WWP2 underlies ROS-induced granulosa cell apoptosis by promoting ubiquitination of BAK in polycystic ovary syndrome

Abstract

Granulosa cell (GC) apoptosis is intrinsically linked to the ovarian dysfunction of polycystic ovary syndrome (PCOS). Although oxidative stress and apoptosis in GCs have been detected in PCOS patients, how reactive oxygen species (ROS) links to GC apoptosis in PCOS remains to be further elucidated. Here, by integrating public single-cell RNA-seq data with clinical GC sample validation, we found that the expression of the E3 ubiquitin ligase WWP2 was significantly reduced, whereas its role in PCOS has not been previously reported. Notably, we first demonstrated that WWP2 overexpression can effectively antagonize mitochondrial apoptosis and ROS in KGNs. Mechanistically, oxidative stress weakened the interaction between WWP2 and BAK and reduced WWP2 expression, thereby suppressing BAK ubiquitination at Lys113. This inhibition impaired proteasomal degradation and consequently increased BAK protein levels. Consistently, disrupting BAK ubiquitination (BAK-K113R mutant) or knocking down WWP2 facilitated KGN apoptosis, and genetic ablation of Wwp2 in PCOS mice further aggravated GC apoptosis and hormonal disturbances. This study elucidates the molecular mechanism by which oxidative stress modulates GC mitochondrial apoptosis through WWP2-mediated BAK ubiquitination, and establishes WWP2 as a potential therapeutic target for PCOS.

Introduction

Polycystic ovary syndrome (PCOS) is a common endocrine and metabolic disorder [1] characterized by menstrual disturbances, androgen excess, insulin resistance, oxidative imbalance, and disordered ovulation, accounting for 80% of women with anovulatory infertility [2–4]. In the ovaries of PCOS patients, the development of small ovarian follicles is arrested at the sinus follicular stage, preventing the formation of dominant follicles and leading to persistent anovulation, a process largely mediated by oxidative stress and abnormal apoptosis of granulosa cells (GCs) [5]. Current treatments, such as combined oral contraceptives (COCs), can regulate androgen levels and menstrual cycles but do not restore fertility or improve polycystic ovarian morphology [6]. Moreover, due to the reduction in oocyte quality, in vitro fertilization (IVF) often exhibits lower fertilization, cleavage, implantation rates, and a higher miscarriage rate in PCOS patients [7]. Given that COCs and IVF cannot effectively improve the intrinsic ovarian abnormality caused by GC apoptosis [8–10], exploring critical molecular mechanisms of GC apoptosis may offer new treatment strategies for PCOS.

Mitochondrial apoptosis is a tightly regulated signaling cascade that can be activated by excessive reactive oxygen species (ROS)-mediated mitochondrial dysfunction [11]. In response to stress stimuli, B-cell lymphoma (BCL)-2 homologous antagonist/killer (BAK) cooperates with BCL-2-associated X protein (BAX) to permeabilize the mitochondrial outer membrane (MOM), thereby resulting in cytochrome c (Cyt C) efflux. This efflux initiates apoptosome assembly, triggers caspase 9 activation, and drives subsequent apoptotic cascades [12–14]. Upstream of these events, the activation of BAK and BAX is initiated by BCL-2 homology 3 (BH3)-only proteins, such as BH3-interacting domain death agonist (BID) and BCL-2 interacting mediator of cell death (BIM), which transiently bind to the canonical hydrophobic groove of BAK and BAX to facilitate their oligomerization. In contrast, prosurvival BCL-2 family members, including BCL2, BCL-xL, and myeloid cell leukemia 1 (MCL1), suppress apoptosis by either directly binding to BAK/BAX to block their oligomerization or sequestering BH3 activators into inactive complexes [15]. Beyond the BCL-2 family, emerging evidence suggests that additional proteins also serve as important regulators of mitochondrial apoptosis. For example, E3 ubiquitin ligases Parkin and MARCHF5 can regulate the apoptotic activity of BAK [16, 17].

Although mitochondrial apoptosis is tightly regulated by multiple BCL-2 family proteins, BAK functions as a central executioner, and its modulation can profoundly impact disease progression. For instance, in age-related hearing loss, ROS mediates BAK upregulation and genetic ablation of BAK reduces cochlear cell apoptosis, highlighting its pivotal role in ROS-driven apoptosis [18]. In PCOS, GCs exhibit mitochondrial dysfunction and ROS-induced apoptosis [5]. However, it remains unknown whether ROS similarly elevates BAK levels in GCs and how the ROS-BAK signaling axis is regulated in PCOS pathogenesis.

Here, by using publicly available human single-cell RNA-seq analysis (scRNA-Seq) data and ovarian GCs of PCOS patients and healthy controls, we identified that differentially expressed genes (DEGs) were enriched in ROS, apoptosis, and the ubiquitin-proteasome pathway (UPP). WW domain-containing E3 ubiquitin protein ligase 2 (WWP2) mRNA and protein levels were significantly downregulated in PCOS patients with oxidative stress, and low WWP2 expression predicted a decline in fertility. Elevated BAK protein levels were observed in PCOS patients, whereas mRNA expression remained unchanged. Importantly, we first identified WWP2 as a novel E3 ligase of BAK, promoting its ubiquitination at Lys113 (K113) and subsequent proteasomal degradation. Disrupting this process through the BAK-K113R mutant or WWP2 knockdown aggravated KGN apoptosis. Consistently, Wwp2 deletion in PCOS mice significantly exacerbated GC apoptosis and reproductive dysfunction. Overall, our findings reveal for the first time that WWP2 counteracts GC apoptosis in PCOS, highlighting it as a potential therapeutic target.

Materials and methods

Subjects and human samples

This research was approved by the Ethics Committee of Shengjing Hospital, China Medical University, Shenyang (Approval No. SZCS2025016). Ovarian GCs were obtained with informed consent from IVF participants at the Center of Reproductive Medicine. PCOS diagnosis adhered to the Rotterdam criteria, and GCs were isolated from follicular fluid through density gradient centrifugation. The fluid was centrifuged at 1500 rpm for 15 min at room temperature (RT) to collect the cell pellets. Following centrifugation, the harvested pellets were resuspended in phosphate-buffered saline (PBS) containing 0.1% hyaluronidase. Then, GCs were purified by using lymphocyte separation medium and red blood cell lysis buffer for the next study.

Animal treatment

Wwp2 knockout (KO) mice were constructed by the Shanghai Model Organisms Center. Female Wwp2-WT and Wwp2-KO mice were each randomly assigned to PCOS or control groups (total n = 48, 12 mice per group). PCOS mouse models were established by daily subcutaneous injection of dehydroepiandrosterone (DHEA, 6 mg per 100 g body weight) combined with a 60% high-fat diet (HFD) for 21 consecutive days in 3-week-old female mice [19]. Following model establishment, a glucose tolerance test was conducted. After 12 h fasting, baseline glycemia was assessed via tail vein blood sampling using the Accu-Chek Performa (Roche, USA). Following intraperitoneal administration of D-glucose at a dose of 2 g/kg, blood glucose levels were monitored at designated time points after injection. All experimental protocols received approval from the Animal Ethics Committee of China Medical University (License number CMU2023296).

Quantitative real-time PCR (qPCR)

Total RNA (500 ng) was reverse transcribed into cDNA in a 20 μL reaction volume using the PrimeScript RT Kit. The qPCR was performed on the Gentier 96 system utilizing TB Green Premix Ex Taq II, with primer sequences listed in Supplementary Table 1.

Western blotting and immunoprecipitation (IP)

Protein extraction was performed using lysis buffer (137 mM NaCl, 10 mM NaF, 50 mM Tris-HCl, pH 7.6, 1 mM ethylenediaminetetraacetic acid (EDTA), 0.1 mM sodium orthovanadate, 10% glycerol, 1% NP-40, 1 mM protease inhibitor) [20] or ONYX lysis buffer (20 mM Tris-pH 7.4, 135 mM NaCl, 1.5 mM MgCl2, 1 mM EGTA, 10% (v/v) glycerol, supplemented with 1% digitonin, 0.5 μg/ml complete protease inhibitor) [17]. For Co-IP and ubiquitination assay, KGN cells were lysed for 30 min on ice in ONYX lysis buffer. The mixture was centrifuged at 12,000 × g for 20 min at 4 °C, and protein concentrations were measured. Lysates were either directly incubated with Anti-Flag Affinity Gel overnight, or first incubated with specific antibodies (1 μg/mg lysate) for 3 h at 4 °C followed by an overnight incubation with Protein A/G Magnetic Beads. IP complexes were washed and eluted by boiling in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer. Proteins from lysates or IP eluates were separated on 8%, 10% or 12% SDS-PAGE gels, transferred to polyvinylidene difluoride membranes, blocked with 5% bovine serum albumin, and probed with primary antibodies overnight at 4 °C. After washing, membranes were incubated with Horseradish Peroxidase (HRP)-conjugated secondary antibodies and detected by chemiluminescence. Antibody information is provided in Supplementary Table 2.

Histology staining

To evaluate ovarian morphology, ovarian specimens underwent 4% paraformaldehyde fixation (48 h) followed by paraffin embedding and microtome sectioning (4–5 μm). Before staining, sections were first deparaffinized in xylene and rehydrated through a graded ethanol series (100%, 95%, 80%, 70%), followed by rinsing in PBS. Hematoxylin-eosin staining was conducted under optimized parameters: 7-min nuclear staining with hematoxylin followed by 2-min cytoplasmic counterstaining using eosin. Cystic follicles and corpus luteum were identified by histological morphology, and their numbers were counted.

Immunofluorescence staining and TUNEL staining of tissue

To assess apoptotic activity and Bak expression in ovarian tissues, antigen retrieval was carried out by boiling ovarian sections in Tris-EDTA buffer after dewaxing, then treated with 0.3% Triton X-100 (15 min) and washed with PBS. Apoptotic activity was quantitatively evaluated using the TUNEL assay (Terminal deoxynucleotidyl transferase dUTP Nick-End Labeling) with a commercial kit following the manufacturer’s protocol. Parallel sections were processed for immunofluorescence: blocking with Ready-to-Use Normal Goat Serum (1 h at RT), primary antibody incubation (4 °C overnight), Coralite 488-conjugated secondary antibody (1 h at RT), and nuclear counterstaining with 4’,6-Diamidino-2-Phenylindole (DAPI). Fluorescent signals were observed with a fluorescence microscope (Nikon, Japan), and images were acquired using NIS-Elements Viewer 5.21 software.

scRNA-seq

The scRNA-seq data from three control and three PCOS GC samples were obtained from the public database GSE240688. The dataset was analyzed using the Seurat v5 package in R. Quality control was performed based on the following criteria: (1) removal of cells with fewer than 200 or more than 7000 detected genes; (2) exclusion of cells in which mitochondrial gene expression exceeded 25% of total gene expression; and (3) removal of cells with unique molecular identifier (UMI) counts below 1000 as well as the top 3% of cells with the highest UMI counts. After filtering, a total of 24,566 cells were retained. Batch effects were corrected using IntegrateLayers with canonical correlation analysis. Integrated data underwent unsupervised clustering. t-SNE visualization revealed seven distinct GC subpopulations (GC0-GC6). DEGs for each subpopulation were identified via FindAllMarkers (expressed in >50% of cluster cells, logFC >0.3). Top 50 DEGs underwent KEGG enrichment, and top 5 DEGs per subpopulation were visualized with DoHeatmap. Focusing on hormone synthesis-related GC0, DEGs between groups were screened (logFC > 0.1, p < 0.01) and subjected to KEGG enrichment via the DAVID database. Enrichment pathway bar plots were generated via https://www.bioinformatics.com.cn/?keywords=pathway. Additional visualizations were performed using DotPlot (Seurat), VlnPlot (Seurat), and ggplot2.

Cell culture

KGN cell line (HyCyte™) was cultured in high-glucose Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. The cells were maintained at 37 °C in a humidified environment with 5% CO₂. All cell lines were authenticated by short tandem repeat profiling and confirmed to be contamination-free.

Generating BAK knockout KGN cells by CRISPR/Cas9 gene editing

BAK gene deletions were achieved using CRISPR-Cas9 via lentiviral transduction. sgRNAs targeting BAK (see Supplementary Table 1) were designed with the Benchling platform (https://benchling.com) for high specificity and on-target efficiency. Viral supernatants (36 h and 72 h post transfection) were gathered, centrifuged, and filtered to remove debris. KGN cells were infected with experimental virus with polybrene. After 48 h post infection, experimental cells were selected with 1 μg/mL puromycin. Surviving clones were expanded and validated as BAK KO KGN cells via molecular assays.

Plasmids and siRNAs construction

Full-length human BAK and its lysine-substituted mutants (BAK-K113R for human) were synthesized and cloned into Flag-tagged destination vectors by Sangon Biotech (Shanghai, China). Similarly, an HA-tagged expression plasmid encoding full-length human WWP2 was obtained from Genechem (Shanghai, China). siRNA targeting human WWP2 was purchased from RIBOBIO (see Supplementary Table 1).

Cell transfection and treatment

Plasmid transfection was conducted with Lipo8000 reagent following the optimized DNA-to-reagent ratio (1 μg/2.4 μL). siWWP2 was transfected with the jetPRIME transfection reagent at a siRNA/transfection reagent ratio of 20 pmol/μL. Cells were collected at 48 or 72 h after transfection, with a final 12 h treatment of 200 µM H₂O₂ applied prior to harvest. For protein stability and degradation assays, KGN cells were exposed to cycloheximide (CHX, 20 μM) and proteasome inhibitor MG132 (10 μM) for the indicated duration.

Mitochondrial extraction

To isolate cytoplasmic and mitochondrial fractions from KGN cells, 2 × 10⁷ KGN cells were resuspended in 1 ml STE buffer (250 mM sucrose, 10 mM Tris-HCl pH 7.2, 1 mM EDTA, 1×protease inhibitor cocktail). The suspension was homogenized by a Dounce homogenizer on ice and transferred to a centrifuge tube. 50 μL of liquid was saved as the total protein for the Western blot. The remaining homogenate was centrifuged (800 × g, 5 min, 4 °C). The supernatant was carefully transferred to a fresh tube and subjected to centrifugation at 15,000 × g for 15 min at 4 °C to isolate the cytoplasmic fraction. The resulting pellet, enriched in mitochondria, was resuspended in 1 mL of STE buffer and centrifuged again at 15,000 × g for 15 min at 4 °C. After removing the supernatant, the washing process was repeated to obtain mitochondria fraction.

Apoptosis analysis of KGN cells

KGN cell apoptosis was assessed using Annexin V/PI flow cytometry and TUNEL assay. For flow cytometry, KGN cells were stained with Annexin V-AbFluor™ 488 and PI in 1× binding buffer (5 × 10⁵ cells/ml), incubated for 15 min in the dark, and analyzed on a NovoCyte system (NovoExpress 1.5.0). For TUNEL staining, cells were fixed in 4% paraformaldehyde for 10 min and then permeabilized with 0.3% Triton X-100 for 15 min. The TdT-labeled reaction buffer was prepared following the manufacturer’s protocol, and cells were incubated at 37 °C in the dark for 60 min. DAPI was used for nuclear counterstaining, applied for 5 min in the absence of light. Images were then captured using a Nikon fluorescence microscope with NIS-Elements Viewer 5.21.

ROS assay

The ROS Assay Kit was used to detect intracellular ROS. KGN cells were washed with PBS three times, followed by incubation in 500 μL of serum-free medium supplemented with 10 μM 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA) at 37 °C for 30 min in the dark. After additional PBS washes, fluorescence imaging was performed using an FITC filter set.

ELISA assays

To evaluate serum hormone levels in mice, the concentrations of testosterone in mouse serum were measured by competitive enzyme-linked immunosorbent assay (ELISA), and the concentrations of follicle-stimulating hormone (FSH) and luteinizing hormone (LH) in mouse serum were detected by double-antibody Sandwich ELISA. Absorbance was read at 450 nm, and sample concentrations were calculated from standard curves.

Statistical analysis

Data are expressed as mean ± SEM or percentage values from ≥3 biological replicates, analyzed using GraphPad Prism 10. Variance homogeneity was assessed through an F-test (two-group comparisons) or Brown–Forsythe test (multi-group analyses). Two-group comparisons utilized Student’s t-test or Welch’s t-test. Multi-group comparisons employed one-way analysis of variance (ANOVA) with Tukey’s post hoc testing, and two-way ANOVA with Bonferroni’s multiple comparison post hoc test. Correlation analyses were performed using the nonparametric Spearman’s rank correlation method. The statistical significance threshold was established at p < 0.05.

Results

GC oxidative stress and apoptosis are associated with PCOS

Given that GCs are pivotal in maintaining normal ovulation [7], we isolated GCs from the follicular fluid of infertility patients undergoing IVF and found a marked reduction in superoxide dismutase 1 (SOD1) mRNA expression in PCOS patients, a key antioxidant enzyme critical for redox balance (Fig. 1A). SOD1 can effectively ameliorate ovarian dysfunction and pathological damage of ovarian tissues in PCOS through its antioxidant properties [21]. Spearman correlation further indicated that SOD1 levels in GCs negatively correlated with antral follicle count (AFC), anti-Müllerian hormone (AMH), and LH/FSH ratios, clinical markers linked to ovulatory dysfunction in PCOS (Figs. 1B, C and S1A). At the protein level, SOD1 was significantly downregulated, accompanied by elevated levels of the oxidative damage marker 3-Nitrotyrosine (3-NT), indicating intensified oxidative stress in ovarian GCs of PCOS patients (Fig. 1D, E). Given the established role of oxidative stress in inducing mitochondrial dysfunction and the previously documented occurrence of apoptosis in PCOS, our study systematically investigated key executors of the mitochondria-dependent apoptotic pathway, and we found that Cleaved Caspase 3 and Cleaved Caspase 9 are increased in GCs derived from PCOS patients (Fig. 1F, G). Moreover, TUNEL assays demonstrated a higher percentage of apoptotic cells within the ovarian tissues in the PCOS-like mouse model (Fig. 1H, I). This concordance across human and murine models suggests dysregulation of oxidative stress and mitochondrial apoptosis as pathological hallmarks of PCOS.

Fig. 1. GC oxidative stress and apoptosis are associated with PCOS.

A qPCR was performed to assess the relative mRNA expression of SOD1 in human ovarian GCs from control subjects and PCOS patients. B Spearman correlation between AFC and SOD1 mRNA expression level (R = -0.5870, p < 0.0001). C Spearman correlation between serum AMH level and SOD1 mRNA expression level (R = −0.4763, p = 0.0001). Western blotting with quantitative analysis of human ovarian GCs revealed increased levels of the pro-oxidative stress marker 3-NT and decreased expression of the antioxidant enzyme SOD1 (D, E), along with upregulated protein levels of pro-apoptotic Cleaved Caspase 3 and Cleaved Caspase 9 (F, G) in PCOS patient-derived GCs relative to control subjects. H, I Representative immunofluorescent staining of ovary sections from each group and quantification of apoptotic levels (TUNEL red, DAPI blue; Scale bar, 100 μm). Student’s t-test or Welch’s t-test are used, and data are presented as mean ± SEM. ^*p < 0.05, ^***p < 0.001.

Oxidative stress inhibits WWP2 expression in GCs and elevates the ubiquitination level of BAK in PCOS

To decipher the mechanism of PCOS-associated GC dysfunction, we analyzed scRNA-seq data of GCs from three PCOS patients and three healthy controls from GSE240688. Following batch effect correction and clustering, t-SNE analysis identified seven GC0-GC6 (Fig. S2A) with distinct top five marker genes (Fig. S2B). KEGG pathway analysis of the top upregulated genes indicated that GC0 was primarily enriched in steroid hormone synthesis pathways, prompting us to focus on this subpopulation for further investigation (Fig. S2C). Differential expression analysis between control and PCOS samples within GC0 revealed significant enrichment in pathways related to oxidative stress, apoptosis, and ubiquitin-mediated proteolysis (Fig. 2A), with apoptosis-related genes specifically shown in (Fig. S2D).

Fig. 2. Oxidative stress inhibits WWP2 expression in GCs and elevates the ubiquitination level of BAK in PCOS.

A Key pathways identified from GC0 cell DEGs in the control and PCOS groups. B Violin plots illustrating WWP2 mRNA expression level in GC0 cells from both PCOS patients and healthy control subjects. C qPCR was used to measure the expression level of WWP2 in control subjects and PCOS patients. D Spearman correlation between AFC and WWP2 mRNA expression level (R = −0.5323, p < 0.0001). E Spearman correlation between AMH level and WWP2 mRNA expression level (R = −0.4732, p = 0.0002). F qPCR was used to measure the expression level of BAK in human ovarian GCs. G Western blotting revealed altered protein expression of WWP2 and BAK in ovarian GCs of PCOS patients versus controls. H Western blotting revealed altered protein expression of WWP2 and BAK in H₂O₂-treated KGN cells compared to untreated controls. I Immunofluorescence staining of ovarian sections was performed to detect Bak protein expression in each group (Bak red, DAPI blue; Scale bar, 100 μm). J, K Ubiquitin Immunoprecipitation (Ub-IP) was performed to assess the ubiquitination levels of BAK in human ovarian GCs from control and PCOS patients, and in KGN cells with or without H₂O₂ treatment. Student’s t-test or Welch’s t-test are used, and data are presented as mean ± SEM. ^**p < 0.01; ns, not significant.

Given our previous findings establishing the critical role of the UPP and the key E3 ubiquitin ligase WWP2 in regulating oxidative stress and apoptosis [22], we further focused on the expression levels of WWP2. Both single-cell analysis and our results showed a significant downregulation of WWP2 expression in GCs from PCOS patients (Fig. 2B, C). Moreover, WWP2 mRNA expression negatively correlated with AFC, AMH, and testosterone, but not with LH/FSH ratio (Figs. 2D, E and S1B, C). WWP2 and SOD1 mRNA expression exhibited a strong positive correlation (Fig. S1D). These results suggest that the WWP2 level in GCs is significantly correlated with oxidative stress and clinical markers of PCOS.

The BCL-2 family proteins are central regulators of apoptosis [23]. We systematically evaluated the mRNA expression of BCL-2 family members in clinical samples from PCOS patients (Figs. 2F and S3A–G). The results revealed significant downregulation of the anti-apoptotic gene BCL2 and upregulation of the pro-apoptotic genes BAD and BAX in PCOS GCs. Most members showed protein expression levels that were consistent with observed mRNA changes (Fig. S3H, I). In contrast, despite comparable BAK mRNA levels, its protein expression was significantly elevated in PCOS GCs and H₂O₂-treated KGN cells (Figs. 2F–I and S4A–D). This post-transcriptional accumulation was associated with a decrease in BAK ubiquitination (Fig. 2J, K), suggesting impaired degradation via the UPP. While E3 ligases like Parkin and HERC1 have been implicated in BAK ubiquitination [16, 24], we observed no significant difference in their gene expression in PCOS GCs compared to controls (Fig. S3J, K). Therefore, the impaired BAK ubiquitination in this disease may be independent of Parkin and HERC1. Notably, Western blot analysis revealed a clear negative correlation between BAK and the E3 ligase WWP2 (Figs. 2G, H and S4A, B). Collectively, these findings suggest that reduced ubiquitination of BAK in PCOS may be induced by decreased WWP2 in PCOS GCs.

WWP2 participates in the proteasome-dependent degradation of BAK and ubiquitinates it via the K113 site

To clarify the connection between WWP2 and BAK, we first conducted both endogenous and exogenous Co-IP assays, which confirmed the interaction between WWP2 and BAK in KGN cells (Fig. 3A, B). Further analysis showed H₂O₂ treatment weakened this binding (Fig. 3C), hinting that oxidative stress may disrupt their interaction. To clarify the specific mechanism of WWP2-BAK regulation, cells were transfected with the HA-WWP2 plasmid or WWP2 siRNA to overexpress or knock down WWP2, and then exposed to the protein synthesis inhibitor CHX and the proteasome inhibitor MG132 for different durations. The data demonstrated that WWP2 overexpression significantly reduced the half-life of BAK (Fig. 3D, E), while WWP2 knockdown had the opposite effect (Fig. 3F, G). Time-dependent MG132 treatment experiments indicated that under proteasome-inhibited conditions, BAK accumulation was faster in the WWP2-overexpressing group than in the control group (Fig. 3H, I). Similarly, BAK protein accumulated faster in the control siRNA group than in the WWP2 siRNA group (Fig. 3J, K). These results suggest that WWP2 controls endogenous BAK protein stability via the proteasome. We hypothesized that the highly conserved lysine 113 residue was the primary target of WWP2-mediated ubiquitination, as this residue is exposed to the cytosol [25] and is spatially consistent with the subcellular localization established by WWP2 [26] (Fig. 3L). Additionally, the ubiquitination assay revealed that the polyubiquitination function of WWP2 on BAK was lost when K113 was mutated to arginine (K113R), and the K113R mutation reduced BAK ubiquitination levels (Fig. 3M, N). Collectively, our data supported that WWP2 interacts with BAK and targets BAK for proteasome-mediated degradation at the K113 site.

Fig. 3. WWP2 participates in the proteasome-dependent degradation of BAK and ubiquitinates it via the K113 site.

A, B Co-IP showed the interaction between WWP2 and BAK in KGN cells. C Co-IP analysis of the interaction between WWP2 and BAK in KGN cells following treatment with H₂O₂ (200 μM, 12 h). Western blot and quantification of BAK protein levels in KGN cells with WWP2 overexpression (D, E, H, I) or knockdown (F, G, J, K) following treatment with CHX (20 μM) or MG132 (10 μM) for the indicated durations. Data are presented as mean ± SEM from three independent experiments (^*p < 0.05, ^**p < 0.01; two-way ANOVA with Bonferroni’s multiple comparison post hoc test). L The Lys113 site of BAK is evolutionarily conserved. M Lys113 is required for BAK ubiquitination. KGN cells were transfected with plasmids expressing Flag-BAK-K113R or Flag-BAK-WT. Ubiquitinated BAK was immunoprecipitated using anti-Flag affinity gel and detected by immunoblotting with anti-ubiquitin (Ub) antibodies. N WWP2 promotes BAK ubiquitination at the Lys113 site. KGN cells were co-transfected with Flag-BAK-WT, Flag-BAK-K113R, and HA-WWP2. BAK ubiquitination levels were detected using anti-Ub antibodies.

To further elucidate the functional role of BAK K113 in mitochondrial apoptosis, we first deleted endogenous BAK in KGN cells using CRISPR/Cas9 gene editing (Fig. 4A). BAK deletion markedly attenuated mitochondrial apoptosis under oxidative stress, as indicated by reduced levels of Cleaved Caspase 3 and Cleaved Caspase 9, along with a decrease in TUNEL-positive cells (Figs. 4B–F and S4E). These results confirm the essential role of BAK in promoting apoptosis in this setting. Next, we reintroduced either wild-type BAK (BAK-WT) or the ubiquitination-defective mutant BAK-K113R into BAK-KO KGN cells. Compared to BAK-WT, expression of BAK-K113R led to a significant increase in apoptosis, as assessed by Western blotting and TUNEL staining (Figs. 4G–K and S4F). To confirm the specific role of WWP2 in regulating BAK ubiquitination, we further overexpressed WWP2 in cells expressing BAK-WT or BAK-K113R. WWP2 overexpression inhibited oxidative stress-induced apoptosis in BAK-WT cells, but this anti-apoptotic effect was attenuated in cells expressing BAK-K113R (Figs. 4L–P and S4G). Together, these findings highlight the essential role of ubiquitination at BAK K113 in controlling apoptosis in KGN cells and identify WWP2 as a key E3 ligase mediating this function.

Fig. 4. Ubiquitination of BAK at K113 by WWP2 attenuates mitochondrial apoptosis.

A Western blot was performed to assess BAK protein levels in KGN cells transfected with sgControl or sgBAK via CRISPR/Cas9. In BAK-proficient (BAK^+/+) and BAK-deficient (BAK^−/−) KGN cells treated with or without H₂O₂ (200 μM, 12 h): TUNEL staining was used to detect apoptosis after H₂O₂ treatment in KGN cells (TUNEL red, DAPI blue; Scale bar, 100 μm) (B); Western blot and quantification of BAK, Cleaved Caspase 3, and Cleaved Caspase 9 in treated KGN cells (C–F). In BAK^−/− KGN cells transfected with Flag-Vector, Flag-BAK-WT, or Flag-BAK-K113R and treated with or without H₂O₂: TUNEL staining was used to detect apoptosis after H₂O₂ treatment in KGN cells (TUNEL red, DAPI blue; Scale bar, 100 μm) (G); Western blot and quantification of BAK, Cleaved Caspase 3, and Cleaved Caspase 9 in treated KGN cells (H–K). In BAK^−/− KGN cells transfected with Flag-BAK-WT or Flag-BAK-K113R with or without HA-WWP2: TUNEL staining was used to detect apoptosis after H₂O₂ treatment in KGN cells (TUNEL red, DAPI blue; Scale bar, 100 μm) (L); Western blot and quantification of BAK, Cleaved Caspase 3 and Cleaved Caspase 9 in treated KGN cells (M–P); Statistical significance was assessed using Data are presented as mean ± SEM, one-way ANOVA followed by Tukey’s post hoc test; ^*p < 0.05, ^**p < 0.01, ^***p < 0.001.

WWP2 overexpression orchestrates BAK downregulation to attenuate apoptosis and ROS production elicited by oxidative stress

We further explored the functional impact of WWP2 overexpression on oxidative stress-induced KGN cells. Exogenous WWP2 overexpression increased BAK ubiquitination and suppressed ROS-induced BAK accumulation (Fig. 5A–C). Given that BAK is a pivotal regulator of the mitochondrial apoptotic pathway, whose oligomerization promotes the release of Cyt C from mitochondria into the cytosol, thereby initiating caspase cascade activation [27], we performed subcellular fractionation assays. The results demonstrated that H₂O₂ treatment markedly reduced mitochondrial Cyt C levels while concomitantly increasing its cytosolic levels, indicative of enhanced mitochondrial outer membrane permeabilization and apoptosis initiation. However, WWP2 overexpression effectively blocked this Cyt C translocation (Figs. 5D and S5A), thus interrupting the subsequent apoptotic signaling, as assessed by the expressions of Cleaved Caspase 3 and Cleaved Caspase 9 (Fig. 5E, F). Similar results were obtained by using TUNEL staining and flow cytometry in KGN cells (Fig. 5G–J).

Fig. 5. WWP2 overexpression orchestrates BAK downregulation to attenuate apoptosis and ROS production elicited by oxidative stress.

A Ub-IP analysis was performed to assess the role of WWP2 in regulating the levels of BAK ubiquitination. KGN cells were transfected with or without the HA-WWP2 plasmid. BAK ubiquitination levels were detected using an anti-Ub antibody. B, C Western blot and quantification of BAK in KGN cells transfected with HA-WWP2 or control plasmid for 36 h, followed by treatment with or without H₂O₂ (200 μM, 12 h). D Western blot was conducted to assess Cyt C levels in the cytoplasmic, mitochondrial, and total cellular fractions under the same treatment conditions. E, F Western blot and quantification of Cleaved Caspase 3 and Cleaved Caspase 9 in treated KGN cells. G, H TUNEL staining and quantification of apoptotic cells after transfection and H₂O₂ treatment in KGN cells (TUNEL red, DAPI blue; Scale bar, 100 μm). I, J Flow cytometry and quantitative analysis of apoptosis under the indicated conditions. K ROS immunofluorescence staining assessed oxidative stress levels in KGN cells (Scale bar, 100 μm). L, M Western blot and quantification of 3-NT and SOD1 in KGN cells after transfection and H₂O₂ exposure. Data are presented as mean ± SEM. Statistical significance was assessed using one-way ANOVA followed by Tukey’s post hoc test. ^*p < 0.05, ^**p < 0.01.

A positive feedback cycle exists between the UPP and oxidative stress. Sustained oxidative stress impairs UPP function and intracellular proteolysis, which in turn promotes mitochondrial dysfunction and pathological ROS overproduction [28, 29]. Based on this, we further examined the impact of WWP2 on oxidative stress regulation. Fluorescence was used to reveal that WWP2 overexpression attenuated abnormal ROS accumulation, an effect associated with decreased levels of the oxidative stress marker 3-NT and increased expression of the antioxidant enzyme SOD1 (Figs. 5K–M and S5B). These findings suggest that WWP2 mitigates oxidative stress-induced damage and apoptosis, at least in part, by modulating the stability of BAK.

WWP2 depletion triggers BAK upregulation, consequently promoting apoptosis and ROS production induced by oxidative stress

The pivotal role of WWP2 in modulating apoptotic signaling was further validated through siRNA-mediated knockdown, which reduced BAK ubiquitination and augmented its protein stability (Fig. 6A). Under oxidative stress induced by H₂O₂, WWP2 deficiency resulted in pronounced BAK accumulation (Fig. 6B, C), concomitant with enhanced mitochondrial Cyt C release (Figs. 6D and S5C) and upregulated expression of Cleaved Caspases 3 and Cleaved Caspases 9 (Fig. 6E, F). Consistent with this pro-apoptotic phenotype, TUNEL staining and flow cytometry further confirmed that WWP2 knockdown exacerbated H₂O₂-induced KGN cell apoptosis (Fig. 6G–J). Moreover, WWP2 deficiency elicited the perturbation of redox homeostasis, potentiating H₂O₂-triggered ROS generation (Figs. 6K and S5D) while concurrently impairing antioxidant defense, as manifested by reduced SOD1 expression and enhanced 3-NT formation (Fig. 6L, M). Collectively, these findings establish that WWP2 acts as a critical molecular mediator linking oxidative stress signaling to mitochondrial apoptotic pathways, primarily through promoting BAK ubiquitination, thereby alleviating oxidative stress and suppressing the activation of intrinsic apoptosis.

Fig. 6. WWP2 depletion triggers BAK upregulation, consequently promoting apoptosis and ROS production induced by oxidative stress.

A Ub-IP analysis was performed to assess the role of WWP2 in regulating the levels of BAK ubiquitination. KGN cells were transfected with control or WWP2-siRNA, and BAK ubiquitination levels were detected by immunoprecipitation using anti-BAK antibodies, followed by immunoblotting with anti-Ub antibody. B, C Western blot and quantification of WWP2 and BAK expression in KGN cells transfected with WWP2-siRNA or control for 60 h, followed by treatment with or without H₂O₂ (200 μM, 12 h). D Western blot was conducted to assess Cyt C levels in the cytoplasmic, mitochondrial, and total cellular fractions under identical treatment conditions. E, F Western blot and quantification of Cleaved Caspase 3 and Cleaved Caspase 9 in treated KGN cells. G, H TUNEL staining and quantification of apoptotic cells after transfection and H₂O₂ treatment in KGN cells (TUNEL red, DAPI blue; Scale bar, 100 μm). I, J Flow cytometry and quantitative analysis of apoptosis under the indicated conditions. K ROS immunofluorescence staining assessed oxidative stress levels in KGN cells (Scale bar, 100 μm). L, M Western blot and quantification of 3-NT and SOD1 in KGN cells after transfection and H₂O₂ exposure. Data are presented as mean ± SEM. Statistical significance was assessed using one-way ANOVA followed by Tukey’s post hoc test. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001.

Wwp2 ablation in mice exacerbates PCOS-induced GC apoptosis and hormonal disorders

To validate the critical role of the ROS-WWP2-BAK signaling axis in GC apoptosis and ovarian dysfunction, we generated Wwp2-knockout (Wwp2-KO) mice and established a PCOS model through combined DHEA administration and HFD induction. Immunofluorescence analysis of ovarian sections from PCOS mouse models revealed that Wwp2 ablation markedly upregulated Bak expression (Fig. 7A, B), consistent with in vitro results indicating that WWP2 knockdown leads to reduced BAK ubiquitination and increased protein stability. Correspondingly, TUNEL staining demonstrated a marked rise in ovarian apoptosis in Wwp2-KO PCOS mouse models (Fig. 7C, D). Serum ELISA demonstrated that Wwp2 deficiency led to elevated testosterone levels and an increased LH/FSH ratio compared with controls (Fig. 7E–H), resembling the exacerbated hyperandrogenism and gonadotropin imbalance observed in clinical PCOS patients. Ovarian histological analysis further revealed that Wwp2-KO mice exhibited collapsed follicular structures (Figs. 7I and S5E), suggesting exacerbated GC apoptosis and ovarian dysfunction. Notably, genetic ablation of Wwp2 did not significantly compromise body weight or glucose tolerance (Fig. S5F, G), indicating that WWP2 primarily influences PCOS phenotypes through regulation of ovarian function rather than metabolic perturbation. Based on these results, we conclude that WWP2 is crucial in PCOS-induced GC apoptosis and hormonal disorders.

Fig. 7. Wwp2 ablation in mice exacerbates PCOS-induced GC apoptosis and hormonal disorders.

A, B Immunofluorescence staining of ovarian sections showing Bak expression in PCOS wild-type (Wwp2-WT) and Wwp2-KO mice (Bak red, DAPI blue; scale bar, 100 μm), with quantification of Bak fluorescence intensity in ovaries. C, D TUNEL staining of ovarian sections and quantification of apoptotic cells in Wwp2-WT and Wwp2-KO PCOS mice (TUNEL red, DAPI blue; Scale bar, 100 μm). Serum testosterone (E), LH (F), and FSH (G) levels were measured by ELISA, and LH/FSH ratio analysis (H) in mice. I Ovarian morphology assessed by H&E staining (Scale bar, 200 μm). Data are expressed as mean ± SEM. Statistical significance was assessed using Student’s t-test. ^*p < 0.05, ^***p < 0.001; ns, not significant.

Discussion

Despite numerous attempts to enable early diagnosis and prevent PCOS-induced impaired reproductive function [30], current treatment strategies primarily focus on symptom relief and fail to fully restore ovarian function. Considering that GC apoptosis and oxidative stress play critical roles in abnormal follicular development and reduced fertility in PCOS patients [31, 32], exploring the signaling molecular pathways that regulate GC apoptosis may provide new strategies for alleviating PCOS. In the current study, we demonstrate for the first time that the E3 ubiquitin ligase WWP2, by targeting BAK for ubiquitination, plays a pivotal role in PCOS-associated ovarian dysfunction. Mechanically, oxidative stress led to WWP2 downregulation, which reduced BAK ubiquitination at the K113 site to increase BAK expression and thus promote mitochondrial apoptosis. Furthermore, Wwp2 knockout in mice further exacerbated reproductive phenotypes in PCOS mice (Fig. 8).

Fig. 8. Working model.

Schematic diagram illustrating the regulation of GC apoptosis in normal (left) and PCOS (right) ovaries. Under normal conditions, WWP2 promotes K113-linked polyubiquitination and degradation of BAK, maintaining mitochondrial integrity and suppressing apoptosis. In PCOS, reduced WWP2 impairs BAK degradation, leading to BAK accumulation, mitochondrial damage, and caspase-dependent apoptosis. WWP2 acts as a key regulator linking oxidative stress to BAK stability, and its downregulation contributes to GC apoptosis and polycystic ovarian changes.

Elevated oxidative stress and BAK-dependent mitochondrial apoptosis in PCOS GCs

GCs provide essential metabolic support and establish a favorable microenvironment for oocyte growth, thereby sustaining follicular development and ovulation [33, 34]. Given their pivotal role in ovarian physiology, we isolated GCs from PCOS patients and found significantly elevated oxidative stress and BAK-dependent mitochondrial apoptosis. Concurrently, the expression of SOD1, a key antioxidant enzyme, was significantly downregulated, and its levels were negatively correlated with AMH and AFC, two critical clinical indicators of ovarian reserve in PCOS. These findings support the concept that GC dysfunction represents a key pathological link contributing to impaired ovarian function in PCOS. Previous studies have demonstrated that ROS accumulation is a major inducer of GC apoptosis and follicular atresia. Chen et al. recently reported that ROS promotes mitochondrial apoptosis in GCs through activation of the MEK/ERK1/2 signaling pathway [5], whereas Gong et al. showed that growth hormone suppresses ROS-induced GC apoptosis by activating the PI3K/AKT pathway [35]. Although ROS accumulation and its associated signaling cascades are recognized as important drivers of GC apoptosis in PCOS, the intricate regulatory networks governing mitochondrial apoptosis are not yet fully understood.

BAK accumulation is mediated by defective ubiquitination in PCOS

We therefore focused on the core regulators of mitochondrial apoptosis, the BCL-2 protein family [36], and systematically evaluated their expression in clinically obtained human GCs. Surprisingly, although the protein level of the essential pro-apoptotic effector BAK was significantly elevated in PCOS GCs, its mRNA expression did not exhibit a significant difference. This suggests that the abnormal accumulation of BAK may be regulated by post-translational modification. Given that DEGs in PCOS GCs were highly enriched in the UPP, we further investigated and confirmed reduced BAK ubiquitination levels in patient GCs.

Ubiquitination, a crucial post-translational modification mediated by E3 ubiquitin ligases, regulates protein stability, localization, or activity through the attachment of distinct chain types to the substrate, contributing to diseases including PCOS [37, 38]. Previous studies have shown that different E3 ubiquitin ligases mediate functionally distinct ubiquitination of BAK. Specifically, Parkin catalyzes non-degradative ubiquitination of BAK at K113, suppressing its pro-apoptotic activity and thereby contributing to neuronal function maintenance [16], whereas HERC1-mediated ubiquitination in certain viral infection models promotes BAK proteasomal degradation [24]. Given reduced BAK ubiquitination and concomitant protein accumulation in PCOS GCs, we investigated the underlying regulatory mechanism. We discovered a significant downregulation of the E3 ubiquitin ligase WWP2 in PCOS patient GCs and confirmed its novel interaction with BAK in KGN cells. In contrast, expression changes of other known BAK regulators, such as Parkin and HERC1, could not adequately account for the observed BAK dysregulation in PCOS GCs.

WWP2-mediated BAK K113 ubiquitination is essential for ovarian GC function

WWP2 is an E3 ubiquitin ligase that modulates diverse physiological and pathological processes by regulating the ubiquitination and degradation of specific substrates [26]. WWP2 could interact with the DNA damage-related protein PARP1 and ubiquitinate its K418 and K249 residues, thereby modulating cardiomyocyte apoptosis [39]. Our previous study has demonstrated that oxidative stress-induced activation of JNK reduces WWP2 expression [22]. Here, we further showed that ROS accumulation weakened the interaction between WWP2 and BAK and reduced WWP2-mediated ubiquitination of BAK at the K113 site, thereby suppressing its degradation through the UPP and leading to abnormal BAK accumulation. In H₂O₂-treated KGN cells, BAK KO alleviated apoptosis. More importantly, using a BAK K113R point mutant cell line, we demonstrated that WWP2 modulates apoptosis by mediating the ubiquitination of BAK at the K113 site.

In addition, accumulating evidence suggests that the UPP engages in a feedback loop with cellular stress, in which severe or sustained oxidative stress impairs UPP function and intracellular proteolysis, subsequently disrupting mitochondrial dynamics and promoting mitochondrial fragmentation and excessive ROS production [28, 29]. In line with this, overexpression of WWP2 attenuated mitochondrial ROS accumulation and apoptosis triggered by H₂O₂ exposure, whereas knockdown further aggravated these effects. Moreover, genetic deletion of Wwp2 further exacerbated reproductive dysfunction in PCOS mice. Our findings establish WWP2 as a novel protective regulator of ovarian function by modulating oxidative stress and apoptosis in GCs.

Limitation

Certain limitations of our work deserve consideration in future studies. First, although we demonstrated the critical function of ROS-WWP2-BAK signaling axis in ovarian GC apoptosis, the potential roles of other BCL-2 family members and E3 ubiquitin ligases in regulating mitochondrial apoptosis in PCOS GCs cannot be ruled out. Second, while we identified ROS-induced downregulation of WWP2, the upstream signaling mechanisms by which oxidative stress suppresses WWP2 expression remain unclear in this disease. Third, although the function of WWP2-mediated BAK ubiquitination in the cell line and the role of WWP2 in mouse models were identified, more direct evidence is needed, such as from BAK point mutant knock-in mice, to provide strong genetic support for our conclusions.

Conclusion

In summary, our study first identifies WWP2 as a novel E3 ubiquitin ligase of BAK in PCOS, mediating its K113 site-specific ubiquitination and degradation. These findings reveal a previously unrecognized ROS-WWP2-BAK axis that regulates GC apoptosis and contributes to ovarian dysfunction in PCOS. Targeting this pathway through pharmacological activation of WWP2 or small-molecule enhancers of BAK ubiquitination may offer a promising therapeutic strategy to restore ovarian function and improve PCOS progression.

Author contributions

Wenke Wang: Writing - original draft, Writing - review & editing, Software, Methodology, Formal analysis, Data curation. Wenjie Wu: Investigation, Software, Formal analysis, Validation, Methodology, Data curation. Mingjun Hao: Writing - original draft, Software, Formal analysis, Validation, Data curation. Shenshen Cui: Writing - review & editing, Methodology, Formal analysis, Data curation. Siqi Zhao: Resources, Investigation, Data curation. Jian-Fei Pei: Writing - review & editing, Visualization, Validation, Supervision, Conceptualization. Naijin Zhang: Writing - original draft, Supervision, Funding acquisition, Conceptualization. Da Li: Writing - review & editing, Writing - original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Funding information

This research was supported by Noncommunicable Chronic Diseases-National Science and Technology Major Project (2025ZD0547800), the National Natural Science Foundation of China (82371647, 82470283, 82271615), National Key Research and Development Program (No. 2024YFC2706900), the Science Foundation for Outstanding Youth of Liaoning Province (No. 2024JH3/50100023), Scientific Research Fund of Liaoning Provincial Education Department (No. LJ222410159094) and the Liaoning Revitalization Talents Program (XLYC2403104).

Data availability

All data relevant to the study are included in the article.

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

All experimental procedures involving human participants and animals were performed in accordance with the relevant guidelines and regulations. The human study was approved by the Ethics Committee of Shengjing Hospital, China Medical University, Shenyang (Approval No. SZCS2025016), and informed consent was obtained from all IVF participants for the collection of ovarian granulosa cells. The animal study protocol was approved by the Animal Ethics Committee of China Medical University (License number CMU2023296).

Supplementary information

The online version contains supplementary material available at 10.1038/s41419-026-08500-y.