Secondary Logo

Journal Logo

Original Articles

Hsp90-associated DNA replication checkpoint protein and proteasome-subunit components are involved in the age-related macular degeneration

Xing, Chen1; Liu, Xiao-Feng2; Zhang, Chun-Feng3; Yang, Liu1

Editor(s): Yin, Yan-Jie; Hao, Xiu-Yuan

Author Information
doi: 10.1097/CM9.0000000000001773



Age-related macular degeneration (AMD) is a disease that affects the macular region of the retina and causes progressive loss of central vision. It is the third leading cause of severe irreversible vision loss worldwide.[1] AMD is clinically classified as early-stage and late-stage. The early-stage AMD is characterized by the accumulation of medium-sized drusen, yellow deposits which form under the retina, and abnormalities of the retinal pigment epithelium (RPE). The late-stage AMD is defined by the presence of choroidal neovascularization (CNV, neovascular or wet AMD) or geographic atrophy (GA, atrophic or dry AMD) in the Beckman classification.[2] Late AMD results in severe and permanent vision impairment and blindness. The global prevalence of AMD has been steadily increasing over the past years and is expected to reach 288 million by 2040.[3] Although the incidence of vision impairment and blindness caused by AMD have decreased since the introduction of vascular endothelial growth factor (VEGF)-targeted therapies,[4] no proven therapies for atrophic disease are currently available. Thus, novel regenerative therapies are required to attenuate the progression of neovascular AMD and GA.

There are several risk factors and biomarkers associated with AMD. Age is an important risk factor for AMD, since nearly all the late AMD cases have been reported in people aged >60 years.[1] Additionally, smoking and diet are known risk factors for AMD.[5] Stress or tissue damage, inflammation, immune response, and pathological angiogenesis have been found to cause AMD through driving CNV.[6,7] However, the exact etiology of AMD remains unclear. Elucidation of the molecular mechanisms involved in the development and progression of AMD will provide new strategies for AMD treatment.

The RPE plays an essential role in maintaining normal retinal health through protecting the retina from systemic insults.[8] RPE acts as a physiological barrier between the photoreceptor cells and the choroidal blood supply, facilitating the delivery of nutrients and ions by the choroidal blood and the transport of photoreceptor-derived waste products to the choroid.[9] The onset of AMD is characterized by initial damage to RPE, resulting in subsequent loss of photoreceptor cells over time.[10] The senescence of RPE induced by the depletion of nicotinamide adenine dinucleotide (NAD+) has been found to play a critical role in the progression of AMD.[8] Phenotypic and functional studies on cultured human RPE cells have found that impaired autophagy contributes to the development of AMD.[11] Consistent with the important role of the RPE in AMD, gene mutations or gene expression alterations in RPE have been shown to contribute to the pathogenesis of AMD. Although the expression profiles of RPE signature genes in mouse and human tissues have been characterized,[12] large-scale functional genomic analysis of the RPE from AMD patients and healthy subjects is still limited.

Given that it is rather difficult to get the RPE biopsies from either AMD patients or healthy people, pre-existing high-throughput gene expression microarray databases, and RNA sequencing from clinical biopsies provide us precious resources to explore the molecular changes in AMD.[13,14] Therefore, we pooled the existing high-throughput gene expression data from RPE/choroid of human AMD and healthy controls to evaluate the differential gene expression and the potential biological pathways involved in the development of AMD.

In the present study, we have downloaded two datasets, GSE99248 and GSE125564, from gene expression omnibus (GEO) database and identified the differentially expressed genes (DEGs) in the RPE from AMD samples and healthy controls. Gene Ontology (GO) enrichment analysis, protein-protein interaction (PPI) network construction, and hub gene analysis were conducted to uncover the key genes involved in the development of AMD. We further examined the expression levels of the top ten hub genes in human RPE cell line, ARPE-19. The key genes altered in the AMD were determined following the induction of senescence in ARPE-19 with FK866, a highly specific inhibitor of nicotinamide phosphoribosyl transferase (NAMPT). Thereafter, we constructed the key gene-miRNA interaction and the key gene-drug interaction networks to explore potential strategies for attenuating the progression of AMD.


Cell culture and FK866 treatment

ARPE-19 cells were obtained from the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences. ARPE-19 cells were grown in DMEM/F12 medium supplemented with 10% fetal bovine serum. For FK866 (Sigma-Aldrich; St Louis, MO, USA) treatment, ARPE-19 cells were serum starved overnight and treated with different doses (0.01–10 μmol/L) of FK866 as described previously.[15] Cells were collected at different time intervals (24, 48, and 72 h) after FK866 treatment and total RNA was extracted for the evaluation of hub gene expression.

Quantitative real-time polymerase chain reaction (RT-qPCR)

Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) from FK866-treated cells. Reverse transcription was performed to synthesize cDNA with 2 μg of total RNA using the Superscript First-Strand Synthesis System (Invitrogen). RT-qPCR was performed to evaluate the expression of the hub genes. The primer sequences for PCR are shown in Supplementary Table S1,

Data collection

The gene expression profiles (GSE99248)[16] of 16 eye samples obtained from the AMD donors and 15 healthy control samples were downloaded from the GEO ( database. The sequencing was previously performed on the GPL11154 Illumina HiSeq 2000 platform (Illumina, San Diego, USA). The mRNA profile for human AMD samples (GSE125564)[17] is available on the GEO database derived from platform GPL23159 (The Human Affymetrix Clariom S Assay).

Identification of DEGs

The Limma package, a prominent tool for analyzing microarray and RNA-Seq data in R software,[18] was used to detect differentially expressed mRNAs by calculating the adjusted P value (adj. P) and the absolute log value of fold change (log|FC|) after extracting the gene expression data from the AMD and control groups. Cut-off values of adj. P < 0.05 and log|FC| > 1 were set to screen the DEGs as described previously.[19] The DEGs from GSE99248 and GSE125564 were pooled to select the common DEGs.

GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis

GO and KEGG pathway enrichment analyses of the common DEGs between GSE99248 and GSE125564 were conducted using the Database for Annotation, Visualization, and Integrated Discovery (DAVID,[20] The biological entities and pathways exhibiting P < 0.05 were regarded as statistically significant in GO analysis, including biological processes (BP), cellular components (CC), and molecular functions (MF) were used for pathway analysis. The GO enrichment analysis and KEGG pathway analysis results were visualized using the GO plot package.[21]

PPI network and hub gene analyses

To analyze PPI network, we first submitted the DEGs to Search Tool for the Retrieval of Interacting Genes (STRING) v11 (,[22] an online database which contains comprehensive information on multiple proteins. The confidence score ≥0.4 was used as the cut-off criteria for selecting significant PPI, yielding PPIs of DEGs as shown in Supplementary Table S2, The PPI networks of DEGs were visualized using Cytoscape software ( (version 3.7.1).[23] To identify the critical DEGs among PPIs, the cytoHubba plug-in was used to calculate the topological properties, including node degree, betweenness, and stress. Finally, the hub genes were selected according to the node degree and the interaction network of DEGs and visualized using Cytoscape software (version 3.7.1).

Construction of target gene-miRNA regulatory network

The regulatory relationships between target genes and miRNAs were identified using online tool miRNet ([24] The target gene-miRNA regulatory network was visualized using Cytoscape (

Drug prediction

The data for netrin family-related drugs were downloaded from PharmacoDB ([25] The target gene-drug regulatory network was visualized using Cytoscape (


Identification of DEGs

To uncover novel molecules involved in AMD, we first analyzed the DEGs between the AMD RPE and healthy samples from datasets GSE99248 and GSE125524 using the cut-off values of adj. P < 0.05 and log|FC| > 1 as previously described.[19] As shown in the volcano plot of GSE99248, 3671 DEGs, including 2873 upregulated and 798 downregulated genes, were identified in the RPE samples from AMD patients compared with those from the healthy controls [Figure 1A]. From GSE125524, 802 DEGs, including 313 upregulated and 489 downregulated genes, were identified in the AMD RPE samples [Figure 1B]. By merging these two DEG datasets, 174 common genes were obtained [Figure 1C].

Figure 1
Figure 1:
Volcano plots of DEGs identified from datasets GSE99248 and GSE125564. (A) DEGs from GSE99248; (B) DE-mRNAs from GSE125564; (C) The Venn diagrams of overlapping DEGs from GSE99248 and GSE125564. A total of 174 overlapping genes were identified. DEGs: Differentially expressed genes; DE-mRNAs: Differentially expressed mRNAs.

GO and KEGG pathway enrichment analyses

Thereafter, the GO and KEGG pathway enrichment analyses were conducted on the selected 174 overlapping genes using the DAVID database. The BP extracted from the GO analysis showed that the DEGs were mainly enriched in the regulation of protein polyubiquitination, mitotic cell cycle regulation, and cellular amino acid metabolism [Figure 2A]. The top three enriched CC were nucleoplasm, mitochondrial membrane, and cytosol [Figure 2B]. The MF analysis showed that the top three enriched functions were protein binding, poly-A RNA binding, and DNA helicase activity [Figure 2C]. We further conducted the KEGG pathway enrichment analysis with the 174 DEGs. It showed that the critical pathways were enriched in the DNA replication, cell cycle regulation, and proteasomal degradation [Figure 2D]. Integration of both GO and KEGG analyses indicated that the AMD-related genes might primarily function in the DNA replication-associated cell cycle regulation and proteasome-mediated protein polyubiquitination.

Figure 2
Figure 2:
Molecular function and pathway enrichment analyses for the 174 overlapping genes. (A) Biological process (BP); (B) cellular component (CC); (C) molecular function (MF); (D) Kyoto Encyclopedia of Genes and Genomes (KEGG).

PPI network and hub gene analysis

To further confirm the key biological pathways involved in the AMD, PPI network was constructed to explore the potential interactions between the 174 DEGs using STRING database and Cytoscape. The PPI networks are shown in the Supplementary Table S2, The hub genes with top ten node degrees in the PPI network were HSP90AA1 (degree = 24), pre-mRNA processing factor 19 (PRPF19) (degree = 18), PSMD4 (degree = 17), PSMD8 (degree = 17), PSMA4 (degree = 17), CHEK1 (degree = 16), SF3B5 (degree = 16), CCT7 (degree = 17), PSMB5 (degree = 15), and PSMB6 (degree = 15) [Figure 3A and Table 1]. The expression levels of these top ten hub genes in the AMD and healthy RPE samples were plotted from GSE99248 and GSE125524. As shown Figure 3B and 3C, most of these genes were upregulated to different extents in the AMD RPEs compared with healthy samples. With the exception that HSP90AA1 has been found to be involved in AMD,[26] the remaining nine hub genes were not directly related to AMD. It is of interest that five out of these nine hub genes, including PSMD4, PSMD8, PSMA4, PSMB5, and PSMB6, are functionally related with proteasomal function [Figure 2D], suggesting that proteasome-mediated protein degradation might play an essential role in the development of AMD. Additionally, the hub gene CHEK1 plays a key role in the DNA replication checkpoint of cell cycle,[27] suggesting that DNA replication checkpoint might be super-activated in the AMD development.

Figure 3
Figure 3:
Identification of hub genes in the overlapping DEGs. (A) Visualization of PPI networks for the top 20 hub DEGs. (B) The expression levels of the top ten hub genes in GSE99248 dataset. (C) The expression levels of the top ten hub genes in GSE125564 dataset. DEGs: Differentially expressed genes; PPI: Protein-protein interaction.
Table 1 - AMD-related top ten hub genes identified from cytoHubba analysis.
Node name MCC DMNC MNC Degree
HSP90AA1 2285 0.29055 23 24
PRPF19 795 0.29953 17 18
PSMD4 3612 0.43715 17 17
PSMD8 3612 0.43715 17 17
PSMA4 3629 0.50256 16 17
CHEK1 344 0.34999 16 16
SF3B5 740 0.41901 10 16
CCT7 1642 0.43918 14 16
PSMB5 3602 0.62588 13 15
PSMB6 1327 0.42792 14 15
DMNC: Density of Maximum Neighborhood Component; MCC: Maximal Clique Centrality; MNC: Maximum Neighborhood Component.

Expression of five hub genes is upregulated in RPE cellular senescence

Since the senescence of RPE cells is critical to the initiation and progression of AMD and ARPE-19 cell line structurally and functionally possesses the properties of RPE cells in vivo,[28] we induced senescence in ARPE-19 cell line and evaluated the expression of the top ten hub genes. Senescence was induced with FK866, a selective NAMPT inhibitor, thereby reducing the NAD+ levels in ARPE-19 cells, as described previously.[8] The mRNA levels of the AMD-related top ten hub genes were evaluated in the FK866-treated ARPE-19 cells. We showed that the expression levels of HSP90AA1, CHEK1, PSMA4, PSMD4, and PSMD8 were upregulated when ARPE-19 cells were treated with 10 μmol/L FK866 [Figure 4A], indicating that these genes may play key roles in the development of AMD. To further confirm the altered expression of these AMD-related key genes, we evaluated the mRNA levels of these genes at different time points upon FK866 treatment. As shown in Figure 4B–F, HSP90AA1, CHEK1, PSMD4, PSMD8, and PSMA4 were upregulated at different time points upon FK866 treatment, confirming that these genes were indeed AMD-related key genes functioning in the RPE senescence.

Figure 4
Figure 4:
Five out of the top ten hub genes were upregulated in the FK866-treated ARPE-19 cells. (A) ARPE-19 cells were treated with 10 μmol/L FK866 for 72 h. Total RNA was extracted and RT-qPCR was performed to evaluate the expression levels of the top ten hub genes. ∗P < 0.05, †P < 0.001. (B–F) ARPE-19 cells were treated with 10 μmol/L FK866 for 0, 24, 48, and 72 h. Total RNA was extracted and the expression levels of HSP90AA1 (B), CHEK1 (C), PSMD4 (D), PSMD8 (E), and PSMA4 (F) were evaluated by RT-qPCR. RT-qPCR: Quantitative real-time polymerase chain reaction.

HSP90AA1 might play a central role in RPE cellular senescence

It has been found that FK866 induces senescence in ARPE-19 cells with a broad range of dose from 0.001 to 10 μmol/L. We wanted to explore whether the AMD-related key genes respond to the low dose of FK866. We showed that only HSP90AA1 was upregulated at low dose (0.1 μmol/L) of FK866, while CHEK1, PSMD4, PSMD8, and PSMA4 were upregulated at the dose of 1.0 μmol/L FK866 or above [Figure 5A–E], indicating that HSP90AA1 might be the most sensitive gene to senescence induction and potentially a central regulator of the RPE senescence. We thereafter analyzed the interaction among the AMD-related key genes. As shown in Figure 5F, HSP90AA1 was indeed at the center of the interaction network.

Figure 5
Figure 5:
Interaction among the key DEGs and prediction of potential drugs targeting the critical DEGs. (A–E) ARPE-19 cells were treated with different doses of FK866 as indicated. Total RNA was extracted and the expression levels of the key genes were evaluated by RT-qPCR. F. The interactions between the proteins encoded by the indicated genes were analyzed by constructing a PPI network. (G) Potential drugs targeting the AMD-related key genes were predicted using PharmacoDB. HSP90AA1, CHEK1, and PSMA4 were found to be directly targeted by small molecules as indicated. AMD: Age-related macular degeneration; DEGs: Differentially expressed genes; PPI: Protein-protein interaction; RT-qPCR: Quantitative real-time polymerase chain reaction.

Prediction of drugs targeting AMD-related key genes

To predict small molecule drugs targeting the AMD-related key genes, we downloaded the list of cancer drugs related to these genes from the cancer pharmacogenomics research database PharmacoDB49 [Figure 5G]. Certain compounds were found to target HSP90AA1 (Geldanamycin, CCT018159, AT13387, and SNX-2112) and CHEK1 (BX795, AZD7762, BX-912, and PD98059). Bortezomib was predicted to target PSMA4. No drugs were found to directly target PSMD4 and PSMD8. Among these drugs, Hsp90 inhibitors have been used in the clinical trials for AMD treatment.[26] BX795 is a multiple kinase inhibitor which suppresses the inflammatory response and is safe for eye treatment.[29,30] Bortezomib and a few other drugs have shown potential anti-tumor activities.[31]

Construction of target gene-miRNA regulatory network

Since gene expression is regulated by miRNAs, we next wanted to determine the specific miRNAs that could target all of the AMD-related key genes. We first analyzed the gene-miRNA network with single genes. HSP90AA1, CHEK1, PSMD4, PSMD8, and PSMA4 were predicted to interact with several miRNAs [Supplementary Table S3,; Figure 6A]. We further constructed the miRNA-gene network to screen the miRNAs simultaneously targeting two genes. As shown in Figure 6B, some miRNAs were found to target two of the five key genes, but the network still included abundant miRNAs. We thus constructed the miRNA-gene network with the miRNAs targeting more than three genes [Figure 6C]. Specifically, hsa-miR-16-5p was identified to interact with four out of the five key genes, including HSP90AA1, CHEK1, PSMD4, and PSMD8, predicting that hsa-miR-16-5p might potentially interfere with the progression of AMD through targeting the RPE senescence-related genes.

Figure 6
Figure 6:
Constructions of the AMD-related gene-miRNA network. (A) Visualization of the indicated gene-miRNA networks. (B) MicroRNAs in (A) targeting two or more genes were selected for network analysis. (C) MicroRNAs in (A) targeting three or more genes were used for network analysis. AMD: Age-related macular degeneration.


AMD is a common retinal degenerative disease in older adults. It is reported that the occurrence of AMD is influenced by individual differences, environmental factors, and genetic factors. However, the risk for AMD is significantly elevated with increasing age. Currently, wet AMD is primarily treated with anti-VEGF drugs, whereas antioxidants are utilized for the management of dry AMD to attenuate the progression of symptoms.[32,33] Therefore, it is important to identify novel genes involved in the development of AMD, which can be targeted to provide more strategies for the treatment of AMD.

In the present study, we set out from database analysis to determine the AMD-related critical genes. Recently, RHO, PDE6A, 3′,5′-cyclic-GMP phosphodiesterase, and G protein alpha have been identified as AMD-related genes by canonical pathway analysis using the Ingenuity Pathway Analysis database.[34] We additionally identified HSP90AA1, PRPF19, PSMD4, PSMD8, PSMA4, CHEK1, SF3B5, CCT7, PSMB5, and PSMB6 as the top ten hub genes associated with AMD. Importantly, in both of the analyzed cohorts, these genes were mostly upregulated in AMD RPE samples compared to healthy control samples. Thus, these genes are likely to promote the progression of AMD. HSP90AA1-encoded protein (Hsp90) has been found to be associated with AMD development.[26,35] However, it remains unknown whether the remaining nine hub genes are related to AMD. As cells age, the synthesis of NAD+ gradually decreases and the resulting depletion of NAD+ may cause certain ageing-related diseases, with some of them resulting in vision impairment.[8] FK866 is an inhibitor of NAMPT, a crucial enzyme catalyzing the synthesis of NAD+.[8,15] Therefore, it is widely used to induce senescence in the RPE cells. We, therefore, evaluated the expression of the top ten hub genes in the FK866-treated ARPE-19 cells. We showed that HSP90AA1, CHEK1, PSMD4, PSMD8, and PSMA4 were upregulated in the senescent ARPE-19 cells, indicating that these genes are AMD-related key genes. However, we cannot rule out the functions of PRPF19, SF3B5, CCT7, PSMB5, and PSMB6 genes in the progression of AMD. PRPF19 is a key splicing factor involved in DNA damage response, ubiquitin-proteasome system, cell proliferation, and apoptosis.[36] Additionally, PRPF19 regulates p53-dependent cellular senescence through modulating MDM4 mRNA splicing.[37] As a spliceosomal protein, SF3B5 might participate in the tumorigenesis and aging.[38,39] Therefore, PRPF19 and SF3B5 might contribute to the pathogenesis of AMD through regulating RPE aging. CCT7/TCP-1 ring complex is involved in the maturation of G protein-coupled receptors (GPCRs).[40] CCT7 might act in the AMD by participating in the GPCR-mediated signaling pathway. However, the precise functional role of PRPF19, SF3B5, and CCT7 in AMD needs to be investigated further. Additionally, the evaluation of hub gene expression in the primary RPE cells obtained from human donor eyes with different types of AMD will provide important functional insights.

Among the proteins encoded by these AMD-related key genes, Hsp90 is a chaperone which is essential for the correct folding and stabilization of various cellular proteins. Importantly, Hsp90 is a potential modulator of cell senescence.[41] Hsp90 plays an essential roles in the retina and Hsp90 inhibitors such as geldanamycin, and its derivatives, such as 17-allylamino-17-demethoxy-geldanamycin and 17-dimethylaminoethylamino-17-demethoxygeldanamycin, can prevent retinal degeneration in the AMD models and have been used in AMD treatment clinical trials.[26] However, prolonged Hsp90 inhibition can also induce the degradation of Hsp90 client proteins, leading to photoreceptor cell death.[42] Thus, combinational therapy might alleviate the side effect by decreasing the dosage or the duration of Hsp90 inhibition. We found that CHEK1 interacts with HSP90AA1 and is upregulated in the senescent RPE cells. CHEK1 encodes the cell cycle checkpoint kinase Chk1, which modulates aging through regulating the DNA replication.[43] Chk1 is an Hsp90 client and requires Hsp90 to acquire its kinase activity.[44] Thus, targeting Chk1 alone or in combination with Hsp90 inhibitors might attenuate the development of AMD through regulating cell senescence and reducing the dose or duration of Hsp90 inhibition to lower the toxicity on the photoreceptor cells. This hypothesis needs to be validated further.

We further analyzed the potential drugs targeting Hsp90 and Chk1. Among the potential Chk1 inhibitors, BX795 is known to suppress the inflammatory response and is safe for eye treatment.[29,30] Thus, BX795 possesses therapeutic potential in AMD treatment. The combination of Hsp90 inhibitors and potential Chk1 inhibitors could improve the AMD treatment landscape.

In addition to HSP90AA1 and CHEK1, we identified proteasomal subunit components PSMD4, PSMD8, and PSMA4 as AMD-related key genes. Since Hsp90 controls the spatial and temporal order of protein interactions and regulates the ubiquitin-mediated proteasomal degradation,[45] Hsp90 might interact with these proteasomal subunit components to regulate the RPE senescence. The PSMD4-encoded protein, proteasome 26S subunit non-ATPase 4, recruits phosphorylated and ubiquitinated nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha (IκBα) to the proteasome, resulting in nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) activation and induction of cellular senescence.[46] PSMD8 is a 19S proteasomal regulatory complex subunit. We propose that Hsp90 might interact with PSMD4 and PSMD8 to regulate the degradation of misfolded proteins during AMD progression.

MicroRNAs play an important role in regulating the development of AMD.[47]In vivo studies found that dysregulated miRNAs might affect AMD progression by targeting genes involved in neurodegeneration and inflammation.[48] As senescent cells cause inflammation by senescence-associated secretory phenotype, miRNAs that inhibit inflammation could be effective in the treatment of AMD. For instance, miRNA-191-5p ameliorates amyloid-β1-40-mediated RPE cell injury by suppressing the NLRP3 inflammation pathway.[49] In this study, we found that miR-16-5p targeted four of the AMD-associated key genes, including HSP90AA1, CHEK1, PSMD4, and PSMD8. Previous studies revealed that miR-16-5p is associated with the regulation of inflammation.[50] Therefore, miR-16-5p might regulate the senescent RPE-induced inflammation and retard the progression of AMD. However, the specific mechanism underlying the role of miR-16-5p in AMD progression requires further investigation.


This study was supported by grant from the National Natural Science Foundation of China (No. 81670841).

Conflicts of interest



1. Mitchell P, Liew G, Gopinath B, Wong TY. Age-related macular degeneration. Lancet 2018; 392:1147–1159. doi: 10.1016/S0140-6736(18)31550-2.
2. Ferris FL, Wilkinson CP, Bird A, Chakravarthy U, Chew E, Csaky K, et al. Clinical classification of age-related macular degeneration. Ophthalmology 2013; 120:844–851. doi: 10.1016/j.ophtha.2012.10.036.
3. Wong WL, Su X, Li X, Cheung CM, Klein R, Cheng CY, et al. Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis. Lancet Glob Health 2014; 2:e106–e116. doi: 10.1016/S2214-109X(13)70145-1.
4. Lim LS, Mitchell P, Seddon JM, Holz FG, Wong TY. Age-related macular degeneration. Lancet 2012; 379:1728–1738. doi: 10.1016/S0140-6736(12)60282-7.
5. Lambert NG, ElShelmani H, Singh MK, Mansergh FC, Wride MA, Padilla M, et al. Risk factors and biomarkers of age-related macular degeneration. Prog Retin Eye Res 2016; 54:64–102. doi: 10.1016/j.preteyeres.2016.04.003.
6. Ambati J, Fowler BJ. Mechanisms of age-related macular degeneration. Neuron 2012; 75:26–39. doi: 10.1016/j.neuron.2012.06.018.
7. Sarks JP, Sarks SH, Killingsworth MC. Morphology of early choroidal neovascularisation in age-related macular degeneration: correlation with activity. Eye (Lond) 1997; 11:515–522. doi: 10.1038/eye.1997.137.
8. Lin JB, Kubota S, Ban N, Yoshida M, Santeford A, Sene A, et al. NAMPT-mediated NAD(+) biosynthesis is essential for vision in mice. Cell Rep 2016; 17:69–85. doi: 10.1016/j.celrep.2016.08.073.
9. Bok D. The retinal pigment epithelium: a versatile partner in vision. J Cell Sci Suppl 1993; 17:189–195. doi: 10.1242/jcs.1993.supplement_17.27.
10. Strauss O. The retinal pigment epithelium in visual function. Physiol Rev 2005; 85:845–881. doi: 10.1152/physrev.00021.2004.
11. Golestaneh N, Chu Y, Xiao YY, Stoleru GL, Theos AC. Dysfunctional autophagy in RPE, a contributing factor in age-related macular degeneration. Cell Death Dis 2017; 8:e2537doi: 10.1038/cddis.2016.453.
12. Bennis A, Gorgels TGMF, Brink JBT, van der Spek PJ, Bossers K, Heine VM, et al. Comparison of mouse and human retinal pigment epithelium gene expression profiles: potential implications for age-related macular degeneration. PLoS One 2015; 10:e0141597doi: 10.1371/journal.pone.0141597.
13. Peck D, Crawford ED, Ross KN, Stegmaier K, Golub TR, Lamb J. A method for high-throughput gene expression signature analysis. Genome Biol 2006; 7:R61doi: 10.1186/gb-2006-7-7-r61.
14. Zhao B, Wang M, Xu J, Li M, Yu Y. Identification of pathogenic genes and upstream regulators in age-related macular degeneration. BMC Ophthalmol 2017; 17:102doi: 10.1186/s12886-017-0498-z.
15. Jadeja RN, Powell FL, Jones MA, Fuller J, Joseph E, Thounaojam MC, et al. Loss of NAMPT in aging retinal pigment epithelium reduces NAD(+) availability and promotes cellular senescence. Aging (Albany NY) 2018; 10:1306–1323. doi: 10.18632/aging.101469.
16. Kim EJ, Grant GR, Bowman AS, Haider N, Gudiseva HV, Chavali VRM. Complete transcriptome profiling of normal and age-related macular degeneration eye tissues reveals dysregulation of anti-sense transcription. Sci Rep 2018; 8:3040doi: 10.1038/s41598-018-21104-7.
17. Gong J, Cai H, Global Stem Cell Array Team NYSCF, Noggle S, Paull D, Rizzolo LJ, et al. Stem cell-derived retinal pigment epithelium from patients with age-related macular degeneration exhibit reduced metabolism and matrix interactions. Stem Cells Transl Med 2020; 9:364–376. doi: 10.1002/sctm.19-0321.
18. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015; 43:e47doi: 10.1093/nar/gkv007.
19. Li W. Volcano plots in analyzing differential expressions with mRNA microarrays. J Bioinform Comput Biol 2012; 10:1231003doi: 10.1142/S0219720012310038.
20. Huang DW, Sherman BT, Lempicki RA. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 2009; 37:1–13. doi: 10.1093/nar/gkn923.
21. Walter W, Sanchez-Cabo F, Ricote M. GOplot: an R package for visually combining expression data with functional analysis. Bioinformatics 2015; 31:2912–2914. doi: 10.1093/bioinformatics/btv300.
22. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 2019; 47:D607–D613. doi: 10.1093/nar/gky1131.
23. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003; 13:2498–2504. doi: 10.1101/gr.1239303.
24. Fan Y, Xia J. miRNet-functional analysis and visual exploration of miRNA-target interactions in a network context. Methods Mol Biol 2018; 1819:215–233. doi: 10.1007/978-1-4939-8618-7_10.
25. Smirnov P, Kofia V, Maru A, Freeman M, Ho C, El-Hachem N, et al. PharmacoDB: an integrative database for mining in vitro anticancer drug screening studies. Nucleic Acids Res 2018; 46:D994–D1002. doi: 10.1093/nar/gkx911.
26. Aguila M, Cheetham ME. Hsp90 as a potential therapeutic target in retinal disease. Adv Exp Med Biol 2016; 854:161–167. doi: 10.1007/978-3-319-17121-0_22.
27. Carr AM, Moudjou M, Bentley NJ, Hagan IM. The Chk1 pathway is required to prevent mitosis following cell-cycle arrest at start. Curr Biol 1995; 5:1179–1190. doi: 10.1016/S0960-9822(95)00234-X.
28. Dunn KC, Aotaki-Keen AE, Putkey FR, Hjelmeland LM. ARPE-19, a human retinal pigment epithelial cell line with differentiated properties. Exp Eye Res 1996; 62:155–169. doi: 10.1006/exer.1996.0020.
29. Yu T, Wang Z, Jie W, Fu X, Li B, Xu H, et al. The kinase inhibitor BX795 suppresses the inflammatory response via multiple kinases. Biochem Pharmacol 2020; 174:113797doi: 10.1016/j.bcp.2020.113797.
30. Yadavalli T, Suryawanshi R, Ali M, Iqbal A, Koganti R, Ames J, et al. Prior inhibition of AKT phosphorylation by BX795 can define a safer strategy to prevent herpes simplex virus-1 infection of the eye. Ocul Surf 2020; 18:221–230. doi: 10.1016/j.jtos.2019.11.011.
31. Mateos MV, Gavriatopoulou M, Facon T, Auner HW, Leleu X, Hajek R, et al. Effect of prior treatments on selinexor, bortezomib, and dexamethasone in previously treated multiple myeloma. J Hematol Oncol 2021; 14:59doi: 10.1186/s13045-021-01071-9.
32. de Guimaraes TAC, Varela MD, Georgiou M, Michaelides M. Treatments for dry age-related macular degeneration: therapeutic avenues, clinical trials and future directions. Br J Ophthalmol 2021; 1–8. doi: 10.1136/bjophthalmol-2020-318452.
33. Gelfman CM, Grishanin R, Bender KO, Nguyen A, Greengard J, Sharma P, et al. Comprehensive preclinical assessment of ADVM-022, an intravitreal anti-VEGF gene therapy for the treatment of neovascular AMD and diabetic macular edema. J Ocul Pharmacol Ther 2021; 37:181–190. doi: 10.1089/jop.2021.0001.
34. Zhang J, Zhou Y. Identification of key genes and pathways associated with age-related macular degeneration. J Ophthalmol 2020; 2020:2714746doi: 10.1155/2020/2714746.
35. Konig S, Hadrian K, Schlatt S, Wistuba J, Thanos S, Bohm MRR. Topographic protein profiling of the age-related proteome in the retinal pigment epithelium of Callithrix jacchus with respect to macular degeneration. J Proteomics 2019; 191:1–15. doi: 10.1016/j.jprot.2018.05.016.
36. Yin J, Zhu JM, Shen XZ. New insights into pre-mRNA processing factor 19: a multi-faceted protein in humans. Biol Cell 2012; 104:695–705. doi: 10.1111/boc.201200011.
37. Yano K, Takahashi RU, Shiotani B, Abe J, Shidooka T, Sudo Y, et al. PRPF19 regulates p53-dependent cellular senescence by modulating alternative splicing of MDM4 mRNA. J Biol Chem 2021; 297:100882doi: 10.1016/j.jbc.2021.100882.
38. Sundberg JP, Berndt A, Sundberg BA, Silva KA, Kennedy V, Smith RS, et al. Approaches to investigating complex genetic traits in a Large-Scale Inbred Mouse Aging Study. Vet Pathol 2016; 53:456–467. doi: 10.1177/0300985815612556.
39. Ouyang Y, Xia K, Yang X, Zhang S, Wang L, Ren S, et al. Alternative splicing acts as an independent prognosticator in ovarian carcinoma. Sci Rep 2021; 11:10413doi: 10.1038/s41598-021-89778-0.
40. Genier S, Degrandmaison J, Moreau P, Labrecque P, Hebert TE, Parent JL. Regulation of GPCR expression through an interaction with CCT7, a subunit of the CCT/TRiC complex. Mol Biol Cell 2016; 27:3800–3812. doi: 10.1091/mbc.E16-04-0224.
41. O’Brien R, DeGiacomo F, Holcomb J, Bonner A, Ring KL, Zhang N, et al. Integration-independent transgenic huntington disease fragment mouse models reveal distinct phenotypes and life span in vivo. J Biol Chem 2015; 290:19287–19306. doi: 10.1074/jbc.M114.623561.
42. Kanamaru C, Yamada Y, Hayashi S, Matsushita T, Suda A, Nagayasu M, et al. Retinal toxicity induced by small-molecule Hsp90 inhibitors in beagle dogs. J Toxicol Sci 2014; 39:59–69. doi: 10.2131/jts.39.59.
43. Jones MJK, Gelot C, Munk S, Koren A, Kawasoe Y, George KA, et al. Human DDK rescues stalled forks and counteracts checkpoint inhibition at unfired origins to complete DNA replication. Mol Cell 2021; 81:426–441.e8. doi: 10.1016/j.molcel.2021.01.004.
44. Arlander SJH, Felts SJ, Wagner JM, Stensgard B, Toft DO, Karnitz LM. Chaperoning checkpoint kinase 1 (Chk1), an Hsp90 client, with purified chaperones. J Biol Chem 2006; 281:2989–2998. doi: 10.1074/jbc.M508687200.
45. Quadroni M, Potts A, Waridel P. Hsp90 inhibition induces both protein-specific and global changes in the ubiquitinome. J Proteom 2015; 120:215–229. doi: 10.1016/j.jprot.2015.02.020.
46. Huber N, Sakai N, Eismann T, Shin T, Kuboki S, Blanchard J, et al. Age-related decrease in proteasome expression contributes to defective nuclear factor-kappaB activation during hepatic ischemia/reperfusion. Hepatology 2009; 49:1718–1728. doi: 10.1002/hep.22840.
47. Intartaglia D, Giamundo G, Conte I. The impact of miRNAs in health and disease of retinal pigment epithelium. Front Cell Dev Biol 2020; 8:589985doi: 10.3389/fcell.2020.589985.
48. Romano GL, Platania CBM, Drago F, Salomone S, Ragusa M, Barbagallo C, et al. Retinal and circulating miRNAs in age-related macular degeneration: an in vivo animal and human study. Front Pharmacol 2017; 8:168doi: 10.3389/fphar.2017.00168.
49. Chen J, Sun J, Hu Y, Wan X, Wang Y, Gao M, et al. MicroRNA-191-5p ameliorates amyloid-(1-40 -mediated retinal pigment epithelium cell injury by suppressing the NLRP3 inflammasome pathway. FASEB J 2021; 35:e21184doi: 10.1096/fj.202000645RR.
50. Tian Y, Cui L, Lin C, Wang Y, Liu Z, Miao X. LncRNA CDKN2B-AS1 relieved inflammation of ulcerative colitis via sponging miR-16 and miR-195. Int Immunopharmacol 2020; 88:106970doi: 10.1016/j.intimp.2020.106970.

Age-related macular degeneration; Retinal pigment epithelium; Cell senescence; HSP90AA1; DNA damage checkpoint; Proteasomal subunit components

Supplemental Digital Content

Copyright © 2020 The Chinese Medical Association, produced by Wolters Kluwer, Inc. under the CC-BY-NC-ND license.