Tet methylcytosine dioxygenase 2(TET2)-dependent epigenetic regulation in the pathogenesis of polycystic ovary syndrome.

Based on the inclusion and exclusion criteria, a total of 24 participants were recruited, including 12 healthy controls and 12 patients diagnosed with PCOS. The baseline characteristics of both groups are shown in Table  1 . There were no significant differences in age and basal follicle-stimulating hormone (FSH) levels between the two groups. However, PCOS patients exhibited significantly higher Luteinizing Hormone (LH) levels ( P  < 0.05). In addition, BMI, levels of LH/FSH ratio, testosterone (T), and anti-Müllerian hormone (AMH) were significantly elevated in the PCOS group compared to controls ( P  < 0.05). Granulosa cells were collected from both groups, and total genomic DNA was extracted using a DNA extraction kit. DNA concentrations were adjusted to 100 ng/μL according to the kit’s instructions. The relative 5hmC modification levels were measured using an ELISA-based assay. ELISA results revealed that the total 5hmC levels in granulosa cells were increased in the PCOS group compared to non-PCOS controls (Fig.  1 A). Furthermore, immunofluorescence staining for 5hmC demonstrated enhanced relative fluorescence intensity in granulosa cells from PCOS patients, consistent with the ELISA findings (Fig.  1 B). To further investigate the distribution of differentially modified 5hmC sites in granulosa cells, we employed the Infinium MethylationEPIC BeadChip to profile DNA from both groups, analyzing over 850,000 CpG sites to detect differences in 5hmC modification between PCOS patients and controls. The results revealed distinct enrichment patterns of 5hmC-modified loci between the two groups. Density plots of β-values indicated a global increase in 5hmC levels in the PCOS group compared to controls (Fig.  1 C). More precisely, 5hmC enrichment was found to be elevated in the granulosa cell DNA of the PCOS group, particularly at proximal promoter elements (TSS1500) and the 5'UTR (Fig.  1 D). In total, 71 hyper-hydroxymethylated probes (DMPs) and 76 hypo-hydroxymethylated probes (DMPs) were identified between the PCOS and control groups (Fig.  1 E). We also analyzed the distribution of transcription factor binding sites in relation to their target genes. Hydroxymethylation trend plots around ± 5 kb of transcription start sites (TSS) revealed significant differences in 5hmC enrichment between the two groups, as confirmed by statistical analysis (Fig.  1 F). These findings demonstrate that ovarian granulosa cells from PCOS patients exhibit elevated levels of 5hmC modification compared to healthy controls. Table 1 Comparison of clinical features between the two groups Control(n = 20) PCOS(n = 20) p Value Age(year) 30.85 ± 2.76 29.75 ± 2.83 0.221 BMI(kg/m 2 ) 21.45 ± 1.41 25.23 ± 1.43  < 0.0001 AMH(ng/ml) 2.75 ± 0.32 5.63 ± 0.74 0.009 FSH(U/L) 6.21 ± 1.46 5.61 ± 0.64 0.100 LH(U/L) 2.58 ± 0.57 4.82 ± 0.89  < 0.0001 T(ng/dL) 44.94 ± 5.63 62.38 ± 26.84  < 0.0001 E2(pg/ml) 52.49 ± 16.84 55.60 ± 14.49 0.535 LH/FSH ratio 0.38 ± 0.09 0.98 ± 0.30  < 0.0001 Data were analysed using unpaired Student’s t-test and presented as mean ± SD. P  < 0.05 is considered significant. T testosterone; AMH anti-Müllerian hormone Fig. 1 Increased DNA 5hmC modification levels in ovarian granulosa cells of PCOS patients. A : Relative expression levels of DNA 5-hydroxymethylcytosine in granulosa cells from PCOS and control groups were measured using ELISA. The y-axis represents the percentage of 5hmC relative to total DNA content. ( n  = 24, * P  < 0.05). B : Immunofluorescence staining of granulosa cells from both groups was performed using a 5hmC-specific antibody. Nuclei were counterstained with Hoechst. The fluorescence intensity of 5hmC (red) was quantified using ImageJ software. Values are presented in relative units (fluorescence intensity per cell). ( n  = 3, * * P  < 0.01, scale bar = 100 μm). C : Standardized β-value density curve. The x-axis of the graph represents the β-value, which characterizes the degree of DNA methylation. The y-axis represents the frequency, where a higher y-value indicates a greater number of methylation sites at a certain methylation level. Typically, the density curve of β-values shows a bimodal distribution, where the majority of methylation sites are either highly methylated or low-methylated. The red line represents the PCOS group, while the blue line represents the control group. D : The pie chart illustrates the overall genomic distribution of 5hmC within DNA. E : Chromosomal distribution of differentially DMPs. The scatter plots illustrate the distribution of genes with significant changes in 5hmC levels across chromosomes. The y-axis represents the Delta Beta value (difference in 5hmC levels between PCOS and Control groups).Top panel (N = 71): Genes showing significantly increased 5hmC levels (Hyper-hydroxymethylation) in the PCOS group (Delta Beta≥  0.2).Bottom panel (N = 76): Genes showing significantly decreased 5hmC levels (Hypo-hydroxymethylation) in the PCOS group (Delta Beta  ≤ -0.2). Each dot represents a specific genomic site/gene, color-coded by chromosome. F : 5hmC levels in relation to transcription start sites. X-axis shows the average number of 5hmCs at CpG sites, Y -axis shows the number of detected 5hmCs in two groups Comparison of clinical features between the two groups Data were analysed using unpaired Student’s t-test and presented as mean ± SD. P  < 0.05 is considered significant. T testosterone; AMH anti-Müllerian hormone Increased DNA 5hmC modification levels in ovarian granulosa cells of PCOS patients. A : Relative expression levels of DNA 5-hydroxymethylcytosine in granulosa cells from PCOS and control groups were measured using ELISA. The y-axis represents the percentage of 5hmC relative to total DNA content. ( n  = 24, * P  < 0.05). B : Immunofluorescence staining of granulosa cells from both groups was performed using a 5hmC-specific antibody. Nuclei were counterstained with Hoechst. The fluorescence intensity of 5hmC (red) was quantified using ImageJ software. Values are presented in relative units (fluorescence intensity per cell). ( n  = 3, * * P  < 0.01, scale bar = 100 μm). C : Standardized β-value density curve. The x-axis of the graph represents the β-value, which characterizes the degree of DNA methylation. The y-axis represents the frequency, where a higher y-value indicates a greater number of methylation sites at a certain methylation level. Typically, the density curve of β-values shows a bimodal distribution, where the majority of methylation sites are either highly methylated or low-methylated. The red line represents the PCOS group, while the blue line represents the control group. D : The pie chart illustrates the overall genomic distribution of 5hmC within DNA. E : Chromosomal distribution of differentially DMPs. The scatter plots illustrate the distribution of genes with significant changes in 5hmC levels across chromosomes. The y-axis represents the Delta Beta value (difference in 5hmC levels between PCOS and Control groups).Top panel (N = 71): Genes showing significantly increased 5hmC levels (Hyper-hydroxymethylation) in the PCOS group (Delta Beta≥  0.2).Bottom panel (N = 76): Genes showing significantly decreased 5hmC levels (Hypo-hydroxymethylation) in the PCOS group (Delta Beta  ≤ -0.2). Each dot represents a specific genomic site/gene, color-coded by chromosome. F : 5hmC levels in relation to transcription start sites. X-axis shows the average number of 5hmCs at CpG sites, Y -axis shows the number of detected 5hmCs in two groups To elucidate the molecular changes in PCOS ovarian granulosa cells, KEGG pathway enrichment analysis was conducted on genes harboring differentially modified 5hmC sites. Results indicated that hyper-hydroxymethylated genes in the PCOS group were predominantly enriched in the MAPK signaling, gap junction, and oocyte meiosis pathways (Fig.  2 A). Conversely, hypo-hydroxymethylated genes were primarily associated with Wnt and ErbB signaling pathways (Fig.  2 B). Gene Ontology (GO) analysis further indicated that the differentially modified genes were mainly involved in metabolism-related pathways (Fig.  2 C). As polycystic ovary syndrome is a disease triggered by gene–environment interactions, we examined the methylation status of known PCOS susceptibility loci. We compared the genes annotated from DMPs with PCOS susceptibility genes listed in the GWAS Catalog database ( https://www.ebi.ac.uk/gwas/ ). This analysis identified four genes—ASIC2, CCDC91, CHEK2, and FTO—corresponding to probes cg24044290, cg08568970, cg04661358, and cg10331105, respectively (Fig.  2 D). We designed primers for the four probe regions and validated the 5hmC modification levels using hydroxymethylated DNA immunoprecipitation followed by quantitative PCR (hMeDIP-qPCR) in granulosa cells from an independent validation cohort of 16 individuals. The results showed significant differences in 5hmC modification levels at the ASIC2 and CCDC91 loci between PCOS and control groups, while CHEK2 and FTO did not reach statistical significance (Fig.  2 E). Receiver Operating Characteristic (ROC) analysis was performed for each of the four candidate loci, and ROC curves were plotted. Individually, these loci exhibited limited predictive power for PCOS (Supplementary Fig.  1 A). However, when combined, the four loci demonstrated high predictive accuracy for PCOS diagnosis, with an area under the curve (AUC) of 0.953 (Fig.  2 F). Fig. 2 Analysis of differential 5hmC sites between the two patient groups. A : KEGG pathway analysis for genes corresponding to hyper-hydroxymethylated 5hmC probes in PCOS. B : KEGG pathway analysis for genes corresponding to hypo-hydroxymethylated 5hmC probes in PCOS. C : GO biological process analysis of genes with differential 5hmC modification. D : Venn diagram showing the overlap between genes annotated from differentially DMPs and known PCOS susceptibility genes from genome-wide association studies (GWAS). E : Bar plots showing the 5hmC levels of four selected genes in granulosa cells from both groups. 5hmC levels were quantified by hydroxymethylated DNA immunoprecipitation followed by q-PCR. ( n  = 8, * P  < 0.05, ns P ≥ 0.05). F: Receiver operating curve (ROC) of the diagnosis model based on DNA 5hmC-qPCR Analysis of differential 5hmC sites between the two patient groups. A : KEGG pathway analysis for genes corresponding to hyper-hydroxymethylated 5hmC probes in PCOS. B : KEGG pathway analysis for genes corresponding to hypo-hydroxymethylated 5hmC probes in PCOS. C : GO biological process analysis of genes with differential 5hmC modification. D : Venn diagram showing the overlap between genes annotated from differentially DMPs and known PCOS susceptibility genes from genome-wide association studies (GWAS). E : Bar plots showing the 5hmC levels of four selected genes in granulosa cells from both groups. 5hmC levels were quantified by hydroxymethylated DNA immunoprecipitation followed by q-PCR. ( n  = 8, * P  < 0.05, ns P ≥ 0.05). F: Receiver operating curve (ROC) of the diagnosis model based on DNA 5hmC-qPCR Since genomic 5hmC levels are mediated by the TET family, we first screened the expression of TET1–3 in PCOS granulosa cells to elucidate the specific enzymatic contributor to the observed epigenetic changes. Western blotting showed that TET1 and TET2—but not TET3—were upregulated in PCOS-derived granulosa cells, with TET2 displaying the most significant elevation (Fig.  3 A and B ). TET2 lacks the CXXC domain of TET1/3, requiring transcription factor interaction for recruitment. This specificity makes targeting TET2 a safer, more precise alternative to toxic, broad-spectrum epigenetic inhibition [ 25 ]. Consequently, TET2 was prioritized as the primary target for our investigation. To further investigate the potential genomic targets of TET2 in ovarian granulosa cells, we conducted ChIP-seq analysis. First, we assessed the specificity of the TET2 antibody used in immunoprecipitation. The IP results confirmed that the antibody had good specificity and met the requirements for subsequent ChIP-seq experiments (Fig.  3 C). Analysis of the sequencing data revealed that TET2 binding sites were predominantly enriched in promoter regions (Fig.  3 D). Examination of the distribution of TET2 binding around transcription start sites (TSS) indicated a clear enrichment in the TET2_IP group at TSS regions, whereas the TET2_Input control group exhibited much weaker signals. This supports the specific binding of TET2 to TSS regions (Fig.  3 E). Motif analysis identified CTAGATCAAATA as the most significantly enriched DNA motif at TET2 binding sites (Fig.  3 F). Additionally, GO and KEGG enrichment analyses of TET2-bound genes revealed that they were primarily enriched in the mTOR signaling pathway and regulation of biological processes (Supplementary Fig.  2 A, 2B). We also highlighted the top five genes showing the most significant TET2 binding enrichment, illustrating potential direct transcriptional targets of TET2 (Fig.  3 G). Fig. 3 Target analysis of TET2 function. A , B : Western blot analysis of the expression levels of the three TET enzymes in granulosa cells from both groups. (n = 3, * P  < 0.05). C : ChIP-seq validation using anti-TET2 antibody. D : Genomic distribution of TET2-binding DNA regions. E : TET2 binding profile around transcription start sites. TET2_IP data show significant enrichment of TET2 near TSS, while TET2_Input serves as a control. F : Motif analysis of TET2 ChIP-seq binding regions. G : IGV screenshots showing the genomic loci of the top five differentially bound genes identified by TET2 ChIP-seq. Annotated genes were prioritized by fold enrichment. The top 10 promoter-associated genes were visualized within a ± 3 kb genomic window surrounding the target regions Target analysis of TET2 function. A , B : Western blot analysis of the expression levels of the three TET enzymes in granulosa cells from both groups. (n = 3, * P  < 0.05). C : ChIP-seq validation using anti-TET2 antibody. D : Genomic distribution of TET2-binding DNA regions. E : TET2 binding profile around transcription start sites. TET2_IP data show significant enrichment of TET2 near TSS, while TET2_Input serves as a control. F : Motif analysis of TET2 ChIP-seq binding regions. G : IGV screenshots showing the genomic loci of the top five differentially bound genes identified by TET2 ChIP-seq. Annotated genes were prioritized by fold enrichment. The top 10 promoter-associated genes were visualized within a ± 3 kb genomic window surrounding the target regions Previous studies have reported that TET2 can promote autophagy through various pathways [ 26 ]. However, whether TET2 regulates autophagy in ovarian granulosa cells requires further experimental validation. To investigate this, we established a TET2-overexpressing human granulosa cell model by transfecting KGN cells with an overexpression plasmid. The efficiency of TET2 overexpression was confirmed by qPCR and Western blot analysis, demonstrating a significant increase in both TET2 mRNA and protein levels (Fig.  4 A and B ). To assess the impact of TET2 on cell proliferation, we performed 5-ethynyl-2'-deoxyuridine (EdU) incorporation assays, which revealed a significant inhibition of granulosa cell proliferation upon TET2 overexpression (Fig.  4 C). While hyperandrogenism is known to induce autophagy in PCOS, its role in regulating TET2 remains controversial, as studies in prostate cancer suggest an inhibitory effect [ 27 ]. To reconcile this inconsistency, we examined the dose-dependent effects of DHEA on TET2 in granulosa cells. Surprisingly, Western blot and q-PCR analysis revealed that TET2 expression remained unchanged across all tested concentrations of DHEA (Fig.  4 D and E ). Then we analyzed autophagy-related protein expression via Western blot to evaluate the impact of TET2 on autophagy. The results showed that TET2 overexpression enhanced DHEA-induced autophagic flux, whereas treatment with a TET2 inhibitor markedly suppressed autophagy (Fig.  4 F). Furthermore, we transfected KGN cells with a GFP-RFP-LC3 dual-labeled plasmid using liposome-mediated transfection to visualize autophagosome formation. The results indicated that TET2 significantly promoted androgen-induced autophagosome formation, while the TET2 inhibitor exerted a suppressive effect (Fig.  4 G). We designed three sets of siRNAs targeting TET2 and identified siRNA-2 as exhibiting the most potent knockdown efficiency at the mRNA level (Supplementary Fig.  1 B). Subsequently, the protein expression of TET2 following knockdown was assessed via Western blot, which confirmed a significant reduction in TET2 protein levels (Supplementary Fig.  1 C). Utilizing this TET2-siRNA, we observed that TET2 silencing suppressed the activation of autophagic flux induced by DHEA, suggesting that TET2 may serve as a potential therapeutic target for improving ovarian function in PCOS, particularly in those with hyperandrogenism (Fig.  4 H). Fig. 4 TET2 overexpression promotes androgen-triggered autophagic flux. A , B : Western blot and RT-qPCR validation of TET2 overexpression in KGN cells following plasmid transfection. ( n  = 3, ** P  < 0.01). C : EdU staining assay to assess the effect of TET2 overexpression. (scale bar = 100 μm). D , E : Granulosa cells were treated with varying concentrations of DHEA (0, 500 nM, 1 uM, and 5 uM). Representative Western blot showing consistent TET2 protein levels across groups, and RT-qPCR analysis confirming no significant change in TET2 mRNA expression relative to control. (n = 3, ns P  ≥ 0.05). F : Representative western blot images showing the protein levels of key autophagy markers, including ATG7, Beclin1, LC3-I/II, and P62, in KGN cells treated with DHEA, DHEA + TET2-OE (overexpression), or DHEA + Bobcat339. ( n  = 3, * P  < 0.05). G : Dual fluorescence RFP-GFP-LC3 system to detect autophagy levels. Yellow puncta represent autophagosomes (GFP + RFP +), and red puncta indicate autolysosomes (RFP + only). Activation of autophagy leads to an increase in both yellow and red puncta. (scale bar = 20 μm). H : Assessment of autophagy levels via Western blot following TET2 knockdown. ( n  = 3, * P  < 0.05, ** P  < 0.01) TET2 overexpression promotes androgen-triggered autophagic flux. A , B : Western blot and RT-qPCR validation of TET2 overexpression in KGN cells following plasmid transfection. ( n  = 3, ** P  < 0.01). C : EdU staining assay to assess the effect of TET2 overexpression. (scale bar = 100 μm). D , E : Granulosa cells were treated with varying concentrations of DHEA (0, 500 nM, 1 uM, and 5 uM). Representative Western blot showing consistent TET2 protein levels across groups, and RT-qPCR analysis confirming no significant change in TET2 mRNA expression relative to control. (n = 3, ns P  ≥ 0.05). F : Representative western blot images showing the protein levels of key autophagy markers, including ATG7, Beclin1, LC3-I/II, and P62, in KGN cells treated with DHEA, DHEA + TET2-OE (overexpression), or DHEA + Bobcat339. ( n  = 3, * P  < 0.05). G : Dual fluorescence RFP-GFP-LC3 system to detect autophagy levels. Yellow puncta represent autophagosomes (GFP + RFP +), and red puncta indicate autolysosomes (RFP + only). Activation of autophagy leads to an increase in both yellow and red puncta. (scale bar = 20 μm). H : Assessment of autophagy levels via Western blot following TET2 knockdown. ( n  = 3, * P  < 0.05, ** P  < 0.01) By intersecting the differentially 5hmC-modified genes in PCOS granulosa cells with TET2 ChIP-seq target genes, we identified USP45 as a potential downstream target of TET2 within the ovarian follicular microenvironment (Fig.  5 A). Analysis of the USP45 promoter region identified multiple TET2 binding motifs (Fig.  5 B). Based on sequencing results, we designed primers for two USP45 promoter sites and performed hMeDIP-qPCR to assess 5hmC levels. Results showed that TET2 overexpression significantly elevated 5hmC modification at both sites (Fig.  5 C and D ). In addition, we designed MSP primers for the USP45 promoter region and performed methylation detection on the promoter sites following bisulfite conversion of DNA from granulosa cells in both PCOS and control groups. The results showed that the methylation level of USP45 in the control group was significantly higher than that in the PCOS group (Fig.  5 E). Furthermore, a comparison of the 5hmC modification levels at the USP45 promoter in granulosa cells revealed a significant increase in the PCOS group compared to the control group (Fig.  5 F). Comparing the mRNA levels of USP45 in granulosa cells between the two groups revealed that USP45 mRNA expression was significantly increased in the PCOS group (Fig.  5 G). Collectively, these findings suggest that TET2 may participate in the progression of PCOS by regulating USP45.As a deubiquitinating enzyme, USP45 selectively removes ubiquitin from target proteins to prevent proteasomal degradation and maintain protein homeostasis [ 28 ]. To identify the downstream targets of USP45, we searched relevant databases and identified PRKAA1 as a potential candidate. To further validate this interaction, we established USP45 knockdown and overexpression models (Fig.  6 A and B , Supplementary Fig.  1 D). Western blot analysis revealed that the expression pattern of PRKAA1 was largely consistent with that of USP45 (Fig.  6 C). Overexpression of TET2 led to a significant increase in USP45 expression at both the mRNA and protein levels (Fig.  6 D and E ). However, overexpression of TET2 had no significant effect on mitochondrial membrane potential or ROS levels in granulosa cells (Supplementary Fig.  1 E, 1F). Western blot analysis revealed that TET2 overexpression reduced overall cellular ubiquitination levels and upregulated PRKAA1, a downstream target of USP45, suggesting that TET2 may promote PRKAA1 deubiquitination through USP45 and thereby enhance granulosa cell autophagy (Fig.  6 F). Furthermore, the effects of TET2 overexpression and USP45 knockdown on downstream proteins were analyzed. It was observed that mTOR activation is inhibited by TET2, thereby promoting the occurrence of pathological autophagy. Conversely, upon the knockdown of USP45, the inhibitory effect of PRKAA1 on mTOR was attenuated, resulting in increased mTOR phosphorylation and a partial suppression of autophagic flux (Fig.  6 G). It is further suggested by our study that cellular autophagy may be promoted by TET2 via the USP45/PRKAA1 axis, leading to further impairment of ovarian function. Fig. 5 USP45 is a downstream regulatory target of TET2. A : Venn diagram showing eight co-expressed genes identified from the results of the 5hmC array and TET2 ChIP-seq data. B : UCSC genome browser view (ChIP-seq signal tracks) showing TET2 binding enrichment at the USP45 gene locus. C : Two 5hmC site-specific primers were designed based on the promoter region of USP45, amplified using an hMeDIP-PCR kit, and detected via agarose gel electrophoresis. D : hMeDIP-qPCR results showing enrichment at the USP45 promoter. ( n  = 3, * P  < 0.05). E : The methylation status of the USP45 promoter was assessed via methylation-specific PCR (MSP). Both Control and PCOS groups exhibited prominent bands in the unmethylated (U) lanes and an absence of bands in the methylated (M) lanes, suggesting that the USP45 promoter is primarily hypomethylated in granulosa cells. F : Bar plots showing the 5hmC levels of USP45 in granulosa cells from both groups. 5hmC levels were quantified by hydroxymethylated DNA immunoprecipitation followed by q-PCR. ( n  = 8, ** P  < 0.01). G : The mRNA expression of USP45 was significantly upregulated in granulosa cells from PCOS patients compared to the control group. ( n  = 8, ** P  < 0.01) Fig. 6 USP45 enhances PRKAA1 expression by promoting its deubiquitination. A : Western blot and RT-qPCRconfirmed the upregulation of USP45 in the overexpression group (USP45-OE). Both mRNA and protein levels of USP45 were significantly increased compared to the control group ( n  = 3, ** P  < 0.01). B : Western blot and RT-qPCR confirmed the downregulation of USP45 in the si-USP45 group (Si-USP45). Both mRNA and protein levels of USP45 were significantly decreased compared to the control group ( n  = 3, ** P  < 0.01). C : Protein levels of USP45 and PRKAA1 were balanced across different treatment groups, as determined by Western blot analysis. Knockdown of USP45 (Si-USP45) led to a significant decrease in PRKAA1 protein levels, whereas overexpression of USP45 (USP45-OE) upregulated PRKAA1 expression. ( n  = 3, * P  < 0.05). D , E : Western blot and RT-qPCR were employed to assess the mRNA and protein expression of USP45 following TET2 overexpression ( n  = 3, ** P  < 0.01). F : Western blot analysis of USP45, PRKAA1, and total intracellular ubiquitination levels in KGN cells after TET2 overexpression and treatment with the TET2 inhibitor Bobcat339. G : Protein levels of USP45, PRKAA1, p-mTOR, P62 and LC-3 were balanced across different treatment groups, as determined by Western blot analysis. ( n  = 3, * P  < 0.05, ** P  < 0.01) USP45 is a downstream regulatory target of TET2. A : Venn diagram showing eight co-expressed genes identified from the results of the 5hmC array and TET2 ChIP-seq data. B : UCSC genome browser view (ChIP-seq signal tracks) showing TET2 binding enrichment at the USP45 gene locus. C : Two 5hmC site-specific primers were designed based on the promoter region of USP45, amplified using an hMeDIP-PCR kit, and detected via agarose gel electrophoresis. D : hMeDIP-qPCR results showing enrichment at the USP45 promoter. ( n  = 3, * P  < 0.05). E : The methylation status of the USP45 promoter was assessed via methylation-specific PCR (MSP). Both Control and PCOS groups exhibited prominent bands in the unmethylated (U) lanes and an absence of bands in the methylated (M) lanes, suggesting that the USP45 promoter is primarily hypomethylated in granulosa cells. F : Bar plots showing the 5hmC levels of USP45 in granulosa cells from both groups. 5hmC levels were quantified by hydroxymethylated DNA immunoprecipitation followed by q-PCR. ( n  = 8, ** P  < 0.01). G : The mRNA expression of USP45 was significantly upregulated in granulosa cells from PCOS patients compared to the control group. ( n  = 8, ** P  < 0.01) USP45 enhances PRKAA1 expression by promoting its deubiquitination. A : Western blot and RT-qPCRconfirmed the upregulation of USP45 in the overexpression group (USP45-OE). Both mRNA and protein levels of USP45 were significantly increased compared to the control group ( n  = 3, ** P  < 0.01). B : Western blot and RT-qPCR confirmed the downregulation of USP45 in the si-USP45 group (Si-USP45). Both mRNA and protein levels of USP45 were significantly decreased compared to the control group ( n  = 3, ** P  < 0.01). C : Protein levels of USP45 and PRKAA1 were balanced across different treatment groups, as determined by Western blot analysis. Knockdown of USP45 (Si-USP45) led to a significant decrease in PRKAA1 protein levels, whereas overexpression of USP45 (USP45-OE) upregulated PRKAA1 expression. ( n  = 3, * P  < 0.05). D , E : Western blot and RT-qPCR were employed to assess the mRNA and protein expression of USP45 following TET2 overexpression ( n  = 3, ** P  < 0.01). F : Western blot analysis of USP45, PRKAA1, and total intracellular ubiquitination levels in KGN cells after TET2 overexpression and treatment with the TET2 inhibitor Bobcat339. G : Protein levels of USP45, PRKAA1, p-mTOR, P62 and LC-3 were balanced across different treatment groups, as determined by Western blot analysis. ( n  = 3, * P  < 0.05, ** P  < 0.01)

A total of 40 women participated in this study, who underwent in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI)treatment after receiving an ovulation induction protocol. A GnRH antagonist protocol was employed for ovulation induction in all patients. In this study, participants had no history of diseases such as severe endometriosis, thyroid dysfunction disease or severe comorbidities (including insulin-dependent diabetes, non-insulin-dependent diabetes or gastrointestinal, cardiovascular, pulmonary, liver or kidney diseases). The Rotterdam criteria was used to diagnose PCOS. The study excluded individuals with congenital adrenal hyperplasia, androgen-secreting tumors, Cushing's syndrome, thyroid disease, or hyperprolactinemia. Healthy controls, characterized by regular menstrual cycles, normal steroid hormone levels, and normal ovarian morphology (antral follicle count, AFC < 12), were women undergoing in vitro fertilization for tubal factor and/or male factor infertility. All participants (patients and controls) had a washout period of at least 3 months without hormone therapy before study inclusion. Ovarian granulosa cells were collected from patients diagnosed with PCOS and control patients at the Reproductive Medicine Center of the Affiliated Hospital of Nanjing University of Chinese Medicine. All participants underwent a standard GnRH antagonist stimulation protocol, followed by ovulation induction using human chorionic gonadotropin (hCG), a GnRH agonist, or a dual trigger. Following granulosa cell collection, the cells were fixed with 4% paraformaldehyde and washed three times with PBS. After permeabilization with Triton X-100, the cells were again washed three times with PBS. Samples were then blocked with an immunofluorescence blocking solution for 1 h, followed by incubation with a specific 5hmC antibody (1:500, Active Motif) overnight. After three washes with PBS, fluorescently labeled rabbit secondary antibodies were applied. Following another three PBS washes, an antifade mounting medium was added, and fluorescence imaging was performed. Granulosa cells were collected and stored at –80 °C. To minimize inter-array variability, all samples were processed simultaneously. Genomic DNA was extracted using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions.A total of 500 ng to 1 μg of genomic DNA was sheared to approximately 300 bp fragments using a Covaris S220 ultrasonicator. For glucosylation, 1 μg of fragmented gDNA was incubated at 37 °C for 1 h with uridine diphosphate glucose (UDP-glucose) and T4 β-glucosyltransferase (T4-βGT, NEB, #M0357S). The reaction products were then denatured at 85 °C for 10 min using formamide and rapidly cooled on ice. The denatured DNA was subjected to deamination by incubation with the APOBEC enzyme (NEB, #E7120S) from the Next Enzymatic Methyl-seq Kit at 37 °C for 3 h. The resulting DNA was then analyzed using a DNA methylation microarray, such as the Illumina Infinium MethylationEPIC BeadChip (850 K array), following the standard protocol provided by the manufacturer. For non-paired samples, the champ DMP function in the ChAMP package can be used for differential methylation analysis of DMPs. For the p-values obtained from the analysis, the Benjamini & Hochberg method can be applied for multiple hypothesis testing to reduce false positives. KGN cells were cultured in high-glucose DMEM/F12 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Cells were maintained in a humidified incubator at 37 °C with 5% CO₂. Small interfering RNA (siRNA) was employed to knockdown TET2 and USP45 in KGN cells. The specific siRNA sequences were designed and synthesized by Genomeditech (Shanghai, China); detailed sequences are provided in Supplementary Table  1 . Cells were transfected with the siRNA using HiPerFect Transfection Reagent (QIAGEN, Germany) according to the manufacturer’s instructions, followed by a 24-h incubation. TET2 overexpression plasmids were purchased from BIOWORLD (Nanjing, China), and USP45 overexpression plasmids were obtained from Genomeditech (Shanghai, China). Cells were transfected with the respective plasmids using Lipofectamine 3000 Transfection Reagent (Invitrogen, USA) according to the manufacturer’s protocol, followed by a 24-h incubation. Cells were collected and lysed using RIPA lysis buffer (Beyotime, China) supplemented with protease inhibitors. Protein concentration was determined using the BCA assay. Proteins were separated by SDS-PAGE and transferred onto PVDF membranes at 400 mA for 40 min. The membranes were blocked at room temperature for 30 min using a rapid blocking solution. After three washes with TBST, membranes were incubated with primary antibodies overnight at 4 °C. Following primary antibody incubation, membranes were washed three times with TBST and incubated with HRP-conjugated secondary antibodies at room temperature for 1 h. Protein bands were visualized using chemiluminescent substrate and imaged. The primary and secondary antibodies used were: TET1(Affinity Biosciences, China, 1:1000), TET2(Proteintech, China, 1:1000), TET3(Signal antibody, China, 1:1000), β-Tubulin(Proteintech, China, 1:5000), ATG7(Proteintech, China, 1:1000), Beclin(Proteintech, China, 1:1000), LC I/II(Proteintech, China, 1:1000), P62/SQSTM1(Proteintech, China, 1:1000), USP45(Affinity Biosciences, China, 1:1000), PRKAA1(Proteintech, China, 1:1000), mTOR(Proteintech, China, 1:1000), p-mTOR(Proteintech, China, 1:1000), UB(Affinity Biosciences, China, 1:1000). Agarose gel electrophoresis was performed to analyze PCR products. A 1% pre-cast agarose gel (Beyotime, China) was placed in the electrophoresis chamber and covered with TEA buffer. For each sample, 5 μL of PCR product were mixed with 1 μL of 6 × loading buffer via gentle vortexing. The resulting mixtures were then carefully loaded into the gel wells using a micropipette, taking care not to damage the gel. Electrophoresis was carried out at a constant voltage of 100 V for 45 min. Following electrophoresis, DNA bands were visualized under UV illumination and digitally captured. This experiment was conducted using the Hydroxymethylated DNA Immunoprecipitation (hMeDIP) Kit from Active Motif (Catalog No. 55010, USA), following the manufacturer's instructions. Genomic DNA obtained using the kit method was divided into two groups: control and TET2-overexpression, with 500 ng of DNA used per sample. For each group, one sample was designated as the immunoprecipitation (IP) group and the other as the IgG negative control.Each reaction tube was prepared by adding 10 μL of Buffer C, 1 μL of PIC solution, 4 μL of 5-hmC antibody (for IP group) or rabbit IgG antibody (for IgG negative control), and sterile water to bring the total volume to 100 μL. Samples were thoroughly mixed in 8-strip PCR tubes and incubated overnight at 4 °C on a rotator. Subsequently, 25 μL of Protein G magnetic beads were added to each tube, followed by 2 h of incubation at 4 °C with rotation. Beads were collected using a magnetic stand, and the supernatant was carefully removed. The beads were washed three times with 200 μL of pre-chilled Buffer C by gentle inversion, followed by three additional washes with 200 μL of pre-chilled Buffer D. After washing, 50 μL of Elution Buffer AM2 was added to each tube, mixed thoroughly, and incubated at 4 °C for 15 min with rotation. Tubes were briefly centrifuged, and beads were magnetically separated. The supernatant was carefully transferred to a new 1.5 mL microcentrifuge tube and mixed with 50 μL of Neutralization Buffer. For DNA purification, 2 μL of proteinase K (0.5 mg/mL) was added to each tube, followed by incubation at 50 °C for 30 min and then at 80 °C for 10 min. The purified DNA was used for subsequent RT-qPCR analysis. Input DNA was diluted to concentrations of 10, 1, 0.1, and 0.01 ng/μL for generating a standard curve, using the appropriate primer sequences. A standard curve was generated by plotting the log₁₀(Input DNA concentration) on the x-axis and corresponding CT values on the y-axis. A standard equation was derived from the curve, and CT values from experimental samples were used to calculate the log-transformed DNA concentration, which was then converted to actual concentration. Fold enrichment was calculated as: Fold enrichment = Amount of enriched sample DNA (ng) / Amount of enriched IgG DNA (ng). DNA bisulfite conversion was performed using the BeyoMag™ DNA Bisulfite Conversion Kit (Beyotime, China) according to the manufacturer’s instructions. Briefly, the conversion reagent was prepared by dissolving the conversion powder in Protection Solution. A total of 2 ug of genomic DNA was mixed with 130 uL of the conversion solution. The mixture underwent thermal cycling in a PCR system, consisting of initial denaturation at 95 °C followed by multiple cycles of denaturation at 95 °C and conversion at 70 °C to achieve chemical modification. Subsequently, the converted DNA was incubated with BeyoMag magnetic beads for 5–15 min for binding. After an initial wash, desulfonation was carried out using a freshly prepared desulfonation solution at room temperature for exactly 20 min. Following two additional washes and air-drying of the beads, the purified DNA was recovered in 20 ul of Elution Buffer. The converted DNA samples were stored at -20 °C or -80 °C for subsequent analysis. Mitochondrial health in KGN cells was assessed by measuring membrane potential changes with JC-1 staining (Beyotime, China). Briefly, cells were treated with 10 μm JC-1 for 20 min at 37 °C in the dark, followed by two rinses with DPBS. Images were captured using a fluorescence microscope (Olympus, Japan). The probe functions as a dual-emission sensor where red fluorescence (aggregates) indicates polarized mitochondria and green fluorescence (monomers) denotes depolarization. Fluorescence quantification was executed through ImageJ analysis. To determine relative differences in mitochondrial polarization, the ratio of red to green fluorescence intensity was calculated for each group. The production of reactive oxygen species (ROS) in granulosa cells was monitored using H2DCFDA (Sigma, USA). After incubating the cells with the probe (5uM) in a dark environment at 37 °C for 30 min, they were collected via 0.05% trypsin–EDTA digestion. The resulting cell suspension was immediately analyzed using a flow cytometer to quantify fluorescence intensity. Continuous experimental data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 7.00. Comparisons between two groups were conducted using Student’s t-tests (independent or paired, as appropriate). Comparisons between three or more groups were performed using one-way analysis of variance (ANOVA), followed by a post hoc test if the ANOVA yielded a significant result. Data normality was assessed using the Kolmogorov–Smirnov test. For normally distributed data, Tukey’s honestly significant difference (HSD) test was used for post hoc multiple comparisons following a significant ANOVA. For non-normally distributed data, the Kruskal–Wallis test was conducted for comparisons across multiple groups, followed by Dunn’s post hoc test for pairwise comparisons. Correlations between categorical variables were evaluated using Pearson’s chi-square test. A P-value of less than 0.05 was considered statistically significant for all analyses.

This study provides evidence for a potential role of 5hmC modifications in PCOS pathogenesis. We observed widespread 5hmC modifications and increased TET1/2 expression in granulosa cells of PCOS patients. Importantly, we found that elevated TET2 expression triggers autophagy, a process implicated in PCOS-related follicular developmental disorders. Our findings further demonstrated that TET2-mediated 5hmC modification of USP45 promotes autophagy. Specifically, USP45 targets PRKAA1 to inhibit mTOR phosphorylation, resulting in an excessive autophagic flux that contributes to the pathological autophagic process in PCOS. These findings suggest that aberrant 5hmC modifications, potentially regulated by TET2, contribute to PCOS and identify TET2 as a potential target for future therapeutic strategies.

PCOS is a common reproductive disorder marked by impaired folliculogenesis and ovulatory dysfunction. Given its complex pathogenesis, the role of epigenetic modifications has become a focal point of research [ 29 , 30 ]. In this study, we sought to elucidate the functional significance and underlying mechanisms of 5hmC in PCOS progression. By employing APOBEC-mediated deamination-based 5hmC profiling in conjunction with the Illumina Infinium MethylationEPIC (850 K) BeadChip, we characterized the 5hmC landscape in ovarian granulosa cells. Additionally, integrated with functional in vitro assays, our findings further elucidate the critical importance of epigenetic modifications in driving PCOS pathogenesis. Current research has predominantly focused on DNA methylation, leaving the potential role of DNA 5hmC largely unexplored. Early research indicated that there is no significant difference in global DNA methylation levels in peripheral blood between PCOS [ 31 ]. Due to the scarcity and limited accessibility of ovarian tissue from patients with PCOS, only a few studies have reported the DNA methylation landscape of human ovarian tissue. Compared to controls, PCOS patients exhibit significantly higher methylation levels at CpG shores (regions 1–2 kb from CpG islands) and promoters with high CpG content outside of core promoter regions, while showing lower methylation levels within gene bodies. Additionally, 342 genes were identified as having specific methylation changes, and these genes are functionally enriched in categories such as protein binding, hormone activity, and transcription factor activity [ 32 ]. Some studies have also shown that the overall DNA methylation level in ovarian granulosa cells of PCOS patients is reduced, with the first intron being the primary genomic region exhibiting hypomethylation in PCOS [ 33 ]. Targeted analysis of DNA methylation at genomic loci—including key genes—in granulosa cells has identified hypomethylation of promoters such as YAP1, LHCGR, NCOR1, and HDAC3, which are associated with ovarian pathology; this hypomethylation may be linked to granulosa cell proliferation, LH receptor overexpression, and dysregulation of hormonal signaling [ 34 ]. Another study employing high-throughput next-generation sulfite sequencing identified substantial alterations in DNA methylation, characterized by hypomethylation at 977 CpG sites affecting 2063 genes and hypermethylation at 2509 CpG sites affecting 1777 genes. Pathway analysis indicated that these differentially methylated genes are linked to ovarian morphology, function, and hormonal regulation [ 35 ]. Our study found that the overall 5hmC modification level of DNA in ovarian granulosa cells from PCOS patients is increased, with differential modification sites located at TSS1500 and 5’UTR regions. The TSS1500 is primarily the promoter region of genes, where transcription factors bind, and it is a key site for DNA methylation modifications. We performed enrichment analysis of differentially modified genes with GWAS genes and found an overlap with four genes: ASIC2, CCDC91, CHEK2, and FTO. CHEK2 is a key serine/threonine protein kinase that can promote DNA repair or initiation by activating the downstream ATM pathway [ 36 , 37 ]. A recent large-scale GWAS study also reported an association between CHEK2 and genetic variations in PCOS [ 38 ]. ASIC2 primarily encodes a subunit of acid-sensing ion channels, which mediate membrane depolarization and intracellular signal transduction through the influx of sodium and calcium ions. It may be associated with insulin resistance in PCOS [ 39 ]. CCDC91 (Coiled-Coil Domain Containing 91) is a gene encoding a protein with a coiled-coil domain, primarily involved in intracellular transport and metabolic regulation [ 40 ]. To date, no studies have reported a direct association between CCDC91 and PCOS. However, considering that granulosa cells in PCOS patients exhibit abnormally increased autophagic activity, leading to mitochondrial dysfunction and metabolic disturbances, abnormalities in CCDC91-mediated Golgi-to-lysosome transport may impair lysosomal maturation or the degradation efficiency of autophagosomes, thereby exacerbating metabolic dysregulation in PCOS, FTO is an RNA N6-methyladenosine (m6A) demethylase that catalyzes the removal of methyl groups from m6A sites on mRNA and lncRNA, thereby influencing RNA stability, translational efficiency, and degradation. Recent studies have reported a strong association between FTO and PCOS. In our study, we constructed a predictive model for polycystic ovary syndrome using four probes, which demonstrated strong predictive value (AUC = 0.952). As the cohort was restricted to Han Chinese individuals from a single center, the role of TET2 might vary across different ethnic groups due to population-specific epigenetic landscapes. Additionally, given the modest sample size, future multi-ethnic and multi-center studies are essential to validate the clinical applicability of our findings. To further investigate the role of TET enzymes in granulosa cells, we analyzed the expression profiles of the TET enzyme family in PCOS and found that TET2 exhibited the most significant difference between the two groups. TET2, a dioxygenase, regulates gene expression by oxidizing 5mCin DNA to generate 5hmC. In contrast to the CXXC domain-containing TET1/3, TET2 recruitment to promoters is mediated indirectly via interactions with transcription factors. This distinct regulatory mechanism bypasses the off-target effects typical of direct enzymatic inhibition. Consequently, disrupting the specific recruitment of TET2 by transcription factors offers a more refined and safer strategy for targeted therapeutic intervention. In ovarian granulosa cells of aged mice, the conversion of 5mC to 5hmC is reduced, accompanied by decreased expression of TET2 [ 24 ]. By generating TET1 and TET2 double-knockout mice, it was found that the knockout resulted in decreased levels of 5hmC and increased expression of 5mC, along with aberrant methylation observed in multiple imprinted genes. Moreover, female knockout mice exhibited reduced ovarian size and impaired fertility, suggesting that TET2 plays a critical role in ovarian function [ 41 ]. Recent studies have highlighted the role of TET2 in promoting autophagy through various pathways, such as suppressing BCL2 expression or inhibiting mTORC1 signaling via mRNA oxidation [ 42 ]. Autophagy is a fundamental catabolic process essential for maintaining cellular homeostasis; however, its dysregulation is increasingly recognized as a hallmark of PCOS [ 43 , 44 ]. In PCOS, granulosa cells frequently exhibit aberrant autophagic activation, which leads to mitochondrial dysfunction, increased apoptosis, and impaired follicular development [ 45 ]. High androgen levels further exacerbate this process, contributing to insulin resistance and endocrine imbalances [ 46 ]. While it is known that Metformin can alleviate PCOS symptoms by modulating the PI3K/AKT/mTOR-mediated autophagy pathway [ 47 ], the precise epigenetic mechanisms driving autophagic dysfunction remain poorly understood. In this study, we employed genome-wide TET2 ChIP-seq and functional assays to investigate its impact on GC proliferation and autophagy. Our findings reveal that TET2 facilitates the 5hmC modification of the USP45 promoter, thereby enhancing its expression. USP45, a deubiquitinating enzyme belonging to the ubiquitin-specific protease family, can specifically remove ubiquitin molecules from ubiquitinated proteins by recognizing and binding to them, thereby reversing their ubiquitination. USP45 maintains intracellular protein homeostasis by regulating the stability and activation of target proteins. In serous ovarian cancer cells, USP45 has been found to promote cell proliferation by deubiquitinating the SNAIL protein, thereby preventing its degradation. We queried the BioGRID database and identified PRKAA1 as a potential downstream target of USP45 [ 48 , 49 ]. Upon activation, PRKAA1 can inhibit the activity of mTORC1, thereby relieving its suppression of the ULK1 complex. This primarily promotes the initiation of autophagy and the early stages of autophagosome formation, rather than directly inducing the expression of autophagy-related genes such as LC3 and ATG7, which are often regulated by other cellular stresses and transcription factors. The activation of ULK1 subsequently accelerates the autophagic flux process [ 50 , 51 ]. Our study found that TET2 can promote the 5hmC modification of the USP45 promoter, thereby enhancing its expression. USP45 regulates the ubiquitination of the downstream target PRKAA1, inhibiting its degradation and promoting the sustained activation of the autophagy pathway. This suggests that epigenetic modifications may exacerbate androgen-induced autophagic responses.

Polycystic ovary syndrome is an endocrine and metabolic disorder characterized by polycystic ovarian morphology, menstrual irregularities, and anovulation [ 1 ]. The estimated global prevalence of PCOS among women of reproductive age ranges from approximately 5% to 15%, with significant variations observed across different ethnic groups and geographic regions. In recent years, the incidence of PCOS has been continuously increasing [ 2 ]. However, the etiology of PCOS remains not fully elucidated. Its onset is likely driven by a multifactorial interplay between genetic predisposition and external triggers, including endocrine-disrupting chemicals (EDCs), chronic systemic inflammation, environmental pollutants, and other emerging factors [ 3 – 5 ]. PCOS impairs folliculogenesis, resulting in ovulatory dysfunction characterized by irregular or absent ovulation. These reproductive aberrations significantly compromise fertility, positioning PCOS as a primary cause of anovulatory infertility in women of reproductive age [ 6 ]. Consequently, further elucidation of the pathophysiological mechanisms of PCOS and the development of targeted therapeutic interventions are imperative. The primary ovarian features in PCOS include an increased number of small antral follicles and disrupted follicular development, often characterized by arrested growth at the pre-antral or early antral stage [ 7 , 8 ]. Research indicates a significantly elevated antral follicle count in the ovaries of women with PCOS (generally defined as 12 or more follicles with diameters of 2–9 mm), with a large proportion of these follicles remaining at an early developmental stage [ 9 ]. Due to granulosa cell dysfunction, these follicles fail to complete the final stages of maturation, resulting in a significantly increased rate of follicular atresia [ 10 , 11 ]. Additionally, follicular development in PCOS patients often arrests at the pre-antral stage, preventing proper transition to the antral follicle stage. Hyperandrogenemia further contributes to this developmental arrest by inhibiting aromatase (CYP19A1) activity in granulosa cells, leading to decreased estrogen production [ 12 , 13 ]. Consequently, understanding the mechanisms that drive abnormal follicular development is crucial for the effective treatment of PCOS. Autophagy is a physiological process through which cells degrade damaged proteins or excess components via the lysosomal pathway, playing a crucial role in maintaining cellular homeostasis [ 14 ]. Autophagy serves as a vital quality-control mechanism that maintains proteostasis by eliminating dysfunctional mitochondria and protein aggregates. However, when pathologically sustained, excessive autophagic flux can compromise cellular integrity and overwhelm compensatory pathways. Consequently, non-selective organelle degradation occurs, leading to the attrition of cellular structure and viability [ 15 ]. Moreover, excessive autophagy may lead to substantial intracellular ATP consumption, resulting in energy depletion and subsequently impairing normal cellular metabolic functions [ 16 ]. The emerging understanding from recent studies emphasizes the critical role of autophagy as a potential key player in both the onset and the progression of PCOS. This recognition is significant as it suggests that modulating autophagic pathways could represent a novel avenue for therapeutic intervention in managing and potentially preventing this complex disorder [ 14 ]. The dysregulation of the autophagic pathway within PCOS granulosa cells triggers excessive cellular self-digestion, contributing to impaired folliculogenesis and premature follicle loss [ 17 ]. Thus, effectively regulating autophagy levels within granulosa cells could be a valuable novel therapeutic approach and a key theoretical foundation for PCOS intervention and treatment. Recent studies have unequivocally demonstrated the involvement of a range of epigenetic mechanisms, including DNA methylation, histone modifications, and non-coding RNAs (ncRNAs) are dysregulated in women with PCOS and contribute to its key features [ 18 ]. Among these epigenetic modifications, 5hmC has emerged as a pivotal DNA modification integral to the facilitation of gene transcription, the modulation of chromatin structural dynamics, and its dynamic regulation during cellular differentiation [ 19 , 20 ]. The formation of 5hmC is primarily catalyzed by the TET family of dioxygenases, comprising TET1, TET2, and TET3, each exhibiting distinct expression profiles across different tissue types [ 21 ]. Advancements in high-throughput sequencing technologies have facilitated the identification of 5hmC as a key regulatory epigenetic mark in both physiological and pathological contexts [ 22 ]. Accumulating evidence suggests that granulosa cells from PCOS patients exhibit elevated levels of 5hmC, accompanied by increased expression of TET enzymes [ 23 ]. Animal model studies further indicate that TET1 expression is downregulated in aging mice, and its overexpression promotes granulosa cell proliferation while inhibiting apoptosis [ 24 ]. However, the functional contributions of TET enzymes and their resultant epigenetic targets in PCOS granulosa cells remain largely undefined, underscoring the need for a more comprehensive investigation. This research investigates the mechanistic involvement of 5hmC in the etiology and clinical advancement of PCOS. Through the integration of APOBEC-assisted 5hmC profiling with the Illumina Infinium HumanMethylation 850 K array, a significant elevation of 5hmC was identified at multiple genomic locations in PCOS ovarian granulosa cells. Genome-wide target gene analysis via TET2 ChIP-seq was conducted to decipher its regulatory functions. Furthermore, the effects of TET2 on the proliferation and autophagy of KGN cells were systematically evaluated. Based on these integrated findings, the pivotal involvement of 5hmC modification in PCOS pathogenesis is highlighted, and TET2 is proposed as a potential therapeutic avenue, suggesting that a novel approach for PCOS intervention could be offered by targeting its activity.

Below is the link to the electronic supplementary material. Supplementary file1 (XLSX 10 KB) Supplementary file2 (DOCX 295 KB) Supplementary file1 (XLSX 10 KB) Supplementary file2 (DOCX 295 KB)