Prenatal Exposure to PFOA Induces Ovarian Function Impairment via the Disruption of the PPARγ/ANGPTL4 Pathway.

To investigate the potential association between PFAS and long-term ovarian aging, we first examined the correlation between serum PFAS levels and age at natural menopause (ANM) using data from the NHANES 1999–2020 March. Following a rigorous inclusion and exclusion process, a total of 2924 premenopausal women and 1785 postmenopausal women were included in the final analysis based on a Cox proportional hazards model ( Figure A). After adjusting for potential confounders including age, race/ethnicity, education, parity, and NHANES cycle, we observed that higher serum concentrations of perfluorooctanoic acid (PFOA), as well as perfluorooctanesulfonate (PFOS) and perfluorohexanesulfonate (PFHxS), were significantly associated with an increased risk of earlier age at natural menopause. Notably, this association exhibited a monotonic, dose–response pattern across increasing exposure tertiles (2nd tertile: medium; third tertile: high) for PFOA, PFOS, and PFHxS ( Figure B). Detailed baseline characteristics of the study population as well as the statistical data were provided in Tables S2 and S3 . The NHANES analysis revealed the association of PFOA exposure with an increased risk of earlier menopause (2nd tertile: HR = 1.220, P = 0.004; third tertile: HR = 1.262, P = 0.002, Figure B and Table S3 ). To further elucidate the contribution of prenatal PFOA exposure to ovarian outcomes, we established a mouse model to assess its long-term effects across the reproductive lifespan. Time-mated pregnant CD-1 mice were administered 1, 2.5, 5, or 10 mg/kg/day PFOA via oral gavage from GD11 to GD20 ( Figure A). The 10 mg/kg group exhibited severe developmental toxicity, including reduced survival rate and decreased number of live pups per litter, and was therefore excluded from further analysis ( Figure S1A–C ). Prenatal PFOA exposure induced ovarian cysts and senescence in middle-aged F1 offspring. (A) Experimental design is shown. Time-mated CD-1 mice received daily oral gavage of PFOA (1, 2.5, 5, or 10 mg/kg/day) and corn oil from gestational day (GD) 11 to postnatal day (PND) 0 ( n = 4–7). F1 female offspring were evaluated at 22 days (prepubertal), 3 months (young, sexual mature stage 1), 6 months (young, sexual mature stage 2), and 9 months (middle-aged, sexual stage 3). The 10 mg/kg group was excluded from further analysis due to severe developmental toxicity. (B) Representative gross images of ovaries from control (CTR-9-month) and PFOA-exposed (5 mg/kg; PFOA-9-month) mice at 9 months. Insets show magnified views highlighting increased ovarian size and vascularization in exposed animals. Hematoxylin and eosin (H&E)-stained ovarian sections (40× and 100×) are showed. An: antral follicle and PA: preantral follicle. Cyst incidence analysis: Fisher’s exact test comparing the prevalence of ovarian cyst between groups. (C,D) Serum concentrations of follicle-stimulating hormone (FSH) and luteinizing hormone (LH) in 9 month old offspring. Data are expressed as mean ± SEM. ** P < 0.01. * P < 0.05. By 9 months of age (middle-aged stage), gross morphological changes were evident in the ovaries of PFOA-exposed F1 females, particularly in the 5 mg/kg group ( Figure B). These ovaries were notably enlarged and highly vascularized compared to the controls. Histopathological examination revealed that 66.7% (6/9) of mice in the 5 mg/kg group developed at least one ovarian cyst, whereas only one animal in the control group (1/11, 9.1%) exhibited cystic structures, a statistically significant difference ( P = 0.0166, Fisher’s exact test; Figure B, table below). H&E staining showed that the cystic structures lacked GC lining and were filled with pale alkaline fluid, contrasting with the multilayered granulosa architecture of normal antral follicles. Surrounding follicles were displaced to the ovarian periphery, indicating local structural disruption. Endocrine profiling at 9 months showed significantly elevated levels of FSH and LH in PFOA-exposed females relative to controls ( Figure C,D), consistent with a dysregulated hypothalamic-pituitary-ovarian (HPO) axis and advanced ovarian function impairment. Histological analysis of the F1 ovaries revealed progressive structural alterations. The ovarian cyst was first identified in one ovary from the 5 mg/kg group ( n = 1/6) as early as at 3 months. Besides, ovaries from PFOA-exposed groups exhibited a tendency toward reduced volume ( Figure A). Concurrently, significant increases in atretic follicles were observed in the 1 mg/kg and 5 mg/kg groups ( P < 0.05 vs control; Figure B,C, enlarged images of typical atretic follicles showed in Figure S2 ). Prenatal PFOA exposure impaired ovulation and caused irregular estrus cycle at sexual mature stages. (A) H&E staining ovarian sections from 3 month and 6 month old F1 female offspring exposed to 0 (CTR), 1, 2.5, or 5 mg/kg/day PFOA. Pa, preantral follicles; An, antral follicles; CL, corpus luteum; Cys, ovarian cyst; arrow (→) indicates atretic follicles of all stages. (B,C) Follicle quantification of 3 month F1 ovaries ( n = 6). (D-E) Follicle quantification of 3 month F1 ovaries ( n = 4–9). (F) Representative vaginal smear of diestrus, proestrus, estrus, and metestrus stages, showing different ratios of cornified epithelial, nucleated epithelial and polymorphonuclear leukocyte cells. Graphic representation of the estrus cycle for 22 consecutive days among 3 month old F1 offspring determined by vaginal lavage cytology ( n = 5–9). The bar plot shows the number of days spent in each stage of estrous cycle and total cycle length. Proestrus (P), estrus (E), metestrus (M), and diestrus (D). (H,I) Serum concentrations of FSH, LH, and AMH at 3 and 6 months across exposure groups. * P < 0.05 and ** P < 0.01 compared with control. Data are expressed as mean ± SEM. By 6 months, isolated cases of ovarian cysts were noted (2.5 mg/kg: n = 1/9; 5 mg/kg: n = 2/8), consistent with earlier pathological observations in middle-aged offspring ( Figure A). The number of corpora lutea was significantly reduced across all PFOA-exposed groups (1 mg/kg: P < 0.05; 2.5 mg/kg: P < 0.05; 5 mg/kg: P < 0.01 vs control; Figure D,E), indicating impaired ovulatory capacity. Estrous cycle monitoring revealed profound disruption: PFOA-exposed mice exhibited significantly prolonged cycles (cycle length: P < 0.05 for 5 mg/kg vs control; Figure F), primarily due to extended diestrus duration. In addition to prolonged diestrus phase, the PFOA 2.5 mg/kg group showed a slight proestrus/estrus extension, which indicated more time spent before ovulation. So, the altered phase distribution suggested dose-dependent progression from subfertility to anovulation. Endocrine profiling showed no significant changes in FSH, LH, estradiol (E 2 ), or testosterone (T) at 3 or 6 months ( Figures G,H and S3A–B ). However, anti-Müllerian hormone (AMH) levels were significantly reduced in the 2.5 mg/kg group at 3 months ( P < 0.05; Figure I), indicating compromised ovarian reserve prior to gross morphological changes. No significant differences were observed in body weight or multiple organ indices among the exposure groups. However, a modest increase in uterus index (uterus weight/body weight) was noted specifically in the 5 mg/kg group at 3 months ( Table S4 ). Granulosa cells constitute the major steroidogenic cell population within ovarian follicles and form an avascular layer surrounding the oocyte. They are essential for hormone synthesis and the secretion of growth factors critical to follicular development. After PFOA exposure, cell proliferation for human KGN granulosa cell and hormone-related genes such as Aromatase, CYP11A1, and StAR were measured. Occupationally exposed individuals have been reported to reach serum PFOA concentrations ranging from 3 to 222 μmol/L. , , Since PFOA levels in serum and follicular fluid are highly correlated and comparable, this range provides a physiologically relevant basis for in vitro experiments. Human KGN granulosa cells exposed to increasing PFOA concentrations (0–500 μmol/L) exhibit reduced cell viability at 500 μmol/L concentration as the lethal dose. Specifically, 200 μmol/L PFOA significantly attenuated cell proliferation ( P 95% cell viability ( Figure A). Cell cycle analysis revealed that 200 μmol/L PFOA induced G0/G1 phase arrest and concurrently suppressed G2/M progression ( Figure D,E), increasing the arrested population from 84.9% to 87.6%, and concurrently suppressed G2/M progression, decreasing this population from 8.8% to 5.1%. In addition, PFOA exposure altered the expression of key steroidogenic genes ( Figure F–H): Aromatase (CYP19A1) was markedly upregulated to 129.0%–153.2% by PFOA with concentrations ≥10 μmol/L, concomitant with a pronounced induction of CYP11A1 (P450scc) expression; whereas StAR expression exhibited selective elevation at 200 μmol/L. PFOA inhibited granulosa cell proliferation by inducing G0/G1 arrest accompanied by dysregulated steroidogenic genes. (A) Optical density (OD) measurements at 450 nm of KGN cells treated with increasing PFOA concentrations (0–500 μmol/L) for 48 h after the Cell Counting Kit-8 assay. (B) Representative fluorescence micrographs showing DAPI-stained nuclei (blue), EdU-labeled proliferating cells (red), and merged images across PFOA doses. Scale bar: 100 μm. (C) Percentage of EdU-positive cells relative to total DAPI-stained nuclei at each concentration. (D) Flow cytometry histograms of propidium iodide (PI)-stained cells showing DNA content distribution after PFOA treatment. (E) Proportions of cells in G0/G1, S, and G2/M phases across treatment groups as determined by PI staining analysis. (F–H) Real-time quantitative PCR detection of Aromatase, CYP11A1, and StAR in KGN granulosa cells exposed to a gradient of PFOA concentrations. Data are expressed as mean ± SEM tested by the one-way ANOVA method. ** P < 0.01. * P < 0.05. To investigate potential molecular mechanisms underlying ovarian phenotypes observed after PFOA exposure, we performed mRNA sequencing on human KGN ovarian granulosa cells treated with 200 μmol/L PFOA for 48 h. A total of 32 differentially expressed ovary-related genes (DEGs) were identified, among which Angiopoietin-like 4 (ANGPTL4) showed the most significant upregulation ( Figure A). A hierarchical heatmap was generated using DEGs ( Figure B). KEGG pathway enrichment analysis revealed that the PPAR signaling pathway and cholesterol metabolism were the most significantly enriched pathways ( Figure C). PFOA activated the PPARγ signaling pathway and upregulated ANGPTL4 expression in ovarian granulosa cells. (A) Transcriptomic sequencing was performed in KGN cells treated with 200 μmol/L PFOA versus control (DMSO) for 48 h. Volcano plot displayed downregulated (blue dots) and upregulated (red dots) DEGs. The threshold was set as |log 2 FC| ≥ 0.5 and adjusted P -value ≤0.2. (B) Heatmap of hierarchically clustered DEGs (rows) between control and the PFOA treatment group. (C) Enriched KEGG pathways of DEGs. (D) Relative mRNA levels of PPARα and PPARγ in granulosa cells exposed to increasing PFOA concentrations (0–200 μmol/L). (E) qRT-PCR analysis of PPAR-related genes (ANGPTL4, FABP3, and PLIN2) mRNA expression in cells treated with PFOA. (F) Molecular docking model visualizing binding poses between PFOA (colored in cyan) and the amino acids within PPARγ ligand-binding domain (colored in magenta). (G) Immunohistochemical staining for ANGPTL4 in ovarian follicles from 6 month old offspring mice prenatally exposed to vehicle or 5 mg/kg PFOA. We further assessed the expression of PPARα and PPARγ in granulosa cells exposed to a gradient of PFOA concentrations. PPARγ expression was markedly increased to 129.6% ± 5.9% and 158.7% ± 6.2% after exposure to 100 and 200 μmol/L for 48 h, while PPARα levels remained unchanged ( Figure D). Among the PPAR target genes, ANGPTL4, FABP3, and PLIN2 were significantly upregulated following PFOA treatment at lower concentrations (10–50 μmol/L), as validated by qRT-PCR ( Figure E). Specifically, consistent with transcriptomic analyses, ANGPTL4 expression was significantly upregulated to 213.6% ± 58.6%, 517.9% ± 62.3%, and 1260.3% ± 199.3% after 48 h exposure to 50, 100, and 200 μmol/L PFOA in a dose-dependent manner, respectively. Molecular docking was performed to evaluate the binding affinity of PFOA to the human PPARγ LBD ( Figure F). Interactions were showed with binding energies of −8.8 kcal/mol for PPARγ LBD. PFOA formed hydrogen bonds with residues His323, Ser289, His449, and Tyr473 of the PPARγ LBD. Immunohistochemistry further confirmed increased ANGPTL4 protein expression in the ovarian follicles of 6 month old F1 female offspring prenatally exposed to 5 mg/kg PFOA ( Figure G). To establish the functional role of ANGPTL4 in PFOA toxicity, we performed siRNA-mediated knockdown in KGN cells. ANGPTL4 silencing significantly attenuated PFOA-induced cell proliferation suppression, increasing EdU + cells from 7.8% ± 0.6% (PFOA + Si-NC) to 10.6% ± 0.9% (PFOA + Si-ANGPTL4, P < 0.05; Figure A,B). Cell cycle analysis revealed that ANGPTL4 depletion rescued PFOA-driven G0/G1 arrest, reducing the arrested population from 81.0% to 73.9%, with concomitant increases in S-phase and G2/M-phase cells ( Figure C,D). At the molecular level, PFOA downregulated cyclin D1 expression, while ANGPTL4 knockdown partially restored its levels ( Figure E), indicating cyclin D1/CDK4 dysregulation as a key mechanism. ANGPTL4 knockdown rescued PFOA-induced cell cycle arrest in granulosa cells. (A,B) Representative images of EdU-labeled proliferating cells in KGN cells treated as follows for 48 h: Ctrl (DMSO), 200 μmol/L PFOA, 200 μmol/L PFOA + Si-NC, and 200 μmol/L PFOA + Si-ANGPTL4. Scale bars: 100 μm. Percentage of EdU-positive cells was quantified and compared across treatment groups. (C,D) Flow cytometry histograms of propidium iodide (PI)-stained cells showing cell cycle profiles. Quantitative statistics of proportions of cells in G0/G1, S, and G2/M phases. (E) Western blot analysis of ANGPTL4, Cyclin D1, and GAPDH protein levels. (F) Forest plot showed Mendelian randomization analysis for causal associations between genetically predicted ANGPTL4 levels and different ovarian diseases. Human genetic evidence further supports the clinical relevance of ANGPTL4 dysregulation in ovarian pathologies. Mendelian randomization analysis demonstrated a significant causal association between genetically predicted ANGPTL4 levels and endometriosis risk (OR = 1.305, P = 0.022; Figure F), a disease characterized by functional ovarian cysts and impaired ovarian function in most circumstances. Given that PFOA induced granulosa cell cycle arrest, we next examined its effects during the prepubertal stagethe critical window of initial follicular recruitment. At this stage, the ovary transitions from quiescent primordial follicles to active follicular growth, making it highly sensitive to environmental insults. A dose-dependent decrease in the ovary index (ovary weight/body weight) was observed in F1 offspring at exposure levels ≥2.5 mg/kg ( Figure A), coinciding with early life growth retardation ( P < 0.001; Figure S1D ), although body weight recovered in adulthood ( Table S4 ). No comparable change was observed in the dams ( Figure B). Prenatal PFOA exposure caused aberrant folliculogenesis via dysregulated PI3K-AKT signaling. (A) Ovary index (ovary weight/body weight) of PND22 F1 females across PFOA exposure groups (1, 2.5, 5 mg/kg/day). (B) Maternal ovary index in dams at PND22 of offspring. (C) Representative HE staining of PND22 ovaries for follicle classification ( n = 4–9). Pri, primary; Pa, preantral; AtPa, atretic preantral; and An, antral. (D–F) Quantification of primordial, primary, preantral, and antral follicles standardized to ovarian section area (units/mm 2 ). (G) Transcriptomic profiling of PND22 ovaries (5 mg/kg vs control); volcano plot shows downregulated (blue) and upregulated (red) DEGs. (H) KEGG analysis of significantly enriched pathways. (I) Gene set enrichment analysis (GSEA) plot of the mTOR signaling pathway. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA; * P < 0.05, ** P < 0.01. Ovarian morphological assessment revealed distinct perturbations in folliculogenesis. Offspring exposed to 5 mg/kg PFOA exhibited increased counts of primary and healthy preantral follicles compared to controls ( P < 0.01 and P < 0.05, respectively; Figure C–E), accompanied by a concurrent rise in atretic preantral follicles ( P < 0.01). In contrast, healthy and total antral follicles were numerically reduced, although the difference narrowly missed statistical significance ( Figure F). These observations collectively indicate hindered preantral-to-antral transition and accelerated early follicular recruitment following in utero PFOA exposure. Transcriptomic profiling of PND22 ovaries (5 mg/kg vs control) identified 788 differentially expressed genes (|log 2 FC| ≥ 0.5, adjusted P -value ≤0.05) ( Figure G). KEGG pathway enrichment analysis highlighted PI3K–AKT–mTOR signaling as a significantly enriched pathway ( Figure H), and gene set enrichment analysis (GSEA) further confirmed activation of this cascade (NES = 1.512, P = 0.006; Figure I), implicating this pathway in aberrant early folliculogenesis after prenatal PFOA exposure. To validate these transcriptomic results, qPCR was performed on representative downstream genes, showing consistent upregulation of PI3K–AKT pathway components (Itga9, Thbs2, Erbb2, Col1a1, Col1a2, Col9a2; Figure S4A ). Hormone signaling genes were also disturbed, with androgen receptor (Ar) and follicle-stimulating hormone receptor (Fshr) significantly altered and a modest increase in Amh expression observed in the PFOA group ( Figure S4B ). These molecular changes correspond to the histological findings of increased preantral follicles and disrupted preantral-to-antral transition, supporting the notion of premature follicular activation followed by developmental arrest.

Data from the NHANES cycles spanning 1999–2000, 2003–2004, 2005–2006, 2007–2008, 2009–2010, 2011–2012, 2015–2016, and 2017–2020 March were analyzed to explore the association between serum PFAS levels and age at natural menopause (ANM). The study included women aged over 20 years who had corresponding PFAS laboratory data, as well as questionnaire data related to reproductive health. The inclusion and exclusion process is depicted in Figure A. Premenopausal women were defined as those who answered “yes” to the question “Have you had at least one menstrual period in the past 12 months?” or answered “no” but indicated that the reason was pregnancy or breastfeeding ( n = 2924). Postmenopausal women were those who answered “no” to the above-mentioned question and knew the age at their last menstrual period ( n = 1785). For women who reported menopause, the age at the last menstrual period was recorded. Association between serum PFAS levels and age at natural menopause (ANM) in NHANES 1999–2020 March. (A) Flowchart illustrating the participant selection process for the analysis of ANM in relation to serum PFAS exposure. (B) Hazard ratios (HRs) and 95% confidence intervals (CIs) for the risk of earlier age at natural menopause according to serum PFAS tertiles (low, medium, high). Models were adjusted for age, race/ethnicity, education level, parity and the NHANES cycle. * P < 0.05, ** P < 0.01, *** P < 0.001. Perfluorooctanoic acid (PFOA), perfluorooctanesulfonate (PFOS), perfluorohexanesulfonate (PFHxS), perfluorodecanoic acid (PFDeA), 2-( N -methylperfluorooctanesulfonamido) acetic acid (Me-PFOSA), and perfluorononanoic acid (PFNA) were selected for the analysis of their associations with ANM because the detection rates of these PFAS were over 50%. Serum PFAS levels were measured, and values below the limit of detection (LLOD) were replaced by LLOD/√2 according to the NHANES analysis guidelines. The serum PFAS levels were then categorized into tertiles based on their distributions within the study population. The Cox proportional hazards regression model, previously described, was used to identify the types of serum PFAS associated with the onset of menopause. Women who had regular menstrual periods in the past 12 months were censored at the age at the interview. The models were adjusted for potential confounding factors including age, race/ethnicity, education level, NHANES cycles, and parity. Perfluorooctanoic acid (PFOA; Cat. No. 33824, purity >98%) was purchased from Sigma-Aldrich. For in vitro experiments, PFOA was dissolved in dimethyl sulfoxide (DMSO, Cat. No. 472301; purity ≥99.9%, Sigma-Aldrich), whereas for in vivo experiments, it was prepared in tocopherol-stripped corn oil (Cat. No. C7031, Solarbio). Female ICR/CD-1 mice were exposed to PFOA at 1, 2.5, and 5 mg/kg/day or to the vehicle control (corn oil). Doses were selected based on previous studies that serum PFOA levels can reach approximately 22 μg/mL (53 μmol/L) in residents living near or working at chemical plants. The 1 mg/kg/day dose approximates the upper limit of environmentally relevant human exposure as estimated by toxicokinetic modeling studies, whereas the 5 mg/kg/day dose was predicted to produce serum concentrations around 50 μg/mL in mice after repeated administration for 10 days, simulating occupational exposure levels. The 2.5 mg/kg/day dose was chosen as an intermediate level based on prior reports showing that similar exposures result in sustained PFOA accumulation in offspring at the sexual mature stage following prenatal exposure in mice. Female ICR/CD-1 mice with confirmed vaginal plugs (gestational day [GD] 1) were purchased from Charles River (Shanghai, China). Mice were housed in a pathogen-free facility at the Zhejiang Center of Laboratory Animals under 12 h light/dark cycles, 21–23 °C, with free access to standard chow and water. After a 1 week acclimation, pregnant mice were randomly assigned to five groups ( n = 4–7 per group). All procedures were conducted in accordance with ICH S5 (R3) reproductive toxicity guidelines and approved by the Institutional Animal Care and Use Committee of the Zhejiang Center of Laboratory Animals (No. ZJCLA-IACUC-20010194). PFOA at 1, 2.5, and 5 mg/kg/day solvated in tocopherol-stripped corn oil or pure corn oil as vehicle control was administered to female ICR/CD-1 mice. Considering the oral ingestion from diet and water as the primary human exposure route for PFOA, pregnant dams were dosed daily by oral pipetting according to their body weight during GD11 to GD20, the critical window for murine ovarian development. The ovary index was calculated as the ovary weight divided by body weight. Litters were standardized to 10 pups on day 4 of postnatalization (PND4) to minimize maternal bias. One ovary per mouse was used for histomorphometry. Ovaries were fixed in 4% paraformaldehyde overnight, dehydrated, paraffin-embedded, and serially sectioned at 4 μm. Every fifth section was stained with hematoxylin and eosin (H&E) for follicle classification according to standard criteria: (1) primordial follicles: oocytes surrounded by a single squamous granulosa layer; (2) primary follicles: oocytes surrounded by a single cuboidal granulosa layer; (3) preantral follicles: two or more granulosa layers without an antrum; (4) antral follicles: multiple granulosa layers with a visible antrum larger than the oocyte; and (5) atretic follicles: ≥5% pyknotic nuclei in the largest cross section, oocyte shrinkage, or germinal vesicle breakdown. Deformed follicles lacking oocytes or showing collapse were included as atretic. Ovarian cysts were identified by large antral cavities, thickened theca layers, and markedly thinned granulosa compartments without oocytes. The total ovary area per section was measured by using ImageJ, and follicle and corpus luteum (CL) counts were normalized by ovary area (units/mm 2 ) to account for size and sectioning variability. Estrous cyclicity was evaluated in 3 month old F1 females by daily vaginal cytology over 22 consecutive days at 10:00 AM. Vaginal smears were stained with crystal violet and classified into four stages based on cell morphology: (1) proestrus: predominantly nucleated epithelial cells; (2) estrus: clusters of cornified squamous epithelial cells; (3) metestrus: leukocytes with cornified cell fragments; and (4) diestrus: predominantly leukocytes. Paraffin sections (4 μm) of 6 month old F1 ovaries were deparaffinized, rehydrated, and subjected to antigen retrieval, followed by 5% BSA blocking. Sections were incubated overnight at 4 °C with rabbit anti-ANGPTL4 (1:100, Proteintech, Cat. No. 18374-1-AP) and then with HRP-conjugated antirabbit secondary antibody (1:1000, FlexAble, Cat. No. KFA005) for 30 min. Visualization was performed with a DAB substrate, and nuclei were counterstained with hematoxylin. Negative controls were prepared by replacing the primary antibody with PBS. The human granulosa-like tumor cell line KGN (Fenghui Biotechnology, Hunan, China) was cultured in F-12 medium (Hyclone) with 10% charcoal-stripped FBS (Gibco) at 37 °C and 5% CO 2 . PFOA stock solutions (1000 mmol/L in DMSO) were diluted to 1–200 μmol/L for 48 h treatments. The DMSO concentration was ≤0.5%. CCK-8 assays confirmed that concentrations below 500 μmol/L did not affect the viability. Experiments were performed independently at least three times. Following PFOA treatment and siRNA transfection for 48 h, KGN cells were fixed and stained using the E-Click EdU Cell Proliferation Imaging Assay Kit (Elabscience, E-CK-A377, Wuhan, China). Fluorescence images were captured using a Leica DMi8 microscope (Nussloch, Germany). To evaluate the role of ANGPTL4, KGN cells at ∼70% confluence were transfected with siRNA targeting ANGPTL4 (si-ANGPTL4) or negative control siRNA (si-NC) (GenePharma, Shanghai, China) using Lipofectamine 3000 (Thermo Fisher Scientific, USA) according to the manufacturer’s instructions. The following sequences for ANGPTL4 knockdown were used: sense 5′-AGGGAAUCUUCUGGAAGACTT-3′ and antisense 5′-GUCUUCCAGAAGAUUCCCUTT-3′. The 3D structure of PFOA (CID:9554) was obtained from the PubChem database ( http://pubchem.ncbi.nlm.nih.gov/ ). The 3D crystal structure of human Peroxisome Proliferator Activated Receptor Gamma (PPARγ) ligand-binding domain (LBD) (PDB ID: 3U9Q , Resolution: 1.52 Å) was downloaded from the RCSB Protein Data Bank ( http://www.rcsb.org/ ). AutoDock Tools were used to prepare the structures of the protein and ligand. Ligand-binding sites of PPARγ were predicted using PrankWeb ( https://prankweb.cz/ ). Molecular docking was conducted to investigate the interaction between PPARγ and PFOA by using AutoDock Vina 1.5.7. The ligand–receptor conformation was evaluated based on the docking score, in which the lowest binding energy was considered as the optimal one. To explore the molecular mechanisms of PFOA-induced reproductive toxicity, PFOA-exposed KGN cells and PND22 prepubertal ovaries (PFOA vs control = 3:3) were subjected to mRNA sequencing. Total RNA (1 μg/sample) was extracted using TRIzol (Invitrogen). Libraries were prepared with the NEBNext Ultra RNA Library Prep Kit and sequenced on an Illumina NovaSeq platform (150 bp paired-end) after cBot clustering with the TruSeq PE Cluster Kit v3-cBot-HS. Raw reads were processed to remove adapters, poly-N, and low-quality reads. Clean reads were aligned to the reference genome using HISAT2 v2.0.5, and gene-level counts were obtained with featureCounts v1.5.0-p3. Differential expression analysis was performed using DESeq2, and P -values were Benjamini–Hochberg adjusted to control the false discovery rate. Functional enrichment analyses, including gene set enrichment analysis (GSEA) and KEGG pathway enrichment, were conducted using clusterProfiler in R v4.3.3. All in vitro and in vivo experiments were analyzed using GraphPad Prism 7.0 (GraphPad Software, USA). Data were presented as the mean ± SD from at least three independent replicates. Normality was assessed by using the Shapiro–Wilk test. One-way ANOVA followed by Tukey’s post hoc test was applied to compare groups. P ≤ 0.05 was considered statistically significant. Statistical analyses for MR, NHANES, and RNA-seq were conducted in R v4.3.3, using packages including TwoSampleMR (v0.6.3), forestploter (v1.1.2), fdrtool (v1.2.17), survival (v3.5-7), DESeq2 (v1.40.2), clusterProfiler (v4.8.3), ComplexHeatmap (v2.16.0), msigdbr (v7.5.1), and enrichplot (v1.20.1). Detailed procedures for reproductive hormone measurement, qPCR, and western blot to quantify gene expression levels, cell cycle assessment, mendelian randomization analyses for ANGPTL4 levels, and ovary-related diseases were provided in the Supporting Information Text S1–S5.

In summary, this study provides novel and integrative evidence for the long-term reproductive consequences of prenatal PFOA exposure, an area that has been largely overlooked in previous research focusing on adult or short-term effects. By leveraging two decades of NHANES data encompassing 4709 women, we demonstrated that elevated PFOA levels are significantly associated with an earlier age at natural menopause, marking an acceleration of reproductive aging. Mechanistically, our experimental findings reveal a previously unrecognized link between ANGPTL4-mediated granulosa cell dysfunction and PFOA-induced ovarian cyst formation, highlighting a critical pathway underlying ovarian function impairment. Together, these findings expand our understanding of the regulatory mechanism of PFAS reproductive toxicity from the perspectives of developmental origins.

Our NHANES analyses revealed significant associations between higher serum PFAS concentrations and the earlier onset of menopause. Complementing these epidemiological findings, our prenatal mouse model revealed that PFOA exposure precipitated progressive ovarian deterioration and cyst formation in aging offspring. In vitro assays showed that PFOA activated the PPARγ–ANGPTL4 signaling axis in granulosa cells, inducing cell-cycle arrest and impairing proliferation. This dysfunction emerged as a central mechanism contributing to cystogenesis and the disruption of the preantral-to-antral follicular transition. NHANES analyses further demonstrated that higher serum PFOA concentrations were associated with an earlier age at natural menopause following a monotonic, dose–response relationship. Menopause, the definitive end point of female reproductive lifespan, typically occurs in midlife and reflects depletion of the ovarian follicle pool, accompanied by elevated circulating FSH and LH. In line with our findings, previous studies have linked elevated PFOA levels to diminished ovarian reserve, premature ovarian insufficiency, and early menopause , , a continuum of disorders indicative of pathological ovarian aging. However, these studies are predominantly cross-sectional, limiting causal inference due to temporal ambiguity. Moreover, NHANES participants represent the general U.S. population under nonoccupational exposure, where the upper tertile of serum PFOA (∼3 ng/mL) is far below internal levels documented in highly contaminated regions, reaching 1.09 μg/mL (2.68 μmol/L), 22 μg/mL (54 μmol/L), and 92 μg/mL (222 μmol/L). , , PFOA is able to cross the placental barrier and detectable in umbilical cord blood of newborns. Mounting evidence supports the ovarian dysgenesis syndrome (ODS) hypothesis, which posits that early life chemical insults can disrupt ovarian programing and manifest as reproductive disorders in adulthood. The potential long-term reproductive outcomes of prenatal PFOA exposure, such as menopausal timing, remain unexplored in humans, largely due to the lengthy follow-up required. To address this gap, we employed a mammalian model and focused on gestational days 11–20, a critical window for ovarian development in mice, to examine the enduring reproductive toxicity of prenatal PFOA exposure. Exposure levels were chosen to encompass a range of human internal concentrations, from low-dose to occupational levels. Our study provides experimental evidence that prenatal PFOA exposure accelerates ovarian function decline and induces ovarian cyst formation in middle-aged F1 offspring. Ovarian cysts first emerged during early sexual maturity (3–6 months), accompanied by irregular estrous cycles, increased atretic follicles, and a reduced number of corpora lutea. Ovarian cysts of uncertain origin are recognized as one of the hallmarks of reproductive aging in mice. Histologically, these cysts were lined with either flattened/attenuated epithelium or cuboidal secretory cells, some of which were classified as being of follicular origin. In CD-1 Swiss mice, spontaneous ovarian cysts typically arise at 16–24 months of age; , therefore, the early onset observed in our study strongly indicates accelerated ovarian function decline in offspring following prenatal PFOA exposure. Elevated serum FSH and LH further supported premature ovarian failure, paralleling the clinical features of early menopause. Previous animal studies similarly demonstrated that prenatal and lactational PFOA exposure compromises female reproductive health through mechanisms such as mitochondrial dysfunction-induced oocyte apoptosis and disruption of transzonal projections and follicular basement membrane integrity. Together with our findings, these results delineate a trajectory of reproductive toxicity, progressing from early structural damage to long-term ovarian deterioration following prenatal PFOA exposure. However, the concepts of ovarian cysts in mice and humans are not fully equivalent. Although the cystic structures observed in this study may be analogous to human follicular cysts, which are often regarded as a manifestation of diminished ovarian function, there are key distinctions. In humans, follicular cysts typically arise when dominant follicles fail to rupture due to excessive FSH stimulation or the absence of the midcycle LH surge just before ovulation. These cysts are generally transient and functional and are capable of spontaneous regression once hormonal balance is restored. In contrast, the ovarian cysts observed in our study were persistent and accompanied by disrupted folliculogenesis and endocrine dysregulation, suggesting a more sustained pathological state. Therefore, they more closely resemble benign ovarian cysts in humans, which, while generally nonmalignant, do not necessarily cause significant loss of ovarian reserve. Benign ovarian cysts are also hypothesized to represent precursor lesions for certain ovarian cancers. , Epidemiological studies have suggested that the presence of ovarian cysts can increase subsequent cancer risk by up to 12-fold. Emerging evidence has further linked PFAS exposure to ovarian cancer. A nested case-control study demonstrated that both individual and mixed plasma PFASs were associated with poorer overall survival in patients with high-grade serous ovarian cancer (HGSOC). Comparative toxicogenomic analyses have identified key genes involved in both PFOA exposure and ovarian carcinogenesis, and Zhong et al. reported that PFNA, a substitute for PFOA, promoted ovarian cancer progression via the TGF-β/SMAD pathway. Moreover, PFOA was shown to stimulate ovarian cancer cell migration and invasion through ERK signaling. Although the cysts observed in our study exhibited no microscopic features of malignancy, abundant surface vasculature was noted. Considering the paucity of direct in vivo evidence, further investigations and better animal models are warranted to clarify the potential link among prenatal PFOA exposure, cyst formation, and ovarian cancer susceptibility. Granulosa cells (GCs) are the predominant somatic cell type within ovarian follicles, which represent the basic functional units of the ovary. During folliculogenesis, the number of GCs progressively increases until the formation of an antral cavity. Excessive GC apoptosis can lead to follicular atresia and fluid accumulation, eventually resulting in thin-walled cystsa morphological feature commonly observed in polycystic ovaries, a disorder characterized by ovulatory dysfunction. To model occupationally relevant exposures, we selected 200 μmol/L PFOA as the in vitro treatment concentration, which approximates internal serum levels reported in fluorochemical production workers. Through integrated transcriptomic profiling, molecular docking, and functional assays, we demonstrated that PFOA activated the PPARγ–ANGPTL4 signaling axis in granulosa cells, which attenuated their proliferation and likely contributed to progressive granulosa cell depletion within follicles. This cellular dysfunction provides a plausible mechanistic basis for the cystic transformation observed in vivo. PPARs are nuclear transcription factors involved in lipid metabolism, steroidogenesis, apoptosis, and cell cycle regulation. Previous studies have demonstrated that PFOA acts as a strong activator of human PPARα but exhibits minimal or negligible activation toward PPARγ or other nuclear receptors in vitro, corroborated by numerous animal studies − as well. Of note, PPARγ expression is restricted to granulosa cells in the ovary and dynamically regulated during follicular development, while PPARα remains stable in ovarian stromal compartments. PPARγ expression rises during follicular maturation but is downregulated after the luteinizing hormone (LH) surge preceding ovulation. Overactivation of PPARγ has been shown to inhibit granulosa cell proliferation in livestock models, suggesting a cell type-specific regulatory axis. , Therefore, persistent or excessive activation of PPARγ may impair ovulatory capacity and has been reported to inhibit GC proliferation. In our study, transcriptomic enrichment analysis highlighted PPAR signaling as a key pathway altered by PFOA exposure and molecular docking confirmed that PFOA can occupy the ligand-binding pocket of human PPARγ. Although substantial upregulation of PPARγ itself was observed only at concentrations of ≥100 μmol/L, downstream target genes of the PPARγ pathway were significantly induced at concentrations as low as 10 μmol/L in a dose–response relationship. This finding suggests that even low-affinity activation of PPARγ by PFOA could elicit subtle yet biologically relevant toxic effects in ovarian granulosa cells, distinct from the canonical PPARα-driven pathways emphasized in prior studies. PFOA exposure significantly increased ANGPTL4 expression in GCs by nearly 10-fold via PPARγ signaling. ANGPTL4 is a multifunctional secreted protein involved in lipid metabolism and angiogenesis. Elevated serum ANGPTL4 levels have been reported in patients with polycystic ovary syndrome (PCOS), and ANGPTL4 overexpression can suppress GC proliferation via activation of the EGFR/JAK1/STAT3 cascade, contributing to PCOS pathogenesis. Conversely, ANGPTL4 exerts beneficial effects in pregnancy by promoting angiogenesis and facilitating spiral artery remodeling, thereby reducing the risk of preeclampsia. Intriguingly, recent findings point to tissue-specific and context-dependent roles of ANGPTL4 under PFOA exposure. A single-cell RNA sequencing study revealed that PFOA disrupted intercellular communication between endometrial stromal and epithelial cells by downregulating ANGPTL4, ultimately impairing embryo implantation. This contrasts sharply with the pronounced upregulation of ANGPTL4 that we observed in granulosa cells, highlighting a paradoxical, tissue-dependent regulation that may underlie divergent reproductive outcomes. While granulosa cell proliferation was inhibited by PFOA-induced ANGPTL4 upregulation, it may not represent the sole mechanism underlying ovarian cyst formation. ANGPTL4 has been reported to promote ovarian cancer progression by multiple pathways, − including the ERK1/2 signaling pathway. In parallel, loss of PKBβ predisposes mice to ovarian cyst development, which has been strongly associated with ERK pathway activation in theca cellsthe somatic cell population adjacent to granulosa cells lining the follicles. Given that ANGPTL4 is secreted, it is plausible that granulosa–theca crosstalk contributes to cyst formation, although further mechanistic studies are warranted. Moreover, our Mendelian Randomization (MR) analysis revealed a significant causal association between genetically predicted ANGPTL4 levels and endometriosis, a condition frequently characterized by functional ovarian cysts and impaired ovarian function. As a contrast, no causal association was observed between ANGPTL4 levels and several types of ovarian cancer. Additionally, PFOA-induced granulosa cell cycle arrest appeared to contribute to the blockade of the preantral-to-antral follicle transition in prepubertal ovaries, suggesting that estrogen synthesis by granulosa cells in the existing follicle pool was insufficient to initiate the first estrous cycle. Consistent with this notion, a three-dimensional follicle culture study demonstrated that while 250 μmol/L PFOA exposure caused minimal disruption during the early growth phase, marked developmental delays emerged once control follicles began forming an antrum. In line with these observations, we found a concurrent increase in primary, healthy, and atretic preantral follicles, consistent with accelerated early follicle recruitment and progressive depletion of the ovarian reserve. This phenomenon is potentially driven by activation of the PI3K–AKT–mTOR signaling pathway, which is well recognized as a critical regulator of early folliculogenesis. Interestingly, both the PI3K–AKT–mTOR and PPARγ–ANGPTL4 pathways appear to be interconnected in maintaining ovarian homeostasis. Previous studies have shown that PPARγ expression in mouse ovaries is low during the neonatal stages (PND5 and 15) but increases sharply around puberty (PND20), coinciding with follicle activation. Inhibition of PPARγ has been reported to activate the PI3K–AKT pathway by downregulating PTEN, thereby accelerating follicle recruitment. Since PTEN acts as a negative regulator of the PI3K–AKT pathway, PTEN deletion in mouse oocytes causes premature AKT activation, accelerated follicular depletion, and early ovarian failure. These findings suggest that PPARγ might act as an upstream regulator of the PTEN–PI3K–AKT axis, balancing follicle activation and quiescence to preserve the ovarian reserve. In our study, prenatal PFOA exposure appeared to disturb this finely tuned equilibrium. F1 offspring exhibited increased preantral follicle counts and upregulation of PI3K–AKT signaling at puberty, followed by an impaired preantral-to-antral transition and ovulatory dysfunction at later stages. This biphasic pattern suggests that PFOA may differentially perturb the PPARγ–PTEN–PI3K–AKT feedback over timeinitially weakening PTEN-mediated inhibition to trigger premature follicle activation but subsequently maintaining excessive PPARγ activity that restrains granulosa cell proliferation and antral maturation. Such asynchronous regulation could underlie the paradoxical coexistence of early follicular activation and later follicular depletion observed in our model. Several limitations should be acknowledged. First, our mouse model involved short-term prenatal exposure to PFOA at low-to-occupational doses. Although this design captured a wide internal exposure range, the lowest dose elicited minimal phenotypic changes, whereas real-world nonoccupational human exposure is typically chronic, low-level, and long-term, which may lead to distinct biological effects. For instance, we observed that even 1 μmol/L PFOA could upregulate CYP11A1 in granulosa cells, suggesting that subtle steroidogenic perturbations may occur at environmentally relevant concentrations. Second, the lack of direct quantification of PFOA in maternal and F1 serum limits the establishment of internal dose–response relationships. Previous studies have demonstrated persistent PFOA retention in offspring following prenatal exposure, supporting in utero accumulation. Future work incorporating LC–MS/MS–based internal dosimetry will be essential to strengthen the exposure–effect linkage. Third, our in vitro experiments relied on acute exposure paradigms, which cannot fully replicate the latent interval between prenatal exposure and the manifestation of offspring phenotypes. Although KGN cells are widely recognized as physiologically relevant surrogates for human granulosa cells, they may not fully recapitulate the complexity of in vivo ovarian physiology. Future validation using conditional granulosa cell-specific PPARγ knockout mice and ANGPTL4 inhibition models will be valuable to further substantiate the causal role of the PPARγ–ANGPTL4 pathway in mediating PFOA-induced ovarian toxicity. Collectively, these limitations highlight the need for long-term, low-dose exposure models and ex vivo systems that better simulate human environmental risk. Future research integrating such approaches and single-cell transcriptomic analyses will be critical to delineating the full spectrum of reproductive hazards posed by PFOA and related PFAS in environmentally relevant contexts.

Per- and polyfluoroalkyl substances (PFASs) are a large group of synthetic fluorinated compounds widely applied in consumer and industrial products such as stain-resistant coatings, food packaging, and cosmetics. , Among PFAS, perfluorooctanoic acid (PFOA) remains the most prevalent contaminant in human serum and aquatic environments. − PFOA exposure has been associated with ovarian dysfunction, including impaired mitochondrial activity in granulosa cells, abnormal folliculogenesis, and reduced ovarian reserve, as revealed by various epidemiological, − in vitro, and in vivo studies. − Despite these observations, the long-term reproductive effects of PFOA exposure during early developmental windows remain insufficiently characterized. The ovary is particularly vulnerable to environmental insults during development, and early life perturbations may only manifest later as reproductive aging or disease. A population-based Danish birth cohort reported that higher maternal serum PFOA levels during pregnancy were linked to delayed menarche in daughters. Animal studies have further shown that prenatal PFOA exposure can delay puberty, impair oocyte maturation, and compromise follicular integrity in early reproductive stages. However, these investigations primarily address short-term outcomes during adolescence or early adulthood, leaving unresolved whether prenatal PFOA exposure contributes to long-term ovarian aging and pathology, such as cyst formation. Given the persistent nature of PFOA and its bioaccumulation potential, it is plausible that early life exposure may predispose the ovary to degenerative changes later in life. The ovary plays a central role in female reproduction by maintaining endocrine homeostasis and orchestrating follicular development. Notably, ovarian aging typically precedes the decline of other organ systems and is characterized by progressive depletion of the ovarian follicle pool, culminating in menopause. In rodents, ovarian cysts are recognized as one of the histological hallmarks of ovarian aging. Two main hypotheses have been proposed regarding the origins of the ovarian cysts. One suggests a follicular origin, as the thin, discontinuous cuboidal cell layer lining the cyst often represents remnants of degenerated granulosa cells. The other attributes cyst formation to intraovarian inclusion cysts, arising from age-related invagination of the ovarian surface epithelium (OSE) during repeated ovulatory cycles. , Collectively, these mechanisms reflect impaired ovulation and defective postovulatory wound healing that accompany reproductive aging. In this study, we investigated the association of PFOA exposure with earlier onset of menopause and probed the mechanism through ovarian morphological examination, cell-based assays, and molecular interaction study. The Cox proportional hazard regression model was leveraged to reveal the association between human PFOA exposure and menopausal timing using population-level data from the National Health and Nutrition Examination Survey (NHANES). Mouse models prenatally exposed to PFOA were used to track progressive changes of ovarian morphology and function. Transcriptomic profiling was conducted to find the candidate disturbed pathway, and the interactions of PFOA with relevant targets involved were further probed by molecular docking. A series of loss-of-function experiments in ovarian granulosa cells were performed to investigate the molecular signaling of PFOA-induced ovarian deterioration. By integrating population-based epidemiological evidence with experimental and molecular approaches, our study bridges human associations with mechanistic insights, thereby addressing critical gaps in understanding the long-term reproductive toxicity of PFOA and offering new perspectives on its role in ovarian aging and dysfunction.