W
,
Li
G
,
Liu
L
.
Circ_PRKDC knockdown promotes skin wound healing by enhancing keratinocyte migration via miR-31/FBN1 axis
.
J Mol Histol
.
2021
;
52
(
4
):
681
91
.
40.
Yang
Y
,
Gong
B
,
Wu
ZZ
,
Shuai
P
,
Li
DF
,
Liu
LL
.
Inhibition of microRNA-129-5p expression ameliorates ultraviolet ray-induced corneal epithelial cell injury via upregulation of EGFR
.
J Cell Physiol
.
2019
;
234
(
7
):
11692
707
.
41.
Yang
M
,
So
KF
,
Lam
WC
,
Lo
ACY
.
Novel programmed cell death as therapeutic targets in age-related macular degeneration
.
Int J Mol Sci
.
2020
;
21
(
19
):
7279
.
42.
Chang
CC
,
Huang
TY
,
Chen
HY
,
Huang
TC
,
Lin
LC
,
Chang
YJ
.
Protective effect of melatonin against oxidative stress-induced apoptosis and enhanced autophagy in human retinal pigment epithelium cells
.
Oxid Med Cell Longev
.
2018
;
2018
:
9015765
.
43.
Hanus
J
,
Anderson
C
,
Wang
S
.
RPE necroptosis in response to oxidative stress and in AMD
.
Ageing Res Rev
.
2015
24
Pt B
286
98
.
44.
Totsuka
K
,
Ueta
T
,
Uchida
T
,
Roggia
MF
,
Nakagawa
S
,
Vavvas
DG
.
Oxidative stress induces ferroptotic cell death in retinal pigment epithelial cells
.
Exp Eye Res
.
2019
;
181
:
316
24
.
45.
Kahlson
MA
,
Dixon
SJ
.
Copper-induced cell death
.
Science
.
2022
;
375
(
6586
):
1231
2
.
46.
Ahmad
I
,
Balasubramanian
S
,
Del Debbio
CB
,
Parameswaran
S
,
Katz
AR
,
Toris
C
.
Regulation of ocular angiogenesis by Notch signaling: implications in neovascular age-related macular degeneration
.
Invest Ophthalmol Vis Sci
.
2011
;
52
(
6
):
2868
78
.
47.
Lai
K
,
Gong
Y
,
Zhao
W
,
Li
L
,
Huang
C
,
Xu
F
.
Triptolide attenuates laser-induced choroidal neovascularization via M2 macrophage in a mouse model
.
Biomed Pharmacother
.
2020
;
129
:
110312
.
48.
Wu
J
,
Chen
J
,
Hu
J
,
Yao
M
,
Zhang
M
,
Wan
X
.
CircRNA Uxs1/miR-335-5p/PGF axis regulates choroidal neovascularization via the mTOR/p70 S6k pathway
.
Transl Res
.
2023
;
256
:
41
55
.