The RNA m5C methylase NSUN4 protects retinal ganglion cells by regulating m5C methylation in the Hedgehog signaling pathway

doi:10.21203/rs.3.rs-4687125/v1

Download PDF

Article

The RNA m5C methylase NSUN4 protects retinal ganglion cells by regulating m5C methylation in the Hedgehog signaling pathway

https://doi.org/10.21203/rs.3.rs-4687125/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Glaucoma, one of the leading causes of blindness worldwide, is characterized by retinal ganglion cell (RGC) damage. NOP2/Sun RNA methyltransferase 4 (NSUN4), a known 5-methylcytosine (m5C) writer, is an essential dual-function mitochondrial protein that facilitates both the process of methylation and the coordination of mitoribosome assembly. However, few studies have focused on its role in RGCs. Methylated RNA immunoprecipitation sequencing (MeRIP-seq) combined with RNA sequencing (RNA-seq) was subsequently conducted to identify differences in the m5C methylome and gene expression profile in the N-methyl-d-aspartate (NMDA)-induced RGC injury model in vivo. We demonstrated that the significantly hypomethylated mRNAs were significantly enriched in the Sonic Hedgehog (SHH) signaling pathway in the NMDA group and that the mRNA expression of the m5C writer Nsun4 was downregulated. Subsequently, we examined the role of NSUN4 in R28 cells in response to glutamate stimulation. m5C dot blot and MeRIP (real-time fluorescence quantitative PCR) qPCR were employed to screen and validate the molecular mechanism of NSUN4 in glutamate-induced R28 cells through m5C regulation. Functionally, NSUN4 suppressed Ca²⁺ overload, mitochondrial dysfunction, and apoptosis of R28 cells in vitro. Mechanistically, NSUN4 increased the global mRNA m5C methylation level of Shh, GLI Family Zinc Finger Protein 1 (Gli1) and Gli2. Our study revealed that NSUN4 can alleviate the death of RGCs, which is associated with increased expression of Shh, Gli1 and Gli2 in the SHH signaling pathway in a m5C-dependent manner in R28 cells. Our findings provide new insights into potential biomarkers of retinal excitotoxity.

Biological sciences/Cell biology/Cell death/Apoptosis

Biological sciences/Molecular biology/Epigenetics/Dna methylation

Biological sciences/Neuroscience/Visual system/Retina

Health sciences/Diseases/Eye diseases/Optic nerve diseases

Biological sciences/Cell biology/Mechanisms of disease

Glaucoma is a multifactorial neurodegenerative disease characterized by progressive optic nerve damage and retinal ganglion cell (RGC) death. It is the second leading cause of irreversible blindness worldwide^1,2. The estimated prevalence of glaucoma in the global population aged 40–80 years is 3.5%, with an expected 111.8 million people affected by glaucoma in 2024¹. Currently available treatments involving drugs or surgery cannot reverse glaucomatous vision loss ¹. Therefore, novel strategies to reduce or reverse vision loss caused by glaucoma must be explored³.

The pathophysiology of glaucoma is multifactorial and complex⁴, with excitotoxicity via the N-methyl-d-aspartate (NMDA) receptor, an ionotropic glutamate receptor, which plays a critical role in the death of RGCs in glaucoma⁵. Excessive glutamate or NMDA can contribute to excitotoxicity in various neurodegenerative diseases by activating NMDA receptors to further induce Ca²⁺ overload, reactive oxygen species (ROS) accumulation and mitochondrial dysfunction^3,6,7. Therefore, NMDA has been demonstrated to be a potential target for the treatment of neurodegenerative diseases such as glaucoma³. However, the exact mechanisms of action of NMDA in glaucoma remain unclear.

The sonic hedgehog (SHH) signaling pathway includes the ligands SHH; two transmembrane protein receptors, namely, Patched (PTCH) and Smoothened (SMO); and the GLI family of zinc finger proteins (GLI1, GLI2, and GLI3)⁸. SHH mediates its action via a receptor complex consisting of PTCH and SMO. In the absence of SHH ligand, PTCH suppresses the formation of coreceptors with SMO, thereby inhibiting the transcription of SHH-targeted genes. After binding to the SHH ligand, PTCH inhibits SMO, triggering downstream transcription of GLI1, GLI2 or GLI3 proteins and other SHH-targeted genes^8,9. Early literature mainly explored the role of the SHH signaling pathway in embryonic development¹⁰. Moreover, it is crucial for maintaining healthy cellular function and tissue homeostasis. Recent studies have revealed that SHH signaling participates in various pathological processes, such as neurogenesis and antioxidation, by counteracting ROS release and preventing mitochondrial dysfunction, inflammation, and apoptosis^11–14. Furthermore, dysfunction of the SHH signaling pathway is associated with the occurrence and development of many cancers and neurodegenerative diseases^11–14. A previous study demonstrated that SHH has neuroprotective effects on damaged RGCs in a rat chronic hypertension model by relieving the inhibition of SMO and subsequently activating GLI1¹⁵. Notably, recent findings have shown that the activation of SHH signaling can modulate the function of primary human trabecular meshwork (TM) cells through GLI1¹⁶, suggesting its potential as a therapeutic target for attenuating cell damage in glaucoma.

The RNA 5-methylcytosine (m5C) modification is a common principal type of mRNA methylation ¹⁷. It is regulated by m5C-related regulatory proteins, including methyltransferases (“writers”), demethylases (“erasers”), and methylated reading proteins (“readers”)¹⁸. Among them, the m5C writers NOP2/Sun RNA methyltransferases (NSUNs) have been shown to methylate multiple sites in various RNA families¹⁹. As a reversible epigenetic modification, m5C modifications of RNA affect the fate of the modified RNA molecules and play critical roles in various RNA processes, including RNA stability, protein synthesis, and transcription. Accumulating evidence indicates that the aberrant modification of m5C is related to many abnormal states, such as mitochondrial dysfunction, abnormal embryogenesis, tumors, nervous system disorders, cancer, and cardiovascular diseases^17,20–23. However, the underlying mechanism of m5C modification in glaucoma is not clearly understood. This study is the first to analyze the relationship between the m5C modification profile and an NMDA-induced excitotoxicty model. NSUN4 is a dual-function mitochondrial protein required for both the methylation and coordination of mitoribosome assembly, forming a complex with mitochondrial transcription termination factor 4 (MTERF4)^24,25. This complex is necessary for mitochondrial ribosomal biogenesis and can further regulate mitochondrial translation ²⁶. It is well accepted that glaucoma is often accompanied by mitochondrial dysfunction. Based on these findings, we hypothesize that NSUN4 may play a crucial role in the development of glaucoma.

To validate our hypothesis, an in vivo NMDA-induced excitotoxity model and an in vitro glutamate-induced R28 cell line were used in this study. MeRIP-seq combined with RNA-seq was used to explore the changes in m5C methylation and mRNA levels in a retinal excitotoxicity model. We demonstrated that the number of hypomethylated mRNAs in the NMDA group was greater than that in the control group. Furthermore, NSUN4 played a critical role in preventing Ca²⁺ overload, mitochondrial dysfunction and cell apoptosis in a m5C-dependent manner. Specifically, it can increase the methylation and mRNA levels of Shh, Gli1, and Gli2 in the SHH signaling pathway. Overall, we found that m5C modification played a protective role in RGCs injured by excitotoxity and provided novel insights into the regulation of glaucoma through m5C methylation-mediated excitotoxity.

Animals and NMDA-induced retinopathy rat model. Forty specific pathogen-free Sprague‒Dawley (SD) rats at 7–8 weeks of age were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The rats were raised under a 12/12 h light/dark cycle at 22–25°C and a relative humidity of 38–50% and were provided ad libitum access to food and water.

All rats were divided into two groups—normal control and NMDA (Sigma, Sigma‒Aldrich, St. Louis, MO, USA) (50 mM)—with twenty rats in each group. All the rats were anesthetized by intraperitoneal injection of pentobarbital sodium, and all efforts were made to minimize suffering. Oxybuprocaine hydrochloride (Santen Pharmaceuticals, Tokyo, Japan) was used for ocular surface anesthesia. A microsyringe (Hamilton, Reno, NV, USA) with a 33 G needle was inserted into the vitreous cavity, and 3 µL of saline (Otsuka Pharmaceutical Co., Ltd., China) or NMDA solution was injected into the right eye. All surgical procedures were performed under a stereomicroscope. Levofloxacin eye drops (Santen Pharmaceuticals, Tokyo, Japan) were used to prevent infection after injection. After 7 days of treatment, the treated rats were euthanized by an anesthetic overdose, and eyeballs of the rats were collected for follow-up research. The euthanasia of experimental animals in this research was performed in accordance with internationally recognized guidelines on animal welfare. Once the animals were deeply anesthetized, cervical dislocation was swiftly performed by a trained technician, ensuring a rapid and painless death.

Methylated RNA immunoprecipitation sequencing (MeRIP-seq). MeRIP-seq and subsequent data analyses were mainly supported by CloudSeq Biotechnologies, Inc. (Shanghai, China). In brief, the retinas of the control and NMDA groups were assessed. Total RNA was extracted, and the quality and quantity of total RNA were assessed by using a NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, MA, USA). Then, m5C RNA immunoprecipitation was performed with the GenSeqTM m5C RNA IP Kit (GenSeq Inc., China) following the manufacturer’s instructions. Both the input samples without immunoprecipitation and the m5C IP samples were subjected to RNA-seq library generation with the NEBNext® Ultra II Directional RNA Library Prep Kit (New England Biolabs, Inc., USA). Library quality was evaluated with a BioAnalyzer 2100 system (Agilent Technologies, Inc., USA). An Illumina HiSeq 4000 sequencer was used to harvest paired-end reads, and Q30 was used to control quality. Then, the clean reads of all the libraries were aligned to the reference genome (UCSC hg19) by HISAT2 software (v2.0.4). Methylated sites on RNAs (peaks) and differentially methylated sites were separately determined by MACS software and diffReps. Peaks overlapping with exons of mRNA were further extracted by customized scripts. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed on the differentially methylated protein-coding genes.

Bulk RNA sequencing. The rRNAs were removed from the total RNA by using Ribo-Zero rRNA Removal Kits (Illumina, San Diego, CA, USA). RNA libraries were constructed by using rRNA-depleted RNAs with a TruSeq Stranded Total RNA Library Prep Kit (Illumina, San Diego, CA, USA). Libraries were controlled for quality and quantified using the BioAnalyzer 2100 system (Agilent Technologies, Inc., USA). Then, an Illumina HiSeq sequencer was used to comprehensively detect mRNA in the samples. Paired-end reads were harvested from the Illumina HiSeq 4000 sequencer and checked using a quality score of Q30. After 3’ adaptor trimming and low-quality read removal by Cutadapt software (v1.9.1), the high-quality trimmed reads (clean reads) were aligned to the rat reference genome (UCSC RN5) with HISAT2 software (v2.0.4). Then, guided by the Ensembl GTF gene annotation file, Cuffdiff software (part of Cufflinks) was used to obtain the gene level RPKM (reads per kilobase of transcript per million mapped reads) as the expression profiles of mRNA, and fold change and p value were calculated based on FPKM (fragments per kilobase of transcript per million mapped reads). mRNAs with a |fold change| ≥ 2, a P value ≤ 1.0, and an FPKM value ≥ 0.1 were considered to be differentially expressed. Gene Ontology (GO) and pathway enrichment analyses were performed on the differentially expressed mRNAs.

R28 cell culture and glutamate-induced R28 cell line model. The R28 cells, an immortalized adherent retinal precursor cell line, were generously provided from the Eye Center of Xiangya Hospital, Central South University (Changsha, China). The cells were cultured in low-glucose Dulbecco's modified Eagle’s medium (DMEM) (Gibco, Waltham, MA, USA) supplemented with 10% fetal bovine serum (Gibco, Waltham, MA, USA) and 1% penicillin‒streptomycin (Gibco, Waltham, MA, USA). The cultured cells were maintained at 37°C with 5% CO₂ in a humidified incubator. According to the literature^27,28, the cells in the glutamate group were treated with 10 mM L-glutamate (ab120049, Abcam, Cambridge, UK) and incubated for 24 h.

Quantitative reverse transcription polymerase chain reaction. Total RNA was isolated from R28 cells using TRIpure Total RNA Extraction Reagent (ELK Biotechnology, Denver, CO, USA). Subsequently, cDNA was synthesized utilizing EntiLink™ 1st Strand cDNA Synthesis Super Mix (ELK Biotechnology, Denver, CO, USA). Quantitative real-time polymerase chain reaction (RT‒PCR) was performed using EnTurbo™ SYBR Green PCR SuperMix (ELK Biotechnology, Denver, CO, USA). The specific primers used were purchased from Wuhan Gene Ctreat Corp., Ltd. (Wuhan, China). The primer information used is presented in Table 1. Relative mRNA levels were normalized to those of β-Actin and were calculated using the 2^−ΔΔCT method.

Table 1

Primer used for quantitative real-time-PCR
Genes
β-Actin	Forward	CGTTGACATCCGTAAAGACCTC
β-Actin	Reverse	TAGGAGCCAGGGCAGTAATCT
Nsun4	Forward	CTCTCAGAGCAGAAGTATGGTGC
Nsun4	Reverse	GATAGAGTTGGAGAGAGGCCACT
Shh	Forward	CGATATGAAGGGAAGATCACAAG
Shh	Reverse	TGCTCGACCCTCATAGTGTAGAG
Gli1	Forward	GACTTGAGCATTATGGACAAGTG
Gli1	Reverse	CTGGATTCTAACGAGAGACAGTCAT
Gli2	Forward	GTTCAGCCTTTGGACATACACC
Gli2	Reverse	TCGCTGTTCTGCTTATTCTGGT
Ptch1	Forward	TACATGTATAACAGGCAATGGAAGT
Ptch1	Reverse	GCAGTCCAAAGGTGTAATGATTAAG
Smo	Forward	CAGATGGCACCATGAGATTTG
Smo	Reverse	GAGCAGGTGGAAATAGGATGTC
Note: Nsun4: NOP2/ Sun RNA methytransferase 4; Shh: Sonic Hedgehog; Gli1: GLI family zinc finger 1; Ptch1: Patched 1; Smo: Smoothened.

Construction of the Nsun4 ORF plasmid. The full-length open reading frame (ORF) plasmid of the rat Nsun4 gene (NM_001106678.1) vector was purchased from ELK Biotechnology Co., Ltd. (Wuhan, China). The Nsun4 plasmid DNA was purified from the bacterial culture using an EndoFree Plasmid Miniprep Kit (EP004, ELK Biotechnology, Wuhan, China). Then, the plasmid was digested through the restriction endonuclease sites KpnI and XhoI and inserted into the pcDNA 3.1 vector by T4 DNA ligase to construct overexpression plasmids. For DNA assembly, positive colonies were cultured in Luria–Bertani (LB) medium overnight. The plasmid DNA was further amplified and extracted by PCR using Entilink™ PCR Master Mix (EQ001, ELK Biotechnology, Wuhan, China) and Sanger sequencing in GENE CREATE.cn (Wuhan, China). Supplementary Table S1 shows the cloning primers used.

Transfection of R28 cells. R28 cells were inoculated in a 6-well plate at a density of 5×10⁵cells/well with antibiotic-free culture medium to 60–70% confluency. The same volume of Opti-MEM™ I Reduced Serum Medium (Gibco, Waltham, MA, USA) was used to dilute the Nsun4 plasmids to 2 µg/cell with Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA). Then, we transfected the prepared Nsun4 plasmids into R28 cells using Lipofectamine 2000 and added transfection solution and basic culture medium, which was changed to complete DMEM at 6 h posttransfection. The R28 cells were divided into 3 groups: control, glutamate, and glutamate + NSUN4.

Cell viability. A CCK-8 assay (Beyotime, Shanghai, China) was performed to determine cell viability. R28 cells were seeded into 96-well plates at a density of 1×10⁴ cells/well and then treated with glutamate and NSUN4 for 24 h. Subsequently, 10 µL of a solution containing 10% CCK-8 reagent was added to each well, and the plates were incubated for an additional 3 h in the dark at 37°C. The absorbance at 450 nm was measured using an enzyme-labeled instrument (Diatek, Wuhan, China) to determine the optical density (OD). The cell viability was calculated using the relative cell formula.

Calcium content. The calcium content of the R28 cells in the three groups was measured as previously described²⁹. Cells seeded in 6-well plates were fully lysed. After centrifugation at 14000 × g for 5 min at 4°C, the supernatant was removed and then analyzed using the o-cresolphthalein complexone method to detect calcium deposition at 575 nm with a microplate reader (USCNK, Hubei, China) in accordance with the instructions of the calcium colorimetric assay kit (Beyotime, Shanghai, China).

Apoptosis assay with Annexin V-FITC/PI. Flow cytometry was utilized to detect cell apoptosis using Annexin V-FITC/propidium iodide (PI) staining (Sangon Biotech, Tianjin, China). The R28 cells were suspended in binding buffer and then labeled with 5 µl of Annexin V-FITC along with 5 µl of PI at room temperature for 15 min in the dark. The percentage of apoptotic cells was quantified and analyzed using a flow cytometer (Beckman Coulter, CytoFLEX, USA).

Analysis of the mitochondrial membrane potential. The mitochondrial membrane potential was analyzed using a mitochondrial membrane potential assay kit with JC-1 (Beyotime, Shanghai, China). R28 cells were seeded in a 6-well plate, washed with phosphate-buffered saline (PBS), incubated with JC-1 solution at 37°C in the dark for 30 min, and washed with PBS. Then, 4′,6-diamidino-2-phenylindole (DAPI) was used to stain the nucleus, which was incubated at room temperature in the dark for 30 min and washed with PBS. The cells were observed using a fluorescence microscope (Nikon, Eclipse Ci-L, Japan). The relative ratio of red/green fluorescence was used to evaluate mitochondrial damage.

Dot blot assay. Total mRNA from R28 cells was purified with TRIpure Total RNA Extraction Reagent (EP013, ELK Biotechnology, Denver, CO, USA). Two microliters of mRNA solution were loaded on a polyvinylidene fluoride (PVDF) membrane (Millipore, Massachusetts, USA) and irradiated with UV for 10 min for sample fixation. Then, the membrane was washed in wash buffer (1x TBS and 0.02% Tween-20) for 5 min. The m5C antibody (Ab214727, Abcam, Cambridge, UK) was diluted 1:1000 in 5% nonfat milk in wash buffer. After overnight incubation at 4°C, the membrane was washed three times gently in wash buffer for 5 min. The secondary antibody (HRP-goat anti-rabbit; AS1107, ASPEN, Wuhan, China) was diluted 1:10,000 in wash buffer and incubated for 1 h at room temperature. Then, an enhanced chemiluminescence (ECL) mixed solution (AS1059, ASPEN, Wuhan, China) was added to the sample surface of the membrane, which was exposed in a dark room. The membrane was stained with 0.02% methylene blue (M8030, Solarbio, Beijing, China).

MeRIP-qPCR. A methylated RNA immunoprecipitation (MeRIP) assay was performed using an EZ-Magna RIP Kit (17–701, Millipore, Germany) following the manufacturer’s instructions. R28 cells were washed with cold PBS and centrifuged for 5 min. Then, R28 cells were lysed using RIP lysis buffer supplemented with protease inhibitor cocktail and RNase inhibitor, followed by incubation on ice for 5 min. The supernatants were collected and incubated with 5 µg of m5C antibody (ab214727, Abcam, Cambridge, UK) or negative control normal anti-rabbit IgG (ab172730, Abcam, Cambridge, UK) to capture the RNAs used for qPCR. Each magnetic bead-antibody complex was added to RNA-binding protein‒RNA complex immunoprecipitation buffer (860 µl of RIP wash buffer, 35 µl of 0.5 M EDTA, and 5 µl of RNase inhibitor for each reaction). Then, the RNA‒protein IP complexes were rotated for 3 h overnight at 4°C and washed 6 times with cold RIP Wash Buffer. All tubes were incubated with proteinase K buffer to digest the proteins. Finally, targeted RNAs were extracted from the immunoprecipitated complex, and the relative gene expression of Shh, Gli1 and Gli2 was measured by qPCR. The primer information used is presented in Table 2. The relative enrichment was normalized to the input = 2^{Ct [IP] – Ct [input]}.

Table 2

Primer used for MeRIP-qPCR
Genes		Sequences (5`-3`)
Shh	Forward	CGATATGAAGGGAAGATCACAAG
Shh	Reverse	TGCTCGACCCTCATAGTGTAGAG
Gli1	Forward	GACTTGAGCATTATGGACAAGTG
Gli1	Reverse	CTGGATTCTAACGAGAGACAGTCAT
Gli2	Forward	GTTCAGCCTTTGGACATACACC
Gli2	Reverse	TCGCTGTTCTGCTTATTCTGGT
Note: Shh: Sonic Hedgehog; Gli1: GLI family zinc finger 1.

Statistical analysis. Data analysis was performed using GraphPad Prism 9.0 software (GraphPad Software, Inc., Boston, MA, USA). The Shapiro–Wilk test confirmed that the measurement data for each group were normally distributed, and the results are presented as the mean ± standard deviation (SD). The Levene test verified that the data had a uniform true variance. Unpaired Student's t tests were used for two group comparisons. One-way ANOVA followed by Tukey's post hoc test was used for multiple comparisons. Differences with P < 0.05 were considered to indicate statistical significance.

MeRIP-seq analysis of the NMDA-treated retinal excitotoxic model. In our previous study, an NMDA-induced excitotoxicity model was established in the retina of Sprague Dawley (SD) rats. More specifically, the arrangement of the cells in the ganglion cell layer (GCL) was disrupted 7 days after the intravitreal injection of 50 mM NMDA, resulting in nuclear pyknosis. Moreover, the number of RGCs in the retina of the NMDA-treated group was significantly lower than that in the control group. Furthermore, NMDA injection caused retinal dysfunction via excitotoxicity³⁰. Then, we used MeRIP-seq and RNA-seq to determine the role of m5C mRNA modification in the NMDA group.

A total of 54.75% and 43.21% of the genes in the control group and NMDA group, respectively, exhibited one m5C methylation peak. A total of 11.89% and 13.37% of the genes in each group had two methylation peaks. Fewer genes had three or more methylation peaks (Fig. 1A). The identified m5C peaks were not uniformly distributed across the entire transcriptome. The three most enriched genomic regions for both groups were coding sequences (CDSs), 3’ untranslated regions (3’ UTRs), and stop codons (Fig. 1B). The number of m5C methylation peaks in the CDS region was the highest, representing more than two-thirds of the methylated mRNAs. Peaks in the stop codon and 3’ UTR regions were found in 8.22% and 7.67% of all m5C-methylated mRNAs, respectively (Fig. 1B). The altered m5C peaks were further transcribed from all chromosomes, especially chromosome 1 (chr1) (Fig. 1C). Specifically, 4877 and 9390 m5C peaks were identified in the control and NMDA groups, respectively. Furthermore, 3516 and 6028 genes were separately methylated in the two groups. Next, 2414 genes were identified as overlapping genes between the two groups (Fig. 1D).

Compared with those in the control group, differentially methylated m5C peaks were identified on all chromosomes (especially chr1) in the NMDA group (Fig. 1E). Compared with the control group, the NMDA group exhibited 601 significantly hypermethylated and 986 significantly hypomethylated m5C peaks, defined by an absolute fold change ≥ 2 and P ≤ 0.0001. Compared with those in the control group, there was an increase in the methylation (hypermethylation) of 569 genes and a decrease (hypo-methylation) in 920 genes after exposure to NMDA (Fig. 1F).

RNA-seq analysis of the NMDA-treated RGC damage model. To further investigate the mechanisms of retinal damage, bulk RNA-seq was performed on isolated retinas from the NMDA-treated rats and the control rats. When the threshold was set as a fold change ≥ 2, 3012 expressed mRNAs were retained, of which 1973 were upregulated (red dots) and 1039 were downregulated (green dots) (Fig. 1G and H). Compared to those in the control group, 617 genes were differentially expressed in the NMDA group (|fold change| ≥2 and P < 0.05), including 358 upregulated genes and 259 downregulated genes (Fig. 1I.). Three types of enzymes implicated in RNA m5C methylation were analyzed: methyltransferases (NSUNs), demethylases (ten-eleven translocation (TETs)), and m5C recognition 'reader' proteins (YTH domain-containing family 2 (YTHDF2), Aly/REF export factor (ALYREF), and Y-box binding protein 1 (YBX1))¹⁹. As shown in Table 3, the expression of the writers Nsun6 and Nsun7, the erasers Tet3 and Tet2, and the readers Ybx1 and Ythdf2 was upregulated, while the expression of the writers Nsun4 and Nsun5 and Nsun2, the eraser Tet1, and the reader Alyref was downregulated in the NMDA group compared to the control group. Because the MeRIP results showed that there were more hypermethylated mRNAs than hypomethylated mRNAs following NMDA exposure (Fig. 1G), the m5C regulators resulting in hypomethylation were further analyzed. Satisfactory m5C modifications were observed for the writers Nsun4, Nsun5, Tet1, and Nsun2 and the erasers Tet3 and Tet2. The two different locations of Nsun4 on chr 5 were shown, and the transcript abundance of Nsun4 in the NMDA-treated retinas was 3.18-fold less than that in the control group (Table 3). Given that NSUN4 exhibited the most dramatic changes in expression in response to NMDA, NSUN4 was selected as a susceptible target for the NMDA-induced excitotoxicity model.

Table 3

The mRNA expression level of m5C regulators in the NMDA group compared to the control group in RNA-seq
Gene	Regulation	Fold change	log₂(fold change)	p_value	Regulation	GeneID	Chromosome
Nsun4	writer	-3.178641	-1.66841	1	down	298426	5
Nsun5	writer	-1.459383	-0.545359	0.46255	down	288595	12
Tet1	eraser	-1.417804	-0.503658	0.3017	down	309902	20
Nsun2	writer	-1.242993	-0.313818	0.57125	down	361191	1
Nsun4	writer	-1.074581	-0.103774	1	down	298426	5
Alyref	reader	-1.008412	-0.012086	0.9778	down	690585	10
Tet3	eraser	1.005111	0.007356	0.99265	up	680576	4
Tet2	eraser	1.015189	0.021749	0.9546	up	310859	2
Ybx1	reader	1.044638	0.063003	0.87215	up	500538	5
Ythdf2	reader	1.059622	0.08355	0.85095	up	313053	5
Nsun6	writer	1.310752	0.390395	0.6226	up	307148	17
Nsun7	writer	1.557653	0.639374	1	up	305339	14
Note: Nsun4: NOP2/Sun RNA methyltransferase family member 4; Tet1: tet methylcytosine dioxygenase 1; Alyref: Aly/REF export factor; Ybx1: Y box binding protein 1; Ythdf2: YTH N(6)-methyladenosine RNA binding protein 2.

GO and KEGG pathway analyses revealed that hypomethylated genes were enriched in the Hedgehog signaling pathway. To reveal the biological significance of NMDA treatment, genes containing significantly decreased m5C peaks (significantly hypomethylated genes) in the NMDA group were analyzed by performing GO and KEGG pathway analyses. GO analysis revealed that hypomethylated mRNAs in the NMDA group were significantly enriched in cell communication, signal transduction, and regulation of cytosolic calcium ion concentration (ontology: biological process); membrane part, cell periphery, and plasma membrane (ontology: cellular component); and molecular transducer activity, transmembrane signaling receptor activity, G protein-coupled receptor activity, and ion channel activity (ontology: molecular function; Fig. 2A-C). Interestingly, GO analysis revealed that Nsun4, a m5C regulator, was enriched in mitochondria and the mitochondrial large ribosomal subunit (cellular component) and in the positive regulation of mitochondrial translation (biological process) (Supplementary Table S2).

Remarkably, KEGG pathway analysis demonstrated that the hypomethylated genes in the NMDA group were enriched in the neuroactive ligand‒receptor interaction, olfactory transduction, and Hedgehog (HH) signaling pathway (Fig. 2D), suggesting that these genes may play regulatory roles in NMDA excitotoxity. The genes enriched in the HH signaling pathway included Indian Hedgehog (ihh), Gli1, Gli2, LDL receptor related protein 2 (Lrp2), wingless-type MMTV integration site family, member 3 (Wnt3), Wnt3a, Wnt5a and Wnt6 (Table 4). The HH signaling pathway includes three ligands, SHH, IHH and Desert Hedgehog (DHH); two transmembrane protein receptors, PTCH and SMO; and the downstream transcription factors GLI1, GLI2, and GLI3⁸. Moreover, mRNA-seq revealed that the expression of the overlapping genes Shh, Ihh, and Gli1, but not Gli2, was downregulated in the NMDA group compared to the control group (Table 5). Previous studies have shown that IHH, a choroid-derived signal, can affect the retinal pigment epithelium (RPE), sclera, neural retina, and further retinal photoreceptors^31,32. SHH protected RGCs from damage in a diabetes model in vivo³³. Therefore, SHH, GLI1 and GLI2, which are involved in the HH signaling pathway, were selected for further exploration and identification of in vitro RGC death models.

Table 4

The m5C information of Ihh, Gli1 and Gli2 mRNAs in Hedgehog signaling pathway in the NMDA group compared to the control group in MeRIP-seq
Genes	Foldchange	P value	FDR	Regulation
Ihh	244.2	6.38E-09	1.40E-06	down
Gli1	2.949446	1.86E-06	0.000126	down
Gli2	7.398165	3.85E-08	6.11E-06	down
Lrp2	3.743295	2.66E-08	4.36E-06	down
Wnt3	6.480799	6.04E-08	8.45E-06	down
Wnt3a	8.878613	2.44E-15	5.5E-12	down
Wnt5a	9.92145	1.09E-07	1.36E-05	down
Wnt6	272.3	8.66E-10	2.8E-07	down
Note: Ihh: Indian Hedgehog; Gli1: GLI family zinc finger 1; Lrp2: LDL receptor related protein 2; Wnt3: wingless-type MMTV integration site family, member 3. FDR: false discovery rate.

Table 5

The mRNAs expression of HH signaling pathway in the NMDA group compared to the control group in RNA-seq
Gene	Fold change	log2(fold change)	p_value	FDR	Regulation	GeneID	chromosome
Shh	-5.541206	-2.4702	0.07815	0.997307	down	29499	4
Ihh	-1.061055	-0.085499	1	1	down	84399	9
Gli1	-1.901752	-0.927329	1	1	down	140589	7
Gli2	1.828649	0.870778	0.10335	0.997307	up	304729	13
Nsun4	-3.178641	-1.66841	1	1	down	298426	5
Note: Shh: Sonic Hedgehog; Ihh: Indian Hedgehog; Gli1: GLI family zinc finger 1; Nsun4: NOP2/Sun RNA methyltransferase family member 4.

NSUN4 has an antiapoptotic effect on glutamate-induced RGC excitotoxity. Since the mRNA level of Nsun4, an important m5C writer, was lower in the NMDA group than in the control group according to RNA-seq (Table 5), we further examined whether NSUN4 exerted regulatory effects on genes involved in the HH signaling pathway in glutamate-induced R28 cells. qRT‒PCR was performed to test the changes in the mRNA level of Nsun4 caused by glutamate. It was significantly decreased (P < 0.001) (Fig. 3A), which was consistent with the in vivo RNA-seq results (Table 5). We also tested the mRNA levels of Shh, Gli1, and Gli2 in the HH signaling pathway in vitro using qPCR. As shown in Fig. 3B-D, the expression levels of these genes were significantly downregulated after treatment with glutamate (P < 0.01). In summary, our results suggest that the expression of Nsun4, Shh, and Gli1 was consistent with the RNA-seq results (Table 5).

To clarify whether NSUN4 can protect R28 cells from apoptosis, we used a CCK-8 kit to test the viability of R28 cells treated with 10 mM glutamate or a combination of 10 mM glutamate and NSUN4 for 24 h. The results showed that cell viability was significantly inhibited by treatment with glutamate; however, NSUN4 significantly restored the cell state and improved the cell survival rate (P < 0.001) (Fig. 3E). Therefore, we demonstrated that NSUN4 reduced cell viability under excessive glutamate conditions.

Furthermore, Annexin V-FITC/PI double staining by flow cytometry was used to evaluate the effect of NSUN4 on cell mortality (Fig. 3F and H). The Annexin V-FITC-positive cell population in the three groups included both early- and late-stage apoptotic cells. We found that R28 cells treated with glutamate underwent apoptosis, and the rate of glutamate-induced cell apoptosis was significantly reduced after 24 h of NSUN4 treatment (P < 0.0001) (Fig. 3I). This demonstrated that NSUN4 mitigated glutamate-induced cell apoptosis beginning at the early stage (Fig. 3F and H).

NSUN4 prevents glutamate-induced mitochondrial dysfunction and Ca ²⁺ overload in R28 cells. A decreased mitochondrial potential is an indicator of the early stage of cell apoptosis³⁴, and the mitochondrial membrane potential (MMP) is believed to be an important parameter for assessing mitochondrial function³⁵. We next evaluated the effect of NSUN4 on mitochondrial function by detecting the MMP in glutamate-treated R28 cells. According to the instructions of the JC-1 Assay Kit, the red and green fluorescence of R28 cells labeled with a JC-1 probe after treatment was observed by a fluorescence microscope (Fig. 4A). The red/green ratio was deemed the relative value of the MMP. As shown in Fig. 4B, the MMP decreased significantly in the glutamate group compared with that in the control group (P < 0.05). However, after NSUN4 treatment, the decrease in the MMP caused by glutamate was partially reversed (P < 0.05) (Fig. 4B). Apparently, NSUN4 has protective effects on mitochondria.

Glutamate-induced excitotoxicity can induce Ca²⁺ overload and mitochondrial dysfunction, which are important factors leading to neuronal death³⁶. Therefore, we further used a calcium colorimetric assay kit to investigate the regulatory effect of NSUN4 on intracellular Ca²⁺ homeostasis. Glutamate significantly increased the intracellular Ca²⁺ concentration, while NSUN4 attenuated the glutamate-induced Ca²⁺ increase and contributed to the maintenance of Ca²⁺ homeostasis in R28 cells (P < 0.0001) (Fig. 4C). Overall, NSUN4 inhibited glutamate-induced excitotoxicity by preventing glutamate-induced mitochondrial dysfunction and Ca²⁺overload. The above evidence indicates that NSUN4 alleviates mitochondrial dysfunction, Ca²⁺overload, and even cell apoptosis induced by glutamate-induced cell excitotoxicity.

NSUN4 increased the global level of m5C mRNA methylation in R28 cells induced by glutamate excitotoxicity. To explore the possible involvement of RNA m5C modification in RGC death caused by excitotoxicity in vitro, we first measured global RNA m5C levels in glutamate-induced R28 cell excitotoxicity. mRNA was extracted from R28 cells that were treated with TRIpure, and then,m5C levels were measured via m5C dot blot (Fig. 5A). Dot blot analysis revealed that global mRNA m5C modification was significantly lower in the NMDA group than in the control group (P < 0.001) (Fig. 5A-D), which suggested that NMDA treatment resulted in a decrease in m5C on mRNA. Additionally, a m5C dot blot assay was applied to examine the role of NSUN4 in modulating m5C modification in R28 cells treated with glutamate. NSUN4 overexpression reversed the glutamate-induced decrease in the m5C level in R28 cells (P < 0.0001) (Fig. 5A-D). Overall, these data suggested that NSUN4 increased RNA m5C modification in glutamate-induced RGCs. Next, we detected whether NSUN4 exerts protective effects through regulating the m5C levels of Shh, Gli1, and Gli2.

NSUN4 protected RGCs by activating the Hedgehog-GLI pathway in a m5C‑dependent manner in a glutamate‑induced RGC excitotoxic model. To substantiate the effect of NSUN4 on the m5C level of these genes, we first conducted MeRIP-qPCR using an anti-m5C antibody in R28 cells. SHH, GLI1, and GLI2 were more strongly enriched in the presence of the m5C antibody than in the presence of IgG (P < 0.0001) (Fig. 5E-G), indicating that m5C modification occurred in R28 cells. In addition, a significant decrease in the m5C-modified mRNAs Shh, Gli1, and Gli2 was observed by MeRIP-qPCR upon glutamate disruption (P < 0.01). However, NSUN4 overexpression increased the levels of the m5C-modified mRNAs of these three genes in R28 cells (P < 0.0001). These results suggested that the protective effects of NSUN4 on glutamate-induced RGCs could occur through regulating the m5C levels of the Shh, Gli1, and Gli2 mRNAs (Fig. 5E-G). Similarly, qRT‒PCR showed that the mRNA levels of Shh, Gli1, and Gli2 were decreased dramatically after glutamate treatment. More importantly, the changes in Shh, Gli1, and Gli2 mRNA expression in the glutamate group were reversed by the addition of NSUN4 to R28 cells (P < 0.05) (Fig. 5H-J). We also determined that the expression of Ptch1 and Smo, two types of receptors in the HH signaling pathway, was significantly greater in the glutamate + NSUN4 group than in the glutamate group (P < 0.01) (Fig. 5K-L). Taken together, our results suggested that the expression of genes in the SHH signaling pathway was regulated by NSUN4-mediated m5C modification in R28 cells (Fig. 6).

Glaucoma, one of the leading causes of irreversible blindness worldwide, is a neurodegenerative disease that is associated with the dysfunction and degeneration of RGCs^1,37. However, currently available treatments for preventing completely RGC loss and visual damage, including decreasing intraocular pressure and protecting the optic nerve, are elusive⁵. Normally, glutamate is critical for mediating visual signal transmission among RGCs, bipolar cells and photoreceptors³⁸. Under pathological conditions, excessive glutamate or NMDA can lead to the overstimulation of NMDA receptors, disturbing intracellular calcium homeostasis and inducing oxidative stress damage, inflammation and RGC death, which are termed glutamate excitotoxicity^5,39. Glutamate excitotoxicity, which also induces mitochondrial dysfunction, is a significant initiator of many neurodegenerative diseases, including glaucoma, diabetic retinopathy, and ischemia‒reperfusion injury^38,40. Therefore, understanding the mechanism of RGC excitotoxicity in the early pathological process of glaucoma patients is crucial for the prevention and treatment of this disease. An NMDA-induced retinal injury model is often used to study the pathogenesis of glaucoma^30,41. Our previous study showed that the number of RGCs and the electrophysiological functions of the retina after NMDA treatment were adversely influenced in vivo³⁰. Furthermore, we determined that the expression of the lncRNA Tsix, which is mainly concentrated in the cytoplasm of the retinal GCL, was significantly decreased in the NMDA-induced retinal excitotoxity model⁴². These changes are consistent with previous studies^5,28.

RNA modifications, as well as DNA and histone modifications, are among the epigenetic-level regulators of biological processes²². mRNA modifications, including N6-methyladenosine (m6A), m5C, N1-methyladenosine (m1A), N7-methylguanosine (m7G), pseudouridine and uridylation, form the epitranscriptome to regulate protein synthesis^18,43–45. In recent years, epigenetic studies, including studies on histone modifications, DNA methylation, noncoding RNAs, and m6A methylation, have been widely reported on the neuroprotective effects and regeneration of glaucoma². For example, bioinformatics analysis of the TM tissues of POAG patients revealed that m6A regulatory genes could serve as potential diagnostic biomarkers for primary open-angle glaucoma (POAG)⁴⁶. Guan et al.⁴⁷ reported that m6A levels in the aqueous humor of patients with pseudoexfoliation glaucoma were significantly greater than those in patients with age-related cataracts. Methyltransferase 3 (METTL3), a m⁶A-related enzyme methylase, has the potential to increase the success rate of glaucoma filtration surgery⁴⁸. Nevertheless, it remains unclear whether m5C participates in the development of glaucoma. Based on the results of high-throughput m5C MeRIP-seq and mRNA-seq experiments, for the first time, we identified potential epigenetic changes in m5C modification and mRNA transcription profiles in the retinas of SD rats treated with NMDA solution. In our study, MeRIP-seq revealed more m5C-hypomodified mRNAs than m5C-hypermodified mRNAs in the NMDA-induced excitotoxity group than in the control group. The number of decreased m5C peaks was greater than that of increased m5C peaks, which was induced by NMDA treatment.

Studies have shown that the functions of m5C writers and readers are believed to regulate gene expression posttranscriptionally and are involved in cellular metabolism and movement⁴⁹. Abnormal m5C levels are significantly linked to various types of human disease, such as cancer, tumor, cardiovascular disease, and mitochondrial dysfunction^17,19. To clarify whether m5C modifications can regulate the RGC injury induced by NMDA, we collected retinas after NMDA treatment for a conjoint analysis between MeRIP-Seq and RNA-Seq. Interestingly, the expression of NSUN4, a known m5C writer, was significantly downregulated in the NMDA-induced excitotoxic rat model. This study aimed to investigate whether NSUN4 contributes to restoring RGCs triggered by glutamate. The R28 retinal precursor cell line, which was obtained from SD rat retinas and immortalized, was used for in vitro experiments related to neuroprotection, cytotoxicity, and physiological function of RGCs. By inducing loss of R28 cells and monitoring changes in intracellular signaling pathways, the pathogenesis of glaucoma can be better understood, and new therapeutic strategies can be developed based on these findings^5,27,28,50. In glutamate-treated R28 cells, we found that glutamate could cause cytosolic Ca²⁺ overload, disruption of the MMP, an early marker of apoptosis, and increased apoptosis. In response to glutamate signals, cytosolic Ca²⁺ can enter mitochondria and regulate their function³⁶. Furthermore, the DNA/RNA level of oxidatively damaged mitochondrial could contribute to glaucomatous neurodegeneration⁵¹. Therefore, we believe that glutamate can mediate mitochondrial dysfunction and apoptosis in R28 cells. These findings are consistent with those of previous studies. Ja FA et al.⁵² reported that mitochondrial dysfunction is an underlying mechanism that drives the neurodegenerative process in glaucoma. Furthermore, we determined that the NSUN4 level was significantly lower in the R28 cell model induced by glutamate than in the control group. Then, we wanted to explore whether NSUN4 could inhibit the apoptosis of R28 cells. The results showed that the overexpression of NSUN4 significantly decreased the Ca²⁺overload, mitochondrial dysfunction, and apoptosis in R28 cells. A previous study showed that NSUN4 can catalyze the m5C modification of Sox9 (SRY (sex determining region Y)-box 9), a critical factor that mediates chondrogenic differentiation, to promote the repair of cartilage defects in vivo²⁶. The expression of NSUN4, a mitochondrial-related gene, was positively associated with both breast cancer and prostate cancer risk⁵³. Therefore, we suggested that NSUN4 improved mitochondrial quality in the excitotoxic model for two reasons, namely, the reduction in Ca²⁺ levels by NSUN4 reduced the MMP, and NSUN4 directly reduced the effect on mitochondrial ribosomal protein occurrence. Overall, we confirmed that NSUN4 may protect R28 cells from glutamate-induced excitotoxity by decreasing early apoptosis in mitochondria and the cellular apoptosis system.

Furthermore, MeRIP-seq revealed that genes with differential hypomodification of m5C, including Ihh, Gli1 and Gli2, were enriched in the HH signaling pathway. RNA-seq further showed that Ihh, Shh and Gli1 were downregulated in the in vivo NMDA-induced excitotoxicity model. Previous studies have shown that IHH signaling can directly affect the RPE and developing scleral mesenchyme³¹. In addition, the expression of Ihh was greater in the choroidal endothelium adjacent to the RPE. The RPE and choroid can provide support for retinal photoreceptors³². Furthermore, IHH signaling can promote the development of the neural retina. In summary, IHH, a choroid-derived signal, can regulate the RPE, sclera and neural retina³¹. These results provided novel insights into effective predictors and therapeutic strategies for retinal vascular diseases and choroid-related blinding disorders³². SHH, the best studied ligand of three hedgehog ligands, is very important in vertebrate eye development⁵⁴. SHH was shown to promote Müller glial reprogramming and further retinal regeneration, as well as the generation of RGCs^55–57. SHH can protect RGCs from damage and promote the survival of RGCs from embryonic stem cell-derived neural progenitors⁵⁶. In this study, the m5C expression of Ihh, but not that of Shh, was significantly lower in the retinas of the NMDA group than in those of the control group. However, SHH was selected for further exploration because the retinas of rats were collected for MeRIP-seq, but only RGCs among whole retina cells were influenced by glaucoma.

Zhao et al.³³ reported that SHH had a neuroprotective effect on damaged RGCs in a diabetes model in vivo. SHH also exerts neuroprotective effects on RGCs in a rat chronic ocular hypertension model by relieving the inhibition of SMO and subsequently activating GLI1¹⁵. Therefore, SHH was selected as the target for further study of glaucoma. Excitotoxicity and oxidative stress have been implicated in the pathogenesis of glaucoma³⁹. Moreover, SHH is critical for inhibiting excitotoxicity, ischemic injury, and oxidative stress. The neuronal nitric oxide synthase (nNOS)-SOX2-SHH axis can protect neurons against early excitotoxicity and ischemic injury in cerebral ischemic stroke pathogenesis⁵⁸. He et al.¹¹ reported that SHH can protect against neurite outgrowth in cortical neurons under H₂O₂-induced oxidative stress, possibly through counteracting ROS release and preventing mitochondrial dysfunction in response to oxidative stress. SHH signaling has recently been shown to be involved in pathological angiogenesis in response to tissue hypoxia and ischemic injury. It was concluded that the SHH inhibitor cyclopamine can effectively inhibit the development of laser-induced experimental choroidal neovascularization⁵⁹. In addition, whole‑exome sequencing and Sanger sequencing confirmed that Shh was differentially expressed and was one of the potential pathogenic mutations in peripheral venous blood samples of glaucoma patients compared with those of the ethnicity‑matched normal control group. This finding suggested that it may be a potential risk factor for glaucoma⁶⁰. Notably, a recent study showed that activation of the SHH signaling pathway can regulate the function of primary human TM cells through Gli1; therefore, Gli1 may be a potential target for attenuating cell damage in glaucoma¹⁶. Similarly, the present study revealed that the expression of Shh, Gli1, and Gli2 in the SHH signaling pathway significantly decreased upon glutamate disruption.

Although NSUN4 can protect R28 cells, we wondered whether it is also responsible for the increase in the m5C levels of Shh, Gli1, and Gli2. A dot blot analysis of total RNA revealed that the m5C level was significantly increased by the overexpression of NSUN4. MeRIP-qPCR further verified that Shh, Gli1, and Gli2 were significantly enriched with the m5C antibody. Consistent with our predictions, a significant decrease in m5C-modified Shh, Gli1, and Gli2 was observed in the NMDA group. The overexpression of NSUN4 in R28 cells considerably increased the m5C level of these three genes. In summary, NSUN4 overexpression increased the global level of m5C mRNA methylation and the levels of m5C-modified Shh, Gli1, and Gli2 mRNA in R28 cells (Fig. 5E-G).

In summary, this study is the first to present a map of RNA m5C modifications in the retinas of SD rats treated with NMDA through the use of advanced technologies, providing a basis for future research. Our work revealed that m5C modification participates in RGC injury triggered by NMDA or glutamate. The m5C methyltransferase NSUN4 has a protective effect on excitotoxic-damaged RGCs by alleviating Ca²⁺ overload, mitochondrial dysfunction, and cell apoptosis, which could be related to the increased m5C level of the genes (Shh, Gli1, and Gli2) in the SHH signaling pathway. The mechanism is shown in Fig. 6. These findings are crucial for understanding the pathophysiological processes of glaucoma and developing new treatment methods, as well as for reducing visual impairment and neuroprotection in glaucoma patients. Our findings broaden the current molecular understanding of RGC damage in glaucoma and highlight the ability of NSUN4 to prevent excitotoxity in glutamate-treated R28 cells. In the future, more mechanistic studies are required to confirm these findings, which are also important. Overall, RNA m5C modification is a potential parameter for RGC death caused by NMDA or glutamate excitotoxity both in vivo and in vitro. The neuroprotective effects of NSUN4 were due to improvements in mitochondrial quality and the cellular apoptosis system in an in vitro glutamate-induced excitotoxicity model. Therefore, NSUN4 may be utilized as a new target or marker of glaucoma therapies.

Acknowledgements

The authors would like to thank DY, WR, LD and GC, who helped me with this paper. We thank WK from Tianjin Medical University Eye Hospital for important scientific suggestions. Additionally, I am thankful for the kind gift of the R28 cells from Jianwei Wei from The Second Xiangya Hospital, Central South University.

Author Contributions

YD and WR performed concept and methodology. LY and GC collected and analyzed the data. LY drafted the manuscript and. LD and GC edited and revised the manuscript. All the authors contributed to the manuscript. All the authors have read and approved the final manuscript.

Competing interests

The authors declare that there are no conflicts of interest or financial disclosures.

Funding This work was supported by Nature Science Foundation of Shanxi Province (202103021223402) to YL, and Nature Science Foundation of Shanxi Province (202103021224345), Health Commission of Shanxi Province (2020004), Research Project Supported by Shanxi Scholarship Council of China (2022-203), Shanxi Bethune Hospital Foundation (2021RC005), and Open Project of Shanxi Key Laboratory of Ophthalmology (2023SXKLOS01) to YD.

Patients consent for publication Not applicable.

Ethics approval These animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Tianjin Medical University (Permit Number: SYXK 2018-0004) and were performed in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. And all experiments in this study adhere to the ARRIVE guidelines (https://arriveguidelines.org).

Data availability

The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Kang, J. M. & Tanna, A. P. Glaucoma. Med. Clin. North Am. 105, 493–510 (2021).
Feng, L., Wang, C., Zhang, C., Zhang, W. & Song, W. Role of epigenetic regulation in glaucoma. Biomed. Pharmacother. Biomedecine Pharmacother. 168, 115633 (2023).
He, S., Stankowska, D. L., Ellis, D. Z., Krishnamoorthy, R. R. & Yorio, T. Targets of Neuroprotection in Glaucoma. J. Ocul. Pharmacol. Ther. Off. J. Assoc. Ocul. Pharmacol. Ther. 34, 85–106 (2018).
Baudouin, C., Kolko, M., Melik-Parsadaniantz, S. & Messmer, E. M. Inflammation in Glaucoma: From the back to the front of the eye, and beyond. Prog. Retin. Eye Res. 83, 100916 (2021).
Liu, M., Li, H., Yang, R., Ji, D. & Xia, X. GSK872 and necrostatin-1 protect retinal ganglion cells against necroptosis through inhibition of RIP1/RIP3/MLKL pathway in glutamate-induced retinal excitotoxic model of glaucoma. J. Neuroinflammation 19, 262 (2022).
Jassim, A. H., Fan, Y., Pappenhagen, N., Nsiah, N. Y. & Inman, D. M. Oxidative Stress and Hypoxia Modify Mitochondrial Homeostasis During Glaucoma. Antioxid. Redox Signal. 35, 1341–1357 (2021).
Luo, X. et al. Tetramethylpyrazine nitrone protects retinal ganglion cells against N-methyl-d-aspartate-induced excitotoxicity. J. Neurochem. 141, 373–386 (2017).
Jing, J. et al. Hedgehog signaling in tissue homeostasis, cancers, and targeted therapies. Signal Transduct. Target. Ther. 8, 315 (2023).
Qi, X. & Li, X. Mechanistic Insights into the Generation and Transduction of Hedgehog Signaling. Trends Biochem. Sci. 45, 397–410 (2020).
Cai, E., Barba, M. G. & Ge, X. Hedgehog Signaling in Cortical Development. Cells 13, 21 (2023).
He, W. et al. Sonic hedgehog promotes neurite outgrowth of cortical neurons under oxidative stress: Involving of mitochondria and energy metabolism. Exp. Cell Res. 350, 83–90 (2017).
Caradu, C. et al. Endogenous Sonic Hedgehog limits inflammation and angiogenesis in the ischaemic skeletal muscle of mice. Cardiovasc. Res. 114, 759–770 (2018).
Chen, S.-D., Yang, J.-L., Hwang, W.-C. & Yang, D.-I. Emerging Roles of Sonic Hedgehog in Adult Neurological Diseases: Neurogenesis and Beyond. Int. J. Mol. Sci. 19, 2423 (2018).
Jia, B. et al. Baicalin attenuates dexamethasone-induced apoptosis of bone marrow mesenchymal stem cells by activating the hedgehog signaling pathway. Chin. Med. J. (Engl.) 136, 1839–1847 (2023).
Wu, J., Zhang, S. & Sun, X. Neuroprotective effect of upregulated sonic Hedgehog in retinal ganglion cells following chronic ocular hypertension. Invest. Ophthalmol. Vis. Sci. 51, 2986–2992 (2010).
Wang, X. et al. Activation of Sonic Hedgehog Signaling Pathway Regulates Human Trabecular Meshwork Cell Function. J. Ocul. Pharmacol. Ther. Off. J. Assoc. Ocul. Pharmacol. Ther. 39, 430–438 (2023).
Balachander, K., Priyadharsini, J. V., Roy, A. & Paramasivam, A. Emerging Role of RNA m5C Modification in Cardiovascular Diseases. J Cardiovasc. Transl. Res. 16, 598–605 (2023).
Barbieri, I. & Kouzarides, T. Role of RNA modifications in cancer. Nat. Rev. Cancer 20, 303–322 (2020).
Zhang, Q. et al. The role of RNA m5C modification in cancer metastasis. Int. J. Biol. Sci. 17, 3369–3380 (2021).
Gao, Y. & Fang, J. RNA 5-methylcytosine modification and its emerging role as an epitranscriptomic mark. RNA Biol. 18, 117–127 (2021).
Delaunay, S. et al. Mitochondrial RNA modifications shape metabolic plasticity in metastasis. Nature 607, 593–603 (2022).
Song, P., Tayier, S., Cai, Z. & Jia, G. RNA methylation in mammalian development and cancer. Cell Biol. Toxicol. 37, 811–831 (2021).
Yuan, Y. et al. Identification of ALYREF in pan cancer as a novel cancer prognostic biomarker and potential regulatory mechanism in gastric cancer. Sci. Rep. 14, 6270 (2024).
Spåhr, H., Habermann, B., Gustafsson, C. M., Larsson, N.-G. & Hallberg, B. M. Structure of the human MTERF4-NSUN4 protein complex that regulates mitochondrial ribosome biogenesis. Proc. Natl. Acad. Sci. U. S. A. 109, 15253–15258 (2012).
Ye, X., Han, Y., Zhang, L., Liu, W. & Zuo, J. MTERF4 regulates the mitochondrial dysfunction induced by MPP(+) in SH-SY5Y cells. Biochem. Biophys. Res. Commun. 464, 214–220 (2015).
Yang, L. et al. Nsun4 and Mettl3 mediated translational reprogramming of Sox9 promotes BMSC chondrogenic differentiation. Commun. Biol. 5, 495 (2022).
Dai, S. et al. Protective activity of tert-butylhydroquinone against oxidative stress and apoptosis induced by glutamate agonizts in R28 cells and mice retina. Biomed. Pharmacother. Biomedecine Pharmacother. 152, 113117 (2022).
Wang, C. et al. The neuroprotective effect of melatonin in glutamate excitotoxicity of R28 cells and mouse retinal ganglion cells. Front. Endocrinol. 13, 986131 (2022).
Zhang, S. et al. Prenatal dexamethasone exposure induces anxiety- and depressive-like behavior of male offspring rats through intrauterine programming of the activation of NRG1-ErbB4 signaling in hippocampal PV interneurons. Cell Biol. Toxicol. 39, 657–678 (2023).
Wang S., Li Y., Liu X. & Zhang Y. The m5C-methylated epi-transcriptomic analysis in a rat model of N-methyl-D-aspartate-induced retinal excitotoxicity. Chin J Ocul Fundus Dis 39, 836–843 (2023).
Dakubo, G. D. et al. Indian hedgehog signaling from endothelial cells is required for sclera and retinal pigment epithelium development in the mouse eye. Dev. Biol. 320, 242–255 (2008).
Lehmann, G. L. et al. Single-cell profiling reveals an endothelium-mediated immunomodulatory pathway in the eye choroid. J. Exp. Med. 217, e20190730 (2020).
Zhao, X. et al. The Effects of Sonic Hedgehog on Retinal Müller Cells Under High-Glucose Stress. Invest. Ophthalmol. Vis. Sci. 56, 2773–2782 (2015).
He, R. et al. Melatonin resists oxidative stress-induced apoptosis in nucleus pulposus cells. Life Sci. 199, 122–130 (2018).
Li, X. et al. Mitochondrial Protection and Against Glutamate Neurotoxicity via Shh/Ptch1 Signaling Pathway to Ameliorate Cognitive Dysfunction by Kaixin San in Multi-Infarct Dementia Rats. Oxid. Med. Cell. Longev. 2021, 5590745 (2021).
Rueda, C. B. et al. Glutamate excitotoxicity and Ca2+-regulation of respiration: Role of the Ca2 + activated mitochondrial transporters (CaMCs). Biochim. Biophys. Acta 1857, 1158–1166 (2016).
Bou Ghanem, G. O., Wareham, L. K. & Calkins, D. J. Addressing neurodegeneration in glaucoma: Mechanisms, challenges, and treatments. Prog. Retin. Eye Res. 100, 101261 (2024).
Boccuni, I. & Fairless, R. Retinal Glutamate Neurotransmission: From Physiology to Pathophysiological Mechanisms of Retinal Ganglion Cell Degeneration. Life Basel Switz. 12, 638 (2022).
Sim, R. H., Sirasanagandla, S. R., Das, S. & Teoh, S. L. Treatment of Glaucoma with Natural Products and Their Mechanism of Action: An Update. Nutrients 14, 534 (2022).
Nakamura, N. et al. Neuroprotective role of sphingolipid rheostat in excitotoxic retinal ganglion cell death. Exp. Eye Res. 208, 108623 (2021).
Zou, J. et al. Melatonin protects against NMDA-induced retinal ganglion cell injury by regulating the microglia-TNFα-RGC p38 MAPK pathway. Int. Immunopharmacol. 118, 109976 (2023).
Li, Y. et al. Expression of lncRNA Tsix in a mouse model of N-methyl-D-aspartic acid-induced retinal excitotoxicity and its significance. Chin. J. Exp. Ophthalmol. 39, 10 (2021).
Wen, K., Zhang, Y., Li, Y., Wang, Q. & Sun, J. Comprehensive analysis of transcriptome-wide m6A methylome in the anterior capsule of the lens of high myopia patients. Epigenetics 16, 955–968 (2021).
Zhao, B. S., Roundtree, I. A. & He, C. Post-transcriptional gene regulation by mRNA modifications. Nat. Rev. Mol. Cell Biol. 18, 31–42 (2017).
Zhu, X., Zhou, C., Zhao, S. & Zheng, Z. Role of m6A methylation in retinal diseases. Exp. Eye Res. 231, 109489 (2023).
zhang, X., Zhou, X. & Duan, X. Exploring diagnostic m6A regulators in primary open-angle glaucoma. Invest. Ophthalmol. Vis. Sci. 64, 3470 (2023).
Guan, J. et al. Critical role of transcriptome-wide m6A methylation in the aqueous humor of patients with pseudoexfoliation glaucoma. Exp. Eye Res. 231, 109473 (2023).
Liu, Y., Gu, C., Li, X., Wang, T. & Yu, L. Involvement of METTL3/m6Adenosine and TGFβ/Smad3 signaling on Tenon’s fibroblasts and in a rabbit model of glaucoma surgery. J. Mol. Histol. 52, 1129–1144 (2021).
Liu, R.-J., Long, T., Li, J., Li, H. & Wang, E.-D. Structural basis for substrate binding and catalytic mechanism of a human RNA:m5C methyltransferase NSun6. Nucleic Acids Res. 45, 6684–6697 (2017).
Gao, Z. et al. Valdecoxib Protects against Cell Apoptosis Induced by Endoplasmic Reticulum Stress via the Inhibition of PERK-ATF4-CHOP Pathway in Experimental Glaucoma. Int. J. Mol. Sci. 23, 12983 (2022).
Gu, L., Kwong, J. M., Caprioli, J. & Piri, N. DNA and RNA oxidative damage in the retina is associated with ganglion cell mitochondria. Sci. Rep. 12, 8705 (2022).
Ja, F.-A. et al. Glaucoma: from pathogenic mechanisms to retinal glial cell response to damage. Front. Cell. Neurosci. 18, (2024).
Li, Y. et al. Mitochondrial related genome-wide Mendelian randomization identifies putatively causal genes for multiple cancer types. EBioMedicine 88, 104432 (2023).
Choy, S. W. & Cheng, S. H. Hedgehog signaling. Vitam. Horm. 88, 1–23 (2012).
Kaur, S. et al. let-7 MicroRNA-Mediated Regulation of Shh Signaling and the Gene Regulatory Network Is Essential for Retina Regeneration. Cell Rep. 23, 1409–1423 (2018).
Jagatha, B. et al. In vitro differentiation of retinal ganglion-like cells from embryonic stem cell derived neural progenitors. Biochem. Biophys. Res. Commun. 380, 230–235 (2009).
Jeon, S., Baek, J. & Lee, W. K. Gli1 Expression in Human Epiretinal Membranes. Invest. Ophthalmol. Vis. Sci. 58, 651–659 (2017).
Zhang, D. et al. nNOS Translocates into the Nucleus and Interacts with Sox2 to Protect Neurons Against Early Excitotoxicity via Promotion of Shh Transcription. Mol. Neurobiol. 53, 6444–6458 (2016).
He, H., Zhang, H., Li, B., Li, G. & Wang, Z. Blockade of the sonic hedgehog signalling pathway inhibits choroidal neovascularization in a laser-induced rat model. J. Huazhong Univ. Sci. Technol. Med. Sci. Hua Zhong Ke Ji Xue Xue Bao Yi Xue Ying Wen Ban Huazhong Keji Daxue Xuebao Yixue Yingdewen Ban 30, 659–665 (2010).
Huang, X. et al. Mutation analysis of the genes associated with anterior segment dysgenesis, microcornea and microphthalmia in 257 patients with glaucoma. Int. J. Mol. Med. 36, 1111–1117 (2015).

No competing interests reported.

Download PDF

Reviewers agreed at journal
07 Aug, 2024
Reviews received at journal
19 Jul, 2024
Reviewers agreed at journal
16 Jul, 2024
Reviewers invited by journal
16 Jul, 2024
Editor assigned by journal
16 Jul, 2024
Editor invited by journal
15 Jul, 2024
Submission checks completed at journal
12 Jul, 2024
First submitted to journal
04 Jul, 2024

You are reading this latest preprint version

The RNA m5C methylase NSUN4 protects retinal ganglion cells by regulating m5C methylation in the Hedgehog signaling pathway

Status:

Version 1

Abstract

Figures

Introduction

Methods

RESULTS

DISCUSSION

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1